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=PMC12548461&blobtype=pdf | # The hemagglutinin proteins of clades 1 and 2.3.4.4b H5N1 highly pathogenic avian influenza viruses exhibit comparable attachment patterns to avian and mammalian tissues
Bingkuan Zhu, Kevin Fung, Hailey Feng, Julia Beatty, Fraser Hill, Anne Tse, Christopher Brackman, Thomas Sit, Agnès Poujade, Nicolas Gaide, Mariette Ducatez, Gilles Foucras, Malik Peiris, Shih-Chieh Ti, John Nicholls, Hui-Ling Yen
## Abstract
The global spread of the A/goose/Guangdong/1/96-lineage H5N1 highly pathogenic avian influenza (HPAI) viruses has been accompanied by an expanded host range and the establishment of sustained viral transmission among dairy cattle. To evaluate if the evolving H5N1 viruses have changed tissue tropism over time, we compared the binding patterns of recombinant hemagglutinin (HA) proteins derived from clade 1 (A/Vietnam/1203/04, H5VN) and circulating clade 2.3.4.4b viruses detected from wild bird (A/Eurasian Teal/Hong Kong/AFCD-HKU-23-14009-01020/2023, H5HK) and dairy cattle (A/bovine/Ohio/B24OSU-439/2024, H5OH). The HA protein of A(H1N1)pdm09 virus was included for comparison. Using bio-layer interferometry, H1 protein prefer entially bound to the α2,6-linked sialoside 6′SLNLN, while H5 proteins preferentially bound to the α2,3-linked sialoside 3′SLN. H5OH showed higher binding affinity to 3′SLN than H5HK and H5VN. The attachment patterns of H1 and H5 proteins to the respiratory tissues of different species and dairy cattle mammary glands were evaluated. Compared to the H1 protein, H5 proteins showed stronger binding to the lung epithelial cells of cat, cattle, chicken, ferret, human, and pig, and the clade 2.3.4.4b H5 proteins exhibited increased binding to pig and cattle bronchial epithelial cells. All H5 proteins were attached to the alveolar and cistern epithelial cells in mammary glands, where α2,3-linked and α2,6-linked sialyl glycans were detected by Maackia amurensis lectin II and Sambucus nigra lectin, respectively. Taken together, the HA proteins of clade 1 and 2.3.4.4b H5N1 viruses generally share comparable attachment patterns to avian and mammalian tissues, despite evolving into antigenically distinct clades over the past 3 decades.
IMPORTANCEThe outbreaks of H5N1 highly pathogenic avian influenza (HPAI) virus among US dairy cattle since 2024 have raised concerns of the potential changes in HA receptor binding specificity and tissue tropism. Using insect-cell-expressed recombinant HA proteins derived from clade 1 and circulating clade 2.3.4.4b H5N1 viruses, we showed that the dairy cattle H5 protein retained binding specificity for the avian-like α2,3-linked sialoside 3′SLN over the human-like α2,6-linked sialoside 6′SLNLN, with higher binding affinity to 3′SLN than the other H5 proteins. Clade 1 and 2.3.4.4b H5 proteins showed comparable attachment patterns to the mammary tissues of lactating dairy cattle, which showed high expression of α2,3-linked and α2,6-linked sialyl glycans. All H5 proteins also showed comparable attachment patterns to the lungs of cat, cattle, chicken, ferret, human, and pig. Our results suggest that the recent H5N1 outbreaks in dairy cattle may be related to ecological factors rather than changes in HA receptor binding specificity.
KEYWORDS highly pathogenic avian influenza (HPAI), clade 2.3.4.4b, recombinant HA protein, tissue tropism S ince the emergence of the A/goose/Guangdong/1/96 (Gs/Gd) lineage of H5N1 highly pathogenic avian influenza virus (HPAI) from Southern China three decades ago, the hemagglutinin (HA) protein has evolved into multiple antigenically distinct clades, and the Gs/Gd-like viruses have spread across continents via migratory birds (1). The emergence and the expanded geographic distribution of clade 2.3.4.4b H5N1 viruses since 2020 has been accompanied by increasing numbers of outbreaks in domestic and wild bird species (2), spillover infections in humans and a wide range of mammalian species (3,4), and the establishment of sustained viral transmission among dairy cattle, a new mammalian host for influenza A viruses (5)(6)(7).
Since March 2024, multiple states in the US have reported outbreaks of genotype B3.13 clade 2.3.4.4b H5N1 virus in dairy cattle (8). Infected cattle presented loss of appetite, massive drop in milk production, and mild respiratory signs (9). Field studies detected higher viral loads in the milk and mammary gland than those detected in the nasal swabs or lungs of infected dairy cattle (6). Experimental studies that used intra-nasal or intra-mammary gland inoculation routes also demonstrated preferential viral replication in the mammary glands over the respiratory tissues (9,10). Importantly, dairy cattle with intra-mammalian inoculation of a genetically distinct clade 2.3.4.4b virus isolated from a wild goose in Europe (genotype euDG) showed comparable clinical signs as those inoculated with a genotype B3.13 dairy cattle isolate, suggesting that the ability to infect dairy cattle may be shared among clade 2.3.4.4b H5N1 viruses (9), a finding supported by the detection of genotype D1.1 of clade 2.3.4.4b in dairy cattle in January 2025 (11). These findings suggest that H5N1 outbreaks in dairy cattle may be attributed to both virological factors (e.g., ability to replicate in mammary gland tissue of dairy cattle) and ecological factors (e.g., shared milking devices and inter-state cattle movements). However, it is not clear if early Gs/Gd-lineage H5N1 viruses also possess the capacity to infect dairy cattle.
The receptor binding profile of HA proteins determines the host range and cell tropism of influenza A viruses. The HA proteins of avian influenza viruses preferentially bind to α2,3-linked sialyl glycans, while the HA proteins of human and swine influenza viruses preferentially bind to α2,6-linked sialyl glycans (2,3). Here, using insect cellexpressed recombinant HA proteins, we compared the HA attachment pattern of clades 1 and 2.3.4.4b viruses to the mammary tissue of dairy cattle, as well as the respiratory tissues of cat, cattle, chicken, ferret, human, and pig.
## RESULTS
## Binding affinity of insect cell-expressed recombinant HA pro teins to Neu5Acα2-3Galβ1-4GlcNAcβ-sp3-PAA-biot (3′SLN) and Neu5Acα2-6Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAcβ-sp3-PAA-biot (6′SLNLN) by bio-layer interferometry assay
To compare the attachment pattern of HA protein of past and circulating H5N1 viruses, we expressed the HA proteins of A/Vietnam/1203/2004 (clade 1, abbreviated as H5VN, GISAID accession number EPI_ISL_21080, https://gisaid.org/), A/Eurasian Teal/ Hong Kong/AFCD-HKU-23-14009-01020/2023 (clade 2.3.4.4b, abbreviated as H5HK, GISAID accession number EPI_ISL_19258154), and A/bovine/Ohio/B24OSU-439/2024 (clade 2.3.4.4b, abbreviated as H5OH, GISAID accession number EPI_ISL_19178083). For comparison, the HA protein from A/California/04/2009 A(H1N1)pmd09 virus (H1CA, GISAID accession number EPI_ISL_393964) was also expressed. The expressed HA proteins were evaluated for their binding affinity to 3′SLN and 6′SLNLN using bio-layer interferometry (BLI). All H5 proteins exhibited preferential binding to 3′SLN over 6′SLNLN, while the H1CA exhibited preferential binding to 6′SLNLN over 3′SLN (Fig. 1A). Further analysis using serially twofold diluted HA proteins (from 200 to 6.25 nM) to determine binding affinity, we observed that H5OH exhibited higher binding affinity (mean K D ± SD = 18.55 ± 1.35 nM) than H5VN (36.45 ± 1.75 nM) and H5HK (46.95 ± 3.45 nM) (one-way analysis of variance [ANOVA], P < 0.01) (Fig. 1B). While the clade 1 H5VN differed by the clade 2.3.4.4b H5OH by 40 amino acids, the two clade 2.3.4.4.b H5HK and H5OH only differed by four amino acids (L111M, L122Q, T199I, V214A, H3 numbering) in HA1 (Fig. 1C). Specifically, the T199I change has been reported to increase the HA binding breadth to α2,3-linked N-acetyllactosamines of the first human H5N1 isolate (A/Texas/37/2024) reported in the dairy cattle outbreak (12).
## Clade 1 and clade 2.3.4.4b H5 proteins bind to the mammary tissues of lactating cows
Infection of mammary tissues has been a unique feature observed from the outbreaks of clade 2.3.4.4b H5N1 in dairy cattle (6,9,10). We compared the attachment pattern of H5VN, H5HK, H5OH, and H1CA to the mammary gland tissues of lactating cows (Fig. 2A). With recombinant proteins diluted to 12.5 µg/mL, we observed no binding of H1CA, but all three H5 recombinant proteins showed comparable binding intensity to the alveolar and cistern epithelial cells (Fig. 2B). It is noteworthy that no binding to the ductal epithelial cells was observed (Fig. 2B). To further characterize α2,3 and α2,6linked sialyl glycans presented in the mammary tissues, we performed lectin staining with Sambucus nigra lectin (SNA) that preferentially binds to NeuAcα2,6Galβ1,4GlcNAc, Maackia amurensis lectin I (MAL-I) that preferentially binds to NeuAcα2,3Galβ1,4GlcNAc in N-glycans or O-glycans, and MAL-II that preferentially binds to NeuAcα2,3Galβ1,3Gal NAc in O-glycans. Both SNA and MAL-II showed apparent binding to the alveolar epithelial cells and minor binding to the cistern epithelial cells. Interestingly, no apparent MAL-I binding was observed, suggesting that the mammary tissue of lactating cows may express low levels of NeuAcα2,3Galβ1,4GlcNAc (Fig. 2A).
## H5 proteins exhibited strong binding to lung epithelial cells of chicken, cat, cattle, ferret, human, and pig
We further evaluated the attachment pattern of recombinant HA proteins to the respiratory tissues of different species. All H5 protein binds to chicken trachea and lung epithelial cells, while no binding of H1CA was observed (Fig. 3A andB). The recombinant HA proteins also exhibited different attachment patterns to the mammalian respiratory tissues (Fig. 4 and5). In the ferret bronchus, H1CA showed patchy (Fig. 4A) but a higher overall binding signal (Fig. 4B) to the bronchial epithelial cells than the H5 proteins, while the clade 2.3.4.4b H5 proteins showed stronger binding to pig and cattle bronchus than H1CA (Fig. 4A andB). All three H5 recombinant proteins showed stronger binding to the lung epithelial cells of cat, cattle, human, and pig than the H1 protein, with the H5OH exhibiting stronger binding to the ferret lung epithelial cells than H5HK or H5VN (Fig. 5A andB). Previous studies have reported that H5N1 and avian influenza viruses generally showed stronger attachment to lung tissues of various species than the human seasonal influenza viruses (13,14). Taken together, the results suggest that the HA proteins derived from clade 1 and clade 2.3.4.4b H5N1 viruses showed minor differences in their attachment patterns to the respiratory tissues of different species.
## DISCUSSION
The expanded host range and the sustained transmission of clade 2.3.4.4b H5N1 viruses among dairy cattle have raised concerns about the potential changes in the HA attachment pattern to sialyl glycans expressed on animal tissues, which may affect viral transmissibility and pandemic potential. Previous studies have shown that dairy cattle viruses exhibited preferential binding for α2-3-linked sialosides and tissue tropism that resembles avian influenza viruses, including early Gs/GD-lineage H5Nx strains (7,12,(15)(16)(17)(18)(19)(20)(21)(22)(23). Similar to our study design, two previous studies used mammalian (18) or insect-cell expressed (7) tissue of cat and ferret, as well as the bronchus tissue of cattle, ferret, and pig. We also observed that the clade 2.3.4.4b H5 proteins exhibited increased binding to pig and cattle bronchial epithelial cells than the H1 protein, suggesting that pigs may be susceptible to infection of clade 2.3.4.4b viruses (24), and further co-infection with other swine influenza viruses may pose a risk for the emergence of novel reassortant viruses.
Considering available literature to date, the results suggest the H5 proteins derived from the past and circulating clade 2.3.4.4b H5Nx viruses generally share a compara ble attachment pattern to the cattle mammary tissues and the respiratory tissues of different hosts. This suggests that the H5N1 outbreaks in dairy cattle may be related to ecological factors rather than changes in HA receptor binding specificity. Additional epidemiological studies and environmental sampling are needed to identify risk factors associated with the introduction of H5N1 into dairy herds. Using BLI, we observed that all H5 proteins preferentially bind to 3′SLN with no detectable binding to 6′SLNLN. H5OH also showed higher binding affinity to 3′SLN than H5HK or HKVN. The BLI assay is sensitive and quantitative, but only allows evaluating HA binding to a specific glycan at a time. On the other hand, the use of recombinant HA proteins allowed assessing viral tropism in tissues presented with diverse glycan structures, although this method is more qualitative than quantitative. It is interesting to note that while H5VN showed lower binding to 3′SLN than H5OH, H5VN appeared to show similar attachment pattern as H5OH to various animal tissues. In agreement with our finding, glycan array analysis (16) showed H5VN demonstrated reduced binding to 3′SLN than the clade 2.3.4.4b A/Texas/37/2024 virus (with identical HA sequence as H5OH). Another study also showed that the H5VN exhibited different binding patterns than the HA of A/bovine/Ohio/B24OSU-432/2024 (with identical HA sequence as H5OH) to bi-antennary N-glycans and O-glycans (19). Collectively, these findings demonstrated the strength of different experimental methods and the need to consider different platforms while assessing HA binding specificity and tissue tropism.
Experimental infection showed that the mammary gland may serve as the main site of replication in dairy cattle by the clade 2.3.4.4b H5N1 viruses (9,10). Since clade 1 and clade 2.3.4.4b H5 proteins all bind to the alveolar and cistern epithe lial cells in the mammary glands, it is likely that clade 1 H5N1 virus may similarly cause infection in mammary gland tissue, given the proper opportunity. The attach ment pattern of H5 proteins in mammary glands is in accordance with the results reported by previous studies (7,18). Using lectin staining, we observed predominant expression of NeuAcα2,6Galβ1,4GlcNAc (detected by SNA) and NeuAcα2,3Galβ1,3Gal NAc (detected by MAL-II) in the mammary gland tissues of lactating dairy cattle, while the expression of NeuAcα2,3Galβ1,4GlcNAc (detected by MAL-I) was low. This result, in combination with the H5 protein attachment pattern, suggests H5 proteins bind to the NeuAcα2,3Galβ1,3GalNAc O-glycans in the mammary tissues (12). Interest ingly, although α2,6-linked sialyl glycans were distributed along the glandular alveo lar and cistern epithelial cells, we did not observe any binding of the H1 protein to these tissues. Our result is consistent with those reported previously, including limited attachment of recombinant H1 protein derived from a mouse-adapted influenza strain A/Puerto Rico/8/34 to dairy cattle mammary tissues (18), and restricted replication of an A(H1N1)pdm09 virus in the ex vivo culture of bovine mammary gland (25).
The major limitation of our study is that the HA attachment pattern alone is insufficient to infer viral replication efficiency in these animal tissues, as viral replica tion is also determined by the viral gene constellation and host-adaptive amino acid changes. As the circulating clade 2.3.4.4b viruses continue to expand their genetic diversity through frequent genetic reassortments with other avian influenza viruses (26)(27)(28)(29), the replication efficiency of the genetically diverse clade 2. Taken together, the data available to date support that the clade 2.3.4.4b retained a comparable receptor binding profile as the early H5N1 viruses. However, it's important to note that H5N1 viruses continue to cause spillover infections in mammals, which provide opportunities for viral adaptation (32). A study reported that a single Gln226Leu mutation may switch the cattle H5N1 virus binding specificity to human-type receptors (15). Continuous efforts on surveillance and monitoring of the evolution of H5N1 viruses isolated from different host species are essential for pandemic preparedness.
## MATERIALS AND METHODS
## Expression of recombinant HA proteins
Soluble recombinant HA proteins were expressed and purified following established protocols (33,34). Genes encoding the ectodomains of the HA protein of the A(H1N1)pdm09 and A(H5N1) (with the multibasic cleavage site removed) viruses were subcloned into the baculovirus transfer vector pFastBac1 (Invitrogen), in frame with an N-terminal gp67 signal peptide, a C-terminal trimerization foldon sequence from bacteriophage T4, followed by a thrombin cleavage site and a His 6 -tag. Transfection and baculovirus amplification were performed in Sf9 cells using the Bac-to-Bac baculovirus expression system (Invitrogen). On day 3 post-infection, supernatants were harves ted, and the HA proteins were purified using Ni-charged immobilized metal affinity chromatography resin (Bio-Rad). Recombinant proteins were stored in PBS with 20% sucrose (Sigma) in aliquots at -80°C.
## HA binding affinity to glycans using bio-layer interferometry
The expressed recombinant HA proteins were evaluated for their binding affinity to biotinylated 3′SLN (Neu5Acα2-3Galβ1-4GlcNAcβ-sp3-PAA-biot) (GlycoNZ) and biotinyla ted 6′SLNLN (Neu5Acα2-6Galβ1-4GlcNAcβ1-3Galβ1-4GlcNAcβ-sp3-PAA-biot) (GlycoNZ) using BLI. Assays were performed using the Octet Red96e system (FortéBio) in 96-well microplates as previously described (35). Dulbecco's PBS with calcium, magnesium, and 0.005% Tween-20 was used as the assay buffer to reconstitute protein and glycan molecules. Biotinylated 3′SLN and 6′SLNLN (GlycoNZ) were preloaded to the strepta vidin-coated biosensors (FortéBio) at 1 µg/mL for 10 min. HA proteins diluted to 67.5 nM were pre-conjugated with the mouse anti-His-tag antibody (Thermo Scientific, clone # MA1-21315) and the horseradish peroxidase (HRP)-conjugated goat anti-mouse secondary antibody (Abcam, clone # ab6789) at the molar ratio of 2:1:2 for 30 min on ice before being added to the wells. The 96-well plates were incubated at 30°C for 30 min, with sample plates agitated at 1,000 RPM. Binding kinetics of H5 proteins to 3′SLN were measured for 20 min at HA concentrations of 6.25, 12.5, 25, 50, 100, and 200 nM. The data were fitted with a 1:1 binding model to evaluate HA binding affinity (K D ) to 3′SLN.
## Immunohistochemical staining with recombinant HA proteins
Formalin-fixed and paraffin-embedded tissue blocks and sections were prepared by the Department of Pathology, Li Ka Shing Faculty of Medicine, The University of Hong Kong or obtained from collaborating laboratories. Immunohistochemistry was performed according to a previously described protocol (18) with minor modifications. Briefly, the sections were dried at 60°C for 20 min, deparaffinized, and rehydrated. Antigen retrieval was achieved by boiling the sections in 10 mM sodium citrate (pH 6.0) for 20 min. Endogenous peroxidase activity was quenched using 3% hydrogen peroxide, and nonspecific binding was blocked with 10% goat serum (Giobco). Histidine-tagged recombinant HA proteins were diluted to 12.5 µg/mL and were pre-conjugated with the mouse anti-His-tag primary antibody and the HRP-conjugated goat anti-mouse secondary antibody (Abcam) as described. This pre-complexed mixture was then applied to the tissue sections and incubated for 2 h at room temperature, followed by washing with PBS buffer containing 0.05% (vol/vol) Tween-20 (PBST). The chromo gen 3-amino-9-ethylcarbazole (AEC) (Sigma-Aldrich) was used as the HRP substrate. Tissues were counterstained with Gill's hematoxylin (Vector Laboratories), mounted with permanent aqueous mounting medium (Bio-Rad), and examined using a Nikon Eclipse Ti-S microscope. The experiments were repeated twice independently.
## Lectin staining
To evaluate the distribution of α2,3-linked and α2,6-linked sialyl glycans in tissue slides, biotinylated Sambucus nigra lectin (SNA), Maackia amurensis lectin I (MAL-I), and Maackia amurensis lectin II (MAL-II) were used for staining. SNA is known to preferentially bind to α2,6-linked terminal sialic acids (SA) (Neu5Acα2,6Galβ1,4GlcNAc). MAL-I and MAL-II preferentially bind to α2,3-linked terminal SA, but MAL-I preferred Neu5Acα2,3Galβ1,4GlcNAc while MAL-II preferred Neu5Acα2,3Galβ1,3GalNAc. Antigen retrieval and endogenous peroxidase blocking were performed as described above. After blocking with 0.1% bovine serum albumin (Sigma-Aldrich), the slides were incubated with 20 µg/mL SNA/MAL-I, or 10 µg/mL MAL-II at room temperature for 1 h, followed by incubation with alkaline phosphatase-conjugated streptavidin (Vector Laboratories) for 45 min. The slides were then developed with Vector Red Substrate Kit (Vector Labora tories). After counterstaining with Gill's hematoxylin (Vector Laboratories), the slides were counterstained with Scott's tap water (Sigma-Aldrich), air-dried, and mounted with Permount (Fisher Scientific).
## Statistical analysis
To compare HA binding intensity to the respiratory and mammary tissues, the chroma gen AEC signal was quantified using the Qupath software (36) from two independently stained tissue slides, as shown. A one-way ANOVA test was used to compare the binding signal of different recombinant HA proteins.
## References
1. Bodewes, Kuiken (2018) "Changing role of wild birds in the epidemiol ogy of avian influenza A viruses" *Adv Virus Res*
2. Adlhoch, Fusaro, Gonzales et al. (2023) "European Food Safety Authority, European Centre for Disease Prevention and Control, European Union Reference Laboratory for Avian Influenza" *EFSA J*
3. Plaza, Gamarra-Toledo, Euguí et al. (2024) "Recent changes in patterns of mammal infection with highly pathogenic avian influenza A(H5N1) virus worldwide" *Emerg Infect Dis*
4. Who (1920) "Cumulative number of confirmed human cases for avian influenza A(H5N1) reported to WHO"
5. Burrough, Magstadt, Petersen et al. (2024) "Highly pathogenic avian influenza A(H5N1) clade 2.3.4.4b virus infection in domestic dairy cattle and cats, United States" *Emerg Infect Dis*
6. Caserta, Frye, Butt et al. (2024) "Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle" *Nature*
7. Song, Hao, Han et al. (2025) "Receptor binding, structure, and tissue tropism of cattle-infecting H5N1 avian influenza virus hemagglutinin" *Cell*
8. Cdc (2025) "Current situation: bird flu in dairy cows"
9. Halwe, Cool, Breithaupt et al. "4b dynamics in experimentally infected calves and cows" *Nature*
10. Baker, Arruda, Palmer et al. (2025) "Dairy cows inoculated with highly pathogenic avian influenza virus H5N1" *Nature*
11. Usda (2025) "APHIS confirms D1.1 genotype in dairy cattle in Nevada"
12. Good, Fernández-Quintero, Rodriguez et al. (2024) "A single mutation in dairy cow-associated H5N1 viruses increases receptor binding breadth" *Nat Commun*
13. Van Riel, Munster, De Wit et al. (2007) "Human and avian influenza viruses target different cells in the lower respiratory tract of humans and other mammals" *Am J Pathol*
14. Shinya, Ebina, Yamada et al. (2006) "Avian flu: influenza virus receptors in the human airway" *Nature*
15. Lin, Zhu, Wang et al. (2024) "A single mutation in bovine influenza H5N1 hemagglutinin switches specificity to human receptors" *Science*
16. Pulit-Penaloza, Belser, Brock et al. (2024) "Transmission of a human isolate of clade 2.3.4.4b A(H5N1) virus in ferrets" *Nature*
17. Santos, Wang, Mcbride et al. (2025) "Bovine H5N1 binds poorly to human-type sialic acid receptors" *Nature*
18. Carrasco, Gröne, Van Den Brand et al. (2024) "The mammary glands of cows abundantly display receptors for circulating avian H5 viruses" *J Virol*
19. Chopra, Ray, Page et al. (2025) "Receptor-binding specificity of a bovine influenza A virus" *Nature*
20. Yang, Qureshi, Kolli et al. (2025) "The haemagglutinin gene of bovine-origin H5N1 influenza viruses currently retains receptorbinding and pH-fusion characteristics of avian host phenotype" *Emerg Microbes Infect*
21. Fabrizio, Kandeil, Harrington et al. (2025) "Genotype B3.13 influenza A(H5N1) viruses isolated from dairy cattle demonstrate high virulence in laboratory models, but retain avian viruslike properties" *Nat Commun*
22. Eisfeld, Biswas, Guan et al. (2024) "Pathogenicity and transmissibility of bovine H5N1 influenza virus" *Nature*
23. Gu, Maemura, Guan et al. (2024) "A human isolate of bovine Full-Length Text Journal of Virology October"
24. "5N1 is transmissible and lethal in animal models" *Nature*
25. Kwon, Trujillo, Carossino et al. (2024) "Pigs are highly susceptible to but do not transmit mink-derived highly pathogenic avian influenza virus H5N1 clade 2" *Emerg Microbes Infect*
26. Imai, Ueki, Ito et al. (2025) "Highly pathogenic avian H5N1 influenza A virus replication in ex vivo cultures of bovine mammary gland and teat tissues" *Emerg Microbes Infect*
27. Fusaro, Zecchin, Giussani et al. (2024) "High pathogenic avian influenza A(H5) viruses of clade 2.3.4.4b in Europe-Why trends of virus evolution are more difficult to predict" *Virus Evol*
28. Signore, Giacinti, Jones et al. (2025) "Spatiotemporal reconstruction of the North American A(H5N1) outbreak reveals successive lineage replacements by descendant reassortants" *Sci Adv*
29. Zhang, Yang, Han et al. (2025) "Unique phenomenon of H5 highly pathogenic avian influenza virus in China: co-circulation of Clade 2.3.4.4b H5N1 and H5N6 results in diversity of H5 Virus"
30. Barman, Turner, Hasan et al. (2025) "Reassortment of newly emergent clade 2.3.4.4b A(H5N1) highly pathogenic avian influenza A viruses in Bangladesh" *Emerg Microbes Infect*
31. Bauer, Leijten, Iervolino et al. (2024) "A 2022 avian H5N1 influenza A virus from clade 2.3.4.4b attaches to and replicates better in human respiratory epithelium than a 2005 H5N1 virus from clade"
32. Carrasco, Lin, Zhu et al. (2025) "The Q226L mutation can convert a highly pathogenic H5 2.3.4.4e virus to bind human-type receptors"
33. Xie, Yang, Jiao et al. (2025) "Clade 2.3.4.4b highly pathogenic avian influenza H5N1 viruses: knowns, unknowns, and challenges" *J Virol*
34. Margine, Palese, Krammer (2013) "Expression of functional recombinant hemagglutinin and neuraminidase proteins from the novel H7N9 influenza virus using the baculovirus expression system" *J Vis Exp*
35. Shi, Zhang, Wang et al. (2013) "Structures and receptor binding of hemagglutinins from human-infecting H7N9 influenza viruses"
36. Du, Wolfert, Peeters et al. (2020) "Mutation of the second sialic acid-binding site of influenza A virus neuraminidase drives compensatory mutations in hemagglutinin" *PLoS Pathog*
37. Bankhead, Loughrey, Fernández et al. (2017) "QuPath: open source software for digital pathology image analysis" *Sci Rep* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724370&blobtype=pdf | # FMDV 3A cooperates with PDCD10 to promote FMDV replication by inhibiting VISA-mediated innate immunity
Qian Li, Xiaofeng Nian, Xiaofen Shang, Zongbo Zeng, Zhikuan Luo, Bo Du, Meihua Ma, Zixiang Zhu, Fan Yang, Jingjing Pei, Weijun Cao, Hongbin Yan, Li Li, Yigang Xu, Xusheng Ma, Haixue Zheng
## Abstract
RIG-I like receptors (RLRs) recognize RNA viruses and induce an innate immune response. Although many host factors strictly regulate the signal transduction of the RLR pathway, the mechanisms remain unclear. In the present study, we dem onstrated that virus infection slightly increased the expression of programmed cell death protein 10 (PDCD10). PDCD10 overexpression inhibited interferon beta (IFN-β) promoter activation after Sendai virus (SeV) infection. Moreover, PDCD10 negatively regulated RNA virus-induced IFN-β secretion. These effects were reversed following PDCD10 gene knockout. PDCD10 also interacted with virus-induced signaling adaptor (VISA) and disrupted the formation of the VISA-IRF3 complex to inhibit IFN-β production. Additionally, PDCD10 promoted foot-and-mouth disease virus (FMDV) replication by inhibiting IFN-β production. FMDV is the causative pathogen of foot-and-mouth disease, one of the most destructive and contagious animal diseases in the world. The FMDV 3A protein plays important roles in viral replication, host tropism, and immune regulation. Our experimental results also showed that full-length 3A cooperated with PDCD10 to inhibit IFN-β production by promoting the binding of PDCD10 to VISA. Collectively, the study findings revealed that PDCD10, as a new negative regulator, cooperated with 3A to inhibit viral-induced IFN-β production.IMPORTANCE Foot-and-mouth disease virus (FMDV) is a pathogen that causes a highly contagious and destructive foot-and-mouth disease in animals with cloven hooves. Although the 3A protein of FMDV is involved in viral replication and host tropism, its function remains unclear. PDCD10 plays critical roles in normal cardiovascular develop ment, cell proliferation, and normal structure and assembly of the Golgi complex. The present study showed that PDCD10 expression was slightly increased by virus infection, while PDCD10 promoted FMDV replication. Our results also demonstrated that PDCD10 inhibited Sendai virus-induced interferon beta (IFN-β) production through interaction with virus-induced signaling adaptor (VISA). PDCD10 also disrupted VISA-IRF3 complex formation to impair IFN-β production induced by RNA virus. The FMDV 3A protein bound with PDCD10 to synergistically promote FMDV replication. This study helped to reveal the potential mechanism of FMDV 3A protein and PDCD10 impact on viral replication. KEYWORDS foot-and-mouth disease virus, 3A, PDCD10, innate immunity, VISA, interferons, virus replication F oot-and-mouth disease virus (FMDV) belongs to the genus Aphthovirus of the Picornaviridae family, and it causes a highly contagious and severe disease with vesicle formation and erosions in the epithelium of animals with cloven hooves (1). The single positive-stranded RNA genome of FMDV encodes four structural proteins (VP1, VP2, VP3, and VP4) and eight nonstructural proteins (L pro , 2A, 2B, 2C, 3A, 3B, 3C, and 3D) (2). The 3A protein of FMDV anchors on the intracellular membrane with its
hydrophobic motif and mediates the localization of the FMDV replication complex on the cell membrane. A previous study demonstrated that 3A interacts with vimentin to negatively modulate FMDV replication (3). The 3A also combines with DDX56 to inhibit IRF3 phosphorylation to enhance FMDV replication (4). We previously observed that 3A suppresses the ANXA1 effect to positively promote FMDV replication (5). Sar1 and Sec12 were hijacked by FMDV 3A for endoplasmic reticulum remodeling in a COPII-independ ent manner (6). Although the 3A protein is one of the critical components of FMDV replication, its underlying function remains unclear.
PDCD10 (also named CCM3 or TFAR-15) is a conserved gene from nematode to human; it is also identified as the third causative gene of CCM and is an apoptosis-related gene (7,8). PDCD10 plays essential roles in cell proliferation, protein synthesis, develop mental disorders, and other diseases (9). The well characterization of PDCD10 is based on its function as a component of the STRIPAK (striatin-interacting phosphatase and kinase) complex that interacts with STK24, MST4 (serine/threonine-protein kinase 26), and STK25 GCKIII kinase subfamily (10)(11)(12)(13). PDCD10 overexpression also decreased NF-κB p65 level and the level of inflammatory factors such as TNF-α and IL-1β (14).
Melanoma differentiation-associated protein 5 (MDA5) and retinoic acid-inducible gene I (RIG-I) are RNA sensors that induce antiviral immune response. After the viral RNA genome is recognized by MDA5 or RIG-I, both sensors use their N-terminal Caspase recruitment domain (CARD) to interact with virus-induced signaling adaptor (VISA, also named mitochondrial antiviral-signaling protein, [MAVS]), which recruits signaling kinases such as TBK1, IκB kinase (IKK) family kinases, to activate IRF3 or NF-κB. The activated NF-κB and IRF3 can translocate into the nucleus and interact with specific promoter regions to initiate the type I interferon (IFN) production, which induces the expression of hundreds of IFN-stimulated genes (ISGs), leading to the establishment of the antiviral state (15,16).
Multiple host factors tightly control signaling by RIG-I-like receptors (RLRs) to prevent excessive inflammatory responses. For instance, tripartite motif protein 25 (TRIM25) interacts with RIG-I to facilitate the K172 site ubiquitination of RIG-I CARD 2 (17). A previous study showed that the RIG-I CARD domain was phosphorylated by pro tein kinase C α (PKC-α) and PKC-β to suppress RIG-I-TRIM25 interaction, thus inhib iting RLRs-mediated antiviral responses (18). Another study demonstrated that Atg5-Atg12 suppressed RIG-I-MAVS interaction to modulate RIG-I signaling transduction (19). Restricting the regulation of RLR-mediated antiviral effects is crucial for the host immune response; however, the mechanism of its regulation is not fully understood.
In the present study, we used the luciferase reporter system to screen for host proteins that regulate type I IFN and found that PDCD10 was a candidate for suppress ing viral RNA-induced IFN-β production. We determined that PDCD10 overexpression decreased IFN-β promoter activity to inhibit the expression of ISGs. These results were reversed after PDCD10 knockdown or knockout during Sendai virus (SeV) or vesicular stomatitis virus (VSV) infection. The underlying mechanism was that PDCD10 disrupted the formation of a complex between VISA and IRF3 during viral infection. FMDV titer was increased after PDCD10 overexpression. During the viral infection process, FMDV also uses PDCD10 to further inhibit IFNβ production and promote FMDV replication. The FMDV intact 3A protein cooperated with PDCD10 to inhibit IFN-β production, while 3A also promoted the binding of PDCD10 to VISA. Taken together, our data suggested that PDCD10 inhibited VISA-mediated innate immune responses, while FMDV 3A cooperated with PDCD10 to promote FMDV replication. The present study expanded our knowledge of novel functions of the 3A protein and PDCD10 in the regulation of immune responses against FMDV.
## MATERIALS AND METHODS
## Cells
Porcine kidney 15 (PK15), baby hamster kidney 21 (BHK21), human embryonic kidney 293, Instituto Biologico-Rim Suino-2 (IBSR-2), and Henrietta Lacks (HeLa) cell lines from ATCC were stored in our lab, while PDCD10-knockout HeLa cells were constructed by our lab. Porcine alveolar macrophages (PAMs) were collected from the lung tissue of 3-month-old pigs. PK15, IBSR-2, and BHK21 cells were maintained in modified Eagle's medium (MEM; Gibco); HeLa and HEK293 cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco). DMEM and MEM were supplemented with 10% fetal bovine serum, 100 U/mL penicillin, 100 µg/mL streptomycin, and the cells were cultured at 37°C under 5% CO 2 atmosphere. The cells were free of mycoplasma contamination.
## Viruses
FMDV (FMDV/O/BY/CHA/2010), SeV, Seneca Valley virus (SVA), enterovirus 71 (EV71), and VSV (VSV-GFP/VSV-RFP) were maintained in our lab. African swine fever virus (ASFV) was stored in our institute. VSV-GFP replication was performed in HEK293 cells. SeV was propagated in embryonated chicken eggs. FMDV and SVA replication or titration was performed in BHK21 cells or IBSR-2.
## Reagents
Monoclonal antibodies against PDCD10 (SAB1412804), Flag (F2555), HA (H3663), and myc (SAB2702192) were purchased from Sigma-Aldrich (Merck Ltd [China], Beijing, China). Primary antibodies against IRF3 (#11904), TRAF6 (#67591), TRAF3 (#33640), TBK1 (#38066), p-IRF3 (#37829), p-TBK1 (#5483), and β-actin (#4970) were purchased from Cell Signaling Technology, Co. Inc. Lipofectamine 2000/3000 was obtained from Invivogene Biotech Co., Ltd., and TRIzol was obtained from Invitrogen (Thermo Fisher Scientific [China] Co., Ltd., China). PDCD10 purified protein (HY-P71190) and VISA (Ag5949) were ordered from MedChemexpress (MCE, USA) and Proteintech (Proteintech Group, Inc., USA). Z-VAD-FMK and Dinaciclib were purchased from MedChemExpress (MCE, USA).
## Plasmid construction and transfection
The DNA fragment of PDCD10 was generated by PCR amplification of the total RNA of PK15 cells. The plasmids containing porcine VISA, STAT1, JAK1, or TBK1 were stored in our laboratory. The plasmids of TRAF3, TRAF6, IRF3, and IRF3-5D (Hongbing Shu' lab) were kept in our laboratory. The plasmid of the 3A gene from FMDV/O/BY/CHA/2010 was also stored in our lab. All the expression plasmids were constructed into the plasmid pcDNA3.1 and sequenced in the Beijing Genomics Institute. Lipofectamine 2000 or 3000 was used for transfection according to the manufacturer's instructions.
## Luciferase reporter assay
HEK293 or PK-15 cells (1 × 10 5 ) were cultured in 48-well plates for 24 h. 20 ng of pRL-TK reporter plasmid and 200 ng of IFN-β or interferon-stimulated response element (ISRE) reporter plasmid were transfected into HEK293 cells (IFN-β, ISRE, and pRL-TK reporter plasmids as gifts were provided by Shu Hong Bing's Lab, Wuhan University, China). After 24 h post-transfection, the cells were either not treated or infected with SeV/FMDV for 12 h. The luciferase activity was measured with the Dual-Luciferase Reporter Assay Kit (Promega, USA) according to the manufacturer's protocol.
## Lentivirus production and infection
The target sequence of PDCD10 short hairpin RNA (shRNA) was 5ʹ-CAGGATGTTGAA TGGGATTAT-3ʹ. Annealed PDCD10 shRNA-synthesized cDNA fragments or a negative control shRNA were digested with EcoRI and BamHI and cloned into the pLVX shRNA expression empty vector (HANBIO, Shanghai, China). The shPDCD10 lenti-vector, psPAX2, and pMD2.G were transfected into HEK293 cells. Culture supernatants were harvested at 48 and 72 h; the culture medium was filtered with a 0.45 µm filter and centrifuged at 72,000 × g for 120 min at 4°C. The shRNA knockdown efficiency was assessed by qPCR and western blot analysis.
## RNA extraction and RT-PCR
Total RNA was extracted using TRIzol reagent (Invitrogen, USA). M-MLV reverse transcriptase (Promega, USA) and random hexamer primers (Takara, Japan) were used to prepare cDNA. The generated cDNA was used as a template for FMDV RNA and cellular mRNA host expression. Real-time quantitative PCR (RT-PCR) was performed to measure the abundance of different mRNAs using Mx3005P qPCR (Agilent Technologies, USA) and SYBR Premix ExTaq reagents (TaKaRa, Japan). The data were normalized to GAPDH expression. The 2 -ΔΔCt method was used to calculate the relative expression of mRNA. RT-qPCR primers are listed in Table 1.
## Co-immunoprecipitation assay
The indicated plasmids were transfected into HEK293 cells, and the cells were cultured in 10 cm 2 plates. After 24 h, the cells were collected and lysed in lysis buffer (20 mM Tris [pH 7.5], 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 10 mg/mL aprotinin, 10 mg/mL leupeptin, and 1 mM PMSF). Lysates were incubated with 1 µg of primary antibody or control IgG and 60 µL of G-Sepharose (GE Healthcare, USA) for 6 h. Sepharose beads were washed three times with 1 mL of lysis buffer containing 500 mM NaCl. Immunoblotting was performed to analyze the precipitates.
## Pull-down assay
Purified VISA and PDCD10 were dissolved in cell lysis buffer and then mixed together, and anti-VISA antibody was added and incubated overnight at 4°C. Protein A/G agarose beads were added into the mixture and incubated for 3 h, then centrifuged at 12,000 rpm for 30 sec, the beads were washed with lysis buffer five times (each for 5 min), and then loading buffer was added, boiled for 10 min, and the samples were detected by western blotting with indicated antibody.
## Western blotting
Cell samples were boiled after lysis with a lysis buffer for 30 min. Cell lysates were subjected to SDS-PAGE (Bio-Rad), and the separated proteins were transferred onto PVDF membranes (Millipore). The membranes were blocked in TBST (Tris-buffered saline with 0.1% Tween 20) with 5% fat-free milk for 1 h at room temperature and then incubated with primary antibodies in 5% BSA overnight at 4°C. The membranes were then washed again with TBST five times and finally incubated with horseradish peroxidase-conjugated secondary antibodies in 5% BSA at room temperature for 1 h. The membranes were then imaged by a ChemiDoc MP Imaging System (Bio-Rad, USA).
## Immunofluorescence microscopy
The cells added with Mito-Tracker (100 mM) were incubated at 37°C for 30 min. The indicated cells were washed with cold phosphate-buffered saline (PBS) three times and fixed with 4% paraformaldehyde at room temperature for 15 min. HeLa cells were permeabilized with 0.1% Triton X-100 at room temperature for 5 min and blocked with 10% goat serum and 0.2% Tween-20 in PBS (PBST). Primary and secondary antibodies were solubilized in 10% goat serum in PBST. The cells were incubated with primary antibodies overnight at 4°C. Subsequently, the cells were washed five times and incubated with secondary antibodies for 1 h. The slides were then stained with DAPI for 15 min and imaged with a laser-scanning confocal microscope (LSCM; SP8, Leica, Solms, Germany).
## Statistical analysis
All tests were reproducible, and similar findings were confirmed by repetition at least three times. Sample variation was determined using Tukey's post hoc test and analyzed by one-way ANOVA or two-way ANOVA. Means are represented with histograms, with error bars representing the standard error of the mean (s.e.m.). The P values < 0.05 were considered statistically significant.
## RESULTS
## PDCD10 expression level was increased after viral infection
To determine whether infection with viruses affects PDCD10 expression, the mRNA and protein expression levels of PDCD10 were assessed after viral infection. As shown in Fig. 1A, PDCD10 protein expression was slightly increased in SeV-infected HEK293 cells and VSV-infected HeLa cells and significantly increased in SVA-infected IBRS2 cells and FMDV-infected PK-15 cells. To determine whether picornaviridae of the same family as FMDV also affect PDCD10 expression, HEK293 cells were infected with EV71, we found that EV71 promotes PDCD10 protein expression (Fig. 1A). Furthermore, to confirm whether the conclusion that viruses promote PDCD10 expression extends to DNA viruses, ASFV was used to infect PDCD10, and the results showed that ASFV also increased the PDCD10 protein level (Fig. 1A). Moreover, as shown in Fig. 1B, the PDCD10 mRNA levels in HEK293, IBRS2, PK-15, or HeLa cells were increased after SeV, VSV, SVA, or FMDV infection. These results suggested that PDCD10 expression was associated with viral infection.
## PDCD10 as a new candidate to inhibit the IFN-β production
IFN-β production is induced through the RLRs signaling pathway during RNA virus infection. To elucidate the regulators that affect IFN-I production, we assessed the IFN-β secretion with enzyme-linked immunosorbent assay (ELISA), and we found that PDCD10 inhibits IFN-β secretion during SeV or VSV infection (Fig. 2A; Fig. S1A). Meanwhile, we performed a luciferase assay to accurately identify the proteins that inhibit the IFN-β promoter. The results of the luciferase assay revealed that PDCD10 could inhibit SeV-induced IFN-β and ISRE promoter activation (Fig. 2A). Poly (I: C) could simulate the RNA virus genome to induce the IFN-I production. We observed that PDCD10 overexpression inhibited IFN-β and ISRE activation after poly (I: C) transfection (Fig. 2B). IFN-β promoter activation was inhibited by PDCD10 in a dose-dependent manner in HEK293 cells after virus infection (Fig. 2C). Phosphorylated IRF3 translocates into the nucleus to induce IFN-β production. We found that PDCD10 overexpression suppressed SeV-induced phosphorylation of IRF3 (Fig. 2D). The expression of ISGs is associated with IFN-I production. PDCD10 was transfected into HEK293 cells, and after 24 h, the cells were infected with SeV or VSV for 12 h. As shown in Fig. 2E and Fig. S1B, PDCD10 antagonized the mRNA levels of IFNB, ISG56, RANTES, and CXCL10. PDCD10 also suppressed the mRNA expression of IFNB, ISG56, RANTES, and CXCL10 after poly (I: C) stimulation (Fig. 2F). These results indicated that PDCD10 was an inhibitor of SeV-triggered IFN-β activation.
## PDCD10 deficiency promoted SeV-induced IFN-I production
To examine the effect of PDCD10 on SeV-induced innate immune response, we used short hairpin RNAs (shRNA) to knock down PDCD10 (Fig. 3A). #3 PDCD10 shRNA overexpressing HEK293 cells were infected with SeV. ELISA results showed that IFN-β secretion was increased in #3 PDCD10 shRNA expressing cells after SeV or VSV infec tion (Fig. 3A; Fig. S1C). As shown in Fig. 3B, the results of the luciferase assay showed that compared to PDCD10-negative control (NC) shRNA, IFN-β promoter activation was increased during PDCD10 knockdown after SeV infection. Consistent with this finding, IFN-β promoter activation was enhanced in PDCD10-knockdown HEK293 cells after poly (I: C) stimulation (Fig. 3C). Next, we examined the phosphorylated IRF3 level in PDCD10-knockdown HEK293 cells during SeV infection. The results showed that IRF3 phosphorylation level was increased after PDCD10 knockdown during SeV infection (Fig. 3D). We also noted that the mRNA levels of IFNB, ISG56, RANTES, and CXCL10 in PDCD10-knockdown cells were increased after viral infection (Fig. 3E).
To further confirm the effect of PDCD10 on innate immune response, we assessed IFN-β promoter activation in PDCD10-knockout cells (PDCD10 KO cells, PDCD10 -/-) during viral infection. As shown in Fig. 3F, PDCD10 expression could not be detected in PDCD10 -/-cells. To verify the IFN-β production in PDCD10 -/-cells, luciferase assay, Western blotting, and qPCR were performed. We found that IFN-β promoter activation was increased after SeV infection or poly (I: C) transfection in PDCD10 -/-HeLa cells (Fig. 3G). PDCD10 deficiency promoted IRF3 phosphorylation (Fig. 3H; Fig. S1D) and enhanced IFNB, ISG56, RANTES, and CXCL10 mRNA expression (Fig. 3I; Fig. S1E). PDCD10 -/- cells were transfected with poly (I: C), and the results (Fig. 3J) showed that PDCD10 deficiency increased IFNB, ISG56, RANTES, and CXCL10 mRNA transcription as compared to PDCD10 +/+ cells (PDCD10 wild-type cells, PDCD10 +/+ ). We transfected PDCD10 into PDCD10 -/-cells and found that the effect on ISGs mRNA transcription was reversed after poly (I: C) stimulation. Our findings establish that PDCD10 deficiency promoted SeV-induced IFN-β activation.
## PDCD10 interacted and co-localized with VISA
To investigate the molecular mechanisms by which PDCD10 influences the innate immune response, we investigated whether PDCD10 was involved in RLR componentmediated IFN-I signaling transduction. First, PDCD10 and RLRs signaling components were transfected separately into HEK293 cells, including RIG-I, MDA5, VISA, TBK1, and IRF3-5D (a constantly activating mutant). As shown in Fig. 4A, we found that RIG-I-, MDA5-, and IRF3-5D-induced IFN-β promoter activation was inhibited by PDCD10. However, PDCD10 had no effect on VISA-and TBK1-induced IFN-β production, thus suggesting that PDCD10 may affect IFN-β production at the VISA or TBK1 level. The co-immunoprecipitation assay (Co-IP) and pull-down assay results showed that PDCD10 interacts with VISA (Fig. 4B andC). Because VISA is predominantly located in mitochondria, we determined whether PDCD10 co-localized with VISA. As shown in Fig. 4D, PDCD10 was co-localized with VISA on mitochondria. These results suggested that PDCD10 interacts with VISA to inhibit the innate immune response.
## PDCD10 inhibits IFN-I production by disrupting the VISA-IRF3 complex
Because PDCD10 binds to VISA, to reveal the mechanism by which PDCD10 interacts with VISA to inhibit IFN-β production, we hypothesized that PDCD10 may affect the function of the VISA-associated complex that acts as a signaling transduction platform. For this purpose, VISA, PDCD10, and RIG-I, TRAF3, TRAF6, TBK1, or IRF3 were transfected into HEK293 cells. After 24 h, the cells were infected with SeV and then lysed. Anti-VISA antibody was used as the IP primary antibody, and competitive Co-IP experiments were performed for the indicated proteins. The results indicated that compared to control, PDCD10 expression inhibited VISA-IRF3 interaction, but not RIG-I, TRAF3, TRAF6, or TBK1 complex (Fig. 5A through E). Furthermore, PDCD10 was transfected into HEK293 cells at different doses, as shown in Fig. 5F; PDCD10 inhibited VISA-IRF3 interaction in a dose-dependent manner during SeV infection.
## PDCD10 promotes FMDV replication
Next, we assessed the effect of PDCD10 on FMDV replication. PDCD10 was transfected into PK15 cells, and the FMDV titers were determined. As shown in Fig. 6A, PDCD10 overexpression increased FMDV mRNA expression and viral titer as compared to those of the control. Additionally, as shown in Fig. 6B, the protein levels of VP0, VP1, and VP3 of FMDV were increased after PDCD10 overexpression. PDCD10 shRNA-overexpressing PK15 cells were infected with FMDV, and the results (Fig. 6C) showed that knockdown of PDCD10 decreased FMDV titer. To investigate the effect of IFN-β on FMDV replication, PK15 cells were treated with porcine IFN-β after FMDV infection. The result showed that IFN-β suppressed FMDV replication (Fig. 6D). To confirm whether PDCD10 affected FMDV replication in an IFN-β-dependent manner, PDCD10 was transfected into IFN-β-deficient BHK21 cells. After 24 h, the cells were infected with FMDV for the indicated time period. We observed that the IFN-β secretion level was not affected (Fig. 5E). PDCD10 or PDCD10 shRNA overexpression did not affect the FMDV replication level in BHK21 cells (Fig. 6F andG). In addition, to determine that the promotion of FMDV replication by PDCD10 correlates with the inhibition of IFNβ production, PK15 cells were transfected with PDCD10, and the cells were then infected with FMDV for 12 h. Subsequently, the medium was collected, and UV light was used to inactivate the FMDV. PK15 cells were pre-treated with the inactivated FMDV medium, and the FMDV titer was assayed. The results showed that in IFN-producing normal PK-15 cells, PDCD10 promoted FMDV replication after treatment with the FMDV inactivated supernatant as compared to control (Fig. 6H). Additionally, PDCD10 knockdown inhibited FMDV replication as compared to EV control after treatment with the inactivated FMDV supernatant (Fig. 6I).
To assess the effect of PDCD10 on antiviral status, PK15 cells were transfected with PDCD10, and the cells were then infected with FMDV for 12 h. Subsequently, the medium was collected, and UV light was used to inactivate the FMDV. PK15 cells were pre-treated with the inactivated FMDV medium, and the cells were then infected with expression. These results suggested that PDCD10-inhibited IFN-β production may affect FMDV replication.
## FMDV 3A cooperated with PDCD10 to inhibit I-IFN production
The effect of FMDV on PDCD10-suppressed IFN-β production was assessed. As shown in Fig. 7A, compared to EV, PDCD10 inhibited FMDV-induced IFN-β promoter activation as determined by luciferase detection. We used low virus titer and high virus titer to investigate the effect of FMDV on PDCD10-mediated innate immune response. We found that using a high titer of FMDV strengthens PDCD10-mediated inhibition of IFN-β promoter activation (Fig. 7A). This finding suggested that FMDV components might be involved in PDCD10-mediated IFNβ inhibition.
To investigate the involvement of FMDV components in PDCD10-mediated innate immune response, different FMDV proteins were separately transfected into PK15 cells, and immunoprecipitation mass spectrometry was performed. The data showed that PDCD10 interacted with 3A (Fig. 7B). To further confirm that 3A interacted with PDCD10, 3A and PDCD10 were transfected into HEK293 cells. We performed a Co-IP assay and observed that 3A interacted with PDCD10 (Fig. 7C). In addition, compared to PDCD10 expression alone, we found that the combination of 3A and PDCD10 showed a higher inhibition effect on IFN-β promoter activation and the mRNA transcription expression of IFNB, ISG54, ISG20, and OAS1 (Fig. 7D andE).
Taken together, the above-mentioned results suggested that 3A cooperated with PDCD10 to inhibit IFN-I production.
## Full length of FMDV 3A interacted with PDCD10 to promote FMDV replication
We co-expressed 3A deletion mutants with PDCD10 to determine the 3A domains responsible for PDCD10 binding. As shown in Fig. 8A, the full-length 3A protein bound to PDCD10, thus suggesting that PDCD10 may interact with the conformational structure of 3A. PK15 cells were transfected separately with 3A, PDCD10, or 3A plus PDCD10. After 24 h, PK15 cells were infected with FMDV. We noted that the promoting effect of 3A plus PDCD10 on FMDV replication was more intensive than 3A or PDCD10 expression alone (Fig. 8B). Next, PK15 cells were transfected with PDCD10 shRNA and 3A and then infected with FMDV. Compared to PDCD10-negative shRNA control, the FMDV titer was decreased after 3A transfection in PDCD10-interfering cells (Fig. 8C). Moreover, co-localization assays were performed to confirm the association between 3A and PDCD10. HeLa cells were transfected with 3A-GFP and PDCD10. As shown in Fig. 8D, either 3A or PDCD10 localized to the cytoplasm and plasma membrane with some highly concentrated spots around the nucleus. The distribution of 3A overlapped with that of PDCD10 (pcc = 0.789). To investigate the mechanism by which 3A cooperated with PDCD10 to promote FMDV, 3A, PDCD10, and VISA were transfected into HEK293 cells, and co-IP results showed that 3A promoted the formation of the VISA-PDCD10 complex (Fig. 8E andF). These results suggested that FMDV replication was promoted by interaction between the full length of 3A protein and PDCD10.
## DISCUSSION
FMDV infection triggers host innate immune response, such as production of IFNs and induction of inflammatory cytokines (20)(21)(22). An increase in IFN-β expression is one of the important approaches to defend against foreign viral infection. RLRs-mediated IFN-β induction is well investigated.
FMDV has a single positive-strand RNA genome. The RNA genome of the virus is recognized by RLRs, which bind to VISA to recruit TBK1. Eventually, the phosphorylated TBK1 activates IRF3 to translocate into the nucleus and induce IFN-I production (23). In the present study, we demonstrated that PDCD10 acted as a candidate repressor for SeV-induced RLR signaling transduction. PDCD10 belongs to the PDCD gene family, which is ubiquitously expressed and conserved (9). In the past decade, several studies have shown that PDCD10 acts as an essential regulator in vasculogenesis, autophagy, and apoptosis (7,24,25). Recent studies have revealed that PDCD10 is associated with immune response, such as B-cell depletion halting the maturation and progression of already formed ectatic blood vessels into multicavernous clinically significant lesions in PDCD10 +/-murine models (26). PDCD10 also has a regulatory effect on other immunerelated genes to improve immune response and suppress the development of inflammation according to the Bayesian gene regulatory network (27). Innate immune response is the first line of defense against foreign pathogens; however, information regarding the influence of PDCD10 on immune responses remains limited. Our present study is the first to identify that PDCD10 altered RNA virus-induced IFN-β production to promote FMDV replication. By using PDCD10 knockdown and knockout models, we demonstrated that PDCD10 negatively regulated RNA virus-induced host antiviral defense. Mechanistically, through direct interaction with VISA, PDCD10 blocked VISA-mediated RLRs signaling transduc tion. VISA is a central platform for RNA nucleic acid sensing. We found that PDCD10 inhibited the formation of the VISA-IRF3 complex. Extensive studies have shown that host proteins play a role in strictly regulating RLRs signaling transduction, for exam ple, lactate directly interacts with VISA trans-membrane domain (TM) to prevent its aggregation (28); UBXN1 interacts with VISA to disrupt its oligomerization and the MAVS-TRAF3-TRAF6 complex formation during viral infection (29). VISA acts as an adapter for activating distinct signaling pathways that lead to IRF3 and NF-κB activation. Siqi Liu et al. found that during virus infection, the phosphorylated MAVS binds to IRF3 and recruits IRF3 for its phosphorylation and activation (30). VISA is associated with TRAF motifs of TRAF2 and TRAF6 to facilitate RIP and NEMO K63-linked poly-ubiq uitination, which leads to NF-κB activation (31,32). VISA interacts with TRAF3, which is essential in virus-induced IRF3 activation (33,34). Our results showed that PDCD10 did not disrupt VISA-TRAF6 interaction; thus, we speculated that PDCD10 might not affect the NF-κB signaling transduction during the RNA virus infection. Phosphorylated VISA interacts with IRF3 to induce IFNβ activation (30). In our present study, we found that PDCD10 inhibited the formation of VISA-IRF3 complex after SeV infection; however, the underlying mechanism requires further research. A previous study has shown that MST1 was associated with TBK1 and IRF3 to impede IRF3 phosphorylation to shut off the antiviral response of the RIG-I-MAVS-IRF3 axis (35). Several studies have shown that in the multicomponent complexes of STRIPAK, the STRN family proteins (STRNs) share a conserved PDCD10-binding region to directly interact with PDCD10, which acts as a protein bridge recruiting GCKIII kinases to STRNs (36)(37)(38)(39). GCKII kinases MST1/2 are associated with IRF3 activation (35). Therefore, we speculated that PDCD10 might affect VISA-IRF3 antiviral response through STRIPAK. This, however, is only a hypothesis that requires further confirmation for the precise mechanism. Next, we attempted to discover the mechanism by which FMDV replication is promoted by PDCD10. PDCD10 was overexpressed in IFN-β-deficient BHK21 cells, and the cells were infected with FMDV. We observed that the FMDV titer in the PDCD10-transfected cells did not increase as compared to EV. These results were further confirmed in PDCD10-knockdown BHK21 cells; this suggested that PDCD10 may promote the replication of FMDV by affecting IFN-β production. These results reinforced the hypothesis that PDCD10 was involved in viral-induced innate immune response.
In the process of evolution, viruses may take advantage of host proteins to coun ter the innate immune response, thereby creating a favorable environment for their replication. For example, H7N9 PB1-F2 induces lysosomal and proteasomal degradation of aggregated VISA (40); the NS1 protein of influenza A viruses targets TRIM25 to prevent ubiquitination of RIG-I, which blocks IFN production (41). FMDV 3A also cooperates with DDX56 to inhibit IRF3 phosphorylation (4). The 3A protein of FMDV plays a critical role in constructing the viral replication complex, which is important for replication. However, the precise mechanisms by which 3A affects FMDV replication remain incompletely understood. In the present study, we found that the full length of the 3A protein interacted and co-localized with PDCD10 to reduce FMDV-induced IFN-I production. A previous study showed that 3A anchors to intracellular membranes to form the replication complex (42). Hence, we speculated that 3A may also anchor to intracellular membranes, such as the mitochondrial membrane, together with PDCD10 to potentiate the inhibition of VISA-mediated innate immune response; further studies are required to confirm this speculation. A previous study showed that FMDV 3A also interacts with VISA (43). In the present study, we found that PDCD10 was associated with 3A and VISA; thus, we speculated that PDCD10-3A, together with VISA, formed a complex to inhibit VISA signaling transduction. This requires further study for confirmation.
Collectively, our results showed that after SeV infection, PDCD10 bound to VISA to disrupt the formation of VISA-IRF3 complex, which suppressed IFN-β activation; PDCD10 also cooperated with the 3A protein to promote FMDV replication (Fig. 9). This study provided new insights into the mechanism by which PDCD10 affected viral-induced IFN-β activation and revealed that FMDV 3A cooperated with PDCD10 to strengthen the inhibition effect on IFN-I production.
## References
1. Rodríguez-Habibe, Celis-Giraldo, Patarroyo et al. (2020) "A comprehensive review of the immunological response against foot-and-mouth disease virus infection and its evasion mechanisms" *Vaccines (Basel)*
2. Grubman, Baxt (2004) "Foot-and-mouth disease" *Clin Microbiol Rev*
3. Ma, Ling, Li et al. (2020) "Cellular vimentin interacts with foot-and-mouth disease virus nonstructural protein 3A and negatively modulates viral replication" *J Virol*
4. Fu, Yang, Ru et al. (2019) "DDX56 cooperates with FMDV 3A to enhance FMDV replication by inhibiting the phosphorylation of IRF3" *Cell Signal*
5. Ma, Zhang, Luo et al. (2022) "FMDV 3A antagonizes the effect of ANXA1 to positively modulate viral replication" *J Virol*
6. Lee, Jiang, Chang et al. (2022) "Foot-and-mouth disease virus 3A hijacks Sar1 and Sec12 for ER remodeling in a COPII-independ ent manner" *Viruses*
7. Bergametti, Denier, Labauge et al. (2005) "Mutations within the programmed cell death 10 gene cause cerebral cavernous malformations" *Am J Hum Genet*
8. Busch, Heath, Hubberstey (2004) "Sensitive genetic biomarkers for determining apoptosis in the brown bullhead (Ameiurus nebulosus)" *Gene*
9. Guan, Lu, Sun et al. (2019) "The molecular evolution and functional divergence of lamprey programmed cell death genes" *Front Immunol*
10. Madsen, Hooper, Tozluoglu et al. (2015) "STRIPAK components determine mode of cancer cell migration and metastasis" *Nat Cell Biol*
11. (2025) *Full-Length Text Journal of Virology*
12. Kück, Radchenko, Teichert (2019) "STRIPAK, a highly conserved signaling complex, controls multiple eukaryotic cellular and develop mental processes and is linked with human diseases" *Biol Chem*
13. Ceccarelli, Laister, Mulligan et al. (2011) "CCM3/PDCD10 heterodimerizes with germinal center kinase III (GCKIII) proteins using a mechanism analogous to CCM3 homodimerization" *J Biol Chem*
14. Fidalgo, Fraile, Pires et al. (2010) "CCM3/ PDCD10 stabilizes GCKIII proteins to promote Golgi assembly and cell orientation" *J Cell Sci*
15. Peng, Wu, Feng et al. (2020) "Cerebral cavernous malformation 3 relieves subarachnoid hemorrhage-induced neuroinflammation in rats through inhibiting NF-kB signaling pathway" *Brain Res Bull*
16. Yoneyama, Kikuchi, Natsukawa et al. (2004) "The RNA helicase RIG-I has an essential function in double-stranded RNA-induced innate antiviral responses" *Nat Immunol*
17. Onoguchi, Yoneyama, Fujita (2011) "Retinoic acid-inducible gene-Ilike receptors" *J Interferon Cytokine Res*
18. Gack, Kirchhofer, Shin et al. (2008) "Roles of RIG-I N-terminal tandem CARD and splice variant in TRIM25-mediated antiviral signal transduction" *Proc Natl Acad Sci*
19. Maharaj, Wies, Stoll et al. (2012) "Conventional protein kinase C-α (PKC-α) and PKC-β negatively regulate RIG-I antiviral signal transduction" *J Virol*
20. Jounai, Takeshita, Kobiyama et al. (2007) "The Atg5 Atg12 conjugate associates with innate antiviral immune responses" *Proc Natl Acad Sci*
21. Langellotti, Quattrocchi, Alvarez et al. (2012) "Foot-and-mouth disease virus causes a decrease in spleen dendritic cells and the early release of IFN-α in the plasma of mice. Differences between infectious and inactivated virus" *Antiviral Res*
22. Pacheco, Smoliga, Donnell et al. (2015) "Persistent foot-and-mouth disease virus infection in the nasopharynx of cattle; tissue-specific distribution and local cytokine expression" *PLoS One*
23. Segundo, Weiss, Pérez-Martín et al. (2012) "Inoculation of swine with foot-and-mouth disease SAP-mutant virus induces early protection against disease" *J Virol*
24. Yoneyama, Onomoto, Jogi et al. (2015) "Viral RNA detection by RIG-I-like receptors" *Curr Opin Immunol*
25. Zhu, Wu, Xu et al. (2010) "Differential angiogenesis function of CCM2 and CCM3 in cerebral cavernous malformations" *Neurosurg Focus*
26. You, Sandalcioglu, Dammann et al. (2013) "Loss of CCM3 impairs DLL4-Notch signalling: implication in endothelial angiogenesis and in inherited cerebral cavernous malformations" *J Cell Mol Med*
27. Shi, Shenkar, Zeineddine et al. (2016) "B-cell depletion reduces the maturation of cerebral cavernous malformations in murine models" *J Neuroimmune Pharmacol*
28. Behdani, Ghaderi-Zefrehei, Rafeie et al. (1748) "RNA-Seq bayesian network exploration of immune system in bovine" *Iran J Biotechnol*
29. Zhang, Wang, Xu et al. (2019) "Lactate is a natural suppressor of RLR signaling by targeting MAVS" *Cell*
30. Wang, Yang, Cheng et al. (2013) "UBXN1 interferes with Rig-I-like receptor-mediated antiviral immune response by targeting MAVS" *Cell Rep*
31. Liu, Cai, Wu et al. (2015) "Phosphorylation of innate immune adaptor proteins MAVS, STING, and TRIF induces IRF3 activation" *Science*
32. Xu, Wang, Han et al. (2005) "VISA is an adapter protein required for virus-triggered IFN-β signaling" *Mol Cell*
33. Zhao, Yang, Sun et al. (2007) "The NEMO adaptor bridges the nuclear factor-κB and interferon regulatory factor signaling pathways" *Nat Immunol*
34. Saha, Pietras, He et al. (2006) "Regulation of antiviral responses by a direct and specific interaction between TRAF3 and Cardif" *EMBO J*
35. Häcker, Redecke, Blagoev et al. (2006) "Specificity in Toll-like receptor signalling through distinct effector functions of TRAF3 and TRAF6" *Nature*
36. Meng, Zhou, Wu et al. (2016) "Mst1 shuts off cytosolic antiviral defense through IRF3 phosphorylation" *Genes Dev*
37. Glatter, Wepf, Aebersold et al. (2009) "An integrated workflow for charting the human interaction proteome: insights into the PP2A system" *Mol Syst Biol*
38. Goudreault, Ambrosio, Kean et al. (2009) "A PP2A phosphatase high density interaction network identifies a novel striatin-interacting phosphatase and kinase complex linked to the cerebral cavernous malformation 3 (CCM3) protein" *Mol Cell Proteomics*
39. Ribeiro, Josué, Wepf et al. (2010) "Combined functional genomic and proteomic approaches identify a PP2A complex as a negative regulator of Hippo signaling" *Mol Cell*
40. Kean, Ceccarelli, Goudreault et al. (2011) "Structure-function analysis of core STRIPAK proteins: a signaling complex implicated in Golgi polarization" *J Biol Chem*
41. Cheung, Lee, Kew et al. (2020) "Virus subtype-specific suppression of MAVS aggregation and activation by PB1-F2 protein of influenza A" *PLoS Pathog*
42. Gack, Albrecht, Urano et al. (2009) "Influenza A virus NS1 targets the ubiquitin ligase TRIM25 to evade recognition by the host viral RNA sensor RIG-I" *Cell Host Microbe*
43. O'donnell, Pacheco, Henry et al. (2001) "Subcellular distribution of the foot-and-mouth disease virus 3A protein in cells infected with viruses encoding wild-type and bovine-attenuated forms of 3A" *Virology (Auckl)*
44. Li, Lei, Xu et al. (2016) "Foot-and-mouth disease virus non-structural protein 3A inhibits the interferon-β signaling pathway" *Sci Rep*
45. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12624634&blobtype=pdf | Amir Razavinia, Abazar Razavinia, Roya Jamshidi, Khalife Lou, Mahlegha Ghavami, Forouzan Shahri, Aida Tafazoli, Bahman Khalesi, Zahra Hashemi, Saeed Khalili
## Abstract
In this article, the authors omitted disclosing that a Generative AI Source had been used in the writing process of the paper. The authors have confirmed that all scientific content, analysis, and conclusions were developed independently by the author group, and that the AI was used solely to correct grammar and improve language clarity. To ensure clarity, the article has been updated to include this statement.The authors have provided a statement to be added to the declerations at the end of the article informing that Generative AI sources were used in the writing process of the paper to correct grammar and improve langauge clarity.The correct version of the decleration should be as below: "Decleration of Generative AI During the preparation of this work, the authors used ChatGPT, Grammarly, QuillBot, DeepL Write, and Microsoft Editor solely for grammar correction, spelling, style improvement, and language refinement. These tools were applied under full human oversight and control. The authors reviewed, edited, and approved all suggestions, and the scientific content, analysis, and conclusions were entirely generated by the authors. The use of these tools was primarily to improve English quality, as the authors are non-native English speakers, and did not affect the scholarly integrity of the work. The authors take full responsibility for the final content of the published article."The authors apologize for the errors. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12608102&blobtype=pdf | # Tiger deaths in Vietnam due to infection with H5N1 highly pathogenic avian influenza virus bearing mutations associated with mammalian host adaptation
Murasaki Amano, Thi Nguyen, Nga, Nguyen Le Khanh Hang, Nguyen Dang Tho, Diep, Thi Dam, Vui, Futoshi Hasebe, Haruka Abe, Le Thi, Quynh Mai
## Abstract
Recently, infections with the H5N1 subtype of highly pathogenic avian influenza virus (H5N1-HPAIV) in mammals have been reported worldwide, including in cows in the United States and successive human cases in Cambodia. In Vietnam, 47 tigers and three leopards died from H5N1-HPAIV infection between August and October 2024. This study aimed to determine the origin of the H5N1 strains that infect tigers in Vietnam and to identify specific mutations associated with mammalian infection. Specimens were collected from tigers that died of suspected H5N1-HPAIV infection in southern Vietnam in September-October 2024. RNA was extracted and subjected to whole-genome sequencing. Time-stamped phylogenetic analysis was performed using H5N1 sequences recently detected in Vietnam and neighbouring countries. Phylogenetic results revealed that the strain found in tigers (Tiger H5N1 strain) belonged to clade 2.3.2.1e and has been genetically close to the H5N1-HPAIV lineage responsible for ongoing human infections in Cambodia since 2023. Tiger H5N1 strains harboured several amino acid substitutions associated with mammalian host adaptation or transmissibility, such as E627K in polymerase basic protein 2, similar to the Cambodian human H5N1 strains. This mammalian-adapted H5N1 lineage should be continuously monitored in poultry and mammals, including humans, in Vietnam to prevent further transmission.
## Main text
Avian influenza virus (AIV) primarily resides in waterfowl and can be transmitted to mammals, including dairy cows, cats, and humans [1], causing a substantial disease burden and economic losses. From October 2024 to May 2025, 2,694 highly pathogenic avian influenza (HPAI) outbreaks were reported in poultry and wild birds worldwide, including 12 mammalian transmissions [2]. Transmission typically occurs from birds to mammals; however, virus-host dynamics have shifted markedly in recent years. For instance, more than 19,000 sea lions reportedly succumbed to H5N1 highly pathogenic AIV (H5N1-HPAIV) infections in 2023 in Peru and Chile [3]. In Cambodia, continuous human H5N1-HPAIV infections have been reported since February 2023, with 27 cases and 12 deaths (mortality rate: 44.4%) as of July 5, 2025 [4].
In Vietnam, a 21-year-old university student died of H5N1-HPAIV infection in Khanh Hoa Province in March 2024 [5]. During poultry surveillance conducted in Vietnam from 2017 to 2022, an average of 3.6% of samples tested positive for the H5 subtype of AIV. In 2022 alone, this rate rose to 5.8%, indicating an increase in poultry infections in recent years [6]. Clade 2.3.2.1e of H5N1-HPAIV has predominantly circulated in southern Vietnam, whereas clade 2.3.4.4b strains have been prevalent nationwide since 2021 [7]. In October 2024, it was reported that at least 50 mammals (47 tigers and three leopards) had died of H5N1-HPAIV infection in Vietnam [8]. Captive mammals are rarely infected with H5N1-HPAIV, and when infected, typically die within a short period. This study aimed to determine the origin of the H5N1 strains that infected tigers in Vietnam (Tiger H5N1 strains) through phylogenetic analysis and to identify any mutations associated with mass mammalian mortality.
Intestinal samples were collected from tigers that died of suspected H5N1-HPAIV infection in the Long An and Dong Nai provinces of Vietnam in September -October 2024 (Supplementary Figure S1). The tiger samples were transported to the National Institute of Hygiene and Epidemiology and National Center for Veterinary Diagnosis and whole-genome sequencing was performed. Time-stamped and maximum-likelihood trees were inferred using complete open reading frame (ORF) sequences. Detailed methods are available in the Supplementary Material.
BLAST analysis of the consensus sequences obtained in this study revealed high homology to previous H5N1 strains detected in Vietnam and neighbouring countries, such as Laos and Cambodia (Supplementary Table S1). Accordingly, time-stamped phylogenetic trees were constructed for the haemagglutinin (HA) and polymerase basic protein 2 (PB2) genes using reference sequences from neighbouring countries to clarify the origin of the H5N1 strain detected in tigers in Vietnam (Tiger H5N1 strain). These strains belonged to clade 2.3.2.1e and were genetically similar to the H5N1-HPAIV lineage, which has caused ongoing human infections in Cambodia since 2023 (Figure 1A). It was inferred that the H5N1-HPAIV lineage causing human infections in Cambodia (Cambodian human H5N1 strain) entered Vietnam at the end of 2023 and spread among poultry in the first half of 2024, subsequently infecting tigers from August to October 2024 (Figure 1A). To validate the time-stamped analysis, a phylogenetic tree was constructed using the HA ORF sequences of all H5N1 strains detected in Vietnam, Cambodia, and Laos from 2020 to the present, and the topology of both phylogenetic trees was found to be consistent (Supplementary Figure S2). Similar results were confirmed for PB2 and all other segments compared to the HA segment (Figure 1B, Supplementary Figure S3A-S3F). Cambodian human H5N1 strains harboured several amino acid substitutions associated with mammalian host adaptation or transmissibility in mammals, such as N158D and T160A in HA, and E627 K in PB2, which persisted in strains that infected poultry and tigers in Vietnam (Figure 1A) [9,10]. HA proteins of the Tiger and Cambodian human H5N1 strains contained PQRERRRKR↓GLF with multiple basic amino acids at the cleavage site, indicating high pathogenicity to poultry and mammals. The neuraminidase (NA) protein of the Tiger H5N1 strain possessed no amino acid substitutions associated with NA inhibitor resistance (Supplementary Table S2) [11]. In contrast, the Tiger H5N1 strain harboured unique amino acid substitutions at positions 188, 189, and 242 in HA and 555 in PB2 (Figure 1A and 1B, Supplementary Table S2). These findings suggest that the Tiger H5N1 strains originally harboured mammalian-adaptive amino acid substitutions and that the lineage evolved uniquely during the outbreak among domestic poultry and tigers in Vietnam.
## Discussion
In this study, the Tiger H5N1 viruses were classified as clade 2.3.2.1e and were found to be closely related to the lineage that has caused human infections in Cambodia in recent years. In fact, the Tiger H5N1 strains and human-associated Cambodian strains possessed amino acid substitutions related to mammalian adaptation (E627 K in PB2) and transmission (N158D and T160A in HA), which are characteristic of mammalian infections. Human infections with H5N1 clade 2.3.2.1e viruses in Cambodia showed a high mortality rate of 44.4% (12 deaths in 27 cases) [4], whereas a total of 70 human infections with H5N1 clade 2.3.4.4b strains were reported in the US during the period spanning from 2024 to mid-February 2025 [12]. Of these, only a single case of fatality was observed in Louisiana in January 2025. There have been no subsequent reports of cases in the US since March 2025 [13]. The mammalian-adapted H5N1 clade 2.3.2.1e lineage may cause human infections in Vietnam soon, as it has been spreading in southern Vietnam and should be continuously monitored to protect public health.
Phylogenetic analysis of the HA gene of all strains detected in Vietnam and neighbouring countries from 2020 to the present showed that most H5N1 strains prevalent in Vietnam in recent years belong to clade 2.3.4.4b, whereas only a small number of clade 2.3.2.1e strains were identified, including the Tiger H5N1 strains, possibly introduced from Cambodia or Laos (Supplementary Figure S2). H5N1-HPAIV of clade 2.3.4.4b also remains prevalent in poultry in Vietnam. This clade has raised significant concerns due to its widespread infection in dairy cattle and livestock workers in the United States and should be closely monitored, although it differs from the Tiger H5N1 strains.
The most recent report investigating the lineage and distribution of the H5N1 influenza virus in Vietnam was published in 2019 [14]. This report indicated that clade 2.3.2.1 was distributed in central and southern Vietnam, while clade 2.3.4.4 was predominantly found in northern Vietnam [14]. However, since 2020, clade 2.3.4.4b has rapidly spread worldwide, resulting in avian influenza outbreaks reported in 84 countries between 2022 and 2023 [15]. In Vietnam, clade 2.3.4.4b has also expanded nationwide since 2020, and as of 2025, clade 2.3.4.4b has become the majority within the country (Supplementary Figure S2). At present, clades 2.3.2.1e and 2.3.4.4b are found in the same geographical area in southern Vietnam. Notably, the identification of clade 2.3.2.1e viruses with mammalian adaptation potential in this study has given rise to concerns regarding the acquisition of mammalian adaptation capacity through genetic reassortment. Acquisition of mammalian adaptation capacity by a highly infectious lineage such as clade 2.3.4.4b could potentially trigger the next pandemic. Consequently, it should be essential to carefully monitor the emergence of reassortants in Vietnam over the forthcoming decade.
## References
1. Butt, Nooruzzaman, Covaleda (2024) "Hot topic: influenza A H5N1 virus exhibits a broad host range, including dairy cows" *JDS Commun*
2. (2025) "Highly pathogenic avian influenza -situation report"
3. Plaza, Gamarra-Toledo, Euguí (2024) "Pacific and Atlantic sea lion mortality caused by highly pathogenic avian influenza A(H5N1) in South America" *Travel Med Infect Dis*
4. (2025) "Avian Influenza A (H5N1)-Cambodia"
5. (2024) "Avian Influenza A(H5N1) -Viet Nam"
6. Nguyen, Sumner, Nguyen (2023) "Avian influenza A(H5) virus circulation in live bird markets in Vietnam, 2017-2022. Influenza Other Respir Viruses"
7. Guardian (2024) "Bird flu outbreak kills dozens of tigers in Vietnam zoos. London: Guardian Media Group"
8. Neumann (2015) "H5n1 influenza virulence, pathogenicity and transmissibility: what do we know?" *Future Virol*
9. Xie, Yang, Jiao (2025) "Clade 2.3.4.4b highly pathogenic avian influenza H5N1 viruses: knowns, unknowns, and challenges" *J Virol*
10. Xu, Luo, Huang (2024) "Influenza neuraminidase mutations and resistance to neuraminidase inhibitors" *Emerg Microbes Infect*
11. (2025) "Global Summary of Recent Human Cases of H5N1 Bird Flu. First H5 bird flu death reported in United States"
12. (2025) "Avian Influenza (Bird Flu). H5 Bird Flu: Current Situation"
13. Nguyen, Firestone, Stevenson (2019) "A systematic study towards evolutionary and epidemiological dynamics of currently predominant H5 highly pathogenic avian influenza viruses in Vietnam" *Sci Rep*
14. (2023) "Avian Influenza (Bird Flu)" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12563749&blobtype=pdf | # HIV-1 Subtype Diversity in Morocco: Signals of Change and Implications for National Surveillance
Maryam Ahmina, Hicham El Annaz, Nada Lamrak, Ahmed Reggad, Mohamed Tagajdid, Rachid Abi, Mohamed Elqatni, Abdelilah Laraqui, Safae Elkochri, Elarbi Bouaiti, Youssef Aadi, Bouchra Mchichi, Nadia Touil, Khalid Ennibi, Idriss Lahlou
## Abstract
Background: Limited molecular surveillance continues to constrain Morocco's HIV response, leaving subtype dynamics largely underreported. Once characterized by a predominance of subtype B, the Moroccan epidemic now appears to reflect shifting patterns shaped by regional and international connectivity. This study aimed to investigate HIV-1 molecular diversity, monitor circulating HIV-1 genetic variants, and inter-gene recombination in a cohort of people living with HIV in Morocco. Methods: We conducted an analysis of individuals diagnosed with HIV-1 infection or receiving follow-up care. Demographic and clinical data were extracted. Genotypic testing was performed on the protease/reverse transcriptase (PR/RT) and integrase (IN) regions of the pol gene using the HIV-1 Genotyping Kit with Integrase. Subtypes were assigned via Stanford HIVdb and HIV Blast, and phylogenetic relationships were analyzed using MEGA 12. Results: Of the 73 individuals enrolled, 64 were successfully sequenced. The median age was 43 years (IQR 35-51.3), with over half aged 25-44, and 85.9% were male. Heterosexual transmission was the main route (87.5%), and 59.4% were ART-naïve. Non-B subtypes predominated (87.5%), led by CRF02_AG (73.4%), followed by B (12.5%), C (7.8%), and A3 (3.1%). The cohort showed significant genetic diversity, including multiple CRFs such as CRF45_cpx (1.6%), CRF01_AE (1.6%), B/CRF02_AG (7.8%), G/CRF02_AG (3.1%), C/CRF02_AG (1.6%), CRF02_AG/CRF45_cpx (1.6%) and CRF02_AG/CRF22_01A1 (1.6%). Conclusions: This study provides updated insight into HIV-1 diversity in Morocco, showing a predominance of non-B subtypes, particularly CRF02_AG, and signals of increasing heterogeneity compared with reports from more than a decade ago that described subtype B predominance. These findings suggest a viral transition shaped in part by regional connectivity and highlight a gap in Morocco's HIV strategy, underscoring the need to implement nationwide molecular surveillance to inform future HIV control efforts.
## 1. Introduction
The global HIV-1 epidemic continues to evolve, not only in its geographic spread but also in its genetic complexity [1]. Nearly 40 million people are currently living with the virus and over 630,000 HIV-related deaths were recorded in 2024 [2,3]. HIV-1 is distinguished by its high genetic diversity, consisting of four groups (M, N, O, and P), with Group M alone accountable for the pandemic [4]. Group M is further classified into multiple subtypes (A-D, F-H, J, K) and more than 150 circulating recombinant forms (CRFs) and unique recombinant forms (URFs) arising from co-infection and recombination events [5][6][7][8][9][10]. While subtype B accounts for less than 15% of global infections, it remains the most studied due to its prevalence in high-income countries. In contrast, non-B subtypes such as CRF02_AG, A1, and C now dominate in regions with the highest HIV burden [6,8,[11][12][13][14][15][16]. HIV-1 genetic diversity results from rapid nucleotide substitution and frequent recombination [4,17,18]. This heterogeneity influences viral virulence, disease progression, and most importantly the emergence of antiretroviral (ARV) resistance. Variability in the env gene enables immune evasion, while evolution in the pol gene, particularly in reverse transcriptase, protease, and integrase, drives resistance mutations that threaten sustained viral suppression [17,[19][20][21][22]. Beyond resistance, this heterogeneity remains a major barrier to effective vaccine development, making molecular surveillance a cornerstone of effective public health strategies [23].
By the end of 2024, an estimated 23,500 people were living with HIV in Morocco, with a national prevalence of 0.08% and around 1000 new infections annually. Morocco has made significant advancement in its HIV response, with 80% of individuals living with HIV aware of their status, 95% receiving antiretroviral therapy (ART), and 95% achieving viral suppression, thereby moving closer to the UNAIDS 95-95-95 targets [24,25]. Recent initiatives to improve genotyping accessibility illustrate the nation's growing commitment to incorporating molecular tools into its national strategy yet contemporary molecular data remain sparse and largely derived from clinically selected resistance-testing cohorts rather than systematic surveillance. Historic reports described subtype B predominance, with later observations of increasing CRF02_AG and other non-B lineages, but most of these data are more than a decade old [26][27][28][29][30][31][32]. Given Morocco's position between sub-Saharan Africa, North Africa, and Europe, ongoing introductions of non-B variants through regional mobility are plausible and underline the need for updated, programme-feasible molecular monitoring to support treatment outcomes, drug resistance management, and epidemic control [12].
In this context, we conducted a molecular epidemiology study of HIV-1 in a Moroccan cohort newly diagnosed with HIV-1 infection and/or receiving routine follow-up care. Using genotypic sequence data from pol regions (PR/RT and IN), we evaluated subtype distribution and identified circulating recombinant forms (CRFs). This study provides an updated molecular snapshot of HIV-1 diversity in a Moroccon referral hospital, addressing gaps in national surveillance and underscoring the need to integrate molecular surveillance into the country's HIV strategy.
## 2. Materials and Methods
## 2.1. Study Population
The study population comprised two groups: ART-naïve individuals, newly diagnosed, with baseline plasma viral load greater than 4 log 10 copies per mL, and ARTexperienced patients in routine follow-up who were referred by their clinicians for genotypic resistance testing owing to persistent viremia above 4 log 10 copies per mL, consistent with virological failure. between August 2024 and May 2025 at the Center of Virology, Infectious and Tropical Diseases, Mohamed V Military Instruction Hospital, Rabat Morocco. Patients were eligible for sequencing if they had sufficient plasma viral load to allow amplification of the PR/RT and integrase regions of the pol gene. Individuals under the age of 18 were excluded. Clinical, demographic, and laboratory data were extracted from medical records. The study was conducted following the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the Faculty of Medicine and Pharmacy at Mohammed V University, Rabat, Morocco (Approval ID: CERB 47/23). Written informed consent was obtained from all participants. All phylogenetic and statistical analyses were performed using de-identified datasets to ensure patient confidentiality.
## 2.2. Genotypic Study and Phylogenetic Analysis
HIV-1 viral load quantification was performed using the COBAS ® 4800 System (Roche Diagnostics, Mannheim, Germany) with a detection limit of 20 copies/mL. CD4+ T-cell counts were measured on a Beckman Coulter NAVIOS flow cytometer (Beckman Coulter Life Sciences, Brea, CA, USA). Genotypic testing was performed retrospectively using stored samples from individuals with ongoing HIV replication. Viral RNA was extracted from 400 µL blood plasma samples and eluted in 60 µL, using Qiagen EZ1 Advanced XL for Nucleic Acid Purification (Qiagen, Hilden, Germany). The viral RNA was used immediately for reverse transcription polymerase chain reaction (RT-PCR), followed by a nested PCR of the protease (PR), reverse transcriptase (RT), and integrase (IN) of the pol gene using the HIV-1 Genotyping Kit with Integrase (Applied Biosystems, Thermo Fisher, Carlsbad, CA, USA) in a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems, Carlsbad, CA, USA). Sequencing was performed on a 3500XL Genetic Analyzer (Applied Biosystems Instruments, Foster City, CA, USA) using POP-7 polymer and capillary electrophoresis. Nucleotide sequences were manually inspected for quality, and ambiguities were resolved by reviewing chromatogram peak profiles. Raw sequence chromatograms were assembled and edited using SeqScape Software version 3.0 (Applied Biosystems, Foster City, CA USA). Consensus sequences were generated and aligned for subtype analysis.
HIV-1 subtypes were assigned in two steps. First, sequences were analyzed with the Stanford University HIV Database (HIVdb) subtyping module (version 9.8) to obtain preliminary classifications. These results were then confirmed using the HIV BLAST tool version 2.13.0 against curated reference sequences from the Los Alamos National Laboratory (LANL) HIV database (https://www.hiv.lanl.gov/content/sequence/BASIC_BLAST/ basic_blast.html (accessed on 23 September 2025)). For phylogenetic reconstruction, one representative reference sequence was selected from LANL for each subtype identified in our cohort, ensuring comprehensive representation while avoiding redundancy. Phylogenetic trees were inferred using the neighbor-joining method based on the Kimura 2-parameter model, with 1000 bootstrap replicates, in MEGA version 12. Bootstrap values ≥ 70 were considered to indicate reliable clustering.
## 2.3. Statistical Analysis
Categorical variables were summarized as frequencies and percentages and compared using Pearson's χ 2 test or Fisher's exact test, as appropriate. Continuous variables were reported as medians with interquartile ranges (IQR1-IQR3) and compared using the nonparametric Mann-Whitney U test. All tests were two-sided, and a p-value of <0.05 was considered statistically significant. Statistical analyses were performed using SPSS Statistics, version 25.0 (IBM Corp., Armonk, NY, USA).
## 3. Results
## 3.1. Cohort Characteristics
Of the 73 individuals initially enrolled, 64 were successfully sequenced and included in the analysis. The remaining 9 samples could not be amplified, due to a combination of suboptimal plasma quality, RNA degradation during storage, and technical amplification failure despite repeated attempts. The demographic and clinical characteristics of the patients included in the study are summarized in Table 1. Overall, 8 patients (12.5%) were infected with HIV-1 subtype B and 56 (87.5%) with non-B subtypes, including CRFs. The median age was 43 years (IQR 35-51.3), with no significant difference between subtype B (43.5 years; IQR 36.5-47.3) and non-B (43 years; IQR 34.8-51.3; p = 0.750). Patients aged 25-44 years represented the most affected group (56.3%), followed by the patients aged >50 years (26.6%). The age category distribution did not differ significantly between subtypes (p = 0.777). The cohort was predominantly male (85.9%). No significant sex differences were noted between groups (p = 0.892). Regarding marital status, 57.8% have been married at some point, while 42.2% remain unmarried (p = 0.891). Heterosexual contact was the main transmission route (96.8%), with only two isolated cases of MSM and bisexual transmission (1.6% each); no significant variation by subtype was observed (p = 0.944). At the time of diagnosis, 43.7% of individuals were classified as CDC stage A, 21.9% as stage B, and 34.4% as stage C. The difference between the groups was not statistically significant (p = 0.255). ART-naïve patients constituted 59.4% of the cohort, whereas ARTexperienced patients comprised 40.6% and exhibited a higher likelihood of being infected with subtype B (p = 0.253). The median plasma viral load at the time of genotyping was 5.10 log 10 copies/mL (IQR 4.51-5.70), with comparable levels in subtype B (5.
## 3.2. Phylogenetic Analysis & Subtype Distribution
HIV-1 subtype analysis was successfully performed in 64/73 participants. The sequences were then analyzed to determine HIV-1 subtype distribution in the cohort. To further investigate the molecular dynamics of HIV-1 in our study, a phylogenetic analysis was conducted based on aligned PR/RT and IN sequences. A detailed description is provided in Figure 1. CRF02_AG was by far the most prevalent variant, identified in 47/64 participants (73.4%), of whom 57.8% (37/47) harbored pure CRF02_AG across both PR/RT and IN regions, while the remaining ten sequences showed discordant subtype assignments between the PR/RT and integrase regions of the pol gene, suggestive of inter-subtype recombination involving CRF02_AG in combination with other subtypes, including B (7.8%), G (3.1%), C (1.6%), CRF45_cpx (1.6%), and CRF22_01A1 (1.6%). The second common subtype was B, observed in 12.5% (8/64), followed by subtype C with 7.8% (5/64) and subtype A3 with 3.1% (2/64). The cohort also exhibited remarkable genetic diversity with multiple CRFs, including CRF01_AE and CRF45_cpx (1.6% each) (Figure 2).
## 4. Discussion
The epidemiological characteristics of the cohort reflect broader national trends and longstanding gaps in the HIV care cascade. In our cohort, men represented the majority of cases (85.9%). This male predominance reflects the nature of the study setting at the Military Training Hospital Mohammed V in Rabat, where many patients are members of the armed forces. Although relatives and civilians are also treated at CVMIT, this sex imbalance contrasts with national data, where women account for approximately 47% of all people diagnosed with HIV in Morocco, reinforcing the feminization of the Moroccan epidemic. This disparity may reflect differences in testing practices, healthcare access, or late diagnosis among women and highlights the need for gender-responsive strategies in HIV screening and care. More than half of our cohort patients were in the 24-44 age category. This age distribution is consistent with national surveillance reports, indicating that individuals aged 25 to 44 remain the most affected in Morocco across both sexes [25]. Heterosexual transmission remained the dominant mode of acquisition (87.5%), a proportion that aligns closely with previous Moroccan reports over the past two decades. Between 2004 and 2015, heterosexual transmission consistently represented the primary route of infection, ranging from 81.25% to 92.3% across multiple cohorts. This epidemiological profile reflects the spread of HIV beyond high-risk groups and into the broader population [26,[28][29][30]33]. Data on men who have sex with men (MSM) remain limited in Morocco, although studies are currently underway to help address this critical gap [34]. Existing evidence indicates ongoing HIV transmission within this population; national surveillance estimates reported a prevalence of 4.2% in 2012, which increased to 5.3% by 2023 [33,35,36]. In our cohort, 57.8% of patients were married or had previously been married, while 42.2% were unmarried. In Morocco, where marriage is institutionally and socially defined as heterosexual, marital status remains an epidemiologically relevant demographic variable. Although it cannot fully capture non-marital partnerships or unreported same-sex practices, integrating marital status into molecular surveillance may provide useful insights into transmission dynamics within couples or stable relationships and inform the design of more targeted interventions [37].
Our cohort consisted of patients referred for genotypic testing due to persistent viral replication, a selection criterion that inherently enriched the study population with ARTnaïve individuals and ART-experienced patients undergoing virologic failure. The ART coverage distribution in our study does not align with national programmatic data, which indicates that over 95% of individuals diagnosed with HIV are receiving treatment, and most attain viral suppression. Our findings identify a subgroup of patients at risk for adverse outcomes, revealing significant deficiencies in the care continuum that molecular surveillance can effectively monitor.
Our cohort showed a median CD4 count of 202 cells/mm 3 and a median viral load of 5.19 log 10 copies/mL. In contrast, over the past two decades in Morocco, median CD4 counts have gradually increased from 116 cells/mm 3 in 2005-2009 to over 400 cells/mm 3 in cohorts from 2014-2015 in ART-naïve patients [28,29] and from 346 cells/mm 3 in 2005-2010 to over 400 cells/mm 3 in cohorts from 2014-2015 in ART-experienced patients [26,27], while median viral loads have decreased from approximately 5.17 log 10 copies/mL to 4.56 log 10 , suggesting that despite national progress, late diagnosis and immune suppression remain common [26,28,30,31]. These immunological and virologic findings are consistent with the CDC stage distribution in our cohort, where 34.4% of patients were classified as stage C, confirming that a considerable proportion of individuals were diagnosed at an advanced stage of HIV infection. In contrast, 43.7% were diagnosed at an asymptomatic stage (stage A), which aligns with national surveillance data showing increased early detection in recent years. Despite substantial advancements in testing expansion and enhanced access to ART, our findings highlight that late diagnosis continues to be a persistent public health challenge [25].
Our study suggests a shift in the molecular epidemiology of HIV-1 in Morocco, with a predominance of non-B subtypes (87.5%) and increasing viral diversity observed within the cohort. In earlier molecular epidemiology studies, HIV-1 subtype B was overwhelmingly predominant in Morocco, representing 93.5% of infections in 1997, while subtypes A and F accounted for just 1.0% and 0.5%, respectively [38]. By 2005, although subtype B remained the most common (76.7%), there was a marked rise in non-B subtypes, particularly CRF02_AG, which constituted 15% of cases [32]. Between 2004 and 2015, subtype B remained dominant across most molecular studies, with proportions ranging from 66% to 74% [26,28,30]. However, during the same period, a gradual increase in non-B subtypes-particularly CRF02_AG-was documented, rising from 9% in 2005-2010 to 25.3% by 2006-2010 [26,30]. The appearance of other subtypes, such as A1, C, and unique recombinants, has also been intermittently reported [26,28,30,31]. Our current findings demonstrate a continued and substantial shift in subtype dynamics, with CRF02_AG now accounting for 73.4% of cases and subtype B declining to just 12.5%, reflecting an ongoing diversification of circulating strains. These results underscore a clear epidemiological transition over the past two decades and highlight the increasing complexity of HIV molecular patterns in Morocco.
Morocco's strategic location at the crossroads of Africa and Europe may contribute to its exposure to diverse HIV-1 variants. The predominance of CRF02_AG (73.4%) in our study aligns with global trends reported in West and Central Africa [14,[39][40][41]; with Cameroon at 60-68% in 2016 [42], where this recombinant form is highly prevalent, unlike much of the MENA region, where subtype B remained the most prevalent (39%) until 2016 in countries like Algeria, Tunisia, and Yemen [43,44]. The persistence of subtype B in our cohort suggests a historical link to earlier European transmission networks where subtype B is still the most widespread viral strain, especially in countries like France with 56%, Germany and Spain, which exceeded 80% [45,46]. As Europe now sees an influx of non-B subtypes due to migration and with globalized transmission, other subtypes have also been introduced; for example, in Spain, subtype F1 was identified in a cohort of MSM and CRF02_AG among immigrant populations [46][47][48][49]. Our study also identified several other subtypes and recombinants, including subtype C, typically associated with Southern and Eastern Africa [50], subtype G, the second most prevalent in West Africa [51], and complex recombinant forms such as CRF04_cpx, originating in Central Africa; CRF45_cpx, previously reported in Cameroon [7,14] and CRF01_AE, common in Southeast Asia [52,53]. Importantly, HIV prevalence among migrants in Morocco has been reported as markedly higher (4.6%) than in the general population (<0.1%), and this group faces significant barriers to accessing healthcare [25, 54,55]. Although our study did not include patient-level migration data, these contextual data support the hypothesis that Morocco's geographic position at the crossroads of Africa and Europe may facilitate exposure to globally circulating variants. Morocco appears to be experiencing a parallel diversification, acting as both a recipient and conduit for recombinant HIV-1 forms. While our data cannot directly establish the drivers of this diversification, these findings are consistent with Morocco's growing exposure to globally circulating variants and reinforce its role as a molecular crossroads in the international HIV epidemic.
An important observation in our study was the variation in subtype classification between the PR/RT and integrase (IN) regions in 10 patients, highlighting subtype discordance. The most frequent pattern was subtype B in PR/RT and CRF02_AG in IN, although other combinations involving CRF02_AG with subtypes C, G, CRF45_cpx, and CRF22_01A1 were also detected. These inter-gene discrepancies are consistent with possible recombination events, a phenomenon well documented in high-diversity settings where co-infection or superinfection facilitates template switching during reverse transcription [18].
The growing complexity of HIV-1 diversity in Morocco underscores the need to strengthen surveillance strategies. At present, genotypic resistance testing is not routinely performed at diagnosis and is generally limited to cases where it is specifically requested or conducted within research settings. By including ART-naïve patients, our study provides baseline molecular data that suggest it may be timely to consider the implementation of baseline resistance testing for all new diagnoses as part of Morocco's HIV response.
Our findings should be interpreted considering several limitations. First, this was a single-center observational study with a modest sample size, which may limit the generalizability of the results. The study population consisted of individuals engaged in clinical care and eligible for genotypic testing, introducing the potential for selection bias. Second, the cohort was predominantly male, reflecting the military setting of the hospital. Third, sequencing was restricted to the pol gene, which prevents definitive characterization of recombinant viruses and may underestimate the true extent of genetic diversity. Finally, our study provides only a snapshot of circulating strains without capturing longitudinal changes. Nevertheless, the study population included not only military personnel but also their relatives and a substantial number of civilian patients referred from different regions of Morocco, offering a valuable window into the molecular diversity observed in clinical practice.
## 5. Conclusions
This study provides one of the most up-to-date molecular snapshots of HIV-1 diversity in Morocco, filling a gap of more than a decade since the last published data, which had reported a predominance of subtype B. Our analysis shows that non-B subtypes, particularly CRF02_AG, now predominate, with indications of increasing heterogeneity within the epidemic. These findings provide signals of changing molecular epidemiology of HIV-1 in Morocco, though broader studies are needed to confirm national trends. Molecular surveillance should be urgently integrated into the national HIV strategy to anticipate subtype dynamics, strengthen public health interventions, and optimize treatment outcomes.
## References
1. (2025) "Consolidated Guidelines on HIV Prevention, Testing, Treatment, Service Delivery and Monitoring: Recommendations for a Public Health Approach. Available online"
2. (2025) "Global HIV & AIDS Statistics-Fact Sheet|UNAIDS"
3. (2024) *World Health Organization. HIV Statistics, Globally and by WHO Region*
4. Smyth, Davenport, Mak (2012) "The origin of genetic diversity in HIV-1" *Virus Res*
5. (2025) "HIV Circulating Recombinant Forms (CRFs). Available online"
6. Williams, Menon, Crowe et al. (2010) "Geographic and Population Distributions of Human Immunodeficiency Virus (HIV)-1 and HIV-2 Circulating Subtypes: A Systematic Literature Review and Meta-analysis" *J. Infect. Dis*
7. Hemelaar, Gouws, Ghys et al. (2006) "Global and regional distribution of HIV-1 genetic subtypes and recombinants in 2004" *AIDS*
8. Nair, Gettins, Fuller et al. (2024) "Global and regional genetic diversity of HIV-1 in 2010-2021: Systematic review and analysis of prevalence" *Lancet Microbe*
9. Eberle, Gürtler (2012) "HIV Types, Groups, Subtypes and Recombinant Forms: Errors in Replication, Selection Pressure and Quasispecies" *Intervirology*
10. Hemelaar (2012) "The origin and diversity of the HIV-1 pandemic" *Trends Mol. Med*
11. Buonaguro, Tornesello, Buonaguro (2007) "Human Immunodeficiency Virus Type 1 Subtype Distribution in the Worldwide Epidemic: Pathogenetic and Therapeutic Implications" *J. Virol*
12. Giovanetti, Ciccozzi, Parolin et al. (1072) "Molecular Epidemiology of HIV-1 in African Countries: A Comprehensive Overview"
13. Njai, Gali, Vanham et al. (2006) "The predominance of Human Immunodeficiency Virus type 1 (HIV-1) circulating recombinant form 02 (CRF02_AG) in West Central Africa may be related to its replicative fitness" *Retrovirology*
14. Hemelaar, Elangovan, Yun et al. (2019) "Global and regional molecular epidemiology of HIV-1, 1990-2015: A systematic review, global survey, and trend analysis" *Lancet Infect. Dis*
15. Hemelaar, Loganathan, Elangovan et al. (2015) "WHO-UNAIDS Network for HIV Isolation and Characterization. Country level diversity of the HIV-1 pandemic between 1990 and" *J. Virol*
16. Bbosa, Kaleebu, Ssemwanga (2019) "HIV subtype diversity worldwide" *Curr. Opin. HIV AIDS*
17. Rambaut, Posada, Crandall et al. (2004) "The causes and consequences of HIV evolution" *Nat. Rev. Genet*
18. Zhang, Foley, Schultz et al. (2010) "The role of recombination in the emergence of a complex and dynamic HIV epidemic" *Retrovirology*
19. Santos, Soares (2010) "HIV Genetic Diversity and Drug Resistance" *Viruses*
20. Shi, Kitchen, Weiser et al. (2010) "Evolution and recombination of genes encoding HIV-1 drug resistance and tropism during antiretroviral therapy" *Virology*
21. Leda, Hunter, De Oliveira et al. (2020) "HIV-1 genetic diversity and divergence and its correlation with disease progression among antiretroviral naïve recently infected individuals" *Virology*
22. Bouman, Venner, Walker et al. (2023) "Per-pathogen virulence of HIV-1 subtypes A, C and D" *Proc. R. Soc. B Biol. Sci*
23. Boomgarden, Upadhyay "Progress and Challenges in HIV-1 Vaccine Research: A Comprehensive Overview. Vaccines 2025"
24. Maroc|onusida (2025)
25. El Annaz, Recordon-Pinson, Tagajdid et al. (2012) "Drug resistance mutations in HIV type 1 isolates from patients failing antiretroviral therapy in Morocco" *AIDS Res. Hum. Retroviruses*
26. Alaoui, El Alaoui, El Annaz et al. (2019) "HIV-1 Integrase Resistance among Highly Antiretroviral Experienced Patients from Morocco" *Intervirology*
27. Annaz, Recordon-Pinson, Baba et al. (2011) "Presence of drug resistance mutations among drug-naive patients in Morocco" *AIDS Res. Hum. Retroviruses*
28. Alaoui, El Alaoui, Touil et al. (2018) "Prevalence of resistance to integrase strand-transfer inhibitors (INSTIs) among untreated HIV-1 infected patients in Morocco" *BMC Res. Notes*
29. Miri, Ouladlahsen, Kettani et al. (2012) "Characterization of protease resistance-associated mutations in HIV type 1 drug-naive patients following the increasing prevalence of the CRF02_AG strain in Morocco" *AIDS Res. Hum. Retroviruses*
30. Bakhouch, Oulad-Lahcen, Bensghir et al. (2009) "The prevalence of resistance-associated mutations to protease and reverse transcriptase inhibitors in treatment-naïve (HIV1)-infected individuals in Casablanca" *Morocco. J. Infect. Dev. Ctries*
31. Akrim, Lemrabet, Elharti et al. (2012) "HIV-1 Subtype distribution in morocco based on national sentinel surveillance data 2004-2005" *AIDS Res. Ther*
32. Mumtaz, Hilmi, Akala et al. (2011) "HIV-1 molecular epidemiology evidence and transmission patterns in the Middle East and North Africa" *Sex. Transm. Infect*
33. Mumtaz, Kouyoumjian, Hilmi et al. (2013) "The distribution of new HIV infections by mode of exposure in Morocco" *Sex. Transm. Infect*
34. Alcs Maroc (2024) "Rapport d'Activite ALCS 2023. Maroc. Available online"
35. Mumtaz, Chemaitelly, Almukdad et al. "Status of the HIV epidemic in key populations in the Middle East and north Africa: Knowns and unknowns"
36. Wand, Morris, Moodley et al. (2002) "Impact of marital status on risk of HIV in South Africa"
37. Elharti, Elaouad, Amzazi et al. (1997) "HIV-1 diversity in Morocco" *AIDS*
38. Lihana, Ssemwanga, Abimiku et al. (2012) "Update on HIV-1 diversity in Africa: A decade in review" *AIDS Rev*
39. Nii-Trebi, Brandful, Ibe et al. (2017) "Dynamic HIV-1 genetic recombination and genotypic drug resistance among treatment-experienced adults in northern Ghana" *J. Med. Microbiol*
40. Janssens, Salminen, Laukkanen et al. (2000) "Near full-length genome analysis of HIV type 1 CRF02. AG subtype C and CRF02. AG subtype G recombinants" *AIDS Res. Hum. Retroviruses*
41. Courtney, Agyingi, Fokou et al. (2016) "Monitoring HIV-1 Group M Subtypes in Yaoundé, Cameroon Reveals Broad Genetic Diversity and a Novel CRF02_AG/F2 Infection" *AIDS Res. Hum. Retroviruses*
42. Sallam, Şahin, Ingman et al. (2017) "Genetic characterization of human immunodeficiency virus type 1 transmission in the Middle East and North Africa"
43. Sallam, Al-Khatib, Sabra et al. (2025) "Challenges in Elucidating HIV-1 Genetic Diversity in the Middle East and North Africa: A Review Based on a Systematic Search" *Viruses*
44. Beloukas, Psarris, Giannelou et al. (2016) "Molecular epidemiology of HIV-1 infection in Europe: An overview" *Infect. Genet. Evol*
45. Visseaux, Assoumou, Mahjoub et al. (2020) "Surveillance of HIV-1 primary infections in France from 2014 to 2016: Toward stable resistance, but higher diversity"
46. Delgado, Benito, Montero et al. (2019) "Diverse Large HIV-1 Non-subtype B Clusters Are Spreading Among Men Who Have Sex with" *Men in Spain. Front. Microbiol*
47. Patiño-Galindo, Domínguez, Cuevas et al. (2018) "Genome-scale analysis of evolutionary rate and selection in a fast-expanding Spanish cluster of HIV-1 subtype F1" *Infect. Genet. Evol*
48. Kostaki, Flampouris, Karamitros et al. (2019) "Spatiotemporal Characteristics of the Largest HIV-1 CRF02_AG Outbreak in Spain: Evidence for Onward Transmissions" *Front. Microbiol*
49. Novitsky, Smith, Gilbert et al. (2002) "Human Immunodeficiency Virus Type 1 Subtype C Molecular Phylogeny: Consensus Sequence for an AIDS Vaccine Design?" *J. Virol*
50. Delatorre, Mir, Bello (2014) "Spatiotemporal dynamics of the HIV-1 subtype G epidemic in West and Central Africa" *PLoS ONE*
51. Angelis, Albert, Mamais et al. (2014) "Global Dispersal Pattern of HIV Type 1 Subtype CRF01_AE: A Genetic Trace of Human Mobility Related to Heterosexual Sexual Activities Centralized in Southeast Asia" *J. Infect. Dis*
52. An, Han, Zhao et al. (2020) "Cross-Continental Dispersal of Major HIV-1 CRF01_AE Clusters in China" *Front. Microbiol*
53. Miri, Wakrim, Kassar et al. (2014) "Impact of immigration on HIV-1 molecular epidemiology in West Africa, Maghreb and southern Europe" *AIDS Rev*
54. Bouri, Najdi (2025) "Migrant healthcare access in Morocco: A narrative review" *J. Public Health Afr*
55. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12691592&blobtype=pdf | # Evolution of antiviral resistance captures a transient interdomain functional interaction between chikungunya virus envelope glycoproteins
Leandro Battini, Sara Thannickal, Malena Cibello, Mariela Bollini, Kenneth Stapleford, Diego Álvarez
## Abstract
Envelope proteins drive virus and host-cell membrane fusion to achieve virus entry. Fusogenic proteins are classified into structural classes that function with remarkable mechanistic similarities. Membrane fusion implies coordinated movements of protein domains through a series of sequential steps. Structures for the initial and final conformations are available for several fusogens, but folding intermediates remain largely unresolved, and the interdependency between regions that drive conformational rearrangements is not well understood. Chikungunya virus (CHIKV) particles display heterodimers of envelope proteins E1 and E2 associated as trimeric spikes that respond to acidic pH to trigger fusion. We followed the experimental evolution of CHIKV under the selective pressure of a novel entry inhibitor. Mutations arising from selection mapped to two residues located in the distal domains of the E2 and E1 heterodimer and spikes. Here, we demonstrate that the antiviral mechanism involves inhibition of membrane fusion. Phenotypic characterization of recombinant viruses indicated that the selected mutations confer a fitness advantage under antiviral pressure, and that the double-mutant virus overcame antiviral inhibition of fusion while single mutants were sensitive. In addition, molecular dynamics simulations suggest that these two residues modulate the conformational rearrangement of the E1-E2 heterodimer. In this line and supporting a functional link between residues, the double-mutant virus displayed a higher pH threshold for fusion than single-mutant viruses. Finally, mutations resulted in distinct replication and spreading outcomes in mice and infection rates in mosquitoes, underscoring the fine-tuning of envelope proteins' function as a determinant for the establishment of infection. Altogether, our approach captured an otherwise unresolved interdomain interaction. IMPORTANCE Chikungunya virus (CHIKV) is a reemergent pathogen that has caused large outbreaks in the last 20 years. There are no available antiviral therapies, and a vaccine has only recently been approved. We describe the mode of action of an inhibitor designed to target CHIKV envelope proteins, blocking entry at the stage of fusion between the virus envelope and host membranes. Fusion is common to the entry of enveloped viruses. Virus envelope proteins drive fusion, undergoing a series of transitions from an initial metastable conformational state to a more stable post-fusion state. Intermediate conformations are transient and have mostly remained inaccessible to structure determination. Here, a selection of viruses that are resistant to antiviral inhibition of fusion uncovered a functional interaction between two residues residing in domains that are apart in both the pre-fusion and post-fusion states. Thus, we provide new insight into the molecular detail of the inner working of virus fusion machinery.
C hikungunya virus (CHIKV) is transmitted to humans by mosquitoes. Infection commonly results in acute fever and joint pain that can progress into chronic polyarthritis (1,2). Since 2004, CHIKV has spread throughout the tropical and subtropical regions around the world, causing outbreaks associated with high socioeconomic cost (3,4). Genomic adaptation of the virus to new mosquito vectors has been pointed out as one of the major causes for the geographical spread of the virus (1,5,6). Therefore, the study of viral determinants of CHIKV fitness in its natural hosts may provide valuable information regarding the continued adaptation of the virus.
CHIKV is an alphavirus of the Togaviridae family (7). It has a 12 kb positive-stranded RNA genome that encodes for 10 proteins in two open reading frames (ORFs). The second ORF in the 3′ of the genome encodes for the structural proteins of the virus: the capsid (C) and the envelope glycoproteins (E3-E2-E1). E2 and E1 are transmembrane proteins that together with E3 form a heterotrimer on the surface of the viral particle, mediating the interaction with cell receptors and fusion between viral and endosomal membranes during entry, respectively (8). E2 belongs to the immunoglobulin superfam ily of proteins and folds into three globular domains (A, B, and C) connected by two antiparallel β-strands, referred to as β-ribbon (9). E1 is a class II viral fusion protein and folds into a β-sheet-rich structure with three β-barrel domains (I, II, and III) bearing the fusion loop in the tip of domain II (9,10). In the viral particle, three E3-E2-E1 trimers fold into a viral spike, with E1 located next to the membrane and E2 positioned over E1, protecting the fusion loop (11).
CHIKV enters the cell by receptor-dependent endocytosis. In the final stage of the entry process, the decrease in endocytic pH induces membrane fusion, triggering a major conformational rearrangement of the envelope proteins that involves dissociation of E1 and E2 and reassociation of E1 into homotrimers (9,(12)(13)(14)(15). The joint action of multiple E1 homotrimers is required to generate the force necessary to curve the opposing membranes (10,16). For this reason, the conformational rearrangement in different spikes must be coordinated, and the timing of each step must be tightly regulated.
We have previously identified compound 11 as a small molecule inhibitor of CHIKV infection following a virtual screening against the envelope proteins of the virus (17). We showed that compound 11 inhibited the internalization of the viral particle and selected a viral variant resistant to the antiviral activity of the compound, which harbors mutations in E1 (Y24H) and E2 (P173S). Interestingly, whereas recombinant viruses carrying both E1-Y24H and E2-P173S showed the resistant phenotype, neither of the single mutants did, suggesting a functional interaction between these two residues located in distant regions of CHIKV envelope glycoproteins.
In this study, we show that compound 11 inhibits the pH-dependent membrane fusion process. Based on molecular dynamics simulations of the pre-fusion conforma tion of the envelope proteins, we propose that residues E1-Y24 and E2-P173 are associated with kinetic barriers that act as checkpoints of the protein conformational change during the fusion process. In this line, characterization of recombinant viruses showed that combined E1-Y24H and E2-P173S mutations altered pH dependence for fusion compared to both WT and single E1-Y24H and E2-P173S mutant viruses. In turn, experimental infection resulted in enhanced replication in mice infected with the double mutant virus but impaired ability to infect mosquitoes compared to the WT virus. Altogether, the results presented in this work support a functional interaction between residues E1-Y24 and E2-P173 that function in a concerted manner to regulate the fusion process and impact on the establishment of infection in vivo.
## RESULTS
## Compound 11 inhibits membrane fusion
We have identified a small molecule inhibitor of CHIKV infection designed to target a pocket located behind the fusion loop of E1 in a cleft formed between domain II of E1 and domains A and B of E2. Time of drug addition assays showed that compound that the effect was gradually lost when addition was delayed, indicating that treatment blocked virus entry but not the release of infectious particles. In line with this observa tion, we found that the compound had no effect on virus attachment and specifically inhibited internalization, reducing relative infection by threefold in Vero cells treated with 50 µM of compound 11 (17). Moreover, the approach suggested that treatment halted viruses at a post-attachment stage and inhibited further progress of infection as the virus was not able to propagate from treated cells to non-treated cells in a focus-forming assay. To further characterize its mode of action, we tested the inhibitory activity of the compound against lentiviruses pseudotyped with CHIKV or VSV envelope proteins using BHK cells as a target for transduction (Fig. 1A). Virus pseudotypes bear CHIKV or VSV envelope proteins surrounding an HIV capsid and allow testing for the specific inhibition of envelope protein function by compound 11 at the stage of entry since no further rounds of virus replication occur after primary infection. While the VSV pseudotype was resistant to the antiviral activity of compound 11, it inhibited CHIKV pseudotyped lentiviruses (EC 50 = 7.6 ± 0.1 µM, Fig. 1A inset) with an EC 50 similar to that obtained against a reporter CHIKV expressing ZsGreen from a subgenomic promoter (EC 50 = 13.77 ± 1.87 µM and Fig. 1B), demonstrating that the compound specifically inhibited CHIKV entry into the host cell. Moreover, both CHIKV and VSV are endocytosed through the clathrin-dependent pathway (18,19). As we have shown that compound 11 did not interfere with virus attachment (17), the fact that inhibition was specific against the CHIKV pseudotype rules out an indirect effect of the compound in endocytosis and suggests that compound 11 may inhibit the function of CHIKV envelope proteins in the membrane fusion process. Then, we performed fusion from without (FFWO) assays in which the fusion of viruses attached on the cell surface is triggered directly at the plasma membrane by lowering the pH of the medium. As a result, the degree of infection is dependent on the fusogenic activity of the viral particles. We first set up the assay by triggering fusion at different pHs after the absorption of CHIKV-ZsGreen (Fig. 1C). Data showed that the fusogenic activity of the envelope proteins followed a sigmoidal curve with a fusion threshold around pH 5.5, similar to the behavior described in literature (20,21). We then evaluated the effect of adding compound 11 at increasing concentrations in FFWO triggered at pH 5.4, 5.5, or 5.6. We found that compound 11 inhibited fusion in a concentration-dependent manner and had a stronger effect at higher pH (Fig. 1D).
As a complementary approach, we set up a cell-cell fusion assay based on the ability of recombinant CHIKV envelope proteins exposed on the plasma membrane to induce fusion with target cells. In BHK cells transfected with a reporter construct encoding for GFP and CHIKV structural proteins separated by a self-cleaving peptide, green multinu clei syncytia formed after treatment with low pH medium (Fig. 1E andF). Fluorescence image analysis was followed to classify GFP-expressing cells based on the number of encompassed nuclei, and the fusogenic activity was estimated using a fusion index (fusion index = 1 -cells/nuclei) that ranges from 1 to 0, with higher values indicating a higher fusogenic activity (Fig. 1G). As in the FFWO assay, compound 11 inhibited the fusogenic activity of CHIKV envelope proteins in a concentration-dependent manner when fusion was triggered at pH 5.6 (Fig. 1H).
In conclusion, complementary approaches using virus pseudotypes, fully infectious reporter CHIKV, and protein expression constructs to evaluate the function of CHIKV envelope proteins demonstrate that compound 11 inhibited entry of CHIKV into the host cell and pinpoint the mechanism of action to the inhibition of the fusion process driven by CHIKV envelope proteins.
## E2-P173S E1-Y24H double mutant resistant phenotype is associated with overcoming of fusion inhibition
Evolution of antiviral resistance generally occurs through stepwise selection of mutations that often result in reduced viral fitness. To study the impact of E2-P173S and E1-Y24H double and single mutants on the resistant phenotype, we used recombinant viruses that were previously constructed in the reporter CHIKV-ZsGreen background (17). We first assessed virus replication kinetics in Vero cells. Growth curves showed that all viruses behaved in the same manner as WT (Fig. 2A, left). Similarly, mutations did not alter growth kinetics in human 293T or Huh-7 cells (Fig. 2A, middle and right). To assess fitness in the presence of compound 11, we measured antiviral activity using reporter CHIKV-ZsGreen in a focus-forming assay. WT and single mutants E2-P173S and E1-Y24H showed similar sensitivity to compound 11, displaying EC 50 of 11.6 ± 3.4 µM, 15.7 ± 1.2 µM, and 11.2 ± 2.1 µM, respectively, while the double mutant displayed EC 50 = 21.4 ± 1.3 µM (Fig. 2B), indicating that only the combined mutations confer a partially resistant phenotype to compound 11 antiviral activity. The phenotype was further confirmed when viral yields were measured following infection of cells treated with increasing compound concentrations (Fig. 2C). As expected, increasing compound concentrations resulted in a gradual decrease in virus yields. The trend for single mutants was similar to WT, showing a ~1,000-fold decrease in viral titers in cells treated at 25 µM of compound compared to non-treated cells. The threefold difference in titer of the E2-P173S mutant compared to WT in the absence of compound was also observed across the concentrations tested, suggesting that the two viruses are similarly sensitive to antiviral treatment. In contrast, the E2-P173S E1-Y24H double-mutant virus demonstrated a clear resistance to the antiviral activity of compound 11, displaying more than 100-fold higher viral yields at 25 µM of compound compared to WT. Taken together, focus forming and virus yield assays indicate that development of antiviral resistance was associated with selection of the combined E2-P173S and E1-Y24H mutations.
To further understand the contribution of selected mutations to antiviral resistance, we evaluated the fitness of WT and mutant viruses in virus competition assays. Untreated cells or cells treated with 50 µM of compound 11 were infected with WT and mutant viruses mixed at an initial 1:1 ratio. Then, the composition of the resulting virus popula tion was assessed by Sanger sequencing of the RT-PCR product amplified from cell culture supernatants by comparing the peak height of the WT and mutant alleles (Fig. 2D). As expected, we observed a 1:1 signal ratio for input populations. In untreated cells, although mutant alleles were detected, in all cases, the ratio favored the WT, suggesting that mutations impaired virus fitness. In turn, in treated cells, both E1-Y24H and the double-mutant viruses completely displaced the WT virus, and the signal ratio for the E2-P173S and WT was inverted compared to competition in untreated cells, indicating that the mutant allele conferred an increment in fitness in the presence of compound 11 (Fig. 2E). Focus-forming and virus yield assays did not anticipate the association of individual mutations to resistance, which became evident in direct competition assays under stringent antiviral pressure. In turn, the assay also suggested that mutations have a cost in virus fitness when viruses compete without antiviral pressure, suggesting that these conflicting forces would have acted together in the selection process that led to the emergence of the double-mutant virus at high antiviral concentrations.
Next, we sought to gain insight into the mechanism by which the double-mutant virus overcomes the antiviral activity of compound 11. First, we evaluated the effect of compound 11 on infection with WT, and single-and double-mutant pseudotyped lentiviruses to directly test the impact of mutations on virus entry (Fig. 3A). At concentra tions greater than 10 µM, only the double-mutant virus displayed increased levels of infection compared to WT, indicating that the combined E2 P173S E1-Y24H mutations in the envelope proteins are sufficient to confer antiviral resistance.
Next, we performed FFWO and cell-cell fusion assays with the WT and single-and double-mutant viruses to determine whether the mutations allow the virus to overcome fusion inhibition by compound 11. In the FFWO assay, the single mutants behaved in the same manner as the WT virus, showing a dose-dependent inhibition of infected cells (Fig. 3B). In contrast, the double mutant infected a higher percentage of cells in the range of concentrations tested. In line with these results, in the cell-cell fusion assay, the fusogenic activity of the WT and both single mutants decreased as drug concentrations increased (Fig. 3C). Contrarily, there were no differences in the fusogenic activity of the double mutant up to 20 µM of compound. Compared to the FFWO assay, resistance of the double mutant virus to the inhibitory effect in the cell-cell fusion assay was not observed at the highest concentrations tested, likely due to methodological differences between approaches based on infection with replicative viruses and overexpression of envelope proteins, respectively.
Altogether, these results show that resistance of E2-P173S E1-Y24H double mutant to the antiviral activity of compound 11 is associated with a weaker inhibition of the fusogenic activity of CHIKV envelope proteins, reinforcing the notion of a functional interaction between E2-P173 and E1-Y24.
## E2-P173S and E1-Y24H are located near two important hinges of CHIKV envelope glycoproteins
It is noteworthy that E2-P173S and E1-Y24H are apart from each other both in the E2-E1 heterodimer (99.2 Å) and in the viral trimeric spike (68.1 Å). To gain deeper insights into the role of E2-P173 and E1-Y24 and the functional interaction between them, we performed molecular dynamics (MD) simulations of the envelope proteins in the pre-fusion conformation (22), and constructed and analyzed the residue interaction network (RIN). To build the RIN, we computed the contact matrix and the dynamic cross correlation matrix (DCCM) of the Cα atoms from the MD trajectory (Fig. 4A) (23). We used RIN to calculate residue communities using the Girvan-Newman algorithm (24) and mapped the resulting communities in the envelope proteins' crystal structure. Communities represent groups of residues that move in a concerted manner. Detected communities were associated with the envelope protein domains but were not identical (Fig. 4B andC).
Next, we analyzed the position of each residue in the protein and in the RIN. E2-P173 is located between E2-B and the β-ribbon. This proline is conserved in viruses of the Semliki fever virus complex and adjacent to E2-P172, which is conserved in all alphavi ruses (Fig. 5A). The PP motif is associated with a break in secondary structure. Addition ally, in the context of the E1-E2 heterodimer, these residues are located next to arch 2 in the complementary strand of the β-ribbon (Fig. 5C). The association between prolines and arches in the complementary strands of the β-ribbon is also present in arch 1 (E2-P269) and arch 3 (E2-P258 and E2-P260). Interestingly, these regions correspond to breakpoints between residue communities and, thus, represent breakpoints in residue connectivity (Fig. 4A and5C). In fact, E2-P173 is located at the hinge of the second Principal Component (PC) of the envelope proteins in a PCA of the MD trajectory, which is the PC associated with the highest RMSF of E2 domain B (Fig. 5D andE). Overall, these data suggest that P172 and P173 would control the flexibility of the β-ribbon and E2 domain B, which is essential in the first steps of the conformational rearrangements of the envelope glycoproteins. Thus, E2-P173S may alter the behavior of the WT protein in this regard.
In turn, in the MD simulations, there was a strong interaction between the loop bearing E1-Y24 and the flexible linker between domains I and III (residues 283-294). Both loops were part of the same community (Fig. 5F, pre-fusion) and showed a high correla tion between each other and with domain E1-III (Fig. 4A black box, Table 1). E1-Y24 is an aromatic residue and establishes π-cation interactions with two nearby arginines (E1-R289 and E1-R21). Interestingly, alphaviruses display an aromatic residue at position 24 (tyrosine or phenylalanine) and an arginine is strictly conserved at positions 289 and 21, indicating that this network of interactions is common across the genus (Fig. 5B). In turn, E1-R21 interacts with E1-E284, E1-F287, and E1-Y1, building an interaction network that may be responsible for the correlated motion of the loop and the flexible linker. Based on residue conservation, alphaviruses would maintain the pattern of π-cation interactions between an aromatic residue in the E1-24 position and positively charged residues at positions E1-289 and E1-21. E1-Y24 loop is separated from E1-III and the flexible linker in the post-fusion conformation of E1 envelope protein (Fig. 5F, post-fusion). This means that the interaction between the E1-Y24 loop and the linker has to break during the conformational change that drives the fusion process. Thus, the stability of this interac tion may be important for the stability of the pre-fusion conformation of E1 and the regulation of the fusion process. Histidine residues become protonated at the acidic pH of the endosome; thus, the E1-Y24H substitution may impact the strength of the interaction network connecting this residue to the linker in the context of the pHtriggered fusion process.
Altogether, MD simulations suggest that while E2-P173 would be associated with the flexibility of domain B, E1-Y24 may modulate the stability of E1 prefusion conformation. In this manner, both residues could be associated with important kinetic barriers in the conformational rearrangement of the envelope proteins and, thus, would play a role in the regulation of the timing and coordination of the different steps in the fusion process. If this were the case, substitutions associated with antiviral resistance may alter the stability and fusogenic function of the E1-E2 heterodimer and would have arisen together as compensatory mutations. Altogether, our analysis provides a mechanistic hypothesis that could explain the molecular basis behind the functional interaction of E1-Y24 and E2-P173 observed in antiviral resistance.
## Impact of E2-P173S and E1-Y24H on viral particle stability and envelope proteins functionality
To test our working hypothesis linking resistance-associated mutations to the coordina ted regulation of the envelope protein complex dynamics, we next studied the thermal stability of the viral particle and the functionality of the envelope proteins of WT and mutant viruses. To study thermal stability, we quantified infectivity after incubation of viruses at 37°C in cell culture medium (Fig. 6A). All viruses showed a similar stabil ity profile, with approximately 25% of the initial infectivity retained after 8 hours of incubation and 2% after 24 hours, indicating that mutations have no major impact on stability.
Initial characterization of virus growth showed no differences in virus yields between WT and mutant viruses (Fig. 2A), suggesting that mutations do not alter CHIKV assembly or release. To further characterize the impact of mutations on protein functionality, we studied their effect on viral attachment, cholesterol dependence for CHIKV entry, and fusion. Virus attachment assays showed no differences between mutant and WT viruses in either BHK or Vero cells (Fig. 6B).
Target membrane cholesterol was shown to promote endocytic uptake of CHIKV, and mutations in E1 were previously linked to an increased cholesterol dependency for fusion (5). To measure the cholesterol dependence for viral infection, we used methyl-β-cyclodextrin (MβCD) to capture cholesterol and lower its levels in the plasma membrane (25) (Fig. 6C). MβCD was thoroughly washed prior to infection to address cholesterol-dependent entry and minimize the impact of cholesterol depletion on nsp1 plasma membrane anchoring and subsequent RNA replication (26,27). All viruses were similarly sensitive to MβCD, suggesting that the WT and mutant viruses display a similar cholesterol dependence. Finally, we assessed the fusogenic activity of WT and mutant envelope proteins. In the FFWO assay (Fig. 6D), the fusion degree against pH followed a sigmoidal curve with a marked fusion threshold for WT and mutant viruses. Both single mutants showed a fusion profile similar to the WT virus. In contrast, there was a clear shift toward a higher pH in the threshold for fusion of E2-P173S E1-Y24H double-mutant virus. We observed a similar behavior in the cell-cell fusion assay. While the fusion index at pH 5.2 was higher for the WT and E1-Y24H mutant, at pH 5.8, the fusion index was higher for the double mutant, indicating a shift of the fusion threshold toward neutral pH for this virus (Fig. 6E).
Altogether, these results show that E1-Y24H and E2-P173S mutations qualitatively change the fusion phenotype, shifting the pH threshold for the double mutant, which confirms the functional interaction between these two residues.
## Mouse and mosquito infections with E2-P173S and E1-Y24H single and double mutants
Previous studies indicated that altered fusion phenotypes impact virus infectivity in vivo (20,21,28,29). Given that mutations associated with antiviral resistance changed the fusion phenotype, we decided to characterize the impact of mutations in vivo. We performed infections in mice with the WT and mutant viruses. Two days post-infection, we quantified footpad viral titers that reflect virus replication at the site of infection and viremia as a proxy of dissemination (Fig. 7A andB). Footpad viral titers were higher for mutant viruses than for the WT, with the double mutant displaying the highest viral yield. In this line, viremia was also higher for mutant viruses, showing a positive correlation with the infection level in the primary infection site.
Next, we evaluated the fitness of mutant viruses in the vector host. Similar to mammalian cell lines, CHIKV entry into mosquito cells is pH dependent. However, differences in receptor usage and alternative entry pathways have been described for the different host cell lines (30). We first carried out growth curves of WT and mutant viruses in the mosquito C6/36 (Ae. albopictus) cell line (Fig. 8A). We found that both E2-P173S and the double mutant reached a lower viral titer than WT (fourfold), and the growth of E1-Y24H was delayed. Profiling of the antiviral activity of compound 11 showed that the WT virus was sensitive to compound 11 (EC 50 = 12.87 µM). As expected, the double mutant virus displayed a resistant phenotype (EC 50 = 48.47 µM). Interestingly, single mutants were partially resistant to treatment with the E2-P173S virus displaying a higher EC 50 (E2-P173S EC 50 = 43.39 µM vs. E1-Y24H EC 50 = 23.71 µM) (Fig. 8B). Finally, we assessed virus fitness following feeding of Ae. aegypti mosquitoes with an infectious blood meal. In line with in vitro data, all mutant viruses infected a lower percentage of mosquitoes than WT (Fig. 8C), with a statistically significant reduction for E2-P173S and the double mutant. Interestingly, viral titers in the bodies of infected mosquitoes were similar for the double mutant and WT (Fig. 8D), suggesting a defect for the mutant virus to overcome the midgut infection barrier to establish an infection (31). In contrast, the viral titer for E2-P173S single mutant was twofold lower than WT, suggesting an impact of the mutation on viral fitness in mosquitoes. Moreover, the result further shows an effect of E1-Y24H on the double-mutant phenotype.
Overall, these results show an opposite impact of E2-P173S and E1-Y24H on viral fitness in the alternate hosts, with a beneficial effect in mice and a detrimental effect in mosquitoes. Noteworthy, while competition experiments indicated that mutations appeared to be associated with loss of fitness in untreated cells in vitro (Fig. 2E), they resulted in increased fitness in mice with the double mutant displaying a more pro nounced phenotype than single mutants. In turn, mutations compromised virus fitness in vitro and in the mosquito vector. In addition, E1-Y24H partially rescued the deleterious effect of E2-P173S in mosquitoes, altogether reinforcing the functional interaction between these two residues.
## DISCUSSION
In this work, we studied the mechanism of action of a previously identified small molecule inhibitor of CHIKV infection and characterized the phenotype of viruses carrying mutations associated with antiviral resistance. Using complementary approaches, we demonstrated that compound 11 specifically inhibits the pH-dependent membrane fusion driven by CHIKV envelope proteins during entry. Characterization of mutant viruses indicated a functional interaction between E2-P173 and E1-Y24: viruses with the combined mutations E2-P173S and E1-Y24H displayed a distinct resistance to compound 11, likely associated with a shift in the fusion threshold toward higher pH, and showed a significantly increased replication in mice and impaired rate of infection in mosquitoes. We propose that E2-P173S and E1-Y24H cooperatively act to engage the heterodimer in its conformational rearrangement at a pH closer to neutral in comparison to WT.
## Functional interaction between E2-P173 and E1-Y24 in the regulation of the fusion process
The results presented here support that E2-P173 and E1-Y24 work together in the regulation of the fusion process despite being distantly located in the envelope proteins of CHIKV. Interestingly, a connection between nearby residues has been previously observed. Infection of mosquitoes with a virus carrying an E1-V80Q mutation, close to E2-P173S, resulted in the selection of pseudorevertants with the second-site mutation E1-N20Y, near E1-Y24H (21). In the same line, the emergence of the epistatic mutations E1-K211T and E1-V156A in the context of the Indian Ocean lineage virus carrying a valine at E1-226 was associated with increased fusion at lower pH and higher sensitivity to NH 4 Cl treatment, suggesting a functional connection between residues 156, 211, and 226 in the entry dynamics (6). Taken together with our results indicating that compensa tory mutations E2-P173S and E1-Y24H modulate the fusogenic function, these results suggest a link between two distant regions of the envelope complex.
Enveloped virus entry culminates with the fusion of viral and cellular membranes and the delivery of the viral genome into the host cell cytosol. Virus envelope proteins anchored to the lipid envelope perform fusion following activation triggered by an external cue such as receptor binding or exposure to acidic pH. A common pathway for fusion involves protein conformational rearrangement from an initial prefusion state to the insertion of a fusion motif in the target membrane in an extended prehairpin state and final folding back into a post-fusion hairpin state (32). Beyond the commonalities in the fusion process, virus fusion proteins diverge in structure and organization on the virus surface. Three canonical classes of fusion proteins are recognized based on their structure. Alphavirus E1 is a class II fusion protein. Cryo-electron tomography has recently allowed us to visualize the pathway of CHIKV membrane fusion, and together with cryo-electron microscopy and crystallography, allowed us to reconstruct conforma tional rearrangements of the mature E1-E2 heterodimer (9,32,33). Still, details on the dynamics of molecular transitions at the residue level remain unresolved. Under acidic conditions, the central arch of E2 β-ribbon becomes disordered and domain B opens to expose the fusion loop at the tip of E1 (12,22,34). E2-P173 is located in a breakpoint in residue connectivity associated with the arch 2 of the β-ribbon, and together with structural analyses, our MD simulations indicate that this structure acts as a hinge for the movement of domain B at the beginning of the fusion process. In turn, E2 dissociates from E1, which rearranges as a homotrimer in the post-fusion state. MD results showed that the E1-Y24 loop strongly interacts with residues in the flexible linker between E1 domains I and III (E1-R289, E1-R21, E1-F287, E1-Y1, and E1-E284) in the pre-fusion conformation. Importantly, this interaction is broken during the conformational change of the envelope proteins toward the post-fusion state. Interestingly, E1-F287 and E1-R289 have been identified as a part of a different interaction network that stabilizes the post-fusion E1 trimer (35), suggesting that the conformational transition of the linker could be important in the regulation of the fusion process. Altogether, we hypothesize that E2-P173 and E1-Y24 are linked in the modulation of important kinetic barriers of the conformational change of the envelope proteins of CHIKV, acting as checkpoints of the fusion process. On the one hand, E2-P173S could alter the flexibility of the arch 2 of the β-ribbon, imposing a kinetic barrier to the opening of domain B. On the other hand, protonation of E1-H24 in E1-Y24H would lower the stability of the pre-fusion interaction network that stabilizes the linker, altering the energetic barrier that E1 needs to overcome in the transition between the pre-fusion and post-fusion states. The fact that the fusion phenotype is only altered in the double mutant suggests a multistep reaction with sequential checkpoints. As the dissociation of E1 and E2, the assembly of the post-fusion trimer, and the joint action of different trimers in the formation of the fusion pore require the coordination between different E1-E2 dimers in the viral particle (10,16), the relative rates of the different checkpoints may be fundamental for the coordination of the heterodimers.
## Impact of E2-P173S and E1-Y24H on in vivo infectivity
E2-P173S and E1-Y24H had opposite effects on fitness in experimental infection of mosquitoes and mice. They resulted in impaired infection rates in mosquitoes but enhanced viral loads in mice relative to the WT virus. Further supporting a connec tion between these residues, in both mosquitoes and mice, the phenotype was more pronounced for the double mutant. Interestingly, the results highlight the evolutionary barrier imposed over viruses that alternate between hosts to develop resistance to an antiviral in the natural infection cycle (36). Mutations that confer resistance in the mammalian host are not necessarily adaptive in the alternate host and may even have a deleterious effect (37), preventing the fixation of the mutation in a natural population.
In further studies, it would be interesting to address the link between the change in the fusion phenotype, an increase in the fusion threshold, and the modulation of the viral fitness in the natural hosts. Notably, an opposite relationship between the fusion phenotype and the fitness in mosquitoes was observed in previous studies, where E1-V80L/Q mutants displayed a decreased fitness in mosquitoes associated with a shift of the fusion threshold toward more acidic pH (20,21). These results suggest a complex relationship between the different functions of the fusion machinery that warrants further investigation.
In conclusion, in this study, we characterized the mechanism of action of a small molecule inhibitor of CHIKV and showed that the compound inhibits the membrane fusion process during CHIKV entry. Additionally, the study of mutations associated with resistance to the antiviral activity of compound 11 identified a functional interaction between two distant residues in the envelope proteins of the virus that, together, impact the fusion phenotype and the in vivo fitness of CHIKV.
## MATERIALS AND METHODS
## Cells and viruses
Vero (Cercopithecus aethiops kidney, ATCC CCL-81), HEK-293T cells (human embryonic kidney cells expressing SV40 T antigen, ATCC CRL-3216), and Huh-7 cells (human hepatoma cells, provided by Apath LLC) were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco). BHK cells (Mesocricetus auratus hamster kidney, ATCC CCL-10) were grown in MEM alpha medium (Gibco). All mammalian cell lines were grown at 37°C in a 5% CO 2 atmosphere in medium supplemented with 10% fetal bovine serum (FBS), 100 U/mL of penicillin, and 100 µg/mL of streptomycin (Gibco). C6/36 cells (Aedes albopictus larvae, ATCC CRL-1660) were grown at 28°C in Leibovitz L-15 medium (Gibco) supplemented with 10% FBS, 10% of tryptose phosphate (29.5 g/L, Britania), 100 U/mL of penicillin, 100 µg/mL of streptomycin, and 250 ng/mL of amphotericin B (Gibco).
WT and mutant CHIKV-ZsGreen of the Indian Ocean Lineage (Eastern Central and South Africa genotype) were derived from infectious cDNA clones as previously described. Viral titers of the final stocks were determined by plaque assay. Viral stocks were stored at -70°C until use. The construction of E1-Y24H and E2-P173S single and double mutants is described elsewhere (17).
## Compound
The synthesis and purification of compound 11 were previously described (17). For biological assays, compound 11 was first dissolved in dimethyl sulfoxide (DMSO) to a final concentration of 10 mM and then diluted in culture medium to the appropriate concentration. In all assays, the DMSO concentration was lower than 1%.
## Pseudotypes production and inhibition assay
The CHIKV ORF encoding for viral structural proteins (C-E3-E2-6K-E1) was cloned into pCI-Neo (Promega) to obtain pCI-neo-CHIKV.
HEK-293T cells at 70% confluence were co-transfected with plasmids psPAX2 (10 µg, Addgene #12260), pLB-GFP (10 µg, Addgene #11619), and pCI-neo-CHIKV or pMD2.G (2 µg, Addgene #12259) using polyethylenimine (PEI) reagent. Pseudotypes were harvested 2 days post-transfection and were concentrated by centrifugation at 3,000 × g 4°C overnight.
For the determination of compound 11 inhibitory activity of pseudotype infection, BHK cells at 50% confluence were treated with increasing concentrations of compound 11 and were infected with pseudotypes of CHIKV or VSV at a multiplicity of infection (MOI) of 0.05. 10 µg/mL of Polybrene (Sigma) was added to enhance pseudotype adsorption. Three days after infection, the percentage of infected cells was quantified by flow cytometry.
## Fusion from without
Confluent BHK cells were treated with increasing concentrations of compound 11 and infected with CHIKV-ZsGreen WT or mutant viruses at an MOI of 0.1 at room temperature. Then, the inoculum was removed, and a phosphate-buffered saline (PBS) and citric acid solution was titrated to an acidic pH, with the corresponding concentration of compound 11 added. After incubating for 2 minutes at room temperature, the acidic medium was removed, cells were washed with PBS, and α-MEM 20 mM NH 4 Cl was added. Cells were analyzed by flow cytometry at 1 day post-infection.
## Cell-cell fusion
BHK cells at 80% confluence grown on a coverslip were transfected with the pCIneo-GFP-CHIKV C-E2-E1 plasmid using PEI reagent. One day post-transfection, cells were treated with increasing concentrations of compound 11 in culture medium for 40 minutes at room temperature. Subsequently, the medium was removed, and compound 11 was added in PBS titrated to an acidic pH with citric acid. After 2 minutes at room temperature, the acidic medium was removed, the cells were washed three times with PBS, and incubated for 4 hours at 37°C before fixation with 4% PFA. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI; Molecular Probes) and fluorescence microscopy images were captured using a Nikon Eclipse 80i Fluorescence microscope equipped with a DS-Qi1Mc camera with a 100× magnification.
Image analysis was followed to quantify syncytium formation in each treatment. Briefly, nuclei and GFP-expressing cells were segmented in every image using local threshold and watershed algorithms. Afterward, GFP-expressing cells were manually classified into syncytium or non-fused cells based on the number of nuclei and morphology. All cells with one nucleus were automatically considered an individual cell. For every image, a fusion index was calculated as 1 -C/N, where C is the number of GFP-expressing cells and N is the number of nuclei within GFP-expressing cells in the image. Six random fields were taken per treatment and three independent experiments were performed.
## Plaque assay
Confluent Vero cells in 24-well plates were infected with serial dilutions of viral samples and were incubated for 1 h at 37°C. Afterward, 1 mL of overlay (DMEM, 2% FBS, 0.4% methylcellulose) was added to each well. Three days post-infection, cells were fixed with 10% formaldehyde and were stained with crystal violet solution (20% ethanol, 0.1% crystal violet in water) to allow plate lysis count.
## Growth curves
Confluent Vero or C6/36 cells grown in 24-well plates were infected with WT or mutant CHIKV-ZsGreen at an MOI of 0.01. After incubating for 1 hour at 37°C for mammalian cells infection or 28°C for mosquito cells infection, the inoculum was removed, and the cells were washed three times with PBS. 750 µL of culture medium was added, and the cells were incubated for 3 days at the corresponding growth temperature (see "Cells and Viruses"). At various time points, a 50 µL aliquot was removed, replaced with culture medium, and stored at -70°C. The viral titer at each time point was determined by plaque assay.
## Viral yield inhibition assay
Confluent Vero cells were treated with increasing concentrations of compound 11 and were infected with CHIKV-ZsGreen WT or mutant viruses at an MOI of 0.01. Cells were incubated for 1 hour at 37°C and then overlaid with culture medium containing the corresponding concentration of compound 11. Two days after the infection, the number of viral particles in the supernatant was titrated by plaque assay.
## Competition assay
Virus stocks were diluted and mixed at a 1:1 ratio to infect confluent Vero cells grown in 24-well plates at an MOI of 0.01. Supernatants of cells treated with compound 11 at 50 µM or mock-treated were collected 3 days after the infection, and RNA was extracted using the Quick RNA Viral kit (Zymo Research). The cDNA corresponding to the regions coding for E2 and E1 was obtained by RT-PCR using SuperScript III reverse transcriptase (Invitrogen) and primers 137 (5′-CGTTTGTAGATAACTGCGG-3′) and 95 (5′-TACTTAATTG TCGAGCTCTTAGTGCCTGCTGA-3′) for first-strand synthesis of E2 and E1, respectively. Then, Pfx Accuprime polymerase (Invitrogen) and primers 136 (5′-GAAGAGTGGAGTCTT GCC-3′) and 137, and 93 (5′-GACTGAAGGGCTCGAGGTCA-3′) and 95 were used for PCR amplification of E2 and E1 fragments, respectively. PCR amplicons were sequenced by the Sanger method and chromatograms analyzed using 4Peaks software (Nucleobytes).
## Molecular dynamics and residue interaction network
We performed molecular dynamics simulations of the CHIKV pre-fusion E1-E2 hetero dimer in solution (PDB 3N42) using GROMACS 2021.2 (38). The protonation state at pH 7 of all protonable residues was defined using PROPKA (39). The Amber99SB*-ILDN force field (40,41) was used to describe the protein, and a cutoff value of 10 Å was used for short-range electrostatic and van der Waals interactions. Long-range electrostatic interactions were treated with PME (Particle Mesh Ewald). The system was solvated using a dodecahedral box of TIP3P water (https://conan.io/) extending 12 Å from the protein's surface. The system was neutralized with 0.15 M NaCl. After, the system was minimized, heated to 310 K, and equilibrated for 200 ps in the NVT ensemble using the V-rescale thermostat (39) with a coupling constant of 0.1 ps. A second equilibration step of 1 ns was carried out in the NPT ensemble using the Berendsen barostat (https:// networkx.org/) with a reference pressure of 1 bar and a coupling constant of 2 ps. Finally, the production run was conducted in the NPT ensemble using the V-rescale thermostat and the Parrinello-Rahman barostat (42). In the equilibration and production runs, the LINCS algorithm (43) was used, and a time step of 2 fs was employed. Four independent runs of 250 ns were carried out with the WT protein. For analysis, we concatenated the last 200 ns of each run. The molecular dynamics simulations were carried out on high-performance computing centers CCAD (https://ccad.unc.edu.ar/) and the High-Per formance Computing Portal at the NYU Grossman School of Medicine.
From the concatenated trajectory, a PCA was conducted with GROMACS tools. The RIN was built from the contact matrix and the DCCM following the guidelines outlined in the work of Sethi and colleagues (23). The contact matrix was obtained using the CONAN tool (https://conan.io/). Two residues were considered to be in contact if they remained within 4.5 Å of each other for more than 75% of the dynamics. The DCCM was derived from the covariance matrix obtained in the PCA with GROMACS. The RIN was constructed using a Python script based on the NetworkX package (https://networkx.org/). A node was defined for each residue, and an edge was established between two nodes if they were in contact during the dynamics. The weight of each edge was defined as the information transfer probability d ij obtained from the correlation of each pair of residues in the dynamics (22). From the RIN, the betweenness centrality of each edge was calculated, and the residue communities were identified using the Girvan-Newman algorithm (24) with NetworkX methods.
## Viral particle stability
WT or mutant CHIKV-ZsGreen stocks were diluted to 5 × 10 4 PFU/mL in serum-free DMEM. At different time points (0, 8, and 24 hours), an aliquot was frozen at -70°C, and the viral titer for each treatment was quantified by plaque assay.
## Viral attachment
Confluent BHK or Vero cells grown in 12-well plates were infected with WT or mutant CHIKV-ZsGreen at an MOI of 0.1. After incubating for 40 minutes at room temperature, cells were washed three times with PBS and harvested in culture medium using a cell scraper. Cells were lysed by three cycles of freezing in liquid nitrogen and thawing at 37°C, and the supernatant was clarified by centrifugation for 10 minutes at 1,000 × g and 4°C. The number of viral particles in the clarified supernatant was determined by plaque assay.
## Cholesterol dependence
A 25 mM MβCD (Thermo Fisher) solution in α-MEM with 25 mM HEPES at pH 7.5 was prepared for immediate use. Confluent BHK cells were treated with increasing concentra tions of MβCD and were incubated for 1 hour at 37°C. Subsequently, cells were washed three times with PBS and were infected with CHIKV-ZsGreen WT or mutant viruses at an MOI of 0.1. After incubating for 1 hour at 37°C, the inoculum was removed, the cells were washed three times with PBS and medium containing 20 mM NH 4 Cl was added to prevent reinfection. One day post-infection, the percentage of infected cells was determined by flow cytometry.
## Experimental infection in mice
Four-to five-week-old male and female C57BL/6J mice (n = 10 per group) were infected subcutaneously in the footpad with 1,000 PFU of WT or each of the mutant CHIKV-ZsGreen. At 2 days post-infection, mice were euthanized to collect blood via cardiac puncture and to harvest the footpad. The footpad was ground in 1 mL of DMEM containing 2% FBS with steel beads using a Tissue-Lyser II (Qiagen), and debris was clarified by centrifugation at 8,000 × g for 10 minutes. Viral titers in the footpad were quantified by plaque assay on Vero cells. For serum, whole blood was centrifuged at 4,000 × g for 15 minutes, and the serum was placed in Trizol. RNA was extrac ted following the manufacturer's instructions, and CHIKV genomes were quantified by RT-qPCR (Applied Biosystems RNA-to-Ct one-step kit) with the following primers targeting CHIKV nsP4: 5′-TCACTCCCTGCTGGACTTGATAGA-3′ and 5′-TTGACGAACAGAG TTAGGAACATACC-3′, and probe: 5′-(6-carboxyfluorescein)-AGGTACGCGCTTCAAGTTCGG CG-(black-hole quencher)-3′. In vitro transcribed CHIKV RNA was used to generate a standard curve. All RT-qPCR samples were run in technical duplicates. All mouse experiments were performed in the biosafety level 3 facility ABSL3 at the NYU Grossman School of Medicine.
## Experimental infection in mosquitoes
Four to seven-7 days post-emergence, female Ae. aegypti mosquitoes (Poza Rica, Mexico; F40) were fed a blood meal containing 10 6 PFU/mL of WT or each of the mutant CHIKV-ZsGreen supplemented with 5 mM ATP for 30 minutes. Engorged females were sorted and incubated at 28°C, 70% humidity, and 12 h diurnal light cycle with 10% sucrose ad libitum for 7 days. After incubation, whole mosquitoes were ground in 300 mL of PBS with ceramic beads using a Tissue-Lyser II (Qiagen), and debris was removed by centrifugation at 8,000 × g for 10 minutes. Viral titers in the bodies were quantified in the bodies by plaque assay. All mosquito studies were performed in the NYU Grossman School of Medicine ABSL3 facility.
## References
1. Weaver, Lecuit (2015) "Chikungunya virus and the global spread of a mosquito-borne disease" *N Engl J Med*
2. Schilte, Staikowsky, Couderc et al. (2013) "Chikungunya virus-associated long-term arthralgia: a 36-month prospective longitudinal study" *PLoS Negl Trop Dis*
3. Puntasecca, King, Labeaud (2021) "Measuring the global burden of chikungunya and Zika viruses: a systematic review" *PLoS Negl Trop Dis*
4. Morrison (2014) "Reemergence of chikungunya virus" *J Virol*
5. Tsetsarkin, Mcgee, Volk et al. (2009) "Epistatic roles of E2 glycoprotein mutations in adaption of chikungunya virus to Aedes albopictus and Ae. aegypti mosquitoes" *PLoS One*
6. Tsetsarkin, Vanlandingham, Mcgee et al. (2007) "A single mutation in chikungunya virus affects vector specificity and epidemic potential" *PLoS Pathog*
7. Strauss, Strauss (1994) "The alphaviruses: gene expression, replication, and evolution" *Microbiol Rev*
8. Holmes, Basore, Fremont et al. (2020) "A molecular understanding of alphavirus entry" *PLOS Pathog*
9. Voss, Vaney, Duquerroy et al. (2010) "Glycoprotein organization of Chikungunya virus particles revealed by X-ray crystallography" *Nature*
10. Guardado-Calvo, Rey (2021) "The viral class II membrane fusion machinery: divergent evolution from an ancestral heterodimer" *Viruses*
11. Sun, Xiang, Akahata et al. (2013) "Structural analyses at pseudo atomic resolution of Chikungunya virus and antibodies show mechanisms of neutralization"
12. Li, Jose, Xiang et al. (2010) "Structural changes of envelope proteins during alphavirus fusion" *Nature*
13. Sánchez-San Martín, Nanda, Zheng et al. (2013) "Cross-inhibition of chikungunya virus fusion and infection by alphavirus E1 domain III proteins" *J Virol*
14. Sánchez-San Martín, Sosa, Kielian (2008) "A stable prefusion intermediate of the alphavirus fusion protein reveals critical features of class II membrane fusion" *Cell Host Microbe*
15. (1016)
16. Sahoo, Gudigamolla, Chowdary (2020) "Acidic pH-Induced conformational changes in chikungunya virus fusion protein E1: a spring-twisted region in the domain I-III linker acts as a hinge point for swiveling motion of domains" *J Virol*
17. Kielian, Vos, Liao (2010) "Alphavirus entry and membrane fusion" *Viruses*
18. Battini, Fidalgo, Álvarez et al. (2021) "Discovery of a potent and selective Chikungunya virus envelope protein inhibitor through computer-aided drug design" *ACS Infect Dis*
19. Sourisseau, Schilte, Casartelli et al. (2007) "Characterization of reemerging chikungunya virus" *PLoS Pathog*
20. Sun, Yau, Briggs et al. (2005) "Role of clathrin-mediated endocytosis during vesicular stomatitis virus entry into host cells" *Virology (Auckland)*
21. Tsetsarkin, Mcgee, Higgs (2011) "Chikungunya virus adaptation to Aedes albopictus mosquitoes does not correlate with acquisition of cholesterol dependence or decreased pH threshold for fusion reaction" *Virol J*
22. Noval, Rodriguez-Rodriguez, Rangel et al. (2019) "Evolution-driven attenuation of alphaviruses highlights key glycoprotein determinants regulating viral infectivity and dissemination" *Cell Rep*
23. Merwaiss, Filomatori, Susuki et al. (2021) "Chikungunya Virus replication rate determines the capacity of crossing tissue barriers in mosquitoes" *J Virol*
24. Mahammad, Parmryd (2015) "Cholesterol depletion using methyl-βcyclodextrin" *Methods Mol Biol*
25. Sethi, Eargle, Black et al. (2009) "Dynamical networks in tRNA:protein complexes" *Proc Natl Acad Sci*
26. Girvan, Newman (2002) "Community structure in social and biological networks" *Proc Natl Acad Sci*
27. Kephart, Hom, Lee (2024) "Visualizing intermediate stages of viral membrane fusion by cryo-electron tomography" *Trends Biochem Sci*
28. Prasad, Blijleven, Smit et al. (2022) "Visualization of conformational changes and membrane remodeling leading to genome delivery by viral class-II fusion machinery" *Nat Commun*
29. Chen, Klose, Sun et al. (2022) "Cryo-EM structures of alphavirus conforma tional intermediates in low pH-triggered prefusion states" *Proc Natl Acad Sci*
30. Cao, Zhang (2013) "Characterization of an early-stage fusion intermediate of Sindbis virus using cryoelectron microscopy" *Proc Natl Acad Sci*
31. Zheng, Sánchez-San Martín, Qin et al. (2011) "The domain Idomain III linker plays an important role in the fusogenic conformational change of the alphavirus membrane fusion protein" *J Virol*
32. Coffey, Vignuzzi (2011) "Host alternation of Chikungunya virus increases fitness while restricting population diversity and adaptability to novel selective pressures" *J Virol*
33. Jvi
34. Pal, Fox, Hawman et al. (2014) "Chikungunya viruses that escape monoclonal antibody therapy are clinically attenuated, stable, and not purified in mosquitoes" *J Virol*
35. Coffey, Vasilakis, Brault et al. (2008) "Arbovirus evolution in vivo is constrained by host alternation" *Proc Natl Acad Sci*
36. Abraham, Murtola, Schulz et al. (2015) "GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers" *SoftwareX*
37. Lindorff-Larsen, Piana, Palmo et al. (2010) "Improved side-chain torsion potentials for the Amber ff99SB protein force field" *Proteins*
38. Best, Hummer (2009) "Optimized molecular dynamics force fields applied to the helix-coil transition of polypeptides" *J Phys Chem B*
39. Jorgensen, Chandrasekhar, Madura et al. (1983) "Comparison of simple potential functions for simulating liquid water" *J Chem Phys*
40. Bussi, Donadio, Parrinello (2007) "Canonical sampling through velocity rescaling" *J Chem Phys*
41. Berendsen, Postma, Van Gunsteren et al. (1984) "Molecular dynamics with coupling to an external bath" *J Chem Phys*
42. Parrinello, Rahman (1981) "Polymorphic transitions in single crystals: a new molecular dynamics method" *J Appl Phys*
43. Hess, Bekker, Berendsen et al. (1997) "LINCS: A linear constraint solver for molecular simulations" *J Comput Chem*
44. Katoh, Rozewicki, Yamada (2019) "MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization" *Brief Bioinformatics*
45. Robert, Gouet (2014) "Deciphering key features in protein structures with the new ENDscript server" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12587602&blobtype=pdf | # Immunocompromised patients with persistent SARS-CoV-2 viral shedding ≥8 weeks, clinical outcomes, and virological dynamics: a retrospective multicenter cohort study, 2020-2024
Clémentine De, La Porte Des Vaux, Nicolas Veyrenche, Kevin Da Silva, Nathalie Chavarot, Marianne Burgard, Olivier Paccoud, Florence Runyo, Margaux Garzaro, Claire Rouzaud, Alexandra Serris, Damien Vimpere, Dany Anglicheau, Luc Mouthon, Olivier Hermine, Marie-Anne Rameix-Welti, Fanny Lanternier, Olivier Lortholary, Cléa Melenotte
## Abstract
Immunocompromised patients (ICPs) infected with SARS-CoV-2 are at higher risk of severe illness. Some experience persistent viral shedding beyond eight weeks, which is associated with increased mortality and invasive fungal infections. However, data on the clinical profile, treatment impact, and standardized management for these patients remain limited. We conducted a retrospective cohort study at Groupe Hospitalier Paris Centre between March 1, 2020, and February 10, 2024. We assessed symptomatic ICPs with persistent SARS-CoV-2 shedding (>8 weeks), analyzing clinical progression, time to viral clearance, and emergence of resistance mutations in rela tion to treatment regimens. Fifty-three patients were included: 53% were solid organ transplant (SOT) recipients, 42% had hematological malignancies (HMs), and 5% had other immunosuppressive conditions. Severe infections occurred in 32%, 91% required hospitalization, and 17% (n = 9) presented invasive mold infections. SOT recipients achieved clinical cure faster than HM patients (P < 0.01). Patients treated with direct antivirals showed significantly faster viral clearance (P = 0.03) than those treated with monoclonal antibodies (mAbs) or convalescent plasma. No resistance mutations emerged against remdesivir or nirmatrelvir/ritonavir. However, 54% of viral strains showed initial or acquired spike protein resistance to mAbs. Direct antiviral therapies, particularly remdesivir and nirmatrelvir/ritonavir, appear safe and effective in promoting faster viral clearance and clinical recovery in ICPs with persistent symptomatic SARS-CoV-2 infection. KEYWORDS immunocompromised patients, persistent SARS-CoV-2 shedding , solid organ transplant (SOT) , direct antivirals , viral resistance T he COVID-19 pandemic has now largely been controlled, thanks to the widespread use of COVID-19 vaccines, with incidence rates drastically decreasing worldwide over the winter of 2024-2025 (1). More than three and a half years after the pandemic, the virus continues to evolve genetically, with current variants now showing high transmis sibility (RX) and resistance to all monoclonal antibody therapies (mAb) (2). The Omi cron JN.1 variant and its descendants are currently dominant worldwide, with ongoing diversification into sub-lineages, including KP.3.1.1 or NB.1.8.1 (3). Immunocompromised patients (ICPs), representing nearly 3% to 6% of the general population in developed countries, remain at high risk for severe and persistent SARS-CoV-2 infections, which are associated with high mortality rates (4, 5). Prolonged SARS-CoV-2 viral sheddinglasting over eight weeks-has been linked to persistent symptoms, spike protein mutations, hospitalization, invasive aspergillosis, and an increased risk of death (6-11). Hence, clinicians continue to face difficulties in managing this immunocompromised
population affected by SARS-CoV-2-including patients with persistent shedding-as there are no standardized, consensus-based treatment guidelines, due to the limited number of prospective studies on this population (12,13). Persistent viral shedding can be observed after exposure to antiviral drugs that fail to eradicate the virus and could potentially induce resistance mutations and emergence of new variants (14)(15)(16)(17). We present here a retrospective cohort study describing the clinical outcome, viral clearance, and virological characteristics, including the incidence of resistance-associated mutations that emerged during persistent SARS-CoV-2 viral shedding (>8 weeks) in ICPs according to the treatments they received.
## MATERIALS AND METHODS
## Study design and definitions
This retrospective cohort study was conducted at the Groupe Hospitalier Paris Centre and included immunocompromised adult patients with persistent SARS-CoV-2 viral shedding lasting over 56 days (>8 weeks) from March 1, 2020, to February 10, 2024.
## Patients definition and inclusion criteria
Persistent viral shedding was defined by at least two positive SARS-CoV-2 PCR tests (nasopharyngeal) over a period exceeding 8 weeks, with no negative tests or confirmed reinfections during this time (unless there was evidence of a positive repeat qPCR test within two days after a first negative nasopharyngeal qPCR, which was thus considered as a false negative). In the absence of a consensual definition of persistent viral shedding in immunocompromised patients, and in line with the definition recently proposed by Machkovech et al. (>30 days), we have chosen to work on an even more specific population with very prolonged persistent viral shedding >8 weeks, which often also presents numerous complications and therapeutic difficulties. This work is a continuation of a previously published study (6).
Included ICPs had conditions such as primary immune deficiencies, HIV with CD4 counts <200/mm³, autoimmune diseases, solid organ transplants (SOT), allogeneic hematopoietic stem cell transplants, chronic lymphoid malignancies, or were receiv ing immunosuppressive therapies. The management of COVID-19 patients within the Paris Centre hospital group was comparable and homogeneous in all the departments participating in this study. It followed the recommendations of the multidisciplinary consultation meetings. Asymptomatic patients without clinical data and those with reinfections were excluded. Severity was categorized based on the 2024 WHO criteria, defining severe infections as those requiring hospitalization with >3L oxygen and/or corticosteroids or tocilizumab (18).
## Data collection and outcome measures
The data collected included clinical and radiological presentations, SARS-CoV-2 variant, qPCR Ct values, underlying conditions, immunosuppressive treatments, and specific anti-COVID-19 therapies. Clinical and radiological outcomes were also collected, as well as kinetics of viral loads and sequencing after treatment (or during the time of infection in untreated patients) and the occurrence of complications: death, hospitalization, and probable or proven invasive mold infections (IMIs), according to EORTC/MSG definitions (18,19). Outcomes included clinical cure (disappearance of symptoms), viral clearance (i.e., qPCR negativity), radiological resolution on CT scan, and complications, such as IMIs, hospitalization, and death. In the case of multiple treatments, two treatments were part of the same line if their initiation did not differ by more than five days.
## Sample preparation and sequencing
Sequencing analysis was performed for patients who had at least two available samples, spaced at least one week apart to ensure robust temporal data. In the case of treatment, we selected available samples before and after receiving anti-SARS-CoV-2 treatments. For optimal sequencing, among the samples collected, before and after treatment, we selected those with the lowest available Ct values in patients with cycle thresh old (Ct) <28. Samples that were not exploitable due to poor sequencing results or poor-quality consensus were substituted by another available sample meeting optimal quality criteria. This selection process ensured reliable temporal consistency, treatment relevance, and data quality across the study cohort. Among the selected samples, nucleic acids were extracted using the NucleoSpin 8 virus Core Kit (Macherey-Nagel), followed by dual RT-PCR with LunaScript Supermix (NEB) and Q5 High-Fidelity DNA polymerase (NEB), using a pool of primers (ARTIC V5.3.2 from https://github.com/artic-network/pri mer-schemes/tree/master/nCoV-2019/V5.3.2). Purified amplicons were sequenced at the Mutualized Platform for Microbiology using the Nextera XT DNA Library Prep kit (Illumina) on the NextSeq 500/2000 systems (Illumina Inc.).
## Mutation and resistance analysis
For each sample, a consensus sequence was generated using an in-house bioinformatics pipeline. This pipeline includes read mapping with minimap2 (v2.26) and BWA (v0.7.17) and variant calling and consensus building with iVar (v1.3.1). The consensus sequence, reflecting the dominant viral population, was constructed by incorporating nucleotide positions with read frequencies above 60%. Minor variants below 40%, which can influence viral evolution and drug resistance, were also noted. For minor variants, a 5% frequency threshold was applied to exclude low-frequency noise. Mutations with read frequencies between 40% and 60% were marked with ambiguous nucleotides to account for intra-host diversity (20). Lineage assignment and phylogenetic annota tion were performed using Nextclade (v3.8.2), following the global SARS-CoV-2 clade system and assessing sequence quality. A phylogenetic tree, with samples color-coded by patient and Pango lineage, was generated using multiple sequence alignment with MAFFT (v7.525), and phylogenies were inferred using IQ-TREE (v2.2.2.2) with the maximum likelihood and general time-reversible model. Drug-resistance mutations were screened in both consensus sequences and minor variants using the Stanford Coronavirus Antiviral & Resistance Database (updated 5/14/2024; https://covdb.stan ford.edu/susceptibility-data/table-mab-susc/). Only mutations identified in at least one cohort sample were reported with their corresponding drugs.
## Statistical analysis
To compare the distribution of continuous and dichotomous variables between two groups, we used the χ 2 test or the two-sided Fisher exact test, respectively. All tests were two-sided, and P < 0.05 was considered significant. Therapeutic impact (e.g., sotrovimab, nirmatrelvir/ritonavir, remdesivir, casirivimab/imdevimab, tixagévimab/cilgavimab, and plasmatherapy) was evaluated using χ² or Fisher's exact tests for categorical data and log-rank tests for survival analyses, with significance set at P < 0.05. Analyses were performed with STATA v17.0.
## RESULTS
We included 53 ICPs with persistent viral shedding (>8 weeks) during the inclusion period. Omicron was the most represented variant (86%) in this cohort.
## Characteristics of patients at diagnosis
Patients' characteristics at diagnosis are displayed in Table 1. The median age was 60 years old (interquartile range [IQR]: [51-71]), and 64% were men. Twenty-eight patients (28/53; 53%) were SOT recipients (two heart, 23 kidney, two lungs, one combined kidney and liver), while 42% had hematological malignancies (HMs) (22/53; including two of the SOT recipients for whom the HM was considered as the main underlying disease). The most represented HMs were non-Hodgkin lymphoma (12/22), chronic lymphocytic leukemia (3/22), and acute myeloid leukemia (2/22, including one patient with hema topoietic stem cell transplantation, HSCT). Other HMs included one case of multiple myeloma, one case of Waldenström's macroglobulinemia, one case of myelodysplastic syndrome (with refractory cytopenia), one case of prolymphocytic leukemia, and one case of tricholeukocytic leukemia. Nine percent (5/53) had another type of immuno suppression. In SOT recipients, the median time from transplantation to SARS-CoV-2 diagnosis was 25 (3-96) months.
Among the 53 patients included, the data on prophylaxis received was not availa ble for eight of them. Among the available data, 16 received no prophylaxis, whether vaccine or monoclonal antibodies. Among the 29 patients who received prophylaxis, three received one dose of vaccine, seven received two doses, 15 received three doses, and four received four doses of vaccine. Among those who received two doses, two also received tixagevimab/cilgavimab as a prophylaxis, while 10 of those who received three doses received monoclonal antibodies prophylactically (seven tixagevimab/cilgavimab
## Clinical evolution and complications
Forty-eight patients were hospitalized during their SARS-CoV-2 infection (Table 2): 13 were admitted to the day hospital for upper or lower respiratory tract infection that did not require oxygen therapy, and 35 were hospitalized in conventional units. Among them, 29 had pneumonia; six presented alterations of the general status, including three with weight loss, two with prolonged fever, and two with acute kidney injury; and six were hospitalized because of clinical symptoms and underlying diseases. Among the 28 patients for whom this information was available, the median time from diagnosis to hospitalization was 46 days . Seventeen of the hospitalized patients had severe infections, defined as requiring hospitalization with >3L oxygen and/or corticosteroids or tocilizumab, including 10 patients with HM (10/17; 59%). Furthermore, most of the patients (28/34; 82%) for whom chest CT scan was available had lesions compatible with SARS-CoV-2 infection at diagnosis, mainly mild (affecting 10% to 25% of the lungs).
During follow-up, 41 of 47 patients (87%) with available data achieved clinical cure. The median time to clinical cure was significantly longer in patients with severe disease than in those with non-severe infection (213 [148-258] days vs. 71 [28-130] days; P = 0.04). SOT recipients reached clinical cure faster than patients with HM (28 vs. 162 [80-227] days; P < 0.01), even after adjusting for disease severity (Fig. 1). Patients receiving CD20/19 inhibitors had a significantly longer duration of symptoms than those without (162 days [103. vs. 37 days [7.5-66]; P < 0.01), even after adjusting for disease severity. Having anti-SARS-CoV-2 IgG levels <260 BAU/mL at diagnosis did not significantly affect time to clinical cure or time to CT scan normalization (P = 0.68 and 0.07, respectively). Seventeen percent of patients (n = 9) experienced IMIs, including eight cases of aspergillosis and one coinfection with Aspergillus spp. and Rhizopus spp. (Table 2). IMIs were more likely to occur in patients with severe SARS-CoV-2 (41% vs. 6%, P = 0.03). IMIs tended to be associated with increased six-month and one-year mortality (P = 0.06). In these patients with persistent viral shedding, one-year mortality was significantly associated with severe SARS-CoV-2 (P = 0.04), defined as requiring the use of dexamethasone and/or tocilizumab.
## Characteristics of the treatment received and impact on clinico-radiological outcomes
Seventy-five percent (40/53) of patients received at least one anti-SARS-CoV-2 treatment. The median time between diagnosis and treatment initiation was three days (1-42). Among the 24 patients who received more than one treatment, 41% (10/24) received all lines of treatments over a 21-day period, while the other 59% were treated sequentially, over a longer period (Fig. S1).
First-line therapies are detailed in Table 3. Patients could have received direct antiviral(s) alone, mab alone, plasma alone, combination of antivirals + mab, or combination of antivirals + plasma. The duration of treatment with nirmatrelvir/ritonavir when involved in first-line treatment was always five days, while the median duration of remdesivir was three days (3)(4)(5). Statistical comparison of outcomes between these groups was not performed due to small sample size in each treatment strategy. However, Table S1 provides descriptive data on time to clinical cure, viral clearance, and CT normalization according to the first-line treatment strategy. Overall, patients treated with direct antivirals seemed to have better outcomes. Thus, analysis was performed to compare outcomes between treatment strategy involving direct antivirals or those that did not. There was no significant impact of first-line treatment with direct antivirals on the time to clinical cure or time to radiological normalization, even after adjustment for confounding variables (i.e., disease severity and underlying condition). Among the hospitalized patients, 12 had previously received outpatient treatment with direct-acting antiviral monotherapy (n = 6), monoclonal antibodies (mAbs) (n = 5), or a combination of both (n = 1). Receiving a first-line treatment involving plasma (n = 12) did not have a significant effect on time to clinical cure (median 68 days without plasma vs. 111 [72-188] days with plasma) or time to CT scan normalization (median 126 days without plasma vs. 321 [127-452] days with plasma), compared to other first-line treatment strategies, after adjusting on disease severity, main underlying disease, and time before treatment initiation (respectively P = 0.11; and P = 0.86). Eight of the patients treated with plasmatherapy had IMIs. Data and outcomes for patients in whom plasma was used as first-line treatment or after are detailed in Table S3.
## Virus dynamics during persistent infection
All patients exhibited persistent viral shedding >8 weeks, and 34 (65%) eventually achieved viral clearance during follow-up. The median time before viral clearance in 3). Notably, the time to viral clearance was not significantly impacted by disease severity, baseline anti-SARS-CoV-2 IgG levels <260 BAU/mL (P = 0.07), or underlying condition (111 [93-271] days in SOT recipients vs. 142 [70-266] days in HM patients, P = 0.67) (Table S2). How ever, first-line treatment involving direct antiviral drugs significantly accelerated viral clearance after adjustment for confounding factors, with a median time of 107 days ([51-118], mean 111 days), compared to 183 days ([97-277], mean 224 days) in patients who did not receive remdesivir or nirmatrelvir/ritonavir (P = 0.04) (Fig. 2). In the 12 patients who received a first-line treatment involving plasmatherapy, the median time to viral clearance was 266 [114-427] days (vs. 115 [70-259] days without plasma; P = 0.14 after adjusting on disease severity, main underlying disease, and time before treatment initiation).
A total of 50 samples from 21 patients were selected for further analysis, as out lined in the Methods section and Fig. S2 ("Sample Preparation and Sequencing" and "Mutation and Resistance Analysis"). Phylogenetic analysis of samples before and after anti-SARS-CoV-2 treatment revealed that the majority of patients had closely related virus strains, with later samples showing longer branches, indicating mutation accumu lation over time (Fig. 3). The emergence of mutations in major variant populations is reported in Fig. 4. Pre-treatment, 13 out of 17 patients for whom pre-treatment sequencing data were available (76%) had already exhibited resistance mutations to anti-SARS-CoV-2 treatments. Among them, only two (15%) had received mAbs as prophylaxis, and eight (62%) had been vaccinated. During persistent viral shedding, 11/21 patients developed mutations, including three (patients 5, 12, and 15) without any prior treatment pressure (even no mAb prophylaxis or vaccine exposure). However, the majority of these patients (8/11; 73%) developed resistance mutations after receiving curative treatments (as opposed to mAb prophylaxis or vaccination). Specifically, four patients developed mutations directly conferring resistance to previously administered treatments, all mAb-based. Mutations included Y453F (resistance to casirivimab, after exposure to casirivimab/imdevimab in patient 9), R346T (resistance to cilgavimab and c135, after exposure to tixagevimab/cilgavimab in patient 37), N450D and F486V (resistance to tixagevimab, casirivimab, bamlanivimab, and etesevimab, after exposure to tixagevimab/cilgavimab and plasmatherapy in patient 48), and K444R (resistance to cilgavimab, after exposure to tixagevimab/cilgavimab in patient 50). Among the six patients treated with curative mAbs, only two did not develop resistance mutations. In patients receiving plasmatherapy, 27% (3/11) developed mutations on the spike protein conferring resistance to mAbs, including two previously exposed to mAbs (curative tixagevimab/cilgavimab in patient 48, prophylactic mAbs before infection in patient 36). Interestingly, the third patient (patient 3) developed N450D and L455S mutations after plasmatherapy despite never having received mAbs or vaccination. Among the five patients treated with remdesivir, four did not develop resistance mutations. One patient (patient 36) developed a mutation on the spike protein conferring resistance to mAbs (P499H) after receiving remdesivir plus plasmatherapy, as well as prior prophylactic mAbs. In the case of a vaccinated patient treated with nirmatrelvir/ritonavir, resistance to cilgavimab emerged after monotherapy, despite no prior mAbs exposure (patient 6). Overall, 19 patients received monoclonal antibodies (mAbs) during their infection, including 17 as first-line treatment. Of these, pre-treatment sequencing data were available for eight patients (not known to the clinician at the time of treatment initiation). Notably, four of them (50%) received curative mAbs despite the presence of preexisting mutations known to reduce SARS-CoV-2 sensitivity to these antibod ies (patients 37, 42, 48, and 52). Minor variant analysis revealed treatment-resistant mutations in five patients, including two who developed resistance to mAbs (Fig. S3). However, the small sample size of this analysis limits any conclusions regarding the impact of these mutations on clinical outcomes or viral clearance.
## DISCUSSION
ICPs with persistent SARS-CoV-2 viral shedding (>8 weeks) represent a highly vulnerable population burdened by significant comorbidities (mostly hypertension and diabetes), low anti-SARS-CoV-2 antibodies at diagnosis (nearly two-thirds had IgG levels <260 BAU/mL), and severe complications including hospitalization (91%) and the need for oxygen therapy and corticosteroid (one-third of the cases). This result reinforces previous findings (4). Almost one-fifth of the patients (17%) developed invasive mold infec tions without necessarily having been hospitalized in an intensive care unit, which is comparable to a previous report (20%) (6).
First-line treatments involving direct antiviral therapies (i.e., remdesivir and/or nirmatrelvir/ritonavir) were associated with faster viral clearance in ICPs compared to treatments involving mAbs or plasmatherapy. Furthermore, neither plasmatherapy nor mAbs demonstrated superiority in reducing the time to viral clearance, even in patients without prior resistance mutations to the mAbs used (mainly in the "pre-Omicron" period). Consistently, recent studies have shown that remdesivir is associated with reduced mortality in hospitalized ICPs, particularly those with HMs or SOT (21). Addition ally, studies have suggested that prolonged or combined courses of direct antivirals could be beneficial for ICPs (12,13,22,23). Thus, it seems crucial to intensify the therapeutic approach for profoundly immunosuppressed patients (particularly SOT and HM), going beyond the guidelines which propose short, early antiviral treatment in severe or comorbid patients. Indeed, a prolonged, systematic treatment regimen with direct antiviral(s) in at-risk patients, repeated in cases of persistence, appears to be warranted.
Recent reports also indicate that some ICPs with persistent SARS-CoV-2 infections may exhibit reduced responses to direct antiviral therapies, such as remdesivir and nirmatrelvir/ritonavir, due to the emergence of resistance mutations (2,24). In our series, we did not identify any of these specific resistance mutations in major SARS-CoV-2 variants, although 85% of the patients whose samples were analyzed showed mutations conferring resistance to mAb(s)-either preexisting or acquired during the course of infection. It has been shown that the emergence of resistance muta tions is not necessarily driven by treatment-induced selective pressure. Indeed, SARS-CoV-2 evolution is largely shaped by immune escape, and some resistance-associated mutations have become lineage-defining. However, resistances to mAb(s) particularly emerged in patients pre-exposed to mAbs, but also in one patient treated with plasmatherapy (1/11; 9%) who developed mutations on spike protein conferring resistance to mAbs (25). These types of mutations were observed in major variants of only one out of five patients receiving remdesivir, but this patient had also been pre-exposed to prophylactic mAbs and curative plasmatherapy (26).
This retrospective study has several limitations. First, it focuses on a highly specific population (patients with persistent viral shedding >8 weeks) and includes a heteroge neous mix of underlying conditions (e.g., HM and SOT) and treatment regimens. These factors may limit the generalizability of the findings, and the low number of cases limits statistical power. However, we have established that remdesivir and nirmatrelvir/ritonavir were the most effective treatments for these hospitalized immunocompromised patients with persistent viral shedding, both in terms of viral clearance efficacy and genetic barrier. This is especially pertinent in the current context, where the Omicron variant has developed resistance to all available therapeutic monoclonal antibodies. Furthermore, even if new monoclonal antibodies theoretically effective against Omicron are devel oped, their superiority over direct antivirals will need to be proven (27). It will also be crucial to closely monitor mutations in immunocompromised patients before and during such treatments or plasmatherapy. Furthermore, these findings should be taken into account when considering therapeutic recommendations for any future viral epidemic or pandemic: monoclonal antibody monotherapy appears to be a riskier option compared to combined or repeated direct antiviral therapies.
## Conclusion
The findings of this study emphasize that the use of direct antiviral therapies (remdesi vir and nirmatrelvir/ritonavir) is the most effective treatment in immunocompromised patients (ICPs), reducing the time to SARS-CoV-2 viral clearance, which has been shown to be associated with invasive mold infections. Moreover, their genetic barriers are adequate.
## References
1. (2025) "COVID-19 dashboard"
2. Nooruzzaman, Johnson, Finkelsztein et al. (2024) "Emergence of transmissible SARS-CoV-2 variants with decreased sensitivity to antivirals in immunocompromised patients with persistent infections" *Nat Commun*
3. Kaku, Uriu, Okumura et al. "Genotype to Phenotype Japan (G2P-Japan) Consortium. 2024. Virological characteristics of the SARS-CoV-2 KP.3.1.1 variant" *Lancet Infect Dis*
4. Evans, Dube, Lu et al. (2023) "Impact of COVID-19 on immuno compromised populations during the Omicron era: insights from the observational population-based INFORM study" *Lancet Reg Health Eur*
5. Martinson, Lapham (2024) "Prevalence of immunosuppression among US adults" *JAMA*
6. Melenotte, Chavarot, Honneur et al. (2024) "Increased risk of invasive aspergillosis in immunocompromised patients with persistent SARS-CoV-2 viral shedding >8 weeks, retrospective case-control study" *Open Forum Infect Dis*
7. Goubet, Dubuisson, Geraud et al. (2021) "Prolonged SARS-CoV-2 RNA virus shedding and lymphopenia are hallmarks of COVID-19 in cancer patients with poor prognosis" *Cell Death Differ*
8. Lee, Shah, Hoyos et al. (2022) "Prolonged SARS-CoV-2 infection in patients with lymphoid malignancies" *Cancer Discov*
9. Nagai, Saito, Adachi et al. (2022) "Casirivimab/imdevimab for active COVID-19 pneumonia which persisted for nine months in a patient with follicular lymphoma during anti-CD20 therapy" *Jpn J Infect Dis*
10. Raglow, Surie, Chappell et al. (2023) "SARS-CoV-2 shedding and evolution in immunocompromised hosts during the Omicron period: a multicenter prospective analysis" *medRxiv*
11. Calderón-Parra, Múñez-Rubio, Fernández-Cruz et al. (2022) "Incidence, clinical presenta tion, relapses and outcome of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection in patients treated with anti-CD20 monoclonal antibodies" *Clin Infect Dis*
12. Brosh-Nissimov, Ma'aravi, Leshin-Carmel et al. (2024) "Combination treatment of persistent COVID-19 in immunocompromised patients with remdesivir, nirmaltrevir/ritonavir and tixegavimab/cilgavimab" *J Microbiol Immunol Infect*
13. Meijer, Paran, Belkin et al. (2024) "Persistent COVID-19 in immunocom promised patients-Israeli society of infectious diseases consensus statement on diagnosis and management" *Clin Microbiol Infect*
14. Choi, Choudhary, Regan et al. (2020) "Persistence and evolution of SARS-CoV-2 in an immunocompromised host" *N Engl J Med*
15. 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*
16. Markov, Ghafari, Beer et al. (2023) "The evolution of SARS-CoV-2" *Nat Rev Microbiol*
17. Gandhi, Klein, Robertson et al. (2022) "De novo emergence of a remdesivir resistance mutation during treatment of persistent SARS-CoV-2 infection in an immunocompromised patient: a case report" *Nat Commun*
18. Maier, Kuan, Saborio et al. (2022) "Clinical spectrum of severe acute respiratory syndrome coronavirus 2 infection and protection from symptomatic reinfection" *Clin Infect Dis*
19. Verweij, Brüggemann, Azoulay et al. (2021) "Taskforce report on the diagnosis and clinical management of COVID-19 associated pulmonary aspergillosis" *Intensive Care Med*
20. Wang, Wang, Zhang et al. "2021. Intra-host variation and evolutionary dynamics of SARS-CoV-2 populations in COVID-19 patients" *Genome Med*
21. Mozaffari, Chandak, Gottlieb et al. (2024) "Remdesivir-associated survival outcomes among immunocompromised patients hospitalized for COVID-19: real-world evidence from the Omicron-dominant era" *Clin Infect Dis*
22. Calderón-Parra, Villanueva, Roca et al. (2024) "Efficacy and safety of antiviral plus anti-spike monoclonal antibody combination therapy vs. monotherapy for high-risk immuno compromised patients with mild-to-moderate SARS-CoV2 infection during the Omicron era: a prospective cohort study" *Int J Antimicrob Agents*
23. Mikulska, Sepulcri, Dentone et al. (2023) "Triple combination therapy with 2 antivirals and monoclonal antibodies for persistent or relapsed severe acute respiratory syndrome coronavirus 2 infection in immunocompromised patients" *Clin Infect Dis*
24. Tamura, Choudhary, Deo et al. (2024) "Emerging SARS-CoV-2 resistance after antiviral treatment" *JAMA Netw Open*
25. Hogan, Duerr, Dimartino et al. (2019) "Remdesivir resistance in transplant recipients with persistent coronavirus disease" *Clin Infect Dis*
26. Lythgoe, Hall, Ferretti et al. (2021) "SARS-CoV-2 within-host diversity and transmission" *Science*
27. Walker, Underwood, Björnsson et al. (2024) "Broadly potent spike-specific human monoclonal antibodies inhibit SARS-CoV-2 Omicron sub-lineages" *Commun Biol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12451297&blobtype=pdf | # A multi-centre study on epidemiology and clinical characteristics of human metapneumovirus infection in Sri Lanka from 2021 to 2024
Shiyamalee Arunasalam, Ishani De Silva, Udeshika Sathgurupathi, Veranja Liyanapathirana, Wasana Kudagammana, Faseeha Noordeen, Thulani Pattiyakumbura, Saranga Sumathipala, Rohitha Muthugala
## Abstract
The authors confirm that all supporting data have been provided within the article. Genomic sequencing data are available in the National Center for Biotechnology Information (NCBI) database (GeneBank accession no. PV178132). A supplementary file summarizing demographic and clinical details of the hMPV positive samples is also provided along with this article.
Human metapneumovirus (hMPV) is a respiratory virus that has been known to cause acute respiratory tract infections (ARTI) since its identification in 2001 [1]. hMPV has gained increased attention after the recent surge in hMPV-associated ARTI cases in different countries, including China, United Kingdom (UK), India and Malaysia. China has reported a rise in hMPV cases, particularly among individuals under the age of 14 by second of January 2025 [2]. In the UK, as of 19 January 2025, hMPV was detected in 4.9% of hospital patients tested for respiratory infections, which is slightly higher than the previous year's peak of 4.18% [3]. In India, seven cases have been confirmed across multiple states, including Karnataka, Gujarat and Tamil Nadu, following the hMPV outbreak reported in China in early 2025 [4]. Malaysia also recorded 327 hMPV cases in 2024, which represents a 45% increase compared to 225 cases in 2023 [5]. However, health officials monitoring the respiratory illnesses at global, regional and country levels through collaborative surveillance systems have not made emergency declarations yet [6].
hMPV is a non-enveloped negative-sense RNA virus of the family Pneumoviridae and genus Metapneumovirus. hMPV is likely to have originated from animals infecting only humans [7]. There are two major types of hMPV, identified as hMPV-A and -B. These are further divided into four subtypes, named as A1, A2, B1 and B2, and two additional subtypes of A2 named as A2a and A2b [1]. hMPV reaches its optimal infectivity within 4-6 days of infection [8]. hMPV is transmitted through direct contact with contaminated body secretions like saliva, aerosol and droplets and through contaminated surfaces. Moreover, there have been reports of nosocomial hMPV infections in infants [9].
hMPV typically causes mild respiratory symptoms similar to the common cold or flu, such as cough, fever and nasal congestion. However, it can cause more severe illness in young children, the elderly and immunocompromised individuals [10]. Experts recommend practising good hygiene, such as frequent handwashing, covering coughs and sneezes and avoiding close contact with sick individuals, to reduce its spread [11]. There is no specific antiviral treatment or vaccine against hMPV infection, but most cases resolve on their own with supportive care [12].
Despite its known prevalence worldwide, there is emerging data on hMPV from different areas of Sri Lanka. Since the first report of hMPV infection in Sri Lanka in 2013 from a 9-month-old girl presented to the outpatient department of a teaching hospital, a few studies have focused on the correlation of hMPV with ARTI, including a mini hMPV wave with severe acute respiratory tract infection (SARI) [13][14][15][16][17]. The present study aims to fill this gap by investigating the prevalence and clinical characteristics of hMPV infection from January 2021 to December 2023 in different locations in Sri Lanka, aiming to enhance public health strategies and preparedness for respiratory infection outbreaks.
## METHODS
## Study design and setting
The study was conducted as a prospective descriptive study in a sample of patients with ARTI (age 12 days to ≤85 years) from four different locations in Sri Lanka: National Hospital, Kandy (NHK) (January 2021-October 2022); Teaching Hospital, Anuradhapura (THA) (March 2021-May 2021); National Cancer Institute Sri Lanka (NCISL) (January 2022-December 2024); and Teaching Hospital, Peradeniya (THP) (November 2023-December 2023). The study was approved by the Ethical Review Committee of the Faculty of Medicine, University of Peradeniya (2021/EC/21, 2022/EC/52), Post Graduate Institute of Science, University of Peradeniya (CEC-PGIS-2021-08) and the Medical Research Institute, Sri Lanka (ERC/ 2025/06). A total of 1,582 patients with ARTI symptoms, including fever (more than or equal to 38 °C) with, cough, cold, sore throat or shortness of breath within the first 7 days of the illness, were selected for the study among the samples received for routine laboratory testing. Demographic and clinical data were extracted from the patients' clinical notes.
## Sample processing and RespiFinder 2Smart assay
The respiratory specimens were subjected to nucleic acid extraction using locally validated commercial kits (QIAGEN, Germany, or SpinStar, Malaysia, or Maxwell® RSC Viral Total Nucleic Acid Purification Kit and using the Maxwell® RSC48, USA). The nucleic acid extracts were tested for respiratory pathogens [(influenza-A, influenza-B, influenza virus H1N1 pdm 09, respiratory syncytial virus-A, respiratory syncytial virus-B, human parainfluenza virus-1, human parainfluenza virus-2 (hPIV-2), human parainfluenza virus-3 (hPIV-3), human parainfluenza virus-4, human coronavirus OC43 (hCoV OC43), human coronavirus 229E, human coronavirus NL63/HKU1, rhinovirus/enterovirus (Rh/EnV), human adenovirus (hAdV), hMPV, human bocavirus type-1 (hBoV-1) and four atypical bacteria such as Mycoplasma pneumonia, Chlamydophila pneumoniae, Legionella pneumophila and Bordetella species] by a commercial multiplex real-time PCR assay (RespiFinder2Smart, PathoFinder, catalogue no: PF2600-2S, Netherlands) according to the manufacturers' guidelines.
## Illumina sequencing
A subset of hMPV-positive samples was then selected for genomic sequencing. Sequencing was performed using the advanced Illumina sequencing platform at the Genomics Laboratory, NCISL. A commercial respiratory virus sequencing kit (Illumina Respiratory ID/AMR Panel, Illumina, USA) employed. The sequencing was performed according to the manufacturer's instructions using the Illumina NextSeq 1000 System, known for its precision in high-throughput sequencing workflows. The forward and reverse sequences were assembled, and consensus contig assembly was performed using an Illumina BaseSpace sequence assembler V.2.0.0.
## RESULTS
A total of 1,582 nasopharyngeal swab samples were tested. Of these, hMPV was detected in 26/1,582 (1.64%) patients. A summary of hMPV detection across different locations in Sri Lanka from 2021 to 2023 is shown in Table 1.
hMPV infections were predominantly detected in children aged <5 years with a child-to-adult ratio of 10:3 among hMPV-positive patients. The infection was more prevalent in males (69.23%) than in females. Of the 26 hMPV-positive patients, 5 were co-infected with other respiratory viruses, including hBoV-1, hCoV OC43, hPIV-2 and Rh/EnV. All co-infected patients were under 5 years of age and two required admissions to the intensive care unit (ICU). The distribution of hMPV infection across different age groups is presented in Fig. 1.
The most common symptoms observed in hMPV-infected patients were fever, cough, cold and sore throat. Lower respiratory tract symptoms, including pneumonia and bronchiolitis, were noted in 12 (46.15%) patients, while 3 (11.53%) patients required ICU admission. All ICU-admitted patients were under 4 months of age and were immunocompromised. Clinical characterization of the hMPV-positive patients is summarized in Table 2.
Of the sequenced samples, only one was successfully sequenced and identified as hMPV type B on L gene analysis (GeneBank accession no. PV178132). in 2019 identified hMPV as the most predominant virus responsible for SARI with a prevalence of 86% during a mini outbreak that occurred in 2019 [16]. It has also to be noted that the prevalence of hMPV in THP is higher than in the other locations in the current study, too. Moreover, the sample collection period of the THP was in the latter part of the Corona Virus Disease 2019 (COVID-19) pandemic. During the COVID-19 pandemic, non-pharmaceutical interventions (NPIs) like wearing face masks, regular hand washing, closure of schools and maintaining social distance would have contributed to the low prevalence in the other three locations, as the sampling in those was done during early or peak time of the pandemic prior to mid-2022. These NPIs were relaxed in mid-2022 in Sri Lanka following the COVID-19 vaccination. The relaxation of NPIs might have also influenced the increase in positivity rate in THP, apart from the higher positivity reported at the same location prior to the pandemic [18].
Low success rate for sequencing could be due to sub-optimal sample storage and transport from the peripheral laboratory to the sequencing laboratory during the COVID-19 pandemic, where laboratories were overwhelmed with a large number of samples.
Prevalence of hMPV infections was higher in males compared to females, and this finding is in agreement with other studies as well [10,19,20]. Most of the hMPV-positive patients were children compared to adults, as has also been reported by Yi et al. in China [10]. However, the wider age distribution of hMPV is considerably varied from RSV infection which usually affects children <2 years of age and older adults >65 years of age [21]. These clinical symptoms ranged from mild upper respiratory tract infection to severe lower respiratory tract infections (LRTI), as discussed [10]. In our study, the most common symptoms observed were fever, cough and sore throat. Lower respiratory tract symptoms included pneumonia and bronchiolitis, and these were noted in 46.15% (n=12) of patients. Additionally, three patients (11.53%) required ICU care. These findings are in agreement with the findings of a case series reported by Jayaweera et al. in 2018 in Sri Lanka. In that study, hMPV infection in children has shown a range of respiratory symptoms with varying severity ranging from common cold to life-threatening LRTIs [14].
hMPV is often co-infected with other respiratory viruses, such as hAdV, RSV, RhV and hPIV [22][23][24]. In this study, the co-infection rate was 0.31%, with hPIV, which had the highest amount of co-infection with hMPV, as also noted by Fathima et al. in Alberta, Canada, in 2012 [25]. According to a case study conducted in Sri Lankan children with ARTI, the co-infection rate of hMPV-RSV infections has been high in Sri Lanka [14]. However, there were no hMPV-RSV co-infections noted in our study. Moreover, it has also to be noted that, in our study, 40% of co-infected patients required ICU care, and this is in agreement with previous studies, which concluded that the co-infection of hMPV with other respiratory viruses can aggravate clinical severity [26,27].
## CONCLUSION
In summary, hMPV infection was prevalent in 1.64% of the patients suspected of ARTI and the majority of the hMPV-infected patients were children less than 5 years of age. hMPV-infected patients showed a range of respiratory symptoms with varying severity, ranging from common cold to life-threatening LRTIs. hMPV co-infections have been noted with hBoV-1, hCoV OC43, hPIV-3 and Rh/EnV. Detailed and continuous screening for hMPV among adults and children will help to mitigate the global burden of hMPV, improve outcomes for high-risk populations and strengthen preparedness against respiratory viral infections leading to outbreaks, epidemics and pandemics.
## Funding information
This study received funding from the Health Sector Development Project (HSEP) and University of Peradeniya, Sri Lanka.
## References
1. Van Den Hoogen, De Jong, Groen et al. (2001) "A newly discovered human pneumovirus isolated from young children with respiratory tract disease" *Nat Med*
2. China (2025) "National sentinel surveillance of acute respiratory infectious diseases"
3. Ukhsa (2025) "UKHSA data dash board"
4. (2025) "India's HMPV count reaches 8, latest case reported in mumbai" *The Economic Times*
5. Young (2025) "HMPV cases are rising across Asia, but experts say not to panic"
6. Who "Trends of acute respiratory infection, including human metapneumovirus"
7. Soto, Gálvez, Benavente et al. (2018) "Human metapneumovirus: mechanisms and molecular targets used by the virus to avoid the immune system" *Front Immunol*
8. Peiris, Tang, Chan et al. (2003) "Children with respiratory disease associated with metapneumovirus in Hong Kong" *Emerg Infect Dis*
9. Panda, Mohakud, Pena et al. (2014) "Human metapneumovirus: review of an important respiratory pathogen" *Int J Infect Dis*
10. Yi, Zou, Peng et al. (2019) "Epidemiology, evolution and transmission of human metapneumovirus in Guangzhou China" *Sci Rep*
11. Al-Tawfiq, Memish (2025) "The surge of human metapneumovirus (hMPV) cases in China and global implications" *New Microbes New Infect*
12. Divarathna, Rafeek, Noordeen (2020) "A review on epidemiology and impact of human metapneumovirus infections in children using TIAB search strategy on PubMed and PubMed Central articles" *Rev Med Virol*
13. Noordeen, Jayaweera, Rayes (2016) "Human metapneumovirus associated pneumonia and severe bronchiolitis in a 9-month-old infant admitted to a Sri Lankan hospital" *Sri Lankan J Infec Dis*
14. Jayaweera, Noordeen, Kothalaweala et al. (2018) "A case series on common cold to severe bronchiolitis and pneumonia in children following human metapneumovirus infection in Sri Lanka" *BMC Res Notes*
15. Jayaweera, Morel, Abeykoon et al. (2021) "Viral burden and diversity in acute respiratory tract infections in hospitalized children in wet and dry zones of Sri Lanka" *PLoS One*
16. Noordeen, Pitchai, Kudagammana et al. (2019) "A mini outbreak of human metapneumovirus infection with severe acute respiratory symptoms in a selected group of children presented to a teaching hospital in Sri Lanka" *Virusdisease*
17. Shapiro, Bodinayake, Nagahawatte et al. (2017) "Burden and seasonality of viral acute respiratory tract infections among outpatients in southern Sri Lanka" *Am J Trop Med Hyg*
18. Heiskanen, Galipeau, Little et al. (2023) "Seasonal respiratory virus circulation was diminished during the COVID-19 pandemic" *Influenza Other Respir Viruses*
19. Cong, Wang, Wei et al. (2022) "Human metapneumovirus in hospitalized children with acute respiratory tract infections in Beijing" *Infect Genet Evol*
20. Devanathan, Philomenadin, Panachikuth et al. (2025) "Emerging lineages A2.2.1 and A2.2.2 of human metapneumovirus (hMPV) in pediatric respiratory infections: insights from India" *IJID Reg*
21. Nguyen-Van-Tam, Leary, Martin et al. (2022) "Burden of respiratory syncytial virus infection in older and high-risk adults: a systematic review and meta-analysis of the evidence from developed countries" *Eur Respir Rev*
22. Pilger, Cantarelli, Amantea et al. (2011) "Detection of human bocavirus and human metapneumovirus by real-time PCR from patients with respiratory symptoms in Southern Brazil" *Mem Inst Oswaldo Cruz*
23. Wang, Wei, Ma et al. (2021) "Epidemiology and genotypic diversity of human metapneumovirus in paediatric patients with acute respiratory infection in Beijing, China" *Virol J*
24. Greensill, Mcnamara, Dove et al. (2003) "Human metapneumovirus in severe respiratory syncytial virus bronchiolitis" *Emerg Infect Dis*
25. Fathima, Lee, May-Hadford et al. (2009) "Use of an innovative web-based laboratory surveillance platform to analyze mixed infections between Human Metapneumovirus (hMPV) and other respiratory viruses circulating in Alberta (AB)" *Viruses*
26. Xiao, Xie, Zhang et al. (2010) "Prevalence and clinical and molecular characterization of human metapneumovirus in children with acute respiratory infection in China" *Pediatr Infect Dis J*
27. Zhang, Liu, Liu et al. (2013) "Epidemiological and clinical features of human metapneumovirus in hospitalised paediatric patients with acute respiratory illness: a cross-sectional study in Southern China" *BMJ Open* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772307&blobtype=pdf | # Impact of prolonged infection on SARS-CoV-2 evolution
Fang Yan, Qiushi Jin, Yuanguo Li, Xuefeng Wang, Chunling Dong, Xianzhu Xia, Yuwei Gao, Jie Zhang, Zhijun Hou, Alexander Bello
## Abstract
Immunocompromised patients with prolonged severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections may serve as reservoirs for viral evolution, with suboptimal immune responses facilitating the accumulation of adaptive mutations. This study aimed to characterize the drivers of SARS-CoV-2 adaptive evolution in such hosts through genomic surveillance. We retrospectively analyzed 24 patients with long-term positive nasopharyngeal reverse transcription-polymerase chain reaction results (symptom onset duration: 7-14 days on average, 1 HIV-positive patient with >20 days of infection). Most infections (April-May 2022) were caused by Omicron variants (predominantly BA.2). Phylogenetic analysis revealed accelerated viral evolution in patients with diverse underlying diseases (e.g., HIV and esophageal cancer). A total of 78 intrahost single-nucleotide variants were identified, with ORF1ab (53.8%) and the Spike protein coding region (20.5%) being hotspots. Notably, the HIV-positive patient's virus developed unique mutations: NSP3-T779I, NSP15-A94T, and Spike double mutations N440K and I794T. Functional assays showed that the N440K/I794T double mutation significantly enhanced infectivity in Hela-hACE2 cells (P < 0.05) but reduced immune evasion (50% neutralizing titer increased ~2-fold vs BA.2, P < 0.001). The I794T mutation was later detected in the JN.1.16 strain, suggesting potential evolutionary persistence. Prolonged SARS-CoV-2 replication in immunocompromised hosts, particularly HIV-posi tive individuals, drives adaptive mutations with altered infectivity and immune evasion. These findings emphasize the need for monitoring such hosts to prevent the spread of potentially transmissible variants. IMPORTANCE This study reveals the characteristics of adaptive mutations and their biological functions of SARS-CoV-2 during long-term infection in immunocompromised patients, emphasizes the importance of close monitoring of the long-term replication of SARS-CoV-2, aiming to timely detect and prevent the spread of newly emerging neutralization-resistant variants in susceptible populations, and provides a key scientific basis for formulating effective public health prevention and control strategies.KEYWORDS accelerated viral evolution in immunocompromised hosts, hotspots of intrahost mutations, unique mutations in the HIV-positive patient
T he severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has been characterized by the regular emergence of new variants. The origin of these new variants is unclear, with some speculating emergence from zoonotic spillover into other vertebrates and spillback into humans (1)(2)(3)(4)(5). An alternative, widely deba ted source of new variants stems from persistent infections in immunocompromised patients. In such hosts, suboptimal immune responses are supposed to facilitate the gradual accumulation of genomic adaptations (4-7), which probably occurred prior to the major SARS-CoV-2 wave in December 2022-January 2023, when China's population exhibited a distinct immunological profile characterized by lower population immunity due to minimal prior infection exposure. This contrasted sharply with nations that had experienced repeated infection waves. China's unique immunological landscape, coupled with prolonged circulation of similar viral strains both domestically and globally, provides a rare opportunity to systematically examine viral evolutionary patterns across divergent immune milieus (8). SARS-CoV-2 infections typically resolve clinically within days, with RNA shedding persisting from several days to weeks (9). However, accumulating case reports describe chronic infections lasting weeks to months, challenging this conventional understanding of viral clearance dynamics (10,11). Chronic SARS-CoV-2 infections are characterized by prolonged detection of replicative virus. To date, all documented cases have occurred exclusively in severely immunocompromised individuals, including those with primary immunodeficiencies, post-transplant immuno suppressive therapy, AIDS, hematological malignancies, and/or associated treatments.
The persistence of replicative virus in these patients likely results from impaired viral clearance mechanisms, particularly involving adaptive immune dysfunction. This contrasts sharply with immunocompetent individuals, who typically clear the infection efficiently (12).
Longitudinal sequencing analyses of chronic infection cases have revealed a greater number of mutations and distinctive mutational patterns compared to those observed in transmission chains among acutely infected individuals (6,13,14). Multiple studies have characterized viral evolutionary dynamics in immunocompromised hosts, demonstrat ing a strong association between immunodeficiency and accelerated within-host viral mutation accumulation (6,7,(13)(14)(15)(16)(17)(18)(19)(20)(21). A recent study revealed that 13.9% of SARS-CoV-2infected B-cell lymphoma patients experienced prolonged infections persisting ≥30 days (22). Patients with B-cell lymphoma often exhibit impaired SARS-CoV-2-neutraliz ing antibody production, which predisposes them to both prolonged infection (23)(24)(25) and diminished vaccine responsiveness (26,27). Notably, one B-cell lymphoma patient maintained persistent SARS-CoV-2 infection for 156 days, during which the virus acquired 16 mutations, including 4 substitutions conferring neutralizing antibody escape (28). In a separate case series, multiple immunocompromised heart transplant recipients independently acquired the immune-evading E484K Spike substitution within a remarkably short 14-day period (18). Current evidence suggests that the predominant selective pressure in these cases likely stems from adaptive advantages conferred by enhanced cell-to-cell transmission within the host (29). Emerging evidence suggests that immunocompromised hosts may serve as reservoirs for accelerated SARS-CoV-2 variant evolution, though the precise dynamics and frequency of this phenomenon remain to be systematically quantified.
In this study, we conducted genomic surveillance of SARS-CoV-2 in 24 immunocom promised patients with suspected chronic infection to track the emergence of intrahost single-nucleotide variants (s). Through longitudinal sequence analysis, we characterized viral evolutionary dynamics within these hosts and compared mutation accumulation rates with those observed in immunocompetent populations.
## RESULTS
## Adaptive evolutionary characteristics in long-term SARS-CoV-2 infections
In immunocompromised patients with long-term chronic severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, the virus often accumulates a large number of adaptive mutations due to continuous replication. These mutations may significantly alter its pathogenic characteristics, including infectivity, pathogenicity, and immune escape ability. This study focuses on this important scientific issue and conducts a retrospective analysis of 24 early patients whose nasopharyngeal reverse transcription quantitative polymerase chain reaction (RT-qPCR) tests remained positive for a long time. The 24 patients are aged between 16 and 70 years old. In this study, we define chronic infection as the persistence of positive SARS-CoV-2 RNA in nasopharyngeal samples from patients more than 7 days after the onset of symptoms. These patients have multiple underlying diseases, including HIV, esophageal cancer, heart disease, cerebral infarction, hypertension, and diabetes mellitus (Table 1). The diverse background of underlying diseases provides a rich sample for exploring the impact of different immune statuses on viral infection.
We performed RT-qPCR and deep sequencing on 14-day samples from each patient, with all selected samples having a cycle threshold value of <32, among which 12 samples were identified as Omicron strains (Table 1). During the period of April-May 2022 when the study was conducted, the dominant circulating SARS-CoV-2 strains in China were BA.1/BA.2 (Fig. 2A, data sourced from Global Initiative on Sharing All Influenza Data [GISAID]). However, phylogenetic tree analysis of 14 patients in this study cohort revealed that under different immune backgrounds, the virus accumulated various mutations during the prolonged infection period, which accelerated the viral evolutionary rate (Fig. 1). Notably, we also identified two WH-01-Hu-like strains, and this finding indicates that due to differences in patients' immune backgrounds, the virus exhibits a greater number of mutations during infection.
To further explore the virus's activity and replication capacity, we attempted to isolate the virus from these clinical samples. The results showed that only the virus from HIV patients' samples could be successfully isolated and amplified in Vero E6-TMPRSS2 cells. The viral sgRNA copy number detected by RT-qPCR and the virus titer determined by PFU were 3 × 10⁵ (Fig. 2B andC). Based on this, we hypothesize that the number of viable virus particles in immunosuppressed patients is higher, making them more likely to persist for a long time. This is consistent with reports in multiple previous studies that SARS-CoV-2 can chronically infect HIV patients for more than 100 days.
## Intrahost genetic diversity during chronic SARS-CoV-2 infection
Through systematic genetic screening of full-genome mutations, a total of 78 mutant/ deletion iSNVs were identified. The allele frequencies of these iSNVs range from 5% to 98% (with no variant reaching 100% fixation) and can be further categorized based on within-host dynamic characteristics: 29.5% (23/78) are low-frequency expanding variants (5%-30%); 61.5% (48/78) are moderate-frequency stable variants (30%-90%); and 9.0% (7/78) are high-frequency unfixed variants (90%-98%). This distribution indicates that all iSNVs are in the dynamic evolutionary stage within the host (not fully fixed), consistent with the characteristics of adaptive mutations accumulating during prolonged infection (Table S1). Further genomic localization analysis showed that these iSNVs are widely distributed across nine open reading frames (ORFs) of SARS-CoV-2 (Fig. 3). To eliminate the influence of ORF length on variant count, we calculated the variant density (number of iSNVs per 1,000 nt) for each ORF and verified differences via Poisson regression (Table S1). Results showed that (i) in terms of absolute iSNV count, ORF1ab had the highest number of variants (42 sites, 53.8% of total) due to its longest reference length (21,552 nt); (ii) in terms of variant density, ORF6 (13.16 sites/1,000 nt; RR = 6.75; P = 0.002), N (6.35 sites/1,000 nt; RR = 3.26; P = 0.002), and Spike (4.19 sites/1,000 nt; RR = 2.15; P = 0.007) had significantly higher densities than ORF1ab (1.95 sites/1,000 nt), while ORF3a, ORF7a, ORF8, and E showed no significant differences (all P > 0.05). This suggests that short ORFs like ORF6, N, and Spike may be under stronger selection pressure during prolonged infection, leading to higher variant accumulation per unit length. Among the above variants, mutations at specific sites show high occurrence frequency and abundance, suggesting that they may have potential selective advan tages in the process of virus adaptation to the host or replication. Among them, the G28423T mutation site in the N protein coding region has the highest occurrence frequency, reaching 10 times, and its allele frequency is maintained above 91% (>91%) but not reaching 100%, indicating that this mutation is nearly fixed in the tested samples and may have a strong adaptive advantage. It is followed by the C29870A mutation in the 3′ untranslated region, which occurs nine times with a wide frequency range (>30% and <95%), reflecting that this site may have differential evolutionary dynamics in different samples. The T26843C mutation in the M protein coding region occurs four times, with a frequency ranging from 70% to 100%, suggesting that it has become a dominant variant in some samples.
iSNVs in the Spike protein coding region and the non-structural protein 14 (NSP14) coding region each occurred twice. Specifically, the C18692T mutation in the NSP14 coding region had frequencies of 97% and 100% in the two detections, indicating that this mutation is nearly completely fixed in the corresponding samples. Mutations in the S protein coding region exhibit diverse frequency characteristics: the frequencies of the A22301C mutation are 49.76% and 38.03%; those of C22480T are 89.63% and 98.82%; those of A23251G are 38.32% and 100%; and those of A24531C are 23.52% and 100%. Notably, mutation sites such as C18692T (NSP14), A22301C (Spike), C22480T (Spike), A23251G (Spike), A24531C (Spike), T26843C (membrane protein), and G28423T (nucleocapsid protein) have become key marker sites for defining subsequent circulating strains due to their multiple occurrences and high frequencies in some samples.
Through comparative analysis with clinical isolates from the same period, it was found that the number of nucleotide and amino acid variations in the S protein of samples from HIV-coinfected patients was significantly higher than that in ordinary clinical isolates. This result is consistent with reports from multiple previous studies, further confirming that the evolutionary rate of the virus is significantly accelerated in the context of immunocompromised hosts (Fig. S1 and Table 1). In addition, deep sequencing results of sample 2 showed that immunosuppressed patients may not only experience persistent long-term infections but also be accompanied by the phenomenon of "hypermutation" in the viral genome, providing important empirical evidence for understanding the adaptive evolutionary mechanism of the virus under immunodeficiency.
## Genomic characterization of persistent SARS-CoV-2 infection in immunocom promised patients
To explore the evolutionary characteristics of the SARS-CoV-2 BA.2 subtype during a specific period, this study systematically retrieved the viral evolutionary patterns of all BA.2-infected patients in China from 1 February 2022 to 30 June 2022 and focused on comparative analysis of the viral evolutionary pattern of patient no. 2 with that of patients in Jilin Province during the same period (Fig. 4A). Through sequence alignment and variation analysis, we identified six iSNVs in the viral genome of patient no. 2 (Fig. 4B). Further functional annotation showed that four of these variations were non-synon ymous substitutions, located at the T779I site of NSP3, the A94T site of NSP15, and the N440K and I794T sites of the Spike protein, with their substitution frequencies in the viral population being 13.9%, 100.0%, 5.37%, and 18.35% in sequence (Fig. 4B). Notably, these variation patterns are not entirely consistent with the evolutionary characteristics presented by SARS-CoV-2 in the early stage of transmission, suggesting that the virus may exhibit a unique evolutionary path under specific host environments or infection conditions. To further clarify the specificity and universality of the above four non-synonymous substitutions, we extensively compared them with viral sequences from other domestic BA.2-infected patients during the same period. The results showed that substitutions at the NSP3-T779I and Spike protein (I794T) sites were not detected in domestic patients during the same period, indicating that they may be unique adaptive variations of patient no. 2; the substitution at NSP15-A94T appeared in some patients during the same period but with a low frequency, accounting for only 18.67%; while the Spike protein (N440K), as a defining site of the BA.2 subtype, underwent a reversion mutation in patient no. 2 (Fig. 4B). This phenomenon suggests that the variation at this site may be dynamically regulated by host immune pressure or viral replication requirements.
To verify whether the aforementioned variations stably exist in the viral population, we performed plaque purification on the viral strain isolated from patient no. 2 and conducted genome sequencing on the purified monoclonal virus. The results showed that none of the aforementioned mutations were detected in the purified viral clones, which suggests that these mutations generated by the virus may be in an unstable state in patients with intact immune function. They cannot survive stably or proliferate effectively in the in vitro culture environment of the Vero E6-TMPRSS2 cell line. It is speculated that such mutations may be closely related to the host's immune selection pressure and are difficult to maintain in cell culture systems lacking corresponding selection pressure. Notably, it is of research value that we found the S protein (I794T) mutation site was detected in the subsequently isolated clinical strain JN.1.16. This finding implies that this mutation may have certain evolutionary potential, being retained and stably inherited in specific transmission chains or evolutionary branches, thus providing important clues for exploring the evolutionary trajectory of the virus.
Statistical analysis of the coding region distribution of amino acid variations in the viral genome of patient no. 2 showed that the coding regions with the largest number of variations were, in order, the Spike protein, nucleocapsid protein, and membrane envelope protein (Fig. 4C). Among them, the Spike protein had as many as 31 mutations, the nucleocapsid protein had seven mutations, and the membrane protein had three mutations. As a key structure for the virus to bind to host cell receptors, the highfrequency mutations in the Spike protein may be related to adaptive evolutionary strategies such as the virus evading host immune recognition and enhancing infection efficiency.
Synthesizing the above research results, we hypothesize that during chronic infection, the evolution of the virus may face a trade-off between multiple selection pressures. This trade-off effect may involve multiple aspects such as viral replication efficiency, immune escape ability, and host adaptability, thereby affecting the ability of viral variants generated during chronic infection to further spread and persistently replicate in the population. This finding provides important experimental basis and theoretical reference for an in-depth understanding of the evolutionary mechanism of SARS-CoV-2 in chronically infected hosts and evaluating the transmission risk of potential variants.
## SARS-CoV-2 Spike mutations confer enhanced infectivity and antibody neutralization escape
To further investigate the impact of specific mutations in the Spike protein of the SARS-CoV-2 BA.2 subtype on viral infectivity and immune escape ability, this study employed a vesicular stomatitis virus (VSV) pseudotyping system for functional verification. We constructed six types of pseudotyped viruses, which express the wild-type D614G mutant Spike protein (WT-Spike-D614G), BA.2 subtype Spike protein (BA.2-Spike), BA.2 subtype N440 mutant Spike protein, BA.2 subtype T794 mutant Spike protein, and BA.2 subtype N440/T794 double-mutant Spike protein BA.2-Spike-N440/T794, respectively. By detecting the neutralizing activity of serum against these pseudotyped viruses, we systematically evaluated the functional effects of the mutation sites. This pseudotyped virus system has been proven to be widely used in studies on viral entry mechanisms and neutralizing antibody evaluation, and its detection results are in good agreement with those of experiments using fully replication-competent viruses. In the detection of viral infectivity, we used reverse transcriptase activity as a quantitative indicator to assess the single-round infection ability of pseudotyped viruses in Hela-hACE2 cells. The results showed that compared with WT-Spike-D614G, the single-round infectivity of BA.2-Spike in Hela-hACE2 cells was significantly reduced, while the single-round infectivity of the BA.2-Spike-N440/T794 double mutant was significantly enhanced, with a statistically significant difference compared with BA.2-Spike (P < 0.05), suggesting that the synergis tic effect of the double mutation can significantly improve the cell invasion ability of the virus. Further comparison revealed that the single-round infectivity of the BA.2-Spike-N440 single mutant in Hela-hACE2 cells was significantly lower than that of the double mutant (Fig. 5A), indicating that the mutation at the T794 site may play a dominant role in enhancing infectivity.
To evaluate the impact of the aforementioned mutations on the virus's immune escape ability, we collected serum from healthy individuals who had been previously vaccinated with inactivated vaccines or adenovirus vector vaccines (targeting the wildtype virus). Using a method where pseudotyped vesicular stomatitis viruses were incubated with serially diluted serum before infecting Hela-hACE2 cells, we calculated the 50% neutralizing titer (NT 50 ) for each Spike variant. The results showed that relative to the ancestral D614G, the neutralizing titers of all BA.2-related mutant strains were significantly reduced (P < 0.001), suggesting that the BA.2 subtype and its mutants all have a certain basis for immune escape. Compared with BA.2-Spike, the neutralizing titer of the BA.2-Spike-N440 single mutant was significantly increased by approximately threefold (P < 0.001), and that of the BA.2-Spike-N440/T794 double mutant was significantly increased by approximately twofold (P < 0.001). Moreover, there were statistically significant differences in neutralizing titers between the single mutants and the double mutant (P < 0.001) (Fig. 5B andC). These results indicate that the mutation at the N440 site can effectively attenuate the virus's immune escape ability, while the
## DISCUSSION
The findings of this study systematically illuminate the evolutionary trajectory of SARS-CoV-2 in immunocompromised hosts with prolonged infections, shedding critical light on the mechanisms driving viral adaptive mutation and its potential public health implications. By focusing on 24 patients-including an HIV-positive individual with an infection duration exceeding 20 days-we identified key patterns of iSNV accumulation, functional alterations in viral proteins, and evolutionary links to subsequent circulating strains, all of which align with and extend prior research on viral persistence in immuno deficient populations.
Based on the comprehensive analysis of the above experimental results, it can be seen that the BA.2-Spike-N440/T794 double mutant can significantly enhance the ability to infect host cells. During the infection of SARS-CoV-2 in patients with impaired immune function, host-adaptive mutations in the Spike protein may reduce the virus's immune escape ability through synergistic effects. Among them, the BA.2-Spike-T794 mutation plays a key role in enhancing cell infection efficiency, while the BA.2-Spike-N440 mutation is the core site for reducing the virus's immune escape ability. The balance between the two may affect the replication and transmission potential of the variant strain in the host.
In conclusion, this study reinforces the critical role of immunocompromised hosts in SARS-CoV-2 evolution, demonstrating that prolonged infection drives the accumu lation of adaptive mutations with context-dependent effects on infectivity, immune evasion, and cell fusion. The link between the HIV patient's I794T mutation and the later JN.1.16 strain highlights the need for targeted surveillance of immunocompromised populations, including those with HIV, hematological malignancies, or post-transplant immunosuppression to detect potentially transmissible variants early. Additionally, our findings emphasize that viral evolution in these hosts is shaped by complex trade-offs between fitness traits, underscoring the importance of integrating genomic surveillance with functional assays to assess the public health risk of emerging mutations. Such approaches will be essential for refining public health strategies, particularly as SARS-CoV-2 continues to circulate and adapt in diverse host populations.
## Mutation sites of spike-pseudotyped virus enhance cell-cell membrane fusion capacity
We conducted cell-cell membrane fusion experiments to further explore the functional differences among different mutants and determined the luciferase activity at 0, 6, 12, and 24 h after the superposition of the two types of cells. At 0 h, fusion had not yet started, and there was no difference among the experimental groups. At 6 h, compared with WT-Spike-D614G, all mutants showed significant differences. At 12 h, compared with WT-Spike-D614G, the BA.2 (P < 0.001) and BA.2-Spike-N440 (P < 0.05) mutants had significant differences, and the difference of the BA.2 mutant was more significant. At 12 h, compared with BA.2, the BA.2-Spike-N440, BA.2-Spike-T794, and BA.2-Spike-N440/T794 mutants all had significant differences, among which the difference in the double-site mutation was more significant. At 24 h, due to the decrease in substrate activity, there was no difference among all experimental groups (Fig. 6A). These results indicate that the BA.2 mutant has the lowest cell-cell membrane fusion capacity, while the cell-cell membrane fusion capacity of the double mutation site is the closest to that of WT-Spike-G614.
To comprehensively characterize the membrane fusion properties of different Spike mutants, in addition to determining luciferase activity, this study also analyzed by observing the size and fluorescence intensity of GFP signal syncytia formed at 24 h. The GFP signal syncytia formed by WT-Spike-G614 were the largest, with the strongest fluorescence signal, while those formed by BA.2 were the smallest, with the weakest fluorescence signal. The GFP signal syncytia formed by BA.2-Spike-N440, BA.2-Spike-T794, and BA.2-Spike-N440/T794 all showed enhanced size and fluorescence intensity compared to BA.2, among which the enhancement effect of BA.2-Spike-N440/T794 was the most significant (Fig. 6B). This result is consistent with the luciferase activity detection results, further confirming the differences in membrane fusion ability among different Spike mutants; that is, the double mutant can significantly restore the membrane fusion function of the virus, making it closer to the level of WT-Spike-G614, while the membrane fusion ability of BA.2 is significantly weaker.
## Conclusion and perspective
In immunocompromised patients with long-term infection of SARS-CoV-2, the virus often accumulates numerous adaptive mutations due to persistent replication. These mutations may significantly alter its pathogenic characteristics, including infectivity, pathogenicity, and immune escape ability. Focusing on this important scientific issue, this study conducted a retrospective analysis of 24 early patients who tested persistently positive in nasopharyngeal RT-qPCR. The time of symptom onset varied among the included patients, averaging 7-14 days, among which the duration of symptoms in patients with HIV coinfection was significantly prolonged, exceeding 20 days. To explore the drivers of viral adaptive evolution in depth, the research team comprehensively and precisely characterized the mutation status of the viral genome by generating wholegenome ultra-deep sequencing. These infection cases all occurred in April-May 2022, and gene sequencing results showed that most of the infected strains were Omicron variants.
Through comparative analysis with viral sequences from patients with ordinary SARS-CoV-2 infection during the same period, the study found that the viral strains from patients with SARS-CoV-2-HIV coinfection had additional specific mutations, specifically including NSP3-T779I, NSP15-A94T, Spike-N440K, and Spike-I794T. To clarify the biological functions of these mutations, pseudotyped vesicular stomatitis viruses carrying the target Spike mutations were constructed for functional verification. The results showed that the pseudoviruses carrying the Spike-T794 mutation and the Spike double mutation sites had significantly enhanced ability to infect cells, while the Spike-N440 mutation could increase the neutralizing activity of serum against the pseudovirus, implying that this mutation might reduce the virus's immune escape ability. The GFP signal syncytia formed by BA.2-Spike-N440, BA.2-Spike-T794, and BA.2-Spike-N440/T794 all showed enhanced size and fluorescence intensity compared to BA.2, among which the enhancement effect of BA.2-Spike-N440/T794 was the most significant. Further analysis of prevalent strains showed that the proportion of the above mutation sites in the SARS-CoV-2 strains prevalent during the same period was low. However, it is noteworthy that the Spike-T794 mutation was detected in subsequent JN.1.16 clinical samples. This finding suggests that adaptive mutations generated by SARS-CoV-2 during long-term replication may have potential transmission risks.
In sample 2 of the immunodeficient samples shown in Fig. S1, the iAAV not only has a high quantity but also includes mutations with significant functional implica tions, namely, Spike-N440/T794 double mutations. Subsequent functional experiments confirmed that this double mutation can significantly enhance viral infectivity (Fig. 5A), indicating that these mutations are not random accumulations but are selected by host immune pressure, representing adaptive evolutionary mutations. This further reflects the characteristic of "directional accelerated evolution" of the virus in an immunodeficient environment.
## MATERIALS AND METHODS
## Clinical-sample collection and high-throughput sequencing
Serial samples were collected from the patient periodically from the upper respiratory tract (throat and nasal swab). Nucleic acid extraction was done from 140 M of the sample, using the QIAamp Viral RNA Mini Kit (Qiagen) according to the manufacturer's instruc tions. All samples were tested for the presence of SARS-CoV-2 with a validated one-step RT-qPCR assay in the Public Health China Clinical Center. Amplification reactions were all performed on a Bio-Rad PCR instrument. Samples with a CT value of ≤36 were consid ered to be positive. For the library construction, amplified products were fragmented (300-500 bp) and subjected to end repair, dA-tailing, and adaptor ligation using Illumina DNA Prep Kit (Illumina, USA); libraries were purified with AMPure XP Beads (Beckman Coulter, USA) and quantified via KAPA Library Quantification Kit (KAPA Biosystems, USA) (concentration ≥2 nM). For the high-throughput sequencing, libraries were sequenced on Illumina NovaSeq 6000 platform (Illumina) with PE150 mode (S4 Reagent Kit), cluster density 1,200-1,400 K/mm², PF cluster % ≥85%. 5. For the read generation, raw signals were converted to base calls via RTA (v.3.4.4); samples were demultiplexed with bcl2fastq v.2.20 (--barcode-mismatches 1); FastQ files were generated and pre-filtered with FastQC (v.0.11.9) (adapter trimming via Cutadapt v.4.0, low-complexity read removal via PRINSEQ-lite v.0.20.4).
## Sequence alignment, genome reconstruction, and variant calling
Sequence reads were first filtered to remove low-quality bases using Trimmomatic (v.0.39) with the following parameters: leading bases with Phred quality of <20 were trimmed (LEADING:20). Trailing bases with Phred qualityof <20 were trimmed (TRAIL ING:20). A sliding window of four consecutive bases was applied, and sequences were truncated if the average quality within the window was <20 (SLIDINGWINDOW:4:20). Reads shorter than 50 bp after trimming were discarded (MINLEN:50). These trimmed sequences were then aligned to an early Wuhan reference sequence (NC_045512.2) using the BWA aligner tool (v.0.7.17). Secondary quality control of the cleaned reads filtered for low-quality bases was performed using FastQC v.0.11.9. The Clean Reads were aligned to the SARS-CoV-2 reference genome (Wuhan-Hu-1, NC_045512.2) using BWA-MEM (v.0.7.17). The SAM files were optimized using Samtools (v.1.10). Based on the optimized BAM files, virus genome sequence reconstruction was completed in conjunction with Bcftools (v.1.10.2-34) and IGV (v.2.16.0). On the basis of the genome reconstruction, iSNVs were identified using Bcftools mpileup call. Variant calling (hSNP identification): intrahost single-nucleotide variants (hSNPs/iSNVs) were identified via bcftools mpileup -f NC_045512.2. fasta -Q 20 -d 10000 (calculating base pileup with minimum base quality Q ≥20, maximum depth capped at 10,000 to avoid PCR duplicate bias) and bcftools call -m -v (multiallelic calling). To ensure reliability, hSNPs were further filtered with the following criteria: minimum read depth at the variant site ≥30× (only high-confidence sites with sufficient coverage were retained); minimum allele frequency of the variant base ≥5% (to exclude random sequencing errors, typically ≤ 1%); average Phred quality of variant bases ≥20 (error rate ≤1%); and no strand bias (ratio of variant bases in forward vs reverse strands ≥0.3).
## Phylogenetic analysis
Phylogenetic analysis was performed using the maximum likelihood (ML) method. For the sequence data set preparation, high-quality genome sequences of 14 patients (CT <35, sequencing coverage ≥98%) were included, along with representative BA.2 sequences from GISAID. The Wuhan-Hu-1 strain (NC_045512.2) and B.1 strain (PX622476) were used as outgroups. The ML tree was constructed using IQ-TREE (v.2.2.0.3) with parameters.
## Virus isolation and plaque purification
The stock sample solution was inoculated onto a monolayer of Vero E6-TMPRSS2 cells. After 48 h of incubation, the cell culture supernatant was collected and subjected to three to four serial passages. Viral RNA was extracted from each passage using the QIAamp Viral RNA Mini Kit (Qiagen), followed by viral copy number quantification using RT-qPCR to confirm positivity. The virus was purified through a combination of limiting gradient dilution and plaque purification. Briefly, confluent cells were seeded in six-well plates and allowed to form a 90% monolayer. After the culture medium was discarded, the virus was serially diluted 10-fold across five gradients. Then, 300 µL of each diluted virus suspension was added to the wells, while 300 µL of DMEM alone was used as a negative control. Following 1 h adsorption at 37°C, the inoculum was removed. A 2% carboxymethylcellulose solution was gently mixed 1:1 with 2× cell maintenance medium. The mixture was then incubated in a 37°C, 5% CO 2 incubator for 48 h. Following incubation, cells were fixed by adding 4-5 mL of 4% paraformaldehyde (PFA) per well, ensuring complete coverage, and incubated at room temperature for 1-2 h. After fixation, plates were gently rinsed under running water to remove residual PFA. Plaques were visualized by staining with 1% crystal violet (300 µL/well) and subsequently photographed. All procedures were conducted under biosafety level 3 containment to ensure safe handling of infectious materials.
## Infectivity of Spike pseudotyped vesicular stomatitis virus with target
Confluent 293T cells were seeded in 12-well plates 24 h prior to transfection. Cells were transfected with 2 µg of pcDNA3. 1-Spike-Δ19 plasmid (or other spike mutant variants) using polyethyleneimine transfection reagent. For pseudovirus infection, VSV-Luciferase-G (1 × 10⁵ TCID 50 ) was diluted in DMEM supplemented with 10% fetal bovine serum (FBS) to a final volume of 300 µL, thoroughly mixed in a 1.5 mL microcentrifuge tube, and added to the transfected cells. The infection proceeded for 6-8 h at 37°C in a 5% CO 2 humidified incubator. Following infection, the culture medium was aspirated and cells were washed twice with phosphate-buffered saline. Fresh DMEM supplemented with 10% FBS (1 mL/well) was added, and cultures were maintained at 37°C in a 5% CO 2 humidified incubator for 36 h. The supernatant was then collected and centrifuged at 3,000 × g for 10 minutes at 4°C to remove cellular debris. Purified viral stocks were aliquoted (500 µL/vial) and stored at -80°C for long-term preservation. All plasmids and viral stocks were archived at the Military Medical College repository.
## Neutralizing titer determination
Test sera were heat-inactivated at 56°C for 30 minutes and serially diluted (starting dilution 1:10) in 50 µL volumes. An equal volume (50 µL) of pseudovirus suspension (100 TCID 50 ) was added to each serum dilution, thoroughly mixed, and incubated at 37°C for 2 h to allow neutralization. Subsequently, 50 µL of Hela-hACE2 cell suspension (1-2 × 10⁵ cells/mL) was added to each well, and plates were incubated at 37°C for 24 h. All samples were tested in duplicate. Each 96-well plate included cell controls (three wells, 50 µL cell suspension + 50 µL DMEM) and pseudovirus controls (three wells, 50 µL cell suspension + 50 µL pseudovirus suspension). After 24 h, relative luminescence units (RLUs) were measured. The neutralization percentage was calculated as [(RLU serum -RLU cell control) / (RLU pseudovirus control -RLU cell control)] × 100%. The serum neutralization titer (NT 50 ) was defined as the reciprocal dilution yielding 50% reduction in pseudovirus infection compared to virus controls.
## References
1. Lu, Velkers, Nieuwenhuijse et al. "Hakze-van der Honing RW, van der Poel WHM, et al. 2021. Adaptation, spread and transmission of SARS-CoV-2 in farmed minks and associated humans in the Netherlands" *Nat Commun*
2. Pickering, Lung, Maguire et al. (2022) "Divergent SARS-CoV-2 variant emerges in white-tailed deer with deerto-human transmission" *Nat Microbiol*
3. Marques, Sherrill-Mix, Everett et al. (2022) "Multiple introductions of SARS-CoV-2 alpha and delta variants into white-tailed deer in Pennsylvania" *mBio*
4. Markov, Ghafari, Beer et al. (2023) "The evolution of SARS-CoV-2" *Nat Rev Microbiol*
5. Munnink, Nijhuis, Worp et al. (2022) "Highly divergent SARS-CoV-2 alpha variant in chronically infected immunocom promised person" *Emerg Infect Dis*
6. Borges, Isidro, Cunha et al. (2021) "Long-term evolution of SARS-CoV-2 in an immunocompromised patient with non-hodgkin lymphoma"
7. Gonzalez-Reiche, Alshammary, Schaefer et al. (2023) "Sequential intrahost evolution and onward transmission of SARS-CoV-2 variants" *Nat Commun*
8. Ma, Fu, Jian et al. (0196) "Distinct SARS-CoV-2 populational immune backgrounds tolerate divergent RBD evolutionary preferences" *Natl Sci Rev*
9. Fu, Han, Zhu et al. (2020) "Risk factors for viral RNA shedding in COVID-19 patients" *Eur Respir J*
10. Corey, Beyrer, Cohen et al. (2021) "SARS-CoV-2 variants in patients with immunosuppression" *N Engl J Med*
11. Moran, Cook, Goodman et al. (2021) "Persistent SARS-CoV-2 infection: the urgent need for access to treatment and trials" *Lancet Infect Dis*
12. Fung, Babik (2021) "COVID-19 in immunocompromised hosts: what we know so far" *Clin Infect Dis*
13. Kemp, Collier, Datir et al. (2021) "SARS-CoV-2 evolution during treatment of chronic infection" *Nature*
14. Choi, Choudhary, Regan et al. (2020) "Persistence and evolution of SARS-CoV-2 in an immunocompromised host" *N Engl J Med*
15. Ko, Yingtaweesittikul, Tan et al. (2022) "Emergence of SARS-CoV-2 spike mutations during prolonged infection in immunocompromised hosts" *Microbiol Spectr*
16. Weigang, Fuchs, Zimmer et al. (2021) "Within-host evolution of SARS-CoV-2 in an immunosuppressed COVID-19 patient as a source of immune escape variants" *Nat Commun*
17. Chen, Zody, Germanio et al. (2021) "Emergence of multiple SARS-CoV-2 antibody escape variants in an immunocompromised host undergoing convalescent plasma treatment"
18. Jensen, Luebke, Feldt et al. (2021) "Emergence of the E484K mutation in SARS-COV-2-infected immunocompromised patients treated with bamlanivi mab in Germany" *Lancet Reg Health Eur*
19. Hensley, Bain, Jacobs et al. (2021) "Intractable coronavirus disease 2019 (COVID-19) and prolonged severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) replication in a chimeric antigen receptor-modified T-cell therapy recipient: a case study" *Clin Infect Dis*
20. Truong, Ryutov, Pandey et al. (2021) "Increased viral variants in children and young adults with impaired humoral immunity and persistent SARS-CoV-2 infection: a consecutive case series" *EBioMedicine*
21. Leung, Chorlton, Tyson et al. (2022) "COVID-19 in an immunocompromised host: persistent shedding of viable SARS-CoV-2 and emergence of multiple mutations: a case report" *Int J Infect Dis*
22. Lee, Shah, Hoyos et al. (2022) "Prolonged SARS-CoV-2 infection in patients with lymphoid malignancies" *Cancer Discov*
23. Hueso, Pouderoux, Péré et al. (2020) "Convales cent plasma therapy for B-cell-depleted patients with protracted COVID-19" *Blood*
24. Malin, Cristanziano, Horn et al. (2022) "SARS-CoV-2-neutralizing antibody treatment in patients with COVID-19 and immunodeficiency due to B-cell non-hodgkin lymphoma" *Blood Adv*
25. Gaitzsch, Passerini, Khatamzas et al. (2021) "COVID-19 in patients receiving CD20-depleting immunochemotherapy for B-cell lymphoma"
26. Terpos, Gavriatopoulou, Fotiou et al. (2021) "Poor neutralizing antibody responses in 132 patients with CLL, NHL and HL afteminst SARS-CoV-2: a prospective study" *Cancers (Basel)*
27. Molica, Giannarelli, Lentini et al. (2022) "Efficacy of the BNT162b2 mRNA COVID-19 vaccine in patients with chronic lymphocytic leukemia: a serologic and cellular study" *Chemotherapy*
28. Khatamzas, Antwerpen, Rehn et al. (2022) "Accumulation of mutations in antibody and CD8 T cell epitopes in a B cell depleted lymphoma patient with chronic SARS-CoV-2 infection" *Nat Commun*
29. Wilkinson, Richter, Casey et al. (2022) "Recurrent SARS-CoV-2 mutations in immunodeficient patients" *Virus Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548497&blobtype=pdf | # Rebound of Respiratory Virus Activity and Seasonality to Pre-Pandemic Patterns
Rainer Gosert, | Klaudia Naegele, | Weiss, | Bingisser, Christian Nickel, Jakob Meyer, Martin Siegemund, Stefano Bassetti, Sarah Dräger, Christoph Berger, Fabian Franzek, Ulrich Heininger, Julia Bielicki, Hans Hirsch, | Peter, M Keller, | Maja Weisser, Nina Khanna, Sarah Tschudin-Sutter, Karoline Leuzinger
## Abstract
The emergence of SARS-CoV-2 and the implementation of non-pharmaceutical interventions (NPIs) profoundly disrupted the transmission dynamics of respiratory viruses, altering their epidemiology and seasonality. However, comprehensive long-term data on these shifts and their post-pandemic implications remain limited. This study analyzed syndromic multiplex panel testing data from 83′823 respiratory specimens collected from 56,519 patients with respiratory tract infections (RTIs) at two tertiary care centers in northwestern Switzerland to systematically assess changes in respiratory virus circulation, seasonality, age distribution, and disease burden across pre-pandemic (2010-2019), pandemic (2019-2022), and post-pandemic (2022-2024) periods. Pre-pandemic, influenza virus (IV), respiratory syncytial virus (RSV), human coronavirus (HCoV), human metapneumovirus (hMPV), and human parainfluenza virus (HPIV) followed distinct seasonal patterns. During the pandemic, SARS-CoV-2 replaced these viruses, leading to a 70-90% decline in their activity (p < 0.001), while rhinovirus/enterovirus and adenovirus were less affected. After NPIs were lifted, substantial off-season activity with markedly higher case numbers and more hospitalizations, especially among pediatric patients, occurred for IV-A/B, RSV, and atypical bacteria. In post pandemic years, virus-specific seasonality is rebounding, with patterns resembling those seen pre-pandemic. However, higher case numbers, increased hospitalizations, and sustained shifts in age distribution persist. The COVID-19 panemic significantly impacted the etiology, seasonality, and age distribution of RTIs. As NPIs were eased, susceptibility to RTIs, particularly among pediatric patients, increased, resulting in more hospitalizations. While post-pandemic periods show a return to pre-pandemic activity patterns, ongoing monitoring is essential to anticipate shifts in respiratory virus dynamics as immunity levels and virus characteristics evolve.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.
## 2 | Materials and Methods
## 2.1 | Patient Cohorts and Inclusion Criteria
The University Hospital Basel is the largest tertiary care facility in Northwestern Switzerland, with a capacity of 700 beds. The University Children′s Hospital Basel, with a capacity of 120 beds, is the leading healthcare provider for children and adolescents in northwestern Switzerland and serves as a specialized referral center. This retrospective study included patients who presented with RTI symptoms to the outpatient clinics or the adult and pediatric emergency departments of both hospitals and received syndromic multiplex panel testing between July 2010 and June 2024. Patients who were admitted following their initial presentation with acute RTI symptoms were included in the analysis. In contrast, hospital-acquired infections, defined as patients who developed RTI symptoms after hospitalization are not included in this analysis. Both, the University Hospital Basel and the University Children′s Hospital Basel, perform among the highest numbers of syndromic multiplex panel tests for patients with RTIs in Switzerland [3,8], providing comprehensive, high-quality epidemiological data for assessing long-term trends in respiratory pathogen dynamics.
## 2.2 | Timeline of Implementation of Non-Pharmaceutical Interventions
In Switzerland, governmental NPIs were introduced in calendar week 11 of 2020. The initial measures included canceling mass gatherings, closing schools and cultural venues. These physical distancing measures were later expanded by banning all gatherings and closing non-essential business. Restrictions were cautiously lifted 7 weeks later but reinstated during the following winter season as Alpha, Delta and Omicron variants surged across Europe. NPIs began to ease in February 2022 and were fully lifted by April 2022 (Supplementary Table 1).
## 2.3 | Syndromic Multiplex Panel Testing
Respiratory specimens were collected using two swabs from the nasopharyngeal and oropharyngeal sites, respectively, and combined into a single universal transport medium tube (Copan, Brescia, Italy). In younger children, only nasopharyngeal swabs were obtained.
All respiratory specimens were sent to the Clinical Virology Unit at the University Hospital Basel, which operates a dedicated emergency diagnostic service available 24 h a day, 7 days a week. Upon receipt, specimens were immediately processed in the diagnostic laboratory, and multiplex panel testing was initiated without delay [3,[9][10][11]. In practice, respiratory samples were analyzed within minutes to a maximum of 1 h after reciept, ensuring a time-to-result of ≤ 4 h from collection. This rapid workflow guaranteed optimal preanalytical conditions, robust diagnostic performance, and timely results to support clinical triage and infection control. Over the study period, three different syndromic multiplex panel testing systems were utilized, targeting both bacterial and viral pathogens. From July 2010 to January 2016, the RespiFinder-22® Respiratory Pathogen Panel (RPP; Patho-Finder, Maastricht, Netherlands) [10] was routinely used, which detects 18 viral and four bacterial respiratory pathogens.
In January 2016, syndromic testing was switched to the NxTAG RPP on the MAGPIX platform (Diasorin, Saluggia, Italy) [10], which detects 18 viral and 2 bacterial pathogens. In November 2016, testing transitioned to the Biofire FilmArray RPP (bioMérieux, Marcy-l′Étoile, France) [3], which detects 18 viral and four bacterial targets. Additionally, since October 2021, the Xpert® Xpress-CoV-2/Flu/RSV-Plus system (Cepheid, CA, USA) [9] has been employed, targeting the three most critical viral respiratory pathogens in tertiary care settings. Since Bordetella pertussis and Bordetella parapertussis were only partially covered by RespiFinder-22® RPP and NxTAG RPP, both bacterial pathogens were detected alongside the syndromic multiplex panel tests using laboratory-developed tests targeting an insertion sequence [11] (Supplementary Table 2).
## 2.4 | Statistics and Data Analysis
All statistical data analyses were performed in R (https://www. r-project.org/), and Prism (version 10; Graphpad Software, CA, USA) was used for data visualization.
For patients with repeated testing, results were consolidated into infection episodes to avoid double-counting. Consecutive positive tests for the same respiratory pathogen within 90 days were classified as a single infection episode, acknowledging that viral shedding can persist for prolonged periods, particularly in immunocompromised individuals. Positive tests for the same respiratory pathogen occurring more than 90 days apart were considered new episodes. Detections of different respiratory pathogens were always treated as separate events regardless of timing. The 90-day interval was chosen as it exceeds typical shedding durations in immunocompetent hosts while also accommodating prolonged persistence occasionally described in immunocompromised patients [12][13][14][15].
In cases of co-detections, each respiratory pathogen was counted independently in pathogen-specific analyses (e.g., a sample positive for both RSV and HAdV contributed to both RSV and HAdV counts). For patient-level outcomes such as unique case numbers or hospitalizations, however, coinfected patients were counted only once to avoid duplication.
Hospitalizations associated with IV-A/B or RSV were defined as cases in which patients presented with acute RTI symptoms to the emergency or outpatient clinics, tested positive for the respective pathogen by syndromic multiplex panel testing, and were admitted to the hospital within 24 h. This definition was applied consistently across pandemic and post-pandemic periods to allow a comparable assessment of virus-associated hospitalization burden.
## 3 | Results
A total of 83′823 respiratory clinical specimens were submitted for routine syndromic multiplex panel testing from 56,519 patients with RTIs (female: 26,382, 46.7%; pediatric patients ≤ 18 years: 13,187, 23.3%) between July 2010 and June 2024 (Supplementary Table 3). Median age of adult patients was 66 years (range: 19 to 111 years), and 6 years (range: 1 to 18 years) for pediatric patients.
A median of 401 syndromic multiplex panel tests for non-SARS-CoV-2 respiratory pathogens were conducted monthly, with a surge in March 2020, coinciding with the initial detection of SARS-CoV-2 in Switzerland. Although syndromic multiplex panel testing decreased after the pandemic, it remained higher than pre-pandemic levels (Figure 1A). The overall test positivity rates for non-SARS-CoV-2 respiratory pathogens were comparable across pre-pandemic periods, ranging from 35.1% to 69.2%. These rates significantly decreased during the pandemic and rebounded in the post-pandemic periods (Figure 1B). Bacterial pathogens like M. pneumoniae and B. pertussis were nearly absent in pandemic years but reemerged in the 2023-24 season, closely aligning with pre-pandemic positivity rates (Figure 1C).
During the pre-pandemic periods, IV-A/B, RSV, seasonal HCoVs, and human metapneumovirus (hMPV) consistently exhibited peak activity during the winter months each year. In contrast, human rhinovirus/human enterovirus (HRV/HEV) displayed biannual seasonal peaks, one in winter and one in summer, while human adenovirus (HAdV) infections were consistently detected throughout the year. Human parainfluenza virus (HPIV) infections were observed year-round, with peak activity noted for HPIV-1, HPIV-2 and HPIV-4 during the autumn and winter months, and for HPIV 3 during the spring and summer months (Figure 2; Supplementary Figure 1). The seasonal patterns of respiratory virus activity were significantly disrupted following the emergence of SARS-CoV-2 in Switzerland in February 2020. Although the first weeks of 2020 were dominated by respiratory viruses other than SARS-CoV-2, the emergence of SARS-CoV-2, followed by the implementation of NPIs led to an unprecedented decline in their activity (Figure 3). Within just 3 weeks, seasonally circulating respiratory viruses were almost entirely replaced by SARS-CoV-2 [3]. During the pandemic winter seasons, positivity rates for IV-A, IV-B, RSV, hMPV, HCoV, HPIV and atypical bacteria dropped substantially, leading to a 70-90% decline in their activity (Table 1). Specifically, IV-A positivity rates dropped to 2-50% of pre-pandemic levels, IV-B to 3-30%, and RSV to 2-30% (Table 1). Notably, the decrease in detection of non-SARS-CoV-2 respiratory pathogens coincided with ongoing SARS-CoV-2 circulation. SARS-CoV-2 cases initially peaked in March 2020, followed by significant surges during the winters of 2020-21 and 2021-22 (Figure 3). In contrast, HRV/HEV and HAdV seemed to be less impacted (Figure 2, Table 1). Their activity maintained an average positivity rate of 13.7% for HRV/ HEV and 4.5% for HAdV during the pandemic 2020-21 season, compared to < 1% for most other respiratory pathogens (Table 1).
Based on these findings, we assessed shifts in peak activity of respiratory pathogens during the pandemic and subsequent post-pandemic seasons. Compiled data on IV-A/B epidemics during the 9 surveillance years preceding the COVID-19 pandemic showed IV-A/B activity typically beginning in November and peaking in January, with an offset until March (Figure 4). Following the emergence of SARS-CoV-2, IV-A/B activity declined in March 2020, remained historically low throughout the summer of 2020, and was absent in the 2020-21 winter season. The 2021-22 influenza epidemic began late in January 2022, peaked at a positivity rate of 10.2% in April, and persisted until June. In the following post-pandemic seasons, seasonal patterns gradually re-established, with IV-A/B epidemics peaking at 29.8% in December 2022 and 25.2% in January 2024 (Figure 4, Supplementary Figure 1).
Cumulative pre-pandemic RSV activity typically began in October, peaked in January, and subsided by March. During the pandemic winter seasons, RSV positivity rates dropped substantially. Following the easing of NPIs, an early epidemic onset occured in May 2021, peaking at 33.8% positivity in July and persisting until December (Figure 3, Figure 4). In the postpandemic winters of 2022-23 and 2023-24, RSV seasonality returned to pre-pandemic patterns, with peak activity in November 2022 and December 2023
When assessing the activity of other respiratory viruses, including hMPV, HPIV types 1-4 and HCoV types -229E, -HKU1, -NL63, and -OC43, we observed similar patterns of significantly reduced positivity rates during the pandemic seasons. This was followed by substantial off-season activity after the lifting of NPIs, and a subsequent return to pre-pandemic seasonality in the post-pandemic periods (Supplementary Figure 1). Notably, bacterial pathogens such as M. pneumoniae and B. pertussis were completely absent during pandemic years, but showed a strong resurge with increased positivity rates in the 2022-2023 and 2023-2024 post-pandemic seasons (Figure 2, Table 1).
We next assessed whether the age distribution of patients with RTIs differed across study periods. In the pre-pandemic period, 25,737 patients were tested (median age 60 years, IQR 27-76), of whom 27.0% were children. During the pandemic, 12,410 patients were tested (median age 59 years, IQR 29-75), with 23.1% pediatric patients. In the post-pandemic period, 18,372 patients were tested (median age 62 years, IQR 36-76), with 20.4% children. Although statistical testing indicated significant differences in median patient age between periods (p < 0.001), this is likely driven by the large sample size analyzed (n = 56,519), as median values and interquartile ranges were highly similar. By contrast, the proportion of pediatric patients declined stepwise, from 27.0% pre-pandemic to 23.1% during the pandemic and 20.4% post-pandemic, indicating that children constituted a progressively smaller proportion of the tested patient population in the later years (Supplementary Table 4).
Analyzing the age distribution of IV-A/B-positive cases (Figure 5), we found a cumulative median patient age of 65 years (IQR: 47-79 years) during the pre-pandemic that significantly decreased to 36 years (IQR: 17-61 years) during the pandemic. This shift was likely driven by a higher impact on pediatric patients, as evidenced by the increased proportion of IV-A/B cases among patients ≤ 18 years, reaching 31.2% in the 2021-22 winter season, following a complete absence of viral activity during the 2020-21 period (Figure 5; Supplementary Figure 2). In the post-pandemic 2023-24 season, the median patient age returned to pre-pandemic levels at 51 years (IQR: 23-70 years), with a decrease of affected pediatric patients to 21.2% (Figure 5; Supplementary Figure 2). Similar trends were observed for HCoV and HPIV infections, with significantly younger patients during the COVID-19 pandemic (p < 0.001; Figure 5) and a significantly higher proportion of pediatric patients affected (p < 0.001; Supplementary Figure 2). After the complete absence of RSV circulation during the pandemic, RSV activity resurged following the lifting of NPIs, leading to an off-season RSV epidemic during the 2020-21 season. This resurgence disproportionately affected pediatric patients, who accounted for a significantly increased 75.1% of cases. However, this proportion decreased to 58% in subsequent seasons, with an increase in RSV infections among adult patients. Correspondingly, the median patient age dropped to 5 years (IQR: 5-36 years) during the pandemic but reverted to 8 years (IQR: 5-58 years) in the post-pandemic 2023-24 season, approaching the pre-pandemic median of 11 years (IQR: 8-33 years) (Figure 5; Supplementary Figure 2). For HRV/HEV, we observed that pediatric and adult patients were alternately more affected during the pre-pandemic periods, as evidenced by shifts in the median patient age and the proportions of pediatric and adult patients affected (Figure 5; Supplementary Figure 2). During the 2020-21 and 2021-22 seasons, a high proportion of up to 71% of pediatric patients were affected, accompanied by a significant decrease in the median patient age. A similar trend was observed for hMPV, while HAdV infections consistently predominated among pediatric patients across all assessed periods. For M. pneumoniae, the median patient age was 37 years (IQR: 20-66 years) during the pre-pandemic. M. pneumoniae activity was absent following the emergence of SARS-CoV-2 in 2020; however, a surge in cases was observed in the 2023-24 post-pandemic season, with a median patient age of 20 years and a significantly higher proportion of pediatric patients affected (p < 0.001; Figure 5; Supplementary Figure 2). Similarily, paediatric patients were primarily affected when B. pertussis/B. parapertussis infections resurged in the 2022-23 season, after its absence during pandemic years. The following 2023-24 post-pandemic season saw a return to higher median patient ages (Figure 5; Supplementary Figure 2). Finally, we assessed hospital admissions among adult patients across pandemic and post-pandemic seasons. SARS-CoV-2 emerged during the typical influenza season in Europe, coinciding with a decline in hospital admissions during the 2019-20 season. No IV-A/B-associated hospitalizations were recorded in the 2020-21 season, while off-season admissions occurred between February and April 2022. In the post-pandemic period, increased IV-A/B activity and higher case numbers were associated with a rise in hospital admissions temporally linked to these infections (R² > 0.98; p < 0.001; Figure 6), peaking at 155 and 157 cases in December and January of the 2022-23 and 2023-24 seasons, respectively (Figure 6). A similar trend was observed for RSV, with off-season hospitalizations peaking in June 2022, followed by a significant increase in RSV infections and related hospital admissions during the 2022-23 season (R 2 > 0.96; p < 0.001; Figure 6).
## 4 | Discussion
This comprehensive study of the viral and bacterial etiologies of RTIs in pediatric and adult populations in northwestern Switzerland across 14 years, including pre-pandemic, pandemic, and post-pandemic periods, showed significant shifts in the etiological landscape of RTIs, providing valuable insights into their evolving epidemiology and clinical impact.
Before the COVID-19 pandemic, respiratory viruses such as IV-A/B, RSV, HCoV and HPIV followed distinct seasonal patterns, peaking in winter, while HRV and HAdV circulated throughout Stringent NPIs included the prohibition of gatherings at private and public events, suspension of global travel, border closures to EU/EFTA countries, closure of non-essential retail stores, mandatory remote work, mask mandates, and enforced isolation or quarantine for COVID-19 cases. Less stringent NPIs involved mandatory mask-wearing in enclosed public indoor spaces (such as public transport, offices, schools, etc.), physical distancing, and recommendations for remote work. For the full timeline of all NPIs, see Supplementary Table 1. Indication for SARS-CoV-2 testing changed over the study period and was conducted using different nucleic acid testing methods (Supplementary Table 5). SARS-CoV-2 vaccination is shown as green bars. The gray bar represents the respective winter season, spanning from November to February of the specified years. e the rate ratio is calculated by comparing the proportion of patients with a positive non-SARS-CoV-2 respiratory virus or bacterial detection in the indicated year to the pre-pandemic level. A rate ratio of 1 indicates no difference in positivity rates between pandemic and pre-pandemic seasons; > 1 indicates a higher positivity rate in pandemic compared to pre-pandemic seasons; < 1 indicates a lower positivity rate in pandemic compared to pre-pandemic seasons.
the year [2], with their relative prevalence decreasing in winter due to interference with other viruses like IV-A/B [16]. The emergence of SARS-CoV-2 and the implementation of NPIs in early 2020 led to a near-complete displacement of seasonally circulating respiratory viruses within just 3 weeks [3][4][5]. Surveillance data revealed a significant increase in SARS-CoV-2 infections, alongside a marked decline in IV-A/B and RSV cases across various climates [6,7,17,18]. Other non-SARS-CoV-2 viruses also exhibited altered circulation patterns. Seasonal HCoV [6,17,19], HPIV [6,20], and hMPV [6,20] declined sharply at the onset of the pandemic, coinciding with the implementation of stringent NPIs. The reduction in respiratory virus circulation was most pronounced during the initial phase of the COVID-19 pandemic when most stringent NPIs were enforced and persisted to varying degrees during subsequent waves of SARS-CoV-2 when novel variants emerged [21,22].
Notably, non-enveloped viruses like HRV and HAdV were less affected by NPIs. Although both declined at the pandemic′s onset [6,17], they rebounded faster than other respiratory viruses. This suggests that their environmental stability, ability to persist on surfaces, and transmission through both direct contact and fomites make them less susceptible to NPIs, though their transmission dynamics are also shaped by shifts in population immunity and the timing of NPIs [23]. Furthermore, their extensive genomic diversity with > 100 distinct serotypes limits cross-neutralization, leaving individuals susceptible to reinfection by different strains. Indeed, HRV infections were consistently detected in both pediatric and adult populations, with shifts between age groups across different seasons, likely reflecting the circulation of distinct HRV types that elicit different immune responses [24]. Our internal validation further supports that HRV/HEV detections in our data set represent HRV circulation in the vast majority of cases ( > 90%; data not shown), consistent with previous reports that respiratory syndromic multiplex panel tests predominantly detect rhinoviruses rather than enteroviruses [8].
By mid-2022, governments eased COVID-19 restrictions. With the potential waning of population immunity due to reduced virus exposure during the pandemic years, and fewer preventive measures in place, there was a subsequent surge in respiratory virus circulation. Our study observed an unusual inter-seasonal surge in RSV infections and hospitalizations in 2021, mirroring a global trend, in which RSV activity peaked in winter instead of summer in Australia and South Africa [7,25,26], and in summer-fall instead of winter in Europe, the U.S. and Canada [27,28]. These findings suggest that the easing of NPIs, along with a rising "immunity debt" in the population after prolonged periods without exposure to certain viruses, contributed to these atypical seasonal trends [29]. In 2022, the RSV season started earlier than usual in the Northern Hemisphere, and RSVassociated hospitalizations peaked between October and December [30,31]. Similarly, IV-A/B resurgence was a major concern following NPI relaxation. Our data align with global trends showing a later occurrence of the 2021-22 influenza season in the Northern Hemisphere as compared to prepandemic years, with shorter seasons in Canada and Europe and a longer season in the U.S [32,33]. By winter 2022-23, IV-A/B activity rebounded with an unusually early onset in Europe, the U.S., and Canada [34][35][36].
Our findings also show shifts in the typical age distribution of RTIs. RSV is known to primarily affect young children, with the highest incidence in infants. However, in our tertiary care hospital that primarily serves adults, a significant number of individuals over 65 years were also affected. IV-A/B was more prevalent among older adults, while HCoV, HPIV, and HRV/ HEV infections remained common across all age groups. During the second and third year of the pandemic, pediatric RTIs due to IV-A/B, RSV, HCoV, and HPIV surged. In the postpandemic period, infections shifted towards older children and adults, reflecting a delayed exposure-driven susceptibility.
Reduced exposure to respiratory viruses early in the COVID-19 pandemic likely contributed to waning population immunity as levels of antivirus specific antibodies declined [37], leading to an "immunity debt" [29]. In the first months of life, infants are partially protected by maternal antibodies transferred transplacentally. Lower levels of maternal antibodies being passed to infants during pregnancy may have resulted in a birth cohort of immunologically naive infants or infants with low levels of maternal antibodies, making them more vulnerable during their first viral exposure. This may have facilitated viral transmission among pediatric populations after the easing of NPIs when schools and childcare facilities reopened, leading to secondary household transmission affecting high-risk groups such as older adults. Immunologically naive children or those with limited prior exposure to RSV during the COVID-19 pandemic may have experienced more severe disease with a higher risk of hospitalization as compared to pre-pandemic years [38,39]. Our study confirms that adult populations were also affected by consecutive major RSV epidemics in the 2022-2023 and 2023-2024 seasons, with increased RSV-associated hospitalizations as compared to pre-pandemic periods.
Similar trends were observed for IV-A/B, with a notable increase in viral activity among paediatric populations during the early COVID-19 pandemic after NPIs were eased, while post-pandemic seasons showed a return to pre-pandemic patterns [40,41]. Early in the pandemic, IV-A/B-associated hospitalizations dropped significantly, and remained at historically low levels until the 2021-22 influenza season. This aligns with surveillance data from Switzerland [42], other European countries [43], the U.S. [32], Canada [6], Australia [20], and South Africa [7], all of which reported declines in IV-A/Bassociated hospitalizations and mortality. By 2021-22, IV-A/B- associated hospitalizations returned to pre-pandemic levels [44]. The subsequent 2022-23 and 2023-24 seasons saw hospitalization rates surpassing pre-pandemic levels, consistent with reports from North America, Europe and Australia [32-35, 40, 41]. This surge likely reflects a combination of increased virus circulation, waning population immunity, reduced vaccine coverage, and antigenic drift of circulating strains, underscoring the urgent need for enhanced influenza surveillance and vaccine adaptation [41,45].
Vaccination dynamics during the study period likely contributed to respiratory pathogen circulation and disease burden. Influenza vaccination uptake rose temporarily by about 5-10% in 2020-21, particularly among older adults, supported by pandemic-related campaigns, but declined again in 2023-24 by 3-10% compared with the two preceding seasons in both the United States and Europe [46][47][48]. This transient rise in influenza vaccine uptake during the 2020-21 season, together with NPIs, likely contributed to historically low influenza incidence during the first pandemic winter, whereas reduced coverage in later years may have contributed to the high case numbers and associated hospitalizations observed after 2022.
Pertussis remains an important respiratory pathogen, particularly in young children, although vaccination uptake in highincome countries has generally been high. In the United States and Europe, infant coverage has consistently exceeded 90% over the past decade and remained above this level through 2023-24 [49,50]. Complementary strategies have further strengthened protection. Maternal immunization, recommended in many European countries for the past 5-12 years, has been associated with significant reductions in hospitalizations among infants [51]. In contrast, uptake of adolescent and adult boosters has remained limited [52,53]. In our epidemiological study, pertussis activity was absent in 2020-21 but re-emerged in 2023-24 at levels comparable to the pre-pandemic period, consistent with U.S. and European surveillance reports [49,54]. This resurgence likely reflects an "immunity gap" resulting from reduced pathogen circulation during COVID-19 restrictions, which limited opportunities for natural boosting of immunity. Waning vaccine-induced protection and persistently low uptake of adolescent and adult boosters may have further contributed to the observed rebound [55,56]. Comparable dynamics were observed for other atypical bacteria. Mycoplasma pneumoniae incidence declined markedly during the COVID-19 pandemic, reaching its lowest incidence in over a decade [57,58]. Following the relaxation of NPIs, cases and associated hospitalizations increased, consistent with the concept of an "immunity gap" caused by reduced exposure during the pandemic, which likely contributed to heightened post-pandemic susceptibility.
At the same time, COVID-19 vaccination markedly reduced the burden of severe courses of COVID-19, altering the relationship between infection incidence and clinical outcomes. In Switzerland, vaccination began in late December 2020 with prioritization of healthcare workers, older adults, and other high-risk groups. By mid-2021, eligibility was expanded to the general adult population, and by the end of 2021 approximately 67% of the population was fully vaccinated. Comparable coverage levels were reported in the United States and Europe [59,60]. The rapid uptake among high-risk groups was associated with a decoupling of infection incidence from severe outcomes, as hospitalization and mortality rates were substantially lower among fully vaccinated individuals, particularly those who had received booster doses, compared with unvaccinated populations [61,62]. The widespread rollout of vaccines, together with the policy-driven relaxation of NPIs in 2022, created the conditions for the re-emergence of other respiratory viruses, reflecting both restored contact patterns and altered population immunity after 2 years of suppressed circulation. In addition, viral interference with widespread SARS-CoV-2 circulation during the early pandemic waves likely further suppressed the activity of other respiratory viruses, including influenza and RSV [3].
Taken together, recent vaccination dynamics, including temporary increases followed by subsequent declines in influenza vaccination coverage among older adults, sustained high infant pertussis coverage with persistent gaps in adolescent and adult booster uptake, and the widespread rollout of COVID-19 vaccines, likely contributed to the post-2022 shifts in susceptibility and disease burden observed across multiple respiratory pathogens in the United States and Europe. These trends underscore the complex interplay of vaccination, waning and cohortspecific immunity, viral interference, and behavioral changes such as mobility patterns and the resumption of social contact, indicating that vaccination alone does not fully account for the observed epidemiological changes.
This study leveraging comprehensive epidemiological data from extensive molecular syndromic multiplex panel testing over a 14-year period at two major tertiary care hospitals in northwestern Switzerland represents one of the longest continuous surveillance periods in comparable research, and allowed us to assess longterm trends in respiratory pathogen dynamics. Following the significant impacts of SARS-CoV-2 interference alongside the implementation of NPIs on respiratory virus circulation, postpandemic seasons reveal a notable resurgence of seasonal patterns. However, this 'return to normal′ is accompanied by a hgher disease burden, and altered age distributions, indicating that a full restortion of pre-pandemic age ditributions patterns and RTI incidence levels may require ongoing viral circulation to rebuild population immunity. Several factors, including differences in the virulence of epidemic IV-A/B and RSV strains and advancements in RSV prophylaxis, such as the monoclonal antibody Nirsevimab® (Sanofi and AstraZeneca) [63], and newly approved vaccines Arexvy® (GSK) [64], and Abrysvo® (Pfizer) [65] may also influence these trends. Given the uncertainties around post-pandemic viral circulation, our findings emphasize the critical need for ongoing surveillance of respiratory virus activity and seasonal trends for refining vaccination strategies for vaccine-preventable RTIs. These adaptive public health strategies are essential to anticipate and mitigate future respiratory virus surges in pediatric and older populations, who may be more vulnerable due to potential immunity gaps following the COVID-19 pandemic, and for reducing overall disease burden.
## 4.1 | Limitations
This study has some limitations. First, syndromic multiplex testing platforms varied over the study period. Given the high sensitivity of nucleic acid testing in symptomatic patients [3,5,9,10], these methodological differences are unlikely to have significantly affected detection rates. Second, variations in testing regimens during the pandemic may have influenced virus detection; though, testing behavior alone rarely explains the magnitude of observed infection waves [66]. Third, reluctance to seek medical care early in the pandemic may have led to underreporting of RTIs. Since other viral illnesses remained low even after COVID-19 restrictions were lifted, healthcare avoidance alone may not fully explain decreased notifications, and emergency physicians involved in this study did not note a major decline in patient numbers during the pandemic. Fourth, pandemic-driven behavioral and procedural changes may have influenced sampling practices. Although the clinical indication for multiplex panel testing remained unchanged during the whole study period and was restricted to patients presenting with clinically relevant RTI symptoms, heightened public awareness during the pandemic may have resulted in some patients with milder illness presenting for hospital care. This may have contributed to the observed surge in testing volumes in early 2020 and the sustained higher testing thereafter. While statistical analyses indicated significant differences in patient age distribution across the pre-pandemic, pandemic, and postpandemic years (Supplementary Table 4), these are likely attributable to the very large sample size analyzed (n = 56,519). Median ages (59-62 years) and interquartile ranges, as well as the proportions of pediatric patients (20-27%), were highly similar across periods, indicating that the overall study population remained demographically stable. This stability provides important context, as the substantial shifts observed in the median age of patients with specific respiratory pathogen detections can therefore be interpreted as true epidemiological changes rather than artifacts of differences in the underlying study population. Fifth, we did not perform medical chart reviews to adjudicate whether hospitalizations were directly attributable to IV-A/B or RSV. Instead, we applied strict criteria, defining hospitalization as the admission of a patient who presented to the emergency department or outpatient clinics with acute RTI symptoms, tested positive by multiplex panel test, and was admitted to the hospital within 24 h of testing.
While our study provides valuable insights into the rebounds of respiratory virus circulation patterns and aligns with current surveillance data [40][41][42], these limitations emphasize the need to consider potential confounding factors when interpreting long-term epidemiological virus data.
## 5 | Conclusion
Our analysis of patients with RTIs over nine pre-pandemic, three pandemic, and two post-pandemic seasons highlights the profound impacts of SARS-CoV-2 emergence and NPIs on the epidemiology of respiratory viruses and bacteria, including shifts in etiology, seasonality, age distribution, disease burden, and related hospitalizations. While the first pandemic year caused a dramatic change in the etiology of RTIs, the subsequent easing of public health measures led to the resurgence of multiple respiratory pathogens, albeit with persistent agerelated disparities. After the pandemic, virus-specific seasonality is gradually returning, with patterns starting to resemble those seen pre-pandemic. However, shifts in affected age groups and fluctuations in virus circulation indicate that full epidemiological normalization has yet to occur. This underscores the need for continued surveillance and adaptive public health strategies to mitigate future disruptions in respiratory virus dynamics.
## References
1. Ison, Hirsch (2019) "Community-Acquired Respiratory Viruses in Transplant Patients: Diversity, Impact, Unmet Clinical Needs" *Clinical Microbiology Reviews*
2. Moriyama, Hugentobler, Iwasaki (2020) "Seasonality of Respiratory Viral Infections" *Annual review of virology*
3. Leuzinger, Roloff, Gosert (2020) "Epidemiology of Severe Acute Respiratory Syndrome Coronavirus 2 Emergence Amidst Community-Acquired Respiratory Viruses" *Journal of infectious diseases*
4. Chow, Uyeki, Chu (2023) "The Effects of the COVID-19 Pandemic on Community Respiratory Virus Activity" *Nature Reviews Microbiology*
5. Leuzinger, Gosert, Søgaard (2021) "Epidemiology and Precision of SARS-CoV-2 Detection Following Lockdown and Relaxation Measures" *Journal of Medical Virology*
6. Park, Seo, Han et al. (2021) "Respiratory Virus Surveillance in Canada During the COVID-19 Pandemic: An Epidemiological Analysis of the Effectiveness of Pandemic-Related Public Health Measures in Reducing Seasonal Respiratory Viruses Test Positivity" *PLoS One*
7. Tempia, Walaza, Bhiman (2020) "Decline of Influenza and Respiratory Syncytial Virus Detection in Facility-Based Surveillance During the COVID-19 Pandemic, South Africa" *Eurosurveillance*
8. Gosert, Koller, Meyer (2024) "Multicenter Evaluation of the QIAstat-DX and the BioFire Multiplex Panel Tests for the Detection of Respiratory Pathogens" *Journal of Medical Virology*
9. Goldenberger, Leuzinger, Sogaard (2020) "Brief Validation of the Novel Genexpert Xpress SARS-CoV-2 PCR Assay" *Journal of Virological Methods*
10. Beckmann, Hirsch (2016) "Comparing Luminex NxTAG-Respiratory Pathogen Panel and RespiFinder-22 for Multiplex Detection of Respiratory Pathogens" *Journal of Medical Virology*
11. Abu Raya, Bamberger, Gershtein et al. (2012) "The Laboratory Diagnosis of Bordetella Pertussis Infection: A Comparison of Semi-Nested PCR and Real-Time PCR With Culture" *European journal of clinical microbiology & infectious diseases: official publication of the European Society of Clinical Microbiology*
12. Tabatabai, Schnitzler, Prifert "Parainfluenza Virus Infections In Patients With Hematological Malignancies or Stem Cell Transplantation: Analysis of Clinical Characteristics, Nosocomial Transmission and Viral Shedding" *PLoS One*
13. Rachow, Lamik, Kalkreuth (2020) "Detection of Community-Acquired Respiratory Viruses In Allogeneic Stem-Cell Transplant Recipients and Controls-A Prospective Cohort Study" *Transplant Infectious Disease*
14. Khanna, Widmer, Decker (2008) "Respiratory Syncytial Virus Infection in Patients With Hematological Diseases: Single-Center Study and Review of the Literature" *Clinical Infectious Diseases*
15. Khanna, Steffen, Studt (2009) "Outcome of Influenza Infections in Outpatients After Allogeneic Hematopoietic Stem Cell Transplantation" *Transplant Infectious Disease*
16. Nickbakhsh, Mair, Matthews (2019) "Virus-Virus Interactions Impact the Population Dynamics of Influenza and the Common Cold" *Proceedings of the National Academy of Sciences*
17. Olsen, Winn, Budd (2021) "Changes in Influenza and Other Respiratory Virus Activity During the COVID-19 Pandemic-United States, 2020-2021" *MMWR. Morbidity and Mortality Weekly Report*
18. Olsen, Azziz-Baumgartner, Budd (2020) "Decreased Influenza Activity During the covid-19 Pandemic-United States" *MMWR. Morbidity and Mortality Weekly Report*
19. Yum, Hong, Sohn et al. (2021) "Trends in Viral Respiratory Infections During COVID-19 Pandemic, South Korea" *Emerging Infectious Diseases*
20. El-Heneidy, Ware, Robson et al. (2022) "Respiratory Virus Detection During the COVID-19 Pandemic in Queensland, Australia" *Australian and New Zealand Journal of Public Health*
21. Wegner, Cabrera-Gil, Tanguy (2024) "How Much Should We Sequence? An Analysis of the Swiss SARS-CoV-2 Surveillance Effort" *Microbiology spectrum*
22. Neves, Walther, Martin-Campos (2023) "The Swiss Pathogen Surveillance Platform-Towards a Nation-Wide One Health Data Exchange Platform for Bacterial, Viral and Fungal Genomics and Associated Metadata" *Microbial Genomics*
23. Russell, Broderick, Franklin (2006) "Transmission Dynamics and Prospective Environmental Sampling of Adenovirus in a Military Recruit Setting" *Journal of infectious diseases*
24. Cui, Xia, Jiang (2024) "Prevalence and Genetic Diversity of Human Rhinovirus Among Patients With Acute Respiratory Infections in China, 2012-2021" *Journal of Medical Virology*
25. Eden, Sikazwe, Xie (2022) "Off-Season RSV Epidemics In Australia After Easing of COVID-19 Restrictions" *Nature Communications*
26. Foley, Yeoh, Minney-Smith (2021) "The Interseasonal Resurgence of Respiratory Syncytial Virus in Australian Children Following the Reduction of Coronavirus Disease 2019-related Public Health Measures" *Clinical Infectious Diseases*
27. Cooksey, Morales, Linde (2020) "Severe Acute Respiratory Syndrome Coronavirus 2 and Respiratory Virus Sentinel Surveillance" *Emerging Infectious Diseases*
28. Meslé, Sinnathamby, Mook et al. (2023) "Seasonal and Inter-Seasonal RSV Activity in the European Region During the COVID-19 Pandemic From Autumn 2020 to Summer 2022" *Influenza and Other Respiratory Viruses*
29. Munro, House (2024) "Cycles of Susceptibility: Immunity Debt Explains Altered Infectious Disease Dynamics Post-Pandemic" *Clinical Infectious Diseases*
30. (2025) "Intensified circulation of respiratory syncytial virus (RSV) and associated hospital burden in the EU/EEA"
31. (2025) "RSV-NET Interactive Dashboard | CDC"
32. (2025) "FluView Interactive -National, Regional, and State Level Outpatient Illness and Viral Surveillance"
33. (2025) "Seasonal influenza -Annual Epidemiological Report for 2021-2022"
34. (2022) "Seasonal influenza -Annual Epidemiological Report for"
35. (2025) "Preliminary Flu Burden Estimates, 2022-23 Season"
36. Reicherz, Xu, Abu-Raya (2022) "Waning Immunity Against Respiratory Syncytial Virus During the Coronavirus Disease 2019 Pandemic" *Journal of infectious diseases*
37. Garcia-Maurino, Brenes-Chacón, Halabi et al. (2024) "Trends in Age and Disease Severity in Children Hospitalized With RSV Infection before and During the COVID-19 Pandemic" *JAMA pediatrics*
38. Bourdeau, Vadlamudi, Bastien (2023) "Pediatric RSV-Associated Hospitalizations before and During the COVID-19 Pandemic" *JAMA Network Open*
39. (2024) "Acute respiratory infections in the EU/EEA: epidemiological update and current public health recommendationswinter"
40. (2025) "Influenza Vaccine"
41. (2024) "Bundesamt Für Gesundheit (BAG) Jahresbericht Zu Den Respiratorischen Viren 2023/2024" *BAG-Bulletin*
42. Alsallakh, Adeloye, Vasileiou (2024) "Impact of the COVID-19 Pandemic on Influenza Hospital Admissions and Deaths In Wales: Descriptive National Time Series Analysis" *JMIR Public Health and Surveillance*
43. Lin, Liang, Guan et al. (2024) "Hospitalized Children With Influenza A Before, During and After COVID-19 Pandemic: A Retrospective Cohort Study" *BMC Pediatrics*
44. Chen, Gilbert, Dubé (2024) "Adult Influenza Vaccination Coverage Before, During and After the COVID-19 Pandemic in Canada" *BMC Public Health*
45. (2021) "FluVaxView: Flu Vaccination Coverage, United States, 2020-21 Influenza Season"
46. (2023) "FluVaxView: Flu Vaccination Coverage, United States, 2022-23 Influenza Season"
47. (2024) "Survey report on national seasonal influenza vaccination recommendations and coverage rates in EU/EEA countries"
48. (2024) "GA: National Center for Immunization and Respiratory Diseases, CDC"
49. (2021) "Estimates of National Immunization Coverage: DTP3"
50. Zumstein, Heininger (2021) "Decline of Pertussis In Hospitalised Children Following the Introduction of Immunisation in Pregnancy-Results From a Nationwide, Prospective Surveillance Study, 2013-2020" *Swiss Medical Weekly*
51. Böhmer, Hellenbrand, Matysiak-Klose et al. (2013) "Pertussis-Impfquoten Bei Erwachsenen in Deutschland"
52. Poethko-Müller, Schmitz (2013) "Impfstatus Von Erwachsenen in Deutschland: Ergebnisse Der Studie Zur Gesundheit Erwachsener in Deutschland (DEGS1)" *Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz*
53. (2024) "Rapid risk assessment: Increase of pertussis cases in the EU/EEA"
54. Wang, Zhang, Liu (2025) "Resurgence of Pertussis: Epidemiological Trends, Contributing Factors, Challenges, and Recommendations for Vaccination and Surveillance" *Human Vaccines & Immunotherapeutics*
55. Nordholm, Emborg, Nørgaard (2023) "Pertussis Epidemic in Denmark" *Eurosurveillance*
56. Tessier, Campbell, Ribeiro (2022) "Impact of the COVID-19 Pandemic on Bordetella Pertussis Infections in England" *BMC Public Health*
57. Dungu, Holm, Hartling (2024) "Mycoplasma Pneumoniae Incidence, Phenotype, and Severity in Children and Adolescents in Denmark Before, During, and After the COVID-19 Pandemic: A Nationwide Multicentre Population-Based Cohort Study" *Lancet Regional Health Europe*
58. (2025) "COVID-19 Vaccine Administration and Coverage, Children and Adults, by Jurisdiction, United States. accessed Aug 13"
59. Heiniger, Schliek, Moser et al. (2022) "Differences in COVID-19 Vaccination Uptake in the First 12 Months of Vaccine Availability In Switzerland-A Prospective Cohort Study" *Swiss Medical Weekly*
60. Collie, Champion, Moultrie et al. (2022) "Effectiveness of BNT162b2 Vaccine Against Omicron Variant in South Africa" *New England Journal of Medicine*
61. Johnson, Amin, Ali (2021) "COVID-19 Incidence and Death Rates Among Unvaccinated and Fully Vaccinated Adults With and Without Booster Doses During Periods of Delta and Omicron Variant emergence-25 US Jurisdictions" *MMWR. Morbidity and Mortality Weekly Report*
62. Hammitt, Dagan, Yuan (2022) "Nirsevimab for Prevention of Rsv In Healthy Late-Preterm and Term Infants" *New England Journal of Medicine*
63. Papi, Ison, Langley (2023) "Respiratory Syncytial Virus Prefusion F Protein Vaccine in Older Adults" *New England Journal of Medicine*
64. Simões, Pahud, Madhi (2025) "Efficacy, Safety, and Immunogenicity of the Matisse (Maternal Immunization Study for Safety and Efficacy) Maternal Respiratory Syncytial Virus Prefusion f Protein Vaccine Trial" *Obstetrics and Gynecology*
65. Van Gordon, Mccarthy, Proctor et al. (2021) "Evaluating COVID-19 Reporting Data in the Context of Testing Strategies Across 31 Low-And Middle-Income Countries" *International Journal of Infectious Diseases* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12687644&blobtype=pdf | # Preliminary assessment of drug repurposing against virusassociated primary effusion lymphoma
Daniela Plaza, Gabriella Chefitz, Emily Mckiernan, Sophie Sandler, Clara Levrero, Kang Kim, Emma Germano, Sarah Braner, Lina Ariyan, Jj Miranda
## Abstract
Drug repurposing uses medicine with a given indication to treat a different disease. Primary effusion lymphoma (PEL), a cancer driven by coinfection with the Kaposi sarcoma-associated herpesvirus and the Epstein-Barr virus, lacks an effective treatment. We optimized a rapid, informative, and educational protocol for quantitatively evaluating repurposed small molecules against PEL. The approach tests measurements of PEL cell growth and viability in culture against known inhibitory concentrations. We demonstrate proper quantitative interpretation of the data by using ethacrynic acid, quizartinib, and darapladib as examples. We hope that this practical experimental pipeline will spread awareness of the potential of drug repurposing, especially for diseases like PEL that have unmet clinical needs. KEYWORDS drug repositioning, inhibitory concentration 50, primary effusion lymphoma, cell proliferation, herpesvirus 8, human, herpesvirus 4, human M icrobial infections cause ~15% of all cancer cases (1). Primary effusion lymphoma (PEL) is driven by coinfection with the Kaposi sarcoma-associated herpesvirus (KSHV) and the Epstein-Barr virus (EBV) (2). Lack of an effective standard treatment yields poor clinical outcomes with median overall survival of less than one year after initiation of chemotherapy (2). New drugs are needed.Repurposing, also referred to as repositioning, is the strategy in which a drug with a given indication is used to treat another disease (3). Given the long process of drug development, repurposing efficiently harnesses existing progress and applies that effort toward additional goals. Repurposing has yielded success against both cancer (4) and infectious disease (5), suggesting particular potential against virus-associated malignan cies. As part of our antiviral cancer research (6-9), we optimized a rapid, informative, and educational protocol for quantitatively evaluating repurposed small molecules against PEL. The framing of the research question and practical experimental protocol were iterated over years of undergraduate independent projects and senior theses. Our approach may be adapted as a laboratory class exercise or course-based undergraduate research experience module for cell biology, microbiology, or cancer biology.
PROCEDURE OverviewPlanning and execution of the project aim to achieve learning objectives at different levels of complexity. Upon completion, students should be able to (i) understand the mechanism of action by which a chosen small molecule inhibits the targeted protein and regulated biological pathway, (ii) apply a known or approximate IC 50 value in selecting appropriate doses for a functional assay, (iii) evaluate the efficacy of a small molecule to reduce cell growth and viability in a cell culture experiment, and (iv) skillfully and accurately perform cell culture maintenance and treatment of immortalized human cells.
We recommend first doing background reading on KSHV-associated primary effusion lymphoma, then choosing a protein, drug, and dose to test, then acquiring cell culture skills, and finally executing the experimental design chosen. Two days in a week, one for cell treatment and one for measurement of growth and viability, would generate results for one biological replicate. A sufficient number of replicates for statistical evaluation can be completed in approximately 1 month. The repetition also allows for mastery of technical cell culture skills. For a laboratory course, a student can test one protein and one small molecule. For a semester-long course-based undergraduate research experience, a student can test multiple combinations, potentially exploring different small molecules that target the same protein or different proteins in the same biological pathway.
## Cell culture
BC-1, a PEL cell line derived from a patient specimen (10), was chosen for practical reasons. This line tends to retain a high viability of >95% and grow with fast doubling times of ~1 day. Other PEL lines behave less reliably. Moreover, BC-1 cells contain both KSHV and EBV, the coinfection found in most clinical cases (2). Many cell lines harbor only KSHV. BC-1 is obtainable from a nonprofit biorepository. We maintained BC-1 cells (ATCC, Manassas, VA) in RPMI-1640 media containing 25 mM HEPES, 2 g/L sodium bicarbonate, and 10% fetal bovine serum (11). Cultures were grown in 5% carbon dioxide at 37°C. Cells were generally passaged at densities between 0.2 × 10 6 and 2.0 × 10 6 cells/mL.
## Hypothesis generation
Assessing small molecule efficacy requires a clear hypothesis with quantitative experimental predictions (Fig. 1A). Logically, identifying a protein suspected to participate in virus or lymphoma biology and then finding an inhibitor for that target is an intuitively stepwise way to frame the question. A publication measuring the IC 50 for biochemical inhibition must then be found. Enzyme assays are preferred because the efficacious concentrations reveal direct inhibition of the protein and not possible pleiotropic off-target effects sometimes observed in cellular assays. This key step is frequently omitted, leading to incorrect interpretations. We recommend choosing two concentrations separated by an order of magnitude near the IC 50 . Testing only one dose is easily inconclusive. If a small molecule is effective within a 10-fold difference from the IC 50 , we consider preliminary repurposing successful and likely occurring through the hypothesized mechanism of action.
## Small molecule treatment
Treatments were performed by adding the small molecule to log-phase BC-1 cells. Compounds were diluted from a 1,000× stock solution. All experiments included vehicle controls of equal volume. Ethacrynic acid (MedChemExpress, Monmouth Junction, NJ) and quizartinib (Tocris, Minneapolis, MN) were dissolved in dimethyl sulfoxide; darapla dib (Selleck Chemicals, Houston, TX) was dissolved in ethanol. Cells were seeded at a density of 0.2 or 0.3 × 10 6 cells/mL. Five milliliters in a T-25 flask is sufficient. Cells were grown for 3 days. Experiments may also be performed for 2 days. Control vehicle-treated cultures often overgrew after 4 days. Many treatments yielded smaller and difficult-tomeasure effects after only 1 day.
Small molecule effects were assessed by measuring cell growth. After incubation, an aliquot was mixed in a 1:1 ratio with 0.4% trypan blue. Cell growth was quantified with a Countess II FL Automated Cell Counter (Thermo Fisher Scientific, Waltham, MA). Alterna tively, manual counting with a hemocytometer is less expensive but slower. When cleaning the hemocytometer and coverslip, properly inactivate potentially infectious virus with 70% ethanol. The seeding cell density was subtracted from the observed cell density to define new cell growth.
We also measured viability using trypan blue exclusion (12). While resazurin (13) and tetrazolium (14) reagents are touted as measuring viability, those assays do not distinguish between fewer cell divisions and cell death. We again used an automated cell counter. If a hemocytometer is substituted, establish a clear threshold of blue staining to reduce observer-dependent variability in classification of dead cells.
Both cell count and viability measurements were normalized relative to the vehi cle-treated control. A complete experiment consisted of 3-4 independent biological replicates. To demonstrate robustness, we also required that replicates come from two distinct batches of cells thawed at different times. Statistically significant changes were determined using a paired t-test.
## Safety issues
Biosafety level 2 facilities are required for work with both KSHV and EBV (15). Although BC-1 cells do not release infectious KSHV, a small proportion releases infectious EBV (16). Following the risk assessment described in the American Society for Microbiology Guidelines for Biosafety in Teaching Laboratories (17), we strongly advise requiring usage of the recommended biological safety cabinet. Consequently, gloves and laboratory coats can be used as personal protective equipment without face protection because of the engineering hazard control provided by the tissue culture hood. Given the routes of infection for KSHV and EBV, we also strongly advise requiring completion of If student access to necessary biosafety level 2 facilities is not available, we suggest, but have not yet rigorously tested, the use of A20 (18) cells. This commercially available immortalized mouse line is also a B-cell lymphoma and may be cultured at biosafety level 1.
## CONCLUSION
We discuss our results with specific examples to illustrate the pipeline. Ethacrynic acid is approved to treat edema and inhibits p-class glutathione S-transferases with an IC 50 of 3 μM (19). Treatment of BC-1 cells yields a statistically significant change in cell growth near this dose (Fig. 1B). We conclude that repurposing warrants further study. Quizartinib is approved to treat acute myeloid leukemia and inhibits the FLT3 receptor tyrosine kinase with an IC 50 of ~1 nM (20). Treatment of BC-1 cells does not yield statistically significant changes in either cell growth or viability near this dose (Fig. 1C). We conclude that repurposing was not successful. We disagree with interpreting growth inhibition by μM doses as supporting an FLT3-dependent mechanism of action in PEL (21). We also present our example where testing relatively high doses leads to growth inhibition likely discordant with the molecular hypothesis. Darapladib is undergoing clinical trials for treating atherosclerosis and inhibits lipoprotein-associated phospholi pase A2 with a subnanomolar IC 50 (22). Treatment with 1 μM darapladib does not yield statistically significant changes in either cell growth or viability (Fig. 1D). Although 10 μM darapladib reduces both BC-1 cell growth and viability (Fig. 1D), this dose is much higher than the IC 50 . We conclude that another mechanism is likely responsible. Indeed, lipoprotein-associated phospholipase A2 knockout produces no or minimal reduction of viability in other cancer cell lines (23).
We can advise on facilitating this research for undergraduates. In our experience, students do exceptionally well at learning cell culture sterile technique and subse quent statistical data analysis. The main challenge is actually achieving the first two learning objectives centered on hypothesis generation. Choosing a protein and small molecule pair to study is daunting with a seemingly unending number of possibilities. We recommend choosing proteins and small molecules from empirically determined lists. Proteins whose knockdown reduces proliferation in the same cell culture model of primary effusion lymphoma (24) would be promising targets. Small molecules that inhibit the chosen proteins in another biological context (25) would be strong candidates for repurposing. Details regarding the mechanism of action and IC 50 may subsequently be uncovered through PubMed searches (26).
Repurposing offers educational opportunities and provides a launching point for future experiments. Background primary literature reading on the mechanism of action of small molecules can lead to deeper dives on the importance of these pathways in cancer biology. Once preliminary data on cell proliferation are obtained, future directions open up. Other small molecules with the same mechanism can be tested. Other PEL models, either in cell culture or mice, could be examined. The molecular mechanism can be validated.
Every combination of protein target, small molecule inhibitor, and treatment dose is a novel experiment. Allowing students to craft the experimental design elevates the inquiry-based learning level of the investigation (27). Given the rarity of opportunities for intellectual ownership of research at the undergraduate level, increased inquiry presents an avenue for heightened engagement (28). We hope that this practical experimental pipeline will spread awareness of the potential of drug repurposing, especially for diseases like PEL that have unmet clinical needs. The advancement of repurposing can be achieved in both the research laboratory and the teaching classroom.
## References
1. Plummer, De Martel, Vignat et al. (2016) "Global burden of cancers attributable to infections in 2012: a synthetic analysis" *Lancet Glob Health*
3. Cesarman, Chadburn, Rubinstein (2022) "KSHV/HHV8-mediated hematologic diseases" *Blood*
4. Pushpakom, Iorio, Eyers et al. (2018) "Drug repurposing: progress, challenges and recommendations" *Nat Rev Drug Discov*
5. Zhang, Zhou, Xie et al. (2020) "Overcoming cancer therapeutic bottleneck by drug repurposing" *Signal Transduct Target Ther*
6. Mercorelli, Palù, Loregian (2018) "Drug repurposing for viral infectious diseases: how far are we?" *Trends Microbiol*
7. He, Miranda (2018) "JQ1 reduces Epstein-Barr virus-associated lymphoproliferative disease in mice without sustained oncogene repression" *Leuk Lymphoma*
8. Fernandez, Miranda (2016) "Bendamustine reactivates latent Epstein-Barr virus" *Leuk Lymphoma*
9. Keck, Moquin, He et al. (2017) "Bromodomain and extraterminal inhibitors block the Epstein-Barr virus lytic cycle at two distinct steps" *J Biol Chem*
10. De Garayo, Liu, Rondeau et al. (2021) "Rationally repurposed nitroxoline inhibits preclinical models of Epstein-Barr virus-associated lymphoproliferation" *J Antibiot (Tokyo)*
11. Cesarman, Moore, Rao et al. (1995) "In vitro establishment and characterization of two acquired immunodeficiency syndrome-related lymphoma cell lines (BC-1 and BC-2) containing Kaposi's sarcoma-associated herpesvirus-like (KSHV) DNA sequences" *Blood*
12. Moquin, Thomas, Whalen et al. (2018) "The Epstein-Barr virus episome maneuvers between nuclear chromatin compartments during reactivation" *J Virol*
13. Strober (2015) "Trypan blue exclusion test of cell viability" *Curr Protoc Immunol*
14. Anoopkumar-Dukie, Carey, Conere et al. (2005) "Resazurin assay of radiation response in cultured cells" *Br J Radiol*
15. Mosmann (1983) "Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays" *J Immunol Methods*
16. Meechan, Potts (2020) "Biosafety in microbiological and biomedical laboratories"
17. Miller, Heston, Grogan et al. (1997) "Selective switch between latency and lytic replication of Kaposi's sarcoma herpesvirus and Epstein-Barr virus in dually infected body cavity lymphoma cells" *J Virol*
18. Byrd, Emmert, Maxwell et al. (2019) "Guidelines for biosafety in teaching laboratories"
19. Kim, Kanellopoulos-Langevin, Merwin et al. (1979) "Establishment and characterization of BALB/c lymphoma lines with B cell properties" *J Immunol*
20. Ploemen, Van Ommen, Van Bladeren (1990) "Inhibition of rat and human glutathione S-transferase isoenzymes by ethacrynic acid and its glutathione conjugate" *Biochem Pharmacol*
21. Zarrinkar, Gunawardane, Cramer et al. (2009) "AC220 is a uniquely Tips and Tools Journal of Microbiology and Biology Education December"
22. "potent and selective inhibitor of FLT3 for the treatment of acute myeloid leukemia (AML)" *Blood*
23. Wu, Wang, Yarchoan (2024) "Pacritinib inhibits proliferation of primary effusion lymphoma cells and production of viral interleukin-6 induced cytokines" *Sci Rep*
24. Blackie, Bloomer, Brown et al. (2003) "The identification of clinical candidate SB-480848: a potent inhibitor of lipoprotein-associated phospholipase A 2" *Bioorg Med Chem Lett*
25. Oh, Jang, Lee et al. (2023) "The lipoprotein-associated phospholipase A2 inhibitor darapladib sensitises cancer cells to ferroptosis by remodelling lipid metabolism" *Nat Commun*
26. Manzano, Patil, Waldrop et al. (2018) "Gene essentiality landscape and druggable oncogenic dependencies in herpesviral primary effusion lymphoma" *Nat Commun*
27. Knox, Wilson, Klinger et al. (2024) "DrugBank 6.0: the DrugBank Knowledgebase for 2024" *Nucleic Acids Res*
28. Sayers, Beck, Bolton et al. (2025) "Database resources of the National Center for Biotechnology Information in 2025" *Nucleic Acids Res*
29. Bell, Smetana, Binns (2005) "Simplifying inquiry instruction: assessing the inquiry level of classroom activities" *Sci Teach*
30. Gunn, Mccauslin, Staiger et al. (2013) "Inquiry-based learning: inflammation as a model to teach molecular techniques for assessing gene expression" *J Microbiol Biol Educ* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645940&blobtype=pdf | # Protein-S-nitrosylation of adenovirus-5 E1A and human papillomavirus 16 E7 limits their ability to inhibit STING activity
Justin Cox, Eain Murphy
## Abstract
All viruses that establish successful infections express proteins that inhibit innate anti-viral pathways such as the stimulator of interferon genes (STING) pathway. In response, cells have evolved mechanisms to limit viruses by modifying these viral proteins via post-translational modifications (PTMs). One potent PTM, protein-S-nitro sylation, inhibits the ability of human cytomegalovirus (HCMV) to undermine the establishment of an anti-viral state. The direct nitrosylation of HCMV tegument protein pp71 at a central cysteine within its pRB binding domain reduces pp71's ability to limit the activity of STING. Two different proteins encoded by unrelated DNA viruses, adenovirus (AdV) E1A and human papillomavirus (HPV) E7, also contain a pRB bind ing domain and inhibit STING like pp71. Herein, we report that E1A and E7 are both protein-S-nitrosylated like pp71. Stable cell lines expressing a WT, or mutants in which the predicted modified cysteine was changed to the closely related serine amino acid, thus blocking protein-S-nitrosylation, revealed that E1A and E7 are both protein-Snitrosylated. Furthermore, induction of the STING pathway promoted IFN-β1 transcript production and the phosphorylation of IRF3, which was limited in E1A and E7 stable cell lines. Mutant stable cell lines exhibited a stronger inhibition of IFN-β1 transcription and reduced IRF3 phosphorylation, suggesting that the PTM limits WT viral protein inhibition of STING. Furthermore, both E1a and E7 can complement the replication of a HCMV that lacks pp71 during times of STING activation. These observations support a model in which protein-S-nitrosylation of viral virulence factors may function as an anti-viral mechanism in DNA virus infections. IMPORTANCE DNA viruses, such as HCMV, AdV, and HPV, have the capacity to cause significant disease. Infection with AdV can cause severe lower respiratory and liver disease in children, and HPV infection is persistent and is a causative agent of cancer. Thus, these infections can be a severe health risk. Host cells have adapted innate responses like protein S-nitrosylation to limit viral replication. Our previous work reported that direct nitrosylation of two HCMV viral proteins, pp65 and pp71, limits their ability to undermine host anti-viral responses. Herein, we investigated whether protein-S-nitrosylation of AdV and HPV proteins inhibits their functions, suggesting that this PTM is an anti-viral mechanism. This may provide insight into the development of broad anti-viral therapeutics for persistent viral infections.
infections in children, resulting in severe liver disease (3). A vaccine for AdV does exist, but it is not currently available to the general public and is provided only to those in the military (4). HPV is a common DNA virus that causes benign warts on the skin. However, specific genotypes of this virus can cause warts in the throat and uterus, which can develop into cancer (5)(6)(7)(8). Although a vaccine is now available to the public, current guidelines suggest that vaccination should not be administered to individuals after 26 years of age, as one may have already been exposed to HPV. Thus, there is a need for the development of broadly neutralizing therapeutics for the control of persistent viral infections, such as AdV or HPV, that are opportunistic, targeting immune-naïve individuals.
The innate immunity of an infected host is the first line of defense for viral infections. Innate immune pathways are responsible for the recognition of molecular signals such as cytoplasmic double-stranded DNA (dsDNA) indicative of a viral intrusion. An important pathway in the recognition of cytoplasmic dsDNA is the stimulator of interferon genes (STING) pathway (9-12). The STING pathway is activated by cyclic guanosine adenosine synthase (cGAS) binding to dsDNA within the cytosol, which then covalently links ATP and GTP to form 2′3′-cGAMP, the activator of STING, resulting in the induction of interferon transcription (13,14). Initiation of this pathway establishes a potent anti-viral response that inhibits the replication of multiple viral infections (15)(16)(17). However, viruses have evolved countermeasure mechanisms to limit the activity of this pathway early in infection, thus undermining the establishment of an anti-viral state within an infected cell. AdV-encoded E1A and HPV-encoded E7 directly inhibit STING by limiting its translocation to the cytosol, resulting in reduced activated TBK1 and a diminished interferon response (18,19). Independent of this function, these proteins also inhibit the retinoblastoma protein (pRB), which is important for the G1 to S phase switch, thus inducing resting cells to initiate the cell cycle, thereby allowing for the replication of the viruses (20)(21)(22)(23)(24). E1A and E7 are both essential proteins for AdV and HPV, and in their absence, viral replication is severely attenuated (25)(26)(27).
Although E1A and E7 function to diminish interferon production within an infected host, the host has evolved multiple countermeasures to regulate the activity of these proteins. AdV E1A levels are regulated both transcriptionally and post-translationally, as a host protein, TRIM-35, ubiquitinates E1A at lysine K48, inducing its proteasomal degradation (28). Additionally, it was observed that in TRIM-35 overexpressing cells, there is a reduction in E1A transcription (28). E7 is regulated by PTMs that lead to reduced protein levels, including ubiquitination by Cullin-1 and Skp2 containing E3 ligase, which leads to its proteasomal degradation (29). Recently, Huang et al. reported that the activation of STING induces the phosphorylation of E7 by TBK1, leading to the degradation of the viral protein (30). PTMs like phosphorylation and ubiquitination regulate the function of viral proteins to establish an anti-viral state. However, there has been increasing evidence that an additional PTM, termed protein-S-nitrosylation, also functions as an anti-viral mechanism. In our previous work, we reported that protein-S-nitrosylation of human cytomegalovirus (HCMV) tegument proteins pp71 and pp65 serves as a potent anti-viral mechanism by blocking the biological functions of these key viral proteins (31,32). pp71 is nitrosylated within its pRB binding domain of "LxCxE, " and when nitrosylation of this site is inhibited, the virus replicates with increased efficiency when compared with wild-type HCMV during conditions where STING is activated (32). Interestingly, pp71 has structural homology to both E1A and E7, in that they all possess this unique pRB binding domain (33). Coupled with the fact that each of these proteins is reported to inhibit STING (18,34), we hypothesized that E1A and E7 are protein-S-nitro sylated like pp71 and protein-S-nitrosylation of E1A and E7 limits their ability to inhibit STING, thereby suggesting this PTM may function as a broad-based anti-viral mechanism.
Herein, we report that E7 and E1A are both protein-S-nitrosylated in the absence of infection. We also observed that simulation of the cGAS/STING pathway by 2′3′-cGAMP in E7 and E1A expressing cell lines results in a marked reduction in IFN-β1 transcripts and IRF3 phosphorylation, which is further inhibited in mutant E1A and E7 stable cell lines in which protein-S-nitrosylation is inhibited, suggesting that nitrosylation of these proteins limits their normal antagonizing abilities. Furthermore, E1A and E7 isoforms that cannot be protein-S-nitrosylated can both complement a pp71-deficient HCMV in the presence of activated STING. In sum, these data suggest that protein-S-nitrosylation may serve as a broad-based anti-viral mechanism for distinct viruses and thus provides a novel therapeutic target to limit viral replication.
## RESULTS
## AdV-5 E1A and HPV-16 E7 are protein-S-nitrosylated independent of viral infection
Stable fibroblast cell lines were generated by transduction of NuFF-1 cells with lentivi ruses that express either WT or serine mutant isoforms of E1A and E7, termed E1A-C124S and E7-C24S. The central cysteine within the pRB binding domain of both viral proteins was mutated to the structurally related serine amino acid (Fig. 1A). Following drug selection, protein expression of E1A and E7 WT as well as the E1A-C124S and E7-C24S mutant isoforms was confirmed by western blotting. We observed similar amounts of protein expression within each of the E1A and E7 stable cell lines (Fig. 1B). To determine if E1A and E7 are protein-S-nitrosylated in NuFF-1 cells, cell lysates from E1A and E7 WT as well as the E1A-C124S and E7-C24S cells were subjected to a biotin switch assay followed by purification with avidin beads. The biotin switch assay changes all nitrosylated sites on a protein to biotin moieties. In the first part of the reaction, nitric oxide groups on cystines are unaltered, whereas the free thiol group on non-nitrosyla ted cystines is blocked by the addition of a chemical moiety. Following this, the nitric oxide groups on the protein-S-nitrosylated cystines are removed and replaced by biotin. Following the biotin switch and avidin purification, lysates were then immunoblotted using antibodies specific for E1A or E7. Actin is naturally nitrosylated at multiple sites and thus serves as a suitable loading control for western blotting after a biotin switch assay on cell lysates (35)(36)(37). We observed bands for WT E1A and E7 from the biotin-enriched lysates, suggesting that these proteins are S-nitrosylated (Fig. 1C andD). Furthermore, we observe slight bands in the mutant groups, but this is to be expected, as there is likely more than one cystine within these proteins that is modified by protein-S-nitro sylation in E1A or E7. The difference in band intensity in the western blots suggests that E7 is nitrosylated at multiple cysteine residues or may suggest that E7 may be endogenously biotinylated within the cell. It is highly probable that there are multiple protein-S-nitrosylated cystines in the proteins being studied (as we previously observed multiple protein-S-nitrosylation sites in both HCMV pp71 and pp65). To this point, E1a has 10 total cystines (including the pRB binding domain), and E7 has 5. Thus, we elected to only mutate the pRB binding domain cystines to serines as the role of this domain in STING inhibition has already been enumerated.
## IFN-β1 transcripts are reduced in AdV-5 E1A-C124S and HPV-16 E7-C24S stable cell lines
We next determined if interferon induction of WT and mutant stable cell lines demon strates reduced STING activation compared with parental NuFF-1 cells following 2′3′-cGAMP treatment by quantifying the levels of IFN-β1 transcripts. To identify the optimal concentration of 2′3′-cGAMP treatment, fibroblasts were treated with 20 µM 2′3′-cGAMP, and protein lysates were collected over an 8 h time course post-treatment (Fig. 2A). Western blot analysis revealed optimal phosphorylation of IRF3 at 4 h post-transfection of 2′3′-cGAMP. To identify the optimal concentration of 2′3′-cGAMP, NuFF-1 fibroblasts were treated with increasing concentrations of the drug, and cell lysates were collected 4 h post-treatment. We observed a stepwise increase in the phosphorylation of IRF3, and EC50 analysis determined that the optimal concentration of 2′3′-cGAMP was 10 µM (Fig. 2B andC). These empirically determined time points and the concentration of 10 µM of 2′3′-cGAMP were used for all the subsequent experiments involving STING activation.
Parental and stable cell lines expressing E1A and E7 were treated with 10 µM of 2′3′-cGAMP for 24 h, and then, RNA was isolated following treatment. As expected, parental cell lines produced IFN-β1 transcripts upon induction of the STING pathway. However, we observed a significant reduction in IFN-β1 transcript accumulation in WT expressing stable cell lines, and importantly, a further reduction of IFN-β1 transcripts for both E1A-C124S and E7-C24S (Fig. 2D). This suggests that E1A-C124S and E7-C24S cell lines that lack the capacity to be protein-S-nitrosylated within the pRB binding domain inhibit STING with better efficiency than their wild-type counterparts, thus blocking an anti-viral state.
## Full-Length Text
## The phosphorylation of IRF3 in AdV-5 E1A-C124S and HPV-16 E7-C24S stable cell lines is inhibited
Next, we wanted to determine the biological impact on the STING pathway of E1A and E7 when they cannot be nitrosylated at their pRB binding domain. To determine if stable cell lines expressing WT or E1A-C124S or E7-C24S induce phosphorylated IRF3 with different efficiencies upon STING activation, parental fibroblast cell lines and stable cell lines were treated with 10 µM of 2′3′-cGAMP for 4 h and then lysed to determine the phosphorylation levels of IRF3. As expected, parental cell lines had an increase in the phosphorylation status of IRF3 following treatment. As predicted, there was a decrease in the phosphorylation of IRF3 in WT E1A or E7 stable cell lines that were drug-treated when compared with non-transduced parental cell lines. Importantly, we observed a further reduction in the phosphorylation of IRF3 in the stable cell lines expressing the isoforms of E1A and E7 that cannot be protein-S-nitrosylated within their pRB-binding domains (Fig. 3A andB). Densitometry confirmed that WT and the serine mutant stable cell lines had a reduction in IRF3 phosphorylation (Fig. 3C). These data suggest that blocking nitrosylation of E1A and E7 provides the proteins with increased ability to antagonize STING in fibroblasts.
## AdV-5 E1A-C124S and HPV-16 E7-C24S stable cells replicate WT HCMV with higher efficiency during STING induction
Our data thus far suggest that the nitrosylation of E1A and E7 impacts their ability to limit STING. Next, we tested if blocking nitrosylation of E1A and E7 had a biological impact on the replication of WT HCMV when the potent anti-viral STING pathway is induced. Parental fibroblasts, E1A or E7 stable cells, were treated with 10 μM of 2′3′-cGAMP, followed by infection at an MOI of 1 with WT HCMV. Viral supernatants were collected 6 days post-infection (dpi) and measured by TCid50. As expected, HCMV replicated in parental cell lines and all stable cell lines in the absence of 2′3′-cGAMP. We observed a significant decrease in the production of HCMV infectious virions in the 2′3′-cGAMPtreated parental NuFF-1 cells, indicating that 2′3′-cGAMP limits HCMV replication as expected. WT HCMV titers were also reduced in E1A and E7 WT stable cell lines, following 2′3′-cGAMP treatment. Importantly, we observed that E1A-C124S and E7-C24S WT HCMV titers were not reduced following 2′3′-cGAMP treatment (Fig. 4). These findings suggest that WT E1A and E7, which are liable to protein-S-nitrosylation by the host cell, are not resistant to the activation of STING, whereas nitrosylation-deficient E1A-C124S and E7-C24S can still allow for efficient viral replication after STING induction. This result supports a model in which blocking the nitrosylation of E1A and E7 allows the proteins to inhibit STING with better efficiency, thus restoring WT HCMV titers.
## AdV-5 E1A-C124S and HPV-16 E7-C24S stable cell lines complement the replication of HCMV that lacks expression of pp71 during STING induction
Our data suggest that E1A-C124S and E7-C24S allow WT HCMV to replicate after the induction of a pro-viral state. Next, we tested if E1A-C124S and E7-C24S complement the STING inhibition functions of pp71 by monitoring replication of a HCMV variant that lacks the capacity to express pp71 (32). Parental and stable cell lines were pre-treated with 10 µM of 2′3′-cGAMP and then infected with a pp71-deficient HCMV virus (Δpp71) at an MOI of 1. After 6 dpi, cell lysates were collected, and DNA was isolated from these cells to measure viral genomes by qPCR. We observed that HCMV genomes in parental NuFF-1 cells were reduced following 2′3′-cGAMP. The WT E1A and E7 stable cell lines were also reduced following 2′3′-cGAMP treatment. The percentage of HCMV genomes in all untreated groups was similar. Importantly, the genomes of Δpp71 in infection of E1A-C124S and E7-C24S stable cell lines were not significantly reduced after inducing STING (Fig. 5A). This suggests that E1A-C124S and E7-C24S stable cell lines are resistant to 2′3′-cGAMP treatment and may complement pp71's function in the inhibition of STING, at least in viral DNA replication. To determine if this blocking of the STING pathway results in complementation of Δpp71 HCMV infectious virus production, parental, E1A, and E7 stable cell lines were pre-treated with 10 µM of 2′3′-cGAMP for 6 h and then infected with Δpp71 HCMV. Cell-associated virus was collected at 6 dpi and then quantified by TCid50. Cell-associated virus was collected to ensure accurate monitoring of viral replication, as HCMV typically spreads cell to cell. Untreated parental and WT E1A and E7 stable cell lines replicated Δpp71 HCMV to similar levels. However, following 2′3′-cGAMP treatment, the parental, E1A, and E7 WT stable cell lines demon strated a significant reduction in HCMV titers, suggesting that E1A and E7 WT stable cell lines, which are still suitable substrates for protein-S-nitrosylation, are still limited by the induction of STING. Importantly, there was no significant reduction of Δpp71 HCMV titers following 2′3′-cGAMP treatment in E1A-C124S and E7-C24S stable cell lines, suggesting that they allowed Δpp71 HCMV replication to levels similar to their untreated counter parts (Fig. 5B). These data suggest that both E1A-C124S and E7-C24S can complement pp71's ability to inhibit STING. In sum, we observed diminished STING induction of an anti-viral response in E1A-C124S and E7-C24S, suggesting that blocking nitrosylation of E1A and E7 allows the proteins to antagonize STING with better efficiency. These findings also highlight that E1A-C124S and E7-C24S can complement pp71's ability to limit STING activity as a Δpp71 HCMV replicated to equal levels in treated and untreated stable cell lines.
## DISCUSSION
Our previous work identified multiple HCMV proteins that were post-translationally modified by protein-S-nitrosylation and characterized a novel mechanism of host regulation in which protein-S-nitrosylation limits the pro-viral functions of pp65 and pp71 during HCMV infection of fibroblasts (31,32). This led us to question whether protein-S-nitrosylation of viral proteins is an anti-viral mechanism in DNA viruses that impacts virulence factors. HCMV pp71 contains a consensus pRB binding domain, LxCxE/D, that is found in two other proteins critical for efficient viral replication, E1A and E7 (18,33). These domains are highly similar-the only known protein S-nitrosylation motif identified in GAPDH, I/LxCxxE/D (38). Furthermore, each of these proteins, in the context of their respective infections, inhibits the activity of STING. This convergence led us to hypothesize that E1A and/or E7 may be protein-S-nitrosylated in a similar fashion as pp71, and this protein modification would impact their biological functions. To this end, stable cell lines expressing WT E1A or E7 revealed that both proteins are protein-Snitrosylated (Fig. 1C andD). This PTM did not require viral infection, suggesting that to a certain level, fibroblasts induce the modification in the absence of a stress-induced event. It should be noted that the complete protein-S-nitrosylation status of E1A and E7 remains unknown. Future mass spectrometry analysis of the two proteins may reveal multiple nitrosylation sites on E1A and E7, as we observed multiple sites on pp71 to be S-nitrosylated (32), and our biotin switch assay suggests that there is more than one nitrosylation site on the proteins (Fig. 1C andD). It may be the case that blocking all sites on E1A or E7 further exaggerates their antagonizing ability of STING. However, we did not observe this effect in pp71, as blocking nitrosylation at the pRB binding domain was sufficient in improving its ability to antagonize STING (32). We did not have the capacity to test whether altering the central cystine to the closely related serine within the pRB binding domain of E1A and E7 impacts their ability to interact with pRB. It remains an attractive possibility that protein-S-nitrosylation is also a regulator of this biological function of these two proteins. Protein-S-nitrosylation of E1A and E7 resulted in a biological outcome that suggests this PTM is inhibitory to the viral proteins. Once STING is activated, it will translocate from the rough endoplasmic reticulum to the Golgi, where it interacts with TBK1 (11,39). This ultimately leads to the activation of transcription factors like IRF3, leading to interferon and inhibitory cytokine production (10,11). Secreted IFN-β1 can then stimulate both the infected cell and adjacent cells, resulting in an anti-viral state through IRF9 and STAT1 activation (40,41). In fact, in our model system, IFN-β1 transcripts were induced in fibroblasts following 2′3′-cGAMP treatment as expected. However, the levels of IFN-β1 were induced to lower levels in stable cell lines that expressed the viral proteins. Importantly, we observed a significantly lower induction of IFN-β1 transcripts in the serine mutant stable cell lines (Fig. 2), indicating that these proteins may antagonize STING with better efficiency in support of a model in which protein-S-nitrosylation of viral proteins is inhibitory to their functions.
Following 2′3′-cGAMP treatment in parental cells, we observed a strong increase in the phosphorylation status of IRF3, which was less pronounced in the WT E1A and E7 stable cells, as these proteins are reported to limit STING induction. E1A-C124S and E7-C24S stable cells had further reduced levels of the phosphorylation of IRF3 compared with WT stable cell lines (Fig. 3A andB). Importantly, there was little difference in total IRF3 protein levels, suggesting that the reduction in phosphorylation of IRF3 is from the reduced activity of STING in the serine mutant cells and not a reduction in total IRF3. This observed reduction in phosphorylation of IRF3 suggests that both E1A and E7, in the absence of additional viral proteins, are sufficient for inhibiting the STING pathway. Importantly, variants of these proteins that are resistant to protein-S-nitrosyla tion within their pRB binding domains demonstrated a more potent inhibition of the STING pathway. Based on these findings, a consistent model emerges in which protein-Snitrosylation supplements the host's anti-viral response.
In our model system, STING is antagonized with higher efficiency in cells that express E1A and E7 isoforms that are limited in their ability to be protein-S-nitrosylated. This next led us to determine if this translates to an advantage for the replicating virus. In the absence of 2′3′-cGAMP, we observed that WT E1A and E7 stable cell lines produced infectious virions similar to levels found in parental NuFF-1 cells. These titers were reduced in parental and WT E1A and E7 cells following STING induction. However, in E1A-C214S and E7-C24S cell lines, we observed that HCMV viral titers were not reduced following 2′3′-cGAMP treatment (Fig. 4). This suggests that blocking nitrosylation of E1A and E7 at their pRB-binding domain provides the proteins an advantage to antagonize STING. These findings support our model that nitrosylation can limit E1A and E7's ability to limit STING activity. It is possible that using infectious HPV and AdV infection may yield different results, but our current results show that expressions of E1A-C124S and E7-C24S alone provide a distinct advantage to the replication of an unrelated virus, HCMV, in an activated anti-viral state.
HCMV tegument protein pp71 possesses several functional similarities to E1A and E7. E1A and E7 both inhibit STING and contain a pRB binding domain similar in structure to pp71 (18,42). To determine if E1A and E7 can functionally complement the lack of pp71 during HCMV infection, NuFF-1 cells were infected with a Δpp71 HCMV. Δpp71 HCMV has a growth defect in the infection of NuFF-1 cells due to its inability to activate viral IE gene expression (42). Parental and E1A and E7 WT expressing NuFF-1 cells infected with Δpp71 HCMV in the absence of STING induction replicated viral genomes to similar levels, and as expected, genomes for the parental cells and stable cell lines were reduced following 2′3′-cGAMP treatment. However, the genomes in the E1A-C124S and E7-C24S stable cell lines were not reduced following 2′3′-cGAMP treatment, suggesting that they can diminish the effects of STING activation (Fig. 5A). Following 2′3′-cGAMP treatment, Δpp71 HCMV titers were similar in the E1A-C124S and E7-C24S stable cell lines com pared with their untreated counterparts (Fig. 5B). These data suggest that E1A and E7, which are still liable to protein-S-nitrosylation, are not able to fully restore the biological functions of pp71, but E1A-C124S and E7-C24S stable cell lines still complement pp71's ability to limit STING. Our findings highlight that blocking protein-S-nitrosylation of E1A and E7 effectively attenuates the proteins' ability to limit STING activity. These results support a model in which blocking nitrosylation of E1A and E7 provides the proteins an advantage in antagonizing STING, but these proteins cannot fully complement pp71 in HCMV infection, as the wild-type E1A and E7 expressing cell lines still resulted in a significant growth defect of Δpp71 HCMV when compared with WT HCMV (Fig. 4 and5). Our data suggest that the key biological functions of pp71 for HCMV infection may be required to support HCMV replication. HCMV pp71 is a multifunctional protein involved in the inhibition of hDAXX, pRB, and STING and stimulates the MIEP in HCMV infection (34,(42)(43)(44)(45) and thus is a critical tegument protein in HCMV infection. We believe that E7 and E1A do not have the capacity to compensate for all the functions of HCMV infection, resulting in the same replication kinetics in parental cell lines. There is evidence that HPV E7 can complement the function of HCMV UL97, but no other evidence suggests that it can compensate for the loss of pp71, further highlighting the importance of our findings that E7-C24S can complement pp71's ability to limit STING (46). In sum, our data suggest that E7 and E1A do not fully complement all pp71's functions but still antagonize STING induction in the absence of pp71. We observe that serine mutant E1A and E7 are resistant to STING activation, highlighting that protein-S-nitrosylation is an anti-viral mechanism in multiple DNA virus infections.
In sum, our work identifies two additional proteins that are regulated by protein-Snitrosylation, suggesting that this modification functions as an anti-viral mechanism. The identification of protein-S-nitrosylation of E1A and E7 represents a novel discovery and may provide evidence for exploiting applications in medicine for increasing protein-Snitrosylation. By blocking protein-S-nitrosylation of E1A and E7 at their pRB-binding domain, we identified that both proteins antagonize STING with higher efficiency, leading to a reduction in IFN-β1 transcripts. Our data further suggest that this may be from the lack of phosphorylation of IRF3 in serine mutant stable cell lines, allow ing the proteins to antagonize STING with higher efficiency (Fig. 6). Importantly, we observed that E1A-C124S and E7-C24S cell lines are resistant to 2′3′-cGAMP stimulation in WT and Δpp71 HCMV infection. Identifying additional factors regulated by protein-Snitrosylation suggests that this may be a conserved and broadly neutralizing anti-viral mechanism that has evolved to limit infection by distinct viral family members. This is supported by the recent report that inhibition of nitric oxide, a key substrate for protein-S-nitrosylation, allows for higher HCMV titers and that treatment with nitric oxide donors attenuates HCMV replication, in support of our model (47). This modification may be exploited therapeutically to control drug-resistant or -persistent viral infections. Identifying additional viral factors that are protein-S-nitrosylated will further provide evidence that protein-S-nitrosylation regulates multiple viral infections and has broadly neutralizing properties.
## MATERIALS AND METHODS
## Cell culture
Newborn fetal fibroblasts (NuFF-1) and HEK-293T cells were maintained in Dulbecco's modified essential medium (Cleveland Clinic) supplemented with 10% fetal bovine serum (FBS) (Millipore-Sigma), 1% penicillin-streptomycin solution (Cleveland Clinic), and 2 mM L-Glutamine (Cleveland Clinic). Cells were maintained at 5% carbon dioxide at 37°C. Cells infected with HCMV for the expansion of the virus were supplemented with complete medium that contains 10% newborn calf serum (Gibco). All cells were split with 0.5% Trypsin/EDTA (Cleveland Clinic) at a dilution of 1:2, and fibroblast passages were recorded to limit cellular senescence and limited to no more than 30 passages for experiments.
## Virus propagation
HCMV genomes were contained within bacterial artificial chromosomes (BAC) in SW105 bacteria. For the isolation of HCMV BACs, bacteria were streaked onto Luria Broth (LB) agar-chloramphenicol plates and grown at 32°C for 24 h. The bacteria were then cultured overnight in LB-chloramphenicol broth, and the next day, BAC DNA was isolated and transfected with electroporation into MRC5 cells that were at 50% confluency (32). The next day, the medium was changed and incubated until the plate demonstrated 100% cytopathic effect (CPE). The virus was then isolated by collecting the supernatant and cells, sonicating, and centrifuging the virus at 72,000 × g for 1.5 h at 18°C in a 20% sorbitol cushion with an SW-28 rotor in a Beckman-Coulter ultracentrifuge. Virus was titered on fibroblasts by TCid50 to measure PFU/mL.
## Lentiviral production and transduction
HEK-293T cells were seeded at 80% confluency 1 day before transfection. To form DNA:Lipofectamine complexes, 5.6 μg of E1A-WT (VB230516-1368xug), E1A-C124S (VB230516-1367tkb), E7-WT (VB240412-1392fuy), and E7-C24S (VB240413-1083mrr) pLV-eGFP-T2A-Puro (Vector Builder), 7.1 μg of p-CMV-VSV-G, and 14.2 μg of pDR 8.91 were incubated with 60 μL of Lipofectamine 2000 (Thermo-Fisher) in Opti-MEM for 30 min at room temperature. Protein expression of the viral proteins is driven by a CMV promoter that is contained between the lentivirus LTRs. DNA:Lipofectamine complexes were overlayed onto HEK-293T cells overnight, and the following day, media were changed to 10% NCS supplemented with 2 mM L-Glutamine. The supernatant was collected at 48 and 72 h post-medijm change, filtered with a 0.45 μm filter (Sigma), and the media were overlayed on NuFF-1 cells at 70% confluency. NuFF-1 cells were then selected with 700 ng/mL of puromycin (Invitrogen) and allowed to grow for two doublings. Protein expression was confirmed with western blotting.
## Measuring viral titers
Human cytomegalovirus was titered by the TCid50 assay. NuFF-1 cells were seeded to 90% confluency in a 96-well plate 24 h prior to infection. Viral supernatant was serially diluted 1:10, and viral plaques were counted 14 dpi by visualization of mCherry expression.
## Western blotting
Protein was isolated by scraping cells with 100 μL of Pierce RIPA lysis and extraction buffer (25 mM Tris-HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS; Thermo Fisher Scientific). Lysates were incubated on ice for 30 min and sonicated 2× for 15 s to lyse cells. Lysates were quantitated by Bradford assay, and 30 µg of protein was separated with an 8% SDS-PAGE. Following separation, the protein was transferred to a nitrocellulose membrane and blocked with 5% BSA for 1 h at room temperature. The membranes were probed with anti-E1A (1:200, 8A8, Santa-Cruz), anti-E7 (1:200, ED17, Santa-Cruz), anti-actin (1:10,000, 13E5, Cell Signaling), anti-pIRF3 (1:500, 4D4G, Cell-Signaling), and anti-IRF3 (1:2000, Thermo Fisher), overnight at 4 degrees Celsius and then washed with TBS-Tween (0.1%) the following day, three times 5 min each wash. Secondary antibodies anti-mouse or rabbit-HRP (1:2000, Cell-Signaling) were probed for 1 h at room temperature and then washed an additional three times for 5 min each wash. All images were processed and imaged on Bio-Rad's Chemidoc with chemiluminescence or by fluorescence.
## Biotin capture of nitrosylated proteins
In total, 100 µg of protein from cell lysates was subjected to a biotin-switch assay following the manufacturer's protocol (Cayman Chemical Company). Biotinyla ted proteins were affinity purified with streptavidin-coated beads M-280 Dynabeads (Thermo-Fisher Scientific) overnight and washed six times with cold PBS using Life Technologies Dynamag-2. Beads were boiled in 2× Laemmli and loaded onto an 8% SDS-PAGE. The protein was separated and probed for antibodies specific to E1A or E7.
## Real-time quantitative PCR
The cells were stimulated with 10 μM of 2′3′-cGAMP, and then 24 h later, lysed with TRIZol. Samples were incubated for 5 min at room temperature, and then, 1′3′-BCP was added to each sample; in addition, the samples were shaken for 15 s. The samples were then incubated for 10 min at room temperature and then centrifuged for 12 min at 12,000 × g. Following centrifugation, the top aqueous layer was extracted, and then, RNA was precipitated with 100% isopropanol and incubated for 5 min at room temperature. Following incubation, the samples were then centrifuged for 10 min at 12,000 × g. Then, the supernatant was aspirated, and RNA pellets were washed with 75% ethanol and centrifuged for 5 min at 7,500 × g. Ethanol was aspirated following centrifugation, and RNA samples were dried for 10 min. The samples were then resuspended in water and incubated at 37 degrees Celsius for 5 min to increase the solubility of RNA. All sample concentrations were measured on a Thermo Fisher nanodrop.
## Statistics
All experiments in this study were analyzed using an unpaired Student t-test. The results were considered significant if the calculated P-value was <0.05, indicated by a 95% CI. All data were graphed and analyzed utilizing the program GraphPad Prism.
## References
1. Hierholzer (1992) "Adenoviruses in the immunocompromised host" *Clin Microbiol Rev*
2. Lion (2014) "Adenovirus infections in immunocompetent and immunocompromised patients" *Clin Microbiol Rev*
3. Zhong, Yi, Xiang et al. (2022) "Hepatitis of unknown etiology in children: current evidence and association" *World J Clin Cases*
4. Gray, Erdman (2018) "Adenovirus vaccines"
5. Husain, Neyaz (2017) "Human papillomavirus associated head and neck squamous cell carcinoma: controversies and new concepts" *J Oral Biol Craniofac Res*
6. Sabatini, Chiocca (2020) "Human papillomavirus as a driver of head and neck cancers" *Br J Cancer*
7. Fakhry, Gillison (2006) "Clinical implications of human papillomavi rus in head and neck cancers" *J Clin Oncol*
8. Hennessey, Westra, Califano (2009) "Human papillomavirus and head and neck squamous cell carcinoma: recent evidence and clinical implications" *J Dent Res*
9. Ishikawa, Barber (2008) "STING is an endoplasmic reticulum adaptor that facilitates innate immune signalling" *Nature*
10. Sun, Li, Chen et al. (2009) "ERIS, an endoplasmic reticulum IFN stimulator, activates innate immune signaling through dimerization" *Proc Natl Acad Sci*
11. Zhong, Yang, Li et al. (2008) "The adaptor protein MITA links virus-sensing receptors to IRF3 transcription factor activation" *Immunity*
12. Jin, Waterman, Jonscher et al. (2008) "MPYS, a novel membrane tetraspanner, is associated with major histocompatibility complex class II and mediates transduction of apoptotic signals" *Mol Cell Biol*
13. Diner, Burdette, Wilson et al. (2013) "The innate immune DNA sensor cGAS produces a noncanonical cyclic dinucleotide that activates human STING" *Cell Rep*
14. (2025) *Full-Length Text Journal of Virology*
15. Ablasser, Goldeck, Cavlar et al. (2013) "CGAS produces a 2'-5'-linked cyclic dinucleotide second messenger that activates STING" *Nature*
16. Lam, Stein, Falck-Pedersen (2014) "Adenovirus detection by the cGAS/STING/TBK1 DNA sensing cascade" *J Virol*
17. Reinert, Lopušná, Winther et al. (2016) "Sensing of HSV-1 by the cGAS-STING pathway in microglia orchestrates antiviral defence in the CNS" *Nat Commun*
18. Bianco, Mohr (2017) "Restriction of human cytomegalovirus replication by ISG15, a host effector regulated by cGAS-STING doublestranded-DNA sensing" *J Virol*
19. Lau, Gray, Brunette et al. (2015) "DNA tumor virus oncogenes antagonize the cGAS-STING DNA-sensing pathway" *Science*
20. Lou, Huang, Zhou et al. (2023) "DNA virus oncoprotein HPV18 E7 selectively antagonizes cGAS-STINGtriggered innate immune activation" *J Med Virol*
21. Wang, Draetta, Moran (1991) "E1A induces phosphorylation of the retinoblastoma protein independently of direct physical association between the E1A and retinoblastoma products" *Mol Cell Biol*
22. Liu, Marmorstein (2007) "Structure of the retinoblastoma protein bound to adenovirus E1A reveals the molecular basis for viral oncopro tein inactivation of a tumor suppressor" *Genes Dev*
23. Dyson, Howley, Münger et al. (1989) "The human papilloma virus-16 E7 oncoprotein is able to bind to the retinoblastoma gene product" *Science*
24. Boyer, Wazer, Band (1996) "E7 protein of human papilloma virus-16 induces degradation of retinoblastoma protein through the ubiquitin-proteasome pathway" *Cancer Res*
25. Jones, Münger (1997) "Analysis of the p53-mediated G1 growth arrest pathway in cells expressing the human papillomavirus type 16 E7 oncoprotein" *J Virol*
26. Ferreira, Mcmillan, Idris (2022) "Genetic deletion of HPV E7 oncogene effectively regresses HPV driven oral squamous carcinoma tumour growth" *Biomed Pharmacother*
27. Saha, Parks (2017) "Human adenovirus type 5 vectors deleted of early region 1 (E1) undergo limited expression of early replicative E2 proteins and DNA replication in non-permissive cells" *PLoS One*
28. Sauthoff, Pipiya, Heitner et al. (2004) "Impact of E1a modifications on tumor-selective adenoviral replication and toxicity" *Mol Ther*
29. Sun, Zhang, Zhang et al. (2023) "Inhibition of human adenovirus replication by TRIM35-mediated degradation of E1A" *J Virol*
30. Oh, Kalinina, Wang et al. (2004) "The papillomavirus E7 oncoprotein is ubiquitinated by UbcH7 and Cullin 1and Skp2-containing E3 ligase" *J Virol*
31. Huang, Huo, Xiao et al. (2024) "Activating STING/TBK1 suppresses tumor growth via degrading HPV16/18 E7 oncoproteins in cervical cancer" *Cell Death Differ*
32. Cox, Nukui, Murphy (2025) "Protein-S-nitrosylation of human cytomegalovirus pp65 reduces its ability to undermine cGAS" *J Virol*
33. Nukui, Roche, Jia et al. (2020) "Protein S-nitrosylation of human cytomegalovirus pp71 inhibits its ability to limit STING antiviral responses" *J Virol*
34. Kalejta, Bechtel, Shenk (2003) "Human cytomegalovirus pp71 stimulates cell cycle progression by inducing the proteasomedependent degradation of the retinoblastoma family of tumor suppressors" *Mol Cell Biol*
35. Fu, Su, Gao et al. (2017) "Human cytomegalovirus tegument protein UL82 inhibits STING-mediated signaling to evade antiviral immunity" *Cell Host Microbe*
36. Bansbach, Guilford (2016) "Actin nitrosylation and its effect on myosin driven motility" *AIMS Mol Sci*
37. García-Ortiz, Martín-Cofreces, Ibiza et al. (2017) "eNOS S-nitrosylates β-actin on Cys374 and regulates PKC-θ at the immune synapse by impairing actin binding to profilin-1" *PLoS Biol*
38. Lu, Katano, Uta et al. (2011) "Rapid S-nitrosylation of actin by NO-generating donors and in inflammatory pain model mice" *Mol Pain*
39. Jia, Arif, Terenzi et al. (2014) "Targetselective protein S-nitrosylation by sequence motif recognition" *Cell*
40. Wang, Liu, Zhong et al. (2010) "WDR5 is essential for assembly of the VISA-associated signaling complex and virus-triggered IRF3 and NF-κB activation" *Proc Natl Acad Sci*
41. Sadzak, Schiff, Gattermeier et al. (2008) "Recruitment of Stat1 to chromatin is required for interferon-induced serine phosphorylation of Stat1 transactivation domain" *Proc Natl Acad Sci*
42. Nan, Wang, Yang et al. (2018) "IRF9 and unphosphorylated STAT2 cooperate with NF-κB to drive IL6 expression" *Proc Natl Acad Sci*
43. Bresnahan, Shenk (2000) "UL82 virion protein activates expression of immediate early viral genes in human cytomegalovirus-infected cells" *Proc Natl Acad Sci*
44. Saffert, Kalejta (2006) "Inactivating a cellular intrinsic immune defense mediated by Daxx is the mechanism through which the human cytomegalovirus pp71 protein stimulates viral immediate-early gene expression" *J Virol*
45. Hofmann, Sindre, Stamminger (2002) "Functional interaction between the pp71 protein of human cytomegalovirus and the PMLinteracting protein human Daxx" *J Virol*
46. Cantrell, Bresnahan (2005) "Interaction between the human cytomegalovirus UL82 gene product (pp71) and hDaxx regulates immediate-early gene expression and viral replication" *J Virol*
47. Kamil, Hume, Jurak et al. (2009) "Human papillomavirus 16 E7 inactivator of retinoblastoma family proteins complements human cytomegalovirus lacking UL97 protein kinase" *Proc Natl Acad Sci*
48. Mokry, Schumacher, Hogg et al. (2020) "Nitric oxide circumvents virus-mediated metabolic regulation during human cytomegalovirus infection" *mBio* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724315&blobtype=pdf | # Rab27a-mediated exosome secretion facilitates classical swine fever virus release and immune evasion
Ning Li, Yuehan Quan, Bihao Luo, Xinxin Chen, Tao Wang, Zifeng Gong, Liang Zhang, Kangkang Guo
## Abstract
Classical swine fever is a highly contagious disease caused by the classical swine fever virus (CSFV), a member of the Flaviviridae family. Exosomes are extracellular vesicles that mediate intercellular communication by transferring membrane compo nents and nucleic acids between cells. They play a role in the progression of various infectious diseases and help viruses evade immune responses. However, the impact of CSFV on exosome secretion and whether exosomes facilitate CSFV's evasion of the host immune response remains unclear. In this study, we first demonstrated that CSFV infection upregulates Rab27a, a small GTPase that regulates exosome secretion by mediating the trafficking and fusion of multivesicular bodies with the plasma membrane. By overexpressing and silencing Rab27a, we assessed its involvement in CSFV replication and release. We also demonstrated an interaction between the CSFV E0 and E2 proteins and Rab27a. Our findings show that Rab27a facilitates the release of exosomes, which, in turn, enhances CSFV spread. These results suggest that CSFV infection upregulates Rab27a, promoting the exosome-mediated release pathway, thereby contributing to viral spread and immune evasion. IMPORTANCE Classical swine fever, caused by the classical swine fever virus (CSFV) from the Flaviviridae family, is highly contagious. Exosomes, extracellular vesicles that transfer membrane components and nucleic acids, play a role in viral progression and immune evasion. This study shows that CSFV infection upregulates Rab27a, and the CSFV E0 and E2 proteins interact with Rab27a. By manipulating Rab27a expression, we found that Rab27a facilitates exosome release, which enhances CSFV spread. These findings suggest that CSFV exploits Rab27a to promote exosome-mediated release, aiding viral spread and immune evasion. KEYWORDS classical swine fever virus (CSFV), exosome, Rab27a, immune evasion, release C lassical swine fever (CSF) is an acute, highly contagious viral disease affecting pigs, caused by the classical swine fever virus (CSFV). This disease poses a significant threat to the global swine industry. Control strategies for CSF typically involve mandatory vaccination combined with the culling of infected animals. However, despite stringent vaccination measures, CSFV continues to re-emerge in multiple outbreaks (1). CSFV is a single-stranded positive-sense RNA virus, and its genome is divided into three major regions: the uncapped 5' untranslated region (containing an internal ribosome entry site), the open reading frame (ORF), and the 3' untranslated region (which lacks a polyA tail) (2). The ORF encodes a polyprotein that is cleaved into four structural proteins (C, Erns, E1, and E2) and eight non-structural proteins (Npro, p7, NS2, NS3, NS4A, NS4B, NS5A, and NS5B) (3).Rab GTPases are a conserved family of proteins critical for regulating vesicular transport in eukaryotic cells (4). Research has shown that the activation or inhibition
of specific Rab GTPases can significantly affect exosome release, including Rab5 and Rab27 (5,6). Among them, Rab27a plays a pivotal role in regulating the secretion of lysosome-related organelles, such as multivesicular bodies (MVBs) (7). Rab27a undergoes isoprenylation at two conserved cysteine residues at its C-terminus, allowing it to anchor MVBs to the plasma membrane, promoting their docking and fusion with the membrane (8). Inhibition of Rab27a expression impedes MVB docking, leading to a significant reduction in exosome secretion (9).
Various Rab GTPases play crucial roles in the cellular entry of CSFV. For example, CSFV employs the classical clathrin-dependent pathway, which requires Rab5 and Rab7, to infect PK-15 cells (10). Alternatively, the virus can enter porcine alveolar macrophage cells through a caveola-mediated endocytic route, involving Rab5, Rab7, and Rab11 (11). Additionally, Rab14 regulates lipid metabolism and enhances CSFV replication (12). CSFV NS4B also interacts with Rab22a and co-localizes with Rab5 in early endosomes of PK-15 cells, potentially forming a ternary complex that aids in viral entry (13).
Exosomes are small extracellular vesicles (EVs) that facilitate complex intercellular communication, with sizes ranging from 30 to 200 nm (14,15). Exosomes originate from intraluminal vesicles within MVBs and are transported to the plasma membrane, where they fuse and release exosomes into the extracellular space (16). These vesi cles carry intracellular contents, including nucleic acids and proteins, facilitating the intercellular exchange of information and materials (15). Increasing evidence points to a close relationship between viruses and exosomes. For instance, retroviruses use the vesicular budding mechanism to enhance viral budding (17), and hepatitis viruses utilize exosomes for intercellular transmission (18). Additionally, exosomes play a key role in activating innate immunity and antiviral responses, contributing to long-term interactions between viruses and host cells (19)(20)(21).
In this study, we investigated how CSFV regulates Rab27a expression and examined its impact on viral replication through overexpression and small interfering RNA (siRNA) knockdown techniques. By focusing on Rab27a's role in exosome secretion, we aim to uncover how CSFV manipulates host cellular machinery to promote immune evasion and its persistence in the host.
## RESULTS
## CSFV infection enhances Rab27a expression
To investigate the impact of CSFV infection on Rab27a expression, PK-15 cells were either infected with CSFV at a multiplicity of infection (MOI) of 1.0 or mock-infected for 12, 24, and 48 hours. The relative mRNA and protein expression levels of Rab27a were examined. We found that both mRNA (Fig. 1A) and protein levels (Fig. 1B) of Rab27a were significantly elevated at all indicated time points. We further investigated whether the impact of Rab27a expression is associated with infection MOI. Cells were infected with CSFV at an MOI of 0.2, 1, or a high dose of 5. We observed that both mRNA (Fig. 1C) and protein levels (Fig. 1D) of Rab27a were significantly elevated at all indicated MOI.
To investigate the cellular universality of CSFV infection on Rab27a expression, validation experiments were further conducted in 3D4/21 and IPEC cell lines. Both cell lines were treated with CSFV at an MOI of 1.0 or mock controls for 12, 24, and 48 hours. The relative mRNA and protein expression levels of Rab27a were examined. The results demonstrated that in 3D4/21 cells, both mRNA (Fig. 1E) and protein levels (Fig. 1F) of Rab27a were significantly elevated at all indicated time points. Similarly, in IPEC cells, Rab27a mRNA (Fig. 1G) and protein (Fig. 1H) expression were significantly elevated at all time points.
## Rab27a knockdown inhibited CSFV release
To investigate the role of Rab27a in CSFV infection, PK-15 cells were transfected with Rab27a-specific siRNA. Knockdown efficiency was assessed using two different siRNA candidates, Rab27a siRNA1 and Rab27a siRNA2, by quantifying Rab27a mRNA and protein levels at 24 and 48 hours post-transfection. The results indicated that both Rab27a siRNA1 and Rab27a siRNA2 efficiently reduced Rab27a expression at both time points and at both the mRNA (Fig. 2A) and protein levels (Fig. 2B) compared to the negative control siRNA (siNC). However, Rab27a siRNA2 exhibited greater knockdown efficiency and was therefore selected for further experiments in this study. To evaluate whether Rab27a knockdown affects cell viability and proliferation, a CellTiter-Glo assay was performed. The results demonstrated no significant changes in cell viability or proliferation following Rab27a knockdown, indicating that the observed effects on CSFV infection were not due to cytotoxicity (Fig. 2C).
Next, we examined the effect of Rab27a knockdown on CSFV proliferation. PK-15 cells were transfected with siRNA for 24 hours, followed by infection or mock infection with CSFV at an MOI of 1.0 for 24 or 48 hours. The relative expression levels of CSFV E2 mRNA and protein were quantified using qPCR and western blot, respectively, while viral titers in the supernatant were measured using the TCID 50 assay. Compared to control cells transfected with siNC, CSFV proliferation was significantly inhibited in Rab27a-knock down cells at both time points, at both the mRNA (Fig. 2D) and protein levels (Fig. 2E).
Additionally, viral titers in the supernatant were significantly reduced in Rab27a-silenced cells compared to siNC-transfected cells (Fig. 2F). These findings suggest that Rab27a plays a crucial role in CSFV replication.
To further elucidate the role of Rab27a in different stages of the CSFV life cycle, we investigated its effects on viral adsorption, invasion, replication, and release. To assess viral adsorption, cells were incubated with CSFV (MOI 1) at 4°C for 2 hours to allow virus attachment. To examine viral invasion, cells were incubated with CSFV (MOI 1) at 37°C for 2 hours to assess viral entry. To analyze viral replication, cells were incubated with CSFV (MOI 1) at 37°C for 6 hours to evaluate intracellular viral replication. Finally, to assess viral release, cells were incubated with CSFV (MOI 1) at 37°C for 9 hours, and viral levels in the culture supernatant were analyzed by qPCR and indirect immunofluorescence assay (IFA) staining.
The results demonstrated that Rab27a knockdown had minimal impact on viral adsorption, invasion, and replication. However, viral release was significantly inhibited, as evidenced by the decreased detection of CSFV in the supernatant via qPCR (Fig. 2G) and IFA (Fig. 2H). These findings suggest that Rab27a is crucial for efficient CSFV release but does not significantly affect early stages of viral infection.
## Full-Length Text
## Rab27a overexpression promoted CSFV release
To further investigate the role of Rab27a in CSFV infection, we evaluated the effect of Rab27a overexpression on CSFV proliferation. PK-15 cells stably transfected with PGK-Rab27a or the control PGK-Flag plasmid were established. The relative expression of Rab27a was significantly higher at both the mRNA (Fig. 3A) and protein levels (Fig. 3B) in the cell line stably transfected with PGK-Rab27a, compared to the control cell line transfected with PGK-Flag plasmid. Similar to the siRNA knockdown results, overexpres sion of Rab27a did not result in any significant impact on cell proliferation (Fig. 3C). To assess the impact of Rab27a overexpression on CSFV proliferation, PK-15 cells stably transfected with PGK-Rab27a or the control PGK-Flag plasmid were infected or mock infected with CSFV at an MOI of 1.0 for 24 or 48 hours. The relative expression levels of CSFV E2 mRNA and protein were quantified using qPCR and western blot, respectively, while viral titers in the supernatant were measured using the TCID 50 assay. Compared to control cells transfected with PGK-Flag plasmid, CSFV proliferation was significantly increased in Rab27a-overexpressing cells at both time points, at both the mRNA (Fig. 3D) and protein levels (Fig. 3E). Additionally, viral titers in the supernatant were significantly higher in Rab27a-overexpressing cells compared to control cells (Fig. 3F).
We also assessed the impact of Rab27a overexpression on different stages of the CSFV life cycle, using the same methods as the siRNA knockdown study. The results demon strated that Rab27a overexpression had minimal impact on viral adsorption, invasion, and replication. However, viral release was significantly increased, as evidenced by the increased detection of CSFV in the supernatant via qPCR (Fig. 3G) and IFA (Fig. 3H). These findings further corroborate the results from the knockdown study, confirming that Rab27a is crucial for efficient CSFV release but does not significantly affect the early stages of viral infection.
## Interaction of Rab27a with CSFV proteins E0 and E2
To investigate the interaction between Rab27a and CSFV proteins, we performed co-immunoprecipitation (Co-IP) experiments. The results showed that Rab27a interacted with the structural proteins E0 and E2 of CSFV, as it was able to co-precipitate with both of these proteins (Fig. 4A). To further validate this interaction, we co-transfected plasmids expressing pEGFP-E0 or pEGFP-E2 with pDsRed1-Rab27a into 293T cells and analyzed the subcellular localization using confocal laser microscopy. The observations revealed that Rab27a significantly co-localized with E0 and E2 in the cytoplasm, providing additional evidence for the interaction between Rab27a and CSFV structural proteins (Fig. 4B).
To assess the specific role of Rab27a during CSFV infection, PK-15 cells were transfec ted with specific siRNA targeting Rab5, Rab7, and Rab11. Knockdown efficiency was assessed by quantifying targeted Rab mRNA and protein levels at 24 and 48 hours posttransfection. The results confirmed that the siRNAs effectively reduced the expression of the targeted Rab mRNA and proteins (Fig. 4C). Subsequently, we evaluated the impact of silencing these Rab proteins on CSFV release by measuring the ratio of virus titers in the cell supernatant to intracellular virus titers. The results showed that silencing Rab5 and Rab7 slightly increased CSFV release compared to the control group, while silencing Rab11 significantly reduced virus release. However, none of these changes reached statistical significance (Fig. 4D). These results indicate that Rab27A specifically interacts with the CSFV structural proteins E0 and E2 and promotes CSFV release through this interaction, while other Rab family members (Rab5, Rab7, and Rab11) do not significantly affect CSFV release.
## Exosomes were involved in CSFV proliferation
Rab27a, a member of the Rab protein family, is a GTPase associated with the exosome membrane, and it regulates the docking and fusion of the exosome membrane with the plasma membrane, thereby influencing exosome release (16). Evidence suggests that exosomes can facilitate virus transmission between cells (22,23), and Rab27a may regulate CSFV proliferation through the exosome pathway. To explore whether exo somes are involved in regulating CSFV replication, we examined the impact of exosome inhibition on CSFV replication by treating cells with the exosome-specific inhibitor GW4869.
To determine the optimal concentration of GW4869, we first assessed its cytotoxicity on PK-15 cells across concentrations ranging from 2.5 to 40 µM. The results showed no significant cytotoxic effects on PK-15 cell viability within this concentration range (Fig. 5A). We then investigated the effect of different GW4869 concentrations on CSFV proliferation. PK-15 cells were treated with varying concentrations of GW4869 or mock treated as a control for 24 hours, followed by CSFV infection (MOI = 1) for 48 hours. The results indicated that GW4869 inhibited CSFV proliferation in a dose-dependent manner (Fig. 5B through E). Using 40 µM GW4869, we assessed its impact on CSFV proliferation at different time points post-infection. The results showed that GW4869 effectively inhibited CSFV proliferation at both 24 and 48 hours post-infection (Fig. 4F through I). These findings suggest that exosomes play a crucial role in CSFV proliferation.
## CSFV infection promoted exosome release
To investigate whether CSFV affects exosome release, we first examined changes in exosome quantity following CSFV infection. PK-15 cells were inoculated with CSFV, and the cell culture supernatants were collected. Exosomes were isolated from the superna tant by ultracentrifugation, as outlined in the protocol in Fig. 6A, resulting in an ultracen trifugation pellet (UC pellet). Nanoparticle tracking analysis (NTA) revealed that the size of the crude exosome preparation ranged from approximately 30-150 nm (Fig. 6B).
Negative staining with 2% phosphotungstic acid followed by transmission electron microscopy (TEM) revealed cup-shaped lipid bilayer vesicles characteristic of exosomes (Fig. 6C), confirming that uniformly sized and stable exosomes were successfully extracted.
To assess changes in exosome release following CSFV infection, we measured the total protein concentration and content of exosomes using a BCA protein assay and Coomassie Brilliant Blue staining. The results indicated that, compared to the control group, the amount of exosomes obtained from the supernatant of CSFV-infected PK-15 cells was significantly increased (Fig. 6D andE). Western blotting was then performed to quantify the exosomal markers CD81 and Alix, and these results were consistent with the protein quantity data (Fig. 6F). These findings suggest that CSFV promotes exosome secretion in PK-15 cells.
To explore whether CSFV is encapsulated within exosomes, we purified the crude exosomes using immunomagnetic beads (Fig. 6G) and analyzed the exosomal nucleic acids by PCR. CSFV nucleic acid fragments were detected in the exosomes derived from infected cells (Fig. 6H). Western blotting also revealed the presence of the E2 protein in the exosomes (Fig. 6I). Furthermore, immunoelectron microscopy showed the presence of CSFV within the exosomes (Fig. 6J).
## Full-Length Text
## Rab27a enhanced CSFV replication by promoting exosome secretion
To investigate whether Rab27a influences CSFV release via the regulation of exosome secretion, we first assessed the impact of Rab27a interference on exosome release following CSFV infection. The results showed that, after Rab27a interference, the total protein content of exosomes in CSFV-infected cells was significantly reduced (Fig. 7A andB), and the levels of exosomal marker proteins Alix and CD81 were also decreased (Fig. 7C), in line with the total protein results.
Next, we examined the effect of Rab27a overexpression on exosome release following CSFV infection. The results indicated that Rab27a overexpression led to an increase in the total protein content of exosomes in CSFV-infected cells (Fig. 7D andE), and the levels of exosomal marker proteins Alix and CD81 were elevated (Fig. 7F). These findings suggest that Rab27a enhances CSFV release via the exosome pathway.
Neutralizing antibodies at a concentration of 10× ND 50 can render 1 MOI of CSFV particles non-infectious, thus preventing free virus particles from infecting cells. In contrast, exosome-mediated viral transmission does not require direct cell-to-cell contact (24). To further investigate whether CSFV release is promoted through the Rab27a-mediated exosome pathway following infection, we designed a Transwell experiment as shown in Fig. 7G. In this experiment, the upper chamber contained PK-15 cells with Rab27a overexpression, Rab27a knockdown (via siRNA), or control groups (either overexpression or siRNA control), all infected with CSFV (MOI = 1.0) for 48 hours. The lower chamber contained PK-15 cells, with a polycarbonate membrane (0.4 µm pores) separating the upper and lower chambers, creating a co-culture system. Excess E2 CSFV neutralizing antibody (10× ND 50 ) was added to the lower chamber to neutralize viral particles and prevent free virus particles from infecting the cells. CSFV infection levels in PK-15 cells from both chambers were assessed by qPCR and IFA. The results showed that, after Rab27a interference, the CSFV content in the lower chamber PK-15 cells was reduced (Fig. 7H andI), whereas Rab27a overexpression increased the CSFV content in the lower chamber PK-15 cells (Fig. 7J andK). These findings further demonstrate that Rab27a promotes CSFV exosome-dependent release. They also suggest that CSFV infection enhances Rab27a expression, which, in turn, increases exosome release and facilitates CSFV release via the exosome pathway.
## DISCUSSION
Following infection, viruses hijack host cellular mechanisms to complete their replication cycle. Rab GTPases play crucial roles in vesicular transport, and many viruses exploit these proteins for invasion, assembly, and release (25). For example, herpes simplex virus type 1 utilizes Rab27a during morphogenesis and release (26). In our previous study, we found that in addition to free propagation, CSFV can also spread between cells through EVs (24). In this study, we investigated the role of Rab27a in CSFV infection and demon strated that Rab27a expression is upregulated in a variety of CSFV-infected different cells. Functional analysis revealed that Rab27a overexpression enhanced CSFV release, whereas Rab27a knockdown significantly reduced viral release. These findings suggest that Rab27a plays a role in the CSFV lifecycle, particularly in viral egress.
We found that Rab27a directly interacted with the CSFV structural proteins E0 and E2, as evidenced by co-immunoprecipitation and co-localization analysis. These interactions Rab5, Rab7, and Rab11. The results showed that knockdown of these Rab proteins had little effect on CSFV release, and only Rab27a significantly regulated the release of exosome-associated viruses, highlighting the central role of Rab27a in this process.
Rab27a undergoes isoprenylation at conserved cysteine residues in its C-terminal region, facilitating vesicle docking at the plasma membrane (27). Rab27a is also essential for exosome secretion, as its inhibition markedly reduces exosome release (28). Exosomes originate from MVBs and contain various biomolecules, including proteins, lipids, nucleic acids, and viral components. Increasing evidence suggests that viruses utilize exosome-mediated transmission to evade immune responses. Flaviviruses such as dengue virus, West Nile virus, hepatitis C virus (HCV), and Zika virus exploit exo some pathways for intracellular spread (29)(30)(31)(32). Moreover, foot-and-mouth disease virus degrades Rab27a to inhibit exosome secretion and evade immune responses (33). Given these findings, we explored whether Rab27a-mediated exosome pathways contribute to CSFV release. Our results indicate that CSFV infection enhances exosome secretion via Rab27a upregulation, and exosome-associated CSFV particles contribute to viral dissemination.
Importantly, we identified intact CSFV particles within exosomes using immunoelec tron microscopy, a phenomenon previously reported in other viruses such as Japanese encephalitis virus (JEV) (34) and hepatitis B virus (35). The presence of CSFV RNA and proteins in exosomes, confirmed by PCR and western blot analysis, further supports the hypothesis that exosomes may facilitate CSFV transmission.
To assess whether exosome-mediated viral transfer allows CSFV to evade neutralizing antibodies, we conducted a Transwell experiment. Excess neutralizing antibodies were added to the lower chamber, where no virus was directly introduced. Despite this, CSFV infection was detected in the lower chamber cells, suggesting that exosome-enclosed virions may bypass antibody-mediated neutralization. Similar mechanisms have been described for HCV, where exosome-mediated transmission partially protects viral RNA and proteins from immune clearance (36). Likewise, JEV-derived exosomes have been shown to evade neutralizing antibodies both in vitro and in vivo (34), aligning with our findings.
In conclusion, this study demonstrates that Rab27a interacts with E0 and E2 and is involved in CSFV release, and CSFV can be transmitted via exosomes, potentially as a strategy for immune evasion. For the first time, we show that CSFV enhances exosome secretion by upregulating Rab27a, and Rab27a interacts with E0 and E2 proteins, thereby promoting virus transmission through the exosome pathway. These findings provide new insights into the role of exosomes in the pathogenesis of CSFV and highlight Rab27a as a potential target for CSF antiviral strategies.
## MATERIALS AND METHODS
## Cell culture and virus
PK-15 and 3D4/21 cell lines were purchased from the American Type Culture Collection (ATCC), and the IPEC cell line is derived from the Veterinary Public Health Laboratory at Northwest A&F University (Shaanxi, China). PK-15 and IPEC were maintained in Dulbecco's Modified Eagle Medium (DMEM; Gibco, Grand Island, NY, USA) containing 1% penicillin-streptomycin solution (Sigma-Aldrich, St. Louis, MO, USA) and 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA). 3D4/21 was maintained in Roswell Park Memorial Institute 1640 (RPMI-1640; Gibco, Grand Island, NY, USA) containing 1% penicillin-streptomycin solution and 10% inactivated FBS. All cell cultures were incubated at 37°C with 5% CO 2 . The CSFV Shimen strain was purchased from the China Veterinary Drug Inspection Institute (Beijing, China) and propagated in PK-15 cells.
## CSFV infection
All experimental work was conducted under appropriate biosafety level (BSL-3) containment in compliance with institutional and national regulations for handling select agents. Cells were infected with CSFV at the indicated MOI and incubated for 2 hours. The inoculum was then removed and replaced with fresh DMEM or RPMI-1640 containing 2% FBS. Supernatants were collected at 9, 24, and 48 hours post-infection.
For exosome collection, cells were first replenished with DMEM containing 2% FBS at 2 hours post-infection. Subsequently, the medium was replaced with serum-free DMEM and incubated for an additional 24 hours.
## Exosome isolation and purification
Exosomes were isolated by differential centrifugation. The cell culture supernatant was first centrifuged at 300 × g for 10 minutes at 4°C to remove live cells, and the supernatant was collected. The supernatant was then centrifuged at 2,000 × g for 10 minutes to eliminate dead cells, and the supernatant was again collected. The sample was further centrifuged at 10,000 × g for 10 minutes to remove cell debris. The resulting supernatant was filtered through a 0.22 µm filter. Exosomes were finally pelleted by ultracentrifu gation at 100,000 × g for 120 minutes, and the pellet was resuspended in 2 mL of pre-cooled PBS to obtain the purified exosomes.
Exosome purification was then performed using immunomagnetic beads. Fifty microliter of Protein A/G immunomagnetic beads (MCE, catalog number HY-K0202) was washed three times with PBS and incubated with 500 µL of Anti-CD81 antibody (10 µg/mL; Proteintech, catalog number 66866-1-Ig) at room temperature for 2 hours to allow antibody binding. After washing the beads three times with PBS, they were incubated with 1 mL of the ultracentrifuged exosome suspension on a shaker at 4°C overnight. Following incubation, the beads were washed three times with PBS, and 100 µL of recombinant CD81 protein solution (500 µg/mL; Proteintech, catalog number 65195-1-Ig) was added to elute the exosomes. The beads were incubated at 25°C for 30 minutes, and the exosome eluate was collected after allowing the beads to settle on a magnetic stand for 2 minutes. The protein concentration of the isolated exosomes was determined using Coomassie Brilliant Blue staining and a BCA protein assay kit (TargetMol, catalog number C0050).
## Nanoparticle tracking analysis
Briefly, exosome samples were diluted prior to analysis, and the relative concentration was calculated based on the dilution factor. The samples were analyzed using gain adjustment and manual shutter control at a speed of 15 or 30 ms, with shutter speeds between 280 and 560. Data were analyzed using NTA 3.2 software (Malvern Panalytical Ltd, Malvern, Worcestershire, UK), and evaluation was performed using the NanoStar II instrument (Malvern Panalytical Ltd).
## Transmission electron microscopy
To observe the morphology of isolated exosomes, a drop of the exosome sample was placed on a disposable glove, and a copper grid was floated on the droplet for 2 minutes. Excess water was carefully removed from the edge of the grid using filter paper. The sample was then stained by placing the grid onto a droplet of phosphotungstic acid solution for 90 seconds. After air-drying, the copper grid was examined under a TEM at an accelerating voltage of 80 kV to assess the exosome morphology.
For immunoelectron microscopy (immuno-EM), the copper grid with the sample was floated on a fixative solution for 10 minutes, followed by three washes in PBS. To block non-specific binding, the grid was incubated with 50 g/L BSA blocking solution for 30 minutes. The grid was then incubated with a primary antibody (anti-E2) solution for 60 minutes, followed by three PBS washes. The grid was subsequently incubated with a secondary antibody solution for 60 minutes, washed three times with PBS, and stained with phosphotungstic acid. After air-drying, the grid was visualized under a TEM at 80 kV.
## Western blot
The cells or exosome samples were lysed on ice with radio immunoprecipitation assay (RIPA) buffer (MedChem Express, catalog number HY-K0010) supplemented with protease inhibitors. The lysates were centrifuged, and the resulting supernatants were mixed with protein loading buffer, followed by heating at 100°C for 5 minutes to denature the proteins. Protein separation was performed using SDS-PAGE, and the proteins were subsequently transferred to a polyvinylidene fluoride membrane (Merck Millipore, catalog number ISEQ00010).
The membrane was blocked at room temperature for 2 hours using Tris-buffered saline with Tween (TBST) containing 5% non-fat milk. The membrane was incubated overnight at 4°C with primary antibodies: rabbit anti-Rab27a (1:4,000; Proteintech, catalog number 17817-1-AP), rabbit anti-Alix (1:2,000; Proteintech, catalog number 12422-1-AP), and mouse anti-CD81 (1:1,000; Proteintech, catalog number 66866-1-Ig). Following three washes with phosphate buffered saline (PBS), horseradish peroxidase (HRP)-conjugated secondary antibodies (goat anti-mouse IgG, 1:10,000; Immunoway, catalog number RS0001; goat anti-rabbit IgG, 1:10,000; Immunoway, catalog number RS0002) were applied for 2 hours at room temperature. Protein signals were visual ized using an enhanced chemiluminescence (ECL) chemiluminescent detection system (mixing solution A and B in equal volumes), and images were captured using a chemilu minescence imaging system. β-actin was used as a loading control.
## Plasmids and siRNA
RNA was extracted from PK-15 cells using TRIzo reagent (Thermo), and cDNA was synthesized using the RevertAid First Strand cDNA Synthesis Kit (Thermo). The pig Rab27a gene was amplified from the cDNA by PCR and cloned into the Piggybacpgk-3p16-iRFP670-6-IB vector to obtain PGK-Rab27a. The siRNA duplex used in this study was siRab27a. The primers used in this study are listed in Table 1, and the siRNAs used in this study are listed in Table 2.
## Real-time PCR analysis
Real-time PCR was performed to quantify the expression of Rab27a, CD81, Alix, and viral genomic copies. Total RNA was extracted from cells or tissues using TRIzol reagent (Thermo, catalog number AM9738) and reverse-transcribed into cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo, catalog number K1622). SYBR Green Real-Time PCR Master Mix (Takara, catalog number CN830A) was used for amplification, following the manufacturer's protocol. The reaction conditions were 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. Primer sequences are available upon request. Relative RNA expression was normalized to β-actin using the comparative Ct method.
For viral entry analysis, PK-15 cells were incubated with CSFV (MOI = 5) at 4°C for 1 hour to allow viral attachment, followed by incubation at 37°C for 2 hours to facilitate
## Construction of Rab27a-overexpressing PK-15 cell line
PK-15 cells were co-transfected with the PGK-Rab27a or PGK-Flag plasmid as a con trol, along with transposase PB, using the jetPRIME transfection reagent (Polyplus, 101000046). After 48 hours, cells were selected in medium containing Blasticidin S HCl (5 mg/mL; Beyotime, ST018) to establish a stable cell line. PGK-Flag vector, expressing the Flag protein, served as a control.
## CSFV titration
PK-15 monolayers in 96-well plates were inoculated with 100 µL of serially diluted CSFV. After 1 PBS wash, cells were fixed with 4% formaldehyde for 20 min, permeabilized with 0.5% Triton X-100, and blocked with 5% BSA. Cells were incubated overnight at 4°C with primary antibody, followed by a fluorophore-conjugated secondary antibody for 1 hour in the dark. Nuclei were stained with DAPI, and fluorescence was detected using a confocal microscope (Nikon). Viral titers were calculated as TCID 50 per half well.
## Cell viability assay
Cell viability was assessed using the Cell Counting Kit-8 (TargetMol, catalog number C0050). Cells were seeded in 96-well plates and cultured for 12, 24, 36, 48, 60, or 72 hours. After the addition of 10 µL of reagent to the culture medium, the mixture was well mixed and incubated at 37°C for 1 hour. The color of the culture medium turned orange, and the absorbance at 450 nm was measured using a microplate reader (Thermo Fisher Scientific). The cell-free group was used as a blank control. The concentration of GW4869 (MCE, catalog number HY-19363) used in this study was determined based on the results of the cell viability assay.
## Cell proliferation assay
Cell proliferation was assessed using the CellTiter-Glo 3D Cell Viability Assay (Promega, catalog number G9681). Cells were seeded in 96-well plates and cultured for 12, 24, 36, 48, 60, or 72 hours. One hundred microliters of CellTiter-Glo 3D Reagent was added to each well. The plates were then gently shaken for 5 minutes to ensure thorough mixing. The plates were incubated at room temperature for 25 minutes. The chemiluminescence values were measured using a luminescence reader (Thermo Fisher Scientific).
## Transwell co-culture assay
Transwell inserts (polycarbonate membrane, 0.4 µm pore size, 24-well plate; Corning, catalog number 3413) containing PK-15 cells with Rab27a overexpression, Rab27a knockdown (via siRNA), or control groups (overexpression or siRNA control) were infected with CSFV (MOI = 1.0) for 48 hours. Uninfected PK-15 cells were seeded in the bottom wells of a 24-well plate. The Transwell inserts with infected cells were then placed into the corresponding wells to establish the co-culture system. The culture medium was supplemented with a neutralizing antibody (10 × ND 50 ) to block viral replication, and the medium was replaced every 24 hours. The system was incubated at 37°C for a total of 72 hours.
## Coomassie brilliant blue staining
After SDS-PAGE electrophoresis, protein gels were transferred to a container, and 50 mL of deionized water was added. The gel was microwaved for 3 minutes, then shaken for 5 minutes. It was stained with 20 mL of Coomassie Brilliant Blue Fast Staining Solution for 10-30 minutes. After staining, the gel was washed with 100 mL of deionized water and destained by shaking and replacing the water every 15 minutes for 30-120 minutes until the protein bands were visible.
## Co-IP experiments
PK-15s transfected with 4 µg of plasmids of 12 proteins of CSFV were harvested at 48 hours with western blot and IP lysis buffer (Beyotime, catalog number P0013). Followed by centrifugation for 30 min at 4°C, a quarter of the supernatant was subjec ted to input assays. The rest were incubated with anti-Flag antibody (1:7,000; Abways, catalog number AB0008) overnight at 4°C, which had been centrifuged and rinsed with TBS. Followed by washing with TBS and boiling in 5× SDS sample buffer, the protein samples were subjected to a western blot with Rabbit anti-Rab27a monoclonal antibody.
## Confocal microscopy
293T cells were seeded onto cell culture coverslips in 24-well plates and incubated for 12 hours at 37°C in a 5% CO 2 incubator. The plasmids of pEGFP and pDsRed were co-transfected and cultured for 48 hours. Cells were then washed three times with PBS and fixed with 4% paraformaldehyde at room temperature. After three washes with PBS, cells were incubated with DAPI (Biosharp, catalog number BL105A) at 37°C for 10 min and washed three times with PBS. Finally, the images were captured using a Leica TCS SP8 laser scanning confocal microscope (LSM510 META, Zeiss, Germany).
## Statistical analysis
Statistical analyses were performed using GraphPad Prism 8 (GraphPad Software). A Student's t-test or two-way analysis of variance (ANOVA) was used to analyze differences between groups. P-values < 0.05 were considered statistically significant.
## References
1. Moennig (2000) "Introduction to classical swine fever: virus, disease and control policy" *Vet Microbiol*
2. Lefkowitz, Dempsey, Hendrickson et al. (2018) "Virus taxonomy: the database of the International Committee on Taxonomy of Viruses (ICTV)" *Nucleic Acids Res*
3. Thiel, Stark, Weiland et al. (1991) "Hog cholera virus: molecular composition of virions from a pestivirus" *J Virol*
4. Chavrier, Parton, Hauri et al. (1990) "Localization of low molecular weight GTP binding proteins to exocytic and endocytic compartments" *Cell*
5. Pfeffer (2001) "Rab GTPases: specifying and deciphering organelle identity and function" *Trends Cell Biol*
6. Butler, Singh, Marnin et al. (2024) "The role of Rab27 in tick extracellular vesicle biogenesis and pathogen infection" *Parasit Vectors*
7. Goishi, Mizuno, Nakanishi et al. (2004) "Involvement of Rab27 in antigen-induced histamine release from rat basophilic leukemia 2H3 cells" *Biochem Biophys Res Commun*
8. Gomes, Ali, Ramalho et al. (2003) "Membrane targeting of Rab GTPases is influenced by the prenylation motif" *Mol Biol Cell*
9. Bobrie, Colombo, Krumeich et al. (2012) "Diverse subpopulations of vesicles secreted by different intracellular mecha nisms are present in exosome preparations obtained by differential ultracentrifugation" *J Extracell Vesicles*
10. 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*
11. Zhang, Liu, Xiao et al. (2018) "Rab5, Rab7, and Rab11 are required for caveola-dependent endocytosis of classical swine fever virus in porcine alveolar macrophages" *J Virol*
12. Liu, Bai, Liu et al. (2023) "The small GTPase Rab14 regulates the trafficking of ceramide from endoplasmic reticulum to golgi apparatus and facilitates classical swine fever virus assembly" *J Virol*
13. Wang, Liu, Sun et al. (2022) "Rab22a cooperates with Rab5 and NS4B in classical swine fever virus entry process" *Vet Microbiol*
14. Simons, Raposo (2009) "Exosomes--vesicular carriers for intercellular communication" *Curr Opin Cell Biol*
15. Ju, Bai, Ren et al. (2021) "The role of exosome and the ESCRT pathway on enveloped virus infection" *Int J Mol Sci*
16. Lin, Anderson, Rahnama et al. (2020) "Exosomes in disease and regeneration: biological functions, diagnostics, and beneficial effects" *Am J Physiol Heart Circ Physiol*
17. Meng, Ip, Abbink et al. (2020) "ESCRT-II functions by linking to ESCRT-I in human immunodeficiency virus-1 budding" *Cell Microbiol*
18. Feng, Hensley, Mcknight et al. (2013) "A pathogenic picornavirus acquires an envelope by hijacking cellular membranes" *Nature*
19. Hong, Truong, Vu et al. (2022) "Exosomes from H5N1 avian influenza virus-infected chickens regulate antiviral immune responses of chicken immune cells" *Dev Comp Immunol*
20. Gong, Kong, Ren et al. (2020) "Exosomemediated apoptosis pathway during WSSV infection in crustacean mud crab" *PLoS Pathog*
21. Kouwaki, Okamoto, Tsukamoto et al. (2017) "Extracellular vesicles deliver host and virus RNA and regulate innate immune response" *Int J Mol Sci*
22. Khabir, Blanchet, Angelo et al. (2024) "Exosomes as conduits: facilitating hepatitis B virusindependent hepatitis D virus transmission and propagation in hepatocytes" *Viruses*
23. Martínez-Rojas, Monroy-Martínez, Ruiz-Ordaz (2025) "Role of extracellular vesicles in the pathogenesis of mosquito-borne flaviviruses that impact public health" *J Biomed Sci*
24. Wang, Zhang, Liang et al. (2022) "Extracellular vesicles originating from autophagy mediate an antibody-resistant spread of classical swine fever virus in cell culture" *Autophagy*
25. (2025) *Full-Length Text Journal of Virology*
26. Hervé, Bourmeyster (2018) "Rab GTPases, master controllers of eukaryotic trafficking" *Small GTPases*
27. Bello-Morales, Crespillo, Fraile-Ramos et al. (2012) "Role of the small GTPase Rab27a during herpes simplex virus infection of oligodendrocytic cells" *BMC Microbiol*
28. Pereira-Leal, Seabra (2000) "The mammalian Rab family of small GTPases: definition of family and subfamily sequence motifs suggests a mechanism for functional specificity in the Ras superfamily" *J Mol Biol*
29. Ostrowski, Carmo, Krumeich et al. (2000) "Rab27a and Rab27b control different steps of the exosome secretion pathway" *Nat Cell Biol*
30. Vora, Zhou, Londono-Renteria et al. (2018) "Arthropod EVs mediate dengue virus transmission through interaction with a tetraspanin domain containing glycoprotein Tsp29Fb" *Proc Natl Acad Sci*
31. Osuna-Ramos, Jesús-González, Palacios-Rápalo et al. (2020) "The regulation of flavivirus infection by hijacking exosomemediated cell-cell communication: new insights on virus-host interactions"
32. Cosset, Dreux (2014) "HCV transmission by hepatic exosomes establishes a productive infection" *J Hepatol*
33. Martínez-Rojas, Monroy-Martínez, Moreno et al. (2024) "Zika virus-infected monocyte exosomes mediate cell-to-cell viral transmission" *Cells*
34. Xu, Xu, Shi et al. (2020) "Foot-and-mouth disease virus degrades Rab27a to suppress the exosome-mediated antiviral immune response" *Vet Microbiol*
35. Xiong, Yang, Zhu et al. (2025) "Extracellular vesicles promote the infection and pathogenicity of Japanese encephalitis virus" *J Extracell Vesicles*
36. Wu, Glitscher, Tonnemacher et al. (2023) "Presence of intact hepatitis B virions in exosomes" *Cell Mol Gastroenterol Hepatol*
37. Gu, Wu, Fang et al. (2020) "Exosomes cloak the virion to transmit Enterovirus 71 non-lytically" *Virulence*
38. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724296&blobtype=pdf | # Herpesvirus-host interactions in neurological diseases: the immunogenetic role of HLA-E
Marianne Graninger, Elisabeth Puchhammer-Stöckl, Hannes Vietzen
## Abstract
Human herpesviruses (HHVs) comprise nine pathogenic members, including herpes simplex virus 1, herpes simplex virus 2, varicella-zoster virus, Epstein-Barr virus, human cytomegalovirus, human herpesvirus 6A/B, human herpesvirus 7, and Kaposi's sarcoma-associated herpesvirus. Clinical manifestations of HHV infection can range from asymptomatic cases to a broad spectrum of neurological complications, spanning from acute conditions such as encephalitis to chronic disorders including Alzheimer's disease and multiple sclerosis. By establishing latency and undergoing repeated reactivation, HHVs maintain lifelong interactions with the human immune system and shape host immune responses, exerting considerable impact on nervous system homeostasis. Individual susceptibility to, and outcomes of, HHV-associated neurological disorders depend on multiple factors, including the infecting HHV strain and host genetics. Recent evidence highlights the pivotal role of the human leukocyte antigen E (HLA-E) pathway-a non-classical major histocompatibility complex class I molecule with immunomodulatory functions-in regulating virus-host interactions. Since some HHVs manipulate HLA-E to evade immune recognition, individual variability in this axis may influence neurological outcomes. In this review, we summarize and discuss current knowledge of the role of HLA-E in herpesvirus-associated neurological diseases.
## HUMAN HERPESVIRUSES AND NEUROLOGICAL DISEASES
T he human herpesvirus (HHV) family comprises nine pathogenic members: her pes simplex virus 1 (HSV-1), herpes simplex virus 2 (HSV-2), varicella-zoster virus (VZV), Epstein-Barr virus (EBV), human cytomegalovirus (HCMV), human herpesvirus 6A (HHV-6A), human herpesvirus 6B (HHV-6B), human herpesvirus 7 (HHV-7), and Kapo si's sarcoma-associated herpesvirus (KSHV) (1). HHV infections may contribute to the development of both (sub)acute and chronic neurological diseases, which result from nervous tissue damage either in the scope of lytic viral replication or from extensive virus-directed or autoreactive immune responses. An overview of the neurological complications of HHV infections is provided in Table 1.
Notably, clinical manifestations of HHV infections range from asymptomatic infection to severe central nervous system (CNS) inflammation with long-term neurological sequelae and progressive neurodegenerative syndromes (27,28). Central to this variability in disease susceptibility and outcome is the complex interplay between the infecting virus and host immune responses. Owing to their ability to establish latent infection and undergo recurrent reactivation, HHVs exert a lifelong influence on the host immune system (29). The long-term interplay between herpesviruses and intrinsic, innate, and adaptive immune pathways shapes host immune functions with significant implications for nervous system homeostasis. Coinfection with multiple herpesviruses further increases the complexity of virus-host interactions and com plicates the investigation of disease associations. Moreover, the risk of developing herpesvirus-related neurological disease depends heavily on individual factors, including the specific viral strain, timing of infection, and host genetic predispositions that influence immune responses and susceptibility to nervous system infection.
In this context, recent research has underscored the importance of the human leukocyte antigen E (HLA-E) pathway. HLA-E is a non-classical major histocompatibility complex (MHC) class I molecule with key immunomodulatory functions, responsible for fine-tuning immune responses by balancing efficient viral control with the pre vention of excessive immunopathology (30). Importantly, HHVs have evolved strat egies to manipulate HLA-E expression and peptide presentation, thereby facilitating immune evasion. Variability in HLA-E expression, peptide-binding capacity, and receptor engagement could therefore influence individual susceptibility to both acute and long-term neurological complications of HHV infections.
In light of recent advances in understanding HHV-related neuroimmunology, the HLA-E axis emerges as a promising focal point for research. In this review, we therefore aim to summarize and discuss current knowledge and future directions regarding the role of HLA-E in herpesvirus-associated neurological diseases, highlighting its potential as both a biomarker and a therapeutic target.
## HUMAN HERPESVIRUS INFECTIONS AND ACUTE NEUROLOGICAL DISEASES
Acute and subacute neurological manifestations of HHV infections are well documented and include a wide spectrum of central and peripheral nervous system disorders (2,13,17,25). HSV-1, HSV-2, and VZV are neurotropic viruses that establish latency in neuro nal ganglia following primary infection and are most frequently associated with acute neurological diseases among the nine HHVs (2,31). Acute inflammatory neurological diseases may arise during either primary infection or viral reactivation. Although HSV-1, HSV-2, and VZV all belong to the Alphaherpesvirinae subfamily, their pathogenesis in the nervous system differs substantially.
HSV-1 infection of the CNS typically results in acute encephalitis, most commonly affecting the temporal and frontal lobes, following viral entry via the olfactory nerve or anterograde transport from the trigeminal ganglia (32)(33)(34). Patients may present with
## HUMAN HERPESVIRUS INFECTIONS AND CHRONIC NEUROLOGICAL DISOR DERS
Over the past decades, growing evidence has linked HHVs to the development of chronic neurological disorders. The most prominent example is the strong epidemiological and mechanistic association between EBV and multiple sclerosis (MS), a chronic neuroinflammatory and autoimmune disease of the CNS. A recent study reported that the risk of developing MS was increased 32-fold following infection with EBV, whereas no increased risk was observed after infection with other viruses, including the similarly transmitted HCMV (14). Additionally, herpesviruses have been implicated in the pathogenesis of neurodegenerative diseases, particularly Alzheimer's disease (AD). HSV-1 is the most extensively studied HHV in this context (4). Post-mortem analyses, in vitro studies, and animal and organoid models have demonstrated that HSV-1 infection is associated with amyloid β (Aβ) deposition in the brain, a hallmark of AD (53)(54)(55)(56)(57). These findings support the hypothesis that Aβ may have protective antiviral functions and accumulates over time as a result of repetitive, subclinical HSV-1 reactivations within the CNS (22). Similar observations have been reported for HHV-6A, HHV-6B, and HHV-7 CNS infections, but their clinical relevance remains less clear (21,22). Emerging data suggest that HCMV, through replication in the gut and subsequent transport via the vagus nerve, may also promote Aβ accumulation in the CNS (18). VZV reactivation has likewise been linked to an increased risk of dementia, and recent studies highlight the protective effect of herpes zoster vaccination in reducing this risk (9,10,58), likely through the prevention of neuroinflammatory processes.
Besides AD, a role of herpesvirus infection in the development of Parkinson's disease (PD) has been proposed. Studies have indicated the possibility of molecular mimicry between α-synuclein and HSV-1, as well as EBV-specific peptides, potentially driving autoimmune responses directed against α-synuclein deposits in dopaminergic neurons, thereby promoting neuronal degeneration in the substantia nigra (5,16). Epidemiolog ical data have linked herpes zoster to an increased risk of PD development (11). The role of HCMV in PD remains controversial. Some studies suggest that chronic neuroin flammation associated with HCMV-driven immunosenescence may contribute to PD progression, while a recent seroprevalence study reported an increased risk of PD in HCMV IgG-seronegative males (17,19). Overall, the role and mechanism of HHV infection in PD remain unclear and deserve further investigation.
## HLA-E: A CONSERVED IMMUNOREGULATOR
The HLA is a highly polymorphic genomic region on chromosome 6 that encodes molecules responsible for pathogen recognition by T cells and the regulation of immune responses. HLA class I genes are broadly divided into classical (HLA-A, HLA-B, and HLA-C) and non-classical genes (HLA-E, HLA-F, HLA-G, HFE, MR1, and MHC-like proteins of the CD1 family) (59,60). While classical HLA class I molecules are highly polymorphic, ubiquitously expressed on nucleated cells, and present a broad array of endogenous peptides to cytotoxic CD8 + T lymphocytes (CTLs), non-classical class I molecules, by contrast, are characterized by restricted tissue expression, limited polymorphism, and a specialized immunoregulatory role.
HLA-E is a non-classical HLA class I molecule with specific immunoregulatory functions. Evolutionary studies indicate that the HLA-E locus is the most highly conserved among all primate MHC class I genes (61). The baseline surface expression of HLA-E is low in most nucleated cells compared to classical MHC class I molecules (62). In vitro studies have reported that HLA-E has a short cell surface half-life, reflecting its residence time at the plasma membrane, of about 12 minutes compared to several hours for classical HLA class I molecules, after which it is rapidly internalized and trafficked from the cell surface to endosomes (60,63,64). There, HLA-E is presumably loaded with lysosome-digested peptides and then recycled to the cell surface (65). Aside from being present in its membrane-bound form, HLA-E can also be released through cleavage by metalloproteinases in a soluble form (sHLA-E), which is thought to protect bystander cells from excessive immune damage (66,67).
While HLA-E transcripts are broadly distributed across tissues, high-level HLA-E protein surface expression is physiologically restricted to resting and activated leuko cytes and endothelial cells (30). Under pathological conditions, such as infections or tumors, HLA-E expression is, however, often upregulated in most tissues. Studies in tumor models indicate that this increase in HLA-E is mainly driven by gene transcription mechanisms influenced by cytokines such as tumor necrosis factor α (TNFα), interleu kin-1β (IL-1β), and interferon γ (IFNγ) via STAT1-dependent or via Class II transactiva tor/SXY regulatory pathways (67)(68)(69).
The HLA-E gene comprises seven exons encoding structural HLA-E components: exon 1 encodes the leader peptide; exons 2, 3, and 4 encode the α1, α2, and α3 domains, respectively; exon 5 encodes the transmembrane region; and exons 6 and 7 encode the cytoplasmic tail (visualized in reference 70) (67). The HLA-E molecule is assembled in the endoplasmic reticulum (ER) into a trimeric complex consisting of the heavy chain (~45 kDa), containing the α1, α2, and α3 domains, which is non-covalently bound to β₂-microglobulin and a nonameric peptide tail (visualized in reference 71) (67). HLA-E has a less flexible binding pocket than classical HLA I molecules, which limits peptide-dependent conformational changes. Peptides are bound via two main anchors (P2 and P9) and three secondary anchor residues at positions P3, P6, and P7 (60,67). HLA-E-bound peptides are preferentially derived from signal leader sequences of other HLA class I molecules and have the canonical sequence VMAPRT(L/V)(V/L/F)L, thus excluding only a few HLA-B and HLA-C allotypes, as well as leader sequences from HLA-F and HLA-E (72,73).
Under physiological conditions, surface-bound, peptide-loaded HLA-E engages heterodimeric CD94/NKG2A receptors expressed on distinct subsets of cytotoxic natural killer (NK), NKT, and some CD8 + T cells (67,74). The interaction between HLA-E and the inhibitory CD94/NKG2A receptor delivers a strong negative signal via the ITIM motifs in the cytoplasmic domain of NKG2A, suppressing NK and T cell cytotoxicity (75,76). This protects healthy, MHC class I-expressing cells from lysis and mediates immune tolerance to autologous cells. HLA-E-mediated immune regulation thus plays a pivotal role in maintaining self-tolerance and preventing autoimmunity.
## HLA-E IN THE CENTRAL NERVOUS SYSTEM
The CNS is considered an immunoprivileged organ, where interactions with the peripheral immune system are tightly coordinated at restrictive barrier sites, namely the blood-brain barrier (BBB) and blood-CSF barrier. The CNS harbors unique immu nological features that support an immunosuppressive environment. Under physiologi cal conditions, HLA-E expression in the CNS is relatively low compared to peripheral tissues; however, HLA-E is detectable on multiple resident cell types, including microglia, astrocytes, and the endothelial cells of the BBB (Human Protein Atlas, proteinatlas.org [77]) (78).
Recent transcriptomic analyses have shown that HLA-E is predominantly expressed in microglia, enabling these resident immune cells to participate in local immune regulation. In contrast, astrocytes show only basal levels of HLA-E expression under healthy conditions. This restricted expression pattern suggests a role of HLA-E in local immune surveillance and homeostasis within the CNS. The engagement of HLA-E with the inhibitory receptor CD94/NKG2A on CNS-resident or infiltrating NK cells and CTLs likely delivers a "do not kill" signal, preventing the lysis of healthy cells (67)-a mecha nism especially critical in the CNS, where excessive immune-mediated cytotoxicity could cause irreversible damage to neurons and glial cells.
HLA-E expression on endothelial cells is thought to protect the BBB from cytotoxicity by circulating NKG2A + NK and T cells. A recent study reported a reduction of HLA-E expression in endothelial cells of the BBB in amyotrophic lateral sclerosis as well as in patients with frontotemporal lobar degeneration, suggesting that downregulation of HLA-E may facilitate BBB breakdown (79). Although baseline HLA-E expression in the CNS is low, it may be upregulated in neuroinflammatory CNS disorders. Studies have shown that TNFα, IL-1β, and IFNγ upregulate cell-surface HLA-E expression on endothelial cells in vitro and induce the release of sHLA-E (66). This upregulation delivers inhibitory signals that protect healthy CNS tissue from bystander damage during infections or autoim mune responses, providing a neuroprotective mechanism during CNS inflammation. In this line, a recently published study demonstrated an increased HLA-E protein expression in endothelial cells of active MS lesions (80).
These diverse disease contexts underscore that HLA-E is a central immunoregulatory node in the CNS. Its expression reflects a broad immunomodulatory response to injury and inflammation in the CNS, restraining excessive immune activation while preserving neural integrity.
## HLA-E POLYMORPHISM
In comparison to the highly polymorphic classical HLA class I molecules, HLA-E exhibits remarkable conservation and low polymorphism. In the human population, HLA-E is predominantly represented by two functional alleles: HLA-E*01:01 and HLA-E0*01:03, which together comprise over 99% of alleles across diverse ethnicities, implying persistent stabilizing selection acting on HLA-E over an extended period (81,82). Globally, HLA-E*01:01 is more frequent in Africa and the western part of South America, while HLA-E*01:03 is more common in European, Southeast Asian, and East Asian populations. Beyond these two common alleles, rare HLA-E variants exist, but they typically differ by synonymous changes or amino acid substitutions outside the peptide-binding groove, and their functional impact remains poorly understood (83).
HLA-E*01:01 and HLA-E*01:03 differ by a single nonsynonymous substitution in exon 3. This base change alters codon 107 in the α2 domain of the heavy chain, resulting in an arginine (Arg) in HLA-E*01:01 or a glycine (Gly) in HLA-E*01:03 (30,84,85). Amino acid 107 is located in an outwardly exposed loop below the α2-helix of the peptide-bind ing platform and does not affect the general structure of peptide-assembled HLA-E molecules (30,86). HLA-E*01:03 appears to be the older allele, as Gly107 is exclusively found in all primate HLA-E orthologs (81).
Although subtle, this amino acid substitution affects several molecular and functional properties. First, expression levels differ between the alleles: HLA-E*01:03 demonstrates greater thermal stability and is generally expressed at higher steady-state levels on the cell surface compared to HLA-E*01:01, likely due to enhanced assembly efficiency with β2-microglobulin, slower ER egress, and improved peptide binding and loading. Second, peptide-binding stability is also affected: HLA-E*01:03 forms more thermally stable peptide-MHC complexes, particularly with canonical leader sequence-derived nonameric peptides from other HLA class I molecules. This increased stability extends the half-life of HLA-E*01:03 at the plasma membrane, prolonging its availability to engage with immune receptors (30,(86)(87)(88)(89)(90)(91).
Functionally, high HLA-E*01:03 expression through binding of viral peptides may dampen cytotoxic responses via enhanced CD94/NKG2A engagement, delaying viral clearance. In contrast, low HLA-E*01:01 expression limits inhibitory signaling, allowing faster immune activation but increasing the risk of bystander tissue damage. In this context, HLA-E*01:03 may act as a neuroprotective factor but also as a potential facilitator of viral immune evasion. Conversely, HLA-E*01:01 may enhance early viral control but pose a higher risk of long-term neuroinflammation and tissue injury. Recently, HLA-E*01:01 has been associated with higher risk for MS, potentially sug gesting that this less inhibitory allele may promote excessive immune activation and predispose individuals to immune-mediated damage in autoimmune or inflammatory CNS disorders (15).
## HLA-E AND HUMAN HERPESVIRUS INTERACTION
Over the course of co-evolution with humans, herpesviruses have developed sophistica ted strategies to evade immune detection, particularly by NK cells and T cells, which are crucial for controlling viral infections (92). One such strategy involves the manipu lation of HLA-E expression and function. Upon HHV infection, HLA-E can bind "self" peptides from stressed cells as well as "non-self, " viral peptides from infected cells. By modulating the HLA-E axis, HHVs exploit natural immunoregulatory mechanisms intended to prevent autoreactive damage, thereby avoiding detection and elimination by the immune system (67).
During acute infection, HHVs typically induce downregulation of classical MHC class I molecules on infected cells, a mechanism that prevents viral antigen presentation to CTLs (93). The lack of MHC class I leader peptides results in the downregulation of surface HLA-E and the loss of an immunological "self" signal, rendering infected cells highly susceptible to NK cell-mediated lysis. To circumvent this form of immune surveillance, HHVs have evolved mechanisms to upregulate HLA-E even in the absence of classical MHC class I molecules. Viral peptides can be loaded onto HLA-E, stabilizing it on the surface of infected cells, which delivers a potent inhibitory signal to NK and T cells via NKG2A and prevents the lysis of virus-infected cells. Additionally, the manipulation of HLA-E may serve the establishment and maintenance of viral latency in various cell types, depending on the specific HHV tropism. An overview of the exploitation of the HLA-E pathway by different HHVs, which markedly impacts host immune responses and, as a consequence, various aspects of neuronal homeostasis, is provided below and shown in Fig. 1. In addition, Table 2 lists published HHV-specific peptide sequences that can bind to and upregulate HLA-E. However, the presence of additional, as of yet unde tected, HLA-E-binding peptides cannot be excluded. Notably, most studies of peptide binding to HLA-E rely on cell surface stabilization assays, in which peptides often induce only modest or borderline upregulation of HLA-E (58). The functional consequences of these interactions, both regarding the inhibition of NKG2A + cells and the activation of HLA-E-restricted T cells or NKG2C + cells, remain unclear and therefore warrant further investigation.
## THE ROLE OF HLA-E IN HHV REACTIVATION
All human herpesviruses undergo recurrent, often subclinical reactivation from latent infection, and several environmental factors have been identified that can trigger viral reactivation (29,102). Reactivation of latent HSV-1 in neurons, for example, can be induced by systemic physical or emotional stress, fever, microbial co-infection, ultraviolet exposure, or hormonal imbalance (102,103). Many of these triggers activate classical signal transduction pathways, including protein kinase C, p38 kinase, c-Jun N-terminal kinase, ERK kinase, and PI3 kinase, which can alter viral and host transcriptional programs to initiate the lytic cycle (102).
Beyond these canonical pathways, the HLA-E pathway has recently emerged as an additional critical regulator of herpesvirus latency and reactivation. In HCMV infection, reactivations frequently occur in immunosuppressed transplant recipients, highlighting the role of impaired NK-cell and CD8 + T cell functions. Recent studies demonstrated that the HLA-E-UL40 axis has a substantial impact on the level of HCMV replication in transplant recipients. Individuals carrying the HLA-E*01:03/01:03 genotype who were infected with HCMV strains exhibiting high-affinity HLA-E-binding UL40 peptide variants showed enhanced stabilization of HLA-E on infected cells. This led to potent inhibition of NKG2A + NK cells and a markedly increased risk of high-level viral reactivation (104,105).
A similar mechanism may be present during EBV infection. EBV reactivation and the associated transition from latency to lytic replication are typically triggered by host cell stress or by B-cell differentiation into plasma cells through B-cell receptor engagement and cytokine signaling (102). However, recent studies have also reported a high prevalence of the HLA-E*01:03/01:03 genotype among patients with EBV-associated lymphomas, suggesting a link between this allele and impaired immune control of EBV-infected cells (97,106). The same studies identified distinct EBV latent membrane protein 1 (LMP-1)-derived peptide variants that strongly upregulate HLA-E and are associated with potent inhibition of NKG2A + NK cells, findings predominantly observed in EBV-associated lymphoma patients (97,106). Interestingly, these peptide variants are also frequently detected in MS patients (15), suggesting that the HLA-E pathway may facilitate more frequent EBV reactivation, which, in turn, could increase the risk of EBV-associated autoimmune and lymphoproliferative diseases.
## EXPLOITING THE HLA-E/NKG2A AXIS: HSV-1 STRATEGY FOR PERSISTENCE
Engagement of the HLA-E-NKG2A pathway as an immune evasion mechanism has so far been identified for several HHVs, and recent observations indicate that this might also account for HSV-1. We have previously identified two conserved HSV-1 peptides that can bind and stabilize HLA-E on the infected cell's surface, likely mediating inhibitory NKG2A signals on cytotoxic immune cells (94). Both of these peptides are expressed during the lytic HSV-1 replication cycle, indicating that HSV-1 must initiate at least some degree of replication in order to engage HLA-E when presenting the identified peptide sequences (107). Furthermore, we observed that the two HSV-1 specific peptides each showed preferential binding to one of the two HLA-E alleles, i.e., HLA-E*01:01 and HLA-E*01:03. In line with this observation, we showed that patients with HSV-1 encephalitis harbored an increased frequency of homozygous HLA-E genotype variants, potentially reflecting the increased binding affinity of the respective peptide sequences to their preferred HLA-E allele, resulting in increased HLA-E stabilization. Considering the presumed association between HSV-1 and AD development, engagement of the HLA-E pathway by HSV-1infected CNS cells could facilitate immune evasion and enable repeated HSV-1 reactiva tions in the brain, leading to amyloid accumulation and initiation of neurodegeneration, particularly in genetically predisposed individuals (53,55). Importantly, the interaction between HSV-1-specific HLA-E stabilization and the NKG2A receptor, both on NK cells and T cells invading the CNS, needs to be functionally confirmed and further investigated to establish its relevance in disease pathogenesis. As a hallmark of Alphaherpesvirinae, HSV-1 establishes latency in neuronal ganglia, where viral replication is restricted by local NKG2A + CTLs in a non-cytolytic manner via the expression of IFNγ and granzyme-mediated cleavage of immediate early HSV-1 replication products (108). Satellite glial cells (SGCs) surrounding neurons act as antigen-presenting cells capable of expressing MHC class I molecules, and SGCs have been proposed to protect HSV-1-infected trigeminal neurons from T cell-mediated cytotoxicity via expression of HLA-E and PD-L1 (109). In this context, the expression of (other) HLA-E-binding peptides during early or abortive replication of HSV-1-or other neurotropic HHVs-deserves further investigation.
## HLA-E-DRIVEN IMMUNE EVASION BY EBV AND ITS ROLE IN CNS AUTOIM MUNITY
Engagement of the HLA-E-NKG2A pathway as an immune evasion mechanism has also been reported for EBV, a ubiquitous herpesvirus that establishes lifelong latency in B cells (96). A central immune evasion mechanism involves the expression of EBV LMP-1derived peptide variants that bind and upregulate surface HLA-E, thereby suppressing NKG2A + NK cells and CD8 + T cell functions and impairing their ability to recognize and eliminate latently EBV-infected cells. Consequently, EBV escapes immune surveillance, which promotes viral latency and contributes to the virus's long-term persistence.
The exploitation of the HLA-E-NKG2A immune evasion axis by EBV has profound implications for the development of MS. EBV infection is a major environmental risk factor for MS, and its capacity to modulate HLA-E through LMP-1 is thought to drive chronic immune activation and influence disease susceptibility as well as progression. Interestingly, LMP-1-derived peptides show a high degree of polymorphism, leading to differential interactions with HLA-E (15,96,97). Specific LMP-1-derived peptide variants, such as GGDPHLPTL and GGDPPLPTL, which strongly upregulate HLA-E surface expression, are predominantly found in MS patients. In contrast, LMP-1derived peptide variants that show only weak upregulation of HLA-E, i.e., GSDPHLPTL and GGDPHLPPL variants, occur more frequently in healthy controls compared to MS patients (15). This results in potent suppression of NKG2A + NK cells and CD8 + T cells in MS patients, impairing their ability to recognize and kill EBV-infected cells. Consequently, EBV may hinder the clearance of infected and potentially autoreactive EBV-infected B cells, facilitating the breakdown of immune self-tolerance and fostering sustained neuroin flammation in MS. Thus, while the HLA-E-NKG2A axis normally serves to protect host tissues from immune-mediated damage, its exploitation by EBV undermines effective immune control and may set the stage for autoimmune pathology.
The strong association of LMP-1 peptide variants with the development of MS further raises the question of whether analyzing LMP-1 peptide variants together with HLA-E alleles represents a suitable strategy to identify patients at risk for MS. A recent retrospective study demonstrated a 260-fold increase in risk among individuals carrying both LMP-1 and HLA-E risk variants, suggesting that this combined analysis is promising for stratifying individual risk (15).
## EBV-SPECIFIC, HLA-E-RESTRICTED T CELLS AND MS
Beyond its canonical role in NKG2A + NK and T cell regulation, HLA-E has emerged in recent years as a pivotal modulator of adaptive immune responses. Under conditions such as viral or bacterial infections, HLA-E can present pathogen-derived peptides to unconventional, HLA-E-restricted CD8 + T cells via their αβ T cell receptor (73,98,99). These HLA-E-restricted T cells can bypass classical MHC presentation pathways and mount robust cytotoxic responses, maintaining the detection and clearance of infected cells despite HLA-E-mediated NKG2A + NK cell evasion. While HLA-E-restricted regulatory CD8 + T cellresponses have been described in human autoimmune disease, such as type 1 diabetes (110,111), such regulatory phenotypes are not well characterized in the context of viral infections. Instead, most reported HLA-E-restricted T cell responses in viral infection appear to adopt cytotoxic or effector functions (98).
EBV is able to induce an HLA-E-restricted CTL response in the human host (98)(99)(100).
A recently published study revealed that HLA-E-restricted CD8 + T cells recognizing the EBV BZLF1-derived SQAPLPCVL peptide are significantly more frequent in individuals with asymptomatic primary EBV infections. In contrast, low frequencies of these cells were found in patients with symptomatic primary EBV infection (infectious mononu cleosis, IM), indicating that these specific T cells play an important part in control of EBV infections. The same study demonstrated that HLA-E*01:03/01:03-expressing cells, compared to HLA-E*01:01/01:01-expressing cells, are associated with especially stable SQAPLPCVL-mediated upregulation of HLA-E, a more pronounced activation and proliferation of HLA-E-restricted, SQAPLPCVL-specific CTLs in response to EBV-infected cells and, subsequently, more efficient inhibition of EBV spread in vitro (97).
Besides primary EBV infection, recent studies have also implicated a controversial role of SQAPLPCVL-specific CD8 + T cells in the pathogenesis of MS. We have previ ously reported lower frequencies of these cells in relapsing-remitting (RR) MS patients compared to healthy individuals, especially in individuals with a history of IM (15). Other groups have, however, reported increased frequencies of SQAPLPCVL-specific CD8 + T cells in RRMS compared to primary progressive MS patients or healthy controls, potentially indicating a role in RRMS disease pathogenesis or excessive EBV reactivation (112). Together, these observations point toward a multifaceted role of SQAPLPCVL-spe cific, HLA-E-restricted CD8 + T cells in EBV immunity and MS, and their impact may depend on the interplay between host HLA-E genotype, EBV infection history, and disease subtype and stage.
## HCMV-SPECIFIC, HLA-E-RESTRICTED NK CELLS AND MS
Besides CD94/NKG2A, peptide-loaded HLA-E can also bind the CD94/NKG2C receptor, which shares 75% amino acid identity with NKG2A and is similarly expressed on NK cells and CTLs (113,114). In general, CD94/NKG2C binds peptide-loaded HLA-E molecules with approximately sixfold lower affinity than CD94/NKG2A (115). In addition, CD94/ NKG2C recognizes a substantially more restricted peptide repertoire, mainly limited to peptides derived from HLA class I leader sequences in the form of VM(A/P)PRT(L/V) (V/L/I/F)L (115)(116)(117).
In contrast to the inhibitory NKG2A, NKG2C functions as an activating receptor. It lacks inhibitory ITIM motifs in its cytoplasmic domain and instead associates with CD94 and the adaptor protein DAP12, the latter containing an activating ITAM motif (118). Engagement of NKG2C with HLA-E/peptide complexes triggers cytolysis of the target cell. NKG2C expression is low in immature NK cells but may increase progressively during maturation, while NKG2A expression declines in parallel. Consequently, mature NK cells typically express either the inhibitory NKG2A or the activating NKG2C, but rarely both (67).
The expression of NKG2C on NK cells is strongly associated with HCMV infection. HCMV infects a large proportion (50%-100%) of the human population and evades T cell recognition by downregulating classical HLA class I molecules. At the same time, it avoids NK cell-mediated lysis by encoding for a UL40-derived peptide that mimics HLA class I leader sequences, thereby promoting HLA-E expression on infected cells and enabling NKG2A-mediated immune evasion. However, HLA-E surface stabilization also permits recognition via the activating CD94/NKG2C receptor and promotes the expansion and activation of virus-specific NKG2C + NK cells. When engaged by the HLA-E/ UL40-derived peptide complex, NKG2C stimulates NK cell cytotoxicity and cytokine production, contributing to the control of HCMV infection (30,119).
NKG2C + NK cells display a distinctive epigenetic profile, with alterations in key transcription factors, signaling adaptors, and surface receptor expression (30,(120)(121)(122)(123)(124). These changes are accompanied by specific clonal expansion and a secondary memory response to HCMV reinfection in adaptive NK cell subsets, resembling clonal T cell responses (125,126). Interestingly, UL40 is a highly polymorphic molecule, and different UL40-derived peptide variants may bind with varying affinities to HLA-E, ultimately affecting the expansion of NKG2C + NK cells (101,119,127).
HCMV infection is a major environmental factor shaping immune responses involved in MS. The role of HCMV in MS pathogenesis and progression remains controversial, as it may contribute through different mechanisms to both the development and prevention of MS (20). Observational studies have demonstrated a somewhat lower MS prevalence in HCMV-seropositive individuals and suggested that HCMV infection may provide some protection against the development of MS (14,(128)(129)(130)(131). Other studies have reported better clinical outcomes in HCMV-seropositive MS patients (132,133), and it has been proposed that HCMV-mediated immunomodulation may convey a protective effect in these individuals.
A recent study by our group revealed that specific immune responses directed against a highly conserved peptide sequence within the EBV nuclear antigen 1 (EBNA-1) region, EBNA-1 381-452 , cross-react with distinct CNS-derived proteins and may elicit autoimmune processes through molecular mimicry (134)(135)(136). However, potent cytotoxic NKG2C + NK cells can eliminate these autoreactive immune cells via recognition of surface HLA-E on activated, autoreactive B cells, CD8 + T cells, and CD4 + T cells, thereby potentially providing, to some extent, protection against MS. Consistent with this finding, recent studies have demonstrated that MS patients exhibit less potent NKG2C + NK cell responses than healthy controls. Furthermore, higher frequencies of NKG2C + NK cells in HCMV-seropositive individuals with MS have been associated with lower disability scores and a decreased risk of disability progression, suggesting a potential protec tive effect mediated through enhanced immunoregulatory functions (137). Mechanisti cally, impaired NKG2C + NK cell responses in MS have been linked to the absence of HCMV infection, a genetic deletion of the NKG2C receptor-which directly correlates with reduced or absent NKG2C surface levels and decreased frequency of NKG2C + NK cells in vivo (138, 139)-or infection with HCMV isolates encoding UL40 peptide variants that induce only weak HLA-E upregulation and limited expansion of NKG2C + NK cells. Ultimately, these factors may reduce NK cell-mediated protection against MS via ineffective control of EBV-related autoimmunity by HCMV-induced NKG2C + NK cells (15). These findings raise the question of whether the induction of high-level NKG2C + NK cells, potentially via therapeutic vaccination with highly potent UL40 peptides or through cellular immunotherapies, could serve as an additional option to limit the pathogenesis and progression of MS.
## HCMV-SPECIFIC, HLA-E-RESTRICTED T CELLS AND MS
While the expansion of NKG2C + cells was initially characterized in the context of NK cell responses, it is now evident that the HCMV-NKG2C-HLA-E axis also extends into adaptive immunity, promoting the expansion of specialized NKG2C + CD8 + and CD4 + T cell subsets (98,99,(140)(141)(142). Studies have reported that increased NKG2C expression in CD8 + T cells correlates with greater disability, as reflected by higher Expanded Disability Status Scale scores in HCMV-seropositive MS patients (143). The pathogenic role of NKG2C T cells in CNS autoimmunity is further supported by post-mortem analyses of MS brain tissue, which show that HLA-E is upregulated on oligodendrocytes in active lesions, where it colocalizes with infiltrating NKG2C + CD4 + T cells (144). This spatial proximity strongly suggests direct interactions between HCMV-imprinted T cells and CNS-resident cells via the HLA-E-NKG2C axis. Such interactions may contribute to immune-medi ated demyelination and neuronal damage, potentially through bystander killing and amplification of chronic inflammation.
## FUTURE DIRECTIONS
The differential functions mediated by HLA-E in the context of HHV infection depend on the presence of specific viral strains and peptides within the host and host genetic risk factors influencing HLA-E peptide binding, peptide presentation, and interaction with NK and T cells. In light of recent research findings, the collective assessment of these factors, together with the timing of infection and consideration of HHV co-infections, may enable risk stratification for the development and prognosis of HHV-associated acute and chronic neuroinflammatory disorders. HLA-E-associated pathways may also serve as therapeutic targets by enhancing or reducing cytotoxic responses via HLA-E, NK cell-directed or T cell-directed treatments, depending on the underlying pathomechan isms of specific diseases. Future studies will need to focus on the impact of immunomo dulatory treatments on these diseases in order to maintain the delicate balance between protective immune activation and excessive, tissue-damaging cytotoxicity. Furthermore, the involvement of HLA-E in the development of neurodegenerative disorders in interplay with HHV infections deserves further investigation, as effective therapeutics for these conditions are limited, and HLA-E-directed interventions could potentially halt progression in these chronic disorders.
## CONCLUSION
HHV infections shape the host immune response over a lifetime, with implications for both acute and chronic neurological disorders. HLA-E is a non-classical MHC class I molecule acting as a ligand for both inhibitory and activating NK and T cell receptors and presenting antigenic peptides to unconventional T cells. Through these roles, HLA-E bridges innate and adaptive immunity, thereby fine-tuning immune activity within the CNS. Under physiological conditions, its low level expression contributes to CNS immune privilege, whereas its upregulation during inflammation may enhance neuroprotection by preventing excessive immune-mediated damage. However, the same mechanisms that preserve tissue integrity can, under certain circumstances, promote pathology. HLA-E surface stabilization via binding of HHV-derived peptides and interaction with the inhibitory CD94/NKG2A receptor on NK and T cells facilitates immune evasion. Con versely, engagement of the activating receptor CD94/NKG2C-particularly in infectiondriven or immune-primed contexts-can induce immunoregulatory pathways but may also shift HLA-E's role toward promoting cytotoxic NKG2C + cell responses, with opposing consequences for neuroinflammatory disease development. Thus, understanding the context-dependent effects of HLA-E may aid in predicting individual risk for neurological disease in HHV-infected individuals and in guiding the design of targeted immunothera pies that protect neural tissue while appropriately modulating neuroinflammation.
## References
1. Encephalitis "acute retinal necrosis; post-infectious autoimmune encephalitis Putative role in Alzheimer's disease, possible role in Parkinson's disease"
2. "Varicella-zoster virus Meningitis, encephalitis, cerebellitis, cranial nerve palsies, myelitis, radiculitis, acute retinal necrosis, vasculopathy Congenital VZV (e.g., microcephaly, cortical atrophy, seizures); postherpetic neuralgia; increased risk of dementia and" *Parkinson's disease*
3. Epstein-Barr Virus Meningitis, Polyradiculopathy "acute cerebellar ataxia; acute disseminated encephalomyelitis Strongly associated with multiple sclerosis; possible role in Alzheimer's disease and Parkinson's disease 13-16 Human cytomegalovirus Meningitis, encephalitis, polyradiculopathy, polyneuropathy Congenital CMV (e.g., cerebral calcification, microcephaly, motor impairments, sensorineural hearing loss); possible role in Alzheimer's disease, Parkinson's disease, controversial role in multiple sclerosis 17-20 Human herpesvirus 6A Unclear Possible role in Alzheimer's disease and multiple sclerosis 21-23 Human herpesvirus 6B Encephalitis (primarily children and immunocompromised individuals"
4. Ictv (2023) "International Committee on the Taxonomy of Viruses 10th Report"
5. Steiner, Kennedy, Pachner (2007) "The neurotropic herpes viruses: herpes simplex and varicella-zoster" *Lancet Neurol*
6. Armangue, Spatola, Vlagea et al. (2018) "Spanish herpes simplex encephalitis study G" *Lancet Neurol*
7. Itzhaki (2021) "Overwhelming evidence for a major role for herpes simplex virus type 1 (HSV1)" *Vaccines (Basel)*
8. Caggiu, Paulus, Arru et al. (2016) "Humoral cross reactivity between α-synuclein and herpes simplex-1 epitope in Parkinson's disease, a triggering role in the disease?" *J Neuroimmunol*
9. Berger, Houff (2008) "Neurological complications of herpes simplex virus type 2 infection" *Arch Neurol*
10. Kristen, Santana, Sastre et al. (2015) "Herpes simplex virus type 2 infection induces AD-like neurodegenera tion markers in human neuroblastoma cells" *Neurobiol Aging*
11. Nagel, Niemeyer, Bubak (2020) "Central nervous system infections produced by varicella zoster virus" *Curr Opin Infect Dis*
12. Eyting, Xie, Michalik et al. (2025) "A natural experiment on the effect of herpes zoster vaccination on dementia" *Nature*
13. Pomirchy, Bommer, Pradella et al. (2025) "Herpes zoster vaccination and dementia occurrence" *JAMA*
14. Zhang, Liu, Xu (2024) "Association between herpes zoster and Parkinson's disease and dementia: a systematic review and metaanalysis" *Front Neurol*
15. Ahn, Park, Hong et al. (2016) "Congenital varicella syndrome: A systematic review" *J Obstet Gynaecol*
16. Zhang, Zuo, Jiang et al. (2021) "Epstein-Barr virus and neurological diseases" *Front Mol Biosci*
17. Bjornevik, Cortese, Healy et al. (2022) "Longitudinal analysis reveals high prevalence of Epstein-Barr virus associated with multiple sclerosis" *Science*
18. Vietzen, Berger, Kühner et al. (2023) "Ineffective control of Epstein-Barr-virus-induced autoimmunity increases the risk for multiple sclerosis" *Cell*
19. Woulfe, Hoogendoorn, Tarnopolsky et al. (2000) "Monoclo nal antibodies against Epstein-Barr virus cross-react with alphasynuclein in human brain" *Neurology (ECronicon)*
20. Sanami, Shamsabadi, Dayhimi et al. (2024) "Association between cytomegalovirus infection and neurological disorders: a systematic review" *Rev Med Virol*
21. Readhead, Mastroeni, Wang et al. (2025) "Alzheim er's disease-associated CD83(+) microglia are linked with increased immunoglobulin G4 and human cytomegalovirus in the gut, vagal nerve, and brain" *Alzheimers Dement*
22. Ma, Liao, Tan et al. (2024) "The association between cytomegalovirus infection and neurodegenerative diseases: a prospective cohort using UK Biobank data" *EClinicalMedicine*
23. Vanheusden, Stinissen, Hart et al. (2015) "Cytomegalovi rus: a culprit or protector in multiple sclerosis?" *Trends Mol Med*
24. Readhead, Haure-Mirande, Funk et al. (2018) "Multiscale analysis of independent alzheimer's cohorts finds disruption of molecular, genetic, and clinical networks by human herpesvirus" *Neuron*
25. Eimer, Kumar, Shanmugam et al. (2018) "Alzheimer's disease-associated β-amyloid is rapidly seeded by herpesviridae to protect against brain infection" *Neuron*
26. Engdahl, Gustafsson, Huang et al. (2019) "Increased serological response against human herpesvirus 6A is associated with risk for multiple sclerosis" *Front Immunol*
27. Eliassen, Hemond, Santoro (2019) "HHV-6-associated neurologi cal disease in children: epidemiologic, clinical, diagnostic, and treatment considerations" *Pediatr Neurol*
28. Berzero, Campanini, Vegezzi et al. (2021) "Human herpesvirus 6 encephalitis in immunocompetent and immunocompro mised hosts" *Neurol Neuroimmunol Neuroinflamm*
29. Mann, Morado-Aramburo, Hasbun (2024) "Emerging shadows: HHV-8-associated encephalitis unveiled in a solid organ transplant recipient" *Transpl Infect Dis*
30. Gilden, Mahalingam, Cohrs et al. (2007) "Herpesvirus infections of the nervous system" *Nat Clin Pract Neurol*
31. Meyding-Lamadé, Strank (2012) "Herpesvirus infections of the central nervous system in immunocompromised patients" *Ther Adv Neurol Disord*
32. Grinde (2013) "Herpesviruses: latency and reactivation -viral strategies and host response" *J Oral Microbiol*
33. Rölle, Jäger, Momburg (2018) "HLA-E peptide repertoire and dimorphism-centerpieces in the adaptive NK cell puzzle?" *Front Immunol*
34. Granerod, Ambrose, Davies et al. (2010) "Causes of encephalitis and differences in their clinical presentations in England: a multicentre, population-based prospective study" *Lancet Infect Dis*
35. Niemeyer, Merle, Bubak et al. (2024) "Olfactory and trigeminal routes of HSV-1 CNS infection with regional microglial heterogeneity" *J Virol*
36. Jennische, Eriksson, Lange et al. (2015) "The anterior commissure is a pathway for contralateral spread of herpes simplex virus type 1 after olfactory tract infection" *J Neurovirol*
37. Doll, Thompson, Sawtell (2019) "Infectious herpes simplex virus in the brain stem is correlated with reactivation in the trigeminal Ganglia" *J Virol*
38. Gnann (2017) "Herpes simplex encephalitis: an update"
39. George, Schneider, Venkatesan (2014) "Encephalitis hospitaliza tion rates and inpatient mortality in the United States, 2000-2010" *PLoS One*
40. Hjalmarsson, Blomqvist, Sköldenberg (2007) "Herpes simplex encephalitis in Sweden, 1990-2001: incidence, morbidity, and mortality" *Clin Infect Dis*
41. Esiri (1982) "Herpes simplex encephalitis. An immunohistological study of the distribution of viral antigen within the brain" *J Neurol Sci*
42. Lundberg, Ramakrishna, Brown et al. (2008) "The immune response to herpes simplex virus type 1 infection in susceptible mice is a major cause of central nervous system pathology resulting in fatal encephalitis" *J Virol*
43. Marques, Cheeran, Palmquist et al. (2008) "Prolonged microglial cell activation and lymphocyte infiltration following experimental herpes encephalitis" *J Immunol*
44. Abbuehl, Hofmann, Hakim et al. (2023) "Can we forecast poor outcome in herpes simplex and varicella zoster encephalitis? A narrative review" *Front Neurol*
45. Moon, Kim, Lee et al. (2014) "Comparison of clinical manifestations, outcomes and cerebrospinal fluid findings between herpes simplex type 1 and type 2 central nervous system infections in adults" *J Med Virol*
46. Kallio-Laine, Seppänen, Kautiainen et al. (2009) "Recurrent lymphocytic meningitis positive for herpes simplex virus type 2" *Emerg Infect Dis*
47. Omland, Vestergaard, Wandall (2008) "Herpes simplex virus type 2 infections of the central nervous system: a retrospective study of 49 patients" *Scand J Infect Dis*
48. Mirakhur, Mckenna (2004) "Recurrent herpes simplex type 2 virus (Mollaret) meningitis" *J Am Board Fam Pract*
49. Sato, Ayabe, Shoji et al. (2005) "Herpes simplex virus type 2 recurrent meningitis (Mollaret's meningi tis): a consideration for the recurrent pathogenesis" *J Infect*
50. Lind, Studahl, Berg et al. (2017) "CXCL11 production in cerebrospinal fluid distinguishes herpes simplex meningitis from herpes simplex encephalitis" *J Neuroinflammation*
51. Momméja-Marin, Lafaurie, Scieux et al. (2003) "Herpes simplex virus type 2 as a cause of severe meningitis in immunocompromised adults" *Clin Infect Dis*
52. Kennedy (2023) "The spectrum of neurological manifestations of varicella-zoster virus reactivation" *Viruses*
53. Wood (2002) "Understanding pain in herpes zoster: an essential for optimizing treatment" *J Infect Dis*
54. Sutherland, Steain, Buckland et al. (2019) "Persistence of a T cell infiltrate in human ganglia years after herpes zoster and during post-herpetic neuralgia" *Front Microbiol*
55. Yuki (2001) "Infectious origins of, and molecular mimicry in, Guillain-Barré and Fisher syndromes" *Lancet Infect Dis*
57. Wozniak, Mee, Itzhaki (2009) "Herpes simplex virus type 1 DNA is located within Alzheimer's disease amyloid plaques" *J Pathol*
58. Martin, Aguila, Araya et al. (2014) "Inflammatory and neurodegeneration markers during asymptomatic HSV-1 reactivation" *J Alzheimers Dis*
59. De Chiara, Piacentini, Fabiani et al. (2019) "Recurrent herpes simplex virus-1 infection induces hallmarks of neurodegeneration and cognitive deficits in mice" *PLoS Pathog*
60. Cairns, Rouleau, Parker et al. (2020) "A 3D human brain-like tissue model of herpes-induced Alzheimer's disease" *Sci Adv*
61. Zhang, Wang, Wu et al. (2025) "A microengineered 3D human neurovascular unit model to probe the neuropathogenesis of herpes simplex encephalitis" *Nat Commun*
62. Taquet, Dercon, Todd et al. (2024) "The recombinant shingles vaccine is associated with lower risk of dementia" *Nat Med*
63. Rodgers, Cook (2005) "MHC class Ib molecules bridge innate and acquired immunity" *Nat Rev Immunol*
64. Gillespie, Quastel, Mcmichael (2025) "HLA-E: immune receptor functional mechanisms revealed by structural studies" *Immunol Rev*
65. Knapp, Cadavid, Watkins (1998) "The MHC-E locus is the most well conserved of all known primate class I histocompatibility genes" *J Immunol*
66. Souza, Adams, Altman et al. (2019) "Casting a wider net: Immunosurveillance by nonclassical MHC molecules" *PLoS Pathog*
67. Barber, Souza, Paterson et al. (2022) "Structure-guided stabilization of pathogen-derived peptide-HLA-E complexes using nonnatural amino acids conserves native TCR recognition" *Eur J Immunol*
68. Wallace, Heunis, Paterson et al. (2024) "Instability of the HLA-E peptidome of HIV presents a major barrier to therapeutic targeting" *Mol Ther*
69. He, Gea-Mallorquí, York et al. (2023) "Intracellular trafficking of HLA-E and its regulation" *J Exp Med*
70. (1084)
71. Coupel, Moreau, Hamidou et al. (2007) "Expression and release of soluble HLA-E is an immunoregulatory feature of endothelial cell activation" *Blood*
72. Rohn, Rebmann (2024) "Is HLA-E with its receptors an immune checkpoint or an antigenic determinant in allo-HCT? Best Practice" *Research Clinical Haematology*
73. Gustafson, Ginder (1996) "Interferon-gamma induction of the human leukocyte antigen-E gene is mediated through binding of a complex containing STAT1alpha to a distinct interferon-gammaresponsive element" *J Biol Chem*
74. Gobin, Van Den Elsen, Hla-E (2000) "Transcriptional regulation of the MHC class Ib genes" *Hum Immunol*
75. Arnaiz-Villena, Suarez-Trujillo, Juarez et al. (2022) "Evolution and molecular interactions of major histocompatibility complex (MHC)-G, -E and -F genes" *Cell Mol Life Sci*
76. O'callaghan, Bell (1998) "Structure and function of the human MHC class Ib molecules" *Immunol Rev*
77. Lee, Goodlett, Ishitani et al. (1998) "HLA-E surface expression depends on binding of TAP-dependent peptides Minireview Journal of Virology December"
78. "HLA class I signal sequences" *J Immunol*
79. Pietra, Romagnani, Manzini et al. (2010) "The emerging role of HLA-E-restricted CD8+ T lymphocytes in the adaptive immune response to pathogens and tumors" *J Biomed Biotechnol*
80. Béziat, Descours, Parizot et al. (2010) "NK cell terminal differentiation: correlated stepwise decrease of NKG2A and acquisition of KIRs" *PLoS One*
81. Kabat, Borrego, Brooks et al. (2002) "Role that each NKG2A immunoreceptor tyrosine-based inhibitory motif plays in mediating the human CD94/NKG2A inhibitory signal" *J Immunol*
82. Long (2008) "Negative signaling by inhibitory receptors: the NK cell paradigm" *Immunol Rev*
83. Uhlén, Fagerberg, Hallström et al. (2015) "Tissue-based map of the human proteome" *Proteomics*
84. Kellogg, Pham, Machalinski et al. (2023) "Microglial MHC-I induction with aging and Alzheimer's is conserved in mouse models and humans" *Geroscience*
85. Pineda, Lee, Ulloa-Navas et al. (2024) "Single-cell dissection of the human motor and prefrontal cortices in ALS and FTLD" *Cell*
86. Durrenberger, Webb, Sim et al. (2012) "Increased HLA-E expression in white matter lesions in multiple sclerosis" *Immunology*
87. Grimsley, Ober (1997) "Population genetic studies of HLA-E: evidence for selection" *Hum Immunol*
88. Kraemer, Blasczyk, Bade-Doeding (2014) "HLA-E: a novel player for histocompatibility" *J Immunol Res*
89. Sauter, Putke, Schefzyk et al. (2021) "HLA-E typing of more than 2.5 million potential hematopoietic stem cell donors: Methods and population-specific allele frequencies" *Hum Immunol*
90. Geraghty, Stockschleader, Ishitani et al. (1992) "Polymor phism at the HLA-E locus predates most HLA-A and -B polymorphism" *Hum Immunol*
92. Grimsley, Kawasaki, Gassner et al. (2002) "Definitive high resolution typing of HLA-E allelic polymorphisms: Identifying potential errors in existing allele data" *Tissue Antigens*
93. Strong, Holmes, Li et al. (2003) "HLA-E Allelic Variants" *Journal of Biological Chemistry*
94. Maier, Grzeschik, Weiss et al. (2000) "Implications of HLA-E allele expression and different HLA-E ligand diversity for the regulation of NK cells" *Hum Immunol*
95. Ulbrecht, Couturier, Martinozzi et al. (1999) "Cell surface expression of HLA-E: interaction with human β2-microglobulin and allelic differences" *Eur J Immunol*
96. Celik, Kraemer, Huyton et al. (2016) "The diversity of the HLA-E-restricted peptide repertoire explains the immunological impact of the Arg107Gly mismatch" *Immunogenetics*
97. Marco, Schuster, Backert et al. (2017) "Unveiling the peptide motifs of HLA-C and HLA-G from naturally presented peptides and generation of binding prediction matrices" *The Journal of Immunology*
98. Ruibal, Franken, Van Meijgaarden et al. (2020) "Peptide binding to HLA-E molecules in humans, nonhuman primates, and mice reveals unique binding peptides but remarkably conserved anchor residues" *J Immunol*
99. De Pelsmaeker, Romero, Vitale et al. (2018) "Herpesvirus evasion of natural killer cells" *J Virol*
100. Verweij, Horst, Griffin et al. (2015) "Viral inhibition of the transporter associated with antigen processing (TAP): a striking example of functional convergent evolution" *PLoS Pathog*
101. Marianne, Vietzen, Puchhammer-Stöckl (2023) "Association between human genetic variants affecting the host NK cell response and the development of herpes simplex virus type 1 encephalitis" *J Med Virol*
102. Graninger, Kühner, Berger et al. (2025) "Distinct NK cell genetic variants are associated with HSV-2 Versus VZV Infection of the CNS" *J Med Virol*
103. Mbiribindi, Pena, Arvedson et al. (2020) "Epstein-Barr virus peptides derived from latent cycle proteins alter NKG2A + NK cell effector function" *Sci Rep*
104. Vietzen, Furlano, Cornelissen et al. (2023) "HLA-E-restricted immune responses are crucial for the control of EBV infections and the prevention of PTLD" *Blood*
105. Joosten, Sullivan, Ottenhoff (2016) "Characteristics of HLA-E restricted T-cell responses and their role in infectious diseases" *J Immunol Res*
106. Romagnani, Pietra, Falco et al. (2002) "Identification of HLA-E-specific alloreactive T lymphocytes: A cell subset that undergoes preferential expansion in mixed lymphocyte culture and displays a broad cytolytic activity against allogeneic cells" *Proc Natl Acad Sci*
107. García, Llano, De Heredia et al. (2002) "Human T cell receptor-mediated recognition of HLA-E" *Eur J Immunol*
108. Hammer, Rückert, Borst et al. (2018) "Peptidespecific recognition of human cytomegalovirus strains controls adaptive natural killer cells" *Nat Immunol*
110. Stoeger, Adler (2018) "Novel" triggers of herpesvirus reactivation and their potential health relevance" *Front Microbiol*
111. Roizman, Whitley (2013) "An inquiry into the molecular basis of HSV latency and reactivation" *Annu Rev Microbiol*
112. Vietzen, Rückert, Hartenberger et al. (2021) "Extent of cytomegalovirus replication in the human host depends on variations of the HLA-E/UL40 Axis" *mBio*
113. Guberina, Da, Nardi et al. (2018) "Susceptibility of HLA-E*01:03 allele carriers to develop cytomegalovirus replication after living-donor kidney transplantation" *J Infect Dis*
114. Vietzen, Staber, Berger et al. (2025) "Puchhammer-Stöckl E. 2023"
115. "Inhibitory NKG2A(+) and absent activating NKG2C(+) NK cell responses are associated with the development of EBV(+) lymphomas" *Front Immunol*
116. Mangold, Rathbun, Renner et al. (2021) "Viral infection of human neurons triggers strain-specific differences in host neuronal and viral transcriptomes" *PLoS Pathog*
117. St, Leger, Koelle et al. (2021) "Local immune control of latent herpes simplex virus type 1 in ganglia of mice and man" *Front Immunol*
118. Van Velzen, Laman, Kleinjan et al. (2009) "Neuron-interacting satellite glial cells in human trigeminal ganglia have an APC phenotype" *J Immunol*
119. Sarantopoulos, Lu, Cantor (2004) "Qa-1 restriction of CD8+ suppressor T cells" *J Clin Invest*
120. Jiang, Canfield, Gallagher et al. (2010) "HLA-E-restricted regulatory CD8+ T cells are involved in development and control of human autoimmune type 1 diabetes" *J Clin Invest*
121. Jørgensen, Livbjerg, Hansen et al. (2012) "Epstein-Barr virus peptide presented by HLA-E is predominantly recognized by CD8bright cells in multiple sclerosis patients" *PLoS One*
122. Braud, Allan, 'callaghan et al. (1998) "HLA-E binds to natural killer cell receptors CD94/NKG2A, B and C" *Nature*
123. Kaiser, Barahmand-Pour, Paulsene et al. (2005) "Interactions between NKG2x immunoreceptors and HLA-E ligands display overlapping affinities and thermodynamics" *J Immunol*
124. Siemaszko, Marzec-Przyszlak, Bogunia-Kubik (2023) "Activating NKG2C receptor: functional characteristics and current strategies in clinical applications" *Arch Immunol Ther Exp*
125. Llano, Lee, Navarro et al. (1998) "HLA-E-bound peptides influence recognition by inhibitory and triggering CD94/NKG2 receptors: preferential response to an HLA-Gderived nonamer" *Eur J Immunol*
126. Lauterbach, Wieten, Popeijus et al. (2015) "HLA-E regulates NKG2C+ natural killer cell function through presenta tion of a restricted peptide repertoire" *Hum Immunol*
127. Lanier, Corliss, Wu et al. (1998) "Association of DAP12 with activating CD94/NKG2C NK cell receptors" *Immunity*
128. Hammer, Rückert, Romagnani (2018) "Natural killer cell specificity for viral infections" *Nat Immunol*
129. Béziat, Liu, Malmberg et al. (2013) "NK cell responses to cytomegalovirus infection lead to stable imprints in the human KIR repertoire and involve activating KIRs" *Blood*
130. Lee, Zhang, Hwang et al. (2015) "Epigenetic modification and antibodydependent expansion of memory-like NK cells in human cytomegalovi rus-infected individuals" *Immunity*
131. Schlums, Cichocki, Tesi et al. (2015) "Cytomegalovirus infection drives adaptive epigenetic diversification of NK cells with altered signaling and effector function" *Immunity*
133. Rückert, Lareau, Mashreghi et al. (2022) "Clonal expansion and epigenetic inheritance of long-lasting NK cell memory" *Nat Immunol*
134. Luetke-Eversloh, Hammer, Durek et al. (2014) "Human cytomegalovirus drives epigenetic imprinting of the IFNG locus in NKG2Chi natural killer cells" *PLoS Pathog*
136. Wu, Sinzger, Frascaroli et al. (2013) "Human cytomegalovirus-induced NKG2C(hi) CD57(hi) natural killer cells are effectors dependent on humoral antiviral immunity" *J Virol*
137. Lopez-Vergès, Milush, Schwartz et al. (2011) "Expansion of a unique CD57+NKG2Chi natural killer cell subset during acute human cytomegalovirus infection" *Proceedings of the National Academy of Sciences*
138. Heatley, Pietra, Lin et al. (2013) "Polymorphism in human cytomegalovirus UL40 impacts on recognition of human leukocyte antigen-E (HLA-E) by natural killer cells" *J Biol Chem*
139. Grut, Biström, Salzer et al. (2021) "Cytomegalovirus seroposi tivity is associated with reduced risk of multiple sclerosis-a presympto matic case-control study" *Eur J Neurol*
140. Alari-Pahissa, Moreira, Zabalza et al. (2018) "Low cytomegalovirus seroprevalence in early multiple sclerosis: a case for the "hygiene hypothesis" *Eur J Neurol*
141. Waubant, Mowry, Krupp et al. "For the US Pediatric MS Network. 2011. Common viruses associated with lower pediatric multiple sclerosis risk" *Neurology (ECronicon)*
142. Sundqvist, Bergström, Daialhosein et al. (2014) "Cytomegalovirus seropositivity is negatively associated with multiple sclerosis" *Mult Scler*
143. Lünemann, Avilés, Tintoré et al. (2024) "Cytomegalovirus immune responses are associated with lower serum NfL and disability accumulation risk at multiple sclerosis onset" *Mult Scler*
144. Zivadinov, Nasuelli, Tommasi et al. (2006) "Positivity of cytomegalovirus antibodies predicts a better clinical and radiological outcome in multiple sclerosis patients" *Neurol Res*
145. Vietzen, Kühner, Berger et al. (2024) "Accumulation of Epstein-Barr virus-induced cross-reactive immune responses is associated with multiple sclerosis" *J Clin Invest*
146. Vietzen, Kühner, Berger et al. (2025) "Early identification of individuals at risk for multiple sclerosis by quantification of EBNA-1 381-452 -specific antibody titers" *Nat Commun*
147. Sattarnezhad, Kockum, Thomas et al. (2025) "Antibody reactivity against EBNA1 and GlialCAM differentiates multiple sclerosis patients from healthy controls" *Proc Natl Acad Sci*
148. Martínez-Rodríguez, Cobo-Calvo, Villar et al. (2016) *Minireview Journal of Virology*
149. "Adaptive natural killer cell response to cytomegalovirus and disability progression in multiple sclerosis" *Mult Scler*
150. Muntasell, López-Montañés, Vera et al. (2013) "NKG2C zygosity influences CD94/NKG2C receptor function and the NK-cell compart ment redistribution in response to human cytomegalovirus" *Eur J Immunol*
151. Muntasell, Pupuleku, Cisneros et al. (2016) "Relationship of NKG2C copy number with the distribution of distinct cytomegalovirus-induced adaptive NK cell subsets" *J Immunol*
152. Pietra, Romagnani, Mazzarino et al. (2003) "HLA-E-restricted recognition of cytomegalovirus-derived peptides by human CD8+ cytolytic T lymphocytes" *Proc Natl Acad Sci*
153. Mazzarino, Pietra, Vacca et al. (2005) "Identification of effector-memory CMV-specific T lymphocytes that kill CMV-infected target cells in an HLA-E-restricted fashion" *Eur J Immunol*
154. Pannemans, Broux, Goris et al. (2014) "HLA-E restricted CD8+ T cell subsets are phenotypically altered in multiple sclerosis patients" *Mult Scler*
155. Perri, Zingaropoli, Pasculli et al. (2024) "The impact of cytomegalovirus infection on natural killer and CD8+ T cell phenotype in multiple sclerosis" *Biology (Basel)*
156. Zaguia, Saikali, Ludwin et al. (2013) "Cytotoxic NKG2C+ CD4 T cells target oligodendrocytes in multiple sclerosis" *J Immunol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12326739&blobtype=pdf | # Long-term impact of the SARS-CoV-2 pandemic on respiratory viruses in Germany
Ralf Eggeling, Rolf Kaiser, Florian König, Lisa Koeppel, Laura-Inés Böhler, Michael Böhm, Norbert Schmeißer, Nico Pfeifer, Virology Network
## Abstract
Background Respiratory viral diseases are one of the greatest challenges facing our healthcare system, with them being one of the main causes of death. This has been demonstrated once again by the impact of the SARS-CoV-2 pandemic in recent years. We study the impact of the SARS-CoV-2 pandemic on the prevalence of respiratory viruses by analysing a subset of the Clinical Virology network database, covering 2,216,198 samples tested for 18 different viral pathogens in the time span from 2010 to 2024.
MethodsWe calculated the prevalence of 17 respiratory viruses before and after onset of the SARS-CoV-2 pandemic and compared the degree of seasonality shift with a newly developed a metric dubbed seasonal disruption index. In addition, we compared coinfection statistics prior to and after the pandemic onset, and also studied the correlation of infection counts with non-pharmaceutical interventions in the time frame from early 2020 to end of 2022.
ResultsWe found that the viral pathogens show a varying degree of seasonality disruption. It is largest among those that are known to show a highly seasonal behavior, namely Influenza and RSV, the latter having the highest seasonal disruption index. Most perennial viruses continued to appear throughout the year. Coinfections occurred before and after the pandemic; patterns before and after pandemic onset are surprisingly similar. The occurrence of most viruses is nonlinearly correlated with the degree of non-pharmaceutical interventions.
ConclusionThe SARS-CoV-2 pandemic had a considerable impact on the occurrence and seasonality of other respiratory viruses. While nearly all seasonality patterns were initially disrupted due to the heavy non-pharmaceutical interventions, viruses are regaining their pre-pandemic seasonality.
## Background
Respiratory viral illnesses are one of the key challenges facing our healthcare systems. Most prominently perceived in the population, each year, an estimated 5% to 15% of all people are affected by respiratory tract infections caused by influenza viruses alone [1]. While some outbreaks such as MERS (Middle East respiratory syndrome coronavirus) in 2012 or SARS (severe acute respiratory syndrome) in 2003 remained fairly localized with a limited number of casualties, others such as the Spanish flu (1918)(1919)(1920) and, more recently, the SARS-CoV-2 pandemic, had a profound impact on human development throughout the world [2][3][4]. In response, considerable efforts are taken to prepare for future pandemics and to limit their societal and economic burdens [5,6].
One aspect of preparedness is the continuous observation of known respiratory virus dynamics to enable quick responses to local and global outbreaks to protect at-risk groups [7,8]. Prior long-term studies revealed predictable seasonality patterns for such viral diseases [9][10][11], for example, enabling preparations of the new influenza vaccines [7,12]. It is also important to assess virus-virus interactions, both synergistic and antagonistic, as coinfected patients might suffer from worse health outcomes [2,10,13].
Several studies suggest that previously identified long-term seasonalities before the pandemic have been distorted, leading to missing seasons or shifts in seasonalities during and after the pandemic [14][15][16]. Such investigations into the altered behavior of respiratory viral strains have been made in several countries [17,18], including Germany [14,[19][20][21]]. An extensively studied example is the respiratory syncytial virus (RSV), with multiple studies observing an entire missing season [15,22,23]. This has been correlated with non-pharmaceuticcal interventions (NPIs), implemented to prevent the SARS-CoV-2 pandemic completely overloading national healthcare systems. The reduction in transmissibility hoped to achieve for SARS-CoV-2 also affected the transmission of other previously analyzed respiratory pathogens [15-17, 19, 23, 24]. There are also a few studies regarding coinfections between SARS-CoV-2 and other pathogens [25,26].
Although many insights regarding the impact of the SARS-CoV-2 pandemic on infection dynamics have been gained, there are still gaps in the literature. Most studies only consider a sub-population [14,19] or small region [17,20] rather than considering the entire population of a country due to a lack of data on a large scale. Many studies are also limited by considering only one [16,23,24] or a small number [17,18,20,22] of viruses, and thus also lack or do not even attempt a systematic comparison. Furthermore, many studies were published relatively early in the pandemic [14,15,17,18,21,22]. Hence, they only focus on a short period compared to the previously identified dynamic patterns and do not permit to assess whether any seasonal shifts and newly identified patterns will remain stable or eventually return to the state prior to the pandemic. Finally, coinfection studies in the literature often involve SARS-CoV-2 and influenza [25], widely neglecting interactions among other viruses.
In this article, we attempt to fill some of the gaps by addressing the aforementioned limitations. We build upon our prior study on the data from the clinical-virology.net [9] in order to cover a large number of contributing sites in Germany and infection counts for 18 different viruses. These data span the period before, during, and after the pandemic, up until April 2024. As a result, four full years after the start of the pandemic are available for a comparative analysis against the prepandemic timeframe. In order to quantify seasonal shifts and other disruptions of viral dynamics during the pandemic and, most importantly, compare the behavior of different viruses, we propose a seasonality disruption index (SDI). We further study the correlation of infection counts with NPIs, also with the primary aim of comparing viruses. In addition, we study possible changes in coinfection patterns among all pairs of viruses before and after the start of the pandemic.
## Methods
This section contains a description of the data, followed by a general outline of our analysis methodology, and a description of our newly developed method for the quantification of seasonality disruption.
## Data collection and preprocessing
Since November 2009, the clinical-virology.net, formerly known as RespVir network, has collected multiplex test records for 17 virus infections from patients who showed respiratory infection symptoms. Since January 2020, SARS-CoV-2 has been included in the list of tested viruses. The records stem from up to 47 sites, although the tested viruses and covered time spans differ greatly across sites. We used the clinical virology network (CVN) database updated until April 2024, using the data from January 2010 onwards to be consistent with our previous work [9], and carried out the following data cleaning and preprocessing steps.
First we removed all biologically implausible data entries. There are 5,201 (of 2,222,843) data points that showed positive results against more than four viruses. Additionally, one site had 1444 positive HBoV test results in early 2018 (January-April), with a nearly 100% positivity rate. Both cases are likely caused by incorrect data entry, for instance by swapping positive and negative labels, so we excluded the affected data points entirely. Furthermore one site marked each of 211 positive test results for Influenza A subtype H1N1 also as positive for subtype H3N2. We excluded these counts for the coinfection analysis specifically.
Second, the tests used by some sites do not differentiate between certain virus types or subtypes, such as FLUAgeneric, which cannot distinguish between influenza A H1N1 and H3N2. In this study, we excluded these combitests and retained only the 18 tests against single virus types. This eliminates 6,771 data points that comprise only combitests.
Third, we filtered out 97,710 data points from non-German test sites, as the one objective of the project was to correlate infection counts with non-pharmaceutical interventions, which needs to be done on the country level. For all countries aside from Germany, samples sizes in the CVN database are too small to draw statistically significant conclusions.
An overview and a quantitative description of the final data set after preprocessing, that is all virus names, abbreviations, absolute and relative infection counts, is included in the supplementary material. Unsurprisingly, the number of tests and positive results for SARS-CoV-2 dwarfs those of the remaining viruses, amounting for more than all others combined. But also the percentage of positive tests results is with more than 24% considerably higher than for all other viruses, where positive percentages range between 1% and 16%.
## Analysis methodology
In our analysis, we first inspected the absolute infection and relative infection counts for each virus aside from SARS-CoV-2 in a time-series fashion for the period of January 2010 to April 2024. Relative counts are defined as the absolute number of positive test results divided by the number of tests conducted.
Subsequently, we investigated how the pandemic affected virus dynamics and analysed alterations in seasonal outbreak behaviour. For this purpose, we developed a statistical measure, the so-called disruption statistics, to quantify the disruption of seasonal patterns. It states the difference between the observed infections and the expected infections based on the median from the past ten years. More details and a mathematical description is provided in the following Quantification of seasonality disruption section. Using the disruption statistics and identical parametrization, we identified disruption profiles, further enabling a clustering of the viruses according to their profile in a dendogram. All of the aforementioned measures allowed an objective comparison of seasonal disruption across viruses.
We further studied the association of infection counts and NPIs. For this purpose, we included SARS-CoV-2 and quantified NPIs using the Stringency Index from the Oxford Covid-19 Government Response Tracker (OxCGRT) [27]. This index combines several factors such as school closings and restrictions on public gatherings. Since the stringency index is available only for the time period January 2020 -December 2022, all analyses regarding NPIs are limited to that time period. Our analysis was performed on a monthly aggregation which is why we averaged the daily stringency index values for each month to suit our format.
Lastly, we investigated coinfections among all 18 viruses before and after the start of the pandemic. Due to the small sample sizes post 2020, it was not possible to repeat the analysis as in a previous study [9], where we analyzed virus pairs with a score that shows statistically significant increase or decrease in coinfections compared to what would be expected by chance. Instead in the present study we solely compared coinfection counts before and after the start of the pandemic.
## Quantification of seasonality disruption
To examine how the pandemic affected virus dynamics and seasonal outbreak behavior, we define statistical measures to quantify the disruption of seasonal virus patterns in the following. For an arbitrary virus, let n id(y,m) denote the absolute number of infections in month m of year y, whereby the subscript id(y, m) = 12 * (y -2010) + m returns the global index of the month in the time period under consideration, i.e., how many months have passed since the beginning of 2010. We calculate smoothed infection counts with a window size of w as
Let d denote the number of past seasons we wish to consider and α denote a small pseudocount to avoid infinity in cases of zero infections in a particular month. We define the disruption statistics
This statistics quantifies how much more/less frequent the infection counts in month i were compared to an expected value, which is based on the median of the past d seasons as the baseline. If not explicitly specified otherwise, we use w = 2, α = 1, and d = 10 for all studies.
Let us consider a time period between months a and b (with b > a). We call the vector of disruption statistics values in the chosen time period disruption profile. The seasonality disruption index (SDI) quantifies the total disruption of infection seasonality between months a and b by
which is essentially a root mean squared deviation of the disruption profile to a straight line at zero (no seasonality disruption). The key advantage of this metric over previously used Kullback-Leibler divergence based quantification of seasonality [9] is that it allows quantification of seasonality disruption for arbitrary time spans and not only whole years.
Further, we considered the disruption statistics for each virus in the given time period, e.g., (DS a , . . . , DS b ) as feature vector and applied agglomerative clustering with average linkage using the feature vectors of all viruses as data points and squared Euclidean distance.
$$ñi = 1 2w + 1 i+w ∑ j=i-w n j for i > m.$$
$$DS i = log2 ( ñi + α median(ñ i-12d , . . . , ñi-12 ) + α )$$
$$SDI(a, b) = 1 b -a b i=a DS 2 i$$
## Results
This Result section first describe infection trajectories over time for a few selected viruses, motivating the need for a unified quantitative evaluation, followed by largescale analyses for all viruses regarding seasonality disruption, correlation with non-pharmaceutical interventions, and coinfections.
## Infection counts over time
We first inspected the absolute infection and relative infection counts for each virus in the time period January 2010 to April 2024. Figure 1 displays four typical examples, identical plots for the remaining viruses are shown in the supplementary material. Despite only using German data here, all plots are consistent up to 2019 with the previous analysis of the CVN data [9]. The extended time period also gives insights on the impact of the SARS-CoV-2 pandemic on the infection dynamics.
As a first example, we use the human respiratory syncytial virus (RSV), which can cause severe infections in premature newborn, young children, elderly and immunocompromised [28,29]. These groups are candidates for either passive immunisation (children) for vaccination (elderly, immunocompromised).
We observed a clear seasonal occurrence pattern with annual peaks in January and February, being disrupted at the onset of the pandemic (Fig. 1) [15,16]. A peak in early 2021, expected based on the typical seasonal pattern in the years before, is missing. The following peak emerged in fall 2021 and the peaks of the subsequent two seasons appeared in December instead of January/February. The RSV occurrence pattern recovered from the pandemic disruption with respect to frequency, however, the seasonal peak is still shifted by two months.
Prior to the pandemic, the seasonality pattern of influenza A subtype H3N2 remained consistent over the years regarding onset and duration. Yet, the size of the individual waves differed. Upon the pandemic influenza A H3N2 could not be observed for nearly two seasons before re-appearing at the end of 2022 in a uni-modal wave. Until the end of the study period, no return to the prepandemic seasonal patterns could be seen.
Prepandemically, the human adenovirus (HAdV) was known to belong to the perennial group showing little preference for a particular season but rather occurred in a fluctuating fashion throughout the year. In contrast to other viruses, almost no disruption of absolute infection numbers could be observed upon the SARS-CoV-2 pandemic. The relative HAdV frequency (regarding number of tests) decreased in recent years. Anyhow, this trend has started before the pandemic.
Human Metapneumovirus (HMPV) is a respiratory pathogen, phylogenetically related to RSV and Parainfluenza 1-4, which is not as intensively observed as Influenza, SARS-CoV-2 or RSV. It is a severe burden not only in children but also in adults, especially those over 65 years of age [30][31][32][33][34]. Infections with HMPV are detected and reported within our respiratory pathogens network, CVN, since the beginning of our activity [9]. More recently, vaccines have been developed and have now entered clinical studies [35,36]. Here, we describe the reemergence of HMPV after the pandemic.
Prior to the pandemic, HMPV showed infection peaks in winter (December-January), although there is at least one season (2011) in which low infection counts were detected and therefore a typical peak can hardly be identified. There is a notable peak in absolute infection counts located at the pandemic onset in early 2020 with an absence in of infections in 2021. The sparsity and sporadic occurrence after the start of the pandemic are in line with prepandemic patterns and may be explained by natural variation inherent to the virus. Notably, relative infection counts remained remarkably low after the start of the pandemic.
## Comparison of seasonality disruption
Prepandemic seasonality trends differed to a large degree among the considered viruses as described in our previous work [9]. In order to obtain a quantitative measure for the disruption of the seasonality patterns, we applied the seasonality disruption index (Quantification of seasonality disruption section). We exemplified the index by RSV, the detailed plots for the other viruses are shown in the supplementary material.
Figure 2A displays smoothed absolute infection counts for the entire time period. It also contains the disruption profile after pandemic onset (March 2020 to March 2024). The disruption profile fluctuates in the range of -6 in early 2021 to +5 in late 2021. Thus, it aligns with the presence and absence of expected and unexpected peaks.
We used the disruption profiles as feature vectors and applied hierarchical clustering. The resulting dendrogram with each disruption profile plotted adjacent to the corresponding leaf in the tree is shown in Fig. 2B. Additionally, the plot also displays the seasonality disruption index (SDI) for each virus, which is the root mean square deviation of the disruption profile from a null vector.
Figure 2B reveals that RSV shows the highest SDI and is also least similar to any other virus in the selection in terms of disruption profile. RSV frequency was unexpectedly low upon the onset of the pandemic (Fig. 2A). Nevertheless, RSV occurrence strongly increased in late 2021 after a period of little detection. It should be noted that HPMV shows both a different original seasonal trend and a different disruption, despite being genetically closely related to RSV and causing similar symptoms [37].
While some viruses such as influenza A (FLUA(H3N2) and FLUA(H1N1)) also exhibit a fluctuating disruption profile and thus a relatively high SDI, others, such as the non-enveloped viruses (Rhinovirus(V), EV, HAdV) less little seasonal disruption, which might be explained by them being perennial [9], not exhibiting a strong seasonality to begin with. In general, viruses that exhibited a stronger seasonality pre-pandemically also have a higher SDI, but the correlation is not perfect: Human parainfluenza 3 (HPIV-3), for instance, shows a relatively high SDI simply due to the absence of infection counts early in the pandemic. The dendrogram resulting from clustering disruption profiles resembles the clustering according to the pre-pandemic seasonal patterns [9], but not perfectly. For RSV, it is remarkable that after the pandemic onset in March 2020, the infection counts remain low as long as the stringency index is above a value of 60. Once restrictions were loosened (late 2021), the RSV wave appeared. A small increase in the stringency index then coincided with the end of aforementioned peak. Finally, in late 2022, when all NPI related public health measures aside from public information campaigns were terminated, the infection counts for RSV rose once again.
## Infection counts vs non-pharmaceutical interventions
Influenza A(H3N2) disappeared with rising stringency index in early 2020 and reappeared at the end of 2022. Unlike RSV, influenza A(H3N2) did not re-emerge within the pandemic period.
We calculated the Spearman (rank) correlation between the stringency index and infection counts for each virus and noted the corresponding p-values. The results in Fig. 3B show that many viruses, such as the RSV and FLUA(H3N2), exhibit a moderate yet significant negative correlation between infection counts and NPIs. However, this is not true for all considered viruses. Most notably, influenza B and A H1N1 have correlations of nearly zero, which is due to very small absolute infection numbers in the considered time period. As the SARS-CoV-2 pandemic reached Germany in March 2020 the first two months of 2020 were excluded for this analysis. If they were taken into account as well, correlations would be higher (Supplement). We also studied Pearson correlations among stringency index and infection counts (Supplement); here the resulting values are smaller in absolute numbers, which indicates that the relationship between NPIs and infection counts is to some degree nonlinear. We additionally repeated all correlation analyses with relative instead of absolute infection counts (Supplement). The results remain almost identical.
We also studied the relationship between the disruption statistics and NPIs (Supplement); here the correlations are even stronger, and there is little difference between Pearson and Spearman correlation.
## Coinfections
We inspected coinfections among viruses for the provided study period and compared coinfection counts before (January 2010 to Feburary 2020) and after the start of the pandemic (March 2020 to April 2024) in a heatmap (Fig. 4). The general pattern appears fairly similar. Frequent coinfections, such as RSV and RV, persist. Most changes pertain human Bocavirus (HBoV) where coinfections became more frequent, in particular with RSV, HadV, and HMPV. Also HMPV/RV coinfections are occurring frequently. The only notable coinfections of SARS-CoV-2 reported in the CNV database are with RSV, RV, and both influenza A subtypes. The numbers are tiny in relation to the SARS-CoV-2 monoinfections, though (Supplement).
## Discussion
In this study, we performed a long-term analysis of respiratory viruses in Germany. We quantified seasonal shifts and compared the behaviour of 17 different viruses using self-developed seasonality disruption statistics.
All viruses showed a lower frequency during the pandemic. The reemergence of the viruses was different. We showed that for viruses that exhibit a specific seasonal outbreak pattern the pandemic had severe impact on the timing and the size of the outbreak waves. RSV constitutes a prime example as the disruption led to a missing wave in the 2020/2021 season followed by a reemergence in fall 2021. This could be interpreted either as a very delayed 2021 infection wave or an advanced 2022 wave. This is a unique behaviour among the considered viruses, further supported by its distinct disruption profile being least similar to any other virus. Therefore, it is plausible that it displays the highest SDI value.
Papenburg and Boivin [37] described different seasonal trends of RSV and HMPV, despite their genetic closeness and causing similar symptoms. Our study emphasizes this by not only showing different prepandemic seasonal trends but also various disruption statistics upon the COVID-19 pandemic onset. Furthermore, our results are in line with Terliesner et al. [14] who suggest that non-seasonal viruses (Rhinovirus/Enterovirus, Adenovirus) remained comparatively more stable upon the COVID-19 outbreak, while others showed out-of-season resurgences. The same holds true for the also closely phylogenetically related Parainfluenza viruses. RSV, HMPV and Parainfluenza belong to the family of Paramyxoviruses but clearly show a different biological behavior as demonstrated here.
Additionally, enveloped viruses such as RSV, FLUA(H3N2) and FLUA(H1N1) exhibit a fluctuating disruption profile and thus a relatively high SDI, whereas non-enveloped viruses, e.g. Rhinovirus or Adenovirus, show none to less seasonal behaviour and thus they show relatively little seasonal disruption. This is in line with Oh et al. [21], who suggests that the viral structure (enveloped versus non-enveloped viruses) might influence viruses dynamics during and after the SARS-CoV-2 prevention measures.
For Influenza A and B the disruption was also very prominent, as the disruption was even longer as compared to RSV and not directly related to the non-pharmaceutical-intervention index. There seems to be a relation as the more stringent the seasonality is, the longer the disruption period and the more difficult it is for the virus to get back into the pre-pandemic seasonality. Due to the vulnerability of the envelope of influenza viruses, they are very vulnerable to environmental influences. If this or other reasons like the reproduction number determine the differences in seasonality can only be speculated as we do not have such information collected in our network.
There is indeed a difference between the spreading of Influenza A H1N1, A H3N2 and the B variants (Yamagata and Victoria line). Different epidemiology between the different influenza variants is clearly visible and is verified by other networks like the Arbeitsgemeinschaft Influenza (AGI, www.RKI.de). Before the SARS-CoV-2 pandemic, different years sometimes had different epidemics showing different variants of influenza A and B. As we do not observe Influenza B Yamagata line derived virus strains after the pandemic, authorities discuss to exclude influenza B Yamagata line like virus vaccine from the recommended vaccine recommendation.
We were able to show statistically significant negative correlations between the case numbers and the implementation of non-pharmaceutical interventions for some viruses. The partially nonlinear correlation could be explained by an exponential increase in the infection counts and indeed Influenza A (H3N2), which has the largest difference between linear and rank correlation, has also the sharpest infection peak. Given this observation the nearly linear correlation between disruption statistics and NPIs is also plausible, as the former is defined on the logarithmic scale.
Aside from SARS-CoV-2, RSV is noted as the primary example for negative correlation between infection counts and NPIs. Yorsaeng et al. [18] could not identify a similar significant increase after implementation of NPIs as others have identified. However, not all viruses show this strong correlation. For some, such as Influenza B and Influenza A, subtype H1N1, it is due to virtually zero infections in the time period under consideration.
We analyzed a composite NPI index with disease incidence. Although individual NPIs, such as mask-wearing, social distancing, or travel restrictions, may vary in their impact, they are often implemented together. As a result, NPIs tend to be highly correlated and may exert synergistic or confounding effects, making it challenging to isolate the specific contribution of any single intervention [22]. Different studies aimed at disentangling these effects. Takeuchi et al. [24] used multiple regression analysis identifying associations between high mask use and high social distancing in 3 and 2 seasons with influenza. In contrast to the individual NPIs within the stringency index, they opted for more granular data for mask use and mobility data. Billard et al. [16] identified associations between school closures and stay-at-home orders in the RSV season using linear mixed models, though effect sizes were small.
We observed that coinfections occurred between different viruses and that this was relatively little affected by the pandemic. Most increases in coinfections are related to HBoV. It is discussed controversially whether HBoV is a mere bystander [38], a true pathogen in its own right [39] or coinfection with HBoV might lead to a more severe course of disease compared to single pathogen infections. We have no detailed clinical data in our network to either support clinical significance of HBOV or to prove the opposite, though. Other coinfection increases pertain RSV, which can be a consequence of the strong RSV wave in fall 2021.
Although our study is fairly comprehensive in terms of the long time period, number of viruses, and contributing hospitals, it has certain limitations. The heterogeneous nature of the CVN data may introduce biases as outlined in the following paragraphs.
Testing policies are not constant across time and location. As a consequence, neither absolute nor relative infection counts are free of bias. The former may overestimate infections and lead to false positive peaks in the infection dynamics with increased testing. Conversely, the latter may underestimate infections and lead to the missing of peaks in the infection dynamic in that situation. Since both forms of bias are possibly problematic, we display long-term trends for both statistics. We used absolute infection counts as basis for our main analysis, since our motivation was to study the possible absence of expected infection peaks post-pandemic onset, for which underestimation of infections is a greater concern than overestimation.
The analysed data set stemmed from the clinical virology network mainly comprising hospital data. Thus, the claims made in this study refers to this specific setting only. Tanislav et al. [40] analysed data from general practitioners and specialists and have shown a decrease in respiratory and gastrointestinal disease occurrence during the pandemic. However, no stratification by virus was performed. Further insight into the viral dynamics regarding surveillance detected by general practitioners would be interesting as a complementing comparison to the presented results. Similarly, there might be agerelated differences in susceptibility or testing bias within the hospital setting.
A further limitation of our study include that the data are restricted to Germany. This may limit the generalizability of our findings. Viral dynamics and the effectiveness of public health measures differ significantly between countries due to variations in health care systems, political decisions, and societal behaviors. As such, our results may not be fully applicable to settings outside the German context. Moreover, pandemic response strategies, including testing regimes, mobility restrictions, and vaccination rollouts, varied widely across countries and over time. These differences make it difficult to draw direct comparisons or apply our findings to other regions without conducting localized analyses.
The disruption statistics and the seasonal disruption index are novel metrics introduced specifically for this study. One limitation is their dependence on the three hyperparameters w, d, and α. Changing one or more of these values drastically will alter the results. While we consider the precise values used to provide a reasonable tradeoff between capturing as much information as possible while also eliminating noise on our CVN data, this does not necessarily generalize to other data sets. A systematic study of the effect of these parameters across different data sets could thus be a topic for future research.
## Conclusion
We provide a basis for data collection and a deeper understanding of virus dynamics and how non-pharmaceutical interventions affect the seasonality and occurrence of viruses is essential for preparing for upcoming seasons. Furthermore, this information is crucial for health policy, as it helps refine or develop new strategies to combat and predict seasonal peaks, as well as to improve diagnostics. Consequently, the challenges of new epidemics and even pandemics can be met.
## References
1. Anton, Bartling, Christian et al. "Jan Philipp Jung"
2. Hauka, Meyer, Ziegler *Uwe Gerd Liebert*
3. Dr, Krause Und Kollegen, Gmbh et al. "Germany [3] SYNLAB Medizinisches Versorgungszentrum Weiden GmbH, Weiden in der Oberpfalz"
4. Groningen "Netherlands [15] Medizinisches Versorgungszentrum (MVZ) Labor Dr. Quade & Kollegen GmbH [16] Labor Krone GbR -Medizinal-Untersuchungsstelle im Regierungsbezirk Detmold"
5. *MVZ Institut für Medizinische Mikrobiologie, Infektiologie, Hygiene und Tropenmedizin GmbH*
6. Graz
7. Zentrum, Aaachen "Medizinisches Versorgungszentrum Dr. Eberhard & Partner Dortmund"
8. Synlab Mvz Leverkusen
9. Petrova, Russell (2018) "The evolution of seasonal influenza viruses" *Nat Rev Microbiol*
10. Piret, Boivin (2021) *Pandemics throughout history. Front Microbiol*
11. Huremović (2019) "Brief history of pandemics (pandemics throughout history)"
12. Cherry, Krogstad (2004) "SARS: the first pandemic of the 21st century" *Pediatr Res*
13. Nicola, Alsafi, Sohrabi et al. (2020) "The socioeconomic implications of the coronavirus pandemic (COVID-19): A review" *Int J Surg*
14. Williams, Jones, Welch et al. (2023) "Outlook of pandemic preparedness in a post-COVID-19 world" *NPJ Vaccines*
15. Tang, Lam, Zaraket et al. (2017) "Global epidemiology of non-influenza RNA respiratory viruses: data gaps and a growing need for surveillance" *Lancet Infect Dis*
16. Naguib, Ellström, Järhult et al. (2020) "Towards pandemic preparedness beyond COVID-19" *Lancet Microbe*
17. Horemheb-Rubio, Eggeling, Schmeiβer et al. (2022) "Respiratory viruses dynamics and interactions: ten years of surveillance in central Europe"
18. Shirreff, Chaves, Coudeville et al. (2010) "Seasonality and Co-Detection of Respiratory Viral Infections Among Hospitalised Patients Admitted With Acute Respiratory Illness-Valencia Region"
19. Moriyama, Hugentobler, Iwasaki (2020) "Seasonality of respiratory viral infections" *Ann Rev Virol*
20. Estrada, Schultz-Cherry (2019) "Development of a universal influenza vaccine" *J Immunol*
21. Calvo, García-García, Pozo et al. (2016) "Infections and coinfections by respiratory human bocavirus during eight seasons in hospitalized children" *J Med Virol*
22. Terliesner, Unterwalder, Edelmann et al. (2022) "Viral infections in hospitalized children in Germany during the COVID-19 pandemic: Association with non-pharmaceutical interventions" *Front Pediatr*
23. Hönemann, Thiem, Bergs et al. (2023) "Indepth analysis of the re-emergence of respiratory syncytial virus at a tertiary Care Hospital in Germany in the summer of 2021 after the alleviation of non-pharmaceutical interventions due to the SARS-CoV-2 pandemic" *Viruses*
24. Billard, Van De Ven, Baraldi et al. (2022) "Wildenbeest JG. International changes in respiratory syncytial virus (RSV) epidemiology during the COVID-19 pandemic: association with school closures. Influenza Other Respir Viruses"
25. Abo, Clifford, Lee et al. (2021) "COVID-19 public health measures and respiratory viruses in children in Melbourne" *J Paediatr Child Health*
26. Yorsaeng, Suntronwong, Thongpan et al. (2020) "The impact of COVID-19 and control measures on public health in Thailand" *PeerJ*
27. Klee, Diexer, Horn et al. (2024) "The impact of nonpharmaceutical interventions on community non-SARS-CoV-2 respiratory infections in preschool children" *BMC Pediatr*
28. Van De Berg, Charles, Dörre et al. (2023) "Epidemiology of common infectious diseases before and during the COVID-19 pandemic in Bavaria, Germany, 2016 to 2021: an analysis of routine surveillance data" *Eurosurveillance*
29. Oh, Buda, Biere et al. (2020) "Trends in respiratory virus circulation following COVID-19-targeted nonpharmaceutical interventions in Germany"
30. Kaiser, Otterbach, Sousa-Poza et al. (2020) "Interventions with Positive Side-Effects: COVID-19 Non-Pharmaceutical Interventions and Infectious Diseases in Europe"
31. Koltai, Krauer, Hodgson et al. (2022) "Determinants of RSV epidemiology following suppression through pandemic contact restrictions" *Epidemics*
32. Takeuchi, Kawashima (2023) "Disappearance and re-emergence of influenza during the COVID-19 Pandemic: Association with infection control measures" *Viruses*
33. Guan, Chen, Li et al. (2021) "Impact of coinfection with SARS-CoV-2 and influenza on disease severity: a systematic review and metaanalysis"
34. Musuuza, Watson, Parmasad et al. (2021) "Prevalence and outcomes of co-infection and superinfection with SARS-CoV-2 and other pathogens: a systematic review and meta-analysis" *PloS ONE*
35. Hale, Angrist, Goldszmidt et al. (2021) "A global panel database of pandemic policies (Oxford COVID-19 Government Response Tracker)" *Nat Hum Behav*
36. Moline, Toepfer, Tannis et al. (2025) "Respiratory Syncytial Virus Disease Burden and Nirsevimab Effectiveness in Young Children From 2023-2024" *JAMA Pediatr*
37. Walter, Munoz (2024) "New Approaches to Respiratory Syncytial Virus Prevention and Treatment" *Ann Rev Med*
38. Wang, Li, Knoll et al. (2021) "Global burden of acute lower respiratory infection associated with human metapneumovirus in children under 5 years in 2018: a systematic review and modelling study" *Lancet Glob Health*
39. Jain, Williams, Arnold et al. (2015) "Community-acquired pneumonia requiring hospitalization among US children" *N Engl J Med*
40. Kurai, Natori, Yamada et al. (2022) "Occurrence and disease burden of respiratory syncytial virus and other respiratory pathogens in adults aged ≥ 65 years in community: a prospective cohort study in Japan. Influenza Other Respir Viruses"
41. Philippot, Rammaert, Dauriat et al. (2024) "Human metapneumovirus infection is associated with a substantial morbidity and mortality burden in adult inpatients" *Heliyon*
42. Falsey, Walsh, House et al. "Risk factors and medical resource utilization of respiratory syncytial virus, human metapneumovirus, and influenza-related hospitalizations in adults-A global study during the 2017-2019 epidemic seasons (hospitalized acute respiratory tract infection [HARTI] study)" *Open Forum Infectious Diseases*
43. Ma, Zhu, Xu et al. (2024) "Development of a novel multi-epitope mRNA vaccine candidate to combat HMPV virus. Hum Vaccines Immunotherapeutics"
44. Daungsupawong, Wiwanitkit (2024) "Multi-epitope mRNA vaccine candidate to combat HMPV virus: Comment. Hum Vaccines Immunotherapeutics"
45. Papenburg, Boivin (2010) "The distinguishing features of human metapneumovirus and respiratory syncytial virus" *Rev Med Virol*
46. Martin, Taylor, Kuypers et al. (2009) "Detection of bocavirus in saliva of children with and without respiratory illness" *J Clin Microbiol*
47. Mohanty, Mishra, Satapathy et al. (2023) "Human Bocavirus infection in childhood acute respiratory infection: Is it an innocent bystander?" *Indian J Med Microbiol*
48. Tanislav, Kostev (2022) "Fewer non-COVID-19 respiratory tract infections and gastrointestinal infections during the COVID-19 pandemic" *J Med Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12300617&blobtype=pdf | # Complete-Genome Analysis of Echovirus-30 Isolated from an Encephalitis Case in India Revealed Distinct Mutations
Rishabh Waghchaure, Jithin Kunjumon, Alfia Ashraf, Ranjana Raju, Anita Shete, Sarah Cherian, Mallika Lavania
## Abstract
Isolated from an Encephalitis Case inIndia Revealed Distinct Mutations.
## 1. Introduction
Enteroviruses are part of the Enterovirus genus within the Picornaviridae family. The major enterovirus groups associated with human illnesses include EV-A, EV-B, EV-C, and EV-D, with E-30 falling under the EV-B group. E-30 has been extensively studied and documented on a global scale, showing strong links to various infectious conditions such as encephalitis, myocarditis, and hand, foot, and mouth disease. Numerous retrospective studies analyzing clinical samples such as feces, serum, and cerebrospinal fluid from various countries have consistently identified E-30 as a leading cause of aseptic meningitis outbreaks in Asia, Europe, and the Americas in recent times [1][2][3][4].
Enteroviruses, particularly Echovirus 30 (E-30), are well known for their recurrent outbreaks, which pose significant public health challenges [5][6][7][8]. E-30 is a non-enveloped virus with a single-stranded, positive-sense RNA genome of approximately 7.4 kilobases. The genome comprises a single open reading frame (ORF) flanked by 5 ′ and 3 ′ untranslated regions (UTRs). This ORF encodes a large polyprotein, which is subsequently cleaved into structural proteins (VP1-VP4) and non-structural proteins. Among these, the VP1 protein is of particular importance for molecular epidemiology, as it contains critical neutralization epitopes and serves as the primary region used for the serotyping and genotyping of enteroviruses [9][10][11].
E-30 outbreaks tend to occur in cyclical patterns, typically peaking every 3 to 5 years, especially in temperate regions [12]. These periodic surges are believed to result from a combination of declining population immunity, viral evolution through mutation and recombination, and favorable environmental conditions that enhance transmission. In China, surveillance studies have identified E-30, alongside EV-A71 and CV-B5, as major contributors to viral meningitis and hand, foot, and mouth disease (HFMD) in children. A notable outbreak in 2014 in Shandong Province saw numerous children presenting with clinical features typical of viral meningitis, such as fever, vomiting, photophobia, and neck stiffness, raising significant health concerns [9]. E-30 continues to be a pressing concern in Europe as well. In 2018, the European Centre for Disease Prevention and Control (ECDC) reported a marked increase in E-30 cases across multiple countries, including Denmark, Germany, and the Netherlands. This prompted detailed molecular investigations, which revealed that many of the circulating strains belonged to newly emerging genogroup VI lineages, indicating active viral evolution and the emergence of novel variants [5]. These findings underscore the importance of continuous molecular surveillance to detect and respond to evolving strains that may lead to widespread outbreaks.
Despite its clinical significance, there is currently no targeted antiviral therapy or approved vaccine for E-30. As a result, clinical management remains largely supportive, focusing on symptom relief and preventing complications. Given the virus's ability to cause CNS infections and its unpredictable outbreak behavior, comprehensive molecular characterization including whole-genome sequencing and phylogenetic analysis is essential. These approaches help clarify transmission dynamics, identify potential recombination events, and detect genetic markers associated with virulence and neurotropism. Moreover, recombination is widely recognized as a major driver of enteroviral genetic diversity, allowing viruses to adapt to new hosts and environmental conditions [11,13].
Against this backdrop, our study provides important insights through the isolation and analysis of an E-30 strain from the cerebrospinal fluid (CSF) of a child diagnosed with aseptic meningitis. By conducting whole-genome sequencing and comparative phylogenetic analysis, we aimed to better understand the virus's genetic makeup, its evolutionary context, and possible recombination patterns. These findings are crucial not only for advancing our understanding of E-30 pathogenesis but also for informing clinical management and strengthening public health surveillance and response strategies in affected regions.
As a designated reference laboratory for viral diagnostics, our department received samples from a two-year-old child in Kerala, India. The child had been experiencing fever for five days along with episodes of seizures. Symptoms included lethargy, irritability, and abnormal movements. Based on clinical assessment and laboratory findings, the case was diagnosed as viral encephalitis. Prior to the illness, the child was healthy, fully vaccinated, and had achieved all age-appropriate developmental milestones.
## 2. Materials and Methods
The clinical specimens including oropharyngeal/nasopharyngeal swabs, CSF, stool and serum, were referred to the laboratory in cold chain to rule out the viral etiology of acute encephalitis syndrome in the patient.
## 2.1. Ethical Clearance
This study was reviewed and approved by the Ethics Committee of the ICMR-National Institute of Virology, Pune (MP-24A-7N), in accordance with established ethical guidelines for biomedical research involving human samples.
## 2.2. Virus Isolation
The isolation of the virus was performed in the rhabdomyosarcoma (RD) cell line following WHO protocols [14][15][16][17][18] for virus characterization. RD cells are large, multinucleated, spindle-shaped cells originally isolated from the muscle tissue of a 7-year-old female patient with pelvic rhabdomyosarcoma refractory to cyclophosphamide and radiation therapy. The RD cell line (ATCC CCL-136) was obtained from the American Type Culture Collection and maintained under standard culture conditions as previously described [18]. RNA was extracted from the cell culture supernatant using a Qiagen kit and tested by qRT-PCR to confirm the presence of enterovirus. The sample tested positive for pan-enterovirus by qRT-PCR with a Ct value of 25. Genotyping was confirmed using semi-nested RT-PCR targeting the VP1 region [19]. Sequence identity was determined through a Nucleotide BLAST search (https://blast.ncbi.nlm.nih.gov/blast/Blast.cgi?PROGRAM=blastn&PAGE_ TYPE=BlastSearch&LINK_LOC=blasthome accessed on 21 April 2025).
## 2.3. Reverse Transcription Polymerase Chain Reaction (RT-PCR), Sequencing, and Typing
RT-PCR, sequencing, and typing were performed following a previously described protocol [20]. Viral RNA was extracted from the supernatants of infected cells using the Body Fluid Viral DNA/RNA Miniprep Kit (Axygen, Union City, CA, USA). RT-PCR was conducted with the PrimeScript One Step RT-PCR Kit Ver.2 (DSS Takara Bio India Pvt. Ltd., New Delhi, India).
To amplify partial VP1 sequences, the primers AN89 and AN88 were used [20]. Complete-genome fragments were amplified and sequenced using multiple primer pairs, as summarized in Additional File 1 (Table S1), on an ABI PRISM 3100 Genetic Analyzer (Applied Biosystems, Waltham, MA, USA). The positive amplification products were sequenced by Tsingke Biological Technology Co., Ltd. (Kunming, China). Enterovirus classification was performed using the Enterovirus Genotyping Tool Version 2.17.
VP1-encoding sequences and complete genomes were compared with publicly available sequences in GenBank using BLAST (https://blast.ncbi.nlm.nih.gov/Blast. cgi?PROGRAM=blastn&PAGE_TYPE=BlastSearch&LINK_LOC=blasthome assessed on 21 April 2025) [21]. The assembly parameters were set at the default values. The algorithm that was best suited based on the sample size was chosen using the auto command.
## 2.4. Whole Genome Assembly and Phylogeny
The whole genome of the isolate was sequenced using the random primer strategy, which utilized the TruSeq Stranded mRNA LT Library Preparation Kit on the Illumina Miniseq platform. For the whole-genome sequencing (WGS) of the positive-sense singlestranded RNA virus Echovirus 30, the TruSeq Stranded mRNA LT Library Preparation Kit (Illumina) was adapted accordingly. Total RNA extracted from the supernatant of virus-infected cell cultures was subjected to random fragmentation, followed by reverse transcription using random primers to achieve comprehensive genome coverage. Strand specificity was maintained by incorporating dUTP during second-strand synthesis, allowing the selective degradation of the second strand. The resulting double-stranded cDNA was then end-repaired, adenylated at the 3 ′ ends, and ligated to indexed sequencing adapters. After library construction, PCR amplification was performed, and the libraries were subsequently quantified and evaluated for quality before being sequenced on an Illumina platform. After RNA library preparation and normalization, the library was loaded onto the platform, and FASTQ data were analyzed with CLC Genomics Workbench version 20, employing de novo assembly and reference mapping.
For the molecular epidemiology study, we retrieved VP1 and 110 whole-genome sequences from GenBank and selected 52 representative strains for genotyping. The aligned sequences were then subjected to phylogenetic reconstruction using the maximum likelihood (ML) approach, implemented via IQ-TREE v2.2.0. The best-fit substitution model was automatically selected by IQ-TREE based on the Bayesian information criterion (BIC). Tree robustness was assessed using 1000 ultrafast bootstrap replicates. All parameters were kept at their default settings unless specified. The final tree was visualized and annotated using Interactive Tree of Life (iTOL version 6).
Phylogenetic analysis of VP1 was conducted using MEGA software version 7, calculating genetic distances using the Kimura two-parameter model [22,23]. The full genome sequence of the Indian isolate (passage 4) of E-30 has been deposited in GenBank under the accession number PQ472410. An initial phylogenetic tree based on neighbor joining (NJ) was constructed based on the VP1 gene to verify the serotype and lineage assignment of the isolate. Given the high resolution required for evolutionary clustering, a subsequent maximum likelihood phylogeny using full genome sequences (n = 110) was conducted to refine the placement of PQ472410.1 within the global Echovirus 30 landscape. The wholegenome IQ-TREE phylogenetic tree forms the basis of all final analyses, interpretations, and conclusions presented in this study.
## 2.5. Mutation Analysis Based on Whole-Genome Sequencing
A comprehensive mutation analysis at the amino acid level was performed on all available whole-genome sequences from 2005 to 2019, using the prototype strain AF162711 (Bastianni strain) as a reference. Nucleotide sequences were aligned with the prototype strain using MAFFT v7.5266 [24] to identify nucleotide substitutions. The aligned FASTA file was then analyzed using an in-house Python v3.10.12 pipeline to detect mutations across the polypeptide regions in a tabular format. To validate the identified mutations, we employed MEGA11, where the aligned FASTA file was translated to assess specific amino acid changes.
## 3. Results
## 3.1. Isolation and Molecular Characterization of Virus
Virus isolation and characterization were carried out using the RD (rhabdomyosarcoma) cell line, which was cultured under standard laboratory conditions. The cells were inoculated and monitored over four successive blind passages (P1 to P4), performed in duplicate. Consistent viral replication was observed throughout the passages, with marked cytopathic effects (CPEs) appearing from day 4 post-inoculation (Figure 1A,B).
The observed CPEs were typical of enteroviral infection and included cellular rounding, aggregation into dense clumps, increased cytoplasmic granularity suggestive of inclusion body formation, and detachment from the culture substrate. The extent of cellular damage intensified with each passage, and monolayers showing over 90-95% CPEs were selected for downstream analysis.
Viral lysates from passage 4 (P4) cultures demonstrating maximum CPEs were harvested for molecular confirmation. RT-qPCR targeting the conserved pan-enterovirus region confirmed the presence of enteroviral RNA in the samples (Figure 1C). Further verification using VP1 gene-specific RT-PCR also yielded a positive result, indicating the successful amplification of the VP1 region, which is critical for enterovirus identification and serotyping. These findings confirm that the virus was successfully propagated in RD cells and reliably identified as an enterovirus. The reproducibility of CPEs, coupled with the molecular detection of viral RNA in both pan-enterovirus and VP1-specific assays, strongly supports the authenticity and integrity of the viral isolate.
## 3.2. Whole-Genome Sequencing and Phylogenetic Analysis
The further positive passages were further investigated for full-genome sequencing. Full-genome sequencing was performed and phylogenetic analyses of this isolate revealed a genotype distributed to E-30 (Enterovirus B) [Figure 2]. The evolutionary history was inferred using the maximum likelihood (ML) method. The whole-genome sequences of the isolate (four different passages) isolated in India in 2023 were determined. These sequences ranged from 7403 to 7426 nucleotides in length and contained an open reading frame (ORF) of 6585 nucleotides, encoding a polyprotein of 2194 amino acids. The ORF was flanked by a noncoding 5 ′ -UTR of 739-754 nucleotides and a noncoding 3 ′ -UTR of 91-113 nucleotides. The whole-genome sequences showed 99.4-99.7% identity at the nucleotide level.
A comprehensive phylogenetic analysis was carried out for the E-30 isolate PQ472410.1 (passage 4) to explore its evolutionary relationships and genetic proximity to globally circulating Echovirus 30 strains. The analysis utilized the complete VP1 coding sequence of the isolate, comparing it with 110 globally representative E-30 sequences across diverse geographical regions and timeframes, with a focus on identifying unique amino acid variations and lineage clustering.
The phylogenetic tree was constructed using the maximum likelihood (ML) method in IQ-TREE, which revealed the clear stratification of global E-30 strains into multiple well-supported clades. The Indian isolate PQ472410.1 was positioned within Clade I, a strongly supported cluster that included 2023 isolates from Nepal (PP621689.1, PP461524.1) and strains from the USA (OQ791513.1, OQ791516.1), Spain (MZ389231.1, MZ389230.1), and the Netherlands (MK815082.1, MK815083.1, MK815095.1, MK815087.1, MK815088.1, MK815090.1, MK815096.1). This grouping received robust bootstrap support exceeding 90%, confirming a high degree of phylogenetic reliability. Of particular interest is that the Indian isolate showed the greatest genetic similarity to a Nepalese strain from 2023, supported by a bootstrap value of 100%. This close relationship suggests a recent common ancestor and points to either shared transmission pathways or concurrent regional evolution in the South Asian context. The extended internal branch leading to the Indian isolate indicates rapid divergence, potentially driven by local selective pressures or undetected circulation in the region prior to detection.
The prototype E-30 strain "Bastianni" occupied a basal position in the tree, serving as an ancestral reference point. This further supports the recent diversification of the Nepal-India sublineage and emphasizes its evolutionary distinction from earlier European (2016-2018), Chinese (2016-2019), and North American (2009-2017) E-30 clusters.
These findings point to the emergence of a genetically distinct and locally evolving E-30 lineage in South Asia. The close clustering of recent South Asian and European isolates underscores the dynamic nature of E-30 circulation and the potential for international spread. Importantly, this analysis highlights the critical need for sustained molecular surveillance and genomic monitoring in the region, especially in areas with limited reporting of enterovirus-related CNS infections. The early detection and characterization of emerging lineages are essential for timely outbreak response and informing public health interventions.
## 3.3. Mutation Analysis Based on Whole Genome Sequencing
Mutation analysis with the complete-genome sequences was performed against the prototype "Bastianni" strain AF162711.1. The comparative genomic investigation based on the dataset of 111 human E-30 isolates encompassed a total of 116,991 mutation records. Further, the mutation analysis revealed a total of six non-synonymous mutations across various genomic regions of the E-30 Indian isolate (Table 1), highlighting the ongoing evolution and adaptation of the virus. We attempted to correlate the six non-synonymous substitutions to possible effects on protein structure or function. Strikingly, the RNA-dependent RNA polymerase (3D) accounts for 50% of these changes, followed by the major capsid protein (VP1-33.33%) and the protease (2A-16.66%). VP1-V43I and VP1-P258L occur in the capsid protein VP1, which plays a key role in capsid stability. In the VP1 capsid protein, the V43I mutation lies within the N-terminal region but outside the canonical BC loop (residues 80-90) implicated in receptor binding [25].
Although not directly within a known antigenic loop, the N-terminal extensions and Cterminal arms of capsid proteins play essential roles in maintaining capsid stability [26]. The P258L mutation, located near the C-terminus of VP1, may similarly affect capsid assembly and viral infectivity, as surface-exposed loops and terminal regions are known to influence receptor attachment and structural integrity. The T60A mutation in the 2A protease lies outside the catalytic triad but may affect enzymatic function, as structurally analogous residues such as V59 and H68 have been shown to be critical for interactions with host factors like SETD3 and for viral replication [25]. Three substitutions were detected in the 3D RNA-dependent RNA polymerase: K22R, H260L, and E431A. The K22R mutation falls within the N-terminal region (residues 1-30), which stabilizes the palm domain and is essential for overall polymerase architecture [27]. The H260L mutation is located in the palm domain, which houses catalytic motifs A-E and governs RNA synthesis. The E431A mutation lies within the thumb domain, which maintains interactions with the template-primer duplex during RNA replication [28].
## 4. Discussion
Enteroviruses are a diverse group of viruses, known to cause a wide array of clinical manifestations, ranging from mild conditions like hand, foot, and mouth disease to more severe diseases such as aseptic meningitis, encephalitis, paralysis, neonatal sepsis-like disease, myocarditis, respiratory infections, and acute hemorrhagic conjunctivitis. Despite their significant impact on public health, the full extent of enterovirus distribution and the associated disease burden remain poorly understood, particularly in regions like Europe and India. This is partly due to inconsistent surveillance systems across different countries, leading to gaps in the understanding of enterovirus epidemiology and the potential risks they pose.
Aseptic meningitis, one of the most severe manifestations of enterovirus infections, is commonly associated with E-30. This virus has been implicated in numerous outbreaks worldwide, especially in Europe and Asia, where it has been a significant cause of viral meningitis. Studies from European countries, notably in 2018, reported an uptick in E30 infections, suggesting the increased circulation of this virus. The higher positivity rates of E-30 detected in enteroviral samples from multiple nations compared to those in previous years highlighted the virus's growing prevalence and its potential to cause large-scale outbreaks [29]. In India, however, systematic surveillance of enterovirus infections remains insufficient, making it difficult to assess the true burden of these infections. A study performed by Mann et al. [30] based on nonpolio enterovirus associated with nonpolioacute flaccid paralysis in Northern India, identified different clusters co-circulating in India.
The last significant outbreak of E-30 in India occurred in 2004, during which a rise in viral meningitis cases was noted, and E-30 was identified as the primary causative agent [31]. Despite these occurrences, the lack of continuous surveillance means that enteroviral infections, including E-30 infections, remain largely underreported. This study presents the first full-genome sequence of E-30 from India, which belongs to Clade I, a strain that is also circulating in Europe (Spain and the Netherlands) and Nepal. This finding is significant as it not only provides genetic insights into the local strain but also contributes to global surveillance efforts aimed at tracking the movement and mutation of enteroviruses.
The whole-genome sequencing of the Echovirus 30 (E-30) strain isolated from a child with encephalitis in Kerala revealed six unique non-synonymous amino acid substitutions not observed in other global isolates. These mutations were located within three functionally crucial viral proteins: the RNA-dependent RNA polymerase (3D), the protease (2A), and the major capsid protein (VP1). Each of these proteins plays a pivotal role in the viral life cycle. 3D is essential for viral RNA synthesis [32], 2A facilitates polyprotein cleavage and the disruption of host cell translation [33], and VP1 mediates receptor binding and immune recognition [34].
Functional annotations and structural modeling suggest that these mutations could influence viral pathogenicity, particularly neurovirulence. Changes in VP1 may affect viral entry and host immune escape, while alterations in 3D and 2A could modulate replication dynamics and host interaction. The exclusive presence of these mutations in a case of encephalitis rather than the more common clinical presentation of aseptic meningitis raises the possibility of a genetic basis for CNS invasion. Although the precise role of these mutations requires confirmation through laboratory models, their potential involvement in enhancing neurotropism is noteworthy.
These findings identify candidate molecular markers that may be linked to virulence, providing a basis for future research into antiviral targets or diagnostic tools. The identification of novel mutations also exposes a broader issue: the lack of comprehensive genomic surveillance of enteroviruses in India and many other resource-limited settings. This genome sequence contributes significantly to the limited database of Indian E-30 isolates and offers a regional perspective on viral diversity and evolution. Given the periodic nature and international spread of enterovirus outbreaks, the inclusion of such data is critical to global surveillance efforts.
In an era of increased global mobility and climate-driven shifts in disease patterns, the real-time tracking of viral evolution is essential. This study adds to the molecular epidemiology of E-30 in South Asia and emphasizes the importance of early detection systems capable of identifying genetic markers linked to severe disease outcomes.
Furthermore, the findings underscore the urgency of strengthening molecular surveillance infrastructure in countries with historically limited diagnostic capacity. Integrating genomic tools such as next-generation sequencing and mutation analysis into routine public health practice can enhance outbreak preparedness and response. The timely identification of emerging variants could enable the rapid implementation of containment strategies, clinical triage protocols, and risk communication.
In conclusion, this work provides a detailed view of a potentially neurovirulent E-30 strain, serving as a bridge between clinical virology, molecular epidemiology, and public health action. It reinforces the essential role of genomic surveillance in recognizing and managing infectious threats in an interconnected world.
## 5. Conclusions
Comprehensive mutational and phylogenetic analyses revealed that the Echovirus 30 (E-30) isolate associated with the 2023 outbreak belongs to a genetically distinct subclade linked to Nepal. This variant is characterized by a cluster of non-synonymous mutations concentrated in essential viral proteins, notably the RNA-dependent RNA polymerase (3Dpol) and the capsid protein VP1. Among these, six amino acid substitutions appear to be unique to this strain, indicating possible adaptive evolution that could affect viral replication, immune escape mechanisms, and tissue specificity. The presence of these mutations suggests a distinct evolutionary pathway, potentially shaped by localized environmental or host-related selective pressures. These molecular changes may influence the virus's antigenic profile and raise concerns regarding cross-protection by existing vaccines. Overall, the findings underscore the importance of sustained genomic monitoring and functional studies to assess the biological implications of emerging variants. Such efforts are essential for predicting outbreak potential, informing the development of diagnostics and therapeutics, and guiding public health strategies tailored to regional viral evolution.
## References
1. Benschop, Broberg, Hodcroft et al. (2021) "Molecular epidemiology and evolutionary trajectory of emerging Echovirus 30" *Europe. Emerg. Infect. Dis*
2. Sousa, Burlandy, Lima et al. (2016) "Echovirus 30 detection in an outbreak of acute myal-gia and rhabdomyolysis" *Clin. Microbiol. Infect*
3. Broberg, Simone, Jansa (2018) "Upsurge in echovirus 30 detections in five EU/EEA countries" *Eurosurveillance*
4. Lee, Seo, Choi et al. (2017) "Seroepidemiology of echovirus 30 in Korean children" *World J. Pediatr.: WJP*
5. (2018) "Enterovirus Detections Associated with Neurological Symptoms in Several EU/EEA Countries"
6. Xiao, Guan, Chen et al. (2013) "Molecular characterization of echovirus 30-associated outbreak of aseptic meningitis in Guangdong in 2012" *Virol. J*
7. Zheng, Ye, Yan et al. (2016) "Laboratory diagnosis and genetic analysis of a family clustering outbreak of aseptic meningitis due to echovirus 30" *Pathog. Glob. Health*
8. Simoes, Hodcroft, Simmonds et al. "Epidemiological and Clinical Insights into the Enterovirus D68 Upsurge in Europe 2021-2022 and Emergence of Novel B3-Derived Lineages, ENPEN Multicentre Study" *J*
9. Tao, Wang, Li et al. (2006) "Molecular epidemiology of enteroviruses associated with aseptic meningitis in Shandong Province" *PLoS ONE*
10. Tan, Huang, Zhu et al. (2013) "The persistent circulation of enterovirus 30 in China: A detailed epidemiological and molecular study" *Arch. Virol*
11. Lukashev (2005) "Role of recombination in evolution of enteroviruses" *Rev. Med. Virol*
12. Pöyry, Stenvik, Hovi (1997) "Circulation patterns of enterovirus 30 and other enteroviruses in Finland during 1983-1993" *Epidemiol Infect*
13. Simmonds, Welch (2006) "Frequency and dynamics of recombination within different species of human enteroviruses" *J. Virol*
14. (2004) "World Health Organization. Polio Laboratory Manual, 4"
15. Bell, Cosgrove (1980) "Routine enterovirus diagnosis in a human rhabdomyosarcoma cell line" *Bull. World Health Organ*
16. Wecker, Ter Meulen (1977) "RD cells in the laboratory diagnosis of enteroviruses" *Med. Microbiol. Immunol*
17. Mcallister, Melnyk, Finkelstein et al. (1969) "Cultivation in vitro of cells derived from a human rhabdomyosarcoma" *Cancer*
18. (2025) "Supplement to the WHO Polio Laboratory Manual an Alternative Test Algorithm for Poliovirus Isolation and Characterization"
19. Oberste, Maher, Pallansch (1999) "Molecular phylogeny and evolutionary analysis of the human enteroviruses" *J. Virol*
20. Nix, Oberste, Pallansch (2006) "Sensitive, semi-nested PCR amplification of VP1 sequences for direct identification of all enterovirus serotypes from clinical specimens" *J. Clin. Microbiol*
21. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J. Mol. Biol*
22. Tamura, Nei (1993) "Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees" *Mol. Biol. Evol*
23. Tamura, Stecher, Kumar (2021) "MEGA11: Molecular Evolutionary Genetics Analysis version 11" *Mol. Biol. Evol*
24. Katoh, Standley (2014) "MAFFT: Iterative Refinement and Additional Methods" *Methods Mol. Biol*
25. Zhou, Jiang, Wang et al. (2020) "Structural insights into the interaction of enterovirus capsid protein VP1 with receptors" *Viruses*
26. Hogle (2002) "Poliovirus cell entry: Common structural themes in viral entry" *Annu. Rev. Microbiol*
27. Hansen, Long, Schultz (1997) "Structure of the RNA-dependent RNA polymerase of poliovirus" *Structure*
28. Thompson, Peersen (2004) "Structural basis for proteolysis-dependent activation of the poliovirus RNA-dependent RNA polymerase" *EMBO J*
29. Georgieva, Stoyanova, Stoitsova et al. (2017) "Echovirus 30 in Bulgaria during the European Upsurge of the Virus"
30. Maan, Dhole, Chowdhary (2019) "Identification and characterization of nonpolio enterovirus associated with nonpolio-acute flaccid paralysis in polio endemic state of Uttar Pradesh, Northern India" *PLoS ONE*
31. Sathish, Scott, Shaji et al. (2004) "An outbreak of echovirus meningitis in children" *Indian Pediatr*
32. Tuthill, Groppelli, Hogle et al. (2010) *Curr. Top Microbiol. Immunol*
33. Kuyumcu-Martinez, Van Eden, Younan et al. (2004) "Cleavage of poly(A)-binding protein by poliovirus 3C protease inhibits host cell translation: A novel mechanism for host translation shutoff" *Mol Cell. Biol*
34. Oberste, Maher, Kilpatrick et al. (1999) "Molecular evolution of the human enteroviruses: Correlation of serotype with VP1 sequence and application to picornavirus classification" *J. Virol*
35. "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=PMC12827366&blobtype=pdf | # Correction: Hodgkin lymphoma: the role of EBV plasma viral load testing in an HIV-endemic setting
J Opie, • Mohamed, • Chetty, J Bailey, • Brown, • Verburgh, • Hardie
The original article can be found online at h t t p s : / / d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 2 3 8 -0 2 4 -0 1 5 2 4 -8. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12585048&blobtype=pdf | # Detection and characterization of human bocaparvovirus in children with and without acute gastroenteritis in African-descendant community of Northern Brazil
Socorro Endrya, Ramos Fôro, Patrícia ☯¤, Lobo Santos, Delana ☯¤, Bezerra Melo, Jane ☯¤, Kaiano Haruko, Consuelo ☯¤, Eliete Silva De Oliveira, Araújo Da Cunha, Danielle ☯¤, Joana Rodrigues De Deus, Arc D', Mascarenhas Pereira, Sylvia ☯¤, De Fátima, Santos Guerra ☯, Luana Da, Silva Soares
## Abstract
Human bocaparvovirus (HBoV) is an emerging virus with worldwide distribution, may be associated with cases of acute gastroenteritis (AGE). To date, four types of HBoV have been characterized: HBoV1, HBoV2, HBoV3 and HBoV4. This study aimed to investigate HBoV in asymptomatic and symptomatic children for AGE from a Quilombola community located in Northern, Brazil, during April 2008 to September 2010.A total of 300 fecal specimens were collected and viral genomic DNA was extracted, amplified using the PCR assay, and subject to sequencing to determine HBoV genotype. HBoV was detected in 11.3% (34/300) of the samples, 12.50% (12/96) from symptomatic and 10.78% (22/204) asymptomatic children. Co-detection with other enteric viruses was reported in 20.6% (7/34) of specimens. Three genotypes of HBoV were detected with the most predominance of HBoV1 (64.7%), followed by HBoV4 (20.6%) and HBoV2 (14.7%). Phylogenetic analysis demonstrated that Brazilian HBoV are closely related with strains from South America, Asia, Africa and Oceania. This is the first description of HBoV in a Quilombola community in Brazil, and this study highlights the ability of the virus to remain in silent circulation in the community, reinforce the need for active monitoring in order to avoid problems public health futures.
## Introduction
Acute gastroenteritis (AGE) is defined as an acute infection in the gastrointestinal tract whose etiology involves different pathogens such as viruses, bacteria, protozoa, helminths and fungi. Viruses being the most prevalent etiological agents, corresponding to 50% of AGE cases [1,2]. Among the viral agents, the most prominent are norovirus (NoV), rotavirus A (RVA), and enteric human adenovirus (HAdV) [3,4]. However, in many cases, around 40% of the etiology of the infection is unknown [1,5].
Human bocavirus (HBoV) was first identified in Stockholm, Sweden in 2005 [6] initially in respiratory tract infections, and then detected in patients with AGE [7][8][9]. Meanwhile, its contribution to the etiology of AGE cases, as well as the circulating types associated with this condition have not yet been elucidated [1,3,10].
HBoV belongs to Parvoviridae family, Parvovirinae subfamily and Bocaparvovirus genus [11]. It is a non-enveloped virus with icosahedral symmetry and a linear genome consisting of single-stranded deoxyribonucleic acid (ssDNA) of approximately 5.3 kb. Its genomes present three Open Reading Frames (ORF) that encode four proteins: NS1, a non-structural viral protein involved in the replication process and with endonuclease function; NP1, a non-structural nuclear phosphoprotein related to the apoptosis process; VP1 and VP2, structural proteins responsible for viral capsid formation. These proteins are used for viral detection, with NS1 and NP1 regions constituting the most conserved region and VP1 and VP2 showing higher variability [1,7,12].
Regarding classification, HBoV are classified into two species, that have the potential to infect different hosts, such as mammals and non-human primates. The species Primate bocaparvovirus 1; includes the HBoV genotypes called HBoV-1 and HBoV-3, the species Primate bocaparvovirus 2; includes HBoV-2 and HBoV-4 [6][7][8][9]. In addition to these species, the species Primate bocaparvovirus 3 has also recently been proposed; which includes non-human bocaparvoviruses primate [13]. Most studies associate HBoV-1 with the respiratory tract while the HBoV-2, HBoV-3 and HBoV-4 species are more related to the gastrointestinal tract. However, there is a need for further studies to elucidate diseases caused by different types of HBoV [9,14].
The present study aimed to describe HBoV species circulating in children, with and without acute gastroenteritis in an African-descendant semi-closed community (Abacatal Quilombola Community) located in Pará State, Northern Brazil.
## Materials and methods
## Ethics
The Ethics Committee on Human Research of the Evandro Chagas Institute (IEC-CEPH) granted ethical approval to our study under number 06199719.9.0000.0019 in April 2019. The ethical consent form was applied to the subjects of this research. Initially, the study team held meetings with community members, such as community leaders, health visitors and school directors, in order to obtain a better understanding of the study area and to inform them about the research objectives. Written informed consent was signed by parents or guardians of the children during the fecal specimen collection. The study was carried out in accordance with the Declaration of Helsinki of 1975 (https://www.wma.net/what-we-do/medical-ethics/declaration-of-helsinki/), revised in 2013.
## Study population and sample collection
Fecal samples were collected from children aged between 0 and 12 years (average 5 years) during April 2008 to September 2010 through regular weekly visits in the community of Abacatal, located in the municipality of Ananindeua, belonging to the metropolitan region of Belém, capital of the state of Pará, Amazon region, Brazil. The map this community located in Ananindeua had demonstrated by Aragão et al [15]. This retrospective study analyzed epidemiological data from May 2019 to May 2020 using a total of 300 fecal samples. Samples from children under 12 years of age were included and samples that had exhausted fecal specimens were excluded from the analysis.. Of these, 63% (189/300) were from children up to 5 years of age, 50.7% (152/300) were male and 68% (204/300) were from the non-diarrheal group. These samples were tested for other gastroenteric viruses like Rotavirus (RVA), Norovirus (NoV) Human Adenovirus and Human Astrovirus (HAstV) in previous studies [15], and the results obtained were used to assess the presence of monoinfection or co-infection with HBoV.
## Detection and molecular characterization of HBoV
Viral DNA was extracted from a 10% fecal suspension (Tris HCl/Ca ++ 0,01M pH 7,2) using a guanidine isothiocyanate/silica method carried out according to the previously described protocol [16]. The nucleic acid was stored at -20°C. Molecular detection was performed using polymerase chain reaction (PCR) followed by nested-PCR, with specific primers, aiming to amplify the gene that encodes the VP1/VP2 protein The nucleic acid was stored at -20°C. Molecular detection was performed by polymerase chain reaction (PCR) followed by nested PCR, with specific primers, aiming to amplify the gene encoding the VP1/VP2 protein. The PCR primers AK-VP-F1 (5'-CGCCGTGGCTCCTGCTCT -3') and AK-VP-R1 (5'-TGTTCGCCATCACAAAAGATGTG-3') were used in the PCR while the primers AK-VP-F2 (5'GGCTCCTGCTCTAG GAAATAAAGAG-3') and AK-VP-R2 (5'-CCTGCTGTTAGGTCGTTGTTGTATGT-3') were used in the nested PCR, using the conditions previously described [8].
The amplicons obtained in Nested-PCR were purified using a commercial kit (PureLink PCR Purification kit, Invitrogen TM , CA, USA), according to the manufacturer recommendations. Subsequently, purified DNA was subjected to a sequencing reaction, in both directions, using a Big Dye Terminator Cycle Sequencing Ready Reaction Kit ® (v.3.1) (Applied Biosystems, CA, USA) and an ABI Prism Model 3130xl DNA Sequencer (Applied Biosystems, Foster City, USA).
The partial sequences of the VP1/VP2 genes were assembled and edited using the Geneious program (v. 9), and then aligned using the Mafft algorithm (v.7) with HBoV 1, 2, 3 and 4 sequences deposited in the database [17,18]. For phylogenetic analysis, the maximum likelihood method was adopted using the FastTree v software. 2.1.11, including the robust GTR+Gamma+F substitution model and with bootstrap reliability testing configuring 1,000 consensus replicates in order to obtain reproducible results and provide greater reliability to the clades [19]. The nucleotide sequences identified in this study were deposited in GenBank (https://www.ncbi.nlm.nih.gov/) with accession numbers OR338847-OR338880.
## Statistical analysis
All data were analyzed using descriptive statistics, processed using the PAST.Uio software through bivariate analysis of variables (age, sex and year of sample collection) in the two research groups, using the chi-square test (χ 2 ) with p value ≤ 0.05 considered statistically significant [20].
## Results
## HBoV detection
Overall, HBoV was identified in 11.3% (34/300) of samples with HBoV-positivity rates of 12.5% (12/96) and 10.8% (22/204) recorded in diarrheic and non-diarrheic children, respectively. Regarding sex, similar HBoV occurrence was observed in females (55.9%, 19/34) and males (44.1%, 15/34). Concerning age group, the most affected were children aged under five years old represented 52.9% (18/34) while those aged 6-12 years old represented 47.1% (16/34). During the years 2008-2010, a significant increase in the identification of HBoV was reported. No statistical significance was observed in the analyzed groups. This data is shown in Table 1 below.
## Co-infection with HBoV and enteric viruses
Co-infection involving HBoV and HAdV were detected in 14.7% (5/34) of the samples, while 5.9% (2/34) presented HBoV and RVA. In all cases, children were symptomatic for AGE (S1 Table ).
## Genotyping and phylogenetic analysis of HBoV
Genotyping of all specimens revealed the circulation of HBoV-1 (64.7%, 22/34), HBoV-2 (14.7%, 5/34) and HBoV-4 (20.6%, 7/34). Based on the phylogenetic inference of the VP1/VP2 partial gene, 22 strains of HBoV-1 genotype were grouped with isolates from Brazil, Thailand, Argentina and European strains that circulated from 2005 to 2018 with bootstrap support values ranging from 78.3% to 100% and with a high nucleotide similarity (99-100%). Five strains were classified as HBoV-2, lineage 2a, clustering with strains from several countries around the world such as Brazil, United States, Australia, Thailand, China, and Tunisia, circulated between 2001 and 2017, with bootstrap support value ranging from 79.1% to 99.6%, and nucleotide similarity ranging 93.6 to 99.8% among them. Concerning HBoV-4 clade, seven "Quilombola" strains clustered with specimens from Brazil, Russia, India, and Ethiopia detected between 1997 and 2016 with nucleotide similarity scores ranging from 98.2% to 99.4%. These results were showed in dendrogram of Fig 1 below.
## Discussion
HBoV has been described in distinct locations around the world in symptomatic and asymptomatic patients for AGE [9,[21][22][23][24][25]. These findings suggest HBoV as an emerging pathogen worldwide, and the investigation of this viral agent is extremely important in order to understand its molecular and epidemiological aspect, given that it can be an important viral agent of gastrointestinal infection [24,25].
In the present study, the occurrence of HBoV was evidenced circulating in children with and without symptoms of AGE from an African descendant community in Brazil in an overall positivity of 11.3%. Others studies conducted in Amazon region reported similar frequencies. Trindade et al. [26] detected HBoV in 10% of children up to 5 years with or without AGE symptoms in Acre. Leitão et al. [27] detected HBoV in 14.2% of younger Amazonian children with AGE. It is worth highlighting that HBoV frequency worldwide in children with AGE is changeable, since low occurrence were related in Korea (0.8%) [28], EUA (1.4%) [21] and China (1.4%) [29] and higher prevalence have already been described in Brazil (41.9%) [30].
With respect HBoV occurrence in symptomatic and asymptomatic AGE groups, in the present analysis the frequencies were similar. Even though, some studies demonstrate a higher detection rate in symptomatic children for AGE [9,31]. This fact could be related to the increase in viral load usually observed in patients with symptoms, which contributes to the dissemination and detection of this pathogen in fecal samples.
Regarding age groups, the presence of HBoV was found in both age groups analyzed in Table 1, with a slight increase in children under 5 years old (52.9%). Similar studies conducted in different countries attribute a higher frequency of HBoV in this age group, in which the majority of AGE cases are also observed and reinforcing that HBoV could plays an important pathogen in childhood diarrhea and indicated that young children are prone to HBoV infection [32][33][34][35][36].
Although some authors associate a higher incidence of HBoV in male sex [24,35,37], no significant differences of HBoV detection among sex were observed in the present study. This highlights the importance of conducting further studies to deepen the understanding of the epidemiological aspects of this viral agent.
HBoV co-infection with other enteric viruses has been reported worldwide. In this study, HBoV was co-detected with other enteric viruses that are involved in AGE in a relatively low frequency compared to data from several studies that correlated HBoV with other enteric viruses [1,38,39]. These findings suggest that this agent may play a collaborative role in the development of gastrointestinal tract infections. However, further research is needed since this pathology can be caused by distinct pathogenic agents.
Phylogenetic analysis revealed a significant predominance of HBoV-1. This specie was also observed by different authors in Brazil [30,36,40]. Although this genotype is more commonly associated with respiratory tract infections, it is also identified in fecal samples due its excretion after infection for months [10,41]. However, there is no evidence confirming whether its presence in feces stems from a gastrointestinal tract infection or arises as a consequence of a respiratory tract infection. Therefore, further research is needed to elucidate the association of HBoV-1 with AGE [23,24].
Besides the predominance of HBoV-1 in this study, other species were found, such as HBoV-2 and HBoV-3, commonly identified in stools samples from children with AGE symptoms [41]. All sequences classified as HBoV-2 clustered within the clade of HBoV-2a, with no sequences attributed to HBoV-2b or HBoV-2c. Furthermore, this study identified a significant rate of HBoV-4 strains (20.6%), once this specie is rarely detected. In Brazil, HBoV-4 was previously identified in fecal samples only in two studies. Sousa et al. [25] detected HBoV-4 in a fecal sample from a hospitalized child with a soft tissue tumor in the submandibular region in 2015 and Viana et al. [36] reported this specie in one fecal sample collected in 1999 understand its natural history in patients with diarrhea.
To date, this is the first report of HBoV circulation in an African descendant community in Brazil. Therefore, it was difficult to compare the present data with other investigations about HBoV in this population. Therefore, it is worth highlighting that HBoV has circulated in this semi-closed community, which may explain the high frequency of HBoV-4 in this population and not provide the propagation of this specie outside the community.
A limitation of the study was the absence of epidemiological characteristics to establish the features of HBoV infection in this population. Another restriction was the lack of screening of other enteric pathogens (e.g., bacteria and parasites) to elucidate the etiologic role of HBoV in AGE end the impact of these other enteric agents on the etiology of AGE.
## Conclusion
This study highlights epidemiological and molecular features of HBoV infection in a Quilombola community. This emphasizes the importance of investigation to understand the distribution of this virus, contributing to knowledge of molecular epidemiology and relevance of circulating types, aiming to prevent potential public health issues.
## References
1. Nora-Krukle, Vilmane, Xu et al. (2018) "Human Bocavirus Infection Markers in Peripheral Blood and Stool Samples of Children with Acute Gastroenteritis" *Viruses*
2. Black, Perin, Yeung et al. (2024) "Estimated global and regional causes of deaths from diarrhoea in children younger than 5 years during 2000-21: a systematic review and Bayesian multinomial analysis" *Lancet Glob Health*
3. Cohen, Platts-Mills, Nakamura et al. (2022) "Aetiology and incidence of diarrhoea requiring hospitalisation in children under 5 years of age in 28 low-income and middle-income countries: findings from the Global Pediatric Diarrhea Surveillance network" *BMJ Glob Health*
4. Justino, Campos, Mascarenhas et al. (2019) "Rotavirus antigenemia as a common event among children hospitalised for severe, acute gastroenteritis in Belém, northern Brazil" *BMC Pediatr*
5. Adam, Wang, Enan et al. (2018) "Molecular Survey of Viral and Bacterial Causes of Childhood Diarrhea in Khartoum State" *Sudan. Front Microbiol*
6. Allander, Tammi, Eriksson et al. (2005) "Cloning of a human parvovirus by molecular screening of respiratory tract samples" *Proc Natl Acad Sci U S A*
7. Kapoor, Mehta, Esper et al. (2010) "Identification and characterization of a new bocavirus species in gorillas" *PLoS One*
8. Kapoor, Simmonds, Slikas et al. (2010) "Human bocaviruses are highly diverse, dispersed, recombination prone, and prevalent in enteric infections" *J Infect Dis*
9. Arthur, Higgins, Davidson et al. (2009) "A novel bocavirus associated with acute gastroenteritis in Australian children" *PLoS Pathog*
10. Guido, Tumolo, Verri et al. (2016) "Human bocavirus: Current knowledge and future challenges" *World J Gastroenterol*
11. Walker, Siddell, Lefkowitz et al. (2021) "Changes to virus taxonomy and to the International Code of Virus Classification and Nomenclature ratified by the International Committee on Taxonomy of Viruses (2021)" *Arch Virol*
12. Soares, Lima, Pantoja et al. (2019) "Molecular epidemiology of human bocavirus in children with acute gastroenteritis from North Region of Brazil" *J Med Microbiol*
13. Ao, Duan (2020) "Novel Primate Bocaparvovirus Species 3 Identified in Wild Macaca Mulatta in China" *Virol Sin*
14. Tymentsev, Tikunov, Zhirakovskaia et al. (2016) "Human bocavirus in hospitalized children with acute gastroenteritis in Russia from 2010 to 2012" *Infect Genet Evol*
15. Aragão, Mascarenhas, Kaiano et al. (2013) "Norovirus diversity in diarrheic children from an African-descendant settlement in Belém, Northern Brazil" *PLoS One*
16. Boom, Sol, Salimans et al. (1990) "Rapid and simple method for purification of nucleic acids" *J Clin Microbiol*
17. Katoh, Misawa, Kuma et al. (2002) "MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform" *Nucleic Acids Res*
18. Kearse, Moir, Wilson et al. (2012) "Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data" *Bioinformatics*
19. Price, Dehal, Arkin (2010) "FastTree 2--approximately maximum-likelihood trees for large alignments" *PLoS One*
20. Hammer, Harper, Ryan (2001) "PAST: Paleontological Statistics Software Package for Education and Data Analysis" *Palaeontol Electron*
21. Chhabra, Payne, Szilagyi et al. (2008) "Etiology of viral gastroenteritis in children <5 years of age in the United States" *J Infect Dis*
22. Lasure, Gopalkrishna (2017) "Molecular epidemiology and clinical severity of Human Bocavirus (HBoV) 1-4 in children with acute gastroenteritis from Pune, Western India" *J Med Virol*
23. Lekana-Douki, Behillil, Enouf et al. (2018) "Detection of human bocavirus-1 in both nasal and stool specimens from children under 5 years old with influenza-like illnesses or diarrhea in Gabon" *BMC Res Notes*
24. Malta, Varella, Guimarães et al. (2020) "Human Bocavirus in Brazil: Molecular Epidemiology, Viral Load and Co-Infections" *Pathogens*
25. De Sousa, Almeida, Fiaccadori et al. (2017) "Identification of Human Bocavirus type 4 in a child asymptomatic for respiratory tract infection and acute gastroenteritis -Goiânia" *Braz J Infect Dis*
26. Trindade, Ramos, Lobo et al. (2023) "Epidemiologic and Clinical Characteristics of Human Bocavirus Infection in Children with or without Acute Gastroenteritis in Acre, Northern Brazil" *Viruses*
27. Leitão, Olivares, Pimenta et al. (2020) "Human Bocavirus genotypes 1 and 2 detected in younger Amazonian children with acute gastroenteritis or respiratory infections, respectively" *Int J Infect Dis*
28. Lee, Chung, Han et al. (2007) "Detection of human bocavirus in children hospitalized because of acute gastroenteritis" *J Infect Dis*
29. Zhang, Ma, Wen et al. (2015) "Clinical epidemiology and molecular profiling of human bocavirus in faecal samples from children with diarrhoea in Guangzhou" *Epidemiol Infect*
30. Campos, Sampaio, Menezes et al. (2016) "Human bocavirus in acute gastroenteritis in children in Brazil" *J Med Virol*
31. Jin, Cheng, Xu et al. (2011) "High prevalence of human bocavirus 2 and its role in childhood acute gastroenteritis in China" *J Clin Virol*
32. Albuquerque, Rocha, Benati et al. (2007) "Human bocavirus infection in children with gastroenteritis" *Brazil. Emerg Infect Dis*
33. Lindner, Karalar, Zehentmeier et al. (2008) "Humoral immune response against human bocavirus VP2 virus-like particles" *Viral Immunol*
34. Tang, Chu, Chou et al. (2015) "Molecular detection of human bocavirus 1 and 2 in children with acute gastroenteritis in Taiwan" *Southeast Asian J Trop Med Public Health*
35. Ali, Hussein, Aufi (2020) "Detecting and Genotyping of Human Bocavirus among Children with Gastroenteritis in Diyala Governorate" *ATMPH*
36. Viana, França, De Azevedo et al. (2024) "Genotypic diversity and long-term impact of human bocavirus on diarrheal disease: Insights from historical fecal samples in Brazil" *J Med Virol*
37. Pará (2010) *Brasil. Rev Pan-Amaz Saude*
38. Bergallo, Daprà, Rassu et al. (2023) "Human Bocavirus in children with acute gastroenteritis in Piedmont" *Italy. Minerva Pediatr (Torino)*
39. Zhirakovskaia, Tikunov, Tymentsev et al. (2009) "Changing pattern of prevalence and genetic diversity of rotavirus, norovirus, astrovirus, and bocavirus associated with childhood diarrhea in Asian Russia" *Infect Genet Evol*
40. Nantachit, Kochjan, Khamrin et al. (2012) "Human bocavirus genotypes 1, 2, and 3 circulating in pediatric patients with acute gastroenteritis in Chiang Mai" *J Infect Public Health*
41. Castro, Costa, Oliveira et al. (2020) "Circulation profile of respiratory viruses in symptomatic and asymptomatic children from Midwest Brazil" *Braz J Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607913&blobtype=pdf | # Molecular insights into nucleocapsid assembly and transport in Marburg and Ebola viruses
Yuki Takamatsu, Olga Dolnik, Ai Hirabayashi, Kenta Okamoto, Tomomi Kurashige, Hu Shangfan, Catarina Harumi, Takeshi Noda
## Abstract
Live-cell imaging enables visualization of the spatiotemporal dynamics of signals in cells. Intracytoplasmic movement of nucleocapsids is crucial during the life cycle of enveloped viruses; however, the molecular mechanisms governing their assembly and transport are not fully understood. Using a Marburg virus (MARV) live-cell imaging system, we identified three nucleocapsid proteins-nucleoprotein (NP), VP35, and VP24-that are necessary and sufficient to form transport-competent nucleocapsidlike structures (NCLSs). These findings are consistent with observations in Ebola virus (EBOV). Interestingly, despite incompatibility among these proteins, VP30 interacts with nucleocapsid proteins from both MARV and EBOV, supporting viral transcription and replication in heterologous systems. Furthermore, we show that the conserved PPxPxY motif at the C-terminus of NP regulates NP-VP30 interactions in both homologous and heterologous contexts and is crucial for VP30 association with NCLSs. Because this motif is conserved across filoviruses, it represents a promising target for antiviral development. Our findings advance the understanding of nucleocapsid formation and offer new avenues for therapeutic intervention against MARV and EBOV.
IMPORTANCEThis study provides crucial insights into the molecular mechanisms of nucleocapsid assembly and transport in filoviruses, specifically Marburg virus (MARV) and Ebola virus (EBOV). Using advanced live-cell imaging, we uncovered how the conserved PPxPxY motif in nucleoprotein (NP) mediates its interactions with VP30, thereby regulating nucleocapsid formation and viral replication. Notably, while VP30's role differs between MARV and EBOV, the underlying mechanism of NP-VP30 interac tion via this motif appears conserved across filoviruses, making it a promising target for broad-spectrum antiviral strategies. These findings deepen our understanding of nucleocapsid protein compatibility and virus-host interactions, offering new avenues for therapeutic intervention against these deadly pathogens. KEYWORDS Marburg virus, Ebola virus, NCLS, live-cell imaging, protein structure-func tion T he viral genome is encapsulated by proteins, forming nucleocapsids (NCs), to protect it from recognition by cellular defense mechanisms (1). In Mononegavirales virus infection, the newly synthesized nucleocapsid associated with viral RNA-dependent RNA polymerase complex is transported to the plasma membrane for virion forma tion and release (2, 3). Cryo-electron microscopy has revealed the high-resolution helical structure of RNA-bound nucleoprotein (NP) of Ebola virus (EBOV), Marburg virus (MARV), Cueva virus, Nipah virus, and measles virus (4-8), which provides the molecular mechanisms driving and stabilizing the basic structure of assembled nucleocapsids. However, the complex structure of nucleocapsid and the molecular mechanisms of NP-phosphoprotein (P) association have been largely unrevealed. Measles virus P is
mostly disordered, while three distinct interacting sites between it and NP contribute to efficient viral genome transcription (9). In paramyxoviruses, polymerase and nucleoprotein are connected via the P protein, and a flexible conformational change may occur during RNA processing (9,10). The P protein is conserved in mononegaviruses, the roles of which are assumed to be divided on VP30 and VP35 in filoviruses (11)(12)(13)(14)(15)(16), whereas the interplay of nucleocapsid proteins during nucleocapsid assembly has been largely concealed.
MARV and EBOV belong to the family Filoviridae and have an approximately 19 kb non-segmented, single-stranded, negative-sense RNA genome. Although temporal epidemics have been reported over several years in central-western Africa, the first MARV epidemic was reported following an EBOV epidemic in 2021 in Guinea, Western Africa (17,18), where the largest ever EBOV epidemic occurred during 2014-2016 with over 11,000 deaths (19). Since antiviral therapeutics for MARV and EBOV diseases are not well established, it is imperative to understand the molecular mechanisms of viral replication to establish countermeasures against these viruses. In this regard, revealing the intracellular dynamics of filoviruses is crucial because nucleocapsid assembly and transport are prerequisites for virion formation.
Oligomerized NPs lead to the formation of inclusion bodies, which are sites for viral genome transcription, replication, and nucleocapsid synthesis (13,(20)(21)(22)(23). The core structure of filovirus nucleocapsids includes a nucleocapsid-like structure (NCLS) composed of NP, encapsulating single-stranded viral genomic RNA (24,25), together with the nucleocapsid protein VP24 and the polymerase cofactor VP35, both of which are essential structural elements that directly interact with NP to build a helical nucleocapsid approximately 800 nm-1,000 nm in length and 50 nm in diameter (20,21,24,26). Viral polymerase L and transcription factor VP30 are also associated with the nucleocapsid (11,27). An immunoelectron microscopy study of nucleocapsids demonstrated that NP, VP24, and VP35 are located from the center in this order, forming NCLS, and phosphory lation-mediated VP30 association was observed at the peripheral region of NCLSs (28,29). Although the structural elements and viral protein properties are suggested to be highly conserved in filoviruses (24,30), a detailed nucleocapsid protein interaction has yet to be clarified, even in a recent cryo-electron microscopy analysis (31,32).
Here, we applied an EBOV live-cell imaging system to MARV and revealed the essential viral factors and compatibility of NC proteins during NC assembly and transport in MARV and EBOV. Moreover, a common and distinct machinery of NC formation was identified between the two viruses. Interestingly, the PPxPxY motif mediates NP-VP30 interactions in filoviruses; however, in EBOV, it had no significant effect on transcrip tion or replication. In contrast, VP30 lost its ability to support MARV transcription and replication.
## RESULTS
## Intracellular transport of MARV NCLSs nucleocapsids
The EBOV viral proteins VP30 and VP35 were used to visualize nucleocapsid and NCLS transport in live-cell imaging systems (2). green fluorescent protein (GFP)-conjugated VP30 and VP35 were analyzed and confirmed to be equivalent to their wild type in expression levels, subcellular localization, minigenomic functions, and colocalization with other NCLS proteins (Fig. S1). Moreover, no differences in nucleocapsid transport characteristics were observed in MARV-VP30 GFP -or MARV-VP35 GFP -expressing cells (Fig. S2; Movies S1 and S2) in accord with previous reports in EBOV infection (3,33).
We previously reported that MARV NCLS transport can be visualized using a viruslike particle (VLP) system expressing NP, VP35, VP24, L, VP40, glycoprotein (GP), and minigenome, together with VP30 GFP (34). Since VP40 and GP were not indispensable for EBOV NCLS transport (2), we transfected only MARV nucleocapsid components in Huh-7 cells, and live-cell imaging was performed from 20 h post-transfection (p.t.) (Fig. 1A). Smooth, long lines were observed in dimethyl sulfoxide (DMSO)-or nocodazoletreated cells, whereas they disappeared in cytochalasin D-treated cells (Fig. 1B). The frequency of NCLSs translocated to a distance of >5 µm, as well as the mean velocity of NCLS transport, was drastically decreased in cytochalasin D-treated cells (Fig. 1C). In summary, NCLSs formed with NPs, VP35, VP30, VP24, L, and the minigenome exhibited long-distance transport via actin filaments, with characteristics similar to those of MARV nucleocapsid transport.
## Identification of viral components necessary for MARV NCLS transport
Next, we applied a reductionist approach to identify viral factors required for NCLS transport. The experiments were repeated with the omission of each viral component (NP, VP35, VP24, L, or minigenome). Using live-cell imaging with MARV-VP30 GFP to track NCLS transport, we observed that transport-competent NCLSs were still formed even when polymerase L or the EBOV-specific minigenome was omitted. In contrast, the omission of NP, VP35, or VP24 resulted in failure to form transport-competent NCLSs (Fig. 2A). Next, the experiments were repeated using MARV-VP35 GFP to monitor the role of VP30 in NCLS transport. Although the presence of NP and VP24 is crucial for NCLS transport, VP30 was dispensable for NCLS transport (Fig. 2B). These data indicate that NP, VP35, and VP24 proteins are essential, whereas polymerase L, VP30, and the minigenome are not necessary for NCLS transport.
## Transport characteristics of NP, VP35, and VP24 forming NCLSs
To confirm that these three components are sufficient to form transport-competent NCLSs, live-cell imaging of NP, VP24, and VP35 together with VP35-GFP-expressing cells was started at 20 h p.t. (Fig. 3A) Here again, we demonstrated that the frequency of NCLSs translocated >5 µm, as well as the mean velocity of NCLS transport in nocodazoletreated cells (Fig. 3B andC; Movie S3), was drastically decreased in cytochalasin D-treated cells (Fig. 3B andC; Movie S4). The characteristics of NCLS transport were similar to those cells transfected with all nucleocapsid components (Fig. 1), and those of nucleocapsid transport in MARV-infected cells (Fig. S2). In summary, NP, VP35, and VP24 nucleocapsid proteins are necessary and sufficient to mediate NCLS transport.
## An exchange of NCLS proteins between MARV and EBOV
The helical structure of NCLS, which is formed by NP, VP35, and VP24, is well-conserved in MARV and EBOV (24,30), although interactions involving nucleocapsid-forming proteins between MARV and EBOV remain largely unexplored. To analyze the compatibility of nucleocapsid proteins between MARV and EBOV, we exchanged each protein in NP, VP35, VP24, and VP30 to observe the assembly and transport of NCLSs. In the MARV livecell imaging system, EBOV NP, VP35, and VP24 were incompatible, whereas MARV-VP30 was exchangeable with EBOV-VP30 (Fig. 4A; Movie S5). EBOV-VP30 was also exchangea ble with MARV-VP30 in the EBOV live-cell imaging system (Fig. 4B; Movie S6). These results indicate that VP30 proteins are structurally compatible with MARV and EBOV proteins.
## Nucleocapsid protein replacement in minireplicon assays
Little information has been published regarding chimeric systems for evaluating genomic RNA replication among filoviruses (35), and the magnitude and mechanisms of possible compatibility are unclear. Here, we replaced either NP, VP35, or VP30 in opti mized minigenome transcription and replication assays (36,37). In the MARV minige nome assay, the exchange of MARV-NP with EBOV-NP and MARV-VP35 with EBOV-VP35 resulted in the abolishment of reporter activity close to the value of the negative control (in the absence of L protein). On the other hand, the exchange of MARV-VP30 with EBOV-VP30 retained approximately 60% reporter activity, which was higher than that in the absence of MARV-VP30, which demonstrated around 40% reporter activity when the value of MARV-VP30 was set as 100% (Fig. 5A). In EBOV minigenome assays, the replace ment of EBOV-VP30 with MARV-VP30 retained approximately 3%-5% reporter activity, which was significantly higher than that in the absence of EBOV-VP30 with less than 1% reporter activity when the value of EBOV-VP30 was set as 100% (Fig. 5B). Interestingly, both MARV-VP30 and EBOV-VP30 were at least partially functional transcriptional activators in the heterologous replicon system.
## Interactions between NP and VP30
To reveal the molecular mechanisms underlying VP30-mediated transcriptional support activity and VP30-NCLS association in heterologous viruses, we performed immunofluorescence and immunoprecipitation assays. Both MARV-NP and EBOV-NP formed perinu clear inclusions, and both MARV-VP30 and EBOV-VP30 were diffusely distributed in the cytoplasmic regions when they were singly expressed (Fig. 6A andB, left lanes) (15,38,39). As previously described, MARV-VP30 accumulated in the MARV-NP-induced inclusions when co-expressed (Fig. 6A, middle lane) (3,39). Similarly, EBOV-VP30 was localized in the EBOV-NP-induced inclusions when co-expressed (Fig. 6B, middle lane) (11,15,38). Notably, EBOV-VP30 also accumulated in MARV-NP-induced inclusions, and MARV-VP30 accumulated in EBOV-NP-induced inclusions (Fig. 6A andB, right lanes).
Next, we performed immunoprecipitation assays using MARV-NP flag and EBOV-NP flag . EBOV-VP30 GFP was precipitated using EBOV-NP flag , and MARV-VP30 GFP was precipitated using MARV-VP30 GFP , as expected (Fig. 6C, lane 2, Fig. 6D, lane 2). Moreover, EBOV-VP30 GFP was precipitated by MARV-NP flag , and MARV-VP30 GFP was precipitated by EBOV-NP flag (Fig. 6C andD, lanes 3). These results demonstrate that direct interactions between NP and VP30 bring about colocalization of these proteins in viral inclusions, suggesting that heterologous VP30 is partially functional in minigenome transcription and replication.
## PPxPxY motif-mediated NP-VP30 interactions
VP30 binds to the PPxPxY motif in NP, which regulates NP-mediated VP30 dephosphory lation (13,40), which is a key modulator of transcriptional support activity in both EBOV and MARV (12,15,41). To reveal interactions between VP30 and the NP PPxPxY motif, mutations were introduced into this motif (MARV-NP ΔVP30 and EBOV-NP ΔVP30 , Fig. 7A). MARV-NP ΔVP30 exhibited reporter activity in the MARV minigenome assay, which was similar to that of the NP wild-type without VP30, regardless of the presence or absence of VP30 (Fig. 7B). In contrast, EBOV-NP ΔVP30 showed a decrease in reporter activity to the same level as that of the negative control when VP30 was not expressed; however, reporter activity was rescued by VP30 expression (Fig. 7C). As reported previously (11), VP30 in EBOV supports transcription and replication activity even when it loses its interaction with NP. Notably, VP30 does not exhibit transcription and replication support activity in MARV when it loses its interaction with NP.
Next, we employed confocal microscopy to visualize the formation of inclusion bodies and the localization of NP and VP30. This analysis demonstrated that MARV-NP ΔVP30 could form inclusion bodies but failed to incorporate either MARV-VP30 or EBOV-VP30 (Fig. 7D). Similarly, EBOV-NP ΔVP30 did not incorporate EBOV-VP30 or MARV-VP30 into the inclusions (Fig. 7E). The recruitment of L and VP35 encapsulated bodies was not affected by the PPxPxY motif, although L of MARV could not be evaluated (Fig. S3).
To clarify the interaction between NP and VP30 mediated by the PPxPxY motif, we conducted immunoprecipitation experiments. MARV-NP ΔVP30 lost its ability to interact with MARV-VP30 (Fig. 7F, lane 4), and EBOV-NP ΔVP30 was unable to precipitate EBOV-VP30 (Fig. 7G, lane 4).
These results indicate that the interactions between NP and VP30 are regulated by the conserved PPxPxY motif in NP, which is well conserved among filoviruses without affecting the formation of inclusions.
## PPxPxY motif and NCLS assembly, and VLP formation
Microscopic analyses were performed to investigate the role of the PPxPxY motif in NCLS assembly and transport. First, to analyze whether the NP mutants could form NCLSs, transmission electron microscopy was conducted. Tubular-like structures with electrondense walls, representing condensed nucleocapsids, were detected in the presence of MARV-NP ΔVP30 (Fig. 8A, arrowheads) and EBOV-NP ΔVP30 (Fig. 8B,arrowheads).
Next, we conducted live-cell imaging analyses of NCLS containing homologous or heterologous VP30-GFP in conjunction with NP or NP ΔVP30 . As shown in Fig. 4, long sequential lines indicating transport-competent NCLSs were observed in MARV-NP-and EBOV-NP-containing NCLSs (Fig. 8C). In contrast, MARV-NP ΔVP30 -and EBOV-NP ΔVP30containing NCLSs demonstrated only diffusely distributed signals from both MARV and EBOV live-cell imaging systems (Fig. 8D).
Subsequently, we examined the process of VLP formation following NCLS transport. In the VLP assay, we transfected cells with NP, VP24, VP35, VP30, L, VP40, GP, a minige nome, and T7, and subsequently collected the VLPs secreted into the cell supernatant (Fig. 8E). Purified VLPs were assessed by western blotting. In both the MARV and EBOV VLP assays, we observed the formation of VP40-containing VLPs, even in the absence of NP, whereas VP30 was undetectable (Fig. 8F andG, lanes 1). However, in the presence of wild-type NP, the incorporation of VP30 was confirmed; this was not observed in the NP mutants (MARV-NP ΔVP30 and EBOV-NP ΔVP30 ) (Fig. 8F andG, lanes 2 and 3).
In summary, mutations in the PPxPxY motif are crucial for VP30 association with NCLSs and for VP30 incorporation into VLPs in both MARV and EBOV.
## PPxPxY motif and NC protein interactions
The association between NP and VP35 was reported upon NP binding peptide (20-48 a.a. of VP35) to the N-terminal side of NP (residues 25-457), and the basic patch of VP35 (residues 220-251) to the central domain of NP (22,40,(42)(43)(44). Regarding the NP-VP24 interaction, two molecules of VP24 bind to two molecules of NP in distinct configurations, and mutations in NP R132A and NP H196A inhibit the formation of NCLS (31). Given that the structure of the C-terminal region of NP and interactions among NCLS proteins are largely unknown, we used AlphaFold2 and AlphaFold3 (45,46) to predict the binding of PPxPxY peptides to VP35 and VP24. Confident structural predictions were obtained only for MARV-VP35, EBOV-VP35, and EBOV-VP24 (pLDDT >80, Fig. S4A through C). All the predicted structures indicated that both MARV and EBOV VP35 interacted with the PPxPxY peptide in a similar structural region. These results suggest that the PPxPxY peptide may interact with the positively charged VP35 surface in both MARV and EBOV (Fig. S4D andE). However, a protein complex was not predicted for the peptide with an alanine substitution. Additionally, we were unable to predict a highly confident binding of the PPxPxY peptide to MARV-VP24; in contrast, both PPxPxY and the alanine substitu tion mutant peptides appeared to interact with EBOV-VP24 (Fig. S4C).
To reveal the NP-PPxPxY motif and VP35 interaction, microscopy and immunoprecipi tation assays were performed. Immunofluorescence microscopy analyses revealed that MARV-NP ΔVP30 recruited MARV-VP35 to these inclusions (Fig. S5A). Similarly, EBOV-NP ΔVP30 recruited EBOV-VP35 (Fig. S5B). Next, we conducted live-cell imaging analyses of NCLSs containing either NP or NP ΔVP30 along with their respective VP35-GFP. Both MARV-NP ΔVP30 and EBOV-NP ΔVP30 exhibited transport-competent NCLSs (Fig. S5C andD). Finally, we performed immunoprecipitation assays, demonstrating that both MARV-NP ΔVP30 and EBOV-NP ΔVP30 successfully precipitated their respective VP35 (Fig. S5E and F, lanes 4). These results suggest that the PPxPxY motif does not govern the interactions among NCLS-forming proteins; rather, it regulates the interaction between NP and VP30 in a localized manner.
In summary, VP30 exhibits compatibility between MARV and EBOV, with transcription and replication activities partially sustained by heterologous VP30 (Fig. 9A). The binding of NP to VP30 is regulated by the PPxPxY motif; when a mutation is introduced, VP30 is unable to bind to NP. Consequently, there was no association between VP30 and NCLS (Fig. 9B). Conversely, when the motif is intact, heterologous VP30 can bind to NCLS, facilitating intracellular transport (Fig. 9C).
## DISCUSSION
In the present study, we revealed that the core structure of nucleocapsids, NCLSs, is formed from NP, VP35, and VP24 (Fig. 1 to 3) in agreement with EBOV, using live-cell imaging systems based on the ectopic expression of fluorescently labeled viral proteins in MARV. Based on these observations, we characterized the nucleocapsid compatibility between MARV and EBOV and sought to reveal the molecular interface between the nucleocapsid-forming proteins.
The characteristic filoviral NP-RNA helical complex provides a scaffold for nucleocap sid formation, which is responsible for the transcription and replication of viral RNAs (4,7,8,24). Previously, MARV-VP30 was reported to be partially functional in the transcription and replication of EBOV chloramphenicol acetyltransferase reporter assays (35). Our p.t., the cells were lysed and protein complexes were precipitated using mouse anti-Flag M2 agarose. An aliquot of the cell lysate (input) was collected before precipitation. Elution was achieved using SDS sample buffer. Western blot analysis was performed using Flag-, GFP-, and α-tubulin-specific antibodies. Lane results demonstrated that EBOV-VP30 also supported transcription and replication of MARV minigenome (Fig. 5B), indicating that filoviruses have conserved machinery for VP30-nucleocapsid associations. On the hand, VP30 plays different roles in recombinant viruses production; MARV was rescued by EBOV VP30 instead of MARV VP30 expressions, whereas EBOV was rescued only by EBOV VP30 (47). VP30 is a phos phoprotein, and its dephosphorylation is crucial for transcriptional support in EBOV rather than in MARV (11,13,41). The LxxIxE and PPxPxY motifs located adjacent to the Cterminus of NP are conserved in filoviruses (13). In EBOV, VP30 binds to NP at the PPxPxY motif, and VP30 is dephosphorylated by PP2A, which is recruited by the LxxIxE motif (13). Notably, mutations involving the PPxPxY motif did not significantly affect the transcrip tional support activity in either the MARV or EBOV minigenome assays (Fig. 7B andC). Defective interactions involving NP-VP30 proteins do not cause defective viral genome transcription/replication in EBOV (11,40), indicating that a high-affinity interaction between NP and VP30 is not strictly required for viral RNA synthesis, and minimal binding of these proteins mediates viral RNA synthesis in EBOV (48). In contrast, the transcription and replication activity of MARV slightly decreased due to mutation of the PPxPxY motif. However, this reduction was consistent with the levels observed in the absence of VP30 in wild-type NP, suggesting that this phenomenon reflects the influence of VP30 rather than the activity of the motif itself in transcription and replication. Multiple studies have demonstrated significant differences in the effects of VP30 on transcription and replica tion in the minigenome systems of MARV and EBOV, primarily attributed to VP30's influence on the polymerase complex (11,41). In EBOV, the absence of VP30 reduces activity to less than 1% of wild-type levels (12,14), whereas in MARV, even without VP30, 60%-70% of transcription and replication activity is maintained (41). This difference may be due to variations in the sensitivity between the two minigenome assays. Specifically, the original reporter activity values for EBOV were reported to be 10 times higher than those for MARV (Fig. S6). This difference in sensitivity may partly explain why the effects of VP30 are more clearly observed; however, further investigation is warranted for another possibility. In filoviruses, nucleocapsids form a left-handed helix with an inner nucleoprotein layer decorated with protruding arms composed of VP24 and VP35 (23,24,30). Even in a recent model, the intrinsically disordered C-terminal region of NP (aa 450-600), which is critical for nucleocapsid formation in the presence of VP24 and VP35, has been unclear due to its flexibility and insufficient resolution of the electron microscopy (EM) map (31,32). This study reveals a common aspect of filoviral nucleocapsid assembly, that is, heterologous VP30 associates with NP and supports transcription/replication in the inclusions in both MARV and EBOV (Fig. 5 and6). Intriguingly, the PPxPxY motif regulates the interaction with NP-VP30 but does not affect the assembly and transport of NCLSs (Fig. 8A andB; Fig. S4C andD). Consequently, introducing mutations in this motif inhibits the association of VP30 to VLPs (Fig. 8F andG). Noteworthy, it has been reported that transcription and replication are inhibited by the addition of competitive peptides The intracellular distribution of proteins noted above the images was visualized using NP-specific antibodies and autofluorescence, and merged images were visualized. The small boxed areas are enlarged at the four corners. Scale bars: 10 µm (scale bar in insets, 2 µm). (F, G) Immunoprecipitation assay in HEK293 cells.
Cells were transfected with the indicated protein-encoding plasmids ([F] MARV and [G] EBOV). NP-encoding plasmids were fused with a FLAG tag, cells were lysed, and protein complexes were precipitated using mouse anti-FLAG M2 agarose at 48 h p.t. An aliquot of cell lysate (input) was collected before precipitation.
Elution was achieved using SDS sample buffer. Western blotting was performed using FLAG-, GFP-, and alpha-tubulin-specific antibodies. against this motif in EBOV minigenome (48). Moreover, PPxPxY motif-bearing proteins, such as RBBP6, hnRNPUL1, and PEG10, modulate EBOV transcription and replication through discrete mechanisms (49), highlighting the importance of this motif in EBOV replication. Given that research on this motif in MARV has not yet been reported, this remains a topic for future investigation.
Using AlphaFold2/3 prediction (45,46), the PPxPxY peptide may bind to the positively charged surface of VP35, both in MARV and EBOV, although no such protein complexes are predicted when peptides hold alanine substitutions (Fig. S3). In EBOV, amino acid residues R225, H240, K248, and K251 of VP35 are reported to be important for its NP interaction (50), which forms the first basic patch (amino acid residues 222-251) (22). Interestingly, residues Q241, Q244, and K248 located in this basic patch have been shown to be part of the PPxPxY-binding motif of NP, but mutations in this motif did not affect the NP-VP35 interaction or the formation of NCLS (Fig. 8A andB; Fig. S4). AlphaFold predictions are computational and speculative; therefore, biophysical interaction assays to validate predicted interactions between the PPxPxY motif and NC proteins are anticipated in the future.
Pseudoviruses are useful for studying the entry process (51); however, they do not recapitulate the complexities of nucleocapsid assembly, protein interactions, and intracellular transport. Therefore, VLPs and minigenome systems were employed in this study. However, these methods have some inherent limitations. Notably, the use of cultured cell lines for microscopic analysis may not fully reflect an in vivo environment. Since antigen-presenting cells such as monocytes, macrophages, dendritic cells, and endothelial cells play prominent roles in filovirus infection (52), analyzing the interactions of NCLS proteins and transport mechanisms in these cell types would provide more clinically relevant insights. While live-cell imaging of NCLS can help mitigate biosafety concerns, recombinant viruses such as conditional recombinant virus systems (53) offer a promising approach to study these processes under more physiological conditions. In particular, future investigations employing VP30-deficient recombinant filoviruses (e.g., EBOV-ΔVP30 or MARV-ΔVP30) are anticipated to clarify the role of the PPxPxY motif of VP30 in NC formation and viral transcription, bridging the gap between reductionist systems and natural infection.
In conclusion, our study demonstrated that the PPxPxY motif not only regulates the binding between NP and VP30 but also influences the association of VP30 with NCLS. Interestingly, the interaction between NP and VP30 through the PPxPxY motif is regula ted in a somewhat permissive manner, exhibiting compatibility between MARV and EBOV. This suggests a potential for developing drugs that inhibit the replication of a wide range of filoviruses by targeting non-specific binding through this motif.
## MATERIALS AND METHODS
## Cells and viruses
Huh-7 and HEK293 cells were maintained at 37°C and 5% CO 2 in Dulbecco's modified Eagle's medium (DMEM; Life Technologies) supplemented with 10% (vol/vol) fetal bovine serum (FBS; PAN Biotech), 5 mM L-glutamine (Q; Life Technologies), 50 U/mL penicillin, and 50 µg/mL streptomycin (PS; Life Technologies). MARV (Musoke accession no. DQ217792.1, GenBank) and recombinant MARV were propagated on VeroE6 cells as described previously (54). All work with infectious viruses was performed in a bio-safety level (BSL)-4 facility at Philipps-Universität Marburg following national legislation and guidelines.
## Plasmids and transfection
Plasmids encoding the MARV proteins pCAGGS-MARV-NP, pCAGGS-MARV-VP35, pCAGGS-MARV-VP30, pCAGGS-MARV-VP24, pCAGGS-MARV-L, pCAGGS-MARV-VP40, and pCAGGS-MARV-GP, as well as the T7-driven MARV minigenome encoding Renilla luciferase and pCAGGS-T7 polymerase, were used as previously described (36,37). Plasmids encoding MARV-VP30 GFP and MARV-VP35 GFP fusion proteins have been described previously (2,3). Plasmids encoding EBOV proteins (pCAGGS-EBOV-NP, pCAGGS-EBOV-VP35, pCAGGS-EBOV-VP30, pCAGGS-EBOV-VP24, and pCAGGS-EBOV-L) and the T7-driven EBOV minigenome encoding Renilla luciferase were prepared as described previously (11,55).
The cloning procedure for plasmids with mutations introduced at the PxPPxY motif with Flag-tagged plasmids (pCAGGS-MARV-NP flag , pCAGGS -MARV-NP ΔVP30 flag , pCAGGS-EBOV-NP flag , and pCAGGS-EBOV-NP ΔVP30 flag ) has been described previously (26). Transfection was performed in Opti-MEM without phenol red (Life Technologies) using TranSIT (Mirus), according to the manufacturer's instructions.
## SDS-PAGE and western blot analysis
SDS-PAGE and western blot analyses were performed as previously described (11,56). Protein detection was performed using Image Lab software (Bio-Rad) or Image Reader LAS-3000 (Fujifilm) for horseradish peroxidase (HRP)-conjugated secondary antibodies, as indicated in the antibodies section below.
## Immunofluorescence analysis and confocal laser scanning microscopy
Immunofluorescence analyses were performed as previously described (57,58). Microscopic images were acquired using a Leica SP5 confocal laser scanning microscope with a 63× oil objective, an Olympus FV3000 microscope with a 100× oil objective, or a Keyence BZ-X810 microscope with a 100× oil objective (Olympus). Cells were grown on µ-Slide 8 or 12 wells (ibidi) and fixed with 4% paraformaldehyde 20 h post-transfection. Nuclear staining was performed using Hoechst 33342 (Dojindo).
## Antibodies
The following primary antibodies were used for immunofluorescence: mouse anti-MARV-NP (37), rabbit anti-MARV-NP (IBT), rabbit anti-MARV-VP30 (37), guinea pig anti-MARV-VP35 (37), rabbit anti-MARV-VP24 (37), chicken anti-EBOV-NP (12), rabbit anti-EBOV-NP (IBT), and rabbit anti-EBOV-VP30 (11). The corresponding secondary antibodies were donkey anti-mouse Alexa488 (Abcam), donkey anti-mouse Alexa594 (Abcam), don key anti-mouse-IRDye 680RD (LI-COR), donkey anti-mouse Alexa488 (Abcam), donkey anti-rabbit Alexa594 (Abcam), donkey anti-rabbit IRDye 680RD (LI-COR), goat anti-guinea pig Alexa594 (Abcam), goat anti-chicken Alexa594 (Thermo Fisher Scientific), and donkey anti-chicken IRDye 680RD (LI-COR).
The following primary antibodies were used for the western blot analysis: mouse anti-MARV-NP monoclonal antibody (33), rabbit anti-MARV-NP antibody (IBT), mouse anti-MARV-VP40 monoclonal antibody (33), chicken anti-EBOV-NP polyclonal antibody (see above), rabbit anti-EBOV antibody (IBT), rabbit anti-EBOV-VP40 antibody (IBT), rabbit anti-GFP antibody (Rockland), mouse anti-FLAG M2 monoclonal antibody (Sigma-Aldrich), rabbit anti-HA monoclonal antibody (ROCKLAND), and rabbit anti-αtubulin antibody (MBL). The corresponding secondary antibodies used were HRP-conju gated goat anti-mouse IgG (Abcam), HRP-conjugated goat anti-rabbit IgG (Abcam), and HRP-conjugated goat anti-chicken IgY (Abcam).
## Minigenome reporter assay
MARV minigenome assays were performed as previously described (37). Briefly, plasmids for the minigenome assay (500 ng of pCAGGS-NP, 100 ng of pCAGGS-VP35, 100 ng of pCAGGS-VP30, and 1,000 ng of pCAGGS-L, 500 ng of a MARV-specific minigenome encoding the Renilla luciferase reporter gene, and 500 ng of pCAGGS-T7 polymerase, and 50 ng of pGL-encoding firefly luciferase reporter gene for normalization) were transfected into HEK293 cells. EBOV minigenome assays were performed as previously described (59). Briefly, plasmids for minigenome assays (125 ng of pCAGGS-NP, 100 ng of pCAGGS-VP35, 100 ng of pCAGGS-VP30, and 1,000 ng of pCAGGS-L, with 250 ng of EBOV-specific minigenome encoding the Renilla luciferase reporter gene, 250 ng of pCAGGS-T7 polymerase, and 50 ng of pGL-encoding firefly luciferase reporter gene for normalization) were transfected into HEK293 cells. Cells were lysed and subjected to a luciferase reporter assay (PJK).
## Live-cell imaging microscopy
A total of 8 × 10 4 , 4 × 10 4 , or 2 × 10 4 Huh-7 cells were seeded onto a µ-Dish 35 mm, a µ-Slide 4 well, or a µ-Slide 8 well (ibidi) and incubated in DMEM/PS/Q with 10% FBS. To observe MARV nucleocapsid transport, cells were infected with a multiplicity of infection of 1. The inoculum was replaced with fresh medium at 1 h post-infection (33). Subse quently, 500 ng of DNA encoding the green fluorescent fusion protein was transfected. To observe MARV NCLS transport, each well was transfected with the following plasmids encoding MARV proteins: (500 ng of pCAGGS-NP, 100 ng of pCAGGS-VP35, 100 ng of pCAGGS-VP30 GFP , and 100 ng of pCAGGS-VP24), together with the MARV minige nome-expressing plasmid and T7 polymerase-coding plasmid (36,37). To observe EBOV NCLS transport, each well was transfected with the following plasmids encoding EBOV proteins: 500 ng of pCAGGS-NP, 200 ng of pCAGGS-VP35, 200 ng of pCAGGS-VP24, and 200 ng of pCAGGS-VP30-GFP. The inoculum was removed at 1 h p.t., and CO 2 -independ ent Leibovitz's medium (Life Technologies) with PS/Q, non-essential amino acid solution, and 3%-20% (vol/vol) FBS was added. Live-cell time-lapse experiments were recorded with a Nikon ECLIPSE TE2000-E using a 63× oil objective, GE Healthcare Delta Vision Elite using a 60× oil objective, Keyence BZ-X810 microscope using a 100× oil objective under BSL-2, and Leica DMI6000B under BSL-4 using a 63× oil objective equipped with a remote-control device to operate the microscope outside the BSL-4 facility (3). To visualize representative trajectories, NCs were detected using a size threshold of 1 µm, and NCLSs using a threshold of 0.5 µm. Signals with a length of at least 5 µm that exhibited directional movement were then traced. While many inclusion bodies are larger than NCs and NCLSs, it is difficult to exclude smaller, nucleus-distant inclusion bodies solely based on size. However, within short observation periods, inclusion bodies typically display random, non-directional movements and can be distinguished from the movements of NCs and NCLSs.
## Treatment of cells with cytoskeleton-modulating drugs
Cells were treated with 15 µM nocodazole (Sigma), 0.3 µM cytochalasin D (Sigma), or 0.15% DMSO (Sigma), as previously described (3). The chemicals were added to the cell culture medium 3 h prior to observation.
## Image processing and analysis
The acquired pictures and movie sequences were processed using the Imaris tracking module (Bitplane; Oxford Instruments, Abingdon, UK) (60). The size of spots >1 µm for nucleocapsids and 0.5 µm for NCLSs was collected. Subsequently, "Quality"-based optimization of the detected spots was performed. The tracking algorithm of "Autore gressive Motion" was applied for tracking with a "Maximum Distance" of 1 and "Max imum Gap" of 1 with filling of gaps with detected objects. Detected trajectories of length >5 µm, duration >15 s, and track straightness >0.2 were processed for the analyses. Moving signals were collected from approximately 10 cells derived from three independent experiments.
## Co-immunoprecipitation analysis
Co-immunoprecipitation assays were performed as previously described (26). For SDS-PAGE, elution was achieved with 70 µL of SDS sample buffer (Fujifilm) and subjected to gel electrophoresis and western blot analysis.
## Ultrathin section electron microscopy
Huh-7 cells were seeded on a 12-well plate and transfected with 1 µg NP or NP mutant-encoding plasmids, 0.5 µg VP24, and 0.5 µg VP35-encoding plasmids. At 48 h post-transfection, cells were fixed with aldehydes, post-fixed with 1% osmium tetroxide, dehydrated in a graded ethanol series, and embedded in EPON 812 (TAAB, Berks, UK). Ultrathin sections were stained with uranyl acetate and lead citrate and observed under a Hitachi HT-7700 microscope operated at 80 kV (Hitachi Hi-Tech, Tokyo, Japan) (31).
## VLP assay
VLP assays were conducted following established protocols (37,55,61), with minor modifications. HEK293 cells were seeded in six-well plates and transfected with plasmids encoding all structural proteins and reporter genes for either MARV (500 ng of pCAGGS-NP, 100 ng of VP35, 500 ng of VP40, 500 ng of GP, 100 ng of VP30, 100 ng of VP24, 1,000 ng of L, 500 ng of pANDY-3M5M, 500 ng of pCAGGS-T7 polymerase) or EBOV (125 ng of pCAGGS-NP, 100 ng of VP35, 250 ng of VP40, 250 ng of GP, 100 ng of VP30, 100 ng of VP24, 1,000 ng of L, 250 ng of pANDY-3E5E, 250 ng of pCAGGS-T7 polymerase). Plasmids encoding the firefly luciferase reporter gene were used for normalization. After 72 h post-transfection, culture supernatants were collected, and VLPs were purified by ultracentrifugation using a 20% sucrose cushion. VLPs were analyzed for VP30 incorpora tion using proteinase K digestion assay (37).
## AlphaFold-multimer prediction
To predict the complex structure of the NP-PPxPxY peptide and VP35 or VP24 in MARV or EBOV, we employed AlphaFold structural predictions. The amino acid sequences of EBOV and MARV proteins were obtained from GenBank (ID: EBOV: NC_002549, MARV: DQ217792.2). The structural complex of EBOV and MARV NPs with each peptide sequence was predicted in AlphaFold2 and AlphaFold3 software (45,46). The most confidently predicted structures (high pLDDT values) were visualized and assessed using UCSF Chimera software (62).
## Statistical analysis
Data represent the mean values and standard deviations from at least three independ ent experiments. Statistical analyses were performed using GraphPad Prism software (version 8.0). Normally distributed samples were analyzed using Student's t-test. Statistically significant differences are indicated by asterisks (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).
## References
1. Akira, Uematsu, Takeuchi (2006) "Pathogen recognition and innate immunity" *Cell*
2. Takamatsu, Kolesnikova, Becker (2018) "Ebola virus proteins NP, VP35, and VP24 are essential and sufficient to mediate nucleocapsid transport" *Proc Natl Acad Sci*
3. Schudt, Kolesnikova, Dolnik et al. (2013) "Live-cell imaging of Marburg virus-infected cells uncovers actin-dependent transport of nucleocapsids over long distances" *Proc Natl Acad Sci*
4. Sugita, Matsunami, Kawaoka et al. (2018) "Cryo-EM structure of the Ebola virus nucleoprotein-RNA complex at 3.6 Å resolution" *Nature*
5. Desfosses, Milles, Jensen et al. (2019) "Assembly and cryo-EM structures of RNA-specific measles virus nucleocapsids provide mechanistic insight into paramyxoviral replication" *Proc Natl Acad Sci*
6. Ker, Jenkins, Greive et al. (2021) "CryoEM structure of the Nipah virus nucleocapsid assembly" *PLoS Pathog*
7. Fujita-Fujiharu, Sugita, Takamatsu et al. (2022) "Structural insight into Marburg virus nucleoprotein-RNA complex formation" *Nat Commun*
8. Hu, Fujita-Fujiharu, Sugita et al. (2023) "Cryoelectron microscopic structure of the nucleoprotein-RNA complex of the European filovirus, Lloviu virus" *PNAS Nexus*
9. Guseva, Milles, Blackledge et al. (2019) "The nucleoprotein and phosphoprotein of measles virus" *Front Microbiol*
10. Cox, Plemper (2015) "The paramyxovirus polymerase complex as a target for next-generation anti-paramyxovirus therapeutics" *Front Microbiol*
11. Biedenkopf, Hartlieb, Hoenen et al. (2013) "Phosphorylation of Ebola virus VP30 influences the composition of the viral nucleocapsid complex: impact on viral transcription and replication" *J Biol Chem*
12. Biedenkopf, Lier, Becker (2016) "Dynamic phosphorylation of VP30 is essential for Ebola virus life cycle" *J Virol*
13. Kruse, Biedenkopf, Hertz et al. (2018) "The Ebola virus nucleoprotein recruits the host PP2A-B56 phosphatase to activate transcriptional support activity of VP30" *Mol Cell*
14. Takamatsu, Krähling, Kolesnikova et al. (2020) "Serine-arginine protein kinase 1 regulates Ebola virus transcription" *mBio*
16. Modrof, Mühlberger, Klenk et al. (2002) "Phosphorylation of VP30 impairs Ebola virus transcription" *J Biol Chem*
17. Martínez, Biedenkopf, Volchkova et al. (2008) "Role of Ebola virus VP30 in transcription reinitiation" *J Virol*
18. Who (2021) "Ebola virus disease. Guinea"
19. Who (2021) "Marburg virus disease"
20. (2016) "Ebola situation report"
21. Watanabe, Noda, Kawaoka (2006) "Functional mapping of the nucleoprotein of Ebola virus" *J Virol*
22. Huang, Xu, Sun et al. (2002) "The assembly of Ebola virus nucleocapsid requires virion-associated proteins 35 and 24 and posttranslational modification of nucleoprotein" *Mol Cell*
23. Miyake, Farley, Neubauer et al. (2020) "Ebola virus inclusion body formation and RNA synthesis are controlled by a novel domain of nucleoprotein interacting with VP35" *J Virol*
24. Noda, Ebihara, Muramoto et al. (2006) "Assembly and budding of Ebolavirus" *PLoS Pathog*
25. Bharat, Noda, Riches et al. (2012) "Structural dissection of Ebola virus and its assembly determinants using cryo-electron tomography" *Proc Natl Acad Sci*
26. Mühlberger, Lötfering, Klenk et al. (1998) "Three of the four nucleocapsid proteins of Marburg virus, NP, VP35, and L, are sufficient to mediate replication and transcription of Marburg virus-specific monocistronic minigenomes" *J Virol*
27. Takamatsu, Kolesnikova, Schauflinger et al. (2020) "The integrity of the YxxL motif of Ebola virus VP24 is important for the transport of nucleocapsid-like structures and for the regulation of viral RNA synthesis" *J Virol*
28. Hartlieb, Modrof, Mühlberger et al. (2003) "Oligomerization of Ebola virus VP30 is essential for viral transcription and can be inhibited by a synthetic peptide" *J Biol Chem*
29. Takamatsu, Yoshikawa, Kurosu et al. (2022) "Role of VP30 phosphorylation in Ebola virus nucleocapsid assembly and transport" *J Virol*
30. Bharat, Riches, Kolesnikova et al. (2011) "Cryo-electron tomography of Marburg virus particles and their morphogenesis within infected cells" *PLoS Biol*
31. Wan, Kolesnikova, Clarke et al. (2017) "Structure and assembly of the Ebola virus nucleocapsid" *Nature*
32. Fujita-Fujiharu, Hu, Hirabayashi et al. (2025) "Structural basis for Ebola virus nucleocapsid assembly and function regulated by VP24" *Nat Commun*
33. Watanabe, Zyla, Parekh et al. (2024) "Intracellular Ebola virus nucleocapsid assembly revealed by in situ cryo-electron tomography" *Cell*
34. Dolnik, Kolesnikova, Welsch et al. (2014) "Interaction with Tsg101 is necessary for the efficient transport and release of nucleocapsids in marburg virus-infected cells" *PLoS Pathog*
35. Takamatsu, Dolnik, Noda et al. (2019) "A live-cell imaging system for visualizing the transport of Marburg virus nucleocapsid-like structures" *Virol J*
36. Mühlberger, Weik, Volchkov et al. (1999) "Comparison of the transcription and replication strategies of marburg virus and Ebola virus by using artificial replication systems" *J Virol*
37. Hoenen, Groseth, De Kok-Mercado et al. (2011) "Minigenomes, transcription and replication competent virus-like particles and beyond: reverse genetics systems for filoviruses and other negative stranded hemorrhagic fever viruses" *Antiviral Res*
38. Wenigenrath, Kolesnikova, Hoenen et al. (2010) "Establishment and application of an infectious virus-like particle system for Marburg virus" *J Gen Virol*
39. Modrof, Möritz, Kolesnikova et al. (2001) "Phosphorylation of Marburg virus VP30 at serines 40 and 42 is critical for its interaction with NP inclusions" *Virology (Auckl)*
40. Becker, Rinne, Hofsäß et al. (1998) "Interactions of Marburg virus nucleocapsid proteins" *Virology (Auckl)*
41. Kirchdoerfer, Moyer, Abelson et al. (2016) "The Ebola virus VP30-NP interaction is a regulator of viral RNA synthesis" *PLoS Pathog*
42. Tigabu, Ramanathan, Ivanov et al. (2018) "Phosphorylated VP30 of Marburg virus is a repressor of transcription" *J Virol*
43. Zhu, Song, Peng et al. (2017) "Crystal structure of the Marburg virus nucleoprotein core domain chaperoned by a VP35 peptide reveals a conserved drug target for filovirus" *J Virol*
44. Leung, Prins, Borek et al. (2010) "Structural basis for dsRNA recognition and interferon antagonism by Ebola VP35" *Nat Struct Mol Biol*
45. Leung, Borek, Luthra et al. (2015) "An intrinsically disordered peptide from Ebola virus VP35 controls viral RNA synthesis by modulating nucleoprotein-RNA interactions" *Cell Rep*
46. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
47. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
48. Enterlein, Volchkov, Weik et al. (2006) "Rescue of recombinant Marburg virus from cDNA is dependent on nucleocapsid protein VP30" *J Virol*
49. Xu, Luthra, Wu et al. (2017) "Ebola virus VP30 and nucleoprotein interactions modulate viral RNA synthesis" *Nat Commun*
50. Batra, Mori, Small et al. (2021) "Non-canonical prolinetyrosine interactions with multiple host proteins regulate Ebola virus infection" *EMBO J*
51. Prins, Binning, Shabman et al. (2010) "Basic residues within the ebolavirus VP35 protein are required for its viral polymerase cofactor function" *J Virol*
52. Takada, Robison, Goto et al. (1997) "A system for functional analysis of Ebola virus glycoprotein" *Proc Natl Acad Sci*
53. Martinez, Leung, Basler (2012) "The role of antigen-presenting cells in filoviral hemorrhagic fever: gaps in current knowledge" *Antiviral Res*
54. Halfmann, Kim, Ebihara et al. (2008) "Generation of biologically contained Ebola viruses" *Proc Natl Acad Sci*
55. Krähling, Dolnik, Kolesnikova et al. (2010) "Establishment of fruit bat cells (Rousettus aegyptiacus) as a model system for the investigation of filoviral infection" *PLoS Negl Trop Dis*
56. Hoenen, Groseth, Kolesnikova et al. (2006) "Infection of naive target cells with virus-like particles: implications for the function of Ebola virus VP24" *J Virol*
57. Kolesnikova, Berghöfer, Bamberg et al. (2004) "Multivesicular bodies as a platform for formation of the Marburg virus envelope" *J Virol*
58. Dolnik, Kolesnikova, Stevermann et al. (2010) "Tsg101 is recruited by a late domain of the nucleocapsid protein to support budding of Marburg virus-like particles" *J Virol*
59. Kolesnikova, Mittler, Schudt et al. (2012) "Phosphorylation of Marburg virus matrix protein VP40 triggers assembly of nucleocapsids with the viral envelope at the plasma membrane" *Cell Microbiol*
60. Hoenen, Jung, Herwig et al. (2010) "Both matrix proteins of Ebola virus contribute to the regulation of viral genome replication and transcription" *Virology (Auckl)*
61. Grikscheit, Dolnik, Takamatsu et al. (2020) "Ebola virus nucleocapsid-like structures utilize Arp2/3 signaling for intracellu lar long-distance transport" *Cells*
62. Biedenkopf, Hoenen (2017) "Modeling the Ebolavirus life cycle with transcription and replication-competent viruslike particle assays" *Methods Mol Biol*
63. Pettersen, Goddard, Huang et al. (2004) "UCSF Chimera-A visualization system for exploratory research and analysis" *J Comput Chem* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12651561&blobtype=pdf | # Emergence of Acquired HIV Drug Resistance Among Individuals Receiving Dolutegravir-Based Antiretroviral Therapy in Uganda: A National Laboratory-Based Survey 2023
Juliet Asio, Christine Watera, Grace Namayanja, Juliana De Fatima Da Silva, Sherri Pals, Deogratius Ssemwanga, Grace Sanyu, Maria Nannyonjo, Kabuga Usher, Kaganda, Ronald Busobozi, Hellen Nansumba, Miriam Nabukenya, Alisen Ayitewala, Mina Ssali, Cordelia Katureebe, Eleanor Magongo, Hudson Balidawa, Esther Nazziwa, Jonathan Ntale, Elliot Raizes, Du-Ping Zheng, Clement Zeh, Stephanie Hackett, Mary Naluguza, Edward Katongole Mbidde, Pontiano Kaleebu, Juliet Nkugwa
## Abstract
Background. As dolutegravir (DTG)-based regimens (DBRs) become more widely used and patients remain on treatment longer, cases of virological failure remain rare. This study presents findings from the second annual round of acquired HIV drug resistance (HIVDR) surveillance in Uganda, among individuals with viral nonsuppression (≥1000 copies/mL) receiving DBRs for ≥9 months. The first round was conducted from February to April 2022.Methods. This nationally representative cross-sectional survey analyzed randomly selected remnant plasma and dried blood spot specimens collected between April and July 2023 from the Central Public Health Laboratories. Genotyping targeted the integrase, protease, and reverse transcriptase regions of the HIV-1 pol gene, resistance was analyzed using the Stanford HIVDR Database. Weighted HIVDR prevalence and 95% confidence intervals (CIs) were calculated for children (0-14 years) and adults (≥15 years).Results. Out of 857 specimens tested, 400 (46.7%) were from children and 457 (53.3%) from adults. Median ages were 11 years for children and 36 years for adults. Median time on DBRs was 1.9 years for children and 2.4 years for adults. Five hundred and fiftynine (65.2%) specimens were successfully genotyped. The prevalence of DTG resistance was 10.1% (95% CI: 6.4%-13.9%) in children and 8.6% (95% CI: 3.9%-13.3%) in adult, higher than the first round with a prevalence of 6.6% (95% CI: 3.5%-9.6%) and 4.4% (95% CI: 0.7%-7.1%), respectivelyConclusions. The increase in DTG resistance among both children and adults highlights the need to strengthen adherence and enhance early identification of individuals at risk of HIVDR through novel and existing programmatic interventions.
The global scale-up of antiretroviral therapy (ART) has substantially reduced new Human Immunodeficiency Virus-1 (HIV) infections and AIDS-related deaths, advancing progress toward the UNAIDS 95-95-95 goals [1]. Dolutegravir (DTG), a second-generation integrase strand transfer inhibitor (INSTI), is widely adapted due to its potent efficacy, favorable safety, and high barrier to resistance [2][3][4]. In 2019, the World Health Organization (WHO) recommended DTG-based regimens (DBRs) as preferred first-line ART and second-line therapy for those failing non-nucleotide reverse transcriptase inhibitor (NNRTI)-based regimens with viral loads (VLs) ≥1000 copies/ mL [5]. In 2020, pediatric DTG formulations were approved for infants from 4 weeks old and weighing ≥3 kg, facilitating use in children [6]. By February 2024, over 25 million people globally were on DBRs [7].
Uganda began transitioning to DBRs in 2018, and by March 2023, 97% of its ∼1.32 million ART recipients were on DBRs, with VL suppression (VLS) exceeding 90% [8]. In Uganda, the VLS among individuals on DBRs is currently at 94% [9]. Similar high VLS rates (>90%) have been reported in Botswana, Zambia, Cuba, Belize, and Vietnam among those on DBRs [7,10,11]. Few studies have reported resistance to DTG among individuals receiving DTG-based ART with viral nonsuppression (VNS) [7] and by July 2023, only 3 countries had completed studies using standardized cross-sectional methods. The prevalence of resistance to DTG was reported in Ukraine at 3.7% and in Malawi at 8.6% among adults ≥18 years; and in Mozambique at 19.6% among those >15 years [7]. In Uganda, the first nationally representative survey was conducted in 2022 using the Cyclical Acquired HIV Drug Resistance (CADRE) methodology [12], and reported a prevalence of DTG resistance of 6.6% (95% CI: 3.5%-9.6%) and 3.9% (95% CI: 0.7-7.1) for children aged 0-14 years and adults aged ≥15 years, respectively [7,13]. All these studies were conducted in the past 2 years, showing that the emergence of DTG resistance has been recorded in <10 years of the global rollout.
WHO recommends that countries scaling-up and maintaining populations on DBRs routinely implement surveys of acquired HIV drug resistance (HIVDR) using standardized methods [7,10]. This will inform modeling work predicting future trends of ADR, which will in turn inform timely interventions for ART programs. Following the first-round CADRE in 2022, we conducted the second nationally representative crosssectional survey to monitor the trends of DTG resistance to inform ART programming.
## METHODS
## Study Design and Setting
This nationally representative cross-sectional survey was conducted in Uganda following CADRE guidance [12] developed by the US Centers for Disease Control and Prevention (CDC) for monitoring DTG resistance. The methods were identical to those of the first CADRE round [7,10,13]. Remnant plasma and dried blood spot (DBS) specimens collected for routine VL monitoring between April and July 2023 were retrieved from the Central Public Health Laboratories (CPHL) in Kampala, the national reference laboratory for HIV VL testing. CPHL receives specimens from all ART-providing health facilities in Uganda through the national specimen transport system. Eligible specimens were from children (<15 years) and adults (≥15 years) who had been on DBRs for ≥9 months and had VNS. All eligible samples were shipped to the Medical Research Council/Uganda Virus Research Institute & London School of Hygiene and Tropical Medicine Uganda Research Unit (MRC/UVRI & LSHTM) in Entebbe for HIVDR genotyping. This is a WHO-designated national and regional reference laboratory for resistance testing using both plasma and DBS specimens.
During the study period, a total of 1 403 705 people were active on ART in Uganda, comprising 1 349 902 (96.2%) on first-line, 51 949 (3.7%) on second line, and 1854 (0.1%) on third-line regimens. Of these, 1 129 766 individuals (80.5%) received VL test within the past 12 months, indicating VL coverage of ∼81% among people on ART. Among those tested, 1 078 689 (95.5%) achieved viral suppression.
## Population and Sampling
A total of 857 specimens, including 400 from children and 457 from adults, were randomly selected from 6896 eligible remnant specimens tested at CPHL during the study period. Eligible specimens comprised 1195 children and 5701 adults, from which samples were selected using simple random sampling without replacement, stratified evenly across the 3-month surveillance period. Specimens were excluded if they lacked accompanying clinical data.
## Data Collection
Demographic, clinical, and ART history data, including age, sex, health facility, region, VL test date, tuberculosis status, and pregnancy or breastfeeding status, were abstracted from the CPHL Laboratory Information System (LIS). The LIS captures information from laboratory requisition forms submitted with VL testing specimens. DTG initiation dates were obtained from health facilities through data calls. All records were deidentified and assigned a unique study identification number (SID) for linkage and analysis.
## Laboratory Procedures
Eligible specimens were retrieved and assessed for suitability at the CPHL repository. Specimens that met the survey criteria included DBS cards with at least 2 full blood spots or nonhemolyzed plasma specimens of ≥1 mL. Before shipment to the genotyping laboratory, selected specimens were stored at -80 °C and transported on dry ice.
## HIVDR Genotyping and Sequence Analysis
At the genotyping laboratory, HIVDR testing was done using a validated HIV-1 Genotyping Kit with integrase (Thermo Fisher Scientific) targeting the protease (codons 6-99), reverse transcriptase (codons 1-251), and integrase (codons 1-288) regions of the HIV-1 pol gene. RNA from DBS or plasma specimens was extracted using the easyMAG system (bioMérieux). The pol region was amplified by reverse transcription-polymerase chain reaction (RT-PCR) and nested PCR and sequenced using the HIV-1 Genotyping Kit with Integrase (Thermo Fisher Scientific). The cycle sequencing products were purified using the magnetic CleanSeq beads (Beckman Coulter). Sanger sequencing was done on the ABI3500xl genetic analyser (Applied Biosystems). Customized RECall software (version 2.7) program was used to edit the raw sequence data and generate consensus sequences [14]. HIVDR mutations were identified using the Stanford HIVDR database with the HIV-1 DB algorithm version 9.5.1 [15]. The laboratory is enrolled in the Virology Quality Assurance program and all sequences generated are assessed for cross-contamination by phylogenetic analysis and the sequence quality using the British Columbia Centre for Excellence (BCCFE) tool-a WHO-recommended software.
## Data Management and Analysis
Client data were de-identified and linked to HIVDR results using the SID to generate a single analytic dataset. Statistical analyses were conducted using SAS version 9.4.
Any INSTI DRM included major, accessory, and other mutations that include highly polymorphic and/or rare nonpolymorphic mutations that may be weakly associated with drug resistance. Major INSTI DRMs were primarily nonpolymorphic mutations that significantly reduced susceptibility, as classified by the Stanford HIVDR Database algorithm. DTG resistance was defined as a score ≥30, with major DTG DRMs comprising level 4 or 5 mutations in the integrase region associated with intermediate-to-high-level resistance.
Weighted HIVDR prevalence estimates with 95% confidence intervals (CIs) were calculated for each drug class. The numerator comprised all specimens with at least 1 major DRM conferring reduced susceptibility to the drug class, and the denominator included all successfully genotyped samples among participants on DBRs at the time of specimen collection.
Weighted HIVDR estimates accounted for unequal selection probabilities across months and amplification success rates. A propensity score model was used to estimate the probability of successful sequencing based on variables including VL category, specimen type, their interaction, age, sex, and region. These probabilities were then used to derive the final nonresponse weights that adjusted for unequal likelihood of amplification success. Final weights combined sampling and nonresponse adjustments to account for differential monthly sampling and amplification failure.
HIVDR analyses focused on: (1) prevalence of acquired drug resistance (ADR) in children (<15 years) and adults (≥15 years); (2) ADR patterns by age group; and (3) characterization of individuals with DTG-associated mutations.
## Patient Consent Statement
Ethical approval was obtained from the UVRI Research Ethics Committee (GC/127/834), the Uganda National Council for Science and Technology (HS 1774 ES), and the US CDC Global Health Center (Atlanta, GA), which determined the activity was not human subjects research, in accordance with applicable US federal law and CDC policy (45 CFR part 46.102(l)(2); 21 CFR part 56; 42 USC §241(d); 5 USC §552a; 44 USC §3501 et seq.). No patient contact occurred. Secondary data were collected from de-identified records using study ID numbers. A waiver of consent to use stored specimens was granted by the UVRI REC.
## RESULTS
During the study period, 10 897 eligible specimens were retrieved from CPHL (Figure 1). After excluding ineligible specimens, 6896 (63.3%) were left, comprising 1195 (17.3%) from children and 5701 (82.7%) from adults. A total of 857 specimens, including 400 (46.7%) from children and 457 (53.3%) from adults, were randomly selected for genotyping. Among these, 64.5% (n = 258) of pediatric and 63.9% (n = 292) of adult specimens were plasma (Table 1).
## Sociodemographic and Clinical Characteristics
Of the 857 participants, 54.7% (n = 469) were female, including 50.3% (n = 201) of children and 58.6% (n = 268) of adults. The median age was 11 years (interquartile range [IQR]: 8-13) among children and 36 years (IQR: 27-46) among adults. Most participants were from the central (32.6%, n = 279) and northern (32.4%, n = 278) regions, while the eastern region was least represented (13.1%, n = 112; Table 1).
Plasma samples comprised 64.2% (n = 550) of specimens, and 86.9% (n = 478) were successfully genotyped in the integrase region. DBS specimens accounted for 35.8% (n = 307), with a lower genotyping success of 26.4% (n = 81). Routine VL testing accounted for 78.2% (n = 670) of specimens, while 11.4% (n = 98) were collected after intensive adherence counseling. Many (58.5%, n = 493) had VLs <10 000 copies/mL.
Most participants (74.9%, n = 642) had been on DBRs for <3 years. Nearly, all (99.3%, n = 851) were receiving first-line ART with tenofovir disoproxil fumarate, abacavir (ABC), or zidovudine (AZT) backbones. Only 0.7% (n = 6) were on third-line regimens including darunavir/ritonavir and DTG. HIV-1 subtype A predominated (65.3%, n = 365), followed by subtype D (27.7%; Table 1).
## HIV Drug Resistance
Of the specimens successfully genotyped in the INT region, 43.8% (n = 245) were from children and 56.2% (n = 314) from adults (Table 1). Major INSTI DRMs were detected in 10.6% (n = 59) of participants, including 13.9% of children (n = 34) and 8.0% of adults (n = 25; Table 2). The weighted prevalence of major INSTI DRMs was 10.7% (95% CI: 6.9%-14.5%) among children and 8.7% (95% CI: 4.0%-13.4%) among adults. For DTG-specific mutations, the weighted prevalence of major DRMs was 10.1% (95% CI: 6.4%-13.9%) in children and 8.6% (95% CI: 3.9%-13.3%) in adults, with intermediateto-high-level resistance detected in 8.4% (95% CI: 5.0%-11.7%) and 8.0% (95% CI: 3.4%-12.6%), respectively.
In the protease/reverse transcriptase region, 473 specimens (55.3%) were successfully genotyped, including 217 from children (54.3%) and 257 from adults (56.2%). Among children, weighted prevalences of PI, NRTI, and NNRTI DRMs were 4.1% (95% CI: 1.7%-6.5%), 51.4% (95% CI: 42.1%-60.8%), and 64.3% (95% CI: 56.1%-72.6%), respectively. Over half of the children with INSTI DRMs also had NNRTI and NRTI mutations. Among adults, PI, NRTI, and NNRTI DRM prevalences were 1.7% (95% CI: 0.1%-3.4%), 23.1% (95% CI: 15.7%-30.5%), and 40.1% (95% CI: 13.0%-49.2%), respectively (Table 2).
Of the 59 participants with INSTI DRMs, 72.9% (n = 43) also had NRTI DRMs, 69.5% (n = 41) had NNRTI DRMs, and 5.1% (n = 3) had PI DRMs (Appendix, participant no: 33, 46, and 59). All 3 with 4-class resistance (PI, NRTI, NNRTI, and INSTI) were children and on regimen ABC + lamivudine + DTG (ABC + 3TC + DTG). One child (participant 59) had previously received zidovudine + lamivudine + lopinavir/ritonavir (AZT + 3TC + LPV/r) for 8 years before switching to a DBR for 17 months. Prior ART history was unavailable for the other 2 children.
## INSTI Drug Resistance Mutations
Among the 59 participants with INSTI DRMs, 25 (42.4%) had a single mutation, 9 (15.3%) had 2, 22 (37.3%) had 3, and 3 (5.1%) harbored 4 DRMs (Appendix). A higher number of DRMs was observed among children than adults (Figure 2). The most frequent mutation occurred at position 138 (n = 31), followed by positions 263 and 188 (each n = 16). The G118 mutation, associated with a strong reduction in DTG susceptibility, was detected in 16 (2.9%) participants (Figure 2).
## DTG Resistance
Most participants (n = 500, 89.4%) were susceptible to DTG, while 10 (1.8%) exhibited low-level, 17 (3.0%) intermediate- level, and 32 (5.7%) high-level resistance (Figure 3). Profiles of those with INSTI DRMs are detailed in Figure 3. DTG susceptibility declined from 96.5% in 2022 to 89.4% in 2023, alongside a general decrease in susceptibility to all INSTI drugs.
Among 49 participants with intermediate-to-high-level DTG resistance, 67.3% also had high-level resistance to NRTIs, 65.3% to NNRTIs, and 1.2% to PIs. Most (n = 37, 75.5%) remained susceptible to PIs, whereas susceptibility to NRTIs and NNRTIs was limited to 12.2% and 14.3%, respectively (Table 3).
## DISCUSSION
This study characterized ADR among persons on DBRs in Uganda, 5 years after the nationwide adoption of DTG. The majority (89.1%) of PLHIV receiving DBRs for ≥9 months remained susceptible to DTG at the time of VF, indicating continued effectiveness of DBRs within the national program.
The study found an overall prevalence of INSTI resistance of 10.1% in children and 8.6% in adults, higher than the ∼3% observed in clinical trial [7,[16][17][18]. Given that most participants had been on DBRs for ≤5 years, these findings warrant concern and point to the need for intensified adherence support and broader HIVDR surveillance to sustain ART effectiveness [7].
In addition, the Uganda Ministry of Health is strengthening differentiated adherence support through structured enhanced adherence counseling, peer-led community models, and digital follow-up for high-risk clients. Programmatic guidance now requires VL-triggered adherence interventions before any regimen switch, ensuring that treatment changes are reserved for confirmed resistance. These actions, embedded in the 2023-2027 National HIV Strategic Plan [19], emphasize adherence support as the cornerstone of sustained viral suppression under DTG-based therapy. These rates are consistent with population-based surveys conducted in Malawi (8.6%), Mozambique (11.4%), and Ukraine (6.6%) [7]. Despite methodological variations, most individuals on DBRs with VNS did not exhibit resistance, supporting WHO guidance on ongoing surveillance to inform ART decisions.
About 1 in 10 individuals with VNS had confirmed DTG resistance, indicating the need for targeted regimen switches. Interventions should prioritize routine VL monitoring, adherence support, resistance testing, and timely regimen changes when resistance is confirmed. Longitudinal studies are needed to assess outcomes among individuals transitioned to PI-based regimens, which carry a higher pill burden and adherence demands.
DTG resistance was more prevalent in treatmentexperienced individuals who transitioned to DBRs (11.5%) compared with ART-naive initiators (4.5%). Notably, 94.9% of those with major DTG DRMs had prior ART exposure. This suggests that archived resistance may be unmasked by functional monotherapy when switching to DTG without baseline resistance testing. High-level NRTI and NNRTI resistance frequently co-occurred, suggesting cumulative resistance burdens. Prior studies have similarly linked accumulated NRTI mutations to an increased risk of DTG resistance via possible synergistic effects [7,17].
The presence of M184V, a mutation associated with NRTI resistance, was common among those with DTG resistance. Though M184V alone does not directly confer resistance to DTG, recent evidence indicates that it has minimal impact on DTG susceptibility or virologic suppression when used in combination regimens such as DTG + 3TC [20,21]. A recent metaanalysis [22], and findings from cohort studies involving individuals with prior NRTI resistance, particularly the M184V/I mutation, demonstrated that prior M184V/I was not associated with an increased risk of virological failure following transition to DBRs [16,17,[21][22][23][24][25][26] and did not significantly affect virological outcomes after switching to DTG + 3TC. This provides reassurance when considering regimen switches among virologically suppressed individuals with incomplete treatment history or limited options.
The presence of thymidine analogue mutations (Appendix) may reflect unreported prior ART exposure particularly to zidovudine, or transmitted resistance, underscoring the potential for pretreatment resistance to DTG [27]. This finding reinforces the need for ongoing baseline resistance surveillance and improved documentation of treatment histories in programmatic settings.
We identified 3 children with major DRMs across all 4 ARV drug classes: NRTIs, NNRTIs, PIs, and INSTIs, despite being on first-line ART (ABC + 3TC + DTG) following transition from PI-based regimens. This highlights a critical treatment gap, considering these children exhibit resistance to all drug classes currently available in Uganda. While nonsuppressed children are eligible for enhanced treatmentsupported services (eg, caregiver treatment literacy, directly observed treatment, and community-based adherence models), our findings underscore the urgent need to scale-up and strengthen these interventions to prevent further resistance and preserve future treatment options. The detection of major DRMs to all 4 antiretroviral drug classes currently available in low-resource settings highlights the urgent need for novel treatment strategies. Clinical trials should investigate the efficacy of third-generation INSTIs and novel agents such as HIV-1 capsid inhibitors (eg, lenacapavir) and broadly neutralizing antibodies. Long-acting injectable regimens, including 3 monthly cabotegravir, currently under evaluation, may be particularly beneficial for children who rely on caregivers for adherence. Major INSTI mutation prevalence rose from 7.1% to 10.1% in children, and from 4.4% to 8.6% in adults between the previous and current surveillance rounds. Although these changes were not statistically significant, increased median DTG duration (from 1.3 to 1.9 years in children; from 1.5 to 2.4 years in adults) [13] may indicate growing drug pressure and persistent VF. A third round is needed to confirm any true increase. Given that most individuals with VF had no resistance, efforts should focus on addressing adherence challenges, reserving regimen changes for those with confirmed resistance. Current Uganda guidelines recommend switching from DTG only when resistance testing indicates intermediate-to-high-level INSTI resistance [28]. adjustments. This represents a promising, resource-sensitive approach to addressing emerging DTG resistance in Uganda.
## Limitations
This study had methodological constraints inherent to laboratorybased HIVDR surveillance using remnant specimens. Low amplification rates and incomplete data led to the exclusion of potentially eligible cases. While both plasma and DBS samples were analyzed, DBS, comprising 35% of specimens, had notably lower amplification success. This reflects known DBS limitations, including reduced RNA yield, particularly at low VLs, and less accurate representation of replicating virus. Suboptimal storage or handling may have further contributed to genotyping failure. To reduce bias, nonresponse weighting using propensity scores based on relevant covariates for amplification and resistance was applied. However, residual bias remains possible. Future efforts will focus on improving amplification success to enhance data quality. Among ART initiators, 4.5% exhibited DTG resistance, which may reflect pretreatment resistance given that most had been on therapy for <2 years. Limited sample size and short duration on DTG precluded stratified analysis by treatment duration, and therefore these findings should be interpreted cautiously. This highlights the potential importance of baseline resistance testing and ongoing surveillance to inform regimen selection and optimize treatment outcomes.
## CONCLUSION
This nationally representative survey identified emerging DTG resistance, among individuals on DBRs for under 3 years. Although the observed increase in DTG resistance was not statistically significant, ongoing surveillance is essential to detect evolving trends. Strengthened VL monitoring, timely resistance testing, and enhanced adherence support remain critical to sustaining DBR efficacy and informing program responses. Another surveillance round will be important to validate these findings and guide future ART policies in Uganda. Uganda and the World Health Organization for technical assistance.
CDC disclaimer. The findings and conclusions in this report are those of the author(s) and do not necessarily represent the official position of the funding agencies.
Potential conflicts of interest. The authors do not have any conflict of interest relevant to this work.
## References
1. "UNAIDS Global AIDS Strategy"
2. Deng, Chen, Lao (2024) "Reasons, efficacy and safety of switching to dolutegravir-based regimens among virologically suppressed PLWH: a retrospective cohort study of 96 weeks" *Infect Drug Resist*
3. Cottrell, Hadzic, Kashuba (2013) "Clinical pharmacokinetic, pharmacodynamic and drug-interaction profile of the integrase inhibitor dolutegravir" *Clin Pharmacokinet*
4. Llibre, Pulido, García et al. (2015) "Genetic barrier to resistance for dolutegravir" *AIDS Rev*
5. (2019) "Policy brief: update of recommendations on firstand second-line antiretroviral regimens. World Health Organization"
6. "Paediatric Abacavir/Lamivudine/Dolutegravir (Pald) Fixed-Dose Combination: Introduction and Rollout Planning ConsiderationsfFor National Programmes"
7. (2024) "HIV drug resistance WHO Brief report 2024"
8. Ministry, Uganda, Moh (2018) "Consolidated Guidelines for the Prevention and Treatment of HIV and AIDS in Uganda"
9. (2023) "Uganda's Electronic Health Information System"
10. (2021) "HIV drug resistance report 2021"
11. (2021) "Surveillance of acquired HIV drug resistance in populations receiving ART 2021"
12. Da Silva, Pals, Chang et al. (2021) "Monitoring emerging human immunodeficiency virus drug resistance in sub-saharan Africa in the era of dolutegravir" *J Infect Dis*
13. Watera, Silva, Namayanja (2022) "HIV-1 drug resistance among people living with HIV receiving dolutegravir-based anti-retroviral regimens in Uganda: a national laboratory-based survey using remnant viral load samples" *J Antimicrob Chemother*
14. Woods, Brumme, Liu (2012) "Automating HIV drug resistance genotyping with RECall, a freely accessible sequence analysis tool" *J Clin Microbiol*
15. (2009) "Stanford University HIV Drug Resistance Database"
16. Paton, Musaazi, Kityo (2022) "Efficacy and safety of dolutegravir or darunavir in combination with lamivudine plus either zidovudine or tenofovir for secondline treatment of HIV infection (NADIA): week 96 results from a prospective, multicentre, open-label, factorial, randomised, non-inferiority trial" *Lancet HIV*
17. Loosli, Hossmann, Ingle (2023) "HIV-1 drug resistance in people on dolutegravir-based antiretroviral therapy: a collaborative cohort analysis" *Lancet HIV*
18. West, Zeeb, Grube (2023) "Sustained viral suppression with dolutegravir monotherapy over 192 weeks in patients starting combination antiretroviral therapy during primary human immunodeficiency virus infection (EARLY-SIMPLIFIED): a randomized, controlled, multi-site, noninferiority trial" *Clin Infect Dis*
19. Commisssion (2020) "The National HIV and AIDS Strategic Plan"
20. Kabiibi, Tamukong, Muyindike et al. (2024) "Virological non-suppression, non-adherence and the associated factors among people living with HIV on dolutegravir-based regimens: a retrospective cohort study. HIV" *AIDS Res Palliative Care*
21. Santoro, Armenia, Teyssou (2022) "Virological efficacy of switch to DTG plus 3TC in a retrospective observational cohort of suppressed HIV-1 patients with or without past M184V: the LAMRES study" *J Glob Antimicrob Resist*
22. Kabra, Barber, Allavena (2023) "Virologic response to dolutegravir plus lamivudine in people with suppressed human immunodeficiency virus type 1 and historical M184 V/I: a systematic literature review and meta-analysis" *Open Forum Infect Dis*
23. Borghetti, Giacomelli, Borghi (2021) "Nucleoside reverse-transcriptase inhibitor resistance mutations predict virological failure in human immunodeficiency virus-positive patients during lamivudine plus dolutegravir maintenance therapy in clinical practice" *Open Forum Infect Dis*
24. Kingwara, Inzaule, Momanyi (2021) "Impact of nucleos(t)ide reverse transcriptase inhibitor resistance on dolutegravir and protease-inhibitor based regimens in children and adolescents in Kenya" *AIDS*
25. Van Wyk, Ajana, Bisshop (2020) "Efficacy and safety of switching to dolutegravir/lamivudine fixed-dose 2-drug regimen vs continuing a tenofovir alafenamide-based 3-or 4-drug regimen for maintenance of virologic suppression in adults living with human immunodeficiency virus type 1: phase 3, randomized, noninferiority TANGO study" *Clin Infect Dis*
26. Ciccacci, Altan, Majid (2025) "HIV dolutegravir resistance and multiclass failure in Mozambique: findings from a real-world cohort" *BMC Infect Dis*
27. Kouamou, Washaya, Ndhlovu et al. (2023) "Low prevalence of Pre-treatment and acquired drug resistance to dolutegravir among treatment naïve individuals initiating on tenofovir, lamivudine and dolutegravir in Zimbabwe" *Viruses*
28. Ministry, Uganda (2022) "Consolidated Guidelines for the Prevention and Treatment of HIV and AIDS in Uganda 2022" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548382&blobtype=pdf | # Discovery of a potent covalent inhibitor that unusually distorts the catalytic dyad of SARS-CoV-2 main protease
Juan Wang, Xiaohong Sang, Wenyan Zheng, Jasper Fuk-Woo, Chan, Jiao Zhou, Yan Xu, Pu Han, Yong Feng, Lifeng Fu, Jessica Oi, Ling Tsang, Shuofeng Yuan, Aaron Ciechanover, Jing An, Kwok-Yung Yuen, Jianxun Qi, Ziwei Huang
## Abstract
Proteases are versatile therapeutic targets for a wide variety of human diseases, including cancer, cardiovascular diseases, and infections by viruses, bacteria, and parasites. The main protease (M pro ) of severe acute respiratory syndrome corona virus 2 (SARS-CoV-2), the virus that causes coronavirus disease 2019 (COVID-19), has been one of the major drug targets for the treatment of COVID-19. In this study, a small-molecule covalent inhibitor, H102, of SARS-CoV-2 M pro was developed from serial structural modifications on a compound previously reported by us to have anti-SARS-CoV-2 activity. H102 exhibited very high potency in SARS-CoV-2 M pro inhibition with IC 50 of 8.8 nM and strongly prevented SARS-CoV-2 replication in VeroE6 cells. The co-crystal structure of H102 bound to M pro determined at 1.50 Å resolution provided a structural mechanism of H102's action and revealed an unusual distortion of the catalytic dyad of the viral enzyme which was caused by the benzyl ring of P2 position of H102 interacting with the side chain of catalytic dyad His41 residue of M pro and pulling His41 side chain away from the other catalytic dyad Cys145 residue. This structural mechanism of H102 is very different from that of other reported covalent M pro inhibitors for which His41 side chain orientation and its close proximity to Cys145 remain unchanged with the binding of a covalent inhibitor. As such, H102 may serve as a biochemical probe for investigat ing further an unusual mechanism of the viral enzyme's catalytic dyad disruption and inhibition, and as a prototype distinct from other reported covalent inhibitors for the development of novel antiviral agents. IMPORTANCE A nanomolar potent small-molecule inhibitor, H102, of SARS-CoV-2 M pro was developed and exhibited strong anti-SARS-CoV-2 infection activity in cells. Co-crystal structure determination of its complex with M pro provided a structural mechanism of H102's action and revealed an interesting structural feature: the benzyl ring at the P2 position of H102 interacts with the reorientated His41 side chain, accompanied by a significant increase of the distance between the catalytic dyad Cys145-His41 residues, which is uncommon in reported covalent inhibitors. Compound H102 may be used as a biochemical probe to further investigate mechanisms of M pro inhibition and potentially different type of lead for developing antiviral agents for treating disease caused by novel coronavirus SARS-CoV-2. KEYWORDS SARS-CoV-2, main protease, covalent inhibitor, catalytic dyad, biochemical probe P roteases are found in many different viruses, including the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that causes the coronavirus disease 2019 (COVID-19) pandemic (1). These viral proteases are responsible for cleaving the viral
precursor polyproteins at specific sites into structural proteins and functional proteins crucial for viral replication (2,3).
The main protease (M pro ) of SARS-CoV-2, also known as 3-chymotrypsin-like (3CL) protease, is a cysteine protease essential for viral replication and one of the major therapeutic targets for treating COVID-19 (4,5). Nirmatrelvir (PF-07321332), a SARS-CoV-2 M pro inhibitor developed by Pfizer (6), has been approved clinically to treat COVID-19 in combination with Ritonavir. Many other small-molecule inhibitors and degraders of SARS-CoV-2 M pro have also been reported and are under development (5,(7)(8)(9)(10)(11)(12)(13)(14), in addition to other targeting and intervention strategies (15).
Previously, we reported the discovery of compound 17 (16), a small-molecule inhibitor of SARS-CoV-2 infection from screening a panel of α-ketoamide analogs using cell viability and plaque reduction assays. Here, in this study, compound 17 was used as the starting template to design and synthesize a series of analogs containing modifications at various moieties of the compound including its warhead. This led to the development of a covalent inhibitor H102, an aldehyde analog possessing potent activities in inhibiting M pro enzyme function and SARS-CoV-2 infection in cells. Co-crystal structure of H102 bound to M pro was determined at 1.50 Å resolution, which revealed very interesting information about the structural mechanism of this novel inhibitor compared to other reported covalent or noncovalent inhibitors.
## RESULTS
## Development of a nanomolar potent M pro inhibitor H102
The previously reported compound 17 (16) with modest M pro inhibitory potency (~45% inhibition at 100 µM) was used as a starting template for synthetic efforts to develop more potent inhibitory analogs. This led to the identification of H96, which contains substitutions at the Cap, P1 (a six-membered lactam Gln mimic), and warhead positions (Table 1). H96 showed over a 10-fold increase in M pro inhibitory potency (IC 50 of 9.4 µM) compared to compound 17.
We changed the Cap of H96 to Boc, resulting in H94, which showed weaker inhibition of SARS-CoV-2 M pro than H96 (Table 1). Nevertheless, H94 was used for further structural optimization because its structure was more amenable to synthesis than H96. Different side chain substitutions were incorporated at the P3 position of H94, resulting in H97-99 and H137 with the substituent tert-Leucine (Tle) for H97 yielding the best effect (Table 1). Thus, this substituent was kept for the P3 position, while we attempted two different P2 modifications, as in H100 and H101, finding that the P2-Phe in H100 was better. Finally, we changed the warhead of H100 from ketoamide to aldehyde while keeping the above-described P2-Phe and P3-Tle with better or best effect. This led to our final lead compound H102 displaying the most potent M pro inhibition among all analogs (IC 50 of 8.8 nM, Table 1). H102 was more potent than another reported aldehyde M pro inhibitor GC376 (IC 50 of 24.5 nM) and Nirmatrelvir (PF-07321332, IC 50 of 22.2 nM), the clinically approved M pro inhibitor drug by Pfizer (6), in M pro inhibition assays (Fig. 1A).
## Kinetic and enzyme inhibition characterization of H102
For further kinetic and enzyme inhibition characterization studies of H102 identified from our screening study as described above, we expressed and purified the SARS-CoV-2 M pro protein in E. coli as reported previously (18). The K m value of Dabcyl-KTSAVLQSGFRKME-Edans for SARS-CoV-2 M pro was determined to be 28.6 ± 3.1 μM (Fig. 1B), which is similar to the previously reported value 28.2 ± 3.4 µM (19). The Ki of H102 was determined to be 2.89 nM (Fig. 1C), which was in line with its high inhibitory potency measured by IC 50 . In order to understand the noncovalent and covalent contributions to high potency of H102, the noncovalent affinity (K I ) and the reaction rate of cova lent modification (k inact ) were calculated. H102 displays high noncovalent affinity (K I = 6.84 nM) and fast reaction rate (k inact = 0.0014 S -1 ), as shown in Fig. 1D. These results indicate both noncovalent and covalent contributions to the high potency of H102. The recovery of M pro activity indicated that both PF-07321332 and GC376 are reversible covalent M pro inhibitors in jump dilution experiments (Fig. 2A). GC376 showed slower M pro activity recovery velocity compared to PF-07321332 (Fig. 2B). H102 with the same aldehyde warhead as GC376 behaved as an irreversible-characteristic inhibitor. The enzymatic activity H102-M pro complex did not recover after 100-fold dilution within 1 h (Fig. 2A). The result suggested that a stable H102-M pro complex was formed. The reorientation of the His41 side chain may result in a reverse reaction barrier. Therefore, H102 displayed slower M pro activity recovery velocity (Fig. 2B). These results show an interesting conversion from a reversible covalent inhibitor to an irreversible inhibitor likely due to the conformational change of a key residue.
To exclude any possible fluorescence quenching effect of H102, we used Fmoc-Glu-EDANS to mimic C-terminal EDANS product Ser-Gly-Phe-Arg-Lys-Met-Glu(EDANS), and the fluorescence intensity of Fmoc-Glu-EDANS solution was measured in the presence or absence of 100 nM of H102 for 1 h. As shown in Fig. S1, the addition of H102 showed no change or effect on the fluorescence intensity of Fmoc-Glu-EDANS in solvent (DMSO) only without the compound, demonstrating that H102 had no fluorescence quenching effect.
## Inhibition by H102 of SARS-CoV-2 replication in cells
The lead compound H102 was examined for anti-SARS-CoV-2 activity in VeroE6 cells by viral load reduction assay. H102 effectively decreased viral replication by 3.55 log 10 at 10 µM and 1.34 log 10 at 2.5 µM, respectively (Fig. 3). As controls, PF-07321332 and GC376 showed 3.48 log 10 and 3.93 log 10 reductions, respectively, in viral RNA load at 10 µM. H102 inhibited viral replication with an EC 50 of 168.9 nM in VeroE6 cells without cytotoxicity at much higher concentrations (over 50 μM, Fig. S2).
## Co-crystal structure determination of H102 in complex with SARS-CoV-2 M pro
To elucidate the mechanism of H102 action, co-crystal structure of H102 bound to SARS-CoV-2 M pro was determined at the resolution of 1.50 Å (Fig. 4A). As revealed in this structure, the overall interactions of H102 with SARS-CoV-2 M pro are similar to those in the published co-crystal structures of other reported M pro covalent inhibitors, except for the orientation of the benzyl ring at the P2 position of H102 and the geometry of M pro catalytic dyad Cys145-His41 residues (Fig. 4B) (6,10,20). In the complex structure of H102 with M pro , the benzene ring of H102's P2 position forms π-π interaction with the side chain of His41 of M pro and causes His41 side chain to undergo a drastic confor mational change. In the inhibitor-free structure of SARS-CoV-2 M pro , His41 and Cys145 constitute the catalytic dyad. The imidazole group of His41, by accepting a proton from Cys145, activates the nucleophilic attack reaction. The distance between the NE2 atom of His41 and the Sγ atom of Cys145 is 3.6 Å in the inhibitor-free structure, whereas this distance is increased to 8.5 Å after H102 binding (Fig. 4C). The torsion angle NCαCβCγ(χ1) of His41 is 78.9° (gauche -, g -) in the inhibitor-free structure but changes to -155.1° (trans, t) after H102 binding (Fig. 4C), which seems to indicate that the side chain of His41 in the inhibitor-free structure is in the least occupied conformational state because the g - side chain conformation has a lower propensity (12.2%) than g + (54.6%) and t (33.2%) for the side chain conformation occurrence of His residues in proteins (21). The benzyl ring at the P2 position of H102 is sandwiched between His41 and Cys145 (Fig. 4B), which completely distorts and blocks the catalytic dyad interactions and provides a structural basis for the highly potent inhibitory activity of H102.
## DISCUSSION
In this study, we attempted to optimize compound 17 previously reported by us to have anti-SARS-CoV-2 activity (16). After a series of structural modifications at various positions of the compound, including the cap, P1, P2, P3, and warhead, we were able to discover a highly potent SARS-CoV-2 M pro aldehyde inhibitor H102 whose M pro inhibitory potency was improved by ~1,000-fold over the starting compound H96, an analog of compound 17. This demonstrated the efficacy of the structural modification strategy. Compound H102 displayed very high potency in inhibiting M pro enzymatic function (IC 50 of 8.8 nM) and was more potent than control compounds PF-07321332 and GC376 in comparative M pro inhibition assays. Furthermore, H102 was effective in blocking SARS-CoV-2 replication in VeroE6 cells, demonstrating its utility as an antiviral agent that merits further development.
The high-resolution co-crystal structure at 1.50 Å resolution of H102 bound to the viral M pro protein revealed an interesting mechanism of action of this inhibitor. Unlike crystal structures of other reported covalent M pro inhibitors with a similar benzyl ring at P2 position, such as 11b and MPI4 (9,20), the benzyl ring of P2 position of H102 was projected in between the catalytic dyad of His41 and Cys145, which is different from that of 11b and MPI4 (Fig. 5B, C andG). This sandwiched interaction between H102's benzyl ring and the enzyme's catalytic dyad, as observed in the crystal structure, caused the reorientation of His41 side chain away from its inhibitor-free state and thus effectively disrupted the enzyme's catalytic dyad geometry and function, which may structurally rationalize the strong M pro inhibitory activity of H102. We further compared the structural mechanism of H102 with other covalent inhibitors of M pro that have other different groups at the P2 position and different warheads. As shown in Fig. 5D through F, the M pro bound crystal structures of three representative covalent inhibitors including 13b (22), PF-07321332 (23), and GC376 ( 17), all displayed little or no distorting effect on the geometry of the catalytic dyad Cys145-His41 residues as shown by the little or no change in the orientation of His41 side chain and distance between His41 side chain and Cys145. Taken together, H102 seems to be different from other reported covalent inhibitors in the unusual and significant distortion of the orientation of His41 side chain and geometry of Cys145-His41 interaction of viral enzyme's catalytic dyad unseen in other covalent inhibitors. Whether the above-described distortion of the catalytic dyad is implicated in the enzyme's functional mechanism awaits further investigation. From other published studies in the literature, conformational variation or flexibility of an enzyme's catalytic residues is commonly observed. About two-thirds of catalytic centers are flexible to perform enzyme functions, based on analysis of more than 60,000 protein structures in the PDB database (24). Specifically for SARS-CoV-2 M pro , since it cleaves polyproteins pp1a and pp1b at 11 distinct sites, this requires that the catalytic pocket of M pro has sufficient flexibility to accommodate various substrates. Most M pro substrates have small-sized Leu and Val in their P2 position, whereas one of the substrates has a bulky Phe in the P2 position (25). Recent studies using dynamical nonequilibrium MD simulations suggested that the S2 subpocket of M pro undergoes substantial conforma tional rearrangement when bound by substrates with large residues such as Phe at the P2 position (26). In light of these observations reported in the literature, the distortion of the catalytic dyad due to the conformational change of catalytic residue His41 reported in our present study is in line with the notion of the enzyme's S2 subpocket flexibility being a part of its functional mechanism.
While the structural mechanism of H102 appears to be uncommon among cova lent inhibitors of M pro as described above, similar observations have been reported for noncovalent inhibitors, such as S-217622, a noncovalent oral SARS-CoV-2 M pro inhibitor with IC 50 of 13 nM (7), and compound 3, a noncovalent SARS-CoV M pro inhibitor with IC 50 of 300 nM (27). In the co-crystal structures of these noncovalent inhibitors bound to M pro , the 2,4,5-trifluorobenzylic moiety of S-217622 and the phenyl ring of compound 3 form face-to-face π interactions with the rotated side chain of His41, similar to the case of H102 (Fig. 5G-I). These compounds-H102, S-217622, and compound 3-share a similar feature of having aromatic groups interacting favorably with a positively charged His41 side chain. This interaction may facilitate the His41 side chain to cross the energy barrier from the less occurring g -conformational state (in the inhibitor-free structure) to the more occurring t state (inhibitor-bound structure), thus allowing the observed conforma tional change to occur. Whether this interesting structural mechanism shared by one covalent inhibitor reported here and two noncovalent inhibitors recently reported by others could be exploited for M pro inhibitor design remains to be further investigated. It is intriguing to postulate that the two noncovalent inhibitors discussed here may take advantage of their distorting effect on the geometry of catalytic dyad Cys145-His41 distance and interaction for more effective blockade of the enzyme function. On the other hand, covalent inhibitors devoid of such a distorting effect on the catalytic dyad, except for the rare case of H102 reported here, use covalent bond formation with Cys145 to gain binding free energy and disable Cys145's catalytic role. Would it be possible that H102, capable of both of these two inhibition mechanisms adopted by noncovalent and covalent inhibitors, respectively, suggests a new and different prototype of more advantageous inhibitors? This possibility seems to be supported by a recently published study by others of the design of SARS-CoV-2 M pro inhibitors based on the concept of dual inhibition (i.e., disrupting the catalytic dyad's His41 while covalently inhibiting Cys145) (28), which is reminiscent of the finding and notion described in our study here.
## Conclusion
A nanomolar potent small-molecule inhibitor, H102, of SARS-CoV-2 M pro was developed from serial structural modifications starting from our previously reported anti-SARS-CoV-2 compound 17 (16). Compound H102 exhibited strong anti-SARS-CoV-2 infection activity in cells. Co-crystal structure determination of its complex with M pro provided a structural mechanism of H102's action and revealed an interesting structural feature involving the benzyl ring of P2 position of H102 interacting with the reorientated His41 side chain and significant increase of the distance between the catalytic dyad Cys145-His41 residues which is uncommon in reported covalent inhibitors. Compound H102 may be used as a biochemical probe to further investigate mechanisms of M pro inhibition and potentially different type of lead for developing antiviral agents for treating disease caused by novel coronavirus SARS-CoV-2.
## MATERIALS AND METHODS
## Compound synthesis and characterization
The details of synthetic methods for preparing target compounds, as well as 1 H NMR and 13 C NMR spectra of intermediates and target compounds, are provided in the supple mental material.
## Protein expression and purification of SARS-CoV-2 M pro
The pET-28b-SARS-CoV-2 M pro plasmid was transformed into E. coli strain BL21(DE3) cells and then cultured in LB medium containing 50 µg/mL kanamycin in a shaking incubator at 37°C. When the cells were grown to an OD 600 of 0.6-0.8, 0.6 mM IPTG was added to the cell culture to induce the protein expression at 16°C. After 18 h, the cells were harvested by centrifugation at 4,000 rpm for 20 min at 4°C. The cell pellets were washed twice by PBS, resuspended in lysis buffer (50 mM HEPES, 300 mM NaCl, 10 mM imidazole, pH 7.5), and lysed by sonication on ice using 3-second ON/5-second OFF cycles for a total of 30 min. The lysate was then clarified by ultracentrifugation at 18,000 rpm at 4°C for 40 min to remove debris. The supernatants were then purified by TALON metal affinity resin (TaKaRa, 635501) and washed with washing buffer (25 mM HEPES, 500 mM NaCl, pH 7.5) to remove unspecific binding proteins. The SUMO-His-tagged SARS-CoV-2 M pro was eluted by elution buffer (25 mM HEPES, 500 mM NaCl, 300 mM imidazole, pH7.5). The SUMO-His-tagged SARS-CoV-2 M pro was then treated overnight at 4°C with His-tagged SUMO protease (home-made) to remove the SUMO-His-tag. The SARS-CoV-2 M pro was further purified by His60 Ni superflow resin (TaKaRa, 635659). The quality of SARS-CoV-2 M pro was checked by Coomassie-stained SDS-PAGE gel. The concentration was determined via BCA Protein Assay Kit. The purified SARS-CoV-2 M pro was stored in (25 mM HEPES, 1 mM DTT, 1 mM EDTA, 10% glycerol, pH 7.5).
## SARS-CoV-2 M pro enzyme inhibition assay
SARS-CoV-2 M pro enzyme inhibition assay for the evaluation of M pro inhibitory potency of compounds was performed according to a commonly used method reported by others (29,30). Specifically, the enzyme inhibition assay was carried out in assay reaction buffer (25 mM HEPES, 1 mM DTT, 1 mM EDTA, 0.01% Triton X-100, pH 7.5) by pre-incubating 85 µL SARS-CoV-2 M pro (final concentration of 150 nM in a total volume of 100 µL) with 5 µL compounds at various concentrations. For H102, the final concentration range was from 0.03 nM to 500 nM (in a total volume of 100 µL). The mixture was incubated at 37°C with gentle shaking for 30 min in blank 96-well plates. While we did not know whether our compounds to be tested had slow onset binding, we adopted the preincubation step reported by others in our enzyme inhibition assays (30,31). Next, 10 µL of 250 µM M pro fluorogenic substrate (Dabcyl-KTSAVLQSGFRKME-Edans, final concentration of 25 µM in a total volume of 100 µL) was added to the reaction mixture, after which the plate was incubated at 37°C for 1 h. The relative fluorescence units (RFU) were measured at a single time point after 1 h of incubation using a PerkinElmer EnVision multimode plate reader with an excitation wavelength of 340 nm and an emission wavelength of 490 nm. Percent inhibition was calculated based on control wells containing no compound (100% activity) and a blank control. The IC 50 values were calculated using GraphPad Prism software. All experiments were performed in triplicate, and the values are presented as mean ± SD.
## Kinetic assay
K m value for substrate Dabcyl-KTSAVLQSGFRKME-Edans was determined at seven different concentrations, ranging from 2.5 to 140 µM (19). The relative fluorescence units (RFU) were monitored continuously by addition of SARS-CoV-2 M pro (final concen tration of 100 nM) for 10 min. K m value was obtained by nonlinear regression using the Michaelis-Menten plot. Then, 85 µL SARS-CoV-2 M pro (final concentration of 10 nM in a total volume of 100 µL) was added to blank 96-well plates, and 5 µL H102 with various concentrations (final concentration in a total volume of 100 µL: 0, 1, 3, 5, 10, 20, and 30 nM) and 10 µL of 250 µM substrate (final concentration of 25 µM in a total volume of 100 µL) were added immediately. The RFU was measured every 2 min for 2 h by a PerkinElmer EnVision multimode plate reader. The K i value was calculated from duplicate measurements by nonlinear regression using Morrison Ki equation (Y = V 0 × ( 1 -((((E t (6,32,33).
$$+ X + (K i × (1 + (S/K m )))) -(((E t + X + (K i × (1 + (S/K m ))))^2) -4 × E t × X)^0.5))/(2 × E t ))))$$
## Reversibility assay
To evaluate the reversibility of SARS-CoV-2 M pro inhibition by the compounds, SARS-CoV-2 M pro at a final concentration of 1 µM was incubated with 1 µM H102, PF-07321332, GC376, or DMSO for 30 min. Then, 1 µL reaction mixture was diluted 100-fold with 99 µL reaction buffer (25 mM HEPES, 1 mM DTT, 1 mM EDTA, 0.01% Triton X-100, pH 7.5) containing fluorogenic substrate at the concentration of 25 µM in blank 96-well plates. The RFU was measured immediately every 1 min for 1 h by a PerkinElmer EnVision multimode plate reader with an excitation wavelength of 340 nm and an emission wavelength of 490 nm. The fractional velocity was determined by dividing the velocity of added compound after 100-fold dilution by the velocity with the solvent DMSO. Three independent experiments were carried out. The values are expressed as the mean ± SD.
## Fluorescence control assay
We conducted a fluorescence control experiment using Fmoc-Glu-EDANS to mimic C-terminal EDANS product Ser-Gly-Phe-Arg-Lys-Met-Glu(EDANS) with or without the compound. Briefly, 5 µL H102 (final concentration of 100 nM in a total volume of 100 µL) and solvent (DMSO) and 10 µL Fmoc-Glu-EDANS (final concentration of 25 µM in a total volume of 100 µL) were added to 85 µL reaction buffer (25 mM HEPES, 1 mM DTT, 1 mM EDTA, 0.01% Triton X-100, pH 7.5) in blank 96-well plates. The RFU was immediately measured every 1 min for 1 h by a PerkinElmer EnVision multimode plate reader with an excitation wavelength of 340 nm and an emission wavelength of 490 nm. Three independent experiments were carried out. The values are expressed as the mean ± SD.
## Cytotoxicity assay
The cytotoxicity of H102 was measured by CellTiter-Glo Luminescent Cell Viability Assay (Promega, G7570). Briefly, VeroE6 cells were seeded into 96-well plates and incubated overnight. On the second day, cells were treated with serially diluted concentrations of the H102 for 48 h. On the day of analysis, CellTiter-Glo reagents were added to induce cell lysis. After incubating at room temperature for 10 min in the dark, luminescent signal was detected using a PerkinElmer EnVision multimode plate reader. The experiments were performed in triplicate, and the values are presented as mean ± SD.
## SARS-CoV-2 viral load reduction assay
Viral load reduction assay was performed for the evaluation of antiviral activity as we described previously (34). Briefly, SARS-CoV-2-infected VeroE6 cells were treated with different concentrations of compounds or dimethyl sulfoxide (DMSO) control. Then, cell culture supernatant samples were collected at 48 h post-inoculation (hpi) for qRT-PCR analysis of viral RNA load. Culture supernatant was lysed with buffer AVL and then extracted for total RNA with the QIAamp viral RNA mini kit (Qiagen). qRT-PCR was used for quantitation of SARS-CoV-2 viral load using the QuantiNova Probe RT-PCR kit (Qiagen) with a LightCycler 480 Real-Time PCR System (Roche). Each 20 µL reaction mixture contained 10 µL of 2 × QuantiNova Probe RT-PCR Master Mix, 1.2 µL of RNase-free water, 0.2 µL of QuantiNova Probe RT-Mix, 1.6 µL each of 10 µM forward and reverse primer, 0.4 µL of 10 µM probe, and 5 µL of extracted RNA as the template. Reactions were incubated at 45°C for 10 min for reverse transcription, followed by 95°C for 5 min for denaturation, and then subjected to 45 cycles of 95°C for 5 s and 55°C for 30 s. Signal detection and measurement were taken in each cycle after the annealing step. The cycling profile ended with a cooling step at 40°C for 30 s. The primers and probe sequences were against the RNA-dependent RNA polymerase/helicase (RdRP/Hel) gene region of SARS-CoV-2: forward primer: 5′-CGCATACAGTCTTRCAGGCT-3′; reverse primer: 5′-GTGTGATGTTGAWATGACATGGTC-3′; specific probe: 5′-FAMTTAAGATGTGGTGCTTGCA TACGTAGAC-IABkFQ-3′. The viral load reduction assay experiments were performed in triplicate.
## Crystallization of SARS-CoV-2 M pro in complex with H102
Concentrations of 5 mg/mL and 10 mg/mL M pro (in a solution containing 20 mM Tris, 150 mM NaCl, 1 mM EDTA, and 1 mM TCEP [pH 7.8]) were incubated with 10 mM H102 at 1:10 vol ratio at room temperature for 2 h. The crystals were obtained by using the sitting-drop vapor diffusion method by mixing 1 µL of protein with 1 µL of reservoir solution and then equilibrating the mixture against 100 µL of the reservoir solution at 18°C. The initial crystallization screenings were carried out using commercially available kits. The complexes were crystallized in a solution containing 0.1 M MES monohydrate pH 6.0 and 20% (wt/vol) polyethylene glycol monomethyl ether 2,000.
## Data collection and structure determination
Diffraction data were collected at the Shanghai Synchrotron Radiation Facility (SSRF) BL17U (wavelength, 0.97918 Å). For data collection, the crystals were cryoprotected by briefly soaking in reservoir solution supplemented with 20% (vol/vol) glycerol before flash-cooling in liquid nitrogen. The data set was processed with HKL2000 software. The structure was determined by the molecular replacement method using Phaser with the previously reported structure (PDB: 7C6S). The atomic models were built using Coot and refined with phenix.refine in Phenix, and the stereochemical qualities of the final models were assessed with MolProbity. Data collection, processing, and refinement statistics are summarized in Table S1.
## Statistical analysis
Each experiment was performed independently at least three times. Statistical analy ses were conducted using GraphPad Prism (version 8.0). Error bars indicate standard deviations. P-values were calculated using Student's t-test to compare each compound candidate with DMSO control. *P < 0.05.
## References
1. Zhou, Yang, Wang et al. (2020) "A pneumonia outbreak associated with a new coronavirus of probable bat origin" *Nature*
2. Zephyr, Yilmaz, Schiffer (2021) "Viral proteases: structure, mechanism and inhibition" *Enzymes*
3. Luan, Huynh, Cheng et al. (2020) "Targeting proteases for treating COVID-19" *J Proteome Res*
4. V'kovski, Kratzel, Steiner et al. (2021) "Coronavirus biology and replication: implications for SARS-CoV-2" *Nat Rev Microbiol*
5. Jin, Du, Xu et al. (2020) "Structure of Mpro from SARS-CoV-2 and discovery of its inhibitors" *Nature*
6. Owen, Allerton, Anderson et al. (2021) "An oral SARS-CoV-2 M pro inhibitor clinical candidate for the treatment of COVID-19" *Science*
7. Unoh, Uehara, Nakahara et al. (2022) "Discovery of S-217622, a noncovalent oral SARS-CoV-2 3CL protease inhibitor clinical candidate for treating COVID-19" *J Med Chem*
8. Qiao, Li, Zeng et al. (2021) "SARS-CoV-2 M pro inhibitors with antiviral activity in a transgenic mouse model" *Science*
9. Dai, Zhang, Jiang et al. (2020) "Structure-based design of antiviral drug candidates targeting the SARS-CoV-2 main protease" *Science*
10. Zhang, Lin, Sun et al. (2020) "Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors" *Science*
11. Sang, Wang, Zhou et al. (2023) "A chemical strategy for the degradation of the main protease of SARS-CoV-2 in cells" *J Am Chem Soc*
12. Desantis, Bazzacco, Eleuteri et al. (2024) "Design, synthesis, and biological evaluation of first-in-class indomethacin-based PROTACs degrading SARS-CoV-2 main protease and with broad-spectrum antiviral activity" *Eur J Med Chem*
13. Alugubelli, Xiao, Khatua et al. (2024) "Discovery of first-in-class PROTAC degraders of SARS-CoV-2 main protease" *J Med Chem*
14. Cheng, Feng, Li et al. (2024) "Development of novel antivrial agents that induce the degradation of the main protease of human-infecting coronaviruses" *Eur J Med Chem*
15. Su, He, Xie et al. (2025) "Enabling the immune escaped etesevimab fully-armed against SARS-CoV-2 Omicron subvariants including KP"
16. Wang, Liang, Chen et al. (2021) "A new class of α-ketoamide derivatives with potent anticancer and anti-SARS-CoV-2 activities" *Eur J Med Chem*
17. Fu, Ye, Feng et al. (2020) "Both boceprevir and GC376 efficaciously inhibit SARS-CoV-2 by targeting its main protease" *Nat Commun*
18. Sang, Wang, Zhou et al. (2023) "A chemical strategy for the degradation of the main protease of SARS-CoV-2 in cells" *J Am Chem Soc*
19. Ma, Sacco, Hurst et al. (2020) "Boceprevir, GC-376, and calpain inhibitors II, XII inhibit SARS-CoV-2 viral replication by targeting the viral main protease" *Cell Res*
20. Yang, Ma, Ma et al. (2021) "A quick route to multiple highly potent SARS-CoV-2 main protease inhibitors*" *Chem MedChem*
21. (2025) *Full-Length Text Journal of Virology*
22. Chakrabarti, Pal (2001) "The interrelationships of side-chain and mainchain conformations in proteins" *Prog Biophys Mol Biol*
23. Zhang, Lin, Sun et al. (2020) "Crystal structure of SARS-CoV-2 main protease provides a basis for design of improved α-ketoamide inhibitors" *Science*
24. Zhao, Fang, Zhang et al. (2022) "Crystal structure of SARS-CoV-2 main protease in complex with protease inhibitor PF-07321332" *Protein Cell*
25. Riziotis, Ribeiro, Borkakoti et al. (2022) "Conformational variation in enzyme catalysis: a structural study on catalytic residues" *J Mol Biol*
26. Shaqra, Zvornicanin, Huang et al. (2022) "Defining the substrate envelope of SARS-CoV-2 main protease to predict and avoid drug resistance" *Nat Commun*
27. Chan, Oliveira, Mulholland et al. (2023) "Substrate recognition and selectivity in SARS-CoV-2 main protease: unveiling the role of subsite interactions through dynamical nonequili brium molecular dynamics simulations" *Biochemistry*
28. Lu, Mahindroo, Liang et al. (2006) "Structure-based drug design and structural biology study of novel nonpeptide inhibitors of severe acute respiratory syndrome coronavirus main protease" *J Med Chem*
29. Zhu, Meng, Feng et al. (2024) "De novo design of SARS-CoV-2 main protease inhibitors with characteris tic binding modes" *Structure*
30. Bai, Ye, Feng et al. (2021) "Structural basis for the inhibition of the SARS-CoV-2 main protease by the anti-HCV drug narlaprevir" *Signal Transduct Target Ther*
31. Riva, Yuan, Yin et al. (2020) "Discovery of SARS-CoV-2 antiviral drugs through large-scale compound repurposing" *Nature*
32. Higashi-Kuwata, Tsuji, Hayashi et al. (2023) "Identification of SARS-CoV-2 M pro inhibitors containing P1' 4-fluorobenzothiazole moiety highly active against SARS-CoV-2" *Nat Commun*
33. Ma, Shang, Yang et al. (2018) "Iminooxazolidin-2-one as a bioisostere of the cyanohydrin moiety: inhibitors of enterovirus 71 3C protease" *J Med Chem*
34. Hu, Lewandowski, Tan et al. (2023) "Naturally occurring mutations of SARS-CoV-2 main protease confer drug resistance to nirmatrelvir" *ACS Cent Sci*
35. Yuan, Yin, Meng et al. (2021) "Clofazimine broadly inhibits coronaviruses including SARS-CoV-2" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12848123&blobtype=pdf | # Nano-synthetic biology for disease diagnosis and treatment
Feng Li, Lizeng Gao, Fang Yang, Ning Gu, Xiyun Yan, Xian-En Zhang
At the beginning of the 21st century, synthetic biology emerged from the development of systems biology [ 1 ]. Synthetic biology follows the philosophy of "Build-to-Learn, Build-to-Use. " It helps deepen our understanding of the laws of life and drives the iterative progress of biotechnology through the simulation, design, synthesis, and modification of biological systems using engineering principles. Multidisciplinary convergence is a hallmark characteristic of synthetic biology. It has demonstrated promising applications across a wide range of fields, such as biomedicine, agriculture and food, chemical engineering and materials, daily chemicals, energy, enzyme engineering, biosensing, and more [ 2 ]. During the same period, with the convergence of nanotechnology and biology, nanobiology emerged as a new frontier [ 3 ]. Nanobiology investigates biological systems from the perspective of nanoscience. By utilizing revolutionary nanotechnology-enabled research tools, it aims to understand the structures, functions, and working mechanisms of biological systems at the microscopic scale, while also exploring the biological effects of nanomaterials. Additionally, nanobiology seeks to create bio-based or biomimetic functional nanostructures and nanodevices by leveraging materials, methods, and tools from biotechnology and nanotechnology. In the context of disciplinary convergence, nanobiology and synthetic biology have merged naturally in recent years, giving rise to a new direction within synthetic biology known as nano-synthetic biology (NSB). Synthetic biology and nanobiology are complementary fields of study. Tools from synthetic biology enhance nanobiology's capability to design and fabricate bio-based nanomaterials and devices. Conversely, nanomaterials and techniques developed through nanobiology enable synthetic biology to achieve functions beyond those found in nature. Endeavors in this field would be devoted to addressing the challenges in modularization, standardization, functional integration, and intelligentialization [ 4 ].
This special issue highlights one of the most active application fields of NSB, i.e., disease diagnosis and treatment, comprising three Reviews and three Full Articles.
Yao et al. [ 5 ] review highlights the significant progress and future potential of nano-oncology in transforming cancer therapy. While the field demonstrates remarkable preclinical success, critical challenges remain in tumor-specific accumulation, immune modulation, and manufacturing standardization. The convergence of AI-driven personalized medicine with nanotechnology presents an exciting frontier. Success in the future will depend on multidisciplinary collaboration to overcome translational bottlenecks through scalable production and robust clinical validation, ultimately bridging the gap between innovative nanotherapies and patient care.
Li et al. [ 6 ] review recent progress in bioinspired artificial ion channels (AICs) based on macrocycles for biomedical applications, highlighting their design, ion transport properties, and applications in antibacterial therapy, anticancer therapy, biosensing, and channelopathy treatment. These AICs could be useful components for constructing hybrid cells with enhanced or reprogrammed functions.
Li et al. [ 7 ] review gene circuit-based cell-free biosensors in the context of synthetic biology, emphasizing their design concepts, construction principles, applications, and advantages in biomanufacturing, environmental monitoring, and medical diagnosis.
Liang et al. [ 8 ] provide an elegant review of the connections between synthetic biology and nanoscience, highlighting how NSB achieves molecular-level precision in bioengineering. By elucidating the design principles of bio-nano hybrid systems from smart nanocarriers to genetic networks, the authors showcase NSB-based applications in precision medicine, biosensing, and biocatalysis. However, key challenges for NSB, such as biocompatibility, in vivo predictability, and ethical considerations, remain imperative. Future efforts should focus on scalable synthesis, stimulus-responsive designs, and regulatory guidelines to accelerate NSB's clinical implementation and maximize its biomedical impact.
Gao et al. Huang et al. [ 10 ] develop a self-assembled nanoparticle based on functionalized peptides to deliver siRNA for antitumor therapy, inhibiting tumor growth in the G1 phase with a p16 minimal inhibitory sequence (p16MIS) and in the G2 phase with siRNA down-regulating the polo-like kinase 1 (PLK1) gene. Such a design not only enhances the penetration and delivery efficiency of siRNA but also synergistically promotes apoptosis of tumor cells. Importantly, such peptide-based siRNA delivery nanoplatform also works in other cell lines, such as macrophages, and thus provides an effective gene delivery system for targeting cancer treatment.
Collectively, the six papers showcase the latest advances in NSB for disease diagnosis and treatment utilizing various representative systems, providing valuable insights into this rapidly evolving field. As guest editors of this special issue, we express our gratitude to all authors for their high-quality contributions, to the editorial board members for their encouraging support, and to all referees for their helpful feedback.
## References
1. Cameron, Bashor, Collins (2014) "A brief history of synthetic biology" *Nat. Rev. Microbiol*
2. Ding, Li, Wang (2020) "Significant research progress in synthetic biology" *Synthet. Biol. J*
3. Zhang (2020) "Nanobiology-symphony of bioscience and nanoscience" *Sci. China-Life Sci*
4. Zheng, Wu, Li (2022) "Integration of synthetic biology and nanobiotechnology for biomedical applications" *Synthet. Biol. J*
5. Teng, Bi, Xing (2025) "Nano-oncology revisited: Insight on precise therapeutic interventions of tumor" *Fundam. Res*
6. Liu, Shi, Li (2025) "Supramolecular macrocyclic artificial ion channels for biomedical applications" *Fundam. Res*
7. Guo, Li, Zuo (2025) "Gene circuit-based sensors" *Fundam. Res*
8. Ji, Li, Guo (2025) "Nanoscale synthetic biology with innovative medicinal applications" *Fundam. Res*
9. Kou, Xu, Yin (2025) "Intranasal delivery of rotigotine to the brain for treating Parkinson's disease" *Fundam. Res*
10. Essola, Yang, Liu (2025) "Tumor suppressor protein-inspired peptide for siRNA delivery and synergistic cancer therapy" *Fundam. Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12403532&blobtype=pdf | # SDS-EDTA-treated chromatography paper strips for sampling and transportation of monkeypox virus, varicella zoster virus and herpes simplex 1 and 2 viruses
M E T H O D O Lo, Nikola Sklenovská, Mandy Bloemen, Graciela Andrei, Katrien Bruyninckx, Elke Wollants, Marc Van Ranst
## Abstract
aches, back pain, and fatigue, followed by lymphadenopathy and a distinctive rash. The rash usually starts on the face and spreads to the palms and soles, evolving through macular, papular, vesicular and pustular stages before crusting over within two to four weeks. Clinical diagnosis of mpox remains challenging as it symptoms closely resemble those of other rash-associated diseases, including varicella-zoster virus (VZV) infections (chickenpox or herpes zoster), measles, and sexually-transmitted diseases such as genital herpes caused mainly by herpes simplex virus 2 (HSV-2) and occasionally herpes simplex virus 1 (HSV-1), as well as syphilis and other conditions [1, 2]. Consequently, laboratory confirmation is essential for accurate diagnosis of monkeypox virus (MPXV) infection.
## Introduction
The recent resurgence of mpox in both endemic and non-endemic regions has drawn renewed attention to its clinical presentation and diagnostic challenges. This viral zoonosis typically manifests with an initial phase of flu-like symptoms such as fever, chills, headache, muscle MPXV-containing materials are classified as Category A (UN2814) infectious substances for transport, requiring the use of triple packaging to ensure biosafety. Furthermore, MPXV specimens must be stored under refrigerated or frozen conditions and ideally transported to the laboratory within seven days of collection [3]. While resource-rich countries are generally able to maintain these stringent cold chain requirements, remote or hard-to-reach forested areas in endemic regions often face significant infrastructure challenges. These include a lack of reliable freezing capabilities, transport delays that lead to prolonged transit times, and difficulties in implementing WHO-recommended biosafety measures during specimen handling.
The use of room-temperature-stable SDS-EDTAtreated chromatography paper strips for sample collection offers a promising solution to logistical challenges and costs associated with mpox testing, especially when combined with appropriate secondary packaging. Our previously validated method has been successfully applied to poliovirus, rotavirus, and norovirus [4][5][6]. In this study, we extend the application of this approach to enable the simultaneous detection of MPXV and common differential diagnosis targets including VZV, HSV-1 and HSV-2.
## Materials and methods
## Preparation of SDS-EDTA-treated chromatography paper strips
Paper strips were prepared as previously described [4]. Briefly, Whatman 17 Chr pure cellulose chromatography paper with a thickness of 0.92 mm, a high absorbency of 870 g water/m2 and a linear flow rate of 190 mm/30 min was used (Cytiva Whatman, Kent, UK). The chromatography paper was cut into strips of 80 mm x 4 mm, while wearing disposable gloves to prevent contamination. The strips were immersed for 2 min in a solution containing 2% (w/v) sodium dodecyl sulfate (SDS), 10 mM ethylenediaminetetraacetic acid (EDTA), and 60 mM Tris-HCl [tris(hydroxymethyl) aminomethane hydrochloride]. Following immersion, the strips were airdried overnight at room temperature (20-25 °C).
## Cells
Human embryonic lung cells (HEL-299, CCL-137 ATCC) were grown in Dulbecco's Modified Eagle's Medium (DMEM, Thermo Fisher Scientific) supplemented with 10% foetal calf serum (FCS), 2 mM l-glutamine, 0.1 mM non-essential amino acids, 1 mM sodium pyruvate, and 10 mM HEPES, at 37 °C in a 5% CO 2 humidified atmosphere.
## Viruses
The following viral strains were used: MPXV (GenBank accession number ON622712.1), successfully isolated from genital lesion swabs from a Belgian patient, HSV-1 (Kos strain, ATCC VR-1493), HSV-2 (G strain, ATCC VR-3393), and VZV (Oka strain, ATCC VR-1832). Viral stocks were prepared and titrated in HEL cells: MPXV (2.5 × 10 7 PFU/mL), HSV-1 (2 × 10 6 PFU/mL), HSV-2 (3 × 10 5 PFU/mL), and VZV (4.5 × 10 4 PFU/mL).
All MPXV-related work was conducted in the highcontainment BSL3 facilities of the KU Leuven Rega Institute under the official permit with reference number AMV/02062020/S88219.2020/0221, and according to the institutional guidelines. VZV and HSV work was conducted in BSL2 facilities at the Rega Institute following biosafety guidelines.
## Sample dilutions, loading on the chromatography paper strips and incubation
All viral stocks (MPXV, VZV, HSV-1 and HSV-2) were serially diluted 1:100, 1:1.000, 1:10.000 and 1:100.000. The SDS-EDTA-treated chromatography paper strips were immersed for approximately 1 min (until the paper was saturated) into the different dilutions of the vital stocks. The strips were allowed to air dry overnight at room temperature. After complete drying, each strip was kept in a separate 50 mL Eppendorf tube and stored under different temperature conditions, -20 °C, 4 °C, room temperature (22 °C), and 37 °C. Storage durations included 1, 7, 14, 60, and 120 days. Negative controls for SDS-EDTA strips (without application of virus) were used for MPXV in each time point.
## Viral inactivation tests
To assess potential viral survival on SDS-EDTA-treated chromatography paper strips, viral clearance studies were conducted using cell-based infectivity assays.
A 1:10 dilution of each virus (MXPV, VZV, HSV-1, and HSV-2) was applied to both SDS-EDTA-treated and untreated chromatography paper strips. The strips were then placed in sterile tubes containing 3 mL Universal Transport Medium (UTM), vigorously mixed and filtered through a 0.45 μm membrane. Subsequently, 0.1 mL of this UTM filtrate was inoculated onto HEL cells (for MPXV and VZV) or E1SM cells (for HSV-1 and HSV-2). Cell cultures were monitored daily for cytopathic effects over a 14-day period. Viral titers were quantified via serial 10-fold dilutions of the samples in 96-well microtiter plates pre-seeded with HEL or E1SM cells.
## Viral DNA extraction
After each storage time interval at the different temperatures, the DNA extraction was performed on the virusladen strips. The SDS-EDTA strips were cut in half and put into an Eppendorf tube with 1000 µl of RNA-free water. After a 10-minute incubation, the SDS-EDTA strips were squeezed thoroughly with the pipette tip to release absorbed viral material. A 400 µL aliquot of the resulting supernatant was used for viral nucleic acid extraction using the MagMAX™ Viral/Pathogen Nucleic Acid Isolation Kit, automated on a on Kingfisher Flex-96 purification system (ThermoFisher Scientific, Europe). Negative controls for DNA extraction step were used for virus at each time point.
## Real-time polymerase chain reaction (qPCR)
Primer and probe sequences are provided in Supplementary Data (Table 1). MPXV primers and probe [7] and HSV-1/HSV-2 primers and probes [8] were published previously. qPCR amplification was conducted on a QuantStudio 7 Flex Real Time PCR system (Applied Biosystems, ThermoFisher). To amplify the targets of interest, a mix for each virus was made using 5 µL TaqMan™ Fast Virus 1-step Master Mix (Applied Biosystems) with 1 µL of each forward and reverse primers (stock concentration 20 µM) and 0.5 µL probe (stock concentration 10 µM). Supplemented with 7.5 µL RNase free water to a total of 15 µL. A total of 5 µL viral DNA was added to the reaction mixes. Thermal cycling conditions were 20 s at 95 °C, followed by 45 cycles of 3 s at 95 °C and 30 s at 60 °C. Analysis was done using QuantStudio Real-Time PCR software (Applied Biosystems, ThermoFisher).
## Result and discussion
The test strips were immersed in a buffer containing SDS and EDTA as primary components. SDS is a potent anionic surfactant that denatures proteins by disrupting non-covalent bonds, effectively neutralizing viral infectivity. EDTA acts as a divalent cation chelator, binding magnesium and calcium ions to inhibit metallonuclease activity that hydrolyses nucleic acids. The buffer also included Tris-HCl, a weak organic base to maintain a stable pH range (8.0-8.5), optimizing EDTA's cation-sequestration capacity. The loss of microbial infectivity was proven by the biosafety tests. The prevention of nucleic acid degradation was proven by qPCR amplification.
In the viral clearance studies, HEL cells and E1SM cells infected with the different sample dilutions were observed for 2 weeks. The cell culture exposed to samples with SDS-EDTA-treated chromatography paper strips stayed negative without detection of cytopathic effect at dilutions ≥ 1:1000, while toxic effects on the cells were observed at dilutions ≤ 1:100. In contrast, cells exposed to the non-treated paper strips showed evidence of viral growth.
To determine whether viral DNA remained present and detectable after storage at different temperatures (-20 °C, 4 °C, 22 °C and 37 °C) for various durations (1 day, 7 days, 30 days, 90 days and 120 days), a qPCR was performed. The number of copies per µL was determined using a standard curve for which mean CT values from duplicate reactions were used (Supplementary data, Fig. 1).
We were able to detect all MPXV and VZV dilutions after all storage times under all storage temperatures. There were two negatives among HSV samples -HSV-1 in dilution 10 -5 stored in -20 °C for 120 days and HSV-2 in dilution 10 -5 stored in 22 °C for 30 days. This could be explained by the expectation of a copy number of fewer than 10 viral copies/µL at this dilution and pipetting errors. HSV-2 in dilution 10 -5 was detected again in all consecutive timepoints (90 and 120 days) (Fig. 1). The viral concentration tested ranged from 27,306 to 3 viral copies/µL for MPXV, 415,767 to 2 viral copies/µL for VZV, 18,484 to 2 viral copies/µL for HSV-1 and 41,729 to 1 viral copies/µL for HSV-2.
The low viral concentrations we were able to detect for MPXV make the strips clinically relevant, especially given the fact that they are meant to be used on skin lesions which are reported to contain the highest number of viral particles among all clinical samples. We were able to detect viral concentrations as low as 2 copies/µL, with some studies reporting high viral loads exceeding 10 9 to 10 12 copies/µL in skin lesions [9,10]. Notably, detection reliability remained unaffected by elevated temperatures and prolonged storage periods. Additionally, the preparation of the SDS-EDTA-treated chromatography strips is very easy, quick and cheap. The approximate cost of goods (chromatography paper, chemicals) to prepare one single SDS-EDTA-treated chromatography strip is 0,10 €. The total processing time to prepare 280 SDS-EDTAtreated chromatography strips by a laboratory technician is approximately one hour.
We were able to detect low viral concentrations after prolonged exposure (120 days) to high temperatures (37 °C). This stability is particularly relevant for countries endemic for mpox, where such conditions are common. These findings could significantly simplify the transportation of samples from remote, hard-to-reach locations to diagnostic laboratories.
## References
1. Jezek, Szczeniowski, Paluku et al. (1988) "Human monkeypox: confusion with chickenpox" *Acta Trop*
2. Wieder-Feinsod, Zilberman, Erster et al. (2023) "Overlooked Monkeypox cases among men having sex with men during the 2022 outbreak -a retrospective study" *Int J Infect Dis*
3. (2024) "Geneva: World Health Organization"
4. Wollants, Maes, Thoelen et al. (2004) "Evaluation of a Norovirus sampling method using sodium Dodecyl sulfate/EDTA-pretreated chromatography paper strips" *J Virol Methods*
5. Rahman, Goegebuer, Leener et al. (2004) "Chromatography paper strip method for collection, transportation, and storage of rotavirus RNA in stool samples" *J Clin Microbiol*
6. Maes, Van Doren, Thoelen et al. (2004) "Poliovirus sampling by using sodium Dodecyl sulfate/EDTA-pretreated chromatography paper strips" *Biochem Biophys Res Commun*
7. Sklenovská, Bloemen, Vergote et al. (2023) "Design and validation of a laboratory-developed diagnostic assay for Monkeypox virus" *Virus Genes*
8. Raymenants, Geenen, Budts et al. (2023) "Indoor air surveillance and factors associated with respiratory pathogen detection in community settings in Belgium" *Nat Commun*
9. Suñer, Ubals, Tarín-Vicente et al. (2023) "Viral dynamics in patients with Monkeypox infection: a prospective cohort study in Spain" *Lancet Infect Dis*
10. Nörz, Brehm, Tang et al. (2022) "Clinical characteristics and comparison of longitudinal qPCR results from different specimen types in a cohort of ambulatory and hospitalized patients infected with Monkeypox virus" *J Clin Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12613913&blobtype=pdf | # Complete genome sequences of four Phikmvirus bacteriophages from Kenyan sewage lytic against multidrug-resistant Pseudomonas aeruginosa
Tweya Meshack, Omwega, Moses Gachoya, Martin Georges, Collins Kigen, James Wachira, Vanessa Natasha, Mikeljon Nikolich, Kevin Kamanyi, Samuel Nyamweya, Erick Odoyo, Lillian Musila
## Abstract
We describe four Phikmvirus Pseudomonas phages isolated from sewage samples in Kenya: vB_PaePA01phi1_RS1, vB_PaePA10145Phi1_RS1, vB_PaePA8132phi1_PS3, and vB_PaePA10145phi1_HR2. Their genomes range from 43,723 to 45,485 bp with 62.24%-62.37% GC content and 65-68 coding sequences. KEYWORDS bacteriophages, bacteriophage therapy, Pseudomonas aeruginosa, Phikmvirus, multi-drug resistance, genomic characterization, Kenya H ospital-acquired Pseudomonas aeruginosa exhibits high resistance to antibiotics, which can cause severe mortality rates and disease severity in hospitalized patients (1, 2). In the context of antibiotic resistance, bacteriophages are promising as potential alternatives to conventional antibiotics (2). We report the complete genome sequences of four lytic phages against P. aeruginosa.Phages vB_PaePA01phi1_RS1, vB_PaePA10145phi1_RS1, vB_PaePA8132phi1_PS3, and vB_PaePA10145phi1_HR2 were isolated in 2022 from four distinct 500 mL grab sewage samples collected on different days from informal settlements in Nairobi, Kenya. The sewage samples were filtered (0.22 µm) to remove bacteria before enrichment in Tryptic Soy Broth (37°C, 180 rpm, 18 h), with the P. aeruginosa strains PA01, MRSN 10145, or MRSN 8132 (Table 1). The enrichment culture was centrifuged and filtered (0.22 µm) to obtain the phages (3) which were then purified through three rounds of sequential single-plaque isolation using the double-layer agar method. High-titer stocks were prepared by infecting host cultures at a Multiplicity of Infection (MOI) of ~0.01 until complete lysis (4-6 h), followed by centrifugation and filtration (4). Prior to phage DNA extraction, the bacterial host DNA and RNA were removed with DNase I and RNase A, respectively. The QIAmp Mini Kit (Qiagen, Germantown, MD, USA) was used following the manufacturer's instructions. DNA purity and quantity were assessed using a Nanodrop One Spectrophotometer and a Qubit fluorometer (Thermo Fisher Scientific Inc., Waltham, MA, USA), respectively. Paired-end sequencing libraries were prepared using the Illumina DNA Prep Kit and whole-genome sequencing on the Illumina NextSeq 1000 (Illumina Inc., San Diego, CA, USA) using a Reagent Kit (300 cycles, 2 × 150 bp reads). FastQC v0.11.9 (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) examined read quality. Raw reads were trimmed using BBduk v38.18 and Dedupe v.38.18 (https://github.com/BioInfoTools/BBMap). Trimmed reads were assembled using SPAdes v4.0.0 (5), and assembly quality assessed by QUAST v5.3.0 (6). The proportion of mapped reads was determined by Bowtie2 v2.5.4 (7). Genomes were annotated with Pharokka v1. 7.4 (8), MMseqs2 v13.45111 (9),. Phage lifestyle was predicted using PhageLeads (https://phageleads.dk/) (13), PhageTerm v1.0.12 (14) was used to determine termini and packaging mechanisms, and CheckV v1.0.3 (15) was used
for genome completeness. Classification based on Mash distance to the top hits was determined against the INPHARED database v1.8 (16). Default parameters were used for all software.
The characteristics of these phages are summarized in Table 1. Comparative genomic analysis (96%-98% similarity NCBI databases) classified the four phages in the genus Phikmvirus. This is consistent with recently published phages by Peters et al. (17), which share close taxonomic placement (Fig. 1A). No tRNAs were detected in these phages, typical of this genus (18,19). Genome comparison using Clinker v0.0.31 (20) shows gene order conservation while also highlighting unique coding sequence variations (Fig. 1B) which may influence infection dynamics (21).
Functional annotation confirmed essential components for virus structure (22, 23) and the lysis cassettes encoding holins, endolysins, and spanins. All four phages lack genes that confer lysogenic lifestyle antibiotic resistance and toxin generation, making them promising candidates for therapeutic applications (24) FIG 1 (A) Whole-genome phylogenetic tree of seven bacteriophages constructed using the Neighbor-Joining method. The analysis includes the four phages isolated in this study together with three closely related phages previously reported by Peters et al. (17). Bootstrap support values are shown at the major nodes.
(B) CDS cluster comparison of the four Phikmvirus phages generated with Clinker (https://github.com/gamcil/clinker). Greyscale links between genomes indicate amino acid identity, while similarity groups are highlighted with unique colors.
## References
1. Edward, Shehawy, Abouelfetouh et al. (2023) "Prevalence of different virulence factors and their association with antimicrobial resistance among Pseudomonas aeruginosa clinical isolates from Egypt" *BMC Microbiol*
2. Pang, Raudonis, Glick et al. (2018) "Antibiotic resistance in Pseudomonas aeruginosa: mechanisms and alternative therapeutic strategies" *Biotechnol Adv*
3. Abedon (2011) *Lysis from without. Bacteriophage*
4. Luong, Salabarria, Edwards et al. (2020) "Standardized bacteriophage purification for personalized phage therapy" *Nat Protoc*
5. Bankevich, Nurk, Antipov et al. (2012) "SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing" *J Comput Biol*
6. Mikheenko, Prjibelski, Saveliev et al. (2018) "Versatile genome assembly evaluation with QUAST-LG" *Bioinformatics*
7. Langdon (2015) "Performance of genetic programming optimised Bowtie2 on genome comparison and analytic testing (GCAT) bench marks" *BioData Min*
8. Bouras, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics*
9. Steinegger, Söding (2017) "MMseqs2 enables sensitive protein sequence searching for the analysis of massive data sets" *Nat Biotechnol*
10. Mcnair, Zhou, Dinsdale et al. (2019) "PHANOTATE: a novel approach to gene identification in phage genomes" *Bioinformat ics*
11. Bouras, Grigson, Papudeshi et al. (2024) "Dnaapler: a tool to reorient circular microbial genomes" *J Open Source Softw*
12. Laslett, Canback (2004) "ARAGORN, a program to detect tRNA genes and tmRNA genes in nucleotide sequences" *Nucleic Acids Res*
13. Yukgehnaish, Rajandas, Parimannan et al. (2022) "PhageLeads: rapid assessment of phage therapeutic suitability using an ensemble machine learning approach" *Viruses*
14. Garneau, Depardieu, Fortier et al. (2017) "PhageTerm: a tool for fast and accurate determination of phage termini and packaging mechanism using next-generation sequencing data" *Sci Rep*
15. Nayfach, Camargo, Schulz et al. (2021) "CheckV assesses the quality and completeness of metagenomeassembled viral genomes" *Nat Biotechnol*
16. Cook, Brown, Redgwell et al. (2021) "INfrastructure for a PHAge REference database: identification of large-scale biases in the current collection of cultured phage genomes" *PHAGE*
17. Peter, Kirillina, Georges et al. (2025) "Complete genome sequences of three Pseudomonas aeruginosa phages of the genus Phikmvvirus" *Microbiol Resour Announc*
18. Alvi, Asif, Tabassum et al. (2020) "RLP, a bacteriophage of the family Podoviridae, rescues mice from bacteremia caused by multi-drug-resistant Pseudomonas aeruginosa" *Arch Virol*
19. Lavigne, Burkal'tseva, Mv et al. (2003) "The genome of bacteriophage φKMV, a T7-like virus infecting Pseudomonas aeruginosa" *Virology (Auckl)*
20. Gilchrist, Chooi (2021) "Clinker & clustermap.js: automatic generation of gene cluster comparison figures" *Bioinformatics*
21. Piel, Bruto, Labreuche et al. (2021) "Genetic determinism of phage-bacteria coevolution in natural populations" *bioRxiv*
22. Cuervo, Fàbrega-Ferrer, Machón et al. (2019) "Structures of T7 bacteriophage portal and tail suggest a viral DNA retention and ejection mechanism" *Nat Commun*
23. Straka, Dubinová, Liptáková (2022) "Phascinating phages" *Microorganisms*
24. Pirnay, Blasdel, Bretaudeau et al. (2015) "Quality and safety requirements for sustainable phage therapy products" *Pharm Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724216&blobtype=pdf | # TDP-43 promotes efficient HSV-1 replication in human DRGderived neurons
Shirley Braspenning, Denise Ohnezeit, Olivia Degulis, Angus Wilson, Ian Mohr
## Abstract
TAR DNA-binding protein 43 (TDP-43) is a versatile nuclear RNA-binding protein that performs important functions in RNA localization, processing, and stability. In the neurodegenerative disease amyotrophic lateral sclerosis (ALS) TDP-43 forms toxic, insoluble cytoplasmic aggregates that ultimately lead to neuronal loss. Although TDP-43 is expressed in every cell type, its function and subcellular localization are particularly important for neuronal homeostasis. However, it is unknown if TDP-43 has a role during herpesvirus infection. Herpes simplex virus type-1 (HSV-1), a ubiquitous neurotropic pathogen, is considered a contributing factor to neurodegenerative disorders. In this study, we tested the requirement for TDP-43 during HSV-1 infection in neuronal and non-neuronal cells. HSV-1 infection of epithelial cells and primary fibroblasts did not change overall TDP-43 abundance, nor did TDP-43 depletion detectably alter HSV-1 productive replication in a multicycle growth experiment. By contrast, when TDP-43 was depleted in neuronally-derived, differentiated HD10.6 cells, HSV-1 infectious virus production was significantly reduced in both single-and multicycle growth experiments. Notably, TDP-43 depletion restricts viral lytic gene expression at the immediate-early phase. Through nanopore direct RNA-sequencing, we uncovered enhanced intron retention in two essential viral genes-ICP0 and UL15-upon TDP-43 depletion. Thus, while depletion of TDP-43 does not detectably affect HSV-1 reproduction in epithelial cells and fibroblasts, TDP-43 is required for efficient replication in HD10.6 cells through modifying the abundance and splicing of viral mRNAs. IMPORTANCE Herpes simplex virus type-1 is a widespread neurotropic pathogen that can cause life-threatening infections of the brain and is increasingly linked to neuro degenerative disease. However, due to the lack of scalable in vitro human neuronal models or small animal models that recapitulate disease, little is known about virus-host interactions in neurons specifically. Using human epithelial cells, primary fibroblasts and a human neuron-derived cell line, we uncovered a cell type specific TDP-43 requirement for efficient HSV-1 virus replication. TDP-43 is a critical neuronal disease factor gene, and we showed it promotes HSV-1 gene expression and splicing of viral mRNAs in neuron-derived cells. This raises the possibility that targeting of TDP-43 could reveal a new antiviral strategy for severe HSV-1 infections. This work further provides valuable insights into the possible etiology of neurodegenerative disease and highlights the importance of studying virus-host interactions in relevant cell types.
Subsequent research identified roles for TDP-43 in regulation of transcription (5,6), miRNA biogenesis (7,8), mRNA splicing (3,(9)(10)(11)(12)(13)(14), nucleocytoplasmic shuttling (15), stress granule formation (16,17), and translation (18)(19)(20) in a variety of cell types. Most recently, TDP-43 has garnered attention for its involvement in several neurodegenera tive diseases, particularly amyotrophic lateral sclerosis (ALS). In ALS, unknown neuronal stress triggers cause TDP-43 to form insoluble proteotoxic aggregates in the cytoplasm that ultimately drive motor neuron loss (21). The extent to which virus infection stress might influence TDP-43 function and whether TDP-43 potentially impacts virus infection biology beyond HIV is incompletely understood and understudied.
A limited number of molecular virus-host interactions involving TDP-43 have been reported, suggesting that TDP-43 may play a role during virus infection. Enterovirus infection was shown to cleave TDP-43 into shorter, more aggregate prone isoforms leading to the formation of TDP-43 inclusion bodies in vivo (22)(23)(24). Similarly, a viral protein in Theiler's murine encephalomyelitis virus was shown to induce the formation of cytoplasmic TDP-43 aggregates during infection (25). HSV-2 infection mildly upregulates TDP-43 in mouse spinal cords, but it is unclear if its intracellular localization is affected (26), and HSV-1 ICP0 reportedly promotes ubiquitination of TDP-43 (27). Finally, TDP-43 was found to bind and stabilize influenza A viral mRNAs, promoting infection (28). This raises the possibility that TDP-43 might impact the replication of other viruses, including neurotropic alphaherpesviruses like herpes simplex virus type-1 (HSV-1).
HSV-1 is a widespread human pathogen that establishes life-long latent infection in sensory ganglionic neurons (29). Reactivation of HSV-1 in response to neuronal stress leads to new infectious virus production and is generally associated with mild pathology such as cold sores but can lead to life-threatening infection in the brain in neonates and immunocompromised individuals. While numerous host factors are effectively hijacked during HSV-1 infection to remodel the intracellular environment, neutralize host defenses, and facilitate virus productive replication, our understanding of how cell type specific roles for host factors, specifically in neuronal-derived cells, might impact HSV-1 reproduction remains limited. In particular, how a pivotal neuronal stress responsive disease gene like TDP-43 that functions in RNA processing might influence HSV-1 gene expression and replication is unknown.
Here, we investigate the role of TDP-43 during HSV-1 infection of human epithelial cells, fibroblasts, and neuron-like cells. Following TDP-43 depletion in epithelial cells and primary fibroblasts, differences in productive HSV-1 replication were not detected. However, TDP-43 depletion was found to restrict virus replication in differentiated human HD10.6 cells, which are derived from immortalized dorsal root ganglia (DRG) sensory neurons and develop neuronal morphology upon differentiation. Depletion of TDP-43 in HD10.6 cells globally reduced virus gene expression and resulted in increased intron retention in mRNAs encoding ICP0, a critical master regulatory protein, and UL15, a key DNA packaging factor. Our data indicate that TDP-43 promotes global lytic gene expression and efficient splicing of key virus genes in a neuronal cell type-specific manner. It further suggests that normal TDP-43 levels are dispensable for viral replica tion in non-neuronal cells but are required for efficient viral replication in a human DRG-derived neuronal cell line.
## RESULTS
## TDP-43 is dispensable for HSV-1 replication in epithelial cells and fibroblasts
To characterize the role of TDP-43 during HSV-1 infection, we first asked if the over all abundance of TDP-43 protein was changed upon HSV-1 infection. ARPE-19 retinal pigmented epithelial cells and primary, normal human dermal fibroblasts (NHDF) were infected with HSV-1 and probed for TDP-43 protein levels by immunoblot at 9 h post-infection (hpi). In both cell types, HSV-1 infection did not detectably alter TDP-43 abundance under these conditions (Fig. 1A). To determine whether TDP-43 is required for efficient HSV-1 replication, ARPE-19s and NHDFs were transfected with two independent siRNAs targeting TDP-43 or a non-targeting control siRNA and infected at low MOI with HSV-1 for 48 hpi. Whereas transfection with either TDP-43 siRNA resulted in effective depletion of TDP-43 in ARPE-19s and NHDFs compared to control, non-targeting siRNA (Fig. 1B), infectious virus titers were not detectably altered during this multi-cycle growth experiment (Fig. 1C andD). Thus, TDP-43 protein abundance is not detectably changed by HSV-1 infection and TDP-43 depletion did not detectably interfere with productive HSV-1 replication in epithelial cells and primary fibroblasts.
## Differentiated HD10.6 cells as a model for lytic HSV-1 infection of neurons
Although TDP-43 is a ubiquitously expressed RNA-binding protein, TDP-43 associated pathology is predominantly limited to the central nervous system. Therefore, we investigated whether there is a unique role for TDP-43 during infection of neuronalderived cells compared to epithelial cells or fibroblasts. To model HSV-1 infection in neurons, we used HD10.6 cells. This cell line was derived from fetal human dorsal root ganglia (DRG)-a natural target cell for HSV-1 infection in vivo-and has been immor talized through tetracycline-regulated v-myc overexpression (30). Upon treatment with neuronal growth factors and doxycycline (dox), HD10.6 cells stop proliferating, express neuronal transcription factors, develop a neuronal morphology including connecting axons, exhibit nociceptive properties, and are permissive for HSV-1 productive replica tion (31). Here, we refer to HD10.6 cells treated with growth factors and dox as neuronlike or differentiated reflecting their immortalized origins and others have referred to this terminal state as "matured. " We confirmed that upon differentiation, HD10.6 cells express elevated β3-tubulin and TrkA compared to proliferating controls, form dense neuronal networks, and can be infected with HSV-1 leading to expression of viral lytic genes of all kinetic classes and ultimately the production of cell-associated infectious virus (Fig. 2).
## TDP-43 depletion restricts HSV-1 replication in a multicycle growth curve in differentiated HD10.6 cells
We first asked if HSV-1 infection of differentiated HD10.6 cells alters TDP-43 mRNA and protein abundance. While a small increase in TDP-43 mRNA was detected by RT-qPCR in HSV-1 infected compared to uninfected (Mock) HD10.6 cells (Fig. 3A), this change was not statistically significant, and corresponding changes in TDP-43 protein levels were not observed (Fig. 3B) in agreement with findings in non-neuronal cells (Fig. 1A). To determine if TDP-43 influences HSV-1 infection in differentiated HD10.6 cells, TDP-43 was depleted using transduction of lentiviruses delivering two independent shRNAs targeting TDP-43 or a non-targeting control shRNA (shNTC). Compared to cultures transduced with shNTC, TDP-43 protein abundance was substantially reduced by similar levels using either of the two TDP-43-specific shRNAs (Fig. 3C andD). Since differentiated HD10.6 cells do not exhibit cytopathic effect upon HSV-1 infection, we used an HSV-1 reporter that expresses eGFP fused to the N terminus of the true late protein Us11 (HSV-1 GFP-Us11) allowing virus replication to be readily monitored in infected cultures (32). Notably, the depletion of TDP-43 in differentiated HD10.6 cells resulted in a marked reduction in the accumulation of GFP-expressing (GFP+) cells at 3 days post-infection (Fig. 3E andF). Accordingly, significantly less-approximately 15-foldinfectious virus was produced as determined by plaque assay from differentiated HD10.6 cells in which TDP-43 was depleted compared to the non-targeting control lentivirus (Fig. 3G). In accordance with reduced viral replication, we confirmed that expression of essential viral lytic genes representing immediate-early (ICP27) or late (UL36) kinetic classes was decreased in differentiated HD10.6 cells depleted for TDP-43 compared to cultures transduced with a non-targeting shRNA lentivirus (Fig. 3H). To ensure that the observed decrease in viral replication was not due to changes in neuronal health upon TDP-43 depletion, we determined ATP levels of NTC shRNA and TDP-43 shRNA transduced cells as a proxy for cell viability. Detectable differences in cell viability were not observed between shNTC, shTDP43-#1 and shTDP43-#2 at 4 days post-transduction (Fig. 3I). Additionally, we probed these cells for β3-tubulin and did not observe differences in signal intensity and neuronal networks (Fig. 3J andK), suggesting that TDP43 knockdown in differentiated HD10.6 cells does not affect neuronal health under these conditions. Together, our data show that while TDP-43 depletion does not detectably interfere with viral replication in epithelial cells and fibroblasts, it restricts HSV-1 gene expression and infectious virus production in a human DRG-derived cell line. This raises the possibility that TDP-43 is differentially required for HSV-1 replication in differentiated HD10.6 cells, but dispensable for productive virus growth in non-neuronal cells.
## HSV-1 viral lytic gene expression is limited upon TDP-43 knockdown
HSV-1 lytic replication is broadly divided into three kinetic phases: immediate-early, early, and late. To establish when in the viral life cycle TDP-43 depletion impacts HSV-1 gene expression, we performed a single-cycle infection. In our hands, differentiated HD10.6 cells were relatively resistant to HSV-1 infection, and only at an MOI of at least 30, based upon virus titered on permissive cells, were the majority of differentiated HD10.6 cells infected at 24 hpi, and viral titers appeared to plateau (Fig. 4A andB). Notably, even under high MOI conditions and in a restricted time frame, TDP-43 depletion significantly reduced the number of GFP+ cells (Fig. 4C andD) and reduced infectious virus production by approximately fourfold (Fig. 4E), suggesting TDP-43 depletion restricts HSV-1 replication within a single replication cycle. Evaluat ing expression of representative virus genes from different kinetic classes by RT-qPCR revealed that viral lytic gene expression was broadly repressed by TDP-43 knockdown (Fig. 4F). This included significant reductions in the abundance of mRNAs encoding the critical immediate-early (IE) regulatory proteins ICP0 and ICP4. Indeed, TDP-43 depletion effectively reduced representative HSV-1 IE mRNAs (ICP0, ICP4, ICP27) abundance in infected cells treated with cycloheximide, which inhibits de novo protein synthesis and only allows virus immediate-early gene transcription (Fig. 4G). To determine if reduced viral gene expression resulted from a defect in virus entry, we determined the number of ICP4-positive (ICP4+) cells in shTDP43-#1 and shNTC transduced cells by indirect immunofluorescence. Consistent with reduced viral gene expression, ICP4+ nuclei were significantly reduced upon TDP43 knockdown at 8 hpi (Fig. 4H andI). However, when we let the infection progress but inhibited virus DNA replication using phosphonoacetic acid (PAA), similar numbers of ICP4+ nuclei were observed (Fig. 4H andJ). Thus, while the kinetics of HSV-1 infection are delayed upon TDP-43 knockdown in differentiated HD10.6 cells, the total number of infected cells is not changed excluding reduced virus entry as an explanation. Together, this suggests that TDP-43 might control HSV-1 infection through changes to both the virus and the host.
## TDP-43 depletion results in a genome-wide reduction of viral gene expres sion
In uninfected cells, several studies have shown a myriad of effects on TDP-43 depletion on RNA biology, including direct changes to the host transcriptome through differential splicing and polyadenylation. To more globally assess whether specific HSV-1 RNAs are changed upon TDP-43 depletion, we employed an unbiased transcriptomelevel sequencing approach. Nanopore direct RNA-sequencing (dRNA-seq) is a powerful technique that allows for sequencing of full length RNAs, capturing transcription start (TSS) and termination sites (TTS) as well as splice junctions (33,34). This is particularly useful when applied to highly dense viral genomes harboring multiple polycistronic transcription units that can be difficult to deconvolute using short-read sequencing approaches (34)(35)(36). In accordance with restricted viral replication, shRNA-mediated depletion of TDP-43 in differentiated HD10.6 cells resulted in the detection of propor tionally fewer (fourfold) viral reads by dRNA-seq (Fig. 5A andB). Next, we quantified the relative abundance of each viral transcript in both conditions (transcripts per million, TPM) and calculated the ratio of knockdown over control. The TPM ratio was below 1 for all viral genes, with no significant difference between genes considered essential or non-essential for viral replication (Fig. 5C). Furthermore, a significant difference in depletion between viral transcripts of different kinetic classes was not observed (Fig. 5D).
Together, this suggests that TDP-43 depletion leads to a genome-wide restriction of viral lytic gene expression.
## Evidence for enhanced intron retention in TDP-43 knockdown cells
To determine whether TDP-43 influences the architecture of the HSV-1 transcriptome through alternative TSS or TTS usage, we extracted the start and end coordinates for all viral reads and filtered out all with an abundance of less than 10 reads, to yield frequency estimates for the 5′ and 3′ ends of polyadenylated viral transcripts. In doing so, we did not observe any unique or differentially used putative TSS or TTS (pTSS or pTTS) for HSV-1 transcripts in TDP-43 depleted differentiated HD10.6 cells, suggesting that TDP-43 does not detectably influence transcription initiation or termination during HSV-1 infection. Additionally, in human neurons, TDP-43 has been shown to control splicing of specific host genes through intron retention or cryptic exon inclusion, thereby altering transcript stability (3,9,10,13,14). Whereas the majority of mammalian genes are spliced, only five HSV-1 transcripts consistently undergo splicing, namely: ICP0, UL15, Us1, Us12, and LAT. Interestingly, when we plotted the read depth across UL15 for both control and TDP-43 depleted differentiated HD10.6 cells, we observed a marked reduction of coverage in TDP-43 depleted cells in the exons, but similar abundance in the intronic region (Fig. 5E). To visualize this better, we calculated the read-depth ratio and plotted the moving average over 10 bases. Above 1, this ratio indicates a higher coverage in the TDP-43 depleted sample than the control at that genomic location, whereas below 1 the coverage is higher for the control sample (Fig. 5F). Interestingly, the coverage ratio ranged from 0.6 to 0.7 in both UL15 exons but was increased sharply across the intron ranging from 0.8 to 1.1 (Fig. 5G). This suggests that upon TDP-43 depletion, UL15 RNAs are less efficiently spliced. We found similar evidence of intron retention in the first ICP0 intron (Fig. 5H), but not for the Us1/Us12 intron implying a level of selectivity. Unfortu nately, as the primary LAT transcript is very unstable and the intron is not polyadenyla ted, we were unable to retrieve reads corresponding to LAT and could, therefore, not assess LAT splicing efficiency. To determine if the unevenness in read-depth through the gene body was a unique signature of spliced genes, we also examined several viral unspliced control genes. Three monocistronic transcripts, UL21, UL30, and UL54 (Fig. 6A through C), did not exhibit similar variation in read depth ratio throughout the gene body, and two polycistronic transcription units Us3/4 and UL6/7(Fig. 6D andE) had a small change of read depth ratio at internal transcription start sites. These data show that large changes of read depth ratio are, indeed, unique to spliced genes. Finally, we assessed splicing of the highly abundant cellular gene ACTG1 (Fig. 6F) and observed efficient splicing in all but the first intron, consistent with earlier reports addressing the effect of TDP-43 on cellular splicing. Together, these data demonstrate that in addition to a global control of viral gene expression, TDP-43 also promotes more efficient splicing of two viral genes, one of which encodes the multifunctional IE regulatory protein ICP0.
## Intron retention in UL15 is dependent on the relative abundance of TDP-43
To confirm enhanced intron retention in ICP0 and UL15 mRNAs, we designed primer sets that can distinguish between spliced and intronic variants of these mRNAs (Fig. 7A). We observed a slight increase (14%) in the relative abundance of the first ICP0 intron upon TDP-43 knockdown (Fig. 7B). However, under CHX treatment, which inhibits protein synthesis and selectively allows only IE gene transcription, the abundance of ICP0 intron-containing transcripts increased by 2.5-fold in differentiated HD10.6 cells depleted for TDP-43 compared to control (Fig. 7C). This could reflect that many intronretaining transcripts are degraded by nonsense-mediated decay (NMD) in the cytoplasm, which is a translation-dependent mechanism. Alternatively, this could indicate a viral factor affecting the abundance of intron containing transcripts. ICP27 is a viral protein known to influence mRNA splicing (37)(38)(39)(40), but the direction of change-greater intron retention rather than less-suggests that reduced ICP27 expression is likely not causing the observed differences. To determine if enhanced intron retention in ICP0 mRNAs affected ICP0 protein levels, we measured ICP0 protein abundance by immunoblotting (Fig. 7D). Indeed, in differentiated HD10.6 cells transduced with a TDP43-targeting shRNA, TDP-43 and ICP0 levels were significantly reduced compared to a non-targeting control (Fig. 7E andF). Similarly, we observed a slight increase in intronic UL15 RNAs upon TDP-43 depletion (Fig. 7G) that significantly correlated to the extent of TDP-43 knockdown in the same sample (Fig. 7H). Together these data show that UL15 splicing efficiency in differentiated HD10.6 cells is influenced by TDP-43 abundance.
## DISCUSSION
A hallmark of several prominent neurodegenerative disorders is proteotoxic accumula tion of key neuronal proteins, such as amyloid-β, tau, α-synuclein, TDP-43 (41)(42)(43). Recent studies implicate herpesvirus reactivation in the onset of these neurodegenerative disorders (44-47) though little is known about virus-host interactions in neuronal cell types. Here, we have shown differential requirement for TDP-43 expression during lytic HSV-1 infection in the human neuronal-derived HD10.6 cell line compared to epithelial cells and primary fibroblasts. We show that the depletion of TDP-43 restricts viral replication in HD10.6-derived neurons during both single-cycle and multi-cycle HSV-1 infection, through generalized transcriptional repression and enhanced intron retention in key virus-encoded mRNAs. As such, this study demonstrates the importance of using neuronal derived cell models to uncover cell type-specific regulatory mechanisms. TDP-43 is an RNA-binding protein that is expressed by all cell types, but of these, neurons are considered uniquely sensitive to TDP-43 depletion, mutations, or mislocal ization. We showed that TDP-43 depletion specifically impairs HSV-1 replication in a model of human DRG-derived neurons (Fig. 3 and4), but not in epithelial cells or fibroblasts (Fig. 1). Limited research into neuron-specific responses and requirements to HSV-1 infection suggests that neurons have a reduced antiviral response and are uniquely sensitive to the depletion of specific proteins (48)(49)(50)(51). The requirement of a neurotropic virus, exemplified here by HSV-1, for a host TDP-43 protein involved in neuronal homeostasis, highlights how viruses repurpose host processes for their own benefit. Despite the fact that TDP-43 was first identified as restricting HIV-1 infection, few studies have addressed its role in promoting or antagonizing replication of other viruses. In this study, we show that HSV-1 mRNA accumulation is reduced upon TDP-43 depletion in HD10.6 cells, even in the absence of viral protein synthesis. This raises the question whether TDP-43 could directly regulate viral gene expression. In HIV-1 infected cells, TDP-43 binds to the TAR DNA sequence and limits its transcription (4) and stabilizes HDAC6 expression-a key HIV restriction factor -further inhibiting viral replication (52,53). More recently, TDP-43 was shown to be recruited by the viral polymerase of influenza A virus, bind and stabilize viral mRNAs (28). Notably, TDP-43 depletion modestly impairs influenza viral growth similar to our observations for HSV-1 in HD10.6 cells. It remains to be explored if HSV-1 mRNAs in HD10.6 cells are stabilized through direct TDP-43 binding, or if TDP-43 controls HSV-1 transcription through interaction with RNA Pol II or viral DNA. Notably, about half of the viral mRNAs contain at least one TDP-43 minimal consensus sequence (54), suggesting that TDP-43 interactions with viral mRNAs are possible. TDP-43 reportedly autoregulates its abundance through a negative feedback loop by binding to the 3′UTR of TDP-43 mRNAs (55,56) and similarly regulates the stability of other mammalian mRNAs through binding to the 3′UTR (11,57,58).
A well-characterized function of TDP-43 in neurons is ensuring correct splicing, through repression of cryptic exons and retained introns in neuronal genes (3, 9-11, 19, 59). Using dRNA-seq to interrogate complete transcript structures in HSV-1 infected cells, we showed that TDP-43 depletion leads to a modest enhancement of intron retention in two viral genes, RL2 (ICP0) and UL15 (TRM3), that are important for efficient viral replication (60). Notably, intron containing viral mRNAs-specifically those of ICP0 and UL15-are thought to accumulate in the nucleus as they are less efficiently exported to the cytoplasm by the viral protein ICP27 (38,61,62). Enhanced intron retention resulting from TDP-43 depletion could, thus, alter the availability of spliced mRNAs for translation and impair efficient replication. Specifically, an early deficit in ICP0 abundance could, in part, contribute to the global reduction in gene expression upon TDP-43 depletion, as antiviral defenses might not be effectively countered. The increased abundance of intron-containing ICP0 transcripts in the absence of viral protein synthesis (Fig. 7C) suggests the mechanism by which TDP-43 regulates viral mRNA splicing is distinct from the ICP27-dependent mechanisms of other host splicing factors (37,40,63,64). Interestingly, a manual search revealed that UL15 pre-mRNA contains five minimal TDP-43 binding sites (54) throughout the intron (Fig. 7I), raising the possibility that UL15 splicing might be regulated similarly to host genes where TDP-43 represses introns through direct binding (11).
In summary, we showed that the critical neuronal disease gene TDP-43 specifically regulates replication of a common, widespread neurotropic HSV-1 in neuron-derived HD10.6 cells but not epithelial cells or primary fibroblasts. This work highlights the importance of investigating virus-host interactions in multiple relevant model systems, as proteins can have cell-type specific functions. Our finding that a protein known for forming proteotoxic aggregates in neurodegenerative disease supports replication of a virus that is increasingly linked to Alzheimer's disease highlights an interesting intersection between viral infection and neuronal health. Future studies should leverage neuronal culture systems to determine whether HSV-1 induced infection stress or virus functions influence neurodegeneration resulting from TDP-43 depletion or aggregation.
## MATERIALS AND METHODS
## Cell culture
Normal human dermal fibroblasts (NHDFs) were cultured in DMEM supplemented with 5% heat-inactivated FBS (HI-FBS, Gibco) and 1% penicillin-streptomycin (PS, Lonza).
Human retinal pigmented epithelial ARPE-19 cells were cultured in DMEM:F12 (Gibco) supplemented with 10% HI-FBS and 1% PS, 15 mM HEPES, and 1 mM sodium pyru vate (Gibco). 293T Lenti-X cells (Takara Bio) were maintained in DMEM supplemented with 10% HI-FBS, 1% PS. Vero cells were maintained in DMEM supplemented with 5% calf-serum and 1% PS.
## Culture and differentiation of HD10.6 cells
Proliferating HD10.6 cells (a gift of Anna Cliffe, University of Virginia School of Medicine) were maintained on 17 µg/mL fibronectin in PBS pre-coated T-75 flasks in Proliferation Media: Advanced DMEM/F12 (Gibco), supplemented with 1× Neurocult SM1 (STEM CELL Technologies), 10 ng/mL prostaglandin E1 (Sigma), 2 mM L-glutamine (Gibco), 1× Primocin (Invivogen), and fresh 0.5 ng/mL bFGF (PeproTech). For differentiation, HD10.6 were plated at 25.000 cells/cm 2 in Proliferation Media on plates precoated with 50 µg/mL poly-L-ornithine hydrobromide (Sigma) in 0.5 M borate buffer pH 8.5 (Boston Bioproducts), and 1 µg/mL fibronectin. One day after plating, media was replaced for Differentiation +/+/+ Media: Neurobasal (Gibco), supplemented with 1× Neurocult SM1, 2 mM L-glutamine, 1× Primocin, 1 µg/mL doxycycline, 50 ng/mL NGF (Alomone Labs), 25 ng/mL CNTF, GDNF, and NT-3 (all Peprotech).
## HSV-1 stocks and infections
Working stocks of wild type HSV-1 strain KOS and HSV-1 GFP-Us11 strain Patton were propagated on Vero or ARPE-19 cells respectively. Virus stocks, supernatants (NHDF, ARPE-19), or cell-associated virus (HD10.6) were titered using plaque assays on Vero cells. For experiments, NHDF, ARPE-19 and Vero cells were infected for 1 h in low serum (1%) media or for HD10.6 cells for 2 h in Differentiation Base Media (Neurobasal with L-glutamine).
## Transfection of siRNAs
NHDFs and ARPE-19s were seed at approximately 30% confluency and transfected with 20 nM siRNA using Lipofectamine RNAiMax (Life technologies) in Optimem (Gibco) 3 days prior to infection. siRNAs used in this study were purchased from Sigma: siTDP-43-#1 (SASI-Hs01-0037054) and siTDP-43-#2 (SASI-Hs01-0037055). AllStars Negative Control siRNA (Qiagen) was used as the non-targeting siRNA control.
## Lentiviral generation and transduction
Lentiviruses were generated by transfecting equimolar ratios of monomeric pMD2.G (VSV-G, derived from Addgene #12259), psPAX2 (Addgene #12260), and transfer plasmids into subconfluent 293T Lenti-X cells (Takara Bio) using a 3:1 ratio of linear polyethylenimine MW 25000 (Polysciences). Media was replaced 1 day later for Optimem (Gibco), and supernatant was collected 3 days post-transfection and passed through 0.45 µm PVDF filter before aliquoting and storage at -80°C. Lentiviral transfer plasmids used in this study were pLKO.1-puro Non-Target shRNA control plasmid as negative control, TRCN0000016040 (shTDP43-#1) and TRCN0000016041 (shTDP43-#2), all from Sigma. To transduce HD10.6 cells, approximately 30 µL lentiviral stock per well was used (24-wells format) and incubated in half the original culturing volume of Differentiation +/+/+ Media overnight without polybrene. The next day, lentivirus containing media was removed and replaced with fresh Differentiation +/+/+ Media. Cells were used for assays 3-4 days post-transduction.
## Immunofluorescence
HD10.6 cells were plated for differentiation on glass coverslips and infected after 10 days. Cells were then fixed with 4% paraformaldehyde, permeabilized for 10 min with 0.1% Triton X-100 in PBS, and blocked with 5% normal goat serum diluted in PBS. Cover slips were incubated with rabbit anti-β3-tubulin (1:2,000, 802001, BioLegend) or mouse anti-ICP4 (1:200, ab6514, Abcam) primary antibody in PBS overnight. The next day coverslips were washed and incubated for 1 h at room temperature with AF555-conjuga ted goat anti-rabbit or goat anti-mouse secondary antibody in PBS (1:500, Thermo Fisher Scientific). Finally, the coverslips were washed and incubated with Hoechst 33342 in PBS (20 µM, Life Technologies) for 5 min, washed, and mounted using Fluorshield mounting media (Sigma). Stained coverslips were analyzed using a Leica DM5000B epifluorescence microscope with 63× objective lens with oil or using a Keyence BZ-X800 high-content imager with 20× objective. Images were adjusted for brightness and contrast using Photoshop CC 2025 software (Adobe).
## Immunoblots
Cells were lysed using 1× Laemmli buffer, and samples were incubated at 95°C for 10 min. Proteins were separated by SDS-polyacrylamide gel electrophoresis (10% gel) and transferred onto nitrocellulose membranes. Membranes were blocked using 5% non-fat dry milk in TBS-T for 1 h and incubated with primary antibodies diluted in TBS-T + 3% BSA + 0.02% NaN 3 overnight at 4°C with agitation. Primary antibodies used were rabbit anti-TDP-43 (10782-2-AP, Proteintech), rabbit anti-GAPDH (2118, Cell Signaling Technologies), rabbit anti-β3-tubulin (802001, BioLegend), mouse anti-ICP8 (ab20193, Abcam), mouse anti-ICP27 (ab31631, Abcam), and mouse anti-ICP0 (ab6513, Abcam). The next day, membranes were washed thrice with TBS-T and incubated with 1:5,000 of HRP-conjugated secondary antibody (Cytiva) for 1 h at RT in TBS-T + 5% milk. Membranes were then washed thrice with TBS-T, and chemiluminescent signal was visualized using Pierce ECL Western Blotting kits at Regular and Pico strengths (Thermo Fisher Scientific) on an iBright CL1000 imaging system.
## RNA isolation and cDNA synthesis
Cells were harvested in Trizol (Thermo Fisher Scientific), mixed with 1:5 chloroform, and centrifuged for 15 min at 12,000 × g at 4°C. From the aqueous phase, RNA was isolated using the RNeasy Mini kit (Qiagen) according to manufacturer's instruc tions. Residual DNA was removed using the Turbo DNA-free kit (Ambion). To generate cDNA, Turbo-DNAse-treated RNA was reversed transcribed with Superscript IV reverse transcriptase (RT+) and random hexamers (ThermoFisher Scientific). The same cDNA synthesis reaction in the absence of reverse transcriptase (RT-) was performed to generate negative controls.
## Quantitative PCR analysis
Quantitative PCR (qPCR) was performed in duplicate on RT+ cDNA using 0.5 µm of each primer (IDT), 1× SsoAdvanced Universal SYBR Green Supermix (1725275, Biorad) on a Bio-Rad IQ5 thermal cycler, and RT-cDNA was run for select targets to exclude residual DNA carryover. An annealing temperature of 58°C was used for all targets. Melt curves were inspected for the presence of secondary peaks prior to data analysis. Primer sequences used in this study are given in Table 1. Relative expression was defined as 2 -(Cq-value target gene -Cq-value housekeeping) and fold change as by normalizing relative expression to the average of control samples. Absolute HSV-1 genome copies were determined by using a 10-fold serial dilution of the UL36 amplicon (gBlock, IDT) in 20 ng/µL salmon sperm DNA (Invitrogen).
## Cell viability
To determine cell viability after shRNA-transduction, the Cell-Titer Glo assay (Promega) was used according to manufacturer's instructions. Briefly, HD10.6 cells were differentiated in black-wall 96-well plates for 10 days and then transduced with lentiviral stocks as described above. Media was refreshed 1 day after transduction, and cells were incubated for 3 additional days. Culture media was removed and replaced with 50 µL assay reagent, mixed for 2 min, incubated for 10 min at room temperature, and measured on the PerkinElmer EnVision 2103 Multilabel Reader.
## Nanopore dRNA-sequencing
To generate dRNA-sequencing libraries, poly(A) RNA was isolated using a Dynabeads mRNA purification kit (Invitrogen). Five hundred nanograms of poly(A) served as input for library preparation using the direct RNA Sequencing (DRS) SQK-RNA004 kit. This resulted in 150 ng of library that was loaded onto FLO-MIN004RA flow cells. Sequencing was performed on a MinION Mk1B for 24 h, yielding 1-2 million reads per sample.
## Nanopore bioinformatic analysis
Basecalling was performed with dorado v.0.8.1 (Oxford Nanopore Technologies), using a high accuracy model (rna004_130bps_hac@v5.0.0) with adapter trimming. FASTQ files were extracted from the unaligned BAM files using BEDtools v.2.30.0 (65). Reads were aligned to the human (hg38) or HSV-1 GFP-Us11 (MF959544.1) genomes using minimap2 v.2.24 (66) with spliced alignment settings (-ax splice) and a k-mer size of 14 (-k14). The resulting SAM files were filtered to exclude unmapped, secondary, and supplementary alignments (-F2308) using SAMtools v.1.16 (67) and then converted to BAM format, sorted, and indexed. For transcriptome alignment, FASTQ files were aligned to the HSV-1 GFP-Us11 and human transcriptomes (Gencode v47) using minimap2, with the setting -ax map -ont for Oxford Nanopore data and a minimum secondary-to-primary score ratio of 0.99 (-p 0.99). The resulting alignments were stored in SAM format. SAM files were filtered to retain primary alignments by removing supplementary, secondary, and unmapped reads (-F2324). Subsequently, reads with hard clipping in the CIGAR string were removed. Only reads that start within 50 bp downstream of annotated start sites were retained. The filtered reads were converted into BAM files, sorted, and indexed with samtools. Transcript abundance was quantified by extracting reference names from the sorted BAM file using samtools view, isolating the third column (reference name) with cut -f3, sorting entries alphabetically with sort, counting unique occurrences of each identifier using uniq -c, and exporting the results to a text file for downstream analysis.
To normalize differences in read depth, each transcript count was divided by the total number of mapped and filtered reads and multiplied by 1 × 10 6 , denoted by transcripts per million (TPM). TPM values from TDP43-KD samples were divided by control samples to calculate a TPM KD/control ratio. Only locations with a minimum raw read depth of 10 were considered for downstream analysis.
pTSS and pTTS were determined through extracting the 5′ and 3′ end of each read using Rsamtools and locations with less than 10 counts were filtered out. pTSS and pTTS between samples were compared by plotting the abundance per location in each area of the genome. Coverage plots were generated using Genomic Ranges in RStudio, and gene models were generated from updated GFF3 files reflecting the HSV-1 GFP-Us11 transcriptome. To calculate the coverage ratio, the read depth for each base in shTDP-43 data set was divided by the read depth in the shNTC data set, and locations with a read depth below 10 in either data set were excluded. To calculate the moving average and smooth out local variations in alignment, the stats::filter function in R was used, with a window size of 10 bases.
## Statistical analysis
Figures show individual data points or mean ± SEM, all statistical analyses were performed with Graphpad Prism 10 software using the statistical test indicated in the figure legends, with * = P < 0.05, ** = P < 0.01, *** = P < 0.001. If no statistics are given, the difference was not significant.
## References
1. "Denise Ohnezeit"
3. Wilson
4. Mohr
5. Grant "Author(s) Nederlandse Organisatie voor Weten schappelijk Onderzoek 452022210"
6. Shirley "Braspenning Deutsche Forschungsgemeinschaft 554758329 Denise Ohnezeit National Institutes of Health R01-AI176335, R01-AI170583 Angus C" *Wilson Ian J. Mohr REFERENCES*
7. Cohen, Lee, Trojanowski (2011) "TDP-43 functions and pathogenic mechanisms implicated in TDP-43 proteinopathies" *Trends Mol Med*
8. Buratti, Baralle (2012) "TDP-43: gumming up neurons through protein-protein and protein-RNA interactions" *Trends Biochem Sci*
9. Tollervey, Curk, Rogelj et al. (2011) "Characterizing the RNA targets and position-dependent splicing regulation by TDP-43" *Nat Neurosci*
10. Ou, Wu, Harrich et al. (1995) "Cloning and characterization of a novel cellular protein, TDP-43, that binds to human immunodeficiency virus type 1 TAR DNA sequence motifs" *J Virol*
11. Morera, Ahmed, Schwartz (2019) "TDP-43 regulates transcrip tion at protein-coding genes and Alu retrotransposons" *Biochim Biophys Acta Gene Regul Mech*
12. Liu, Russ, Cali et al. (2019) "Loss of nuclear TDP-43 is associated with decondensation of LINE retrotranspo sons" *Cell Rep*
13. Buratti, Conti, Stuani et al. (2010) "Nuclear factor TDP-43 can affect selected microRNA levels" *FEBS J*
14. Kawahara, Mieda-Sato (2012) "TDP-43 promotes microRNA biogenesis as a component of the Drosha and Dicer complexes" *Proc Natl Acad Sci*
15. Buratti, Dörk, Zuccato et al. (2001) "Nuclear factor TDP-43 and SR proteins promote in vitro and in vivo CFTR exon 9 skipping" *EMBO J*
16. Ma, Prudencio, Koike et al. (2022) "TDP-43 represses cryptic exon inclusion in the FTD-ALS gene UNC13A" *Nature*
17. Polymenidou, Lagier-Tourenne, Hutt et al. (2011) "Long pre-mRNA depletion and RNA missplicing contribute to neuronal vulnerability from loss of TDP-43" *Nat Neurosci*
18. Ling, Pletnikova, Troncoso et al. (2015) "TDP-43 repression of nonconserved cryptic exons is compromised in ALS-FTD" *Science*
19. Klim, Williams, Limone et al. (2019) "ALSimplicated protein TDP-43 sustains levels of STMN2, a mediator of motor neuron growth and repair" *Nat Neurosci*
20. Brown, Wilkins, Keuss et al. (2022) "TDP-43 loss and ALSrisk SNPs drive mis-splicing and depletion of UNC13A" *Nature*
21. Ayala, Zago, 'ambrogio et al. (2008) "Structural determinants of the cellular localization and shuttling of TDP-43" *J Cell Sci*
22. Liu-Yesucevitz, Bilgutay, Zhang et al. (2010) "Tar DNA binding protein-43 (TDP-43) associates with stress granules: analysis of cultured cells and pathological brain tissue" *PLoS One*
23. Mcdonald, Aulas, Destroismaisons et al. (2011) "TAR DNA-binding protein 43 (TDP-43) regulates stress granule dynamics via differential regulation of G3BP and TIA-1" *Hum Mol Genet*
24. Neelagandan, Gonnella, Dang et al. (2019) "TDP-43 enhances translation of specific mRNAs linked to neurodegenerative disease" *Nucleic Acids Res*
25. Fiesel, Weber, Supper et al. (2012) "TDP-43 regulates global translational yield by splicing of exon junction complex component SKAR" *Nucleic Acids Res*
26. Russo, Scardigli, Regina et al. (2017) "Increased cytoplasmic TDP-43 reduces global protein synthesis by interacting with RACK1 on polyribosomes" *Hum Mol Genet*
27. (2025) *Full-Length Text Journal of Virology*
28. Neumann, Sampathu, Kwong et al. (2006) "Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis" *Science*
29. Fung, Shi, Deng et al. (2015) "Cytoplasmic translocation, aggregation, and cleavage of TDP-43 by enteroviral proteases modulate viral pathogenesis" *Cell Death Differ*
30. Zhang, Yang, Li et al. (2023) "Enterovirus D68 infection induces TDP-43 cleavage, aggregation, and neurotoxicity" *J Virol*
31. Xue, Ruller, Fung et al. (2018) "Enteroviral infection leads to transactive response DNAbinding protein 43 pathology in vivo" *Am J Pathol*
32. Masaki, Sonobe, Ghadge et al. (2019) "TDP-43 proteinopathy in Theiler's murine encephalomyelitis virus infection" *PLoS Pathog*
33. Cabrera, Rodríguez-Izquierdo, Jiménez et al. (2020) "Analysis of ALS-related proteins during herpes simplex virus-2 latent infection" *J Neuroinflammation*
34. Harrell, Davido, Bertke (2023) "Herpes Simplex Virus 1 (HSV-1) Infected Cell Protein 0 (ICP0) targets of ubiquitination during productive infection of primary adult sensory neurons" *Int J Mol Sci*
35. Dupont, Krischuns, Gianetto et al. (2024) "The RBPome of influenza A virus NP-mRNA reveals a role for TDP-43 in viral replication" *Nucleic Acids Res*
36. Cohen (2020) "Herpesvirus latency" *J Clin Invest*
37. Raymon, Thode, Zhou et al. (1999) "Immortalized human dorsal root ganglion cells differentiate into neurons with nociceptive properties" *J Neurosci*
38. Thellman, Botting, Madaj et al. (2017) "An immortalized human dorsal root ganglion cell line provides a novel context to study herpes simplex virus 1 latency and reactivation" *J Virol*
39. Benboudjema, Mulvey, Gao et al. (2003) "Association of the herpes simplex virus type 1 Us11 gene product with the cellular kinesin light-chain-related protein PAT1 results in the redistribution of both polypeptides" *J Virol*
40. Garalde, Snell, Jachimowicz et al. (2018) "Highly parallel direct RNA sequencing on an array of nanopores" *Nat Methods*
41. Depledge, Srinivas, Sadaoka et al. (2019) "Direct RNA sequencing on nanopore arrays redefines the transcriptional complexity of a viral pathogen" *Nat Commun*
42. Braspenning, Sadaoka, Breuer et al. (2020) "Decoding the architecture of the Varicella-Zoster virus transcriptome" *mBio*
43. Braspenning, Verjans, Mehraban et al. (2021) "The architecture of the simian varicella virus transcriptome" *PLoS Pathog*
44. Sciabica, Dai, Sandri-Goldin (2003) "ICP27 interacts with SRPK1 to mediate HSV splicing inhibition by altering SR protein phosphoryla tion" *EMBO J*
45. Hardy, Sandri-Goldin (1994) "Herpes simplex virus inhibits host cell splicing, and regulatory protein ICP27 is required for this effect" *J Virol*
46. Bryant, Wadd, Lamond et al. (2001) "Herpes simplex virus IE63 (ICP27) protein interacts with spliceosomeassociated protein 145 and inhibits splicing prior to the first catalytic step" *J Virol*
47. Tang, Patel, Krause (2019) "Hidden regulation of herpes simplex virus 1 pre-mRNA splicing and polyadenylation by virally encoded immediate early gene ICP27" *PLoS Pathog*
48. Ross, Poirier (2004) "Protein aggregation and neurodegenerative disease" *Nat Med*
49. Soto (2003) "Unfolding the role of protein misfolding in neurodegener ative diseases" *Nat Rev Neurosci*
50. Goedert (2015) "Alzheimer's and Parkinson's diseases: the prion concept in relation to assembled Aβ, tau, and α-synuclein" *Science*
51. De Chiara, Piacentini, Fabiani et al. (2019) "Recurrent herpes simplex virus-1 infection induces hallmarks of neurodegeneration and cognitive deficits in mice" *PLoS Pathog*
52. Dutton, Turnbaugh, Patel et al. (2025) "Asymptomatic neonatal herpes simplex virus infection in mice leads to persistent CNS infection and long-term cognitive impairment" *PLoS Pathog*
53. Taquet, Dercon, Todd et al. (2024) "The recombinant shingles vaccine is associated with lower risk of dementia" *Nat Med*
54. Eyting, Xie, Michalik et al. (2025) "A natural experiment on the effect of herpes zoster vaccination on dementia" *Nature*
55. Yordy, Iijima, Huttner et al. (2012) "A neuron-specific role for autophagy in antiviral defense against herpes simplex virus" *Cell Host Microbe*
56. Chan, Liu, Bastard et al. (2024) "Human TMEFF1 is a restriction factor for herpes simplex virus in the brain" *Nature*
57. Dai, Idorn, Serrero et al. (2024) "TMEFF1 is a neuron-specific restriction factor for herpes simplex virus" *Nature*
58. Zimmer, Ewaleifoh, Harschnitz et al. (2018) "Human iPSC-derived trigeminal neurons lack constitutive TLR3-dependent immunity that protects cortical neurons from HSV-1 infection" *Proc Natl Acad Sci*
59. Fiesel, Voigt, Weber et al. (2010) "Knockdown of transactive response DNAbinding protein (TDP-43) downregulates histone deacetylase 6" *EMBO J*
60. Valenzuela-Fernández, Alvarez, Gordon-Alonso et al. (2005) "Histone deacetylase 6 regulates human immunodeficiency virus type 1 infection" *Mol Biol Cell*
61. Lukavsky, Daujotyte, Tollervey et al. (2013) "Molecular basis of UG-rich RNA recognition by the human splicing factor TDP-43" *Nat Struct Mol Biol*
62. Ayala, Conti, Vázquez et al. (2011) "TDP-43 regulates its mRNA levels through a negative feedback loop" *EMBO J*
63. Avendaño-Vázquez, Dhir, Bembich et al. (2012) "Autoregulation of TDP-43 mRNA levels involves interplay between transcription, splicing, and alternative polyA site selection" *Genes Dev*
64. Colombrita, Onesto, Megiorni et al. (2012) "TDP-43 and FUS RNA-binding proteins bind distinct sets Full-Length Text Journal of Virology December"
65. "of cytoplasmic messenger RNAs and differently regulate their posttranscriptional fate in motoneuron-like cells" *J Biol Chem*
66. Gu, Wu, Xu et al. (2017) "TDP-43 suppresses tau expression via promoting its mRNA instability" *Nucleic Acids Res*
67. Koike, Pickles, Ayuso et al. (2023) "TDP-43 and other hnRNPs regulate cryptic exon inclusion of a key ALS/FTD risk gene, UNC13A" *PLoS Biol*
68. Pellet, Roizman (2013) "Herpesviridae"
69. Sandri-Goldin (1998) "ICP27 mediates HSV RNA export by shuttling through a leucine-rich nuclear export signal and binding viral intronless RNAs through an RGG motif" *Genes Dev*
70. Phelan, Dunlop, Clements (1996) "Herpes simplex virus type 1 protein IE63 affects the nuclear export of virus intron-containing transcripts" *J Virol*
71. Ellison, Rice, Verity et al. (2000) "Processing of alpha-globin and ICP0 mRNA in cells infected with herpes simplex virus type 1 ICP27 mutants" *J Virol*
72. Wang, Liu, Lu et al. (2016) "Serine/arginine-rich splicing factor 2 modulates herpes simplex virus type 1 replication via regulating viral gene transcriptional activity and pre-mRNA splicing" *J Biol Chem*
73. Quinlan, Hall (2010) "BEDTools: a flexible suite of utilities for comparing genomic features" *Bioinformatics*
74. Li (2018) "Minimap2: pairwise alignment for nucleotide sequences" *Bioinformatics*
75. Li, Handsaker, Wysoker et al. "Genome Project Data Processing S. 2009. The sequence alignment/map format and SAMtools"
76. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645918&blobtype=pdf | # Differential expression of viral entry protein neuropilin 1 (NRP1) and neuropilin 2 (NRP2) in fatal COVID-19
A Dette, F Moers, T Mayr, S Stillfried, M Bernhardt, S Förster, C Werlein, M Ackermann, M Muders, G Kristiansen, P Boor, I Gütgemann
## Abstract
Coronavirus disease 2019 (COVID-19) is associated with hyperinflammation, endothelialitis, hypoxemia, and hypercoagulation, contributing to thrombosis in acute severe and long COVID. While ACE2 is the primary severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) receptor, its low expression in certain infected cell types suggests alternative co-receptors. Neuropilins (NRP1 and NRP2), widely expressed, have been proposed as co-factors for viral entry. We analyzed NRP1 and NRP2 expression in autopsy tissues from heart, lung, and hematolymphoid organs using immunohistochem istry (n = 20) and compared findings with public single-cell RNA sequencing (scRNAseq) data. Selected cases were further examined by spatial multiplex immunofluorescence (CODEX). In vitro binding of NRP1/NRP2 to SARS-CoV-2 spike fragments S1 and S1′ was assessed by immunofluorescence microscopy. NRP1 was abundantly expressed in myocardial capillary endothelial cells (ECs) and macrophages in the heart and lung; NRP2 was found in alveolar macrophages and mast cells. scRNAseq re-analysis confirmed these in situ patterns. In vitro, NRP1 exclusively bound S1, while NRP2 bound both S1 and S1′. SARS-CoV-2 RNA was detected in neuropilin-positive, ACE2/TMPRSS2-negative vascular EC and mast cells. The detection of SARS-CoV-2 RNA in neuropilin-positive but ACE2/TMPRSS2-negative cell clusters supports that neuropilins are involved in systemic viral dissemination. NRP1 on vascular EC may contribute to angiogenesis, vascular damage, and microangiopathy, while NRP2 represents a potential immunomodulatory target to regulate macrophage activity, resolve inflammation, and potentially prevent the progression of pulmonary fibrosis and limit excessive mast cell activation in long COVID.
IMPORTANCEThe well-known severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) receptor, angiotensin-converting enzyme 2 (ACE2), exhibits low expres sion in key cell types implicated in coronavirus disease 2019 (COVID-19) pathology, such as endothelial cells and B cells, macrophages, and mast cells. In contrast, neuro pilins, identified as co-receptors for SARS-CoV-2, are abundantly expressed in these cells under physiological conditions and may be involved in virus-host interactions. This study presents a detailed in situ analysis of Neuropilin 1 (NRP1) and Neuropilin 2 (NRP2) expression in fatal COVID-19 cases using immunohistochemistry and spa tial multiplex immunofluorescence phenotyping, complemented by single cell RNA sequencing. Additionally, it demonstrates differential binding affinities of NRP1 and NRP2 to SARS-CoV-2 spike protein fragments S1 and S1′ in vitro, suggesting distinct roles for these neuropilins in viral recognition. This study highlights the impact of the unique furin cleavage site in SARS-CoV-2, which may contribute to increased pathogenicity through its interaction with NRP1.
S evere acute respiratory syndrome coronavirus 2 (SARS-CoV-2) caused the coronavi rus disease 2019 (COVID-19) pandemic and emerging variants remain a persistent health threat. SARS-CoV-2 infection triggers a cascade of pathophysiological mecha nisms, including endothelial dysfunction, hyperinflammation, and platelet activation, which collectively contribute to immunothrombosis (1). Although SARS-CoV-1 and SARS-CoV-2 genome sequences are highly homologous, SARS-CoV-2 exhibits a 10-20fold higher binding affinity to the angiotensin-converting enzyme 2 (ACE2) receptor compared with SARS-CoV-1 due to differences in its spike protein. This enhances cell entry and results in a more vigorous pro-inflammatory cytokine response ("cytokine storm") in monocytes and macrophages, accompanied by a dysregulated myeloid cell compartment (2)(3)(4). Monocytes also play a central role in inducing COVID-19-associated microvascular thrombosis by initiating intrinsic and extrinsic coagulation cascades (5). While fatal COVID-19 cases have declined, post-acute sequelae of COVID-19 (PASC) (6) were estimated to affect at least 10% of infected people-65 million individuals globally, including symptoms related to microvascular damage, e.g. dyspnea, "brain fog," and thrombotic complications (1,7). Survivors of severe COVID-19 often experience variable degrees of health impairment, with pulmonary fibrosis as a significant long-term complication (8).
SARS-CoV-2 typically enters cells via its spike protein binding to the ACE2 receptor, with subsequent processing by transmembrane protease serine subtype 2 (TMPRSS2; Fig. S1) (9). While ACE2 expression has been extensively studied (10,11), the protein is absent in several cell types affected by the virus, including neurons (12) and nasal and respiratory epithelia (13), and is low or absent in capillary endothelial cells (ECs) (13,14). Neuropilin-1 (NRP1) has been identified as a co-factor that facilitates SARS-CoV-2 entry via furin-processed S1 spike protein fragments (13,15). Following cleavage at the S1/2 site ( 16), a C-terminal sequence (-RRAR*) is exposed, conforming to the C-end rule (CendR) motif (-RXXR*) (13,15,17). Binding of the S1 fragment to the b1/b2 domains of NRP1 and NRP2 has been shown to enhance viral infectivity and replication (15,18). While several viruses use neuropilins for cell entry, only Epstein-Barr virus (EBV) and human T-lymphotropic virus 1 (HTLV1) are known to interact with neuropilins via the CendR motif (19).
Neuropilins are transmembrane proteins that lack intrinsic signaling activity and function as co-receptors for signaling molecules like vascular endothelial growth factors (VEGF) and class 3 semaphorins. They are involved in angiogenesis, lymphangiogenesis, immune responses, axon guidance (20,21), as well as cancer progression and therapy resistance (22). Organ-specific SARS-CoV-2 pathogenesis depends on the expression of viral entry factors. NRP1 is predominantly expressed in vascular ECs, vascular smooth muscle cells, mesenchymal stem cells, neurons, epithelial cells of the respiratory and gastrointestinal tracts (23), as well as pancreatic β cells (24). In addition to its neuronal expression (25), NRP2 is expressed in smooth muscle cells of the intestine and bladder (26) and is essential for the development of smaller lymphatic vessels (20). The absence or low levels of ACE2 expression and presence of neuropilins on respiratory and olfactory epithelia and vascular ECs (10,13,14), combined with the upregulation of neuropilin transcripts in SARS-CoV-2 infected lungs (27), highlight the critical role of neuropilins in systemic and acute COVID-19. Despite organ-specific studies on ACE2 and other SARS-CoV-2 co-factors (28), detailed characterization of NRP1 and NRP2 expression in COVID-19-infected tissues remains limited.
This study aims to provide a comprehensive assessment of NRP1 and NRP2 protein expression in the heart, lung, and hematolymphoid organs of COVID-19 autopsies using immunohistochemistry (IHC). Selected cases also underwent in situ multiplex immuno fluorescence staining (CODEX) (29) to visualize rare cell types at a spatial level. The findings were compared with single-cell RNA sequencing (scRNAseq) data from fatal COVID-19 heart and lung samples (30). In addition, SARS-CoV-2 viral RNA was detected and localized in lymph nodes and spleen using RNAscope in situ hybridization. Finally, the role of neuropilins in viral recognition was further explored in vitro by analyzing their differential binding to spike protein fragments using immunofluorescence microscopy.
## RESULTS
## Histopathologic changes in COVID-19 autopsies with a focus on neuropilins
Pulmonary tissue from fatal COVID-19 autopsies revealed alveolar damage in fulmi nant cases, as well as early-phase organizing chronic pneumonia in subacute cases. Histopathological features included prominent intra-alveolar edema, microthrombi, hyaline membranes, intra-alveolar fibrin deposits, and mixed inflammatory infiltrates. The myocardium demonstrated scattered mixed inflammatory cells alongside either normal or hypertrophic myocardial fibers. Similar histomorphological findings have been described previously (27,31,32).
To investigate NRP1 and NRP2 expression patterns in situ, standard IHC was per formed on COVID-19 autopsy tissues, as summarized in Table 1. Neuropilins were expressed on epithelial, mesenchymal stromal, and immune cells in all organs examined (Table 1) with some notable differences: NRP1 was strongly expressed in macrophages and Kupffer cells, splenic white pulp and lymph node follicular lymphocytes, capillary and liver sinusoidal ECs, as well as in proximal tubules and glomeruli (Fig. S5). In contrast, NRP2 expression was more restricted and was moderately detected in intestinal smooth muscle cells and in macrophages within the lung, trachea, spleen, lymph nodes, and adrenal gland (Fig. S6).
## Increased NRP1 expression in capillaries of fatal COVID-19 hearts
Myocardial small capillaries, arterioles, and postcapillary venous vessels showed strong NRP1 expression by IHC in COVID-19 (Fig. 1B), whereas NRP2 was not detected (Fig. S6B). NRP1 localization on vascular EC was confirmed by CD31 co-expression using CODEX multiplex imaging (Fig. 1A). scRNAseq analysis demonstrated abundant NRP1 transcripts in vascular EC and pericytes, with NRP2 mainly restricted to lymphatic EC (Fig. 1H through K). Although NRP1 mRNA was expressed in cardiomyocytes, neither NRP1 nor NRP2 protein was detected by IHC or CODEX. In the myocardium, NRP1-positive mononuclear cells along larger vascular beds (Fig. 1D) corresponded to CD68-positive macrophages in serial sections (Fig. 1E). scRNAseq analysis confirmed strong expression a Identification of tissue and cell type was based on histopathological and morphological evaluation of tissues from patients 1-3 (Table S1A) by two independent pathologists. Representative images of corresponding morphological structures are listed.
of NRP1 and NRP2-but not ACE2-in cardiac macrophages (Fig. 1G through K). ACE2 transcripts were largely confined to cardiomyocytes and pericytes, consistent with prior findings (33), and barely detectable in vascular EC (Fig. 1G). Compared with influenza pneumonitis and non-infected control myocardium, NRP1-positive capillaries were more frequent in COVID-19 (F (2, 23) = 0.91, P = 0.417); Fig. S7).
## Expression patterns suggest an important role for neuropilins in macro phages in fatal COVID-19 lungs
NRP1 and NRP2 proteins were heterogeneously expressed in larger vessel EC of the trachea and lung of COVID-19 patients (Fig. 2B inset; Fig. S5I andS6I), while absent from small capillaries. Their transcripts were detected in vascular EC (Fig. 2L andM). Intense NRP1 expression in alveolar macrophages was not specific to COVID-19 but was also noted in other causes of death (n = 28, Fig. S2). However, we identified COVID-19-associated syncytial macrophages with high phagocytic activity in the lung that were positive for NRP1 (Fig. 2A inset, 2D) and CD68 (Fig. 2A inset,2E) by CODEX and IHC. NRP1/CD68-positive macrophages, predominantly interstitial, expressed the microglial and macrophage marker ionized calcium-binding adapter molecule 1 (IBA1) in their cytoplasm, with ruffled membrane appearance (Fig. 2F through H), suggestive of migratory behavior (34). NRP2 protein expression was also detected in alveolar macrophages (Fig. 2C; Fig. S6A), consistent with scRNA-seq analysis (Fig. 2M andP). NRP2 transcripts were mainly expressed in lymphatic ECs (Fig. 2P), while NRP1 mRNA was predominantly expressed by vascular ECs, fibroblasts, and myofibroblasts (Fig. 2P). Alveolar type II (AT2) and goblet cells also expressed NRP1 mRNA (Fig. 2P), although NRP1 protein was largely absent in lung epithelial cells (Fig. S5A). Focal, weak NRP2 protein expression in epithelial cells, detected by IHC, was confirmed by transcript data (Fig. 2P).
ACE2 and TMPRSS2 mRNA were largely restricted to lung epithelial cells (Fig. 2K, O and P), while ECs, stromal cells, as well as lymphocytes and macrophage populations lacked expression. No overlap was observed between SARS-CoV-2-infected cell types (Fig. 2N) and those co-expressing ACE2 and TMPRSS2 (Fig. 2O). Only low levels of SARS-CoV-2 RNA were detected in lung epithelial cells from fatal COVID-19 patients (Fig. 2N).
We detected a significant upregulation of NRP1 transcripts in vascular ECs in COVID-19 lungs compared with noninfectious controls (P < 7.5E-6, Fig. S8A), while ACE2 and TMPRSS2 expression remained unchanged. NRP1 and the vascular endo thelial marker FLT1/VEGFR1 (fms-related receptor tyrosine kinase 1/VEGF receptor 1) transcript expression predominated in vascular ECs, while lymphatic ECs showed high NRP2 expression along with the lymphatic markers KDR/VEGFR2 (kinase insert domain receptor/VEGF receptor 2) and FLT4/VEGFR3 (fms related receptor tyrosine kinase 4/ VEGF receptor 3) (Fig. 2P). The activating ligand for VEGFRs, VEGFA, was produced and upregulated (P < 7.5E-6) by lung epithelial cells (Fig. 2P).
Both vascular and lymphatic ECs expressed high levels of the transforming growth factor beta receptor 2 (TGFBR2) (Fig. 2P), suggesting endothelial repair in response to viral infection (35). TGFBR2 was significantly upregulated in vascular ECs (P < 7.5E-6, Fig. S8A) in fatal COVID-19, while upregulation of the neuropilin-binding ligand TGFB1 (20) was observed in monocytes (Fig. S8B). Monocytes also showed upregulation of the TGFB1 inducing factor C-C motif chemokine ligand 2 (CCL2) (P < 7.5E-6, Fig. S8B) (36).
## SARS-CoV-2 RNA and NRP1 expression in lymphocytes in hematolymphoid organs
Similar to the lung, macrophages and sinus histiocytes in the lymph nodes of COVID-19 autopsies were strongly positive for NRP1 (Fig. S5F) and NRP2 (Fig. S6F). Follicular lymphocytes in lymph nodes and spleen were NRP1-positive and NRP2-negative (Fig. 3B andC; Fig. S5E). NRP1 expression in lymphocytes was not specific to COVID-19 but was also observed in influenza and noninfectious causes of death. CODEX multiplex analysis revealed that a subset of CD20-positive splenic B lymphocytes co-expressed NRP1 (Fig. 3A), while CD8-and CD4-positive lymphocytes were negative for NRP1 (Fig. 3E). Lymphocytes in the spleen and lymph nodes showed antisense SARS-CoV-2 RNA signals, as demonstrated by RNAscope FISH (Fig. 3G).
## NRP1 and NRP2 expression in rare cell types linked to COVID-19 phenotype
Severity of COVID-19 inversely correlates with the plasmacytoid dendritic cell (pDC) response (37). Tissue DCs in autopsy samples, identified by CD123 expression by CODEX (Fig. 3F), were positive for NRP1 in lymph nodes (Fig. 3E) but were NRP2-negative. NRP1 expression was also detected in tracheal DCs by IHC (Fig. S5I). Transcript analysis revealed NRP1 but not NRP2 expression in pDCs (Fig. 2P). Although NRP1 transcripts were present in mast cells (Fig. 2L andP), protein expression was not detected in a subset of mast cells that instead exhibited strong NRP2 expression by CODEX (Fig. 4E through G).
## NRP1 and NRP2 differentially bind to SARS-CoV-2 cleavage sites S1 and S1′
To explore the role of soluble S1 and S1′ in neuropilin binding (Fig. S1), NRP2-negative HEK293 cells, which endogenously express NRP1, were transfected to express NRP2. These cells express low to undetectable levels of the SARS-CoV-2 (co-)receptors ACE2, FURIN, and TMPRSS2 (Fig. S3). HEK293 cells not expressing NRP2 (Fig. S4I) served as internal negative controls to quantify NRP2 binding to spike fragments. As previously shown (13,15), endogenous NRP1 on HEK293 cells bound the soluble S1 spike fragment (Fig. 5A through C). However, S1′ was not bound by NRP1 (Fig. 5E through G). Only recombinant expression of NRP2 enabled HEK293 cells to additionally bind S1′ soluble spike fragment (Fig. 5N through P), with preserved S1 binding (Fig. 5I through L). Manual cell counts revealed NRP2 expression and S1′ binding in a similar fraction of NRP2-trans fected cells (5%; 4%).
## DISCUSSION
This study presents a comprehensive analysis of NRP1 and NRP2 protein expression in COVID-19 in situ by IHC and CODEX, alongside transcriptomic data from publicly available scRNAseq data sets (30). Our findings highlight the critical role of neuropilins in SARS-CoV-2 infection within the lung, heart, and hematolymphoid tissues. We observed concordant NRP1 and/or NRP2 protein and transcript expression, especially in cell types with minimal ACE2 expression, including vascular ECs, macrophages, B lymphocytes, and mast cells. Primary targets of SARS-CoV-2 infection are nasopharyngeal, bronchial, and lung epithelial cells, with ACE2 and TMPRSS2 as the primary entry factors (Fig. 2O) (9). However, minimal ACE2 and TMPRSS2 expression in SARS-CoV-2-positive cell clusters (Fig. 2N andO) suggests neuropilins, among others, as alternative entry factors. It should be noted that productive infection of epithelial cells can induce severe cytopathic effects, potentially hindering their detection in scRNA-seq analyzes. Our findings demonstrate the presence of SARS-CoV-2 RNA within vascular ECs, macrophage, B lymphocyte, and mast cell clusters expressing neuropilins, further supporting their role in viral entry (Fig. 2L andM). While monocytes and lymphocytes are reported to lack ACE2 expression (38), consistent with our own findings (Fig. 2K andP), we observed strong neuropilin expression in B lymphocytes and macrophages (Fig. 1 to 3). We detected antisense SARS-CoV-2 RNA in lymphocytes in spleen and lymph nodes. While the precise S1A) and influenza-related death (patient 5; Table S1B) with corresponding areas in H&E (left column), using serial sections. In COVID-19 cases, lymphocytes within the spleen and (Continued on next page) identity and localization of the signal remain uncertain, B lymphocytes are known to be permissive to productive SARS-CoV-2 infection (39), and NRP1 expression provides a potential mechanism for the severe lymphopenia associated with poor outcomes in COVID-19 patients (40). Nrp1 is also a surface marker of CD4+ CD25+ T regulatory cells (Tregs) and is co-regulated with Foxp3 (forkhead box P3), an important regulator for Tregs (41). The potential role of Tregs in severe COVID-19 and PASC has been discussed recently (42), including findings on T-cell dysregulation in PASC (43). However, contrary to previous results (41), we did not detect relevant NRP1 mRNA in Tregs in COVID-19 lungs (Fig. 2P).
The importance of alternative docking receptors aligns with previous reports of SARS-CoV-2 virus particles in vascular ECs despite low ACE2 levels (27). Biochemical studies (15) indicate that NRP1 enhances infectivity in the presence of ACE2 by binding the S1 peptide to its b1 domain but does not facilitate viral attachment independently. These results were later extended, showing that NRP1 interacts with the receptor binding domain (RBD) and full trimeric spike protein with higher binding frequency than ACE2, thus enhancing viral attachment in vitro (44). Similarly, NRP2 has been shown to interact with the RBD of SARS-CoV-2 spike protein in vitro (45), indicating that neuropilins may serve roles beyond simple co-factors. Our in vitro studies (Fig. 5) confirm the binding of NRP1 to the novel and SARS-CoV-2-specific S1 spike fragment (Fig. S1A), whose C-terminus (-PRRAR*; ProArgArgAlaArg*) complies with the CendR consensus sequence S1A).
Scale bars correspond to 50 µm. CODEX color code: DAPI (blue), CD8 (cyan), NRP2 (green), mast cell tryptase (yellow), NRP1 (red), and CD16 (magenta). (-RXXR*) (17). Proteolytic cleavage at the S1′/S2′ site generates a non-canonical CendR motif at the C terminus of S1′ (-KPSKR*; LysProSerLysArg*), which is conserved in SARS-CoV-1 ( 16) and is only bound in the presence of NRP2 (Fig. 5). NRP1 lacked binding capacity for S1′. This suggests greater flexibility of NRP2 towards CendR-like peptides, consistent with its binding to the C-terminal end of mature VEGFC protein (46), which does not conform to the CendR consensus (17). The emergence of the S1/S2 cleavage site in SARS-CoV-2 resulted in S1 peptide and de novo NRP1 binding. After deletion of the S1/S2 cleavage site, there is still evidence of S2′ spike proteolytic fragments, which suggests the existence of S1′ in vitro (47). Additionally, Frolova et al. (2022) showed that after deletion of the S1/S2 cleavage site, syncytia formation is still observed in cells overexpressing TMPRSS2. This study also reported the presence of a band reminiscent of S2′, suggesting that cleavage at S1/S2 (furin cleavage site) is not an absolute requirement for cleavage at S1′/S2′ (48). Our data indicate that, apart from a differential expression of neuropilins throughout the body, differential recognition motifs for viral peptides may also contribute to COVID-19 pathology. The ability of S1′ to bind via NRP2, but not NRP1 alone, suggests differential interactions between the two neuropilins and CendR-com patible proteins. This may also apply to other viruses exploiting neuropilin-mediated cellular uptake (19). More abundant NRP1 expression in situ compared with NRP2 (Fig. 1 and2; Table 1; Fig. S5 andS6) may reflect differences relevant to clinical severity and viral dissemination. It is possible that not only the cell tropism based on NRP1 and NRP2 expression contrib utes to the pathology of acute COVID-19 and PASC, but also the differential binding of neuropilins to proteolytic spike fragments. Future studies should address whether soluble circulating spike proteins, observed in PASC patients ( 6), interact with neuropilins on ECs, potentially contributing to chronic microangiopathy. Although neuropilin expression patterns were similar in COVID and non-COVID-asso ciated death, we identified a trend toward increased NRP1 protein levels in myocardial capillary ECs (Fig. S7) and transcriptional upregulation in pulmonary vascular ECs in COVID-19 (Fig. S8A), consistent with prior findings (27). High NRP1 expression in cardiac microcapillaries (Fig. 1B, H andK), combined with scarce ACE2 expression (Fig. 1G andK), supports earlier reports of low or negligible ACE2 transcripts in ECs (10,13,14,33). Occasional detection of ACE2 protein despite minimal transcript detection (49,50) may reflect potential methodological challenges due to different antibodies.
SARS-CoV-2 infection triggers a cytokine storm, complement over-activation, and thrombosis, leading to endothelial damage with tissue hypoxia, angiogenesis (51,52), and vascular remodeling (27,32). Werlein et al. (32) reported intussusceptive angiogen esis in COVID-19 hearts alongside increased CD11b/TIE2 positive macrophages near vessels, suggesting a promoting role in driving angiogenesis. Interestingly, we observed a subset of cardiac macrophages strongly positive for NRP1 in the vicinity of larger vessels (Fig. 1D). Also in the lung, intussusceptive angiogenesis in COVID-19 has been demonstrated, accompanied by an upregulation of NRP1, NRP2, and VEGFA (27), in partial agreement with our own findings of NRP1 and NRP2 protein and transcript expression in the lung (Fig. 2; Fig. S8A). Autopsy results revealed small vessel occlusion and microangi opathy alongside endothelial damage to be involved in the pathogenesis of COVID-19 (5). Endothelial Nrp1 is essential for sprouting angiogenesis (53). Even without replicating virus, spike protein binding to neuropilins may initiate signaling. The interplay between virus-host interactions and NRP1 expression may contribute to initial micro-thrombotic events, intussusceptive angiogenesis, acute and chronic hypoxemia, and subsequent tissue remodeling and fibrosis (54), linking acute infection to long-term sequelae such as PASC.
Another observed complication of acute COVID-19 with long-term consequences is new-onset hyperglycemia and diabetes. Wu et al. (2021) (24) found that pancreatic β cells exhibit selectively high expression of NRP1, with low ACE2 and TMPRSS2 expression at both mRNA and protein levels. They also showed that SARS-CoV-2 specifically infects pancreatic β cells and induces apoptotic cell signaling, which was reduced by NRP1 inhibition, supporting NRP1-mediated β cell targeting.
NRP2 is upregulated in macrophages and during pro-inflammatory states in the heart and lung (55). Therapeutic modulation of NRP2 has been discussed for myelofibrosis (56) and has shown promise in late-stage clinical trials for pulmonary fibrosis in sarcoidosis patients (55), where the possible mechanism of action of the NRP2 agent Efzofitimod is to reduce pro-inflammatory macrophages and to prevent the progression of pulmo nary fibrosis (55). We observed abundant NRP1 and NRP2 expression in alveolar and interstitial lung macrophages (Fig. 2; Fig. S2, S5A and S6A), extending previous findings (20). IBA1 expression at the leading edge of these cells (Fig. 2F through H) suggests migratory activity. Macrophage overactivation, in particular by syncytial macrophages expressing NRP1 and NRP2 (Fig. 2A andD), may contribute to cytokine storm (57,58), and syncytia formation is associated with severe disease (59).
Pro-inflammatory pathways transition to fibrosis-related pathways in prolonged COVID-19 (54), with TGFβ playing a key role (60). We detected significant upregula tion of the neuropilin ligand TGFB1 and chemokine CCL2 in monocytes from fatal COVID-19 cases (Fig. S8B). The significant upregulation of TGFBR2 in vascular ECs of COVID-19 patients (Fig. S8A) suggests that TGFβ signaling may play a role in the endothelial response to viral infection. This is consistent with previous findings of active TGFβ signaling through TGFBR2 in infected endothelium (35). Endothelial TGFβ signaling increases vascular permeability (61), contributing to vascular leakage and locally compromising the blood-brain barrier in PASC patients, who exhibit elevated serum TGFβ levels (62). Mast cells, also activated by TGFβ (63), express NRP2 (Fig. 4), contain SARS-CoV-2 RNA (Fig. 2N), and are devoid of ACE2 (Fig. 2K), extending previous findings (20,30). Mast cell activation has been implicated in PASC pathophysiology (64). Additionally, we observed CD16 (IgG Fc receptor III, FcγIIIb) expression on NRP2positive mast cells (Fig. 4H), which has been linked to systemic mastocytosis (65). As the mechanism of NRP2 action in chronic mast cell activation is still unclear, further investigation into NRP2 as a potential therapeutic target in PASC is warranted.
## Limitations
The cohorts in this study, comprising COVID-19 patients (n = 20) and influenza/ noninfectious controls (n = 13), were relatively small. NRP2 expression detected by IHC appeared low across samples, with variable intensity likely due to differences in post-mortem intervals. Such variability may introduce noise into the data analysis and obscure biological differences. Further preclinical studies are necessary to evaluate NRP2 modulation as a potential therapeutic approach.
## Conclusion
Combining mRNA and protein expression data in situ, we identified NRP1 and NRP2 on cell types exhibiting minimal ACE2 and TMPRSS2 expression in the lung, heart, and hematolymphoid organs. In summary, NRP1 was detected in myeloid cells, B lympho cytes, and vascular ECs, while NRP2 expression was largely restricted to macrophages and mast cells. Our study suggests that primary infection with SARS-CoV-2 in the upper respiratory tract depends on ACE2, while neuropilins are important for systemic viral dissemination, monocyte/macrophage-induced vascular damage, microangiopathy, and thrombosis. In this context, neuropilins might play a multifaceted role in SARS-CoV-2 pathogenesis, contributing to both acute and chronic complications. Neuropilins, particularly NRP2, represent promising druggable targets in the treatment of pulmonary fibrosis. As macrophages and mast cells are critically involved in viral spreading, excessive cytokine storm, and fibrotic remodeling of infected tissue, modulation of these cells by targeting NRP2 might offer a new therapeutic strategy in acute COVID-19 as well as in PASC. Further research is necessary to fully understand the functional roles of NRP1 and NRP2 in disease progression.
## MATERIALS AND METHODS
## Patient samples
Autopsy tissues were analyzed using tissue microarrays (TMA) from three different institutions in Germany: University Hospital Bonn (UKB), University Hospital Aachen, and Hannover Medical School (MHH). A total of 20 patients who died from COVID-19 in 2020 were compared with seven patients who succumbed to influenza type H1N1, seasonal A or B. Additionally, six cases of noninfectious deaths were used as controls. The COVID-19 cohort included 11 females and 9 males (mean age: 70.25 ± 11.22 years). The influenza cohort comprised five females and two males (mean age: 50.86 ± 15.24 years), while the uninfected cohort included two females and four males (mean age: 71.25 ± 13.91 years). Infection with SARS-CoV-2 or influenza was confirmed by nasopharyngeal swabs (RT-PCR for SARS-CoV-2 or influenza RNA, respectively). Patient characteristics are provided in Table S1.
## Immunohistochemistry
TMAs of formalin-fixed paraffin-embedded organ samples were assembled using tissue cores 3-4 mm in diameter. Standard paraffin sections (3-4 µm) following standard embedding (fixation: 4% buffered formalin) were used. Morphology was evaluated by hematoxylin-eosin staining. For IHC staining of CD34 and CD68, slides were processed as previously described (66). Serial section staining for CD34 (QBEnd10, Agilent, Santa Clara, USA; 1:200) and CD68 (PG-M1, Agilent; 1:100) was performed using the semiauto matic Autostainer 480S (Medac, Wedel, Germany; for staining conditions see Table S2). Serial section staining for NRP1 (EPR3113, Abcam, Cambridge, UK; 1:200) was performed after deparaffinization with EZ prep (Roche, Basel, Switzerland) and treatment with CC1 buffer (pH 8, Roche) using the automatic BenchMark Ultra staining platform with the OptiView detection kit (Roche; Table S2). TMAs from COVID-19 patients 1-3 (Table S1A) were stained manually in the case of NRP2 with 15 µg/mL (aNRP2-36v2, aTyr Pharma; for protocol see reference 67). Photomicrographs were acquired using either a BX51 microscope (Olympus, Hamburg, Germany) with Zeiss AxioCam MRc5 and Axiovision software (Carl Zeiss, Oberkochen, Germany) or digitally scanned on a Leica scanner (Aperio GT 450 DX, Leica Biosystems, Wetzlar, Germany) and analyzed with QuPath v0.4.3 (68).
For quantification of NRP1-positive myocardial capillaries (Fig. S7), three representa tive squares in each tissue core were selected in QuPath (each 0.0625 mm 2 ) and positive capillaries were counted manually. A mean of three squares was calculated per patient. If multiple cores per patient were available, means were averaged across cores. A total of 35 myocardium samples were analyzed (COVID-19 n = 13 patients, influenza n = 7 patients, noninfectious controls n = 6 patients), and an analysis of variance (ANOVA) was conducted. Insufficient tissue samples were excluded from analysis and staining results confirmed by two independent observers (A.D., I.G.). Variable interpatient NRP1 intensity of staining was observed, possibly reflecting differences in autolytic degradation of autopsy tissue.
Semiquantitative scoring of NRP1 intensity in alveolar macrophages in lungs was conducted in a blinded, independent, and randomized manner. Representative peripheral lung areas per core were photographed in QuPath and evaluated blindly. A total of 56 lung samples were analyzed (COVID-19 n = 16 patients, influenza n = 7 patients, noninfectious controls n = 5 patients). One noninfectious control case (no. 3) was excluded due to lymphangiosis carcinomatosa in the lung. Staining intensities were scored twice independently as follows: 0, negative; 1, low; 2, moderate; and 3, strong (Fig. S2). Discrepant core results were reviewed for a final consensus score.
## CODEX multiplexed tissue imaging
1-3 µm sectioned FFPE TMAs of COVID-19 patients 2 and 3 (Table S1A) were prepared and stained for CODEX-enabled multiplexed tissue imaging following manufacturer's instructions. The full protocol can be found in (69). A list of antibodies is provided in Table S3. Nuclei were detected using DAPI. Images were analyzed using the Enable Medicine platform (Menlo Park, CA, USA).
## SARS-CoV-2 RNA detection with fluorescence in situ hybridization (FISH)
We performed FISH on 1-µm TMA sections of spleen and lymph node from two COVID-19 cases (patients 2 and 3, Table S1A) with the RNAscope Multiplex Fluorescent Reagent Kit v2 assay (Advanced Cell Diagnostics, Inc., Hayward, CA, USA), following the protocol previously described by reference 31.
## In vitro binding experiment
## Cloning of soluble spike fragments
Cloning of soluble SARS-CoV-2 spike protein fragments was performed by PCR-based site-directed mutagenesis with Phusion Green Hot Start II High-Fidelity (ThermoFisher, Waltham, USA) using pEXPR TO FRT SARS-CoV-2 S d18-SH as template. Mutagenesis primers (Eurofins, Ebersberg, Germany) used in this approach are listed in Table S4. In brief, coding sequences for spike protein fragments S1 and S1′ were generated by introducing a STOP codon at amino acid positions 686 (S1; Ser686Ter) and 816 (S1′; Ser816Ter) and concordantly deleting 3′ sequences of the original spike protein cDNA (NCBI accession number YP_009724390.1). PCR conditions were as follows: 98°C/ 30 s, 25× (98°C/10 s, 58°C/10 s, 72°C/90 s), 72°C/10 min. PCR fragments were purified using QIAprep Spin Miniprep Kit (Qiagen, Hilden, Germany), A-tailed by incubation with Kapa2G fast HS genotyping mix (Merck, Darmstadt, Germany) in the absence of primers for 20 min at 68°C and cloned into pCR4 using the pCR4 TOPO sequencing kit (Life Technologies, Bleiswijk, Netherlands). Re-cloning into the pCMV6-Entry mamma lian expression vector (Origene Technologies, Rockville, USA) used flanking SfaAI and NotI (ThermoFisher) restriction sites. Sequences were verified by Sanger sequencing (Eurofins). To increase secretion, cDNAs for S1 and S1′ fragments were modified by replacing the endogenous by the NRP2 signal peptide in pCMV6-S1 and pCMV6-S1′ (data not shown), followed by a N-terminal hemagglutinin (HA) sequence, using pCMV6-HA-NRP2 as template (72). Primer sequences are listed in Table S4.
## Production of soluble spike fragments
HEK293 and CaCo-2 cells were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ), Braunschweig, Germany. HEK293 cells were cultivated in DMEM/F12 media supplemented with 10% fetal bovine serum (FBS) and penicillin/strep tomycin (Life Technologies). CaCo-2 cells were grown in MEM/20% FBS and non-essential amino acid supplement. The cells were routinely checked for mycoplasma infection and were free of any contamination. HEK293 and CaCo-2 cDNA were used for real-time quantitative PCR (qRT-PCR) as described in Fig. S3. Expression vectors containing S1 and S1′ fragments were transfected in HEK293 cells using the Effectene transfection method (Qiagen). Stable HEK293 clones were obtained by selection with 400 µg/mL G418 sulfate (Life Technologies). Expression of spike protein fragments in HEK293 clones was determined by Western blotting from cell lysates. Secretion of spike protein fragments was determined from serum-free cell culture supernatant in the presence of 1 mM ACE2 inhibitor (MLN-4760, MedChemExpress, Monmouth Junction, USA) and 10 mM NRP1 inhibitor (EG01377, MedChemExpress) using SARS-CoV-2 spike antibody (ab277628, Abcam, Amsterdam, NL). Secreted spike protein fragments were purified from cell culture supernatants by exploiting their binding to heparin using HiTrap heparin HP affinity columns (Cytiva, Freiburg, Germany) with batch elution with increasing sodium chloride concentration in phosphate buffered saline (pH 7.4, PBS, ThermoFisher). HEK293 cells stably expressing NRP2v2 (RC220706, Origene Technologies) were generated accordingly and checked for expression by Western blot of cell lysates using anti-NRP2 antibody (AF2215, R&D Systems, Abingdon, UK).
## Immunofluorescence
A total of 50,000 cells per 12-mm cover slips were seeded in complete media. After at least 6 h, cells were treated according to the experimental condition and incubated overnight. Immunofluorescence was performed as previously described (73) with the following modifications: cells were fixed on ice for 10 min with 4% formaldehyde in PBS and treated with 1% BSA and 0.5% Triton X-100 in PBS for 20 min prior to incubation with primary and secondary antibodies (Table S5). NRP1 and NRP2 on HEK293 cells were detected using specific antibodies (EPR3113 and EPR23808-72, respectively; Abcam Ltd., Cambridge, UK). HEK293 cells not expressing the NRP2 construct were used as internal negative controls for binding specificity. C29F4 antibody (Cell Signaling Technology, Leiden, Netherlands) was used for HA-tagged spike protein fragments. Rabbit isotype controls were purchased from Cell Signaling Technology. Controls are depicted in Fig. S4. Cell membrane/F-actin staining was performed using phalloidin (A30104, Thermo Fisher Scientific, Langerwehe, Germany) and cells were embedded using Fluoromount-G mounting medium with DAPI (Thermo Fisher Scientific).
## References
1. Jing, Wu, Xiang et al. (2022) "Pathophysiologi cal mechanisms of thrombosis in acute and long COVID-19" *Front Immunol*
2. Wrapp, Wang, Corbett et al. (2020) "Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation" *Science*
4. Jafarzadeh, Chauhan, Saha et al. (2020) "Contribution of monocytes and macrophages to the local tissue inflammation and cytokine storm in COVID-19: Lessons from SARS and MERS, and potential therapeutic interventions" *Life Sci*
5. Schulte-Schrepping, Reusch, Paclik et al. (2020) "Severe COVID-19 is marked by a dysregulated myeloid cell compartment" *Cell*
6. Iba, Levy, Maier et al. (2024) "Four years into the pandemic, managing COVID-19 patients with acute coagulopathy: what have we learned?" *J Thromb Haemost*
7. Proal, Vanelzakker, Aleman et al. (2023) "SARS-CoV-2 reservoir in post-acute sequelae of COVID-19 (PASC)" *Nat Immunol*
8. Davis, Mccorkell, Vogel et al. (2023) "Long COVID: major findings, mechanisms and recommendations" *Nat Rev Microbiol*
9. Alrajhi (2023) "Post-COVID-19 pulmonary fibrosis: an ongoing concern" *Ann Thorac Med*
10. Jackson, Farzan, Chen et al. (2022) "Mechanisms of SARS-CoV-2 entry into cells" *Nat Rev Mol Cell Biol*
11. Li, Li, Zhang et al. (2020) "Expression of the SARS-CoV-2 cell receptor gene ACE2 in a wide variety of human tissues" *Infect Dis Poverty*
12. Zhou, Niu, Jiang et al. (2020) "SARS-CoV-2 targets by the pscRNA profiling of ACE2, TMPRSS2 and furin proteases"
13. Lindskog, Méar, Virhammar et al. (2022) "Protein expression profile of ACE2 in the normal and COVID-19-affected human brain" *J Proteome Res*
14. Cantuti-Castelvetri, Ojha, Pedro et al. (2020) "Neuropilin-1 facilitates SARS-CoV-2 cell entry and infectivity" *Science*
15. Mccracken, Saginc, He et al. (2021) "Lack of evidence of angiotensin-converting enzyme 2 expression and replicative infection by SARS-CoV-2 in human endothelial cells" *Circulation*
16. Daly, Simonetti, Klein et al. (2020) "Neuropilin-1 is a host factor for SARS-CoV-2 infection" *Science*
17. 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*
18. Teesalu, Sugahara, Kotamraju et al. (2009) "C-end rule peptides mediate neuropilin-1-dependent cell, vascular, and tissue penetration" *Proc Natl Acad Sci*
19. Ishitoku, Mokuda, Araki et al. (2023) "Tumor Full-Length Text Journal of Virology November"
20. "necrosis factor and interleukin-1β upregulate NRP2 expression and promote SARS-CoV-2 proliferation" *Viruses*
21. Balistreri, Yamauchi, Teesalu (2021) "A widespread viral entry mechanism: the C-end Rule motif-neuropilin receptor interaction" *Proc Natl Acad Sci*
22. Roy, Bag, Singh et al. (2017) "Multiface ted role of neuropilins in the immune system: potential targets for immunotherapy" *Front Immunol*
23. Schellenburg, Schulz, Poitz et al. (2017) "Role of neuropilin-2 in the immune system" *Mol Immunol*
24. Islam, Mishra, Bodas et al. (2022) "Role of neuropilin-2-mediated signaling axis in cancer progression and therapy resistance" *Cancer Metastasis Rev*
25. Abebe, Ayele, Muche et al. (2021) "Neuropilin 1: a novel entry factor for SARS-CoV-2 infection and a potential therapeutic target" *Biol Targets Ther*
26. Wu, Lidsky, Lee et al. (2021) "SARS-CoV-2 infects human pancreatic β cells and elicits β cell impairment" *Cell Metab*
27. Assous, Martinez, Eisenberg et al. (2019) "Neuropilin 2 signaling mediates corticostriatal transmission, spine maintenance, and goal-directed learning in mice" *J Neurosci*
28. Lambrinos, Cristofaro, Pelton et al. (2022) "Neuropilin 2 is a novel regulator of distal colon contractility" *Am J Pathol*
29. Ackermann, Verleden, Kuehnel et al. (2020) "Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in Covid-19" *N Engl J Med*
30. Sarabipour, Gabhann (2021) "Targeting neuropilins as a viable SARS-CoV-2 treatment" *FEBS J*
31. Goltsev, Samusik, Kennedy-Darling et al. (2018) "Deep profiling of mouse splenic architecture with CODEX multiplexed imaging" *Cell*
32. Delorey, Ziegler, Heimberg et al. (2021) "COVID-19 tissue atlases reveal SARS-CoV-2 pathology and cellular targets" *Nature*
33. Wong, Klinkhammer, Djudjaj et al. (2021) "Multisystemic cellular tropism of SARS-CoV-2 in autopsies of COVID-19 patients" *Cells*
34. Werlein, Ackermann, Stark et al. (2023) "Inflammation and vascular remodeling in COVID-19 hearts" *Angiogenesis*
35. Chen, Li, Chen et al. (2020) "The ACE2 expression in human heart indicates new potential mechanism of heart injury among patients infected with SARS-CoV-2" *Cardiovasc Res*
36. Ohsawa, Imai, Kanazawa et al. (2000) "Involvement of Iba1 in membrane ruffling and phagocytosis of macrophages/microglia" *J Cell Sci*
37. Zhao, Xue, Weiner et al. (2024) "TGF-βR2 signaling coordinates pulmonary vascular repair after viral injury in mice and human tissue" *Sci Transl Med*
38. Gharaee-Kermani, Denholm, Phan (1996) "Costimulation of fibroblast collagen and transforming growth factor β1 gene expression by monocyte chemoattractant protein-1 via specific receptors" *J Biol Chem*
39. Venet, Ribeiro, Décembre et al. (2023) "Severe COVID-19 patients have impaired plasmacytoid dendritic cell-mediated control of SARS-CoV-2" *Nat Commun*
40. Radzikowska, Ding, Tan et al. (2020) "Distribution of ACE2, CD147, CD26, and other SARS-CoV-2 associated molecules in tissues and immune cells in health and in asthma, COPD, obesity, hypertension, and COVID-19 risk factors" *Allergy*
41. Pontelli, Castro, Martins et al. (2022) "SARS-CoV-2 productively infects primary human immune system cells in vitro and in COVID-19 patients" *J Mol Cell Biol*
42. Huang, Pranata (2020) "Lymphopenia in severe coronavirus dis ease-2019 (COVID-19): systematic review and meta-analysis" *J Intensive Care*
43. Bruder, Probst-Kepper, Westendorf et al. (2004) "Neuropilin-1: a surface marker of regulatory T cells" *Eur J Immunol*
44. Haunhorst, Bloch, Javelle et al. (2022) "A scoping review of regulatory T cell dynamics in convalescent COVID-19 patients -indications for their potential involvement in the development of Long COVID?" *Front Immunol*
45. Yin, Peluso, Luo et al. (2024) "Long COVID manifests with T cell dysregulation, inflammation and an uncoordinated adaptive immune response to SARS-CoV-2" *Nat Immunol*
46. Hou, Cao, Kim et al. (2023) "Biophysical investigation of interactions between SARS-CoV-2 spike protein and neuropilin-1" *Protein Sci*
47. Husain, Yuen, Sun et al. (2022) "Cell-based receptor discovery identifies host factors specifically targeted by the SARS CoV-2 spike" *Commun Biol*
48. Parker, Linkugel, Goel et al. (2015) "Structural basis for VEGF-C binding to neuropilin-2 and sequestra tion by a soluble splice form" *Structure*
49. Johnson, Xie, Bailey et al. (2021) "Loss of furin cleavage site attenuates SARS-CoV-2 pathogenesis"
50. Frolova, Palchevska, Lukash et al. (2022) "Acquisition of furin cleavage site and further SARS-CoV-2 evolution change the mechanisms of viral entry, infection spread, and cell signaling" *J Virol*
51. Xu, Ilyas, Weng (2023) "Endothelial dysfunction in COVID-19: an overview of evidence, biomarkers, mechanisms and potential therapies" *Acta Pharmacol Sin*
52. Hikmet, Méar, Edvinsson et al. (2020) "The protein expression profile of ACE2 in human tissues" *Mol Syst Biol*
53. Shweiki, Itin, Soffer et al. (1992) "Vascular endothelial growth factor induced by hypoxia may mediate hypoxia-initiated angiogenesis" *Nature*
54. Barbosa, Gonçalves, De Araujo et al. (2021) "Endothelial cells and SARS-CoV-2: an intimate relationship" *Vascul Pharmacol*
55. (2025) *Full-Length Text Journal of Virology*
56. Fantin, Vieira, Plein et al. (2013) "NRP1 acts cell autonomously in endothelium to promote tip cell function during sprouting angiogenesis" *Blood*
57. Ackermann, Kamp, Werlein et al. (2022) "The fatal trajectory of pulmonary COVID-19 is driven by lobular ischemia and fibrotic remodelling" *EBioMedicine*
58. Dhupar, Powers, Eisenberg et al. (2024) "Orchestrating resilience: how neuropilin-2 and macrophages contribute to cardiothoracic disease" *J Clin Med*
59. Vosbeck, Förster, Mayr et al. (1924) "Neuropilin2 in mesenchymal stromal cells as a potential novel therapeutic target in myelofibrosis" *Cancers (Basel)*
60. Felkle, Zięba, Kaleta et al. (2023) "Overreactive macrophages in SARS-CoV-2 infection: the effects of ACEI" *Int Immunopharmacol*
61. Milde, Ritter, Tennent et al. (2015) "Multinucleated giant cells are specialized for complement-mediated phagocytosis and large target destruction" *Cell Rep*
62. Rajah, Bernier, Buchrieser et al. (2022) "The mechanism and consequences of SARS-CoV-2 spike-mediated fusion and syncytia formation" *J Mol Biol*
63. Zhang, Phan (1996) "Cytokines and pulmonary fibrosis" *Biol Signals*
64. Lee, Kayyali, Sousa et al. (2007) "Transforming growth factor-β1 effects on endothelial monolayer permeability involve focal adhesion kinase/Src" *Am J Respir Cell Mol Biol*
65. Greene, Connolly, Brennan et al. (2024) "Blood-brain barrier disruption and sustained systemic inflammation in individuals with long COVID-associated cognitive impairment" *Nat Neurosci*
66. Haque, Frischmeyer-Guerrerio (2022) "The role of TGFβ and other cytokines in regulating mast cell functions in allergic inflammation" *Int J Mol Sci*
67. Sumantri, Rengganis (2023) "Immunological dysfunction and mast cell activation syndrome in long COVID" *Asia Pac Allergy*
68. Teodosio, Mayado, Sánchez-Muñoz et al. (2015) "The immunophenotype of mast cells and its utility in the diagnostic work-up of systemic mastocytosis" *J Leukoc Biol*
69. Koerber, Schneider, Pritchard et al. (2023) "Nestin expression in osteocytes following myeloablation and during bone marrow metastasis" *Br J Haematol*
70. Förster, Chong, Siefker et al. (2023) "Development and characterization of a novel neuropilin-2 antibody for immunohistochemical staining of cancer and sarcoidosis tissue samples" *Monoclon Antib Immunodiagn Immunother*
71. Bankhead, Loughrey, Fernández et al. (2017) "QuPath: open source software for digital pathology image analysis" *Sci Rep*
72. Black, Phillips, Hickey et al. (2021) "CODEX multiplexed tissue imaging with DNA-conjugated antibodies" *Nat Protoc*
73. Van Rg, Drake (2010) "The Python language reference Release 3.0.1 [Repr.]. Python Software Foundation"
74. Wolf, Angerer, Theis (2018) "SCANPY: large-scale single-cell gene expression data analysis" *Genome Biol*
75. Dutta, Polavaram, Islam et al. (2022) "Neuropilin-2 regulates androgenreceptor transcriptional activity in advanced prostate cancer" *Oncogene*
76. Becker, Förster, Gielen et al. (2019) "Paucimannosidic glycoepitopes inhibit tumorigenic processes in glioblastoma multiforme" *Oncotarget*
77. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607758&blobtype=pdf | # WDR81 represses IKK-mediated expression of pro-survival genes to regulate apoptosis
Sannoong Hu, Pranav Danthi
## Abstract
Apoptosis is a common host response to virus infection. The extent and timing of apoptosis following infection is controlled by the balance between the strength of signals that activate death-inducing and survival-promoting pathways in cells. In many cell types, infection with mammalian orthoreovirus (reovirus) results in induction of cell death by apoptosis late in infection. In this study, we uncovered that WD repeat-containing protein 81 (WDR81) is required for apoptosis induction after reovirus infection. The requirement for WDR81 for apoptosis induction is not unique to reovirus because cells lacking WDR81 are also resistant to apoptosis induced by other agonists. We find that in cells deficient in WDR81, expression of several pro-survival genes is upregulated. The expression of these genes is controlled by the inhibitor of κB kinase (IKK) complex-nuclear factor of kB (NFκB) signaling pathway. When IKK signaling is blocked in WDR81-deficient cells, pro-survival gene expression is restored to normal levels, and the cells regain their susceptibility to cell death triggers. Our work uncovers a new function for WDR81 in controlling apoptosis. Additionally, it reveals a previously unknown link between an endosomally localized protein, WDR81, and IKK-NFκB signaling.IMPORTANCE Virus infection often results in the death of the infected cells. Cell death prior to generation of virus progeny limits the spread of infection to neighboring cells and therefore can be beneficial to the host. However, cell death might also cause tissue destruction and could contribute to viral disease. It is therefore important to understand how cell death is controlled. Here, we uncover a cell death-regulating role for WD repeat-containing protein 81 (WDR81)-a cellular protein that has not been previously implicated in affecting cell death. We find that when this protein is absent, cells express a much greater level of survival signals. These survival signals prevent efficient induction of cell death. By investigating how these survival signals are expressed, we reveal a new link between WDR81 and nuclear factor of kB (NFκB), a well-known cellular survival pathway.
KEYWORDS reovirus, cell death, apoptosisA poptosis is a well-studied form of programmed cell death. Apoptosis is character ized by activation of a cascade of proteases called caspases (1, 2). The activation of these caspases ultimately leads to the death of the cell in a non-inflammatory manner (2). Apoptosis can be triggered by extrinsic or intrinsic pathways. Extrinsic signals such as tumor necrosis factor (TNF) can bind to their receptors and trigger the activation of an initiator caspase, caspase-8. In its activated form, caspase-8 cleaves effector caspases, caspase-3 and caspase-7. The intrinsic pathway involves the mitochondria and results in the activation of caspase-9, which, in turn, also activates effector caspases. Activated effector caspases cleave numerous cellular components, ultimately leading to cell death (2).
Cells express proteins that prevent the spontaneous activation of caspases (2)(3)(4). The expression of these pro-survival gene products is usually mediated by the inhibi tor of κB kinase (IKK) complex via the nuclear factor of kB (NFκB) transcription factor (3,4). Thus, in the absence of a cell death trigger, the cell continues to grow and divide without activating caspases. Apoptosis can be triggered to overcome survival signals by extrinsic ligands such as TNF or TNF-related apoptosis-inducing ligand (TRAIL) or by intrinsic imbalances such as stress or intracellular pathogens like viruses (5)(6)(7). Virus infection and replication within a cell is often detected by pathogen-associated molecular patterns, which trigger a variety of cellular immune responses, including cell death (5,8). Mammalian orthoreovirus (reovirus) infection triggers apoptosis (9)(10)(11). While reovirus-activated apoptosis is not associated with limiting viral replication within the infected cell, it has been associated with disease pathogenesis in mouse models (12)(13)(14). Reovirus-induced apoptosis requires activation of both the extrinsic and the intrinsic apoptotic pathways (11,(15)(16)(17). While a few viral and cellular factors are implicated in this process (10,(18)(19)(20)(21)(22), the exact manner in which apoptotic pathways are activated in reovirus-infected cells is not understood.
We previously identified a function for a host protein, WD repeat-containing protein 81 (WDR81), in reovirus infection (23). WDR81 regulates endosomal maturation (24). It is present in the phosphoinositide-3-kinase (PI3K) complex on the outer leaflet of early endosomes and inhibits the function of class III PI3Ks, thus allowing the maturation of the early endosome to late endosome (24). In the absence of WDR81, reovirus virions attach to cell surface receptors, reach the endosome, and are disassembled to generate entry intermediates called infectious subvirion particles (ISVPs) by the action of endosomal proteases (23). However, the ISVPs are trapped in the endosomes and fail to enter the cytoplasm and launch infection. When infection is launched with in vitro generated ISVPs, which do not require endosomal uptake, infection proceeds normally (23). These data implicate WDR81 in regulating productive trafficking of reovirus particles through the endosomal pathway during the early stages of reovirus infection.
In this study, we analyzed the role of WDR81 in reovirus-induced programmed cell death. We demonstrate that even though WDR81 is dispensable for infection by reovirus ISVPs, it is required for ISVP-induced apoptosis. We show that in the absence of WDR81, other death agonists also fail to trigger apoptosis. Our data show that the absence of WDR81 results in increased expression of pro-survival genes. We demonstrate that blocking the IKK-NFκB pathway normalizes the expression of these pro-survival genes and restores the capacity of cells to succumb to apoptosis inducers. Together, our study describes a novel function of WDR81 in apoptosis. Furthermore, it reveals a previously unknown connection between WDR81 and the IKK-NFκB signaling pathway.
## RESULTS
## WDR81 is required for reovirus-induced programmed cell death
WDR81, a host protein critical for early-to-late endosome maturation, plays a key role in cellular trafficking (24). Previous work demonstrated that WDR81 is essential for infection by reovirus virions but dispensable for infection by ISVPs, an entry inter mediate generated via in vitro chymotrypsin digestion of virions (23). To confirm this phenotype, control cells (WT mouse embryo fibroblast [MEFs]) and cells deficient in WDR81 (ΔWDR81 MEFs) were infected with reovirus virions or ISVPs. Virus infectivity was measured 24 h post-infection by indirect immunofluorescence. Consistent with previous work (23), reovirus virions showed lower infectivity in ΔWDR81 MEFs compared to WT MEFs, while infection with reovirus ISVPs showed no difference in infectivity between WT and ΔWDR81 MEFs (Fig. 1A). Virus production was also quantified by measuring viral titer 24 h post-infection by plaque assay. Consistent with previous results and indirect immunofluorescence, in comparison to WT MEFs, reovirus virions produced ~0.7 log 10 lower infectious virus in the ΔWDR81 MEFs, while infection with reovirus ISVPs showed no difference in the amount of infectious titer produced in both cell types (Fig. 1B).
To study the role of WDR81 in reovirus-induced programmed cell death, WT MEFs and ΔWDR81 MEFs were infected with reovirus virions or ISVPs and incubated for 48 h. Permeability of host cell membranes to Sytox Green nucleic acid dye, which only stains nuclei of dead or dying cells (25), was evaluated as a measure of cell death. This later time point was chosen as MEFs infected with reovirus do not show detectable cell death at 24 h post-infection (data not shown). When infected with virions, in comparison to WT MEFs, ΔWDR81 MEFs showed significantly fewer dead cells (Fig. 1C). This result was expected as reovirus virions are unable to establish infection or replicate in ΔWDR81 MEFs (Fig. 1A andB). Interestingly, upon infection with reovirus ISVPs, ΔWDR81 MEFs still showed a significantly lower number of Sytox Green-positive cells in comparison to the WT MEFs (Fig. 1C). This result was unexpected since reovirus ISVPs are capable of successful replication in ΔWDR81 MEFs, as shown in Fig. 1A andB. To confirm the in vitro generation of ISVPs from purified virions, virions and ISVPs were resolved on an SDS-PAGE gel (Fig. 1D). The cleavage of µ1C protein to δ and the loss of σ3 protein indicate the successful generation of ISVPs. Together, these results suggest that WDR81 plays a role in reovirus-induced programmed cell death. Because ISVPs efficiently infected ΔWDR81 MEFs, for the remainder of this study, we used ISVPs as a tool to dissect the role of WDR81 in cell death induction.
## WDR81 is required for reovirus-induced caspase-3/7 activity
Reovirus induces cell death either through apoptosis, involving both the intrinsic and the extrinsic pathways, or necroptosis, mediated by receptor-interacting protein kinase 3 (RIPK3) and mixed lineage kinase domain like pseudokinase (MLKL) activation (9,26). Reovirus infection of MEFs produces morphological and biochemical changes in cells that resemble apoptosis (22). To determine whether reovirus infection induces apoptosis in MEFs, we assessed the effect of ZVAD-fmk (27), a broad-spectrum caspase inhibitor, on cell death 48 h post-infection. Consistent with data shown in Fig. 1, a minimal difference was observed between infection of WT and ΔWDR81 cells with ISVPs. ZVAD-fmk treatment did not reduce ISVP infectivity in either cell type to a level that affects infectious virus output (Fig. 2A and not shown). We then assessed the ability of ZVAD-fmk to inhibit cell death induced by ISVPs and found that treatment with ZVADfmk significantly reduced the activation of cell death in WT MEFs (Fig. 2B). These data suggest that MEFs undergo cell death by apoptosis, following infection with reovirus ISVPs. Because ΔWDR81 cells fail to undergo cell death following infection by ISVPs, our results suggest that WDR81 is required for efficient induction of apoptosis. A hallmark of apoptosis is the activation of effector caspases-caspase-3 and caspase-7. To determine whether effector caspases are activated, WT and ΔWDR81 MEFs were infected with ISVPs. The activity of effector caspases in cell lysates at 48 h post-infection was measured using a caspase-3/7 substrate which shows enhanced luminescence upon cleavage. ISVP infection of WT MEFs resulted in a ~2.5-fold increase in caspase-3/7 activity, compared to only a ~1.5-fold increase in ΔWDR81 MEFs (Fig. 2C). Caspase-3/7 activation was also quantified using a cell-permeable substrate that fluoresces upon cleavage by active caspases. Upon infection of WT MEFs with reovirus ISVPs, we detected a large number of cells positive for caspase-3/7 activity. In contrast, ΔWDR81 MEFs exhibited markedly fewer cells with active caspase-3/7 (Fig. 2D). These data suggested that WDR81 is required for ISVP-induced caspase-3/7 activation. To confirm the effectiveness of ZVAD-fmk, we also assessed caspase activation by ISVPs in the presence of ZVAD-fmk. ZVAD-fmk was efficiently able to inhibit caspase-3/7 activation (Fig. 2C andD). Together, these data suggest that ISVPs trigger caspase-mediated apoptosis in MEFs and that WDR81 is required for this process.
## WDR81-deficient cells express genes that are also induced by TNF treatment
We hypothesized that the resistance of ΔWDR81 MEFs to ISVP-induced apoptosis might result from altered expression of genes regulating cell death and survival. Gene expres sion differences between WT and ΔWDR81 MEFs were compared by RNA-seq. RNA-seq analysis, using a false discovery rate (FDR) cutoff of <0.0001, revealed numerous genes differentially expressed between WT and ΔWDR81 MEFs (Fig. 3A). Ingenuity pathway analysis (28) of differential gene expression revealed that WDR81 deficiency alters gene expression across multiple cellular pathways, including-but not limited to-those governing pathogen responses and cell-based immunity (data not shown). The largescale, bidirectional changes in gene expression between WT and ΔWDR81 cells made it challenging to use these analyses to identify specific pathways driving the resistance of ΔWDR81 cells to ISVP-induced apoptosis. Many constitutively expressed host genes promote cell survival by repressing the activation of pathways of apoptosis (29). One reason for the reduced ability of ΔWDR81 MEFs to undergo apoptosis following ISVP infection could be because such pro-survival genes are expressed at a higher level in these cells at a basal level. In most cells, including MEFs, pro-survival genes can be upregulated by treatment with TNF (29). We therefore identified those genes whose expression is differentially regulated by TNF treatment (Fig. 3B). We identified 107 TNFresponsive genes in WT cells (using the same stringent FDR cutoff). We then asked if these genes were also differentially expressed in ΔWDR81 MEFs. Notably, 66 of 107 TNFinduced genes were differentially expressed at a basal level in untreated ΔWDR81 MEFs compared to untreated WT MEFs (Fig. 3C). To validate the differential gene expression analysis, transcript levels of A20 and Bcl2-two TNF-induced genes-were assessed by RT-qPCR (30,31). RT-qPCR analysis confirmed that ΔWDR81 MEFs exhibit ~10-fold higher A20 mRNA levels and ~2.5-fold higher Bcl2 mRNA levels compared to WT MEFs (Fig. 3D andE). These findings indicate that ΔWDR81 MEFs exhibit elevated basal expression of genes typically induced by TNF. We have previously reported that expression of human WDR81 in ΔWDR81 MEFs restores virus infection and virus-induced cell death (23). While these data indicated that the phenotype observed in these cells is related to the absence of WDR81 and not an off-target effect, we sought to confirm that the unexpected effect of WDR81 loss on the expression of pro-survival genes is also restored by re-expression of WDR81. Trans-complementation of human WDR81 in ΔWDR81 cells significantly reduced the levels of A20 mRNA compared to control ΔWDR81 cells (Fig. 3F). Together, these data suggest that in the absence of WDR81, cells have higher levels of pro-survival genes that are typically upregulated by TNF.
## Pro-survival gene expression in ΔWDR81 MEFs depends on IKK activity
After engaging its receptor on the cell surface, TNF signals via the IKK complex to activate NFκB and drive expression of its genomic targets (29). Given the similarity between TNF-induced genes and those upregulated in ΔWDR81 MEFs, we hypothesized that IKK signaling may regulate pro-survival gene expression observed in WDR81-deficient cells. Treatment with an IKK inhibitor significantly reduced A20 mRNA levels in ΔWDR81 MEFs (Fig. 4A). These data suggest that at least some of the genes that are differentially expressed in the absence of WDR81 are controlled by IKK signaling. WDR81 negatively regulates endosomal type III PI3K, which generates phosphatidylinositol 3-phosphates (PtdIns3P) involved in endosomal signaling (24). Consequently, ΔWDR81 cells are expected to exhibit elevated PI3K activity. To assess if higher PI3K activity plays a role in higher pro-survival gene expression, WT and ΔWDR81 MEFs were treated with broad-spectrum PI3K inhibitor LY294002 (32). Treatment of ΔWDR81 MEFs with LY294002 in gene expression is included. (D) MEFs were grown in a 24-well plate, and total mRNA was extracted using Bio-Rad Aurum Total RNA mini kit according to manufacturer's protocol. cDNA was synthesized using Applied Biosystems High-Capacity cDNA Reverse Transcription kit. qPCR for A20 gene expression was done using Applied Biosystems StepOnePlus Real-Time PCR system. A20 gene expression was calculated using the 2 -ΔΔCT method using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as internal control. Error bars represent SD. ***, P < 0.0005 by Student's t-test. (E) MEFs were grown in a 24-well plate, and total mRNA was extracted using Bio-Rad Aurum Total RNA mini kit according to manufacturer's protocol. cDNA was synthesized using Applied Biosystems High-Capacity cDNA Reverse Transcription kit. qPCR for Bcl2 gene expression was done using Applied Biosystems StepOnePlus Real-Time PCR system. Bcl2 gene expression was calculated using the 2 -ΔΔCT method using GAPDH as internal control. Error bars represent SD. *, P < 0.05 by Student's t-test. (F) MEFs were grown in standard media as described in Materials and Methods. Total mRNA was extracted using Bio-Rad Aurum Total RNA mini kit according to manufacturer's protocol. cDNA was synthesized using Applied Biosystems High-Capacity cDNA Reverse Transcription kit. qPCR for A20 gene expression was done using Applied Biosystems
StepOnePlus Real-Time PCR system. A20 gene expression was calculated using the 2 -ΔΔCT method using GAPDH as internal control. Error bars represent SD. ***, P < 0.0005, **, P < 0.005, by one-way analysis of variance. at functional concentrations (data not shown) did not reduce A20 mRNA levels (Fig. 4B), indicating that elevated PI3K activity does not contribute to increased pro-survival gene expression in the absence of WDR81. These findings together indicate that signaling via IKK but not PI3K drives the upregulation of certain pro-survival genes in WDR81-deficient cells.
## WDR81 controls the activation of the multiple pathways of apoptosis
TNF signaling via IKKs drives the expression of NFκB-dependent genes that enhance cell survival (29). We hypothesized that the upregulated pro-survival genes in ΔWDR81 cells may confer resistance to apoptosis induced by other triggers. ABT-737 inhibits the Bcl2 family of antiapoptotic proteins (33). Since the Bcl2 family of proteins typically inhibits the activation of intrinsic apoptotic pathways, ABT-737 treatment results in mitochon drial cytochrome c release, caspase-9 activation, and cell death. To test if WDR81 is required for apoptosis via the intrinsic pathway, we treated WT and ΔWDR81 MEFs with ABT-737 for 24 h. Treatment with ABT-737 resulted in comparable levels of cell death (Fig. 5A) and caspase-3/7 activation (Fig. 5B) in both WT and ΔWDR81 MEFs. These results indicate that in MEFs, WDR81 is dispensable for apoptosis induced by the intrinsic pathway. We next tested if WDR81 is required for extrinsic trigger-induced apoptosis. To evaluate this, we used TNF and cycloheximide (CHX) to induce death in WT and ΔWDR81 MEFs. This combination is a well-characterized inducer of death receptor-mediated, extrinsic apoptosis (34). While neither TNF nor CHX alone altered cell survival (data not shown), upon treatment with the combination for 24 h, WT MEFs succumbed to cell death. In contrast, ΔWDR81 MEFs showed significantly lower numbers of dead cells (Fig. 5C). Similarly, TNF-CHX treatment induced caspase-3/7 activity in WT MEFs, whereas it failed to do so in ΔWDR81 MEFs (Fig. 5D). Thus, in MEFs, WDR81 is required for apoptosis induced by extrinsic triggers.
Since IKK and NFκB drive expression of pro-survival genes in ΔWDR81 cells, we assessed the impact of IKK inhibition on TNF-CHX-induced cell death. We found that treatment with the IKK inhibitor restored TNF-CHX-induced cell death and caspase-3/7 activation in ΔWDR81 MEFs to levels comparable to WT MEFs (Fig. 5C andD). Thus, the resistance to extrinsic agonist-triggered apoptosis in ΔWDR81 cells is driven by elevated IKK activity. Treatment with the PI3K inhibitor LY294002 to reduce elevated PI3K signaling in ΔWDR81 cells did not restore TNF-CHX-induced cell death or caspase-3/7 activation in ΔWDR81 MEFs (Fig. 5C andD), even under conditions where the inhibitor was confirmed FIG 4 Higher A20 mRNA levels in WDR81-deficient cells are driven by IKK but not PI3K signaling. (A) MEFs were incubated in media with either DMSO control, TNF (50 ng/mL), or IKK inhibitor (5 µM) for 3 h. Total mRNA was extracted using Bio-Rad Aurum Total RNA mini kit according to manufacturer's protocol. cDNA was synthesized using Applied Biosystems High-Capacity cDNA Reverse Transcription kit. qPCR for A20 gene expression was done using Applied Biosystems
StepOnePlus Real-Time PCR system. A20 gene expression was calculated using the 2 -ΔΔCT method using GAPDH as internal control. Error bars represent SD. ***, P < 0.0005 by one-way analysis of variance (ANOVA). (B) MEFs were incubated in media with DMSO control or PI3K inhibitor (25 µM) for 3 h. Total mRNA was extracted using Bio-Rad Aurum Total RNA mini kit according to manufacturer's protocol. cDNA was synthesized using Applied Biosystems High-Capacity cDNA Reverse Transcription kit. qPCR for A20 gene expression was done using Applied Biosystems StepOnePlus Real-Time PCR system. A20 gene expression was calculated using the 2 -ΔΔCT method using GAPDH as internal control. Error bars represent SD. **, P < 0.005, *, P < 0.05 by one-way ANOVA.
to be effective (data not shown). These results indicate that elevated PI3K activity does not contribute to the resistance of ΔWDR81 MEFs to extrinsic apoptosis. Furthermore, our results suggest that the elevated PI3K activity in the absence of WDR81 does not appear to impact IKK activity.
To assess the impact of WDR81 in other cell types, we compared cell death induc tion by ABT-737 and TNF-CHX in control and WDR81-deficient HT1080 cells, a human fibrosarcoma cell line. While control HT1080 cells succumbed to ABT-737 at 48 h post-treatment, ΔWDR81 HT1080 cells were resistant to cell death induced by this drug in this time frame (Fig. 5E). Similarly, whereas control HT1080 cells undergo cell death after TNF-CHX treatment, ΔWDR81 HT1080 cells remain viable (Fig. 5F). These data indicate that in HT1080 cells, WDR81 controls apoptosis by both the intrinsic and the extrinsic pathways. Collectively, our evidence indicates that WDR81 is required for efficient cell death in multiple cell types.
## WDR81-deficient cells support ISVP-induced apoptosis when IKK activity is blocked
Apoptosis induction by reovirus requires death receptor signaling via TRAIL (17). Additionally, for efficient cell death induction in MEFs, the death signals need to be amplified by tBid-mediated activation of the intrinsic apoptotic pathway (15). Given that IKK inhibition restores cell death in TNF-CHX-treated cells, we asked if it would also restore cell death in ISVP-infected cells. For these experiments, WT and ΔWDR81 MEFs were infected with reovirus ISVPs and subsequently incubated in the presence of IKK inhibitor. Infectivity measurements using indirect immunofluorescence revealed no significant differences in infection levels between dimethyl sulfoxide (DMSO)-trea ted and IKK inhibitor-treated cells (Fig. 6A), confirming that the IKK inhibitor does not impact viral replication. We measured cell death and caspase-3/7 activation by ISVPs 48 h post-infection in the presence of an IKK inhibitor. In control, DMSO-treated cells, ISVP infection induced a substantial level of apoptotic death in WT MEFs, while ΔWDR81 MEFs showed minimal cell death, consistent with our earlier observations (Fig. 6B). Although IKK inhibition had no effect on ISVP-induced apoptosis in WT MEFs, it fully restored apoptosis in ΔWDR81 MEFs to levels comparable to WT MEFs (Fig. 6B). Similarly, in the presence of IKK inhibitor, ISVPs induced caspase-3/7 activation in an equivalent number of cells to WT MEFs (Fig. 6C). Together, these results demonstrate that IKK-driven pro-survival gene expression in ΔWDR81 MEFs prevents ISVP-induced caspase-3/7 activation and apoptosis.
## DISCUSSION
In this study, we uncover a novel role of the protein WDR81 in regulating the apoptosis pathway in cell culture. We show that WDR81 is required for the induction of apopto sis. In the absence of WDR81, cells infected with reovirus or treated with inducers of apoptosis fail to activate caspase-3/7 and succumb to cell death (Fig. 1 and5). These data suggest that WDR81 broadly regulates apoptosis. In the absence of WDR81, although caspase-3/7 can be activated (Fig. 5), their activation is impaired due to the upregulation of pro-survival signals (Fig. 4). Our data indicate that a large majority of these upregula ted pro-survival signals are those that are regulated by the IKK-NFκB pathway (Fig. 3 to 6). Inhibition of the IKK pathway is sufficient to overcome WDR81-dependent deficiencies in caspase-3/7 activation and death (Fig. 5 and6). Based on these data, we propose a model (Fig. 7) where WDR81 is responsible for suppressing the expression of IKK-NFκBdriven pro-survival genes. Our data suggests that in the absence of WDR81, upregulated pro-survival genes prevent efficient induction of apoptosis. This work highlights a novel association between WDR81 and the IKK-NFκB-mediated cell survival pathway.
WDR81 is a cellular protein that inhibits class III PI3Ks on the early endosomal membrane (24). This function regulates the efficient conversion of early to late endo somes and thus cellular endosomal trafficking. Cells lacking WDR81 have delayed earlyto-late endosome conversion and lysosomal turnover of plasma membrane proteins such as epidermal growth factor receptor is impaired (24). Since PI3K inhibitors were not able to rescue cell death in ΔWDR81 MEFs, we think that the cell survival controlling function of WDR81 is independent of its role as a PI3K inhibitor. WDR81 can interact with p62 and LC3C promoting autophagy (35). In the absence of WDR81, there is an increase in protein aggregation and reduced autophagic clearance (35). It is possible that in the absence of WDR81, there is decreased autophagic turnover of an activator of the IKK-NFκB pathway. While we have shown the role of WDR81 in regulating cell death in two cell types (MEFs and HT1080 cells), it is possible that in some cell types, the role of WDR81 in regulating death is absent or even reversed. Mutations in WDR81 have been associated with increased cell death in Purkinje cells of mice, resulting in improper motor abilities (36).
Previous work has shown that NFκB signaling mediated through the IKK complexes is required for reovirus-induced death (37,38). These data were obtained using cells deficient in IKK or NFκB subunits. Reovirus affects NFκB in two phases. In the first phase, infection triggers NFκB activation, and in the second stage, NFκB activity is inhibited (39). It is thought that NFκB inhibition later in infection decreases the concentration of survival factors in the cell, thereby sensitizing them to cell death (39). Since WDR81deficient cells have constitutively high IKK-NFκB activity, they allow us to explore the biological effect of not being able to effectively block NFκB later in infection. Our work demonstrating that WDR81-deficient cells are resistant to apoptosis supports the idea that the second stage inhibition of NFκB is required for cell death. It is possible NFκB inhibition later in infection occurs in a cell type-specific manner. Such a possibility may account for the organ-specific role of NFκB in reovirus pathogenesis (40).
Reovirus triggers death via TRAIL-mediated signaling. In reovirus-infected cells, TRAIL signaling promotes caspase-8 activation (17). But caspase-8 activation is not sufficient to induce apoptosis. Apoptosis induction after reovirus infection requires caspase-8mediated cleavage of Bid (15). Cleaved Bid (tBid) engages the mitochondrial apoptotic pathway to induce sufficient effector caspase activation to induce cell death (41). TNF-CHX shares some similarities with cell death signaling with reovirus since it also induces caspase-8 activation. Following TNF-CHX treatment, whether amplification of death signals via the intrinsic mitochondrial apoptotic pathway is needed for cell death may vary with cell type (42). Our results also indicate that WDR81-deficient MEFs are capable of undergoing cell death following direct activation of the intrinsic pathway through inhibition of the Bcl2 family of anti-apoptotic proteins. Thus, in MEFs, the effect of WDR81 on cell death appears to be restricted to events in death receptor signaling that lead to caspase-8 activation. In contrast, we found that in HT1080 cells, WDR81 is required for the activation of both the intrinsic and the extrinsic apoptotic pathways (Fig. 5E andF). Consistent with other work (29,43), our RNA-seq data indicate that TNF drives the expression of multiple pro-survival genes. Many of these genes are also expressed at higher levels in cells lacking WDR81. It is possible that the different impact of WDR81 deficiency on the intrinsic apoptotic pathway in MEFs and HT1080 cells relates to differences in the set of pro-survival genes expressed in these two cell types or the extent to which they are expressed. It is also possible that the threshold of pro-survival gene expression needed to block cell death in each of these cell types is different. Currently, it remains unknown which pro-survival genes impact cell death induction during infection with reovirus, TNF, or Bcl2 inhibitor treatment. In some cases, a single target, A20, is sufficient to protect cells from TNF-induced cell death (29). While our data demonstrate that A20 is expressed at a higher level in the absence of WDR81, whether the enhanced expression of A20 alone is sufficient to prevent activation of reovirus-or TNF-induced extrinsic apoptotic pathways in WDR81-deficient cells remains unknown. Whether a single pro-survival gene can prevent the induction of the intrinsic apoptotic pathway also is unidentified. These studies remain a subject of our investigation.
## MATERIALS AND METHODS
## Cells and viruses
Murine L929 cells (spinner cells) were grown at 37°C in Joklik's minimal essential medium (Lonza) supplemented with 5% fetal bovine serum (FBS) (Life Technologies), 2 mM L-glutamine (Invitrogen), 0.5 U/mL penicillin, 50 µg/mL streptomycin (Sigma Aldrich), and 25 ng/mL amphotericin B (Sigma-Aldrich). MEFs containing non-targeting plasmid (WT MEFs) and cells containing lentiCRISPRv2 vector targeting WDR81 (ΔWDR81 MEFs) were maintained at 37°C in Dulbecco's modified Eagle medium (DMEM) (Gibco) supplemented with 10% FBS, 2 mM L-glutamine, and 2 µg/mL puromycin (Invivogen). WT MEFs transduced with vector (WT-vec), ΔWDR81 MEFs transduced with vector (ΔWDR81-vec), and ΔWDR81 MEFs transduced with human WDR81 (ΔWDR81-hWDR81) were maintained at 37°C in DMEM (Gibco) supplemented with 10% FBS, 2 mM L-glu tamine, 2 µg/mL puromycin (Invivogen), and 8 µg/mL blasticidin (Invivogen). ACE2expressing HT1080 cells (WT HT1080) and their WDR81-deficient counterparts (ΔWDR81 HT1080), a gift from Dr. Paul Bieniasz, Rockefeller University, were maintained at 37°C in DMEM (Gibco) supplemented with 10% FBS, 2 mM L-glutamine, and 1.25 µg/mL puromycin (Invivogen) (44). All reovirus experiments were performed with Type 3 Dearing from the Cashdollar laboratory (T3D CD ), a stock of which was obtained from Dr. John Parker.
## Reovirus purification
Reovirus T3D CD was propagated in spinner cells. Spinner cells infected with a second passage of reovirus stock were lysed by sonication. Virions were extracted from lysates using Vertrel-XF specialty fluid (Dupont). The extracted particles were layered onto 1.2 to 1.4 g/cm 3 CsCl step gradients. Gradients were centrifuged at 187,000 × g for 4 h at 4°C. Bands corresponding to purified virions (1.36 g/cm 3 ) were isolated and dialyzed into virus storage buffer (10 mM Tris [pH 7.4], 15 mM MgCl 2 , and 150 mM NaCl). After dialysis, the particle concentration was determined by measuring the optical density at 260 nm (OD 260 ) of the purified virion stocks (one unit at OD 260 = 2.1 × 10 12 particles/mL). The purification of virions was confirmed by SDS-PAGE (45).
## Generation of reovirus ISVPs
Purified T3D CD virions (2 × 10 11 particles/mL) were digested with 200 µg/mL Na-p-tosyl-Llysine chloromethyl ketone-treated chymotrypsin (Worthington Biochemical) in a total volume of 100 µL for 30 min at 32°C. The reactions were then quenched by adding 1 mM phenylmethylsulfonyl fluoride (Sigma-Aldrich). ISVP generation was confirmed by SDS-PAGE (46).
## Reovirus infectivity by indirect immunofluorescence
MEFs were seeded in 96-black-well plates (Greiner) at 2 × 10 4 cells/well. Twenty-four hours after seeding, cells were adsorbed with T3D CD virions or ISVPs at an MOI of 10 PFU/cell for 1 h at room temperature (RT) using phosphate buffered saline (PBS) for mock infections. Inoculum was then replaced with media and incubated at 37°C in CO 2 incubator. Twenty-four hours post-infection, cells were washed with PBS and fixed with 100 µL methanol for 30 min at -20°C. Cells were then washed with 0.5% PBS-Tween 20 (PBS-T) and blocked using 1% PBS-bovine serum albumin (PBS-BSA). Subsequently, cells were stained for 1 h with reovirus polyclonal serum (1:1,000 in 1% PBS-BSA). Cells were then washed with PBS-T three times before staining for 1 h with LI-COR donkey anti-rabbit IRDye-800 CW (1:1,000 in 1% PBS-BSA) and Draq-5 (1:10,000 in 1% PBS-BSA). Cells were then washed three times with PBS-T and 50 µL water added to each well. The plate was imaged using the LI-COR Odyssey CLX Scanner to capture infrared (IR) images at 800 nm (reovirus) and 700 nm (cells). Infectivity was measured by signal at 800 nm (reovirus) normalized to signal at 700 nm (cells).
## Single-step reovirus growth assay
MEFs were seeded in 24-well plates (Greiner) at 1 × 10 5 cells/well. Twenty-four hours after seeding, cells were adsorbed with T3D CD virions or ISVPs at an MOI of 10 PFU/cell for 1 h at RT. Inoculum was then replaced with media as described above and incubated at 37°C in a CO 2 incubator. Twenty-four hours post-infection, cells were harvested by 3× freeze-thaw cycle, and virus titer was measured by plaque assay on spinner cells. Briefly, confluent six-well plates of spinner cells were adsorbed with 100 µL of 10-fold serially diluted, freeze-thawed lysate for 1 h. Cells were then overlaid with Media 199 supple mented with 2 mM L-glutamine (Invitrogen), 0.5 U/mL penicillin, 50 µg/mL streptomycin (Sigma Aldrich), and 25 ng/mL amphotericin B (Sigma-Aldrich), and 1% Difco Bacto Agar (BD). Plates were incubated until countable plaques were visible and fixed with 3.7% formaldehyde in PBS. Fixed plates were then stained with 1% crystal violet in 5% ethanol solution to visualize and count plaques.
## Cell death assays
MEFs were seeded in 96-black-well plates (Greiner) at 2 × 10 4 cells/well. Twenty-four hours after seeding, cells were adsorbed with T3D CD virions or ISVPs at MOI of 10 PFU/ cell for 1 h at RT using PBS for mock infections. Inoculum was replaced with media as described above and supplemented with 50 nM Sytox Green nucleic acid dye (Invitro gen). Forty-eight hours post-infection, plates were imaged by fluorescence microscopy using IncuCyte (Essen Biosciences). Sytox Green-positive cells per field were quantified using the associated software.
## Caspase-3/7 assays
For fluorescence microscopic analysis of caspase-3/7, MEFs were seeded in 96-blackwell plates (Greiner) at 2 × 10 4 cells/well. Twenty-four hours after seeding, cells were adsorbed with T3D CD ISVPs at MOI of 10 PFU/cell for 1 h at RT using PBS for mock infections. Inoculum was replaced with media as described above, supplemented with 2.5 µM IncuCyte Caspase-3/7 Green (Sartorius). Forty-eight hours post-infection, plates were imaged by fluorescence microscopy using IncuCyte S3 imager (Essen Biosciences). Caspase-3/7-positive cells per field were quantified using the associated software.
For luminescence-based relative caspase-3/7 assay, MEFs were seeded in 96-blackwell plates (Greiner) at 2 × 10 4 cells/well. Twenty-four hours after seeding, cells were adsorbed with T3D CD ISVPs at an MOI of 10 PFU/cell for 1 h at RT using PBS for mock infections. Inoculum was replaced with media as described above. Forty-eight hours post-infection, caspase-3/7 assay was measured using Promega Caspase-3/7 Glo kit according to manufacturer's instructions. Luminescence was measured using Biotek Synergy LX multimode reader. Relative caspase-3/7 activity was calculated by normaliz ing luminescence signal of sample to that of mock, which was set to 1.
## Analysis of host gene expression by RT-qPCR
Total RNA was extracted from cells using Bio-Rad Aurum Total RNA mini kit according to manufacturer's instructions. Eluted RNA was converted into cDNA using Thermo Fisher High-Capacity cDNA Reverse Transcription Kit using random hexamers. A20 and Bcl2 gene expression was estimated by qPCR using Applied Biosystems StepOne Real-Time PCR machine. Relative mRNA levels were calculated using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as internal control using the 2 -ΔΔCT method (47). Primers used for qPCR are as follows: GAPDH, forward ACCCAGAAGACTGTGGATGG and reverse GGATGCA GGGATGATGTTCT; A20, forward CTGGATGTCAATCAACAATGGGA and reverse ACTAGGGT GTGAGTGTTTTCTGT; Bcl2, forward ATGCCTTTGTGGAACTATATGGC and reverse GGTATGC ACCCAGAGTGATGC.
## Library preparation for RNA-seq and gene expression analysis
Total RNA was extracted from cells using Bio-Rad Aurum Total RNA mini kit according to manufacturer's instructions. RNA was submitted to Indiana University Center for Genomics and Bioinformatics for RNA-seq processing. Briefly, cDNA library construction was done using a TruSeq Stranded mRNA Low-Throughput Sample Prep Kit (Illumina) following the manufacturer's protocol. Sequencing was performed using an Illumina NextSeq 500 platform with a 75 bp sequencing module, generating 38 bp pairedend reads. After the sequencing run, demultiplexing was performed with bcl2fastq v.2.20.0.422.
The sequenced reads were adapter-trimmed and quality-filtered using Trimmomatic v.0.38 (48), setting the cutoff threshold for average base quality score at 20 over a window of 3 bases, with a minimum read length of 20 bases after trimming (parame ters: LEADING:20 TRAILING:20 SLIDINGWINDOW:3:20 MINLEN:20). Cleaned reads were mapped to the mouse genome sequence GRCm39 using STAR RNA-seq aligner v.2.7.11a (49) using default parameters. Concordantly mapped read pairs aligning to the exon regions of annotated genes on the sense strand (Ensembl release 109) were counted using the featureCounts tool v.2.0.0 (50) of subread package (parameters: -t exon -g gene_id -s 2 -p -B -C). The differential expression analysis was conducted using DESeq2 version 1.44.0 (51).
## Chemicals and reagents
Pan-caspase inhibitor ZVAD-fmk (Cayman Chemicals) was used at a final concentration of 20 µM (26). IKK inhibitor BAY-65-1942 (Bayer) was used at a final concentration of 5 nM (26). PI3K inhibitor LY294002 (Cayman Chemicals) was used at a final concentration of 25 nM. Human tissue necrotic factor alpha (TNF) (Sigma-Aldrich) was used at a final concentration of 50 ng/mL. Cycloheximide was used at a final concentration of 10 µg/mL. Bcl2 inhibitor ABT-737 (Cayman Chemicals) was used at a final concentration of 5 µM.
## References
1. Taylor, Cullen, Martin (2008) "Apoptosis: controlled demolition at the cellular level" *Nat Rev Mol Cell Biol*
2. Bertheloot, Latz, Franklin (2021) "Necroptosis, pyroptosis and apoptosis: an intricate game of cell death" *Cell Mol Immunol*
3. Liu, Zhang, Joo et al. (2017) "NF-κB signaling in inflammation" *Signal Transduct Target Ther*
4. Luo, Kamata (2005) "IKK/NF-κB signaling: balancing life and death -a new approach to cancer therapy" *J Clin Invest*
5. Verburg, Lelievre, Westerveld et al. (2022) "Viral-mediated activation and inhibition of programmed cell death" *PLoS Pathog*
6. Zhou, Jiang, Liu et al. (2017) "Virus infection and death receptor-mediated apoptosis" *Viruses*
7. Hardwick (2001) "Apoptosis in viral pathogenesis" *Cell Death Differ*
8. Danthi (2016) "Viruses and the diversity of cell death" *Annu Rev Virol*
9. Connolly, Dermody (2002) "Virion disassembly is required for apoptosis induced by reovirus" *J Virol*
10. Coffey, Sheh, Kim et al. (2006) "Reovirus outer capsid protein μ1 induces apoptosis and associates with lipid droplets, endoplasmic reticulum, and mitochondria" *J Virol*
11. Kominsky, Bickel, Tyler (2002) "Reovirus-induced apoptosis requires mitochondrial release of Smac/DIABLO and involves reduction of cellular inhibitor of apoptosis protein levels" *J Virol*
12. Oberhaus, Smith, Clayton et al. (1997) "Reovirus infection and tissue injury in the mouse central nervous system are associated with apoptosis" *J Virol*
13. Brown, Short, Stencel-Baerenwald et al. (2018) "Reovirus-induced apoptosis in the intestine limits establishment of enteric infection" *J Virol*
14. Debiasi, Robinson, Sherry et al. (2004) "Caspase inhibition protects against reovirus-induced myocardial injury in vitro and in vivo" *J Virol*
15. Danthi, Pruijssers, Berger et al. (2010) "Bid regulates the pathogenesis of neurotropic reovirus" *PLoS Pathog*
16. Kominsky, Bickel, Tyler (2002) "Reovirus-induced apoptosis requires both death receptor-and mitochondrial-mediated caspasedependent pathways of cell death" *Cell Death Differ*
17. Clarke, Meintzer, Gibson et al. (2000) "Reovirus-induced apoptosis is mediated by TRAIL" *J Virol*
18. Wisniewski, Werner, Hom et al. (2011) "Reovirus infection or ectopic expression of outer capsid protein μ1 induces apoptosis independently of the cellular proapoptotic proteins Bax and Bak" *J Virol*
19. Danthi, Hansberger, Campbell et al. (2006) "JAM-A-independent, antibody-mediated uptake of reovirus into cells leads to apoptosis" *J Virol*
20. Danthi, Coffey, Parker et al. (2008) "Independ ent regulation of reovirus membrane penetration and apoptosis by the μ1 φ domain" *PLoS Pathog*
21. Knowlton, Dermody, Holm (2012) "Apoptosis induced by mammalian reovirus is beta interferon (IFN) independent and enhanced by IFN regulatory factor 3-and NF-κB-dependent expression of Noxa" *J Virol*
22. Tyler, Squier, Rodgers et al. (1995) "Differences in the capacity of reovirus strains to induce apoptosis are determined by the viral attachment protein sigma 1" *J Virol*
23. Snyder, Abad, Danthi (2022) "A CRISPR-Cas9 screen reveals a role for WD repeat-containing protein 81 (WDR81) in the entry of late penetrating viruses" *PLoS Pathog*
24. Liu, Jian, Sun et al. (2016) "Negative regulation of phosphatidylinositol 3phosphate levels in early-to-late endosome conversion" *J Cell Biol*
25. Berger, Hiller, Thete et al. (2017) "Viral RNA at two stages of reovirus infection is required for the induction of necroptosis" *J Virol*
26. Berger, Danthi (2013) "Reovirus activates a caspase-independent cell death pathway" *mBio*
27. Lotem, Sachs (1996) "Differential suppression by protease inhibitors and cytokines of apoptosis induced by wild-type p53 and cytotoxic agents" *Proc Natl Acad Sci*
28. Krämer, Green, Pollard et al. (2014) "Causal analysis approaches in Ingenuity Pathway Analysis" *Bioinformatics*
29. Oeckinghaus, Ghosh (2009) "The NF-κB family of transcription factors and its regulation" *Cold Spring Harb Perspect Biol*
30. Martens, Van Loo (2020) "A20 at the crossroads of cell death, inflammation, and autoimmunity" *Cold Spring Harb Perspect Biol*
31. Genestier, Bonnefoy-Berard, Rouault et al. (1995) "Tumor necrosis factor-α up-regulates Bcl-2 expression and decreases calcium-dependent apoptosis in human B cell lines" *Int Immunol*
32. Vlahos, Matter, Hui et al. (1994) "A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1benzopyran-4-one (LY294002)" *J Biol Chem*
33. Oltersdorf, Elmore, Shoemaker et al. (2005) "An inhibitor of Bcl-2 family proteins induces regression of solid tumours" *Nature*
34. Nio, Zighelboim, Berek et al. (1990) "Cycloheximide-induced modulation of TNF-mediated cytotoxicity in sensitive and resistant ovarian tumor cells" *Cancer Chemother Pharmacol*
35. Liu, Li, Wang et al. (2017) "The BEACH-containing protein WDR81 coordinates p62 and LC3C to promote aggrephagy" *J Cell Biol*
36. Traka, Millen, Collins et al. (2013) "WDR81 is necessary for purkinje and photoreceptor cell survival" *J Neurosci*
37. Connolly, Rodgers, Clarke et al. (2000) "Reovirus-induced apoptosis requires activation OF transcrip tion factor NF-κB" *J Virol*
38. Hansberger, Campbell, Danthi et al. (2007) "IκB kinase subunits α and γ are required for activation of NF-κB and induction of apoptosis by mammalian reovirus" *J Virol*
39. Clarke, Meintzer, Moffitt et al. (2003) "Two distinct phases of virus-induced nuclear factor κB regulation enhance tumor necrosis factor-related apoptosis-inducing ligand-mediated apoptosis in virusinfected cells" *J Biol Chem*
40. O'donnell, Hansberger, Connolly et al. (2005) "Organ-specific roles for transcription factor NF-κB in reovirus-induced apoptosis and disease" *J Clin Invest*
41. Gross, Yin, Wang et al. (1999) "Caspase cleaved BID targets mitochondria and is required for cytochrome c release, while BCL-X L prevents this release but not tumor necrosis factor-R1/Fas death" *J Biol Chem*
42. Galluzzi, Vitale, Aaronson et al. (2018) "Molecular mechanisms of cell death: recommendations of the Nomenclature Committee on Cell Death 2018" *Cell Death Differ*
43. Marques-Fernandez, Planells-Ferrer, Gozzelino et al. (2013) "TNFα induces survival through the FLIP-L-dependent activation of the MAPK/ERK pathway" *Cell Death Dis*
44. Poston, Weisblum, Hobbs et al. (2022) "VPS29 exerts opposing effects on endocytic viral entry" *mBio*
45. Berard, Coombs (2009) "Mammalian reoviruses: propagation, quantification, and storage" *Curr Protoc Microbiol*
46. Snyder, Danthi (2018) "Infectious subviral particle to membrane penetration active particle (ISVP-to-ISVP*) conversion assay for mammalian orthoreovirus" *Bio Protoc*
47. Schmittgen, Livak (2008) "Analyzing real-time PCR data by the comparative C T method" *Nat Protoc*
48. Bolger, Lohse, Usadel (2014) "Trimmomatic: a flexible trimmer for Illumina sequence data" *Bioinformatics*
49. Dobin, Davis, Schlesinger et al. (2013) "STAR: ultrafast universal RNA-seq aligner" *Bioinformatics*
50. Liao, Smyth, Shi (2014) "featureCounts: an efficient general purpose program for assigning sequence reads to genomic features" *Bioinformatics*
51. Love, Huber, Anders (2014) "Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2" *Genome Biol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12542352&blobtype=pdf | # First detection of echovirus 18 associated with aseptic meningitis in a child in Niger Republic, 2024: a case report
Adamou Lagare, Balkissa Abdoulaye, Fatou Thiaw, Hadiza Ousmane, Ndack Ndiaye, Balki Aoula, Fatimata Hassane, Santou Mamadou, Habibatou Amadou, Hanatou Hamidou, Mahamadou Doutchi, Martin Faye
## Abstract
Background Viral meningitis is an inflammation of the meninges that cause substantial morbidity and mortality in young children worldwide. Echovirus 18 (E18) has been recognized as an important causative pathogen of aseptic meningitis and recombinant strains have been associated with severe cases. Herein, we report the detection of E18 associated with aseptic meningitis in a child with a sickle cell anemia from Niger Republic in December 2024.Case presentation On November 19th, 2024, a 13-year-old boy returning from travel in a neighboring country was admitted at the National Reference Center for Sickle Cell Anemia (CNRD) in Niamey, following a severe anaemia combined with renal failure and hepatic dysfunction. Despite the efforts of medical care and compensatory treatment, he died six days after the onset of symptoms including a constantly high fever with persistent headaches and febrile convulsions, suggesting an aseptic meningitis. Further laboratory tests on blood sample using whole genome sequencing at Pasteur Institute of Dakar, confirmed the echovirus 18 virus as the etiological agent.
ConclusionThis is the first reported case of E18-associated with aseptic meningitis in Niger and the patient's death could be associated with the underlying co-morbidity of sickle cell anemia SS form. Our findings point the need for strengthening the surveillance of viral etiologies of meningitis in Niger. More studies are needed to predict and mitigate the re-emergence risk and gain a better insight into the prevalence and molecular epidemiology of E18 over time.
## Background
Meningitis is a devastating disease with a high case fatality rate due to the inflammation of the meninges [1]. It can affect people of all ages and young children, people with weak immunity, such as those undergoing chemotherapy, those with chronic illnesses, and elderly individuals, are high-risk groups [2]. Meningitis can be fatal and requires immediate medical care. According to the World Health Organization (WHO), it is caused by infections from viruses, bacteria, parasites or fungi [3]. Bacterial meningitis is of particular concern and can be fatal within 24 h. However, there are effective treatments and vaccines against some of the main bacterial causes of meningitis. Unlike, viral meningitis (VM) can be less severe and cause substantial morbidity and mortality in children [4,5]. The transmission of VM is from person to person through direct contact or droplets and can suddenly appear with fever, headache, chills, vomiting, lethargy, nausea, abdominalgia and confusion [5,6]. Nonpolio human enteroviruses (NPEVs), especially Echovirus 18 (E18) has mostly been reported as an important pathogen that has caused many cases and outbreaks of aseptic meningitis throughout the world [6]. Meanwhile, the virus exhibited a low worldwide genetic diversity with only three genogroups or genotypes [7]. Since its first identification in 1955, the virus was reported in Australia, the United States, Europe, Asia [2,5]. In Africa, E18 has been previously reported in Tunisia [8], Nigeria [9], Ethiopia [10], Cameroon (Genbank accession number: PQ889359) and Senegal [11] during surveillance of the circulating enteroviruses. So far, there has been no report of an E18-associated aseptic meningitis in Africa, and little is known on its epidemiology in the region.
Herein, we reported a case of E18-associated with aseptic meningitis in a child in the Republic of Niger in 2024.
## Case presentation
This was a retrospective cross-sectional study reporting an anonymous case for whom parent's oral consent was obtained. For instance, on November 19th, 2024, a 13-year-old boy with an unmonitored sickle cell anemia SS form, consulted at the National Reference Center for Sickle Cell Anemia (CNRD) in Niamey, Niger Republic, three days after the onset of symptoms including fever, persistent headaches and febrile convulsions. The patient returned from Zamfara State in the Federal Republic of Nigeria a month ago and declared a non-documented previous admission for gingivolabial hemorrhage a year ago. He started self-medication with analgesics at home one day before consultation, but saw a non-evolving health condition. Based on his clinical condition, the patient was admitted and sedated. At entry consultation, the patient had an axillary fever of 39 °C, a bodyweight of 25 kg and 95% of ambient oxygen saturation. First, sickle cell anemia test was carried out and confirm the presence of the hemoglobin SS. Prior to blood transfusion, blood typing, blood cell count and biochemistry were requested by the physicians. Preliminary clinical diagnosis showed that the patient belongs to the O + blood group and suffered with a severe anemia, renal failure and hepatic dysfunction (Table 1). On the same day, the patient was urgently blood-transfused. Considering the CRP test results antibiotics including Ceftriaxone (100 mg/kg per day in 1 to 2 60-minute IV injections) and Gentamicin (5 mg/kg once daily) were administered. However, the glycemia was normal. Following signs of shortness of breath, mechanical ventilation was also carried out.
In view of the persistent febrile convulsions, the physician first suspected for a probable bacterial meningitis or a cerebral malaria infection. The malaria test was negative for the four Plasmodium species and the CSF analysis showed colorless and hematic aspects at the macroscopic analysis, while the cytological analysis revealed a low number of white blood cells of <5/mm 3 and an elevated number of red blood cells of 1250/mm 3 . Thus, sweeping out a possible bacterial meningitis infection.
His vital signs deteriorated over-time with the appearance of epistaxis in addition to a constantly high fever (39-40 °C) and convulsions. Unfortunately, on November 21st, 2024, he died after two days of treatment.
## Advanced laboratory tests
Considering the occurrence of hemorrhagic signs before the patient's death, blood sample previously collected has Last, in a final effort to confirm the cause of this death, an aliquot of the serum was sent to the to the virology department at the Institut Pasteur de Dakar (IPD) on December 11th, 2024. Genomic sequencing technics using an enrichment library preparation method was carried out to determine the implicated pathogen [12]. The experiment was performed on the Illumina Iseq100 instrument and data were analyzed using the Chan Zuckerberg ID platform (https://czid.org/). Obtained sequences were analyzed using a Maximum-likelihood (ML) phylogenetic inference built with the IQ-TREE web-server for 1000 ultra-fast bootstrap replicates [13]. Similarity plot, bootscan method with the Simplot program (version 3.5.1) and RDP5 Beta (version 5.64) were used to identify potential recombination of the sequenced virus genome [14].
Interestingly, the sequencing analysis confirmed the presence of the Echovirus 18 virus with a genome coverage of 95%. The newly characterized E18 sequence from Niger belonged to the genotype B (Genbank accession number: PV679817) and was more closely related to an isolate identified from sewage in Nigeria in 2018 (Genbank accession number: MW373961). In addition, it clustered with 3 sequences associated with acute flaccid paralysis in India between 2008 and 2009 (Genbank accession numbers: JN203849, JN203850 and JN203852) and an isolate identified from human's stools in Ethiopia in 2018 (Genbank accession number: MF990301), underlying the risk of feco-oral transmission (Fig. 1). No evidence for recombination on the newly characterized E18 sequence from Niger was found.
## Discussion
Meningitis epidemics occur annually in Niger Republic due to its location within the African meningitis belt which extends from Ethiopia to Senegal [15]. However, since the introduction of the Men A conjugate vaccine against Neisseria meningitidis serogroup A in 2011 [16], and the subsequent vaccine campaign using the quadrivalent meningococcal (MenACWY) conjugate vaccine [17], the incidence of bacterial meningitis has substantially decreased. Viral meningitis is a significant yet often underreported public health concern in Sub-Saharan Africa, where diagnostic limitations and surveillance gaps hinder accurate case detection [18]. Human E18 is most commonly known to cause aseptic meningitis [19]. Interestingly, clinical signs identified from this patient were similar to those reported from cases of aseptic E18-associated meningitis in China in 2015 [5]. The possible E18 infection of the patient's from Nigeria highlights the potential for E18 spread through cross-border transmission [20]. Therefore, a multi-syndromic surveillance approach will be important to unravel the underlying reason(s) of this emergence and may help to predict and mitigate the re-emergence risk. The lack of data on enterovirus-associated meningitis in Niger could be probably due to limited detection resources, poor knowledge on the disease and misdiagnosis with severe cerebral malaria or bacterial meningitis. Therefore, it is important to raise awareness among healthcare professionals for the rapid identification and diagnosis of cases of VM in Niger by considering analysis of the cerebrospinal fluid in cases with persistent signs (high fever, headache, neck stiffness) and non-conclusive bacterial meningitis identification.
In addition, the sequence from Niger clustered with isolates from the Republic of India where a total of 4 confirmed cases of E18-associated infection was recently reported, including 1 death [5]. It highlights the potential for virus spread through international travel and could be attributable to long maintained the bilateral trade and investment relations between India and the West African countries such as Nigeria and Niger.
The detection of E18-associated aseptic meningitis in West Africa also argues for more research on this disease and should be considered as a warning sign. The newly characterized E18 sequence could be useful not only in the development and evaluation of novel medical countermeasures against enteroviruses, but also in studies focusing on better understanding of virulence factors of E18-associated aseptic meningitis and the phylodynamic of E18 worldwide and particularly in Africa.
This study presents some limitations that warrant the discussion. First, the weak medical care capacities of the CNRD did not allow to overcome the severe anemia presented by the patient all over his admission and therefore, complicated his condition and led to death. Second, this study was based on a single case, reducing the capacity in obtaining detailed clinical data which could have hindered robust statistical analyses to evaluate the association between the E18 infection and the clinical outcomes. Last, the use of a single strains for phylogeny analysis may have hindered our ability to assess the infection origin and probable introduction pathways.
## Conclusion
Although E18 has been circulating in Africa for years, induced E18-associated aseptic meningitis was only found in other continents previously. To the best of our knowledge, this is the first report of E18 infection in children from Niger, highlighting the need for an adapted medical care and future close surveillance. From our study, the patient's death could be associated with the underlying co-morbidity of sickle cell anemia SS form. Considering the good experience on medical care and vaccination strategy for bacterial meningitis outbreak containment, additional effort should be directed towards the investigation of viral etiologies of meningitis in Niger.
## References
1. Putz, Hayani, Zar (2013) *Meningitis. Prim Care*
2. Takáts, Balázs, Boros et al. (2024) "A meningoencephalitis outbreak associated with echovirus type 18 (E18) in south-western Hungary in mid-2023" *Arch Virol*
3. Who, Meningitis (2025) "Disponible sur: h t t p s"
4. Cdc, Meningitis (2025) "About Viral Meningitis"
5. Chen, Li, Guo et al. (2015) "An outbreak of echovirus 18 encephalitis/meningitis in children in Hebei province, china" *Emerg Microbes Infect Juin*
6. Lavania, Viswanathan, Bhardwaj et al. (2022) "Detection of Echovirus-18 in children suspected with SARS-CoV-2 infection with multisystem inflammatory syndrome: A case report from India"
7. Jiang, Xu, Li et al. (2023) "Case report: Clinical and virological characteristics of aseptic meningitis caused by a recombinant echovirus 18 in an immunocompetent adult"
8. Othman, Mirand, Slama et al. (2015) "Enterovirus Migration Patterns between France and Tunisia" *Jin DY, éditeur. PLOS ONE. 28 déc*
9. Osundare, Opaleye, Akindele et al. (2016) "Detection and characterization of human enteroviruses, human cosaviruses, and a new human parechovirus type in healthy individuals in Osun state" *Viruses*
10. Altan, Aiemjoy, Phan et al. (2018) "Enteric virome of Ethiopian children participating in a clean water intervention trial" *PLOS ONE. 16 août*
11. Fernandez-Garcia, Kebe, Fall et al. (2017) "Identification and molecular characterization of non-polio enteroviruses from children with acute flaccid paralysis in West africa" *Sci Rep*
12. Keita, Koundouno, Faye et al. (2021) "Resurgence of Ebola virus in 2021 in Guinea suggests a new paradigm for outbreaks" *Nat 23 Sept*
13. Trifinopoulos, Nguyen, Haeseler et al. (2016) "W-IQ-TREE: a fast online phylogenetic tool for maximum likelihood analysis" *Nucleic Acids Res 8 Juill*
14. 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 20 Janv*
15. Sibomana, Hakayuwa (2024) "The meningitis outbreak returns to Niger: Concern, efforts, challenges and recommendations"
16. Oduoye, Hb, Muzammil et al. (2023) "Meningitis in Niger Republic amidst COVID-19: current issues and novel recommendations" *Ann Med Surg*
17. Terranella, Cohn (2011) "Meningococcal conjugate vaccines: optimizing global impact" *Infect Drug Resist*
18. Geteneh, Kiros, Tamrat et al. (2025) "Viral meningitis in Sub-Saharan africa: trends in prevalence, etiologies, and diagnostic approaches" *Virol J 15 Avr*
19. Pallansch, Oberste "Cardiomyopathies and heart failure" *Biomol Infect Immune Mech. 2003 DICM*
20. Grunnill, Eshaghi, Damodaran et al. (2024) "Inferring enterovirus D68 transmission dynamics from the genomic data of two 2022 North American outbreaks. Npj Viruses 2" *Août* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12660759&blobtype=pdf | # Correction: Determinants of indiscriminate antimicrobial use in commercial chicken farms in Bangladesh and their impact on food safety and public health
Md Hasan, Tiasha Ali, A Dawlat Khan, Monjurul Islam, Abu Khan, Ibne Sayeed, Abdullah Noman, Shariful Al Mamun, Mohammad Islam, Cameron Hassasn, Tahmina Clark, Ariful Shirin, Islam, Mehedi Hasan, Arif Khan |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607598&blobtype=pdf | # Genome-wide mapping of EBV-induced genomic variations identifies the role of MUC19 in EBV latency
Jingwen Yu, Yaohao Wang, Qirong Liu, Xiaohui Zhou, Erle Robertson, Yonggang Pei
## Abstract
Epstein-Barr virus (EBV) infects over 95% of the world's population and is tightly associated with multiple human malignant diseases. As the first discovered human oncovirus, EBV is known to induce genomic instability by promoting various types of genomic modifications in host chromosomes. However, the mechanisms through which EBV interacts with the host genome and regulates cellular gene expression in genomic modifications are not yet fully elucidated. In this study, we conducted primary EBV infection in B cells and performed the analyses of copy number variants using whole genome sequencing. The results showed genomic regions susceptible to EBV-induced mutations and unveiled MUC19 to be a critical host factor in EBV latency, which was distinctively activated transcriptionally upon EBV infection. Finally, we identified the intrinsic tandem repeats in MUC19 to be its functional domain, which promotes cell survival and cell cycle through activating mechanistic target of rapamycin (mTOR) signaling in EBV-positive cells. Further results indicate that EBV nuclear antigen 1 binds to the promoter of the MUC19 gene and enhances its expres sion. In conclusion, these results provide novel insights into the roles of MUC19 in EBV latency, highlighting its potential as a promising therapeutic target for the treatment of EBV-associated lymphomas. IMPORTANCE Genomic instability is a hallmark of cancer. EBV contributes to host genomic instability after primary infection. This study maps the EBV-induced genomic variations using deep whole genome sequencing and identifies the critical factor MUC19, which is one of the most understudied genes, with a genomic sequence exceeding 177 kbp that encodes a protein over 800 kD. In this study, we revealed that EBV induced the duplicated copy number variants of the MUC19 gene and enhanced its expression, which further promotes cell survival and cell cycle via mTOR signaling. Overall, this study maps the genomic perturbations induced by EBV primary infection and offers new insights into the critical role of MUC19 in EBV latency.
lymphocytes (7). For instance, the latent EBV episome causes genomic instability by frequently interacting with the host genome (8). The advancement of high-throughput sequencing methods facilitates the identification of distinct chromosome modifications with higher precision (9,10). For instance, the genome-wide association study on single-nucleotide mutations from 681 clinical non-Hodgkin lymphoma (NHL) patients and 749 controls identified key mutations in diffuse large B-cell lymphoma (DLBCL) within LOC283177 and confirmed mutations associated with chronic lymphocytic leukemia at chromosomes 13q14, 11q22-23, 14q32, and 22q11.22 (11). Compared to small genomic variations on a single-nucleotide level, structural variations (SVs) on a copy number level bear the potential to further identify the chromosomal variations induced by viral antigens. Copy number variants (CNVs) typically refer to genomic fragment variations over 1 kb in the human genome and are known to contribute to multiple malignant cancers, including lymphomas (11)(12)(13)(14). Therefore, studying the phenotypic impact of CNVs in EBV-positive cells has the potential to reveal novel variations and deepen our understanding of EBV-associated oncogenesis.
Previous studies in at least 12 lymphoblastoid cell lines (LCLs) showed that EBV infection induces chromosome modifications across more than 70 chromosomal bands (P < 0.05) with shared integration at 1p31, 1q43, 2p22, 3q28, 4q13, 5p14, 5q12, and 11p15 (15). In Burkitt's lymphoma, EBV integration leads to large genomic deletions (dels) in the viral genome, including regions that contain the LMP and EBER genes. Meanwhile, the integrated viral genome disrupts chromosome stability within infected cells, leading to translocations and deletions on chromosomes 11 and 19 (8). Spectral karyotyping using the primary EBV infection model identified widespread chromatin and telomere abnormalities on the host genome that occurred 4 weeks after post-infec tion (16). Moreover, EBV-targeting genomic perturbations contribute to oncogenesis via disrupting genes, activating oncogenes, and increasing host genomic instability (17). Collectively, these studies confirm that EBV is capable of interacting with the host genome in B lymphocytes and inducing genomic variations to reshape the B-cell genomic landscape. However, the mechanisms by which EBV induces these mutations and their oncogenic pathways remain largely unexplored.
EBV proteins such as BNRF1 and EBV nuclear antigen 1 (EBNA1) are shown to mediate host chromosomal structural variations. BNRF1 induces SVs and is essential for the replication and maintenance of EBV latent infection (7). EBNA1 attaches EBV episomes to host genome through OriP during latency by forming a replication-dependent crossstructure with host DNA (18). A recent study revealed that the aggregation of EBNA1 around the host chromosome at 11q23 causes double-strand breakage in EBV-positive B lymphocytes, potentially inducing genomic mutation (19). The following results showed that EBV episomes attach to the human genome through clusters of 18 bp palindromic repeats homologous to the sequence on host chromosome 11q23, disrupting chromo somal stability (19). This study provides novel insights into EBV-induced mutations on chromosome 11q23, while EBV's ability to induce perturbations across the human genome suggests the involvement of additional mechanisms.
Here, we conducted an integrated analysis of EBV-associated genomic variations across different B-cell lines using whole genome sequencing (WGS) and highlighted the specific genomic CNVs induced by EBV primary infection. We discovered that EBV infection triggers the copy number duplication (dup) of the MUC19 repeat region and enhances MUC19 expression to promote cell survival and cell cycle, suggesting its oncogenic role. These findings provide new insights into EBV-induced genomic variations and highlight MUC19 as a novel target for anti-EBV therapeutics.
## RESULTS
## EBV can induce specific CNVs in B cells
To investigate the EBV-induced genomic changes, we performed EBV primary infection in B cells and collected the infected (EBVinf ) or uninfected B cells (total-B) for WGS. Then, CNV analysis revealed 53 distinct dups and 469 dels, highlighting EBV-induced genomic instability (Fig. 1a andb). Frequent CNV duplications were observed at chr17p11.2 (six dups), chr10p11.1 (four dups), and chr22q11.1 (four dups), while frequent deletions were detected at chr11p15.5 (nine dels), chr21q22.3 (eight dels), and chr3p24.3 (eight dels). We next sequenced two LCLs and identified shared CNVs of these three EBVpositive samples (Fig. 1c). Interestingly, the shared CNVs are found to overlap with known EBV-specific chromosomal fragile regions in BL and NPC (20)(21)(22), indicating the potential to study EBV-induced genomic variations in EBV-mediated diseases (Fig. 1c). To examine the characteristics of EBV-mediated CNVs, we mapped the CNV regions to the high-throughput chromosome conformation capture (Hi-C) results from EBV-positive GM12878 cells, which defined the topologically associating domains (TADs) with active transcription (23). This analysis revealed that the majority of EBV-induced CNVs overlap with TADs, with most deletions occurring within these regions (Fig. 1d), demonstrat ing that EBV infection can induce genomic variations and modulate TAD architecture. Additionally, the shared segmental duplications are generally within non-coding and intergenic regions (Fig. 1e), which suggests the potential functions of non-coding regions in EBV-induced oncogenesis.
The identified CNVs included known oncogenes, such as DMBT1 and FOXO1 (Fig. 1f andg). Rearrangements within DMBT1 are frequently discovered in multiple tumors (24,25), and mutated FOXO1 is a known marker for decreased overall survival in DLBCL patients associated with failure to achieve event-free survival at 24 months (EFS24) after diagnosis (14,26). Primary infection induced copy number duplication within DMBT1 and deletion in FOXO1, highlighting their essential functions during EBV infection. Besides CNV analyses, we performed the single-nucleotide polymorphism (SNP) analyses on EBV-infected primary B cells, which revealed 4,369,022 EBV-induced SNP sites (Fig. S1a). After normalizing the SNP counts based on the lengths of their CNVs, we further defined the 100 most genomically unstable CNV regions, which overlap with 233 genes. These selected genes were involved in specific virus infection and host defense (Fig. S1b). Therefore, we identified EBV-induced CNVs in host chromosomes from EBV primary infection using WGS analyses, which may play a crucial role in EBV-mediated diseases.
## CNV duplication of MUC19 is associated with its enhanced expression
To further identify key genes associated with the EBV-specific CNVs, we performed a comparative analysis between EBV-positive (EBVinf, LCL1, and LCL2) and EBV-negative (BL41, BJAB, and total-B) samples. The results revealed that only one gene, MUC19, was specifically duplicated by EBV infection (Fig. 2a), while seven other genes were found to be partially deleted (Fig. S2a). Moreover, our results suggest that EBV infection in B cells introduced a 17,368 bp duplication located within the middle of the MUC19 gene (Fig. 2b). Additionally, this duplication was determined by targeting specific regions of the potential CNV duplication across several EBV-associated cell lines. The results indicated that EBV-positive cells exhibit a higher copy number of this region compared to EBV-negative cells (Fig. 2c andd). We then propose that such duplication may lead to upregulation of MUC19 in EBV-positive B cells, and the following results showed that MUC19 expression is significantly higher in EBV-positive cell lines (LCL1, LCL2, and GM12878) compared to the EBV-negative cell lines (Ramos, BJAB, and DG75) (Fig. 2e), which suggests that MUC19 may be specifically upregulated in EBV latency.
Duplicated genes within CNV regions are shown to regulate pathogenesis by overexpression through the dosage effect (27). To further explore the expression changes of CNV-associated genes during EBV primary infection, we analyzed the transcriptional profile of EBV primary infection in B cells (Fig. S2b) (28). The results showed that MUC19 expression was upregulated in EBV-infected B cells 14 days post-infection (Fig. 2f). Enrichment analysis of differentially regulated genes from transcriptome further demonstrated the critical signaling pathways during EBV infection, including cell cycle, NF-κB signaling, and tumor necrosis factor (TNF) signaling (Fig. 2g), which are well-established pathways associated with EBV infection and its ability to drive proliferation (29). Additionally, several EBV-associated transcription factors also colocalize at the LRRK2/MUC19 region, highlighting its significance in EBV-mediated pathogenesis (Fig. S2c). Therefore, these findings indicate that EBV activates MUC19 expression upon infection, suggesting a contributing role of MUC19 in EBV latency.
## MUC19 promotes cell cycle in EBV-positive B cells
Human mucin gene MUC19 (chr12: 40,393,394-40,570,832) is over 177 kbp in length and encodes a protein of 8,384 amino acids with mucin-like threonine/serine-rich repeats (30). MUC19 has three von Willebrand factor D domains near its N-terminus, and a von Willebrand factor C (VWC) domain followed by a C-terminal cystine-knot (CTCK) domain (Fig. 3a). As a member of the mucin family, MUC19 shares close homology with MUC2, MUC6, MUC7, MUC5A, and MUC5B (Fig. S3a). Previous studies showed MUC19 secretion in glandular tissues and epithelial cells, including major salivary glands (31), lacrimal glands (32), middle ear epithelium (33), and airway tissues (34). However, our results indicate that MUC19 localizes in the cytoplasm (Fig. 3b) and shows no detectable secretion in EBV-associated cell lines (Fig. S3b), demonstrating that MUC19 primarily functions within the cytoplasm in these cells.
To explore the potential oncogenic role for MUC19 in B lymphoma cells, we generated MUC19 knockdown cells using CRISPRi and CRISPR/Cas9 systems (35,36), both of which inhibit MUC19 expression (Fig. 3c ande). Then, the following assays showed that MUC19 downregulation significantly reduced cell viability (Fig. 3d andf). This abrogated viability suggests an important role for MUC19 in maintaining cell growth and survival. Further more, MUC19 suppression in LCL1 caused cell cycle arrest and triggered apoptosis (Fig. 3g). The indicated cells are marked by a significantly elevated proportion of apoptotic cells (Sub-G1) (Fig. 3h). These findings were consistent with those observed in Namalwa cells using the CRISPRi system (Fig. S3c andd). In summary, we characterized the localization of MUC19 in EBV-positive lymphoma cells and identified its role in promoting cell survival and cell cycle using distinct CRISPR systems.
## MUC19 can activate the mechanistic target of rapamycin pathway through its repeat region
Previous studies have indicated that the C-terminus of MUC1 triggers Wnt/β-catenin pathway and increases SNAIL transcription to activate epithelial-mesenchymal transition (EMT) in cancer cells (37). To reveal the mechanisms underlying MUC19's role in promoting cell cycle, we determine how MUC19's C-terminal domains (VWC and CTCK) regulate the potential signaling pathways. However, after detecting the expression of key factors among Wnt/β-catenin, EMT, PI3K, and the mitogen-activated protein kinase (MAPK) pathways, we failed to observe any significant expression changes in the presence of the VWC and CTCK domains of MUC19 protein (Fig. S4a through d).
Then, a recent CNV analysis identifies MUC19 as a highly mutated gene in hepatoid adenocarcinoma of the stomach (HAS) and a promising novel diagnostic marker for HAS (38). They found that overexpression of MUC19 leads to activated Wnt signaling exemplified by Wnt factors cyclin D1 (encoded by CCND1) and c-Myc. Therefore, we performed CRISPR-mediated knockout of MUC19 in Daudi cells and demonstrated that MUC19 modulates cyclin D1 expression but not c-Myc (Fig. 4a andb), indicating that MUC19 may utilize signaling pathways other than canonical Wnt signaling. Notably, weeks of EBV infection. Dots denote shared duplications (red) and deletions (blue) of all three EBV-positive cell samples. CNVs for BL and NPC are obtained from studies describing EBV-specific chromosomal fragile sites (20)(21)(22). (d) EBV-induced CNVs overlapped with the topologically associating domains in EBV-positive GM12878 cells. EBV-induced CNVs are defined as the CNVs shared between EBVinf and LCLs that are not observed in uninfected primary B cells. (e) EBV-induced CNVs were mapped to the Hg38 refGene annotation to reveal the characteristics of these CNV regions. (f and g) The scatter plots display the mutated oncogenes as a result of EBV infection. DMBT1 is shown to be duplicated (f), and FOXO1 contains EBV-induced segmental deletion (g). The red line indicates the average copy ratio per segment. A copy ratio (log2) exhibits the log2 transformed value of CNV in tumors by reflecting deviations from diploid coverage. cyclin D1 expression has been previously reported to depend on mTORC1 signaling (39,40), implying a potential connection between MUC19 and mTORC1-regulated processes.
To further explore the role of MUC19 in EBV latency, we noticed the repeat region in the MUC19 protein, which contains tandem copies of "GVTGTTGPSA" (Fig. 3a). Motif analysis revealed approximately 351 copies of this basic repeat unit in MUC19 DNA (5′-GGAGTGACAGGGACAACTGGACCATCAGCT-3′, referred to as Re) (Fig. 4c). Of these, 338 copies are located within the EBV-induced duplicated CNV regions. To validate the biological functions of Re, we first overexpressed Re in HEK293T cells and found it also localized in the cytoplasm (Fig. 4d), which is similar to MUC19's localization in B cells (Fig. 3b). Following the confirmation of Re expression (Fig. S4e), it was observed that Re overexpression upregulated several downstream factors of the mechanistic target of rapamycin (mTOR) signaling pathway, which further supports that Re can activate mTOR signaling (Fig. 4e andf). MUC19 knockdown in EBV-positive LCL1 cells leads to suppressed phosphorylation of mTOR and pan-PI3K, as well as cyclin D1 expression, demonstrating the critical role of MUC19 in mTOR signaling (Fig. 4g). Furthermore, Re overexpression in both HEK293T and Akata cells confirmed its role in mTOR activation through enhancing the phosphorylation of mTOR and pan-PI3K and increasing cyclin D1 expression (Fig. 4h). The mRNA expression of mTOR downstream factors RPS6KB1, CCND1, and SGK1 was elevated following the increase of Re copies (Fig. 4i through k). Moreover, Re expression was found to elevate cell viability, underscoring its significance in promoting cell cycle progression and cell survival (Fig. 4l), suggesting its pro-survival role, which may have further oncogenic implications (40,41). Therefore, these results indicate that the tandemly linked repeats within MUC19 have overlaying effects on mTOR activation, supporting its pro-survival role following partial duplication from EBV infection.
## EBNA1 binds to the MUC19 gene and enhances its expression
EBV nuclear antigens are shown to disrupt host genomic stability, potentially leading to oncogenic mutations (42). Among them, EBNA1 is a multi-functional EBV-encoded latent protein that is important for maintenance of the genome and also serves as a transcription activator (18,43). EBNA2 and EBNALP are also EBV-encoded transcription factors (44). Overexpression of these antigens revealed that EBNA1 significantly activates MUC19 expression (Fig. 5a), which is also validated in EBV-positive LCL1 cells (Fig. 5b). EBNA1 can activate cyclin D1 expression that may be associated with MUC19 (Fig. 5c). To explore whether EBNA1 regulates MUC19 expression on transcriptional initiation, we performed chromatin immunoprecipitation (ChIP) assays in EBV-positive LCL1 and Akata cells with overexpressed EBNA1 and demonstrated that EBNA1 can bind to the MUC19 promoter region and facilitate its transcription (Fig. 5d ande).
EBNA1 is shown to aggregate around chromosomal regions and trigger doublestrand breakage (19). Additionally, the ChIP assay revealed that EBNA1 actively interacts with the MUC19 repeat region (Fig. 5f). Together, these findings imply that EBNA1 may induce MUC19 CNVs in EBV infection. Furthermore, multiple sequence alignment indicates homology between the MUC19 repeat unit and the EBV genome (Fig. 5g). This homologous sequence specifically entailed 19 bp of the total 30 bp sequence and is the same as the intrinsic repeats among the EBV genome, such as BNRF1, EBNAs, show that MUC19 is duplicated by EBV primary infection when comparing primary B cells (total-B) to EBV-infected B cells (EBVinf). (c) A schematic of designed primers to validate the identified CNVs in MUC19. (d) The copy number of identified CNVs within MUC19 was determined in EBV-positive or EBV-negative cell lines. ****, P < 0.0001. (e) Real-time PCR was conducted to detect MUC19 mRNA expression in EBV-positive or EBV-negative cell lines. (f) RNA-seq analysis shows that multiple mucin factors were differentially expressed during EBV primary infection. The samples on days 0 and 14 were obtained from the GEO data set GSE125974, with a P value threshold of 0.05. (g) Kyoto Encyclopedia of Genes and Genomes analysis on the differentially expressed genes (DEGs) reveals the important cellular signaling pathways regulated by EBV primary infection. An absolute fold change of 5 was used as a threshold for differentially expressed genes. MAPK, mitogen-activated protein kinase. BHLF1, BMRF1, and LF3 genes. In the MUC19 DNA sequence, 351 motif occurrences were identified within the MUC19 gene, while 51 homologous occurrences were found in the EBV genome (Fig. 5h). This interesting finding suggests homologous recombination of EBV with the host genome, specifically MUC19, during EBV infection.
Finally, to explore the functions of the MUC19 repeat unit on EBV dynamics, we developed sgRNAs targeting the indicated 30 bp repeat unit in EBV-positive Daudi cells (Fig. 6a). Disruption of the MUC19 repeats results in reduced MUC19 expression (Fig. 6b). Interestingly, disruption of the 30 bp repeat unit induces intense expression of viral immediate-early genes BZLF1 and BRLF1, but not viral early gene (BALF5) or late gene (BLLF1) (Fig. 6c). Similarly, targeting the 30 bp repeat unit does not result in increased EBV replication or progeny virus production (Fig. 6d). These findings collectively indicate that MUC19 can inhibit the initiation of EBV reactivation. The tandem repeats of MUC19 are essential for this function, offering novel perspectives for developing targeted therapies against persistent EBV infection.
To conclude, we demonstrate that EBV infection promotes genomic instability via EBNA1, which triggers duplication within MUC19's repeat region and activates its expression. The expressed MUC19 then activates mTOR signaling, thus promoting cell survival and cell cycle, finally contributing to EBV latency and EBV-mediated diseases (Fig. 6e).
## DISCUSSION
This study explored EBV-induced genomic perturbations in B cells on a copy number level and identified a specific CNV duplication of MUC19 in EBV-positive cell lines, indicating the critical role of MUC19 in EBV-mediated pathogenesis. These results showed that MUC19 expression is upregulated by EBV infection and promotes cell cycle through modulating the mTOR-cyclin D1 signaling pathway via its tandem repeats. Then, we confirmed that EBNA1 can bind to the promoter of the MUC19 gene and enhance its expression in B lymphoma cells. Finally, we showed that the MUC19 repeat region can inhibit the initiation of EBV reactivation.
Our study highlights the CNVs introduced by EBV infection throughout the human genome. Besides MUC19, there are other interesting CNVs shown from our primary infection model, including IRF4/DUSP22 duplication and Notch homolog 2 N-terminallike protein A (NOTCH2NLA) deletion (Table S1; Fig. S5a andb). Interestingly, IRF4 and dual specificity phosphatase 22 (DUSP22) were identified to be within the same EBV super enhancer (45). These results indicate that DUSP22 was activated in EBV-positive B cells, suggesting its potential role in promoting EBV latency (Fig. S5a). NOTCH2NLA was downregulated in LCLs (Fig. S5b), possibly due to EBV-induced copy number deletion. However, NOTCH2NLA activates the Notch signaling pathway, which further contributes to EBV oncogenesis (46,47). This change in expression brought by CNV cannot provide a functional explanation for its role in EBV latency through Notch signaling. However, it might still be important from angles other than Notch activation. Further experiments are required to further explore the role of NOTCH2NLA in EBV-related malignancies.
Furthermore, our findings suggest an EBV-specific MUC19 at its C-terminus, which could link to maintenance of EBV episome during latency. However, overexpression of its C-terminus fails to impact key pro-survival signaling pathways (Fig. S4a through d). Nevertheless, the C-terminus of MUC19 could have functional roles but requires further investigation. Although MUC19 is the only mucin factor identified from our screen with copy number duplication, mucin factors in general were differentially regulated upon EBV infection (Fig. 2g). For instance, the membrane-tethered MUC12 exhibited a similar expression pattern to MUC19 after EBV primary infection. The oncogenic role of MUC12 in renal cell carcinoma is related to the c-Jun/TGF-β signaling pathway (48). As the mucin family is tightly associated with numerous inflammatory pathways and immune responses in various pathologies, these genes may also be involved in EBV-associated biological processes. Future efforts may explore the roles of other mucin factors in regulating EBV-mediated diseases.
EBV primary infection has demonstrated that the observed genomic variations are EBV specific and identified EBNA1 as the underlying contributor, but the precise mechanisms have yet to be fully elucidated. We observed that EBNA1 overexpression in EBV-associated cells facilitates its binding to multiple regions of MUC19; thus, further investigation may be needed to determine whether the endogenous expression of EBNA1 operates similarly. Further studies could employ fluorescence staining techniques to observe interactions between viral proteins and specific genomic regions, as well as chromosomal breakages in adjacent regions, to confirm the occurrence of double-strand breaks (DSBs) and DNA repair. Additionally, we can explore the mechanisms underly ing mutations during transcription. Previous studies found that the CHAMP1 complex promotes the assembly of heterochromatin at multiple chromosomal loci, including centromeres and telomeres, and facilitates homologous recombination repair of DSBs in these regions (49). Moreover, the CHAMP1 complex plays a crucial role in heterochro matin assembly and DSB repair in highly specialized chromosomal regions. This study suggests that DSBs may originate from secondary structures formed by single-stranded DNA and its binding to heterochromatin. The studies of heterochromatin organization in EBV primary infection could enhance our comprehension of EBV-host interactions on chromosomal modifications.
In conclusion, our study combined WGS data from EBV primary infection and other LCLs to characterize EBV-specific copy number variations. Integrative analysis and expression assays identified MUC19 as a critical factor for EBV infection. These findings unveil a novel role for MUC19 in regulating cell cycle through mTOR activation to promote EBV latency. Furthermore, we revealed that the repeat region is essential for MUC19-mediated mTOR activation and is closely connected with the EBV-encoded EBNA1. Therefore, the study elucidates genomic mutations that are induced by EBV primary infection and highlights the potential of MUC19 as a novel therapeutic target for EBV-associated malignancies.
## MATERIALS AND METHODS
## Plasmids
EBV latent genes (EBNA1, EBNA2, and EBNALP) are cloned into the pEGFP-C1 vector (Takara, Japan). The VC plasmid in the study was generated by cloning the VWC and CTCK domains (NM_173600.2: 24,529-25,205) into the pCMV-Myc-N vector (Takara) that was digested by EcoRI and XhoI. The Re plasmid was generated by cloning the FIMO identified a total of 351 copies of the 30 bp repeat in MUC19, with 338 located within the duplicated region. A P value threshold of 0.00002 was applied.
(d) The localization of the basic repeat unit (with Myc tag) was determined in HEK293T cells. Re was stained with Alexa Fluor 488 (green), and cell nuclei were stained using DAPI (blue). The yellow dotted line denotes the cell membrane of a single HEK293T cell. Re, the 30 bp basic repeat unit in MUC19 DNA. Scale bars, 10 µm. (e) HEK293T cells were transfected with Re plasmid or not, and then qPCR was performed to determine mRNA expression of the indicated targets.
(f) The schematic illustrates the mTOR-related signaling pathway. (g) The expression of key factors in the mTOR signaling pathway was examined after MUC19 downregulation in LCL1 cells. The antibody for pho-PI3K targets P85α/β/P55γ, whereas the antibody for PI3K targets P85α. (h) The expression of indicated cellular factors was determined following Re overexpression in HEK293T and Akata cells at 48 h post-transfection. (i-k) The expression of RPS6KB1 (i), CCND1 (j), or SGK1 (k) was determined 48 h following the expression of multiple copies of Re. Re (×3), 3 copies of the 30 bp repeat; Re (×10), 10 copies of the 30 bp repeat. ***, P < 0.001; ****, P < 0.0001. (l) CCK-8 was utilized to assess cell viability after Re overexpression in HEK293T cells. basic repeat unit of MUC19 (5′-GGAGTGACAGGGACAACTGGACCATCAGCT-3′) into the pCMV-Myc-N vector (Takara), digested by EcoRI and XhoI. pMD2.G (Addgene #12259) and PsPAX2 (Addgene #12260) are used for lentivirus packaging. For CRISPR systems, the plasmid dCas9-KRAB-MeCP2-green fluorescent protein (GFP) is modified from lenti_dCas9-KRAB-MeCP2 (Addgene #122205) by replacing the BSD with GFP, which is a generous gift from Dr. Ruilin Tian (Southern University of Science and Technology, China). The sgRNA backbone plasmid pLG15-NC is also a gift from Dr. Ruilin Tian. Transient knockout using CRISPR/Cas9 is performed using lentiCRISPRv2 (Addgene #98290). sgRNA primers used for CRISPR systems are designed using the CHOPCHOP online tool (chopchop.cbu.uib.no/) and are listed in Table S2.
## Cells and antibodies
HEK293T (human embryonic kidney cell line) cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS), 50 U/mL penicillin, 50 µg/mL streptomycin, and 2 mM L-glutamine. EBV-negative cells (total-B, BL41, BJAB, DG75, and Ramos) and EBV-positive cells (EBVinf, Akata, Daudi, GM12878, LCL1, LCL2, Namalwa, and Raji) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 50 U/mL penicillin, 50 µg/mL streptomycin, and 2 mM L-glutamine. All the cells were cultured in a 37°C incubator at 5% CO 2 . Transfections of HEK293T were performed using a Calcium Phosphate Cell Transfection Kit (C0508; Beyotime, China), and transfections into LCL1 and Akata were conducted using GenePulser Xcell Electropo ration Systems (Bio-Rad, USA).
MUC19 antibody was purchased from R&D Systems (MAB8245). β-Actin and cyclin D1 antibodies were obtained from Proteintech (66009-1-Ig and 60186-1-Ig). pho-PI3K P85α/β/P55γ antibody was purchased from Beyotime (AF5905). mTOR, pho-mTOR, and PI3K p85α antibodies were obtained from Abmart (T55306, T56571, and T40115).
## EBV primary infection and whole genome sequencing
Approximately 10 million primary B cells were infected with EBV (B95.8, multiplicity of infection = 1) for 14 days, while 5 million B cells were stored at -140°C as a negative control. Then, both the infected and uninfected B cells were collected, and the genomic DNA was extracted from these obtained cells to establish a library using Nextera DNA Flex Library Preparation kits (Illumina, USA). Subsequently, the libraries were submitted to the Genome Technology Access Center at Washington University in St. Louis for whole genome sequencing using the NovaSeq 6000 system (Illumina). Besides collecting cells in EBV primary infection, other EBV-negative or EBV-positive cells were also sent for whole genome sequencing with the same procedures.
## Multi-omics sequencing data analysis
CNVkit (v.0.9.4) (50) was used to call CNVs from WGS data. The WGS data for primary B cells were used as a reference for the CNV analysis of the EBV-transformed cells. A comparative analysis was performed using bedtools (v.2.30.0) and R scripts. Visuali zations were made using the R package circlize (v.0.4.16). The WGS data underwent standard processes for quality control and alignment to human genome assembly Hg38.
Somatic SNVs and small insertions and deletions were performed using GATK (51). Joint analysis of CNV and SNV was done by intersecting SNV loci with CNV segments using the R package GenomicRanges (v.1.52.0). The CNV regions were sorted in descending order based on the number of SNVs contained within each region. To normalize the data, mutation scores were calculated by dividing the number of SNVs and normalized by the length of the corresponding CNV region. The top 100 CNV segments with the highest mutation scores were selected for enrichment analysis.
RNA-seq data were analyzed using the sequencing results from EBV-infected primary B cells on days 0 and 14, which are available at the Gene Expression Omnibus (GEO) under accession number GSE125974 (28). The raw data of gene expression were quantified using Salmon (v.0.8.2). Differentially expressed genes were obtained using ESeq2 (v.1.40.2) in R, with a P value of <0.05 and absolute fold change of >5. Kyoto Encyclopedia of Genes and Genomes analysis was carried out using the R package clusterProfiler (v.4.8.3) and visualized by ggplot2 (v.3.5.1). ChIP-seq visualizations were performed using WashU Epigenome Browser (52). The identified Hi-C loops from GM12878 cells were acquired from GEO under data set GSE63525 (23).
## CRISPR systems
For the CRISPRi system, the dCas9-KRAB-MeCP2-GFP plasmid was transfected into Namalwa cells to establish stable cell lines through GFP fluorescence-activated cell sorting. Subsequently, sgRNAs targeting promoter regions of interest were designed using the CHOPCHOP online tool (chopchop.cbu.uib.no/) and cloned onto the pLG15-NC vector. These sgRNA constructs were then transfected into target cells, followed by puromycin selection at 2 µg/mL for 3 days and subsequent cell culture in the normal medium.
For the CRISPR/Cas9 system, the lentiCRISPR (v.2) plasmid was transfected using electroporation in both LCL1 and Daudi cell lines. LCL1 cells then underwent puromycin selection at 0.5 µg/mL for 3 days, followed by cell culture with 20% FBS, while Daudi cells underwent puromycin selection at 3 µg/mL for 3 days before cell culture with 20% FBS.
## Quantitative real-time PCR
Quantitative real-time polymerase chain reaction (qRT-PCR) was utilized to measure the expression of target genes. Total RNA was extracted using an RNA extraction kit (Beyotime). Samples with high-quality RNA (A260/A280 ratio between 1.8 and 2.0) were used for cDNA synthesis. The synthesized cDNA was used as a template for PCR amplification using gene-specific primers with SYBR Green Master Mix (Yeason, China). The qRT-PCR reactions were run on QuantStudio (v.5; Thermo Fisher Scientific, USA) with the following cycling conditions: initial denaturation at 95°C for 5 minutes, followed by 40 cycles of denaturation at 95°C for 15 seconds, annealing at 55°C for 30 seconds, and extension at 72°C for 30 seconds. The relative expression levels of the target genes were calculated using the 2 (-ΔΔCt) method. The qPCR primers designed in this study are listed in Table S2.
## Chromatin immunoprecipitation assay
Chromatin immunoprecipitation is performed using a ChIP assay kit (Beyotime). Ten million cells were collected 48 h post-transfection and were cross-linked using formalde hyde to a final concentration of 1% (0.68 mL of 37%/25 mL media) for 10 minutes. The cross-linking was stopped using glycine solution with a final concentration of 125 mM. Next, the cells were collected using centrifugation (700 × g, 4 min) and lysed using ChIP lysis buffer. After sonication, the lysates were incubated with salmon sperm DNA/protein A agarose beads, together with 5 µg EBNA1 antibody or normal IgG antibody, and placed on a rotating platform overnight at 4°C. The next day, the formaldehyde cross-links were reversed, and the collected DNA was purified with a DNA purification kit (TIANGEN, China).
The primers used for targeting the promoter (-323 bp) were 5′-TCTGGGTTTGGTATGA GCTGG-3′ and 5′-GTCTGCCATCATGGGCTAGG-3′; those used for targeting the promoter (-226) were 5′-CACAGGGTGGCAAGAAGACA-3′ and 5′-TGGGCTAGGTGTGGTAAACT-3′. The primers used for the repeat region were 5′-GGACAACTGGACCATCAGCT-3′ and 5′-TCC AGTTGTCCCTGTCACTCC-3′.
## Cell viability assay
Cell viability was measured using the Cell Counting Kit-8 (Beyotime). Briefly, the indicated cells were transferred to a 96-well plate at a density of 5,000 cells per well in a 100 µL medium. Ten microliters of CCK-8 solution was then added to each well. Then, the optical density at 450 nm was measured 2 h after the addition of the CCK-8 reagent using BioTek Synergy H1 microplate reader (Agilent Technologies, USA).
## Immunofluorescence
The B cells are centrifuged and air-dried on coverslips. Then they were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 for 10 minutes, and blocked with 5% bovine serum albumin (BSA) for 20 minutes at room temperature. Primary antibodies against the target protein were diluted at a ratio of 1:1,000 using phos phate-buffered saline (PBS)-B (4% BSA) and incubated overnight at 4°C. After washing with PBS, the cells were incubated with fluorochrome-conjugated secondary antibodies (1:1,000 diluted) for 1 h at room temperature in the dark. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole. Fluorescent images were taken using a confocal microscope LSM 900 (Zeiss, Germany). Image processing was performed using Zen blue (v. 3.4.91).
## Western blot and enzyme-linked immunosorbent assay
The cells were harvested and lysed for protein extraction. Then the proteins were resolved on SDS-PAGE gel and transferred to a polyvinylidene fluoride (PVDF) membrane. This membrane was blocked with 5% skim milk, followed by overnight incubation with a specific primary antibody. After washing the membranes, they were incubated with horseradish peroxidase (HRP)-labeled secondary antibody for 1 h, and the membrane was washed again. Finally, the signals were detected through chemiluminescence, and the intensity of protein bands was quantitatively measured using the image analysis software ImageJ (v.1.54). Additionally, the enzyme-linked immunosorbent assay (ELISA) was performed to evaluate the secretion of MUC19 in B lymphoma cells using a commercially available ELISA kit (CUSABIO, China).
## Flow cytometry assay
This assay was determined using the Cell Cycle and Apoptosis Analysis Kit (Beyotime), which employs the propidium iodide staining method for cell cycle and apoptosis analysis. After staining the cellular DNA with propidium iodide, flow cytometry was performed using FACSCanto SORP (BD Biosciences, USA), and the results were analyzed with FlowJo (v.10.6.2).
## Motif discovery
The motif discovery was conducted using FIMO (v.5.5.5) (53). The P value was defined as the probability of a random match for the desired sequence with as good or better a score. The score was computed by summing the appropriate entries from each column of the position-dependent scoring matrix that represents the motif. R package ggseqlogo (v.0.2) was used to visualize the motifs.
## Figure S5 (mBio02055-25-s0005.tif). EBV-induced CNV deletions may cause differential expression in various B cells.
Legends (mBio02055-25-s0006.docx). Legends for supplemental materials. Table S1 (mBio02055-25-s0007.xlsx). The identified CNVs induced by EBV primary infection. Table S2 (mBio02055-25-s0008.xlsx). The primers used in this study. Table S3 (mBio02055-25-s0009.xlsx). Source data.
## References
1. Zhou, Schmidt, Jiang et al. (2015) "Epstein-Barr virus oncoprotein super-enhancers control B cell growth" *Cell Host Microbe*
2. Soldan, Lieberman (2023) "Epstein-Barr virus and multiple sclerosis" *Nat Rev Microbiol*
3. Damania, Kenney, Raab-Traub (2022) "Epstein-Barr virus: biology and clinical disease" *Cell*
4. Negrini, Gorgoulis, Halazonetis (2010) "Genomic instability--an evolving hallmark of cancer" *Nat Rev Mol Cell Biol*
5. Hanahan, Weinberg (2011) "Hallmarks of cancer: the next generation" *Cell*
6. Mesri, Feitelson, Munger (2014) "Human viral oncogenesis: a cancer hallmarks analysis" *Cell Host Microbe*
7. Shumilov, Tsai, Schlosser et al. (2017) "Epstein-Barr virus particles induce centrosome amplification and chromosomal instability" *Nat Commun*
8. Jox, Rohen, Belge et al. (1997) "Integration of Epstein-Barr virus in Burkitt's lymphoma cells leads to a region of enhanced chromosome instability" *Ann Oncol*
9. Mahmoud, Gobet, Cruz-Dávalos et al. (2019) "Structural variant calling: the long and the short of it" *Genome Biol*
10. Alkan, Coe, Eichler (2011) "Genome structural variation discovery and genotyping" *Nat Rev Genet*
11. Robbe, Ridout, Vavoulis et al. (2022) "Whole-genome sequenc ing of chronic lymphocytic leukemia identifies subgroups with distinct biological and clinical features" *Nat Genet*
12. Spielmann, Lupiáñez, Mundlos (2018) "Structural variation in the 3D genome" *Nat Rev Genet*
13. Conde, Riby, Zhang et al. (2014) "Copy number variation analysis on a non-Hodgkin lymphoma case-control study identifies an 11q25 duplication associated with diffuse large B-cell lymphoma" *PLoS One*
14. Novak, Asmann, Maurer et al. (2015) "Whole-exome analysis reveals novel somatic genomic alterations associated with outcome in immunoche motherapy-treated diffuse large B-cell lymphoma" *Blood Cancer J*
15. Lestou, Braekeleer, Strehl et al. (1993) "Non-random integration of Epstein-Barr virus in lymphoblastoid cell lines" *Genes Chromosomes Cancer*
16. Lacoste, Wiechec, Silva et al. (2010) "Chromosomal rearrangements after ex vivo Epstein-Barr virus (EBV) infection of human B cells" *Oncogene*
17. Vojtechova, Tachezy, 2025 "Genome integration of human DNA oncoviruses" *J Virol*
18. Dheekollu, Wiedmer, Ayyanathan et al. (2021) "Cell-cycle-dependent EBNA1-DNA crosslinking promotes replication termination at oriP and viral episome maintenance" *Cell*
19. Li, Abbasi, Kim et al. (2023) "Chromosomal fragile site breakage by EBV-encoded EBNA1 at clustered repeats" *Nature*
20. Sall, Shmakova, Karpukhina et al. (2023) "Epstein-Barr virus reactivation induces MYC-IGH spatial proximity and t(8;14) in B cells" *J Med Virol*
21. Thomas, Dreval, Gerhard et al. (2023) "Genetic subgroups inform on pathobiology in adult and pediatric Burkitt lymphoma" *Blood*
22. Bruce, To, Lui et al. (2021) "Whole-genome profiling of nasophar yngeal carcinoma reveals viral-host co-operation in inflammatory NF-κB activation and immune escape" *Nat Commun*
23. Rao, Huntley, Durand et al. (2014) "A 3D map of the human genome at kilobase resolution reveals principles of chromatin looping" *Cell*
24. Mollenhauer, Herbertz, Holmskov et al. (2000) "DMBT1 encodes a protein involved in the immune defense and in epithelial differentiation and is highly unstable in cancer" *Cancer Res*
25. Nakahara, Shiraishi, Okamoto et al. (2004) "Detrended fluctuation analysis of genome-wide copy number profiles of glioblastomas using array-based comparative genomic hybridization" *Neuro Oncol*
26. Trinh, Scott, Morin et al. (2013) "Analysis of FOXO1 mutations in diffuse large B-cell lymphoma" *Blood*
27. Weischenfeldt, Symmons, Spitz et al. (2013) "Phenotypic impact of genomic structural variation: insights from and for human disease" *Nat Rev Genet*
28. Wang, Li, Zhang et al. (2019) "RNA sequencing analyses of gene expression during Epstein-Barr virus infection of primary B lymphocytes" *J Virol*
29. Yu, Wang, Wang et al. (2024) "Decoding critical targets and signaling pathways in EBV-mediated diseases using large language models" *Viruses*
30. Culp, Robinson, Cash et al. (2015) "Salivary mucin 19 glycoproteins: innate immune functions in Streptococcus mutans-induced caries in mice and evidence for expression in human saliva" *J Biol Chem*
31. Chen, Zhao, Kalaslavadi et al. (2004) "Genome-wide search and identification of a novel gel-forming mucin MUC19/MUC19 in glandular tissues" *Am J Respir Cell Mol Biol*
32. Yu, Chen, Han et al. (2008) "MUC19 expression in human ocular surface and lacrimal gland and its alteration in Sjögren syndrome patients" *Exp Eye Res*
33. Kerschner, Khampang, Erbe et al. (2009) "Mucin gene 19 (MUC19) expression and response to inflammatory cytokines in middle ear epithelium" *Glycoconj J*
34. Rose, Voynow (2006) "Respiratory tract mucin genes and mucin glycoproteins in health and disease" *Physiol Rev*
35. Tian, Gachechiladze, Ludwig et al. (2019) "CRISPR interference-based platform for multimodal genetic screens in human iPSC-derived neurons" *Neuron*
36. Sanjana, Shalem, Zhang (2014) "Improved vectors and genomewide libraries for CRISPR screening" *Nat Methods*
37. Gnemmi, Bouillez, Gaudelot et al. (2014) "MUC1 drives epithelial-mesenchymal transition in renal carcinoma through Wnt/β-catenin pathway and interaction with SNAIL promoter" *Cancer Lett*
38. Zhu, Yu, Xu et al. (2022) "Genomic profiling and the impact of MUC19 mutation in hepatoid adenocarci noma of the stomach" *Cancer Commun*
39. Panwar, Singh, Bhatt et al. (2023) "Multifaceted role of mTOR (mammalian target of rapamycin) signaling pathway in human health and disease" *Sig Transduct Target Ther*
40. Huang, Dai, Mou et al. (2009) "Overproduction of cyclin D1 is dependent on activated mTORC1 signal in nasopharyngeal carcinoma: implication for therapy" *Cancer Lett*
41. Tashiro, Tsuchiya, Imoto (2007) "Functions of cyclin D1 as an oncogene and regulation of cyclin D1 expression" *Cancer Sci*
42. Gruhne, Sompallae, Masucci (2009) "Three Epstein-Barr virus latency proteins independently promote genomic instability by inducing DNA damage, inhibiting DNA repair and inactivating cell cycle checkpoints" *Oncogene*
43. Kennedy, Sugden (2003) "EBNA-1, a bifunctional transcriptional activator" *Mol Cell Biol*
44. Portal, Zhou, Zhao et al. (2013) "Epstein-Barr virus nuclear antigen leader protein localizes to promoters and enhancers with cell transcription factors and EBNA2" *Proc Natl Acad Sci*
45. Jiang, Zhou, Liang et al. (2017) "The Epstein-Barr virus regulome in lymphoblastoid cells" *Cell Host Microbe*
46. Fiddes, Lodewijk, Mooring et al. (2018) "Human-specific NOTCH2NL genes affect notch signaling and cortical neurogenesis" *Cell*
47. Saha, Robertson (2011) "Epstein-Barr virus-associated B-cell lymphomas: pathogenesis and clinical outcomes" *Clin Cancer Res*
48. Gao, Yin, Zhang et al. (2020) "The oncogenic role of MUC12 in RCC progression depends on c-Jun/TGF-β signalling" *J Cell Mol Med*
49. Li, Zhang, Syed et al. (2025) "CHAMP1 complex directs heterochro matin assembly and promotes homology-directed DNA repair" *Nat Commun*
50. Talevich, Shain, Botton et al. (2016) "CNVkit: genome-wide copy number detection and visualization from targeted DNA sequenc ing" *PLoS Comput Biol*
51. Mckenna, Hanna, Banks et al. (2010) "The Genome Analysis Toolkit: a MapReduce framework for analyzing nextgeneration DNA sequencing data" *Genome Res*
52. Li, Hsu, Purushotham et al. (2019) "WashU epigenome browser update" *Nucleic Acids Res*
53. Grant, Bailey, Noble (2011) "FIMO: scanning for occurrences of a given motif" *Bioinformatics*
54. Pei, Banerjee, Jha et al. (2017) "An essential EBV latent antigen 3C binds Bcl6 for targeted degradation and cell proliferation" *PLoS Pathog* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584710&blobtype=pdf | # Genomic and pathological insights into the first identified genotype IIIb chicken anemia virus strain in Bangladesh
Marjana Akter, Roni Mia, S Nazmul Hasan, Anandha Mozumder, Nurejunnati Jeba, Raduyan Farazi, Farzana Akter, Sharmin Akter, Md Rahman, Sukumar Saha, Tofazzal Islam, Md Golzar Hossain, Md Hossain
## Abstract
Chicken anemia virus (CAV) is a highly infectious pathogen that causes severe immunosuppression and significant economic losses in poultry, with limited genomic data in Bangladesh hindering effective disease control. This study character izes a CAV strain from a field outbreak in Bangladesh through clinical, pathological, and molecular analyses, including genome sequencing, genotyping, and assessment of viral protein structure. Clinically suspected chickens exhibiting anemia, depression, pale combs, and cyanotic wings underwent gross and histopathological examinations. Viral DNA was extracted from bone marrow samples and analyzed using PCR, followed by complete genome sequencing with next-generation sequencing. Phylogenetic analysis, genotyping, mutational profiling, and structural predictions of viral proteins (VP1, VP2, and VP3) were performed to assess evolutionary relationships and pathogenic potential. Histopathology confirmed severe lymphoid depletion in the thymus, spleen, and bursa of Fabricius, consistent with CAV-induced immunosuppression. Molecular detection confirmed CAV presence in eight out of ten samples. Whole-genome analysis revealed a 2,330 bp genome with 42% GC content, clustering within genotype IIIb and showing close genetic relatedness to a Chinese strain. Mutational analysis revealed several nucleotide substitutions in the viral genomes, with the highest number of amino acid changes observed in the VP3 protein. Computational modeling revealed minor structural variations in VP1 and VP2, which may affect antigenicity and phosphatase activity. This study provides the complete genome characterization of a Bangladeshi CAV strain, revealing critical genetic variations that may influence viral virulence. The findings underscore the need for enhanced surveillance and targeted vaccine strategies to mitigate CAV-related losses in poultry.
IMPORTANCEThis study provides the first complete genomic characterization of a genotype IIIb chicken anemia virus (CAV) strain in Bangladesh. By integrating clinical, pathological, and molecular analyses, the research identifies critical genetic variations that could influence viral virulence, immune evasion, and disease severity. The findings highlight the close genetic relationship between this Bangladeshi strain and a previously reported Chinese strain, suggesting potential epidemiological links. Furthermore, the study underscores the need for enhanced surveillance and targeted vaccine strategies to mitigate CAV-induced immunosuppression and economic losses in poultry farming. The insights gained from this research contribute to a deeper understanding of CAV evolution and could inform future diagnostic and control measures to protect poultry populations.
## Inoculum preparation
Five grams of each bone marrow sample was collected, finely minced using sterile scissors and forceps, and homogenized in a mortar with sterile sand using a pestle (16). The homogenate was suspended in phosphate-buffered saline to prepare a 10% (wt/vol) solution. This suspension underwent three freeze-thaw cycles to facilitate cell disruption, followed by two centrifugation cycles at 3,000 rpm for 10 minutes each. The supernatant was collected and treated with antibiotics and antimycotic agents to prevent contamination. The processed samples were then used for DNA extraction and subsequent molecular virus detection.
## Gross and histopathological investigation
The thymus, spleen, bursa of Fabricius, lungs, liver, and heart were examined for gross pathological changes and submitted to the Histopathology Laboratory in the Department of Surgery and Obstetrics at Bangladesh Agricultural University for detailed histopathological analysis. Each tissue sample was dissected, fixed in 10% formalin, embedded in paraffin wax, and processed using standard histological protocols, including hematoxylin and eosin staining (17). The stained sections were examined microscopically under an OLYMPUS CX41 microscope, and high-resolution photomi crographs were captured at the Department of Physiology, Bangladesh Agricultural University, Mymensingh, for further evaluation and documentation.
## DNA extraction and PCR amplification
DNA extraction from the prepared inoculum was performed using the TIANamp Virus DNA/RNA Kit (China), following the manufacturer's instructions. The extracted DNA was either immediately used for PCR or stored at -20°C for later use. The partial VP1 gene was amplified using PCR with the following primer set: VP1-F: 5′-ATG GCA AGA CGA GCT CGC-3′ and VP1-R: 5′-TCA GGG CTG CGT CCC CCA-3′, producing a 1,350 bp fragment (18). PCR amplification was performed in a 20 µL reaction mixture containing 1.0 µL (500 nM final concentration) of each forward and reverse primer, 6 µL of DNA template, 10 µL of Master Mix (TaKaRa Taq Version 2.0 plus dye), and 2 µL of double-distilled water. The reaction conditions were as follows: initial denaturation at 94°C for 5 minutes, 35 cycles of denaturation at 94°C for 1 minute, annealing at 59°C for 1 minute, and extension at 72°C for 1.5 minutes, a final extension at 72°C for 10 minutes, followed by storage at 4°C. For amplification of the partial VP2 gene, the primer set CAV-1 (5′-CTA AGA TCT GCA ACT GCG GA-3′) and CAV-2 (5′-CCT TGG AAG CGG ATA GTC AT-3′) was used to generate a 419 bp DNA fragment, following a previously published protocol (19). PCR-amplified products were separated on a 1.5% agarose gel prepared in 1 × TBE buffer and stained with ethidium bromide. Electrophoresis was performed at 100 V for 120 minutes, and the amplified products were visualized and analyzed using a GelDoc Go system (BioRad, USA).
## Whole-genome sequencing
The complete CAV viral genome was sequenced using next-generation sequencing (NGS) technology (Illumina NovaSeq 6000). The sequencing methodology was previ ously reported (20). Briefly, genomic DNA libraries were prepared using the Rapid Plus DNA Lib Prep Kit for Illumina (Cat# RK20208). Sequencing reads underwent quality assessment (FastQC v0.11.9), adapter and low-quality base trimming (Trimmomatic v0.39), and host DNA removal. Filtered reads were aligned to the reference genome of chicken anemia virus isolate CAVDL21 (NCBI accession no. OQ749509.1) using BWA (21). Circular genome confirmation was achieved through de novo assembly (Unicycler) and annotation with Prokka (v1.14.6) (22). The final assembly showed ~300 × coverage with overlapping termini, confirming circularity. The complete genome, designated CAV/ Narsingdi-1-MGH-BD, has been deposited in the NCBI databases.
## Phylogenetic tree construction and genotyping
A total of 38 whole-genome sequences of CAV strains were retrieved from GenBank using NCBI BLAST. Multiple sequence alignment of the reference sequences and CAV/ Narsingdi-1-MGH-BD was performed using MEGA11 software. A phylogenetic tree was constructed using pairwise distance matrices and the neighbor-joining method (23). The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) was annotated next to the branches, indicating confidence levels of inferred clusters (24). Genotyping was conducted following a previously published protocol (25). Complete coding regions of the VP1 gene of various CAV genotypes were retrieved from GenBank and used for genotyping.
## Mutational analysis and protein structure prediction
Reference sequences (accession nos.NC_001427.1 and AF395114.1) were retrieved from the NCBI database (https://www.ncbi.nlm.nih.gov/). Nucleotide substitutions in the VP1, VP2, and VP3 gene sequences of CAV/Narsingdi-1-MGH-BD were aligned with the retrieved reference sequences using CLC Sequence Viewer (Version 8.0). To identify amino acid substitutions, the translated sequences of VP1, VP2, and VP3 of CAV/ Narsingdi-1-MGH-BD were aligned with the reference sequences using CLC Sequence Viewer. The tertiary structures of VP1, VP2, and VP3 proteins were predicted using the Swiss-Model online server (https://swissmodel.expasy.org/) with default parameters. Protein sequences were submitted to the server, and models were built based on the highest identity and GMQE values. The structures were visualized using PyMOL software and validated using Procheck tools from SAVESv6.0 (https://saves.mbi.ucla.edu/) and the ProSA-web server (https://prosa.services.came.sbg.ac.at/prosa.php) by analyzing the Ramachandran plot and Z-score, respectively.
## RESULTS
## Clinical, gross, and histopathological findings
The affected chickens showed typical clinical symptoms of CAV infection, including anemia, depression, pale combs, and blue or cyanotic wings (Fig. 1A). The morbidity rate of the chickens was recorded at approximately 10%, as documented in the farm's record book. Gross pathological examinations revealed significant changes in the infected chicks compared to healthy ones. The infected chicks displayed hemorrhage in the subcutaneous layer, lungs, and scattered hemorrhagic spots in the heart (Fig. 1B, F, andH). A pale and slightly atrophic bursa of Fabricius was observed (Fig. 1E). Additionally, the infected chicks exhibited hypertrophic thymuses and pale-colored, swollen spleens and livers, whereas healthy chickens showed no such lesions (Fig. 1C, D andE). Histopa thological examinations revealed massive lymphoid depletion in the spleen, thymus, and bursa (Fig. 2A, B andC). Necrosis in the spleen, cortical thinning of the thymus, and atrophy with structural disorganization of bursal follicles were observed (Fig. 2A, B andC). Necrosis and degeneration of hepatocytes in the liver and myocardial tissue in the heart were noted (Fig. 2D andF). The lungs exhibited severe congestion and hemorrhage (Fig. 2E).
## Molecular detection and whole-genome sequencing
Viral DNA extracted from the bone marrow of suspected CAV-infected and healthy chickens was analyzed using gene-specific primers for the VP1 and VP2 genes. Gel electrophoresis and subsequent UV transillumination confirmed that eight out of ten bone marrow samples exhibited target-specific bands of 1,350 bp and 419 bp, respectively. As expected, no viral DNA bands were detected in the bone marrow of healthy chickens. One PCR-positive sample, exhibiting strong VP1 (1,350 bp) and VP2 (419 bp) bands indicative of high viral load, was selected for whole-genome sequenc ing using NGS; the corresponding chicken displayed typical clinical signs and severe histopathological lesions of CAV. The complete CAV genome was 2,330 bp with 42% GC content and contained three coding regions encoding VP1 (1,350 bp), VP2 (651 bp), and VP3 (366 bp), consistent with the complete CAV genomes previously reported in the literature (2,298-2,319 bp) (10). The complete genome sequence of the CAV isolate (CAV/ Narsingdi-1-MGH-BD) has been assigned the GenBank, BioSample, and SRA accession nos. PQ412955, SAMN44002588, and PRJNA1167429, respectively (20).
## Phylogenetic analysis and genotypic characterization
A phylogenetic tree was constructed based on multiple sequence alignments of the complete nucleotide sequence of CAV/Narsingdi-1-MGH-BD and 37 reference strains retrieved from GenBank. The tree indicated that the identified CAV strain was closely related to the Chinese strain (OQ749509.1) and formed an independent cluster (Fig. 3A). Genotyping based on VP1 gene sequences identified CAV/Narsingdi-1-MGH-BD as genotype IIIb, clustering within the CAV genotype IIIb group (Fig. 3B).
## Mutational analysis of CAV/Narsingdi-1-MGH-BD compared to reference strains
Mutational analysis of the complete genome was performed in comparison with the NCBI reference strain and BD-3, the only previously reported complete CAV genome from Bangladesh, identified in 2004. Various synonymous and nonsynonymous nucleotide substitutions were detected in the VP1, VP2, and VP3 genes of CAV/Narsingdi-1-MGH-BD compared to both the NCBI reference strain and BD-3 (Fig. 4A through C). For VP1, multiple nucleotide substitutions were identified (Fig. 4A), but these were all synony mous mutations compared with the NCBI reference strain (NC_001427.1), resulting in no amino acid changes. However, comparison with BD-3 revealed nine nonsynonymous mutations resulting in amino acid substitutions (Fig. 5A). For VP2, nucleotide variations were identified that resulted in a single nonsynonymous mutation (A153V) in compar isons with the reference strain and BD-3 (Fig. 5B). VP3 showed the highest rate of nonsynonymous mutations with five amino acid substitutions (E1V, P2A, T9M, F26L, and K44E) compared to the NCBI reference strain, while only two nonsynonymous changes (F26L and A38V) were observed relative to BD-3 (Fig. 5C).
## Physicochemical properties and predicted tertiary structures of viral proteins
The ProtParam program was used to characterize the physicochemical properties of the viral proteins, including the number of amino acids, molecular weight (MW), isoelectric point (pI), instability index (II), aliphatic index (AI), and grand average of hydropathicity (GRAVY). The findings indicated slight variations in the VP1, VP2, and VP3 proteins of CAV/Narsingdi-1-MGH-BD compared to reference sequences, except for VP3, which exhibited greater stability (Table 1). Comparative analysis of amino acid sequences between CAV/Narsingdi-1-MGH-BD and BD-3 revealed significant differences that may affect protein function. The nine amino acid substitutions in VP1 compared to BD-3 alter the physicochemical properties, with BD-3 showing a higher AI (57.92 vs 54.31), suggesting differences in thermostability. The single A153V mutation in VP2 is particu larly significant as valine substitution introduces a more hydrophobic residue that may affect the protein's phosphatase activity. For VP3, the two mutations (F26L and A38V) compared to BD-3 result in dramatic differences in protein stability, with BD-3 VP3 showing an II of 145.4 compared to 42.68 in our strain (Table 1), indicating our strain's VP3 is significantly more stable. The Ramachandran plot analysis (Table 3) shows that our strain's VP3 has only 66.90% of residues in the most favorable region compared to 100% for BD-3, suggesting conformational differences that may affect apoptin function. The predicted tertiary structures of VP1, VP2, and VP3 proteins in CAV/Narsingdi-1-MGH-BD displayed minor structural differences compared to reference strains due to amino acid substitutions (Fig. 6, Tables 2 and3).
## DISCUSSION
CAV is a significant pathogen affecting poultry, primarily targeting hematopoietic and immune system components (7,26,27). This study presents the clinical, pathologi cal, and molecular characterization of an identified CAV strain, including complete genome sequencing, genotyping, and structural analyses of viral proteins. These findings underscore the mechanisms of viral infection and the genomic evolution of CAV within the Bangladeshi poultry population.
CAV primarily infects hemocytoblasts in the bone marrow and precursor T cells in the thymus, resulting in aplastic anemia and immunosuppression (28). Clinical signs such as anemia, depression, pale comb, and cyanotic wings indicate severe hemato poietic disruption due to viral replication in bone marrow cells. CAV typically induces anemia and immunosuppression in young chickens, characterized by pallor, atrophy of lymphoid organs (thymus, spleen, and Bursa of Fabricius), and a swollen, mottled liver (29). However, in this study, hypertrophic pallor spleens and thymuses were observed atypically. Concurrent infections with other pathogens, such as fowl adenoviruses, can alter the typical disease progression of CAV by immunosuppression (30). Although CAV is classified as a single serotype, genotyping remains epidemiologically and clinically relevant for several reasons. Different genotypes exhibit varying levels of pathogenic ity; for example, genotype III strains have been associated with more severe clinical disease, even in the presence of maternal antibodies (25,31,32). A Chinese study further demonstrated that mortality rates can vary significantly among different CAV strains under experimental conditions (33). Experimental studies have demonstrated that genotype III CAV infections can cause swollen, pale livers and spleens in infected chicks (34). Additionally, hemorrhagic lungs and hearts observed in CAV-infected chickens in this study suggest the possibility of concurrent infections, although specific pathogens were not investigated. Variations in the virulence of CAV strains could also contribute to differing pathological outcomes (34). The identification of genotype IIIb in our study, reported for the first time in Bangladesh and distinct from the previously described genotype II (BD-3), highlights the need to evaluate whether current vaccine strains provide optimal protection against circulating field viruses. Despite the availability of vaccines, the high seroprevalence reported in Bangladesh ( 14) suggests that vaccination alone may not be sufficient for effective CAV control. Our detection of a genotype IIIb strain with unique mutations, particularly in the stable VP3 protein, underscores the importance of integrating vaccination strategies with stringent biosecurity measures. Histopathological examinations confirmed significant lymphoid depletion in the thymus, bursa of Fabricius, and spleen, indicating impaired T-cell maturation and reduced immune competence (29). The VP3 protein, also known as apoptin, plays a crucial role in inducing apoptosis in hematopoietic precursor cells, leading to reduced erythropoiesis and subsequent anemia (35,36). Apoptin-mediated apoptosis is triggered via the mitochondrial pathway, interacting with cellular apoptotic regulators such as Bcl-2, caspase-9, and caspase-3 (37). The observed histopathological changes in the spleen and thymus are consistent with apoptosis-mediated lymphoid depletion, confirming the immunosuppressive nature of CAV infection.
The molecular detection of viral DNA in the bone marrow of infected chickens confirms the hematopoietic tropism of CAV. Complete genome sequencing of CAV/ Narsingdi-1-MGH-BD revealed a genome length of 2,330 bp, which is distinctly different from the previously identified BD-3 strain (2,298 bp) from Bangladesh. The 2,330 bp genome size falls within the typical range for complete CAV genomes (2,298-2,319 bp) (10). Variations in reported CAV genome sizes primarily result from differences in non-coding regulatory regions, particularly the region between the polyadenylation signal and the start of VP1. Despite these differences, both CAV/Narsingdi-1-MGH-BD and BD-3 encode the three key structural proteins-VP1, VP2, and VP3. Phylogenetic analysis indicated that CAV/Narsingdi-1-MGH-BD is closely related to a Chinese strain (OQ749509.1) and clusters within genotype IIIb, whereas BD-3 belongs to genotype II (25). This finding suggests a potential epidemiological link between Bangladeshi and Chinese CAV strains, likely facilitated by poultry trade and movement. The genotypic Mutational analysis revealed several synonymous and nonsynonymous nucleotide substitutions in viral genes, along with amino acid changes in viral proteins, compared to both the NCBI reference strain and BD-3, the only previously sequenced CAV strain from Bangladesh (38). The nine amino acid differences in VP1 between our strain and BD-3 represent substantial genetic divergence within Bangladeshi CAV populations. These mutations may indicate adaptation to local conditions or the introduction of divergent strains. Notably, our strain shows no VP1 amino acid changes compared with the NCBI reference strain, whereas BD-3 exhibits multiple differences, suggesting that our strain may be closer to the ancestral CAV lineage in terms of VP1 sequence. In addition, computational analysis of physicochemical properties indicated that the VP1, VP2, and VP3 proteins of CAV/Narsingdi-1-MGH-BD exhibit minor variations in MW, and hydropathicity compared to reference strains. The marked difference in VP3 stability between our strain (II 42.68) and BD-3 (145.4) is particularly noteworthy. VP3 (apoptin) plays a key role in inducing apoptosis in infected cells, and the higher stability observed in our strain may contribute to more efficient apoptosis induction and potentially greater pathogenicity (39). Predicted tertiary structures of VP1, VP2, and VP3 revealed slight conformational differences, likely influenced by amino acid substitutions. Structural variations in VP1 may affect antigenicity and immune recognition, while changes in VP2 could alter its phosphatase activity, potentially impacting immune modulation (10,11,29). The VP3 mutations may enhance apoptotic efficiency, exacerbating immunosup pression and disease severity in infected chickens (29,35).
In conclusion, this study highlights the clinical, pathological, and molecular char acteristics of a newly identified CAV strain in Bangladesh. The findings confirm that CAV/Narsingdi-1-MGH-BD exhibits significant genetic variation compared to previously reported strains. Phylogenetic analysis establishes its close relation to a Chinese strain, placing it within genotype IIIb. The observed mutations, particularly in VP3, may have implications for viral pathogenesis and immune evasion. Further investigation is needed to conduct large-scale epidemiological studies and assess the functional impact of these genetic variations on virulence and vaccine efficacy. These findings contribute to a deeper understanding of CAV evolution and its epidemiological significance in poultry health management.
## References
1. Rosenberger, Cloud (1998) "Chicken anemia virus" *Poult Sci*
2. Schat (2009) "Chicken anemia virus"
3. Jørgensen, Otte, Nielsen et al. (1995) "Influence of subclinical virus infections and other factors on broiler flock perform ance" *Br Poult Sci*
4. Sun, Yu, Jiang et al. (2023) "Molecular characterization of chicken infectious anaemia virus (CIAV) in China during 2020-2021" *Avian Pathol*
5. Okay, Aşkar (2021) "Molecular characterization of VP2 and VP3 proteins of chicken anemia virus isolates in Turkey" *Turkish Journal of Veterinary and Animal Sciences*
6. Techera, Marandino, Tomás et al. (2021) "Origin, spreading and genetic variability of chicken anaemia virus" *Avian Pathol*
7. Fatoba, Adeleke (2019) "Chicken anemia virus: a deadly pathogen of poultry" *Acta Virol*
8. Van Santen, Li, Hoerr et al. (2001) "Genetic characteriza tion of chicken anemia virus from commercial broiler chickens in Alabama" *Avian Dis*
9. Rosario, Breitbart, Harrach et al. (2017) "Revisiting the taxonomy of the family Circoviridae: establish ment of the genus Cyclovirus and removal of the genus Gyrovirus" *Arch Virol*
10. Noteborn, De Boer, Van Roozelaar et al. (1991) "Characterization of cloned chicken anemia virus DNA that contains all elements for the infectious replication cycle" *J Virol*
11. Peters, Jackson, Crabb et al. (2002) "Chicken anemia virus VP2 is a novel dual specificity protein phosphatase" *J Biol Chem*
12. Noteborn (2004) "Chicken anemia virus induced apoptosis: underlying molecular mechanisms" *Vet Microbiol*
13. Goryo, Sugimura, Matsumoto et al. (1985) "Isolation of an agent inducing chicken anaemia" *Avian Pathol*
14. Kabir, Saha, Hossain et al. (2021) "Serological survey on the prevalence of chicken infectious anemia virus in broiler breeder and layer farms in some selected areas of Bangladesh" *J Adv Vet Anim Res*
15. Ou, Lin, Liu et al. (2018) "Epidemiology and molecular characterization of chicken anaemia virus from commercial and native chickens in Taiwan" *Transbound Emerg Dis*
16. Mou, Hasan, Mozumder et al. (2025) "Distinct amino acid substitutions in the EEV glycoprotein and DNA-dependent RNA polymerase of lumpy skin disease virus identified in wetland areas of Bangladesh" *Res Vet Sci*
17. (1016)
18. Hossain, Pathan, Hasan et al. (2024) "Molecular detection and genetic characterization of avian leukosis virus from field outbreaks in Bangladesh" *Vet Med Sci*
19. Sreekala, Gurpreet, Dwivedi (2020) "Detection and molecular characterization of chicken infectious anaemia virus in young chicks in Punjab region of north-western India" *Braz J Microbiol*
20. Tongkamsai, Lee, Cheng et al. (2019) "Persistent Infection with chicken anemia virus in 3-week-old chickens induced by inoculation of the virus by the natural route" *Pathogens*
21. Akter, Hasan, Jeba et al. (2025) "Complete genome sequence of chicken anemia virus from a field outbreak in Bangladesh" *Microbiol Resour Announc*
22. Li, Durbin (2009) "Fast and accurate short read alignment with Burrows-Wheeler transform" *Bioinformatics*
23. Seemann (2014) "Prokka: rapid prokaryotic genome annotation" *Bioinformatics*
24. Saitou, Nei (1987) "The neighbor-joining method: a new method for reconstructing phylogenetic trees" *Mol Biol Evol*
25. Felsenstein (1985) "Confidence limits on phylogenies: an approach using the bootstrap" *Evolution*
26. Kim, Kwon, Bae et al. (2010) "Molecular characteriza tion of chicken infectious anemia viruses detected from breeder and broiler chickens in South Korea" *Poult Sci*
27. Wani, Dhama, Malik (2016) "Impact of virus load on immunocyto logical and histopathological parameters during clinical chicken anemia virus (CAV) infection in poultry" *Microb Pathog*
28. Fang, Jia, Hu et al. (2023) "Molecular characterization and pathogenicity study of a highly pathogenic strain of chicken anemia virus that emerged in China" *Front Cell Infect Microbiol*
29. Yan, Song, Zhang et al. (2025) "Chicken anemia virus inhibits hematopoiesis and development of chicken embryo" *Poult Sci*
30. Adair (2000) "Immunopathogenesis of chicken anemia virus infection" *Dev Comp Immunol*
31. Abghour, Mouahid, Darkaoui et al. (2021) "Pathogenicity of field strain of fowl aviadenovirus serotype 11 isolated from chickens with inclusion body hepatitis in Morocco" *PLoS One*
32. Song, Kim, Kwon et al. (2024) "Genetic characterization of chicken infectious anaemia viruses isolated in Korea and their pathogenicity in chicks" *Front Cell Infect Microbiol*
33. Abdel-Mawgod, Zanaty, Elhusseiny et al. (2024) "Genetic heterogeneity of chicken anemia virus isolated in selected Egyptian provinces as a preliminary investigation" *Front Vet Sci*
34. Li, Yan, Wang et al. (2021) "Molecular evolution and pathogenicity of chicken anemia virus isolates in China" *Arch Virol*
35. Elsamadony, Mekky, Fedawy et al. (2025) "Genetic differences and pathogenicity of chicken anemia virus strains in broiler's baby chicks" *Egyptian Journal of Veterinary Sciences*
36. Malla, Arora, Khan et al. (2020) "Apoptin as a tumor-specific therapeutic agent: current perspective on mechanism of action and delivery systems" *Front Cell Dev Biol*
37. Feng, Liang, Teodoro (2020) "The role of apoptin in chicken anemia virus replication" *Pathogens*
38. Kang, Reynolds (2009) "Bcl-2 inhibitors: targeting mitochondrial apoptotic pathways in cancer therapy" *Clin Cancer Res*
39. Islam, Johne, Raue et al. (2002) "Sequence analysis of the full-length cloned DNA of a chicken anaemia virus (CAV) strain from Bangladesh: evidence for genetic grouping of CAV strains based on the deduced VP1 amino acid sequences" *J Vet Med B Infect Dis Vet Public Health*
40. Craig, Rimondi, Delamer et al. (2009) "Molecular characterization of chicken infectious anemia virus circulating in Argentina during 2007" *Avian Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645960&blobtype=pdf | # African swine fever virus hijacks host pyrimidine metabolism to promote viral replication
Zebu Song, Yilin Chen, Hui Guo, Guihong Zhang, Lang Gong, Zezhong Zheng
## Abstract
African swine fever (ASF) is a highly contagious disease of pigs caused by the African swine fever virus (ASFV), posing a significant threat to global swine produc tion. As an obligate intracellular parasite, ASFV relies on host metabolic networks to fulfill its replication requirements. However, the precise mechanisms by which it manipulates nucleotide metabolism remain unclear. In this study, untargeted metabolomic analysis of ASFV-infected porcine alveolar macrophages revealed significant perturbations in purine and pyrimidine metabolism, glycolysis, the pentose phosphate pathway (PPP), and the glutamate and aspartate metabolic pathways. Functional validation demonstrated that ASFV depends on de novo pyrimidine biosynthesis for viral genome replication. Notably, ASFV employs a dual strategy to sustain the supply of nucleotide precursors: (i) it hijacks the PPP to generate ribose-5-phosphate and NADPH for redox balance, and (ii) it enhances glutamine uptake and catabolism to provide the nitrogen and carbon needed for nucleotide biosynthesis and tricarboxylic acid cycle replenishment. Furthermore, although aspartate is essential for pyrimidine synthesis, ASFV circumvents dependence on extracellular aspartate by activating a cytosolic GOT1-mediated synthesis pathway. Collectively, these findings elucidate how ASFV reprograms host nucleotide metabolism to support its replication, offering new insights into virus-host metabolic interactions and identifying potential targets for antiviral therapy. IMPORTANCE African swine fever (ASF) is a devastating disease that causes substantial economic losses in the global pig industry. This study demonstrates that the African swine fever virus (ASFV) reprograms host cell metabolism to produce the essential building blocks required for its replication. Specifically, ASFV manipulates host nucleo tide biosynthetic pathways to secure both the substrates for DNA synthesis and the reducing power necessary to mitigate oxidative stress. Elucidating these metabolic interactions not only deepens understanding of ASFV pathogenesis but also highlights promising metabolic targets for antiviral therapy. By elucidating how ASFV hijacks nucleotide biosynthesis within infected cells, our findings pave the way for innovative strategies to combat ASF.
ASFV manipulates host metabolism is critical not only for elucidating its replication mechanisms but also for identifying new antiviral targets.
Nucleotide metabolism is a prime example of a critical host pathway exploited by viruses. Nucleotides, the basic building blocks of both DNA and RNA, are indispensa ble for viral genome replication. Several viruses have been shown to enhance their replication by activating either the de novo nucleotide biosynthesis or the salvage pathways (5). However, it remains unclear whether ASFV modulates host nucleotide metabolism to meet its replication requirements and which mechanisms or key regulatory nodes are involved.
Nucleotides consist of a nitrogenous base, a pentose sugar, and phosphate groups. Glucose is funneled into the pentose phosphate pathway (PPP), which produces ribose-5-phosphate (R5P), a key precursor for nucleotide biosynthesis. PPP comprises an oxidative branch that generates NADPH-crucial for maintaining cellular redox homeostasis-and a non-oxidative branch that interconverts sugars (6). Several viruses have been reported to hijack the oxidative PPP (oxPPP) to secure the necessary building blocks for replication (7)(8)(9). Moreover, NADPH generated by the oxPPP enhances viral antioxidant capacity, thereby facilitating viral replication (10). Importantly, our prior work revealed that intracellular reactive oxygen species (ROS) levels increase significantly following ASFV infection (11), suggesting a potential role for oxPPP-derived NADPH in counteracting oxidative stress during ASFV replication.
In addition to glucose metabolism, amino acid metabolism plays a pivotal role in nucleotide biosynthesis. Glutamine is a key amino acid and the second major energy substrate after glucose. As a vital source of both carbon and nitrogen, glutamine can directly donate its γ-nitrogen for nucleotide synthesis once internalized. Furthermore, it can be converted into glutamate and subsequently into α-ketoglutarate (α-KG) to fuel the tricarboxylic acid (TCA) cycle or donate nitrogen via transamination for the synthesis of other non-essential amino acids (12). Cellular glutamine uptake is mediated by solute carrier (SLC) transporters, including SLC1A5, SLC38A1, and SLC38A2 (13). Recent studies have shown that viral infection can activate these glutamine transporters, thereby accelerating glutamine uptake (14) and upregulating relevant metabolic enzymes to enhance glutamine catabolism for nucleotide biosynthesis (15). However, whether ASFV similarly enhances glutamine uptake and catabolism remains to be fully elucidated.
Aspartate, a non-essential amino acid, is an indispensable substrate for constructing the pyrimidine ring and thus plays a key role in nucleotide synthesis and cell proliferation (16). Aspartate can be acquired through extracellular uptake or de novo synthesis. De novo aspartate synthesis is primarily driven by TCA cycle replenishment: mitochondrial oxaloacetate (OAA) is produced and converted into aspartate by mitochondrial GOT2, or alternatively, cytosolic GOT1 converts OAA into aspartate through reductive glutamine metabolism. When mitochondrial-dependent aspartate biosynthesis is impaired, GOT1 plays a crucial role in maintaining aspartate production and supporting cell proliferation (17). Previous studies have reported that defects in endogenous aspartate synthesis can suppress tumor growth (18), and Liu et al. found that foot-and-mouth disease virus enhances aspartate uptake via the SLC38A8 transporter to promote its replication (19). However, the role of aspartate in ASFV replication remains unclear, and further investiga tion is needed to elucidate how ASFV utilizes aspartate to support its life cycle.
In this study, we investigate how ASFV reprograms host nucleotide metabolism, focusing on the PPP, glutamine uptake and catabolism, and aspartate biosynthesis from a metabolic perspective. Our findings provide new insights into the metabolic interplay between ASFV and host cells and identify potential targets within nucleotide metabo lism.
## RESULTS
## ASFV infection remodels nucleotide metabolic networks to potentially facilitate replication
Nucleotides play a critical role in viral replication, and the metabolic pathways governing their synthesis have become key targets in antiviral drug development (20). To inves tigate ASFV-induced metabolic perturbations, we reanalyzed our previous untargeted metabolomics data from ASFV-infected porcine alveolar macrophages (PAMs) across different infection stages (see the supplemental material). Notably, ASFV infection caused stage-specific accumulation and depletion of nucleotide precursors, including R5P and aspartate, as well as intermediates associated with nucleotide biosynthesis (Fig. 1A). These findings suggest that ASFV may facilitate its replication by modulating host nucleotide metabolic pathways. To further examine these changes, we performed pathway enrichment analysis using MetaboAnalyst version 5.0 (www.metaboanalyst.ca) to systematically identify pathways affected by ASFV infection. As shown in Fig. 1B, purine and pyrimidine metabolism pathways were consistently enriched across all infection stages (3, 12, and 24 hours post-infection [hpi]), along with related pathways such as the PPP and aspartate metabolism. Collectively, this spatiotemporal metabolic rewiring positions nucleotide metabolism as a central element of ASFV replication, highlighting potential targets for therapeutic intervention.
## ASFV replication selectively depends on pyrimidine nucleotide synthesis
To determine whether ASFV replication relies on de novo synthesis of purine, pyrimi dine, or both types of nucleotides, we evaluated the effects of AVN-944, an inhibi tor of inosine monophosphate dehydrogenase (IMPDH, a key enzyme in the purine pathway), and brequinar, an inhibitor of dihydroorotate dehydrogenase (DHODH, a key enzyme in the pyrimidine pathway), on ASFV replication. As shown in Fig. 2C and Fig. S2A andB, various concentrations of AVN-944 did not affect the replication of the ASFV, whereas brequinar inhibited ASFV replication in a dose-dependent manner (Fig. 2D; Fig. S2C andD). To further determine whether brequinar suppresses ASFV replication by depleting intracellular pyrimidine levels, we supplemented infected cells with exogenous uridine to evaluate whether ASFV replication could be rescued via the pyrimidine salvage pathway. Notably, exogenous uridine restored ASFV replication reduced by brequinar treatment, suggesting that brequinar inhibited ASFV replication by depleting intracellular pyrimidine pools (Fig. 2E; Fig. S2E). These findings demonstrate that ASFV replication critically depends on pyrimidine nucleotide biosynthesis while being independent of de novo purine synthesis under the conditions tested.
## PPP coordinates antiviral redox balance and nucleotide metabolism to promote ASFV replication
Our results demonstrate that ASFV replication depends on de novo pyrimidine nucleo tide synthesis, which involves two essential components: the provision of R5P and the construction of the pyrimidine ring (Fig. 3A). The PPP supplies R5P, a key precursor for nucleotide synthesis. To investigate the role of the PPP in ASFV replication, we treated infected cells with RRx-001 (Fig. 3B), an inhibitor of glucose-6-phosphate dehydrogenase (G6PD), the rate-limiting enzyme of the PPP. Dose-response assays revealed a significant reduction in ASFV-P30 protein expression (Fig. 3C), indicating that PPP activity is essential for ASFV replication. In addition, the oxPPP generates NADPH, which enhances viral antioxidant capacity and supports DNA synthesis (10). Our previous study found that ASFV infection increases intracellular ROS levels (11); therefore, we hypothesized that ASFV selectively hijacks the oxPPP to generate NADPH and establish a favorable environment for DNA synthesis under oxidative stress. Consistent with this hypothesis, infected cells exhibited elevated NADPH/NADP + ratios (Fig. 3D), while RRx-001 treatment abrogated this increase (Fig. 3D), suggesting that viral replication requires oxPPP-derived Full-Length Text NADPH to maintain redox homeostasis. To determine whether PPP supports ASFV replication via NADPH-mediated antioxidant defense, nucleotide precursor provision, or both, we performed functional rescue experiments with N-acetylcysteine (NAC, an antioxidant) and exogenous nucleosides. Both treatments partially restored viral replication following RRx-001 inhibition (Fig. 3E andF), indicating that the PPP facilitates ASFV replication through dual mechanisms: NADPH-dependent redox regulation and nucleotide precursor supply.
## ASFV replication depends on glutamine uptake and induces dynamic remodeling of the host glutamine pool
De novo pyrimidine nucleotide biosynthesis requires not only R5P, supplied by the PPP, but also glutamine, which donates the γ-nitrogen for pyrimidine ring assembly (Fig. 3A). This positions glutamine as a key metabolic checkpoint in ASFV replication.
To investigate the role of glutamine in ASFV replication, we cultured infected cells in glutamine-depleted medium, then supplemented them with varying concentrations of glutamine and assessed ASFV replication by measuring viral titers. The results showed a strong dependence on glutamine: ASFV titers were significantly reduced under glutamine-free conditions but restored upon glutamine supplementation (Fig. 4A). This confirms glutamine as an essential metabolic substrate for ASFV replication.
To further elucidate the impact of ASFV infection on glutamine utilization, we measured glutamine levels in both the culture supernatant and within cells following ASFV infection of PAMs. Our data showed that ASFV infection dose-dependently reduced glutamine concentrations in the culture supernatant (Fig. 4B). However, intracellular glutamine levels initially increased 3 hpi, followed by a marked decline (Fig. 4C). Collec tively, these findings suggest that ASFV infection not only enhances the uptake of extracellular glutamine but also promotes glutamine catabolism during its replication.
## ASFV exploits dual glutamine metabolic pathways to sustain nucleotide biosynthesis and TCA cycle replenishment for viral replication
Glutamine serves as a critical source of both carbon and nitrogen for biosynthesis. Upon cellular uptake, its γ-nitrogen is directly utilized for nucleotide biosynthesis, while its carbon skeleton is converted into glutamate and further metabolized into α-KG to replenish the TCA cycle for energy and biosynthesis (Fig. 4D). To determine whether ASFV-induced glutamine catabolism operates through these pathways, we used glutamine-deficient medium as a baseline and supplemented it separately with either nucleosides or α-KG. Notably, nucleoside supplementation restored ASFV replication (Fig. 4E), demonstrating that glutamine-derived nitrogen supports de novo nucleotide synthesis essential for viral propagation. Concurrently, α-KG supplementation also restored viral replication efficiency (Fig. 4F), suggesting that glutamine-derived carbon flux may sustain TCA cycle activity, which could contribute to meeting biosynthetic and energetic demands. These findings collectively suggest that ASFV may redirect gluta mine catabolism through two distinct pathways: one that directly incorporates nitrogen into nucleotide precursors and another that channels carbon into the TCA cycle via α-KG, thereby facilitating viral replication and maintaining host metabolic support.
## ASFV hijacks SLC1A5 to enhance glutamine-fueled nucleotide synthesis and drive viral proliferation
While our previous findings established ASFV's reliance on glutamine catabolism, the mechanism underlying its enhanced glutamine uptake remained unclear. Given that glutamine transport is mediated by SLC proteins-membrane-bound transporters critical for amino acid homeostasis (21)-we hypothesized that ASFV might selectively upregulate glutamine-specific SLC transporters to meet its biosynthetic demands, similar to strategies employed by other viruses (22,23). The primary glutamine transporters include SLC1A5, SLC38A1, and SLC38A2 (24). To determine which of these are involved in ASFV-induced glutamine metabolism, we evaluated the mRNA expression levels of these transporters during ASFV infection. Our results showed that SLC1A5 mRNA expression was significantly upregulated, whereas SLC38A1 and SLC38A2 expression remained unchanged (Fig. 5A). Moreover, treatment with the SLC1A5-specific inhibitor V-9302 or siRNA-mediated knockdown of SLC1A5 significantly inhibited ASFV replication (Fig. 5B; Fig. S3A andB). Importantly, supplementation with exogenous nucleoside restored viral replication under SLC1A5 inhibition (Fig. 5C), directly linking SLC1A5-driven glutamine uptake to nucleotide biosynthesis. Taken together, these findings suggest that ASFVinduced upregulation of SLC1A5 facilitates increased glutamine uptake, thereby supporting the nucleotide biosynthesis essential for viral replication.
## ASFV replication is unaffected by exogenous aspartate deprivation
Virus-induced alterations in amino acid metabolism provide insight into how viruses reprogram host resources to support replication. Previous studies have reported that ASFV disrupts normal amino acid metabolism, including that of aspartate (4). Aspartate is also a crucial substrate for pyrimidine ring assembly (Fig. 3A), and it can be supplied either through extracellular uptake or intracellular de novo synthesis. To preliminarily examine ASFV's aspartate utilization strategy, we cultured infected cells in aspartatedeficient medium and performed supplementation experiments using varying concen trations of aspartate. Notably, the absence of exogenous aspartate did not affect the replication of ASFV (Fig. 6A; Fig. S3C through E), indicating that ASFV replication is resilient to extracellular aspartate deprivation. These findings suggest that ASFV may rely primarily on intracellular de novo aspartate synthesis to meet its replication needs.
## De novo aspartate synthesis sustains ASFV replication through compartmen talized metabolic regulation
Given ASFV's independence from extracellular aspartate, we investigated whether viral replication relies on endogenous de novo aspartate biosynthesis. Two major pathways contribute to intracellular aspartate production. Specifically, glutamine is converted by glutaminase into glutamate and subsequently into α-KG. Within mitochondria, α-KG is processed through the TCA cycle to generate OAA, which is then converted to aspartate by mitochondrial GOT2. Alternatively, α-KG can undergo reductive carboxylation to form citrate, which is exported to the cytosol, converted into OAA, and finally transformed into aspartate by cytosolic GOT1 (Fig. 6B).
To examine the contribution of these pathways, we initially treated infected cells with aminooxyacetic acid hemihydrochloride (AOA), a pan-transaminase inhibitor. Unexpect edly, AOA at non-toxic concentrations did not significantly inhibit viral replication (Fig. 6C), suggesting that broad-spectrum transaminase inhibition alone is insufficient to disrupt the compensatory mechanisms regulating aspartate biosynthesis in infected cells. This suggests the presence of redundant transaminase activity or activation of alternative pathways, ensuring the continued aspartate production to support viral replication. To further dissect compartment-specific roles, we employed siRNA-medi ated knockdown of GOT1 or GOT2. Knockdown of GOT1 markedly suppressed ASFV replication (Fig. 6D andE), and aspartate supplementation partially restored replication efficiency (Fig. 6F), confirming that cytosolic aspartate synthesis is critical for viral propagation. Paradoxically, silencing GOT2 enhanced viral replication (Fig. 6G andH), suggesting that inhibition of mitochondrial aspartate flux may promote viral prolifera tion through currently undefined mechanisms. Collectively, these findings demonstrate that ASFV replication critically depends on de novo aspartate synthesis, primarily via the cytosolic GOT1 pathway. However, the counterintuitive enhancement of replication following GOT2 suppression highlights a regulatory role for mitochondrial aspartate metabolism, underscoring the need for further investigation into the compartmentalized regulation of GOT1 and GOT2 during ASFV infection.
## Temporal reciprocal regulation of GOT1 and GOT2 directs aspartate flux essential for ASFV replication
To elucidate ASFV's regulatory strategy over GOT1 and GOT2, we mapped the kinet ics of their protein expression at various time points post-infection. As shown in Fig. 7A, GOT1 exhibited a biphasic expression pattern: levels gradually increased during early-to-mid infection (3-12 hpi), peaked at 12 hpi, and declined to basal levels during late infection (18-24 hpi). In contrast, GOT2 displayed an inverse trend, with its expression decreasing following infection and subsequently increasing after 12 hpi (Fig. 7B). This reciprocal expression pattern suggests that ASFV strategically prioritizes cytosolic aspartate synthesis via GOT1 during active genome replication (318 hpi) while simultaneously suppressing mitochondrial aspartate flux through GOT2 to minimize metabolic competition. Given that GOT1 knockdown inhibited ASFV replication, whereas GOT2 knockdown enhanced it, we further investigated potential functional crosstalk between these enzymes. Our results showed that siRNA-mediated silencing of GOT2 led to a compensatory upregulation of GOT1 (Fig. 7C andD), indicating that when mitochon drial aspartate production is impaired, ASFV may rebalance aspartate synthesis through the cytosolic route. This compensatory mechanism likely explains why GOT2 depletion paradoxically enhances ASFV replication, as enhanced GOT1 activity amplifies cytosolic aspartate production to support viral nucleotide biosynthesis. Finally, to directly link GOT1-derived aspartate to viral nucleotide biosynthesis, we performed an exogenous nucleoside supplementation assay following GOT1 knockdown using siRNA. As shown in Fig. 7E, exogenous nucleoside supplementation partially restored ASFV replication. This result confirms that GOT1-mediated aspartate synthesis primarily supports de novo nucleotide production rather than general metabolic homeostasis, underscoring its importance in ASFV replication.
## DISCUSSION
In this study, we elucidated how ASFV hijacks host nucleotide metabolism to support its replication by reprogramming key metabolic pathways. Our untargeted metabolo mics analysis of ASFV-infected PAMs revealed significant perturbations in purine and pyrimidine metabolism, glycolysis, the PPP, and amino acid pathways, including those of glutamine and aspartate. Functional experiments confirmed that ASFV replication critically depends on de novo pyrimidine biosynthesis. Notably, our findings show that ASFV employs a dual strategy to secure nucleotide precursors: (i) it activates the PPP to generate R5P and maintain NADPH-mediated redox homeostasis, and (ii) it promotes glutamine uptake and catabolism to supply both the nitrogen and carbon required for nucleotide synthesis and TCA cycle replenishment. Furthermore, our results indicate that ASFV induces metabolic adaptations in aspartate synthesis, favoring intracellu lar production via cytosolic GOT1 while remaining resilient to extracellular aspartate deprivation. Collectively, these findings provide novel insights into the metabolic interplay between ASFV and its host cells and identify potential metabolic vulnerabilities that could be targeted for antiviral therapy.
Our results are consistent with numerous studies demonstrating that viruses reprogram host metabolic pathways to meet the demands of replication. Research on viruses such as Epstein-Barr virus, Kaposi's sarcoma-associated herpesvirus, coxsackie virus B3, vaccinia virus, and has shown that these pathogens hijack host nucleotide biosynthesis for their replication (25)(26)(27)(28)(29). Similarly, ASFV appears to selectively manipu late host pyrimidine nucleotide metabolism. The observed dependence of ASFV on de novo pyrimidine synthesis is supported by our data showing that inhibition of DHODH with brequinar, but not inhibition of purine synthesis with AVN-944, significantly reduces viral replication. As for why ASFV replication tends to utilize the synthesis of pyrimidine nucleotides, we speculate that de novo purine nucleotide biosynthesis is more energyintensive due to the complex assembly of cyclic structures and multi-step regulatory processes, whereas pyrimidine biosynthesis is relatively more energy-efficient. This distinction could play a pivotal role in the metabolic "tug-of-war" between ASFV and its host. Notably, our previous study demonstrated that brequinar inhibits ASFV replication by activating ferroptosis, a regulated form of cell death, highlighting its potential as an antiviral agent (30). However, that study did not address the metabolic prerequisites that render ASFV susceptible to DHODH inhibition. In contrast, the current work systemati cally dissects the virus-induced reprogramming of host pyrimidine metabolism, revealing how ASFV sustains de novo nucleotide biosynthesis through host pathways, thereby creating a state of metabolic dependency exploitable by brequinar. Together, these two findings support a "metabolic-cell death" collaborative model for brequinar's antiviral mechanism against ASFV, providing a theoretical foundation for dual-targeted therapeu tic strategies.
Moreover, the role of the PPP in maintaining redox balance and supplying nucleotide precursors has been well documented in the literature (31). Our study extends these observations to ASFV by demonstrating that inhibition of the PPP rate-limiting enzyme G6PD via RRx-001 not only diminishes nucleotide precursor production but also reduces the NADPH/NADP + ratio, which is crucial for mitigating oxidative stress. Previous studies have indicated that an elevated NADPH/NADP + ratio reflects a reductive intracellular environment that maintains redox homeostasis and facilitates DNA synthesis (28). Our results show that exogenous nucleoside or antioxidant supplementation partially rescues viral replication under G6PD inhibition, underscoring the dual role of the PPP in supplying nucleotide precursors and maintaining redox balance.
While our data support the contribution of oxPPP-derived NADPH to oxidative stress defense and suggest a role for the PPP in nucleotide synthesis, they do not conclusively demonstrate that ASFV specifically enhances nucleotide precursor flux through the oxPPP. Notably, [1,2-13 C 2 ]-glucose was used in this study to trace the metabolic flow of 13 C in mock-infected and ASFV-infected cells. However, there was no significant difference in (M + 1)-labeled R5P flux between mock-infected and ASFV-infected cells (data not shown). This discrepancy may arise from insufficient detection sensitivity or compensa tory R5P production via the non-oxPPP branch or salvage pathways, which merits further investigation.
Glutamine is a crucial amino acid, serving as both an important energy substrate and a key source of carbon and nitrogen for biosynthetic reactions. Our results are consistent with recent reports that viral infections can upregulate specific amino acid transporters to meet their biosynthetic and energetic demands (32). We observed that ASFV infection upregulates the glutamine transporter SLC1A5, which enhances glutamine uptake. This increased uptake directs glutamine's nitrogen toward pyrimidine synthesis and its carbon toward TCA cycle replenishment to sustain metabolic flux. Notably, the rebound in intracellular glutamine levels at 24 hpi (Fig. 4C) may reflect a compensatory metabolic response. Initially, ASFV infection accelerates the uptake and catabolism of extracellular glutamine to satisfy high metabolic demands for nucleotide biosynthesis and TCA cycle replenishment, thereby depleting intracellular glutamine. However, the sudden increase observed at 24 hpi may indicate an upregulation of alternative glutamine uptake pathways or a reduction in glutamine consumption during the late stages of infection. This shift might serve to restore intracellular metabolic balance and ensure the availability of glutamine for critical biosynthetic processes, ultimately supporting sustained viral replication during advanced stages of infection. In addition, Dai et al. reported that during the late phase of ASFV infection, host cells produce abundant phenyllactic acid, which accumulates and inhibits the utilization of glutamine by ASFV, indirectly explaining the increase in intracellular glutamine levels observed at 24 hpi (33). Overall, our findings are consistent with previous reports linking elevated glutamine uptake to enhanced viral replication (34,35).
Aspartate is recognized as a critical metabolite for cell proliferation due to its essential role in nucleotide synthesis (36). Although previous studies have highlighted the importance of extracellular aspartate in supporting viral replication (19), our data reveal that ASFV bypasses this dependency by activating GOT1-mediated, cell-autono mous aspartate synthesis. Additionally, we observed that suppression of mitochondrial GOT2 paradoxically enhances viral replication. Previous reports have demonstrated that cytosolic GOT1 and mitochondrial GOT2 are co-regulated by the mitochondrial electron transport chain (ETC). Specifically, impairment of ETC complex I (NADH dehydrogenase) disrupts mitochondrial aspartate synthesis, thereby triggering a compensatory increase in cytosolic aspartate production via GOT1 (16,17). Mitochondria, particularly complexes I and III of the ETC, are major sources of ROS, and excessive ROS can inflict oxidative damage on key ETC components, including complex I (37). Studies have shown that ASFV infection induces significant ROS production and oxidative stress (38,39). Our findings align with this model: GOT2 knockdown results in a compensatory increase in GOT1 expression, rerouting aspartate synthesis to the cytosol. ASFV, which replicates in cytoplasmic viral factories, likely benefits from localized aspartate pools that directly feed viral nucleotide biosynthesis.
In conclusion, our findings support a model in which ASFV orchestrates coordinated reprogramming of host metabolism, specifically targeting nucleotide biosynthesis to create an environment favorable to viral replication. By simultaneously enhancing the PPP, upregulating glutamine uptake and metabolism, and rerouting aspartate synthesis to the cytosol, ASFV ensures a robust supply of both energy and essential biosynthetic precursors (Fig. 8). Furthermore, our work extends the understanding of ASFV metabolic hijacking by highlighting its selective reliance on pyrimidine synthesis, which may reflect evolutionary pressures favoring energy-efficient pathways under conditions of high metabolic competition. This dual strategy-targeting both the supply of nucleo tide precursors and the maintenance of redox balance-underscores the sophistication of ASFV's metabolic adaptations and provides a compelling rationale for therapeutic targeting of these pathways.
## MATERIALS AND METHODS
## Virus, cell lines, antibodies, and chemical reagents
The ASFV strain GZ201801_2 (GenBank accession number ON263123) was isolated from clinical specimens during the early ASF outbreaks in 2018 and is preserved at the Infectious Diseases Laboratory of South China Agricultural University. PAMs were isolated from 28-day-old specific-pathogen-free pigs and maintained at 37°C with 5% CO 2 in RPMI 1640 medium (Gibco, Billings, MT, USA) supplemented with 10% fetal bovine serum (Gibco).
The antibodies used in this study included ASFV-P30 mouse monoclonal antibody (produced by our laboratory), β-tubulin (M20005; Abmart), GOT1 (14886-1-AP; Protein tech), and GOT2 (67738-1-Ig; Proteintech). The chemical reagents used were AVN-944 (HY-13560), brequinar (HY-108325), NAC (HY-B0215), V-9302 (HY-112683), AZD-8055 (HY-10422), uridine (HY-B1449), and AOA (HY-107994), all from MedChemExpress. Glutamine (C0212) and aspartate (ST1476) were obtained from Beyotime. RRx-001 (S8405) was purchased from Selleck, and nucleosides (ES-008-D) from Sigma-Aldrich.
## siRNA transfection
siRNA transfection was performed using Lipofectamine RNAiMAX (Thermo Fisher Scientific) following the manufacturer's instructions. The siRNA used in this study was designed and synthesized by Tsingke Biotech Co., Ltd. (Beijing, China). Briefly, 5 µL siRNA and 3 µL Lipofectamine RNAiMAX were each diluted in 100 µL Opti-MEM (31985070, Gibco). After a 5 min incubation at room temperature (RT), the two mixtures were combined, thoroughly mixed, and incubated for an additional 20 min at RT. The resulting transfection complex (200 µL/well) was added to 12-well culture plates and incubated for 24 h before proceeding with subsequent experimental procedures. The siRNA sequences are provided in Table 1.
## Virus infection and drug treatment
PAMs were seeded into cell culture plates for 4 h to allow adherence, then washed with RPMI 1640 and incubated with an ASFV suspension (MOI = 1) for 2 h. After incubation, the supernatant was discarded and replaced with drugs at the appropriate concentrations depending on the experimental condition. At the indicated time points, cell samples were collected for real-time quantitative PCR (RT-qPCR), Western blotting, or other analyses. Unless otherwise indicated, PAMs were infected with ASFV at MOI = 1, and samples were collected at 24 hpi for Western blotting, RT-qPCR, and viral titer analysis.
## Hemadsorption assay
The hemadsorption assay was used to determine virus titers. The virus of unknown titer was serially diluted 10-fold in RPMI 1640 and inoculated onto PAMs seeded in 96-well plates. After 24 h, 15 µL of a 1% suspension of washed porcine erythrocytes was added to each well, and hemadsorption was recorded for up to 72 h. Endpoint titers were calculated using the Reed-Muench method.
## RT-qPCR
Total RNA was extracted from cell samples using the Total RNA Rapid Extraction Kit (220010; Fastagen) and reverse transcribed using the StarScript III RT Kit (A232-10; Genstar). qPCR was conducted using the ChamQ SYBR qPCR Master Mix (Q311-02; Vazyme) on a Bio-Rad CFX96 system. The primer sequences are provided in Table 2.
## Western blotting
Cell samples were collected and lysed in ice-cold RIPA lysis buffer (P0013B; Beyotime) for 10 min. Lysates were centrifuged at 4°C for 15 min, and total protein concentration in the supernatant was determined using the BCA Protein Assay Kit (P0012; Beyotime). Protein samples were separated using 10% SDS-PAGE and transferred onto a nitrocellulose (NC) membrane (Merck KGaA, Darmstadt, Germany). The NC membrane was blocked with 5% skim milk for 45 min and washed with Tris-buffered saline with 0.1% Tween 20 (TBST), followed by incubation with the primary antibody at 4°C for 8 h. The membrane was then incubated with the secondary antibody at RT for 1 h. Proteins were detected using a Tanon-5200 multi-infrared imaging system (Shanghai Tianneng Technology Co., Ltd., Shanghai, China). Each Western blot was repeated on lysates from three independent cell cultures (biological replicates).
## Cell viability, glutamine, and NADPH determination
Cell viability was assessed using the Cell Counting Kit-8 (C0038; Beyotime) after drug treatment. Glutamine levels were measured with the Glutamine Content Detection Kit (ml092936; MLBio), and the NADPH/NADP + levels were determined using the CheKine Micro Coenzyme II NADP(H) Assay Kit (KTB1010; Abbkine). All experimental procedures were performed according to the manufacturer's instructions.
## Statistical analysis
Metabolic pathway enrichment analysis of the metabolomics data was conducted using the online software MetaboAnalyst version 5.0 (www.metaboanalyst.ca). All data were analyzed using GraphPad Prism version 8.0 (GraphPad Software), and results are expressed as mean ± standard deviation from at least three independent experiments. Statistical significance was determined using Student's t-test. *P < 0.05, **P < 0.01, and ***P < 0.001 were considered statistically significant.
## References
1. Dixon, Sun, Roberts (2019) "African swine fever" *Antiviral Res*
2. Bappy, Asim Mdm, Ahasan et al. (2024) "Virus-induced host cell metabolic alteration" *Rev Med Virol*
3. Zandi, Shokri, Mahmoudvand et al. (2022) "Interplay between cellular metabolism and DNA viruses" *J Med Virol*
4. Xue, Liu, Zhu et al. (2022) "African swine fever virus regulates host energy and amino acid metabolism to promote viral replication" *J Virol*
5. Thaker, Ch'ng, Hr (2019) "Viral hijacking of cellular metabolism" *BMC Biol*
6. Chen, Zhang, Hoshino et al. (2019) "NADPH production by the oxidative pentose-phosphate pathway supports folate metabolism" *Nat Metab*
7. Eisenreich, Rudel, Heesemann et al. (2019) "How viral and intracellular bacterial pathogens reprogram the metabolism of host cells to allow their intracellular replication" *Front Cell Infect Microbiol*
8. Lévy, Bartosch (2016) "Metabolic reprogramming: a hallmark of viral oncogenesis" *Oncogene*
9. Fontaine, Sanchez, Camarda et al. (2015) "Dengue virus induces and requires glycolysis for optimal replication" *J Virol*
10. Tang, Chen, Zhao et al. (2023) "Newcastle disease virus manipulates mitochondrial MTHFD2-mediated nucleotide metabolism for virus replication" *J Virol*
11. Chen, Song, Chang et al. (2023) "Dihydromyricetin inhibits African swine fever virus replication by downregulating toll-like receptor 4-dependent pyroptosis in vitro" *Vet Res*
12. Wang, Bai, Ruan et al. (2019) "Coordina tive metabolism of glutamine carbon and nitrogen in proliferating cancer cells under hypoxia" *Nat Commun*
13. Yoo, Yu, Sung et al. (2020) "Glutamine reliance in cell metabolism" *Exp Mol Med*
14. Hirabara, Gorjao, Levada-Pires et al. (2021) "Host cell glutamine metabolism as a potential antiviral target" *Clin Sci*
15. Zhu, Li, Da Silva et al. (2017) "A critical role of glutamine and asparagine γ-nitrogen in nucleotide biosynthesis in cancer cells hijacked by an oncogenic virus" *mBio*
16. Lucas, Gui, Hosios et al. (2015) "Supporting aspartate biosynthesis is an essential function of respiration in proliferating cells" *Cell*
17. Birsoy, Wang, Chen et al. (2015) "An essential role of the mitochondrial electron transport chain in cell proliferation is to enable aspartate synthesis" *Cell*
18. Luengo, Danai, Bush et al. (2018) "Aspartate is an endogenous metabolic limitation for tumour growth" *Nat Cell Biol*
19. Liu, Zhu, Xue et al. (2023) "Picornavirus infection enhances aspartate by the SLC38A8 transporter to promote viral replication" *PLoS Pathog*
20. Hoffmann, Kunz, Simon et al. (2011) "Broadspectrum antiviral that interferes with de novo pyrimidine biosynthesis" *Proc Natl Acad Sci*
21. Kandasamy, Gyimesi, Kanai et al. (2018) "Amino acid transporters revisited: new views in health and disease" *Trends Biochem Sci*
22. Liu, Tian, Hao et al. (2024) "] infections differentially promote PEDV replication by reprogramming glutamine metabolism" *PLoS Pathog*
23. Liu, Tang, Meng et al. (2022) "SLC1A3 facilitates newcastle disease virus replication by regulating glutamine catabolism" *Virulence*
24. Scalise, Pochini, Galluccio et al. (2017) "Glutamine transport and mitochondrial metabolism in cancer cell growth" *Front Oncol*
25. Wang, Shen, Nobre et al. (2019) "Epstein-barr-virus-induced one-carbon metabolism drives B cell transformation" *Cell Metab*
26. Wan, Tavakoli, Wang et al. (2024) "Hijacking of nucleotide biosynthesis and deamidation-mediated glycolysis by an oncogenic herpesvirus" *Nat Commun*
27. Qin, Rao, Yuan et al. (2022) "SARS-CoV-2 couples evasion of inflammatory response to activated nucleotide synthesis" *Proc Natl Acad Sci*
28. (2025) *Full-Length Text Journal of Virology*
29. Nouwen, Breeuwsma, Zaal et al. (2024) "Modulation of nucleotide metabolism by picornaviruses" *PLoS Pathog*
30. Dsouza, Pant, Pope et al. (2025) "Vaccinia growth factordependent modulation of the mTORC1-CAD axis upon nutrient restriction" *J Virol*
31. Chen, Guo, Chang et al. (2023) "Brequinar inhibits African swine fever virus replication in vitro by activating ferroptosis" *Virol J*
32. Goodwin, Xu, Munger (2015) "Stealing the keys to the kitchen: viral manipulation of the host cell metabolic network" *Trends Microbiol*
33. Nguyen, Lim, Nguyen et al. (2018) "Hepatitis C virus modulates solute carrier family 3 member 2 for viral propagation" *Sci Rep*
34. Dai, Ma, Wubshet et al. (2024) "The accumulation of phenyllactic acid impairs host glutamine metabolism and inhibits African swine fever virus replication: a novel target for the development of anti-ASFV drugs" *Viruses*
35. Mitra, Chaudhuri, Sarkar et al. (2024) "Rotavirus rewires host cell metabolic pathways toward glutamine catabolism for effective virus infection" *Gut Microbes*
36. Chambers, Maguire, Alwine (2010) "Glutamine metabolism is essential for human cytomegalovirus infection" *J Virol*
37. Lane, Fan (2015) "Regulation of mammalian nucleotide metabolism and biosynthesis" *Nucleic Acids Res*
38. Zhao, Jiang, Zhang et al. (2019) "Mitochondrial electron transport chain, ROS generation and uncoupling (review)" *Int J Mol Med*
39. Xia, Wang, Liu et al. (2020) "African swine fever virus structural protein p17 inhibits cell proliferation through ER stress-ROS mediated cell cycle arrest" *Viruses*
40. Liu, Li, Zhai et al. (2025) "Resveratrol inhibits African swine fever virus replication via the Nrf2-mediated reduced glutathione and antioxidative activities" *Emerg Microbes Infect*
41. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12816819&blobtype=pdf | # Forecasting Influenza Epidemics and Pandemics in the Age of AI and Machine Learning
Oleksandr Kamyshnyi, Iryna Halabitska, | Oksenych, Iryna Kamyshna, Pavlo Petakh, | Denis, E Kainov
## Abstract
Influenza's rapid evolution, driven by its segmented RNA genome, high mutation rate, and extensive animal reservoirs, underpins its capacity to cause recurring epidemics and unpredictable pandemics. Recent advances in artificial intelligence (AI) and machine learning (ML) are transforming influenza forecasting by enabling the prediction of viral evolution and the optimisation of public health preparedness. This review synthesises insights from historical data (1890-2025) and contemporary research to examine the evolving role of AI in influenza prediction. It highlights major developments including transformerbased models for viral evolution, real-time integration of mobility and environmental data, hybrid quantum, which are classical algorithms, and multimodal data fusion frameworks, it also consideres critical risk modifiers such as meteorological variation, armed conflict, and host genetics. Importantly, the review distinguishes between retrospective, proof-of-concept analyses and prospective, real-time forecasting applications, clarifying their respective contributions to operational public health preparedness and informed decision-making.
## 1 | Introduction
Accurate and timely forecasting is crucial for the effective management of influenza outbreaks, facilitating a shift from reactive responses to proactive public health interventions [1]. Advances in artificial intelligence (AI) and machine learning (ML) have revolutionised epidemiological modeling, enabling the prediction of epidemic trajectories, real-time monitoring of viral evolution, and the rapid deployment of targeted control measures [2]. These technologies leverage complex data streams to capture the multifaceted nature of influenza transmission, incorporating biological determinants such as viral genetics and host immunity, environmental influences including meteorological variables and ultraviolet radiation, as well as social factors like human mobility and behavioural patterns (Figure 1) [3]. This review examines the integration of AI and ML methodologies within influenza forecasting frameworks, focussing on the fusion of heterogeneous data types through advanced predictive analytics. Notably, recent developments in quantum computing and multimodal data integration have demonstrated significant potential to enhance computational efficiency and model accuracy. These approaches enable the simultaneous analysis of genomic sequences, environmental parameters, and epidemiological indicators, thereby strengthening the spatiotemporal precision of outbreak predictions. This schematic illustrates a multifaceted machine learning framework that integrates antigenic drift prediction, antigen mapping, immunity dynamics, host factors (age, sex, genetics), meteorological data, and conflict zone analysis to support the development of antiviral agents, AI-guided vaccine strategies, and the selection of candidate molecules for clinical trials.
Recent progress in computational epidemiology underscores the capacity of AI systems to integrate diverse data sources, encompassing genomic surveillance, real-time mobility, and climate metrics [4]. The amalgamation of these datasets has led to refined temporal and spatial resolution in forecasting models, enabling earlier detection of transmission hotspots and more precise estimation of epidemic peaks [5]. Furthermore, the deployment of adaptive learning algorithms facilitates continual model recalibration, allowing forecasts to dynamically incorporate new information about pathogen evolution and shifting population susceptibility [6]. Such responsiveness is crucial in the context of rapidly evolving viral landscapes and diverse public health interventions, ultimately leading to more effective disease control and mitigation strategies [7].
This review provides a distinct contribution by systematically differentiating between retrospective, proof-of-concept AI models and prospective, real-time forecasting applications relevant to public health decision-making. In contrast to prior reviews that primarily focus on algorithmic performance, we integrate biological, environmental, social, and conflict-related determinants within a unified analytical framework. By critically evaluating the operational readiness, limitations, and contextual applicability of AI-driven influenza forecasting, this work aims to bridge the gap between methodological innovation and real-world public health implementation.
## 1.1 | Literature Search and Selection
Relevant publications were identified through searches in PubMed, Scopus, and Web of Science covering the period from 2000 to 2025, using the keywords 'influenza forecasting', 'machine learning', 'artificial intelligence', 'epidemiological modeling', 'viral evolution', and 'predictive analytics'. Additional sources were retrieved from reference lists and preprint repositories (medRxiv, arXiv) to include emerging research. Studies were included if they applied or evaluated AI or ML approaches for influenza prediction, surveillance, or vaccine design.
Emphasis was placed on studies describing model architecture, data sources, and performance metrics, with explicit notation of whether analyses were conducted retrospectively or prospectively. This strategy ensured balanced coverage of both conceptual and applied advances in AI-driven influenza forecasting.
Studies were excluded if they lacked sufficient methodological detail, relied solely on simulated data without epidemiological grounding, or did not report the outcomes of model validation. Given the heterogeneity of study designs, a formal meta-analysis was not performed; instead, a structured narrative synthesis was applied. This approach allowed for a balanced assessment of methodological rigor, data sources, and practical relevance across diverse AI-based forecasting studies.
## 2 | Historical Perspective: Lessons Shaping
## Modern Forecasting
Historical pandemics have informed today's AI-driven forecasting tools. The 1890 epidemic revealed early transmission patterns [8], while the 1918 H1N1 pandemic, with over 50 million deaths [9], spurred basic surveillance systems-ancestors of modern predictive models. Later outbreaks (1957 H2N2, 1968 H3N2, 2009 H1N1pdm09) refined these approaches by exposing viral spread dynamics [10]. The drastic decline in influenza cases during COVID-19's non-pharmaceutical interventions (NPIs) in 2020-21 [11] underscored the power of real-time data, now supercharged by AI.
Recent studies have leveraged historical pandemics to enhance the performance of AI models. Recent studies employing graph neural networks (GNNs) have demonstrated improved spatial influenza nowcasting, though direct re-analysis of early pandemic data remains limited [12]. The 2009 H1N1pdm09 pandemic highlighted the role of global air travel, prompting machine learning (ML) models to incorporate flight data, which reduced prediction errors [13]. The 2020-21 influenza decline during COVID-19 NPIs provided a dataset for training LSTM models, achieving an accuracy boost in low-transmission scenarios [14]. Additionally, transfer learning from historical pandemics has enabled models to generalise across diverse influenza strains, with evidence showing an improvement in predicting the dynamics of novel strains [15].
## 3 | Biological Drivers in Forecasting Models
Influenza's biology underpins predictive algorithms: antigenic drift involves mutations in hemagglutinin (HA) and neuraminidase (NA) at approximately 2-3 × 10 -3 substitutions per site per year, which challenge existing immunity [16]. Recent transformer-based and sequence-learning approaches have shown promise for predicting antigenic drift from genomic data, which may help anticipate vaccine mismatch [17][18][19]. Antigenic shift, resulting from reassortment events in animal reservoirs, leads to the emergence of novel influenza strains, with deep learning techniques identifying shift risks through mapping of co-infection zones [20]. Additionally, immunity dynamics, characterised by short-lived immune protection especially pronounced in older adults [21], are incorporated into age-specific forecasting models to improve prediction accuracy.
Biological factors play a crucial role in influenza forecasting through the analysis of virus-host protein interactions and pathogenicity prediction. For instance, the use of XGBoost to predict influenza A virus-human protein-protein interactions has demonstrated high accuracy [22]. Machine learning models based on the predicted structure of hemagglutinin enable the assessment of avian influenza pathogenicity [23]. Furthermore, models focused on antigenic distance prediction [17,24] and zoonotic potential assessment [25] underscore the importance of biological drivers in forecasting.
These biological mechanisms represent key uncertainties in forecasting, primarily due to the non-linear nature of viral evolution across geographic and host boundaries [26]. Advances in computational virology have enabled the integration of hostpathogen interactions into dynamic models, capturing the probabilistic nature of antigenic events [27]. Predictive tools now incorporate protein structural data to detect functionally significant mutations with immunological consequences. The inclusion of zoonotic interface data from agricultural systems has expanded forecasting scope beyond human populations [28]. Furthermore, real-time genomic surveillance pipelines have shortened the response window for identifying threats related to drift or shifts [29].
Genomic sequencing advancements have revolutionised biological forecasting. Recent ML models integrating sequencebased features have achieved high accuracy in predicting antigenic distances, offering insights into antigenic drift [17,18]. Another study in 2024 introduced a multi-modal ML approach integrating metagenomic data from environmental samples (e.g., poultry farms), which improved shift predictions [30]. Immunity dynamics models now incorporate longitudinal serology data, with a 2024 study using Bayesian machine learning (ML) to predict waning immunity in elderly populations, thereby boosting age-specific forecast accuracy [31]. 4 | Environmental Inputs: Meteorology and Solar-
## UV in Predictions
Environmental determinants are discussed in this section in an integrated manner to avoid repetition across subsequent methodological and application-focused sections. Meteorological and solar ultraviolet factors are therefore addressed here as crosscutting modifiers of influenza transmission, informing rather than duplicating later discussions of AI model architectures and performance.
They represent critical environmental components of influenza forecasting models and play a critical role in refining forecast accuracy. Low absolute humidity, approximately 4 g/m 3 , enhances viral survival and transmission potential [32]. AI models incorporate real-time weather data, such as humidity and temperature, to more precisely predict transmission peaks. Solar ultraviolet (UV) radiation also significantly influences influenza dynamics; a UV-Index below 2 is associated with increased influenza surges, while each unit increase in UVI reduces transmission rates by 7%-10% [33]. Neural network models leverage these UVI trends to adjust seasonal forecasts, projecting an extension of influenza seasons by 2050 in response to anticipated climate change [34,35].
Incorporating meteorological data and solar ultraviolet (UV) radiation into influenza forecasting models has been shown to enhance the precision of epidemiological predictions significantly. While the number of studies explicitly addressing these environmental determinants remains limited in the existing body of literature, recent advances in predictive modeling underscore the importance of integrating climate-related variables, such as temperature, humidity, and solar UV indices, into surveillance frameworks. This approach not only improves the temporal and spatial resolution of outbreak forecasts but also contributes to a more comprehensive understanding of the complex interplay between environmental conditions and viral transmission dynamics [12,36,37].
Recent advances in atmospheric data resolution have improved the spatial precision of influenza forecasting, particularly in regions with previously limited monitoring capabilities [38]. Machine learning frameworks now integrate not only meteorological indicators but also microclimatic fluctuations to capture shortterm variability in transmission risk. The inclusion of dynamic environmental baselines enables more robust adaptation of models to climate anomalies and extreme weather patterns [39].
Integration of meteorological and satellite data in convolutional neural network frameworks has been shown to improve shortterm influenza forecasts in tropical regions [40,41]. Emerging evidence indicates that increased levels of ultraviolet radiation are associated with decreased influenza transmission and supports the application of machine learning models to predict UV-driven seasonal dynamics with substantial accuracy [17]. NeuralGCM models project a 10%-15% extension of influenza seasons by 2050 due to climate change [35,42]. Air pollution data (e.g., PM2.5 levels) have also been incorporated into AI systems, resulting in an increase in transmission observed in high-pollution zones, which enhances urban-focused forecasting [43]. Additionally, ensemble machine learning models that combine meteorological, UV, and climate variables have improved long-term forecasting performance. However, data latency in remote and low-resource regions remains a limiting factor [38]. 5 | AI and Machine Learning: Core of Influenza Forecasting AI powers advanced surveillance and prediction through several key methodologies. Nowcasting models, such as LSTM and gradient-boosting algorithms, integrate syndromic, laboratory, and mobility data to reduce forecasting errors by 10%-25% over 1-8-week horizons [27,30]. The inclusion of real-time smartphone mobility data further enhances the precision of these short-term forecasts [44]. Early warning systems employ natural language processing (NLP) techniques to analyse news reports and travel data, enabling the detection of outbreaks like the 2019 H1N1pdm resurgence prior to official confirmation [45]. Transformer models are utilised to predict viral evolution by analysing genomic sequences, facilitating anticipatory responses [19]. In antigenic mapping, deep learning approaches account for approximately 80% of the variance in H3N2 strains, thereby accelerating vaccine strain selection processes [17]. Additionally, graph neural networks contribute by forecasting crossstrain immunity patterns, which informs broader vaccine design strategies [46].
A range of machine learning architectures, including deep recurrent, graph-based, and transformer models, have been applied to influenza prediction tasks using diverse datasets. Their comparative features and reported accuracies are summarised in Table 1.
Recent developments in quantum machine learning (QML) and multi-modal data fusion have pushed the boundaries of influenza forecasting [52]. Emerging hybrid quantum-classical algorithms have been proposed to accelerate high-dimensional modeling tasks in infectious disease research, though their application to influenza forecasting remains largely conceptual [52,53]. Enhanced natural language processing models, utilising multilingual social media datasets, have enabled the detection of outbreaks several days earlier than traditional surveillance systems. Graph-based deep learning architectures have shown potential for modeling cross-strain immune interactions [12,28], though their validation in influenza forecasting is still preliminary [12,18]. Multi-modal fusion models combining genomic, syndromic, and social media data have achieved significant reductions in nowcasting errors, underscoring the value of integrated approaches. These advancements underscore the increasing potential of AI to provide timely, precise, and context-aware influenza forecasts.
## Model/Algorithm
Primary data Type(s) Forecast time horizon
## Representative reference
LSTM (long short-term memory)
Epidemiological, mobility, meteorological 1-8 weeks ahead [47] Graph neural network (GNN) Spatial-temporal case data, mobility, climate
Real-time and 1-4 weeks forecasts [48] Transformer-based sequence model Genomic (HA/NA), antigenic data Seasonal to multi-year [18] CNN (convolutional neural network)
Meteorological + satellite data 1-6 months [49] XGBoost ensemble Epidemiological + host factors 2-4 weeks [50] Reinforcement learning (RL) Epidemiological + intervention data Scenario simulation [51] Multimodal fusion model Genomic + social media + mobility Real-time [37] Collectively, these models highlight the complementary strengths of neural, statistical, and hybrid approaches in enhancing temporal and spatial forecasting precision across diverse data environments. Continued innovation in algorithmic methods and data integration will be crucial to meeting the challenges of evolving viral threats [54].
Together, the summarised models (Table 2) and the end-to-end workflow (Figure 2) outline how AI frameworks transform complex, multi-source data into actionable public health intelligence.
The diagram illustrates six interconnected stages -from data acquisition and integration to decision support -highlighting the transformation of heterogeneous datasets into actionable insights for public health and vaccine strategy optimisation.
It is essential to differentiate between retrospective model evaluations, where historical datasets are used to benchmark algorithmic accuracy, and prospective, real-time implementations designed to guide ongoing public health decisions. Most of the AI and ML models summarised in Table 1 have been developed and validated retrospectively, using previously collected genomic, epidemiological, or environmental data. These studies often demonstrate high accuracy in back-testing scenarios but may overestimate actual performance in realworld settings.
In contrast, several prospective implementations have applied AI forecasting in operational settings. Real-time nowcasting systems that integrate mobility and syndromic data, as well as NLP-based outbreak detection tools, have supported situational awareness in CDC and WHO surveillance frameworks [59,60]. However, these systems typically display higher uncertainty due to data latency, underreporting, and evolving epidemic conditions [61,62]. Clarifying whether models were used retrospectively or prospectively is crucial for assessing the actual public health value and readiness of AI-driven forecasting technologies [62]. This distinction also highlights the importance of transparency in model validation and evaluation standards for realworld applications.
## 5.1 | Common Datasets and Data Preprocessing
## Challenges in Influenza Forecasting
AI-based influenza forecasting relies on the integration of heterogeneous datasets that capture viral, environmental, and behavioural dynamics [63]. Table 2 summarises the major data sources used in recent studies and highlights typical preprocessing challenges. Harmonising these diverse inputs remains a crucial step for improving model reliability and reproducibility.
Standardisation of metadata, development of federated learning frameworks, and automated feature extraction pipelines are emerging as key strategies to overcome these challenges and facilitate large-scale, real-time forecasting.
Reproducibility remains a major challenge in AI-based influenza forecasting, as many studies rely on proprietary datasets or lack publicly available code and standardised evaluation benchmarks. Greater adoption of open data practices, shared preprocessing pipelines, and federated learning frameworks would substantially improve transparency and facilitate independent validation across diverse settings.
## 5.2 | Limitations and Challenges of Machine
## Learning in Influenza Forecasting
Despite significant progress, several limitations constrain the practical implementation of AI and ML in influenza forecasting.
First, overfitting remains a major issue, particularly when models are trained on limited or region-specific datasets. Highly complex architectures such as deep neural networks can achieve excellent retrospective performance but may fail to generalise to unseen outbreaks or geographic regions [64,65].
Second, interpretability challenges hinder the adoption of blackbox models in public health decision-making. Many forecasting systems provide accurate outputs without transparent reasoning, making it difficult for epidemiologists to assess the reliability of predictions or to justify interventions based on them [66]. Third, regional data scarcity and heterogeneity limit global model transferability. Forecasting systems trained on data from high-income countries often perform poorly in low-and middleincome regions due to inconsistent reporting, lack of genomic sequencing, and differences in health infrastructure [67].
Additionally, data latency, privacy constraints, and biases from social media or mobility datasets can distort early warning outputs. Addressing these issues requires hybrid frameworks that combine mechanistic and data-driven models, robust validation pipelines, and explainable AI methods to ensure interpretability and reproducibility across diverse settings [68].
A growing body of evidence suggests that the most reliable influenza forecasts are derived from the integration of AIdriven, statistical, and mechanistic models, rather than relying on a single methodological paradigm. Ensemble frameworks that combine outputs from machine learning algorithms with compartmental (e.g., SEIR-type) or autoregressive models enhance forecast accuracy, stability, and uncertainty quantification, especially during atypical transmission periods. These hybrid approaches unite the interpretability and theoretical grounding of mechanistic models with the adaptive learning capacity of AI, yielding complementary strengths for real-time prediction and policy planning.
However, the published literature often overrepresents successful or high-performing AI models, introducing a potential publication bias that obscures null or less favourable results. Notably, several studies have shown that machine learning approaches do not consistently outperform traditional models when applied prospectively, particularly in data-sparse settings or during irregular influenza seasons [69,70]. In such contexts, simpler autoregressive or compartmental frameworks may achieve comparable, or even superior, real-time accuracy. Acknowledging these limitations is essential to maintain a balanced perspective on AI's role and to ensure that future research builds on realistic expectations of predictive performance.
Beyond technical limitations, ethical and governance challenges remain central to the deployment of AI-driven influenza forecasting. The use of mobility data, social media signals, and satellite imagery raises concerns regarding privacy, data ownership, and potential misuse of surveillance. Moreover, unequal access to high-quality data and computational resources may exacerbate existing global health disparities, underscoring the need for transparent governance frameworks and ethically grounded deployment strategies.
## 6 | Translating Forecasts Into Vaccine and
## Antiviral Strategies
Influenza forecasting not only anticipates epidemic trends but also directly informs the design and update of vaccines and virus-directed antivirals [71,72]. Predictive models that track antigenic drift and reassortment guide the timely selection of vaccine strains, improving alignment with circulating viruses [73]. Machine learning algorithms integrating genomic and epidemiological data enable real-time evaluation of vaccine effectiveness and prediction of resistance patterns [74].
By linking genomic forecasts with immunological and epidemiological datasets, AI systems can identify emerging mutations that reduce vaccine efficacy, providing early warnings for formulation updates [75,76]. These approaches also support the anticipation of antiviral resistance, optimising therapeutic strategies before widespread clinical failure occurs [77].
Recent advances in transformer-based and multimodal models have strengthened this connection between forecasting and response, allowing rapid adaptation of vaccine composition and deployment strategies [78,79]. In this sense, AI-driven innovation serves as a practical extension of forecasting-transforming predictive insights into actionable public health interventions that enhance preparedness and resilience against evolving influenza threats.
## 7 | Forecasting Amid Conflict
War disrupts traditional disease surveillance systems, yet artificial intelligence demonstrates adaptability in these challenging environments. By leveraging satellite imagery, mobility data, and natural language processing (NLP) of open-source information, AI tools effectively monitor influenza transmission among displaced populations. This integration enables the timely forecasting of outbreaks, providing critical guidance for humanitarian aid and public health interventions in conflict zones [44,80].
Forecasting influenza outbreaks in conflict zones poses challenges due to data instability and dynamic conditions. Although direct studies on this issue are scarce in the current literature, the use of flexible and multi-source data models, such as graph neural networks capturing temporal, geographical, and functional spatial features, may enhance forecast robustness in such complex environments [12].
These challenges are exemplified by the ongoing Russia-Ukraine war, where the degradation of healthcare infrastructure has further complicated disease monitoring and response efforts. The Russia-Ukraine war, escalating in February 2022, has severely disrupted formal influenza surveillance [81]. This led to a nearcomplete cessation of formal infectious disease reporting, including influenza, in conflict-affected regions [81]. AI-driven open-source intelligence platforms have mitigated this by analysing multilingual news and social media [81]. These platforms have enabled the early detection of outbreaks in displaced populations, with reports indicating a decrease in traditional influenza reporting, necessitating a reliance on alternative data sources [82]. AI models that integrate satellite imagery and mobility data have improved outbreak predictions in war-torn areas, although challenges remain due to restricted access and data sparsity [83].
One such region where these challenges became particularly evident is southern Ukraine, following the collapse of the Kakhovka Dam. The collapse of the Kakhovka Dam released approximately 18 km 3 of water, flooding over 620 km 2 and affecting more than 100,000 people. This event mobilised an estimated 83,000 tonnes of heavy metals from reservoir sediments, contaminating water sources. Disrupted sewage and flooding increased waterborne disease risks, while dislodged landmines restricted access to care and mobility data. AI models integrating satellite and water data improved outbreak forecasting. Damaged irrigation systems have also exacerbated food insecurity and influenza vulnerability [84][85][86][87][88].
In response to these compounding vulnerabilities-including rising rates of chronic illness, comorbidities, depression, and disrupted gut microbiota-recent AI-driven innovations are increasingly being deployed to support disease surveillance and prediction in high-risk, resource-limited settings [89,90]
## 8 | Host Factors Enhancing Forecast Precision
Demographic and genetic factors significantly enhance the accuracy of influenza forecasting models. Age and sex are critical determinants, as children and elderly populations often drive both transmission and disease severity [95]. Sex-based immune differences, including hormonal influences such as estrogenmediated effects, further refine these models by capturing variations in immune response between males and females. Additionally, genetic variants like IFITM3 rs12252-C have been linked to a higher risk of severe influenza outcomes [96]. The integration of multi-omics data-comprising single-nucleotide polymorphisms (SNPs) and transcriptomic profiles-through machine learning methods improves predictions of individual susceptibility and disease progression, thereby contributing to more precise and personalised forecasting [97].
Host-related factors, including laboratory parameters, significantly improve diagnostic and predictive accuracy. Machine learning models utilising laboratory data have shown efficacy in predicting influenza A and B infections [98] and forecasting hospitalisations [99], highlighting the value of host-specific information in surveillance.
Multi-omics has advanced host factor integration. Recent research combining genomic, proteomic, and metabolomic data has improved susceptibility predictions [100]. Sex-specific models that account for hormonal influences have enhanced severity predictions in females [101]. Additionally, data from wearable devices, such as heart rate variability, have increased the accuracy of early infection detection [102]. Incorporating lifestyle factors and comorbidities into machine learning models shows promise for further refining risk assessments. These real-world limitations highlight the importance of adaptive model recalibration, uncertainty quantification, and integration with mechanistic epidemiological models. Continuous validation under operational conditions is therefore critical to ensure that AI-derived forecasts provide tangible, reliable benefits for decision-makers during active outbreaks.
Accurate influenza forecasts and early outbreak detection underpin effective public health decision-making. Deep learning models analysing search engine data and other multisource inputs enable real-time epidemic trend monitoring [105,106]. Ensemble machine learning methods provide precise hospitalisation predictions [99], while comparative studies demonstrate the utility of machine learning algorithms in assessing hospitalacquired influenza risk [50].
A reinforcement learning framework for optimising NPIs demonstrated a reduction in case numbers in simulated environments [107,108]. Vaccine allocation models that integrate socioeconomic indicators have improved equitable distribution in resource-limited settings [109,110]. Digital twin models enable real-time NPI simulations, facilitating dynamic policy adjustments to health outcomes [111]. Federated learning approaches have promoted global data sharing and collaboration, enhancing model robustness [112].
## 10 | Conclusion
Artificial intelligence and machine learning can enhance influenza preparedness by enabling earlier detection and more adaptive forecasting. However, their value in practice remains limited by inconsistent data quality, model transparency, and algorithmic bias. Current evidence shows that AI methods do not consistently outperform established statistical or mechanistic models, highlighting the need for rigorous and balanced evaluation.
Future research should prioritise the prospective validation of real-time, interpretable systems and explore their integration with ensemble and mechanistic frameworks to strengthen reliability and policy relevance. Transparent and ethically governed data infrastructures are crucial for ensuring equitable access, especially in settings with limited digital capacity.
While AI and machine learning offer substantial potential to enhance influenza forecasting, current evidence suggests that their real-world performance remains highly context-dependent.
In several prospective applications, AI-based approaches do not consistently outperform traditional statistical or mechanistic models, particularly in data-scarce or rapidly evolving settings. Recognising these limitations is essential to avoid overreliance on algorithmic outputs and to promote responsible, evidence-based integration into public health practice.
Developing standardised benchmarks, open data sharing, and cross-sector collaboration will help move the field beyond proofof-concept studies. These efforts are crucial to ensuring that predictive technologies evolve into trusted, evidence-based tools that support global public health decision-making.
Future progress in AI-driven influenza forecasting will depend on prospective validation, standardised benchmarking, and the development of ethically governed data infrastructures. Strengthening reproducibility, interpretability, and cross-sector collaboration will be critical to transforming AI models from experimental tools into trusted components of global influenza preparedness and response.
## References
1. Sharan, Vijay, Yadav et al. (2023) "Surveillance and Response Strategies for Zoonotic Diseases: A Comprehensive Review" *Science in One Health*
2. Zhao, Li, Cao (2024) "AI for Science: Predicting Infectious Diseases" *Journal of Safety Science and Resilience*
3. Long, Mistry, Haslam et al. (2019) "Host and Viral Determinants of Influenza A Virus Species Specificity" *Nature Reviews Microbiology*
4. Vargas-Santiago, León-Velasco, Maldonado-Sifuentes et al. (2025) "A State-of-the-Art Review of Artificial Intelligence (AI) Applications in Healthcare: Advances in Diabetes, Cancer, Epidemiology, and Mortality Prediction" *Computers*
5. Chen, Chen, Zhao (2024) "High-Resolution Short-Term Prediction of the COVID-19 Epidemic Based on Spatial-Temporal Model Modified by Historical Meteorological Data" *Fundamental Research*
6. Ye, Pandey, Bawden (2025) "Integrating Artificial Intelligence With Mechanistic Epidemiological Modeling: A Scoping Review of Opportunities and Challenges" *Nature Communications*
7. Alahmari, Almuzaini, Alamri et al. (2024) "Strengthening Global Health Security Through Health Early Warning Systems: A Literature Review and Case Study" *Journal of Infection and Public Health*
8. Johnson, Mueller (2002) "Updating the Accounts: Global Mortality of the 1918-1920 'Spanish' Influenza Pandemic" *Bulletin of the History of Medicine*
9. Olsen, Azziz-Baumgartner, Budd (2020) "Decreased Influenza Activity During the COVID-19 Pandemic -United States" *MMWR Morbidity and Mortality Weekly Report*
10. Taubenberger, Kash (2010) "Influenza Virus Evolution, Host Adaptation, and Pandemic Formation" *Cell Host & Microbe*
11. Bedford, Riley, Barr (2015) "Global Circulation Patterns of Seasonal Influenza Viruses Vary With Antigenic Drift" *Nature*
12. Luo, Wang, Fan (2025) "A Novel Graph Neural Network Based Approach for Influenza-like Illness Nowcasting: Exploring the Interplay of Temporal, Geographical, and Functional Spatial Features" *BMC Public Health*
13. Jefferson, Del Mar, Dooley (2020) "Physical Interventions to Interrupt or Reduce the Spread of Respiratory Viruses"
14. Davis, Mott, Olsen (2022) "The Role of Non-Pharmaceutical Interventions on Influenza Circulation During the COVID-19 Pandemic in Nine Tropical Asian Countries" *Influenza and Other Respiratory Viruses*
15. Gawande, Zade, Kumar et al. (2025) "The Role of Artificial Intelligence in Pandemic Responses: From Epidemiological Modeling to Vaccine Development" *Molecular Biomedicine*
16. Brownstein, Freifeld (2007) "Healthmap: The Development of Automated real-time Internet Surveillance for Epidemic Intelligence"
17. Shah, Palomar, Barr et al. (2024) "Seasonal Antigenic Prediction of Influenza A H3N2 Using Machine Learning" *Nature Communications*
18. Li, Li, Shang et al. (2024) "A Sequence-based Machine Learning Model for Predicting Antigenic Distance for H3N2 Influenza Virus" *Frontiers in Microbiology*
19. Thadani, Gurev, Notin (2023) "Learning From Prepandemic Data to Forecast Viral Escape" *Nature*
20. Balcan, Gonçalves, Hu et al. (2010) "Modeling the Spatial Spread of Infectious Diseases: The Global Epidemic and Mobility Computational Model" *Journal of Computational Science*
21. Nogareda, Ghiselli, Velandia-González (2024) "Seasonal Influenza Vaccination Programs in the Americas: A Platform for Sustainable Life-Course Immunization and Its Role for Pandemic Preparedness and Response" *Vaccines*
22. Li, Li, Li et al. (2025) "Prediction of Influenza A Virus-Human Protein-Protein Interactions Using XGBoost With Continuous and Discontinuous Amino Acids Information" *PeerJ*
23. Shin, Kim, Kim et al. (2024) "Machine Learning Using Template-Based-Predicted Structure of Haemagglutinin Predicts Pathogenicity of Avian Influenza" *Journal of Microbiology and Biotechnology*
24. Jia, Xia, Dong et al. (2024) "Metafluad: Meta-learning for Predicting Antigenic Distances Among Influenza Viruses" *Briefings in Bioinformatics*
25. Kim, Kim, Kim (2025) "Machine Learning Assessment of Zoonotic Potential in Avian Influenza Viruses Using PB2 Segment" *BMC Genomics*
26. Tan, Van Dorp, Balloux (2024) "The Evolutionary Drivers and Correlates of Viral Host Jumps" *Nature Ecology & Evolution*
27. Barkan, Siddiqui, Cheng (2025) "Leveraging Large Language Models to Predict Antibody Biological Activity Against Influenza A Hemagglutinin" *Computational and Structural Biotechnology Journal*
28. Krokidis, Koumadorakis, Lazaros (2025) "Alpha-Fold3: An Overview of Applications and Performance Insights" *International Journal of Molecular Sciences*
29. Struelens, Ludden, Werner et al. (2024) "Real-Time Genomic Surveillance for Enhanced Control of Infectious Diseases and Antimicrobial Resistance" *Frontiers in Science*
30. Kieran, Sun, Maines et al. (2024) "Machine Learning Approaches for Influenza A Virus Risk Assessment Identifies Predictive Correlates Using Ferret Model in Vivo Data"
31. Cuesta-Herrera, Pastenes, Arencibia et al. (2024) "Dynamics of Activation and Regulation of the Immune Response to Attack by Viral Pathogens Using Mathematical Modeling" *Mathematics*
32. Shaman, Kohn (2009) "Absolute Humidity Modulates Influenza Survival, Transmission, and Seasonality" *Proceedings of the National Academy of Sciences of the United States of America*
33. Ianevski, Zusinaite, Shtaida (2019) "Low Temperature and Low UV Indexes Correlated With Peaks of Influenza Virus Activity in Northern Europe During 2010-2018" *Viruses*
34. Zarei, Nikoo, Al-Rawas (2025) "Hybrid Deep Learning Downscaling of GCMs for Climate Impact Assessment and Future Projections in Oman" *Journal of Environmental Management*
35. Kochkov, Yuval, Langmore (2024) "Neural General Circulation Models for Weather and Climate" *Nature*
36. Musa, Nia, Bragazzi et al. (2024) "Avian Influenza: Lessons From past Outbreaks and an Inventory of Data Sources, Mathematical and AI Models, and Early Warning Systems for Forecasting and Hotspot Detection to Tackle Ongoing Outbreaks" *Healthcare*
37. Ray, Wang, Wolfinger et al. (2025) "Flusion: Integrating Multiple Data Sources for Accurate Influenza Predictions" *Epidemics*
38. Tsang, Du, Cowling et al. (2024) "An Adaptive Weight Ensemble Approach to Forecast Influenza Activity in an Irregular Seasonality Context" *Nature Communications*
39. Sevgin (2025) "Machine Learning-Based Temperature Forecasting for Sustainable Climate Change Adaptation and Mitigation" *Sustainability*
40. Zhu, Chen, Qin (2024) "Study on the Impact of Meteorological Factors on Influenza in Different Periods and Prediction Based on Artificial Intelligence RF-Bi-LSTM Algorithm: To Compare the COVID-19 Period With the non-COVID-19 Period" *BMC Infectious Diseases*
41. Turtle, Riley, Ben-Nun et al. (2021) "Accurate Influenza Forecasts Using Type-Specific Incidence Data for Small Geographic Units" *PLoS Computational Biology*
42. Ruan, Liang, Sun et al. (2025) "Climate Warming and Influenza Dynamics: The Modulating Effects of Seasonal Temperature Increases on Epidemic Patterns" *Climate and Atmospheric Science*
43. Rosca, Carbureanu, Stancu (2025) "Data-Driven Approaches for Predicting and Forecasting Air Quality in Urban Areas" *Applied Sciences*
44. Zvyagin, Brace, Hippe (2022) "GenSLMs: Genome-Scale Language Models Reveal SARS-CoV-2 Evolutionary Dynamics"
45. Smith, Lapedes, De Jong (2004) "Mapping the Antigenic and Genetic Evolution of Influenza Virus" *Science*
46. Arevalo, Bolton, Le (2022) "A Multivalent nucleoside-modified mRNA Vaccine Against all Known Influenza Virus Subtypes" *Science*
47. Koge, Wagatsuma (2025) "Long Short-Term Memory-Based Forecasting of Influenza Epidemics Using Surveillance and Meteorological Data in Tokyo, Japan" *Frontiers in Public Health*
48. Lira, Martí, Sanchez-Pi (2022) "A Graph Neural Network With Spatio-Temporal Attention for Multi-Sources Time Series Data: An Application to Frost Forecast" *Sensors*
49. Yu, Kong, Leung (2024) "A 1D Convolutional Neural Network (1D-CNN) Temporal Filter for Atmospheric Variability: Reducing the Sensitivity of Filtering Accuracy to Missing Data Points" *Applied Sciences*
50. Cho, Lee, Kim (2024) "Prediction of Hospital-Acquired Influenza Using Machine Learning Algorithms: A Comparative Study" *BMC Infectious Diseases*
51. Tian, Li, Wang et al. (2026) "Optimizing Spatiotemporal Nonpharmaceutical Interventions for Influenza: An Adaptive Reinforcement Learning Approach for Regional Heterogeneity" *Infectious Disease Modelling*
52. Moon, Shim, Kim et al. (2025) "MIFlu: Large Language Model-Based Multimodal Influenza Forecasting Scheme" *IEEE Journal of Biomedical and Health Informatics*
53. Kumar, Yadav, Mukherjee et al. (2024) "Recent Advances in Quantum Computing for Drug Discovery and Development" *IEEE Access*
54. Duan, Xiong, Li et al. (2024) "Deep Learning Based Multimodal Biomedical Data Fusion: An Overview and Comparative Review" *Information Fusion*
55. Shu, Mccauley (2017) "GISAID: Global Initiative on Sharing all Influenza Data -From Vision to Reality"
56. Chen, Yu (2024) "Importance of Integrating Epidemiological and Genomic Surveillance of Seasonal Influenza Viruses to Monitor Global Circulation" *Clinical and Translational Medicine*
57. Venkatramanan, Sadilek, Fadikar (2021) "Forecasting Influenza Activity Using machine-Learned Mobility Map" *Nature Communications*
58. Alessa, Faezipour (2018) "A Review of Influenza Detection and Prediction Through Social Networking Sites" *Theoretical Biology and Medical Modelling*
59. Liscano, Anillo Arrieta, Montenegro et al. (2025) "Early Warning of Infectious Disease Outbreaks Using Social Media and Digital Data: A Scoping Review" *International Journal of Environmental Research and Public Health*
60. Mendes, Mendes, Moura (2025) "Harnessing Artificial Intelligence for Enhanced Public Health Surveillance: A Narrative Review" *Frontiers in Public Health*
61. Adewumi (2025) "Critical Analysis of Infectious Disease Surveillance and Response System in Nigeria" *Discover Public Health*
62. Liu, Cao (2022) "Modeling Time Evolving COVID-19 Uncertainties With Density Dependent Asymptomatic Infections and Social Reinforcement" *Scientific Reports*
63. Adeoye, Onifade, Bayode (2025) "Artificial Intelligence and Computational Methods for Modelling and Forecasting Influenza and Influenza-like Illness: A Scoping Review" *Beni-Suef University Journal of Basic and Applied Sciences*
64. Hsu, Chan, Weng et al. (2025) "Environmental PM(2.5) Exposure: An Ignored Factor Associated With Blood Cadmium Level in Hemodialysis Patients" *Therapeutics and Clinical Risk Management*
65. Nurtas, Altaibek, Ydyrys et al. (2025) "Analyzing Historical Seismic Data for Region-Specific Earthquake Prediction Through Deep Neural Networks" *Journal of Seismology*
66. Villanueva-Miranda, Xiao, Xie (2025) "Artificial Intelligence in Early Warning Systems for Infectious Disease Surveillance: A Systematic Review" *Frontiers in Public Health*
67. Fatumo, Chikowore, Choudhury et al. (2022) "A Roadmap to Increase Diversity in Genomic Studies" *Nature Medicine*
68. Banad, Sharif, Rezaei (2025) "Artificial Intelligence and Machine Learning for Smart Grids: From Foundational Paradigms to Emerging Technologies With Digital Twin and Large Language Model-Driven Intelligence" *Energy Conversion and Management X*
69. Marquez, Barrón-Palma, Rodríguez et al. (2023) "Supervised Machine Learning Methods for Seasonal Influenza Diagnosis" *Diagnostics*
70. Schimit, Sergio, Fontoura (2025) "Vaccination as a Game: Behavioural Dynamics, Network Effects, and Policy Levers-A Comprehensive Review" *Mathematics*
71. Morris, Gostic, Pompei (2018) "Predictive Modeling of Influenza Shows the Promise of Applied Evolutionary Biology" *Trends in Microbiology*
72. Taaffe, Lambach, Gsell et al. (2025) "Next-Generation Influenza Vaccines and the Pandemic Horizon: Challenges, Innovations, and the Road Ahead" *Vaccines*
73. Agor, Özaltın (2018) "Models for Predicting the Evolution of Influenza to Inform Vaccine Strain Selection" *Human Vaccines & Immunotherapeutics*
74. Olawade, Teke, Fapohunda (2024) "Leveraging Artificial Intelligence in Vaccine Development: A Narrative Review" *Journal of Microbiological Methods*
75. El Arab, Alkhunaizi, Alhashem et al. (2025) "Artificial Intelligence in Vaccine Research and Development: An Umbrella Review" *Frontiers in Immunology*
76. Wang, Chavda, Prajapati (2023) "An Amalgamation of Bioinformatics and Artificial Intelligence for COVID-19 Management: From Discovery to Clinic" *Current Research in Biotechnology*
77. Aw, Zhang, Vignuzzi (2025) "Strategies and Efforts in Circumventing the Emergence of Antiviral Resistance Against Conventional Antivirals"
78. Elfatimi, Lekbach, Prakash et al. (2025) "Artificial Intelligence and Machine Learning in the Development of Vaccines and immunotherapeutics-yesterday, Today, and Tomorrow" *Frontiers in Artificial Intelligence*
79. Hayawi, Shahriar, Alashwal et al. (2024) "Generative AI and Large Language Models: A New Frontier in Reverse Vaccinology"
80. Echeverri-De La Hoz, Martínez-Bravo, Gastelbondo-Pastrana (2025) "Genomics of Novel Influenza A Virus (H18N12) in Bats, Caribe Colombia" *Scientific Reports*
81. Kannan, Chen, Akhtar (2022) "Use of Open-Source Epidemic Intelligence for Infectious Disease Outbreaks"
82. Macintyre, Lim, Quigley (2022) "Preventing the next Pandemic: Use of Artificial Intelligence for Epidemic Monitoring and Alerts" *Cell Reports Medicine*
83. Teitelbaum, Ferraz, De La Cruz et al. (2024) "The Potential of Remote Sensing for Improved Infectious Disease Ecology Research and Practice" *Proceedings Biological sciences*
84. Hryhorczuk, Levy, Prodanchuk (2024) "The Environmental Health Impacts of Russia's War on Ukraine" *Journal of Occupational Medicine and Toxicology*
85. Petakh, Huber, Kamyshnyi (2018) "Geographical Factors and Air Raid Alarms Influence Leptospirosis Epidemiology in Ukraine"
86. Symochko, Pereira, Demyanyuk et al. (2024) "Resistome in a Changing Environment: Hotspots and Vectors of Spreading With a Focus on the Russian-Ukrainian War" *Heliyon*
87. Dzhemiliev, Antunez, Kizub (2024) "Bridging Medical Expertise in Crisis: The Development and Implementation of a Novel Mobile Application for Ukrainian Physicians During Wartime" *Journal of Global Health*
88. Petakh, Behzadi, Oksenych et al. (2024) "Current Treatment Options for Leptospirosis: A Mini-Review" *Frontiers in Microbiology*
89. Liu (2025) "Bracing the Artificial Intelligence Technology in Viral Infectious Disease Control" *Infectious Medicine*
90. Halabitska, Petakh, Kamyshna et al. (2024) "The Interplay of Gut Microbiota, Obesity, and Depression: Insights and Interventions" *Cellular and Molecular Life Sciences: CMLS*
91. Zhang, Zhou, Zheng et al. (2023) "Predicting Influenza With Pandemic-Awareness via Dynamic Virtual Graph Significance Networks" *Computers in Biology and Medicine*
92. Faheem, Wassif, Bayomi et al. (2024) "Improving Neural Machine Translation for Low Resource Languages Through Non-Parallel Corpora: A Case Study of Egyptian Dialect to Modern Standard Arabic Translation" *Scientific Reports*
93. Yagoub, Tesfaldet, Alsumaiti et al. (2024) "Estimating Population Density Using Open-Access Satellite Images and Geographic Information System: Case of Al Ain City, UAE" *Remote Sensing Applications: Society and Environment*
94. Williams, Taylor, Thompson (2024) "Blockchain-Based Secure Data Sharing Framework for Healthcare Information Systems" *International Journal of Information Engineering and Science*
95. Rio, Caldarelli, Miccoli (2025) "Sex Differences in Immune Responses to Infectious Diseases: The Role of Genetics, Hormones, and Aging" *Diseases*
96. Sciarra, Campolo, Franceschini et al. (2023) "Gender-Specific Impact of Sex Hormones on the Immune System" *International Journal of Molecular Sciences*
97. Ali (2023) "Artificial Intelligence in Multi-Omics Data Integration: Advancing Precision Medicine, Biomarker Discovery and Genomic-Driven Disease Interventions" *International Journal of Science and Research Archive*
98. Hu, Liu, Dong (2025) "Evaluation of a Machine Learning Model Based on Laboratory Parameters for the Prediction of Influenza A and B in Chongqing, China: Multicenter Model Development and Validation Study" *Journal of Medical Internet Research*
99. Gantenberg, Mcconeghy, Howe (2023) "Predicting Seasonal Influenza Hospitalizations Using an Ensemble Super Learner: A Simulation Study" *American Journal of Epidemiology*
100. Wang, Zhao, Zhang (2024) "Emerging Trends and Hot Topics in the Application of Multi-Omics in Drug Discovery: A Bibliometric and Visualized Study" *Current Pharmaceutical Analysis*
101. Aristu-Zabalza, Andrés-Rozas, Boyer-Díaz (2025) "Sex-Specific Differences in Preclinical Models of Advanced Chronic Liver Disease and Portal Hypertension" *Biology of Sex Differences*
102. Sanches, Silva, Librantz et al. (2023) "Wearable Devices to Diagnose and Monitor the Progression of COVID-19 Through Heart Rate Variability Measurement: Systematic Review and Meta-Analysis" *Journal of Medical Internet Research*
103. Volkova, Ayton, Porterfield et al. (2017) "Forecasting Influenza-Like Illness Dynamics for Military Populations Using Neural Networks and Social Media" *PLoS One*
104. Jang, Kim, Thompson et al. (2025) "Modeling Vaccination Prioritization Strategies for Post-Pandemic COVID-19 in the Republic of Korea Accounting for Under-reporting and age-structure" *Journal of Infection and Public Health*
105. Yang, Li, Yang (2023) "Deep-Learning Model for Influenza Prediction From Multisource Heterogeneous Data in a Megacity: Model Development and Evaluation" *Journal of Medical Internet Research*
106. Morris, Hayes, Cox et al. (2023) "Neural Network Models for Influenza Forecasting With Associated Uncertainty Using Web Search Activity Trends"
107. Chand, Ravi (2024) "A Novel Reinforcement Learning Framework for Disassembly Sequence Planning Using Q-learning Technique Optimized Using an Enhanced Simulated Annealing Algorithm" *Analysis and Manufacturing*
108. Peng, Xu, Chen et al. (2023) "Using Reinforcement Learning for Multi-Objective Cluster-Level Optimization of Non-pharmaceutical Interventions for Infectious Disease"
109. Watkinson, Williams, Gillibrand et al. (2023) "Evaluating Socioeconomic Inequalities in Influenza Vaccine Uptake During the COVID-19 Pandemic: A Cohort Study in Greater Manchester, England" *PLoS Medicine*
110. Mondal, Dutta, Barua (2023) "Vaccinet: Towards a Reinforcement Learning Based Smart Framework for Predicting the Distribution Chain Optimization of Vaccines for a Pandemic"
111. Niarakis, Laubenbacher, An (2024) "Immune Digital Twins for Complex Human Pathologies: Applications, Limitations, and Challenges" *NPJ Systems Biology and Applications*
112. Yurdem, Kuzlu, Gullu et al. (2024) "Federated Learning: Overview, Strategies, Applications, Tools and Future Directions" *Heliyon* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12826971&blobtype=pdf | # Improvement in UV resistance of entomopathogenic viruses by coating silica nanoparticles on occlusion bodies surface
Jia Wang, Zhongqiang Wu, Xiaoqin Yang, Chengfeng Lei, Xiulian Sun, Virologica Sinica
## Abstract
Entomopathogenic viruses, such as baculoviruses and cypoviruses, have been employed as biological pesticides against agricultural and forestry pests. However, their susceptibility to inactivation under field UV radiation has hindered their broader application. In this study, we effectively improved the UV resistance of insect virus occlusion bodies (OBs) by coating their surfaces with silica nanoparticles (SiO 2 NPs). Monodisperse SiO 2 NPs with uniform size distribution and excellent colloidal stability were synthesized using the St€ ober method. Subsequent amination modification of the SiO 2 NPs with a silane coupling agent shifted their isoelectric point from 3.2 to 8.1. This modification imparted a strong positive charge to the NPs within the pH range of 4.5-5.5, while the OBs of insect viruses remained negatively charged in this range. Consequently, the amino-functionalized SiO 2 NPs were successfully coated onto the surfaces of OBs of three representative insect viruses: nucleopolyhedrovirus, granulovirus, and cypovirus, through electrostatic interactions. Laboratory bioassays confirmed that Mamestra brassicae multiple nucleopolyhedrovirus (MbMNPV) coated with SiO 2 -NH 2 NPs retained its native viral pathogenicity against Spodoptera exigua larvae under normal laboratory condition, while it demonstrated 2.299-2.712 folds higher potency than MbMNPV physically mixed with unmodified SiO 2 NPs after UV irradiation. Outdoor trials revealed that SiO 2 -NH 2 NPs coating significantly improved the survival time of MbMNPV, with the median survival time increased from 1.43 days to 5.15 days. This nanoparticle coating strategy establishes a robust platform for developing photostable biopesticides while preserving their ecological safety profiles. The modular nature of this approach suggests its broad applicability across different entomopathogenic virus formulations.
## INTRODUCTION
Numerous viruses in the families Baculoviridae, Reoviridae, Densoviridae, and Entomopoxvirinae are capable of inducing epizootics in natural insect populations. This epidemiological potential has led to the global commercialization of viral agents as bioinsecticides. Among these, baculoviruses represent a prominent family of arthropod-specific pathogens characterized by circular double-stranded DNA genomes packaged within enveloped rod-shaped nucleocapsids (Rohrmann, 2019). Baculovirus occlusion bodies (OBs) serve as stable environmental reservoirs for viral transmission. The biocontrol applications of these viruses are derived from multiple advantageous characteristics: a narrow host range restricted to specific insect species, environmental compatibility with minimal non-target effects on vertebrates and plants, and high infectivity sufficient to initiate pest population collapse (Lapointe et al., 2012). However, the practical deployment of baculovirus-based insecticides faces several biological constraints. One key limitation is the rapid ultraviolet (UV)-induced inactivation of OBs, which results in short persistence on plant surfaces in fields where they are exposed to solar radiation (Behle and Birthisel, 2014). Among the solar spectrum, the mediumwave or erythermal UV band (UVB, 280-320 nm) is the most important photoinactivator of baculoviruses (Burges and Jones, 1998).
Baculoviruses exhibit pronounced sensitivity to UV radiation, with unformulated viruses typically losing their activity within 24-48 h on plant surfaces (Shapiro and Domek, 2002). This degradation is accelerated in tropical agroecosystems, where intense solar exposure may reduce viral activity to less than 24 h (Lacey et al., 2015). Extensive researches have been conducted to enhance the UV resistance of baculoviruses, including the incorporation of additives, such as carbon black, dyes, fluorescent brighteners, lignin, and nanoparticles (Behle and Birthisel, 2014). While these interventions have demonstrated varying degrees of success in laboratory settings by providing effective protection against UV-induced viral inactivation, field trials have yielded inconsistent results, with no definitive evidence of improved efficacy (Behle and Popham, 2012). This discrepancy may be attributed to the application method; in field applications where baculovirus OBs were formulated with UV protectants, initial leaf surface spraying provided partial UV shielding of OBs through liquid-phase protection. However, as water evaporates, most OBs are exposed to UV radiation again, resulting in failure to effectively prolong viral persistence (Lacey et al., 2008).
Another critical challenge in formulating UV protectants with biopesticides is the substantial economic burden imposed by additive requirements. When applied at recommended concentrations alongside the elevated water volumes characteristic of biopesticide application (typically 400-1000 L/ha), the cumulative quantities and associated costs of these additives escalate substantially (Leggett et al., 2011). Notably, UV protectant dosage requirements ranging from 5 to 20 kg/ha frequently result in formulation costs that often exceed the cost of the active ingredient by an order of magnitude (Reid et al., 2023). This cost disparity renders these combinations economically unsustainable for most commercial agricultural operations. To overcome this limitation, numerous studies have employed encapsulation strategies to enhance the UV resistance of baculovirus OBs (Burges and Jones, 1998). Despite their demonstrated efficacy in laboratory settings, these approaches have not yet been transitioned to commercial adoption. This implementation gap can be attributed to multiple technical barriers, including, but not limited to, prohibitive production costs, phytotoxic effects of encapsulation materials, incompatibility with long-term storage protocols, and operational challenges such as clogging issues in spraying equipment (Lacey et al., 2015).
Nano-SiO 2 is widely regarded as a promising material due to its advantageous properties, including low cost, high specific surface area, chemical inertness, good mechanical properties, thermal stability, and nonpolluting nature. Furthermore, nano-SiO 2 exhibits very low photocatalytic activity, as well as high UV and infrared ray shielding and reflecting abilities (Cai et al., 2021;Gao et al., 2020). In addition, nano-SiO 2 has many hydroxyl groups on its surface, known as silanols, which impart hydrophilicity and serve as reactive sites for further alteration into other useful functionalities. The modified functional groups on the surfaces of silica particles play a critical role in the specific properties of composite materials, such as hydrophilicity, hydrophobicity, chemical binding ability (Lee and Yoo, 2016). Moreover, the effect of nanoparticles addition on the final properties of the polymeric composite depends on both particle size and surface modifications (Bracho et al., 2012). Owing to these characteristics, nano-SiO 2 is particularly suitable for applications that require ultraviolet resistance and thermal aging protection.
In this study, monodisperse SiO 2 NPs were synthesized using the St€ ober method, followed by surface modification with amino groups using a silane coupling agent. The amino-functionalized silane (SiO 2 -NH 2 ) NPs were subsequently applied to the OBs, forming a UVprotective core-shell architecture. Comparative UV stability tests revealed that SiO 2 -NH 2 -coated OBs exhibited significantly enhanced UV resistance compared to uncoated OBs and OBs suspensions physically mixed with unmodified SiO 2 NPs. This surface engineering approach not only establishes a novel strategy for improving the field durability of baculovirus-based insecticides against solar UV degradation but also offers a solution to address the critical limitation of baculoviruses in agricultural application scenarios.
## RESULTS
## Silica nanoparticles were synthesized and amino-functionalized
For the aim of improving the UV resistance of baculoviruses, we attempted to coat the SiO 2 NPs on the OBs surface. The overall construction strategy of amino-functionalized SiO 2 NPs-coated OBs is shown in Fig. 1.
The St€ ober method is regarded as one of the simplest and most effective approaches for synthesizing monodispersed silica spheres (Zou et al., 2008). The SiO 2 NPs were synthesized using the St€ ober method exhibited excellent dispersibility in water. Scanning electron microscopy (SEM) showed that the synthesized SiO 2 NPs were spherical with uniform morphology and a particle diameter of approximately 53.9 ± 6.7 nm (Fig. 2A). The SiO 2 NPs were further functionalized with amino groups using the silane coupling agent (3-trimethoxysilylpropyl)diethylenetriamine (DETA). SEM analysis confirmed that amino-functionalization did not alter the dispersibility and surface morphology of the SiO 2 NPs (Fig. 2B). Furthermore, we measured the zeta potentials of both the unmodified and amino-modified SiO 2 NPs under different pH conditions. The unmodified SiO 2 NPs exhibited an isoelectric point (pI) at pH 3.4 (Fig. 2C), whereas the amino-functionalized SiO 2 NPs (SiO 2 -NH 2 NPs) showed a significantly higher pI at pH 8.1 (Fig. 2D). This upward shift in the pI confirms the successful amino-functionalization of the nanoparticle surface.
## Baculovirus OBs were successfully coated with amino-silica nanoparticles by electrostatic interactions
To determine the appropriate coating conditions, we measured the zeta potentials on the surfaces of OBs from four virus species under different pH conditions and found that the relationship between the zeta potential of OBs from all four viruses and the pH values followed a logistic curve model. The pIs on the OBs surfaces of Autographa californica multiple nucleopolydrovirus (AcMNPV), Mamestra brassicae multiple nucleopolyhedrovirus (MbMNPV), Dendrolimus punctatus cypovirus 1 (DpCPV-1), and Pieris rapae granulovirus (PiraGV) were 4.2 (Fig. 3A), 3.8 (Figs. 3B), 4.0 (Fig. 3C), and 3.2 (Fig. 3D), respectively.
Based on the difference in pIs between the OBs and SiO 2 -NH 2 NPs, the surface of the OBs might be coated with SiO 2 -NH 2 NPs via electrostatic interactions. Notably, the pH of the MES buffer was adjusted to 5.0 to ensure that SiO 2 -NH 2 NPs with higher pIs carried sufficient positive charges, while OBs with lower pIs carried sufficient negative charges, thereby promoting effective coating. SEM observations showed that AcMNPV, MbMNPV, DpCPV1, and PiraGV OBs (Fig. 4A-D) were successfully coated with SiO 2 -NH 2 NPs in the MES buffer system (Fig. 4E-H). Uncoated SiO 2 -NH 2 NPs were removed from the suspension by centrifugation. The SiO 2 -NH 2 coated MbMNPV OBs (Mb@SiO 2 -NH 2 ) were used in further experiments.
## Silica nanoparticles coated OBs showed structural stability post extensive washing and prolonged storage
Following three centrifugal washing cycles, Mb@SiO 2 -NH 2 retained the SiO 2 -NH 2 nanoparticles coating on the viral OB surfaces, conclusively validating the electrostatic interaction methodology as a robust strategy for achieving a stable nanoparticle coating (Fig. 5A). In addition, we evaluated the stability of Mb@SiO 2 -NH 2 in aqueous dispersions under prolonged storage conditions using SEM. The results revealed that the SiO 2 -NH 2 NPs on viral OBs retained coating integrity after storage for 30 days (Fig. 5B), 1 year (Fig. 5C) and even after 3 years (Fig. 5D). The above results indicate that the SiO 2 NPs can be stably coated on the surfaces of baculovirus OBs by electrostatic interaction.
## Silica nanoparticles coated OBs showed improved UV resistance
To verify whether coating SiO 2 -NH 2 NPs onto the surface of OBs through electrostatic interaction can enhance UV resistance of OBs, laboratory bioassays against S. exigua 2nd instar larvae were carried out using Mb@SiO 2 -NH 2 , MbMNPV alone and MbMNPV/SiO 2 NP mixture with or without UV treatment. The results showed that there is no significant difference in LC 50 values between MbMNPV alone and Mb@SiO 2 -NH 2 without UV exposure groups, indicating that SiO 2 -NH 2 NPs coated MbMNPV retained its native viral pathogenicity against the host larvae, while the MbMNPV/SiO 2 NP mixture exhibited 2.124-2.147 folds higher insecticidal potency with a lower LC 50 value than MbMNPV control. After UV irradiation, both MbMNPV/SiO 2 NPs mixture and Mb@SiO 2 -NH 2 exhibited significantly lower potency than the non-irradiated treatments, whereas Mb@SiO 2 -NH 2 retained superior bioactivity with 2.299-2.712 fold higher potency than the MbMNPV/ SiO 2 NP mixture (Table 1). These results indicate that the SiO 2 -NH 2 coating provided superior UV protection for the OBs.
## Survival time of silica nanoparticles coated OBs was greatly prolonged under natural conditions
Based on the laboratory bioassays results, we further conducted outdoor pot experiments on MbMNPV and Mb@SiO 2 -NH 2 to analyze whether coating SiO 2 -NH 2 NPs could improve the UV resistance of the OBs under natural conditions. First, we analyzed the relationship between the known density of MbMNPV/Mb@SiO 2 -NH 2 OBs on the leaf surface and the mortality of S. exigua larvae under laboratory conditions. Because the slopes and intercepts of the probit regression for MbMNPV and Mb@SiO 2 -NH 2 were not significantly different (χ 2 = 4.59, d.f = 2, P = 0.101), the data from the two tests were pooled to generate a regression equation: y = 1.334x -1.329, where y represents the probit of mortality of the tested larvae and x represents the log 10 -transformed OB density on the leaf surface.
Under outdoor conditions, virus suspensions containing MbMNPV or Mb@SiO 2 -NH 2 were sprayed on the leaves of pak choi, and the activities of OBs were measured by taking the leaves of pak choi to feed S. exigua 2nd instar larvae at different time points of sunlight exposure. The percentage of the original activity remaining (%OAR) of MbMNPV OBs might fit a typical exponential decay pattern, whereas Mb@SiO 2 -NH 2 maintained 100% OAR in the first 4 days, followed by an exponential decay pattern thereafter (Fig. 6). For the Mb@SiO 2 -NH 2 treatment, there were too few points to establish an exponential formula, and we established the median survival times (τ 1/2 ) of both treatments using interpolation by the %OAR line charts. The results showed that the average median survival time (τ 1/2 ) of Mb@SiO 2 -NH 2 reached 5.15 ± 0.05 days, which was significantly longer than MbMNPV's 1.43 ± 0.20 days (t = 31.678, P = 0.001). These results indicate that, by coating SiO 2 -NH 2 NPs, the survival time of MbMNPV under field conditions was greatly prolonged, and its ability to resist UV was significantly improved. AcMNPV@SiO 2 -NH 2 (E), MbMNPV@SiO 2 -NH 2 (F), DpCPV1@SiO 2 -NH 2 (G), PiraGV@SiO 2 -NH 2 (H). Scale bars, 1 μm. Mb@SiO 2 -NH 2 after 30-days storage. C Mb@SiO 2 -NH 2 after 1-year storage. D Mb@SiO 2 -NH 2 after 3-years storage. Scale bars, 1 μm. a Indicates that the potency of the treatments was significantly different from that of the control treatment (MbMNPV or MbMNPV/SiO 2 +UV), based on whether the 95% confidence interval of the potency ratio was 1.0. b Mortality of larvae inoculated with pure MbMNPV at the highest concentration was lower than 50% after UV irradiation, thus the LC 50 of this treatment was not applicable.
## DISCUSSION
Entomopathogenic virus-based insecticides are an environmentally sustainable pest control solution. However, their susceptibility to solar UV radiation remains a critical limitation to broader field applications. In this study, we developed a novel strategy to enhance the UV stability of entomopathogenic virus OBs by coating them with SiO 2 -NH 2 NPs that possess inherent UV-shielding properties. This protective coating significantly prolongs viral persistence under UV exposure, while maintaining bioactivity, thereby offering a promising approach to effectively address the inherent photosensitivity of baculoviruses.
The surface of baculovirus OBs is primarily composed of a glycoprotein matrix known as a polyhedrin membrane. In aqueous environments, the exposed regions of this protein shell contain hydrophilic amino acid residues, whose side chains often contain reactive functional groups such as amino (-NH 2 ) and carboxyl (-COOH). These functional groups provide viable anchoring sites for chemical conjugation and surface modification (Chen et al., 2012). Previous studies have reported covalent conjugation of proteins to carboxylated nanomaterials via carbodiimide-HCl/N-hydroxysuccinimide (EDC/NHS) coupling (Wang et al., 2008). Based on these findings, we initially carboxylated SiO 2 NPs using carboxyethylsilanetriol sodium salt (CES) and attempted to bind them to OBs via EDC/NHS-mediated linkage. Despite extensive optimization of reaction conditions, such as time, temperature, pH, and crosslinker concentration, effective coating of the OBs could not be achieved.
This failure is likely due to electrostatic repulsion. The pI of baculovirus OBs ranges from pH 3 to 4, and at neutral pH, both OBs and carboxylated SiO 2 NPs carry negative surface charges, which inhibits conjugation (Rohrmann, 2019). To overcome this issue, we modified SiO 2 NPs (pI = 3.4) with DETA, introducing amine groups and generating positively charged SiO 2 -NH 2 NPs (pI = 8.1). These nanoparticles efficiently coated the OBs, forming OB@SiO 2 -NH 2 composites even without EDC, indicating that electrostatic interactions dominated the coating process. Nevertheless, at neutral pH, excess unbound SiO 2 -NH 2 NPs caused OB aggregation, impairing the colloidal stability and usability of the resulting OB@SiO 2 -NH 2 composites.
We further optimized the coating process using MES buffer (pH 4.5-5.5), which enhanced the positive charge of SiO 2 -NH 2 NPs and minimized aggregation. OBs maintained their biological activity across a pH range of 4-9 (Braga and Moscardi, 2002). SEM observations confirmed that SiO 2 -NH 2 NPs remained stably anchored on the OB surface after extensive washing and prolonged storage, validating the robustness of the coating.
While MbMNPV/SiO 2 NPs mixtures demonstrated enhanced potency compared to the virus alone in laboratory bioassays, SiO 2 -NH 2 coated OBs showed efficacy comparable to that of pure MbMNPV. This discrepancy may arise from the substantially higher NP concentrations in simple OBs and SiO 2 NP mixtures compared with our precisely coated preparations, where unbound NPs were removed. Previous studies have indicated that foliar SiO 2 NP application (1 mg/cm 2 ) can induce 58%-85% larval mortality in Plutella xylostella through physical mechanisms such as cuticular abrasion and spiracle blockage (Shoaib et al., 2018). Of course, in practical application, it is not necessary to remove the unbound NPs post coating, which may provide extra sunlight protection of the OBs and enhance the insecticidal efficacy as well.
The environmental safety profile of silica nanomaterials warrants further consideration. Certain metallic nanoparticles raise ecotoxicological concerns (e.g., silver NPs causing DNA damage) (Banumathi et al., 2017;Benelli, 2018). Application of SiO 2 NPs on the leaf and stem surfaces did not alter either photosynthesis or respiration of horticultural and crop plants; they also did not alter gene expression in the insect trachea. Furthermore, amorphous silica maintains an established safety profile and is classified by the World Health Organization (WHO) as a generally recognized safe nanomaterial for agricultural applications (Athanassiou et al., 2018).
In this study, we successfully established a robust and broadly applicable method for anchoring SiO 2 -NH 2 NPs to baculovirus OBs. Notably, the functionalized NPs coating maintained structural integrity and UV shielding functionality even upon suspension destabilization, ensuring prolonged viral protection. Li et al. (2015) developed a recombinant AcMNPV by fusing a nano-ZnO-binding peptide to the NM domain of the polyhedron envelope protein (PEP). This genetic modification enabled the specific binding of commercially available nano-ZnO particles to the surface of occlusion bodies (OBs), resulting in significantly enhanced UV resistance of the engineered baculovirus (Li et al., 2015). However, it should be noted that prior to potential agricultural applications, such recombinant baculovirus strains must undergo rigorous biosafety evaluations to ensure environmental safety and regulatory compliance.
The objective of amino-functionalizing the SiO 2 NPs in this study was to impart a positive surface charge, thereby enabling them to coat the OBs through electrostatic interaction. Our results demonstrate that this intentional surface modification not only achieved this aim but also preserved the core functional advantages of the silica nanoparticles, such as their UV-shielding properties and storage stability.
Our coating methodology demonstrated universal applicability across major entomopathogenic viral types, including Mamestra brassicae multiple nucleopolyhedrovirus (MbMNPV) (Alphabaculovirus mabrassicae, genus Alphabaculovirus), Autographa californica multiple nucleopolydrovirus (AcMNPV) (Alphabaculovirus aucalifornicae, genus Alphabaculovirus), Dendrolimus punctatus cypovirus 1 (DpCPV-1) (Cypovirus altineae, genus Cypovirus), and Pieris rapae granulovirus (PiraGV) (Betabaculovirus arrapae, genus Betabaculovirus). This technique significantly expands the operational scope of baculovirus-based insecticides by overcoming critical limitations of recombinant technologies. Field persistence tests revealed remarkable UV protection: SiO 2 -NH 2 coating extended MbMNPV's environmental half-life from 1.43 to 5.15 days. The observed efficacy decline after day 4 may reflect cumulative UV penetration through inter-NP gaps, eventually damaging the OB-encapsulated viral DNA. This suggests opportunities for further optimization through tighter NP packing or secondary protective layers.
## CONCLUSIONS
In this study, we established a novel strategy of coating OBs with SiO 2 nanoparticle by shifting the isoelectric point through aminomodification. The SiO 2 -NH 2 NPs coated OBs showed structural stability post extensive washing and prolonged storage. Additionally, these modified OBs showed improved UV resistance in laboratory and greatly prolonged the OBs survival time under natural conditions. This strategy establishes a robust platform for developing photostable biopesticides.
## MATERIALS AND METHODS
## Viruses, insects, and plants
MbMNPV Chb1 strain, AcMNPV E2 strain, DpCPV-1, andPiraGV were preserved in our laboratory. S. exigua larvae were commercially acquired from Henan Keyun Biological Co., Ltd., and reared under controlled environmental conditions (27 ± 1 • C, 70% RH) using a nutritionally optimized artificial diet. For the outdoor trials, Brassica rapa var. chinensis (pak choi) seeds obtained from Harbin Nongxin Seed Co., Ltd. were germinated in standardized growth chambers and transplanted into 170mm diameter circular pots containing sterilized substrates 28 days before experimental implementation. 2-Morpholinoethanesulfonic acid (MES) was purchased from Shanghai Aladdin Bio-Chem Technology Co., Ltd. (Shanghai, China). DETA was purchased from Shanghai Acmec Biochemical Technology Co., Ltd. (Shanghai, China). Anhydrous ethanol, tetraethyl orthosilicate (TEOS), and glacial acetic acid were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Ammonium hydroxide was purchased from Shanghai Macklin Reagent Co., Ltd. (Shanghai, China).
## OBs amplification
Baculovirus and cypovirus OBs were amplified through in vivo propagation using the 3rd-instar of their corresponding host larvae. Post-mortem purification involved collecting liquefied cadaver homogenates, followed by multi-step differential centrifugation as per the standardized baculovirus purification protocol (Gross et al., 1994).
## Synthesis of SiO 2 NPs
SiO 2 nanoparticles were synthesized using a modified St€ ober protocol (St€ ober et al., 1968). The reaction system was established in a borosilicate Erlenmeyer flask by sequentially introducing 50 mL anhydrous ethanol, 2 mL ammonium hydroxide, and 1 mL deionized water, followed by homogenization and thermal equilibration at 40.0 ± 1 • C in an orbital shaker (150 rpm) for 30 min. Subsequently, TEOS (1.5 mL) was rapidly injected into the preheated solution to initiate the silica particle synthesis. The reaction proceeded under continuous agitation (150 rpm) at 40.0 ± 1 • C for 12 h to form monodisperse spherical silica particles by hydrolysis and condensation of TEOS in the presence of anhydrous ethanol and ammonia. Post-synthesis purification involved centrifugation (18500×g, 10 min) for primary particle isolation. The pellet was subjected to three successive purification cycles using anhydrous ethanol and deionized water with centrifugation (18500×g, 10 min) to remove the residual reactants and byproducts. The final nanoparticle suspension was prepared in sterile ultrapure water and preserved under aseptic conditions at 4.0 • C for subsequent functionalization procedures.
## Amino modification of SiO 2 NPs
Surface amino-modification of SiO 2 NPs was performed following established silanization protocols with process optimization (Wang et al., 2007). Approximately 10 mg of SiO 2 NPs was dispersed in 1 mL of deionized water with sonication and mixed with 20 μL glacial acetic acid and 20 μL DETA. The reaction mixture was subjected to continuous rotational incubation (26 ± 1 • C, 50 rpm) for 4 h to ensure complete silane coupling. After the amination process, the particles were washed and dispersed in 10 mM MES buffer (pH 5.5 adjusted by 0.1 mol/L NaOH) and stored at 4 • C for subsequent bioconjugation applications.
## Determination of zeta potential of NPs and OBs surface
Zeta potential characterization was performed using a Zetasizer Nano ZSP analyzer (Malvern Panalytical Ltd., UK) (Garg et al., 2024). Sample preparation involved loading 1 mL of 1 mg/mL nanoparticle dispersion in MES buffer or 1 × 10 9 OBs/mL virus suspension into precision-folded capillary zeta cells, both preconditioned at target pH levels (2.0-10.0, adjusted with 0.1 mol/L HCl or NaOH). Between the measurements, the flow cell was rigorously rinsed with deionized water to eliminate cross-contamination. The tests were conducted in triplicate.
## Coating of OBs with SiO 2 -NH 2 NPs
The coating protocol was initiated by dispersion of 4 mg SiO 2 -NH 2 NPs in 0.5 mL 0.1 mol/L MES buffer (pH 5.0) through sonication, 0.5 mL of OBs stock solution (5 × 10 8 OBs/mL in 0.1 mol/L MES buffer, pH 5.0) was introduced and mixed by vortexing. The reaction system underwent continuous rotational conjugation (25 • C, 50 rpm) for 2 h to facilitate nanoparticle assembly on OB surfaces. Post-conjugation purification involved centrifugation (5000×g, 5 min) to isolate OBs@SiO 2 -NH 2 complexes. Three successive wash cycles with 10 mL of deionized water were performed under identical centrifugal conditions to eliminate the unbound nanoparticles. The final composite bioparticles were resuspended in deionized water and maintained at 4 • C before the bioactivity evaluation.
## SEM analysis
Synthesized SiO 2 NPs, amino-functionalized SiO 2 NPs, surfacecoated OB@SiO 2 -NH 2 complexes, and pure OBs were subjected to morphological characterization by field-emission SEM (SU-8010; Hitachi, Japan) at 3 kV, as described previously (Kuang et al., 2017). The diameter of NPs was measured using ImageJ software, wherein 100 representative NPs were measured across the electron micrographs to determine the mean particle dimensions.
## Structural stability detection
To test the structural stability, the OB@SiO 2 -NH 2 composite was subjected to three cycles of centrifugation (100×g, 5 min for the supernatant, followed by 5000×g, 10 min for the pellet) with deionized water before SEM characterization of its surface coating morphology.
The storage stability was evaluated through SEM imaging after dispersing the OB@SiO 2 -NH 2 particles in deionized water and storing the suspension at 4 • C for 30 days, 1 year and 3 years.
## Laboratory bioassay
The relationship between virus concentration and insect mortality was determined using a modified droplet-feeding method (Hughes and Wood, 1981). Enough 2nd instar S. exigua larvae were starved in a 27 ± 1 • C incubator for 16 h. The OBs of MbMNPV and Mb@SiO 2 -NH 2 were counted using a hemocytometer, and the virus suspension was diluted to 1 × 10 4 , 3 × 10 5 , 1 × 10 5 , 3 × 10 5 , and 1 × 10 6 OBs/mL with 0.01% Tween 20-PBS solution (containing 10% sucrose). An appropriate amount of erioglaucine disodium salt (food coloring) was added to obtain a deep-blue virus suspension. The blank control was a 0.01% Tween 20-PBS solution containing only sucrose and erioglaucine disodium salt. The starved larvae were dosed with a dilution series of OBs for 10 min in a Petri dish and the blue-turned larvae were transferred to a 24-well plate containing fresh artificial diets and incubated at 27 ± 1 • C.
For UV irradiation treatments, 1 mL of MbMNPV or Mb@SiO 2 -NH 2 suspension at a concentration of 1 × 10 8 OBs/mL was placed in a 35 × 12 mm cell culture dish and irradiated with one BOT UVB 20 W G13 T8 germicidal bulb (BOT, Beijing, China) at 303 nm for 30 min. The UV light dose was determined using a UV light meter (UV-340A, Lutron, Taipei, China). To make the MbMNPV/SiO 2 NPs mixture, 4 mg SiO 2 NPs with a diameter of 50 ± 5 nm (Macklin Reagent Co., Ltd., Shanghai, China) were mixed with 1 mL of OBs at a concentration of 1 × 10 8 OBs/mL in an Eppendorf tube and agitated until homogeneous. MbMNPV, Mb@SiO 2 -NH 2 , and MbMNPV/SiO 2 NP mixture were serially diluted with PBS to 1 × 10 5 , 3 × 10 5 , 1 × 10 6 , 3 × 10 6 , and 1 × 10 7 OBs/mL. All bioassay experiments, with or without UV irradiation, were conducted in duplicate. The median lethal concentration (LC 50 ) and potency ratio data were analyzed by probit analysis using Polo-Plus (Robertson et al., 2007).
## Pot experiments under sunlight irradiation
Plant preparation and virus application: Pak choi plants were cultivated in 170-mm round pots for four weeks before the experiments. Suspensions of MbMNPV and Mb@SiO 2 -NH 2 were prepared by dilution with 0.05% (v/v) Tween 20-PBS to a final concentration of 1 × 10 7 OBs/ mL. The suspensions were evenly sprayed onto the leaves of potted pak choi plants placed in the open field every two days from November 24, 2021, to December 2, 2021, in Wuhan, Hubei, China (30 • 44 ′ N, 114 • 26 ′ E). Virus applications were conducted between 7:00 and 8:00 h with a spray deposition rate of 3.35 μL/cm 2 . During this period, the weather conditions were consistently sunny, with midday (13:00 h) UV irradiance of 900 ± 15 μW/cm 2 . Leaf samples were collected on December 2, 2021, resulted the sample exposed for 8, 6, 4, 2, and 0 days post spray, and 6-mm leaf discs were excised using a hole punch. These discs were placed in 24-well plates containing 1% agar (0.5 mL), and a single 2nd instar S. exigua larva was introduced into each well. After a 24 h feeding period at 27 ± 1 • C, larvae that had consumed the leaf discs were transferred to a fresh artificial diet in new 24-well plates. Mortality was recorded every 48 h until all larvae died or pupated. The experiment was repeated thrice, with 48 larvae per treatment in each replicate.
Calibration of virus efficacy in the laboratory: To establish a doseresponse relationship, MbMNPV/Mb@SiO 2 -NH 2 suspensions at concentrations of 1 × 10 4 , 3 × 10 5 , 1 × 10 5 , 3 × 10 5 , and 1 × 10 6 OBs/mL were applied to pak choi leaves under controlled laboratory conditions. The bioassay protocol was similar to that used in the field experiments. To determine the actual OBs density on the leaf surfaces, six pak choi plants were sprayed with 0.05% Tween 20 under conditions identical to those used in the outdoor pot experiments. The leaves were weighed before and after spraying to calculate the deposited suspension volume per unit leaf area (μL/cm 2 ). The OB density (OBs/cm 2 ) on the leaf surface was derived by multiplying the applied virus concentration by the measured deposition volume. A standard calibration curve was established by fitting the log 10 -transformed OB density on the leaf surface to the probit values, corresponding to larval mortality. Equality of the probit regressions (equal slopes and equal intercepts) for MbMNPV and Mb@SiO 2 -NH 2 were test by a χ 2 test using Polo-Plus (Robertson et al., 2007).
Persistence and half-life analysis: The infectious OB density on leaves exposed to sunlight for varying durations was inferred from larval mortality data using the calibration curve. The %OAR was calculated, with the 1-h post-spray value normalized to 100% (Akhanaev et al., 2017). The median survival time (τ 1/2 ) was determined by interpolating the time required for the %OAR to decrease to 50% (Thompson, 1947).
A t-test was used to assess differences in the half-lives between MbMNPV and Mb@SiO 2 -NH 2 .
## References
1. Akhanaev, Belousova, Ershov et al. (2017) "Comparison of tolerance to sunlight between spatially distant and genetically different strains of Lymantria dispar nucleopolyhedrovirus" *PLoS One*
2. Athanassiou, Kavallieratos, Benelli et al. (2018) "Nanoparticles for pest control: current status and future perspectives" *J. Pest. Sci*
3. Banumathi, Vaseeharan, Suganya et al. (2017) "Toxicity of camellia sinensisfabricated silver nanoparticles on invertebrate and vertebrate organisms: morphological abnormalities and DNA damages" *J. Cluster Sci*
4. Behle, Birthisel (2014) "Formulations of entomopathogens as bioinsecticides"
5. Behle, Popham (2012) "Laboratory and field evaluations of the efficacy of a fast-killing baculovirus isolate from Spodoptera frugiperda" *J. Invertebr. Pathol*
6. Benelli (2018) "Mode of action of nanoparticles against insects" *Environ. Sci. Pollut. Res. Int*
7. Bracho, Dougnac, Palza et al. (2012) "Functionalization of silica nanoparticles for polypropylene nanocomposite applications" *J. Nanomater*
8. Braga, Moscardi (2002) "Field efficacy of the nucleopolyhedrovirus of Anticarsia gemmatalis Hübner (Lepidoptera: Noctuidae): effect of formulations, water pH, volume and time of application, and type of spray nozzle" *Neotrop. Entomol*
9. Burges, Jones (1998) "Formulation of bacteria, viruses and protozoa to control insects"
10. Cai, Gao, Chen et al. (2021) "An effective, lowcost and eco-friendly method for preparing UV resistant silk fabric" *J. Nat. Fibers*
11. Chen, Cao, Liu et al. (2012) "Rotavirus capsid surface protein VP" *Biomaterials*
12. Gao, Bao, Cai et al. (2020) "Multifunctional silk fabric via surface modification of nano-SiO2" *Textil. Res. J*
13. Garg, Patel, Gupta et al. (2024) "Pharmaceutical applications and advances with zetasizer: an essential analytical tool for size and zeta potential analysis"
14. Gross, Russell, Rohrmann (1994) "Orgyia pseudotsugata baculovirus p10 and polyhedron envelope protein genes: analysis of their relative expression levels and role in polyhedron structure" *J. Gen. Virol*
15. Hughes, Wood (1981) "A synchronous peroral technique for the bioassay of insect viruses" *J. Invertebr. Pathol*
16. Kuang, Zhang, Wang et al. (2017) "Three conserved regions in Baculovirus Sulfhydryl oxidase P33 are critical for enzymatic activity and function" *J. Virol*
17. Lacey, Grzywacz, Shapiro-Ilan et al. (2015) "Insect pathogens as biological control agents: back to the future" *J. Invertebr. Pathol*
18. Lacey, Thomson, Vincent et al. (2008) "Codling moth granulovirus: a comprehensive review" *Biocontrol Sci. Technol*
19. Lapointe, Thumbi, Lucarotti (2012) "Recent advances in our knowledge of baculovirus molecular biology and its relevance for the registration of baculovirusbased products for insect Pest population control"
20. Lee, Yoo (2016) "Advanced silica/polymer composites: materials and applications" *J. Ind. Eng. Chem*
21. Leggett, Leland, Kellar et al. (2011) "Formulation of microbial biocontrol agentsan industrial perspective" *Can J Plant Pathol*
22. Li, Zhou, Lei et al. (2015) "Improvement in the UV resistance of baculoviruses by displaying nano-zinc oxide-binding peptides on the surfaces of their occlusion bodies" *Appl. Microbiol. Biotechnol*
23. Reid, Malmanche, Chan et al. (2023) "Production of entomopathogenic viruses"
24. Robertson, Savin, Preisler (2007) "Bioassays with Arthropods"
25. Rohrmann (2019) *Baculovirus Molecular Biology*
26. Shapiro, Domek (2002) "Relative effects of ultraviolet and visible light on the activities of corn earworm and beet armyworm (Lepidoptera: noctuidae) nucleopolyhedroviruses" *J. Econ. Entomol*
27. Shoaib, Elabasy, Waqas et al. (2018) "Entomotoxic effect of silicon dioxide nanoparticles on Plutella xylostella (L.) (Lepidoptera: plutellidae) under laboratory conditions" *Toxicol. Environ. Chem*
28. St€ Ober, Fink, Bohn (1968) "Controlled growth of monodisperse silica spheres in the micron size range" *J. Colloid Interface Sci*
29. Thompson (1947) "Use of moving averages and interpolation to estimate medianeffective dose: I. Fundamental formulas, estimation of error, and relation to other methods" *Bacteriol. Rev*
30. Wang, Zhao, O'donoghu et al. (2007) "Fluorescent nanoparticles for multiplexed bacteria monitoring" *Bioconjug. Chem*
31. Wang, Zhao, Tan (2008) "Bioconjugated silica nanoparticles: development and applications" *Nano Res*
32. Zou, Wu, Shen (2008) "Polymer/silica nanocomposites: preparation, characterization, properties, and applications" *Chem. Rev* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607797&blobtype=pdf | # Comprehensive metabolomics combined with machine learning for the identification of SARS-CoV-2 and other viruses directly from upper respiratory samples
Catherine Hogan, Anthony Le, Afraz Khan, Linghui Su, Chunhong Huang, Malaya Sahoo, Chieh-Wen Lo, Marwah Karim, Karin Stein, Shirit Einav, Tina Cowan, Benjamin Pinsky
## Abstract
Metabolic profiling of respiratory samples from individuals infected and uninfected with respiratory viral infections may identify biomarker signatures that complement routine clinical diagnostic testing and offer unique insights into patho physiology. We used liquid chromatography quadrupole time-of-flight mass spectrome try to generate untargeted metabolomic profiles and identified top biomarker signatures differentiating severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) positive from negative samples via machine learning. We then adapted these signatures to liquid chromatography-tandem mass spectrometry for targeted profiling and assessed classification performance, including samples positive for other respiratory viruses and negative for viral testing. A total of 1,226 samples were tested, including 521 positive samples for SARS-CoV-2, 97 for influenza A, 96 for respiratory syncytial virus (RSV), 211 for other respiratory viruses, and 301 negative samples. The top-performing model was the Light Gradient Boosting Model, which showed an area under the receiver operating characteristic curve (AUC) of 0.99 (95% confidence interval [CI], 0.99-1.00), sensitivity of 0.96 (95% CI, 0.91-0.99), and specificity of 0.95 (95% CI, 0.90-0.97). A separate machine learning analysis investigating the performance by viral subtype showed high perform ance for the identification of influenza A virus with an AUC of 0.97 (95% CI, 0.94-0.99) and RSV with an AUC of 0.99 (95% CI, 0.97-1.00). The two features with the highest ranking were identified as 3-oxo-heneicosanoic acid and 2-(4-hydroxyphenyl) ethanol. These findings extend our understanding of the metabolic impact of respiratory viral infections and support the potential of metabolomics to complement routine clinical diagnostic methods.IMPORTANCE Molecular testing has greatly improved how viruses are diagnosed; however, gaps remain, including limited sensitivity directly from specimens and inability to differentiate active from resolved infection. In this study, we investigated the use of a distinct diagnostic approach, mass spectrometry for detection of metabolites (small molecules) combined with machine learning analysis, for the diagnosis of SARS-CoV-2 and other respiratory viruses. We demonstrated strong performance of this approach directly from upper respiratory swab samples to differentiate SARS-CoV-2-infected versus uninfected individuals. Extension of this approach to influenza and RSV maintained a high level of performance. This research suggests that mass spectrometry-based infectious disease diagnostic testing has clinical potential and that these metabolomic features may reveal novel host-pathogen interactions and therapeutic targets. Applying a similar approach to prospective, multisite cohorts of patients with other infectious diseases carries potential to extend our understanding of the metabolic pathways involved in the host response to infection.
KEYWORDS metabolomics, nasopharyngeal, SARS-CoV-2, COVID-19, variants of concern, machine learning, respiratory viruses T he clinical presentations of respiratory viral infections overlap substantially, and accurate diagnostic testing is required to identify the causal etiological agent for better clinical management. This is particularly important in an era of increased accessibility of targeted antiviral therapy, including for influenza and coronavirus disease 2019 (COVID-19) (1,2). Respiratory viruses have been associated with distinct metabolo mic profiles that may differentiate infected from uninfected individuals directly from respiratory samples, serum, and plasma (3)(4)(5). Direct testing of upper respiratory tract samples is of clinical interest as it is less invasive than phlebotomy, aligns with current routine testing practice for respiratory virus infections, and may closely correlate with metabolomic changes at the site of viral infection as investigated previously for influenza (3).
For severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2), a consistent trend across studies has been the finding of reduced circulating serotonin levels in acute infection, including as a marker of clinical severity (6,7) and in individuals with long COVID-19 (8), likely driven by several mechanisms, including upregulation of interferon genes. These studies were performed during a period where SARS-CoV-2 dominated the epidemiology of respiratory viruses and largely in the absence of circulating influenza virus. Now that SARS-CoV-2 is part of the landscape of routine respiratory viruses, there is a need to investigate the performance of metabolomics for the identification of SARS-CoV-2 in the context of other circulating respiratory viruses.
In this study, we used liquid chromatography/quadrupole time-of-flight (LC/Q-TOF) on upper respiratory samples combined with machine learning analysis to identify a top biomarker feature signature as a potential discriminator between SARS-CoV-2-infec ted and uninfected individuals. This signature was then adapted to simpler and more sensitive testing by targeted liquid chromatography-tandem mass spectrometry (LC/ MS-MS, referred to as tandem mass spectrometry) in an independent validation cohort. We subsequently combined virus culture with tandem mass spectrometry for metabolic profiling of virally infected and uninfected cells. Finally, we pursued chemical synthesis of the top feature and performed tandem mass spectrometry testing to further confirm identification.
## MATERIALS AND METHODS
## Study population and sample collection
We identified individuals infected with SARS-CoV-2, influenza A, respiratory syncy tial virus (RSV), human metapneumovirus (hMPV), adenovirus (AdV), rhinovirus (RV), and parainfluenza viruses (PIVs) 1-4 from respiratory swab samples (including nasal, mid-turbinate and nasopharyngeal swabs) collected in viral transport medium at Stanford Health Care, Stanford Tri-Valley, Stanford Children's Health and affiliated clinics and outpatient centers in the San Francisco Bay Area. All samples were selected through convenience sampling, whereby available samples with sufficient residual volume were included. Samples from the biomarker signature discovery cohort were collected from symptomatic inpatient and outpatient individuals between 22 December 2021 and 14 January 2022. Negative samples in the discovery cohort were tested only for SARS-CoV-2 by nucleic acid amplification testing (NAAT). Samples for the validation cohort were collected from symptomatic inpatient and outpatient individuals between 23 December 2021 and 31 May 2023. Negative samples in the validation cohort were collected from symptomatic inpatient and outpatient individuals who tested negative for SARS-CoV-2, influenza A, influenza B, and RSV by NAAT. All samples were stored at -80°C until thawed for processing for metabolomics testing, and no formal freeze-thaw or sam ple stability testing was performed. SARS-CoV-2 NAAT was performed using one of several emergency-use authorized commercial SARS-CoV-2 tests, including the Aptima SARS-CoV-2 and Panther Fusion SARS-CoV-2 Assays (both from Hologic Inc., Marlbor ough, MA, USA), Xpert Xpress SARS-CoV-2, SARS-CoV-2 plus, SARS-CoV-2/Flu/RSV, and SARS-CoV-2/Flu/RSV plus (all from Cepheid, Sunnyvale, CA, USA), SARS-CoV-2 reverse transcription quantitative PCR (RT-qPCR) Reagent Kit (PerkinElmer, Shelton, CT, USA), New Coronavirus Nucleic Acid Detection Kit (Revvity, Waltham, MA, USA), and the ePlex Respiratory Pathogen Panel 2 (GenMark Diagnostics, Carlsbad, CA, USA). Genotyping was performed by RT-qPCR as part of laboratory variant surveillance, and SARS-CoV-2 whole-genome sequencing was performed on a subset of these samples, as previously described (9). NAAT for the other respiratory viruses was carried out using the Food and Drug Administration-cleared Hologic Panther Fusion FluA/B/RSV, AdV/hMPV/RV, and Paraflu Assays (Hologic, Inc.).
## Metabolic profiling by LC/Q-TOF on the discovery cohort
Metabolic profiling of samples was performed by LC/Q-TOF using an in-line two-col umn reversed-phase and ion-exchange method, as previously described (10). In brief, a volume of 50 µL of VTM was mixed with 200 µL of isopropanol and centrifuged at 4°C for 15 min at 17,000 × g. Chromatographic separation was performed using a series of two columns: column 1, an HSS T3, 1.8 µm, 2.1 mm ID × 50 mm (Waters Corporation, Milford, MA, USA); column 2, Intrada Amino Acid, 3 µm, 2 mm ID × 30 mm (Imtakt, Portland, OR, USA), on a Waters class H quaternary pump (Waters Corporation). Three eluants were used: eluant A: 4 mL formic acid, 0.1 mL ammonium hydroxide in 1 L water; eluant B: 2.7 mL formic acid, 0.74 mL ammonium hydroxide in 1 L methanol; and eluant C: 10 mL formic acid, 11 mL ammonium hydroxide in 1 L water. Detection was performed on a Waters Cyclic Ion Mobility Separation Quadrupole Time of Flight Mass Spectrometer (CIMS Q-TOF) with electrospray ionization (ESI). An injection volume of 5 µL was used with a runtime of 20 min. Samples were analyzed in positive ion mode in a randomized manner using a mass range of m/z 50-1,200 with the CIMS Q-TOF in V-mode, with a lock mass of leucine enkephalin (m + H/z = 556.2771 and m -H/z = 554.2615). Peak picking, peak alignment, and qualitative analysis were performed using Progenesis QI v.1.0.0.1 (Waters Corporation) using D5-pyroglutamate as an internal standard. The resulting table of retention time, accurate mass, peak area, and peak height was then directly exported from Progenesis QI to Microsoft Excel.
## Fragmentation
Fragmentation experiments of the leading 20 features were performed using a pooled sample consistent with all samples run in the LC/Q-TOF on the discovery cohort. The same LC/Q-TOF method as above was used but with a collision energy of 10, 20, and 30 V at the transfer cell of the Q-TOF. The 20 V collision energy produced the best results and was retained for the subsequent phase of the study. This collision energy is equivalent to a Q2 collision cell on a typical quadrupole time-of-flight mass spectrometer. Multiple fragments were identified for each biomarker using the known retention time and molecular ion of the desired biomarkers.
## Targeted testing by LC/MS-MS on the validation cohort
Using the Q1 molecular ion and identified Q3 fragment ions, the two-column method was repeated using tandem mass spectrometry. After confirmation of the Q1/Q3 selective reaction monitoring (SRM), a modified analytical method was used to reduce analysis time. Chromatographic separation was performed using a WARP 2.7 µm 90A C18 tapered ID × 50 mm column (Premier LCMS) on a Class H quaternary pump (Waters Corporation) for the full cohort. Two eluants were used: eluant A: 0.1% formic acid water; eluant B: 0.1% formic acid methanol. The flow rate was at 0.45 mL/min with a gradient formation of initial conditions of 99% A, 1% B; at 0.3 min, 99% A, 1% B; at 1.5 min, 82% A, 18% B; at 1.7 min, 82% A, 18% B; and at 2 min, 99% A, 1% B. Detection was performed on a Xevo TQ-XS Triple Quadrupole Mass Spectrometer with ESI (Waters Corporation). An injection volume of 5 µL was used with a runtime of 3.5 min. Samples were analyzed in positive ion mode.
## Identification of the top differentiating SARS-CoV-2 biomarkers
To determine the chemical structures of the top differentiating compounds, the mass-to-charge ratios (m/z) of the intact and fragmented metabolites were used to query the METLIN Gen2 database (https://metlin.scripps.edu/). Candidate biomarkers were then confirmed by obtaining the commercially available compounds, 2-(4-hydrox yphenyl) ethanol and 4-ethoxyphenol (Sigma-Aldrich, St. Louis, MO, USA), or, if not commercially available, synthesizing the compound internally (3-oxo-heneicosanoic acid, synthesis protocol). We pursued this approach with purchased or synthesized com pounds to confirm the biomarker signature and enable the most robust tier 1 match of the top feature of interest. Biomarker identification by mass spectrometry was performed using the same method as for the targeted testing by LC/MS-MS described above, with the exception of using an orthogonal column for confirmation. Chromatographic separation was performed using a WARP 2 2.7 µm 90A C18 tapered ID × 50 mm column (Premier LCMS) on a class H quaternary pump (Waters Corporation). The SRM pairs and retention times of these standards were compared to those observed in clinical samples.
## Virus culture
Once the biomarker signature was identified through untargeted and targeted metabolomics as above, we pursued virus culture combined with tandem mass spectrometry of the supernatant from SARS-CoV-2-infected and SARS-CoV-2-uninfected cells. A549-ACE2 (BEI Resources, NR-53821) and Vero E6-TMPRSS2 (JCRB Cell Bank, #JCRB1819) were maintained in Dulbecco's modified Eagle medium (DMEM) supplemen ted with 10% fetal bovine serum (FBS), 1% penicillin-streptomycin, and 1 mg/mL G418 (all cell culture reagents from Gibco Scientific, Waltham, MA, USA). All cell lines were maintained in a humidified incubator at 37°C with 5% CO 2 and tested negative for Mycoplasma by MycoAlert (Lonza, Morristown, NJ, USA). SARS-CoV-2/wild-type (USA-WA1/2020) viral stock was generated in Vero E6-TMPRSS2 cells, as previously described (11). Whole-genome sequencing confirmed no deletions in the spike multibasic cleavage domain (9).
A549-ACE2 cells were seeded at 10E4 cells per well. The next day, cells were infected with SARS-CoV-2 (USA-WA1/2020) at a multiplicity of infection of 0.05 in DMEM containing 2% FBS. After 2 hours of incubation at 37°C, the viral inoculum was removed, cells were washed with PBS, and 100 µL DMEM containing 10% FBS was added. Culture supernatants were harvested 24 hours post-infection, and the virus was inactivated by adding isopropanol to a final concentration of 80%, followed by a 10 min incubation at room temperature (12). Inactivated supernatants from cell culture with and without SARS-CoV-2 infection (SARS-CoV-2 infected, n = 2; SARS-CoV-2 uninfected, n = 2) were tested by tandem mass spectrometry as described above.
## Machine learning analysis
Data were exported from Excel to Python for machine learning (ML) analysis. Data preprocessing was performed using quantile scaling to output a normal distribution. To identify potential biomarkers for SARS-CoV-2 infection, ML analysis was first performed on the preprocessed data of the discovery cohort (LC/Q-TOF). The full data set was randomly divided into a training set (80% of samples) and a test set (20% of samples) to provide balance between the model learning effectively and reserving a separate subset for testing. The training set was used to build the machine learning models, and the test set was used to evaluate the test performance of the biomarker signature using the area under the receiver operating characteristic curve (AUC). Shapley additive explanations (SHAP) summary plots were generated for each machine learning model to highlight the top 20 features and their individual contribution, as is standard practice with this analysis to better understand the impact of each feature toward model prediction (13).
Four models were used to assess the top 20 features of SARS-CoV-2 infection: two statistical models (Lasso and logistic regression) and two ML models (Random Forest [RF] and Light Gradient Boosting Model [LGBM]). The two traditional statistical models were used to investigate test performance of models assuming a linear relationship between features and outcome. In addition, the two machine learning models were investigated to draw on these models' robustness for complex datasets and ability to capture nonlinear relationships, as described in more detail previously (3). Analyses were performed using Python 3.9.16, with LGBM v.3.1.0 for gradient-boosted decision trees and scikit-learn v.0.23.2 for RF. Stratified k-fold cross-validation and grid search were used for hyperparameter tuning. SHAP v.0.36.0 was utilized to compute feature importance.
The abovementioned method was then applied to the preprocessed validation cohort data (LC/MS-MS) to further evaluate test performance. In addition to the binary classification pipeline described above, an adaptation was made to investigate the use of four separate ML models (three neural networks [NNets 1, 2, and 3] and one support vector machine) for multiclass classification to predict infection status into four groups (SARS-CoV-2, influenza A, RSV, and negative) based on the same data set. The same proportions for training and testing were used. Neural networks were prepared on TensorFlow v.2.15.0.
## Statistical analysis
Descriptive data analysis was performed using the chi-squared and Mann-Whitney U tests for continuous variables, implemented in R v.4.0.2. The impact of potential confounders determined a priori, including age, sex, and machine learning model output, was investigated with multivariable analysis, as previously described (3).
## RESULTS
## Discovery cohort description
A total of 325 samples were tested in the discovery cohort (Fig. 1) (Table S1). All samples were successfully tested, and no sample was excluded from analysis. Of these, 254 were positive for SARS-CoV-2, and 71 were negative for SARS-CoV-2 by NAAT. Wholegenome sequences were obtained for 99.6% (253 out of 254), all of which were Omicron subvariants. Overall, the median age was 41 years (29-55 years), and 55.4% of the participants were female. Most samples were nasopharyngeal swabs (66.5%), followed by nasal (31.1%) and mid-turbinate swabs (2.5%) (Table S1).
## Test performance of the discovery cohort by LC/Q-TOF
The top model, LGBM, achieved an AUC of 1.00 (95% confidence interval [CI], not available) for classifying SARS-CoV-2 infection versus negative from respiratory swabs (Fig. S1). Other models also showed high performance, including RF with an AUC of 0.99 (95% CI, 0.97-1.00) and Lasso with an AUC of 0.99 (95% CI, 0.98-1.00). SHAP analysis based on the LGBM and RF models identified 341.3050 at 12.29 (written as m/z at retention time) as the top-ranking feature for classification performance of SARS-CoV-2 infection status (Fig. S2). This feature was associated with higher risk of SARS-CoV-2 infection. The top 20 compound signatures identified through SHAP analysis were subsequently incorporated into the targeted metabolomics approach for the validation cohort.
## Validation cohort description
A total of 1,226 samples were tested in the validation cohort (Table 1). All samples were successfully tested, and no sample was excluded from analysis. Of these, 521 were positive for SARS-CoV-2; 404 samples were positive for other respiratory viruses; and 301 were negative for the viruses above. Whole-genome sequences were obtained for 91.0% (474 out of 521), and all were Omicron subvariants (BA.1.X, n = 70; BA.2.X, n = 94; BA.5.X, n = 34; BQ.1.X, n = 168; XBB0.10, n = 9; XBB.1.X, n = 42; XBB.2.X, n = 4; other, n = 53). The validation cohort included 404 samples positive for other respiratory viruses (influenza A, n = 97; RSV, n = 96; hMPV, n = 50; AdV, n = 51; RV, n = 50; PIV-1, n = 11; PIV-2, n = 9; PIV-3, n = 22; and PIV-4, n = 18). Six cases of viral co-infection were also characterized, including two SARS-CoV-2/influenza A, three SARS-CoV-2/RSV, and one SARS-CoV-2/rhinovirus. The baseline demographic and clinical characteristics of the validation cohort are described in Table 1. The median age was 42 years (14-65 years), and 49.8% of the participants were female. All samples included were nasopharyngeal swabs.
## Test performance of the validation cohort by LC/MS-MS
All models achieved an AUC of 0.98 or greater for classifying SARS-CoV-2 infection versus negative when including the full data set with viral coinfections. The top-performing model was LGBM with an AUC of 0.99 (95% CI, 0.99-1.0), sensitivity of 0.96 (95% CI, 0.91-0.99), and specificity of 0.95 (95% CI, 0.90-0.97) (Table 2; Fig. 2). Feature ranking by SHAP analysis based on LGBM showed that two compounds accounted for most of the test performance such that biomarker identification efforts were focused on these compounds (Fig. 3). Compound 341 > 88.2 (written as precursor ion >fragment ion, which corresponds to compound 341.3050 at 12.29, written as m/z at retention time from the discovery set) was the top-ranking feature and was subsequently identified as 3-oxo-heneicosanoic acid (Fig. 3). More specifically, high concentrations of this compound showed the greatest impact on model performance and was the strongest predictor for SARS-CoV-2 infection in this data set. The second most important feature was 139 > 77, which was identified as 2-(4-hydroxyphenyl) ethanol when compared with commercially available material. High concentrations of this compound predicted the absence of SARS-CoV-2 infection. Median concentrations of the top 20 features in SARS-CoV-2-positive and SARS-CoV-2-negative samples are presented separately (Table S2). Exclusion of the six cases of viral co-infection showed similar overall model performance, with an AUC of 1 (95% CI, 0.99-1.00) for the top-performing model, LGBM, for the main analysis of classification of SARS-CoV-2-positive versus SARS-CoV-2-nega tive samples. A multivariable machine learning model incorporating a priori defined potential confounders (age and sex) demonstrated that only the outcome of the machine learning model was significantly associated with SARS-CoV-2 infection status (Table S3). The same biomarker signature originally derived for SARS-CoV-2 was then applied to the same dataset, including viral coinfection but using multiclass analysis to include separate categories for SARS-CoV-2-positive, influenza A-positive, RSV-posi tive, and negative samples for any of these three viruses. This signature showed high performance for the classification of influenza A with an AUC of 0.97 (95% CI, 0.94-0.99) and RSV with an AUC of 0.99 (95% CI, 0.97-1.00). The relative importance of each feature in the multiclass model is illustrated in Fig. 4. The 3-oxo-henecosanoic acid compounds (main compound and related isomers B, C, and D) showed a trend toward greater relative feature importance for SARS-CoV-2 and RSV, compared to influenza A (Fig. 4).
## Biomarker feature identification
The top-ranking SARS-CoV-2 feature was identified as a compound with a m/z ratio of 341.3050, corresponding to oxo-heneicosanoic acid (C21H40O3), though the position of the oxo-group can vary from C2 to C20. Tandem mass spectrometry demonstrated major fragments at m/z 69.1 and 88.1 in the positive cases, suggesting the presence of multiple oxo-isomers. Based on machine learning analysis, the SRM pair 341.3 > 88.1 was identified as a key marker for SARS-CoV-2 detection, prompting further focus on this compound. The fragment ion at m/z 88.1 suggested that the oxo group is located at the C3 position, forming the fragment O = CCH2CO 2 H + H + , which was later confirmed greater concentration of 3-oxo-heneicosanoic acid compared to the uninfected cell supernatant (Fig. S4). Additionally, the chromatogram of the synthetic 3-oxo-heneicosa noic acid matched those of SARS-CoV-2-positive patient samples and supernatant from SARS-CoV-2-infected cells.
The second-ranked SARS-CoV-2 biomarker had a m/z ratio of 139.073. Tandem mass spectrometry (LC-MS/MS) revealed a major fragment ion at m/z 77.04, which is consistent with the benzenium ion in either 4-ethoxyphenol or 2-(4-hydroxyphenyl) ethanol. The 139.073 > 77.04 peak was present by LC/MS-MS in the uninfected cell supernatant and absent in the infected cell supernatant. Compared with the top candidate biomarker, standard 4-ethoxyphenol produced no signal for the SRM pair 139.1 > 77. However, 2-(4-hydroxyphenyl) ethanol displayed an identical SRM pair of 139.1 > 77 and a nearly identical retention time, confirming the biomarker as 2-(4-hydroxyphenyl) ethanol. This compound likely presents the hydroxyl group in the para position relative to the ethanol group, and clinical samples may contain a mixture of para-, meta-, and ortho-isomers.
## DISCUSSION
In this study, we performed comprehensive metabolic profiling of upper respiratory samples combined with machine learning analysis to identify a top biomarker signa ture as a potential discriminator between SARS-CoV-2-infected and SARS-CoV-2-uninfec ted individuals. Through a discovery cohort of 325 samples from adult and pediatric inpatients and outpatients from a single health center, we identified a 20-feature biomarker signature by LC/Q-TOF with the potential to classify individuals with and without SARS-CoV-2 infection directly from upper respiratory swab samples. We further investigated the test performance of this signature in an independent cohort of 1,226 samples and demonstrated >0.95 AUC, sensitivity, and specificity for the top ML model, LGBM. Model performance was driven by two main compounds of interest, which were identified as 3-oxo-heneicosanoic acid and 2-(4-hydroxyphenyl) ethanol. Importantly, although the signature was derived from a discovery cohort selected for SARS-CoV-2 positive versus negative, it maintained a high level of performance when extended to the differentiation of influenza A virus and RSV, supporting the potential use of this signature for other respiratory viruses.
The biomarker signature, which includes 20 distinct features, highlighted 3-oxoheneicosanoic acid and 2-(4-hydroxyphenyl) ethanol as the top two compounds contributing to the robust classification performance of this method. The first com pound, 3-oxo-heneicosanoic acid, is a 21-carbon, long-chain saturated oxo fatty acid. That such a compound might differentiate SARS-CoV-2 from other respiratory viruses was not entirely surprising, given that SARS-CoV-2-infected individuals have been shown to have altered lipid metabolism (14,15). Nevertheless, the biological role of 3-oxoheneicosanoic acid is not well understood, and odd-chain fatty acids are uncommon in human metabolism. Interestingly, individuals with COVID-19 have elevated levels of odd-chain fatty acids in their sebum lipidome compared to healthy controls (16). The important role of the respiratory microbiota in SARS-CoV-2 infection also raised concerns that the 3-oxo-heneicosanoic acid detected in these respiratory samples may be derived from bacteria in the upper respiratory tract (17)(18)(19). However, the presence of 3-oxo-heneicosanoic acid in the supernatant of human cell culture, at higher levels in the supernatant of SARS-CoV-2-infected cells but present in the supernatant of uninfected cells as well, strongly supports an endogenous source.
The compound 3-oxo-heneicosanoic acid can be formed from disrupted β-oxida tion or cells under oxidative stress (20,21). Mitochondria and peroxisomes are the major organelles for the metabolism and regulation of long-chain fatty acids and very long-chain fatty acids and are the locations within the cell where β-oxidation takes place (22,23). SARS-CoV-2 has been shown to interfere with mitochondrial and peroxisomal functions (24,25). The increase in 3-oxo-heneicosanoic acid we observed may result from SARS-CoV-2-induced dysregulation of β-oxidation pathways and alterations in mitochon drial and peroxisomal functions. However, no direct evidence currently supports this hypothesis, and the origin of this odd-chain fatty acid remains unknown. Further investigation is required to elucidate the mechanisms underlying the accumulation of 3-oxo-heneicosanoic acid in SARS-CoV-2-infected cells and its potential role in COVID-19 pathophysiology.
The second most important compound identified for differentiating SARS-CoV-2 infection from other respiratory virus infections was 2-(4-hydroxyphenyl) ethanol, also known as tyrosol. Endogenous tyrosol is a metabolite of the amino acid tyrosine, and this compound is also found in various dietary sources, such as olive oil (26). Tyrosol possesses relatively weak antioxidant properties, though its stability and capacity for intracellular accumulation render tyrosol a durable antioxidant, especially during inflammation (27,28). Tyrosol can inhibit H 2 O 2 -induced cell death and regulate key signal transduction pathways (29). Additionally, tyrosol and its metabolites have been shown to reduce the level of reactive oxygen species (ROS) and prevent the activation of NF-κB in tumor necrosis factor alpha (TNF-α)-treated human endothelial cells (30). ROS production and TNF-α are the key players in the host antiviral immune response, and excessive activation of these responses can lead to tissue damage (31,32). Hydroxy tyrosol, a derivative of tyrosol, can reduce the expression of the SARS-CoV-2 papain-like protease and apoptosis in infected epithelial cells (33). Additional experiments will be required to investigate how SARS-CoV-2 infection results in decreased levels of 2-(4hydroxyphenyl) ethanol.
The performance of this biomarker signature suggests potential utility as a diagnos tic test for respiratory virus infection, and though mass spectrometry is complex, the method described in this study has several advantages over respiratory virus NAAT. The LC/MS-MS method requires limited preanalytical processing (the addition of isopropanol followed by centrifugation), and analysis on the instrument requires less than 5 min per sample. Mass spectrometry is also inexpensive on a per-sample basis (reagent cost: ~10 cents per sample) compared to NAAT (reagent cost: ~10 USD to >150 USD, depending on panel size). In addition, an LC/MS-MS instrument is similar in price to an automated NAAT system, such as the Hologic Panther. Notably, SHAP feature importance analysis demonstrated that only a few compounds account for most of the test performance. This supports adaptation of this diagnostic approach to a simpler modality that could be performed in the near-care setting, including existing portable mass spectrometers.
The testing approach presented in this study is valuable as it draws from minimally invasive sample collection and benefits from two large sets of samples separated into discovery and validation sets. The data presented directly extend our understanding of the key metabolites that may differentiate SARS-CoV-2-infected from SARS-CoV-2-unin fected individuals and pave the way for an alternative diagnostic approach for viral identification that may complement current testing strategies. Furthermore, the analysis spans a comprehensive range of two statistical and two machine learning models, which demonstrated the reproducibility of findings across models and supported the robustness of the approach. Nonetheless, several limitations should be highlighted. First, these samples were collected retrospectively after a variable period of freezer storage and without controlling for several variables that may confound results that a prospective study would have been better suited to address. This includes timing since symptom onset, disease severity, vaccination status, and impact of therapeutics. Nonetheless, based on the demographic data collected and due to the large cohort studied, this impact may have been lessened through the distribution of these factors between the groups. Second, these data were generated from Stanford Health Care, a single health care system, and further work is required to assess generalizability of this diagnostic approach. Third, the positive sample set included only Omicron subvariants, and was selected based on successful sequencing results for the SARS-CoV-2-positive samples, resulting in a limited distribution of viral burden as estimated by cycle threshold values. Future work should incorporate additional SARS-CoV-2 variants, as well as a wider range of cycle threshold values to evaluate potential dose-response relationships and to confirm signature performance in samples with a low viral burden. More comprehensive analysis of additional respiratory sample types, including lower respiratory tract samples, may also be valuable. Fourth, our data set did not incorporate bacterial coinfection data, which may have confounded results. Last, due to the exploratory nature of this work, the mechanistic explanations for the metabolomic changes detected in SARS-CoV-2 samples remain largely speculative. Additional studies are needed to investigate these metabolic pathways in depth and determine how their changes contribute to the differentiation between SARS-CoV-2-positive and SARS-CoV-2-negative samples.
In summary, we identified a biomarker signature differentiating SARS-CoV-2-posi tive versus SARS-CoV-2-negative samples directly from upper respiratory samples. We subsequently demonstrated the performance of this signature based on an independ ent validation cohort by targeted mass spectrometry combined with machine learning analysis. This work suggests that mass spectrometry-based infectious disease diagnostic testing has clinical potential and that these metabolomic features may reveal novel host-pathogen interactions and therapeutic targets. In addition to SARS-CoV-2, we believe that applying a similar approach to prospective, multisite cohorts of patients with other infectious diseases can greatly extend our understanding of the metabolic pathways involved in the host response to infection.
## References
1. Bhimraj, Morgan, Shumaker et al. (2020) "Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients with COVID-19" *Clin Infect Dis*
2. Uyeki, Bernstein, Bradley et al. (2019) "Clinical Practice Guidelines by the Infectious Diseases Society of America: 2018 update on diagnosis, treatment, chemoprophylaxis, and institutional outbreak management of seasonal influenzaa" *Clin Infect Dis*
3. Hogan, Rajpurkar, Sowrirajan et al. (2021) "Nasopharyngeal metabolomics and machine learning approach for the diagnosis of influenza" *EBioMedicine*
4. (1016)
5. Shen, Yi, Sun et al. (2020) "Proteomic and metabolomic characterization of COVID-19 patient sera" *Cell*
6. Le, Wu, Khan et al. (2022) "Targeted plasma metabolomics combined with machine learning for the diagnosis of severe acute respiratory syndrome virus type 2" *Front Microbiol*
7. Roberts, Muelas, Taylor et al. (2021) "Untargeted metabolomics of COVID-19 patient serum reveals potential prognostic markers of both severity and outcome" *Metabolomics (Los Angel)*
8. Oliveira, Mwangi, Sartim et al. (2022) "Metabolomic profiling of plasma reveals differential disease severity markers in COVID-19 patients" *Front Microbiol*
9. Wong, Devason, Umana et al. (2023) "Serotonin reduction in post-acute sequelae of viral infection" *Cell*
10. Wang, Miller, Verghese et al. (2021) "Multiplex SARS-CoV-2 genotyping reverse transcriptase PCR for population-level variant screening and epidemiologic surveillance" *J Clin Microbiol*
11. Le, Mak, Cowan (2020) "Metabolic profiling by reversed-phase/ionexchange mass spectrometry" *J Chromatogr B Analyt Technol Biomed Life Sci*
12. Saul, Karim, Ghita et al. (2023) "Anticancer pan-ErbB inhibitors reduce inflammation and tissue injury and exert broad-spectrum antiviral effects" *J Clin Invest*
13. Karim, Pohane, Lo et al. (2024) "Chemical inactivation strategies for SARS-CoV-2-infected cells and organoids" *STAR Protoc*
14. Lundberg, Lee (2017) "A unified approach to interpreting model predictions"
15. Theken, Tang, Sengupta et al. (2021) "The roles of lipids in SARS-CoV-2 viral replication and the host immune response" *J Lipid Res*
16. Thomas, Stefanoni, Reisz et al. (2020) "COVID-19 infection alters kynurenine and fatty acid metabolism, correlating with IL-6 levels and renal status" *JCI Insight*
17. Spick, Longman, Frampas et al. (2021) "Changes to the sebum lipidome upon COVID-19 infection observed via rapid sampling from the skin" *EClinicalMedicine*
18. Ortega-Peña, Rodríguez-Martínez, Me et al. (2022) "Staphylococcus epidermidis controls opportunistic pathogens in the nose, could it help to regulate SARS-CoV-2 (COVID-19) infection?" *Life (Basel)*
19. Ji, Won, Gil et al. (2021) "The nasal symbiont Staphylococcus species restricts the transcription of SARS-CoV-2 entry factors in human nasal epithelium"
20. Kumar, Pandit, Sharma et al. (2022) "Nasopharyngeal microbiome of COVID-19 patients revealed a distinct bacterial profile in deceased and recovered individuals" *Microb Pathog*
21. Erlemann, Rokach, Powell (2004) "Oxidative stress stimulates the synthesis of the eosinophil chemoattractant 5-oxo-6,8,11,14-eicosate traenoic acid by inflammatory cells" *J Biol Chem*
22. Houten, Wanders (2010) "A general introduction to the biochemistry of mitochondrial fatty acid β-oxidation" *J Inherit Metab Dis*
23. Ali, Kobayashi, Morito et al. (2023) "Peroxisomes attenuate cytotoxicity of very long-chain fatty acids" *Biochim Biophys Acta Mol Cell Biol Lipids*
24. Violante, Achetib, Van Roermund et al. (2019) "Peroxisomes can oxidize medium-and long-chain fatty acids through a pathway involving ABCD3 and HSD17B4" *FASEB J*
25. Cook, Moreno, Beltran et al. (2019) "Peroxisome plasticity at the virus-host interface" *Trends Microbiol*
26. Koyuncu, Purdy, Rabinowitz et al. (2013) "Saturated very long chain fatty acids are required for the production of infectious human cytomegalovirus progeny" *PLoS Pathog*
27. Marković, Torić, Barbarić et al. (2001) "Hydroxytyrosol, tyrosol and derivatives and their potential effects on human health" *Molecules*
28. Benedetto, Varì, Scazzocchio et al. (2007) "Tyrosol, the major extra virgin olive oil compound, restored intracellular antioxidant defences in spite of its weak antioxidative effectiveness" *Nutr Metab Cardiovasc Dis*
29. Plotnikov, Plotnikova (2021) "Tyrosol as a neuroprotector: strong effects of a "Weak" antioxidant" *Curr Neuropharmacol*
30. Lee, Hur, Lee et al. (2018) "Protective effects of tyrosol against oxidative damage in L6 muscle cells" *FSTR*
31. Muriana, Paz, Lucas et al. (2017) "Tyrosol and its metabolites as antioxidative and anti-inflammatory molecules in human endothelial cells" *Food Funct*
32. Zawawi, Kalyanasundram, Zain et al. (2023) "Prospective roles of tumor necrosis factor-Alpha (TNF-α) in COVID-19: prognosis, therapeutic and management" *Int J Mol Sci*
33. Wieczfinska, Kleniewska, Pawliczak (2022) "Oxidative stress-related mechanisms in SARS-CoV-2 infections" *Oxid Med Cell Longev*
34. Crudele, Smeriglio, Ingegneri et al. (2022) "Hydroxytyrosol recovers SARS-CoV-2-PLpro-dependent impairment of interferon related genes in polarized human airway, intestinal and liver epithelial cells" *Antioxidants (Basel)* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607897&blobtype=pdf | # Inhibition of HIV-1 replication by primer RNA packaging inhibitors
Marc Mirande, Frédéric Subra, Clémence Richetta, Eric Deprez, Olivier Delelis
## Abstract
An essential step in the life cycle of human immunodeficiency virus type 1 (HIV-1) is the packaging of tRNA 3Lys during viral assembly. This step, which requires the assembly of a ternary complex consisting of GagPol, mitochondrial lysyl-tRNA synthe tase, and tRNA 3 Lys , is essential for the production of infectious viruses. Indeed, it allows the initiation of reverse transcription in the particles. In this work, we used an in vitro assay to isolate molecules from a chemical library that disrupt the interaction between the integrase domain of GagPol and mitochondrial lysyl-tRNA synthetase. The three selected molecules also inhibit HIV-1 replication in an ex vivo assay. The number of HIV-1 particles produced in the presence of the drugs is greatly reduced, with a halfmaximal inhibitory concentration obtained at a drug concentration of 5 µM. In addition, we have shown that the few particles produced in the presence of the drugs exhibit significantly reduced infectivity, due to their inability to synthesize minus-strand strong stop cDNA. In parallel, the tRNA 3Lys content of the particles obtained in the presence of drugs decreased significantly. The inability to package tRNA 3Lys into virions results in replication-defective particles. IMPORTANCE Several families of inhibitors have been selected to tackle the human immunodeficiency virus (HIV). Cell entry, reverse transcriptase, integrase, or protease is the most popular targets along the viral life cycle. During the budding step of the virus, the packaging of tRNA 3Lys into particles, which serves as a primer for the initiation of reverse transcription in the virus, has been shown to be essential for the production of infectious virions. In this study, we isolated molecules that inhibit viral replication by inhibiting the incorporation of tRNA 3Lys into virions. The inhibitors of HIV-1 replication described in this work demonstrate that the HIV-1 tRNA packaging step can be exploited for the development of a new family of drugs with novel resistance profiles. KEYWORDS human immunodeficiency virus, integrase, host-pathogen interactions, lysyl-tRNA synthetase, lacidipine, azelastine H uman immunodeficiency virus type 1 (HIV-1) is a lentivirus belonging to the Retroviridae family that requires reverse transcription of its single-stranded RNA genome into a double-stranded cDNA that integrates into the genome of infected cells. To achieve the initiation step of reverse transcription into the nucleocapsid of the virus shortly after budding (1), the primer RNA must be packaged into the newly formed viral particles. This primer RNA is cellular tRNA 3 Lys (2, 3). Human lysyl-tRNA synthetase (LysRS) has been shown to be involved in the recruitment of tRNA 3Lys into the virions through its interaction with Gag or GagPol (4-7). In humans, cytoplasmic and pre-mitochondrial LysRSs are encoded by the KARS1 gene through alternative splicing (8). The mature, active form of mitochondrial lysyl-tRNA synthetase (mLysRS) is produced in mitochondria after cleavage of its N-terminal mitochondrial targeting sequence (9). The catalytic domain of mLysRS interacts with the integrase (IN) domain of the GagPol
polyprotein precursor to form the tRNA 3
Lys packaging complex (10,11). Another group previously proposed that the LysRS incorporated into HIV-1 originates from the cytosolic multisynthetase complex (MSC) (5). According to this model, after dissociation from the complex following phosphorylation at Ser207 by MAPK, cytosolic LysRS (cLysRS) would associate with Gag for packaging into the viral particles (12). The cytosolic and mature mitochondrial LysRSs have specific N-terminal sequences of 21 and 19 amino acid residues, respectively (9). Monospecific antibodies directed against these extra-pep tides identified the human mLysRS species only in extracts of HIV-1 particles (6).
Effective inhibitors of HIV-1 replication have been described. They target different steps of the viral life cycle and include (i) co-receptor antagonists and fusion inhibitors that prevent cell infection (13), (ii) nucleoside analog reverse transcriptase inhibitors and non-nucleoside reverse transcriptase inhibitors that target the reverse transcription step (14), (iii) IN strand transfer inhibitors (INSTIs) that target the chromosomal integration step (15), (iv) allosteric IN inhibitors which block the IN-directed localization of viral RNA into the capsid (16), and (v) protease inhibitors that block the maturation step (17). Due to the high propensity for the emergence of resistance resulting from the high level of mutations accumulated during the reverse transcription step, multi-therapies are used to block viral replication (18). Since primer RNA incorporation into virions also controls viral infectivity (19), we hypothesized that blocking tRNA 3
Lys incorporation into viral particles might be another means of controlling viral replication.
We have previously shown that the mitochondrial species of human lysyl-tRNA synthetase interacts with the GagPol polyprotein precursor of HIV-1 (10) and that the C-terminal IN domain of GagPol is the major contributor to this interaction (4). More specifically, the C-terminal domain (CTD) of IN is the major domain interacting with mLysRS. We first isolated molecules from a chemical library that inhibit the mLysRS:IN association in vitro. These compounds were then used in ex vivo assays to test their ability to inhibit HIV-1 replication. In this work, we have isolated three molecules that render HIV-1 particles defective for replication by decreasing the uptake of tRNA 3
Lys into virions and inhibiting reverse transcription initiation. Our data provide strong proof of concept for the selection of the tRNA 3
Lys packaging complex as a new operational target for HIV-1 replication inhibition.
## RESULTS
## Isolation of molecules inhibiting mLysRS:IN interaction
High-throughput screening of a chemical library was performed using an HTRF assay developed to monitor the interaction between human mLysRS and HIV-1 IN (4). The "Prestwick Chemical Library" contains 1280 FDA-approved drugs. Three molecules were isolated for their ability to inhibit the interaction of mLysRS with IN (Fig. 1), but not with p38, another protein that does interact with LysRS in vitro, and associates cLysRS to the MSC in vivo (11). The half-maximal inhibitory concentration (IC50) in the mLysRS:IN assay ranged from 64 ± 8 µM for 08B10, 310 ± 30 µM for 15G09, to 340 ± 50 µM for 08C10. In this experiment, the background level of the HTRF signal (in %) is around 40%-50% of the value obtained in the absence of inhibitor. The potency of these inhibitors was also determined in the mLysRS:IN-CTD222 interaction and the mLysRS:Pol interaction. IN-CTD222 corresponds to the C-terminal β-barrel domain of IN identified as the domain of IN interacting with mLysRS (4), and the Pol polyprotein corresponds to the C-terminal moiety of the GagPol polyprotein precursor containing IN at the very C-terminal extremity, which associates with mLysRS and tRNA Lys to form the tRNA-pack aging complex of HIV-1. As shown in Fig. S1, the three selected molecules also impede these interactions with similar IC50. This suggested that the three inhibitors might also be able to challenge the interaction between mLysRS and GagPol, the species of IN present at the budding stage of HIV-1 infection.
08B10 and 15G09 are lacidipine and cilnidipine (Fig. 2), respectively, and are built around a dihydropyridine ring. They have been reported as calcium channel blockers and display antihypertensive properties (20). 08C10 is azelastine (Fig. 2), a benzyl phthala zone derivative with antihistaminic activity (21).
A number of dihydropyridine derivatives have been described (20). In a search for compounds more potent than lacidipine or cilnidipine, 18 dihydropyridine derivatives were assayed. Diludine, nimodipine, nifedipine, felodipine, celvidipine, or levamlodipine did not significantly alter the mLysRS:IN interaction in the HTRF assay. Amlodipine, nicardipine, isradipine, nitrendipine, or nivaldipine inhibit the interaction with an IC50 greater than 500 µM. Lercanidipine, azelnidipine, nisoldipine, benidipine, and manidipine display inhibitory profiles similar to lacidipine and cilnidipine (Fig. S2). These compounds preferably carry a bulky, cyclic, or branched group appended to positions R1 or R2.
## Inhibition of HIV-1 replication
According to our model, tRNA 3
Lys packaging requires the assembly of a ternary complex consisting of GagPol, mLysRS, and tRNA 3
Lys . Thus, disruption of the GagPol:mLysRS interaction is expected to impair tRNA 3
Lys packaging into the virus, resulting in a defective HIV-1 replication step due to the absence of the primer RNA required for initiation of reverse transcription.
To test this hypothesis, azelastine, lacidipine, and cilnidipine were used in an ex vivo assay to determine their effect on HIV-1 proliferation. MT4 cells were infected with NLENG1-IRES-GFP, a virus that allows for multiple cycles of infection, and the molecules were added to the medium 36-48 h after infection, before the completion of the first round of viral infection; 72 h after the drugs were added, when two rounds of HIV-1 infection were underway, the level of GFP fluorescence was determined by flow cytometry, as a marker of HIV-1 infection (Fig. 3, left panel). HIV-1 replication was inhibited in a dose-dependent manner. When drugs were added at a final concentration of 10 µM, GFP fluorescence was reduced to 15.2%, 27.2%, or 18.9% of the value observed when DMSO was added alone for azelastine, lacidipine, and cilnidipine, respectively (Fig. 3, left panel). The half-maximal inhibitory concentration was obtained for a drug concentration of approximately 5 µM. Furthermore, although the number of GFP-positive cells is reduced, the intensity of GFP fluorescence is not affected, with a maximum fluorescence intensity of approximately 10 4 units (Fig. S4). This indicates that viral DNA integration into the host genome is not affected, suggesting that replication does not proceed without integration, as observed in the case of HIV resistance to some INSTI drugs (22,23). At the same time, no drug toxicity was observed, as determined by the MTT assay (Fig. 3, right panel). These data clearly demonstrate that the three selected molecules inhibit HIV-1 replication and are not toxic to cells. The selectivity index (SI) (24) of the three drugs, defined as the ratio of 50% cytotoxic concentration (CC50; >20 µM) to 50% inhibitory concentration (IC50; ~5 µM), >4.
We verified that the three molecules did not interfere with the initial stages of HIV-1 replication. Using envelope-defective viruses that are only capable of a single cycle of infection, we observed that none of the three molecules inhibited the first round of replication, as shown by the level of GFP expression in MT4 cells infected with NLENG1-ES-IRES-GFP viruses (Fig. 4). These drugs also did not inhibit the in vitro aminoacylation activity of LysRS. When added up to 1 mM, azelastine and cilnidipine did not inhibit Lysyl-tRNA Lys synthesis, and lacidipine resulted in a 30% inhibition (Fig. S5).
Of the five other dihydropyridine derivatives that showed IC50s greater than 500 µM in vitro (Fig. S2), nisoldipine did not inhibit HIV-1 replication in the ex vivo assay, lercanidipine and azelnidipine had higher cellular toxicity, and benidipine and manidi pine were similar to lacidipine and cilnidipine (Fig. S3a andb). They were not used for the remainder of the study.
We used the CPRG method developed in HeLa P4 cells grown in 96-well plates to measure the infectivity of virus particles, as low levels of virus were recovered after exposure of infected MT4 cells to different drug concentrations. In this assay, expression of the LacZ gene is under the control of the HIV-1 LTR, and its level of trans-activation correlates with the level of expression of HIV-1 Tat. HeLa P4 cells were incubated with normalized amounts of virus as measured by the P24 assay, and the expression of β-galactosidase was determined by the CPRG assay (Fig. 5). As expected for viruses depleted in the tRNA 3
Lys primer and consistent with the data of Fig. 3, viruses recovered after MT4 cells were exposed to 10 µM drugs exhibited low infectivity: less than 4%, 22%, and 34% for azelastine, lacidipine, and cilnidipine, respectively, compared with wild-type viruses.
## Effect of inhibitors on the synthesis of minus-strand strong-stop viral DNA
Inhibition of HIV-1 replication following exposure of infected cells to 10 µM of the drugs was determined after infection of MT4 cells with the NL4-3 virus, and treatment of infected cells with trypsin to eliminate viruses remaining in solution, which could contaminate de novo produced particles recovered after different times of infection. Drugs were added 4 h after infection (Fig. 6); 48, 72, or 96 h after infection, the extent of HIV-1 replication was monitored by quantifying the expression of Gag after incubation with a fluorescently labeled IgG (KC57-RD1) and FACS analysis (Fig. 6a). The level of Gag expression after 48 h of infection was not affected by the presence of drugs. By contrast, the effect of drugs on multi-cycle virus replication is clearly visible after 72 or 96 h of infection, when expression of Gag is reduced by 80% compared with controls (Fig. 6a). This is consistent with a mechanism of inhibition affecting the late steps of HIV replica tion that would limit tRNA 3
Lys packaging, resulting in particles that are unable to initiate reverse transcription and are thus deficient for the second round of infection. The amount of virus released into the culture medium after 48, 72, or 96 h of HIV-1 infection was quantified after concentration by ultracentrifugation. After 72 or 96 h of infection in the presence of drugs, only 10%-20% of viruses were recovered, compared with controls (Fig. 6b), consistent with the lowest Gag expression observed above. The infectivity of these viruses was determined by the CPRG assay after incubation of HeLaP4 cells with normalized amounts of virus as measured by the P24 assay. After 48 h of infection in the presence of drugs, the few viruses recovered after ultracentrifugation showed an infectivity reduced by more than 80% (Fig. 6c). The infectivity of viruses recovered after 72 or 96 h of infection remained low. Thus, when MT4 cells infected with NL4-3 viruses are incubated in the presence of any of these three inhibitors, the amount of virus produced is greatly reduced (Fig. 6b), and their infectivity is impaired (Fig. 6c). Taken together, these results show that the molecules reduce the infectivity of newly formed viruses. This results in reduced virus production after two or more cycles of replication.
We used digital RT-PCR and digital PCR to determine the viral mRNA and minusstrand strong stop viral cDNA content of the particles (Fig. 7; Fig. S6). The amount of early viral cDNA produced in the HIV-1 particles is a measure of the efficiency of the initiation step of reverse transcription, which requires the packaging of tRNA 3
Lys at the budding step, the primer for cDNA synthesis.
After 48 h of infection, corresponding to a single replication cycle, the number of viral particles produced was very low (see P24, Fig. 6b), precluding accurate quantification of viral DNA and RNA. In contrast, viral DNA and RNA could be nicely quantified by digital PCR after 72 or 96 h of infection (Fig. S6). The amount of minus-strand strong stop cDNA present in the viruses, normalized to the amount of viral RNA present, was significantly reduced 72 and 96 h after infection in the presence of 10 µM of either azelastine, lacidipine, or cilnidipine (Fig. 7). Thus, the ability of viral particles to promote the reverse transcription initiation step is altered in the presence of these drugs, as expected for tRNA 3 Lys -deficient viral particles. The low levels of minus-strand strong stop cDNA detected in the viruses produced in the presence of the drugs explain the low infectivity of these particles.
## tRNA 3
Lys deficiency in viruses obtained after exposure to inhibitors MT4 cells were infected with the NL4-3 virus. After 72 h, the cells were exposed to 10 µM of azelastine, lacidipine, or cilnidipine. The viruses were then recovered from the cell culture supernatant and purified by centrifugation through a sucrose cushion. Recovery of the viruses was quantified using the P24 assay. The tRNA was extracted with phenol-ether and precipitated with ethanol. Samples corresponding to 1, 0.5, 0.25, and 0.125 ng of P24 were spotted onto nitrocellulose membranes and hybridized with a radiolabeled oligonucleotide probe complementary to tRNA 3 Lys (Fig. 8). The amount of tRNA 3
Lys was significantly reduced in the samples obtained after incubation with the drugs, compared with the DMSO control. Quantification using pure tRNA 3
Lys as a control showed the presence of 10-15 tRNA molecules per viral particle. In samples obtained in the presence of drugs, the tRNA content was reduced 4-fold to 5-fold.
## DISCUSSION
We have previously shown that the human mitochondrial species of lysyl-tRNA synthetase (mLysRS) is hijacked by the virus during HIV-1 particle assembly (6). It interacts with the Pol domain of the GagPol polyprotein precursor (10) and more specifically with the C-terminal β-barrel domain of IN (4) to form the tRNA 3 Lys packaging complex (11). In this work, we have isolated inhibitors of the interaction between human mLysRS and HIV-1 IN. These molecules, which were selected in vitro after screening of a chemical library for their ability to inhibit the interaction between the two purified proteins, were also shown to efficiently block HIV-1 replication in an ex vivo assay. The half-maximal inhibitory concentration (IC50) of these molecules was estimated to be around 50-300 µM in the in vitro assay (Fig. 1). However, due to the poor solubility of IN, which can form multiple oligomers (25), we suspected that these values were underestimated. Indeed, when these molecules were used in an ex vivo assay, they showed an operational IC50 of about 5 µM and did not elicit cell toxicity (Fig. 3). Two of these molecules, lacidipine and cilnidipine, have a dihydropyridine backbone, but none of the other 16 derivatives tested showed a better inhibitory profile in the in vitro and ex vivo assays (Fig. S2 andS3).
The molecules isolated in this work do not inhibit LysRS tRNA-aminoacylation activity, nor do they inhibit the first stages of HIV-1 infection, from fusion to integration. However, new HIV-1 particles formed and released in the presence of the drugs are replication-defective. We used quantitative PCR to analyze the RNA and DNA content of the viruses exposed to the drugs. In particular, the viral content of minus-strand strong stop cDNA is greatly reduced, whereas the viral RNA content is unaffected, indicating a defective reverse transcription initiation mechanism within the particles. The number of particles produced in the presence of drugs is significantly reduced (Fig. 6b), making it challenging to obtain a sufficient amount of virus for the precise quantification of tRNA 3
Lys in these particles. Nevertheless, the tRNA 3 Lys content appears to be substantially diminished (Fig. 8). These data are consistent with a reduced level of tRNA 3
Lys packaging. These results show that the minus-strand strong stop cDNA is present in newly formed particles, but not after treatment with drugs. This strongly supports the idea that reverse transcription begins in newly made virions before they infect new cells.
The three molecules isolated and characterized here clearly demonstrate that the HIV-1 tRNA packaging step can be targeted by a new family of drugs with novel resistance profiles. Our data provide strong proof of concept for the use of the tRNA 3 Lys packaging complex as a novel target for HIV-1 replication inhibition.
## MATERIALS AND METHODS
## HTRF assay
Homogeneous time-resolved fluorescence (HTRF) assays were performed in black, flat-bottom, half-area, 96-well microplates (Corning #3694). Proteins were isolated as described previously (4). Several batches of IN were used throughout this study. It was determined that IN, IN-CTD222, and Pol proteins were greater than 90% homogene ous. Human mLysRS with a C-terminal HA-tag (mLysRS-HA) was incubated at a dimer concentration of 1.5 nM with HIV-1 IN carrying a C-terminal His-tag at a dimer concen tration of 25 nM, in 10 mM Tris-HCl pH 7.5, 50 mM NaCl, 10 mM 2-mercaptoethanol, and BSA at 1 mg/mL, in the presence of 0.2 mM of the molecules of the library (Pre stwick Chemical Library, 1280 compounds) (https://www.prestwickchemical.com/screen ing-libraries/prestwick-chemical-library/). The molecules are dissolved in DMSO at a final concentration of 10 mM. After a 1 h incubation on ice, antibodies (Cisbio) directed to the His-tag and conjugated with Eu 3+ cryptate (Cisbio #61HISKLB) and the HA-tag conjugated with XL665 (Cisbio #610HAXLB) were added, and incubation was continued for 30 min. After the addition of 50 mM KF, fluorescence of Eu 3+ cryptate and of XL665 was recorded at 620 nm (I 620 ) and 665 nm (I 665 ), respectively, following excitation of Eu 3+ cryptate at 317 nm, in an Infinite M1000 PRO microplate reader (TECAN). The results were expressed as the ratio of I 665 /I 620 .
## Cells and viruses
MT4 cell line (26) was maintained in RPMI 1640. HEK293T and HeLa-P4 cells (27) were maintained in Dulbecco's modified Eagle medium (DMEM). All media were supplemen ted with Glutamax and with 10% heat-inactivated fetal calf serum (Hyclone) and 1% penicillin/streptomycin (100 units/mL) (Gibco). All media were purchased from Gibco (Life Technologies Co.). All cell lines used here were incubated at 37 °C, under 5% CO 2 atmosphere. HIV-1 stocks were prepared by calcium phosphate-mediated transfection of HEK293T cells, as previously described (22), with shuttle vector plasmids encoding HIV-1 NL4-3 (GenBank: AF324493.1) or NLENG1-IRES-GFP (28). The latter vector comes from the HIV NL4-3 strain and contains a gfp-IRES-nef cassette at the nef locus. For clarity reasons, we designate this vector here as "NL4-3-GFP. " For single-round infection assays, the envelope-defective NLENG1-ES-IRES-GFP virus, pseudotyped with VSV-G protein, was used.
The HIV-1 p24 gag antigen contents in viral inoculates were determined by enzymelinked immunosorbent assay (Perkin-Elmer Life Sciences).
## HIV infectivity and toxicity assays
Replication of the NL4-3 virus was determined either by the ELISA technique or in HeLa-P4 cells by the CPRG method as described previously (29). These are HeLa CD4 LTR-lacZ cells in which the expression of lacZ is induced by the HIV transactivator Tat, allowing precise quantification of the infectivity of HIV-1. The viral titer was determined by quantifying β-galactosidase activity in HeLa-P4 lysates in a colorimetric assay based on the cleavage of chlorophenol red-β-D-galactopyranoside (CPRG) by β-galactosidase. The cytotoxicity is evaluated by the MTT assay.
For NL4-3-GFP vectors, viral infection is followed by GFP expression (i.e., the percentage and geometric mean fluorescence intensity [MFI] of GFP + cells). Infectivity was estimated by flow cytometry using a FACS Celesta flow cytometer (BD Biosciences). Toxicity is also assessed by flow cytometry (side and forward scatter).
For NL4-3 vectors, viral replication is followed by intracellular Gag staining with anti-gag antibody KC57-FITC or KC57-RD1 (Beckman Coulter), as described (22).
## Quantification of viral RNA and DNA
MT4 cells (2 × 10 6 cells) were infected with the NL4-3 viruses. After 4-5 h of infection at 37°C, cells were twice washed with PBS, incubated 5 min at 37°C in 0.25% trypsin-EDTA solution (Gibco) to eliminate viruses remaining in the medium, incubated 5 min at 37°C in complete RPMI medium, twice washed with PBS, resuspended in RPMI medium at a cell density of 200 × 10 3 cells/mL and incubated with 10 µM of azelastine (25 mM stock solution in H 2 0), or lacidipine and cilnidipine (25 mM stock solution in DMSO). A second aliquot of the inhibitors was added to the culture medium 48 h after infection. Cultures were collected 48, 72, or 96 h after infection; the cells and cellular debris were removed by centrifugation; and the supernatant was subjected to high-speed centrifugation (SW41 rotor, 20,000 rpm, 2 h at 10°C) to pellet the viruses. Viruses were resuspended in 200 µL of PBS, and 50 µL samples were mixed with a pellet contain ing 500 × 10 3 uninfected MT4 cells to serve as RNA and DNA carriers. Viral RNA and DNA were purified with the RNeasy minikit (Qiagen) and QIAamp DNA blood minikit (Qiagen), respectively, according to the manufacturer's instructions. All RNA and DNA quantifications were performed by digital PCR and RT-qPCR on a QIAcuity instrument (QIAGEN) using 26K 24-well nanoplates and the QIAcuity Probe PCR and One-Step Viral RT-PCR kits. Primers and probes were 5′-ATCTGAGCCTGGGAGCTCTCT, 5′-CTGCTAGAGAT TTTCCACACTGAC, and 5′-FAMAAGTAGTGTGTGCCC for the quantification of minus-strand strong-stop cDNA, and 5′-CTGAAGCGCGCACGGCAA, 5′-GACGCTCTCGCACCCATCTC, and 5′-FAMTAGCCTCCGCTAGTCAAAATTTTTGG CGT for viral RNA quantification.
## Quantification of tRNA 3
Lys MT4 cells were infected with NL4-3 viruses (200 ng of p24 antigen per 10 6 cells). Twenty-four hours after infection at 37°C, the cells were washed three times with PBS and resuspended in RPMI medium at a cell density of 500 × 10 3 cells/ml, and plated in six-well plates (4 mL per well) in the presence of inhibitors at a final concentration of 10 µM. The concentration of DMSO in the cultures is 0.1%. Seventy-two hours post-infection, the cells were centrifuged for 5 min at 400 × g and 12°C. The supernatant was then subjected to a second centrifugation step for 20 min at 1,300 × g and 12°C to remove cell debris. The clear supernatant was subjected to ultracentrifugation on a 20% sucrose cushion (2.5 h at 20,000 rpm in a SW55Ti rotor, at 12°C). The virus pellets were resuspended in 100 µL of PBS and stored at -80°C. The amount of virus was determined by the P24 assay.
After two phenol-ether extractions, the tRNA was ethanol-precipitated, and the resulting pellet was dissolved in 20 µL of H 2 O. The samples were serially diluted in H 2 O, and 5 µL of each dilution was spotted onto Nytran membranes (Nytran 0.45, Schleicher & Schuell). The tRNAs were fixed with a three-minute UV exposure. Pure in vitro-transcribed tRNA 3
Lys (23) was used as an internal standard. The membrane was hybridized with an oligonucleotide probe (5'-cctggaccctcagattaaaagtctga-3') radiolabeled using T4-polynu cleotide kinase (Roche) and [γ-32 P]-ATP. Radioactivity was quantified using a Typhoon (GE Healthcare).
## tRNA aminoacylation assay
Initial rates of tRNA aminoacylation were measured at 25°C in 0.1 mL of 20 mM Imidazole-HCl buffer (pH 7.5), 100 mM KCl, 0.5 mM DTT, 12 mM MgCl 2 , 2 mM ATP, 180 µM 14 C-labeled lysine (NEN; 16.66 Ci/mol), and saturating amounts of global yeast tRNA (30). The incubation mixture contained catalytic amounts (2 nM) of mLysRS appropriately diluted in 10 mM Tris-HCl (pH 7.5), 10 mM 2-mercaptoethanol, containing bovine serum albumin at 4 mg/mL. Inhibitors were added at final concentrations ranging from 3.9 to 1,000 µM. One unit of activity is the amount of enzyme producing 1 nmol of lysine-tRNA Lys /min, at 25°C.
## References
1. Abbink, Berkhout (2007) "HIV-1 reverse transcription: close encounters between the viral genome and a cellular tRNA" *Adv Pharmacol*
2. Seif, Niu, Kleiman (2015) "In virio SHAPE analysis of tRNA(Lys3) annealing to HIV-1 genomic RNA in wild type and protease-deficient virus" *Retrovirology (Auckland)*
3. Wain-Hobson, Sonigo, Danos et al. (1985) "Nucleotide sequence of the AIDS virus" *LAV. Cell*
4. Phongsavanh, Al-Qatabi, Shaban et al. (2020) "How HIV-1 integrase associates with human mitochondrial Lysyl-tRNA synthetase" *Viruses*
5. Cen, Khorchid, Javanbakht et al. (2001) "Incorporation of lysyl-tRNA synthetase into human immunodeficiency virus type 1" *J Virol*
6. Kaminska, Shalak, Francin et al. (2007) "Viral hijacking of mitochondrial lysyl-tRNA synthetase" *J Virol*
7. Kovaleski, Kennedy, Hong et al. (2006) "In vitro characterization of the interaction between HIV-1 Gag and human lysyl-tRNA synthetase" *J Biol Chem*
8. Tolkunova, Park, Xia et al. (2000) "The human lysyl-tRNA synthetase gene encodes both the cytoplasmic and mitochondrial enzymes by means of an unusual alternative splicing of the primary transcript" *J Biol Chem*
9. Dias, Octobre, Kobbi et al. (2012) "Activation of human mitochondrial lysyl-tRNA synthetase upon maturation of its premitochondrial precursor" *Biochemistry*
10. Kobbi, Octobre, Dias et al. (2011) "Association of mitochondrial Lysyl-tRNA synthetase with HIV-1 GagPol involves catalytic domain of the synthetase and transframe and integrase domains of Pol" *J Mol Biol*
11. Khoder-Agha, Dias, Comisso et al. (2018) "Characterization of association of human mitochondrial lysyl-tRNA synthetase with HIV-1 Pol and tRNA3Lys" *BMC Biochem*
12. Duchon, St Gelais, Titkemeier et al. (2017) "HIV-1 exploits a dynamic multi-aminoacyl-tRNA synthetase complex to enhance viral replication" *J Virol*
13. Esté, Telenti (2007) "HIV entry inhibitors"
14. Gu, Zhu, Wang et al. (2020) "Recent discoveries in HIV-1 reverse transcriptase inhibitors" *Curr Opin Pharmacol*
15. Pommier, Johnson, Marchand (2005) "Integrase inhibitors to treat HIV/AIDS" *Nat Rev Drug Discov*
16. Singer, Dinh, Levintov et al. (2023) "The drug-induced interface that drives HIV-1 integrase hypermultimerization and loss of function" *mBio*
17. Ghosh, Osswald, Prato (2016) "Recent progress in the develop ment of HIV-1 protease inhibitors for the treatment of HIV/AIDS" *J Med Chem*
18. Saag, Benson, Gandhi et al. (2018) "Antiretroviral drugs for treatment and prevention of HIV infection in adults: 2018 recommendations of the International Antiviral Society-USA panel" *JAMA*
19. Gabor, Cen, Javanbakht et al. (2002) "Effect of altering the tRNA(Lys)(3) concentration in human immunodeficiency virus type 1 upon its annealing to viral RNA, GagPol incorporation, and viral infectivity" *J Virol*
20. Velena, Zarkovic, Gall Troselj et al. (2016) "2016. 1,4-dihydropyridine derivatives: dihydronicotinamide analogues-model compounds targeting oxidative stress" *Oxid Med Cell Longev*
21. Bernstein (2007) "Azelastine hydrochloride: a review of pharmacology, pharmacokinetics, clinical efficacy and tolerability" *Curr Med Res Opin*
22. Richetta, Subra, Malet et al. (2022) "Mutations in the 3'-PPT lead to HIV-1 replication without integration" *J Virol*
23. Richetta, Tu, Delelis (2022) "Different pathways conferring integrase strand-transfer inhibitors resistance" *Viruses*
24. Ho, Liu, Yeh et al. (2018) "Micafungin is a novel anti-viral agent of chikungunya virus through multiple mechanisms" *Antiviral Res*
25. Deprez, Tauc, Leh et al. (2000) "Oligomeric states of the HIV-1 integrase as measured by time-resolved fluorescence anisotropy" *Biochemistry*
26. Foley Ge, Farber, Bg et al. (1965) "Continuous culture of human lymphoblasts from peripheral blood of a child with acute leukemia" *Cancer*
27. Charneau, Alizon, Clavel (1992) "A second origin of DNA plus-strand synthesis is required for optimal human immunodeficiency virus replication" *J Virol*
28. Levy, Aldrovandi, Kutsch et al. (2004) "Dynamics of HIV-1 recombination in its natural target cells" *Proc Natl Acad Sci*
29. Delelis, Thierry, Subra et al. (2010) "Impact of Y143 HIV-1 integrase mutations on resistance to raltegravir in vitro and in vivo" *Antimicrob Agents Chemother*
30. Francin, Kaminska, Kerjan et al. (2002) "The N-terminal domain of mammalian Lysyl-tRNA synthetase is a functional tRNAbinding domain" *J Biol Chem* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724324&blobtype=pdf | # Avian coronavirus IBV-induced activation of the NLRP3-Caspase-1-IL-1β axis in renal collecting ducts contributes to nephropathogenesis
Min Huang, Chengyin Liukang, Rong Liang, Yingfei Li, Ruihua Yang, Mingwei Zhang, Ye Zhao, Jing Zhao, Guozhong Zhang
## Abstract
Avian coronavirus infectious bronchitis virus (IBV) induces severe renal inflammation and urate deposition in chickens, but the underlying mechanisms remain incompletely understood. Here, we used both in vivo chicken infection models and in vitro primary renal epithelial cell cultures and found that IBV activates the NLRP3 (NOD-like receptor family, pyrin domain containing 3) inflammasome, leading to the maturation and secretion of IL-1β (interleukin-1 beta). Pharmacological inhibition of NLRP3 in an in vivo chicken model alleviates renal pathology without affecting viral replication, implicating host inflammatory responses in disease progression. These findings indicate that host inflammasome activation, rather than direct viral replica tion or cytotoxicity, plays a central role in IBV-associated renal pathology. Notably, we observed that NLRP3 is predominantly expressed in the renal collecting duct, primarily within AQP2 (Aquaporin-2)-positive epithelial cells-an anatomic localization of inflammasome activity not previously reported. Single-cell transcriptomic profiling further revealed that IBV infection reprograms collecting duct cells from an electrolyteregulating phenotype into a pro-inflammatory state, accompanied by disrupted urate metabolism and impaired ion transport. These findings identify collecting duct-targeted NLRP3 activation as a central mechanism of IBV-induced nephropathy and provide a theoretical foundation for developing targeted anti-inflammatory therapies.
IMPORTANCEIn vivo studies in chickens demonstrate that the activation of the NLRP3 inflammasome is a key driver of renal injury during infectious bronchitis virus (IBV) infection. Pharmacological inhibition of NLRP3 significantly alleviates renal inflammation and tissue damage without affecting viral replication, highlighting the central role of host inflammatory responses in disease progression. Importantly, we report for the first time that NLRP3 activation is predominantly localized to AQP2-positive collecting ducts, a nephron segment essential for uric acid excretion and electrolyte balance. IBV infection reprograms these epithelial cells into a pro-inflammatory, metabolically dysregulated state, promoting urate crystal formation and amplifying tissue injury. These findings reveal a spatially confined epithelial-immune axis of coronavirus-induced renal pathology and suggest new avenues for targeted intervention. KEYWORDS infectious bronchitis virus, NLRP3 inflammasome, renal inflammation, collecting duct epithelial cells, inflammation-mediated pathology I nfectious bronchitis virus (IBV) is an enveloped, positive-sense single-stranded RNA virus and one of the earliest identified coronaviruses in the genus Gammacoronavirus (1, 2). In addition to its canonical respiratory manifestations, the currently predominant GI-19 IBV strains can cause acute nephritis with tubular lesions and urate deposition,
particularly severe in young chickens (3). Although the clinical manifestations and histopathological characteristics of IBV infection have been well characterized (4), the underlying molecular mechanisms driving renal inflammation and tissue damage remain incompletely understood.
The innate immune system represents the host's primary defense against viral infection, in which members of the NOD-like receptor (NLR) family, a group of nucleo tide-binding domain and leucine-rich repeat-containing receptors, play a pivotal role in sensing intracellular danger signals, including invading viruses (5). Among them, the NLRP3 (NOD-like receptor family, pyrin domain containing 3) inflammasome is particu larly critical. The NLRP3 inflammasome is a cytosolic multiprotein complex classically activated through the canonical "two-signal model" (6). The initial "priming" signal upregulates the transcription of NLRP3 and key pro-inflammatory mediators, includ ing pro-IL-1β, via pattern recognition receptor pathways, particularly Toll-like receptors (TLRs) (7). The subsequent "activation" signal, elicited by cellular stressors such as ionic imbalance, mitochondrial dysfunction, or lysosomal rupture, facilitates NLRP3 oligomeri zation and the formation of the active inflammasome complex (8). Upon maturation, the inflammasome activates Caspase-1, which mediates the cleavage and secretion of the pro-inflammatory cytokines IL-1β and IL-18 (interleukin-18), ultimately driving robust inflammatory responses (7). Although inflammasome activation is central to antiviral immunity, dysregulated or sustained activation can trigger a cytokine storm, leading to severe tissue damage and immune-mediated pathology (9,10). Notably, excessive inflammasome activation has been implicated in the pathogenesis of severe lung injury and increased mortality associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection (11). Moreover, viral infections such as those caused by severe acute respiratory syndrome coronavirus (SARS-CoV), Middle East respiratory syndrome coronavirus (MERS-CoV), and influenza A virus (IAV) are associated with cytokine overproduction, including TNF-α, IL-1β, and type I/II interferons, which disrupt immune homeostasis and contribute to host tissue pathology (12)(13)(14). Clinical and experimental studies have shown that IBV infection is associated with marked upregulation of IL-1β in renal tissue, suggesting a potential role for the NLRP3 inflammasome signaling axis in IBV-induced renal injury (15,16). However, the mechanisms by which IBV induces kidney inflammation and the contribution of the NLRP3 inflammasome to this process have yet to be characterized.
In this study, we investigated the role of the NLRP3 inflammasome in IBV-induced kidney injury using an in vivo chicken infection model and in vitro primary renal epithelial cell culture. Our results demonstrate that IBV infection induces NLRP3 activation in renal tissues, particularly within collecting duct epithelial cells. This localized inflammasome activity is accompanied by pro-inflammatory transcriptional reprogramming, altered urate metabolism, and impaired ion transport, contributing to urate crystal formation and exacerbation of tissue damage. Pharmacological inhibition of NLRP3 significantly attenuates renal inflammation and histological injury without affecting viral replica tion. These findings provide new insight into the cellular and molecular mechanisms underlying IBV-induced nephritis and highlight the collecting duct as a central site of immune-metabolic dysfunction during coronavirus infection.
## RESULTS
## IBV infection promotes cytokine-mediated renal inflammation in chickens
To investigate the molecular basis of IBV-induced renal pathology, we established an animal model by infecting 1-day-old specific-pathogen-free (SPF) chickens with IBV. Gross examination of kidneys from IBV-infected chickens revealed marked renal enlargement and urate deposition (Fig. 1A). Histopathological assessment further revealed severe renal lesions, characterized by widespread vacuolar degeneration, necrosis, and desquamation of renal tubular epithelial cells, as well as a prominent infiltration of inflammatory cells within the renal interstitium (Fig. 1B). These findings collectively indicated that IBV infection results in profound structural and cellular damage to renal tissues.
In order to characterize the inflammatory responses associated with IBV-induced renal injury, we quantified the expression of pro-inflammatory cytokines in serum. Enzyme-linked immunosorbent assay (ELISA) demonstrated a significant and sustained increase in circulating IL-1β and IL-18 levels during 5-14 days post-infection (dpi), consistent with a systemic inflammatory response (Fig. 1C). To gain a comprehensive view of host transcriptional changes, we performed high-throughput RNA sequenc ing (RNA-seq) on kidney tissues and primary chicken embryonic kidney (CEK) cells following IBV infection. Differential gene expression analysis revealed robust upregula tion of a broad range of inflammation-associated genes (Fig. 1D), including pro-inflammatory interleukins (e.g., IL1B, IL6, and IL12A), chemokines (CCL5 and CXCL13), and cognate receptors (IL1R1 and CXCR5), suggesting the activation of multiple inflammatory signaling pathways. These transcriptomic changes were further corroborated by quantitative reverse transcription PCR (qRT-PCR), which confirmed significant induction of IL1B, IL18, and IL8 expression in renal tissues (Fig. 1E). These findings indicate that IBV infection induces renal inflammatory responses both in vivo and in vitro, which are associated with the upregulation of pro-inflammatory cytokines and chemokines.
## IBV infection induces NLRP3 inflammasome activation in chicken kidneys
To elucidate the molecular mechanisms underlying inflammasome activation in response to IBV infection, we conducted transcriptomic profiling using RNA-seq on kidney tissues obtained from IBV-infected chickens. Differential gene expression analysis was subse quently performed to identify host genes that were significantly modulated during infection. Pathway enrichment analysis of these differentially expressed genes (DEGs) indicated a significant activation of signaling cascades associated with the NLRP3 inflammasome. Specifically, enrichment was observed in pathways involved in inflammasome complex assembly, upstream signal transduction events, and canonical inflammasome activation processes. Taken together, these findings indicate the involvement of NLRP3-mediated signaling during IBV pathogenesis (Fig. 2A).
In parallel, gene set enrichment analysis (GSEA) further corroborated these findings, revealing significant enrichment of gene sets related to inflammasome activation, NOD-like receptor signaling, Toll-like receptor signaling, cytokine-mediated communica tion, and general inflammatory responses (Fig. 2B). These transcriptomic results provide compelling evidence that IBV infection in chickens leads to the activation of the NLRP3 inflammasome in renal tissues. Moreover, the data suggest that IBV triggers a widespread activation of innate immune signaling pathways, highlighting a potential mechanistic link between viral infection and host inflammatory responses.
We next performed qRT-PCR analysis to assess the expression levels of key NLRs in infected kidney tissues, aiming to clarify the molecular basis of IBV-induced inflammasome activation. Among the NLR family members analyzed, NLRP3 mRNA expression was the most significantly upregulated following IBV infection, indicating that NLRP3 is likely the primary inflammasome sensor involved in the renal immune response to IBV (Fig. 2C; Fig. S1). To validate these findings at the protein level and determine the spatial distribution of NLRP3 expression, immunohistochemistry (IHC) was performed on kidney sections from infected chickens. Enhanced NLRP3 immunoreactivity was observed in distinct renal tubular structures, indicative of spatially restricted activation of the inflammasome within renal tubular compartments (Fig. 2D).
Western blot (WB) further confirmed inflammasome activation in infected kidneys relative to uninfected controls, showing significant increases in the protein levels of NLRP3, cleaved Caspase-1, and mature IL-1β at 5 and 7 dpi (Fig. 2E). In agreement with protein-level findings, immunohistochemical analysis at 5 dpi showed markedly increased IL-1β staining in kidney tissues, consistent with enhanced IL-1β expression in infected kidney tissue (Fig. 2F). IHC and immunofluorescence analyses were performed on kidney samples collected at 5 dpi, a time point selected based on peak histopatholog ical changes observed in previous experiments.
These results indicate that NLRP3 is involved in the renal immune response to IBV infection, likely through the activation of canonical inflammasome pathways and subsequent IL-1β maturation and secretion. These findings suggest that NLRP3 may act as an upstream regulator of renal inflammation during IBV infection.
## IBV infection activates the NLRP3-Caspase-1-IL-1β pathway in CEK cells
Transcriptomic analysis of IBV-infected primary CEK cells demonstrated a marked enrichment of signaling pathways associated with inflammasome activation, innate immune recognition, and proinflammatory responses (Fig. 3A through C). These transcriptional alterations are consistent with the immune signatures observed in renal tissues from IBV-infected chickens, thereby providing consistent evidence of inflammasome involvement in host antiviral defense. Notably, inflammasome activation is characterized by NLRP3 oligomerization and the formation of punctate cytoplasmic aggregates, which represent hallmark features of inflammasome assembly (17,18). To establish an experimental reference for NLRP3 activation, positive control conditions were generated in CEK and HD11 cells using a classical two-signal priming protocol, in which cells were first exposed to Poly(I:C), followed by stimulation with the potassium ionophore BMS. This sequential treatment reliably induced the formation of punctate NLRP3 aggregates (Fig. 3D), thereby validating the assay system. The subcellular distribution of NLRP3 during IBV infection was further examined by immunofluorescence microscopy. Co-staining of NLRP3 with the IBV nucleocapsid (N) protein, a marker of productive infection, revealed a striking redistribution of NLRP3 from a diffuse cytoplasmic pattern in uninfected cells to concentrated perinuclear puncta in IBV-N-positive cells (Fig. 4A). This relocalization is indicative of inflammasome complex assembly within the infected host cell cytoplasm.
These microscopic findings were validated by immunoblotting and enzymatic activity measurements. At 24 hours post-infection (hpi), western blot analysis revealed marked upregulation of NLRP3 together with increased expression of pro-IL-1β in CEK cells, which was accompanied by enhanced secretion of mature IL-1β into the culture supernatant (Fig. 4B). Consistently, Caspase-1 enzymatic activity assays demonstrated significant activation of Caspase-1 following IBV infection, accompanied by elevated IL-1β release (Fig. 4C andD). To further confirm the roles of NLRP3 and Caspase-1 in IBV-induced IL-1β secretion, CEK cells were treated with the NLRP3 inhibitor CY-09 and the Caspase-1 inhibitor Ac-YVAD-CMK. Both inhibitors significantly reduced IL-1β secretion without affecting CEK cell viability (Fig. 4E through H). These findings together indicate that IBV infection activates the NLRP3 inflammasome in CEK cells through a canonical NLRP3-Caspase-1 axis, leading to IL-1β maturation and release, and under score the central role of inflammasome signaling in the innate immune response to IBV infection. To assess strain dependence, CEK cells were infected with the M41 strain, which is confined to the respiratory tract. Western blot analysis showed no increase in NLRP3 expression, indicating that M41 does not activate the NLRP3 inflammasome (Fig. S2A).
## NLRP3 mediates IBV-induced renal injury
To assess the contribution of the NLRP3 inflammasome to IBV-induced renal injury, chickens infected with IBV were treated with the specific NLRP3 inhibitor MCC950 (Fig. 5A). At 5 dpi, MCC950 treatment markedly decreased renal expression of NLRP3, cleaved Caspase-1, and mature IL-1β proteins compared with IBV-infected control chickens (Fig. 5B through E). Consistently, ELISA analysis of serum samples showed significantly lower IL-1β concentrations in MCC950-treated groups (Fig. 5F). In addition, qRT-PCR analysis revealed that MCC950 treatment substantially reduced the transcription of multiple inflammasome-associated cytokines and chemokines in renal tissues (Fig. 5G). These findings demonstrate that MCC950 effectively suppresses NLRP3 inflammasome activation and attenuates downstream proinflammatory responses triggered by IBV infection.
In vivo inhibition of NLRP3 by MCC950 markedly attenuated disease progression, as evidenced by decreased clinical scores and reduced mortality in infected chickens (Fig. 5H andI). Gross pathological examination revealed lesions in the trachea of both IBV-infected and MCC950-treated groups, whereas no conspicuous macroscopic changes were observed in the lungs of any group. The major gross differences were renal, with MCC950 treatment reducing mottling, swelling, and urate deposition (Fig. 5J). Histopathological analysis further confirmed that MCC950 treatment alleviated microscopic lesions, including extensive inflammatory cell infiltration and tubular necrosis (Fig. 5K andL). Importantly, these protective effects occurred without significant changes in renal viral load (Fig. 5M), indicating that the protective efficacy of MCC950 is mainly attributable to modulation of host inflammatory responses rather than direct effects on viral replication. MCC950 treatment failed to alleviate IBV-induced tracheal ciliostasis in infected chickens (Fig. S3A). Similarly, histopathological and viral load analyses showed no improvement in tracheal or lung lesions compared with the IBV-infected group (Fig. S3B through D). Taken together, these findings indicate a central role of NLRP3 inflammasome activation in IBV-induced kidney injury and point to NLRP3 inhibition as a potential therapeutic approach to mitigate renal pathology and mortality in infected chickens.
## Collecting duct-specific NLRP3 inflammasome activation by IBV leads to inflammation and uric acid accumulation
Our earlier work provided evidence that IBV infects both distal tubular and collecting duct epithelial cells in the chicken kidney (19). To explore IBV-induced renal inflammation, kidney tissues collected at 5 dpi were analyzed by immunofluorescence for IBV nucleocapsid (N) and NLRP3 proteins. NLRP3 expression was upregulated and colocal ized with IBV-N in specific renal tubules (Fig. 6A), indicating inflammasome activation in tubular epithelial cells. Dual staining with AQP2 and CALB1 showed NLRP3 upregulation predominantly in AQP2 (Aquaporin-2)-positive collecting duct cells, but not in CALB1 (Calbindin)-positive distal tubules (Fig. 6B), that NLRP3 activation occurs predominantly in collecting duct epithelia. In contrast, qPCR and immunofluorescence analyses (IFAs) showed no increase in NLRP3 expression in tracheal tissues after IBV infection, indicating that NLRP3 inflammasome activation does not occur in the trachea (Fig. S2B andC).
To clarify IBV-induced inflammation's role in kidney injury, single-cell RNA sequenc ing (scRNA-seq) of collecting duct cells revealed a strong innate immune response. GSEA showed upregulation of viral sensors (TLR3, IFIH1, and EIF2AK2) and down stream pathways (IRF signaling, JAK/STAT, AP-1; Fig. S4A andB). Key transcription factors (STAT1/3, IRF1/7/8) were elevated, likely driving pro-inflammatory cytokine and chemokine expression (Fig. 6C). Immune regulatory and adhesion molecules (IL1RAP, F2RL1, and THBS1-ITGB6) were differentially expressed, suggesting enhanced immune cell recruitment and inflammation via altered cell interactions (Fig. S4B). Notably, negative regulators (SOCS1, DUSP10, and ZC3H12A) were also upregulated, indicating feedback to limit inflammation (Fig. 6C; Fig. S4C).
In chickens, uric acid is the primary nitrogenous waste excreted mainly via collecting ducts as urate salts (20,21). Our data show that IBV infection disrupts urate metabolism and transport in collecting duct cells. GSEA revealed upregulation of purine metabo lism and the rate-limiting enzyme xanthine dehydrogenase (XDH) (Fig. 6D), indicating increased uric acid synthesis and accumulation. Simultaneously, membrane transport pathways were activated, including elevated expression of the urate transporter SLC2A9 (GLUT9; Fig. 6E), suggesting enhanced tubular urate reabsorption contributing to renal urate buildup. Conversely, ion and small-molecule transport pathways crucial for cellular homeosta sis were significantly downregulated. GSEA showed suppression of ion transport, small molecule metabolism, and V-ATPase complex pathways vital for tubular acidification and urine pH regulation. Key transport genes, including AQP2, SCNN1G, SCNN1B, and V-ATPase subunits ATP6V0D1 and ATP6V0D2, were markedly decreased (Fig. 6F; Fig. S4D andE). In chickens, impaired tubular regulation of water, ions, and acid-base balance may reduce urate solubility and facilitate crystal formation during hyperuricemic states.
In conclusion, IBV infection induces a pathological state in collecting duct epithelia characterized by persistent inflammasome activation, altered uric acid metabolism, and impaired electrolyte and acid-base transport. These dysfunctions create a pro-inflammatory, metabolically imbalanced microenvironment that promotes urate crystal forma tion and hinders renal excretion. Urate deposition further amplifies inflammation via a positive feedback loop, contributing to progressive kidney injury and increased mortality in infected chickens.
## DISCUSSION
IBV-induced renal pathology is strain dependent. Mutations in the S1 subunit of the spike glycoprotein largely determine kidney tropism by modulating receptor binding and viral entry (22). Nephropathogenic lineages (e.g., GI-19 and GI-7) are associated with severe renal lesions (3), while accessory and nonstructural proteins further influence virulence and host immune responses (23,24).
Proinflammatory cytokines are key mediators of the innate immune response, initiating and modulating inflammation during viral infection (25,26). A range of viral pathogens can upregulate the expression of proinflammatory cytokines, such as interleukin (IL)-1β, IL-18, and IL-8, thereby activating inflammatory signaling pathways, promoting immune cell recruitment, and potentially causing tissue damage (27,28). Viruses such as SARS-CoV-2, MERS-CoV, and influenza virus have been reported to induce marked pathological changes in host tissues by promoting excessive cytokine production (13,14,29). In this study, GI-19 genotype IBV infection in chickens induced a marked inflammatory response in renal tissues. The major pathological features included multifocal mottled renal lesions, disruption of tubular architecture, and extensive infiltration of inflammatory cells. Consistently, the expression of IL1B, IL18, and IL8 mRNA was significantly upregulated in renal tissues from infected chickens, further supporting the role of GI-19 IBV infection in mediating renal inflammation.
The production of proinflammatory cytokines is largely mediated by inflammasome activation (30). Inflammasomes, as cytosolic multiprotein complexes, orchestrate pathogen-induced inflammatory signaling through caspase-1 activation, leading to the maturation and secretion of IL-1β and IL-18 (31). Multiple studies have demonstrated that viruses such as dengue virus, IAV, and enterovirus 71 (EV71) activate the NLRP3 inflammasome (32)(33)(34). NLRP3 activation is a key feature of coronavirus infections. It contributes to cytokine storms in severe SARS-CoV-2 infection, promotes inflammation and tissue damage during MERS-CoV infection, and drives inflammatory responses in SARS-CoV (35). However, whether the infectious bronchitis virus activates the NLRP3 inflammasome to induce inflammation and thereby contribute to tissue damage remains unclear. In this study, we demonstrate that IBV activates the NLRP3 inflammasome both in vivo and in vitro, thereby promoting the maturation and release of proinflammatory cytokines through the NLRP3-caspase-1-IL-1β signaling axis. To our knowledge, this is the first study to demonstrate that IBV infection activates the NLRP3 inflammasome in animal models, leading to IL-1β secretion and subsequent inflammatory responses. Similar to SARS-CoV-2 and MERS-CoV, where NLRP3 activation is triggered by mitochondrial ROS, K + efflux, and lysosomal destabilization (36), viral proteins such as ORF3a, E, and N can further enhance activation via ion flux or interaction with inflammasome components (37). IBV may employ comparable pathways, and elucidating these mechanisms will advance our understanding of coronavirus-induced inflammasome regulation. We confirmed the activation of the NLRP3 inflammasome, underscor ing its key role in IBV-induced renal inflammation. To further assess this, we used MCC950, a selective inhibitor of NLRP3 assembly and Caspase-1 activation via ATPase inhibition. Mechanistically, MCC950 directly engages the NACHT domain of NLRP3-at or near the Walker-B/ATPase site-thereby blocking ATP hydrolysis and locking NLRP3 in a closed, inactive conformation that prevents oligomerization and ASC speck forma tion (38). MCC950 specifically inhibits NLRP3 without affecting other inflammasomes (i.e., it does not inhibit AIM2, NLRC4, or NLRP1 under conditions where NLRP3 activa tion is potently suppressed) (39) and has shown efficacy in reducing inflammation and improving survival in SARS-CoV-2 models (40). In this study, MCC950 treatment in IBV-infected chickens did not alter renal viral load but significantly reduced the protein levels of NLRP3, cleaved Caspase-1, IL-1β, serum proinflammatory cytokines, and chemokine mRNA levels. Histopathology showed marked attenuation of renal inflammation, accompanied by improved clinical signs and survival. These results indicate that NLRP3 hyperactivation drives renal inflammation and pathology in IBV infection, and MCC950 provides protection by targeting this pathway without affecting viral replica tion.
Although NLRP3 plays a critical role in renal inflammation, MCC950 treatment failed to mitigate IBV-induced respiratory damage, including tracheal ciliostasis and lung lesions, and did not significantly reduce viral loads in these tissues (Fig. S3). These findings suggest that NLRP3 contributes to IBV pathogenesis in a tissue-spe cific manner, being more important in kidney injury than in respiratory impairment. This study demonstrates that in CEK cells, IBV infection activates the NLRP3 inflammasome cascade, promoting maturation and release of IL-1β. Notably, avian and mamma lian NLRP3 inflammasomes differ significantly in molecular architecture and activation mechanisms (41). In mammals, the activation of NLRP3 involves protein oligomeriza tion, a process that facilitates the recruitment of ASC and pro-Caspase-1, ultimately triggering inflammasome assembly (42). ASC is typically considered a morphological marker of NLRP3 assembly (43). Previous studies have reported that chickens, unlike mammals, lack the ASC-encoding gene (35). This evolutionary divergence leads to the absence of canonical ASC specks, posing challenges for elucidating the mechanisms of IBV-induced NLRP3 activation in avian models. Although ASC is absent, previous studies have demonstrated that chicken NLRP3 can mediate IL-1β maturation and secretion, indicating a potentially ASC-independent mechanism involving oligomerization-driven recruitment of downstream signaling components (44,45). Supporting this, treatment with the NLRP3 inhibitor CY-09 or Caspase-1 inhibitor Ac-YVAD-CMK significantly reduced IL-1β secretion, implicating the NLRP3/Caspase-1 axis. Thus, the perinuclear punctate NLRP3 aggregates observed in chicken cells may serve as alternative morpho logical markers of inflammasome activation in the absence of ASC.
Previous studies on the NLRP3 inflammasome have focused primarily on immune cells such as macrophages (46), but recent evidence highlights the role of epithelial cells in pathogen sensing and inflammation (47). In renal disease, tubular epithelial cells have been identified as key targets of NLRP3 activation (48). Experimental data show that CEK cells activate the NLRP3 inflammasome upon IBV infection, suggesting that renal epithelial cells are both targets of infection and active contributors to inflammation. In vitro models further reveal species-specific differences in avian inflammatory signaling compared to mammals. In mammalian cells, canonical NLRP3 activation typically follows a two-signal model: a priming step (Signal 1, e.g., LPS) and an activation step (Signal 2, e.g., nigericin) (8). However, this dual-signal protocol fails to activate NLRP3 in CEK cells, likely due to inefficient lipopolysaccharide (LPS) sensing via chicken TLR4 and attenuated TRIF-dependent signaling, leading to weak NF-κB activation and insufficient NLRP3 priming (49,50). Stimulation with the TLR3 agonist Poly(I:C) (51) and the smallmolecule NLRP3 activator BMS (46) induced NLRP3 puncta in CEK cells, but western blotting showed no significant increase in NLRP3 protein levels. These findings suggest that the current stimulation protocols may be inadequate for robust NLRP3 activation in CEK cells, representing a limitation of this study.
While human coronaviruses, including SARS-CoV-2, are known to activate the NLRP3 inflammasome in renal tissues and contribute to acute kidney injury, the spatial and temporal dynamics of this activation in specific tubular epithelial subtypes remain poorly understood (52)(53)(54). In this study, dual immunofluorescence staining revealed that IBV-induced NLRP3 expression is predominantly localized to collecting duct epithe lial cells in chickens-a previously unreported anatomic distribution, pointing to the collecting duct as a potential site of virus-driven inflammation. Given its essential role in water and electrolyte homeostasis, collecting duct dysfunction may destabilize renal physiology (55). Our data show that IBV infection induces a pro-inflammatory pheno typic shift in collecting duct epithelial cells, implicating this segment as a key contributor to renal inflammation. Notably, chickens excrete nitrogen as uric acid, primarily via urate crystals in the collecting ducts (20,21), making them especially vulnerable to urate-rela ted injury. IBV infection upregulates XDH and SLC2A9, increasing uric acid production and reabsorption, which raises uric acid levels and promotes urate crystal formation in the collecting ducts. These crystals act as damage-associated molecular patterns (DAMPs) to activate the NLRP3 inflammasome and amplify inflammation (56). Our findings suggest that IBV triggers a "urate crystal-inflammation" feedback loop that drives renal injury progression, akin to gout-associated nephropathy in humans (57,58). Thus, IBV-induced NLRP3 activation in collecting duct epithelial cells establishes a pathogenic link between urate metabolism disruption and inflammation, offering new insights into IBV-related renal pathogenesis.
## MATERIALS AND METHODS
## Viruses and cells
The SD and M41 subtype of IBV used in this study was previously isolated and identified by our team. HD11 cells were maintained in RPMI 1640 medium (Gibco, USA) supplemen ted with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. CEK cells were isolated from 18-day-old SPF chicken embryos. All cells were incubated at 37°C in a humidified atmosphere containing 5% CO 2 .
## Reagents and antibodies
Chicken IL-1β ELISA kit (SEA563Ga) and IL-18 ELISA kit (SEA064Ga) were purchased from Cloud-Clone Product, China. Total RNA Isolation Kit was purchased from Magen, Beijing, China. M5 HiPer Real-Time PCR Super Mix was purchased from Mei5bio, Beijing, China. Caspase 1 Activity Assay Kit, Immunol Staining Fix Solution, Immunostaining Permeabili zation Buffer containing Triton X-100, Immunol Staining Blocking Buffer, CCK-8 solution, and Cell Lysis Buffer for Western were purchased from Beyotime, China. Chemicals used included CY-09 (MedChemExpress, HY-103666), Ac-YVAD-cmk (Selleck, S9727), MCC950 Sodium (Selleck, S7809), BMS-986299 (Selleck, S9899), and polyinosinic-polycytidylic acid (poly[I:C]; Sigma, P1530). Antibodies were mouse mab IBV-N protein (HyTes, 3BN1), rabbit pab NLRP3 (ABclonal, A24297), rabbit pab IL1β (ABclonal, A16288), rabbit pab Cleaved-caspase1 (Wanlei Biotechnology, WL03450), mouse mab β-actin (ABclonal, AC004), mouse mab GAPDH (ABclonal, AC033), rabbit pab AQP2 (Abmart, PK58037), and mouse mab calb1 (Boster Bio, BM0203). Alexa Fluor 488-conjugated anti-mouse IgG (H+L; Cell Signaling Technology, 4408) and Alexa Fluor 555-conjugated anti-rabbit IgG (H+L; Cell Signaling Technology, 4413).
## Enzyme-linked immunosorbent assay
The concentrations of IL-1β and IL-18 in cell culture supernatants and serum were quantified by ELISA according to the manufacturer's instructions. Briefly, 100 µL of standards and test samples were added to each well of a pre-coated 96-well microplate and incubated at 37°C for 1 hour. After discarding the contents, the plate was gently blotted dry on absorbent paper. Then, 100 µL of biotin-labeled anti-IL-1β antibody was added to each well and incubated at 37°C for an additional hour, followed by three washes with wash buffer. Subsequently, 100 µL of horseradish peroxidase (HRP)-conju gated streptavidin was added, and the plate was incubated at 37°C for 30 minutes, followed by five washes. Thereafter, 90 µL of TMB substrate was added to each well and incubated in the dark at 37°C for 10-20 minutes. The reaction was terminated by adding 50 µL of stop solution, and the optical density was measured at 450 nm.
## Caspase-1 activity assay
Caspase-1 activity in cell lysates was determined using a Caspase-1 Activity Assay Kit (Beyotime, China). Briefly, 50 µL of cell lysate was mixed with 40 µL of assay buffer, followed by the addition of 10 µL of Ac-YVAD-pNA (2 mM) substrate, yielding a final volume of 100 µL. For the blank control, the lysis buffer was used in place of the sample. The reaction mixture was incubated at 37°C until visible color development, and the absorbance was measured at 405 nm using a microplate reader.
## Real-time quantitative PCR
Total RNA was extracted from cells or tissues using an RNA extraction kit and reversetranscribed to generate cDNA. The resulting cDNA was amplified by real-time quanti tative PCR (RT-qPCR) using the M5 HiPer Real-Time PCR Super Mix. Thermal cycling conditions were initial denaturation at 95°C for 30 s; 40 cycles of 95°C for 5 s and 60°C for 30 s; followed by melt-curve analysis from 60°C to 90°C at a ramp rate of 1°C/s. The expression levels of target genes were normalized to β-actin mRNA expression. Detailed primer sequences are listed in Table S1.
## Detection of viral RNA in tissue samples
Total RNA was extracted from cells or tissue homogenates using a commercial RNA extraction kit with on-column DNase treatment, and first-strand cDNA was synthesized using a standard reverse-transcription kit. IBV RNA was quantified by RT-qPCR (SYBR Green chemistry, M5 HiPer Real-Time PCR Super Mix) using primers targeting the conserved 5′-UTR (forward 5′-GTTGGGCTACGTTCTCGC-3′; reverse 5′-AAGCCATGTTGTCA CTGTCTAT-3′; amplicon 130 bp). Each 20 µL reaction contained 2 × M5 HiPer SYBR Premix 10 µL, forward primer 0.4 µL (10 µM; 0.2 µM final), reverse primer 0.4 µL (10 µM; 0.2 µM final), cDNA 2 µL, and nuclease-free water 7.2 µL.
## Dual immunofluorescence of kidney sections
Dual immunofluorescence staining of kidney tissues was performed using the mIHC Dual Immunofluorescence Kit (Panovue, China). Briefly, paraffin-embedded kidney tissues were sectioned at a thickness of 4 µm and mounted on slides. After deparaffinization in xylene and rehydration through a graded ethanol series, heat-induced antigen retrieval was carried out in citrate-EDTA buffer (Beyotime, China) for 10 minutes. Sections were then blocked with blocking buffer at room temperature for 1 hour, followed by incubation with the first primary antibody at 4°C overnight. The next day, sections were washed three times with TBST and further rinsed according to the kit instructions. Fluorophore-conjugated secondary antibodies were applied and incubated at room temperature for 20 minutes, followed by TSA amplification for 10 minutes, and then washed. Antibody stripping was performed using the stripping buffer provided in the kit for 15 minutes, followed by TBST washing. The second round of staining was carried out by repeating the primary antibody incubation and subsequent steps as described above.
After staining, the sections were coverslipped, and fluorescence signals were acquired using a Nikon A1 confocal microscope (Nikon, Tokyo, Japan).
## Indirect IFA and confocal microscopy
Cell samples were collected at predetermined time points post-transfection or infec tion and fixed using Immunol Staining Fix Solution. Subsequently, the cells were permeabilized with Immunostaining Permeabilization Buffer containing Triton X-100 and blocked with Immunol Staining Blocking Buffer. The cells were then incubated with specific primary antibodies at 4°C for 12 hours. For staining, Alexa Fluor 488-conjugated anti-mouse IgG (H+L) and/or Alexa Fluor 555-conjugated anti-rabbit IgG (H+L) were added and incubated at room temperature in the dark for 1 hour. Nuclei were stained with DAPI at room temperature for 10 minutes. The cells were then washed five times with phosphate-buffered saline (PBS) containing Tween 20 (PBST), with each wash lasting 5 minutes. Finally, the cells were observed and imaged using a Nikon A1 fluorescence microscope (Nikon, Tokyo, Japan). Colocalization analysis was performed using Fiji ImageJ software. The observed correlation coefficient (R[obs]) represents the Pearson correlation coefficient calculated using ImageJ, ranging from -1 to +1, where +1 indicates a perfect positive correlation (complete colocalization), 0 indicates no correlation (random distribution), and -1 indicates a perfect negative correlation.
## Drug treatment and virus infection
CEK cells were seeded into culture plates and grown to 80%-90% confluence. Prior to infection with IBV at a multiplicity of infection (MOI) of 0.01, cells were pretreated with small-molecule inhibitors at the indicated concentrations for 2 hours. Following pretreatment, the drug-containing medium was removed, and cells were infected with IBV for 2 hours at 37°C to allow viral adsorption. After infection, the viral inoculum was discarded, and cells were washed twice with PBS. Fresh maintenance medium containing the corresponding concentrations of the inhibitors was then added, and cells were incubated for an additional 24 hours at 37°C in a 5% CO 2 incubator. After treatment, cells or supernatants were collected for downstream analyses, including IL-1β and Caspase-1 activity assays. The inhibitors used in this study included CY-09 and Ac-YVAD-cmk, all of which were administered following the same treatment protocol. Stock solutions were prepared by dissolving each compound in dimethyl sulfoxide or sterile water. Solvent control groups were prepared using equivalent volumes of the respective solvents to account for solvent-related effects.
## In vivo infection experiments and sample collection
Two independent in vivo infection experiments were conducted. In the first experiment, 1-day-old SPF chickens were randomly assigned to two groups: the infectious bronchitis virus strain SD (IBV-SD) group and the control group. Birds in the IBV-SD group were inoculated with 10 5.5 TCID₅₀ of IBV-SD via the intranasal and ocular routes, while the control group received an equal volume of PBS.
In the second experiment, another cohort of 1-day-old SPF chicken was randomly assigned to four groups: the IBV-SD group, IBV-SD group treated with high-dose MCC950 (50 mg/kg/day), IBV-SD group treated with low-dose MCC950 (20 mg/kg/day), and a negative control group. All virus-inoculated groups received 10 5.5 TCID₅₀ of the corre sponding viral strain via the same administration route. MCC950 was administered daily via intraperitoneal injection. The dosages of MCC950 (50 mg/kg/day and 20 mg/kg/day) were selected based on previous studies involving chickens and other animals, which demonstrated effective inhibition of NLRP3 inflammasome activation at these doses (59,60). The intraperitoneal injection route was chosen to ensure efficient drug delivery and uniform distribution in the body.
Throughout the experimental period, all chickens were monitored daily for clinical signs, including sneezing, tracheal rales, and somnolence. On designated days postinfection, three chickens from each group were randomly selected for euthanasia and necropsy. Macroscopic lesions in the trachea, lungs, and kidneys were recorded. Tracheal ciliary activity and mean lesion scores were assessed as previously descri bed (24). Peripheral blood and tissue samples were collected aseptically. Portions of tissue samples were fixed in 10% neutral-buffered formalin for immunohistochemical or histopathological analysis, and lesion severity was evaluated according to criteria described in previous studies (61). Remaining tissues were snap frozen in liquid nitrogen and stored at -80°C for analysis of gene transcription and protein expression levels. Blood samples were centrifuged to obtain serum, and cytokine concentrations were measured using ELISA.
## References
1. Schalk, Hawn (1931) "An apparently new respiratory disease of baby chicks" *J Am Vet Med Assoc*
2. Fabricant (1998) "The early history of infectious bronchitis" *Avian Dis*
3. Zhao, Zhao, Zhang (2023) "Key aspects of coronavirus avian infectious bronchitis virus" *Pathogens*
4. Chong, Apostolov (1982) "The pathogenesis of nephritis in chickens induced by infectious bronchitis virus" *J Comp Pathol*
5. Elinav, Strowig, Henao-Mejia et al. (2011) "Regulation of the antimicrobial response by NLR proteins" *Immunity*
6. Schroder, Tschopp (2010) "The inflammasomes" *Cell*
7. Fu, Wu (2023) "Structural mechanisms of NLRP3 inflammasome assembly and activation" *Annu Rev Immunol*
8. (2025) *Full-Length Text Journal of Virology*
9. Paik, Kim, Shin et al. (2025) "Updated insights into the molecular networks for NLRP3 inflammasome activation" *Cell Mol Immunol*
10. Waldstein, Varga (2022) "Respiratory viruses and the inflammasome: the double-edged sword of inflammation" *PLoS Pathog*
11. Davis, Wen, Ting (2011) "The inflammasome NLRs in immunity, inflammation, and associated diseases" *Annu Rev Immunol*
12. Jose, Manuel (2020) "COVID-19 cytokine storm: the interplay between inflammation and coagulation" *Lancet Respir Med*
13. Shi, Nabar, Huang et al. (2019) "SARS-Coronavirus open reading frame-8b triggers intracellular stress pathways and activates NLRP3 inflammasomes" *Cell Death Discov*
14. Zhou, Chu, Li et al. (2014) "Active replication of Middle East respiratory syndrome coronavirus and aberrant induction of inflammatory cytokines and chemokines in human macrophages: implications for pathogenesis" *J Infect Dis*
15. Yoshizumi, Ichinohe, Sasaki et al. (2014) "Influenza A virus protein PB1-F2 translocates into mitochondria via Tom40 channels and impairs innate immunity" *Nat Commun*
16. Jang, Koo, Jeon et al. (2013) "Altered proinflammatory cytokine mRNA levels in chickens infected with infectious bronchitis virus" *Poult Sci*
17. Li, Huang, Li et al. (2022) "Nephropathogenic infectious bronchitis virus mediates kidney injury in chickens via the TLR7/NF-κB signaling axis" *Front Cell Infect Microbiol*
18. Broz, Dixit (2016) "Inflammasomes: mechanism of assembly, regulation and signalling" *Nat Rev Immunol*
19. Lin, Lv, Sun et al. (2022) "TRIM50 promotes NLRP3 inflammasome activation by directly inducing NLRP3 oligomerization"
20. Liukang, Zhao, Tian et al. (2024) "Deciphering infected cell types, hub gene networks and cell-cell communication in infectious bronchitis virus via single-cell RNA sequencing" *PLoS Pathog*
21. Scanes, Dridi (2022) "Chapter 27 -protein metabolism"
22. Dantzler (2005) "Challenges and intriguing problems in comparative renal physiology" *J Exp Biol*
23. You, Liu, Huang et al. (2023) "Identification and comparison of the sialic acid-binding domain characteristics of avian coronavirus infectious bronchitis virus spike protein" *J Virol*
24. Laconi, Van Beurden, Berends et al. (2018) "Deletion of accessory genes 3a, 3b, 5a or 5b from avian coronavirus infectious bronchitis virus induces an attenuated pheno type both in vitro and in vivo" *J Gen Virol*
26. Zhao, Sun, Zhao et al. (2021) "Coronavirus endoribonuclease ensures efficient viral replication and prevents protein kinase R activation" *J Virol*
27. Bloom, Zinkernagel (1996) "Immunity to infection" *Curr Opin Immunol*
28. Ganaie, Wang, Su et al. (2023) "Editorial: Virus-induced innate immune response and inflammation" *Front Microbiol*
29. Chaudhary, Meher, Krishnamoorthy et al. (2023) "Interplay of host and viral factors in inflammatory pathway mediated cytokine storm during RNA virus infection" *Curr Res Immunol*
30. Alhamlan, Aa (2024) "Pro-inflammatory and anti-inflammatory interleukins in infectious diseases: a comprehensive review" *Trop Med Infect Dis*
31. Vora, Lieberman, Wu (2021) "Inflammasome activation at the crux of severe COVID-19" *Nat Rev Immunol*
32. Yao, Sterling, Wang et al. (2024) "The role of inflammasomes in human diseases and their potential as therapeutic targets" *Sig Transduct Target Ther*
33. Latz, Xiao, Stutz (2013) "Activation and regulation of the inflammasomes" *Nat Rev Immunol*
34. Shrivastava, Visoso-Carvajal, Garcia-Cordero et al. (2020) "Dengue virus serotype 2 and its non-structural proteins 2A and 2B activate NLRP3 inflammasome" *Front Immunol*
35. Allen, Scull, Moore et al. (2009) "The NLRP3 inflammasome mediates in vivo innate immunity to influenza A virus through recognition of viral RNA" *Immunity*
36. Wang, Lei, Yang et al. (2015) "Reciprocal regulation between enterovirus 71 and the NLRP3 inflammasome" *Cell Rep*
37. Shah (2020) "Novel coronavirus-induced NLRP3 inflammasome activation: a potential drug target in the treatment of COVID-19" *Front Immunol*
38. Yang, Wang, Kouadir et al. (2019) "Recent advances in the mechanisms of NLRP3 inflammasome activation and its inhibitors" *Cell Death Dis*
39. Guarnieri, Angelin, Murdock et al. (2023) "SARS-COV-2 viroporins activate the NLRP3inflammasome by the mitochondrial permeability transition pore" *Front Immunol*
40. Coll, Hill, Day et al. (2019) "MCC950 directly targets the NLRP3 ATP-hydrolysis motif for inflammasome inhibition" *Nat Chem Biol*
41. Coll, Robertson, Chae et al. (2015) "A small-molecule inhibitor of the NLRP3 inflammasome for the treatment of inflammatory diseases" *Nat Med*
42. Pan, Shen, Yu et al. (2021) "SARS-CoV-2 N protein promotes NLRP3 inflammasome activation to induce hyperinflammation" *Nat Commun*
43. Billman, Hancks, Miao (2024) "Unanticipated loss of inflammasomes in birds" *Genome Biol Evol*
44. Yu, Matico, Miller et al. (2024) "Structural basis for the oligomerization-facilitated NLRP3 activation" *Nat Commun*
45. Andreeva, David, Rawson et al. (2021) "NLRP3 cages revealed by full-length mouse NLRP3 structure control pathway activation" *Cell*
46. Gao, Chen, Fan et al. (2020) "Newcastle disease virus RNAinduced IL-1β expression via the NLRP3/caspase-1 inflammasome" *Vet Res*
47. Mao, Liu, Wu et al. (2024) "Duck hepatitis a virus 1-encoded 2B protein disturbs ion and Full-Length Text Journal of Virology December"
48. "organelle homeostasis to promote NF-κB/NLRP3-mediated inflammatory response" *Int J Biol Macromol*
49. Zito, Buscetta, Cimino et al. (2020) "Cellular models and assays to study NLRP3 inflammasome biology" *Int J Mol Sci*
50. Rao, Zhou, Deng et al. (2023) "Activation of NLRP3 inflammasome in lung epithelial cells triggers radiation-induced lung injury" *Respir Res*
51. Vilaysane, Chun, Seamone et al. (2010) "The NLRP3 inflammasome promotes renal inflammation and contributes to CKD" *J Am Soc Nephrol*
52. Keestra, Van Putten (2008) "Unique properties of the chicken TLR4/MD-2 complex: selective lipopolysaccharide activation of the MyD88-dependent pathway" *J Immunol*
53. Neerukonda, Katneni (2020) "Avian pattern recognition receptor sensing and signaling" *Vet Sci*
54. Hong, Lee, Vu et al. (2021) "Immunomodula tory effects of poly(I:C)-stimulated exosomes derived from chicken macrophages" *Poult Sci*
55. Chauhan, Walle, Lamkanfi (2020) "Therapeutic modulation of inflammasome pathways" *Immunol Rev*
56. Zhang, Gerzanich, Cruz-Cosme et al. (2024) "SARS-CoV-2 ORF3a induces COVID-19-associated kidney injury through HMGB1-mediated cytokine production"
57. Falcón-Cama, Montero-González, Acosta-Medina et al. (2023) "Evidence of SARS-CoV-2 infection in postmortem lung, kidney, and liver samples, revealing cellular targets involved in COVID-19 pathogenesis" *Arch Virol*
58. Pearce, Soundararajan, Trimpert et al. (2015) "Collecting duct principal cell transport processes and their regulation" *Clin J Am Soc Nephrol*
59. Wen, Yang, Ma et al. (2021) "The roles of NLRP3 inflammasomemediated signaling pathways in hyperuricemic nephropathy" *Mol Cell Biochem*
60. Maiuolo, Oppedisano, Gratteri et al. (2016) "Regulation of uric acid metabolism and excretion" *Int J Cardiol*
61. Braga, Forni, Correa-Costa et al. (2017) "Soluble uric acid activates the NLRP3 inflammasome" *Sci Rep*
62. Shi, Wang, Zhang et al. (2022) "Dihydromyricetin alleviates Escherichia coli lipopolysaccharide-induced hepatic injury in chickens by inhibiting the NLRP3 inflammasome" *Vet Res*
63. Zeng, Xie, Feng et al. (2022) "Specific inhibition of the NLRP3 inflammasome suppresses immune overactivation and alleviates COVID-19 like pathology in mice" *EBioMedicine*
64. Xu, Ma, Cheng et al. (2021) "An attenuated TW-like infectious bronchitis virus strain has potential to become a candidate vaccine and S gene is responsible for its attenuation" *Vet Microbiol*
65. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12587137&blobtype=pdf | # Southern African Journal of HIV Medicine
Natasha O'connell, Tara-Lee Von Mollendo, Gert Van Zyl, Stephen Korsm, James Nuttall
## Abstract
Patient presentationA 4-month-old female full-term infant, breast-and formula-fed since birth, was diagnosed with HIV when admitted to intensive care for severe pneumonia due to cytomegalovirus (CMV) and Pneumocystis jirovecii. Her CD4 count was 153 cells/µL (30%). As per the national guidelines, 1 no baseline HIV viral load (VL) was done; however, the positive HIV total nucleic acid test, with cycle threshold value of 16.7, suggested a high concentration of either HIV-1-RNA, -DNA, or both. Abacavir/lamivudine/dolutegravir (ABC/3TC/DTG) was initiated along with ganciclovir, followed by valganciclovir for 6 weeks, prednisone, and cotrimoxazole for 21 days, followed by daily cotrimoxazole prophylaxis.
She was discharged after prolonged oxygen treatment and difficulty establishing oral feeds, and was re-admitted within 2 weeks with acute gastroenteritis. An ophthalmology review 3 weeks later showed left eye CMV retinitis; valganciclovir was reinitiated. The mother reported poor feed and medication tolerance. At 7 months of age (3 months after ART initiation) she was re-admitted with septic shock resulting from Pseudomonas aeruginosa bacteraemia. Her HIV VL at that time was > 10 million copies/mL (> 7 log 10 ) and CD4 count 41 cells/µL (17%). Despite continued valganciclovir treatment, CMV retinitis progressed to bilateral disease with left macula involvement threatening her vision and requiring regular intravitreal ganciclovir injections.
HIV drug resistance (HIVDR) testing 10 days after re-admission revealed high-level resistance to ABC, 3TC, and DTG. ART was switched to zidovudine/lamivudine/lopinavir/ritonavir (AZT/3TC/LPV/r), < 4 months after starting the primary ART regimen. Within 3 weeks of initiating AZT/3TC/LPV/r, Mycobacterium bovis Bacillus Calmette-Guérin (BCG) disease, with right axillary and chest wall abscesses and pneumonia, was diagnosed on induced sputum. She was treated with rifampicin, isoniazid, ethambutol, and levofloxacin for 9 months, and LPV/r was boosted with additional ritonavir. Dolutegravir resistance is predominantly reported in antiretroviral therapy-experienced individuals. We describe an infant who developed high-level resistance to abacavir, lamivudine, and dolutegravir within 97 days of initiation, despite initial wild-type infection. Causative factors likely include probable sub-therapeutic antiretroviral drug levels, poor tolerance, severe immunocompromise, and a high pre-treatment viral load.
## Rapid development of drug resistance during initial dolutegravir-based antiretroviral therapy of an infant with HIV
## Read online:
Scan this QR code with your smart phone or mobile device to read online.
## What this study adds:
The evolution of DTG resistance from a confirmed transmitted wild-type virus in this infant is more rapid than previously described. Identification of risk factors for developing DTG resistance in infants requires further investigation.
http://www.sajhivmed.org.za Open Access
Both parents were newly diagnosed with HIV at the time of the infant's diagnosis. The mother reported having tested HIV negative during a previous pregnancy. Both parents reported being ART-naïve and were started on tenofovir/ lamivudine/dolutegravir (TLD), achieving early viral suppression. Alternative mechanisms for HIV infection in the infant, such as blood transfusions, scarification and nonmaternal breastmilk exposure, were not reported.
Since DTG resistance is unexpected so soon after starting first-line ART, we aimed to establish whether drug resistance was transmitted from the mother to the child, or developed rapidly in the infant.
## Methods
HIV drug resistance testing of the HIV polymerase (pol) gene was performed on ethylenediaminetetraacetic acid blood samples from the mother and repeated for confirmation on the infant 2 months after initiation of AZT/3TC/LPV/r. The father was virologically suppressed on TLD, and HIVDR testing was not performed.
The infant's VL was 13 355 copies/mL, enabling bulk 3 kb pol gene sequencing 2 covering integrase (int), reverse transcriptase (rt), and protease (pro). The mother's VL of 63 copies/mL required two separate assays to amplify shorter regions: a published assay to amplify the HIV-1 group M integrase gene, 3 and a published assay to amplify HIV-1 gag p6, protease and reverse transcriptase (p6-pro-rt). 4 Assays were performed at limiting dilution, yielding eight amplicons for integrase and three for p6-pro-rt.
We also sequenced the original dried blood spots (DBS) collected from the infant at diagnosis to determine whether HIVDR was present at diagnosis.
Sequencing libraries were prepared using Oxford Nanopore Technologies Native Barcoding Kit 96 V14 (SQK-NBD114.96), and sequenced on R10 ONT Flongle flow cells. The Nano-RECall pipeline was used for sequence alignment and correction. Relatedness between infant and maternal sequences was assessed with neighbour joining phylogenetic analysis. Drug resistance interpretation was with the Stanford HIV Drug Resistance Database.
## Ethical considerations
Written consent for the publication of this case report was obtained from the mother and ethical approval from the University of Cape Town (reference number: HREC REF: 322/2025).
## Results
Neighbour joining phylogenetic analysis revealed high (> 98%) sequence identity between the mother and infant across all HIV enzymes (Figure 1), with similar polymorphisms, corroborating mother-to-child transmission. The infant's sample from 2 months after switching to AZT/3TC/LPV/r showed no protease inhibitor mutations, a nucleoside reverse transcriptase inhibitor (NRTI) mutation, M184V, and two major DTG-associated mutations, G118R and E138K, which were absent in the mother's sample and the infant's original DBS. This supports the rapid acquisition of HIVDR during 97 days of DTG-containing ART.
## Discussion
Dolutegravir-based ART is now the preferred first-and second-line treatment across all populations because of superior tolerability, few drug interactions, and high genetic barrier to resistance. South Africa adopted DTG-based regimens in 2019. 1 In this case, HIVDR testing revealed several mutations: M184V/I, conferring high-level resistance to 3TC and lowto-intermediate resistance to ABC; G118R, associated with virologic failure on DTG and causing a > 10-fold reduction in DTG susceptibility; and E138K, linked to integrase strand transfer inhibitor use, reducing drug efficacy further with other mutations. 5 Dolutegravir resistance is rare in treatment-naïve individuals who start treatment on DTG-based ART; however, high baseline VL, severe immunosuppression and sub-therapeutic drug levels are key risk factors. 6 This infant was virally unsuppressed and severely immunosuppressed after 97 days on ART (VL > 7 log 10 ; CD4 count 41 cells/µL). High viral replication increases the chance of mutations, and poor feed tolerance may have led to sub-therapeutic ART levels, further contributing to developing resistance. Advanced HIV disease with severe immunosuppression and active opportunistic infections can be associated with malabsorption and impaired ART absorption. 6 At this stage the infant was not on rifampicin, which can further reduce DTG drug levels. Zinc was prescribed for 10 days, but details on timing of administration in relation to ART to prevent reduction in DTG levels were not documented.
High-level baseline DTG resistance has been reported in two children who were newly diagnosed with HIV (14 and 18 months of age, respectively). 7,8 In both cases (R263K in one, and E138K and G118R mutations in the other), the mothers were on DTG-based ART during pregnancy and/or breastfeeding with intermittently unsuppressed VLs resulting from treatment interruptions, but HIVDR testing was not performed on them. Since paired mother-child HIVDR testing at the time of infant diagnosis was not done, HIVDR in these children was unexplained and may have been acquired because of prolonged low-level DTG exposure through breastmilk. As the mothers ultimately had a favourable VL response to DTG treatment, transmitted resistance is unlikely.
This report provides strong evidence of rapidly acquired DTG resistance and, in contrast to previous reports, we were able to show the acquisition of drug resistance between an initial DBS sample and a subsequent sample 3 months later. Additionally, paired maternal-infant HIVDR testing corroborates acquisition of drug resistance after transmission.
## Conclusion
These findings support the rapid acquisition of DTG resistance in an infant treated with ABC/3TC/DTG. Likely contributing factors are the high VL, probable reduced ART drug levels because of poor tolerance, and severe immunosuppression, which favour accelerated viral replication and resistance evolution.
## References
1. "South African National Department of Health. National consolidated guidelines for the management of HIV in adults, adolescents, children, and infants and prevention of mother-to-child transmission"
2. Coetzee, Woods, Delaney (1099) "Full pol-gene PCR and rapid ONT library preparation for accurate drug resistance sequencing"
3. Van Laethem, Schrooten, Covens (2008) "A genotypic assay for the amplification and sequencing of integrase from diverse HIV-1 group M subtypes" *J Virol Methods*
4. Van Zyl, Katusiime, Wiegand (2017) "No evidence of HIV replication in children on antiretroviral therapy" *J Clin Invest*
5. "HIV drug resistance database: INSTI and NRTI resistance notes"
6. Stanford (2025)
7. Cevik, Orkin, Sax (2020) "Emergent resistance to dolutegravir among INSTI-naïve patients on first-line or second-line antiretroviral therapy: A review of published cases" *Open Forum Infect Dis*
8. Francois, Van Onacker, Jordan (2023) "First case report of a perinatally HIVinfected infant with HIV resistance to dolutegravir associated with tenofovir/ lamivudine/dolutegravir use in mothers" *AIDS*
9. Anderson, Van Zyl, Hsiao (2024) "HIV drug resistance in newly diagnosed young children in the Western Cape, South Africa" *Pediatr Infect Dis J* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12606788&blobtype=pdf | # M AT T E R S A R I S I N G
Guido Bendezu-Quispe
## Abstract
Background We read with great interest Wen et al. 's article on HPV infection in Huizhou men, commending this pioneering research as the first systematic analysis of HPV epidemiology in the region. The findings, including a 30.53% positivity rate and prevalent genotypes like HPV6, HPV52, HPV11, and HPV16, offer valuable insights for developing effective prevention strategies.
analyze the epidemiological characteristics of HPV infections in men in the Huizhou region. The findings, such as the overall HPV positivity rate of 30.53% and the prevalence of HPV6, HPV52, HPV11, and HPV16 genotypes, offer valuable insights for prevention strategies [1]. However, we would like to express our concern regarding the methodology for classifying HPV infection types employed in the study.
## Main body
The authors categorized infections into "low-risk HPV, " "high-risk HPV, " and "mixed HPV. " Specifically, the "mixed HPV" category includes any combination of high and low-risk genotypes [1]. This grouping may inadvertently obscure the true incidence and clinical implications of high-risk HPV infections in the studied population.
## Background
We read with great interest the recent article by Wen et al., "HPV infection incidence and genotype distribution among male patients visiting outpatient departments in Huizhou from 2014 to 2023" [1], published in your distinguished journal. We commend the authors for their pioneering research, as it is the first study to systematically Main body We are concerned, however, about the study's HPV classification methodology. Categorizing infections into "low-risk, " "high-risk, " and "mixed" (which includes any combination of high and low-risk genotypes) might obscure the true incidence of high-risk HPV infections. High-risk HPV types are strongly linked to malignant transformations. Hence, a single high-risk genotype poses a significant health risk, irrespective of low-risk co-infection. Grouping highrisk infections with low-risk types in the "mixed" category could therefore underestimate the proportion of patients with high-risk HPV.
## Conclusion
For future research, we suggest presenting high-risk HPV prevalence by distinguishing between solely low-risk, solely high-risk, and high-risk mixed infections (multiple infections that include at least one high-risk genotype). This offers a more accurate understanding of the burden, aiding public health efforts, screening, and vaccination programs.
Diagnostic criteria for HPV infection classification: a comment on Wen et al.
Elian Analí Camero-Ortiz 1 , Silvia Karen Delgado-García 2 , Guido Bendezu-Quispe 3* and Gustavo A. Grandez-Castillo 4 According to the National Cancer Institute (NIH), high-risk carcinogenic HPV genotypes include 12 types: HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59 [2]. Among these, HPV 16 and HPV 18 are recognized as the most frequent carcinogenic HPV-related genotypes. In men, the most frequent genotypes are low-risk HPV 6 and 11, with HPV 16 being high-risk. These types are strongly related to carcinogenic diseases, such as cervical cancer and other malignancies in men, such as anal, penile, and oral cancer [3]. The presence of a single highrisk HPV genotype already confers a significant health risk, regardless of coinfection with low-risk types. However, in the Wen et al. study, by combining high-risk infections with any low-risk type in the "mixed" category, the study could underestimate the proportion of patients with high-risk HPV infections. For example, classifying a patient coinfected with HPV6 (low-risk) and HPV16 (high-risk) as "mixed, " rather than primarily recognizing it as a high-risk HPV infection, could lead to an underestimation of the true proportion of patients with high-risk HPV infections in the cohort.
Therefore, for future epidemiological studies, we suggest that the prevalence of HPV infections be presented in a manner that clearly distinguishes between: (i) infections solely of low-risk HPV, (ii) infections solely of highrisk HPV, and (ii) high-risk mixed infections (multiple infections that include at least one high-risk genotype). This proposed distinction, which has been reported in previous studies [4], would facilitate a more accurate understanding of the burden of high-risk HPV infections in the population. Consequently, it would enable better prioritization of public health efforts, screening programs, and vaccination strategies [5].
## Conclusion
Categorizing HPV infections as "low-risk", "high-risk", and "mixed" (considering any HPV genotype combination) may underestimate the true incidence and clinical implications of high-risk HPV. We suggest that presenting the prevalence of "solely high-risk HPV", "solely lowrisk HPV", and "high-risk mixed infections" could enhance HPV genotyping reports. We appreciate the authors' contribution and hope that this feedback will aid future research.
## References
1. Wen, Yang, Wu (0272) "HPV infection incidence and genotype distribution among male patients visiting outpatient departments in Huizhou from 2014 to 2023" *Virol J*
2. Posada, Acevedo, Arredondo et al. (2020) "High-risk human papillomavirus infection and associated factors in the anal canal of HIV-positive patients in Medellín, 2017-2018" *Rev Saúde Pública*
3. 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*
4. Sendagorta-Cudós, Burgos-Cibrián, Rodríguez-Iglesias (2009) "Infecciones genitales por el virus del papiloma humano. Enfermedades Infecc Microbiol Clínica [Internet]" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12839329&blobtype=pdf | # miR-204 Negatively Regulates HIV-Tat-Mediated Inflammation in Cervical Epithelial Cells via the NF-κB Axis: Insights from an In Vitro Study
Kadambari Akolkar, Vandana Saxena
## Abstract
Despite antiretroviral therapy, HIV proteins, such as Tat, persist in tissues, driving chronic inflammation. Cervical inflammation in females not only accelerates HIV progression but also increases the risk of other STIs; hence, understanding the underlying factors/regulators is vital. However, Tat-induced cervical inflammation and its regulation are hitherto poorly understood, which we investigated using TZM-bl cells. Tat stimulation in these cervical epithelial cells significantly increased the expression of various inflammatory mediators, including cytokines (IL-1β, TNF-α, IL-6, IL-17a, GM-CSF), chemokines (MIP-1α, MIP-1β), adhesion molecules (ICAM-1, P-Selectin, E-Selectin), and ROS. Further upregulation of inflammatory mediators (NF-κB, IRAK-4) along with TLR7 was observed in Tat-stimulated cells. Interestingly, Tat stimulation decreased miR-204-5p expression in these cells, suggesting a role in regulating Tat-mediated inflammatory processes. Using a gain-of-function approach, we further observed that the overexpression of miR-204-5p reduced the expression of IL-1β, TNF-α, IL-6, MIP-1α, MIP-1β, ICAM-1, P-Selectin, and ROS in the Tat-stimulated TZM-bl cells, along with NF-κB, IRAK-1, and IRAK-4. Using Western blotting and luciferase assays, miR-204-5p was further shown to directly target NF-κB. Here, we report that HIV-1 Tat stimulation in cervical epithelial cells downregulates hsa-miR-204-5p, thereby activating the pro-inflammatory TLR7/NF-κB axis, highlighting its relevance to understanding mechanisms underlying cervical inflammation.
## 1. Introduction
Chronic immune activation, characterized by increased levels of pro-inflammatory cytokines and chemokines, is a pathognomonic feature of progressive HIV infection [1]. HIV acquisition and spread predominantly occur at mucosal surfaces, where mucosal inflammation further increases the risk of infection and transmission [2]. Among females, the cervicovaginal mucosa serves as the main portal of HIV and secondary transmission during heterosexual contact. Inflammation in the female genital tract creates an environment that favors HIV replication and establishment of a productive infection. Although combination antiretroviral therapy (cART) results in effective viral suppression, persistent immune activation and inflammation remain significant challenges. Even in the presence of cART, the early viral proteins, including HIV-1 Tat (Transactivator of Transcription), persist in tissues [3,4], contributing to chronic inflammation. Tat, an essential viral protein for HIV transcription and replication [5,6], is secreted from HIV-infected cells and perturbs both HIVinfected and bystander cells [7,8]. In context with the cervical microenvironment, it has been shown that extracellularly released HIV-1 Tat protein is internalized by the cervical cells and alters the intracellular milieu, resulting in increased progression of cervical carcinoma [9]. Further, several reports have highlighted that HIV proteins, including Tat, induce epithelialto-mesenchymal transition (EMT) and promote the development of epithelial malignancies in HPV-16-immortalized anal, cervical, and oral epithelial cells [10]. Inflammation is one of the key drivers of EMT, where pro-inflammatory cytokines activate EMT, inducing transcription factors [11]. Cellular uptake of HIV-Tat has been shown to activate transcription factors, such as NF-κB [12], leading to pro-inflammatory cytokines, such as TNF-α, IL-6, CCL2, IL-8, and IL-1β, among others [6,[13][14][15][16]. The role of HIV-1 Tat in inducing inflammation during neuroHIV and in the peripheral system has been widely studied, where Tat-induced NLRP3 inflammasome activation, releasing cytokines and chemokines in microglia [17,18] and enteric neuronal cells [19]. In contrast, less is known about HIV-Tat-mediated regulation of cervical inflammation and needs investigation. One of the important regulators of various cellular processes, including inflammation, is microRNAs. MicroRNAs (miRNAs/miRs) are a class of ~21-25 nucleotide-long, small non-coding RNA molecules that can downregulate gene expression by binding to a complementary sequence in the 3 ′ untranslated region (3 ′ UTR) of target mRNAs. During HIV immunopathogenesis, miRNAs play multifaceted roles by regulating key aspects of immunopathogenesis, including viral replication, immune activation, and inflammation [20][21][22]. Numerous reports have indeed shown HIV-associated miRNA dysregulation that regulates inflammatory responses [23][24][25][26]. In microglia, HIV-1 Tat protein altered the profiles of various miRNAs and contributed to an increased inflammatory environment [27,28].
In our previous study, we observed dysregulation of miR-204-5p in cytobrushderived cervical cells from HIV-infected females, accompanied by elevated levels of proinflammatory cytokines [29]. In a recent study, Kannan et al. also reported that miR-204-5p reduces HIV-Tat-mediated ferroptosis and the release of pro-inflammatory cytokines in microglial cells by targeting Acyl-CoA synthetase long-chain family member 4 (ASCL4) [30]. In other inflammatory conditions, miR-204-5p-mediated regulation of inflammation was facilitated by downregulating NF-κB [31,32]. Further, in various cancers, including gastric cancer [33], breast cancer [34], renal cell carcinoma [35], non-small cell lung cancer (NSCLC) [36], and glioma [37], miR-204 exerts regulatory roles. MiR-204-5p inhibited proliferation and invasion and induced apoptosis in cervical cancer cells by targeting different mediators, such as ATF2 and EphB2 [38][39][40]. Taken together, these findings underscore a probable role for miR-204 in regulating cellular functions in cervical cells, but whether it plays a role in the HIV-Tat-driven inflammatory response in these cells remains unknown. Given that during chronic HIV infection, Tat induces inflammation and could therefore be an important factor associated with the risk of cervical cancer, understanding the regulatory mechanism(s) associated with HIV-Tat-driven inflammation in cervical cells is warranted. In this study, we first examined the role of HIV-1 Tat in inducing inflammation in cervical epithelial cells under in vitro conditions, followed by an examination of the regulatory mechanisms involved. This study reveals that HIV-1 Tat-induced inflammation and oxidative stress are regulated by miRNA-204, which targets the NF-κB axis in HeLa cell-derived TZM-bl cells. https://doi.org/10.3390/cells15020117
## 2. Materials and Methods
## 2.1. Reagents
Antibodies were acquired from the Cell Signaling Technology (Danvers, MA, USA): NF-κB (1:1000 dilution); IL-1β (1:1000 dilution); IRAK1 (1:1000 dilution); IRAK4 (1:1000 dilution); Anti-Rabbit IgG (1:2000 dilution). Cell culture growth medium, Dulbecco's modified Eagle's medium (DMEM), was purchased from Gibco (Waltham, MA, USA). Recombinant HIV-1 Tat protein and cellular ROS assay kit DCFDA/H2DCFDA were obtained from Abcam (Cambridge, UK). miRCURY LNA RT kit; miRCURY LNA miRNA PCR Assays for U6 and miR-204-5p; miRCURY LNA SYBR Green PCR Kit; HiPerfect Transfection Reagent were purchased from Qiagen (Hilden, Germany). TaqMan™ MicroRNA Reverse Transcription Kit and TaqMan™ MicroRNA assays for miR-204-5p and RNU-44 were purchased from Applied Biosystems (Waltham, MA, USA). miRIDIAN miRNA-204-5p mimic/inhibitor, miRIDIAN miRNA mimic negative control were procured from Dharmacon (Lafayette, CO, USA). TRIzol reagent and magnetic bead-based multiplex assay (ProcartaPlex) kit were from Invitrogen (Waltham, MA, USA). PowerUP SYBR Green Master Mix was procured from Applied Biosystems (Waltham, MA, USA).
## 2.2. RNA Extraction from Human Cervical Epithelial Cell Line
TZM-bl, a HeLa cell-derived human cervical epithelial cell line, was used in this study (NIH AIDS Research and Reference Reagent Program). TZM-bl cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Waltham, MA, USA), 200 µg/mL L-glutamine, and 100 U/mL penicillin-streptomycin at 37 • C with 5% CO 2 .
For assay setup, 50,000 cells were stimulated with 50 ng recombinant HIV-1 Tat protein (Abcam, UK) for 24 h unless otherwise stated, followed by harvesting the cells and the supernatant. Supernatants were stored at -80 • C until further use in ELISA/bioplex assay. Cells were stored in TRIzol reagent (Invitrogen, Waltham, MA, USA), and RNA was isolated as per the manufacturer's instructions. The quality of the samples was evaluated by estimating the A260 nm/A230 nm and A260 nm/A280 nm ratios for phenol and protein contamination, respectively, using a NanoDrop (DeNovix, Wilmington, DE, USA USA). Further, 250 ng of RNA was used for cDNA synthesis (Takara, California, USA), followed by real-time PCR.
## 2.3. Real-Time qPCR (RT-qPCR) Analysis
Gene expression profiling of various cytokines (TNF-α, IFN-β, IL-1β, IL-6), transcription factors (NF-κB), and intermediate signaling molecules (IRAK1 and IRAK4) was carried out by RT-PCR using PowerUp SYBR Green Master Mix (Applied Biosystems TM , Waltham, MA, USA). The β-actin gene was used as a housekeeping gene to determine the relative expression of the gene of interest. PCR reactions were performed on an AB7500 Fast (Applied Biosystems TM ). Table 1 illustrates the primer sequences used in this study. For the expression levels of miRNAs, miRCURY LNA miRNA Custom PCR Panels (Qiagen) and TaqMan™ MicroRNA assays (Applied Biosystem) were used in RT-qPCR, and the data were normalized to the reference gene (U6 snRNA/RNU44). The fold change was calculated by the 2 -∆∆Ct method.
## Genes
Primer Sequence Reference
(#-primer designed from the NCBI database using the NCBI reference sequence).
$$β-actin F-5 ′ -TCGTCCACCGCAAATGCTTCTAG-3 ′ R-5 ′ -ACTGCTGTCACCTTCACCGTTCC-3 ′ [41] IL-1β Fq-5 ′ -ATGCACCTGTACGATCACTG-3 ′ Rq-5 ′ -ACAAAGGACATGGAGAACACC-3 ′ [42] IL-6 Fq-5 ′ ACCCCCAATAAATATAGGACTGGA-3 ′ Rq-5 ′ -GCTTCTCTTTCGTTCCCGGT-3 ′ [43] TNF-α Fq-5 ′ -CTGGGGCCTACAGCTTTGAT-3 ′ Rq-5 ′ -GGCTCCGTGTCTCAAGGAAG-3 ′ [44] IFN-β F-5 ′ -GGTTACCTCCGAAACTGAAGA-3 ′ R-5 ′ -CCTTTCATATGCAGTACATTAGCC-3 ′ [45] NF-κB Fq-5 ′ -TCTCCCTGGTCACCAAGGAC-3 ′ Rq-5 ′ -TCATAGAAGCCATCCCGGC-3 ′ [46] IRAK1 5 ′ -ACGGACACCTTCAGCTTTGG-3 ′ 5 ′ -TCCACCAGGTCTTTCAGATACTTG-3 ′ # (NM_001025242.2) IRAK4 F-5 ′ -TCATAGGCGGCAGGAACTTA-3 ′ R-5 ′ -ACCCAAACACTTCCCATCAG-3 ′ # (NM_001145257.2) TLR7 F-5 ′ -GTTACCAGGGCAGCCAGTTC-3 ′ R-5 ′ -ATGAGCCTCTGATGGGACAA-3 ′ [47]$$
## 2.4. Transfection Assays
TZM-bl cells were seeded into 48-well plates (50,000 cells per well) or six-well plates (1 × 10 6 cells per well) for gene expression analysis by RT-PCR and protein analysis by Western blotting, respectively. Cells were transiently transfected with 50 ng of miRNA-204-5p mimic/miR-204-5p inhibitor using HiPerfect Transfection Reagent (Qiagen, Germany) according to the manufacturer's instructions. Transfection with scrambled miRNA served as a control for the assay. Briefly, miRNA-204-5p mimic/inhibitor and a scrambled miRNA negative control (mock) were incubated with HiPerfect reagent in DMEM (without FBS) for 10 min. This complex was slowly added to the single-cell suspension and incubated in DMEM supplemented with 10% FBS for 16 h. Following transfection, cells were exposed to HIV-1 Tat (50 ng/mL) for another 24 h. Total RNA and protein were extracted for further investigations.
## 2.5. Western Blotting
A total of 1 × 10 6 TZM-bl cells were seeded in a six-well plate and harvested after 24 h of Tat stimulation. Briefly, the control, Tat-treated, and miR-204-mimic-transfected cells were lysed using 300 µL of Pierce TM RIPA buffer (Thermo, Waltham, MA, USA). Lysates were incubated at 4 • C for 30 min and then centrifuged at 12,000× g for 10 min at 4 • C. Bradford assay (Bio-Rad, Herculus, CA, USA) was used to determine the protein concentrations as per the manufacturer's instructions. A total of 100 µg of soluble proteins was resolved in 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by blotting onto a 0.2 µm Immun-Blot polyvinylidene fluoride membrane (Bio-Rad, USA). Then the membranes were blocked with 5% non-fat dry milk in 1× Tris-Buffered Saline with Tween-20 (TBS-T) buffer for 1 h at room temperature, washed three times with TBS-T, and incubated overnight with primary antibodies at 4 • C. After three washes, the membranes were incubated with a secondary antibody for 1 h at room temperature. Next, the protein signals were developed using Clarity Western ECL substrate (Bio-Rad, USA) and visualized on ChemiDoc XRS+ System (Bio-Rad). Each band intensity was normalized to the internal control, GAPDH, and the data were presented as relative fold change by using Image Lab software (version 3.0.1) analysis.
## 2.6. Pro-Inflammatory Markers' Analysis
Cytokine concentrations within the culture supernatants were quantified utilizing the magnetic bead-based ProcartaPlex™ Multiplex Immunoassay, Human Inflammation Panel I (Invitrogen, USA), in accordance with the manufacturer's protocol. Briefly, 50 µL of cell supernatants were incubated with 50× simplex beads in a 96-well plate for 2 h at room temperature on a shaker. Subsequently, after washing the beads, a detection antibody was added for 30 min at room temperature. Following the washing, the beads were incubated with Streptavidin-PE for another 30 min. The samples were resuspended in reading buffer and analyzed using the LABScan 100 TM xPONENT system. The concentration of cytokines was derived from a standard curve generated by the standard provided in the kit. Calculations were performed using a four-parameter logistic (4PL). For the concentrations that were below the detection limit, the lowest obtained value was considered.
## 2.7. Target Identification and Validation by 3 ′ UTR Luciferase Assay
miRNA-mRNA target prediction was carried out using miRWalk. RNA hybrid server (accessed on 20 September 2024; https://bibiserv.cebitec.uni-bielefeld.de/rnahybrid) was used to predict miRNA-mRNA interactions [48]. The NF-κB's 3 ′ UTR sequence was PCR-amplified with primers Fq-3 ′ CTAACTAGTCCTGCTGACAATTT-5 ′ and Rq-3 ′ TATCAAGCTTTAAGCAACCTCATC-5 ′ and cloned downstream of the luciferase gene in the pMIR-REPORT luciferase vector. HEK 293 cells were transfected with reporter plasmid (500 ng) and miR-204-5p or scrambled oligonucleotide (gifted by Dr. Rajesh Gacche, Savitribai Phule Pune University, Department of Biotechnology, Pune, India) using jetPRIME (Polyplus, S.A, Illkirch, France) transfection reagent. After 24 h, cells were lysed using the Bright-GloTM Luciferase Assay System (Promega, Madison, WI, USA), and luciferase activity was measured on a PerkinElmer Victor X2 Multilabel Microplate Reader (Waltham, MA, USA).
## 2.8. Reactive Oxygen Species Production Analysis by DCFDA/H2DCFDA
To assess ROS production, TZM-bl cells were resuspended in DCFDA/H2DCFDA solution (Abcam, Cambridge, UK) at a concentration of 20 µM and incubated for 30 min at 37 • C in the dark. After incubation, the cells were stimulated with HIV-1 Tat protein for 15 min, 30 min, 60 min, and 120 min. Fluorescence was measured by flow cytometry. To evaluate the effect of hsa-miR-204-5p, ROS levels were measured in unstimulated control, Tat-treated, and hsa-miR-204-5p mimic-transfected or mock cells after 2 h of Tat stimulation.
## 2.9. NF-kB Nuclear Translocation Analysis by Fluorescence Microscopy
The translocation of NF-kB in TZM-bl cells was analyzed using a fluorescence microscope. Glass Coverslips (diameter 18 mm) were cleaned with ethanol, followed by UV exposure for 15 min, and then placed into a 12-well cell culture plate. In total, 0.1 × 10 6 cells were seeded in a plate, grown overnight, and transiently transfected with either miRNA-204-5p mimic, miRNA-204-5p inhibitor, or scrambled miRNAs. They were then stimulated with 50 ng of HIV-1 Tat protein. Afterward, the cells were washed with PBS and fixed with fresh 4% paraformaldehyde solution for 20 min at room temperature. They were then washed three times with PBS and incubated for 15 min in blocking buffer containing 3% BSA and 0.1% Triton™ X-100 at room temperature. Subsequently, the cells were incubated with a primary monoclonal rabbit antibody against phospho-NF-kB (Cat. no. ab3033) for 1 h, diluted 1:1000 in antibody dilution buffer containing 3% BSA. The cells were washed three times with PBS and incubated for 30 min with secondary antibodies, goat anti-rabbit IgG CF@647 (2 µg), diluted in antibody dilution buffer (Biolegend, San Diego, CA, USA, Cat. no. 392111). After washing with PBS, they were incubated with DAPI (5 µg/mL; Invitrogen, Cat. No. D1306) for 5 min. Coverslips from a 12-well plate were carefully removed with the help of forceps and mounted on a glass slide with ProLong™ RapidSet™ Antifade Mountant (Thermo Fisher, Cat. no. P38931), and visualized using a fluorescence microscope (Zeiss Cell Discoverer 7.0, Carl Zeiss Microscopy GmbH, Jena, Germany, magnification 40).
## 2.10. Statistical Analysis
All the experiments were performed in triplicate and repeated three to four times. The data were expressed as mean ± SEM. An unpaired Student's t-test was used to derive the statistical difference between two conditions/groups, while one-way ANOVA followed by post hoc Dunnett's test was used to compare the difference between more than two experimental conditions. Statistical significance was determined using Graph-Pad Prism version 8.4.2 (San Diego, CA, USA). A p-value of ≤0.05 was considered to be statistically significant.
## 3. Results
## 3.1. HIV-1 Tat Primes the Inflammatory Responses in Cervical Epithelial Cells
The HIV-Tat protein has been shown to induce pro-inflammatory factors in various cells, including monocytes, microglia, and enteric neurons [16][17][18][19]. However, its role in driving inflammatory processes in the cervical cells is largely undefined. Further, HIV-Tat has been shown to promote EMT in cervical cells, resulting in cancer progression. Since cervical inflammation facilitates EMT, here we examined the role of HIV-Tat in inducing inflammation in the cervical epithelial cells using in vitro assays. TZM-bl cervical epithelial cells were first stimulated with 50 ng/mL HIV-1 Tat for 6-48 h, followed by gene expression profiling of inflammatory mediators. A significant increase in mRNA expression of TNF-α (p = 0.02), NF-κB (p = 0.04), and IRF7 (p = 0.01) was noted at 24 h in the Tat-stimulated cells compared to the non-stimulated cells (Figure 1A-C). In subsequent assays, the 24 h time point post Tat stimulation was therefore used.
Furthermore, the inflammatory profiles of various mediators were determined using RT-PCR and a magnetic bead-based multiplex immunoassay. Tat-stimulation in TZM-bl cells increased the expression of TNF-α (p = 0.0007) and IFN-β (p = 0.015) gene, while no difference was noted in IL-1β and IL-6 gene expression between the Tat-stimulated cells and that of non-stimulated cells (Figure 1D). Secreted cytokine profile showed increased levels of IL-1β (p = 0.03), TNF-α (p = 0.04), IL-6 (p = 0.01), IL-17a (p = 0.01), MIP-1α (p = 0.01), MIP-1β (p = 0.01), and GM-CSF (p = 0.04) post Tat stimulation as compared to the unstimulated cells (Figure 1E). Other inflammatory markers, such as intracellular adhesion molecule (ICAM)-1 (p = 0.02), P-Selectin (p = 0.04), and E-Selectin (p = 0.04), were also significantly upregulated upon Tat stimulation (Figure 1F). Since one of the direct consequences of the inflammatory process is the production of reactive oxygen species (ROS), and Tat-mediated induction of ROS has previously been reported in microglial cells [48], contributing to neuroinflammation, we determined whether Tat induces ROS production in cervical cells using an in vitro DCFDA assay. A significant increase in ROS levels was observed gradually in Tat-stimulated cells from 30 min onwards (p < 0.05; Figure 1G). These in vitro findings collectively suggest that HIV-1 Tat primes the release of inflammatory mediators in the cervical epithelial cells.
## 3.2. Tat-Mediated Inflammatory Responses Involved TLR7/NF-κB Signaling in Cervical Cells
TLRs are known to stimulate expression of pro-inflammatory cytokines and induce inflammatory responses during the majority of infections [49,50]. Moreover, to explore the mechanism of Tat-mediated cervical inflammation, we examined the involvement of TLRs, intermediate proteins (IRAKs), and transcription factor (NF-κB) by RT-PCR and Western blotting. It was observed that among the various TLRs examined, TZM-bl cells exhibited the increased expression of the TLR7 gene to >8-fold post-Tat-stimulation (p = 0.01) than those of unstimulated cells (Figure 2A). Furthermore, increased expressions of NF-κB (p = 0.02), interleukin-1 receptor-associated kinase IRAK1 (p = 0.07), and IRAK4 (p = 0.04) were also observed in Tat-stimulated cells (Figure 2B) than in the unstimulated cells. In addition to unstimulated cells, heat-inactivated HIV-Tat was also used as a negative control. Indeed, TZM-bl cells did not elicit a measurable response to heat-inactivated Tat (Figure S1).
Protein expression analysis using Western blotting showed that HIV-1 Tat significantly increased the expression of IL-1β (p < 0.001) and NF-κB (p = 0.004), no significant change was observed in the expression of IRAK1 (p > 0.05) in the TZM-bl cells compared to the unstimulated controls (Figure 2C), suggesting the plausible role of TLR7/NF-κB signaling during Tat-mediated inflammatory responses in these cells.
## 3.3. miR-204-5p Regulates Tat-Mediated Inflammation by Targeting NF-κB in the Cervical Cells
Since we noted that HIV-Tat induces inflammation in cervical cells, we next sought to examine the regulatory factors involved. Previous reports showed that miR-204-5p is a negative regulator of inflammation [32,49]. In the microglial cells [30], hsa-miR-204-5p has previously been shown to regulate Tat-induced inflammatory processes. Further, its downregulation has been associated with metastasis of various cancers, including cervical cancer [38,50,51]. Hence, we first determined whether Tat stimulation alters miR-204-5p expression in vitro in cervical epithelial cells. A significant downregulated expression of miR-204-5p was noted in the TZM-bl cells post Tat stimulation (p = 0.005) than in the non-stimulated cells (Figure 3A).
To further investigate the potential role of miR-204-5p in cervical cells, TZM-bl cells were transfected with miR-204-5p mimic, while a scrambled miR-mimic was used as a mock. Transfection conditions were confirmed in independent experiments using concentrations of 50 nm miR-204-5p and scrambled miR-mimic (mock) for 24 h. Cells transfected with 50 nM miR-204-5p mimic significantly increased the expression of miR-204-5p at 24 h (~53-fold; p < 0.0001) than mock-transfected cells (Figure S2). Role of miR-204-5p during HIV-Tat-induced cervical inflammation was assessed by overexpressing miR-204-5p in the TZM-bl cells, followed by quantifying the expression profile of various cytokines post Tat-stimulation using RT-PCR, magnetic bead-based multiplex assay, or Western blotting. It was found that TZM-bl cells transfected with miR-204-5p mimic had a significantly reduced expression of TNF-α (p = 0.002) and IFN-β (p = 0.0009) genes post Tat stimulation as compared to the cells transfected with scrambled miR-mimic/mock (Figure 3B). It is noteworthy that TZM-bl cells transfected with miR-204-5p mimic alone did not increase the inflammatory mediators (TNF-α, IFN-β, IL-1β, NF-κB, IRAKs) as compared to the mock-transfected cells (p > 0.05; Figures 3 and4). In the multiplex cytokine assay, inflammatory markers-TNF-α (p = 0.056), IL-6 (p = 0.01), MIP-1α (p = 0.03), MIP-1α (p = 0.002) ICAM-1 (p = 0.04), and P-Selectin (p = 0.01)-were also downregulated in the miR-204-5p mimic-transfected cells upon Tat stimulation (Figure 3C,D) than the mock cells. Moreover, overexpression of miR-204-5p reduced Tat-induced ROS production (p = 0.03) compared with mock cells (Figure 3E). These findings collectively envisage that miR-204-5p negatively regulates HIV-1 Tat-induced inflammatory responses in the TZM-bl cells. Given that miRNAs exert their activity by targeting specific proteins in the signaling cascades, we further aimed to understand the mechanism of miR-204-5p-mediated regulation of inflammation in the cervical cells post HIV-Tat stimulation. Both IRAK-4 and NF-κB are intermediate molecules involved in TLR-mediated inflammatory cascades; hence, we examined their expression profiles. Since in our previous assay (Figure 2B), we observed upregulation of these mediators post-Tat-stimulation, we next sought to determine whether miR-204-5p exerts any regulatory role on these intermediate proteins using loss-and gainof-function approaches. Interestingly, it was observed that miR-204-5p overexpression https://doi.org/10.3390/cells15020117
Cells 2026, 15, 117 significantly decreased the expression of IRAK1 (p < 0.0001), IRAK4 (p = 0.05), and NF-κB (p < 0.0001) genes in the TZM-bl cells post Tat stimulation (Figure 4A). Using Western blotting, it was further noted that overexpression of miR-204-5p in these cells decreased IL-1β (p = 0.001) and NF-κB (p = 0.009) expression in response to Tat stimulation than that of mock cells (Figure 4B), while IRAK1 did not show any significant difference (p > 0.05). We also checked the effect of miR-204-5p inhibitor on the expression of cytokines and intermediate molecules in Tat-stimulated TZM-bl cells. An increase in protein expression levels of NF-kB and IRAK-1 was observed in miR-204-5p inhibitor-transfected and Tat-stimulated TZM-bl cells. Although the expression levels were not significant (Figure S3). To reveal the role of TLR7 in Tat-mediated inflammatory signaling in TZM-bl cervical cells, we treated the cells with the TLR7 inhibitor M5049 (25 nM). Interestingly, a significant reduction in the expression of IFN-β (p = 0.0007), NF-κB (p = 0.001), IRAK1 (p = 0.003), and IRAK4 (p = 0.0001) was observed (Figure S4). No significant change in TNF-α expression was observed.
Using in silico tools-bioinformatics analysis with miRWalk and RNA hybrid serverwe identified NF-κB as a putative target of miR-204-5p and calculated the binding energy (-25.9 kcal/mol) of miRNA-mRNA, respectively (Figure 4C). To confirm NF-κB as a direct target of miR-204-5p, we used a 3 ′ UTR luciferase reporter system. Specifically, the 3 ′ UTR of NF-κB was cloned into the reporter plasmid downstream of the luciferase gene in the pMIR-REPORT vector, then co-transfected with miR-204-5p mimic. It was found that NF-κB is indeed a direct target of miR-204-5p, as evidenced by a significant decrease in luciferase activity (p = 0.003, 24 h) following transfection with the miR-204-5p mimic (Figure 4D). Overall, these findings suggest a potent role of miR-204-5p in regulating cervical inflammation mediated through the TLR/NF-κB axis in the TZM-bl cells.
To provide functional evidence of NF-κB activation in Tat-stimulated TZM-bl cells, we assessed the nuclear translocation of phospho-NF-κB using a fluorescence microscope. The results indicated that Tat stimulation increased nuclear localization of phospho-NF-κB compared with non-stimulated cells. The reduction in phospho-NF-κB translocation was observed when cells were transfected with miR-204-5p mimic prior to Tat stimulation. Further, upon transfection with miR-204-5p inhibitor, an increased nuclear translocation of phospho-NF-κB was observed in Tat-stimulated TZM-bl cells compared to the cells transfected with scrambled miRNA (mock). These results provide direct evidence of NF-κB activity being altered in a miR-204-dependent manner in TZM-bl cells upon Tat stimulation (Figure 5). Nuclei were stained with DAPI (blue). Phospho-NF-kB was stained with CF@647 (red).
## 4. Discussion
Among females, persistent cervical inflammation is a critical factor for increased viral shedding, thereby increasing the risk of HIV transmission. However, unlike peripheral and neuro-inflammatory responses, little is known about HIV mediated regulation of cervical inflammation. HIV proteins, particularly Tat and gp120, have been shown to induce chronic inflammation in microglia, monocytes, and epithelial cells, among others [17,[52][53][54]. Although the cervicovaginal mucosa is an immunologically distinct site, HIV-Tat likely regulates inflammatory responses in cervical cells, as examined here using TZM-bl cells derived from HeLa cells under in vitro conditions.
In line with previous findings, our data showed that HIV-Tat drives inflammatory responses in cervical cells, resulting in the release of potent inflammatory cytokines (IL-1β, TNF-α, IL-6, IL-17a, and GM-CSF) [2,16,55]. We also observed that HIV-Tat upregulated the expression of chemotactic cytokines, including MIP-1α and MIP-1β, in TZM-bl cells. Elevated levels of these chemokines (MIP-1α and MIP-1β) have been detected in the genital secretions of HIV-infected women and have been linked to an increased risk of HIV acquisition [56]. GM-CSF has previously been shown to exert pro-inflammatory roles in many conditions [57][58][59]. During HIV infection, alveolar macrophages have been shown to increase the levels of GM-CSF, leading to pathogenic AIDS-associated interstitial lung disease [57]. Additionally, HIV proteins, including Tat, gp120, Nef, and Vpr, altered mitochondrial function and generated reactive oxygen species in brain tissues, leading to neuroinflammation [60][61][62]. In line with this, we also observed increased ROS levels in cervical epithelial cells post HIV-Tat stimulation, which might exacerbate inflammatory responses in these cells, together with other pro-inflammatory cytokines and chemokines. We also observed that Tat-stimulated cells increased the expression of adhesion molecules, including ICAM-1, P-Selectin, and E-Selectin. Tat-induced upregulation of these adhesion molecules has previously been reported in brain tissues, including microglia and endothelial cells [63,64]. ICAM and the Selectins are important regulators of cell-cell interaction and known to be involved in metastasis and angiogenesis [63,64]. ICAM-1 has been shown to promote inflammatory responses during HIV infection, leading to accelerated disease progression [65]. Taken together, these findings underscore the potent role of Tat-mediated inflammation across different cell types.
Inflammatory responses are largely initiated in response to TLR activation [66,67]. To understand the mechanism of Tat-induced inflammation in TZM-bl cells, we examined the expression of various TLRs and the potential intermediates involved. Among other TLRs, we observed increased expression of TLR7, along with IRAK4 and NF-κB, in TZM-bl cells post Tat-stimulation. TLR-mediated immune dysregulation has been reported in T cells, pDCs, and monocytes [68,69] during HIV immunopathogenesis, and blocking TLR7 has been shown to reduce immune activation [70,71]. In a murine model, HIV-1 ssRNA has been shown to elevate pro-inflammatory cytokines through TLR7 activation [72]. HIV proteins have also been shown to trigger cytokine production through distinct TLR signaling pathways in various cell types. HIV-1 structural proteins (p17, p24, and gp41) increased IL-8 by activating TLR2 [73] and TLR10 [74] and involving the NF-κB axis in primary T cells, as well as TZM-bl cells. Gp120 stimulated TLR-2 and TLR-4 mediated induction of TNF-α and IL-8 via activation of NF-κB in the primary genital epithelial cells [54]. Unlike other HIV proteins, less is known about the role of the Tat protein in modulating TLR-mediated responses. A single study shows that Tat directly binds to TLR4 and its co-receptor MD-2, leading to the induction of pro-inflammatory cytokines such as TNF-α and IL-10 [75], contributing to immune dysregulation in immune cells. TLR signaling pathways, including TLR7, culminate in the activation of NF-κB, a transcription factor crucial for inducing inflammation and other immune responses and regulating HIV disease progression [76][77][78].
Evidence indicates that Tat modulates key enzymes in NF-κB signaling. HIV-Tat interacts with IκB-α and p65, thereby increasing NF-κB transcriptional activity in HeLa cells. Tat has also been shown to inhibit SIRT-1, a negative regulator of T cell activation, leading to hyperimmune activation [79,80]. In corroboration, we also noted upregulation of NF-κB in response to HIV-Tat in the TZM-bl cells. Further blocking of NF-κB using BAY 11-7082 (40 µM) significantly decreased the expression of pro-inflammatory cytokines (TNF-α, IL-1β, and IFN-β) (Figure S5). These findings suggest that NF-κB is a crucial factor driving inflammatory responses in cervical cells. Besides NF-κB, Tat stimulation also increased the expression of IRAKs, a crucial component of TLR-mediated NF-κB activation and also reported during HIV immunopathogenesis. Although we did not observe a significant change in the expression of IRAK1 in Tat-stimulated TZM-bl cells. In a previous study carried out on HIV infected microglial cells it was observed that IRAK blocking led to induce a pro-inflammatory milieu, highlighting the role of TLR-7 and TLR-8 [81]. We also observed that, upon inhibiting the TLR7 by its inhibitor (M5049), the expression of intermediate molecules NF-κB, IRAK-1/-4 and cytokine IFN-β was significantly reduced. Collectively, these findings underscore that TLR7-mediated inflammatory responses might be a critical factor during cervical inflammation; however, the present study could not establish a direct link between TLR7 to Tat-stimulated inflammation, which needs detailed investigations.
Cervical inflammation not only speeds up the disease progression in HIV infected women, decreases the efficacy of the local antiretrovirals, but can also influence EMT, leading to cancer progression, particularly in HIV-HPV co-infected women. Hence, understanding the factors involved in regulating inflammation is warranted. miRNAs are one such regulator, known to mitigate the inflammatory processes. The role of miRNA-204-5p has been reported in various inflammation-associated conditions, including neuroinflammation, corneal inflammation, and synovial inflammation, among others [32,49,82,83]. In microglial cells, miR-204-5p has been shown to suppress Tat-mediated neuroinflammation through ferroptosis [30,32]. In line with previous reports, we also observed downregulation of miR-204 in TZM-bl cells, along with increased levels of inflammatory cytokines, in response to Tat stimulation. Using the gain-of-function approach-i.e., transfecting the cells with a specific mimic-we further showed that overexpression of miR-204-5p indeed suppressed the pro-inflammatory mediators, such as TNF-α, IL-6, MIP-1α, MIP-1β, ICAM-1, and P-Selectin, as well as ROS. miR-204-5p has been previously shown to regulate oxidative stress associated with neuroinflammatory changes and depression phenotypes in the murine model by targeting via the NF-κB axis RGS12 [32]. miR-204-5p also regulates myopia development by targeting TXNIP [84]. MiRNAs exert their regulatory effect by targeting specific proteins. Several reports show that miR-204-5p suppresses NF-κB activity [31,32]. We noted an inverse profile of miR-204 and inflammatory mediators-i.e., cytokines and NF-κB-upon Tat stimulation, indicating their probable association. We confirmed these observations by overexpressing miR-204-5p in TZM-bl cells, which resulted in reduced NF-κB and inflammatory cytokine expression after Tat stimulation. Using miRWalk, an in silico tool, we predicted that the putative target site of miRNA-204-5p was located within the 3 ′ UTR of NF-κB, which we further validated using a luciferase assay, underscoring that miR-204-5p targets the NF-κB axis in the regulation of inflammatory responses.
A dynamic interplay reportedly occurs between NF-κB signaling and reactive oxygen species, and ROS-mediated HIV LTR activation through NF-κB has been reported in Jurkat cells [85]. HIV-Tat has been shown to potentiate NF-κB-induced inflammatory responses by altering cellular redox state in HeLa cells [86,87]. We also observed Tat-induced ROS and NF-κB upregulation in TZM-bl cells, which was downregulated in miR-204-5p-transfected cells. Together, these findings reveal that oxidative stress production and NF-κB activation are interlinked in Tat-mediated cervical inflammatory responses. However, their regulatory roles need to be explored further. Previous studies revealed that Tat activates NF-κB in HEK, HL3T1, and Jurkat cells, among others [12,88]. In line with these studies, we also observed NF-κB activation in Tat-stimulated TZM-bl cells. Nuclear translocation of phospho-NF-κB was increased in Tat-stimulated TZM-bl cells. It was also noted that miR-204-5p mimic reduced the nuclear localization of phospho-NF-κB in these Tat-stimulated cells. In contrast, miR-204-5p inhibitor increased phospho-NF-κB in TZM-bl cells upon Tat stimulation. These results indicate that NF-κB activation in Tat-stimulated TZM-bl cervical cells is miR-204-5p dependent manner.
Our study provides mechanistic insights into HIV-1 Tat-induced inflammatory processes in the cervical cells. In previous studies, we [89] and others [2,[90][91][92] have reported elevated cervical inflammation in cytobrush-derived cervical cells and also in cervicovaginal lavage samples of HIV-infected women. However, the regulatory mechanisms driving inflammation in cervical cells remain poorly understood. A major challenge in studying these mechanisms is the limited availability of cervical cells and the unavailability of suitable animal models. To overcome these limitations and gain detailed molecular insights, TZM-bl cells have been widely used as an in vitro model for cervical studies [93,94], which was also used in the current exploratory study. However, our findings are based solely on in vitro assays using TZM-bl cells, which is one of the study's limitations. TZM-bl cells are transformed cancer-derived epithelial cells that may not fully represent the physiological state of normal cervical epithelial cells. Further, these cells lack the cellular diversity of the cervicovaginal mucosa; hence, the findings presented here need to be confirmed using a primary cervical cell line, cervical explant culture and/or in vivo models. To the best of our knowledge, this exploratory study is the first to provide evidence that HIV-1 Tat increases inflammation in cervical epithelial cells (TZM-bl) via the miR-204-5p/NF-κB axis (Figure 6). The mechanism of the upstream of NF-κB cascading and TLR7 activation by HIV-1 Tat protein needs to be explored further. Our findings offer functional insights into the interaction between HIV-1 Tat and the miR-204-5p/NF-κB signaling axis, underscoring its possible relevance to cervical inflammation and mucosal immunity, which needs further investigation.
## References
1. Douek, Roederer, Koup (2009) "Emerging Concepts in the Immunopathogenesis of AIDS" *Annu. Rev. Med*
2. Arnold, Burgener, Birse et al. (2016) "Increased Levels of Inflammatory Cytokines in the Female Reproductive Tract Are Associated with Altered Expression of Proteases, Mucosal Barrier Proteins, and an Influx of HIV-Susceptible Target Cells" *Mucosal Immunol*
3. De Souza Rios, Mapekula, Mdletshe et al. "HIV-1 Transactivator of Transcription (Tat) Co-Operates with AP-1 Factors to Enhance c-MYC Transcription" *Front. Cell Dev. Biol*
4. Cafaro, Schietroma, Sernicola et al. (1704) "Role of HIV-1 Tat Protein Interactions with Host Receptors in HIV Infection and Pathogenesis" *Int. J. Mol. Sci*
5. Dayton (1986) "The Trans-Activator Gene of the Human T Cell Lymphotropic Virus Type III Is Required for Replication" *Cell*
6. Li, Yim, Lau (2010) "Role of HIV-1 Tat in AIDS Pathogenesis: Its Effects on Cytokine Dysregulation and Contributions to the Pathogenesis of Opportunistic Infection" *AIDS*
7. Debaisieux, Rayne, Yezid et al. (2012) "The Ins and Outs of HIV-1 TAT" *Traffic*
8. Clark, Nava, Caputi (2017) "Tat Is a Multifunctional Viral Protein That Modulates Cellular Gene Expression and Functions" *Oncotarget*
9. Barillari, Palladino, Bacigalupo et al. (2016) "Entrance of the Tat Protein of HIV-1 into Human Uterine Cervical Carcinoma Cells Causes Upregulation of HPV-E6 Expression and a Decrease in P53 Protein Levels" *Oncol. Lett*
10. Lien, Mayer, Herrera et al. "HIV-1 Proteins Gp120 and Tat Promote Epithelial-Mesenchymal Transition and Invasiveness of HPV-Positive and HPV-Negative Neoplastic Genital and Oral Epithelial Cells"
11. Islam, Morshed, Babu et al. (2022) "The Role of Inflammations and EMT in Carcinogenesis" *Adv. Cancer Biol.-Metastasis*
12. Li, Liu, Fujinaga et al. (2024) "Enhanced NF-κB Activation via HIV-1 Tat-TRAF6 Cross-Talk" *Sci. Adv*
13. Bennasser, Badou, Tkaczuk et al. (2002) "Signaling Pathways Triggered by HIV-1 Tat in Human Monocytes to Induce TNF-α" *Virology*
14. Eugenin, Dyer, Calderon et al. (2005) "HIV-1 Tat Protein Induces a Migratory Phenotype in Human Fetal Microglia by a CCL2 (MCP-1)-dependent Mechanism: Possible Role in NeuroAIDS" *Glia*
15. Ott, Lovett, Mueller et al. (1998) "Superinduction of IL-8 in T Cells by HIV-1 Tat Protein Is Mediated Through NF-κB Factors" *J. Immunol*
16. Yang, Wu, Lu (2010) "Mechanism of HIV-1-TAT Induction of Interleukin-1beta from Human Monocytes: Involvement of the Phospholipase C/Protein Kinase C Signaling Cascade" *J. Med. Virol*
17. Chivero, Guo, Periyasamy et al. (2017) "HIV-1 Tat Primes and Activates Microglial NLRP3 Inflammasome-Mediated Neuroinflammation" *J. Neurosci*
18. Sheng, Hu, Hegg et al. (2000) "Activation of Human Microglial Cells by HIV-1 Gp41 and Tat Proteins" *Clin. Immunol*
19. Ngwainmbi, De, Smith et al. (2014) "Effects of HIV-1 Tat on Enteric Neuropathogenesis" *J. Neurosci*
20. Trobaugh, Klimstra (2017) "MicroRNA Regulation of RNA Virus Replication and Pathogenesis" *Trends Mol. Med*
21. Chinniah, Adimulam, Nandlal et al. (2022) "The Effect of miRNA Gene Regulation on HIV Disease" *Front. Genet*
22. Das, Rao (2022) "The Role of microRNAs in Inflammation" *Int. J. Mol. Sci*
23. Cuesta-Sancho, Márquez-Ruiz, Illanes-Álvarez et al. "Expression Profile of microRNAs Related with Viral Infectivity, Inflammatory Response, and Immune Activation in People Living with HIV"
24. Morando, Rosenzvit, Pando et al. "The Role of MicroRNAs in HIV Infection" *Genes*
25. Fulcher, Koukos, Koutsioumpa et al. (2017) "Unique microRNA Expression in the Colonic Mucosa during Chronic HIV-1 Infection" *AIDS*
26. Ballegaard, Ralfkiaer, Pedersen et al. (2017) "MicroRNA-210, MicroRNA-331, and MicroRNA-7 Are Differentially Regulated in Treated HIV-1-Infected Individuals and Are Associated with Markers of Systemic Inflammation" *J. Acquir. Immune Defic. Syndr*
27. Sardo, Vakil, Elbezanti et al. (2016) "The Inhibition of microRNAs by HIV-1 Tat Suppresses Beta Catenin Activity in Astrocytes" *Retrovirology*
28. Periyasamy, Thangaraj, Bendi et al. (2019) "HIV-1 Tat-Mediated Microglial Inflammation Involves a Novel miRNA-34a-NLRC5-NFκB Signaling Axis" *Brain Behav. Immun*
29. Akolkar, Sonar, Rao et al. (2025) "Unraveling Cervical Inflammation in HIV-Infected Women: The Regulatory Role of miR-204-5p and miR-3691-3p" *Life Sci*
30. Kannan, Sil, Oladapo et al. (2023) "HIV-1 Tat-Mediated Microglial Ferroptosis Involves the miR-204-ACSL4 Signaling Axis" *Redox Biol*
31. Wa, Huang, Pan et al. (2019) "miR-204-5p Represses Bone Metastasis via Inactivating NF-κB Signaling in Prostate Cancer" *Mol. Ther.-Nucleic Acids*
32. Lan, Li, Fan et al. (2021) "MicroRNA-204-5p Reduction in Rat Hippocampus Contributes to Stress-Induced Pathology via Targeting RGS12 Signaling Pathway" *J. Neuroinflamm*
33. Zhang, Yin, Hu et al. (2015) "MicroRNA-204-5p Inhibits Gastric Cancer Cell Proliferation by Downregulating USP47 and RAB22A" *Med. Oncol*
34. Li, Jin, Zhang et al. (2014) "Decreased Expression of miR-204 Is Associated with Poor Prognosis in Patients with Breast Cancer" *Int. J. Clin. Exp. Pathol*
35. Munari, Marchionni, Chitre et al. (2014) *Clear Cell Papillary Renal Cell Carcinoma: Micro-RNA Expression Profiling and Comparison with Clear Cell Renal Cell Carcinoma and Papillary Renal Cell Carcinoma. Hum. Pathol*
36. Shi, Zhang, Sun et al. (2014) "MiR-204 Inhibits Human NSCLC Metastasis through Suppression of NUAK1" *Br. J. Cancer*
37. Ying, Li, Wu et al. (2013) "Loss of miR-204 Expression Enhances Glioma Migration and Stem Cell-like Phenotype" *Cancer Res*
38. Shu, Zhang, Cai (2017) "MicroRNA-204 Inhibits Cell Migration and Invasion in Human Cervical Cancer by Regulating Transcription Factor 12" *Oncol. Lett*
39. Duan, Wu, Chen et al. (2018) "Q. miR-204 Regulates Cell Proliferation and Invasion by Targeting EphB2 in Human Cervical Cancer" *Oncol. Res*
40. Li, Guo, Liu et al. (2019) "Molecular Mechanism of miR-204 Regulates Proliferation, Apoptosis and Autophagy of Cervical Cancer Cells by Targeting ATF2" *Artif. Cells Nanomed. Biotechnol*
41. Cai, Zhou, Huang et al. (2012) "Reduced Expression of Krüppel-like Factor 17 Is Related to Tumor Growth and Poor Prognosis in Lung Adenocarcinoma" *Biochem. Biophys. Res. Commun*
42. Ali, Dasari, Van Keulen et al. (2017) "Canonical Stimulation of the NLRP3 Inflammasome by Fungal Antigens Links Innate and Adaptive B-Lymphocyte Responses by Modulating IL-1β and IgM Production" *Front. Immunol*
43. Ecker, Ledur, Da Silva et al. (2017) "Chalcogenozidovudine Derivatives with Antitumor Activity: Comparative Toxicities in Cultured Human Mononuclear Cells" *Toxicol. Sci*
44. Zong, Kimura, Kinoshita et al. (2019) "Exposure to 1,2-Dichloropropane Upregulates the Expression of Activation-Induced Cytidine Deaminase (AID) in Human Cholangiocytes Co-Cultured with Macrophages" *Toxicol. Sci*
45. Vasilishina, Kropotov, Spivak et al. (2019) "Relative Human Telomere Length Quantification by Real-Time PCR"
46. Eiró, González, González et al. (2012) "Relationship between the Inflammatory Molecular Profile of Breast Carcinomas and Distant Metastasis Development" *PLoS ONE*
47. Zheng, Zhang, Yu et al. (2010) "Expression of Toll-like Receptors 7, 8, and 9 in Primary Sjögren's Syndrome. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endodontol"
48. Rehmsmeier, Steffen, Höchsmann et al. (2004) "Fast and Effective Prediction of microRNA/Target Duplexes" *RNA*
49. Wu, Zhang, Mo et al. (2022) "Identification of Novel Rheumatoid Arthritis-Associated MiRNA-204-5p from Plasma Exosomes"
50. Zhang, Wang, Chen (2013) "MiR-204 down Regulates SIRT1 and Reverts SIRT1-Induced Epithelial-Mesenchymal Transition, Anoikis Resistance and Invasion in Gastric Cancer Cells" *BMC Cancer*
51. Liu, Long, Du et al. (2016) "miR-204 Regulates the EMT by Targeting Snai1 to Suppress the Invasion and Migration of Gastric Cancer" *Tumor Biol*
52. Ben Haij, Planès, Leghmari et al. (2015) "HIV-1 Tat Protein Induces Production of Proinflammatory Cytokines by Human Dendritic Cells and Monocytes/Macrophages through Engagement of TLR4-MD2-CD14 Complex and Activation of NF-κB Pathway" *PLoS ONE*
53. Levast, Barblu, Coutu et al. (2017) "HIV-1 Gp120 Envelope Glycoprotein Determinants for Cytokine Burst in Human Monocytes" *PLoS ONE*
54. Nazli, Kafka, Ferreira et al. (2013) "HIV-1 Gp120 Induces TLR2-and TLR4-Mediated Innate Immune Activation in Human Female Genital Epithelium" *J. Immunol*
55. Zhou, Liu, Gao et al. (2019) "HIV-1 Tat Enhances Purinergic P2Y4 Receptor Signaling to Mediate Inflammatory Cytokine Production and Neuronal Damage via PI3K/Akt and ERK MAPK Pathways" *J. Neuroinflamm*
56. Masson, Passmore, Liebenberg et al. (2015) "Genital Inflammation and the Risk of HIV Acquisition in Women" *Clin. Infect. Dis*
57. Agostini, Trentin, Zambello et al. (1992) "Release of Granulocyte-Macrophage Colony-Stimulating Factor by Alveolar Macrophages in the Lung of HIV-1-Infected Patients. A Mechanism Accounting for Macrophage and Neutrophil Accumulation" *J. Immunol*
58. Petrina, Martin, Basta "Granulocyte Macrophage Colony-Stimulating Factor Has Come of Age: From a Vaccine Adjuvant to Antiviral Immunotherapy" *Cytokine Growth Factor Rev. 2021*
59. Shiomi, Usui, Mimori (2016) "GM-CSF as a Therapeutic Target in Autoimmune Diseases" *Inflamm. Regener*
60. Price, Ercal, Nakaoke et al. (2005) "HIV-1 Viral Proteins Gp120 and Tat Induce Oxidative Stress in Brain Endothelial Cells" *Brain Res*
61. Banerjee, Zhang, Manda et al. (2010) "HIV Proteins (Gp120 and Tat) and Methamphetamine in Oxidative Stress-Induced Damage in the Brain" *Potential Role of the Thiol Antioxidant N-Acetylcysteine Amide. Free. Radic. Biol. Med*
62. Ivanov, Valuev-Elliston, Ivanova et al. (2016) "Oxidative Stress during HIV Infection: Mechanisms and Consequences" *Oxid. Med. Cell. Longev*
63. Pu, Tian, Flora et al. (2003) "HIV-1 Tat Protein Upregulates Inflammatory Mediators and Induces Monocyte Invasion into the Brain" *Mol. Cell. Neurosci*
64. Cota-Gomez, Flores, Cruz et al. (2002) "The Human Immunodeficiency Virus-1 Tat Protein Activates Human Umbilical Vein Endothelial Cell E-Selectin Expression via an NF-κB-Dependent Mechanism" *J. Biol. Chem*
65. Yu, Shang, Jiang (2020) "ICAM-1 in HIV Infection and Underlying Mechanisms" *Cytokine*
66. Vijay (2018) "Toll-like Receptors in Immunity and Inflammatory Diseases: Past, Present, and Future" *Int. Immunopharmacol*
67. Duan, Du, Xing et al. (2022) *Toll-Like Receptor Signaling and Its Role in Cell-Mediated Immunity. Front. Immunol*
68. Funderburg, Luciano, Jiang et al. (1915) "Toll-like Receptor Ligands Induce Human T Cell Activation and Death, a Model for HIV Pathogenesis" *PLoS ONE*
69. Meier, Alter, Frahm et al. (2007) "MyD88-Dependent Immune Activation Mediated by Human Immunodeficiency Virus Type 1-Encoded Toll-like Receptor Ligands" *J. Virol*
70. Rappe, Finsterbusch, Crotta et al. (2021) "A TLR7 Antagonist Restricts Interferon-Dependent and -Independent Immunopathology in a Mouse Model of Severe Influenza" *J. Exp. Med*
71. Barrat, Meeker, Chan et al. (2007) "Treatment of Lupus-prone Mice with a Dual Inhibitor of TLR7 and TLR9 Leads to Reduction of Autoantibody Production and Amelioration of Disease Symptoms" *Eur. J. Immunol*
72. Baenziger, Heikenwalder, Johansen et al. (2009) "Triggering TLR7 in Mice Induces Immune Activation and Lymphoid System Disruption, Resembling HIV-Mediated Pathology" *Blood*
73. Henrick, Yao, Rosenthal (2015) "the INFANT study team. HIV-1 Structural Proteins Serve as PAMPs for TLR2 Heterodimers Significantly Increasing Infection and Innate Immune Activation" *Front. Immunol*
74. Henrick, Yao, Zahoor et al. (2019) "TLR10 Senses HIV-1 Proteins and Significantly Enhances HIV-1 Infection" *Front. Immunol*
75. Ben Haij, Leghmari, Planès et al. (2013) "HIV-1 Tat Protein Binds to TLR4-MD2 and Signals to Induce TNF-α and IL-10" *Retrovirology*
76. Liu, Zhang, Joo et al. (2017) "NF-κB Signaling in Inflammation" *Signal Transduct. Target. Ther*
77. Mogensen, Melchjorsen, Larsen et al. (2010) "Innate Immune Recognition and Activation during HIV Infection" *Retrovirology*
78. Li, Sohn, Choi et al. (2013) "Roles of TLR7 in Activation of NF-κB Signaling of Keratinocytes by Imiquimod" *PLoS ONE*
79. Fiume, Vecchio, De Laurentiis et al. (2012) "Human Immunodeficiency Virus-1 Tat Activates NF-κB via Physical Interaction with IκB-α and P65" *Nucleic Acids Res*
80. Kwon, Brent, Getachew et al. (2008) "Human Immunodeficiency Virus Type 1 Tat Protein Inhibits the SIRT1 Deacetylase and Induces T Cell Hyperactivation" *Cell Host Microbe*
81. Campbell, Rawat, Teodorof-Diedrich et al. (1858) "IRAK1 Inhibition Blocks the HIV-1 RNA Mediated pro-Inflammatory Cytokine Response from Microglia" *J. Gen. Virol*
82. Abbasi, Amini, Moustardas et al. (2024) "Effects of miR-204-5p Modulation on PAX6 Regulation and Corneal Inflammation" *Sci. Rep*
83. Wang, Fan, Teng et al. "Mir204 and Mir211 Suppress Synovial Inflammation and Proliferation in Rheumatoid Arthritis by Targeting Ssrp1. eLife 2022"
84. Jiang, Hong, Guo et al. (2024) *Regulate Oxidative Stress in Myopia. Sci. Rep*
85. Pyo, Yang, Yoo et al. (2008) "Reactive Oxygen Species Activate HIV Long Terminal Repeat via Post-Translational Control of NF-κB" *Biochem. Biophys. Res. Commun*
86. Westendorp, Shatrov, Schulze-Osthoff et al. (1995) "HIV-1 Tat Potentiates TNF-Induced NF-Kappa B Activation and Cytotoxicity by Altering the Cellular Redox State" *EMBO J*
87. Zhang, Sang, Ruan et al. (2011) "Akt/Nox2/NF-κB Signaling Pathway Is Involved in Tat-Induced HIV-1 Long Terminal Repeat (LTR) Transactivation" *Arch. Biochem. Biophys*
88. Demarchi, Di Fagagna, Falaschi et al. (1996) "Activation of Transcription Factor NF-kappaB by the Tat Protein of Human Immunodeficiency Virus Type 1" *J. Virol*
89. Saxena, Ghate, Bichare et al. (2018) "Increased Degranulation of Immune Cells Is Associated with Higher Cervical Viral Load in HIV-Infected Women" *AIDS*
90. Thakar, Patil, Shukre et al. (2014) "Short Communication: Genital Tumor Growth Factor-B1 Levels in HIV-Infected Indian Women Are Associated with Reduced Levels of Innate Antimicrobial Products and Increased HIV Shedding" *AIDS Res. Hum. Retroviruses*
91. Nkwanyana, Gumbi, Roberts et al. (2009) "Impact of Human Immunodeficiency Virus 1 Infection and Inflammation on the Composition and Yield of Cervical Mononuclear Cells in the Female Genital Tract" *Immunology*
92. Bélec, Bélec, Gherardi et al. (1995) "Proinflammatory Cytokine Expression in Cervicovaginal Secretions of Normal and HIV-Infected Women" *Cytokine*
93. Naushad, Okeoma, Islam et al. (2024) "A Dual Anti-HIV and Anti"
94. Peters, Gonzalez-Perez, Musich et al. (2015) "Infection of Ectocervical Tissue and Universal Targeting of T-Cells Mediated by Primary Non-Macrophage-Tropic and Highly Macrophage-Tropic HIV-1 R5 Envelopes" *Retrovirology*
95. "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=PMC12645993&blobtype=pdf | # Porcine hemagglutinating encephalomyelitis virus VW572 (not Gent/PS412 and Labadie) uses the CD81 receptor and MVBderived exosomal pathway for efficient entry and spread in neuronal cells
W Zaib, C Kaviani, X Kang, Y Gao, F Broucke, W Van Den Broeck, S Coppens, S Theuns, H Nauwynck, K Laval
## Abstract
Porcine hemagglutinating encephalomyelitis virus (PHEV) is considered a neurotropic coronavirus that invades the peripheral (PNS) and central (CNS) nervous system of the pig and causes acute encephalomyelitis, also known as "vomiting and wasting disease. " Recently, PHEV has been proposed as a potential surrogate virus model to further elucidate the neuropathogenesis of other betacoronaviruses. In this study, we compared key steps in the replication cycle of three distinct PHEV isolates (VW572, Gent/PS412, and Labadie) in mouse neuronal (N2a) cells. We found that PHEV-VW572 replicates more efficiently in these cells compared to the other two isolates. Interest ingly, PHEV-VW572 showed high intracellular virus titers without efficient extracellular release. Further investigation revealed that PHEV-VW572, but not PHEV-Gent/PS412, mainly uses multivesicular body (MVB)-derived exosomes for viral egress. Transmission electron microscopy confirmed the presence of complete PHEV-VW572 virions within intracellular vesicles and the release of fused PHEV-exosome structures near the plasma membrane. Finally, we showed that PHEV binding is restricted for all isolates. Still, we demonstrated that only PHEV-VW572 entry into cells is mediated by the tetraspanin CD81 receptor. Overall, these results suggest that PHEV-VW572 uses the MVB-derived exosomal pathway as a strategy to promote efficient infection and overcome the early restriction in neuronal cells. In addition, these findings highlight isolate-specific differences in PHEV neurotropism.
IMPORTANCEThe neuropathogenesis of betacoronaviruses remains largely unclear despite the global impact of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic. While these viruses are primarily known for their respiratory effects, mounting evidence suggests they can also cause significant neurological complications, ranging from mild symptoms such as headaches to severe outcomes, such as encephalitis and neurological diseases. The exact mechanisms by which coronaviruses affect the nervous system are still not fully understood, which hampers the development of adequate treatments and prevention strategies for these neurological disorders. In this study, we used the porcine hemagglutinating encephalomyelitis virus (PHEV) as a surrogate model for SARS-CoV-2 to further unravel the neuropathogenesis of betacoronaviruses.KEYWORDS PHEV, neuropathogenesis, CD81 receptor, MVB-derived exosomal pathway C oronaviruses (CoVs) are common pathogens of humans and animals. While most cause mild to severe respiratory infections, CoVs are also known to exhibit neurotropic and neuroinvasive capabilities in several of their hosts, including rodents,
swine, and humans (1)(2)(3). Since the coronavirus disease (COVID-19) pandemic, there is a growing interest in the pathogenesis of betacoronaviruses and their impact on society. Specifically, numerous COVID-19 cases associated with neurological manifesta tions have been reported over the years worldwide (4). Still, the exact role of the virus in this type of complication is not known. Therefore, a better understanding of the neuropathogenic potential of betacoronaviruses is urgently needed. So far, the mouse hepatitis virus has been used as an experimental platform to study the pathobiology of respiratory and neurological manifestations, similar to those of human-CoVs (5,6). Under-studied, but equally relevant, the swine betacoronavirus, porcine hemagglutinat ing encephalomyelitis virus (PHEV), has recently been proposed as a surrogate model to study severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection (7). Like SARS-CoV-2, PHEV can similarly invade the peripheral nervous system (PNS) and spread to the central nervous system (CNS) in mice via the olfactory pathway, therefore causing neuroinflammation and neurological symptoms, suggesting that these two viruses may share common viral mechanisms for neuropathologies in their hosts.
PHEV is a pathogen of veterinary importance, causing acute encephalomyelitis, also known as "vomiting and wasting disease" in pigs (8). The virus circulates subclinically at a high prevalence in most swine herds worldwide. The virus is transmitted primarily through close contact and respiratory droplets and enters via the oral or nasal route. The incubation period is about 4 to 6 days. It can infect naïve pigs of any age, but clinical signs are age-dependent. Neonatal pigs that are born to naïve sows and become infected reach a mortality rate of 100%. The clinical signs include anorexia accompanied or followed within a few hours by vomiting. At 1 day post-infection (dpi), the nasal mucosa, tonsils, and lungs serve as primary replication sites. Alternatively, the virus can also replicate in the epithelial cells of jejunal villi after oral inoculation. From 2 to 3 dpi, PHEV spreads to the PNS. The virus progresses via the peripheral nerves innervating the primary replication site. Several pathways for viral dissemination can be used. The first viral spread pathway is from the nasal mucosa and tonsils to the trigeminal ganglion. A second pathway is from the lungs to the vagal nerve. A third pathway is from the small intestine to the solar ganglion. At 4 dpi, PHEV finally reaches the CNS (mainly pons and medulla). Viral replication and inflammation in the brain lead to death (9). To date, there is no vaccine available (10).
Here, we investigated the neuropathogenesis of PHEV by comparing the replication kinetics of three distinct PHEV isolates in mouse neuroblastoma (N2a) cells. We aimed to identify key similarities and/or differences in the replication cycle between isolates that may influence their neurotropic and neuroinvasive potential.
## RESULTS
## PHEV-VW572 infects N2a cells, but without efficient release of extracellular virus particles
We first characterized the replication kinetics of three PHEV isolates in N2a cells to determine their neuroinvasive potential compared to fully susceptible porcine kidney cells (RPD). As shown in Fig. 1A, the percentage of viral-infected RPD cells was significantly higher (approx 70%) at 24 hours post-inoculation (hpi) for PHEV-VW572 than for PHEV-Gent/PS412 and -Labadie isolates (approx 30%). At 48 hpi, the percentage of infected RPD cells reached >80% for PHEV-VW572 and approximately 60% and 40% for both PHEV-Gent/PS412 and Labadie isolates, respectively. In contrast, the percentage of PHEV-positive N2a cells remained low at 24 hpi and was comparable (approx 10%) between isolates. While the percentage of infected N2a cells increased to approximately 70% for the PHEV-VW572 isolate at 48 hpi, it did not significantly increase (approx 10%-20%) for PHEV-Gent/PS412 and Labadie isolates. At 72 hpi, most RPD cells inoculated with PHEV-VW572 detached due to infection. It was therefore not possible to perform counting. Still, the percentage of Gent/PS412 and Labadie infected N2a cells remained similar between 48 and 72 hpi (Fig. S1). A rate-limiting step of infection was observed in N2a cells infected with PHEV-Gent/PS412 and Labadie at a higher multiplicity of infection (MOI), but not for the PHEV-VW572 isolate (Fig. S2). Representative immunofluorescence (IF) pictures of the PHEV infection are depicted in Fig. 1B andD. No signal was detected in the mock condition. The PHEV S protein was expressed in the cytoplasm of RPD and N2a cells, as single infected cells and syncytia.
We also quantified the extracellular and intracellular PHEV titers in both inoculated cell types to confirm viral infectivity. As shown in Fig. 1C, an increase of intra-and extracellular PHEV titers in RPD cells was detected at 24 hpi for all isolates. Maximal titers were observed between 48 and 72 hpi. A decrease in PHEV intracellular titer starting from 72 hpi was noticed as most cells had been infected by that time. In N2a cells, newly replicated virus was detected for both PHEV-VW572 and Gent/PS412 isolates starting from 3 hpi. Intracellular viral titers increased over the course of the infection for all three isolates and reached comparable levels to those observed in RPD cells. In contrast, lower extracellular virus titers were detected in N2a cells than in RPD cells. While the PHEV extracellular titer increased from 24 to 72 hpi for both Gent/PS412 and Labadie, no significant increase was found for PHEV-VW572 in the neuronal cells throughout the course of the infection (<10 3 50% tissue culture effective dose [TCID 50 ]/mL). Further more, quantitative PCR measurements of PHEV-VW572 N genomic RNA in cell lysates and supernatants confirmed viral titration results. Intracellular viral loads were significantly higher than extracellular ones in N2a cells at 48 and 72 hpi (Fig. S3).
Overall, we concluded that all three PHEV isolates replicated more efficiently in RPD than in N2a cells. While comparable intracellular PHEV titers were reached in both cell types at the late stage of infection, a significantly lower extracellular virus titer was observed in N2a compared to RPD cells throughout the course of the infection. Interest ingly, a high intracellular titer with no increase of extracellular virus was observed in N2a cells inoculated with the VW572 isolate. These findings suggested that new progeny virions accumulate inside the cells with no efficient egress.
## Both PHEV-VW572 and PHEV-Gent/PS412 isolates do not spread cell-to-cell in N2a cells
Based on the above findings, we hypothesized that PHEV-VW572 spreads cell-to-cell in neuronal cells and does not release extracellular virus particles as part of an immune evasive strategy. Spike-induced cell-to-cell fusion is known to be important for efficient cell-to-cell spread of betacoronaviruses (11). To test this hypothesis, we compared the ability of PHEV-VW572 and PHEV-Gent/PS412 isolates to spread from cell to cell in the presence of PHEV-neutralizing antibodies. As PHEV-Gent/PS412 and PHEV-Labadie isolates showed similar viral replication kinetics, we only performed these experiments using PHEV-Gent/PS412 for direct comparison with the PHEV-VW572 isolate. As shown in Fig. 2A, treatment with viral neutralizing antibodies significantly inhibited PHEV spread in N2a cells in a dose-dependent manner. The percentage of PHEV-VW572-infected cells decreased significantly from approximately 70% (control) to approximately 3.5% (with treatment). A similar downward trend was observed for the Gent/PS412 isolate, where the percentage of infected cells went from approximately 20% to 3%. Representative IF pictures are provided in Fig. 2B andC. These findings demonstrate that PHEV does not spread cell-to-cell in N2a cells, independently of the viral isolate used. Therefore, these results do not support our hypothesis that PHEV-VW572 virions accumulate intracellu larly and mainly spread cell-to-cell to evade antibody detection.
## PHEV-VW572, but not PHEV-Gent/PS412, uses MVB-derived exosomal pathway for egress in N2a cells
Betacoronaviruses are known to exploit either the lysosomal pathway or secretory vesicles for release into the extracellular environment (12)(13)(14). Recently, it was shown that PHEV can use the encapsulation of multiple viral components and host factors within multivesicular bodies (MVB) for efficient cell-to-cell communication (11). Still, it is not known whether this strategy is shared among all PHEV isolates or whether specific PHEV isolates use it only for efficient spread in neurons. Hence, we next aimed to examine whether PHEV-VW572 and Gent/PS412 isolates use the MVB-derived exosome pathway for viral egress in neuronal cells or not. Prior to viral inoculation, cells were pre-treated with GW4869, a specific inhibitor of neutral sphingomyelinase (nSMase), essential for exosome biogenesis and release. As shown in Fig. 3A, treatment of cells with 10 µM GW4869 significantly reduced the percentage of PHEV-infected cells from 75% to 30% at 48 hpi for the VW572 isolate. This coincided with a significant decrease in virus titer from 10 4.6 to 10 3.5 TCID 50 /mL (Fig. 3B). In contrast, GW4869 treatment did not affect PHEV infection with the Gent/PS412 isolate. Comparative IF pictures of treated versus untreated infected cells are provided in Fig. 3C andD. Additionally, cells treated with GW4869 after the 1 h viral inoculation showed a similar reduction in the percentage of infection for PHEV-VW572, demonstrating that the inhibitor did not affect the initial entry of viral particles to cells upon pre-treatment (Fig. S4). These findings indicate that PHEV-VW572, but not Gent/PS412 isolate, uses the MVB-derived exosomes for viral exit in neuronal cells.
Using transmission electron microscopy (TEM), we further investigated and compared the viral assembly and egress steps between both isolates. PHEV-VW572 was found to assemble in the ER-Golgi intermediate compartment (ERGIC) with the apposition of nucleocapsids along membranes of the budding compartment as particles developed and budded (Fig. 3E, upper panel). In addition, we observed the presence of complete virions within intracellular vesicles. PHEV-Gent/PS412 assembly also occurred in the ERGIC, but in addition to vesicles containing virions, some naked virus particles were seen, as well in the cytoplasm (Fig. 3E, lower panel). At the egress level, no clear signs of viral exit were detected for PHEV-VW572. No exocytosis, budding of virions at the plasma membrane, nor extracellular vesicles (EVs) containing virions were seen. Instead, we mainly observed fused virus-EV structures in the vicinity of the plasma membrane (Fig. 3F, upper panel). In contrast, clear viral exit sites and spike formation were seen at the plasma membrane of PHEV-Gent/PS412 infected cells (Fig. 3F, lower panel). Furthermore, the plasma membrane integrity was assessed in cells upon PHEV infection by TEM. While cells infected with PHEV-VW572 showed an intact plasma membrane, we found clear signs of plasma membrane disruption and compromised membrane continuity (indica tive of cell lysis) in PHEV-Gent/PS412 infected cells (Fig. S5A). No significant change in the permeability of the plasma membrane of N2a cells was detected during viral infection following trypan blue and propidium iodide exclusion staining (Fig. S5B). Overall, these results supported the concept that PHEV-VW572, but not PHEV-Gent/PS412, uses the MVB-derived exosomal pathway for efficient egress and spread in N2a cells.
## PHEV-VW572 uses exosomal marker tetraspanin CD81 as a receptor for entry into N2a cells
We previously showed that only 10% of neuronal cells were infected with PHEV at 24 hpi. Therefore, we hypothesized that a restriction at the binding and/or entry level(s) may occur in these cells. To test this, we characterized and compared the kinetics of PHEV binding to N2a cells between both isolates using Dio-labeled viral particles. As shown in Fig. 4A, a maximum of approximately 15%-20% of cells showed viral particles bound at their plasma membrane at 60 min post-binding, and this was true for both isolates. No significant difference in the number of bound virus particles (approx 2 particles/per cell) was observed between the two isolates at that time (Fig. 4B). Representative IF pictures of cells with bound PHEV particles are depicted in Fig. 4C. As viral particle aggregates were observed binding to cells for PHEV-VW572 by IF, we further characterized the entry mechanisms of PHEV by TEM. A picture of a PHEV virion is provided in Fig. 4D. As expected, the virion shows a spherical shape of approximately 81 nm in diameter with the presence of a helical nucleocapsid with a typical "crown-like" appearance. We found that the virus mainly enters cells via either engulfment via coated pits, indicative of clathrin-mediated endocytosis, or via direct fusion (Fig. 4E andF). While no major differences in entry mechanisms were observed between the two isolates by TEM, the presence of viral aggregates for VW572 may suggest a possible entry via multi-vesicular bodies. Overall, these findings indicate that PHEV binds to neuronal cells, independently of the viral isolate used.
Next, we determined if PHEV uses a specific receptor or set of receptors, present only on a restricted number of cells, for efficient entry, and if this strategy is shared by both isolates. The angiotensin-converting enzyme 2 (ACE2) has been identified as the main cellular receptor for SARS-CoV-2 (15). Still, N2a cells are known to express very low levels of ACE2, and transfection of these cells with an ACE2-GFP fusion plasmid is often required to study ACE2 cell localization and activity. In line with this, we did not observe a decrease in PHEV infection in cells treated with a function-blocking antibody against ACE2 compared to the untreated condition, independently of the isolate used (Fig. S6A, C, and D). As ß1 integrin has been proposed to serve as an alternative entry receptor for betacoronaviruses, we tested whether PHEV may similarly use it for entry into N2a cells (16). However, no significant decrease in PHEV infection was observed after blocking antibody treatment (Fig. S6B, E, andF). As PHEV-VW572 employs the MVB-exosomal pathway and releases fused PHEV-EVs into the extracellular environment, we tested whether it may use an exosomal receptor for entry into cells. Several tetraspanins, including CD81, CD63, and CD9, have been widely used as markers of exosomes for trafficking viral entry (17)(18)(19). Interestingly, CD81-enriched microdomains, along with CD9, are preferred entry sites for coronaviruses in neuronal cells (20,21). Pretreatment of cells with anti-CD81 blocking antibody significantly inhibited PHEV infection of neuronal cells at 24 hpi, though only for the VW572 isolate (Fig. 4G). Comparative IF pictures of PHEV infection between both isolates and upon CD81 antibody treatment are provided in Fig. 4H andI. Finally, we tested whether small interfering RNA (siRNA) targeting CD81 could efficiently reduce PHEV infection in neuronal cells. At 24 h post-transfection with siRNA sequences, including CD81-targeting and non-targeting control siRNA, cells were inoculated with PHEV for 24 h, and viral protein expression was assessed by IF staining. We demonstrated a significant decrease in PHEV infection in CD81 siRNA-treated cells compared to non-treated ones (Fig. S7). Overall, these findings confirmed that PHEV-VW572 entry into neuronal cells is mediated by the CD81 receptor.
## DISCUSSION
Neuroinfectious diseases represent a major threat to public health worldwide. Respi ratory viruses, such as betacoronaviruses, have been shown to be associated with a broad range of acute, as well as chronic and long-term neurological manifestations (22)(23)(24)(25). To date, the pathophysiological mechanisms associated with these viral infections remain poorly understood. Using PHEV, a surrogate model to study betacoronaviruses, we investigated and compared key steps in the replication cycle of three distinct viral isolates in mouse neuronal cells. First, we compared the replication kinetics of these isolates and demonstrated that viral infection is restricted at early time points in neuronal cells compared to control porcine kidney cells (RPD). No significant increase in the percentage of infected cells was observed for both PHEV-Gent/PS412 and -Labadie isolates throughout the course of infection. In contrast, PHEV-VW572 seemed to overcome the early restriction as demonstrated by an increased percentage of infected cells at 48 hpi. These results suggest that PHEV-VW572 replicates more efficiently in neuronal cells than PHEV-Gent/ PS412 and -Labadie isolates. These results also correlate with the clinical outcome of the infection in pigs. The Belgian PHEV-VW572 was isolated from a pig with vomiting and wasting disease. The presence of infectious PHEV antigen was confirmed in the trigeminal and vagal ganglia following oronasal inoculation in pigs (8). In contrast, PHEV-Gent/PS412 was isolated from a pig that showed respiratory but no neurological symptoms. This may explain why this isolate does not replicate efficiently in neurons. For the PHEV-Labadie isolate, limited information was provided about its isolation from a French farm. Therefore, we performed a complete genomic characterization of this isolate, and a phylogenetic tree was constructed based on the whole-genome sequences of PHEV-VW572, -Gent/PS412, and -Labadie (Fig. S8). The phylogenetic analysis shows that there is a closer relationship between PHEV-Gent/PS-412/2020 and VW572, while the PHEV Labadie is more divergent from those two. The closest relatives for Gent/ PS-412/2020 are LJ/2021 and VW572. There is more genetic divergence between isolate Labadie and PHEV-Gent/PS-412/2020, as illustrated by the branch lengths in the tree. This is in contrast to our data showing similar replication kinetics between PHEV Labadie and Gent/PS412 in N2a cells. It is possible that PHEV-Labadie restricts viral expression and replication in neurons as a strategy for viral persistence, a common feature shared among other coronaviruses (26)(27)(28). A comparative in vivo neuropathogenesis study of these isolates in pigs is needed to further investigate this process.
Second, we demonstrated that PHEV-VW572 infection resulted in a high intracellu lar virus titer alongside an extracellular virus titer, which did not increase. We initially proposed that new progeny virions accumulated inside the cells upon infection as part of an immune-evasive strategy for efficient viral spread. However, we found that neutralizing antibodies abrogated PHEV infection in N2a cells, suggesting that the virus spreads extracellularly rather than cell-to-cell. These results are in striking contrast with previous literature showing that SARS-CoV mediates neuronal cell-to-cell spread to enhance its neuroinvasiveness (11). In recent years, exosomes have been reported to play an important role in facilitating betacoronavirus transmission and modulating host immune responses. Exosomes are small extracellular vesicles (EVs) that are released into the extracellular space when the plasma membrane fuses with an MVB formed via endocytosis. These vesicles are classified based on their size, which can range from 30 to 100 nm (exosomes) to 100-1,000 nm (microvesicles) (29). Upon treatment with an inhibitor of exosome biogenesis (GW4869), we demonstrated that PHEV-VW572, but not Gent/PS412 isolate, uses the MVB-derived exosomal pathway for viral egress in neuronal cells. These results were further confirmed by TEM, where the presence of PHEV-VW572 virions was seen within intracellular vesicles. Nevertheless, no clear signs of exocytosis or release of virions within EVs were seen at the cell surface. In contrast, we observed the presence of fused PHEV-EV structures close to the plasma membrane. These results are partially in agreement with the study from Li et al., which demonstrated the presence of PHEV-modified exosomes in supernatants of PHEV-infected cells by EM (30). The authors concluded that PHEV mainly uses exosomes for cell communica tion, specifically for mediating the transfer of immunostimulatory cargo to uninfected neuroimmune cells. Still, this study only included one PHEV strain (PHEV-CC14), and the authors failed to generalize their conclusions to other PHEV strains or isolates. In contrast, we demonstrated that PHEV assembly and egress are isolate-dependent, and PHEV-VW572 uses the MVB-derived exosomal pathway as a strategy to promote efficient infection and overcome the early restriction in N2a cells. Notably, the inhibitor GW4869 resulted in a significant but partial reduction (approx 50%) in PHEV infection, therefore suggesting the involvement of alternative viral dissemination mechanisms independent of the GW4869-sensitive pathway.
Third, we showed that PHEV binds to approximately 15%-20% of N2a cells, inde pendently of the PHEV isolate used. These results suggest that the restricted PHEV infection observed at the early stage of infection (24 hpi) may occur at a binding level. Betacoronavirus entry is initiated by the binding of viral spike (S) protein to a cellular receptor, followed by proteolytic cleavage of S protein and release of the fusion peptide, allowing for host-cell entry (31). As the use of specific cell receptors is often responsible for viral tropism, we screened potential receptor candidates for PHEV entry into cells by performing function-blocking antibody experiments. Blocking ACE2 and ß1 integrin receptors had no effect on viral infection, suggesting that these two receptors are not involved in PHEV entry into N2a cells. For ACE2, this is not surprising as very low levels of mRNA and protein expression have been detected in these cells (32,33). In contrast, a study from Lv et al. (34) showed that ß1 integrin mediates PHEV entry into cells. It is important to note that the authors used a different viral isolate than those examined in our current study. This underlines again the isolate-specific differences in PHEV pathogenesis. Interestingly, we demonstrated that PHEV-VW572, but not PHEV-Gent/PS412 isolate, uses the tetraspanin CD81 receptor for entry into N2a cells. Tetraspanins play a crucial role in this process by acting as scaffolding proteins that facilitate coronavirus infections by forming specialized microdomains on cell membranes. For instance, it was shown that MERS-CoV enters cells using an entry complex that includes a receptor, a protease, and CD9 tetraspanin (18). The inhibition of PHEV-VW572 infection following neutralizing antibody or siRNA treatment suggests that CD81 is the main receptor for the virus in neuronal cells. The exploitation of tetraspaninenriched microdomains for efficient cell entry has also been observed for other viruses, such as HIV-1 (35). Additionally, the enrichment of CD81 at the cell membrane and its colocalization with structural proteins in neuronal cells may explain why it serves as a preferred entry site for PHEV (36). The fact that the CD81 receptor is not involved in the entry of PHEV-Gent/PS412 isolate further confirms that isolate-specific differences exist in cellular receptor use. Strikingly, CD81 is the most highly enriched protein in EVs, even though it is primarily localized inside the plasma membrane (37). In accordance with the literature, we therefore hypothesize that CD81 may also be expressed on fused PHEV-EVs and facilitate exosomal fusion with the target cell, thus enhancing the overall infection (38). This mechanism may explain why PHEV-VW572 can overcome the early block of infection and efficiently spread in N2a cells over time, while PHEV-Gent/PS412 is not able to do so. A hypothetical model of PHEV entry and spread into N2a cells is provided in Fig. 5. Finally, we showed that both isolates could enter N2a cells either by direct fusion at the plasma membrane or by clathrin-mediated endocytosis, as shown by TEM. These results are consistent with previously reported mechanisms for PHEV (39). Still, it is important to mention that accurate identification of coronavirus particles by TEM remains challenging due to morphological similarities with subcellular structures. Clathrin-coated vesicles have a size similar to coronavirus, but lack nucleocapsid cross-sections and occur freely in the cytoplasm rather than within vacuoles (40,41). These diagnostic pitfalls highlight the necessity of stringent morphological criteria to differentiate coronaviruses, such as PHEV, from cellular mimics in TEM studies. In conclusion, this study reveals new mechanisms by which PHEV efficiently infects and spreads within neuronal cells. It may also contribute to furthering our understanding of betacoronavirus neuropathogenesis in order to develop new treatments against neurological manifestations.
## MATERIALS AND METHODS
## Cells
Porcine kidney cells (Rein de Porc, or RPD) were maintained in RPMI glutamax medium with 10% fetal calf serum, 1% streptomycin, and 0.5% gentamicin, and incubated at 37°C with 5% CO 2 . Murine neuroblastoma (N2a) cells (ATCC CCL-131) were cultured in Dulbecco's modified Eagle medium (DMEM) with supplements and incubated at 37°C with 5% CO 2 .
## Virus
PHEV-VW572 (accession no. DQ011855) (stock titer of 10 6,8 TCID 50 /mL) was isolated in Belgium in 1972 from the tonsils of two diseased pigs suffering from vomiting and wasting disease (42). PHEV-Labadie (accession no. PV820711.1) (stock titer of 10
## Viral inoculation
Cells were (mock) inoculated with either PHEV-VW572, Gent/PS412, or PHEV-Labadie isolates at an MOI of 1 (or 5 when indicated) and incubated for 1 h at 37°C with 5% CO 2 .
After washing with warm DMEM (N2a), 1 mL of culture medium was added, and the cells were further incubated at 37°C with 5% CO 2 . Cell viability was checked by propidium iodide (10 µg/mL, Hello Bio) and trypan blue (0.4%, ThermoFisher) staining. Cell viability was >90%. At corresponding time points, the cell supernatant and lysate were collected and stored at -80°C for viral titration. Cells were fixed with 4% PFA for 10 min, washed with phosphate-buffered saline (PBS), and stored in PBS at 4°C for immunofluorescence (IF) staining.
## Virus titration
To quantify the PHEV infectious virus, both intracellular and extracellular virus titers were determined. The supernatant and cell lysate were collected at different hpi. Viral titration was performed on RPD cells, which are known to be fully susceptible to PHEV infection. The virus titer was calculated as 50% tissue culture effective dose (TCID 50 ) according to the Reed and Muench formula (43).
## Genomic RNA quantification
Total intracellular and extracellular genomic RNA was extracted using the RNeasy Mini Kit (QIAGEN) and the IndiSpin Pathogen Kit (Indical Bioscience), respectively. PHEV N gene RNA was quantified by reverse transcription quantitative PCR using the Takyon One-Step Kit Converter with previously described PHEV N gene-specific primers (7).
A standard curve was generated using serial dilutions of a purified 334 bp amplicon of the PHEV-N gene, and viral RNA copy numbers were calculated and expressed as copies/μL. Amplification was performed with Takyon Low Rox SYBR MasterMix dTTP Blue (Eurogentec) on a QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific).
## Antibodies
The following primary antibodies were used at optimized concentrations: swine polyclonal anti-PHEV biotinylated (1/20, produced in-house), swine anti-PHEV serum (0.86 and 2.15 mg/mL, produced in-house), polyclonal IgG CD81 (Novus Biologicals, NBP2-20564), monoclonal IgG1 anti-integrin ß1 antibody (Abcam, ab24693), and ACE2 antibody polyclonal IgG (Bio-Techne, AF933). The following secondary antibodies were used: streptavidin-FITC (1/200, ThermoFisher Scientific), streptavidin-TR (1/200, ThermoFisher Scientific), goat anti-rabbit AF594 (1/200, ThermoFisher Scientific), and goat anti-rabbit AF488 (1/200, ThermoFisher Scientific).
## IF staining and microscopy
Fixed cell coverslips were washed with PBS and permeabilized using 0.1% Triton X-100 for 2 min, followed by two additional washes with PBS. For staining with PHEV biotinyla ted antibody, an extra step was performed in which cell coverslips were pre-incubated with an avidin/biotin solution (ThermoFisher Scientific) for
## Transmission electron microscopy
Cells were gradually fixed in situ by adding 2.5% glutaraldehyde in 0.05 M sodium cacodylate buffer (EMS #11654) to the culture medium at a 1:1 vol ratio, incubated at 37°C for 5 min. The final 1:0 vol ratio was incubated at 37°C for 20 min, followed by 40 min at room temperature (RT). After fixation, cells were washed four times with 0.1 M sodium cacodylate buffer. Cells were post-fixed using 1% osmium tetroxide (EMS #19150) for 1 h at RT. Cells were subsequently washed four times with UP water, and an en bloc staining step was performed using 1% uranyl acetate (EMS #22400-1) in UP water at 4°C for 1 h in the dark, followed by four washes. Cells were slowly dehydrated in a graded ethanol series for 15 min each. The dehydrated samples were then infiltrated in a 2:1 vol ratio mixture of 100% ethanol and Spurr resin (EMS #14300) for 2 h at RT, followed by a 1:2 vol ratio mixture and fully infiltrated with 100% Spurr resin overnight at 4°C. The next day, two additional changes of Spurr resin were made, ending with a final 100% infiltration overnight at 4°C. The coverslip samples were embedded in fresh Spurr resin and polymerized at 60°C for a minimum of 8 h. Ultrathin sections (~70 nm) were cut using a Leica EM UC6 ultramicrotome equipped with a diamond knife (Ultra 45°, 2.5 mm DiATOME #DU4525). Sections were collected on copper formvar support single slot grids (EMS FF2010-Cu-50) and stained with 1% uranyl acetate at 37°C for 30 min in the dark, followed by a rinse, and further stained once dried with lead citrate for 5 min at RT, followed by a final rinse. Once dried, the sections were examined using a transmission electron microscope (JEOL JEM-1400 plus BF-TEM). Digital images were captured with a Quemesa device camera (Olympus Soft Imaging Solutions, Germany) at various magnifications to visualize subcellular structures and any virus-like particles or organelles of interest.
## Antibody neutralizing assay
To determine whether PHEV can spread cell-to-cell, PHEV-inoculated cells were incubated with swine anti-PHEV neutralizing antibody serum for 48 h at 37°C with 5% CO 2 . Cell coverslips were fixed with 4% PFA for 10 min, and IF staining was performed, as previously described.
## Neutral sphingomyelinase inhibitor GW4869
Cells were pretreated with GW4869 (10 µM) for 24 h, followed by inoculation with PHEV at an MOI of 1 for 1 h. When indicated, cells were first inoculated with PHEV at an MOI of 5, the inoculum was removed, and cells were washed before adding the inhibitor. After 48 h of incubation, cells were fixed and stained for PHEV S protein, as described above.
## PHEV purification and Dio-labeling
Viral purification was performed as previously described (44). Purified PHEV-VW572 and -Gent/PS412 isolates were labeled with Dio (1:100, Vybrant DiO Cell-Labeling, Solution Thermo Fisher) dissolved in DMSO (Molecular Probes), by vigorous vortexing followed by a 1 h incubation at RT. After filtration using a Sephadex G-50 column, the infectivity of Dio-labeled virus remained significantly unchanged. The purity of the PHEV suspensions was assessed using lipophilic labeling and IF staining. Confocal microscopy was used to analyze the staining by randomly selecting five regions.
## PHEV binding assay
Cells were cooled down on ice for 5 min and inoculated with 250 µL of Dio-labeled VW572 and Gent/PS412 virus at an MOI of 1 for 1 h on ice. Afterward, cells were washed twice with cold DMEM to remove unbound particles and were further incubated for 5, 10, 15, 30, and 60 min. Cell coverslips were fixed with 1% PFA for 10 min at 37°C and counterstained with Hoechst. The percentage of PHEV-positive cells was determined by counting the number of cells with viral particles bound on the plasma membrane of 15 randomly selected cells for each isolate. The mean number of virus particles bound per cell was calculated based on the number of virus particles attached to the plasma membrane of five randomly selected PHEV-positive cells using z-stack imaging.
## Antibody blocking assay
Cells were pre-incubated for 2 h at 37°C with the following function-blocking antibod ies: polyclonal IgG anti-CD81 (5 µg/mL, Novus Biologicals, NBP2-20564), monoclonal IgG1 anti-integrin ß1 (5 µg/mL, Abcam, ab24693), polyclonal IgG anti-ACE2 (5 µg/mL, Bio-Techne, AF933), or appropriate isotype controls. After washing, cells were inoculated with PHEV for 1 h, as described above. Function-blocking antibodies were further kept in the medium for 48 h at 37°C. IF staining was performed as previously described.
## siRNA transfection
N2a cells were seeded into a 24-well plate one day before transfection. Cells were transfected with CD81-targeting siRNA (Ambion) at a final concentration of 100 nM using Lipofectamine RNAiMAX (Invitrogen), following the manufacturer's instructions. The siRNA sequences used were sense: 5′-GUACCUCAUUGGAAUUGCAtt-3′, antisense: 5′-UGCAAUUCCAAUGAGGUACag-3′. The 5′-AATCGGGCAGTTGTTTGAGAT-3′ (siCRK) sequence, corresponding to positions 1,023 to 1,043 of Leishmania CRK1, a protein kinase, was chosen as a negative control. siRNA and lipid-based transfection reagent were diluted in Opti-MEM, incubated to allow complex formation, and then added to the cells in complete growth medium. After 24 h of transfection, cells were inoculated with PHEV-VW572 at an MOI of 1. Cells were fixed at 24 hpi, and IF staining was performed, as previously described.
## Statistical analysis
Data were analyzed with GraphPad Prism 9.3.0 software (GraphPad Software Inc.). For statistical significance, data were subjected to a one-way or two-way analysis of variance (ANOVA) followed by a t-test and Tukey's multiple comparisons, respectively. All presented results represent the means and standard deviation (SD) of three independent experiments. *, P value < 0.05; **, P value < 0.01; ***, P value < 0.001; ****, P value < 0.0001; ns, not significant.
## References
1. Jacomy, St-Jean, Brison et al. (2010) "Mutations in the spike glycoprotein of human coronavirus OC43 modulate disease in BALB/c mice from encephalitis to flaccid paralysis and demyelination" *J Neurovirol*
2. Amruta, Ismael, Leist (2022) "Mouse adapted SARS-CoV-2 (MA10) viral infection induces neuroinflammation in standard laboratory mice" *Viruses*
3. Sun, Perlman (1995) "Spread of a neurotropic coronavirus to spinal cord white matter via neurons and astrocytes" *J Virol*
4. Bonhenry, Charnley, Gonçalves (2024) "SARS-CoV-2 infection as a cause of neurodegeneration" *Lancet Neurol*
5. Bender, Weiss (2010) "Pathogenesis of murine coronavirus in the central nervous system" *J Neuroimmune Pharmacol*
6. Chakravarty, Sarma (2021) "Murine-β-coronavirus-induced neuropathogenesis sheds light on CNS pathobiology of SARS-CoV2" *J Neurovirol*
7. Shi, Li, Zhang et al. (2022) "PHEV infection: A promising model of betacorona virus-associated neurological and olfactory dysfunction" *PLoS Pathog*
8. Andries, Pensaert (1980) "Virus isolated and immunofluorescence in different organs of pigs infected with hemagglutinating encephalo myelitis virus" *Am J Vet Res*
9. Andries, Pensaert (1980) "Immunofluorescence studies on the pathogenesis of hemagglutinating encephalomyelitis virus infection in pigs after oronasal inoculation" *Am J Vet Res*
10. Mora-Díaz, Piñeyro, Houston et al. (2019) "Porcine hemagglutinating encephalomyelitis virus: a review. Front Vet Sci"
11. Zeng, Evans, King et al. (2022) "SARS-CoV-2 spreads through cell-to-cell transmission" *Proc Natl Acad Sci*
12. Ghosh, Dellibovi-Ragheb, Kerviel et al. "Altan-Bonnet N. 2020. βcoronaviruses use lysosomes for egress instead of the biosynthetic secretory pathway" *Cell*
13. Wang, He, Li et al. (2023) "The role of lysosomes as intermedi ates in betacoronavirus PHEV egress from nerve cells" *J Virol*
14. Eymieux, Rouillé, Terrier et al. (2021) "Ultrastructural modifications induced by SARS-CoV-2 in Vero cells: a kinetic analysis of viral factory formation, viral particle morphogenesis and virion release" *Cell Mol Life Sci*
15. Li, Moore, Vasilieva et al. (2003) "Angiotensin-converting enzyme 2 is a functional receptor for the SARS coronavirus" *Nature*
16. Park, Myint, Appiah et al. (2021) "The spike glycoprotein of SARS-CoV-2 binds to β1 integrins expressed on the surface of lung epithelial cells" *Viruses*
17. Gräßel, Fast, Scheffer et al. (2016) "The CD63-syntenin-1 complex controls post-endocytic trafficking of oncogenic human papillomaviruses" *Sci Rep*
18. Earnest, Hantak, Li et al. (2017) "The tetraspanin CD9 facilitates MERS-coronavirus entry by scaffolding host cell receptors and proteases" *PLoS Pathog*
20. Florin, Lang (2018) "Tetraspanin assemblies in virus infection" *Front Immunol*
21. Sims, Gu, Krendelchtchikov et al. (2014) "Neural stem cellderived exosomes mediate viral entry" *Int J Nanomedicine*
22. Cone, Zhou, Mcnamara et al. (2024) "CD81 fusion alters SARS-CoV-2 spike trafficking" *MBio*
23. Messlinger, Neuhuber (2022) "Activation of the trigeminal system as a likely target of SARS-CoV-2 may contribute to anosmia in COVID-19" *Cephalalgia*
24. Sarubbo, Haji, Vidal-Balle et al. (2022) "Neurological consequences of COVID-19 and brain related pathogenic mechanisms: a new challenge for neuroscience" *Brain Behav Immun Health*
25. Li, Liu, Lin et al. (2022) "COVID-19 and risk of neurodegenerative disorders: a mendelian randomization study" *Transl Psychiatry*
26. Stein, Ramelli, Grazioli et al. (2022) "SARS-CoV-2 infection and persistence in the human body and brain at autopsy" *Nature*
27. Sturman (1983) "The molecular biology of coronaviruses" *Adv Virus Res*
28. Li, Yang, Li (2024) "Exosomes and SARS-CoV-2 infection" *Front Immunol*
29. Kalluri, Lebleu (1979) "The biology, function, and biomedical applications of exosomes" *Science*
30. Doyle, Wang (2019) "Overview of Extracellular Vesicles, Their Origin, Composition, Purpose, and Methods for Exosome Isolation and Analysis" *Cells*
31. Li, Mu, Tian et al. (2023) "Porcine hemagglutinating encephalomyelitis virus co-opts multivesicularderived exosomes for transmission" *mBio*
32. Letko, Marzi, Munster (2020) "Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronavi ruses" *Nat Microbiol*
33. Qiao, Li, Bao et al. (2020) "The expression of SARS-CoV-2 receptor ACE2 and CD147, and protease TMPRSS2 in human and mouse brain cells and mouse brain tissues" *Biochem Biophys Res Commun*
34. Deshotels, Xia, Sriramula et al. (2014) "Angiotensin II mediates angiotensin converting enzyme type 2 internalization and degradation through an angiotensin II type I receptor-dependent mechanism" *Hypertension*
35. Lv, Li, Guan et al. (2019) "Porcine hemagglutinating encephalomyelitis virus activation of the integrin α5β1-FAK-cofilin pathway causes cytoskeletal rearrangement to promote its invasion of N2a cells" *J Virol*
36. Gordón-Alonso, Yañez-Mó, Barreiro et al. (2006) "Tetraspanins CD9 and CD81 modulate HIV-1-induced membrane fusion" *J Immunol*
37. Martins, Correia, Dias et al. (2019) "CD81 promotes a migratory phenotype in neuronal-like cells" *Microsc Microanal*
38. Jeppesen, Fenix, Franklin et al. (2019) "Reassessment of exosome composition" *Cell*
39. Van Dongen, Masoumi, Witwer et al. (2016) "Extracellular vesicles exploit viral entry routes for cargo delivery" *Microbiol Mol Biol Rev*
40. Li, Zhao, Lan et al. (2017) "Porcine hemagglutinating encephalomyelitis virus enters neuro-2a cells via clathrin-mediated endocytosis in a Rab5-, Cholesterol-, and pH-dependent manner" *J Virol*
41. Hopfer, Herzig, Gosert et al. (2021) "Hunting coronavirus by transmission electron microscopy -a guide to SARS-CoV-2-associated ultrastructural pathology in COVID-19 tissues" *Histopathology*
42. Goldsmith, Tatti, Ksiazek et al. (2004) "Ultrastructural characterization of SARS coronavirus" *Emerg Infect Dis*
43. Vijgen, Keyaerts, Lemey et al. (2006) "Evolutionary history of the closely related group 2 coronaviruses: porcine hemagglutinating encephalo myelitis virus, bovine coronavirus, and human coronavirus OC43" *J Virol*
44. Reed, Muench (1938) "A simple method of estimating fifty per cent endpoints12" *Am J Epidemiol*
45. Laval, Favoreel, Van Cleemput et al. (2016) "Entry of equid herpesvirus 1 into CD172a+ monocytic cells" *J Gen Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12540439&blobtype=pdf | Anna Garbuglia, Athanasios Kossyvakis, Victor Huber, Lei Zhang, Shuang Jin, Dabao Ma, Zhiqiang Liu, Jinsheng Ye, Qingquan Liu
## Abstract
Objectives: Influenza co-infection, characterized by concurrent or sequential infection with influenza and other pathogens, lacks comprehensive quantitative analysis. This study evaluates the status, key hotspots, and clinical advancements in influenza co-infection research from 2005 to 2025 to guide future investigations. Methods: We analyzed articles from 2005 to 2025 sourced from the Web of Science database using R, VOSviewer, and CiteSpace. Concurrently, we extracted clinical trials from PubMed within the same timeframe to assess advancements in the field.
Results:The study analyzed 3,058 articles, noting a consistent rise in publications on influenza co-infection from 2005 to 2025, with a significant spike between 2020 and 2021. The United States led in publication numbers, followed by China, Germany, the United Kingdom, and France. Among these, the United Kingdom exhibited the highest international collaboration. Key collaborative centers included the Centers for Disease Control and Prevention, Emory University, and St. Jude Children's Research Hospital. "PLOS ONE" and "BMC Infectious Diseases" published the most articles, while "Journal of Virology" and "Journal of Infectious Diseases" were the most cited. Keywords such as "infection", "virus", "COVID-19", "children", and "respiratory syncytial virus" highlighted research hotspots and emerging trends in influenza co-infection. The study of pathogenic mechanisms and immune interactions in influenza-bacterial co-infection remains crucial. The COVID-19 pandemic has intensified research on the epidemiological shifts and clinical impacts of co-infection. Emphasis has also been placed on the significance of pediatric populations in influenza and respiratory viral co-infections. Clinical trials have mainly targeted preventive strategies for high-risk groups and the effects of influenza vaccination on the respiratory microbiome.
Conclusion:This study comprehensively analyzes the current research landscape and identifies key hotspots in influenza co-infection. The findings offer crucial guidance for future studies in this field.
The global trends and clinical progress in influenza co-infection: a visualization and bibliometric analysis 1 Introduction
Influenza represent a persistent and formidable global public health threat, responsible for approximately 1 billion cases annually, including 3-5 million severe cases and 290,000-650,000 deaths worldwide, according to the World Health Organization (World Health Organization, 2025). The virus's substantial genetic variability enables it to trigger recurrent seasonal epidemics and poses unpredictable pandemic risks (Smyk et al., 2022;Morens et al., 2023). While influenza itself causes significant global morbidity and mortality, its impact is often magnified through coinfection with other respiratory pathogens (Bartley et al., 2022;Yan et al., 2023).
Co-infection with influenza occurs when a host is simultaneously or sequentially infected with influenza virus and various pathogens, including Streptococcus pneumoniae, Staphylococcus aureus (including methicillin-resistant strains) (Bartley et al., 2022), Haemophilus influenza (Arranz-Herrero et al., 2023), as well as respiratory syncytial virus (RSV) (Haney et al., 2022), SARS-CoV-2 (Yan et al., 2023), and fungi (Feys et al., 2022). It is crucial to recognize that the microbiological diagnosis of co-infection exhibits significant variability depending on the types of pathogens involved. Viral co-infections are generally detected using RT-qPCR or multiplex PCR techniques (Yan et al., 2023), while bacterial co-infections are predominantly verified through quantitative or semi-quantitative sputum and blood cultures (Bartley et al., 2022;Arranz-Herrero et al., 2023). Fungal complications are typically evaluated through a comprehensive methodology that includes culture, antigen or biomarker assays, alongside imaging evidence (Feys et al., 2022). Concurrently, research varies in its application of quantitative thresholds (such as Ct values, viral copy numbers, CFU/mL, antigen indices), sampling intervals, and specimen types (including oropharyngeal/nasopharyngeal swabs, sputum, blood, bronchoalveolar lavage, etc.), and this methodological diversity has a direct impact on detection rates and the clinical interpretation of co-infections (Bartley et al., 2022;Yan et al., 2023;Arranz-Herrero et al., 2023;Feys et al., 2022). Consequently, in the process of defining influenza co-infection, it is essential to take into account the range of pathogens involved as well as the necessary diagnostic criteria and timing considerations. For instance, it is important to classify "concurrent" co-infection as detection occurring within 48 h (Yan et al., 2023), to categorize infections identified during hospital days 1-3 as community-acquired, and to label those identified on days 4-14 as hospital-acquired (Bartley et al., 2022). Cross-study comparability is certainly limited by methodological heterogeneity, but more detailed, pathogen-specific assessments of influenza co-infections have also been made possible by continuous improvements in multiplex diagnostics and improved classification. Concurrently, these developments have resulted in a more precise identification of their clinical detriments.
Abbreviations: RSV, respiratory syncytial virus; Ifs, Impact Factors; JCR, Journal Citation Reports; WoSCC, Web of Science Core Collection; SCP, single-country publication; MCP, multi-country publication; CDC, Centers for Disease Control and Prevention; LAIV, live attenuated influenza vaccine.
These co-infections have clinical significance as they are associated with increased severity of influenza, leading to complications such as severe pneumonia, acute respiratory distress syndrome, and multiple organ failure (Yan et al., 2023). Numerous research studies have highlighted that co-infections significantly raise the risks of hospitalization, admission to intensive care units, the requirement for mechanical ventilation, and mortality rates (Bartley et al., 2022;Yan et al., 2023). Some combinations of coinfections have been shown to result in several-fold increases in mortality rates (Arranz-Herrero et al., 2023). Additionally, coinfections often prolong hospital stays, increase healthcare costs, and present challenges in clinical diagnosis due to overlapping symptoms, thereby impeding prompt identification (Bartley et al., 2022;Haney et al., 2022;Chotpitayasunondh et al., 2021). The challenge of influenza co-infection persists, particularly post-COVID-19 pandemic (Chotpitayasunondh et al., 2021). Consequently, an extensive and growing body of research has expanded, reflecting the growing academic interest and progress in this area. Regrettably, the absence of hotspot and frontier analyses in this domain impedes researchers' ability to swiftly and precisely pinpoint future research trajectories.
Bibliometrics serves as a quantitative tool for analyzing scientific literature to development trends, focal points, and frontiers within specific research domains (Lian et al., 2023). A combined quantitative and qualitative bibliometric assessment of influenza co-infection literature elucidates various publication characteristics, such as key contributing countries, journals, authors, and institutions, prominent studies, common keywords, and collaborative networks among countries, institutions, and authors (Song et al., 2019). This analysis offers new researchers an overview of the field's evolution and development trends (Chen et al., 2018). To address this need, the present study utilized R software, VOSviewer, and CiteSpace for a comprehensive analysis of influenza co-infection literature from 2005 to 2025. The study aims to delineate shifts and patterns in research focal points within this domain and pinpoint potential areas for future investigation. Understanding the current landscape and prospects of influenza co-infection is crucial for its sustainable advancement.
## 2 Materials and methods
## 2.1 Data collection
The data for this study were retrieved from the Web of Science (Capital Medical University Edition) and PubMed databases on June 4, 2025. Details of the search strategy are provided in Figure 1. Articles and review articles were included, without distinguishing pathogen types, sample collection time, and other such factors during the inclusion or analysis phases. Duplicate records were excluded. The remaining articles were saved in plain text format, and their references were exported as full records. Clinical trial results were additionally exported in PubMed format. Subsequently, two researchers undertook a manual screening of clinical trial entries, meticulously examining titles, abstracts, and keywords to systematically eliminate studies that were not pertinent to influenza or co-infection, thus maintaining a high level of topical A flow chart of publication retrieval.
relevance. Discrepancies among reviewers were addressed through discussion, and, when required, adjudicated by a third researcher.
## 2.2 Data analysis
Origin 2018 was used to analyze annual publication trends. Further data visualization and scientific knowledge mapping were performed using R software (version 4.5.0) (http://www.bibliometrix.org) (Aria andCuccurullo, 2017), VOSviewer (version 1.6.20) (van Eck andWaltman, 2010), and CiteSpace (version 6.4.R1) (Chen, 2006), in combination with the bibliometrix package. To maintain accuracy and reliability, two independent authors conducted all data extraction and analytical procedures separately.
VOSviewer was employed to visualize co-authorship networks among countries and institutions, conduct co-citation analysis of sources, and explore keyword co-occurrence. For co-authorship analysis, the minimum number of documents was set at five for each country and 15 for each organization. In the co-citation analysis of sources, only those with at least 104 citations were included. For keyword co-occurrence, the minimum occurrence threshold was set at twenty-five, excluding general terms such as "Influenza" and "Co-Infection." The analysis was conducted using Journal Impact Factors (IFs) from the 2024 edition of the Journal Citation Reports (JCR).
## 3 Results
## 3.1 General landscapes of global publications
A total of 3,058 publications were retrieved from the Web of Science Core Collection (WoSCC) database, including 2,729 articles and 329 reviews. We constructed a line chart to illustrate the annual trend in the number of publications on influenza coinfection from 2005 to 2025 (Figure 2A). Based on the yearly growth rate of publications, the study period was divided into the following four phases: the first phase (2005)(2006)(2007)(2008)(2009) was characterized by slow growth; the second phase (2010-2019) exhibited steady growth; the third phase (2020-2021) demonstrated rapid increase; and the fourth phase (2021-2025) represented a period of high-level fluctuation, primarily influenced by the By examining the countries of corresponding authors, it was observed that the USA (n = 713) was the leading contributor in this field, followed by China (n = 577), Germany (n = 141), the United Kingdom (n = 138), and France (n = 126). Among the top five most productive countries, the USA ranked first both in total articles and single-country publications (SCP, n = 555), with a multi-country publication (MCP) rate of 22.2%. China held the second position in both total publications and SCPs (n = 496); however, only 14.0% of its articles were MCPs, which was notably lower than the other leading countries (Figure 2B, Table 1). In contrast, while the United Kingdom exhibited a lower total number of publications, the percentage of MCPs attained 40.6%, representing the highest figure among the leading five nations. Ireland, Belgium, and Saudi Arabia, despite producing a lower volume of publications overall, exhibited notable MCP rates of 75.0%, 57.1%, and 54.3%, respectively. Furthermore, Figure 3A reveals extensive international collaboration among nations within this domain. The collaboration analysis also identified the Centers for Disease Control and Prevention (CDC, n = 66), Emory University (n = 62), and St. Jude Children's Research Hospital (n = 61) as prominent collaboration centers (Figure 3B, Table 2).
## 3.2 Journals and co-cited journals
Using R software (version 4.5.0) with the Bibliometrix and ggplot2 packages, together with VOSviewer (version 1.6.20) for cocitation journal analysis, a total of 3,058 articles were identified across 681 academic journals (Annex 1). As shown in Table 3 and Figure 4A, "Plos One" published the highest number of articles (n = 129, IF = 2.6), followed by "BMC Infectious Diseases" (n = 85, IF = 3.0), "Viruses-Basel" (n = 81, IF = 3.5), "Journal of Medical Virology" (n = 79, IF = 4.6), and "Journal of Virology" (n = 63, IF = 3.8). Furthermore, Table 4 and Figure 4B present the most frequently cited journals, with "Journal of Virology" (n = 4,719, IF = 3.8) leading, followed by "Journal of Infectious Diseases" (n = 4,231, IF = 4.5), "Plos One" (n = 4,104, IF = 2.6), "Clinical Infectious Diseases" (n = 3,503, IF = 7.3), and "Journal of Immunology" (n = 2,434, IF = 3.4). The co-citation journal map indicates that "Journal of Virology", "Journal of Infectious Diseases", and "Plos One" serve as key collaborative hubs within the field (Figure 5). These results suggest that "Journal of Virology" and "Plos One" may represent influential journals in influenza co-infection research. Additionally, the data reveal a notable scarcity of publications in prestigious journals within this domain, emphasizing the necessity to improve both the depth and quality of associated research.
## 3.3 Analysis of cited-references
We utilized the bibliometrix package in R software to identify the top 25 most cited publications in the domain of influenza co-infection (Table 5). Each of these references has been cited at least 303 times and they are distributed across 20 different journals, indicating that substantial breakthroughs are still lacking in this area. Notably, there is no single dominant journal among the top 25 cited works. The three most cited articles are "Coinfections in patients with COVID-19: A systematic review and meta-analysis, " "Pneumonia and respiratory failure from swineorigin influenza A (H1N1) in Mexico, " and "Bacterial co-infection and secondary infection in patients with COVID-19: a living rapid review and meta-analysis." Upon further analysis, we observed that these publications mainly provide general overviews concerning infection types, epidemiology, and clinical characteristics in the context of influenza co-infection.
To systematically evaluate the major research themes and emerging issues in the field of influenza co-infection, we employed CiteSpace software to identify 435 references with marked citation bursts, according to predetermined criteria (top 25; status count: 2; minimum duration: 2). A representative selection of 25 is depicted in Figure 6. The full list of these references, together with their corresponding DOIs, is provided in Annex 2. Notably, the three references with the highest citation burst strengths were: "Bacterial complications during pandemic influenza infection" (strength: 39.77), "Emergence of a novel swineorigin influenza A (H1N1) virus in humans" (strength: 33.81), and "Pneumonia and respiratory failure from swine-origin influenza A (H1N1) in Mexico" (strength: 32.07). In addition, the most recent citation bursts were associated with the following articles:
(1) "Coinfection with influenza A virus enhances SARS-CoV-2 infectivity, " (2) "SARS-CoV-2 co-infection with influenza viruses, RSV, or adenoviruses, " and (3) "The effects of the COVID-19 pandemic on community respiratory virus activity". In summary, based on the citation burst analysis, we have identified three principal research hotspots in the field of influenza co-infection: (1) The epidemiological and clinical characteristics of influenza co-infection; (2) The pathological and immunological interplay between influenza viruses and co-infecting pathogens; (3) The interaction between COVID-19 and influenza.
## 3.4 Keyword clusters and evolution
Keyword cluster analysis is an effective tool for identifying research hotspots and developmental trends within academic fields. In this study, we applied VOSviewer software to extract 7,990 keywords from the current literature. Table 6 provides a comprehensive overview of keyword frequency, indicating that 20 terms appeared more than 151 times. Notably, "infection" ranked first with 694 occurrences, followed by "virus" (n = 636), "covid-19" (n = 495), "children" (n = 428), "respiratory syncytial virus" (n = 420), "pneumonia" (n = 332), "influenza a virus" (n = 323), and "streptococcus pneumoniae" (n = 279).
In addition, we identified 195 keywords with a minimum occurrence of 25 times and constructed a keyword cluster map using these terms (Figure 7). The map presents four distinct clusters, each represented by a different color. Utilizing cluster analysis, we identified four major thematic clusters that reflect the current research frontiers in influenza co-infection. Cluster 1 (red dots) focuses on the fundamental mechanisms underlying co-infections between influenza viruses and bacteria, particularly Streptococcus pneumoniae and Staphylococcus aureus. Key terms include infection, influenza A virus, Streptococcus pneumoniae, pandemic influenza, and bacteria. Cluster 2 (green dots) centers on pediatric populations, exploring the epidemiological features, clinical diagnostic methods, and disease burden associated with co-infections caused by influenza and other respiratory viruses, such as RSV. This cluster underscores the susceptibility of children to multiple viral co-infections and the significance of this group as a priority in clinical research. Relevant terms include children, RSV, epidemiology, disease, and respiratory tract infection. Cluster 3 (blue dots) highlights research on the evolutionary trajectories, cross-species transmission, and global surveillance of influenza, especially zoonotic viruses, such as avian influenza. Key terms in this cluster include virus, transmission, surveillance, evolution, and identification. Cluster 4 (yellow dots) captures the changing landscape of multi-pathogen co-infection during the COVID-19 pandemic, with a particular focus on variations in the incidence, severity, and mortality of communityacquired pneumonia involving influenza and other respiratory pathogens. Core keywords encompass COVID-19, pneumonia, community-acquired pneumonia, United States, and mortality. The full list of keywords corresponding to these four clusters is provided in Annex 3.
In order to connect these indicators with disease burden, patient characteristics, and decision-making in clinical care and epidemiology, our analysis of keywords underscores the following significant areas of research in influenza and co-infection: (1) Priority populations encompass infants, children, hospitalized patients, and those who are critically ill (e.g., "children, " "young children, " "infants, " "hospitalized children, " "critically-ill patients"). Individuals with compromised immune systems and those with comorbid conditions represent significant populations (e.g., "HIV, " "tuberculosis"). Outcome metrics concentrate on the aspects of hospitalization and mortality. (2) Co-infection constellations focus on influenza A and its various subtypes, such as "H1N1, " "H3N2, " and "H5N1." High-frequency patterns encompass the following combinations: "influenza + streptococcus pneumoniae, " "influenza + staphylococcus aureus, " "influenza + RSV, " "influenza + human metapneumovirus, " "influenza + rhinovirus, " "influenza + coronavirus (including SARS-CoV-2), " "influenza + human bocavirus, " "influenza + adenovirus, " "influenza + parainfluenza virus, " and "influenza + newcastle-disease virus." (3) Geographic hotspots suggest that China and the United States are prominent regions for high-output research and serve as critical settings for epidemiological studies and priority populations, which aligns with national analyses. (4) Populations with significant exposure associated with poultry and livestock represent a concurrent area of interest. Risk is concentrated in occupational and environmental contexts associated with poultry, swine, pigs, chickens, and wild birds, which is indicative of their close relationship with outbreaks of highly pathogenic avian influenza and instances of cross-species transmission. (5) Interventions and prevention strategies encompass vaccination (e.g., "influenza vaccination, " "conjugate vaccine"), antiviral and antimicrobial management (e.g., "oseltamivir, " "procalcitonin, " "c-reactive protein"), multiplex molecular diagnostics (e.g., "real-time PCR, " "multiplex PCR"), as well as public health surveillance and infection control measures (e.g., "surveillance, " "seasonality, " "guidelines"). The integration of vaccination strategies alongside the focused management of highrisk populations is identified as a critical area for both clinical and policy intervention (Annex 3).
Additionally, to elucidate temporal shifts and forecast trends in influenza co-infection research, we utilized the bibliometrix toolkit in R to create a visual trend topic chart (Figure 8). From 2007 to 2013, research primarily focused on identifying and monitoring respiratory pathogens, with an emphasis on assay development, viral origins, and clinical specimen analysis. Foundational studies during this period concentrated on influenza virus subtypes, such as H1N1, and zoonotic reservoirs, particularly birds, setting the stage for future research. Between 2013 and 2017, the focus shifted toward the epidemiology and clinical impact of pediatric respiratory infections, including RSV, human metapneumovirus, and co-infection patterns in infants and young children. During this phase, terms related to diagnostic innovations and communityacquired pneumonia emerged. From 2017, studies increasingly focused on bacterial co-infections, antibacterial resistance, and the immunological interactions between influenza viruses and bacterial pathogens like Streptococcus pneumoniae. Since 2022, interest has surged in COVID-19, SARS-CoV-2, and the clinical impact of viral co-infections, including nonpharmaceutical interventions and the global effects of pandemics. Our analysis suggests that research on influenza co-infection will likely intensify, concentrating on copathogenesis mechanisms, clinical risk assessment, and interactions between emerging respiratory viruses and bacterial agents, with particular attention to vaccine development and populationlevel management.
## 3.5 Clinical progress analysis
Six clinical trials were identified from the PubMed database (Annex 4). The following work presents a concise overview of the principal arguments: (1) The examination of host genetic factors indicated that low-expression variants of the mannosebinding lectin 2 gene did not demonstrate a significant correlation with heightened overall susceptibility or severity of critical influenza in pediatric populations. In the subgroup of influenza associated with methicillin-resistant staphylococcus aureus coinfection, an increased frequency of the B allele among carriers was noted. (2) In pediatric patients diagnosed with hematologic malignancies and receiving chemotherapy, community-acquired influenza and parainfluenza viruses have been recognized as prevalent pathogens associated with lower respiratory tract infections, which primarily demonstrated self-limiting trajectories. Under conditions of immunosuppression, the association of viralbacterial co-infection with rapid and severe disease progression No clinically significant adverse events were documented. (6) A further investigation revealed that, although LAIV resulted in a lower incidence of symptoms following bacterial exposure when compared to the inactivated influenza vaccine, it was linked to markedly elevated rates and density of pneumococcal colonization. These trials primarily investigate two key areas in this field: (1) Unique host types and strategies for preventing influenza co-infection;
(2) The influence of influenza vaccines on the microbiota of the respiratory tract and their correlation with co-infection. 4 Discussion
## 4.1 General information
The present study conducted a comprehensive bibliometric and visual analysis of 3,058 publications on influenza co-infection from 2005 to 2025. The results reveal a distinct upward trend in the volume of literature over this period. Specifically, the study period can be divided into four phases based on the annual publication growth rates: (1) slow growth from 2005 to 2009, (2) steady increases from 2010 to 2019, (3) rapid expansion from 2020 to 2021, (4) sustained high-level fluctuations from 2021 to 2025. These publication trends closely mirror two major public health events: the 2009 H1N1 influenza pandemic (Fineberg, 2014) and the onset of the COVID-19 pandemic in 2019 (Yan et al., 2023;Chotpitayasunondh et al., 2021). It is important to note that the data for 2025 are incomplete, as the search was conducted on June 4, 2025.
The United States is the global leader in published research on influenza co-infection, with 713 articles, demonstrating the country's substantial research interest and contributions in this field. China ranks second with 577 publications, while Germany, the United Kingdom, and France have also made significant advances. Among the top 20 most prolific institutions, the U.S.based CDC, Emory University, and St. Jude Children's Research Hospital lead with 66, 58, and 48 publications, respectively, underscoring the U.S. dominance. However, institutions from China, Australia, France, Spain, and the United Kingdom are also well-represented, reflecting the global attention and collaborative efforts in this research domain. Additionally, the CDC has historically taken the initiative or partnered in the establishment of comprehensive nationwide programs for influenza sentinel surveillance, hospitalization surveillance, pathogen spectrum surveillance, and co-infection surveillance. The accumulation of high-quality, reusable large-scale clinical and molecular epidemiological datasets has established a robust data foundation for multicenter studies and expedited publication processes, resulting in a significant increase in research output. From an Research Hospital, which is well-known for its research in pediatrics and severe infections, and is ranked among the top three institutions based on publication count. This indicates a favorable relationship between the areas of collaboration hubs and thematic hotspots.
The extant literature on influenza co-infection is extensive, spanning 3,058 articles published across 681 academic journals. The journals with the highest publication output in this domain are PLOS ONE, BMC Infectious Diseases, and Viruses-Basel, underscoring their substantial contributions to the development of this field. Furthermore, Journal of Virology and PLOS ONE emerge as the most frequently cited journals, serving as central hubs of journal collaboration and thus highlighting their representative and influential role within this research area.
## 4.2 Hotspots and development trends
As mentioned above, by conducting a comprehensive analysis of literature clustering, keyword frequency, keyword co-occurrence, and research topic evolution, we identified emerging research hotspots in influenza co-infection. The findings emphasize three primary areas: First, the pathogenic mechanisms and immune interactions in influenza-bacterial coinfections remain a core focus, highlighting how synergistic effects worsen clinical outcomes. Second, the COVID-19 pandemic has caused epidemiological shifts and added complexity to the clinical burden of influenza coinfections, emphasizing the need for updated surveillance and management. Third, the epidemiology and Co-citation analysis of journals in influenza co-infection.
precision management of influenza-respiratory virus coinfection in children require special attention due to their increased susceptibility and unique clinical needs. It should be noted that the research findings are not prescriptive inferences but rather the result of a systematic examination of the body of existing knowledge.
## 4.2.1 Pathogenic mechanisms and immune interactions in influenza-bacterial coinfection
According to bibliometric analysis, the topic of "pathogenic mechanisms and immune interactions" has been identified as a significant research hotspot within the domain of influenza coinfection. The interaction between influenza viruses and bacterial pathogens is crucial in infection biology. Influenza infection increases susceptibility to secondary bacterial infections, worsening disease severity and mortality (Aguilera and Lenz, 2020;Morris et al., 2017). Streptococcus pneumoniae and Staphylococcus aureus are the most common secondary bacterial pathogens, while Haemophilus influenzae, Klebsiella pneumoniae, and Moraxella catarrhalis also play important roles among Gram-negative bacteria (Morris et al., 2017;Sender et al., 2021;Huo et al., 2025).
Numerous aspects reveal the synergistic effects of influenza viruses and bacteria. Firstly, influenza viruses can promote the adhesion of bacteria to respiratory epithelial cells during respiratory tract infections, resulting in a higher bacterial load in the tissues (Rowe et al., 2019). Secondly, influenza-induced damage to the respiratory epithelium impairs mucociliary clearance and exposes additional bacterial adhesion sites, greatly increasing bacterial attachment and invasion (Sender et al., 2021;Huo et al., 2025;Oliva and Terrier, 2021). Thirdly, type I and type III interferons induced by influenza viruses delay epithelial repair by inhibiting cell proliferation and differentiation and promoting apoptosis, increasing the risk of secondary bacterial infections and worsening disease severity (Oliva and Terrier, 2021;Major et al., 2020). In addition, the dysregulation of the host immune response also plays a crucial role (Aguilera and Lenz, 2020). While type I interferons are vital for antiviral defense, their excessive activation impairs antibacterial responses by reducing neutrophil recruitment and disrupting alveolar macrophages, promoting bacterial survival and spread. Co-infection induces a synergistic cytokine stormwith elevated IL-6, IL-1β, TNF-α, and MCP-1-further increasing susceptibility to secondary bacterial infections and worsening tissue damage and disease severity (Sender et al., 2021;Oliva and Terrier, 2021;Zangari et al., 2021). At the same time, the adaptive immune response, involving CD8+/CD4+ T cells and antibodies, is essential for clearing secondary bacterial infections, but factors like weakened Th17 responses, inhibited T-cell function, and reduced antibacterial cytokines can compromise this protection and raise the risk of secondary infections (Sender et al., 2021;Oliva and Terrier, 2021). Furthermore, following co-infection with influenza virus and Staphylococcus aureus, mitochondrial autophagy in lung epithelial cells inhibits apoptosis, promoting viral and bacterial proliferation, worsening inflammation and pneumonia, and ultimately lowering survival rates (Huo et al., 2025). Additionally, influenza and bacterial neuraminidases synergistically cleave sialic acid on host cells, enhancing bacterial adhesion and colonization in the respiratory tract and worsening the pathological damage of secondary infections and co-infections (Alshammari et al., 2025). Ultimately, secondary bacterial pathogens exploit virusinduced immunosuppression and tissue damage to increase their proliferation and pathogenicity. Influenza-altered alveolar macrophages have reduced phagocytic and bactericidal functions, leading to higher bacterial loads and worse clinical outcomes (Lee et al., 2018;Deinhardt-Emmer et al., 2020). Future research will focus on the synergistic mechanisms between influenza subtypes and diverse bacteria, leveraging advanced omics and modeling technologies to enable precise prevention and treatment. Approaches such as combined vaccination and immune-response modulation will be important for better managing co-infections. In addition, differences in specimen types, sampling strategies, and diagnostic methods can significantly impact detection yield. It is crucial to address these sources of variability to minimize misclassification and enhance the reliability of conclusions regarding clinical burden, pathogenic mechanisms, and immune interactions in studies of influenza-bacterial coinfection (Xue et al., 2024;Zhang et al., 2025). Furthermore, future research will necessitate thorough diagnostic stewardship (Elbehiry and Abalkhail, 2025;Yi et al., 2025). Metagenomic and next-generation sequencing have the potential to enhance the detection of pathogens in influenza-associated coinfections; however, mere detection does not suffice to establish a causal relationship (Elbehiry and Abalkhail, 2025). Respiratory specimens with low biomass often harbor oral commensals and opportunistic organisms, with results being affected by the burden of host DNA, background contamination, and the specific thresholds of the analytical platform used (Elbehiry and Abalkhail, 2025;Yi et al., 2025). Standardized pre-analytical workflows, along with quantitative or semiquantitative reporting and interpretation frameworks that incorporate clinical context, host-response signals, and traditional microbiology, are crucial for differentiating between colonization and genuine bacterial pathogenesis in the context of influenza (Elbehiry and Abalkhail, 2025). Examples include the enhanced sensitivity of metagenomic next-generation sequencing for challenging pathogens like Nocardia when specific positivity criteria and background subtraction are utilized, as well as 16S-based algorithms that distinguish pathogenic streptococci from oral flora (Kurniawan et al., 2025;Chen et al., 2025). The methodological safeguards in place serve to prevent the misidentification of incidental microbes as causative agents of disease, thereby enhancing the precision of mechanistic inferences in research pertaining to influenza-bacterial coinfection (Elbehiry and Abalkhail, 2025;Kurniawan et al., 2025;Chen et al., 2025). Simultaneously, the advancement of biomarkers is anticipated to enhance early detection and treatment methodologies.
## 4.2.2 Epidemiological shifts and clinical burden of influenza coinfection during the COVID-19 pandemic
Our analysis indicates that "epidemiological shifts and the clinical burden of influenza co-infection during the COVID-19 pandemic" represents another significant research interest in the field of influenza co-infection. In the early COVID-19 pandemic, widespread non-pharmaceutical interventions like mask-wearing, social distancing, and travel restrictions significantly reduced global seasonal influenza activity and co-infection cases (Pun et al., 2024). As these measures were eased, influenza activity rebounded in several regions, often coinciding with continued SARS-CoV-2 transmission, particularly during the typical influenza season.
Bacterial co-infection in COVID-19 patients occurs at a lower rate than historically observed in influenza, with a prevalence of 7%, thus routine antibiotic use is not advised (Lansbury et al., 2020). Respiratory viral co-infection in COVID-19 patients is about 5.01%, with influenza viruses comprising 1.54% (Krumbein et al., 2023). Among these, 73.6% are due to influenza A and 25.1% to influenza B (Varshney et al., 2023). Although less frequent than single infections, co-infection with influenza and SARS-CoV-2 results in poor outcomes, such as deterioration or death, in 15.7% of cases, posing significant clinical challenges (Varshney et al., 2023).
SARS-CoV-2, as a novel pathogen in the setting of influenza co-infection, can infect pulmonary epithelial cells concurrently with the influenza virus (Zarkoob et al., 2022). This dual infection often leads to more severe pulmonary inflammation and tissue damage, heightening the risk of intensive care units admission and invasive mechanical ventilation (Swets et al., 2022), thus increasing mortality risk (Alosaimi et al., 2021;Yu et al., 2021;Stowe et al., 2021). Reports have also highlighted a heightened risk of longterm complications and adverse outcomes in these patients (Yue et al., 2020). The emergence of SARS-CoV-2 has complicated the immunological response to co-infection. Influenza infection can temporarily suppress SARS-CoV-2 replication (Zarkoob et al., 2022), while prior SARS-CoV-2 infection might inhibit influenza virus entry (Kuriakose and Kanneganti, 2023). For instance, a test-negative case-control study showed that individuals with influenza had a 58% lower risk of testing positive for SARS-CoV-2, suggesting competitive inhibition between the viruses (Stowe et al., 2021). Nonetheless, co-infection with SARS-CoV-2 and influenza can worsen immune dysregulation and provoke cytokine storms, exacerbating clinical outcomes (Bai et al., 2021;Vilas et al., 2022).
Patients with co-infections often exhibit symptoms indistinguishable from those with singular infections of COVID-19 or influenza, complicating clinical diagnosis and treatment (Varshney et al., 2023). This overlap particularly endangers vulnerable populations. The similarity in clinical presentations delays timely diagnosis and appropriate therapeutic decisions, underscoring the need for enhanced biomarker identification and nucleic acid detection methods (Carbonell et al., 2023). To address co-infection risks, robust surveillance of both influenza and SARS-CoV-2 is crucial, alongside ongoing vaccination efforts. Developing vaccines effective against both pathogens could be a cost-effective strategy to alleviate public health burdens and reduce complications from co-infections (Sanchez-Martinez et al., 2024).
## 4.2.3 Epidemiology and management of influenza-respiratory virus coinfection in pediatric populations
Our bibliometric analysis highlights "epidemiology and management of influenza-respiratory virus co-infection in pediatric populations" as the third prominent research hotspot. Children, due to their developing immune systems and regular exposure to pathogens in environments like daycare facilities and educational institutions, are highly vulnerable to viral co-infections. A study by Mandelia et al. (2021) revealed that approximately 10.8% of samples from respiratory viral infections exhibited co-infections. Interestingly, the incidence of coinfections was significantly higher in children, accounting for 18% of cases, compared to only 2.8% in adults, indicating a nearly six-fold contrast. Notably, the majority of these co-infections were observed in children under the age of five (Mandelia et al., 2021). Some studies incredibly suggest that as many as 93% of such occurrences happen among children (Weidmann et al., 2023). This underscores the importance of giving priority to children in studies on co-infections linked to influenza. In outbreaks of influenza, around 26.3% of individuals face co-infections, commonly with rhinovirus, adenovirus, or RSV, with children being the most affected group (Torner et al., 2024). These co-infections in children often lead to more serious consequences, such as heightened need for intensive care, prolonged hospitalization, and intricate treatment regimens, emphasizing the significance of pediatric viral co-infection as a notable public health issue (Mandelia et al., 2021).
Our analysis of keyword clusters revealed cluster 2, emphasizing "childre" and "RSV" (Table 6), highlighting crucial research areas. RSV is a primary cause of acute lower respiratory tract infections in children under five, potentially resulting in pneumonia and increasing the likelihood of long-term respiratory problems such as asthma (Pacheco et al., 2021). Despite the significant socio-economic impact of human RSV, there is currently no approved vaccine (Pacheco et al., 2021). A Danish study revealed that influenza cases with various respiratory pathogens, including RSV, are more common in children under 5 years old and decline with increasing age (Schneider et al., 2020). Furthermore, the presence of SARS-CoV-2 presents a notable challenge in cases of pediatric influenza co-infections. In the U.S. during the 2021-2022 influenza season, 6% of pediatric influenza hospitalizations involved co-infections with SARS-CoV-2, and 16% of influenza-related pediatric fatalities were associated with co-infections. These instances often required invasive mechanical ventilation or noninvasive respiratory support such as BiPAP/CPAP (Adams et al., 2022).
Accurate diagnosis and management play a critical role in preventing influenza co-infection among children. Evidence suggests that timely identification of viral infections through pointof-care testing, based on clear clinical indications, can reduce unnecessary antibiotic use and shorten hospital stays (Schneider et al., 2020). This approach is particularly advantageous for children under 5 years old and outside of influenza/RSV seasons (Schneider et al., 2020). Prompt initiation of antiviral treatment is also crucial in preventing co-infections associated with influenza (Xu et al., 2023). Live attenuated vaccines are indispensable for safeguarding vulnerable populations, such as children (Ryan et al., 2022). It is imperative to enhance the research, development, and distribution Keyword co-occurrence map for influenza co-infection publications.
of vaccines targeting influenza and other respiratory viruses, with a special focus on children (Adams et al., 2022).
## 4.3 Clinical progress
This research synthesizes and critically evaluates six pivotal clinical trials, elucidating a tripartite framework for contemporary clinical investigations on influenza co-infection: targeted protection for vulnerable populations; the nuanced ecological implications of vaccination and its clinical significance; and the imperative to transform future research paradigms.
Initially, in prevention strategies and safeguarding highrisk populations, clinical practice is advancing toward risk stratification to implement precision public health measures that are customized according to varying degrees of vulnerability. In immunosuppressed individuals, such as pediatric patients postchemotherapy, a multifaceted strategy is essential: enhancing active immunization through vaccination, bolstering passive protection via improved infection prevention and control measures, and implementing early and accurate diagnostic techniques, such as multiplex PCR, to address the dangerous interplay between viral and bacterial pathogens (Tantawy et al., 2015). For healthcare professionals facing elevated exposure risks, empirical evidence underscores the efficacy of N95 respirators over medical masks in mitigating bacterial colonization and co-infection; thus, N95 should be established as the standard for protection in high-risk environments (MacIntyre et al., 2014).
Secondly, concerning the interplay between influenza vaccination and microbial ecology, a cutting-edge subject, our assessment uncovers a complex and dual perspective. The LAIV typically exhibits transient and mild effects on the nasopharyngeal microbiome (Peno et al., 2025); however, under specific circumstances (e.g., co-infection with other viruses), it may temporarily facilitate the proliferation of opportunistic bacteria such as streptococcus pneumoniae (Peno et al., 2021). This indicates a phenomenon that possesses both advantageous and disadvantageous aspects. On one hand, baseline microbiome structures may enhance mucosal immune responses (Peno et al., 2025). Conversely, alterations in the local immune microenvironment induced by vaccines may unintentionally establish brief periods that promote bacterial colonization (Hales et al., 2020;Peno et al., 2021). Nevertheless, the existing conclusions are constrained by methodological limitations, such as inadequate resolution of 16S rRNA and qPCR, absence of transmission and disease endpoints, as well as issues related to sample Trending topics in influenza co-infection.
representativeness. The identified tensions suggest a distinct research agenda: the adoption of metagenomics, longitudinal sampling, and the connection to clinical endpoints to clarify mechanisms and quantify absolute risk.
In conclusion, the existing evidence substantiates the need for tailored protection strategies for high-risk groups and calls into question conventional models for assessing the benefitrisk ratio of vaccines. This highlights the necessity for research driven by mechanisms at broader population scales to enhance clinical practice.
## 4.4 Limitations
This study outlines emerging research directions and identifies key hotspots in the field of influenza co-infection, providing valuable insights for future exploration. However, several limitations must be acknowledged. First, our bibliometric analysis primarily utilized the WoSCC database. Although this may have led to the exclusion of some relevant publications, the Web of Science platform is widely recognized for its rigorous curation, quality standards, and reliability in bibliometric studies, offering a robust foundation for our analysis. To address this limitation, we supplemented our assessment with clinical progress data from PubMed-based clinical trials. Second, our analysis was restricted to English-language publications, excluding potentially relevant research in other languages. Third, There exists significant variability in the operational definition of "coinfection" among the primary studies incorporated within our corpus. There are additional variations in laboratory criteria for pathogen confirmation, such as qPCR cycle-threshold cut-offs, microbiological culture interpretation standards, and the kinds of clinical specimens examined (e.g., upper vs. lower respiratory tract samples). The variability in definitions and methodologies leads to the potential for misclassification bias. Consequently, it is imperative to exercise caution when engaging in cross-study comparisons and synthesizing findings. While we have emphasized these issues in the discussion part, they continue to represent a fundamental limitation on the strength and comparability of our conclusions. Lastly, we did not conduct an in-depth author analysis. A significant portion of influenza co-infection research originates from China, where common surnames complicate accurate author disambiguation, potentially affecting authorspecific metrics. Despite these limitations, our study provides a comprehensive overview of the current research landscape, effectively mapping prevailing hotspots and emerging trends in influenza co-infection research.
## 5 Conclusion
Our study identifies the main research hotspots and frontiers in influenza co-infection. The key findings are:
A. Influenza co-infection research has garnered significant global interest, with the United States, China, Germany, the United Kingdom, and France as leading contributors. These countries engage in extensive and in-depth collaboration. B. "PLOS ONE" and "BMC Infectious Diseases" are leading journals in publishing studies in this field, with "Journal of Virology" being the most frequently cited. "Journal of Virology" and "PLOS ONE" likely have a significant impact on influenza co-infection research. C. The investigation into the pathogenic mechanisms and immune interactions of influenza-bacterial co-infection is a major research focus. D. The epidemiological shifts and clinical impact of influenza co-infection during the COVID-19 pandemic have gained prominence in recent studies. E. Research has also concentrated on pediatric populations as a key focus for influenza and respiratory viral co-infections. F. Clinical trials in this area primarily address the increased risk of severe co-infections in certain groups, such as immunocompromised individuals or those with highrisk genetic profiles, emphasizing the need for targeted prevention. Additionally, studies on the transient effects of influenza vaccination on the respiratory microbiome and bacterial colonization offer valuable insights into coinfection risk.
In conclusion, our study provides valuable insights into the research trends and focal points within the realm of influenza co-infection. These results enhance researchers' understanding and scholarly discourse on this subject. Additionally, our analysis underscores emerging trends and potential avenues for future exploration. By elucidating the present research landscape and pinpointing areas warranting additional scrutiny, this study imparts crucial insights and recommendations for researchers. It aids in informed decision-making and fosters innovation in forthcoming investigations on influenza co-infection.
## References
1. Adams, Tastad, Huang et al. (2022) "Prevalence of SARS-CoV-2 and influenza coinfection and clinical characteristics among children and adolescents aged <18 years who were hospitalized or died with influenza -United States, 2021-22 Influenza Season" *MMWR Morb. Mortal. Wkly. Rep*
2. Aguilera, Lenz (2020) "Inflammation as a modulator of host susceptibility to pulmonary influenza, pneumococcal, and co-infections" *Front. Immunol*
3. Alosaimi, Naeem, Hamed et al. (2021) "Influenza co-infection associated with severity and mortality in COVID-19 patients" *Virol. J*
4. Alshammari, Maina, Blanchard et al. (2025) "Understanding the molecular interactions between influenza a virus and streptococcus proteins in co-infection: a scoping review" *Pathogens*
5. Aria, Cuccurullo (2017) "bibliometrix: an R-tool for comprehensive science mapping analysis" *J. Informetr*
6. Arranz-Herrero, Presa, Rius-Rocabert et al.
7. Lalueza (2023) "Determinants of poor clinical outcome in patients with influenza pneumonia: A systematic review and meta-analysis" *Int. J. Infect. Dis*
8. Bai, Zhao, Dong et al. (2021) "Coinfection with influenza A virus enhances SARS-CoV-2 infectivity" *Cell Res*
9. Bartley, Deshpande, Yu et al.
10. (2022) "Bacterial coinfection in influenza pneumonia: rates, pathogens, and outcomes" *Infect. Control Hosp. Epidemiol*
11. Carbonell, Moreno, Martin-Loeches et al. (2023) "The role of biomarkers in influenza and COVID-19 community-acquired pneumonia in adults" *Antibiotics*
12. Chen (2006) "CiteSpace II: Detecting and visualizing emerging trends and transient patterns in scientific literature" *J. Am. Soc. Inform. Sci. Technol*
13. Chen, Xie, Wang et al. (2018) "Clinical features of pulmonary nocardiosis and diagnostic value of metagenomic next-generation sequencing: a retrospective study" *BMC Med. Inform. Decis. Mak*
14. Chotpitayasunondh, Fischer, Heraud et al. (2021) "Influenza and COVID-19: what does co-existence mean? Influenza Other Respir" *Viruses*
15. Deinhardt-Emmer, Rennert, Schicke et al. (2020) "Co-infection with Staphylococcus aureus after primary influenza virus infection leads to damage of the endothelium in a human alveolus-ona-chip model" *Biofabrication*
16. Elbehiry, Abalkhail (1991) "Metagenomic next-generation sequencing in infectious diseases: clinical applications, translational challenges, and future directions" *Diagnostics*
17. Feys, Goncalves, Khan et al. (2022) "Lung epithelial and myeloid innate immunity in influenza-associated or COVID-19associated pulmonary aspergillosis: an observational study" *Lancet Respir. Med*
18. Fineberg (2009) "Pandemic preparedness and response-lessons from the H1N1 influenza of" *N. Engl. J. Med*
19. Hales, Jochems, Robinson et al. (2020) "Symptoms associated with influenza vaccination and experimental human pneumococcal colonisation of the nasopharynx" *Vaccine*
20. Haney, Vijayakrishnan, Streetley et al. (2022) "Coinfection by influenza A virus and respiratory syncytial virus produces hybrid virus particles" *Nat. Microbiol*
21. Huo, Li, Tang et al. (2025) "Vital role of PINK1/Parkin-mediated mitophagy of pulmonary epithelial cells in severe pneumonia induced by IAV and secondary staphylococcus aureus infection" *Int. J. Mol. Sci*
22. Krumbein, Kummel, Fragkou et al. (2023) "Respiratory viral co-infections in patients with COVID-19 and associated outcomes: a systematic review and meta-analysis" *Rev. Med. Virol*
23. Kuriakose, Kanneganti (2023) "Pyroptosis in antiviral immunity" *Curr. Top. Microbiol. Immunol*
24. Kurniawan, Alia, Shiraishi et al. (2020) "systematic algorithm using 16S ribosomal RNA for accurate diagnosis of pneumonia pathogens" *Sci. Rep.-UK*
25. Lee, Morris-Love, Cabral et al. (2018) "Coinfection with influenza A virus and Klebsiella oxytoca: an underrecognized impact on host resistance and tolerance to pulmonary infections" *Front. Immunol*
26. Lian, Li, Lan et al. (2023) "Bibliometric and visual analysis in the field of tea in cancer from 2013 to 2023" *Front. Oncol*
27. Macintyre, Wang, Rahman et al. (2014) "Efficacy of face masks and respirators in preventing upper respiratory tract bacterial colonization and co-infection in hospital healthcare workers" *Prev. Med*
28. Major, Crotta, Llorian et al. (2020) "Type I and III interferons disrupt lung epithelial repair during recovery from viral infection" *Science*
29. Mandelia, Procop, Richter et al. (2021) "Dynamics and predisposition of respiratory viral co-infections in children and adults" *Clin. Microbiol. Infect*
30. Morens, Taubenberger, Fauci (2023) "Rethinking nextgeneration vaccines for coronaviruses, influenzaviruses, and other respiratory viruses" *Cell Host Microbe*
31. Morris, Cleary, Clarke (2017) "Secondary bacterial infections associated with influenza pandemics" *Front. Microbiol*
32. Oliva, Terrier (2021) "Viral and Bacterial Co-Infections in the Lungs: Dangerous Liaisons" *Viruses*
33. Pacheco, Galvez, Soto et al. (2021) "Bacterial and viral coinfections with the human respiratory syncytial virus" *Microorganisms*
34. Peno, Armitage, Clerc et al. (2021) "The effect of live attenuated influenza vaccine on pneumococcal colonisation densities among children aged 24-59 months in The Gambia: a phase 4, open label, randomised, controlled trial" *Lancet Microbe*
35. Peno, Jagne, Clerc et al. (2025) "Interactions between live attenuated influenza vaccine and nasopharyngeal microbiota among children aged 24-59 months in The Gambia: a phase 4, open-label, randomised controlled trial" *Lancet Microbe*
36. Pun, Tao, Yam et al. (2024) "Respiratory viral infection patterns in hospitalised children before and after COVID-19 in Hong Kong" *Viruses*
37. Rowe, Meliopoulos, Iverson et al.
38. Rosch (2019) "Direct interactions with influenza promote bacterial adherence during respiratory infections" *Nat. Microbiol*
39. Ryan, Schewe, Crowe et al. (2022) "Sequential delivery of live attenuated influenza vaccine and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in the ferret model can reduce SARS-CoV-2 shedding and does not result in enhanced lung pathology" *J. Infect. Dis*
40. Sanchez-Martinez, Alpuche-Lazcano, Stuible et al. (2024) "SARS-CoV-2 spike-based virus-like particles incorporate influenza H1/N1 antigens and induce dual immunity in mice" *Vaccine*
41. Schneider, Holm, Bang et al. (2018) "Point-of-care tests for influenza A and B viruses and RSV in emergency departments -indications, impact on patient management and possible gains by syndromic respiratory testing" *Euro Surveill*
42. Sender, Hentrich, Henriques-Normark (2021) "Virusinduced changes of the respiratory tract environment promote secondary infections with streptococcus pneumoniae" *Front. Cell Infect*
43. Smyk, Szydlowska, Szulc et al. (2022) "Evolution of influenza viruses-drug resistance, treatment options, and prospects" *Int. J. Mol. Sci*
44. Song, Chen, Hao et al. (2019) "Exploring two decades of research on classroom dialogue by using bibliometric analysis" *Comput. Educ*
45. Stowe, Tessier, Zhao et al. (2021) "Interactions between SARS-CoV-2 and influenza, and the impact of coinfection on disease severity: a test-negative design" *Int. J. Epidemiol*
46. Swets, Russell, Harrison et al.
47. Girvan (2022) "SARS-CoV-2 co-infection with influenza viruses, respiratory syncytial virus, or adenoviruses" *Lancet*
48. Tantawy, Barakat, Adly et al. (2015) "One-year prospective study of community acquired influenza and parainfluenza viral infections in hospitalized Egyptian children with malignancy: single center experience" *Pediatr. Hematol. Oncol*
49. Vilas, Peters, Van Dijken et al. (2022) "Influenza infection in ferrets with SARS-CoV-2 infection history"
50. Weidmann, Green, Berry et al. (2023) "Assessing respiratory viral exclusion and affinity interactions through co-infection incidence in a pediatric population during the 2022 resurgence of influenza and RSV" *Front. Cell. Infect. Microbiol*
51. (2025) "Influenza (seasonal)"
52. Xu, Cai, Yue et al. (2023) "Comparative effectiveness of oseltamivir versus peramivir for hospitalized children (aged 0-5 years) with influenza infection" *Int. J. Infect. Dis*
53. Xue, Zhu, Lei et al. (2023) "Evaluation of the FPMC respiratory panel for detection of respiratory tract pathogens in nasopharyngeal swab and sputum specimens" *Int. J. Infect. Dis*
54. Yi, Tan, Long et al. (2021) "Lopinavir/ritonavir is associated with pneumonia resolution in COVID-19 patients with influenza coinfection: a retrospective matched-pair cohort study" *J. Med. Virol*
55. Yue, Zhang, Xing et al. (2020) "The epidemiology and clinical characteristics of co-infection of SARS-CoV-2 and influenza viruses in patients during COVID-19 outbreak" *J. Med. Virol*
56. Zangari, Ortigoza, Lokken-Toyli et al. (2021) "Type I interferon signaling is a common factor driving streptococcus pneumoniae and influenza A virus shedding and transmission" *Commun. Biol*
57. Zhang, Jia, Hu et al. (2025) "Exploring the standardized detection and sampling methods of human nasal SARS-CoV-2 RBD IgA" *Front. Immunol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671144&blobtype=pdf | # Global, regional, and national burden of HIV and other sexually transmitted infections among women of childbearing age from 1990 to 2021
Xiaoyu Zhang, Chenglong Hu, Yao Liang, Wanguo Dong, Jian Gao, Yu Ji, Chang Cao, Wei Shi, Shuaijie Zhu, Heng Guo, Tianfeng Hua, Hui Li, Min Yang
## Abstract
Sexually transmitted infections (STIs) represent a significant public health burden, particularly among women of childbearing age. This study aims to analyze the global, regional, and national age-standardized rates of incidence, prevalence, mortality, and disability-adjusted life years (DALYs) for STIs among women aged 15-49 from 1990 to 2021. Data were sourced from the Global Burden of Disease 2021 database, and age-standardized rates were calculated using direct standardization methods. Temporal trends were evaluated through average annual percentage change (AAPC) via jointpoint regression analysis. In 2021, compared with other STIs, HIV/AIDS had the highest age-standardized mortality rate (12.98 per 100,000; 95%
S TIs remain a significant global public health concern, encompassing a wide range of diseases such as HIV and syphilis, which are primarily transmitted through sexual contact and, in some cases, through vertical transmission routes (1)(2)(3). The burden of STIs is particularly severe among individuals under 49 years of age (4)(5)(6). Women of childbearing age, especially those who are sexually active, are at increased risk of these infections due to their unique anatomical features. The vaginal mucosa is thin and easily penetrated by pathogens, and the female reproductive tract allows for easier upward spread of infections, increasing the risk of complications such as pelvic inflammatory disease and infertility (7). In addition to the immediate health risks, STIs can lead to severe consequences, including infertility, preterm birth, and the vertical transmission of infections (8)(9)(10). Despite the recognized importance of this issue, there is a lack of comprehensive assessment regarding the global and regional burden of STIs among women of childbearing age. This gap in understanding hinders effective disease control efforts. Furthermore, while HIV/AIDS has received substantial attention, other STIs may not be as prominently addressed (11,12).
The Global Burden of Disease, Injuries, and Risk Factors Study 2021 (GBD 2021) is a comprehensive database widely used in epidemiological research to assess the burden of disease, age-standardized incidence rates, and trends over time (13,14).
In this study, we extracted data on STIs, including HIV/AIDS, syphilis, chlamydial infection, gonococcal infection, genital herpes, and trichomoniasis, from the GBD 2021 database. We analyzed age-standardized rates of incidence, prevalence, mortality, and disability-adjusted life years (DALYs) and examined temporal trends in women of childbearing age from 1990 to 2021 at global, regional, and national levels.
## MATERIALS AND METHODS
## Data acquisition
GBD 2021 provides a comprehensive evaluation of the health impacts associated with 369 diseases, injuries, and disabilities, along with an analysis of 88 risk factors across 204 countries and territories over recent decades. At the regional level, the GBD 2021 categorizes regions into five quintiles based on the sociodemographic index (SDI) values. This aggregate measure reflects social and economic factors influencing health outcomes across different geographical areas. The database also identifies 21 geograph ically proximate regions for more localized analysis (5). The statistical methods and standardization techniques used in GBD 2021 have been previously described in other studies (5,15). In this study, we focused on data related to STIs, including HIV/AIDS, syphilis, chlamydial infection, gonococcal infection, genital herpes, and trichomoniasis. These rates are reported per 100,000 individuals. Our analysis specifically targeted females aged 15-49 years, a group categorized by the World Health Organization (WHO) as women of childbearing age [https://www.who.int/data/gho/indicator-metadata-regis try/imr-details/women-of-reproductive-age-(15-49-years)-population-(thousands)]. Data on incidence, prevalence, mortality, and DALYs were extracted from GBD 2021, covering the years 1990 to 2021. The data set includes global, regional, and national data from the five SDI regions, 21 GBD regions, and 204 countries and territories.
## Data analysis
To enhance data comparability and exclude the potential influence of population structure, age standardization was performed. In this study, we applied a direct standardization method using the age structure of the standard world population. This method adjusts for population differences to analyze age-standardized rates of the incidence, prevalence, mortality, and DALYs. The estimation methods have been described in previous studies (16)(17)(18). To evaluate trends in disease burden over time, we calculated the AAPC between 1990 and 2021 using jointpoint regression models. An increasing trend was identified when both the estimated AAPC and its lower 95% confidence interval (CI) exceeded 0. Conversely, a decreasing trend was identified when both the estimated AAPC and its upper 95% CI were below 0. If the AAPC did not meet either criterion, the trend was considered stable throughout the study period. Data analysis was performed using R software (version 4.3.2).
## RESULTS
## HIV/AIDS
In 2021, the estimated age-standardized rates for incidence, prevalence, mortality, and DALYs in women of childbearing age were as follows: 36.99 (32. S1). Between 1990 and 2021, the AAPC in the age-standardized incidence rate was -1.88 (-2.17 to -1.59), suggesting a decline in incidence over the period. Conversely, the AAPCs for prevalence, mortality, and DALYs were 3.50 (3.35 3) for DALYs. Conversely, Burundi saw the most significant drop in the incidence rate (AAPC: -14.38 [-14.98 to -13.78]), while Burkina Faso had the most notable declines in the other three metrics, with AAPCs of -5.96 (-6.09 to -5.83) for prevalence, -8.21 (-8.86 to -7.55) for mortality, and -8.13 (-8.75 to -7.50) for DALYs (Table S3; Fig. 1).
## Syphilis
In 2021, the global age-standardized rates for syphilis were as follows: incidence 112.28 (57.32 to 190.60), prevalence 1,121.17 (684.72 to 1,721.54), mortality 0.01 (0.01 to 0.02), and DALYs 2.10 (1.55 to 2.85) (Table S1). Over the period from 1990 to 2021, the agestandardized incidence and prevalence rates remained relatively stable, with AAPCs of 0.26 (-0.10 to 0.63) and 0.09 (-0.03 to 0.21), respectively. In contrast, both age-standar dized DALYs and mortality rates declined significantly, with AAPCs of -1.14 (-1.32 to -0.96) for DALYs and -0.70 (-0.91 to -0.48) for mortality (Table S2).
At the regional level in 2021, central Sub-Saharan Africa exhibited the highest rates of incidence and prevalence, with values of 1,210.05 (636.64 to 1,997.65) and 4,689.77 (2,830.28 to 7,249.79), respectively. Eastern Sub-Saharan Africa had the highest mortality and DALY rates at 0.059 (0.030 to 0.127) and 8.41 (5.70 to 12.95), respectively (Table S1). From 1990 to 2021, tropical Latin America saw the most significant increase in agestandardized incidence and prevalence, with AAPCs of 1.27 (0.91 to 1.64), and 1.26 (0.96 to 1.56), respectively. The Caribbean showed the largest increase in age-standardized mortality and DALYs, with AAPCs of 1.19 (0.55 to 1.83) for mortality and 0.75 (0.32 to 1.18) for DALYs (Table S2).
In 2021, the highest age-standardized incidence rate was observed in Equatorial Guinea (1,506.26 S3). Between 1990 and 2021, the countries with the largest increases in age-standardized incidence, prevalence, mortality, and DALY rates were Brazil, Mongolia, Kuwait, and Dominica, respectively. The corresponding AAPCs were as follows: incidence 1.29 (0.94 to 1.65), prevalence 3.37 (1.99 to 4.77), mortality 13.06 (8.49 to 17.83), and DALYs 4.29 (4.06 to 4.53). In contrast, the most notable decreases in age-standardized incidence, preva lence, mortality, and DALY rates were observed in Malawi, Mozambique, the Northern Mariana Islands, and Armenia, respectively. The corresponding AAPCs were as follows: incidence -2.54 (-2.85 to -2.23), prevalence -3.52 (-5.42 to -1.59), mortality -7.18 (-9.01 to -5.30), and DALYs -6.02 (-7.85 to -4.15) (Table S3; Fig. 2).
## Chlamydial infection
In 2021, the global age-standardized rates of incidence, prevalence, mortality, and DALYs in women of childbearing age were as follows: incidence 5,179.06 (2,938.63 to 8,417.46), prevalence 4,570.25 (2,705.08 to 7,271.25), mortality 0.025 (0.015 to 0.037), and DALYs 4.58 (3.09 to 6.76) (Table S1). Between 1990 and 2021, the age-standardized incidence and prevalence showed no significant change (AAPCs: 0.10 [-0.11 to 0.30]; 0.10 [-0.09 to 0.30]), while the mortality and DALY rates declined (AAPCs: -1.12 [-1.30 to -0.95]; -0.27 [-0.35 to -0.20]) (Table S2).
In 2021, among the 21 GBD regions, Oceania had the highest age-standardized rates of incidence 13,457.05 (7,805.12 to 21,350.70) and prevalence 11,416.49 (6,706.78 to 17,990.07) (Table S1). eastern Sub-Saharan Africa exhibited the highest rates of mortality 0.112 (0.055 to 0.237) and DALYs 11.00 (6.54 to 19.81) (Table S1). From 1990 to 2021, high-income North America saw the most pronounced increases in age-standardized incidence (AAPC: 0.59 [0.38 to 0.81]) and prevalence (AAPC: 0.37 [0.16 to 0.57]). Some regions, including the Caribbean and tropical Latin America, saw increases, with the most S2).
At the national level in 2021, Fiji reported the highest age-standardized incidence and prevalence rates of chlamydial infection, at 17,317.70 (10,149.13 to 27,142.43) for incidence and 14,617.66 (8,645.71 to 22,791.91) for prevalence. South Sudan recorded the highest mortality and DALY rates, at 0.180 (0.060 to 0.520) for mortality and 15.65 (6.58 to 37.42) for DALYs. Between 1990 and 2021, the United States experienced the most significant increase in age-standardized incidence rates (AAPC: 0.63 [0.44 to 0.81]), while the United Kingdom led in the growth of prevalence rates (AAPC: 0.85 [0.74 to 0.96]). Kuwait showed the largest increase in mortality rates (AAPC: 11.72 S4; Fig. 3).
## Gonococcal infection
In 2021, the global age-standardized rates of gonococcal infection in women of child bearing age were estimated as follows: incidence 1,430.11 (857.81 to 2,239.55), preva lence 1,008.52 (625.98 to 1,548.20), mortality 0.009 (0.005 to 0.013), and DALYs 1.24 (0.85 to 1.84) (Table S1). Between 1990 and 2021, all metrics exhibited declining trends, as reflected in their AAPCs: -0.46 (-0.49 to -0.42) for incidence, -0.46 (-0.49 to -0.43) for prevalence, -1.16 (-1.34 to -0.97) for mortality, and -0.75 (-0.84 to -0.67) for DALYs (Table S2).
In 2021, Oceania recorded the highest age-standardized rates for both incidence (8,770.75 [4,521.47 S1). From 1990 to 2021, Oceania reported the largest increases in incidence (AAPC: 0.34 [0.32 to 0.36]) and prevalence (AAPC: 0.33 [0.30 to 0.35]). The Caribbean exhibited the most pronounced rise in mortality (AAPC: 1.18 [0.55 to 1.83]), while tropical Latin America showed the highest growth in DALY rates (AAPC: 1.20 [1.01 to 1.39]) (Table S2).
Across nations, the highest age-standardized rates for gonococcal infection in 2021 were observed in Papua New Guinea for both incidence 10,495.59 (5,402.89 to 18,253.83) and prevalence 7,041.67 (3,647.84 to 12,214.27). South Sudan recorded the highest rates for mortality and DALYs, 0.063 (0.019 to 0.179) and 5.64 (2.54 to 13.34), respectively. From 1990 to 2021, the most significant growth in age-standardized rates was recorded in Norway for incidence (AAPC: 0.41 [0.38 to 0.45]), Sri Lanka for prevalence (AAPC: 0. S4; Fig. 4).
## Trichomoniasis
Globally, in 2021, the estimated age-standardized rates of trichomoniasis were 6,709.73 (3,676.25 to 10,839.25) for incidence, 5,552.21 (2,918.06 to 9,234.17) for prevalence and 10.54 (3.51 to 24.28) for DALYs (Table S1). Over the period from 1990 to 2021, all three metrics exhibited upward trends, with AAPCs of 0.27 (0.25 to 0.30) for incidence, 0.24 (0.19 to 0.28) for prevalence, and 0.24 (0.19 to 0.28) for DALYs (Table S2).
Regionally, southern Sub-Saharan Africa reported the highest age-standardized rates in 2021, with incidence at 21,438.37 (12,145.47 to 33,847.43), prevalence at 19,947.23 S1). From 1990 to 2021, most regions showed stable or declining trends in these metrics. Western Sub-Saharan Africa showed the most notable increase in age-standardized incidence rates (AAPC: 0.18 [0.03 to 0.32]). However, no upward trends in prevalence or DALY rates were observed across any GBD regions during this period (Table S2).
In 2021, Zimbabwe reported the highest age-standardized incidence rate of trichomoniasis (22,426.64 [13,111.81 to 34,757.76]), while Zambia recorded the highest rates for both prevalence (21,910.47 [12,391.83 to 34,278.10]) and DALYs (41.39 [14.77 to 94.31]). From 1990 to 2021, the most significant decline in the age-standardized incidence rate was observed in Iraq (AAPC: -0.87 [1.06 to -0.67]). Among the other metrics, Burkina Faso showed the most notable decline, with the AAPC of -1.34 (-1.43 to -1.25) for prevalence and -1.36 (-1.44 to -1.28) for DALYs. Conversely, Kenya experi enced the most prominent increases across these three metrics, with incidence rising at an AAPC of 0.56 (0.38 to 0.73), prevalence at 0.74 (0.53 to 0.95), and DALYs at 0.72 (0.58 to 0.85) (Table S5; Fig. 5).
## Genital herpes
Globally, in 2021, the estimated age-standardized rates for genital herpes were 1,222.71 (861.48 to 1,650.15) for incidence, 17,137.09 (13,485.32 to 21,121.75) for prevalence, and 4.78 (2.02 to 9.63) for DALYs (Table S1). Between 1990 and 2021, these rates showed increasing trends, with AAPCs 0.24 (0.21 to 0.27) for incidence, 0.18 (0.13 to 0.22) for prevalence, and 0.18 (0.16 to 0.21) for DALYs (Table S2).
Regionally, central Sub-Saharan Africa recorded the highest age-standardized rates in 2021, with incidence at 3,099.41 (2,294.47 S2).
Equatorial Guinea had the highest age-standardized rates in terms of incidence, prevalence, and DALYs, with values of 3,124.69 (2,304.16 to 4,072.00), 57,776.05 (48,415.58 to 67,132.42), and 15.54 (6.45 to 31.19), respectively. From 1990 to 2021, countries such as Czechia, Eswatini, and India showed the most significant increases in these metrics over the years, with incidence AAPC of 0.53 (0.46 to 0.59) for Czechia, prevalence AAPC of 0.39 (0.37 to 0.40) for Eswatini, and DALY AAPC of 0.40 (0.34 to 0.46) for India. On the other hand, the Republic of Korea exhibited the most notable decline across the three metrics. The AAPCs were as follows: incidence -1.62 (-1.71 to -1.52), prevalence -2.12 (-2.23 to -2.02), mortality -7.18 (-9.01 to -5.3), and DALYs -2.06 (-2.19 to -1.94) (Table S5; Fig. 6).
## SDI regions
Over the past several decades, low-SDI regions have borne the highest burden of STIs in terms of incidence, prevalence, mortality, and DALYs, despite gradual improvements over time. In 2021, the age-standardized rates of incidence (9,418.29 vs 19,719.63), prevalence (21,057.11 vs 46,638.54), mortality (0.54 vs 36.54), and DALYs (54.66 vs 2,326.23) in high-and low-SDI regions, respectively, showed persistent disparities (Fig. 7). From 1990 to 2021, STI metrics generally decreased in low-SDI regions, while incidence and prevalence increased in low-middle SDI regions, and both mortality and DALYs increased in middle-SDI regions and high-middle SDI regions, respectively.
Regarding the distribution of specific STIs, trichomoniasis accounted for the highest proportion of new cases globally, comprising 46.91% of all STI cases, with rates ranging from 43.49% to 60.80% across different SDI regions. However, in high-middle SDI regions, chlamydial infection was the most prevalent infection in terms of new cases, accounting for 44.79% of cases in 2021 (Fig. S1). Genital herpes represented the highest proportion of cases globally (59.25%) and across the five SDI regions, with rates ranging from 56.94% to 68.78%. Additionally, HIV/AIDS accounted for the largest share of deaths and DALYs, both globally and across all SDI regions (Fig. S2). The proportion of deaths attributed to HIV/AIDS ranged from 98.95% to 99.74%, and the proportion of DALYs ranged from 73.14% to 98.17% (Fig. S3 andS4).
## DISCUSSION
This study estimated the age-standardized incidence, prevalence, DALYs, and mortal ity rates of HIV/AIDS and five other STIs-syphilis, chlamydial infection, gonococcal infection, trichomoniasis, and genital herpes-among women of childbearing age using data from the GBD 2021. The analysis spans from 1990 to 2021, providing a comprehen sive overview at global, regional, and national levels, along with an assessment of how these rates have evolved.
Women of childbearing age with STIs face not only the risk of vertical transmission but also an increased likelihood of infertility, gestational hypertension, and adverse pregnancy outcomes, such as preterm delivery (8,19,20). While the mortality and DALYs associated with STIs other than HIV/AIDS are generally lower, the incidence and prevalence of these infections remain high globally, warranting continued attention and intervention for women of reproductive age. Between 1990 and 2021, the disease burden of HIV/AIDS, trichomoniasis, and genital herpes showed an upward trend, with the exception of HIV/AIDS incidence rates, which remained stable or decreased. Conversely, syphilis, chlamydia, and gonococcal infections either alleviated or stabilized over the same period. Among all STIs, genital herpes had the highest prevalence in women of childbearing age. Genital herpes, unlike trichomoniasis, is a leading cause of neonatal transmission, and its management remains a challenge due to the lack of effective treatment to prevent vertical transmission (21,22). Furthermore, many cases of genital herpes are asymptomatic, complicating efforts at early detection and prevention. As genital herpes is associated with high mortality rates among neonates, its prevention, especially in the context of maternal transmission, remains a significant challenge (23).
By 2021, significant disparities in the burden of STIs are evident across regions, countries, and territories. Low-SDI regions experience a particularly acute disease burden, with many metrics showing increasing trends, consistent with findings in other populations (15,24). These patterns may stem from limited access to prevention and treatment measures, such as HIV pre-exposure prophylaxis, pregnancy testing, and timely interventions to prevent vertical transmission (25,26). The prevalence and predominant types of STIs vary across regions, reflecting distinct socioeconomic and healthcare challenges.
Between 1990 and 2021, HIV/AIDS metrics showed significant increases in several regions, particularly in South Asia, where the rising burden may be attributed to factors such as drug abuse, sex trade, and inadequate healthcare infrastructure (27). Conversely, metrics for other STIs, such as gonococcal infection, demonstrated a declining trend in many SDI regions, likely due to improved detection methods and preventive measures (28). Given that early-stage STIs are often asymptomatic, expanding targeted screen ing programs, particularly in resource-limited regions, is essential for effective disease control (27).
This study provides a comprehensive quantitative analysis of the burden of STIs among women of childbearing age at global, regional, and national levels, addressing a critical gap in prior research. While previous studies often focused on specific regions or individual STIs, this analysis highlights trends in multiple infections over time (29,30). Different regions need to prioritize prevention and control measures, as well as screening for STIs among women of childbearing age, based on the burden and changing trends of STIs in their respective regions.
This study has some limitations. First, the analysis is limited to six STI diseases, including HIV, and does not encompass the full spectrum of STIs. Second, data on gender minorities were not included, such as transgender and non-binary individuals, as the GBD 2021 only provided binary sex data (male/female), although these groups may face distinct STI risks. Third, the disease burden in less economically developed regions may have been underestimated due to gaps in healthcare data and infrastructure. Highincome countries often utilize high-sensitivity diagnostic modalities, such as PCR testing and genomic sequencing, whereas low-income countries primarily rely on syndromic management or rapid diagnostic tests with limited accuracy. Inadequate testing capacity and restricted access to healthcare services contribute to a high proportion of undiag nosed and unreported cases (31). Moreover, sociocultural factors further exacerbate these surveillance gaps, including limited sexual health education, persistent stigma surrounding STIs, and structural gender inequalities that hinder healthcare-seeking behavior (32). Finally, the term "global rates" refers to estimates based on the 204 countries and territories covered by the GBD 2021, which, although broadly representa tive, may not include every region of the world.
In conclusion, this study provides age-standardized rates and trends for several common STIs among women of reproductive age from 1990 to 2021. The findings reveal substantial regional variation in the burden of these infections, underscoring the need for targeted screening strategies tailored to local conditions. Enhanced prevention and control efforts, particularly in resource-limited areas, are critical to mitigating the spread and impact of STIs globally.
## References
1. Ghosn, Taiwo, Seedat et al. (2018) "HIV" *Lancet*
2. Korenromp, Rowley, Alonso et al. (2019) "Global burden of maternal and congenital syphilis and associated adverse birth outcomes-estimates for 2016 and progress since 2012" *PLoS One*
3. Muzny (2018) "Why does trichomonas vaginalis continue to be a "neglected" sexually transmitted infection?" *Clin Infect Dis*
4. Zheng, Yu, Lin et al. (2022) "Global burden and trends of sexually transmitted infections from 1990 to 2019: an observational trend study" *Lancet Infect Dis*
5. Vos, Lim, Abbafati et al. (2019) "Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study" *The Lancet*
6. Chemaitelly, Finan, Racoubian et al. (2024) "Estimates of the incidence, prevalence, and factors associated with common sexually transmitted infections among Lebanese women" *PLoS One*
7. Van Gerwen, Muzny, Marrazzo (2022) "Sexually transmitted infections and female reproductive health" *Nat Microbiol*
8. Gao, Liu, Yang et al. (2021) "Association of maternal sexually transmitted infections with risk of preterm birth in the United States" *JAMA Netw Open*
9. Taylor, Adekanmbi, Zhang et al. (2023) "The impact of Neisseria gonorrhoeae mono-and coinfection on adverse pregnancy outcomes" *Open Forum Infect Dis*
10. Tsevat, Wiesenfeld, Parks et al. (2017) "Sexually transmitted diseases and infertility" *Am J Obstet Gynecol*
11. Grabowski, Mpagazi, Kiboneka et al. "Rakai Health Sciences Program. 2022. The HIV and sexually transmitted infection syndemic following mass scale-up of combination HIV interventions in two communities in southern Uganda: a population-based crosssectional study" *Lancet Glob Health*
12. Ong, Baggaley, Wi et al. (2019) "Global epidemiologic characteristics of sexually transmitted infections among individuals using preexposure prophylaxis for the prevention of HIV infection: a systematic review and metaanalysis" *JAMA Netw Open*
13. Cao, He, Sang et al. (2019) "Age-standardized incidence, prevalence, and mortality rates of autoimmune diseases in women of childbearing age from 1990 to" *Autoimmun Rev*
14. Dai, Alsalhe, Chalghaf et al. (2020) "The global burden of disease attributable to high body mass index in 195 countries and territories, 1990-2017: an analysis of the global burden of disease study" *PLoS Med*
15. Du, Liu, Liu (2019) "Global, regional, and national disease burden and attributable risk factors of HIV/AIDS in older adults aged 70 years and above: a trend analysis based on the global burden of disease study" *Epidemiol Infect*
16. Ferrari, Santomauro, Aali et al. (2024) "Global incidence, prevalence, years lived with disability (YLDs), disability-adjusted life-years (DALYs), and healthy life expectancy (HALE) for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021" *Lancet*
17. Li, Liang, Wang et al. (2024) "Global cancer statistics for adolescents and young adults: population based study" *J Hematol Oncol*
18. Huang, Wang, Liu et al. (2024) "Global trends in adolescent and young adult female cancer burden, 1990-2021: insights from the global burden of disease study" *ESMO Open*
19. Felske, Castillo, Kuret et al. (2022) "Comparing adverse neonatal and maternal outcomes of chlamydia, gonorrhoea, and syphilis infections and co-infections in pregnancy" *Paediatr Perinat Epidemiol*
20. Rac, Revell, Eppes (2017) "Syphilis during pregnancy: a preventable threat to maternal-fetal health" *Am J Obstet Gynecol*
21. Fu, Sun, Han et al. (2019) "Incidence trends of five common sexually transmitted infections excluding HIV from 1990 to 2019 at the global, regional, and national levels: results from the global burden of disease study" *Front Med*
22. Bhatta, Keyal, Liu et al. (2018) "Vertical transmission of herpes simplex virus: an update" *J Dtsch Dermatol Ges*
23. (2020) "Management of genital herpes in pregnancy" *Obstetrics Gynecol*
24. Zhang, Ma, Han et al. (2019) "Global, regional, and national burdens of HIV and other sexually transmitted infections in adolescents and young adults aged 10-24 years from 1990 to 2019: a trend analysis based on the global burden of disease Study" *Lancet Child Adolesc Health*
25. Chavula, Zulu, Hurtig (2022) "Factors influencing the integration of comprehensive sexuality education into educational systems in low-and middle-income countries: a systematic review" *Reprod Health*
26. Logie, Mackenzie, Malama et al. (2024) "Sexual and reproductive health among forcibly displaced persons in urban environments in low and middle-income countries: scoping review findings" *Reprod Health*
27. Shannon, Strathdee, Goldenberg et al. (2015) "Global epidemiology of HIV among female sex workers: influence of structural determinants" *Lancet*
28. Pérez, Fernández, Lago et al. (2024) "Addressing sexually transmitted infections due to Neisseria gonorrhoeae in the present and future" *Microorganisms*
29. Rowley, Vander Hoorn, Korenromp et al. (2016) "Chlamydia, gonorrhoea, trichomoniasis and syphilis: global prevalence and incidence estimates" *Bull World Health Organ*
30. Dev, Adhikari, Dongol et al. (2021) "Prevalence assessment of sexually transmitted infections among pregnant women visiting an antenatal care center of Nepal: Pilot of the World Health Organization's standard protocol for conducting STI prevalence surveys among pregnant women" *PLoS One*
31. Wi, Ndowa, Ferreyra et al. (2019) "Diagnosing sexually transmitted infections in resource-constrained settings: challenges and ways forward" *J Int AIDS Soc*
32. Nguyen, Eaton (2022) "Trends and country-level variation in age at first sex in sub-Saharan Africa among birth cohorts entering adulthood between 1985 and 2020" *BMC Public Health* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12669569&blobtype=pdf | # Combined pathological, microbiological and virological evaluation of vitreous aspirates: a retrospective evaluation of 374 vitrectomy specimens with non-neoplastic disorders
Mihaly Sulyok, Kristina Schmauder, Markus Schneider, Jonas Neubauer, Tina Ganzenmueller, Irina Bonzheim, Daniela Süsskind, Karl Ulrich Bartz-Schmidt, Silke Peter, Falko Fend
## Abstract
BACKGROUND: Intraocular infections pose substantial diagnostic challenges due to their varied aetiologies and complex inflammatory responses. The value of combining cytological, microbiological and virological assessments remains underexplored. This extensive retrospective study conducted at a German tertiary care university hospital aims to elucidate the correlation between different inflammatory patterns and detected pathogens in vitreous fluid samples. METHODS: A retrospective analysis of 374 non-neoplastic vitreous samples from 353 patients was performed, integrating cytological, microbiological and virological data with final clinical diagnoses. RESULTS: Cytological analysis showed 284 instances of lymphocytic uveitis and 76 of granulocytic endophthalmitis. Pathogens were identified in 46 out of 181 microbiologically tested samples; notable pathogens included Staphylococcus epidermidis, Staphylococcus aureus, Streptococcus pneumoniae, Tropheryma whipplei, Candida albicans and Toxoplasma gondii. Virological tests on 188 samples detected viral DNA in 42 cases, predominantly varicella zoster virus and Epstein-Barr virus, correlating well with clinical suspicions of retinitis. Interestingly, no pathogens were found in 67% of the lymphocytic uveitis cases, and multiple pathogens were detected simultaneously in seven instances, suggesting potential latent infections or reactivations. A significant pattern emerged correlating increased neutrophil counts with pathogen detection, highlighting a notable association (p =∠0.03) in a subset analysed for both neutrophil levels and pathogen presence. CONCLUSION: This study offers critical insights into the epidemiology of intraocular infections in Germany, underscoring the importance of comprehensive pathological assessments. It emphasizes the diagnostic value of the underlying inflammatory patterns for predicting pathogen presence and identifies notable cases of infections, including rare pathogens like Tropheryma whipplei.
## BACKGROUND
Uveitis is characterised by intraocular inflammation, frequently due to immune-mediated mechanisms or an infectious aetiology [1]. Despite the fact that approximately 30% of uveitis cases remain without identified aetiology, ophthalmologists need to exclude infectious uveitis as the cause of inflammation, as this directly impacts treatment options, and delay in diagnosis of infections may have catastrophic consequences [1][2][3]. Diagnosis of intraocular infections is often a challenging task, since numerous microorganisms may be causative, with protean manifestations [1]. The epidemiology of infectious uveitis, which comprises approximately 10-30% of all cases, varies across different regions and populations. The prevalence and distribution of certain causative pathogens can differ based on geographic location, socioeconomic factors and healthcare access [4,5]. Of note, data on the epidemiology of infectious uveitis/ endophthalmitis in Germany are scarce [6,7].
Vitreous fluid aspirate, obtained via vitrectomy, is crucial for diagnosing infectious uveitis, enabling targeted antimicrobial therapy and improving patient outcomes. By pinpointing specific pathogens, it supports definitive diagnoses and guides treatment decisions. Additionally, analysing the pathogen profile in uveitis cases enhances epidemiological understanding, aiding in disease surveillance and management [8].
Microbiological and virological laboratory analysis involves the identification of infectious agents such as bacteria, viruses, fungi or parasites, using mainly culture-based or molecular technologies such as PCR, while cytopathological examination assesses the cellularity and cellular composition and may allow direct provisional identification of some infectious agents. Data on the relative contributions of these approaches and their potential added value, however, are sparse.
In this large, retrospective single-centre study, we analysed the association between microbiological, virological and pathological findings and provided descriptive statistics on the epidemiology of infectious uveitis and endophthalmitis at a large German tertiary care university hospital. Furthermore, it was assessed whether morphological evaluation of vitreous aspirates may help to predict the presence of infectious uveitis.
## METHODS Patients
We identified uveitis cases from the electronic case record system (PAS-NET) of the Institute for Pathology and Neuropathology at the University of Tuebingen, using 'Glaskörperpunktat' (vitreous fluid aspirate) as the material type on the 10JAN2023. Cases with a pathologically confirmed diagnosis of vitreoretinal lymphoma or other malignancy were excluded. Basic demographic parameters and diagnoses were extracted, along with data on microbiological and virological findings obtained from the records of the Institutes of Medical Microbiology and Medical Virology of the University Hospital Tuebingen.
## Vitrectomy procedure
The indication for vitrectomy was based on the clinical and ophthalmological assessment by the treating physicians, usually for suspicion of infectious uveitis/endophthalmitis, lymphoma or uveitis not responding to appropriate therapy. After obtaining written consent, a 23 G vitrectomy was performed to collect a vitreous sample. The sample was taken either undiluted under air or diluted under balanced salt solution infusion, depending on the surgeon's decision. The central vitreous was removed during the sample collection, and at the end of the surgery, the formation of peripheral holes was excluded using circular identification.
## Diagnostic procedures for bacterial, fungal and parasitological detection
Microbiological analysis of vitreous aspirates consisted primarily of culture-based diagnostics. Vitreous specimens were cultivated in nutrient broth with whole liver for unselective enrichment of microorganisms. Additionally, specimens were streaked on 5% sheep blood agar (Becton Dickinson, Heidelberg, Germany) with incubation at room air and brainheart infusion agar plates (in-house production) with incubation under microaerophilic and anaerobic conditions. Agar plates and enrichment broth were incubated at 37 °C for 72 h. Suspected samples for fungal infections were cultivated at 30 °C on Sabouraud agar (in-house production; since 2021 Becton Dickinson, Heidelberg, Germany) for 14 days. Single colonies of grown microorganisms were identified by MALDI-TOF (Matrix-assisted laser desorption/ionisation-Time of flight) using the Microflex LT instrument (Bruker Daltonics, Germany). Culture negative specimens were analysed further by 16S-rRNA-PCR (details below) if sufficient specimen volume was available. Additional molecular diagnostics were performed depending on the differential diagnosis provided by the clinician. Toxoplasma gondii was detected by a conventional molecular in-house PCR assay until 2018 and afterward by a real-time PCR assay (T. gondii RT-PCR, Sacace Biotechnologies, Como, Italy) conducted on a Lightcycler 480 II instrument (Roche, Basel, Switzerland). Tropheryma whipplei was detected by a conventional molecular in-house assay until 2022 and afterward by real-time PCR (RealCycler® TRWH-GX, Progenie Molecular, Valencia, Spain). For undirected detection of bacterial DNA, an eubacterial conventional in-house PCR was performed targeting the V1-V3 region of the 16S rRNA genes (forward primer: AGAGTTTGATCCTGGCTCAG; reverse primer: TTACCGCGGCTGCTGGCAC), resulting in a fragment of approximately 500 base pairs in length. Subsequent Sanger sequencing was conducted following state-of-the-art standard diagnostic protocols. Analysis of 16S sequences was performed using the curated EzBioCloud database in the current updated version, respectively.
## Real-time PCR for virus detection
Real-time PCR assays for detection of HSV-, CMV-, VZV-or EBV-DNA were conducted on a Lightcycler 480 II (Roche) using laboratory-developed inhouse PCR assays (until 2021) or the CMV, EBV or HSV-1/2&VZV Argene® assay (Biomerieux, for samples collected after 2021), depending on the differential diagnosis and assay requests by the clinician. All assays included an internal control for sufficient nucleic acid extraction and absence of PCR inhibition. Results were reported qualitatively, but the cycle threshold (Ct) values (as far as available) can be used as a measure to semi-quantitatively assess the virus concentration in a sample.
## Cytopathological analysis
Cytopathological analysis of vitreous fluid aspirates was conducted following standard procedures. The native samples were centrifuged with the Shandon Cytospin4 (4 min, speed 1100 rpm). Depending on the volume, two to four cytocentrifugates were prepared on coated glass slides. A minimum of 100 μl and a maximum of 400 μl per glass slide were pipetted, depending on the cell density of the sample (volume in μl noted on the glass slide). Following fixation, one preparation was routinely stained with May-Grünwald-Giemsa after air drying, while the remaining preparations were left to air dry. Cytopathological analysis was performed by experienced ocular cytopathologists. The slides were examined under a light microscope at various magnifications to assess cellular morphology, presence of inflammatory cells, microorganisms and any evidence of atypical or malignant cells.
## Statistical analysis
Descriptive statistics were performed using R software (version 4.0.3). A subgroup analysis of samples with both microbiological, virological and cytopathological results were performed using ordinal logistic regression to assess the the association between presence/absence of detected pathogens (as the dependent variable) and the semiquantitatively assessed neutrophilic granulocytes. Visualisation was performed using the ggplot2 package and GraphPad Prism (version 10.1.1).
The study was conducted under the ethics approval no.639/2023BO2, issued by the local ethics committee of the University Hospital Tübingen. The study was conducted in accordance with institutional regulations; all patients had given consent permitting the use of their clinical data and samples. Anonymised data and the script of statistical analysis are available from the corresponding author on request.
## RESULTS
## Findings in cytopathological diagnostics
A total of 374 non-neoplastic vitreous fluid samples obtained from 353 patients between 1JAN2010 and 10JAN2023, were included in the analysis (Fig. 1), after exclusion of vitreoretinal lymphoma or other malignancy (n =∠67; mostly lymphomas and one melanoma metastasis). All cases were analysed cytologically. These samples included 9 cases in which a specific pathogen was detected by cytology and also microbiologically, including 2 cases of T. whipplei infection and one case of Leishmania infection, each with granulocytic inflammation patterns, and 4 cases of intraocular toxoplasmosis with lymphocytic inflammation patterns. Moreover, 2 intraocular Candida infections diagnosed by pathology (one case confirmed by microbiology) displayed a granulocytic inflammation pattern.
Altogether 284 cases of lymphocytic uveitis and 76 cases of granulocytic endophthalmitis were identified. The remaining 14 samples presented intraocular bleeding, predominantly histiocytic or granulomatous inflammation as cytopathological finding.
## Pathogen detection in microbiological and virological diagnostics
Among the 181 microbiologically analysed samples, in 46 (25%) positive samples the following microorganisms were identified: Staphylococcus epidermidis (n =∠15), Staphylococcus aureus (n =∠3), Streptococcus pneumoniae (n =∠2), T. whipplei (n =∠3), Streptococcus oralis (n =∠1), Streptococcus dysgalactiae (n =∠1), Klebsiella pneumoniae (n =∠1), Prevotella spp. (n =∠1), Candida albicans (n =∠6), and T. gondii (n =∠13). Additionally, 135 samples were found to be microbiologically negative (Table 1).
Of 374 samples, a total of 188 were sent for virological assessment. Viral nucleic acids were found in 42 specimens (22%).
In 93% of those samples, a lymphocytic inflammation pattern was determined by pathology. PCR detected varicella zoster virus (VZV, n =∠22), herpes simplex virus 1 (n =∠3), herpes simplex virus 2 (n =∠5), cytomegalovirus (n =∠7) and/or Epstein-Barr virus (n =∠10) genomes. For virus positive cases, the initially suspected diagnosis upon sample submission by the treating physicians to the virology lab frequently (n =∠18) was retinitis and/or retinal necrosis and we detected CMV-, HSV-or VZV-DNA in these samples. Among two patients with VZV-DNA detection in the vitreous fluid, preceding history of facial herpes zoster was reported. EBV was detected at low viral loads in most of the cases (see Supplementary Table 1), likely reflecting the presence of latent EBV infection of infiltrating immune cells or rather minor reactivations in seropositive individuals.
## Correlation of cytopathology, microbiology and virology
In 67% (105/158) of cases of lymphocytic uveitis diagnosed by pathology and further analysed by either microbiology or virology, no evidence of an infectious agent was found. In the other 53 (33%) specimens classified as lymphocytic uveitis a causative organism could be detected. Notably, in 72% (n=38) of those positive cases, viruses were detected. Among the remaining cases within this subgroup, T. gondii (n =∠10), T. whipplei (n =∠1), Prevotella spp. (n =∠1), S. epidermidis (n =∠1) and C. albicans (n =∠1) were identified. All detected pathogens are shown in Fig. 2 and Table 1.
In total, 140 samples were assessed by both microbiology and virology diagnostics. Multiple (i.e. means two) organisms have been detected from the same specimen in seven cases (5% of this subgroup) and consisted of the combinations of any pathogen with EBV, i.e. C. albicans ∫∠EBV (n =∠2), T. gondii ∫∠EBV (n =∠1), VZV ∫∠EBV (n =∠3) and CMV ∫∠EBV (n =∠1).
We semiquantitatively categorised neutrophil abundance in this subgroup into four distinct categories: 0, 1∫, 2∫, and 3∫. From the 140 samples, 128 were available for this subgroup analysis. Of those, the majority (100 samples; 78,1%) were classified in category 0 and in 37 out of the 100 cases (37%) pathogens were detectable. Category 1∫∠ included 4 samples (3.1% of total samples), and in 3 of these samples, pathogens have been detected. Category 2∫∠comprised 4 samples (3.1% of total samples), and all of these (100% of category 2∫) were tested positive for pathogens. The category with the highest neutrophil counts, 3 ∫∠, included 20 samples (15.6% of total samples), with 8 (40% of category 3∫) showing pathogen presence. Statistical analysis indicated a significant association between the level of neutrophil abundance and the detection of pathogens, confirmed by chi-squared test (p =∠0.03). The ordinal logistic regression analysis did not demonstrate a significant association between the semiquantitative categorisation of neutrophils and the likelihood of pathogen detection.
## Endophthalmitis and associated pathogens in our study
Out of the 25 cases with granulocytic endophthalmitis with positive microbiology, the clinical diagnosis was intravitreal surgical administration of medication and postoperative endophthalmitis in seven cases, mostly at external sites. Pathogens detected in those cases were S. epidermidis, S. aureus and S. oralis (Supplementary Fig. 1). Two cases were clinically classified as endogenous endophthalmitis, one due to pneumococcus endocarditis. Out of the six intraocular candida infections, two were considered endogenous endophthalmitis, panuveitis in one case, and endophthalmitis without further specification in the three remaining cases. Remarkably, all six individuals exhibited a predisposing condition, including five patients with cytotoxic or immunosuppressive therapy or radiotherapy and one with a history of intravenous drug abuse.
## Toxoplasmosis and M. Whipple in our study
Regarding the 14 cases of intraocular toxoplasmosis, one patient was HIV positive, and two others were suspected of having toxoplasmosis based on clinical findings. In the three patients with intraocular Whipple's disease (Fig. 3 and Supplementary Fig. 2), the diagnosis was not suspected clinically, but found by untargeted eubacterial 16S rRNA PCR in all patients and by PAS staining in two of them. In one of these cases, initially, CMV retinitis had been suspected clinically.
## Patient demographics
There was no significant difference between age/sex and aetiologic agent (p =∠0.44 Kruskal Wallis test and p =∠0.7 chi square test). The median patient age of our cohort was 66.4 years without major deviation between the different pathologic aetiologies of granulocytic endophthalmitis (median age 66.4 years) and lymphocytic uveitis (median age 66.8 years).
## DISCUSSION
The findings of our retrospective, single-centre analysis of a large cohort of patients with vitrectomy provide valuable insights into the spectrum of aetiological agents encountered in infectious uveitis in a large tertiary care ophthalmology centre in Germany. Furthermore, the correlation between cytopathology and virological and microbiological findings underlines the importance of a synoptic diagnostic approach and provides clues for an optimal handling in non-neoplastic vitrectomy specimens.
The differential diagnosis for uveitis/vitritis and endophthalmitis includes inflammatory, infectious, immunological and neoplastic disorders, notably vitreoretinal lymphoma. Vitrectomy serves diagnostic and therapeutic roles, with specimens often analysed cytologically and tested for microbes and viruses if infection is suspected. However, systematic studies on the diagnostic value of such a multi-pronged analytical approach in non-neoplastic uveitis/endophthalmitis are sparse, and there are only limited data available regarding the aetiology of nonneoplastic uveitis and endophthalmitis in Germany.
By conducting a large retrospective single-centre study, we aimed to address this knowledge gap and provide descriptive statistics on these conditions. Our findings shed light on the prevalence and distribution of various infectious agents, such as bacteria (e.g. S. epidermidis, S. aureus), fungi (e.g. C. albicans), viruses (e.g. VZV) and parasites (e.g. T. gondii). Our findings are in line with the literature [9,10], especially with regard to the dominant role of toxoplasmosis; on the other hand [11], other reportedly prevalent causes of endophthalmitis such as syphilis and tuberculosis were not identified in our series. The lack of ocular infections due to Mycobacterium tuberculosis in our cohort likely is a reflection of the epidemiological situation, rather than due to methodological reasons. Indeed, national tuberculosis incidence in the examined time period varied between 5.0-7.2/ 100.000 inhabitants and was below average in our region [12]. Moreover, tuberculous uveitis represents a rare manifestation among clinical findings in tuberculosis patients.
Infectious uveitis contributes approximately from 10 to 28% of uveitis cases; greatly depending on the geographic region and several other factors [1,[12][13][14]. This is roughly in line with findings in our study, where in 33% of vitreous aspirates classified as lymphatic uveitis, a pathogen could be detected. In the United States, one study identified histoplasmosis followed by viral agents (mainly HSV) and toxoplasmosis as frequently detected pathogens [13]. Another large study from India showed viral agents (HSV, VZV and CMV) as leading causes, followed by tuberculosis and only a low prevalence of toxoplasmosis (with a declining proportion from 1.69 to 0.66%). Increasing numbers of Toxocara infections, endogenous uveitis and syphilis were also reported [14]. In immunocompromised patients, opportunistic infections like invasive mould (e.g. Aspergillus), Pneumocystis, atypical mycobacteria may also cause devastating intraocular infections. Among cases of granulocytic endophthalmitis where an infectious pathogen could be identified in our study, some were associated with antecedent intravitreal surgical procedures, emphasising the need for vigilant aseptic techniques during and following such interventions [15]. In this subpopulation, coagulase-negative Staphylococci were notably prevalent. The presence of endocarditis-related Pneumococcus and other endogenous sources, such as systemic candida infections, further illustrates the rare but potential origins of ocular infections by septic spreading. Proactive assessment for ocular involvement has to be taken into account when diagnosing systemic infections and concomitant visual impairment [16,17]. Intraocular Candida infections present a unique diagnostic challenge. This underscores the necessity for comprehensive clinical assessments, including consideration of underlying systemic conditions.
Interestingly, in three cases, we detected the bacterium T. whipplei, responsible for a rare systemic condition that predominantly affects the gastrointestinal tract but can manifest in various extraintestinal sites, including the eyes [18]. The pathophysiology of Whipple-uveitis remains poorly understood, but it is hypothesised that bacterial infiltration and subsequent immune response contribute to ocular inflammation. Epidemiological data from Germany indicate an incidence of Whipple's disease at approximately 0.5-1.0 cases per million per year, with uveitis being a less commonly reported manifestation [19]. In a retrospective study analysing German patient data, ocular involvement was observed in about 10% of Whipple's disease cases [20]. Our findings underscore the importance of Whipple uveitis. Early diagnosis and treatment are crucial to prevent irreversible ocular damage and systemic complications [21]. Future research should focus on elucidating the molecular mechanisms driving ocular involvement in Whipple's disease and optimising therapeutic strategies to improve patient outcomes.
Our study emphasises the importance of combined pathological, virological and microbiological examination of vitreous aspirates in the diagnosis of intraocular infections. The absence of granulocytes in vitreous aspirate specimens was identified as a highly indicative marker for a negative bacteriological or fungal result, thereby assisting in the differentiation between bacterial/ fungal and non-bacterial/non-fungal causes of endophthalmitis. In contrast, a pattern of granulocytic inflammation identified by histopathology resulted in the detection of causative bacterial or fungal organisms in only around 50% of cases, either by culture or by PCR within our cohort (27 cases with bacterial pathogens, 5 cases with Candida, 5 cases with viruses). Neutrophilic granulocytes play a crucial role in averting the expansion of mainly bacterial or fungal infections and are seen frequently in cases of postoperative endophthalmitis [22][23][24]. However, as findings from this cohort reveal, herpes simplex virus 1 or 2 infections can lead to invasion of granulocytes as well. A retrospective cytopathologic study of acute retinal necrosis cases also found polymorphonuclear cells alongside lymphocytes and macrophages in nearly one-third of vitrectomy samples in confirmed HSV or VZV infection [25]. This indicates both innate and adaptive immune activation, highlighting the diagnostic challenge of granulocyte-rich inflammation in viral uveitis. Moreover, recruitment of neutrophilic granulocytes in sterile tissue injury or noninfectious inflammatory diseases are commonly known and might play a confounding role when using granulocytic inflammation as a marker for bacterial or fungal infections [26,27].
Concurrent high numbers of neutrophils (3∫) appear to reduce successful pathogen detection in our cohort and are most likely reflecting intensified inflammation processes resulting in degradation of vital organisms, as well as DNA digestion, aggravating thereafter cultural and molecular diagnostics. Moreover, purulent inflammation often necessitating dilution of specimens before vitreous extraction reduces pathogen load and therefore detection sensitivity. Indeed, French data on postoperative bacterial endophthalmitis reveal culture-positive detection of microorganisms in up to 64.7% in directly obtained, undiluted vitreous aspirate specimens [28]. Obtained sample volume remains a further known key factor for sensitive pathogen detection, which can be challenging considering the anatomically small extent within the ocular bulb. To further address this diagnostic issue in sensitive pathogen detection, untargeted molecular strategies like evolving Next-generation sequencing technologies could be of help and have already shown to be of beneficial use in uncovering intraocular infections, which however, need further contextual evaluation in the future [29].
While the majority of cases with lymphocytic uveitis did not exhibit an infectious aetiology, our study revealed cases with, probably, reactivating virus infections were identified. These were predominantly caused by VZV, followed by CMV and HSV infections, which is in line with the literature [30,31]. In a number of vitreous specimens, EBV DNA was detected. However, high Ct-values in EBV real-time PCR suggest very low viral loads and likely represent the presence of rare latently infected lymphocytes in uveitis of other cause, rather than active disease. Notably, EBV was frequently detected together with other pathogens, e.g. concurrent toxoplasmosis. While minor EBV reactivation is possible, especially in a single case with a Ct-value of 24.8, EBV uveitis is only rarely reported in the literature [32]. This highlights that virus detection, especially in the case of EBV, does not necessarily indicate causality.
Besides viral infections, T. gondii was identified as a frequent pathogen presenting with lymphatic infiltration of the vitreous body. 1 out of 14 of the discordant lymphocytic uveitis cases with positive microbiological result and without identified pathogen in the pathological report initiated reviewing of the pathology slides and led to retrospective identification of a Toxoplasma bradycyst. This highlights the importance of considering infectious causes in a multidisciplinary fashion, even in cases presenting as subacute, suspectedly non-infectious uveitis.
Vitreous opacities encompass a broad diagnostic range, from benign to life-threatening conditions. As emphasised by Coupland et al. [33], accurate aetiological classification and diagnostic yield from vitreous biopsies depend critically on proper specimen handling, timely fixation, and the application of specialised staining and molecular techniques. A prospective study using appropriate mucolytic pre-treatment and fixation in PreservCyt demonstrated that diagnostic material could be obtained in over 90% of vitreous biopsies, enabling both morphological and immunocytochemical characterisation of inflammatory and neoplastic processes [34].
## References
1. Agrawal, Thng, Gupta et al. (2023) "Infectious uveitis: conversations with the experts" *Ocul Immunol Inflamm*
2. Callegan, Parke, Jett et al. (2002) "Bacterial endophthalmitis: epidemiology, therapeutics, and bacterium-host interactions" *Clin Microbiol Rev*
3. Sugita, Ogawa, Shimizu et al. (2013) "Use of a comprehensive polymerase chain reaction system for diagnosis of ocular infectious diseases" *Ophthalmology*
4. García-Aparicio, De Yébenes, Otón et al. (2021) "Prevalence and incidence of uveitis: a systematic review and meta-analysis" *Ophthalmic Epidemiol*
5. Lin (2015) "Infectious uveitis" *Curr Ophthalmol Rep*
6. Leuchtenberger, Fleckenstein, Heger et al. (2005) "Epidemiology of secondary uveitis in Germany" *Invest Ophthalmol Vis Sci*
7. Ness, Boehringer, Heinzelmann (2017) "Intermediate uveitis: pattern of aetiology, complications, treatment and outcome in a tertiary academic center" *Orphanet J Rare Dis*
8. Damato, Angi, Romano et al. (2012) "Vitreous analysis in the management of uveitis" *Mediators Inflamm*
9. Tsirouki, Dastiridou, Symeonidis et al. (2018) "A focus on the epidemiology of uveitis" *Ocul Immunol Inflamm*
10. Keynan, Finkelman, Lagacé-Wiens (2012) "The microbiology of endophthalmitis: global trends and a local perspective" *Eur J Clin Microbiol Infect Dis*
11. Koch-Institut "SurvStat@RKI 2.0"
12. Grajewski, Caramoy, Frank et al. (2015) "Spectrum of uveitis in a German tertiary center: review of 474 consecutive patients" *Ocul Immunol Inflamm*
13. Choi, Rivera-Grana, Rosenbaum (2019) "Reclassifying idiopathic uveitis: lessons from a tertiary uveitis center" *Am J Ophthalmol*
14. Dogra, Singh, Agarwal et al. (2017) "Epidemiology of uveitis in a tertiary-care referral institute in North India" *Ocul Immunol Inflamm*
15. Stem, Rao, Lee et al. (2019) "Predictors of endophthalmitis after intravitreal injection. ophthalmol" *Retina*
16. Kullberg, Verweij, Akova et al. (2011) "European expert opinion on the management of invasive candidiasis in adults" *Clin Microbiol Infect*
17. Breazzano, Bond, Bearelly et al. (2022) "American academy of ophthalmology recommendations on screening for endogenous candida endophthalmitis" *Ophthalmology*
18. Rasool (2024) "Ophthalmic manifestations of Whipple's disease" *Curr Opin Ophthalmol*
19. Desnues, Moussawi, Fenollar (2010) "New insights into Whipple's disease and Tropheryma whipplei infections" *Microbes Infect*
20. Schneider, Moos, Loddenkemper et al. (2008) "Whipple's disease: new aspects of pathogenesis and treatment" *Lancet Infect Dis*
21. Salman, Salomao, Dalvin et al. (2023) "Ocular Whipple disease: cases diagnosed over four decades" *Ocul Immunol Inflamm*
22. Leveziel, Knoeri, Errera et al. (2022) "Anterior chamber tap cytology in acute postoperative endophthalmitis: a case-control study" *Br J Ophthalmol*
23. Urban, Backman (2020) "Eradicating, retaining, balancing, swarming, shuttling and dumping: a myriad of tasks for neutrophils during fungal infection" *Curr Opin Microbiol*
24. Honda, Uehara, Matsumoto et al. (2016) "Neutrophil left shift and white blood cell count as markers of bacterial infection" *Clin Chem Acta*
25. Hojjatie, Shantha, Keefe et al. (2022) "Cytopathology of vitreous specimens in acute retinal necrosis" *Ocul Immunol Inflamm*
26. Wang, Hossain, Thanabalasuriar et al. (2017) "Visualizing the function and fate of neutrophils in sterile injury and repair" *Science*
27. Wang, Arase (2014) "Regulation of immune responses by neutrophils" *Ann N Y Acad Sci*
28. De Liano, Ventura, Salcedo-Villanueva et al. (2017) "Sensitivity, specificity and predictive values of anterior chamber tap in cases of bacterial endophthalmitis"
29. Doan, Wilson, Crawford et al. (2016) "Illuminating uveitis: metagenomic deep sequencing identifies common and rare pathogens" *Genome Med*
30. Takase, Kubono, Terada et al. (2014) "Comparison of the ocular characteristics of anterior uveitis caused by herpes simplex virus, varicella-zoster virus, and cytomegalovirus" *Jpn J Ophthalmol*
31. Radosavljevic, Agarwal, Chee et al. (2022) "Epidemiology of viral induced anterior uveitis" *Ocul Immunol Inflamm*
32. Silpa-Archa, Sriyuttagrai, Foster (2022) "Treatment for Epstein-Barr virus-associated uveitis confirmed by polymerase chain reaction: efficacy of anti-viral agents and a literature review" *J Clin Virol*
33. Coupland (2008) "The pathologist's perspective on vitreous opacities" *Eye*
34. Van Ginderdeuren, Calster, Stalmans (2014) "Van den Oord J. A new and standardized method to sample and analyse vitreous samples by the Cellient® automated cell block system" *Acta Ophthalmol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12618335&blobtype=pdf | # Correction: The presence of human polyomavirus JC (JCPyV) in pediatric brain tumors: a plausible trigger in Wnt/β-catenin pathway
Valeria Pietropaolo, Sara Passerini, Sara Messina, Marta De Angelis, Lucia Nencioni, Francesca Gianno, Manila Antonelli |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12805835&blobtype=pdf | # Quantification and diagnostic relevance of blood and hememediated inhibition of prion detection by RT-QuIC
Robert Piel, David Schneider
## Abstract
Prion diseases are characterized by misfolding of prion protein (PrP) from correctly folded PrP C to a disease-associated form, PrP D . Real-time quaking-induced conversion (RT-QuIC) detects prions by "seeding" reaction mixtures, which contain recombinant PrP, with samples suspected to contain prions, resulting in amplification of misfolded PrP. The assay is sensitive to inhibition by tissue constituents, including blood. Heme, a cofactor of hemoglobin (Hb), has been shown to bind PrP in an isoform-specific manner and to affect the stability of other pathogenic amyloids. Herein, tissue samples from scrapie-positive sheep were used to seed RT-QuIC reactions in the presence of heme-as free hemin, as a cofactor of Hb, and as present in whole blood. At equivalent heme concentrations, the inhibitory action of free heme was the least and that of blood the greatest, suggesting other components of Hb and whole blood have additional inhibitory actions. We also demonstrate that this inhibition of RT-QuIC acts through disruption of the recombinant PrP assay substrate, rather than destruction of PrP D seeds. Lastly, heme concentrations were measured in several ruminant tissues. Heme levels exceeded inhibitory thresholds in nearly all types of intact tissue but were reduced below inhibitory levels at a 1:1,000 dilution of most tissue types, with whole blood being one of a few notable exceptions. Our results suggest that detection of PrP D seeding activity is not precluded by exposure to heme in tissue samples, but that the final heme concentration introduced into the RT-QuIC assay mixture is the critical factor that impacts detection sensitivity. IMPORTANCE Real-time quaking-induced conversion (RT-QuIC) is an ultrasensitive amplification assay for the detection of prions. The assay has shown exceptional performance in optimal laboratory conditions, on par with bioassay, and far surpassing current immunoassay diagnostics. However, efforts to apply RT-QuIC as a real-world diagnostic have been hampered by inconsistencies and unexpectedly low sensitivity in some field samples. This study aims to quantify and characterize the mechanism of inhibition from blood and its constituent parts, hemoglobin and heme-omnipresent components of most sample types. Such systematic evaluations of RT-QuIC inhibitory factors represent necessary steps toward the consistent and sensitive performance necessary for a field-applicable diagnostic assay. KEYWORDS hemoglobin, heme, blood, real-time quaking-induced conversion, prions I n prion diseases, which include scrapie in sheep and goats, chronic wasting disease (CWD) in cervids, and various others in both agricultural animals and humans, initial transmission is followed by a prolonged asymptomatic incubation period, which can often last years. As infectious prions are shed during preclinical infection (1, 2), these animals represent a critical management challenge to efforts targeting highly transmis sible prion diseases. These include the final stages of eradication for classical scrapie in the USA as well as responses to the continuing expansion of CWD across North
America. Currently, the diagnosis of prion disease in small ruminants and cervid species is accomplished by immunoassay using lymphoid tissues typically collected postmor tem. A reliable and highly sensitive assay for early-stage prion disease using peripheral tissues or blood would greatly aid in disease management efforts.
A defining feature of prion infection is the replication of the infectious unit (the prion) by templated misfolding of natively expressed prion protein (PrP) from its normal "cellular" form (PrP C ) to a misfolded "disease" form (PrP D ) (3). Amplification assays such as real-time quaking-induced conversion (RT-QuIC) exploit this inherent mechanism of replication to enable detection of minuscule amounts of misfolded protein that may otherwise elude conventional methods such as immunoassay. In RT-QuIC, enhanced detection is accomplished by exposing a reaction mixture containing monomeric, metastable recombinant prion protein (rPrP) substrate to test samples that may contain prions (4,5). In the case of a positive sample, PrP D in the sample acts as a template or "seed" for the conversion of the rPrP substrate to a misfolded form. The assay then proceeds in cycles of shaking, through which the newly misfolded substrate can also participate in seeding further conversion, thereby amplifying the amount of misfolded prion protein in the reaction mixture to detectable levels.
The RT-QuIC assay has shown exceptional sensitivity for the detection of prions, where seeding activity can be detected in positive samples that have been diluted more than 10 billion-fold from the starting tissue (5). However, a significant hurdle facing the assay is that RT-QuIC is inhibited by various constituents naturally present in tissues and bodily fluids. A common practice has been to dilute tissue samples 1,000-fold (10 -3 ) before use in the assay (5)(6)(7). While this technique does succeed in circumventing a large part of tissue-mediated inhibition, it also inherently limits the potential sensitivity of the assay, thereby limiting the potential of amplification assays to enhance the detection of early infections.
In addition to "tissue" generally, constituents of blood and, in particular, red blood cells are known to strongly inhibit the RT-QuIC assay (8)(9)(10)(11). This is problematic both because blood is a near-ubiquitous contaminant of most tissues and because blood itself is known to harbor infectious prions (12)(13)(14)(15)(16) and is thus a very attractive sample type for live animal testing that could be repeated over time. While some progress has been made in detecting bloodborne prions by RT-QuIC with pre-analytic processing techniques such as selective precipitation (12) and affinity purification (10), it is clear that some level of inhibition is still present relative to non-blood-exposed samples (11). In contrast to other cell types, red blood cells, or erythrocytes, are unique in that >95% of their dry mass is made up of hemoglobin (Hb) (17,18). Critical to its oxygen-carrying function, each tetrameric Hb protein contains four heme molecules as cofactors. Free heme (hemin in its oxidized form) is a highly reactive porphyrin molecule that is known to cause oxidative damage (19,20) when its normal cellular niche is disrupted, such as after hemolysis or ischemia-reperfusion injury, conditions that are variably present during tissue sampling, storage, and homogenization. In addition to its potential to cause oxidative damage, heme also binds to PrP in an isoform-sensitive manner (21)(22)(23). While physiologic functions of PrP C -heme binding have been proposed (21,22), the implications for misfolding assays are not entirely clear. In addition to erythrocyte-medi ated inhibition of RT-QuIC (8), hemin itself has previously been shown to inhibit the detection of prions by another type of misfolding assay, the protein misfolding cyclic amplification (PMCA) assay (24). Furthermore, hemin has also been shown to influence the structures and stability of other disease-associated amyloids, including amyloid-β of Alzheimer's disease (25,26), α-synuclein of Parkinson's disease (27), and lysozyme of lysozyme amyloidosis (28).
The present study aims to elucidate the extent to which, and processes by which, blood and its major components, hemoglobin and heme, inhibit the RT-QuIC assay in the sensitive detection of prions in tissue samples. This knowledge may ultimately allow for mitigation of this inhibition and the more sensitive detection of prions in blood-contain ing samples and/or blood directly.
## MATERIALS AND METHODS
## Tissue preparation
The tissue samples used in this study are from a frozen archive of tissues collected from naturally infected and experimentally infected small ruminants. All tissue samples were collected after humane euthanasia and were same-day processed and frozen (-80°) until use.
Sheep brain homogenates for RT-QuIC seeding (#4789 [pos], #4799 [neg]) were prepared as 10% (wt/vol) homogenates in 1× phosphate-buffered saline (PBS) using a rotor stator homogenizer (GLH-01, Omni International) with single-use plastic probe tips.
Blood samples for RT-QuIC inhibition and rPrP binding studies were collected from a scrapie-naïve sheep (#5047) in 10 mL EDTA Vacutainer tubes (Becton Dickinson). Tissue samples for heme quantification were collected from two sheep (#4645, #4649) and one goat (#5010G). Tissues collected included brainstem, cerebellum, tonsil, retropharyngeal lymph node (RPLN), spleen, kidney, liver, diaphragm, skeletal muscle, rectal mucosa, cerebrospinal fluid (CSF), and whole blood collected with ACD anticoagulant (8:60 mL ACD Formula A; Fenwal). Sheep placental cotyledons were also collected, with five individual cotyledons gathered from each of 10 placentas (#1266-1275). Tissues for heme quantification experiments, excluding blood and CSF, were prepared as 10% (wt/ vol) homogenates in 1× PBS using 0.7 mm Zirconia beads (BioSpec 11079107zx) in a bead-beating grinder (Fast Prep 24; MP bio). CSF was prepared by centrifugation at 500 × g for 10 min, and the resulting supernatant was collected. Tissue homogenate, CSF supernatant, and whole blood stocks were stored at -80°C, with working sub-aliquots stored short term at -20°C.
Homogenate concentrations in this manuscript are described as dilutions relative to intact tissues, i.e., a 10% (wt/vol) homogenate is represented as a 10 -1 dilution.
## Heme and Hb solution preparation
Hemin (Hemin-Cl; EMD Millipore) and Hb (Hemoglobin from bovine blood; Sigma Aldrich) stock solutions were prepared gravimetrically from dry reagents.
Hemin was solubilized in 0.1 M NaOH and subsequently diluted at least 100-fold in seed dilution (SD) buffer (1× PBS pH 7.4 + 0.1% SDS + 1× N-2 media supplement [Gibco-Fisher]) to amend pH; hemin was found to be soluble in this buffer to at least 200 µM.
Hb was prepared in 1× PBS pH 7.4 for quantification standards or in SD buffer for RT-QuIC spiking. For Hb, concentrations are calculated and reported as monomeric Hb.
Apohemoglobin (apoHb) was prepared from Hb stock solutions by acid-acetone extraction (29,30). A solution of 20 mM Hb was added dropwise with constant stirring to approximately 30 volumes of acidified acetone (0.2% vol/vol 12 N HCl) at -20°C. The resulting precipitate was collected by centrifugation at 1,000 × g for 15 min and resuspended in ddH 2 O. The resulting solution was then successively dialyzed against ddH 2 O, 1.6 mM sodium bicarbonate, and 1× PBS pH 7.4. Following dialysis, residual precipitate was removed by centrifugation at 3,000 × g for 10 min. ApoHb concentrations were estimated by UV-Vis absorbance at 280 nm using ε 280nm = 0.0162 M -1 (29,31). ApoHb dilutions prepared using this extinction coefficient were also compared to (wt/ vol) holohemoglobin (holoHb) standards by SDS-PAGE and Coomassie staining.
## Heme quantification in blood and tissues
Heme concentrations in whole blood were measured by alkaline detergent hematin (ADH) assay (32,33). Dilutions of whole blood and Hb standards were prepared in 1× PBS. A total of 20 µL of each was then added to 180 µL of a buffer consisting of 0.1 M NaOH + 2.5% wt/vol Triton X-100. UV-Vis absorbance spectra were recorded, and blood-heme concentrations were calculated by comparison to the Hb standard curve at 575 nm.
Heme concentrations in sheep and goat tissues were measured by oxalic acid fluorescence assay (34,35). Dilutions of tissue homogenates and Hb standards were prepared in 1× PBS. A total of 20 µL of each was then added to 980 µL of 2 M oxalic acid. The sample preparations were then split, and 500 µL was incubated at 100°C for 30 min while the other half was maintained at room temperature. Samples were then measured for fluorescence using an excitation wavelength of 400 nm and emission of 662 nm. For each sample, the room temperature measurements were subtracted from the 100°C incubated measurements to control for any non-heme-derived fluorescence present in the samples. Measurements from the Hb standard curve were then used to calculate the concentration of heme for tissue samples.
UV-Vis and fluorescence measurements were performed using a CLARIOstar microplate reader (BMG Labtech).
## RT-QuIC
RT-QuIC was performed using hamster-sheep chimeric rPrP substrate (Syrian hamster residues 23 to 137 [accession no. K02234] followed by sheep [R154, Q171] residues 141 to 234 [accession no. AJ567988]). Protein was expressed in DE3 Escherichia coli using the pET41 vector and Overnight Express Autoinduction System 1 (Novagen, Madison, WI). rPrP was purified from inclusion bodies as described by Orrú et al. (4). Briefly, inclusion bodies were solubilized in guanidine hydrochloride, purified using nickel immobilized metal affinity chromatography, and refolded on the resin with a gradient of guanidine hydrochloride using an AKTA Pure FPLC (Cytivia). Following elution by imidazole gradient, rPrP was dialyzed into 10 mM Na 2 PO 4 (pH 5.8) and stored at -80°C.
RT-QuIC reactions, consisting of 2 µL seed material and 98 µL RT-QuIC assay buffer (10 mM NaPO 4 pH 7.4, 300 mM NaCl, 1 mM EDTA, 10 µM ThT, 0.1 mg/mL rPrP) per well, were carried out at 42°C with alternating cycles of 1 min double orbital shaking at 700 rpm and 1 min rest for 100 h total using FLUOstar or CLARIOstar microplate readers (BMG Labtech). ThT fluorescence was measured using 20 flashes per well, bottom read, with an excitation wavelength of 450 ± 10 nm and emission wavelength of 480 ± 10 nm, fixed gain of 1,800, and 15 min read intervals.
For RT-QuIC, seed homogenate dilutions were prepared in SD buffer (1× PBS pH 7.4 + 0.1% SDS + 1 × N-2 media supplement [Gibco-Fisher]). In RT-QuIC reactions contain ing hemin, Hb, or blood, inhibitors were spiked into the seed homogenate dilutions such that the concentrations named represent the final concentrations present in the seed material. For blood, reported concentrations represent the heme/monomeric Hb concentration present in each dilution of whole blood.
## Methods used to evaluate RT-QuIC data
Each 100 h record of ThT fluorescence measurements was exported to an Excel spreadsheet. A custom script (Python version 3.10.9) was written to import, merge, and export as one comma-delimited data file all Excel files relevant to a given experi ment. Each experiment's data file was then imported into SAS (SAS version 9.4), where various SAS procedures (PROCs) were applied to transform data, detect and characterize reactions, perform statistical analyses, and produce graphs for presentation.
RT-QuIC reaction data, including Excel exports of raw ThT fluorescence measurements and graphs depicting raw ThT fluorescence curves, are available from the National Agricultural Library Ag Data Commons database (https://doi.org/10.15482/USDA.ADC/ 28836296). Examples of raw ThT fluorescence curves can be seen in Fig. 8A of this manuscript.
Given some extreme effects of the conditions of this study on the morphology and variability in the fluorescence data, in place of a traditional plate-wide fluorescence threshold, custom algorithms were created and uniformly applied to each individual well's data to detect reactions and evaluate morphologic features. The fluorescence readings were first regressed over time (PROC TRANSEG) to produce its penalized B-spline and its upper 99.9% confidence limit. This resulted in the local reduction in variability and a liberal upper confidence limit. Subsequently, a moving estimate of the fluorescence trend (ThT trend ) was calculated as the leading 1 h median of the penalized B-spline (PROC EXPAND). Similarly, a moving estimate of a critical threshold value (ThT crit ) was calculated as the preceding 5 h median of the fluorescence upper confidence limit offset by an additional 30 min. Thus, at a given time point, a relatively short forward-biased estimate of ThT trend was compared to a local window estimate of the preceding baseline fluorescence (ThT crit ). A positive amplification signal and the accompanying lag time for a given reaction was determined by whether and when ThT trend first exceeded ThT crit . The reaction height at a given time point was defined as the baseline (ThT crit )-subtracted fluorescence.
Because of the inhibitory conditions being tested, reactions frequently could not be detected in all replicates. These instances were captured graphically by assigning them to a not-detected (ND) reference line placed at an arbitrary post-assay time of 110 h. For statistical analysis, all times (lag and ND times) were first ranked (PROC RANK) and then analyzed using a generalized linear model based on the gamma distribution (PROC GLIMMIX). Occasional extreme outliers were identified and removed from analyses based on a panel of studentized residual plots. All final models were well fit by the gamma distribution. Post hoc analyses consisted of pre-planned comparisons of interest using the modeled least squares means, variation, and the Kenward-Roger method of degrees of freedom estimation for unbalanced data. The family-wise error rate of multiple comparisons was controlled using the adjustment method of Holm (i.e., stepdown Bonferroni). Significance was accepted at a P < 0.05.
## Heme-rPrP interaction
Hemin-rPrP interactions were quantified by UV-Vis spectral shift. RT-QuIC buffer without ThT (10 mM NaPO 4 pH 7.4, 300 mM NaCl, 1 mM EDTA, 0.1 mg/mL rPrP) was spiked with mock sample buffer (1× PBS + 0.1% SDS) or 1× PBS containing hemin at varying concentrations. Due to rPrP precipitation observed upon addition of the SDS+ buffer at 0 µM hemin, the 1× PBS buffer condition was used for binding ratio experiments. Matched hemin-only solutions were also prepared where hemin solutions were spiked into RT-QuIC buffer without ThT or rPrP (10 mM NaPO 4 pH 7.4, 300 mM NaCl, 1 mM EDTA). Differential spectra were taken, subtracting hemin-only spectra from hemin-rPrP spectra. Evolution of a differential peak at 416 nm and a valley at 385 nm was observed. The difference between these wavelengths was plotted against the hemin:rPrP molar ratio to show dose-dependent evolution of the shift and eventual saturation of rPrP with hemin. Interaction of rPrP with Hb or blood was tested by identical methods to investigate the possibility of heme transfer and/or binding to rPrP from either source.
## Seed exposure to Hb and blood
A 10% brain homogenate from a scrapie-positive sheep was mixed 1:1 with whole blood, Hb solution, or PBS and incubated at 4°C for 24 h or 7 days. Prior to incubation, blood was lysed by sonication for two pulses of 30 seconds each in a water bath sonicator (Qsonica) at 180 W. Hb solution concentration was matched to that of whole blood as measured by ADH assay. Following incubation, mixtures were diluted in SD buffer and used as RT-QuIC seed material as described above.
## rPrP stability in RT-QuIC buffer
RT-QuIC assay buffer (-ThT) containing 0.1 mg/mL rPrP was incubated with mock seed samples containing hemin, Hb, or blood. UV-Vis spectra were then recorded on a CLARIOstar microplate reader (BMG Labtech). Samples were subsequently incubated for 24 h at 42°C. Following incubation, samples were centrifuged at 21,000 × g for 10 min, and the supernatant was assayed for rPrP by SDS-PAGE and Coomassie staining.
## RESULTS AND DISCUSSION
## Relative inhibitory effects of blood, hemoglobin, and hemin on prion RT-QuIC
To quantify the inhibitory effects of heme, Hb, and whole blood, a matrix of RT-QuIC assay conditions was assembled where scrapie-positive sheep brain homogenates in dilutions ranging from 10 -3 to 10 -6 were spiked with heme, Hb, or whole blood at concentrations ranging from 0 to 200 µM (Fig. 1). These concentrations represent the amounts of heme or monomeric Hb present in the seed material prior to RT-QuIC analysis, where 2 µL of seed material is then introduced to 98 µL of reaction buffer. For whole blood, the reported concentrations describe the amount of monomeric Hb the dilution contains.
The obex was chosen as the source of prion seed for several reasons. The obex is the richest source of prions relative to other tissues and, as collected and measured in this study, is very low in natural heme content (≤98 nM for a 10 -3 dilution; Fig. 2). Use as a simply diluted homogenate also avoids the inherent potential of seed purification methods to perturb the authentic state of prions-whether that be alterations to conformation, solubility, or bound ligands, or to introduce selection bias for subpopula tions of prion content. The addition of heme, Hb, or blood to the seed homogenates resulted in dosedependent increases in lag times at all seed dilutions (Fig. 1B). At higher inhibitor concentrations, prion detection was completely ablated (Fig. 1B). Of the three inhibitor sources tested, blood appeared to produce the greatest inhibition, followed by Hb, and then heme.
In addition to the longer lag times and lost detection of seeded reactions, the morphology of the RT-QuIC reaction curves (ThT fluorescence vs time) was changed by the addition of heme-containing inhibitors. In the presence of inhibitors, dose-depend ent decreases were observed in the fluorescence signal, both during the pre-reaction period (Fig. 1A) and following amplification (Fig. 1C). Neither hemin nor Hb absorb photons strongly at the excitation or emission wavelengths of ThT (hemin A max : 385 nm, Hb A max : 411 nm, ThT ex : 450 nm, ThT em : 480 nm), and others have demonstrated that the fluorescence of unbound ThT is not impacted by the presence of heme (25). To examine this more closely in our specific reaction conditions, we added the heme-con taining inhibitors to RT-QuIC reaction buffer both with and without a standard amount (0.1 mg/mL) of rPrP. Interestingly, while ThT fluorescence is known to increase sharply when bound to amyloid (36,37), we did observe increased baseline fluorescence in the presence of rPrP (Fig. 3A). The addition of heme-containing inhibitors showed minimal impact on ThT fluorescence in the rPrP-free buffers but did suppress the rPrP-dependent ThT fluorescence increase (Fig. 3B). The contributions of other reaction factors were also analyzed, and the results were summarized as Fig. S1 at https://doi.org/10.15482/ USDA.ADC/28836296. One interpretation of the results is that while hemin or Hb do not substantially interact with ThT alone, both molecules appear to interact with rPrP. This interaction may be competitive, i.e., displacing ThT from rPrP, or they may bind concurrently. In the latter scenario, the close proximity of the two bound molecules could allow the heme macrocycle to act as an acceptor for energy transfer, thereby suppressing ThT fluorescence (38).
## Blood/heme does not destroy PrP Sc seeding activity
Previous research has demonstrated that exposure to heme results in the destruction or restructuring of other pathogenic amyloids (25)(26)(27)(28). To test this potential with prions from a natural host, homogenate of a scrapie-positive sheep brain was mixed 1:1 with either lysed whole blood, concentration-matched Hb solution, or 1× PBS and incubated at 4°C for 1 week. These conditions were chosen to mimic those of a severely hemo lyzed blood sample or heavily blood-contaminated tissue with subsequent refrigerated storage delays (24 h or 1 week) prior to analysis. The limiting dilution for detection was then determined using serial dilutions of each seed material. Except for the 10 -3 dilutions in which Hb was still present at >100 µM in the reaction mixture, significant differences in lag time or limits of detection were not observed (Fig. 4). These results demonstrate that exposure to whole blood or hemoglobin in the tissue homogenate does not disrupt the seeding activity of PrP D in the assay.
These encouraging results align with those of another study (39) that examined bloodborne prion detection by a different assay, PMCA. In that study, prion-spiked blood samples did show some changes to detectability in early PMCA rounds but ultimately yielded unhindered limits of detection by serial dilution. As part of its mechanism, the PMCA assay includes serial dilutions between "rounds. " Similar to our conditions at the 10 -3 dilution, concentrations of blood components in the reaction mixture would be relatively high in these early rounds and be diluted away as the assay progresses, ultimately allowing for detection of the apparently undisrupted prions.
## Inhibitors act on the rPrP assay substrate
Another fundamental component of amplification assays is the recombinant prion protein substrate. Prior research has shown that rPrP binds to heme in an isoform-specific manner, where the heme:PrP binding ratio is higher for monomeric rPrP than for a misfolded form (23). To confirm similar heme:rPrP binding under RT-QuIC conditions, we spiked RT-QuIC reaction buffer (-ThT) (10 mM NaPO4 pH 7.4, 300 mM NaCl, 1 mM EDTA, 0.1 mg/mL rPrP) with mock seed samples containing from 0 to 200 µM hemin. Binding of rPrP and hemin was observed as a red shift in the maximum absorbance peak of heme (Fig. 5A, B, andD). In addition to the spectral shift, visible turbidity was evident in the spiked samples; this was observed as an overall increase in the UV-Vis spectra absorbance baseline (Fig. 5A B). Interestingly, significant turbidity was also observed in the 0 heme condition, where the reaction buffer was spiked with buffer (1× PBS + 0.1% SDS) alone (Fig. 5A C). To prevent interference from SDS-induced turbidity, heme spiking was also performed using a mock sample buffer of 1× PBS lacking SDS. These conditions did eliminate the turbidity observed at 0 µM hemin (Fig. 5B C) while maintaining the heme binding-induced red shift (Fig. 5B). For this reason, the following experiments examining the binding ratio of hemin:rPrP were performed in the absence of SDS.
To quantify the spectral shift, differential spectra were calculated where the spectra of the hemin solutions alone were subtracted from those of the hemin + rPrP mixtures (Fig. 5D). Evolution of a differential peak at 416 nm and a valley at 385 was observed (Fig. 5D). The difference of these absorbance values was then plotted against the heme:rPrP molar ratio (Fig. 5E). When rPrP was exposed to hemin at the previously tested inhibitory concentrations of 12.5 µM-200 µM, a dose-dependent spectral shift was observed; however, saturation was not reached. Subsequently, hemin was added at 5×, 10×, and 20× the inhibitory concentration series (i.e., 62.5 µM-1,000 µM, 125 µM-2,000 µM, and 150 µM-4,000 µM) to reach saturation. Ultimately, extensive precipitation of rPrP precluded precise quantification of the binding ratio; however, loss of linearity in the red-shift increase was seen beginning at molar ratios greater than ~5:1 hemin:rPrP (Fig. 5E). This broadly agrees with the previously reported 7:1 binding ratio for heme to rPrP (23).
RT-QuIC reaction buffer (-ThT) was also spiked with Hb and whole blood at 12.5 µM-200 µM. While binding of hemin to rPrP displays a strong shift in the differential (heme-subtracted) UV-Vis spectrum prior to incubation (Fig. 6B), initial exposure of rPrP to Hb or whole blood results in a much smaller initial spectral shift (Fig. 6C D). This suggests the majority of heme in these conditions remains bound to Hb or that any transfer of heme to rPrP may occur more gradually over the assay run time. An example figure comparing the differential spectra to the raw and inhibitor-only spectra at the 200 µM conditions can be found in the supporting information (Fig. S2 at https://doi.org/ 10.15482/USDA.ADC/28836296). To examine the inhibitor impacts at later assay time points under reaction conditions, RT-QuIC reaction buffer was spiked with sample buffer (SDS+) containing hemin, Hb, or whole blood and incubated at 42°C for 24 h. These mixtures were then centrifuged at 21,000 × g for 10 min to remove insoluble material, and the resulting supernatants were analyzed by SDS-PAGE and Coomassie staining. In agreement with the initial spectroscopic measurements, even in the absence of heme, a significant portion of the rPrP substrate is lost after 24 h at reaction conditions (Fig. 6A). For the hemin-exposed samples, additional, dose-dependent rPrP loss was observed (Fig. 6B). After 24 h, the Hb or blood spiked buffers unexpectedly showed more rPrP remaining in the buffer than in the 0 heme condition (Fig. 6C andD), indicating that the presence of Hb or blood in the reaction mixture actually stabilizes the solubility of rPrP. These results suggest that both hemin and Hb interact with rPrP in the assay substrate and impact its solubility in the reaction buffer, albeit through different mechanisms.
In response to the evident changes to rPrP substrate stability following exposure to heme, heme-free RT-QuIC reactions were prepared using a range of substrate concentra tions from 0.01 to 0.10 mg/mL (10%-100%) so that the impact of rPrP depletion alone could be assessed. Reactions performed with depleted rPrP, particularly those below 0.05 mg/mL (50%), exhibited longer lag times, lost detections, and diminished fluorescence signals (Fig. 7) similar to the heme-mediated effects on the standard reaction mixture containing 0.1 mg/mL rPrP.
Lastly, a reaction condition matrix was prepared testing reactions containing 0.10, 0.15, and 0.20 mg/mL (100, 150, 200%) rPrP spiked with 0, 50, 100, and 200 µM Hb. As readily observed in Fig. 8A, reactions with elevated rPrP concentrations showed dose-dependent rescue of lag times (Fig. 8B), reaction frequency, and fluorescence intensity (Fig. 8C), albeit with some delay in lag times in the absence of Hb (Fig. 8B, 0 µM Hb). Together, these data demonstrate that interactions with the rPrP substrate are the primary factor by which heme-containing inhibitors disrupt prion detection by RT-QuIC.
As a caveat to mitigating this type of assay inhibition through increased sub strate concentration, unseeded (spontaneous) misfolding reactions in the absence of hemoglobin began somewhat earlier at 0.15 mg/mL [rPrP] as compared to that occurring in the standard 0.10 mg/mL [rPrP] buffer (Fig. S3 at https://doi.org/10.15482/USDA.ADC/ 28836296). This suggests assay cutoff times should be re-evaluated before attempting to apply simple substrate supplementation as a counter to potential heme contamination in a diagnostic setting.
## Heme and protein constituents of Hb each contribute to inhibition of RT-QuIC
Given the apparent mechanistic differences between free hemin-and Hb-mediated inhibition, assays were also performed in the presence of apoHb to extricate the respective contributions of the bound heme and globin protein components of Hb. ApoHb was prepared from the holoprotein (Hb) by acidified acetone extraction. Following reconstitution, apoHb concentrations were estimated using the previously published extinction coefficient ε 280nm = 0.0162 M -1 (29,31). As apoHb is inherently unstable due to the lack of its native cofactor, absorbance-based protein quantification can be inexact. As a method of secondary confirmation, solutions containing equal amounts of Hb and apoHb (as estimated by 280 nm absorbance) were prepared and analyzed by SDS-PAGE and Coomassie staining. This comparison demonstrated grossly equivalent concentrations between the two forms (Fig. 9A). Though the addition of apoHb did result in assay inhibition relative to 0 µM controls (Fig. 9B), the overall effectiveness in prolonging lag time was generally less for the apoprotein (apoHb) as compared to the holoprotein (Hb) (main effect of Hb type: at 10 -4 , P < 0.0001; at 10 -5 , P = 0.0316). This was most evident when the reaction was seeded with 10-fold greater PrP D , where significant differences between Hb types were detected at 12.5 µM-50 µM Hb (seeded at 10 -4 , P Holm ≤0.0034), but only at 12.5 µM Hb when seeded at a dilution of 10 -5 (P Holm ≤0.0211). These data suggest that both the globin protein component and the bound heme cofactor contribute to the inhibitory effects of Hb on RT-QuIC.
## Heme quantification in diagnostic tissues
To gauge the potential impact of heme/Hb-mediated inhibition on diagnostic efforts using RT-QuIC, various tissues from small ruminants were assayed for total heme concentration by oxalic acid fluorescence assay. The results of these measurements are summarized in Fig. 2, along with calculated extrapolations for various tissue homogenate dilutions. Briefly, while heme concentrations exceed inhibitory levels in nearly all intact tissues, by the time a 10 -3 dilution was reached, only the blood, spleen, and placenta approached inhibitory levels of heme. However, several tissues maintained micromolar levels of heme if only diluted 10 -2 , which may still be capable of disrupting detection in peripheral tissues or during the earliest stage of infection where tissue samples are more likely to bear low prion titers.
## Conclusions
In this study, we demonstrate that each of heme, Hb, and whole blood inhibits the RT-QuIC assay. Delayed lag times were observed with seed material samples containing as low as 12.5 µM heme and, in the case of lower prion-titer samples, a substantial loss of replicate detections was observed at concentrations as low as 50 µM for Hb or whole blood. Heme itself, the globin component of Hb, and the remaining constitu ents of whole blood are each contributing factors to the observed inhibition. When prion-containing samples were incubated in the presence of Hb or whole blood prior shown as bar graphs where # denotes significant differences between Hb-containing and Hb-free reactions at 0.1 mg/mL rPrP (all P Holm <0.0001), and * denotes significant differences between elevated (0.15, 0.20 mg/mL) and standard (0.1 mg/mL) rPrP concentrations within a given Hb concentration (all P Holm ≤0.005).
to being assayed, seeding activity was not lost, suggesting that inhibition occurs at the level of the assay itself. Hemin was confirmed to be able to bind rPrP in RT-QuIC reaction buffer, and both free hemin and Hb induced solubility changes in the protein sub strate. When heme-free reactions were performed with limited quantities of substrate, outcomes mirrored those of reactions inhibited by heme, Hb, or blood. Finally, increas ing the starting rPrP concentration was able to rescue Hb-inhibited prion detection, reinforcing the conclusion that inhibition from Hb-or blood-containing samples is a result of depletion of available rPrP substrate in the assay mixture. When heme was measured in small ruminant tissues, levels were, not unexpect edly, shown to vary between tissue types. Of interest, nearly all types of intact tissue contain >200 µM heme; however, given our data suggesting pre-reaction exposure does not degrade PrP D seeding activity, the concentrations at the actual testing dilutions are more relevant. At the most commonly tested tissue dilution, 10 -3 , only the blood, spleen, and placenta still contained millimolar quantities, and even these tissues would be at the low end of the inhibitory range. It should be noted that the tissues in this study were collected from freshly euthanized animals which were then exsanguinated and necropsied in a timely manner. It is therefore possible that field-collected samples may have a considerably greater degree of blood contamination or hemolysis. Even so, given the results from whole blood and splenic tissue, it remains unlikely that heme-mediated inhibition would be more than a minimal factor at a 10 -3 dilution or greater. While Asterisks denote significant differences between reactions containing equivalent concentrations of Hb or apoHb (all P Holm ≤0.034). these levels of sample dilution are common practice, they are inherently limiting to the theoretical sensitivity of the assay. In recent years, a variety of efforts have been made to enrich or purify prions from tissues (10,(40)(41)(42), some of which are only semi-selective for PrP D and may result in increased inhibitor concentrations along with the enriched prions. Other strategies aim to allow for testing of less diluted samples (43,44). In these cases, it is possible that heme or Hb levels may again be present at inhibitory levels. Somewhat conveniently, the ability to see the color of a solution of aqueous Hb with the naked eye roughly coincides with the inhibitory ranges described in this study (Fig. 10). As a broad observation, if the final seed material to be introduced to the RT-QuIC assay is visibly red/pink/brown colored, Hb-mediated inhibition may influence the reaction lag times, fluorescence signal maxima, and frequency of detection, and thus should be considered when interpreting results.
## References
1. Plummer, Wright, Johnson et al. (2017) "Temporal patterns of chronic wasting disease prion excretion in three cervid species" *J Gen Virol*
2. Mathiason, Hays, Powers et al. (2009) "Infectious prions in pre-clinical deer and transmission of chronic wasting disease solely by environmental exposure" *PLoS One*
3. Prusiner (1998) *Prions. Proc Natl Acad Sci U S A*
4. Orrù, Groveman, Hughson et al. (2017) "RT-QuIC Assays for Prion Disease Detection and Diagnostics"
5. Wilham, Orrú, Bessen et al. (2010) "Rapid end-point quantitation of prion seeding activity with sensitivity comparable to bioassays" *PLoS Pathog*
6. Dassanayake, Orrú, Hughson et al. (2016) "Sensitive and specific detection of classical scrapie prions in the brains of goats by real-time quaking-induced conversion" *J Gen Virol*
7. Holz, Darish, Straka et al. (2021) "Evaluation of real-time quaking-induced conversion, ELISA, and immunohistochemistry for chronic wasting disease diagnosis" *Front Vet Sci*
8. Cramm, Schmitz, Karch et al. (2016) "Stability and reproducibility underscore utility of RT-QuIC for diagnosis of creutzfeldtjakob disease" *Mol Neurobiol*
9. Nonaka, Iwasaki, Horiuchi et al. (2024) "Detection limitations of prion seeding activities in blood samples from patients with sporadic prion disease" *BMC Neurol*
10. Orrú, Wilham, Raymond et al. (2011) "Prion disease blood test using immunoprecipitation and improved quaking-induced conversion" *mBio*
11. Thomas, Salamat, De Wolf et al. (2023) "Development of a sensitive realtime quaking-induced conversion (RT-QuIC) assay for application in prion-infected blood" *PLoS One*
12. Elder, Henderson, Nalls et al. (2013) "In vitro detection of prionemia in TSE-infected cervids and hamsters" *PLoS One*
13. Mathiason, Powers, Dahmes et al. (2006) "Infectious prions in the saliva and blood of deer with chronic wasting disease" *Science*
14. Houston, Mccutcheon, Goldmann et al. (2008) "Prion diseases are efficiently transmitted by blood transfusion in sheep" *Blood*
15. Hunter, Foster, Chong et al. (2002) "Transmission of prion diseases by blood transfusion" *J Gen Virol*
16. Dassanayake, Truscott, Zhuang et al. (2015) "Classical natural ovine scrapie prions detected in practical volumes of blood by lamb and transgenic mouse bioassays" *J Vet Sci*
17. Weed, Reed, Berg (1963) "Is hemoglobin an essential structural component of human erythrocyte membranes?" *J Clin Invest*
18. Kaza, Ojaghi, Robles (2021) "Hemoglobin quantification in red blood cells via dry mass mapping based on UV absorption" *J Biomed Opt*
19. Suliman, Carraway, Velsor et al. (2002) "Rapid mtDNA deletion by oxidants in rat liver mitochondria after hemin exposure" *Free Radic Biol Med*
20. Miller, Shaklai (1994) "Oxidative crosslinking of LDL protein induced by hemin: involvement of tyrosines" *Biochem Mol Biol Int*
21. Lee, Raymond, Schoen et al. (2007) "Hemin interactions and alterations of the subcellular localization of prion protein" *J Biol Chem*
22. Tripathi, Singh (2016) "Prion protein-hemin interaction upregulates hemoglobin synthesis: implications for cerebral hemorrhage and sporadic creutzfeldt-jakob disease" *JAD*
23. Pato, Célier, Rezaei et al. (2004) "Heme as an optical probe of a conformational transition of ovine recPrP" *Protein Sci*
24. Soutyrine, Yogasingam, Huang et al. (2015) "Effects of heme-PrP complex on cell-free conversion and peroxidase-linked immunode tection of prions in blood-based assays" *Res Vet Sci*
25. Atamna, Boyle (2006) "Amyloid-beta peptide binds with heme to form a peroxidase: relationship to the cytopathologies of Alzheimer's disease" *Proc Natl Acad Sci*
26. Sankar, Donegan, Shah et al. (2018) "Heme and hemoglobin suppress amyloid β-mediated inflammatory activation of mouse astrocytes" *J Biol Chem*
27. Hayden, Kaur, Williams et al. (2015) "Heme stabilization of α-synuclein oligomers during amyloid fibril Full-Length Text Journal of Clinical Microbiology"
28. *Biochemistry*
29. Sonavane, Haider, Kumar et al. (2017) "Hemin is able to disaggregate lysozyme amyloid fibrils into monomers" *Biochimica et Biophysica Acta (BBA) -Proteins and Proteomics*
30. Pires, Belcher, Palmer (2017) "Quantification of active apohemo globin heme-binding sites via dicyanohemin incorporation" *Biochemis try*
31. Izadi, Henry, Haladjian et al. (1997) "Purification and characterization of an extracellular heme-binding protein, HasA, involved in heme iron acquisition" *Biochemistry*
32. Vasudevan, Mcdonald (1997) "Spectral demonstration of semihe moglobin formation during CN-hemin incorporation into human apohemoglobins" *J Biol Chem*
33. Zander, Wolf (1984) "Alkaline haematin D-575, a new tool for the determination of haemoglobin as an alternative to the cyanhaemiglobin method. I. description of the method" *Clin Chim Acta*
34. Heuck, Reinauer, Wood (2008) "The alkaline haematin detergent (AHD575) method for the determination of haemoglobin in blood--a candidate reference measurement procedure" *Clin Lab*
35. Morrison (1965) "Fluorometric microdetermination of heme protein" *Anal Chem*
36. Marcero, Piel, Iii et al. (2016) "Rapid and sensitive quantitation of heme in hemoglobinized cells" *BioTechniques*
37. Krebs, Bromley, Donald (2005) "The binding of thioflavin-T to amyloid fibrils: localisation and implications" *J Struct Biol*
38. Xue, Lin, Chang et al. (2017) "Thioflavin T as an amyloid dye: fibril quantification, optimal concentration and effect on aggregation" *R Soc Open Sci*
39. Koga, Yoshihara, Bando et al. (2013) "Development of a heme sensor using fluorescently labeled heme oxygenase-1" *Anal Biochem*
40. Kramm, Pritzkow, Lyon et al. (2017) "Detection of Prions in Blood of Cervids at the Asymptomatic Stage of Chronic Wasting Disease" *Sci Rep*
41. Safar, Wille, Itri et al. (1998) "Eight prion strains have PrP(Sc) molecules with different conformations" *Nat Med*
42. Henderson, Denkers, Hoover et al. (2020) "Progression of chronic wasting disease in white-tailed deer analyzed by serial biopsy RT-QuIC and immunohistochemistry" *PLoS One*
43. Denkers, Henderson, Mathiason et al. (2016) "Enhanced prion detection in biological samples by magnetic particle extraction and real-time quaking-induced conversion" *J Gen Virol*
44. Christenson, Li, Rowden et al. (2023) "Nanoparticleenhanced RT-QuIC (Nano-QuIC) diagnostic assay for misfolded proteins" *Nano Lett*
45. Piel, Veneziano, Nicholson et al. (2024) "Validation of a real-time quakinginduced conversion (RT-QuIC) assay protocol to detect chronic wasting disease using rectal mucosa of naturally infected, pre-clinical whitetailed deer (Odocoileus virginianus)" *PLoS One* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724189&blobtype=pdf | # SUMOylation-dependent degradation of nucleocapsid is responsible for Pestivirus uncoating
Lin-Ke Zou, Ji-Shan Bai, Rui-Cong Sun, Han-Fei Yang, Ming-Yue Wan, Bing-Qian Zhao, Bo-Tao Sun, Jin-Xia Chen, Jing Chen, Yan Cheng, Bin Zhou
## Abstract
Classical swine fever virus (CSFV), a highly virulent member of the Pestivirus genus, is one of the most significant pathogens within this group. Although uncoating is a prerequisite for productive infection, the molecular determinants orchestrating this process remain obscure. The Core protein, functioning as the viral nucleocapsid, plays a pivotal role in the uncoating cascade. This study delineates a SUMOylation-depend ent, ubiquitin-independent proteolytic mechanism essential for CSFV uncoating. The valosin-containing protein (VCP/p97) preferentially associates with SUMO1-modified Core, directing it toward degradation via the 26S proteasome, specifically through engagement with PSMB2 and PSMD2 subunits. Site-directed mutagenesis of the SUMOylation motif abolishes VCP-mediated degradation, substantiating its functional indispensability. Fluorescent tracking of CSFV virions using molecular beacon and quantum dot labeling further reveals that VCP governs the endosomal trafficking of viral particles from early to late endosomes-an essential step for capsid disassembly and genome release. Moreover, VCP operates in concert with NPL4 and UFD1, enabling the translocation of SUMOylated Core toward the proteasomal machinery. Collectively, these findings uncover a previously uncharacterized SUMO1-VCP-PSMB2/PSMD2 axis that couples intracellular trafficking with proteasomal disassembly of the CSFV Core, providing mechanistic insights into Pestivirus uncoating and nominating host factors as promising antiviral targets.
IMPORTANCEThe fusion of the viral membrane and genome release are hallmark events of enveloped virus infections. However, the related dynamic mechanisms of most viruses remain poorly understood. Here, we demonstrate that VCP directly interacts with the CSFV core protein, and the core protein undergoes proteasomal degradation mediated by the PSMD2 and PSMB2 subunits, with VCP acting as a critical mediator. Surprisingly, this degradation process is independent of ubiquitination but exhibits a strong correlation with the SUMOylation of the nucleocapsid protein. In addition, we found that CSFV genome uncoating occurred in late endosomes, a process regulated by the host VCP. Depletion of VCP prevents viral trafficking to late endosomes and thereby disrupts uncoating efficiency. This is the first evidence implicating SUMOylation in viral uncoating. Deciphering the molecular intricacies governing viral uncoating is pivotal for propelling the development of broad-spectrum antiviral therapeutics aimed at the Pestivirus genus.
chronic disease (1). Even in nations where CSF was previously eradicated, the disease has reemerged (2,3). CSFV is a positive-sense, single-stranded RNA virus about 12.3 kb in length, consisting of an open reading frame (ORF) surrounded by two untranslated regions (UTRs) (4). This genome encodes a polyprotein that is cleaved into four structural proteins (capsid protein Core, envelope glycoproteins Erns, E1, and E2) and eight non-structural proteins (Npro, p7, NS2, NS3, NS4A, NS4B, NS5A, and NS5B) (5,6). Over the past decades, substantial insights have been gained into the molecular mechanisms governing viral genome replication (7)(8)(9)(10). Nevertheless, the processes involved in viral genome assembly and uncoating remain poorly understood. As the virus traverses the endosomal pathway, conformational changes in the envelope glycoproteins mediate membrane fusion, enabling the release of the viral nucleocapsid into the cytoplasm (11). Nevertheless, the precise mechanism by which the viral genome is subsequently released to initiate protein synthesis during infection remains unclear in the CSFV life cycle.
The CSFV Core protein has been considered a structural component essential for nucleocapsid formation and RNA packaging. The CSFV core protein is a small protein rich in basic amino acids located at the N-terminus of the viral polyprotein (after the Npro). It requires processing at the N and C termini in order to be released from the polyprotein. These cleavages are performed by the virus-encoded autoprotease Npro and the host cellular SP between Cys 168 /Ser 169 and Ala 267 /Asp 268 , respectively. The final mature CSFV core protein is about 80-90 aa long (12). Similar to other members of the Pestivirus genus, the CSFV capsid protein is a small, highly basic polypeptide with strong affinity but low specificity for binding to the viral genome (13). It also interacts with viral non-structural proteins and host cell proteins. These interactions play a crucial role in modulating viral virulence, vitality, and replication. Through the application of yeast two-hybrid, several host proteins interacting with the CSFV Core protein have been identified, further substantiating its involvement in viral pathogenicity (14)(15)(16). Despite these advancements, studies on the CSFV Core have made limited progress, particularly concerning its connection with CSFV uncoating.
VCP, a member of the AAA+ family of ATPase proteins, mediates a broad spectrum of cellular functions (17). It functions as a segregase protein that directly recognizes ubiquitin signals on target proteins with the help of adaptor proteins (18). Utilizing ATP hydrolysis, VCP unfolds ubiquitinated proteins and dislodges them from larger molecular complexes (19,20). Beyond its canonical role in recognizing ubiquitinated substrates, accumulating evidence suggests that VCP also engages SUMOylated proteins, particularly in the context of SUMO-ubiquitin hybrid signals generated by SUMO-tar geted ubiquitin ligases such as TOPORS and RNF4 (21). These combinatorial modifications facilitate substrate unfolding and proteasomal targeting by the UFD1/NPL4/VCP complex, thereby mitigating the accumulation of misfolded or potentially cytotoxic protein aggregates. Co-factors such as CLIP-cohibin complex further modulate substrate specificity, collectively positioning VCP as a key integrator of SUMO-and ubiquitin-medi ated proteostasis pathways (22). VCP has been involved in numerous viral contexts, particularly those within the Flaviviridae family; these positive-sense RNA viruses exploit VCP during early stages of infection. Our previous studies have demonstrated that VCP is a critical host factor essential for CSFV replication (23). Specifically, suppressing VCP expression or enzymatic activity disrupted the transport of virions from early endosomes to lysosomes. Similarly, studies utilizing a replication-defective reporter virus system demonstrated that DBeQ and NMS-873, two small-molecule inhibitors of VCP, effectively impeded the uncoating of Yellow Fever virus (YFV) (24). Likewise, DBeQ and NMS-873 sensitivities were observed in Dengue virus (DENV) (25) and Zika virus (ZIKV) (26), indicating that VCP may be universally required across Flaviviruses (27,28). Furthermore, experiments using small interfering RNAs (siRNAs) targeting VCP inhibited early stage of West Nile virus (WNV) infection (29). Recent findings with Japanese encephalitis virus (JEV) indicate that the VCP inhibitor successfully blocked viral replication in both tissue culture and a mouse model (30). Given that VCP impacts multiple cellular processes, additional studies have corroborated the requirement of VCP for the replication of diverse viruses (31)(32)(33)(34)(35)(36)(37)(38). Considering its pivotal role in viral replication, VCP emerges as a highly promising target for the development of antiviral therapeutics, with broad implications for treating a variety of viral infections.
SUMOylation is a ubiquitination-like post-translational modification that plays a pivotal role in regulating protein functions. Recent studies have shown that proteins from both RNA and DNA virus families can be modified by SUMO conjugation, and this modification appears to be crucial for the functionality of viral proteins (39)(40)(41). SUMOylation has widespread effects on various cellular processes, including transcriptional regulation, apoptosis, stress responses, and cell cycle control, making it an attractive target for viral dysregulation. The capsid structural protein of human papillomaviruses (HPV) is also SUMOylated, and this modification enhances their stability (42). The DENV envelope protein binds to Ubc9, and overexpression of Ubc9 leads to reduced plaque formation (43).
In this study, we identified that the VCP played a pivotal role in CSFV uncoating and CSFV Core protein degradation. Mechanistically, VCP plays a pivotal role in the translo cation of Core from the early endosome to the late endosome. Moreover, CSFV Core protein undergoes SUMOylation in a SUMO1-dependent manner, with VCP interacting with the SUMOylated Core to mediate its subsequent degradation. The SUMOylated Core is then degraded through the ubiquitin-independent PSMB2 and PSMD2 26S protea some system. Collectively, our findings highlight VCP as a key host factor involved in CSFV uncoating, presenting significant potential for the development of broad-spectrum antiviral strategies targeting Flaviviridae.
## RESULTS
## VCP interacts with Core and degrades it
Our previous study has demonstrated that VCP plays a crucial role in the replication of CSFV; however, the association between VCP and CSFV viral proteins remains insufficiently investigated. Based on this knowledge gap, the present study aimed to explore the relationship between VCP and CSFV viral proteins. Specifically, PK-15 cells were transfected with a series of CSFV protein-expressing plasmids and subsequently performed co-immunoprecipitation (Co-IP) assays. The results confirmed a specific interaction between VCP and the CSFV Core protein (Fig. 1A). Additionally, confocal microscopy was used to examine this interaction. The results showed that at 12 hours post-infection (hpi), there was no observable co-localization between Core and VCP (Fig. 1B). At 24 and 48 hpi, a discrete population of VCP co-localized with Core in infected cells, reflecting its targeted recruitment to Core-containing micro-environments, whereas the remaining VCP pool exhibited widespread cytoplasmic distribution (Fig. 1B). Next, the impact of VCP on Core stability was evaluated, and VCP overexpression resulted in a reduction in Core protein, with the Core protein inversely correlating with cellular VCP expression (Fig. 1C). In contrast, siRNA-mediated knockdown of VCP caused an increase in Core protein (Fig. 1D). These findings demonstrate an inverse relationship between the Core protein and VCP expression. Additionally, cells co-transfected with plasmids expressing Core and VCP were collected at specific time points, and western blotting data showed that VCP overexpression accelerated Core degradation (Fig. 1E), whereas the knockdown of VCP delayed the degradation process compared with the siCtrl (Fig. 1F). To further elucidate the critical role of VCP in the degradation of the CSFV Core protein, cycloheximide (CHX) chase experiments were conducted. PK-15 cells were transfected with pEGFP-VCP and subsequently infected with CSFV for 24 h. The cells were then treated with CHX for different time points to inhibit protein synthesis. These results revealed that VCP expression shortened the half-life of the Core protein, compared with that in cells transfected with the empty vector (Fig. 1G). In contrast, knockdown of VCP extended the half-life of the Core protein (Fig. 1H), indicating that VCP accelerated Core degradation. Collectively, these findings demonstrate that VCP directly interacts with the CSFV Core protein and facilitates its degradation.
## Core is degraded by VCP via the proteasome pathway
The proteasomal and autophagy-lysosome pathways represent the two principal mechanisms of intracellular protein degradation in eukaryotic cells. To identify the primary pathway through which VCP facilitates degradation of the CSFV Core, PK-15 cells were transfected with plasmids encoding VCP and subsequently infected with CSFV for 24 h. The cells were then treated with the proteasome inhibitor MG132, the autophagy activator rapamycin, or the autophagy inhibitor bafilomycin A1 (BafA1). Western blotting analysis revealed that treatment with MG132 effectively reversed the VCP-mediated degradation of the Core protein, whereas neither rapamycin nor BafA1 had a compara ble effect (Fig. 2A). Notably, MG132 treatment enhanced the interaction between VCP and Core (Fig. 2B). This observation was further confirmed using Bortezomib-another proteasome inhibitor, which also blocked Core degradation (Fig. 2C)-confirming that Core degradation occurred in a proteasome-dependent manner. These findings strongly suggest that Core degradation is predominantly dependent on the proteasome pathway. To further validate the involvement of the proteasome in Core degradation, siRNAs targeting three proteasome subunits (PSMF1, PSMD2, and PSMB2) were employed to evaluate their roles in this process. Western blotting analysis demonstrated that knockdown of PSMD2 and PSMB2 significantly suppressed CSFV replication (Fig. 2D). Furthermore, infected cells were fixed and subjected to confocal microscopy using specific antibodies. As shown in Fig. 2E, VCP co-localized with PSMD2 and PSMB2 after CSFV infection, with Pearson's correlation coefficient analysis supporting these data. Subsequently, the roles of PSMF1, PSMD2, and PSMB2 in VCP-mediated Core degradation were investigated. PK-15 cells with PSMF1, PSMD2, or PSMB2 knockdown were transfected with plasmids expressing VCP or an empty vector and infected with CSFV for 24 h. The results indicated that knockdown of PSMD2 and PSMB2 reversed Core degradation compared with control cells transfected with siCtrl, whereas PSMF1 knockdown had no effect (Fig. 2F), indicating that Core degradation was specifically mediated by the PSMD2 and PSMB2 subunits. Finally, the cells transfected with pFlag-Core were fixed and stained with the indicated antibodies against PSMF1, PSMD2, and PSMB2 for confocal microscopy. The results showed co-localization of PSMD2 or PSMB2 with VCP and Core (Fig. 2G), whereas PSMF1 did not co-localize (Fig. 2G), as further confirmed by Pearson's correlation coefficient analysis. Collectively, these findings suggest that VCP facilitates Core degradation by recruiting PSMD2 and PSMB2 to form a functional proteasome complex.
## SUMO1 interacts with Core and is essential for the degradation of Core
Proteasome-mediated degradation is traditionally associated with ubiquitination (44). To determine whether ubiquitination contributes to the degradation of the CSFV Core protein, PK-15 cells were exposed to increasing concentrations of the ubiquitination inhibitor Pyr-41. Core protein expressions were subsequently assessed by western blotting analysis. Notably, inhibition of ubiquitination by Pyr-41 did not prevent Core protein degradation (Fig. 3A). To investigate the ubiquitination status of Core, HEK-293T cells were co-transfected with pFlag-Core and pHA-ubiquitin for 24 h. METTL14, a known ubiquitination substrate (45), served as a positive control. Co-immunoprecipitation (Co-IP) assays revealed robust ubiquitination of METTL14, whereas no ubiquitin signal was detected for Flag-Core, the empty vector, or IgG controls (Fig. 3B). These results indicate that Core evades ubiquitin conjugation, reinforcing the conclusion that its proteasomal degradation proceeds via a ubiquitin-independent pathway. Given that SUMOylation, a small ubiquitin-like modification, is implicated in protein degradation (46), and a (G) PK-15 cells transfected with pFlag-Core were fixed and stained with rabbit anti-PSMF1, anti-PSMD2 or anti-PSMB2 antibody
previous study has highlighted a potential interaction between Core and SUMO1 ( 14), we hypothesized that SUMOylation would be involved in Core degradation. To investigate this, cells were treated with the SUMOylation inhibitor 2-D08, which impaired Core degradation even at a relatively low dose (Fig. 3C). In Fig. 3D, Co-IP of Flag-Core followed by immunoblotting with anti-Flag antibody revealed a prominent band with a higher molecular weight compared with the unmodified Flag-Core. This higher-molecularweight band was consistent with the expected size of Flag-Core conjugated to SUMO (SUMOylated Flag-Core), confirming that VCP specifically interacted with the SUMOyla ted form of Core protein rather than its unmodified counterpart. Additionally, overex pression of SUMO1 and SUMO2, but not SUMO3, differentially suppressed Core protein expression, with SUMO1 exhibiting the strongest effect and SUMO2 showing a more modest reduction, revealing a clear gradient of efficacy among these SUMO paralogs (Fig. 3E). Conversely, siRNA-mediated knockdown of SUMO1 led to the stabilization of Core protein compared to the siCtrl group (Fig. 3F). More importantly, we also explored the involvement of Ubc9, the E2-conjugating enzyme in the SUMOylation pathway and confirmed that Core interacted with Ubc9 (Fig. 3G), with an inverse correlation between Ubc9 expression and Core protein (Fig. 3H). Knockdown of Ubc9 increased the abun dance of Core protein expressions (Fig. 3I). Moreover, knockdown of Ubc9 effectively blocked the degradation Core caused by overexpression of SUMO1 (Fig. 3J), and cotransfection of SUMO1 and Ubc9 demonstrated a synergistic effect on Core degradation (Fig. 3K). Collectively, these findings suggest that SUMOylation, orchestrated by SUMO1 and Ubc9, plays a pivotal role in regulating the degradation of the CSFV Core protein. To further assess whether SUMOylation contributes to the proteasomal degradation of the CSFV Core protein, PK-15 cells were pretreated with the proteasome inhibitor MG132, followed by transfection with a pEGFP-SUMO1 overexpression plasmid and infected with CSFV. Cell lysates were harvested at 24 hpi and analyzed by western blotting. Notably, MG132 treatment abrogated the SUMO1-induced degradation of the Core protein (Fig. 3L), indicating that SUMOylation facilitates Core degradation through the proteasomal pathway. Collectively, these findings demonstrate that the CSFV Core protein undergoes proteasome-mediated degradation through a ubiquitin-independent but SUMO1/Ubc9dependent SUMOylation mechanism.
## VCP-mediated degradation requires Core SUMOylation
To ascertain the relationship between SUMOylation-mediated degradation of the Core protein and VCP, as well as to elucidate the underlying molecular mechanism, we first silenced SUMO1 expression in PK-15 cells using siRNA, followed by transfection with a pEGFP-VCP plasmid and subsequent CSFV infection. As anticipated, knockdown of SUMO1 attenuated Core degradation induced by VCP overexpression (Fig. 4A), suggest ing that VCP facilitated Core degradation in a SUMOylation-dependent manner. Previous studies have identified amino acid residues K179, K180, and K221 of the CSFV Core protein as mediators of interaction with SUMO1, whereas residue K220 mediates interaction with Ubc9 (14). Therefore, lysine residues at positions 179, 180, 220, 221, and 246 were individually or collectively substituted with arginine (Fig. 4B). The stability of these Core mutants was assessed under VCP overexpression conditions, revealing that only mutants with all five sites altered, as well as the K220/221 group, remained resistant to degradation (Fig. 4C). This indicates that lysine residues K220/221 are critical for VCPmediated recognition and degradation of Core. We next investigated whether K220/221 served as SUMOylation sites of Core. The results showed that after mutating K220/221 and K5 sites, Core no longer interacted with SUMO1, whereas the wild-type (WT) Core and Core with K179/180 mutation still retained the interaction with SUMO1 (Fig. 4D). Similarly, Core with K220/221 and K5 mutations also lost the interaction with Ubc9 (Fig. 4E). These findings suggest that K220/221 are SUMOylation sites of Core. Furthermore, confocal microscopy was employed to assess the co-localization of VCP with the Core mutants. In alignment with our hypothesis, VCP failed to co-localize with the Core-K5 and K220/221 mutants (Fig. 4F), whereas the Core-WT and other mutants maintained colocalization with VCP (Fig. 4F), reinforcing the notion that K220/221 were essential for VCP-mediated degradation. Collectively, these findings demonstrate that SUMOylation of Core at residues K220/221 is crucial for its recognition and subsequent degradation mediated by VCP.
## VCP facilitates CSFV uncoating by transporting virus to the late endosome
The results presented above have demonstrated that VCP interacts with the CSFV Core protein and mediates its degradation, and the CSFV Core protein serves as the nucleo capsid protein of this virus. This observation has further extended our interest in investigating whether VCP is involved in the uncoating of CSFV. Building upon estab lished JEV genome tracking methodologies. The envelopes of purified CSFV were labeled with streptavidin-modified quantum dots (SA-QDs) by intercalating biotinylated lipids into the lipid layer. In addition, a multivalent fluorescence amplification strategy was used as a facile and competent tool for highly sensitive RNA imaging to label the RNA genome while maintaining the integrity of the viral RNA (47). To determine whether the dual-labeled viruses specifically labeled CSFV, we first evaluated their co-localization with the viral E2 protein using an immunofluorescence assay (IFA). Confocal microscopy revealed that nearly all QD625 and MMBs signals co-localized with E2 signals, whereas negligible signals were observed in the negative control group (Fig. 5A), confirming the specificity of the dual-labeling strategy. We further assessed the impact of dual labeling on viral infectivity. Compared with unlabeled viruses, labeling CSFV with QD625 and MMBs had a minimal effect on its infectivity (Fig. 5B). This efficient dual-labeling approach enabled us to monitor the release behavior of individual RNA genomes at the single-virus level. To observe viral uncoating, dual-labeled viruses were first incubated with PK-15 cells at 4°C for 30 min to allow membrane binding without endocytosis. Subsequently, the temperature was raised to 37°C to initiate infection, and the cells were fixed at different time points. After 15 min, dual-labeled viruses accumulated in the perinuclear region of cells, and yellow signals-representing the overlap of envelope (QD625) and RNA genome (MMBs)-were partially co-localized with Rab5, whereas some remained dispersed in the cytoplasm (Fig. 5C-a). At this stage, minimal yellow fluorescence was observed in Rab7-positive compartments (Fig. 5C-f signals accumulated in Rab7 compartments (Fig. 5C-i andj), marking the completion of genome release. However, upon VCP knockdown, a large number of virus particles remained trapped within Rab5-positive compartments (Fig. 5C-k through o), consistent with our previous observations (23). Notably, strong yellow fluorescence signals were retained within Rab5-positive, but not Rab7-positive compartments (Fig. 5C-p through t), indicating that genome-envelope separation was inhibited. These findings suggest that CSFV uncoating occurs in the late endosomal stage, where the viral RNA genome separates from the envelope. VCP facilitates this uncoating process by promoting endosomal maturation; its depletion restricts virus particles to early endosomes, thereby impairing CSFV uncoating and subsequent infection.
## K220/221 is essential for CSFV transport to the late endosome for uncoating
To probe the functional significance of these residues in viral replication, K-to-R substitu tions were introduced into an infectious full-length cDNA clone of the highly virulent CSFV strain Shimen. The rescued mutant viruses, named CSFV-ΔK220/221 and CSFV-ΔK4, contain a combination of K-to-R substitutions at residues K179, K180, K220, and K221 in the Core protein (Fig. 6A). PK-15 cells were subsequently infected with recombinant viruses harboring site-directed mutations in the Core protein: CSFV-ΔWT, CSFV-ΔK4, or CSFV-ΔK220/221 (MOI = 1). Viral titers were measured at the indicated time points. Notably, mutant viruses CSFV-ΔK4 and CSFV-ΔK220/221 exhibited significantly reduced titers compared with the parental CSFV-ΔWT virus (Fig. 6B), indicating impaired viral replication. Furthermore, to assess whether restoration of SUMOylation sites could rescue viral fitness, a compensatory mutation was introduced into the CSFV-ΔK4 background, generating the CSFV-ΔRK179/180 revertant virus (Fig. 6A). However, the replication capacity of this revertant remained similar to that of CSFV-ΔWT (Fig. 6C), demonstrating that restoration of K220/221 alone is sufficient to fully recover viral replication. PK-15 cells were treated with the proteasome inhibitor MG132 and subse quently infected with CSFV as outlined in Fig. 6D. Administration of MG132 during the first 0-1 hpi and 1-2 hpi markedly suppressed viral replication (Fig. 6D), indicating that proteasomal activity played a pivotal role in the CSFV uncoating process. In parallel, confocal microscopy was employed to assess whether a CSFV Core mutant affects viral trafficking from Rab5-positive early endosomes to Rab7-positive late endosomes. The analysis demonstrated that substitution of the K220/221 residues significantly impaired the translocation of CSFV from Rab5 to Rab7 compartments (Fig. 6E). These findings establish that the SUMOylation sites within Core are critical determinants of CSFV's endosomal trafficking from early to late compartments, thereby facilitating efficient uncoating. In summary, our findings delineate the critical role of lysine residues K220/221 in the degradation of the Core protein, mediated through a SUMOylation-dependent mechanism. Furthermore, the disruption of these residues impairs viral propagation, highlighting their indispensable role in the life cycle of CSFV.
## Core degradation requires the VCP-NPL4-UFD1 complex
VCP exhibits limited intrinsic substrate specificity on its own and relies on various cofactors to execute its diverse functions (48). To identify the potential cofactors of VCP involved in Core degradation, we employed confocal microscopy to assess the interac tions between VCP, Core, and NPL4 or UFD1. In cells transfected with an empty vector, VCP co-localized with both NPL4 and UFD1 (Fig. 7A). In cells transfected with pFlag-Core, both NPL4 and UFD1 co-localized with VCP and Core (Fig. 7A). To further validate these interactions, the Co-IP assay was performed. Cells were transfected with pFlag-Core for 24 h and subjected to Co-IP assay using a Flag monoclonal antibody. The results confirmed that Core interacted with VCP, UFD1, and NPL4 (Fig. 7B). The results above indicate that VCP forms a complex with NPL4 and UFD1 to bind Core. Furthermore, siRNAs targeting NPL4 or UFD1 were transfected into PK-15 cells, which were then infected with CSFV. At 24 hpi, the cells were harvested and analyzed by western blotting.
The results revealed a significant decrease in CSFV Npro protein expression in both NPL4and UFD1-knockdown cells (Fig. 7C), suggesting that NPL4 and UFD1 were involved in regulating CSFV replication. To investigate whether NPL4 and UFD1 participate in the VCP-mediated degradation of Core, we observed that knockdown of NPL4 or UFD1 impeded VCP-mediated Core protein degradation (Fig. 7D). These results confirm that VCP forms a trimeric complex with NPL4 and UFD1 to mediate the degradation of the Core protein. Furthermore, we also investigated the role of UFD1 and NPL4 in the trafficking process of CSFV. The results showed that knockdown of UFD1 and NPL4 impaired the transport of CSFV from Rab5 to Rab7 (Fig. 7E).
## DISCUSSION
VCP is a multifunctional ATPase involved in a variety of cellular processes, regulating protein homeostasis, and has been implicated in the life cycle of multiple viruses (38). In the Flaviviridae family, particularly the Flavivirus genus, VCP plays a pivotal role during early infection stages. It supports the replication cycle of flaviviruses, including ZIKV (26), YFV (24), WNV (29), and DENV (25), facilitating key processes such as viral uncoating, replication organelle formation, and the disassembly of stress granules. Building on our previous work that demonstrated the importance of VCP in CSFV infection through pharmacological inhibition, overexpression, and RNA interference (23). Our current findings highlight VCP's essential role in Core protein degradation and viral uncoating (Fig. 8). First, we demonstrate that VCP physically interacts with the CSFV Core protein, and modulating the VCP expression alters Core protein expression. Second, VCP coordinates endosomal trafficking: VCP knockdown impairs CSFV transit from early to late endosomes, blocking uncoating. Collectively, these data indicate that VCP facilitates viral uncoating by promoting Core degradation, establishing a functional link between VCP-mediated disassembly of viral components and the progression of productive CSFV infection.
Proteasomal degradation is typically initiated via ubiquitination. However, CSFV Core degradation deviates from this paradigm: Core lacks ubiquitination, and its stability is unaffected by the ubiquitin-activating enzyme inhibitor Pyr-41-consistent with previous reports (49). Instead, our results substantiated that the non-ubiquitinated Core engaged the SUMOylation pathway, interacting with SUMO1 and Ubc9, which aligns with findings in macrophages (14). SUMOylation is increasingly recognized as a mechanism hijacked by viruses to regulate host-pathogen interactions and promote replication (50). Our study identified SUMOylation as a critical mediator of CSFV Core uncoating and degradation-a distinct mechanism for Flaviviridae. Specifically, we demonstrated that Core SUMOylation was a prerequisite for its recognition and processing by VCP and the proteasome. RNAi-mediated knockdown of SUMO1 or Ubc9 confirmed that Core degradation depends on SUMOylation, suggesting that this modification may represent a broader viral strategy to bypass the canonical ubiquitin-proteasome pathway. We further mapped the key residues driving Core SUMOylation: lysines K220 and K221 were indispensable for SUMO1/Ubc9-mediated modification, which in turn enabled recogni tion and degradation by VCP. This aligns with prior reports that SUMOylation acts as a proteasomal degradation signal in a ubiquitin-independent manner, particularly in the context of viral protein quality control. Critically, mutating K220/221 not only stabilized Core but also severely impaired CSFV replication, underscoring these residues' essential role in the viral life cycle. Notably, compensatory restoration of K220/221 failed to rescue replication, suggesting that SUMOylation requires context-dependent (and potentially sequential) modification, involving cooperative action of multiple lysine residues. These results highlight a previously unappreciated mechanism by which host SUMOylation machinery and VCP coordinate to control CSFV protein turnover and infectivity, offering potential targets for antiviral intervention.
Previous reverse genetics studies have demonstrated that although the CSFV Core protein is not strictly required for in vitro virion assembly, it is pivotal for efficient replication and in vivo pathogenicity. For instance, deleting most of the Core gene (VP447ΔC) severely impairs viral proliferation, which can be partially rescued by a compensatory N2177Y point mutation in the NS3 protein (51). Although this mutation restored particle production in vitro, the resulting virus was completely attenuated in pigs, suggesting that NS3 partially substitute Core in virion assembly, but not in mediating full infectivity or pathogenicity. Recent work further underscores that genomic RNA harboring NS3 compensatory mutations can be packaged without Core, emphasizing the cis-acting role of NS3 in genome encapsidation (52). These findings align with those of our current study, in which we demonstrate that Core not only contributes to viral infection but also undergoes VCP-and SUMOylation-dependent degradation during uncoating, a process critical for efficient CSFV infection. Although NS3 may structurally compensate for Core in certain contexts, the regulated degradation of Core is likely required for proper uncoating and early infection events-functions that NS3 cannot fully replicate. Together, these observations underscore the multifaceted role of the Core protein in both structural integrity and dynamic post-entry processes essential for viral propagation.
VCP activity is primarily modulated through its interaction with distinct cofactors. We identified that UFD1 and NPL4 bind to VCP, forming a complex involved in the degrada tion and transport of Core. Knockdown of VCP, UFD1, or NPL4 inhibited CSFV replication. Notably, UFD1 or NPL4 knockdown prevented VCP-induced Core degradation. These findings align with previous reports showing that VCP plays a role in the degradation of other viral proteins, such as the coronavirus nucleocapsid N protein (34) and the ZIKV capsid protein through a ubiquitin-dependent pathway (26).
Despite defining the role of SUMOylation in Core degradation during uncoating, our study has several limitations. First, the lack of suitable Core-specific antibodies prevented direct evaluation of how SUMOylation alters the RNA-binding affinity of Core; instead, we relied on indirect readouts such as viral replication, protein stability, and mutant phenotypes. Second, although the K220/221 mutations likely disrupt SUMOylation, potential impacts on RNA binding cannot be excluded. To address this, we examined mutations at a distinct lysine site (K179/180), which had no impacts on replication, supporting the specificity of the K220/221 phenotype. Finally, SUMOylation appears to function post-entry, as modifications observed during ectopic expression may not reflect physiological regulation during assembly. Clarifying the timing and dynamics of Core SUMOylation will be an important direction for future work.
In conclusion, our study is the first to demonstrate that SUMOylation of the Core protein plays a pivotal role in CSFV uncoating. We propose a model where CSFV utilizes VCP to extract SUMOylated Core, recruit proteasome subunits to form a proteolytic complex, and release its genome following uncoating (Fig. 8). These findings not only deepen our understanding of the CSFV life cycle but also reveal potential therapeutic targets for antiviral strategies. Targeting the VCP-SUMOylation pathway could offer a novel approach for inhibiting CSFV replication, contributing to the control and eradica tion of CSF. Future studies should explore the broader implications of SUMOylation in viral uncoating and evaluate its potential as a therapeutic target across various viral families. objective. Sequential scanning was performed to prevent spectral overlap. Imaging parameters-including laser intensity, exposure time, and detector gain-were carefully optimized to prevent signal saturation and ensure accurate fluorescence representation. All images within the same experiment were captured under identical settings. Postacquisition processing was limited to linear adjustments of brightness and contrast using Nikon NIS-Elements software. Co-localization was quantitatively assessed using the Nikon A1 software, and the results were expressed as Pearson's correlation coefficients. Co-localization analysis was performed using NIS-Elements software (Nikon Instruments, Japan). For quantitative evaluation, regions of interest (ROIs) were delineated around areas exhibiting co-localized signals from the two fluorophores, and the Pearson's correlation coefficient was calculated across all pixels within these ROIs. Values of PCC were obtained from at least three independent experiments, and the representative data are presented.
## Coimmunoprecipitation assay and western blotting
Cells co-transfected with pEGFP-VCP and pFlag-E2, -Core, -NS3, -NS4B, -NS5A, or -NS5B, were lysed in NP-40 lysis buffer. A 25% aliquot of the supernatant (whole cell lysate, WCL) was removed from all samples for later use. The remaining 75% of the lysates were treated with mouse anti-Flag antibody or rabbit anti-VCP antibody and incubated with rotation for 6 h at 4°C, and then, 40 µL of a protein A/G Plus-agarose slurry (sc-2003; Santa Cruz) was added to the lysate for 4 h at 4°C with rotation. The agarose beads were washed with NP-40 by centrifugation at 1,000 × g for 5 min at 4°C four times. The agarose beads were collected by centrifugation and resuspended in 2 × SDS loading buffer for SDS-PAGE and western blotting. The cell samples were washed three times with ice-cold PBS and then lysed in RIPA lysis buffer (R0020, Solarbio) supplemented with protease inhibitors (Sigma) for 30 min at 4°C. Lysates were clarified by centrifugation at 12,000 × g for 10 min at 4°C. A 120 µL aliquot of the supernatant was removed from all samples for later use, then resuspended in a 5 × SDS loading buffer. Protein samples were separated by SDS-PAGE and transferred to nitrocellulose membranes. After blocking with 5% skim milk for 1 h at 37°C, the membranes were incubated with primary antibodies overnight at 4°C, followed by incubation with corresponding horseradish peroxidase-conjugated secondary antibodies for 1 h at 37°C. The immunolabeled protein complexes were visualized using the ECL Plus kit (Jacob enzyme Biotech, SQ201) and using the Tanon 5200 scanner system (Tanon, China). β-actin was used as a loading control. To determine the expression of indicated proteins, the corresponding protein/actin quantity was used to calculate the grayscale values using ImageJ 7.0 software. medium, the cells were washed with PBS twice and cultured in the maintenance medium for 3 days.
## Statistical analysis
All data were presented as means ± standard deviations (SD) as indicated. Student's t-test was used to compare the data from pairs of treated and untreated groups. Statistical significance is indicated by asterisks in the figures. All statistical analyses and calculations were performed using Prism 9 (GraphPad Software, Inc., La Jolla, CA)
## References
1. Sun, Shi, Guo et al. (2010) "Changes in the porcine peripheral blood mononuclear cell proteome induced by infection with highly virulent classical swine fever virus" *Journal of General Virology*
2. Postel, Austermann-Busch, Petrov et al. (2018) "Epidemiology, diagnosis and control of classical swine fever: recent developments and future challenges" *Transbound Emerg Dis*
3. Ganges, Crooke, Bohórquez et al. (2020) "Classical swine fever virus: the past, present and future" *Virus Res*
4. De Smit (2000) "Laboratory diagnosis, epizootiology, and efficacy of marker vaccines in classical swine fever: a review" *Veterinary Quarterly*
5. Johns, Bensaude, Rocca et al. (2010) "Classical swine fever virus infection protects aortic endothelial cells from pIpC-mediated apoptosis" *Journal of General Virology*
6. Lamp, Riedel, Sosa et al. (2011) "Biosynthesis of classical swine fever virus nonstructural proteins" *J Virol*
7. Liu, Liu, Zhou et al. (2022) "Cellular ESCRT components are recruited to regulate the endocytic trafficking and RNA replication compartment assembly during classical swine fever virus infection" *PLoS Pathog*
8. Cheng, Lou, Liu et al. (2021) "Microfilaments and microtubules alternately coordinate the multistep endosomal trafficking of classical swine fever virus" *J Virol*
9. Chen, Song, Hu et al. (2024) "Classical swine fever virus non-structural protein 5B hijacks host METTL14-mediated m6A modification to counteract host antiviral immune response" *PLoS Pathog*
10. Guo, Zhang, Liu et al. (2023) "Attachment, entry, and intracellular trafficking of classical swine fever virus" *Viruses*
11. White, Delos, Brecher et al. (2008) "Structures and mechanisms of viral membrane fusion proteins: multiple variations on a common theme" *Crit Rev Biochem Mol Biol*
12. Heimann, Sosa, Martoglio et al. (2006) "Core protein of pestiviruses is processed at the C terminus by signal peptide peptidase" *J Virol*
13. Xiao, Bai, Xu et al. (2008) "Effect of NS3 and NS5B proteins on classical swine fever virus internal ribosome entry site-mediated translation and its host cellular translation" *J Gen Virol*
14. Gladue, Holinka, Fernandez-Sainz et al. (2010) "Effects of the interactions of classical swine fever virus Core protein with proteins of the SUMOylation pathway on virulence in swine" *Virology (Auckl)*
15. Gladue, Donnell, Fernandez-Sainz et al. (2014) "Interaction of structural core protein of classical swine fever virus with endoplasmic reticulum-associated degradation pathway protein OS9" *Virology (Auckl)*
16. Gladue, Holinka, Fernandez-Sainz et al. (2011) "Interaction between Core protein of classical swine fever virus with cellular IQGAP1 protein appears essential for virulence in swine" *Virology (Auckl)*
17. Erzberger, Berger (2006) "Evolutionary relationships and structural mechanisms of AAA+ proteins" *Annu Rev Biophys Biomol Struct*
18. Stach, Freemont (2017) "The AAA+ ATPase p97, a cellular multitool" *Biochem J*
19. Van Den Boom, Meyer (2018) "VCP/p97-mediated unfolding as a principle in protein homeostasis and signaling" *Mol Cell*
20. Ji, Li, Peterle et al. (2022) "Translocation of polyubiquitinated protein substrates by the hexameric Cdc48 ATPase" *Mol Cell*
21. Liu, Ackermann, Hoffmann et al. (2024) "Concerted SUMO-targeted ubiquitin ligase activities of TOPORS and RNF4 are essential for stress management and cell proliferation" *Nat Struct Mol Biol*
22. Capella, Mandemaker, Caballero et al. (2021) "Nucleolar release of rDNA repeats for repair involves SUMO-mediated untethering by the Cdc48/p97 segregase" *Nat Commun*
23. Sun, Hu, Li et al. (2022) "Valosincontaining protein (VCP/p97) is responsible for the endocytotic trafficking of classical swine fever virus" *Vet Microbiol*
24. Ramanathan, Zhang, Douam et al. (2020) "A sensitive yellow fever virus entry reporter identifies valosin-containing protein (VCP/p97) as an essential host factor for flavivirus uncoating" *mBio*
25. Mazeaud, Pahmeier, Sow et al. (2021) "The biogenesis of dengue virus replication organelles requires the ATPase activity of valosincontaining protein" *Viruses*
26. Gestuveo, Royle, Donald et al. (2021) "Analysis of Zika virus capsid-Aedes aegypti mosquito interactome reveals pro-viral host factors critical for establishing infection" *Nat Commun*
27. Tamura, Fukuhara, Uchida et al. (2018) "Characterization of recombinant flaviviridae viruses possessing a small reporter tag" *J Virol*
28. Arakawa, Tabata, Ishida et al. (2022) "Flavivirus recruits the valosin-containing protein-NPL4 complex to induce stress granule disassembly for efficient viral genome replication" *J Biol Chem*
29. Phongphaew, Kobayashi, Sasaki et al. (2016) "Valosin-containing protein (VCP/p97) plays a role in the replication of West Nile virus" *Virus Res*
30. Sehrawat, Khasa, Prajapat et al. (2021) "Valosin-containing protein/p97 plays critical roles in the Japanese encephalitis virus life cycle" *J Virol*
31. Arita, Wakita, Shimizu (2012) "Valosin-containing protein (VCP/p97) is required for poliovirus replication and is involved in cellular protein secretion pathway in poliovirus infection" *J Virol*
32. Yi, Fang, Zou et al. (2016) "Affinity purification of the hepatitis C virus replicase identifies valosincontaining protein, a member of the ATPases associated with diverse cellular activities family, as an active virus replication modulator" *J Virol*
33. Yi, Yuan (2017) "Aggregation of a hepatitis C virus replicase module induced by ablation of p97/VCP" *J Gen Virol*
34. Wong, Kumar, Tay et al. (2015) "Genomewide screen reveals valosin-containing protein requirement for coronavirus exit from endosomes" *J Virol*
35. Lin, Prendergast, Grey (2017) "The host ubiquitin-dependent segregase VCP/p97 is required for the onset of human cytomegalovirus Full-Length Text Journal of Virology December"
36. *PLoS Pathog*
37. Panda, Rose, Hanna et al. (2013) "Genome-wide RNAi screen identifies SEC61A and VCP as conserved regulators of Sindbis virus entry" *Cell Rep*
38. Carissimo, Chan, Utt et al. (2019) "VCP/p97 is a proviral host factor for replication of chikungunya virus and other alphaviruses" *Front Microbiol*
39. Das, Dudley (2021) "How viruses use the VCP/p97 ATPase molecular machine" *Viruses*
40. Ren, Wang, Zong et al. (2024) "TRIM28-mediated nucleocapsid protein SUMOylation enhances SARS-CoV-2 virulence" *Nat Commun*
41. Wang, Zhao, Zhou et al. (2022) "PIAS1mediated SUMOylation of influenza A virus PB2 restricts viral replication and virulence" *PLoS Pathog*
42. Wilson (2017) "Viral interplay with the host sumoylation system" *Adv Exp Med Biol*
43. Marusic, Mencin, Licen et al. (2010) "Modification of human papillomavirus minor capsid protein L2 by sumoylation" *J Virol*
44. Su, Tseng, Yu et al. (2016) "SUMO modification stabilizes dengue virus nonstructural protein 5 to support virus replication" *J Virol*
45. Sorokin, Kim, Ovchinnikov (2009) "Proteasome system of protein degradation and processing" *Biochemistry (Mosc)*
46. Chen, Song, Hu et al. (2024) "Classical swine fever virus non-structural protein 5B hijacks host METTL14-mediated m6A modification to counteract host antiviral immune response" *PLoS Pathog*
47. Son, Kim, Lim et al. (2023) "SUMOylation-mediated PSME3-20S proteasomal degradation of transcription factor CP2c is crucial for cell cycle progression" *Sci Adv*
48. Liu, Wang, Hu et al. (2021) "In-situ quantitation of genome release of Japanese encephalitis viruses by quantum dot-based single-virus tracking" *Nano Today*
49. Braxton, Southworth (2023) "Structural insights of the p97/VCP AAA+ ATPase: how adapter interactions coordinate diverse cellular functionality" *J Biol Chem*
50. Chen, Zhu, Fan et al. (2019) "Important roles of C-terminal residues in degradation of capsid protein of classical swine fever virus" *Virol J*
51. Fan, Li, Zhang et al. (2022) "SUMOylation in viral replication and antiviral defense" *Adv Sci (Weinh)*
52. Riedel, Lamp, Heimann et al. (2012) "The core protein of classical swine fever virus is dispensable for virus propagation in vitro" *PLoS Pathog*
53. Lamp, Barth, Reuscher et al. (2025) "Essential role of cis-encoded mature NS3 in the genome packaging of classical swine fever virus" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12781936&blobtype=pdf | # Low-replication influenza virus mediates high pathogenicity through an inflammation-driven lung-heart-brain axis in mice
Wenfei Zhu, Zhuoya Xu, Xinglian Wang, Xiyan Li, Zi Li, Guolin Dong, Lei Yang, Ye Zhang, Ruina You, Yousong Peng, Dayan Wang
## Abstract
The outcomes of viral infections typically correlate with viral load in host tissues. In this study, we identified a H3N2 strain A/Environment/Guangxi/44461/2019 (GX19) that induced rapid mortality in mice by 4 days post-infection despite exhibiting low pulmonary replication capacity. Pathological analysis revealed that GX19 at 10 6 TCID 50 (GX19-6) caused more severe lung damage than GX19 at 10 5 TCID 50 (GX19-5), while inducing pulmonary pathology comparable to a H3N8 virus A/Changsha/1000/2022 at 10 6 TCID 50 (CS-6). Both GX19-6 and CS-6 triggered greater cardiac damage than GX19-5. Notably, GX19-6 displayed unique neurovirulence, eliciting significantly more severe brain damage than GX19-5 and CS-6, accompanied by evident cerebral haemorrhage. Gene Set Variation Analysis (GSVA) revealed distinct cardiac gene expression profiles among viral infections. Specifically, GX19-5 up-regulated gene sets associated with arrhythmia, whereas GX19-6 triggered pathways involved in cardiac arrest. Neither of these effects was present in CS-6 infection. In the brain, GX19-6 specifically induced stronger upregulation of cerebral venous thrombosis and acute ischaemic stroke gene sets compared to other groups, consistent with its pronounced neuropathology. Transcriptomic profiling demonstrated significant alterations across all three organs in GX19-6infected mice, showing suppression of T-cell immunity in the lungs and brain alongside elevated systemic inflammation. In the heart, increased inflammation and apoptosis were accompanied by impaired energy metabolism and reduced cardiac function, potentially contributing to the observed hypoxic responses in the heart, lungs, and brain. Collectively, these findings reveal an inflammation-driven lung-heart-brain axis in influenza virus pathogenicity.
## Introduction
The influenza virus is a major infectious agent posing significant threats to global public health. Historically, influenza pandemics have caused catastrophic mortality, as exemplified by the 1918 Spanish flu pandemic, which claimed at least 50 million lives worldwide [1,2]. Seasonal influenza viruses infect approximately a billion people annually, including 3-5 million cases of severe illness [3]. The virus primarily invades the host through the respiratory tract, with clinical outcomes ranging from asymptomatic infection or mild respiratory illness to severe systemic disease and fatal outcomes [4,5]. This broad spectrum of disease severity stems from complex virus-host interactions: viral pathogenicity is determined by immune evasion mechanisms (e.g. NS1 proteinmediated suppression of host antiviral responses) and replication efficiency, while host susceptibility depends on the robustness of innate and adaptive immune defences [6][7][8].
Numerous studies have shown that influenza virus infections extend beyond pulmonary involvement, triggering systemic effects that can induce neuroinflammation and cardiac injury [9,10]. The severity of neurological complications caused by influenza infection varies. Mild cases present with confusion and severe headache, while severe cases may develop into conditions such as encephalitis and Guillain-Barré syndrome [11,12]. Experimental studies in murine models have demonstrated that even in the absence of detectable live virus in brain tissue, influenza virus infections can activate microglia and astrocytes and lead to elevated levels of pro-inflammatory cytokines [13,14]. Epidemiological studies have established a clear association between influenza virus infection and an increased risk of acute myocardial infarction, myocarditis, and heart failure [9,[15][16][17]. Consistent with this, clinical observations frequently report elevated cardiac biomarkers (e.g. cardiac troponin) and electrocardiographic abnormalities during influenza virus infections [18,19]. Further studies using mouse models have shown that the myocardial inflammation and functional abnormalities caused by viral infection are mainly a type of immune pathological damage mediated by an excessive immune response [20].
Typically, higher viral replication correlates with increased tissue damage and worse clinical outcomes [21,22]. Intriguingly, we identified a distinct viral strain with low replicative potential but pronounced pathogenicity during studies of the biological features of H3 subtype influenza viruses in a mouse model. Subsequent typing confirmed this strain as H3N2, which was isolated from live poultry-related environments. In an effort to uncover the molecular mechanisms underlying its unique "low replication but high pathogenicity" phenotype, we conducted a series of experiments, encompassing assessments of tissue tropism, replication ability, histopathological alterations, as well as transcriptome sequencing. The result revealed that this strain may drive lethal outcomes through a previously unrecognized mechanism -a pathogenic lung-heart-brain axis that precipitates fatal disease independent of high viral titres.
## Results
## GX19 virus causes rapid death of mice with limited replication both in vivo and ex vivo
As shown in Figure 1A, when inoculated with viral strains A/Environment/Guangxi/44461/2019 (H3N2) or A/Changsha/1000/2022 (H3N8) at 10 6 TCID 50 (denoted as GX19-6 and CS-6, respectively), all mice died during the observation period. However, GX19-6 inoculated mice exhibited more body weight loss than CS-6 inoculated mice at 1 d post infection (dpi) and 2 dpi, with p-value <0.05 and p-value <0.01, respectively, and thus all 5 mice died 3 days earlier than those of CS-6. By comparison, body weight loss occurred at the first two days post inoculation but was continuously increasing at the following observation period for the mice inoculated with A/Environment/Guangxi/44461/2019 at 10 5 TCID 50 (denoted as GX19-5). The maximum body weight loss was 17%, and all mice survived. In accordance with the body weight change, 5/5, 5/5, and 0/5 mice were dead when inoculated with GX19-6, CS-6, and GX19-5, respectively (Figure 1B). Tissue tropism of both A/ Environment/Guangxi/44461/2019 (GX19) and A/ Changsha/1000/2022 (CS) was investigated by both TCID 50 and RT-qPCR assays. Viruses can only be detected in respiratory tissues. Other tissues, including brains, heart, spleen, liver, eyes, kidneys, and intestinal tissues, were negative for influenza A viruses. Replication ability was also investigated both in vivo and ex vivo for both viral strains (Figure 1C-E). As expected, the CS exhibited robust replication both in respiratory tissues and the MDCK cell line. However, the GX19 virus can only be detected in lung tissue, but not in the nasal turbinate or trachea (Figure 1C&D). In addition, the viral load in lung tissue or MDCK cells for GX19 virus was significantly lower than that for CS viruses (p-value < 0.05 in the t-test). This result indicated that the GX19 virus causes rapid death of mice but with limited replication both in vivo and ex vivo.
## Pathological changes in the mouse lungs, hearts, and brains after viral infections
Subsequent pathological examination of three major organs (lungs, heart, and brain) in GX19-6 infected mice revealed distinct tissue-specific damage patterns associated with the rapid mortality phenotype. Pulmonary histopathology (Figure 2A) demonstrated characteristic influenza-induced damage, with both GX19-6 and CS-6 infections causing severe oedema and inflammatory infiltration at comparable levels, significantly exceeding the moderate pathology observed in GX19-5 infected mice (Figure 2D & Table S1). Cardiac sections exhibited dose-dependent injury (Figure 2B&E and Table S2), where GX19-5 induced only focal inflammation and minor structural disruption, while both GX19-6 and CS-6 infections provoked extensive myocardial damage with dense inflammatory infiltrates. Most strikingly, neuropathological analysis (Figure 2C) uncovered GX19-specific virulence, with both doses inducing cerebral haemorrhage and oedema -effects absent in CS-infected mice. Quantitative assessment (Figure 2F & Table S3) confirmed this neurotropism, showing GX19-6-induced brain injury scores (only two brain samples for GX19-6 due to severe degradation of a mouse brain sample in the H&E experiment) nearly double those of GX19-5 (p-value <0.01), while CS-6 caused minimal neurological damage.
Transcriptomic profiling via RNA-Seq was conducted on lung, cardiac, and brain tissues to systematically characterize host responses to viral infection. Gene Set Variation Analysis (GSVA) of acute disease-associated pathways revealed organ-specific dysregulation patterns. Pulmonary transcriptomes showed significant upregulation of gene sets related to acute respiratory distress syndrome (ARDS), COPD exacerbation, asthma exacerbation, and multiple organ dysfunction syndrome across all three viral infections (GX19-5, GX19-6, and CS-6), with no virus-specific differences in magnitude. Cardiac analysis demonstrated distinct viral strain-dependent effects: while CS-6 infection down-regulated gene sets associated with cardiac arrest, arrhythmias, heart failure, and myocardial infarction, GX19-6 uniquely activated gene sets associated with heart arrest, and GX19-5 specifically induced upregulation of arrhythmia-related gene sets. Cerebral transcriptional changes were particularly striking, with evaluation of five neurological disease pathways (intracerebral haemorrhage, cerebral infarction, cerebral venous thrombosis, acute ischaemic stroke, and brain neoplasms) revealing that GX19-6 infection induced the most pronounced upregulation of cerebral venous thrombosis and acute ischaemic stroke gene sets -a finding consistent with the severe haemorrhagic pathology observed in GX19-6 infected brains. These results demonstrate that while all three virus infections induce acute pulmonary injury pathways, GX19-6 drives unique transcriptional programmes in both cardiac and neural tissues that correlate with its distinct pathological manifestations.
## Transcription analysis of GX19-6 infection
To elucidate the molecular mechanisms of GX19 pathogenicity in mice, we performed a comprehensive transcriptomic analysis. Lung tissue profiling revealed 835 differentially expressed genes (DEGs) post GX19-6 infection when compared to the PBS group (636 upregulated and 199 down-regulated; Figure 3A). These DEGs were mainly enriched in innate and adaptive immune responses, and inflammation, such as imune response-regulating cell surface receptor signalling pathway, tumour necrosis factor superfamily cytokine production (Figure S1A). Functional analysis of DEGs by direction showed that up-regulated DEGs were also enriched in biological processes associated with innate immune responses and inflammation, such as myeloid leukocyte activation and acute inflammatory response (Figure 3B). Notably, down-regulated genes were mainly enriched in immunological processes, especially T-cell immunity, such as T-cell differentiation and lymphocyte development (Figure 3C), suggesting complex immunomodulatory effects of GX19 infection.
Cardiac transcriptomic analysis identified 99 DEGs following GX19-6 infection (83 up-regulated and 16 down-regulated; Figure 3A). These DEGs were enriched in diverse functions, including immune response, steroid hormone secretion, lymphocyte proliferation, apoptotic process, hyperoxia, cell cycle, ATP metabolic process, and so on (Figure S1B). Functional analysis of DEGs by direction showed that upregulated genes demonstrated significant enrichment in immune-related pathways, including humoral immune response and response to tumour necrosis factor, as well as apoptosis (such as hepatocyte apoptotic process), ATP metabolism (such as negative regulation of ATP metabolic process and mitochondrion organization), and oxygen-related processes (such as cellular response to oxygen levels) (Figure 3D). Intriguingly, down-regulated genes were predominantly associated with corticosteroid hormone secretion, regulation of endocrine processes, and modulation of apoptotic signalling pathways (Figure 3E), suggesting potential cardiac-endocrine crosstalk during viral infection.
Transcriptional profiling of brain tissue showed minimal gene expression alterations, with only 12 DEGs detected (11 up-regulated and 1 down-regulated; Figure 3A). These DEGs were enriched in only a few biological processes (Figure S1C). More sensitive Gene Set Enrichment Analysis (GSEA) uncovered subtle but coordinated changes in ribosomal biogenesis and protein translation pathways (Figure 3F), suggesting potential modulation of neuronal protein synthesis machinery following viral infection.
## Specific transcriptome changes of GX19-6
Next, we identified GX19-6-specific transcriptome changes by comparing DEGs identified in GX19-6 and those in GX19-5 and CS-6. In the lung, 157 GX19-6-specific DEGs were identified (140 up-regulated and 17 down-regulated) (Figure 4A). These DEGs were enriched in diverse functions, including proteolysis, peptidase activity, interferon response, and muscle contraction (Figure S2). Functional analysis of DEGs by direction showed that up-regulated DEGs were also enriched in diverse biological processes, including peptidase activity, interferon response, and negative regulation of hydrolase activity (Figure 4C), while the down-regulated DEGs were primarily associated with several biological processes that seem to be irrelevant to viral infections, such as tube size maintenance and muscle cell proliferation (Figure 4D). In the heart, 46 GX19-6-specific DEGs were identified, and all were up-regulated (Figure 4B). These DEGs were enriched in the immune system (such as lymphocyte proliferation), apoptosis (such as cardiac muscle cell apoptotic process), and energy metabolism regulation (such as negative regulation of mitochondrial membrane permeability, negative regulation of ATP metabolic process, and response to hypoxia) (Figure 4E). These transcriptional changes aligned with the more severe pathological manifestations observed in the heart of GX19-6-infected mice.
In the brain, we identified only two GX19-6-specific up-regulated DEGs (Figure S3), which were primarily associated with circadian rhythm regulation and cellular differentiation processes. To elucidate potential mechanisms underlying GX19-6-induced neuropathology, we performed focussed analysis of the top 100 GX19-6 exclusive up-and down-regulated genes. The up-regulated genes were associated with diverse functions such as long-term memory, cellular response to hypoxia, positive regulation of cell activation, and so on (Figure 4F), while the down-regulated genes were predominantly related to immune functions, particularly T-cell immunity (Figure 4G). This suggests that GX19-6 infection may suppress immune responses in the mouse brain.
Immune cells in tissues may have a large influence on the DEG analysis. Thus, we computationally inferred the immune cell compositions in mice's lungs, hearts, and brains using the bulk RNA-Seq data. As shown in Figure S4, in the heart and brain, only minor differences were observed between virus infection groups and PBS group; In the lung, the virus infection groups had higher ratios of monocytes and M1 macrophage, and a lower ratio of CD4 naive T cells compared to the PBS group, suggesting that the observed DEGs in the lung may be influenced by the immune cell composition. However, when targeting GX19-6, we observed overall similar immune cell compositions across all three tissues compared to both CS-6 and GX19-5. This similarity suggests that the immune cell composition may have a limited impact on the GX19-6-specific DEG analysis.
## Compensatory innate immune activation by GX19-6
Given the observed T-cell immunity suppression (Figure 3C & Figure 4G), we hypothesized compensatory innate immune activation. Accordingly, we analysed the fold changes of expression of inflammation-related genes in the lungs, heart, and brain after viral infections. As shown in Figure 4H&I, most inflammatory factors in GX19-6 virus infections were up-regulated in the lung and heart. The fold changes of inflammatory factors in GX19-6 infections were higher than those in GX19-5 and CS-6 in the heart and were comparable to those in CS-6 in the lung and brain. Some factors, such as Cd14 and Msr1 (marked by stars), showed significantly larger fold changes in GX19-6 infections than those in GX19-5 and CS-6 infections in all three organs. These results suggest systemic innate immune hyperactivation as a potential compensatory mechanism for T-cell suppression.
Then, we analysed the mechanisms underlying the evaluated expressions of inflammation-related genes. We firstly analysed the expression changes of three kinds of pattern recognition receptors (PRRs) after viral infections, including TLR (Toll-like receptor), RLR (RIG-I-like receptor), and NLR (NOD-like receptor) (Figure S5). In the lung, both GX19-6 and CS-6 showed significant up-regulations in the TLR and NOD-associated pathways compared to the PBS group. For the RLR, only GX19-6 showed some extent of up-regulation. In both heart and brain, three PRRassociated pathways did not show up-regulation in viral infections. In most cases, GX19-6 showed similar or lower fold changes compared to the other two viruses.
The high viral inoculum in virus infections of mice may lead to cell death, which can result in inflammation independent of PRRs. Thus, we compared the gene expression levels of four cell-death-related pathways, including apoptosis, ferroptosis, necroptosis, and pyroptosis, between three viruses in three tissues. As shown in Figure S6, for the apoptosisassociated pathway, all three viruses had no obvious up-regulations in the three tissues. For the ferroptosis and pyroptosis-associated pathways, GX19-6 showed the highest fold changes among the three viruses in all tissues, especially in the lung and heart. For the necroptosis-associated pathway, GX19-6 showed comparable (lung) or lower (heart and brain) fold changes than the other two viruses.
## Specific transcriptome changes of CS-6
We also identified CS-6-specific transcriptome changes by comparing DEGs identified in CS-6 and those in GX19-5 and GX19-6 (Figure 4A&B, Figure S3). In the lung, 447 CS-6 specific DEGs were identified (345 up-regulated and 102 down-regulated) (Figure 4A). These DEGs were mainly enriched in the immune system and response to viruses (Figure S7A). The up-regulated DEGs also showed significant enrichment in immune processes and response to viruses, such as regulation of immune effector process, regulation of interleukin-6 production, and response to viruses (Figure S7B), while no biological processes were enriched in the down-regulated DEGs. In the heart, 1464 CS-6-specific DEGs were identified (1065 up-regulated and 399 down-regulated) (Figure 4B). These DEGs were mainly enriched in adaptive immunity, cell cycle, inflammation, and response to viruses (Figure S7C). The up-regulated DEGs were enriched in the immune system, such as regulation of immune effector processes and regulation of T-cell activation (Figure S7D), while the down-regulated DEGs were enriched in the muscle system and development (Figure S7E). In the brain, 16 CS-6-specific DEGs were identified (6 up-regulated and 10 down-regulated) (Figure S3). They were enriched in only one biological process (defence response to viruses) (Figure S7F). The up-regulated DEGs were also mainly enriched in innate immunity and response to viruses, such as response to type 1 interferon, interferon-alpha production, and defence response to viruses (Figure S7G), while no enriched functions were observed for the down-regulated DEGs.
## Discussions
The severity of influenza virus infection is generally associated with viral load in the host. Here, we observed that the GX19 virus caused significantly greater damage in mice at high doses compared to low doses. Although GX19 exhibited much weaker replication in the lungs than the CS virus, with substantially lower viral titres, it led to faster mortality and more severe pathology in multiple organs. This suggests that GX19 may employ a novel mechanism to induce fatal outcomes in mice. Pathological analysis revealed that GX19-6 infection caused severe damage to the lungs, heart, and brain, which was more pronounced than that induced by GX19-5 or CS-6. Notably, brain damage in GX19-6-infected mice was particularly severe, with significant haemorrhage and oedema, indicating that rapid death may result from acute brain injury. Transcriptomic data showed that GX19-6 suppressed T-cell immunity in the lungs and brain while elevating systemic inflammation, possibly due to cell death (ferroptosis or pyroptosis). In the heart, increased inflammation and apoptosis were accompanied by reduced energy metabolism and weakened cardiac function. Hypoxic responses were observed in the heart, lungs, and brain (Figure 4E&F and Figure S6). Based on these findings, we propose the following hypothesis (Figure 4J): GX19 infections in the mice lung trigger excessive systemic inflammation, leading to cardiac inflammation, apoptosis, and metabolic dysfunction. This impairs heart function, causing cerebral hypoxia, thrombosis, and stroke, ultimately resulting in the rapid death of mice.
Previous studies have reported that influenza virus infection in the lungs can lead to damage in other organs such as the heart, kidneys, and brain [23][24][25]. However, these studies typically involved high-dose viral infections where the virus directly entered the bloodstream to affect multiple organs [26]. For example, research by Zheng et al. demonstrated that severe influenza infection could lead to viremia, allowing the virus to directly infect various organs and cause sepsis [27]. Alternatively, some neurotropic strains, such as A/Vietnam/1203/04 (H5N1), can invade the central nervous system through neural pathways [28]. In contrast, our study found that GX19 exhibited poor replication in the lungs, and no virus was detected in the heart and brain by TCID 50 analysis (data not shown). This suggests that GX19 induces rapid mortality through a mechanism distinct from previously reported pathways. As outlined earlier, we propose that GX19 triggers inflammationmediated lung-heart-brain axis dysfunction, ultimately leading to fatal outcomes without inducing a classic cytokine storm. However, the precise mechanisms by which pulmonary inflammation causes cardiac injury and subsequent cerebral haemorrhage remain unclear and warrant further investigation.
The dissociation between viral replication and pathogenicity observed with GX19 challenges conventional paradigms of influenza virulence and highlights the need to explore alternative pathogenic mechanisms beyond viral load-dependent models. Our findings suggest that certain influenza strains may cause severe disease through indirect, inflammationdriven multi-organ failure rather than direct viral cytopathic effects or classical cytokine storms. This novel pathogenic mechanism warrants further investigation into the specific viral factors responsible for triggering such disproportionate inflammatory responses despite limited replication, as well as the molecular pathways connecting pulmonary inflammation to remote organ damage. Understanding these alternative virulence mechanisms could lead to new therapeutic strategies targeting inflammationmediated organ dysfunction in severe influenza cases.
## Conclusion
This study represents the first identification of a unique H3N2 influenza strain capable of inducing rapid mortality in mice despite demonstrating attenuated pulmonary replication. Through systematic pathological evaluation coupled with multi-organ transcriptomic profiling, we have elucidated a previously unrecognized pathogenic mechanism wherein the virus triggers fatal outcomes via a cascade of inflammation-mediated events along the lung-heartbrain axis, culminating in catastrophic cerebral injury. These findings not only expand our understanding of influenza virus pathogenesis beyond conventional replication-dependent models but also provide critical insights for developing targeted therapeutic interventions against severe influenza cases.
## Materials and methods
## Cell culture, viral propagation, and growth kinetics
Madin-Darby canine kidney (MDCK) cells were maintained in Dulbecco's Modified Eagle's Medium (DMEM; Invitrogen, Carlsbad, CA, USA) supplemented with 10% foetal bovine serum (FBS; Invitrogen), penicillin (100 U/mL), and streptomycin (100 μg/mL; Invitrogen). Two H3 subtype influenza virus strains -A/Environment/Guangxi/44461/2019 (GX19, H3N2) and A/Changsha/1000/2022 (CS, H3N8) [29] -were propagated in 9-to 11-day-old embryonated chicken eggs. Viral titres were quantified by determining the 50% tissue culture infectious dose (TCID 50 ) in MDCK cells. For viral growth kinetics analysis, MDCK cells were infected with either GX19 or CS virus at a multiplicity of infection (MOI) of 0.001. Following infection, cells were incubated at 37°C in infection medium containing 2 mg/mL N-ptosyl-L-phenylalanine chloromethyl ketone (TPCK)treated trypsin (Sigma, St. Louis, MO, USA). Culture supernatants were harvested at 24, 48, 72, and 96 h post-infection (hpi), and viral titres were determined via TCID₅₀ assay on MDCK cells as previously described [30].
## Pathogenicity and replication kinetics of GX19 in mice
Specific-pathogen-free (SPF) female C57BL/6 mice (8-10 weeks old) were obtained from Vital River Laboratories (Beijing, China). Mice (n = 5 per group) were anesthetized with isoflurane and inoculated intranasally with 50 μL of either 10 5 TCID 50 of GX19 (H3N2), 10 6 TCID 50 of GX19, or 10 6 TCID 50 of CS (H3N8). A control group (n = 5) received 50 μL of phosphate-buffered saline (PBS). Body weight and clinical signs were monitored daily for 14 days postinfection (dpi). Mice exhibiting >25% weight loss were humanely euthanized and recorded as fatalities. For tissue tropism analysis and viral replication analysis, groups of mice were intranasally challenged with 10 6 TCID 50 of the respective viruses. At 1 and 4 dpi, three mice per group were euthanized, and tissues, including the nasal turbinate, trachea, lung, brain , heart, spleen, liver, eyes, kidney, and intestinal tissues, were collected. Viral titers in homogenized tissues were quantified by TCID₅₀ assay [30], with all lung tissues weighed prior to homogenization and the resulting TCID₅₀ values expressed as per gram of tissue. Virus detections were verified additionally by RT-qPCR assay. Briefly, RNA standards (10¹ -10⁷ copies/μL) were prepared from M gene recombinant plasmids to generate a standard curve (R²≥0.99). Viral RNA was extracted using the RNeasy Mini Kit (Qiagen). TaqMan-based RT-qPCR targeting the influenza M gene was performed on a real-time PCR system: 42°C for 30 min, 95°C for 5 min, followed by 40 cycles of 95°C for 15s and 60°C for 30s. Viral copy numbers were calculated via the standard curve equation and adjusted for elution volume and dilution factor to obtain the original sample's viral load.
## Hematoxylin and eosin (H&E) staining
At 4 dpi, lung, heart, and brain tissues were collected from C57BL/6 mice and immediately fixed in 4% paraformaldehyde at 4°C for 24 h. The fixed tissues were dehydrated in a graded ethanol series (70%, 80%, 90%, 95%, and 100%, 1 h each), cleared in xylene (twice, 15 min each), and embedded in paraffin. Paraffin blocks were sectioned at a thickness of 3∼4 μm using a rotary microtome (Leica RM2235, Leica Biosystems, Germany). One tissue section was randomly obtained for each mouse's lung, heart, and brain. Tissue sections were deparaffinized in xylene (twice, 10 min each), rehydrated through a descending ethanol series (100%, 95%, 90%, 80%, and 70%, 5 min each), and rinsed in distilled water. They were immersed in haematoxylin for 5 min, differentiated in 1% acid ethanol for 30 s, rinsed in running tap water for 5 min, and then counterstained with eosin for 2 min.
## Lung injury assessment
Microscopic images were captured using an Olympus BX53 microscope equipped with a DP74 digital camera at 20× magnification. Uniform settings for brightness, contrast, and thresholding were applied during image acquisition and analysis to ensure consistency across samples. Lung injury was semi-quantitatively assessed on H&E-stained lung tissue sections by two blinded pathologists based on three histological features: alveolar oedema, structural damage, and inflammatory cell infiltration. Each parameter was scored on a scale from 0 to 4, where 0 indicated no abnormality, 1 indicated minimal changes, 2 indicated mild changes, 3 indicated moderate changes, and 4 indicated severe or extensive damage. This results in a total injury score ranging from 0 to 12. For each H&E-stained section, at least five non-overlapping high-power fields (40×magnification) (ROI) were selected from representative regions, including the alveolar area, perivascular space, and bronchiolar surroundings, to ensure a comprehensive assessment of lung pathology [31].
## Cardiac injury assessment
Cardiac injury was assessed semi-quantitatively on H&E-stained heart tissue sections based on two key histopathological parameters: inflammatory cell infiltration and structural damage. For each parameter, severity was scored on a scale from 0 to 4. The total cardiac injury score ranged from 0 to 8. Histological evaluation was performed independently by two blinded pathologists to ensure objectivity and consistency. Sections were examined in at least five non-overlapping high-power fields (HPFs, 40× magnification) selected from representative regions of the left ventricular wall, interventricular septum, and subendocardial area to ensure comprehensive assessment of myocardial injury.
## Brain injury assessment
Brain tissue sections (4 μm thick) were stained with haematoxylin and eosin (H&E) according to standard protocols. Histopathological evaluation was independently performed by two blinded investigators under 40× objective magnification. Three key parameters were semi-quantitatively assessed within predefined regions of interest (ROIs): (1) Perivascular inflammatory cell infiltration, scored on a scale of 0-3 (0 = none; 1 = mild, scattered cells; 2 = moderate, partial or complete perivascular cuffing; 3 = severe, multilayered cuffing or diffuse infiltration), with the final score calculated as the mean from two representative ROIs; (2) Cerebral oedema, evaluated by interstitial sparsity and scored from 0 to 3 (0 = compact tissue; 1 = mild focal spacing; 2 = moderate diffuse spacing; 3 = severe vacuolization or rarefaction), excluding regions adjacent to large vessels or haemorrhage; and (3) Haemorrhage and vascular abnormalities, assessed by counting the number of haemorrhagic foci per section and determining the proportion (%) of vessels exhibiting wall thickening or red blood cell (RBC) extravasation, with at least 10 vessels evaluated per ROI.
## RNA extraction, library preparation, and RNA-Seq
Total RNA was extracted from homogenized lung, brain, and heart tissues, which were collected at 4 dpi, using the TRIzol reagent (Qiagen), and were purified using a RNeasy Plus Universal Mini kit (Qiagen) with DNase I digestion, according to the manufacturer's instructions. The RNA integrity was verified using an Agilent 2100 Bioanalyzer (RNA Integrity Number, RIN≥7.0) (Table S4). The RNA quality was assessed using a NanoDrop spectrophotometer (A260/A280 ≥ 1.8. A260/A280 ≥ 2.0). The cDNA library was prepared using NEBNext® Ultra™ II RNA Library Prep Kit for Illumina (NEB) according to the manufacturer's protocol. Briefly, mRNA was purified from 2 ug of total RNA using oligo (dT) magnetic beads. Divalent cations were used to fragment the purified mRNA into small pieces at 94°C for 5 min; thereby priming bias was avoided when synthesizing the cDNA. The cleaved RNA fragments were used for double-stranded cDNA synthesis with random hexamer (N6) primers. The synthesized cDNA was subjected to end repair and a-Tailing processes before ligation of the adaptors. The end products were enriched by PCR to create the final cDNA library with sequences of approximately 300 bp. The libraries were analysed for size using Agilent 5200 (Agilent Technologies, Santa Clara, CA), and the concentration was determined using a Qubit™ 3 Fluorometer with the Qubit™ 1X dsDNA HS Assay Kit and Qpcr. The final DNA libraries were sequenced on the Illumina Novaseq 6000 platform (Illumina, San Diego, CA, USA) using V4 sequencing chemistry to generate approximately 30 million reads per sample.
## RNA-Seq data processing and differential gene expression analysis
Raw sequencing reads were trimmed for adapters and low-quality bases using fastp (v0.23.2) [32]. Clean reads were aligned to the Mus musculus reference genome (GRCm39) from NCBI using HISAT2 (v2.2.1) with default parameters [33]. Gene-level read counts were obtained using FeatureCounts (v2.0.1) [34]. To evaluate global transcriptomic variation across samples and to identify potential outliers or batch effects, principal component analysis (PCA) was performed using the prcomp function in R (v4.3.1). Raw count data were directly analysed using the DESeq2 package (v1.44.0) [35]. Differentially expressed genes (DEGs) were identified based on an adjusted p-value < 0.05 and an absolute log₂(fold change) ≥ 1. The DEGs of each virus in different tissues were obtained by comparing RNA-Seq data of virus infection samples with those of the PBS group.
## Gene function enrichment analysis
Gene Ontology (GO) enrichment analysis was conducted using the clusterProfiler package (v4.12.6) in R [36]. GO terms with p-value < 0.05 and q-value < 0.05 were considered significantly enriched. To further examine the functional relationships among enriched biological process (BP) terms, clustering was performed using rrvgo (v1.16.0), which groups GO terms based on their semantic similarity [37].
GSEA analysis of KEGG pathways was performed using the clusterProfiler (v4.12.6) R package.
Genes associated with several acute diseases in the lung, heart, and brain were curated from the DisGe-NET database [38]. Using these disease gene sets, we performed Gene Set Variation Analysis (GSVA) with the R package GSVA (v1.52.3) [39] to transform the gene expression matrix into a disease signature enrichment score matrix. Subsequently, the differential activity of these disease signatures between experimental groups and PBS controls was assessed using linear models and empirical Bayes moderation implemented in the limma package (v3.64.3).
## Inflammation response genes and hypoxia response genes
Inflammation response genes were curated from the predefined HALLMARK_INFLAMMATORY_RE-SPONSE gene set in the Molecular Signatures Database (MSigDB). Hypoxia response genes were obtained from genes reported by Ebersole et al., 2018[40].
## Cell death-related genes and PRR-related genes
Genes associated with four distinct cell death pathways were derived from the MSigDB database, including apoptosis (HALLMARK_APOPTOSIS), pyroptosis (REACTOME_PYROPTOSIS), necroptosis (GOBP_NECROPTOTIC_SIGNALING_PATH-WAY), and ferroptosis (GOBP_FERROPTOSIS). The genes related to pattern recognition receptors (PRRs) were sourced from the Gene Ontology (GO) database, specifically encompassing three key signalling pathways: the Toll-like receptor (TLR) signalling pathway (GO:0002224), the RIG-I signalling pathway (GO:0045089), and the NOD-like receptor (NLR) signalling pathway that merged genes from nucleotidebinding oligomerization domain containing 1 signalling pathway (GO:0070427) and nucleotide-binding oligomerization domain containing 2 signalling pathway (GO:0070431).
## Statistical analysis
All statistical analyses were conducted in R (version 4.3.1). Comparison of mice body weight loss or virus titres and comparison of injury scores between viruses were conducted with the Student's t-test using the t.test() function in R. Comparison of fold changes of inflammation response genes in mice lung, heart, and brain between different viruses was conducted with a paired Wilcoxon rank-sum test using the pairwise.wilcox.test() function in R. A p-value of less than 0.05 was considered statistically significant.
## References
1. Johnson, Mueller (2002) "Updating the accounts: global mortality of the 1918-1920" Spanish" influenza pandemic" *Bull Hist Med*
2. Patterson, Pyle (1991) "The geography and mortality of the 1918 influenza pandemic" *Bull Hist Med*
3. Organization, Influenza (2025)
4. Flerlage, Boyd, Meliopoulos (2021) "Influenza virus and SARS-CoV-2: pathogenesis and host responses in the respiratory tract" *Nat Rev Microbiol*
5. Koo, Lim, Choe (2018) "Radiographic and CT features of viral pneumonia" *Radiographics*
6. Soares, Teixeira, Moita (2017) "Disease tolerance and immunity in host protection against infection" *Nat Rev Immunol*
7. Iwasaki, Medzhitov (2015) "Control of adaptive immunity by the innate immune system" *Nat Immunol*
8. Das, Ma (2008) "Structural basis for suppression of a host antiviral response by influenza A virus" *Proc Natl Acad Sci*
9. Filgueiras-Rama, Vasilijevic, Jalife (2021) "Human influenza A virus causes myocardial and cardiacspecific conduction system infections associated with early inflammation and premature death" *Cardiovasc Res*
10. Hosseini, Wilk, Michaelsen-Preusse (2018) "Longterm neuroinflammation induced by influenza A virus infection and the impact on hippocampal neuron morphology and function" *J Neurosci*
11. Sivadon-Tardy, Orlikowski, Porcher (2009) "Guillain-Barré syndrome and influenza virus infection" *Clin Infect Dis*
12. Steininger, Popow-Kraupp, Laferl (2003) "Acute encephalopathy associated with influenza A virus infection" *Clin Infect Dis*
13. Sadasivan, Zanin, Brien (2015) "Induction of microglia activation after infection with the non-neurotropic A/CA/04/2009 H1N1 influenza virus" *PLoS One*
14. Wang, Zhang, Li (2008) "Apoptosis and proinflammatory cytokine responses of primary mouse microglia and astrocytes induced by human H1N1 and avian H5N1 influenza viruses" *Cell Mol Immunol*
15. Brown, Pittman, Iii (2011) "Right and left heart failure in severe H1N1 influenza A infection" *Eur Respir J*
16. De Boer, Riezebos-Brilman, Van Hout (2024) "Influenza infection and acute myocardial infarction" *NEJM Evidence*
17. Ni, Cao, Kinnamon (2025) "Antecedent Flu-like illness and onset of idiopathic dilated cardiomyopathy: The DCM precision medicine study" *Circ Heart Fail*
18. Ison, Campbell, Rembold (2005) "Cardiac findings during uncomplicated acute influenza in ambulatory adults" *Clin Infect Dis*
19. Lippi, Sanchis-Gomar (2021) "Cardiac troponin elevation in patients with influenza virus infections" *Biomed J*
20. Filgueiras-Rama, Vasilijevic, Jalife (2021) "Human influenza A virus causes myocardial and cardiacspecific conduction system infections associated with early inflammation and premature death" *Cardiovasc Res*
21. Jong, Simmons, Thanh (2006) "Fatal outcome of human influenza A (H5N1) is associated with high viral load and hypercytokinemia" *Nat Med*
22. Legge, Braciale (2005) "Lymph node dendritic cells control CD8+ T cell responses through regulated FasL expression" *Immunity*
23. Silverman, Walsh, Santoro (2025) "Influenza-Associated acute necrotizing encephalopathy in US children" *JAMA*
24. Pettilä, Webb, Bailey (2011) "Acute kidney injury in patients with influenza A (H1N1) 2009" *Intensive Care Med*
25. Kenney, Aron, Gilbert (2022) "Influenza virus replication in cardiomyocytes drives heart dysfunction and fibrosis" *Sci Adv*
26. Kalil, Thomas (2019) "Influenza virus-related critical illness: pathophysiology and epidemiology" *Crit Care*
27. Zheng, He, Zuo (2025) "Influenza A virus dissemination and infection leads to tissue resident cell injury and dysfunction in viral sepsis" *EBioMedicine*
28. Kim, Hetman (2020) "Zika virus infects pericytes in the choroid plexus and enters the central nervous system through the blood-cerebrospinal fluid barrier" *PLoS Pathog*
29. Yang, Sun, Gao (2022) "Human infection of avian influenza A H3N8 virus and the viral origins: a descriptive study" *Lancet Microbe*
30. Reed (1938) "A simple method of estimating fifty per cent endpoints" *Am J Hyg*
31. Milross, Hunter, Mcdonald (2024) "Distinct lung cell signatures define the temporal evolution of diffuse alveolar damage in fatal COVID-19" *EBioMedicine*
32. Chen, Zhou, Chen (2018) "Fastp: an ultra-fast allin-one FASTQ preprocessor" *Bioinformatics*
33. Kim, Langmead, Salzberg (2015) "HISAT: a fast spliced aligner with low memory requirements" *Nat Methods*
34. Liao, Smyth, Shi (2014) "Featurecounts: an efficient general purpose program for assigning sequence reads to genomic features" *Bioinformatics*
35. Love, Huber, Anders (2014) "Moderated estimation of fold change and dispersion for RNA-Seq data with DESeq2" *Genome Biol*
36. Yu, Wang, Han (2012) "Clusterprofiler: an R package for comparing biological themes among gene clusters" *Omics: A Journal of Integrative Biology*
37. Sayols (2023) "Rrvgo: a bioconductor package for interpreting lists of gene ontology terms" *MicroPubl Biol*
38. Piñero, Bravo, Queralt-Rosinach (2017) "DisGeNET: a comprehensive platform integrating information on human disease-associated genes and variants" *Nucleic Acids Res*
39. Hänzelmann, Castelo, Guinney (2013) "GSVA: gene set variation analysis for microarray and RNA-Seq data" *BMC Bioinformatics*
40. Ebersole, Novak, Orraca (2018) "Hypoxia-inducible transcription factors, HIF1A and HIF2A, increase in aging mucosal tissues" *Immunology*
41. Chen, Chen, Zhang (2021) "The genome sequence archive family: toward explosive data growth and diverse data types" *Genomics Proteomics Bioinformatics*
42. (2025) "Database Resources of the National Genomics Data Center" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645914&blobtype=pdf | # Single-cell RNA sequencing highlights the role of epithelialimmune dual features of proximal tubule cells in BK polyomavirus nephropathy
Feng Yang, Xutao Chen, Hui Zhang, Shicong Yang, Huifei Yang, Peisong Chen, Guodong Zhao, Yingzhen He, Siyan Meng, Dongfeng Yin, Qian Li, Jiang Qiu, Gang Huang
## Abstract
Up to 10% of renal allograft failures are caused by BK polyomavirus nephropathy (BKPyVN). However, there is no specific antiviral agent for BKPyVN. The only treatment is to reduce the levels of immunosuppression, which is not always practical and increases the risk of rejection. Since targeting the microenvironment is a promising approach, we performed single-cell RNA sequencing (scRNA-seq) on BKPyVN samples and stable allografts to obtain BKPyVN microenvironmental atlases. Interestingly, we identified a novel subpopulation of proximal tubule cells (annotated as IGKC+ PT) with epithelial-immune dual features that may contribute to the progression of BKPyVN through T-cell exhaustion. Additionally, we determined that the IGKC+ PT subpopulation might serve as a non-invasive diagnostic marker through scRNA-seq of urine samples and co-immunofluorescence staining. These results improve our understanding of the BKPyVN microenvironment and may guide the development of new therapeutic and diagnostic approaches for a wide range of patients. IMPORTANCE BKPyVN severely threatens kidney transplant recipients. Due to the lack of effective drugs against BK polyomavirus (BKPyV), reducing immunosuppressant therapy is the only treatment. Unfortunately, this approach is not always effective and increases the acute rejection risk. A growing body of research suggests that potential therapeu tic targets may be identified by studying the disease microenvironment. However, traditional methods have not explained why the large number of infiltrating T cells in the BKPyVN microenvironment does not effectively clear BKPyV. Newly available large-scale scRNA-seq technology can be used to study gene expression at a single-cell resolu tion, offering a new way to investigate the BKPyVN microenvironment. By combining scRNA-seq with experimental analysis, we found a novel subpopulation of proximal tubule cells (annotated as IGKC+ PT) with epithelial-immune dual features that may contribute to the progression of BKPyVN through T-cell exhaustion.
A growing body of research suggests that the disease microenvironment can be targeted as a therapeutic strategy (5,6). Previous studies observed that a large number of T cells infiltrate the BKPyVN microenvironment, but these T cells are ineffective in clearing BKPyV (7,8). A recent study demonstrated that BKPyV overlapping peptide pools can induce PD-1 expression in T cells, a well-established biomarker of T-cell exhaustion (9). Another study documented a significant correlation between T-cell exhaustion and BKPyV clearing time in renal transplant patients (8). In the BKPyVN microenvironment, T cells may be in an exhausted state, resulting in the failure to eliminate virus-infected cells. Furthermore, proximal tubule (PT) cells serve as the natural target cells of BKPyV (10)(11)(12), and they may inhibit T-cell activation via PD1/PD-L1 under inflammatory conditions (13). Therefore, a comprehensive depiction of the ecosystem and immune phenotypes in BKPyVN at a single-cell resolution is urgently needed.
Single-cell RNA sequencing (scRNA-seq) is an unbiased method of studying the transcriptional profiles associated with kidney disease (14,15). In this study, we used scRNA-seq data from kidney and urine samples of BKPyVN in combination with experiments to reveal the single-cell atlases of the BKPyVN microenvironment. Our findings may help facilitate the development of novel therapeutic and diagnostic modalities.
## RESULTS
## Identifying cell types in BKPyVN samples via scRNA-seq
We performed scRNA-seq on renal allograft samples from three patients with BKPyVN and two stable allograft (STA) samples (Fig. 1A). After quality control and doublet removal (Fig. S1 andS2), 12,856 single cells were clustered into 14 cell clusters (Fig. 1B). With marker-based annotations, eight major cell types were identified: PT cells (SLC13A1+, ALDOB+, and LRP2+), loop of Henle cells (SLC12A1+), fibroblast (FL) cells (ACTA2+ and COL1A1+), endothelial cells (EC) (PECAM1+ and PODXL+), T and natural killer (T/NK) cells (CD3D+, CD3E+, and CD3G+), B cells (MS4A1 + and CD79A+), myeloid (MD) cells (LYZ+), and mast (MT) cells (TPSB2+ and TPSAB1+) (Fig. 1C). Among these cells, PT and T cells were the two most abundant cell types in the disease microenvironment. Further comparison of the cellular composition in different groups revealed that the number and proportion of PT cells were significantly lower in the BKPyVN group than in the STA group, whereas the number and proportion of T/NK cells were dramatically higher. The number and proportion of other immune cells (B, MD, and MT) were also higher in the BKPyVN group (Fig. 1D).
To provide additional evidence of differences in the proportions of various cell types between the BKPyVN and STA groups, we calculated the ratio of cell types in bulk RNAseq data from 137 samples (28 BKPyVN and 109 STA samples) using MuSiC (v ..1.1). The proportions of major immune cell types (T/NK, MD, MT, and B) were significantly higher in the BKPyVN group than in the STA group (P < 0.05), whereas the proportion of PT cells was significantly lower (P < 0.05) (Fig. S3). Together, these results suggest that PT and T/NK cells might play an essential role in the progression of BKPyVN.
## Epithelial-immune dual features of PT cells in BKPyVN samples
We performed a subpopulation analysis of PT cells to further investigate their role in BKPyVN. To confirm the absence of other cell types, PT cells were defined by the expression of SLC13A1, ALDOB, and LRP2. In total, 4,453 PT cells were clustered into three distinct cell subpopulations (Fig. 2A). We annotated these subpopulations based on the expression of canonical markers: GPX3+ PT (GPX3+ and SLC7A8+), DCXR+ PT (DCXR+ and UPP2+), and IGKC+ PT (IGKC+, IGKM+, VIM+, EPCAM+, SERPINA1+, CLU+, and IL32+) cells (Fig. 2B). The IGKC+ PT subpopulation was only found in the BKPyVN group, and its proportion tended to increase with the progression of BKPyVN (Fig. 2C). LRP2 and IGKC were co-expressed in the same cells (Fig. S4), and the IGKC+ PT subpopulation was present only in BKPyVN samples, according to the results of co-immunofluorescence staining with anti-LRP2 and anti-IGKC (Fig. 2D).
To determine the role of the IGKC+ PT subpopulation in BKPyVN, we performed gene set variation analysis (GSVA) based on BKPyVN-related signaling pathways. According to the 2019 guidelines for BKPyVN published by the American Society of Transplantation (1), the progression of BKPyVN primarily relies on evaluation of inflammation, fibrosis, and BKPyV infection. Therefore, we obtained gene sets related to inflammation, fibrosis, and BKPyV infection from hallmark gene sets and a previous study (16) (Tables S1 andS2). Compared with the DCXR+ PT and GPX3+ PT subpopulations, signaling pathways associated with BKPyVN progression were significantly enriched in the IGKC+ PT subpopulation, indicating that the IGKC+ PT subpopulation was associated with BKPyVN progression (Fig. 2E). To further illustrate the correlation between the proportion of IGKC+ PT cells and BKPyVN progression, we calculated the cell fraction of the IGKC+ PT subpopulation in the bulk RNA-seq data of BKPyVN from 28 samples using MuSiC (v.0.1.1). Based on the mean proportion of IGKC+ PT cells, we categorized BKPyVN patients into high-and low-IGKC+ PT groups. Gene set enrichment analysis (GSEA) results showed that BKPyVN progression-related signaling pathways were significantly enriched in the high-IGKC+ PT group (Fig. 2F; Tables S3 andS4). Using SCENIC (Supple mental Material S1), we also found that the transcription factor RELB, a component of NF-κB that has been reported to regulate BKPyV gene expression and replication (17,18), acted as a specific regulator in the IGKC+ PT cells (Fig. S5). Taken together, these results indicate that the epithelial-immune dual features of IGKC+ PT cells may contribute to the progression of BKPyVN.
## Signaling pathways in the IGKC+ PT subpopulation involved in BKPyVN progression
Based on pseudotime analysis, the direction of PT evolution was consistent with observed changes in groups and subpopulations (Fig. 3A). Therefore, along the pseudotime axis, significantly upregulated genes were involved in promoting the progression of BKPyVN through the IGKC+ PT subpopulation, while downregulated genes were associated with PT function in the normal state (Fig. 3B). Based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, the upregulated genes were enriched in viral replication, immune response, and cell proliferation, while the downregulated genes were enriched in metabolism (Fig. 3B right). Interestingly, the upregulated genes exhibited enrichment in the PD-L1 expres sion and PD-1 checkpoint pathway, indicating that the IGKC+ PT cells might promote T-cell exhaustion in patients with BKPyVN.
Although prior clinical observations and our findings showed a significant increase in the number of T cells in patients with BKPyVN, it remains unclear why these T cells cannot effectively clear BKPyV. Therefore, we investigated the status of T cells in BKPyVN samples by performing a subpopulation analysis of 8,320 T/NK cells. Four distinct cell subpopulations (Fig. 4A) were identified: NK cells (NKG7+ and GNLY+), CD+ T cells (IL7R+ and CD4+), CD8+ T cells (CD8+), and MIX cells (MKI67+ and PCNA+) (Fig. 4B). In CD4+ and CD8+ T cells, we detected expression of markers representing cytotoxicity and exhaustion (Fig. 4B). We further assessed T-cell status by scoring cytotoxicity and cellular depletion and found that T cells in the BKPyVN group had lower cytotoxicity scores and higher cellular depletion scores than those in the STA group (Fig. 4C), suggesting that T cells in the BKPyVN group were in a condition of cellular exhaustion.
## Experimental validation of T-cell exhaustion in BKPyVN samples
We collected 18 blood samples (6 healthy samples, 6 STA samples, and 6 BKPyVN samples) to assess the proportion of exhausted T cells in BKPyVN samples using flow cytometry. The lymphocyte population was initially distinguished using forward corner scatter (FSC-A) and skewed corner scatter (SSC-A) (Fig. 5A). Next, individual cells were defined based on forward scatter area (FSC-A) and height (FSC-H) (Fig. 5B). CD3+ T cells were then circled based on the negative staining control (Fig. 5C), where CD4+ T cells and CD8+ T cells were identified (Fig. 5D). Finally, the proportion of PD-1-positive cells was determined based on the isotype control. The proportion of PD-1-positive cells among CD4+ T cells (Fig. 5E) and CD8+ T cells (Fig. 5F) was significantly higher in the BKPyVN group than in healthy individuals (P < 0.05) and the STA group (P < 0.05). This finding further demonstrates T-cell exhaustion in the BKPyVN immune microenviron ment.
## Regulatory network of cell-cell communication in BKPyVN
We used CellPhoneDB (v.2.1.6) to identify interactions between IGKC+ PT cells and other cell subpopulations in BKPyVN samples. To highlight the role of the IGKC+ PT subpopula tion, we combined GPX3+ PT and DCXR+ PT into PTN cells. According to CellPhoneDB quantitative analyses (Table S5), PDCD1-FAM3C and TIGIT-NECTIN2 exhibited statistically significant interactions within T|IGKC+ PT cells (P < 0.01). The interaction score for PDCD1-FAM3C was 1.016 in the CD4|IGKC+ PT group and 1.121 in the CD8|IGKC+ PT group. The scores for TIGIT-NECTIN2 in the CD4|IGKC+ PT and CD8|IGKC+ PT groups were 0.228 and 0.250, respectively (Fig. 6). These results indicate that the IGKC+ PT subpopula tion suppressed the immune response by exhausting CD4+ and CD8+ T cells via PDCD1-FAM3C and TIGIT-NECTIN2 ligand-receptor pairs. To further validate the relationship between IGKC+ PT and T cells, we demonstrated the expression of ligand-receptor pairs associated with T-cell depletion (Fig. S6). Taken together, these findings indicate that the IGKC+ PT subpopulation could induce T-cell exhaustion via PDCD1-FAM3C and TIGIT-NECTIN2.
## Expression profiling and major cell types in BKPyVN urine samples
To confirm the value of the IGKC+ PT subpopulation as a non-invasive diagnostic marker in BKPyVN urine samples, we performed scRNA-seq using urine samples from patients with BKPyVN (BKU, n = 3), urine samples from healthy volunteers (HHVU, n = 7), and STA kidney samples (n = 2) (Fig. 7A). We used the sctransform and harmony integration methods to annotate 11 cell clusters and then identified 9 major cell types based on classical cell-type markers (Fig. S7; Fig. 7B): PT (ALDOB+, LRP2+, GPX3+, and MT1G+), CD (DEFB1+ and AQP3+), FL (CLOL1A2+, COL3A1+, and ACTA2+), EC (EMCN+ and PECAM1+), urothelial cells (KRT19+), bladder epithelial cells (KRT13+, KRT4+, and KRT17+), T/NK (NKG7+, GNLY+, CD3D+, and CD3E+), B (MS4A1 + and CD79A+), and MD (LYZ+, C1QB+, and C1QA+) (Fig. 7C). Notably, the transcriptional signature of PT cells in urine samples (BKU and HHVU) and STA kidney samples was remarkably similar and nearly identical (Fig. 7C, bottom); these PT cells were found in the same cell cluster (Cluster 0), suggesting their presence in urine. A noticeable number of PT cells were present in urine samples from patients with BKPyVN (Fig. 7D). Additionally, we confirmed that the PT cells in urine samples were derived from the kidneys by integrating single-cell transcriptomic data and kidney spatial transcriptomic data using multimodal intersection analysis (Fig. S8; Material S2). Our analysis of 788 urinary PT cells revealed two distinct subpopulations in urine (Fig. 8A). The MT1F+ PT subpopulation was predominantly found in HHVU urine samples, whereas the IGKC+ PT subpopulation was present only in urine samples from patients with BKPyVN (Fig. 8B andC). Immunofluorescence co-staining with antibodies against LRP2 and IGKC revealed double-positive PT cells (IGKC+ PT) in urine samples from patients with BKPyVN but not in HHVU urine samples (Fig. 8D). The proportion of IGKC+ PT cells was significantly associated with the urinary BKPyV viral load (Pearson correlation = 0.99, P = 0.008; Table S6). The results indicated that the IGKC+ PT subpopulation present in urine might be useful as a non-invasive diagnostic marker for BKPyVN.
## DISCUSSION
Managing BKPyVN remains a clinical challenge (19,20). However, understanding its microenvironmental heterogeneity could help identify new therapeutic targets. We demonstrated the presence of epithelial-immune dual features of the IGKC+ PT subpo pulation in BKPyVN patients and showed that this subpopulation could contribute to the progression of BKPyVN by exhausting T cells. Our findings suggest that new therapeutic approaches targeting the IGKC+ PT subpopulation could be developed.
Determining the cell-type composition in patients with BKPyVN is crucial for under standing disease progression (21,22). We observed a significant decrease in the number of PT cells and a significant increase in the number of T cells in the BKPyVN group compared with those in the STA group. This finding aligns with established clinical features of BKPyVN, which include the renal tubular epithelial cell damage and extensive immune cell infiltration (20). We also discovered significant differences in the proportions of PT and T cells between the BKPyVN and STA groups, suggesting that these two cell types play an essential role in BKPyVN.
## Full-Length Text
We identified novel epithelial-immune dual features of the IGKC+ PT subpopulation in the BKPyVN group through PT subpopulation analysis and immunofluorescence costaining assay. Previous scRNA-seq studies of BKPyV (16,23) systematically characterized the gene signature of BKPyV-infected PT cells and found that genes in these cells are mainly enriched in cell proliferation and immune signaling pathways, in alignment with our current observations. Specifically, we observed HSP90AA1 as a characteristic gene in the IGKC+ PT subpopulation, and this gene was identified in a previous study as a marker of BKPyV infection in PT cells. The IGKC gene encodes the constant domain of the light chain in antibodies. Initially, this gene was believed to be present only in B cells. However, recent studies have shown that it is also expressed in non-B cells, such as various types of tumor cells (24,25), central neurons, cells in the placenta, and testis, and mammary epithelial cells during the proleptic phase (26,27). Previous studies have shown that non-B-cell-expressed IGKC not only functions as a natural antibody (28,29) but also promotes cell proliferation, cancer development, and metastasis (30), in contrast to B-cellexpressed IGKC. BKPyV replication is known to rely host cell proliferation (12,16).
We assessed the association between IGKC+ PT subpopulation and the progression of BKPyVN by analyzing BKPyV reads. However, obtaining renal biopsy specimens that meet the samples requirements of the Smart-seq2 platform is difficult. The combination of BD platform limitations (low sequencing depth) and the small genome size of BKPyV (5.0 kb) results in minimal detection sensitivity. Based on these considerations and an estab lished methodology (31), we assessed viral infection by applying GSVA based on the gene signatures in BKPyV-infected samples reported previously (16). Subsequently, we aimed to clarify the role of IGKC+ PT subpopulation in BKPyVN progression. However, the limited cohort size constrained robust comparisons, while large-scale sequencing was cost prohibitive. As an alternative, we employed a computational deconvolution strategy (21). Given the lack of detailed clinical annotations in published data sets (especially for BKPyVN progression), we could not directly explore the role of IGKC+ PT cells in the progression of BKPyVN. GSVA and GSEA results demonstrated that the IGKC+ PT subpopulation may contribute to BKPyVN progression.
Notably, the gene set in the IGKC+ PT subpopulation that changed significantly with pseudotime was significantly enriched in the immune checkpoint-associated signaling pathway, suggesting that the IGKC+ PT subpopulation, similar to cancer-associated fibroblasts (32,33), might induce T-cell exhaustion. Subsequently, we demonstrated via T-cell functional analysis and flow cytometry that T cells in the BKPyVN immune microen vironment were in a state of exhaustion. This explains why the large number of infiltrating T cells in the BKPyVN microenvironment does not effectively clear BKPyV (7). Our results suggested that T-cell exhaustion plays a vital role in BKPyVN and indicated that immunotherapeutic approaches targeting T-cell exhaustion could be applied as a BKPyVN treatment strategy. We further demonstrated via cell-cell communication analysis that the IGKC+ PT subpopulation suppressed the immune response by causing CD4+ and CD8+ T-cell exhaustion through the PDCD1-FAM3C and TIGIT-NECTIN2 ligandreceptor pairs. PDCD1-FAM3C and TIGIT-NECTIN2 are previously identified immune checkpoints involved in T-cell depletion (34,35). Current in vitro experiments cannot fully mimic the complex microenvironment of renal cells in vivo (36,37). Moreover, BKPyV is species specific (38), which makes it challenging to validate the detailed molecular mechanisms by which the IGKC+ PT subpopulation regulates T-cell depletion.
To further investigate the potential utility of the IGKC+ PT subpopulation in noninvasive diagnostics, we demonstrated the presence of the IGKC+ PT subpopulation in urine samples from patients with BKPyVN. The high cost of scRNA-seq might limit its clinical applications (39,40). Developing bioinformatics methods for scRNA-seq (41,42), such as the deconvolution algorithm used to calculate the proportions of different cell types in urine samples (21,43,44), could address this issue. Nevertheless, the deconvolu tion method requires high-quality single-cell transcriptome data (45). Our study may encourage more researchers to participate in BKPyVN urine single-cell transcriptome research.
In conclusion, we demonstrated that the IGKC+ PT subpopulation promoted BKPyVN progression via T-cell exhaustion. The IGKC+ PT subpopulation might be a promising target for BKPyVN treatment and a novel non-invasive diagnostic marker.
## MATERIALS AND METHODS
## Tissue and urine sample collection
We performed scRNA-seq on three BKPyVN biopsy tissues and two STA samples. Additionally, we profiled the single-cell transcriptome in three urine samples from patients with BKPyVN and seven urine samples from healthy volunteers to identify potential non-invasive biomarkers. The detailed clinical characteristics are summarized in Table S7.
## scRNA-seq and data processing
The kidney biopsy tissues were processed as described in prior studies (46,47), and established protocols were used to analyze the urine samples (48,49). Briefly, kidney samples were minced and then digested with an enzyme to obtain single-cell suspen sions (Supplemental Material S3). We used the BD Rhapsody Whole Transcriptome assay analysis pipeline to acquire raw gene expression matrices. Subsequent analyses were performed using Seurat (v.4.0.2). The SCT method (v.0.3.5) was used for normalization. The R package "Harmony" (v.0.1.1) was used for batching correction. The following criteria were applied: cells with >500 genes, <4,000 genes, and <30% mitochondrial gene expression in unique molecular identifier (UMI) counts (for urine samples, <60% mitochondrial gene expression in UMI counts) (33). We employed a graph-based clustering approach with a 0.2 resolution to detect cell clusters and subsequently identified major cell types using canonical marker genes from prior studies (46,47,50,51).
## Correlation with public data sets
The microarray cohort contained gene expression profiles associated with BKPyVN (n = 28) and control groups (n = 109) that were downloaded from the GEO database (GSE75693, GSE47199, and GSE72925; accessed on 6 June 2020) (52)(53)(54). The ComBat algorithm (55) was used to eliminate batch effects. To evaluate the relative abundance of each cell type identified in the present study, MuSiC (v.0.1.1) (42) was used, along with the method of pre-grouping cell types. scRNA-seq of kidney tissue samples served as a reference for estimating cell-type proportions in bulk data. Subsequently, we summarized the set of genes related to the progression of BKPyVN based on the Molecular Signatures Database (56) and literature reports (16). Then, the BKPyVN data cohort was divided into high (50%) and low (50%) groups according to the relative cell abundance of the IGKC+ PT subpopulation. Finally, we employed GSEA to explore the relationship between the abundance of IGKC+ PT cell clusters and the progression of BKPyVN. According to the 2019 guidelines of the American Society of Transplantation (1), the progression of BKPyVN primarily relies on an evaluation of inflammation, fibrosis, and BKPyV infection. Thus, we obtained gene sets related to inflammation, fibrosis, and BKPyV infection from hallmark gene sets and previous studies. (57)(58)(59). The BKPyV REPLICATION is a gene signature of BKPyV-infected PT cells, as identified in a previous study (16).
## Pseudotime trajectory analysis
We used Monocle2 (v.2.18.0) (60) to investigate the mechanisms by which PT cells promoted disease progression. Briefly, we reconstructed the cellular differentiation trajectory of PT cells and identified the genes involved in the transition from a normal state to a disease state. We employed the "differentGeneTest" function to select signature genes. Then, we reduced the dimensionality of the data using Discriminative Dimension ality Reduction via Learning a Tree and ordered the cells in pseudotime. Finally, we performed KEGG pathway enrichment analysis using KOBAS (v.3.0) (61).
## T-cell functional analysis
To evaluate the functional status of CD8 +and CD4+T cells, a set of genes associated with cytotoxicity (NKG7, PRF1, GZMA, GZMB, GZMK, IFNG, CCL4, and CST7) and exhaustion (LAG3, TIGIT, PDCD1 [also known as PD1], HAVCR2, and CTLA4) were used to calculate cytotoxicity and exhaustion scores, based on previous studies (6, 62).
## References
1. Hirsch, Randhawa, Astidco (2019) "BK polyomavirus in solid organ transplantation-Guidelines from the American society of transplantation infectious diseases community of practice" *Clin Transplant*
2. Furmaga, Kowalczyk, Zapolski et al. (2021) "BK polyomavirus-biology, genomic variation and diagnosis" *Viruses*
3. Mayr, Nickeleit, Hirsch et al. (2001) "Polyomavirus BK nephropathy in a kidney transplant recipient: critical issues of diagnosis and management" *Am J Kidney Dis*
4. Hirsch, Babel, Comoli et al. (2014) "European perspective on human polyomavirus infection, replication and disease in solid organ transplantation" *Clin Microbiol Infect*
5. Fu, Akat, Sun et al. (2019) "Single-cell RNA profiling of glomerular cells shows dynamic changes in experimental diabetic kidney disease" *J Am Soc Nephrol*
6. Jin, Li, Chen et al. (2020) "Single-cell transcriptomic analysis defines the interplay Full-Length Text Journal of Virology November"
7. "between tumor cells, viral infection, and the microenvironment in nasopharyngeal carcinoma" *Cell Res*
8. Kotla, Kadambi, Hendricks et al. (2021) "BK polyomaviruspathogen, paradigm and puzzle" *Nephrol Dial Transplant*
9. Stervbo, Nienen, Weist et al. (2019) "BKV clearance time correlates with exhaustion state and T-cell receptor repertoire shape of BKV-specific Tcells in renal transplant patients" *Front Immunol*
10. Wilhelm, Kaur, Geng et al. (2025) "Donor variability and PD-1 expression limit BK polyomavirus-specific T-cell function and therapy" *Transplantation*
11. Moriyama, Marquez, Wakatsuki et al. (2007) "Caveolar endocytosis is critical for BK virus infection of human renal proximal tubular epithelial cells" *J Virol*
12. Low, Humes, Szczypka et al. (2004) "BKV and SV40 infection of human kidney tubular epithelial cells in vitro" *Virology (Auckl)*
13. Yang, Chen, Zhang et al. (2023) "Single-cell transcriptome identifies the renal cell type tropism of human BK polyomavirus" *Int J Mol Sci*
14. Starke, Lindenmeyer, Segerer et al. (2010) "Renal tubular PD-L1 (CD274) suppresses alloreactive human T-cell responses" *Kidney Int*
15. Li, Dixon, Wu et al. (2022) "Comprehensive single-cell transcriptional profiling defines shared and unique epithelial injury responses during kidney fibrosis" *Cell Metab*
16. Li, Ferdinand, Loudon et al. (2022) "Mapping single-cell transcriptomes in the intra-tumoral and associated territories of kidney cancer" *Cancer Cell*
17. An, Cantalupo, Zheng et al. (2021) "Single-cell transcriptomics reveals a heterogeneous cellular response to BK virus infection" *J Virol*
18. Liang, Tikhanovich, Nasheuer et al. (2012) "Stimulation of BK virus DNA replication by NFI family transcription factors" *J Virol*
19. Yang, You (2020) "Regulation of polyomavirus transcription by viral and cellular factors" *Viruses*
20. Kant, Dasgupta, Bagnasco et al. (2022) "BK virus nephrop athy in kidney transplantation: a state-of-the-art review" *Viruses*
21. Imlay, Baum, Brennan et al. (2022) "Consensus definitions of BK polyomavirus nephropathy in renal transplant recipients for clinical trials" *Clin Infect Dis*
22. Wang, Park, Susztak et al. (2019) "Bulk tissue cell type deconvolution with multi-subject single-cell expression reference" *Nat Commun*
23. Chung, Goldstein, Chen et al. (2020) "Single-cell transcrip tome profiling of the kidney glomerulus identifies key cell types and reactions to injury" *J Am Soc Nephrol*
24. Weissbach, Follonier, Schmid et al. (2024) "Single-cell RNA-sequencing of BK polyomavirus replication in primary human renal proximal tubular epithelial cells identifies specific transcriptome signatures and a novel mitochondrial stress pattern" *J Virol*
25. Sheng, Liu, Qin et al. (2016) "IgG is involved in the migration and invasion of clear cell renal cell carcinoma" *J Clin Pathol*
26. Qiu, Zhu, Zhang et al. (2003) "Human epithelial cancers secrete immunoglobulin G with unidentified specificity to promote growth and survival of tumor cells" *Cancer Res*
27. Huang, Sun, Mao et al. (2008) "Expression of immunoglobulin gene with classical V-(D)-J rearrange ment in mouse brain neurons" *Int J Biochem Cell Biol*
28. Li, Korteweg, Qiu et al. (2014) "Two ultrastructural distribution patterns of immunoglobulin G in human placenta and functional implications" *Biol Reprod*
29. Jiang, Ge, Liao et al. (2015) "IgG and IgA with potential microbial-binding activity are expressed by normal human skin epidermal cells" *Int J Mol Sci*
30. Shao, Hu, Ma et al. (2016) "Epithelial cells are a source of natural IgM that contribute to innate immune responses" *Int J Biochem Cell Biol*
31. Tang, Zhang, Liu et al. (2018) "Lung squamous cell carcinoma cells express non-canonically glycosylated IgG that activates integrin-FAK signaling" *Cancer Lett*
32. Marjanovic, Hofree, Chan et al. (2020) "Emergence of a highplasticity cell state during lung cancer evolution" *Cancer Cell*
33. Li, Sun, Peng et al. (2022) "Single-cell RNA sequencing reveals a pro-invasive cancer-associated fibroblast subgroup associated with poor clinical outcomes in patients with gastric cancer" *Theranostics*
34. Chen, Zhou, Liu et al. (2020) "Singlecell RNA sequencing highlights the role of inflammatory cancerassociated fibroblasts in bladder urothelial carcinoma" *Nat Commun*
35. Breuer, Foroushani, Laird et al. (2013) "InnateDB: systems biology of innate immunity and beyond--recent updates and continuing curation" *Nucleic Acids Res*
36. Garcia-Alonso, Lorenzi, Mazzeo et al. (2022) "Single-cell roadmap of human gonadal development" *Nature*
37. An, Robles, Duray et al. (2019) "Human polyomavirus BKV infection of endothelial cells results in interferon pathway induction and persistence" *PLoS Pathog*
38. Popik, Khatua, Fabre et al. (2019) "BK virus replication in the glomerular vascular unit: implications for BK virus associated nephropathy" *Viruses*
39. Barth, Solis, Kack-Kack et al. (2016) "In vitro and in vivo models for the study of human polyomavirus infection" *Viruses*
40. Suvà, Tirosh (2019) "Single-cell RNA sequencing in cancer: lessons learned and emerging challenges" *Mol Cell*
41. Balzer, Rohacs, Susztak (2022) "How many cell types are in the kidney and what do they do?" *Annu Rev Physiol*
42. Charytonowicz, Brody, Sebra (2023) "Interpretable and context-free deconvolution of multi-scale whole transcriptomic data with UniCell deconvolve" *Nat Commun*
43. Fan, Lyu, Zhang et al. (2022) "MuSiC2: cell-type deconvolution for multi-condition bulk RNA-seq data" *Brief Bioinform*
44. (2025) *Full-Length Text Journal of Virology*
45. Dong, Thennavan, Urrutia et al. (2021) "SCDC: bulk gene expression deconvolution by multiple single-cell RNA sequencing references" *Brief Bioinform*
46. Baron, Veres, Wolock et al. (2016) "A single-cell transcriptomic map of the human and mouse pancreas reveals interand intra-cell population structure" *Cell Syst*
48. Im, Kim (2023) "A comprehensive overview of RNA deconvolution methods and their application" *Mol Cells*
49. Liu, Hu, Liu et al. (2020) "Single-cell analysis reveals immune landscape in kidneys of patients with chronic transplant rejection" *Theranostics*
50. Wu, Malone, Donnelly et al. (2018) "Single-cell transcriptomics of a human kidney allograft biopsy specimen defines a diverse inflammatory response" *J Am Soc Nephrol*
51. Abedini, Zhu, Chatterjee et al. "TRIDENT Study Investigators. 2021. Urinary single-cell profiling captures the cellular diversity of the kidney" *J Am Soc Nephrol*
52. Wang, Zhao, Zhao et al. (2021) "Single-cell RNA-seq analysis identified kidney progenitor cells from human urine" *Protein Cell*
53. Wilson, Wu, Kirita et al. (2019) "The single-cell transcriptomic landscape of early human diabetic nephropathy" *Proc Natl Acad Sci*
54. Liao, Yu, Chen et al. (2020) "Single-cell RNA sequencing of human kidney" *Sci Data*
55. Sigdel, Gao, He et al. (2016) "Mining the human urine proteome for monitoring renal transplant injury" *Kidney Int*
56. Lubetzky, Bao, Broin et al. (2014) "Genomics of BK viremia in kidney transplant recipients" *Transplantation*
57. Sigdel, Nguyen, Liberto et al. (2019) "Assessment of 19 genes and validation of CRM gene panel for quantitative transcriptional analysis of molecular rejection and inflammation in archival kidney transplant biopsies" *Front Med (Lausanne)*
58. Zhang, Parmigiani, Johnson (2020) "ComBat-seq: batch effect adjustment for RNA-seq count data" *NAR Genom Bioinform*
59. Liberzon, Birger, Thorvaldsdóttir et al. (2015) "The Molecular Signatures Database (MSigDB) hallmark gene set collection" *Cell Syst*
60. Emanuele, Enrico, Mouery et al. (2020) "Complex cartography: regulation of E2F transcription factors by cyclin F and ubiquitin" *Trends Cell Biol*
61. Needham, Perritt, Thompson (2024) "Single-cell analysis reveals host S phase drives large T antigen expression during BK polyomavirus infection" *PLoS Pathog*
62. Caller, Davies, Antrobus et al. (2019) "Temporal proteomic analysis of BK polyomavirus infection reveals virus-induced G 2 arrest and highly effective evasion of innate immune sensing" *J Virol*
63. Qiu, Hill, Packer et al. (2017) "Single-cell mRNA quantification and differential analysis with Census" *Nat Methods*
64. Bu, Luo, Huo et al. (2021) "KOBAS-i: intelligent prioritiza tion and exploratory visualization of biological functions for gene enrichment analysis" *Nucleic Acids Res*
65. Lee, Hong, Etlioglu et al. (2020) "Lineagedependent gene expression programs influence the immune landscape of colorectal cancer" *Nat Genet*
66. Garcia-Alonso, Handfield, Roberts et al. (2021) "Mapping the temporal and spatial dynamics of the human endome trium in vivo and in vitro" *Nat Genet* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12425896&blobtype=pdf | # Editorial: Unraveling diarrheic virus-host interactions: mechanisms and implications
Ahmed Ghonaim, Yanrong Zhou, Gaopeng Hou, Yinxing Zhu, Wentao Li, Curtis Brandt
## Abstract
Editorial on the Research TopicUnraveling diarrheic virus-host interactions: mechanisms and implications Understanding the complex interplay between diarrheic viruses and their hosts is central to combating the global burden of viral gastroenteritis in both humans and animals. The Research Topic "Unraveling Diarrheic Virus-Host Interactions: Mechanisms and Implications" presents seven diverse articles that span molecular virology, host immune modulation, diagnostic innovation, gut microbiota dysbiosis, and outbreak investigation. Together, these studies provide fresh insights into virus-host dynamics and point toward translational strategies for diagnosis, prevention, and treatment.Viruses and the gut ecosystem: from microbiota to metabolomics Lv et al. reviewed the diverse factors shaping piglet gut microbiota, such as host genetics, maternal influences, feeding environment, diet, and the pathogenic challenge of porcine epidemic diarrhea virus (PEDV). The authors' analysis highlighted how PEDV disrupts the intestinal barrier and microbial balance. They also discussed the potential of Chinese herbal medicine, particularly Qiwen Huangbai San, to restore mucosal immunity and promote microbial homeostasis.To investigate the role of the gut microbiota in human viral gastroenteritis, Wang et al. conducted a metagenomic study of children infected with norovirus. They found persistent dysbiosis, enrichment of Bacteroides uniformis and Veillonella, and altered carbohydrate and lipid metabolic pathways that correlated with disease severity. These findings offer candidate biomarkers for diagnosis and therapy.Frontiers in Cellular and Infection Microbiology frontiersin.org 01
## Host-targeted antivirals and mechanistic insights
The antiviral of Saxifraga stolonifera was explored by Lu et al., who demonstrated that this plant disrupts the interaction between the PEDV nucleocapsid protein and host p53. The researchers identified quercetin and other bioactive components as key effectors and, through network pharmacology and molecular docking, linked these compounds to modulation of p53-related signaling pathways, highlighting a host-targeted antiviral approach.
Wakeford et al. reported that mutation of the K48 ubiquitin linkage site in host cells markedly reduced the replication of murine norovirus. This was achieved by creating a non-permissive, proinflammatory environment, which revealed the importance of ubiquitination dynamics in viral propagation.
## Surveillance, epidemiology, and diagnostic innovation
Long-term epidemiological monitoring of norovirus in Shenzhen was carried out by Wang et al., who analyzed seven years of surveillance data. Their study uncovered genotype shifts, recombination breakpoints, and mutations associated with viral evolution, with infections concentrated among children under three years of age and peaking during the winter in more developed districts.
In an outbreak investigation, Li et al. identified airborne transmission from vomitus exposure as the main route of spread for sapovirus in a Shenzhen school. The authors showed that prompt decontamination and adherence to standard vomit cleanup protocols significantly reduced the number of cases.
Wang et al. developed a multiplexed TaqMan MGB qPCR assay for the simultaneous detection of four major feline viruses. The assay demonstrated high sensitivity, specificity, and throughput, enabling rapid diagnostics in multi-pathogen infection scenarios, which are becoming increasingly common in clinical veterinary practice.
## Perspectives and future directions
Taken together, these articles underscore the complexity and translational potential of research into diarrheic virus-host interactions. Several unifying themes emerge:
Microbiota-immune crosstalk is both a driver and a consequence of viral infections.
Host-directed antivirals, whether herbal-derived or targeting ubiquitin pathways, show promise in reducing resistance development.
Integrated surveillance and innovative diagnostics remain central to early detection and containment, especially in highdensity or zoonotic contexts.
Looking ahead, integrating systems biology, multi-omics, structural virology, and interactomics will be crucial for identifying universal host targets and novel therapeutics. Bridging laboratory discoveries with field-ready applications will accelerate progress in controlling viral gastroenteritis across species. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12572328&blobtype=pdf | # Editorial Expression of Concern to: HDAC5 is required for maintenance of pericentric heterochromatin, and controls cellcycle progression and survival of human cancer cells
P Peixoto, V Castronovo, N Matheus, C Polese, O Peulen, A Gonzalez, M Boxus, E Verdin, M Thiry, F Dequiedt, D Mottet
## Abstract
Chief are issuing an editorial expression of concern to alert readers that after the publication of this article, it was brought to the attention of the publisher that two blots in figures 2G and 5D have been reused, illustrating different conditions. The authors proposed a correction but due to the age of the article, the data could not be provided for verification. Readers are urged to take caution when interpreting the content and conclusions of this article. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802147&blobtype=pdf | # Optimizing hepatitis C virus testing in the era of point-of-care RNA diagnostics
Emily Helm, H Kim, Alexander Greninger
## Abstract
Major advances in hepatitis C virus (HCV) treatment have made timely and accurate diagnosis a critical determinant for United States elimination goals. Recently authorized point-of-care (POC) HCV RNA testing enables faster diagnosis than the traditional two-step algorithm but comes with higher laboratory costs. We analyzed HCV testing data from 2017 to 2024 across three medical centers in Seattle, Washington, to evaluate strategies for integrating new rapid direct detection tests. HCV antibody testing volumes increased 72% over the study period, with outpatient settings accounting for 76.0% of tests and a 2.7% positivity rate. Emergency department testing increased by 682% to 5,654 tests in 2024, with a 10.3% positivity rate, such that one-third of all HCV diagnoses in our medical system now originate from the public county hospi tal emergency department. Adoption of sample-to-answer HCV RNA testing in 2024 reduced median collection-to-result turnaround times for antibody-positive specimens from 84 h (IQR 58-120) to 45 h (IQR 28-57). Using a viral load cut-off of 10,000 IU/mL, HCV antigen testing was estimated to detect 98% of infections. Converting all HCV testing to POC RNA would increase laboratory costs by 260% (+$6,439 per HCV infection detected), while restricting POC RNA to the public county hospital emergency depart ment would increase costs by 22.3% (+$552 per HCV infection detected). Reflexing antibody-positive samples to antigen testing slightly reduced costs. These findings highlight the significant laboratory costs associated with POC HCV RNA testing and the need for specific reimbursement and funding mechanisms for new HCV testing algorithms.
IMPORTANCE Hepatitis C virus (HCV) elimination in the UnitedStates requires rapid and reliable diagnosis, yet current testing pathways are too slow for treatment initiation in a single clinical visit. Analysis of more than 325,000 HCV test results from 2017 to 2024 from our academic medical system in Seattle, Washington, highlights the growing role of emergency departments, particularly those serving safety-net popula tions, in making new HCV diagnoses. While point-of-care HCV RNA testing can enable connection to treatment, it substantially increases laboratory costs when implemented broadly. Targeted use of point-of-care HCV RNA in high-yield settings such as safety-net emergency departments is essential to maximize public health impact while preserving laboratory resources. These findings highlight the need for policy and reimbursement frameworks that support cost-effective deployment of new HCV diagnostic technologies. KEYWORDS reimbursement, point-of-care, HCV, antibody, antigen, RNA H epatitis C virus (HCV) remains a major public health concern, with an estimated prevalence of 1.6% among adults in the United States (1). With the opioid epidemic, rates of acute hepatitis C cases in the United States tripled from 2009 to 2018 (2) as injection drug use became the most identified risk factor for HCV infection (3). The treatment landscape for HCV has undergone a dramatic transformation over the past
decade. The introduction of direct-acting antivirals has revolutionized care, offering short, all-oral regimens that cure over 90% of patients regardless of genotype or liver disease stage (4)(5)(6). Despite the availability of curative therapy, the CDC estimates that 69,000 acute HCV infections occurred in 2023, along with 101,525 newly reported chronic HCV cases (7).
The pace of HCV elimination now hinges on improvements in case-finding and diagnostics that link patients to care. Recent modeling estimates that only 60% of people infected with HCV in the United States are aware of their status (2). Among those who received HCV testing in the United States between 2013 and 2021, only 34% had cleared infection, indicating that many patients are not being successfully connected to treatment (8). These data underscore the urgent need for an effective testing paradigm that identifies active infections and facilitates connection of patients to treatment.
The current recommended HCV screening algorithm begins with antibody testing, followed by reflex HCV RNA testing if the antibody result is reactive (9). Because antibody testing costs roughly one-third that of RNA testing based on current Centers for Medicare & Medicaid Services (CMS) fee schedules (10), this approach offers a cost-effective method for broad population screening but results in delayed detection as well as operational complexity. Positive antibody results require confirmation with RNA testing before treatment, as approximately 37% of infected individuals may clear the virus spontaneously (11), and more than 1.8 million Americans have been treated for HCV and remain antibody-positive (12). Positive antibody tests can also be confirmed using the slightly less sensitive HCV core antigen test (13,14). In 2023, the World Health Organiza tion endorsed either assay for confirming a positive antibody screen (15); however, no HCV antigen test has yet received FDA authorization (16).
While HCV screening ideally occurs in primary care settings where appropriate follow-up can be arranged, many individuals lack access to routine primary care and instead rely on emergency departments for care (17). Factors associated with not having a primary care physician-including younger age, non-private insurance status, experiencing homelessness, and having a substance use disorder-often overlap with risk factors for HCV infection (18,19). The traditional two-step algorithm of HCV testing generally does not allow confirmatory testing to be completed during a single emer gency department encounter. Recently, the FDA approved the first point-of-care (POC) HCV RNA test for HCV infection screening (20), enabling the detection of active HCV infection and potential initiation of therapy during the emergency department visit. A recent large randomized controlled trial of HCV antibody screening in emergency departments found that only 16% of patients with new HCV diagnoses initiated antiviral therapy when only antibody results were available during the emergency department encounter, underscoring the importance of having RNA results available in real time (21). However, POC HCV RNA testing is limited by cost and testing capacity (22). Additionally, current POC HCV testing is only approved by the FDA to provide a qualitative result with a limit of detection approximately 10-fold higher than available HCV RNA tests performed in the laboratory (23). Determining how to implement POC HCV RNA testing effectively is an important question for medical systems and clinical microbiology labs.
Here, we analyzed HCV testing from 2017 to 2024 across our medical system to inform the implementation of HCV RNA screening. Our academic medical system includes three hospitals with distinct patient populations as well as outpatient services and a cancer center, thus allowing us to analyze testing patterns across different settings. In addition, we performed cost modeling to more accurately estimate overall laboratory costs associated with different testing approaches.
## MATERIALS AND METHODS
## Clinical and testing setting
Located within King County, Washington, University of Washington (UW) Medicine comprises three hospitals, a cancer center, and a network of clinics. These hospitals include a 660-bed academic medical center hospital, a 500-bed public county hospi tal, and a 281-bed community hospital, each with its own emergency department. UW Medicine operates 24 primary care clinics and multiple outpatient sites, reporting 2,231,037 clinic visits in 2024 (24). Overall, King County reported 1,254 newly diagnosed cases of HCV infection in 2019 and 755 in 2024 (25). In 2019, the highest rates of HCV diagnoses per 100,000 population were among individuals aged 60-69, followed by those aged 30-39, 50-59, and 40-49. By 2024, the age groups with the highest rates remained 60-69, followed by 40-49, 30-39, and 50-59 (25).
## UW medicine HCV testing
Both HCV antibody and viral load tests are offered through the UW Virology laboratory at a location that is at least one mile off-site from UW Medicine hospitals and clinics. No onsite POC testing was offered for either antibody or viral load tests during this period. A courier system is in place to bring patient samples from other sites to the laboratory. HCV testing can be ordered as standalone antibody or viral load testing, or as reflex viral load testing following a reactive antibody test based on the ordering provider's preference. If an order is placed for HCV antibody with a reflex to viral load testing following a reactive antibody test, the protocol is for two samples to be drawn at the initial blood draw with the second sample used for reflex testing. If this order is placed by the clinician, reflex PCR testing will be performed if the antibody screen is reactive. Antibody tests are performed using the Abbott Architect i2000. Viral load testing was performed on the Abbott m2000 from 2017 to 2024, switching to the Hologic Panther on 1 November, 2024.
## HCV testing data
HCV testing data from 1 January 2017 to 31 December 2024 were obtained through the UW Medicine Department of Laboratory Medicine and Pathology Data Warehouse. Hepatitis C antibody and RNA testing data for both reflexed and unreflexed testing were identified using test order codes. The data set was filtered to exclude proficiency testing, research testing, and non-patient samples. Orders without results were also excluded to avoid analyzing canceled or reordered samples. For analysis of ordering locations, location codes were grouped into seven categories: employee health, cancer center, outside (reference laboratory services for external non-UW clients), emergency department (ED), inpatient, outpatient, and other. The "other" category primarily includes infusion clinics and outpatient procedure locations that did not fit into other groups. All analyses were performed using R version 4.5.1.
## Cost projections
We used two approaches to estimate laboratory costs associated with HCV testing. The first approach utilized reimbursements from the 2025 CMS Clinical Laboratory Fee Schedule: $14.27 per HCV antibody test and $42.84 per HCV RNA test (10). The second approach relied on internal cost accounting data from the UW Department of Laboratory Medicine and Pathology of $23.09 per HCV antibody test and $49.16 per HCV RNA viral load test. For HCV POC RNA testing, the cost was modeled based on the Cepheid Xpert HCV test as this is the only HCV POC test FDA-approved in the United States. The cost was estimated at $90.69 per test based on reagent costs ($50) combined with internal staffing costs for nucleic acid amplification tests (NAAT) performed one specimen at a time. HCV antigen test costs were modeled based on either a one-step method such as the Roche combination antibody-antigen test or a two-step method such as the Abbott antigen test. Potential HCV antigen test costs were approximated using reagent costs for qualitative hepatitis B surface antigen testing ($4.64 per test). For combined antigen/antibody testing, this value was added to the antibody test cost of $27.73 per test, assuming no additional staffing time. For antibody testing reflexed to antigen testing, an additional $5.03 per test was added to antibody test cost to cover reagents and the minimal extra staff time for processing the reflex test. In all scenarios, we assumed that testing would be performed in an existing laboratory with existing instrumentation and used internal cost calculations for general overhead costs.
Projected testing volumes for different algorithms were based on HCV testing volumes and positivity rates observed at UW Medicine in 2024. To accurately capture orders used for HCV screening, only tests ordered as "HCV antibody with reflex to PCR" were included. Cost differences between algorithms were calculated by comparing the total costs of each scenario to the cost of maintaining the traditional two-step algorithm. To simplify modeling of the sensitivity of the HCV antigen test, we assumed 100% sensitivity at viral loads >10,000 IU/mL and no sensitivity at viral loads <10,000 IU/mL (13,14) with the difference in test positivity modeled in Fig. S1.
## RESULTS
## Hepatitis C antibody utilization
As the CDC currently recommends a two-step algorithm for HCV screening, we first analyzed ordering trends for HCV antibody testing across our academic medical system. HCV antibody testing volumes increased from 26,188 tests in 2017 to 45,010 tests in 2024, including a 1.9-fold increase between 2019 and 2021 corresponding both to a UW initiative to increase ED-based HCV testing as well as updated 2020 CDC guidance recommending one-time HCV screening for all adults (26) (Fig. 1A). With the increased testing, the positivity rate declined from 7.2% in 2019 to 4.2% in 2024 (Fig. 1B). The HCV antibody testing population became noticeably younger between 2017 and 2024, the proportion of individuals age 20-39 years almost doubling from 26.9% to 51.8% (Table 1). Between 2017 and 2024, outpatient settings accounted for 76.0% of antibody tests, followed by inpatient (8.5%) and the emergency department (6.9%) settings (Fig. 1C). Notably, emergency department testing rose 7.8-fold from 723 tests in 2017 to 5,654 tests in 2024 (Fig. 1C). Positivity rates were lowest in outpatient testing (2.7%) and highest in inpatient (17.0%) and emergency department (13.7%) testing (Fig. 1D).
Given the increase in HCV antibody testing from the emergency department, we further characterized testing at this location. Between 2017 and 2024, the public county hospital emergency department accounted for 73.3% of all ED HCV antibody tests, followed by the academic medical center (18.7%) and community hospital (8.0%) (Fig. 2A). Since the public county hospital contributed nearly three-quarters of ED tests, we focused subsequent analyses on this site. From 2021 to 2023, the public county hospital emergency department rose 3.7-fold, from 1,004 tests in 2021 to 3,751 tests in 2023, reaching 4,013 tests in 2024 (Fig. 2B). During this time period, the positivity rate declined from 17.0% in 2021 to 12.5% in 2023 (Fig. 2C).
## HCV viral load utilization
HCV antibody testing is only the first step in the CDC algorithm, as all reactive antibody tests should be confirmed with viral load testing to quantify HCV RNA levels (26). Unlike antibody testing, HCV RNA tests decreased by 35.3% from 2017 to 2024, with steady declines during the 2020 COVID-19 pandemic year followed by stabilization from 2021 to 2024 (Fig. 3A). In 2017, outpatient settings accounted for 64.4% of viral load tests, decreasing to 37.3% in 2024 (Fig. 3A). Alongside the overall decrease in viral load testing, the proportion of tests with detectable HCV RNA fell from 38.5% in 2017 to 15.1% in 2024 (Fig. 3B). Conversely, inpatient and emergency department viral load testing increased from 19.4% of all HCV RNA tests in 2020 to 38.6% by 2024 (Fig. 3A). Viral load testing in the public county hospital emergency department rose 6.6-fold between 2017 and 2024 (Fig. 3C). Despite the increased volume, the positivity rate for detectable viral loads from the public county hospital emergency department remained high at 30.7% in 2024 (Fig. 3C).
## Turnaround time analysis
We next examined turnaround times for reflex HCV antibody and viral load tests. From January 2017 to October 2024, HCV viral load testing was performed twice a week using the multi-step Abbott m2000 platform, resulting in a median collection-to-RNA-result turnaround time of 84 h (IQR 58-120) for reactive screening tests (Fig. 4). In November 2024, viral load testing switched to the Hologic Panther platform, enabling three runs per week. However, true random-access testing remained limited due to quality control costs associated with quantitative viral load testing. This change reduced the median turn around time to 45 h (IQR 28-57) (Fig. 4).
## Cost projections
Based on this testing data, we conducted cost analyses for different algorithms for HCV testing (Fig. 5). We examined five scenarios for implementing POC HCV RNA testing instead of initial antibody testing, using 2024 data: (i) all HCV testing, (ii) inpatient and emergency department testing only, (iii) emergency department testing only, (iv) public county hospital emergency department testing only, and (v) maintaining all testing remaining as the traditional two-step algorithm. To simplify projections, we assumed all algorithms would eventually identify all 446 HCV infections detected in 2024. To estimate costs, we first used the 2025 Medicare reimbursement rates for HCV antibody testing and HCV viral load quantification (Table 2). If no changes were made and all testing remained as a two-step algorithm, the total laboratory cost was $704,848, with a laboratory cost per HCV infection detected of $1,580. Switching to single-step POC HCV RNA testing for all screening would increase costs by an additional $1,172,744. Restricting single-step POC HCV RNA testing to inpatient and emergency department settings would increase costs by $211,703; to emergency departments at all three hospitals by $135,450; and to the public county hospital emergency department by $94,051. Corresponding costs per HCV infection detected would be $4,210 for universal single-step RNA testing; $2,055 for inpatient and emergency department testing; $1,884 for all emergency department testing; and $1,791 for the public county hospital emergency department.
While these calculations provide a framework for reimbursement costs, they do not fully capture laboratory expenses, as currently authorized POC HCV RNA testing is more costly than traditional RNA methods due to higher reagent prices and increased staffing needs. Additionally, reimbursement rates do not always reflect the actual costs of performing tests. Therefore, we conducted a second round of laboratory cost projections using internal UW cost accounting data for HCV antibody and traditional RNA testing, combined with estimated reagent and staffing costs for POC RNA testing (Table 2). Using these internal estimates, the total laboratory cost for HCV testing in 2024 was $1,103,127. We projected added costs of $2,871,634 if all screening shifted to POC HCV RNA testing, While HCV RNA testing is currently the only FDA-authorized method to confirm active HCV infection in the United States, HCV core antigen testing is used in other countries for this purpose (15,(27)(28)(29). We, therefore, projected the costs of implementing either combination HCV antigen/antibody or reflex HCV antibody-to-antigen testing, which could offer lower-cost alternatives if FDA-authorized tests become available in the United States (Fig. 5). Because sensitivity is a major concern with antigen testing, we first assessed how many infections would be detectable based on literature indicat ing antigen tests reliably detect viral loads above 10,000 IU/mL (13,14). Restricting analysis to viral loads ordered reflexively after positive antibody screening in 2024, 98.1% of detectable HCV RNA tests had viral loads above this threshold and would likely be identified by antigen testing (Fig. S1). For combination antigen/antibody testing, we assumed the test would be performed upfront, with any positive antibody and negative antigen results reflexed to RNA testing to ensure infection detection (28). Using Hepatitis B surface antigen reagent costs as a proxy for HCV antigen costs, we estimated that initiating all screening with combination HCV antigen/antibody tests would add $183,193 in laboratory costs, compared to $25,573 if restricted to inpatient and emergency department testing, $17,169 if restricted to all emergency department testing, and $11,494 if restricted to public county hospital emergency department testing (Table 3). For reflex antibody-to-antigen testing, we assumed that any specimen with an antibody-positive, antigen-negative result would be reflexed to an HCV viral load test to not miss infections. Based on this scenario, we estimated a cost savings of $10,844 if all positive screening antibody tests were reflexed to antigen testing vs a savings of $10,811 for inpatient and emergency department testing, $5,933 for all emergency department testing, and $4,652 for the public county hospital emergency department testing only (Table 4).
## DISCUSSION
The recent FDA authorization of the first point-of-care HCV RNA test has generated significant enthusiasm among providers and policymakers as a potential catalyst for HCV elimination in the United States. However, questions remain about whether and how this test should be implemented across medical systems. To address this from a clinical laboratory perspective, we conducted a comprehensive analysis of recent HCV testing within our medical system. Overall, we observed a substantial increase in outpatient HCV antibody testing between 2019 and 2021, which was temporally associated with a local initiative to broadly screen for HCV as well as CDC's 2020 recommendation of screening all adults for HCV in 2020, along with a corresponding decline in positivity rates. Despite this increase in antibody testing, viral load testing volumes have declined markedly since 2017, primarily in the outpatient setting, consistent with consolidation of earlier guideline changes that recommend viral load testing only at baseline and post-treatment cure assessment (30). With the availability of highly effective curative therapies and shifting risk factors for HCV, testing in the emergency department for HCV has gained increased attention nationwide (21,31). In our medical system, emergency department testing rose in 2022 for both HCV antibody and viral load tests corresponding to implementation of a program in late 2021 to promote screening in the emergency department. The public county hospital emergency department accounted for the majority of these tests, reflecting a patient population with higher risk behaviors and less presentation to outpatient primary care. The antibody positivity rate in the emergency department was five times higher than the outpatient rate, indicating a significantly higher-risk popula tion. Notably, in 2024, the positivity rate for reflex HCV RNA testing in the public county hospital emergency department was only 30.7%, underscoring the high proportion of previously treated individuals in this population, given spontaneous clearance rates are estimated to be between 15% and 40% for HCV infection (12).
While testing is the first step toward curing HCV infection, patients must be linked to treatment to achieve improved clinical outcomes. A major limitation of the current testing algorithm is that HCV antibody tests reflexed to viral load confirmation do not yield results within the timeframe of an emergency department visit, as reflected by our median turnaround time of 84 h from 2017 to 2024. No doubt this approach was originally designed for an outpatient-first testing model, rather than the current scenario where over one-third of HCV infections are diagnosed from the public county emergency department, as seen in our system. A recent study of HCV antibody screening in the emergency department found that fewer than 10% of newly diagnosed HCV infections were cured when only the antibody result was available during the emergency department visit, highlighting the potential importance of having the HCV RNA result during the visit (21). To address turnaround time, we transitioned HCV viral load testing to the random access Hologic Panther platform, reducing median turnaround time to 45 h, consistent with 1.6-4.0 day median turnaround time seen in laboratory-reflex based HCV testing (32)(33)(34). However, despite the instrument's random-access capability, samples were still batched and run only three times weekly due to quality control costs associated with quantitative viral load testing. While this represents a significant improvement, the turnaround remains too long for a typical emergency department visit, indicating the need for alternative testing strategies.
In June 2024, the FDA authorized the first POC HCV RNA test, enabling results within the timeframe of a clinical visit. However, the costs associated with this testing approach remain high, particularly given strained hospital and Medicaid budgets. For example, we found that converting all HCV testing to this platform would increase laboratory costs Full-Length Text by 260%, or approximately $2.87 million, across our three-hospital system. Restricting POC HCV RNA implementation to the public county hospital emergency department could substantially reduce turnaround time-other onsite Cepheid tests have median real-world turnaround times under 3 h-for the 149 individuals diagnosed with HCV infection there. However, implementing this testing in just one hospital unit would increase annual laboratory costs by roughly a quarter of a million dollars, with likely increases as emergency department caseloads grow. These expenses may come at the cost of testing other patients amid constrained budgets.
While no HCV antigen tests are currently FDA-approved, antigen testing is used in other countries to confirm active HCV infection (28,29,35). We, therefore, modeled the same scenarios applied to POC HCV RNA testing but using either upfront combination HCV antigen/antibody testing or reflex antibody-to-antigen testing. These approaches resulted in much lower laboratory costs, and if on-site chemistry line testing is available, could potentially provide turnaround times equivalent to or better than POC HCV RNA testing. For example, implementing combination antigen/antibody testing in the public county hospital emergency department would add only $11,494 in costs-less than one-twentieth the additional cost of POC RNA testing-while reflex antibody-to-antigen testing was overall cost-saving from the laboratory perspective.
Our cost projections suggest that HCV antigen testing could be a cost-effective strategy for confirming active infection during an emergency department visit. While earlier smaller studies raised concerns about the analytical sensitivity of antigen tests (36,37), our data indicate that antigen testing would miss only 2% of infections based on reflexed HCV viral loads. This aligns with previous reports showing 3% (38), 5% (35), and 6% (39) of antigen tests resulting as negative despite detectable HCV viral loads. To avoid missing infections, we modeled an algorithm where all positive antibody, but negative antigen tests are reflexed to HCV RNA testing, allowing eventual identification of these few cases, albeit with a longer turnaround time. While this reflex testing ensures no infections are overlooked due to lower antigen sensitivity, eliminating this reflex step could further reduce laboratory costs. While our cost projections aimed to reflect realistic laboratory expenses, it has several limitations. We focused exclusively on laboratory costs and did not include costs related to clinical care associated with this testing. Additionally, we modeled testing not as a true POC test but rather performed in a hospital clinical laboratory due to quality assurance and compliance reasons, which can reduce overall staffing and overhead costs. We also did not factor in additional costs or savings on either healthcare or the broader societal level following cured HCV infections that do not progress to further complications. Our projections are based on 2024 testing data and do not include dynamic forecasting of future testing volumes. Because infectious disease dynamics are nonlinear, successful or unsuccessful implementation of HCV test-and-treat strategies could result in different future infection trends, impacting future testing volumes and institutional costs (40). Since no FDA-approved HCV antigen tests exist, we estimated costs using Hepatitis B surface antigen tests as a proxy, which may not be accurate as Hepatitis B surface antigen tests have been used for decades such that HCV antigen reagents may have higher costs, resulting in anslight overestimation of cost savings for this testing strategy. A key limitation of our analysis is that our model considers positive test results rather than clinical outcomes. Linking patients to HCV treatment and cure remains a major challenge in emergency departments (21,41) and must be addressed for the clinical benefits of testing to be realized. While the country of Georgia was successful in increasing HCV infection diagnoses through the use of the more cost-effective antigen test, they reported difficulties in translating this to treatment initiation, highlighting the importance of connecting laboratory testing to treatment simultaneously (42). Despite these limitations, our findings align with a recent CDC study showing cost savings for an antibody-to-antigen reflex algorithm, while combined antigen/antibody testing or POC RNA testing was associated with higher costs (43). These findings are also supported by studies from other countries showing cost savings when using antigen testing compared to RNA tests (44,45), with the greatest cost savings resulting from eliminating RNA testing for antibody-positive, antigen-negative specimens (46). Another limitation of our study is that it focuses on a single medical system within one city. However, our findings are consistent with other published results, supporting their broader generalizability. For example, a study from Baltimore reported a 13.8% HCV antibody positivity rate in their emergency department similar to our observed rate (31). Additionally, our data show that 3.4% of emergency department tests resulted in confirmed HCV infections, which agree with rates reported in multi-city studies (1.9%) (21) and a study from Baltimore (1.7%) (36).
In summary, the recent availability of POC HCV RNA testing raises important questions about capacity, cost-effectiveness, and evolving HCV testing trends. Our data show that while emergency department testing is increasing within our system, it still represents only a small fraction of total HCV screening and can substantially increase laboratory costs. Thus, any testing strategy must balance the need for rapid results in the emergency department with the need for low-cost screening in outpatient populations. We demonstrate that an antibody-to-antigen reflex algorithm could result in laboratory cost savings while providing rapid results, assuming on-site testing is available. If an automated chemistry line with this capability is not available on-site, targeted POC RNA testing in high-positivity, rapid-turnaround settings offers an alternative approach. Furthermore, recent data from a large randomized control trial of emergency depart ment HCV antibody screening-in which fewer than 10% of individuals achieved HCV cure over 18 months of follow-up-highlights the significant challenges in translating testing into successful health outcomes. Given the substantially higher costs of direct HCV RNA testing, enthusiasm for POC HCV RNA testing must be matched with rigorous studies evaluating its impact on HCV cure rates.
## References
1. Medicare $704
2. "Cost accounting $1,103,127 $1,349,270 $1,453,525 $1,658,250 $3,974,761 Increased laboratory costs compared to two-step algorithm"
3. "Medicare $0 $94,051 $135,450 $211,703 $1"
4. "The cost difference between using a two-step algorithm for HCV screening compared to single-step HCV RNA testing was compared. The first value listed is based on current Medicare reimbursement rates for HCV antibody and viral load testing and the second value listed is based on estimated lab costs for HCV antibody testing, traditional HCV RNA testing, and point-of-care HCV RNA testing. All values are in US dollars. All values are based on 2024 data for reflex test orders, and costs are estimated for one year"
5. Hall, Bradley, Barker et al. (2025) "Estimating hepatitis C prevalence in the United States, 2017-2020" *Hepatology*
6. Gnanapandithan, Ghali (2023) "Self-awareness of hepatitis C infection in the United States: A cross-sectional study based on the National Health Nutrition and Examination Survey" *PLoS One*
7. (2019) "Hepatitis surveillance in the United States, 2017 | CDC"
8. Bourlière, Gordon, Flamm et al. (2017) "Sofosbuvir, velpatasvir, and voxilaprevir for previously treated HCV infection" *N Engl J Med*
9. Forns, Lee, Valdes et al. (2017) "Glecaprevir plus pibrentasvir for chronic hepatitis C virus genotype 1, 2, 4, 5, or 6 infection in adults with compensated cirrhosis (EXPEDITION-1): a singlearm, open-label, multicentre phase 3 trial" *Lancet Infect Dis*
10. Puoti, Foster, Wang et al. (2018) "High SVR12 with 8-week and 12-week glecaprevir/pibrentasvir therapy: an integrated analysis of HCV genotype 1-6 patients without cirrhosis" *J Hepatol*
11. (2025) "Hepatitis C surveillance | 2023"
12. Wester, Osinubi, Kaufman et al. (2023) "Hepatitis C virus clearance cascade -United States, 2013-2022" *MMWR Morb Mortal Wkly Rep*
13. (2026) *Full-Length Text Journal of Clinical Microbiology*
14. Cartwright, Patel, Kamili et al. (2023) "Updated operational guidance for implementing CDC's recommendations on testing for hepatitis C virus infection" *MMWR Morb Mortal Wkly Rep*
15. "Centers for Medicare & Medicaid Services. 2025 CLFS Files"
16. Aisyah, Shallcross, Hully et al. (2018) "Assessing hepatitis C spontaneous clearance and understanding associated factors-A systematic review and meta-analysis" *J Viral Hepat*
17. Westbrook, Dusheiko (2014) "Natural history of hepatitis C" *J Hepatol*
18. Bui, Brown, Brown et al. (2024) "Comparison of a dual antibody and antigen HCV immunoassay to standard of care algorithmic testing" *J Clin Microbiol*
19. Mixson-Hayden, Dawson, Teshale et al. (2015) "Performance of ARCHITECT HCV core antigen test with specimens from US plasma donors and injecting drug users" *J Clin Virol*
20. Who (2023) "New recommendation on hepatitis C virus testing and treatment for people at ongoing risk of infection Policy brief"
21. Liaw, Petterson, Rabin et al. (2014) "The impact of insurance and a usual source of care on emergency department use in the United States" *Int J Family Med*
22. Cummings, Martinez, Mourad (2022) "Primary care gap: factors associated with persistent lack of primary care after hospitalisation" *BMJ Open Qual*
23. Levine, Linder, Landon (2002) "Characteristics of Americans with primary care and changes over time" *JAMA Intern Med*
24. Fda (2024) "FDA Permits Marketing of First Point-of-Care Hepatitis C RNA Test"
25. Haukoos, Rothman, Galbraith et al. "DETECT Hep C Screening Trial Investigators. 2025. Hepatitis C screening in emergency departments: The DETECT Hep C randomized clinical trial" *JAMA*
26. Reipold, Shilton, Donolato et al. (2024) "Molecular point-of-care testing for hepatitis c: available technologies, pipeline, and promising future directions" *J Infect Dis*
27. Cepheid (2025) "Cepheid-Xpert-HCV-GPM-reference-sheet-US-IVD-10119-English"
28. (2025) "The UW medicine family"
29. (2025) "Hepatitis C Data dashboards"
30. Schillie, Wester, Osborne et al. (2020) "CDC recommendations for hepatitis C screening among adults -United States" *MMWR Recomm Rep*
31. (2020) "Electronic address: easloffice@easloffice.eu, Clinical Practice Guidelines Panel: Chair:, EASL Governing Board representative:, & Panel members" *J Hepatol*
32. Mandel, Underwood, Masterman et al. (2023) "Province-toprovince variability in hepatitis C testing, care, and treatment across Canada" *Can Liver J*
33. Handanagic, Shadaker, Drobeniuc et al. (2024) "Lessons learned from global hepatitis C elimination programs"
34. Bhattacharya, Aronsohn, Price et al. (2023) "Update: AASLD-IDSA recommendations for testing, managing, and treating hepatitis C virus infection" *Clin. Infect. Dis. Off. Publ. Infect. Dis. Soc. Am:ciad*
35. Hsieh, Rothman, Laeyendecker et al. (2016) "Evaluation of the centers for disease control and prevention recommendations for hepatitis C virus testing in an urban emergency department" *Clin Infect Dis*
36. Tao, Tang, Fajardo et al. "2022 Reflex hepatitis C virus viral load reflex testing following an initial positive HCV antibody test: a global systematic review and meta-analysis" *SSRN Journal*
37. López-Martínez, Arias-García, Rodríguez-Algarra et al. (2019) "Significant improvement in diagnosis of hepatitis C virus infection by a one-step strategy in a central laboratory: an optimal tool for hepatitis C elimination?" *J Clin Microbiol*
38. Thompson, Fenton, Charlton (2022) "HCV reflex testing: a singlesample, low-contamination method that improves the diagnostic efficiency of HCV testing among patients in Alberta" *Canada. J Assoc Med Microbiol Infect Dis Can*
39. Pérez-García, Aguinaga, Navascués et al. (2019) "Hepatitis C core antigen: diagnosis and monitoring of patients infected with hepatitis C virus" *Int J Infect Dis*
40. Prostko, Rothman, Hsieh et al. (2024) "Performance evaluation of the Abbott Alinity Hepatitis C antigen next assay in a US urban emergency department population" *J Clin Virol*
41. Wong, Gan, Mohamed et al. (2020) "Hepatitis C core antigen testing to diagnose active hepatitis C infection among haemodialysis patients" *BMC Nephrol*
42. Gunsolus, Prostko, Pearce et al. (2024) "Comparison of a hepatitis C core antigen assay to nucleic acid amplification testing for detection of hepatitis C viremia in a US population" *Microbiol Spectr*
43. Van Tilborg, Marzooqi, Wong et al. (2018) "HCV core antigen as an alternative to HCV RNA testing in the era of direct-acting antivirals: retrospective screening and diagnostic cohort studies" *The Lancet Gastroenterology & Hepatology*
44. Fraser, Martin, Brummer-Korvenkontio et al. (2018) "Model projections on the impact of HCV treatment in the prevention of HCV transmission among people who inject drugs in Europe" *J Hepatol*
45. Hernandez-Con, Wilson, Tang et al. (2023) "Hepatitis C cascade of care in the direct-acting antivirals era: a metaanalysis" *Am J Prev Med*
46. Averhoff, Shadaker, Gamkrelidze et al. (2026) *Full-Length Text Journal of Clinical Microbiology*
47. (2020) "Progress and challenges of a pioneering hepatitis C elimination program in the country of Georgia" *J Hepatol*
48. Hall, Sandul, Kamili et al. (2025) "Cost-effectiveness analysis of testing approaches for diagnosis of hepatitis C among US adults" *Clin Infect Dis ciaf*
49. Wasitthankasem, Vichaiwattana, Auphimai et al. (2017) "HCV core antigen is an alternative marker to HCV RNA for evaluating active HCV infection: implications for improved diagnostic option in an era of affordable DAAs" *PeerJ*
50. Juárez-Figueroa, Iracheta-Hernández, Medina-Islas et al. (2014) "Comparison of two diagnostic algorithms for the identification of patients with HCV viremia using a new HCV antigen test" *Ann Hepatol*
51. Jülicher, Galli (2018) "Identifying cost-effective screening algorithms for active hepatitis C virus infections in a high prevalence setting" *J Med Econ* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802276&blobtype=pdf | # Antibody responses following COVID-19 vaccination and breakthrough infections in naïve and convalescent individuals suggest imprinting to the ancestral strain of SARS-CoV-2
Siddhartha Mahanty, Emily Eriksson, Peta Edler, Francesca Mordant, Nicholas Kiernan-Walker, David Price, Sabine Braat, Eamon Conway, Vanessa Bryant, Honghua Ding, Leo Yi, Yang Lee, Louise Randall, Ramin Mazhari, Ivo Mueller, Kanta Subbarao
## Abstract
The binding and neutralizing activity of SARS-CoV-2 antibodies are important correlates of protection of current COVID-19 vaccines. SARS-CoV-2 exposure status and COVID-19 vaccine types can influence these responses and the breadth of cross-reactivity to variants. In this longitudinal cohort study, we used SARS-CoV-2specific multiplex Luminex antibody assays and live virus neutralization of ancestral (VIC01/2020), Delta and Omicron (BA1, BA2, and BA5) SARS-CoV-2 variants to com pare antigen-specific binding, and neutralizing antibody (nAb) responses to primary vaccination (two doses) of adenovirus vectored (AdVV) or mRNA vaccines followed by a booster dose of mRNA vaccine in convalescent (n = 51) and infection-naïve individu als (n = 47). In a subset of individuals, we performed additional analysis of antibody responses following breakthrough infection. We found that titers of anti-SARS-CoV-2 nAb following primary vaccination (two doses) with AdVV vaccine were significantly lower than those following mRNA vaccine, irrespective of prior SARS-CoV-2 infection status. However, an mRNA vaccine booster dose resulted in equivalent binding and nAb titers to the ancestral virus in all individuals, irrespective of primary vaccine type. Notably, vaccinated infection-naïve, but not convalescent individuals required the third dose of vaccine (mRNA) to induce nAbs to Omicron subvariants BA1, BA2, and BA5 though titers against the variants were lower than those against the ancestral strain. Importantly, breakthrough infection with Omicron strains induced higher nAb titer rises against the ancestral strain than against Omicron variants consistent with imprinting of the immunologic response and recall of pre-existing immunity to the ancestral strain. IMPORTANCE Studies on binding and neutralizing antibody responses to COVID-19 vaccines and breakthrough infections were frequently confounded by unsuspected exposure to intercurrent natural infections with SARS-CoV-2 in the community particu larly during year 1 and 2 of the pandemic. This study is extraordinary in that it was conducted in Australia, where SARS-CoV-2 circulation was largely contained by public health and social measures for the first 2 years of the pandemic. We followed welldefined study populations who received vaccines in a natural infection-free setting or a separate subgroup who had natural infection. Thus, the study provides unique insights in infection-naïve, vaccinated individuals and those with breakthrough infections with Omicron variants. In this setting, we had an opportunity to demonstrate evidence of antigenic imprinting, with neutralizing antibody responses to the ancestral vaccine antigen being higher than responses to the infecting Omicron variant.
T he COVID-19 pandemic has presented varying challenges across the globe, with Australia experiencing a unique trajectory. With public health measures, nota bly, stringent border closures, mandatory quarantine for arriving travelers and social distancing, Australia maintained a low case count during the initial phases of the pandemic (1)(2)(3)(4). This period, preceding the emergence of the Omicron variants in late 2021, allowed for widespread vaccination efforts, leveraging mRNA and adenovi rus-vectored vaccines, achieving a remarkable vaccination rate of >90% among adults. Consequently, with the re-opening of the borders and emergence of Omicron strains, Australia witnessed a low incidence of severe COVID-19 cases within a largely immu nized population (5). This distinctive scenario provided an exceptional opportunity to investigate immune responses to SARS-CoV-2 vaccination without the confounding influence of background immunity from prior infections. In addition, it facilitated a comparative analysis of antibody responses against emerging variants among vaccina ted individuals with and without previous exposure to the virus.
Early reports have highlighted the emergence of anti-SARS-CoV-2 antibodies shortly after infection, with subsequent dynamics characterized by a rapid initial decline followed by a more gradual decay in titers (6)(7)(8)(9). Studies investigating vaccine-induced antibody kinetics, particularly in response to mRNA, protein, and vector-based COVID-19 vaccines, have demonstrated robust responses, albeit with considerable variation in peak levels and decay rates across vaccine types. Notably, individuals with prior COVID-19 infection exhibit significantly elevated antibody levels post-vaccination, with a slower decline over time (10).
Neutralizing antibodies (nAb) generated by vaccination with the spike protein from the ancestral virus prevent virus entry into host cells efficiently, but levels of cross-reac tivity with SARS-CoV-2 variants vary, irrespective of whether the nAbs are generated by vaccination or by natural infection (11)(12)(13). However, the majority of serological studies have been conducted in settings with high community transmission of the virus (14), making it difficult to evaluate neutralizing activity generated by COVID vaccines over extended infection-free periods that were characteristic of low-transmission environ ments like Australia (3,15,16).
In this cohort study, we explored SARS-CoV-2 antibody responses among COVID-19naïve and convalescent individuals in Australia following vaccination with BNT162b2 (mRNA) or ChAdOX1 (AdVV) vaccines. Our investigation includes evaluation of antibody levels and nAb against ancestral, Delta and Omicron (BA1, BA2, and BA5) variants, following primary immunization, booster doses given months after the primary vaccines, and around breakthrough infections with variant viruses. Our findings support recent reports of the dominance of neutralizing activity against the ancestral strain post-vacci nation (17)(18)(19)(20) and yield insights into the development of heterologous (AdVV primary vaccine and mRNA booster) vs homologous (all vaccines were mRNA type) immunity against SARS-CoV-2 variants following vaccination and breakthrough infections.
Our results illuminate the complex interplay between vaccination, prior infection, and emerging variants, shedding light on the dynamics of antibody responses crucial for informing ongoing pandemic response strategies.
## MATERIALS AND METHODS
## Study subjects
This study is an immunological sub-study from two study cohorts-a DISCOVER-HCP-Vaccine cohort (Peter Doherty Institute for Infection and Immunity, Melbourne, Australia) (N = 95) and a COVID PROFILE cohort (21) (The Walter and Eliza Hall Institute, Melbourne, Australia) (N = 171). From the two cohorts, a total of 126 sera and plasma samples were collected from 57 participants in the DISCOVER-HCP-Vaccine cohort and from 69 participants in the COVID PROFILE cohort. Study cohort demographics are detailed in supplementary data (Table S1a). At study entry, participants were categorized into infection-naïve controls (Naive; n = 72) or COVID-recovered (Convalescent; n = 54). COVID-19 infection status was determined based on available results of PCR tests for SARS-CoV-2 RNA in nasal/oral swabs and further verified by SARS-CoV-2-specific antibody levels at baseline (pre-vaccination). Ninety eight participants (51 Naïve and 47 Convalescent individuals) who had received 2 doses of COVID-19 vaccines at the start of the current immunological sub-study and also had pre-vaccination samples collected and another 4 (Naïve) participants prior to the booster vaccine were included in the analysis of responses to the primary vaccines and the booster dose. The analysis of pre-and post-infection samples included data from 61 participants (36 Naïve and 25 Convalescent participants). Demographics for this subpopulation are summarized in Table S1b.
## Vaccination, sample collection, and breakthrough infection
All participants received either two doses of BTN162b2 (mRNA) or ChAdOX1 (AdVV) under an Australian government-supported vaccination program. When a third (booster) dose was recommended, a subset of participants (n = 115) received one dose of mRNA vaccine. Among the mRNA recipients, a small subgroup (n = 8; four in the naïve and four in the convalescent group) received a Moderna mRNA vaccine (mRNA 1273) when it was an available alternative for the third dose. Ninety-eight participants entered the study prior to receiving the first vaccine dose, 4 entered before the booster vaccine, and 24 entered after the primary vaccine series and had a breakthrough infection. Depending on the timepoint relative to vaccination when participants were enrolled in the study, serum (in serum separation tubes) or plasma samples (in tubes with EDTA) were collected prior to vaccination (Pre-Vax) as a baseline sample, and 2-4 weeks after the second dose of vaccine (Post-2nd Vax) and/or 2-4 weeks after the third vaccine dose (Post-3rd Vax). The last sample collected prior to the third vaccine dose served as pre-3rd Vax. A number of participants (N = 61) had a breakthrough infection during follow-up. The last sample collected before the breakthrough diagnosis served as pre-infection sample and post-infection samples were collected 2-4 weeks post-diagnosis of the infection (Fig. 1A). Sample flow for SARS-CoV-2-specific binding antibody and nAb titer analyses is summarized in Fig. S1 and S2, respectively.
## SARS-CoV-2-specific antibody multiplex assay
Plasma antibody levels specific for SARS-CoV-2 antigens S1, S2, receptor-binding domain (RBD), Spike, and nucleoprotein (NP), based on sequences from the ancestral strain, Delta RBD, Omicron BA1 and BA2 RBD were measured using a multiplex serological assay employing the Luminex platform as previously described (22). Antibody levels to seasonal coronavirus antigens (NL63 NP, OC43 Spike, 229E S1, and HKU1 Spike), Influenza A antigen (H1N1 hemagglutinin), and tetanus toxoid were also measured in the assay. For each individual, total IgG, IgM, and IgA levels were measured for each ancestral strain-derived antigen and total IgG levels were measured for variant antigens. For standardization between plates data were normalized using an algorithm which adjusted for plate-to-plate variation based on standard curves.
## Micro-neutralization assay
SARS-CoV-2 isolates, including CoV/Australia/VIC01/2020 (the ancestral strain) (23) were passaged in Vero cells and Omicron variants (specific strains BA.1, BA.2, and BA.5) were passaged on TMPRSS-expressing Vero cells and stored at -80°C. Sera from study participants were collected in tubes containing spray-dried K₂EDTA (1.8 mg/mL) and tested for neutralizing activity against SARS-CoV-2 variants.
Samples were heat-inactivated at 56°C for 30 minutes. Serial dilutions of serum, ranging from 1:20 to 1:10,240, were prepared before the addition of 100 TCID50 of the respective SARS-CoV-2 variant in MEM/0.5% BSA. The mixtures were incubated at room temperature for 1 h. Residual virus infectivity in the serum/virus mixtures was assessed in quadruplicate wells of Vero cells or TMPRSS-expressing Vero cells, as appropriate. The cells were incubated in serum-free media containing 1 µg/mL of TPCK trypsin at 37°C and 5% CO2, and viral cytopathic effect was evaluated on day 5. The nAb titer was calculated using the Reed-Muench method as previously described (24,25).
## Statistical analysis
The sample sizes used in our analyses were constrained by the number of individuals with available data in each of the two sources of participants, the DISCOVER-HCP and the COVID Profile studies. Among the mRNA vaccine recipients, the small number of Moderna mRNA vaccine recipients (n = 8) limited the power of any comparisons with Pfizer mRNA vaccine recipients. Antibody measures that were below the limit of detection were assumed to be the limit of detection (i.e., a titer of 10 in the microneutral ization assay), such that fold-rise measures were conservative, and the corresponding estimates provide a lower-bound on the true fold change.
A linear mixed-effects regression model was used to calculate the geometric means at each time point and the fold-change of the geometric means between two specified time points (with corresponding 95% confidence intervals [CI]). The outcome was the log antibody levels (RBD binding or micro neutralization assay titer [MNT]) and the models included fixed effects for timepoint, vaccination type (AdVV or mRNA), pre-study infection status (naïve or convalescent) and for MNT outcomes, and the virus variant (ancestral, Delta, BA1, BA2, or BA5). Repeated measures of individuals were accounted for via a random effect (intercept) for each participant. To obtain estimates by the vaccine type received and pre-infection status (and COVID-19 variant for MNT outcomes), all two-, three-, and four-way interaction terms were also included in the model.
The emmeans function (emmeans package in R [26]) was used to estimate the geometric mean at each time point from the model as a marginal mean effect. The margins command (margins package in R [27]) was used to estimate the fold-change in antibody levels between two time-points as the marginal effect. Corresponding two-sided 95% confidence intervals (95% CI) and P-values for pre-infection status, vaccine type, and antibody type combination from the linear mixed-effects model are reported. 95% CI will be quoted herein as (95% CI [lower limit, upper limit]). Statistical significance was assigned to P-values ≤ 0.05. No adjustment for multiple testing was applied to the confidence intervals or P-values given that the outcomes were not powered for.
Correlation of binding antibody to neutralizing antibody levels was calculated using Spearman rank correlation, with 95% CIs calculated via z-transformation.
## RESULTS
## Demographic distribution across study cohort groups
Overall, there was a higher number (68%-72%) of females in all study groups, except the Convalescent AdVV group, where only 40% were female (Table S1a). The median age for participants receiving the AdVV vaccine (52.0 years for Naïve and 59.0 years for Convalescent) was higher compared to participants who received the Pfizer primary vaccine (39.0 years for Naïve and 46.0 years for Convalescent). This is as expected, as Australians over 50 years of age were eligible for only the AdVV vaccine as part of the initial vaccine rollout in Australia (28). On average, around 40% of participants who received an AdVV primary vaccine had a breakthrough infection, as opposed to 52%-57% of participants who received the mRNA vaccine. However, the overall percentage of participants who experienced a breakthrough infection was similar for the Naïve group compared to Convalescent group (50% compared to 46%; Table S1b).
## Robust SARS-CoV-2-specific antibody levels after two doses of COVID-19 vaccination in both previously uninfected and convalescent individuals
Vaccine-induced antibody responses have been the subject of several studies which have provided valuable insights into the immune response to COVID-19 vaccination (29)(30)(31)(32). To determine if the vaccine-induced immune responses in our low-transmission study cohort align with previous observations, where robust antibody responses have been described after two COVID-19 vaccine doses, we measured IgG, IgM, and IgA antibody levels to several Spike protein-derived SARS-CoV-2 antigens including the receptor binding domain (RBD; Fig. 1A), S1, S2, Spike trimer, and the nucleoprotein (Fig. S3) in previously SARS-CoV-2 uninfected (naïve) individuals and in individuals who had recovered from SARS-CoV-2 infection (convalescent). Antibody levels for all isotypes assessed prior to vaccination (Pre-Vax) in convalescent individuals were tested a median of 233 days (range 153-422) days after initial diagnosis and median of 27 days (range 8-73) after a second dose of COVID-19 vaccination (Post-2nd vax). The naive and conva lescent groups were further stratified based on which COVID-19 vaccine was received in the primary vaccination, AdVV, or mRNA. Compared to pre-vaccination levels, we found that two doses of vaccine resulted in an increase in antibody levels to ancestral RBD-specific IgG antibody levels in all individuals (Fig. 1A, top panel). As expected, the relative change in RBD-specific IgG levels from pre-vaccination to post-2nd vaccine was greater in naïve participants that received AdVV (pre-vaccination geometric mean MFI (GM [95% CI]) 176.1 [118.7, 261.2] pre-vaccine, 11,154.8 [7,519.6, 16,547.5] post-vaccine) or mRNA (GM: 111.1 [81, 152.6] pre-vaccine, 16,704.5 [12,169.3, 22,929.8] post-vaccine) compared to convalescent participants who received an equivalent primary vaccina tion of either AdVV (MFI GM 3,388.3 [2,363.9,4,856.5] pre-vaccine, 17,988 [12,549.6, 25,782.4] post-vaccine) or mRNA (MFI GM 3,426.1 [2,371.9, 4,948.9] pre-vaccine, 23,603.4 [16,340.5, 34,094.5] post-vaccine; Fig. 1; Table S2a). The higher relative increase in IgG to RBD in participants with no prior antigen-exposure before vaccination resulted in IgG antibody levels to SARS-CoV-2 RBD post-2nd vaccine in naïve participants that were not substantially different from the corresponding levels in convalescent participants, who were vaccinated after recovery from infection. This observation was true for both AdVV and mRNA vaccines (Table S2a). Ancestral-RBD-specific IgM antibody levels (Fig. S2a). In contrast, no change in the IgM antibody levels was observed after two doses of vaccine in naïve participants that received AdVV (MFI GM 512. S2a).
While vaccination generally resulted in an increase in ancestral-RBD-specific IgA levels after two doses of vaccine (Fig. S4 S2a).
Collectively, these findings established that antigen-specific IgG binding antibod ies are induced by COVID-19 vaccination irrespective of vaccine type, but antibody responses following two doses of mRNA vaccine were higher than following two doses of AdVV vaccine. Notably, relative change in antibody levels between pre-and post-vac cination timepoints was most prominent in the SARS-CoV-2 naïve population, leading to antibody levels post-second vaccine being comparable between naïve and convalescent participants (Table S2b). Furthermore, increases in levels of antigen-specific IgA and IgM binding antibodies were observed to a lesser extent than for IgG and were not statistically significant for IgM.
## Vaccination induces neutralizing antibodies against the wildtype ancestral strain
While total antigen-specific antibody levels are a good measure of an overall humoral response elicited by COVID-19 vaccines, neutralizing antibodies (nAb) have been reported to be a correlate of protective immunity to SARS-CoV-2 (33). Therefore, in addition to binding antibody levels, we also determined neutralizing activity against live wildtype (ancestral) virus, in sera from each individual before vaccination (pre-vax) and after two doses of COVID-19 vaccine (post 2nd vax; Fig. 1B top panel). NAbs were detected after vaccination in both convalescent and naïve individuals, but the vaccine response was higher among the convalescent individuals. Average (geometric mean) titers of nAbs against the ancestral strain increased 4.7-fold (from GMT 10 [below the limit of detection of the assay] pre-vaccine to 46.9 [33.5, 65.7] post-vaccine) in the naïve group and 17.5-fold (GMT 29.8 [22.1, 40.2] to 519. 7 [385.1, 701.3]) in the convalescent group following two doses of AdVV vaccine. Naïve individuals who were vaccinated with two doses of mRNA vaccine had an average increase of 10.1-fold in GMT from pre-vac cination levels (from 10 [below limit of detection in the assay] to 100.9 [77.5, 131.3]; Fig. 1B top panel, Table S3). The corresponding GM fold-increase in the convalescent group was 68.1 (GMT from 22 [16.2, 29.9] to 1,496.4 [1,101.7, 2,032.6]). Of note, the fold-change in nAb titers from pre-vaccination to post-second vaccine was 3.7 times (95% CI [2.4, 5.8]) greater in convalescent participants than naïve participants for AdVV vaccine recipients and 6.7 times (95% CI [4.5, 10.1]) for mRNA recipients. Furthermore, after a second vaccine dose, the average (geometric mean) nAb titers in participants who received mRNA vaccines were 2.1 times (95% CI [1.4, 3.3]) higher for naïve (GMT for mRNA 100.9 [77.5, 131.3], AdVV 46.9 [33.5, 65.7]) and 2.9 times (95% CI [1.9, 4.4]) higher for convalescent individuals (GMT for mRNA 1496. 4 S3).
While high antigen-specific antibody levels do not always equate to high nAb titers, we found that 2-5 weeks after two doses of COVID-19 vaccines total RBD-specific IgG levels was strongly correlated with nAb titers in naïve (r = 0.65, 95% CI [0.44, 0.79]) and convalescent individuals (r = 0.85, 95% CI [0.76, 0.91]; Fig. S6).
## Two doses of COVID-19 vaccines induce lower neutralizing activity to the Delta variant than to the ancestral strain of SARS-CoV-2 despite high binding antibody levels
By December 2021, the Delta variant of SARS-CoV-2 had been circulating in Australia for ~6 months and the Omicron variant replaced it as the dominant circulating variant (5,34). At this time, approximately 87% of eligible Australians had received two doses of a COVID-19 vaccine (35). All of the participants in our study had received two doses of COVID-19 vaccines. Analysis of antibodies to the variants in our study cohort revealed that, like antibody-binding to the wildtype-derived RBD, there was a robust rise in total IgG antibody levels that bound to the Delta-derived RBD after two doses of COVID-19 vaccines in both vaccine groups (Fig. 1A, bottom panel). However, the GM fold-change in levels from pre-vaccination to post-2nd vaccine naïve participants was 4.6 times greater than in convalescent participants for both the AdVV (95% CI [2.4, 8.6]) and mRNA vaccine groups (95% CI [2.6, 8.2]) (Table S4) which reflects higher average pre-vaccination levels due to prior exposure in the convalescent group. Consequently, despite the greater fold increase, the average MFI antibody level post-2nd vaccine for naïve participants was, as expected, about half of the average level for convalescent participants (Table S4).
When comparing whether there were differences in binding capacity to Delta RBD post-2nd vaccine between individuals receiving mRNA or AdVV vaccine, we found that average Delta-RBD-binding antibodies were 1.9 (95% CI [1.1, 3.1]) times higher in naïve mRNA recipients compared to naïve AdVV vaccine recipients and 1.7 (95% CI [1.0, 2.7]) times higher in convalescent participants who received mRNA compared to the corresponding AdVV vaccine recipients (Table S4).
As with the neutralizing titers against the ancestral strain, increases in cross-reactive neutralizing titers to the Delta variant were also observed after two doses of vaccine (Fig. 1B, bottom panel). The average fold-change in GMT against the Delta variant was greater in convalescent participants compared to naïve individuals for both AdVV (4.0-fold, 95% CI [2.5, 6.2]) and mRNA vaccine recipients (11.4-fold, 95% CI [7.7, 17.1]). However, the GMT of vaccine-induced neutralizing activity against the wildtype variant was 1.6-2.9 times higher compared to the Delta variant (Table S5).
## Primary vaccination and a third vaccine dose result in equivalent IgG levels to Omicron subvariants BA1 and BA2
The Australian government recommended a third dose of COVID-19 vaccine to boost immunity in the population preceding the Omicron variant infection wave in Australia. Irrespective of the primary vaccine received (mRNA or AdVV), participants in our study received a third dose of COVID-19 vaccine using the mRNA formulation containing the S-protein sequence from the ancestral strain. We evaluated antibody responses to Omicron subvariants in our study participants 2-5 weeks after their third vaccine dose.
Here, we assessed the levels of IgG Abs binding to the RBD protein antigen derived from either BA1 or BA2 subvariants after primary vaccination (post-2nd vax), prior to vaccine dose three (pre-3rd vax) and after the third vaccine dose (post-3rd vax) in uninfected individuals that were enrolled as SARS-CoV-2 naïve and infected individuals who were convalescent from an infection with the ancestral strain (Fig. 2A).
We found that the levels of IgG binding to RBD from both subvariants were higher overall in naïve individuals who received two doses (primary vaccination) of the ancestral strain-derived mRNA vaccine (BA1 MFI GM 9102, 95% CI [7,159,11,573] [2,006, 3,606] and BA2 (MFI GM 4067, 95% CI [3,034,5,453]; Table 1). However, the vaccine subgroups (AdVV vs mRNA recipients) did not differ within the convalescent cohort (Table 1) apart for binding to BA1 RBD (AdVV MFI GM 5346 95% CI [4,044, 7,068 and mRNA MFI GM 9713 95% CI [7,347,12,841]). After primary vaccination at the timepoint between dose 2 and 3 (pre-3rd vax), the overall antigen-specific antibody levels had declined. However, average BA2 binding antibody levels were generally higher than BA1 binding antibody levels at this timepoint (Table 1).
Interestingly, in homologous vaccine recipients (mRNA for both primary vaccine series and third dose), for both naïve and convalescent groups, the average binding antibody levels for both Omicron sub-variants BA1 and BA2 were similar after two doses compared to average levels after three vaccine doses (Fig. 2A; Table 1). However, in recipients of heterologous vaccine (AdVV for primary vaccine series and mRNA for the third dose), binding IgG levels were approximately twofold higher after the third dose compared to levels after two doses of vaccine in naïve participants for BA1 (2.1 GM fold change 95% CI [1.5, 2.9]) and BA2 (1.8 GM fold change 95% CI [1.3, 2.5]; Table 1). For the corresponding convalescent subgroups, the fold-change between the two vaccine events was slightly lower for BA1 (1.8-fold change 95% CI [1.3, 2.5]; Fig. 2A; Table 1).
Binding antibody levels to both subvariants after three mRNA doses (homologous vaccination) in naïve participants were similar to those seen in convalescent participants. In participants receiving heterologous vaccination, binding antibodies to both subvar iants were lower in naïve participants compared to convalescent (Table 1). These data indicate that, for participants receiving AdVV vaccine, hybrid immunity against the ancestral strain is associated with higher binding antibody levels to Omicron subvariants compared to immunity generated by vaccine alone (naive).
## Neutralizing antibody titers against SARS-CoV-2 variants are significantly boosted by a third vaccine dose in naïve individuals
Since our study participants were vaccinated with the ancestral strain-derived S protein, we investigated whether primary vaccination followed by a booster (third) dose generated neutralizing activity against emerging new Omicron subvariants BA1, BA2, and BA5. Sera collected after primary vaccination (post-2nd vax), pre-vaccine dose 3 (pre-3rd vax), and post third vaccine dose (post 3rd vax) from naïve and convalescent individuals was tested in a MNT assay utilizing the ancestral strain or one of the Omicron variants as target (Fig. 2B). After primary vaccination in naïve individuals, there was a complete absence of detectable neutralizing activity to all Omicron subvariants tested (Fig. 2B). However, the third vaccine dose generated detectable neutralizing activity against Omicron subvariants in this cohort (Fig. 2B; Table 2). In convalescent participants, who had detectable nAb levels after two doses, we found that three doses of mRNA vaccine (homologous vaccination; post-3rd vax) resulted in similar neutralizing activity against all variants as was observed after two vaccine doses (post-2nd vax), indicating a restoration of antibody levels without significant boosting by the third dose. In contrast, heterologous vaccination in convalescent participants induced 2-3 times higher nAb titers against all variants after the third vaccine dose (post-3rd vax, Fig. 2B; Table 2) compared to titers after two doses (post-2nd vax).
After the third vaccine dose, nAb titers were similar for heterologous and homologous vaccination for the BA2 subvariant (Table 2) for both naïve and convalescent groups. In contrast, for subvariants BA1 and BA5, in convalescent participants, heterologous vaccine recipients had higher nAb titers (GMT for BA1 43.1 [31.4, 59.3]; for BA5 77.9 [56.7, 107.2]) compared to homologous vaccine recipients (GMT for BA1 23.3 [16.8, 32.2] and for BA5 30.4, 95% CI [21.8, 42.5]; Table 2). Notably however, after three vaccine doses, neutralizing activity against all the Omicron variants tested was markedly lower than the corresponding neutralizing activity against the ancestral strain for all individuals (Fig. 2B; Table 2).
## Breakthrough infections induce higher nAb titers against the ancestral strain than against Omicron subvariants
A subset of our study cohort (n = 61) acquired SARS-CoV-2 infection after enrollment. While genomic data related to the virus were not collected at the time of infection to identify the variant responsible for infection, the breakthrough infections coincided with the disappearance of the Delta variant and emergence and surge of Omicron variants in the community. The infections occurred between January 2022 and December 2022 when Australian Department of Health data indicate that ≥93% of reported infections were attributed to Omicron variants (Fig. S5).
We measured binding antibody levels in plasma and nAb titers in sera collected 2-4 weeks after breakthrough infections in vaccinated individuals who had previously recovered from infection with the ancestral variant (n = 25) and previously SARS-CoV-2 naïve (n = 26) individuals. Upon measuring IgG binding antibodies to RBD derived from different variants (ancestral, BA1, BA2), we found that post-infection RBD-binding levels for BA1 and BA2 were significantly increased compared to the corresponding pre-infec tion sera in the naïve participants who received either mRNA or AdVV primary vaccina tion (Fig. 3A; Table 3). For convalescent participants, irrespective of primary vaccination, the increase in binding IgG antibody levels to BA1 or BA2 RBD was not significant in individuals with breakthrough infections. Except for a few individuals, there was no significant increase in Ancestral RBD binding levels after a breakthrough infection.
In contrast to binding antibodies, nAb titers against all variants (ancestral, BA1, BA2, and BA5) were boosted after infection (post-infection) in all previously naïve individuals with a few exceptions (Fig. 3B; Table 4). In convalescent participants, the changes in nAb titers for the variants tested varied, exhibiting an increase in some and a decrease in others after infection. Additionally, pre-infection nAb titers against the ancestral strain in convalescent AdVV recipients (GMT 132, 95% CI [77,228]) and mRNA recipients (GMT 303, 95% CI [191,481]) were on average higher than corresponding titers in naïve individuals (AdVV recipients: GMT 52, 95% CI [31,88] and mRNA recipients: GMT 44, 95% CI [31,62]). However, despite having breakthrough infections with an Omicron subvar iant, there was a boost in nAb titers against the ancestral strain. This was particularly evident in the previously naïve group (Fig. 3B; Table 4).
Stratification of both the naïve and convalescent individuals based on receiving homologous or heterologous vaccination showed that average fold-increase in nAb titers from pre-infection to post-infection timepoints did not differ between the two vaccine groups (Table 4). As noted above, infection occurring during high levels of Omicron transmission in the community was associated with a boosting of nAb titers against both the ancestral virus and Omicron subvariants in the naïve participants (Fig. 3B; Table 4). These observations are consistent with enhanced response to the original vaccine antigen suggesting that antigen imprinting by vaccination with the ancestral strain had occurred.
## DISCUSSION
In this study, we used sera/plasma samples collected from individuals living in Australia, a low SARS-CoV-2 transmission country in the first two years of the COVID-19 pandemic, to perform comprehensive analysis of binding and neutralizing antibody levels in response to COVID-19 vaccines and breakthrough infections. The two salient findings of our study were, first, the need of a booster (third dose) of vaccine for the generation of neutralizing activity against Omicron variants, and second, the dominant boosting of neutralizing activity against the ancestral strain following infection with Omicron variants, indicative of imprinting of the immune response to the original antigen. If imprinting does occur in the context of vaccines, studies investigating the antigenic "distance" required to circumvent the imprinting would be of great interest for the design of future vaccines against SARS-CoV-2.
Several studies of vaccine responses to two-dose vaccination have been reported. However, most have been conducted in the setting of high (or unknown) levels of viral infections in the community or of work-related exposure in healthcare workers. This factor can influence immune responses to viral antigens. Here, we confirm previous findings of a robust rise in levels of binding antibodies to ancestral Spike antigens after two vaccine doses (29,36,37) in a population with low levels of community transmission. We also found that in the absence of exposure to the virus in a region with absent or extremely low levels of community transmission, the antibody responses induced by vaccination differed between the infection-naive and COVID-19 convalescent individuals as has been reported in conditions of ambiguous transmission (37,38).
Previous studies have shown that a third booster dose increases nAb titers and binding antibodies above the levels achieved by two doses of vaccine (29,39). Among the infection-naive individuals, the third vaccine dose boosted neutralizing activity against the variant virus strains as well as the ancestral strain. However, the third vaccine dose did not increase binding antibody levels in this group. The discordance between the binding antibody and nAb response to the vaccine likely reflects the broad targeting of the former to a range of Spike protein antigens, compared to the narrower subset of neutralizing Ab targets. Furthermore, in convalescent individuals, the booster dose produced minimal change in binding and nAb titers overall, presumably due to the higher absolute neutralizing levels already achieved by hybrid immunity in convalescent individuals who were subsequently vaccinated. Since the primary vaccinations in our study population were consistently either two doses of AdVV or two doses of mRNA vaccine, followed by an mRNA booster as the third dose in all participants, we were able to compare antibody responses after heterologous (AdVV/AdVV/mRNA) vs homologous (mRNA/mRNA/mRNA) vaccination. Previous reports demonstrate that heterologous vaccination induces broader and more durable antibody responses (40,41). Of note in the present study, in infection-naïve individuals, a third vaccine dose was required to generate neutralizing activity against Omicron subvariants, irrespective of primary vaccination type. Our data show that in naïve individuals, heterologous vaccination induced a greater boost in neutralizing antibody levels to the ancestral strain of SARS-CoV-2 than was observed after homologous vaccination. In addition, when we stratified the convalescent cohort by vaccine received (AdVV or mRNA), there was a significant boost of neutralizing activity to both the ancestral virus and variant strains in the AdVV recipients after the third vaccine dose (mRNA, heterologous vaccination) which supports previous reports that heterologous vaccination has the capacity to induce better crossvariant neutralization (42).
In individuals previously vaccinated and boosted with antigens derived from the ancestral strain of SARS-CoV-2, breakthrough infections with variants stimulate a de novo expansion of B cells targeting the altered viral spike glycoprotein but, at the same time, also an expansion of cross-reactive B and T cells previously sensitized to shared epitopes (17,43,44). "Imprinting" of immune responses refers to the concept whereby following first exposure to an antigen, immune responses to subsequent exposure to a closely related new antigen predominantly targets epitopes that are shared with the original antigen. Evidence for immunological imprinting has been found with Omicron infections (17,(45)(46)(47)(48)(49)(50). In these studies, in individuals previously vaccinated with the spike protein from the ancestral strain, Omicron infections were associated with a boost in neutralizing Ab titers against the ancestral strain as well as the infecting Omicron strain (17,19,48).
Our data are consistent with these reports. We demonstrated that Omicron infections boosted nAb titers against the ancestral strain as well as, or better than, nAb titers against the infecting Omicron strains (Fig. 3B). Importantly, our study was conducted during a period with virtually no community transmission of SARS-CoV-2 virus that could have an impact on de novo immune reeponses in our participants. In contrast to our study population, almost all the previous reports of immunological imprinting sourced samples from cohorts in settings of high virus transmission, with potential confounding effects on the immune responses observed (17,19,(43)(44)(45)(46)(47)(48)(49)(50).
The mechanisms underlying the phenomenon of imprinting await conclusive explanation; however, it has been proposed that epitope masking and feedback inhibition by pre-existing Abs may impede the recruitment of naive B cells specific to novel epitopes on variant spike proteins (19,51,52). Interestingly, in a study reported by Yisimayi and colleagues (53), robust variant-specific responses were seen after Omi cron infections in individuals who previously received inactivated SARS-CoV-2 vaccine, suggesting that inactivated virus vaccines may leave fainter immunological imprints compared with mRNA or vectored vaccines. The clinical significance of immunological imprinting is as yet uncertain. Booster vaccines containing spike proteins from BA.5 and XBB.1.5 remain very effective in preventing severe disease and deaths caused by these variants (54)(55)(56)(57)(58)(59)(60), suggesting that Abs directed against shared epitopes with the strain that imprints the immune response contribute to protection provided by the variant booster vaccines against severe outcomes.
Our study, that leveraged access to an increasingly rare COVID-19-naïve popula tion has some limitations. While the findings of the study give a unique longitudinal perspective of antibody responses following primary and booster vaccination and after breakthrough infection, the sample size is relatively small. This reflects the limitations of conducting research in a rapidly changing environment as a result of the pandemic as well as specific local factors, such as repeated lockdowns that prevented travel and visits to the clinic. As a result, we could not explore potential differences in antibody titers between males and females. The study did not include the elderly or children. This may limit the scope and generalizability of our results, but the consistency of the responses within each subgroup supports internal validity of the data and lends strength to our conclusions. Because the reporting of SARS-CoV-2 infections in Australia changed from PCR in centralized laboratories to self-testing with Rapid Antigen Tests (RATs), we do not have definitive information on the infective variant of the breakthrough infections. However, based on transmission data in Australia and locally in Melbourne, the timing of breakthrough infections that occurred in this cohort coincides with epidemiological data where 99% of the infections were caused by Omicron variants. B cell responses were not characterized in this study because the rapid implementation of research studies in the early phases of the SARS-CoV-2 pandemic did not allow us time to establish the necessary protocols. However, analysis of B cell proliferation in a subset of our participants using a mathematical model of in-host immune cell kinetics estimated that mRNA vaccines induced 2.1 times higher memory B cell proliferation than AdVV vaccines after adjusting for age, interval between doses and priming dose. Additionally, extending the duration between the priming dose and second vaccine dose beyond 28 days boosted neutralizing antibody production per plasmablast concentration by 30% (61). Additional analyses could have provided validation of our data and potentially elucidated underlying mechanisms for the observed patterns of immune responses to the vaccines and infections.
In conclusion, our study of vaccine-induced immune responses is unique because it was conducted in COVID-19-naïve and post-COVID-19 infected individuals in a setting where the confounding effects of community transmission and unintentional exposure to SARS-CoV-2 infections were circumvented. Thus, we characterized de novo antibody responses to three doses of vaccines as well as responses to infection with SARS-CoV-2 variants. We have demonstrated that, in these conditions, two doses of vaccines were insufficient for generation of nAb responses to the variant viruses that a 3rd dose of either a heterologous or homologous vaccine induced equivalent neutralizing antibody responses in both infection-naïve and convalescent individuals and that infection of vaccinated individuals with SARS-CoV-2 boosts levels of nAbs against the infecting variant in addition to the original vaccine virus, indicative of immune imprinting. Immune imprinting needs to be addressed in vaccine design and vaccination programs because the first experience with SARS-CoV-2 in different populations around the world varies greatly, as does the context and nature of subsequent exposure to the virus.
## References
1. *BA1 Naive AdVV*
2. (5326) "BA1 Convalescent AdVV"
3. (9028) "BA1 Naive"
4. (8118) *BA1 Convalescent*
5. "BA2 Naive Ad VV 11 5,256 15"
6. Convalescent (1541)
7. "BA2 Naive mRNA 25 6"
8. Ba "Convalescent mRNA 15 11,264 11"
9. "a MFI, median fluorescence intensity; CI, confidence interval"
10. "Using a regression model (detailed in the "Statistical analysis" section), P-value for difference from 1.0, signifying no change, in MFI values"
11. "BA1 Convalescent Ad VV"
12. "BA5 Convalescent AdVV 10 18"
13. "a GMT, geometric mean neutralizing titers; CI, confidence interval"
14. "Using a regression model (detailed in the "Statistical analysis" section), P-value for difference from 1.0, signifying no change"
15. Price, Shearer, Meehan et al. (2020) "Early analysis of the Australian COVID-19 epidemic"
16. Adekunle, Meehan, Rojas-Alvarez et al. (2020) "Delaying the COVID-19 epidemic in Australia: evaluating the effectiveness of international travel bans" *Aust N Z J Public Health*
17. Bloomfield, Ngeh, Cadby et al. (2022) "SARS-CoV-2 vaccine effectiveness against Omicron variant in infection-naive population" *Emerg Infect Dis*
18. Milazzo, Giles, Parent et al. (2022) "The impact of non-pharmaceutical interventions on COVID-19 cases in South Australia and Victoria" *Aust N Z J Public Health*
19. (2022) "COVID-19 Australia: epidemiology report 61"
20. Loesche, Karlson, Talabi et al. (2022) "Longitudinal SARS-CoV-2 nucleocap sid antibody kinetics, seroreversion, and implications for seroepidemio logic studies" *Emerg Infect Dis*
21. Perez-Saez, Zaballa, Lamour et al. (2023) "Long term anti-SARS-CoV-2 antibody kinetics and correlate of protection against Omicron BA.1/BA.2 infection" *Nat Commun*
22. Thevarajan, Nguyen, Koutsakos et al. (2020) "Breadth of concomitant immune responses prior to patient recovery: a case report of non-severe COVID-19" *Nat Med*
23. Rössler, Knabl, Laer et al. (2022) "Neutralization profile after recovery from SARS-CoV-2 Omicron infection" *N Engl J Med*
24. Gianfagna, Veronesi, Baj et al. (2022) "Anti-SARS-CoV-2 antibody levels and kinetics of vaccine response: potential role for unresolved inflammation following recovery from SARS-CoV-2 infection" *Sci Rep*
25. Akerman, Milogiannakis, Jean et al. (2023) "Emergence and antibody evasion of BQ, BA.2.75 and SARS-CoV-2 recombinant sub-lineages in the face of maturing antibody breadth at the population level" *EBioMedicine*
26. Cromer, Steain, Reynaldi et al. (2023) "Predicting vaccine effectiveness against severe COVID-19 over time and against variants: a meta-analysis" *Nat Commun*
27. Bánki, Seekircher, Falkensammer et al. (2022) "Six-month follow-up of immune responses after a rapid mass vaccination against SARS-CoV-2 with BNT162b2 in the district of Schwaz/Austria" *Viruses*
28. Rössler, Riepler, Bante et al. (2022) "SARS-CoV-2 Omicron variant neutralization in serum from vaccinated and convales cent persons" *N Engl J Med*
29. Gidding, Machalek, Hendry et al. (2021) "Seropreva lence of SARS-CoV-2-specific antibodies in Sydney after the first epidemic wave of 2020" *Med J Aust*
31. Macartney, Quinn, Pillsbury et al. (2020) "NSW COVID-19 Schools Study Team"
32. Addetia, Piccoli, Case et al. (2023) "Neutralization, effector function and immune imprinting of Omicron variants" *Nature*
33. Cao, Jian, Wang et al. (2023) "Imprinted SARS-CoV-2 humoral immunity induces convergent Omicron RBD evolution" *Nature*
34. Johnston, Li, Painter et al. (2024) "Immunological imprinting shapes the specificity of human antibody responses against SARS-CoV-2 variants" *Immunity*
35. Pérez-Alós, Hansen, Armenteros et al. (2023) "Previous immunity shapes immune responses to SARS-CoV-2 booster vaccination and Omicron breakthrough infection risk" *Nat Commun*
36. Eriksson, Hart, Forde et al. (2023) "Cohort PROFILE: A longitudinal Victorian COVID-19 cohort (COVID PROFILE)" *medRxiv*
37. Mazhari, Ruybal-Pesántez, Angrisano et al. (2021) "SARS-CoV-2 multi-antigen serology assay" *Methods Protoc*
38. Caly, Druce, Roberts et al. (2020) "Isolation and rapid sharing of the 2019 novel coronavirus (SARS-CoV-2) from the first patient diagnosed with COVID-19 in Australia" *Med J Aust*
39. Houser, Gretebeck, Ying et al. (2016) "Prophylaxis with a middle east respiratory syndrome coronavirus (MERS-CoV)-Specific human monoclonal antibody protects rabbits from MERS-CoV infection" *J Infect Dis*
40. Subbarao, Mcauliffe, Vogel et al. (2004) "Prior infection and passive transfer of neutralizing antibody prevent replication of severe acute respiratory syndrome coronavirus in the respiratory tract of mice" *J Virol*
41. Lenth (2023) "_emmeans: estimated marginal means, aka least-squares means_R package version 1"
42. Leeper (2021) "margins: marginal effects for model objects"
43. (2022) "Australia's COVID-19 vaccine rollout"
44. Belik, Jalkanen, Lundberg et al. (2022) "Comparative analysis of COVID-19 vaccine responses and third booster dose-induced neutralizing antibodies against Delta and Omicron variants" *Nat Commun*
45. Cho, Muecksch, Schaefer-Babajew et al. (2021) "Anti-SARS-CoV-2 receptor-binding domain antibody evolution after mRNA vaccination" *Nature*
46. Jalkanen, Kolehmainen, Häkkinen et al. (2021) "COVID-19 mRNA vaccine induced antibody responses against three SARS-CoV-2 variants" *Nat Commun*
47. Lucas, Vogels, Yildirim et al. (2021) "Impact of circulating SARS-CoV-2 variants on mRNA vaccine-induced immunity" *Nature*
48. Khoury, Cromer, Reynaldi et al. (2021) "Neutralizing antibody levels are highly predictive of immune protection from symptomatic SARS-CoV-2 infection" *Nat Med*
49. (2021) "A48A7CA2587320081F7BF/$File/covid_19_ australia_epidemiology_report_60_reporting_period_ending_10_april_ 2022.pdf 35. Australian Government Operation COVID Shield C-VR-O. 2021. Australian government operation COVID Shield, COVID-19 vaccine roll-out"
50. Liu, Sánchez-Ovando, Dowson et al. (2023) "Superior immunogenicity of mRNA over adenoviral vectored COVID-19 vaccines reflects B cell dynamics independent of anti-vector immunity: Implications for future pandemic vaccines" *Vaccine (Auckl)*
51. Wei, Pouwels, Stoesser et al. (2022) "Antibody responses and correlates of protection in the general population after two doses of the ChAdOx1 or BNT162b2 vaccines" *Nat Med*
52. Luczkowiak, Labiod, Rivas et al. (2022) "Neutralizing response against SARS-CoV-2 variants 8 months after BNT162b2 vaccination in naive and COVID-19-convales cent individuals" *J Infect Dis*
53. Yamamoto, Tanaka, Oshiro et al. "CoV-2 Seroepidemiological Study among NCGM staff. 2023. Antibody responses and correlates after two and three doses of BNT162b2 COVID-19 vaccine" *Infection*
54. Barros-Martins, Hammerschmidt, Cossmann et al. (2021) "Immune responses against SARS-CoV-2 variants after heterologous and homologous ChAdOx1 nCoV-19/BNT162b2 vaccination" *Nat Med*
55. Liu, Munro, Wright et al. (2023) "Persistence of immune responses after heterologous and homologous third COVID-19 vaccine dose schedules in the UK: eight-month analyses of the COV-BOOST trial" *Journal of Infection*
56. Hammerschmidt, Bosnjak, Bernhardt et al. (2021) "Neutralization of the SARS-CoV-2 delta variant after heterologous and homologous BNT162b2 or ChAdOx1 nCoV-19 vaccination" *Cell Mol Immunol*
57. Cao, Yisimayi, Jian et al. (2022) "BA.2.12.1, BA.4 and BA.5 escape antibodies elicited by Omicron infection" *Nature*
58. Kared, Wolf, Alirezaylavasani et al. (2022) "Immune responses in Omicron SARS-CoV-2 breakthrough infection in vaccinated adults" *Nat Commun*
59. Cao, Wang, Jian et al. (2022) "Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies" *Nature*
60. Koutsakos, Lee, Reynaldi et al. (2022) "The magnitude and timing of recalled immunity after breakthrough infection is shaped by SARS-CoV-2 variants" *Immunity*
61. Kuhlmann, Mayer, Claassen et al. (2022) "Break through infections with SARS-CoV-2 omicron despite mRNA vaccine booster dose" *Lancet*
62. Painter, Johnston, Lundgreen et al. (2023) "Prior vaccination promotes early activation of memory T cells and enhances immune responses during SARS-CoV-2 breakthrough infection" *Nat Immunol*
63. Wang, Zhou, Muecksch et al. (2022) "Memory B cell responses to Omicron subvariants after SARS-CoV-2 mRNA breakthrough infection in humans" *J Exp Med*
64. Weber, Dähling, Rose et al. (2023) "Enhanced SARS-CoV-2 humoral immunity following breakthrough infection builds upon the preexisting memory B cell pool" *Sci Immunol*
65. Bergström, Xu, Heyman (2017) "Epitope-specific suppression of IgG responses by passively administered specific IgG: evidence of epitope masking" *Front Immunol*
66. Schaefer-Babajew, Wang, Muecksch et al. (2023) "Antibody feedback regulates immune memory after SARS-CoV-2 mRNA vaccination" *Nature*
67. Yisimayi, Song, Wang et al. (2024) "Repeated Omicron exposures override ancestral SARS-CoV-2 immune imprinting" *Nature*
68. Carr, Wu, Gahir et al. (2023) "Neutralising immunity to Omicron sublineages BQ.1.1, XBB, and XBB.1.5 in healthy adults is boosted by bivalent BA.1-containing mRNA vaccination and previous Omicron infection" *Lancet Infect Dis*
69. Lin, Xu, Gu et al. (2023) "Durability of bivalent boosters against Omicron subvariants" *N Engl J Med*
70. Lin, Xu, Gu et al. (2023) "Effectiveness of bivalent boosters against severe Omicron infection" *N Engl J Med*
71. Shrestha, Burke, Nowacki et al. (2023) "Effectiveness of the coronavirus disease 2019 bivalent vaccine" *Open Forum Infect Dis*
72. Tan, Chiew, Pang et al. (2023) "Protective immunity of SARS-CoV-2 infection and vaccines against medically attended symptomatic omicron BA.4, BA.5, and XBB reinfections in Singapore: a national cohort study" *Lancet Infect Dis*
73. Tan, Chiew, Pang et al. (2023) "Effectiveness of bivalent mRNA vaccines against medically attended symptomatic SARS-CoV-2 infection and COVID-19-related hospital admission among SARS-CoV-2-naive and previously infected individuals: a retrospective cohort study" *Lancet Infect Dis*
74. Wang, Bowen, Tam et al. (2023) "SARS-CoV-2 neutralising antibodies after bivalent versus monovalent booster" *Lancet Infect Dis*
75. Hodgson, Liu, Mahanty et al. (2025) "Memory B cell proliferation drives differences in neutralising responses between ChAdOx1 and BNT162b2 SARS-CoV-2 vaccines" *Front Immunol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724334&blobtype=pdf | # A short intrinsically disordered domain of MCPyV ALTO regulates TBK1 signaling during MCPyV infection
Taylor Senay, Xiaomei Li, Sneha Shirhattikar, Tiana Luo, Jianxin You
## Abstract
Merkel cell polyomavirus (MCPyV) is an oncogenic human polyomavirus that latently infects most adults. Although the causative link between MCPyV and Merkel cell carcinoma is well established, the molecular mechanisms that govern viral latency and prevent oncogenic progression remain poorly understood. We previously reported that the MCPyV early protein ALTO is a key modulator of the STING-TBK1 signaling axis, enabling the virus to co-opt innate immune pathways to suppress excessive viral replication and promote latency over transformation. In this study, we expand on this model by identifying a short, essential domain within ALTO that is required for TBK1 activation. This domain, which we term LIT (Lost in Tau), is necessary for ALTO-TBK1 interaction but dispensable for ALTO trafficking and its interactions with STING or Src. When expressed alone, the LIT domain functions as a dominant negative inhibitor of wild-type ALTO, competitively blocking TBK1 activation through a novel TBK1 interaction domain. Deletion of the LIT domain from ALTO not only abolishes TBK1 interaction and downstream phosphorylation but also eliminates TBK1-mediated suppression of MCPyV replication during early infection of human dermal fibroblasts. These findings provide mechanistic insight into how ALTO promotes viral persistence and immune evasion. More broadly, they highlight the functional importance of intrinsically disordered regions in modulating host-virus interactions and suggest that MCPyV latency is actively maintained through a finely tuned balance of pro-and anti-viral signaling. Identification of domains such as LIT provides new insight into TBK1 regulation and informs future strategies to study viral persistence. IMPORTANCE Merkel cell polyomavirus causes lifelong, latent infections in the skin of most people. When this latency is perturbed, the virus can give rise to Merkel cell carcinoma (MCC), an aggressive and difficult-to-treat skin cancer. Efforts to prevent this cancer depend on understanding what controls viral latency and persistence. We previously reported that MCPyV stimulates the host's STING-TBK1 signaling axis to limit its own replication. In this work, we identify and characterize a short amino acid motif within the ALTO protein's intrinsically disordered region that is required for this immunestimulating activity. This region appears to be critical for helping the virus maintain latency by fine-tuning the host's response. Our findings provide new insight into how MCPyV may regulate early infection dynamics and suggest that identifying functional domains such as LIT could help guide future approaches to study viral persistence and its contribution to MCC.
KEYWORDS viral latency, TBK1, ALTO, MCPyVT he viral family Polyomaviridae includes approximately 117 species of non-enveloped viruses carrying small (~5 kb) circular double-stranded DNA genomes (1). Of these, around 16 are thought to be human-specific (2), and only one, Merkel cell polyomavirus (MCPyV or MCV), has been conclusively demonstrated to be carcinogenic (3). MCPyV
is ubiquitous in the human population and, for most individuals, exists as a silent member of the skin virome, establishing lifelong episomal latency in cutaneous cells (4)(5)(6)(7)(8). However, in a subset of cases, MCPyV integrates into the host genome, where it drives oncogenesis of the rare but aggressive neuroendocrine skin cancer Merkel cell carcinoma (MCC) through constitutive high-level expression of its Small Tumor (sT) antigen and a truncated form of its Large Tumor (LT) antigen (3,9,10).
Nearly two decades after its discovery as the major causative agent in approximately 80% of MCCs (3,11,12), most mechanistic studies of MCPyV have focused on LT and sT. These viral effectors have homologs in other well-characterized polyomaviruses, such as SV40, and their roles in MCPyV replication and transformation are well established (13)(14)(15). However, the upstream events that initiate viral genome truncation and integration, as well as the cellular conditions that distinguish persistent infection from oncogenic progression, remain poorly understood. These represent critical gaps in our knowledge of MCPyV biology. Elucidating the regulatory mechanisms that govern this transition is essential for improving early detection and prevention of MCC, which currently has limited and often non-durable treatment options.
There is currently no consensus on how MCPyV establishes, maintains, or exits latency. Clinical observations reveal that immunosuppressed individuals carry higher MCPyV viral burdens and are at substantially increased risk of developing MCC, suggesting that innate immune control is a key determinant of viral latency (16)(17)(18). The cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS-STING) pathway, a cytosolic DNA-sensing system that activates type I interferon and proinflammatory signaling, has emerged as particularly important in this context (19)(20)(21)(22)(23)(24)(25). We have previously reported that STING expression is dramatically reduced in virus-positive MCC tumors, and that reactivation of STING signaling promotes MCC cell death and T-cell infiltration (26). Conversely, during early infection, MCPyV strongly activates the STING pathway and its downstream proinflammatory effectors, which in turn restricts viral replication and supports the establishment of a balanced, persistent infection (27)(28)(29). Together, these findings suggest that MCPyV latency is not a passive process, but rather one that is tightly regulated through interactions between viral effectors and host innate immune pathways. Disruption of this equilibrium, whether through viral mutation or host immune suppression, may be a critical inflection point that drives malignant progression.
While LT and sT have been the primary focus of MCPyV research, less attention has been paid to another early region protein, the alternative large tumor antigen open reading frame (ALTO). First identified in 2014, ALTO was initially thought to be dispensa ble for replication in transfection-based assays (30). However, recent studies, including our own, have revealed its active role in immune signaling. ALTO has been shown to modulate Src-family kinases, PLCγ1 signaling (31), and the NFκB pathway (31,32). In our recent work, we demonstrated that ALTO acts as a modulator of the STING-TBK1 axis, forming a signaling complex with STING, TBK1, and Src to induce TBK1 autophos phorylation, leading to its degradation and resulting in viral replication suppression and latency promotion (29). This function places ALTO in the unique position of acting as an "anti-viral viral protein" that dampens its own replication, evades immune detection, and persists in the host.
These findings reshape our understanding of ALTO as a multifunctional effector in MCPyV biology, one that fine-tunes host responses through structurally flexible, context-dependent interactions. In this study, we build upon our prior work to dis sect the structure-function relationship of ALTO in greater detail. Using a systematic truncation approach, we identify a short, previously uncharacterized region within ALTO's intrinsically disordered domain that is required for its interaction with TBK1 and its immunomodulatory function. Through this analysis, we aim to illuminate how ALTO contributes to latency maintenance, how disruption of its function may pro mote oncogenesis, and how ALTO-targeted mechanisms could be leveraged to better understand or intervene in MCPyV-associated disease.
## RESULTS
## Identification of the TBK1-activating domain within ALTO's N-terminal intrinsically disordered region
As shown in our previous study published in 2024, one of ALTO's key features is its ability to stimulate TBK1 autophosphorylation (29). To identify the region(s) of ALTO responsible for stimulating TBK1 autophosphorylation, we began by evaluating the current understanding of ALTO's structural features. The ALTO protein remains poorly characterized at the structural level, and even its precise length is debated: while early studies described a 250-amino acid (aa) product encoded across two exons, more recent annotations favor a 248-aa version encoded within a single exon (30)(31)(32). As no conclusive functional differences between these isoforms have been reported, we included both in our in vitro studies and observed no phenotypic distinctions between them.
Multiple reports of in silico analyses agree that ALTO's most stable structural feature is a C-terminal alpha helix (approximately aa 200-250) that contains a 22-aa trans membrane or hydrophobic domain (aa 225-246). In contrast, the N-terminal region, representing roughly 80% of the protein, is intrinsically disordered (Fig. 1A) (33)(34)(35). Although intrinsically disordered regions (IDRs) are known to be highly plastic and capable of a wide range of functions (36)(37)(38)(39), their lack of fixed globular shape makes structure-based functional prediction based on modeling difficult. We therefore took a methodical but exploratory approach using progressive N-terminal focused truncation. Each truncation shortened the protein by 30 aa compared to the one before, allowing us to examine region-specific phenotypic variation with fine resolution (Fig. 1B). In addition to the progressive N-terminal truncation mutants, we constructed a C-terminal ALTO truncation variant lacking the hydrophobic or transmembrane domain (ALTOΔTM). These variant sequences were generated using the NanoBiT platform, a structural complementation-based split enzyme system, which quantifies tagged protein-protein interactions in live cells via relative luminescence. We have previously found that ALTO favors N-terminal functional tags (29), so all constructs were generated with N-terminal LgBiT and SmBiT tags-the large and small subunits, respectively, of a reconstitutable NanoLuc luciferase enzyme-enabling protein detection and downstream protein-pro tein interaction analysis (Fig. 1B).
To test which region or regions of ALTO govern its TBK1-stimulatory activity, we transfected HEK 293T cells with constructs encoding LgBiT-tagged ALTO progressive truncation mutants or an empty vector control. Expression of most variants was roughly equivalent, and all migrated to positions consistent with their predicted molecular weights (Fig. 1C). Although structural prediction of ALTO remains largely inconclusive, AlphaFold modeling suggested that all variants preserved the characteristic N-terminal disordered region and C-terminal alpha helix (Fig. S1). Since ALTO is also known to form homo-oligomers (29,31), we used ALTO-ALTO interactions as a functional readout to assess the structural integrity of the truncation mutants. Self-interaction assays con firmed that variants retained appropriate homo-and hetero-oligomerization capability (Fig. S2), further supporting the functional integrity of these constructs.
We then tested whether the ALTO truncation mutants retained the ability to stimulate TBK1 autophosphorylation. As observed with wild-type (WT) ALTO, the first four truncation mutants (ALTOΨ, ALTOχ, ALTOΦ, and ALTOυ) retained the ability to stimulate TBK1 autophosphorylation (Fig. 1C, lanes 1-5). In contrast, the ALTOτ truncation (aa 151-250) and all shorter N-terminal truncated variants completely lost this activity despite comparable protein expression levels (Fig. 1C, lanes [6][7][8]. ALTOΔTM also failed to stimulate TBK1 autophosphorylation and was expressed at a relatively lower level (Fig. 1C, lane 9). This, taken with its overall lowered homo-and hetero-oligomerization capability (Fig. S2), supports previous reporting that the putative transmembrane or hydrophobic domain is required for robust expression of ALTO (32).
## The LIT domain is specifically required for ALTO-TBK1 interaction but not for binding STING or src
The loss of TBK1 autophosphorylation in ALTOτ and shorter variants may indicate a disruption in the formation of the ALTO-STING-TBK1-Src "signal complex" we previously described (29). To determine whether the loss of TBK1-stimulatory activity in certain truncation mutants was due to disrupted complex formation-and to identify which specific protein-protein interactions might underlie the observed loss of function-we co-transfected HEK 293T cells with SmBiT-tagged ALTO truncation mutants and LgBiTtagged STING, TBK1, or Src (Fig. 2A through C). Kinetic monitoring of these interactions showed that after a short equilibration period, all readings remained relatively stable over the observation window; therefore, single representative timepoints are presen ted throughout this study (Fig. S3). For all three interactors, interaction strength was substantially reduced or eliminated at different truncation points, suggesting that STING, Src, and TBK1 each rely on distinct ALTO regions for binding.
Despite its inability to stimulate TBK1 autophosphorylation as we observed in Fig. 1C, ALTOτ showed the strongest interaction with STING (Fig. 2A) and retained approximately half-maximal interaction with Src (Fig. 2B), indicating it remains structurally and functionally competent in other respects. ALTO-TBK1 interaction peaked with the ALTOΦ variant, declined substantially with ALTOυ, and was reduced to near-background levels at ALTOτ (Fig. 2C). These data suggest that ALTOτ is properly folded and localized but lacks a domain specifically required for ALTO-TBK1 engagement. Taken together, the TBK1 activation and protein-protein interaction assays identify the 30 aa region absent in ALTOτ (aa 121-150) as a critical functional segment. We therefore refer to this region as the Lost in Tau (LIT) domain, a designation specific to this short motif within MCPyV ALTO and unrelated to the neuronal microtubule protein Tau. Although direct Western blot validation of every truncation variant was not possible, Fig. 1C demonstrates that LgBiT fusions are expressed at comparable levels, and each interaction assay includes a LgBiT: Negative control condition as an internal reference for nonspecific signal.
## The LIT domain functions as a dominant negative inhibitor of ALTO-induced TBK1 autophosphorylation
To further investigate whether the LIT region is sufficient to confer TBK1-related activity, we tested whether the domain alone could recapitulate or inhibit full-length ALTO function. We first cloned the LIT sequence into the NanoBiT vectors for use in sensitive live cell interaction assays, then subcloned the same reading frame with the addition of the N-terminal LgBiT tag into the pcDNA4c vector, adding an Xpress tag and placing the construct under the control of a highly active CMV promoter for applications requiring more robust expression (Fig. 3A). NanoBiT protein-protein interaction studies of LIT's interactions with WT ALTO, STING, Src, and TBK1 were initiated using the standard LgBiT-LIT construct; however, we observed substantial differences in expression levels between WT ALTO and LIT, as well as high variability and background luminescence readings in both experimental and negative control conditions (Fig. S4). Because higher LIT expression would also produce proportionally higher background signal, we concluded that the NanoBiT platform was not suitable for reliably examining LIT interactions.
Since the LIT domain appeared to be critical for ALTO-TBK1 interactions (Fig. 1C and 2C), we hypothesized that it might act as a dominant negative inhibitor by competitively interfering with TBK1 activation by WT protein. To test this, we co-transfected our Xpress-LgBiT-LIT construct with WT ALTO and assessed TBK1 autophosphorylation as well as total TBK1 protein levels. As expected, modest expression of WT ALTO alone strongly stimulated TBK1 autophosphorylation, whereas even much higher expression of Xpress-LgBiT-LIT failed to do so (Fig. 3B). Co-transfection of the two constructs showed a marked decrease in TBK1 autophosphorylation, confirming a dominant negative phenotype. To both quantify the reduction caused by the dominant-negative effect of LIT and test its effect in normal human dermal fibroblasts (HDFs), we repeated our co-transfection experiment and performed In-Cell Western analysis (40) The ratio of phosphorylated TBK1 to total TBK1 was decreased by around 70% when LIT and WT ALTO were cotransfected as compared with WT ALTO alone (Fig. 3C; Fig. S5).
An alternate ALTO mutant containing the N-terminal 150 aa of ALTO-terminating with the LIT region-was tested as a more stable truncation construct than LIT alone, without requiring a bulky N-terminal tag (Fig. S6). This variant, termed ALTO150, was stably and detectably expressed; however, it had only a modest inhibitory effect on TBK1 autophosphorylation when co-transfected with WT ALTO. Co-transfection also resulted in a modest increase in WT ALTO abundance, suggesting that the N-terminal domain of ALTO mediates its dimerization, which in turn contributes to protein stabilization. While ALTO150 exhibited some dominant negative activity, its stabilization of WT ALTO may offset this effect, creating a self-limiting interaction that blunts its ability to fully suppress WT ALTO function.
## Deletion of the LIT domain abolishes TBK1 activation without disrupting ALTO structure, oligomerization, or other partner binding
While truncation mutants helped define the boundaries of the LIT domain, we next asked whether targeted deletion of this region would phenocopy the loss of function seen in ALTOτ. To determine whether the observed loss of interaction and TBK1 stimula tion was due to a sequence-specific function of the LIT region or simply a consequence of reduced length or flexibility, we generated an ALTO variant lacking the LIT coding sequence, termed ALTO deletion of LIT (ALTOdelLIT) (Fig. 4A). The removal of a random 30-aa segment was not expected to produce a phenotype, as a deletion of this length in the ALTOΨ truncation did not impair TBK1 autophosphorylation or interaction (Fig. 1C and2C).
Consistent with previous results, LgBiT-tagged WT ALTO robustly stimulated TBK1 autophosphorylation and degradation in transfected HEK 293T cells. In contrast, ALTOdelLIT failed to induce either autophosphorylation or degradation of TBK1, despite comparable protein expression levels (Fig. 4B). NanoBiT interaction assays confirmed that this loss of function was not due to impaired ALTO oligomerization, as ALTOdelLIT retained equal or even enhanced capability to oligomerize with itself and with WT ALTO (Fig. 4C). Similarly, interactions between ALTOdelLIT and STING (Fig. 4D), as well as ALTOdelLIT and Src (Fig. 4E), were not significantly different from those of WT ALTO. Interaction between ALTOdelLIT and TBK1, however, was reduced by more than 30-fold compared with WT ALTO (Fig. 4F), indicating a specific requirement for the LIT region in TBK1 engagement. Protein expression across all interaction assays was largely equal, further supporting the reliability of these findings (Fig. S7). The preserved interactions between ALTOdelLIT and ALTO, STING, and Src suggest that ALTOdelLIT is structurally and functionally intact, and that its loss of TBK1 engagement reflects a specific require ment for the LIT region. The LIT region lies within ALTO's N-terminal intrinsically disordered region (IDR), and as such, no confident structure-based predictions can be made about its function (Fig. 1A). To further investigate whether specific functional domain(s) of ALTO within the LIT region might drive the ALTO-TBK1 interaction, we analyzed its amino acid sequence for short linear motifs (SLiMs), also known as eukaryotic linear motifs (ELMs), using the ELM Resource prediction tool (http://elm.eu.org/search.html). The analysis was performed without specifying a subcellular compartment, as ALTO's subcellular localization remains poorly defined. The resulting predictions were then filtered to include only those located within or adjacent to the LIT region (Fig. S8). Although SLiMs are inherently degener ate and therefore not considered high-confidence predictions when found, no motifs corresponding to known TBK1-binding sites or interaction domains with other signal complex components were identified.
## The LIT domain does not influence ALTO subcellular localization or trafficking
Given that ALTOdelLIT retains structural integrity and continues to interact with Src, STING, and itself, we next hypothesized that differences in intracellular trafficking might account for the phenotypic divergence between WT ALTO and ALTOdelLIT. To test this, we generated HDFs expressing either WT ALTO or ALTOdelLIT under a doxycy cline (Dox)-inducible promoter (inALTOWT and inALTOdelLIT HDFs, respectively). After several hours of Dox induction, cells were fixed and immunostained for ALTO and various intracellular membrane markers. Substantial colocalization was observed only with the cis-Golgi marker GM130, where both WT ALTO and ALTOdelLIT appeared to accumulate during early expression (Fig. 5A). At later time points, both proteins were observed to traffic outward from the Golgi, potentially via the secretory pathway (Fig. S9). Further immunofluorescence analysis revealed partial colocalization of distal ALTO and ALTOdelLIT foci with the early endosome marker Rab5 (Fig. 5B), suggesting ALTO may transiently localize to the plasma membrane and re-enter the cell via the endocytic pathway, a finding consistent with prior reports of ALTO co-fractionating with lipid raft markers (31).
Although we did not identify a specific motif or altered trafficking pattern that could explain the loss of ALTO-TBK1 interaction in the absence of the LIT domain, these findings suggest that ALTO's trafficking may involve dynamic cycling through the secretory and endocytic pathways. Further investigation into ALTO's trafficking and intracellular lifecycle may provide important insights. Despite the absence of clear structural or localization changes, our data demonstrate that the LIT domain is specifically required for ALTO-induced TBK1 autophosphorylation.
## The LIT domain is essential for ALTO-mediated suppression of MCPyV replication in infected cells
Finally, to place our protein-level observations into their full viral context, we evaluated whether WT ALTO and ALTOdelLIT could suppress MCPyV replication by using our HDF model of early MCPyV infection (41). Donor-matched inALTOWT and inALTOdelLIT HDFs (Fig. 6A) were infected with equal titers of ALTO null MCPyV, a genetically intact virus containing a point mutation in the ALTO start codon that eliminates ALTO expression (29). Following infection, cells were either mock-induced or induced to express their respective ALTO variant, and viral replication was assessed 5 days post-infection.
As we previously reported, ALTO null virus replicates significantly more efficiently in uninduced HDFs compared with HDFs induced to express WT ALTO. In contrast, induc tion of ALTOdelLIT failed to significantly reduce viral replication (Fig. 6B). Because the suppression of viral replication by ALTO is known to occur through a TBK1-driven negative feedback mechanism (29), this loss of function demonstrates that the LIT domain is required for TBK1 stimulation by ALTO during viral infection.
## DISCUSSION
MCPyV remains the only conclusively identified oncogenic human polyomavirus. Despite its widespread presence in the general population and its clear role in the pathogenesis of MCC, major gaps persist in our understanding of key aspects of MCPyV biology. Among the most critical of these are the mechanisms by which MCPyV establishes persistent infection, maintains latency, and ultimately exits latency to drive oncogenesis.
This study builds on our previous work proposing an MCPyV latency mechanism involving ALTO-mediated modulation of the STING-TBK1 axis. To further define this mechanism, we used a progressive N-terminal truncation approach spanning ALTO's intrinsically disordered region (IDR) and alpha-helical domain to identify regions critical for TBK1 activation (Fig. 1A andB). Through this strategy, we discovered a short, previ ously uncharacterized region of ALTO that is specifically required for TBK1 stimulation. The ALTOτ mutant, which retains only the C-terminal 100 aa of ALTO (aa 151-250), was the first truncation to completely lose the ability to stimulate TBK1 autophosphorylation in cellulo (Fig. 1C) and showed almost no detectable interaction with TBK1 in proteinprotein interaction assays (Fig. 2C). In contrast, its immediate upstream predecessor, ALTOυ, retained both activity and binding, indicating that the 30 aa present in ALTOυ but absent in ALTOτ (aa 121-150) are essential for this function. We therefore designated this region the LIT (Lost in Tau) domain. Located within ALTO's IDR, the LIT domain is predicted to localize at the interface of ALTO, STING, TBK1, and Src within the multiprotein signal complex (Fig. 6C). When expressed in isolation, the LIT domain acted as a dominant negative inhibitor of WT ALTO, competitively blocking TBK1 activation (Fig. 3B andC). To confirm that the LIT domain is necessary for ALTO's immune modulatory activity, we engineered a deletion construct (ALTOdelLIT) and found that it failed to stimulate TBK1 autophosphorylation (Fig. 4B). This loss of function appears to be sequence-specific: ALTOdelLIT maintained expression, proper folding, and interactions with other signal complex proteins such as STING and Src, but showed a dramatic reduction in TBK1 binding (Fig. 4F). This loss was not attributable to the absence of a canonical TBK1-interaction motif (Fig. S8) or to altered protein trafficking, as ALTO and ALTOdelLIT both localized to the cis-Golgi at early time points and trafficked through both the secretory and endocytic pathways (Fig. 5A andB). While our data demonstrate that the LIT domain is specifically required for ALTO-mediated TBK1 activation, it remains to be determined whether LIT can also modulate TBK1 phosphorylation in response to other upstream stimuli such as poly I/C or viral infection. Finally, we demonstrated that the LIT domain is functionally required for TBK1mediated suppression of MCPyV replication during early infection (Fig. 6A). Notably, ALTOdelLIT induction failed to reduce viral replication to the extent observed with direct pharmacological inhibition of TBK1 (29) in our prior study. These results support the conclusion that ALTO, through the LIT domain, plays a central role in modulating hostvirus interactions during early infection.
It is important to note that limitations inherent to cell culture modeling of MCPyV infection prevent straightforward, direct examination of viral latency. Our examinations of exogenously expressed viral proteins, and even of early viral infection, can only reveal general dynamics in limited temporal windows. We therefore cannot exclude the possibility that chronic MCPyV infection may behave differently than we are able to observe in vitro. Accordingly, while our findings support a model in which ALTO helps restrain replication through TBK1 activation, they do not directly demonstrate regulation of viral latency, and definitive proof of ALTO's role in latency will require more complex systems that capture long-term infection dynamics. Nevertheless, we feel that the experimental observations we have described reflect a possible mechanism of latency, since a means to limit viral replication during early infection would also likely serve a similar function during persistent infection.
The identification of a novel immune interaction domain within ALTO's IDR builds on prior reports implicating other regions of ALTO in Src-family kinase binding (29,31), PLCγ1 signaling, and NFκB modulation (31,32). Collectively, these studies revise the earlier view of ALTO as a dispensable accessory protein and instead establish it as a key viral effector that actively shapes innate immune responses through structurally dynamic interactions. As ALTO-encoding polyomaviruses continue to emerge as a phylogenetic subset of mammalian polyomaviruses (32), it will be important to assess whether the LIT domain is conserved across species or represents a specialized adaptation of MCPyV to its human host. Given that many polyomaviruses establish persistent infections, conservation of LIT-like motifs would support the idea that latency is a regulated process driven by active host-virus interactions, rather than a passive consequence of viral quiescence.
MCPyV ALTO is not unique among viral or oncogenic proteins in its use of intrinsic disorder to regulate complex signaling networks. Other viral effectors, including HPV E7 (42) and KSHV ORF57 (43), also use disordered regions to engage host pathways with high flexibility and specificity. As the functional importance of intrinsically disordered viral proteins becomes more widely appreciated, new opportunities are emerging to map these regions and understand their mechanistic roles in immune evasion, persistence, and oncogenesis (44)(45)(46).
Several key questions remain regarding the lifecycle and regulation of ALTO during MCPyV infection. Although we observed initial accumulation of ALTO at the cis-Golgi, followed by trafficking through secretory and endocytic pathways (Fig. 5; Fig. S9), and prior studies have reported association with lipid rafts (31), the complete subcellular dynamics of ALTO, particularly during long-term infection, are not yet well defined. Given the known importance of spatial regulation for many signaling proteins, a deeper understanding of ALTO's localization and trafficking may reveal new insights into its diverse functions and regulatory partners.
A particularly intriguing and unresolved observation is the frequent loss of ALTO expression in virus-positive MCC, even in cases where the ORF is retained (32). While some loss may result from stochastic truncation events during viral integration, ALTO protein is undetectable in several virus-positive MCC lines that retain its intact coding sequence. From an evolutionary perspective, this may be a form of immune evasion. ALTO is highly immunostimulatory, and its continued expression after integration could trigger immune clearance of the host cell. Thus, the absence of ALTO in MCC tumors may reflect a form of survivorship bias. Only those cells that have lost ALTO expression may escape immune detection and progress to cancer. Since ALTO can be robustly expressed even in the absence of all other viral proteins, this suggests that host-medi ated regulatory mechanisms, rather than viral factors, may suppress ALTO post-integra tion. Identifying such factors could reveal novel therapeutic targets or biomarkers for MCC progression.
Finally, the discovery of a compact domain within ALTO that selectively activates TBK1 opens the door to potential translational applications. ALTO itself or a small molecule mimic of its TBK1 activating domain could serve as a noncanonical TBK1 agonist for boosting antiviral immunity in immunocompromised individuals, poten tially suppressing MCPyV replication and reducing cancer risk. Beyond direct antiviral applications, TBK1 is a known druggable target for a variety of conditions associated with its dysregulation, as well as a target of interest in the development of novel cancer immunotherapies (47)(48)(49). Identification and characterization of the noncanonical mechanism by which ALTO is able to stimulate TBK1 autophosphorylation could also enable the development of novel agonists and antagonists. Conversely, the LIT domain itself, or derivatives thereof, could act as a dominant negative inhibitor of WT ALTO by blocking its ability to activate TBK1. This raises the possibility of repurposing LIT-mimick ing compounds to modulate ALTO-TBK1 signaling in MCPyV-associated conditions, such as latency control or early-stage infection.
## MATERIALS AND METHODS
## Cell culture
All cells were maintained at 37°C with 5% CO 2 in a humidified incubator. HEK 293T cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, Gibco) supplemented with 10% fetal bovine serum (FBS, Cytiva). HDFs were isolated from neonatal foreskin as previously described (50) and maintained in DMEM supplemented with 10% tetracy cline-screened FBS (Cytiva).
## Chemicals and reagents
Doxycycline (Dox) and puromycin were dissolved in sterile water to stock concentrations of 0.5 and 10 mg/mL, respectively, and stored at -20°C. Unless otherwise noted in the text or figure legends, Dox was used at a final concentration of 0.5 µg/mL. Puromycin was used during selection for stable cell lines as described below. Hexadimethrine bromide (Polybrene, Aldrich) was dissolved in water to a 12 mg/mL stock, sterilized by 0.2-µm syringe filtration, and used fresh for each experiment.
## Recombinant plasmid construction
NanoBiT constructs expressing WT ALTO, TBK1, and STING were generated in our previous study (29). The human c-Src coding sequence (hereafter "Src") was amplified from the pDEST40-ZXHA-WtSrc plasmid and cloned using XhoI and NheI sites. LgBiT-and SmBiT-tagged truncated ALTO mutants (Ψ, χ, Φ, υ, τ, σ, ρ, ΔTM, and LIT) were PCRamplified from the pBiT1.1-N_ALTO (L-ALTO) backbone and subcloned into pBiT1.1-N[TK/ LgBiT] or pBiT2.1-N[TK/SmBiT] vectors using XhoI and NheI. Deletion constructs pBiT1.1-N_ALTOdelLIT (L-ALTOdelLIT) and pBiT2.1-N_ALTOdelLIT (S-ALTOdelLIT) were constructed by linear PCR omitting the LIT region, followed by Bpu10I digestion and self-ligation.
Constructs for pcDNA4c-ALTO150 and pcDNA4c-LgBiT-LIT were prepared by amplifying their respective coding sequences from their respective NanoBiT plasmids and subcloning into the pcDNA4c vector (Invitrogen) using BamHI and XhoI or EcoRI, respectively.
The pTRIPz-RFP vector (Open Biosystems) and pTRIPz-ALTO construct were generated as previously described (29). pTRIPz-ALTOdelLIT was derived from pBiT1.1-N_ALTOdelLIT and cloned into the pTRIPz backbone using the AgeI and MluI sites to replace the RFP cassette. All oligonucleotides were obtained from Integrated DNA Technologies (IDT); cloning primer sequences are provided in Table S1.
## Structural prediction
Structural modeling of WT ALTO (Fig. 1A) was performed as previously described (29,35). Progressive ALTO truncation mutants and the STING-TBK1-ALTO-Src complex were modeled using the AlphaFold3 server (builds 2025.5.23 and 2025.06.10) with default settings (34). Models were visualized and recolored using NIH iCn3D (51,52).
## Generation of inducible stable HDFs
Inducible HDF lines were generated as previously described (27,50). Briefly, lentiviral particles were produced in HEK 293T cells by co-transfecting pTRIPz transfer plasmids (encoding WT ALTO or ALTOdelLIT), psPAX, and pMD2.G using Lipofectamine 2000 (Invitrogen). Supernatants were collected 48 h post-transfection, centrifuged, 0.2-µm filtered, and supplemented with 6 µg/mL Polybrene prior to transduction of early-pas sage HDFs (≤2 passages post-isolation). After 48 h, culture medium was replaced, and cells were selected with 4 µg/mL puromycin for 5 days, followed by maintenance in 1 µg/mL puromycin.
## NanoBiT protein-protein interaction assays
HEK 293T cells were seeded into opaque white 96-well plates (3 × 10⁴ cells/well) in 100 µL DMEM + 10% FBS. Warm PBS was added to outer wells and between wells to reduce edge effects. The following day, cells were transfected with 100 ng total plasmid DNA per well (50 ng each of LgBiT and SmBiT constructs or SmBiT control) using FuGENE HD (Promega) in Opti-MEM I (Gibco). After 20-24 h incubation (20 h for TBK1 assays), culture medium was replaced with fresh Opti-MEM I ~30 min before reading. NanoGlo Live Cell Substrate (Promega) was diluted to 5× in Live Cell Buffer and added under low-light conditions. Luminescence was measured using a Luminoskan Ascent (Thermo) every 15 min for 90 min. Representative time points were selected based on signal strength and stability.
## Western blotting
Cells were lysed in ice-cold buffer (10 mM HEPES pH 7.5, 500 mM NaCl, 1 mM EDTA, 1 mM DTT, 0.5% Triton X-100) supplemented with PhosSTOP (Roche) and cOmplete protease inhibitors (Roche). Lysis proceeded on ice for 30-60 min with intermittent vortexing. Lysates were cleared by centrifugation at 12,000 × g for 10 min at 4°C. Protein concentration was determined by Bradford assay. Samples were mixed with Laemmli buffer, boiled for ≥2 min, flash frozen, and stored at -80°C. Proteins were resolved by SDS-PAGE and transferred to 0.45-µm PVDF membranes. Blots were blocked and incubated overnight at 4°C with primary antibodies diluted in 5% milk in PBST (0.1% Tween-20), followed by secondary antibodies for 1 h at room temperature. Signal was developed using SuperSignal West Pico PLUS (Thermo) and imaged on an Amersham Imager 600 or 680 (GE Healthcare). Primary antibodies: anti-GAPDH (1:5,000, CST 2118S), anti-TBK1 (1:1,000, CST 3504S), anti-phospho-TBK1 S172 (1:1,000, CST 5483S), anti-ALTO (1:10,000, courtesy of the Galloway Lab), anti-STING (1:1,000, CST 13,647S), anti-LgBiT (1:500, Promega N710A), anti-Xpress (1:1,000, Invitrogen 44-0528), and anti-Src (1:1,000, CST 2109S). Secondary antibodies: HRP-conjugated anti-rabbit IgG (1:2,500, CST 7074S) and anti-mouse IgG (1:2,500, CST 7076S).
## Immunofluorescent staining
HDF-inALTO and -inALTOdelLIT cells were seeded on coverslips and treated with Dox (0.5 µg/mL) for durations indicated in the figures. Cells were fixed in 3% paraformalde hyde (PBS, 20 min), permeabilized in 0.5% Triton X-100/3% BSA (1 h, RT), and incubated with primary antibodies in the same buffer for 1 h in humidified chambers. Following PBS washes, secondary antibodies with DAPI were applied under low-light conditions. Cells were imaged using an Olympus IX81 inverted microscope. Primary antibodies: anti-ALTO (1:1,000, courtesy of the Galloway Lab), anti-Rab5 (1:200, CST 3547T), and anti-GM130 (1:200, BD 610822). Secondary antibodies: AlexaFluor 594-conjugated goat anti-rabbit IgG (1:500, Invitrogen A11012) and AlexaFluor 488-conjugated goat anti-mouse IgG (1:500, Invitrogen A11029).
## MCPyV infection
Infections were performed using ALTO null MCPyV in HDF-inALTO and -inALTOdelLIT lines as previously described (29). Briefly, HDFs were seeded in serum-free infection medium, to which virus was added directly. After two days, Tet-free FBS was added to stimulate viral replication; the next day, Dox was added to a final concentration of 0.5 µg/mL to induce ALTO or ALTOdelLIT expression. Cellular and viral DNA were collected on day 5 post-infection, and viral replication was assessed by qPCR using primers specific to the non-coding regulatory region (NCRR) of the viral genome, normalized to genomic GAPDH, following our published methods (27,29). Oligonucleotides were obtained from Integrated DNA Technologies (IDT); primer sequences are provided in Table S2.
## In-Cell western blot
In-Cell Western assays were performed in 96-well plates using normal primary HDFs. Cells were seeded at 20,000 cells per well, which we determined fell within the linear range of TBK1 detection by our primary antibody. Cells were transfected in OptiMEM (Gibco) with 100 ng total DNA per well-85 ng pcDNA4c-LgBiT-LIT (or empty vector) plus 15 ng LgBiT-ALTO (or empty vector)-using TransIT-2020 (Mirus). At 20 h post-transfection, cells were washed with cold PBS, then fixed with cold 100% ethanol, and permeabilized in 0.2% Triton X-100 in PBS. Wells were then washed with cold PBS containing 0.05% Tween-20. Primary and secondary antibodies were diluted in LI-COR Intercept blocking buffer containing 0.05% Tween-20 in PBS. Cells were incubated with diluted primary antibody at room temperature for 1.5 h, then washed with cold PBS containing 0.05% Tween-20 and stained with secondary antibodies at room temperature with shaking for 1 h in darkness. Images were developed on a Licor Odyssey CLx. Fold change was calculated by first correcting for signal noise by subtracting the average signal measured in wells stained only with secondary antibody. Signal for TBK1 pS172 in each well was then compared with the signal for total TBK1 in that same well, and all values were compared with the average signal in WT ALTO-empty vector, which was set to 1. Primary antibodies: anti-TBK1 (1:200, ProteinTech 67211-1-Ig) and anti-phospho-TBK1/NAK (Ser172) D52C2 (1:100, CST 5483T). Secondary antibodies: goat anti-rabbit IgG IR800CW (1:100, LI-COR 926-32211) and AlexaFluor 680-conjugated goat anti-mouse IgG (1:100, Invitrogen A-21057).
## Statistical analysis
All statistical analyses were performed using GraphPad Prism. For single comparisons, unpaired two-tailed t-tests with Welch's correction were used. For multiple group comparisons, Brown-Forsythe and Welch's ANOVA tests were applied where indicated.
Significance thresholds were: P < 0.05 (*), P < 0.01 (**), P < 0.001 (***). Confidence intervals were set at 95%.
## References
1. Moens, Calvignac-Spencer, Lauber et al. (2017) "ICTV virus taxonomy profile: polyomaviridae" *J Gen Virol*
2. Fazeli, Sirat, Malekshahi (2025) "Novel human polyomaviruses discovered from 2007 to the present: an update of current knowledge" *Rev Med Virol*
3. (2025) *Full-Length Text Journal of Virology*
4. Feng, Shuda, Chang et al. (2008) "Clonal integration of a polyomavirus in human Merkel cell carcinoma" *Science*
5. Tolstov, Knauer, Chen et al. (2011) "Asymptomatic primary Merkel cell polyomavirus infection among adults" *Emerg Infect Dis*
6. Schowalter, Pastrana, Pumphrey et al. (2010) "Merkel cell polyomavirus and two previously unknown polyomaviruses are chronically shed from human skin" *Cell Host Microbe*
7. Tolstov, Pastrana, Feng et al. (2009) "Human merkel cell polyomavirus infection II. MCV is a common human infection that can be detected by conforma tional capsid epitope immunoassays" *Int J Cancer*
8. Foulongne, Kluger, Dereure et al. (2010) "Merkel cell polyomavirus in cutaneous swabs" *Emerg Infect Dis*
9. Bopp, Wieland, Hellmich et al. (2021) "Natural history of cutaneous human polyomavirus infection in healthy individuals" *Front Microbiol*
10. Chang, Moore (2012) "Merkel cell carcinoma: a virus-induced human cancer" *Annu Rev Pathol*
11. Busam, Jungbluth, Rekthman et al. (2009) "Merkel cell polyomavirus expression in merkel cell carcinomas and its absence in combined tumors and pulmonary neuroendocrine carcinomas" *Am J Surg Pathol*
12. Kassem, Schöpflin, Diaz et al. (2008) "Frequent detection of Merkel cell polyomavirus in human Merkel cell carcinomas and identification of a unique deletion in the VP1 gene" *Cancer Res*
13. Rodig, Cheng, Wardzala et al. (2012) "Improved detection suggests all Merkel cell carcinomas harbor Merkel polyomavi rus" *J Clin Invest*
14. Rapchak, Yagobian, Moore et al. (2022) "Merkel cell polyomavirus small T antigen is a viral transcription activator that is essential for viral genome maintenance" *PLoS Pathog*
15. Shuda, Feng, Kwun et al. (2008) "T antigen mutations are a human tumor-specific signature for Merkel cell polyomavirus" *Proc Natl Acad Sci*
16. Googins, An, Gauthier et al. (2025) "Polyomavirus large T antigens: unraveling a complex interactome" *Tumour Virus Res*
17. Engels, Frisch, Goedert et al. (2002) "Merkel cell carcinoma and HIV infection" *Lancet*
18. Wieland, Kreuter (2011) "Merkel cell polyomavirus infection and Merkel cell carcinoma in HIV-positive individuals" *Curr Opin Oncol*
19. Ma, Brewer (2014) "Merkel cell carcinoma in immunosuppressed patients" *Cancers (Basel)*
20. Burdette, Monroe, Sotelo-Troha et al. (2011) "STING is a direct innate immune sensor of cyclic di-GMP" *Nature*
21. Yum, Li, Fang et al. (2021) "TBK1 recruitment to STING activates both IRF3 and NF-κB that mediate immune defense against tumors and viral infections" *Proc Natl Acad Sci*
22. Shang, Zhang, Chen et al. (2019) "Cryo-EM structures of STING reveal its mechanism of activation by cyclic GMP-AMP" *Nature*
23. Jiang, Chen, Wang et al. (2020) "cGAS-STING, an important pathway in cancer immunotherapy" *J Hematol Oncol*
24. Hopfner, Hornung (2020) "Molecular mechanisms and cellular functions of cGAS-STING signalling" *Nat Rev Mol Cell Biol*
25. Abe, Harashima, Xia et al. (2013) "STING recognition of cytoplasmic DNA instigates cellular defense" *Mol Cell*
26. Ahn, Konno, Barber (2015) "Diverse roles of STING-dependent signaling on the development of cancer" *Oncogene*
27. Liu, Kim, Krump et al. (2020) "Selective reactivation of STING signaling to target Merkel cell carcinoma" *Proc Natl Acad Sci*
28. Krump, Wang, Liu et al. (2021) "Merkel cell polyomavirus infection induces an antiviral innate immune response in human dermal fibroblasts" *J Virol*
29. Wang, Yang, Senay et al. (2023) "Characterization of the impact of merkel cell polyomavirus-induced interferon signaling on viral infection" *J Virol*
30. Wang, Senay, Luo et al. (2024) "Merkel cell polyomavirus protein ALTO modulates TBK1 activity to support persistent infection" *PLoS Pathog*
31. Carter, Daugherty, Qi et al. (2013) "Identification of an overprinting gene in Merkel cell polyomavirus provides evolutionary insight into the birth of viral genes" *Proc Natl Acad Sci*
32. Peng, Abere, Shi et al. (2023) "Membrane-bound Merkel cell polyomavirus middle T protein constitutively activates PLCγ1 signaling through Src-family kinases" *Proc Natl Acad Sci*
33. Salisbury, Amonkar, Vinueza et al. (2024) "Polyomavirus ALTOs, but not MTs, downregulate viral early gene expression by activating the NF-κB pathway" *Proc Natl Acad Sci*
34. Lanclos, Radulovic, Bland et al. (2024) "Implications of intrinsic disorder and functional proteomics in the merkel cell polyomavirus life cycle" *J Cell Biochem*
35. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
36. Mirdita, Schütze, Moriwaki et al. (2022) "ColabFold: making protein folding accessible to all" *Nat Methods*
37. Dyson, Wright (2005) "Intrinsically unstructured proteins and their functions" *Nat Rev Mol Cell Biol*
38. Wright, Dyson (1999) "Intrinsically unstructured proteins: reassessing the protein structure-function paradigm" *J Mol Biol*
39. Cortese, Uversky, Dunker (2008) "Intrinsic disorder in scaffold proteins: getting more from less" *Prog Biophys Mol Biol*
40. Holehouse, Kragelund (2024) "The molecular basis for cellular function of intrinsically disordered protein regions" *Nat Rev Mol Cell Biol*
41. Lu, Salvino (2023) "The In-Cell Western immunofluorescence assay to monitor PROTAC mediated protein degradation" *Methods Enzymol*
42. Liu, Yang, Payne et al. (2016) "Identifying the target cells and mechanisms of merkel cell polyomavirus infection" *Cell Host Microbe*
43. (2025) *Full-Length Text Journal of Virology*
44. Lee, Russo, Pavletich (1998) "Structure of the retinoblastoma tumour-suppressor pocket domain bound to a peptide from HPV E7" *Nature*
45. Majerciak, Pripuzova, Chan et al. (2015) "Stability of structured Kaposi's sarcoma-associated herpesvirus ORF57 protein is regulated by protein phosphorylation and homodimerization" *J Virol*
46. Dyson (2023) "Vital for viruses: intrinsically disordered proteins" *J Mol Biol*
47. Mishra, Verma, Rao et al. (2020) "Intrinsically disordered proteins of viruses: Involvement in the mechanism of cell regulation and pathogenesis" *Prog Mol Biol Transl Sci*
48. Kast-Woelbern, Martinho, Julio et al. (2025) "The use of intrinsic disorder and phosphorylation by oncogenic viral proteins to dysregulate the host cell cycle through interaction with pRb" *Viruses*
49. Xiang, Song, Tang et al. (2021) "TANKbinding kinase 1 (TBK1): An emerging therapeutic target for drug discovery" *Drug Discov Today*
50. Sun, Revach, Anderson et al. (2023) "Targeting TBK1 to overcome resistance to cancer immunotherapy" *Nature*
51. Sun, Maggs, Panda et al. (2025) "TBK1 targeting is identified as a therapeutic strategy to enhance CAR T-cell efficacy using patient-derived organotypic tumor spheroids" *Cancer Immunol Res*
52. Liu, Krump, Buck et al. (2019) "Merkel cell polyomavirus infection and detection" *J Vis Exp*
53. Wang, Youkharibache, Zhang et al. (2020) "iCn3D, a web-based 3D viewer for sharing 1D/2D/3D representations of biomolecular structures"
54. Wang, Youkharibache, Marchler-Bauer et al. (2022) "D: from web-based 3D Viewer to structural analysis tool in batch mode" *Front Mol Biosci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12729479&blobtype=pdf | # Development of a Next-Generation Sequencing Protocol for Assessing Lenacapavir Resistance in HIV-1 Capsid
Omar Khalili, | Collins, Ambe Chenwi, Daniele Spalletta, Greta Marchegiani, Luca Carioti, | Hossein, Eizadi Moghadam, Ada Bertoli, Vincenzo Spagnuolo, Maria Santoro, Francesca Ceccherini-Silberstein, Maria Bellocchi, Collins Ambe, Maria Concetta, Antonella Castagna, Daniele Armenia, Stefano Bo- Nora, Leonardo Calza, Anna Cattelan, Giovanni Cenderello, Adriana Cervo, Laura Comi, Antonio Biagio, Emanuele Focà, Roberta Gagliardini, Andrea Giacomelli, Filippo Lagi, Giulia Marchetti, Stefano Rusconi, Francesco Saladini, Maurizio Zazzi
## Abstract
Lenacapavir (LEN) is a first-in-class capsid inhibitor (CAI) that targets multiple stages of the HIV-1 lifecycle, showing efficacy in heavily treatment-experienced (HTE) individuals with multidrug-resistance (MDR) and in pre-exposure prophylaxis (PrEP). This study aimed to characterize a novel in-house next-generation sequencing (NGS) protocol targeting HIV-1 capsid (CA) region using both HIV-1 RNA from plasma and HIV-1 DNA from peripheral-blood-mononuclear-cells (PBMCs). A total of 60 samples (41 HIV-1 RNA and 19 HIV-1 DNA) with various HIV-1 subtypes and viremia levels were tested. Overall, molecular amplification was successful in 83.3% of cases (75.6% for HIV-1 RNA and 100% for HIV-1 DNA), while high quality sequences were obtained in 76.7% of samples (65.9% for HIV-1 RNA and 100% for HIV-1 DNA). Among RNA samples with viremia ≥ 500 copies/mL, sequencing success reached 92.6%, showing a statistically significant association with viral load. Subtype-specific analysis showed amplification and sequencing rates of 86.0% and 79.1% for subtype B, and 76.5% and 70.6% for non-B subtypes, with no significant difference. Reproducibility was fully confirmed by pairwise similarity analyses at 10% and 20% frequency cutoff, upon reprocessing 13 HIV-1 RNA samples. This protocol provides an important tool, primarily for subtype B, for personalized HIV-1 treatment with CAI-based strategies, enabling efficient characterization of LEN resistance mutations in the CA region, using both DNA and RNA samples.
## 1 | Introduction
Despite the progress made in improving the efficacy and safety of antiretroviral therapy (ART), HIV-1 cure is not feasible and therefore continues to be a major global public health issue [1]. HIV-1 infection remains a serious concern in a proportion of heavily treatment experienced (HTE) people with HIV-1 (PWH) who harbor a multidrug-resistance (MDR) virus, often as a result of exposure to suboptimal treatment and multiple treatment failures [2]. For this fragile population ART options can also be limited.
In this context, strategies for prevention of new infections and treatment are fundamental for reducing the individual and societal burden of HIV-1. In this light, a continuous development of new agents active against resistant variants of HIV-1 and targeting novel mechanisms of action is required, in order to provide simpler and more efficacious treatment options to all PWH, irrespective of their prior treatment history. Among new antiretrovirals, lenacapavir (LEN) is a novel injectable first-inclass HIV-1 capsid inhibitor (CAI). It inhibits selectively multiple stages of capsid function, by directly binding to the interface between the capsid protein subunits, preventing the nuclear import of proviral DNA, hindering virus assembly and release by interfering with the function of Gag/Gag-pol genes, and thereby generating malformed capsids [3]. LEN use is indicated, in combination with other antiretrovirals, for the treatment of HTE individuals with MDR and has recently shown high efficacy for prevention (more than 99%) when used as pre-exposure prophylaxis (PrEP) in diverse populations (cisgender women, cisgender gay, bisexual, and other men, transgender women, transgender men, and gender-nonbinary persons) [4][5][6].
Integrating into clinical routine new pharmacological targets, such as the capsid, is crucial for optimizing treatment strategies and improving clinical outcomes. Due to the novelty of this target, a method for characterizing the resistance associated mutations in the capsid is mandatory. Margot et al. in a recent study used the Sanger sequencing method to evaluate the capsid resistance mutations in 27 HIV-1 RNA plasma samples [7]. However, there are limitations of Sanger sequencing in detecting low-frequency variants and providing adequate resistance profiles [8]. This has led to the need for the implementation of the genotypic test for the capsid, by using next-generation sequencing (NGS) technologies, which offers greater sensitivity and accuracy in identifying resistance-associated mutations. The advent of NGS has redefined genome sequencing techniques. This technology sequences millions of fragments simultaneously per run, enabling the detection of minority quasispecies with mutations occurring at frequencies below 20%, with an accuracy greater than 99%. Evidence suggests that some mutations present in HIV-1 resistant minority variants may be clinically relevant [9,10].
The aim of this study was to develop a novel NGS protocol to sequence HIV-1 capsid region for assessing resistance to LEN by using both HIV-1 RNA from plasma samples and HIV-1 DNA from peripheral blood mononuclear cells (PBMCs) matrices, with diverse subtypes and viremia levels.
## 2 | Materials and Methods
## 2.1 | Clinical Samples
For the protocol development, residual anonymized specimens were used, originating from routine clinical practice and/or research activities within two cohorts of PWH living in Italy: the PRESTIGIO Registry (https://trials-ice2.advicepharma.com/ PRESTIGIO/) and the ICONA Foundation (https://www. fondazioneicona.org/). Ethic Committee approval was deemed unnecessary under Italian law for residual anonymized samples used for diagnostic purposes since this was not considered a clinical trial of medicinal products for clinical use (Art. 6 and Art. 9, Law Decree 211/2003). For residual anonymized samples obtained from research activities, approval was obtained by the Ethic Committee of each participating center involved in the above-mentioned Italian cohorts.
## 2.2 | HIV-1 Extraction and Amplification
Viral RNA was extracted from 1 mL of plasma using the QIAamp Viral RNA Mini Kit (Qiagen GmbH, Hilden, Germany), after ultracentrifugation at 23,000 g for 2 h at 4°C, following the manufacturer's instructions, while DNA was extracted from a pellet of 5 × 10⁶ PBMCs using the High Pure PCR Template Preparation Kit (Roche Diagnostics, Basel, Switzerland).
HIV-1 RNA from plasma and HIV-1 DNA from PBMCs were subjected to reverse transcription/amplification and amplification, respectively, using the SuperScript III One-Step RT-PCR system for long templates (Invitrogen, Carlsbad, CA, USA). In detail, each 50 µL reaction contained: 10 µL of extracted viral genome, 25 µL of 2× reaction mix, 8 µL of MgSO₄ (5 mmol/L), 3 µL of DNase/RNase-free water, 0.75 µL each of forward and reverse primers (10 pmol/µL), 1 µL of RNase Out (40 U/µL; replaced with 1 µL of water for HIV-1 DNA samples) and 1.5 µL of RT/Taq enzyme. Thermal cycler conditions were 50°C for 30 min, 94°C for 2 min, followed by 45 cycles of 95°C for 30 s, 53°C for 30 s, and 72°C for 2 min; final extension at 72°C for 10 min. When no visible amplification product was detected after agarose gel electrophoresis of the first-round PCR, a nested PCR was performed. The 50 µL reaction mix contained: 5 µL of firstround PCR product, 33 µL of DNase/RNase-free water, 5 µL of PCR buffer (10×), 3.5 µL of MgCl 2 (25 mM), 1 µL of dNTPs (10 mmol/L), 0.9 µL each of forward and reverse primers (10 pmol/µL) and 0.7 µL of AmpliTaq Gold DNA polymerase (Life Technologies, Carlsbad CA, USA). Thermal cycler conditions were 93°C for 12 min, 40 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 2 min, with a final extension at 72°C for 10 min. Amplicons were analyzed by agarose gel electrophoresis to confirm band sizes (See Figure 1). Gag specific primers and thermal profiles were adapted from Soria et al. [11] (See Table 1).
## 2.3 | HIV-1 Next Generation Sequencing
Each amplified sample was purified using (0.8× ratio) Ampure XP Beads (Beckman Coulter, Pasadena, CA, USA), then quantified using Qubit dsDNA HS Assay Kit (Invitrogen, Carlsbad, CA, USA) with Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). For each sample, 1 ng of amplicon was involved in a tagmentation reaction by Nextera XT DNA Library Kit (Illumina Inc., San Diego, CA, USA) and uniquely indexed with Nextera XT Index Kit v2 (Illumina Inc., San Diego, CA, USA) following the manufacturer's instructions. After a second purification (0.6x ratio) and second quantification, the libraries were diluted at 4 nM and pooled. Finally, 15 pM of the denatured pool was sequenced paired-end with MiSeq Reagent Kits v2 (2 × 250) (Illumina Inc., San Diego, CA, USA) with 6%-10% of PhiX Control V3 library to monitor sequencing quality [12].
## 2.4 | Bioinformatics Analyses, Mutational Pattern and Subtype Assignment
NGS data obtained as FASTQ files were analyzed using HIVdb Stanford algorithm (version 9.8; https://hivdb.stanford.edu/) to determine the mutational pattern of each sample. Sequences were considered valid with at least 100 coverage reads per position, based on commonly used parameters in the literature [13][14][15]. The consensus sequences for all samples were generated with prevalence cutoffs of 5%, 10%, and 20% mutation detection threshold (MDT) as defined by the Stanford HIDVB website (https://hivdb.stanford.edu/hivdb-capsid/byreads/). The HIV-1 subtype was assessed using two automated tools, COMET (https://comet.lih.lu/) and Stanford (https://hivdb. stanford.edu/) and the results for each sample were compared with those obtained from a previous subtyping analysis on protease/reverse transcriptase. In addition, a quality control of the raw data obtained in the FASTQ format was performed with Trimmomatic [16] software in order to remove adapters, PCR primers and poor-quality reads. FASTQ files were analyzed with VirVarSeq software version 1 [17] using K03455.1 as reference sequence. This biostatistical analysis was performed using a higher coverage threshold of 1,000, to increase the robustness of variant detection for each sample; a consensus sequence with prevalence cutoff of 5%, 10% and 20% was generated using quasitools (https://github.com/phac-nml/quasitools).
## 2.5 | Efficiency Assessment
The efficiency assessment of the protocol was conducted, considering the main variables that could influence the success rate: the different compartment (HIV-1 RNA from plasma and HIV-1 DNA from PBMCs), the HIV-1 subtype (B vs. non-B), and viremia levels for the HIV-1 RNA samples. In particular, viremia levels were stratified according to the following strata: ≥ 50-500; ≥ 500-1000; ≥ 1000-10,000; ≥ 10,000 copies/mL. The efficiency of the two main steps of the process (molecular amplification and sequencing on the Miseq platform) were considered separately.
## 2.6 | Reproducibility Assessment
For both precision and reproducibility testing, the variability in the result is determined based on nucleotide sequence similarity by comparison with the consensus sequences derived from the replicates (at 5%, 10%, and 20% cutoffs, as described before.) Briefly, sequences were aligned with the HXB2 reference for Gag and CA regions using MAFFT v7.475. The CA region (positions 397-1089 of Gag or 1186-1878 relative to HXB2, resulting in 693 nucleotides and 231 residues) was manually extracted with BioEdit v7.7. Pairwise sequence similarity for each sample across two runs was assessed using the Needleman-Wunsch algorithm implemented in EMBOSS Needle [18]. Acceptance criteria was more than 90% of pairwise comparisons with at least 98% identical (with non-matching mixtures counted as a difference) [19].
## 2.7 | Statistical Analysis
Descriptive statistics were expressed as median values and the interquartile range (IQR) for continuous variables and the number (percentage) for categorical variables. To evaluate the impact of HIV-1 subtype (B vs. non-B) and viral load stratification (< 500 vs. ≥ 500 copies/mL) on amplification and sequencing success, the Fisher exact test was applied for categorical variables. All statistical analyses were performed using the SPSS software package for Windows (version 23.0, SPSS Inc., Chicago, IL). A two-sided p-value of less than 0.05 was considered statistically significant.
## 3 | Results
## 3.1 | Sample Characteristics
Sixty samples were analyzed, of which 41 (68.3%) were HIV-1 RNA, obtained from plasma samples, and 19 (31.7%) were ,070] HIV-1 RNA copies/ mL. Among plasma samples, 18 (43.9%) had HIV-1 RNA levels in the range ≥ 50-1000 copies/mL (≥ 50-500 copies/mL: 14 samples; ≥ 500-1000 copies/mL: 4 samples), 13 (31.7%) had HIV-1 RNA levels in the range ≥ 1000-10,000 copies/mL, and 10 (24.4%) presented HIV-1 RNA levels ≥ 10,000 copies/mL (See Table 2). All HIV-1 DNA samples were collected from 10 individuals receiving LEN-based antiretroviral therapy at different time points. Seven PBMC samples were obtained from viremic individuals (with a median HIV-RNA of 23,554 [1447-32,200] copies/ml), while 12 were collected during virological suppression (plasma HIV-RNA ≤ 50 copies/ml).
Overall, HIV-1 subtype B was detected in 43 samples (71.7%), of which 25 were HIV-1 RNA samples and 18 were HIV-1 DNA samples. Non-B subtypes were identified in the remaining 17 samples (28.3%), predominantly among HIV-1 RNA (n = 16) and in only one HIV-1 DNA sample (see Table 2). The non-B subtypes included 4 samples with CRF02_AG (6.7%), 3 with subtype F1 (5.0%), 2 with the DF recombinant form (3.3%, one in the HIV-1 DNA sample), 2 with subtype C (3.3%), and 1 with subtype G (1.7%). The remaining 5 samples (8.3%) harbored other recombinant forms: CRF09_cpx (n = 1), CRF12_BF (n = 1), CRF41_CD (n = 1), CRF42_BF (n = 1), and CRF60_BC (n = 1).
## 3.2 | Efficiency of the Developed NGS Protocol
Overall, high-quality CA region sequences were obtained from 46 out of 60 samples, corresponding to a global sequencing success rate of 76.7%, with 1600 minimum reads per position and 100% gag coverage found. The main factors potentially affecting the success rate of the developed NGS protocol were carefully evaluated. These included the viral genome specimen (HIV-1 RNA vs. HIV-1 DNA), the HIV-1 subtype, and the level of viremia in HIV-1 RNA samples.
The analysis was performed by evaluating amplification efficiency of the CA region, sequencing efficiency among successfully amplified samples, and the overall sequencing success rate across all samples. The results are summarized in Table 3.
## 3.2.1 | Efficiency by Sample Compartment: HIV-1 RNA and HIV-1 DNA Samples
The first variable evaluated to assess the efficiency of the developed NGS protocol was the viral genome compartment, considering different sample types: HIV-1 RNA from plasma sample and HIV-1 DNA extracted from PBMCs.
Considering the compartment, successful sequences were obtained from 27 of 41 HIV-1 RNA samples (65.9%) and in all 19 HIV-1 DNA samples (100%) (see Table 3).
In terms of amplification efficiency, the CA region was successfully amplified in 31 of 41 HIV-1 RNA samples (75.6%), and in all 19 HIV-1 DNA samples (100%). Among the 31 amplified RNA samples, 27 (87.1%) yielded analyzable sequences, while 4 (12.9%) failed sequencing due to low coverage (< 100 reads).
## 3.2.2 | Efficiency by HIV-1 Subtype-Associated Variability: B Subtype versus Non-B Subtypes
The second variable analyzed was the impact of HIV-
## 3.2.3 | Efficiency by Viremia Levels of HIV-1 RNA Samples
Efficiency was additionally assessed by categorizing the 41 HIV-1 RNA plasma samples according to viral load ranges.
Amplification and sequencing success were both 100% (10/10) in the ≥ 10,000 copies/mL range and 92.3% (12/13) in the ≥ 1000-10,000 copies/mL range. To further investigate performance at lower viral loads, a sub-analysis focusing on samples with viral loads below 1000 copies/mL was conducted. In the ≥ 500-1000 copies/mL range, both amplification and sequencing success were 75.0% (3/4). In the ≥ 50-500 copies/ mL range, amplification success decreased to 42.9% (6/14), and sequencing success was 14.3% (2/14), (see Table 3).
Notably, samples with viral loads above 500 copies/mL showed high reliability, with 25 out of 27 (92.6%) successfully amplified and sequenced.
Statistical analysis confirmed a significant association between viral load and both amplification and sequencing success (Fisher's exact test, both p < 0.001), as shown in Figure 3.
## 3.3 | Evaluation of NGS Reproducibility
To evaluate the reproducibility of the developed protocol, 13 representative HIV-1 RNA samples with successful amplification and analyzable sequences were reprocessed, starting from the same extraction sample. All reprocessed samples showed consistent performance, with successful amplification and sequencing achieved in every case. Using the pairwise similarity sequence test with EMBOSS Needle, nucleotide sequences obtained from the two processes were compared for each sample at NGS cut-offs of 5%, 10%, and 20% (see Table 4).
All samples demonstrated very high similarity values. Reproducibility acceptance criteria were met at all NGS analysis cut-offs, excluding the most stringent 5% threshold. The proportion of samples with at least 98% pairwise similarity was 100% (13/13) at both the 20% and 10% cut-offs. The lowest pairwise similarity observed was 96.7% at the 5% cut-off. The specific amino acid changes identified at the three NGS cut-offs (5%, 10%, and 20%) for each of the 13 sample pairs are detailed in Supplementary Table S1.
## 3.4 | LEN Resistance Profile
Overall, within the 46 CA sequences obtained, no major mutations associated with resistance to LEN were detected, except for two sequences derived from a single individual with a virological failure during LEN-based therapy. In this case, the resistance-associated mutation K70H (97.8%) and Q67K (98.5%) were identified in the HIV-1 RNA sequence, and were also confirmed in the corresponding proviral HIV-1 DNA at frequencies of 75.6% and 75.7%, respectively. This was the only individual among those who had both HIV-1 RNA and HIV-1 DNA samples who harbored LEN-resistance mutations.
Furthermore, the accessory mutation T107A was observed in three sequences from three distinct individuals, with variant frequencies of 98.2% and 66.9% in two HIV-1 RNA sequences, and 5.9% in a HIV-1 DNA sequence, respectively.
Finally, the accessory mutation A105T, at a frequency of 6.0%, was detected in only one sequence of the replicated HIV-1 RNA samples used to assess protocol reproducibility at 5% cutoff (see Table S1).
## 4 | Discussion
In this study, a novel and highly effective NGS protocol was developed for sequencing the HIV-1 CA region to assess resistance to LEN, a recently approved CA inhibitor for antiviral treatment, with significant potential applicability to both current and future antiretroviral drugs targeting this protein. The previously published protocol based on Sanger sequencing (Soria et al., 2016) was modified and optimized for the Illumina NGS platform to enable efficient amplification and deep sequencing of the HIV-1 CA region across various sample matrices (HIV-1 DNA and HIV-1 RNA), different subtypes, and viremia levels. The method is versatile and robust and can be implemented in both diagnostic and research settings, particularly for assessing LEN resistance in HTE PWH with MDR, either at baseline before initiating LEN therapy or in cases of virological failure [20]. To date, the novelty of this protocol lies in its ability to detect low- frequency variants through the implementation of an NGS-based approach, and to efficiently analyze both HIV-1 RNA samples and HIV-1 DNA derived from PBMCs.
The protocol achieved a 100% success rate with HIV-1 DNA samples and about 66% with HIV-1 RNA samples. Amplification failures were predominantly observed in HIV-1 RNA samples with low-level viremia, particularly within the ≥ 50-500 copies/mL range, where a marked reduction in both amplification and sequencing success rates was observed. Accordingly, to better characterize performance in this critical range, a sub-analysis was conducted on samples with viral loads below 1000 copies/mL. Amplification and sequencing success were substantially decreased in the ≥ 50-500 copies/mL range, with only 2 out of 14 samples (14.3%) successfully amplified and sequenced. Conversely, performance improved considerably in the ≥ 500-1000 copies/mL range, where 3 out of 4 samples (75.0%) yielded successful results. Notably, a statistically significant association was found between viral load ≥ 500 copies/ mL and sequencing success, with 92.6% of these samples successfully amplified and sequenced. This highlights the utility of the protocol, especially considering that most commercial kits are not validated for viral loads below 1,000 copies/mL.
The higher success rate observed with HIV-1 DNA samples can likely be attributed to the type of starting material. In the study, proviral DNA was extracted from PBMCs, resulting in higher nucleic acid yield and purity compared to whole blood. This excellent performance will enable genotypic resistance testing (GRT) in the HIV-1 DNA context, for individuals with undetectable or with low-level viremia.
Regarding the subtype issue, similar success rates were observed for both HIV-1 B and non-B subtypes (79.1% and 70.6%, respectively), with no statistically significant difference.
In the study population, this was likely due to the consistent binding efficiency of the primers within the viral gag region. Amplification success appeared to be influenced more by other factors, particularly low-level viremia (< 500 copies/mL), which substantially reduced amplification efficiency.
Concerning the resistance, the population included also samples from individuals treated with LEN. One of them, who experienced a virologic failure, showed the major resistance mutation K70H with the Q67K at high prevalence (> 75%), in both HIV-1 RNA and HIV-1 DNA matrices. Furthermore, the accessory mutation T107A was identified in three distinct individuals: two in HIV-1 RNA samples (at 98.2% and 66.9% prevalence, respectively) and one in a HIV-1 DNA sample (5.9% prevalence). These results show that the protocol can effectively detect resistance-associated mutations, highlighting its relevance for monitoring the effectiveness of LEN and other future CA inhibitors. Notably, even low-prevalence mutations were identified, although their clinical significance remains to be established.
No recurrent mutations were observed in specific regions of the capsid across the study population. The distribution of differences appeared random, with no indication of sequencing bias. The observed mutational patterns were largely attributable to subtype variability relative to the reference sequence used for alignment.
Finally, to evaluate the reproducibility and sensitivity of the protocol, 13 representative HIV-1 RNA samples with different subtypes and viral loads were reprocessed. The amplification and sequencing results were confirmed. Using pairwise similarity comparison, all sequences at 10% and 20% frequency cutoff achieved the 100% criteria of reproducibility (≥ 98% similarity) [18]. While 76.9% reproducibility was achieved at the 5% threshold (with 10 out of 13 pairs showing ≥ 98% similarity). This result is likely due to technical limitations, and is in line with other NGS protocols, since a 10% frequency threshold is currently recommended for NGS-based GRT analyses [15].
A limitation of this study is the relatively small sample size (60 samples), particularly regarding the distribution of non-B subtypes in HIV-1 DNA samples. Although 28.3% of the study population carried non-B subtypes, only one HIV-1 DNA sample with a non-B subtype was available for analysis. This could limit the interpretation of results obtained from HIV-1 DNA samples, particularly with respect to non-B subtypes. Overall, no statistically significant differences were observed between B and non-B subtypes in terms of amplification or sequencing success, suggesting that subtype variability did not substantially impact assay performance. However, the small number of non-B samples may have reduced the statistical power of the analysis. Therefore, further studies with larger and more balanced subtype representation are warranted. Indeed, the subtypes analyzed were limited to the most common ones circulating in Italy [21]. Even if, non-B subtypes are increasing among newly diagnosed cases in Italy, LEN is primarily used in patients with long treatment histories and multidrug resistance, the majority of whom are infected with subtype B. Therefore, this protocol is particularly applicable to this setting and other European settings with similar HIV-1 molecular epidemiology.
It should be highlighted that all the amplification failures in samples with non-B subtypes occurred with HIV-1 RNA levels below 500 copies/mL. Additional data from non-B subtype samples with higher viral loads would be useful to further assess protocol performance. Having access to a larger number of non-B samples, especially those that failed amplification with low viremia (such as CRF02_AG or subtype C), would have the opportunity to better investigate whether the observed failures are attributable to low viral load, or if subtype-specific factors may also play a role. In cases of low-level viremia, increasing the input amount of viral RNA or modifying PCR cycling parameters may improve the performance.
Finally, regarding the HIV-1 DNA analysis, the use of PBMCs instead of whole blood could potentially limit the adoption of the protocol in diagnostic routine, as Ficoll separation involves additional steps, increased costs, and longer processing times.
## 5 | Conclusions
The study presents a novel highly effective NGS protocol for analyzing the HIV-1 CA region, viral target of LEN and other future CA inhibitors. The protocol achieved high success rates for both HIV-1 DNA and RNA samples with viremia levels ≥ 500 copies/ml. High amplification and sequencing success rates were achieved in HIV-1 subtype B samples, with similarly favorable outcomes observed in non-B subtypes, despite the limited number of cases. The overall reproducibility and robustness of the protocol remained very high, supporting its reliability and implementation in both clinical and research
## References
1. Team, Bank, Galli et al. "CATANIA: Bruno Cacopardo, Maurizio Celesia, Michele Salvatore Paternò Raddusa, Carmen Giarratana"
2. "Joint United Nations Programme on HIV/AIDS (UNAIDS). 2024. The Urgency of Now: AIDS at a Crossroads. Geneva: Joint United Nations Programme on HIV/AIDS; 2024. Licence: CC BY-NC-SA 3.0 IGO"
3. Brizzi, Cable, Patel et al. (2024) "Heavily Treatment-Experienced Patients with HIV: Are New Mechanisms of Action Enough?" *Journal of International Medical Research*
4. Dvory-Sobol, Shaik, Callebaut et al. (2022) "Lenacapavir: A First-in-Class HIV-1 Capsid Inhibitor"
5. Vardanega, New, Mezzio et al. (2024) "US Cost-Utility Model of Lenacapavir Plus Optimized Background Regimen (OBR) vs Fostemsavir Plus OBR and Ibalizumab Plus OBR for People with HIV with Multidrug Resistance" *Journal of Managed Care & Specialty Pharmacy*
6. Bekker, Das, Karim (2024) "Twice-Yearly Lenacapavir or Daily F/TAF for HIV Prevention in Cisgender Women" *New England Journal of Medicine*
7. Kelley, Acevedo-Quiñones, Agwu (2025) "Twice-Yearly Lenacapavir for HIV Prevention in Men and Gender-Diverse Persons" *New England Journal of Medicine*
8. Margot, Naik, Nekkalapudi et al. (2023) "Rapid HIV-1 Genotyping Assay for the Detection of Capsid Mutations" *Journal of Medical Virology*
9. Li, Yu, Song (2022) "HIV-1 Genotypic Resistance Testing Using Sanger and Next-Generation Sequencing in Adults With Low-Level Viremia in China" *Infection and Drug Resistance*
10. Armenia, Santoro, Bellocchi (2022) "Viral Resistance Burden and APOBEC Editing Correlate with Virological Response in Heavily Treatment-Experienced People Living with Multi-Drug Resistant HIV" *International Journal of Antimicrobial Agents*
11. Stella-Ascariz, Arribas, Paredes et al. (2017) "The Role of HIV-1 Drug-Resistant Minority Variants in Treatment Failure" *Journal of Infectious Diseases*
12. Soria, Alteri, Scarlatti (2016) "Occupational HIV Infection in a Research Laboratory with Unknown Mode of Transmission: A Case Report" *Clinical Infectious Diseases*
13. Bellocchi, Scutari, Carioti (2023) "Frequency of Atypical Mutations in the Spike Glycoprotein in SARS-CoV-2 Circulating From July 2020 to July 2022 in Central Italy: A Refined Analysis by Next Generation Sequencing" *Viruses*
14. Biba, Fiaschi, Varasi (2024) "A Comparison of Sanger Sequencing and Amplicon-Based Next Generation Sequencing Approaches for the Detection of HIV-1 Drug Resistance Mutations" *Viruses*
15. Taylor, Lee, Nykoluk (2019) "A MiSeq-HyDRA Platform for Enhanced HIV Drug Resistance Genotyping and Surveillance" *Scientific Reports*
16. Armenia, Carioti, Micheli (2024) "Comparison of Different HIV-1 Resistance Interpretation Tools for Next-Generation Sequencing in Italy" *Viruses*
17. Bolger, Lohse, Usadel (2014) "Trimmomatic: A Flexible Trimmer for Illumina Sequence Data" *Bioinformatics*
18. Verbist, Thys, Reumers (2015) "VirVarSeq: A Low-Frequency Virus Variant Detection Pipeline for Illumina Sequencing Using Adaptive Base-Calling Accuracy Filtering" *Bioinformatics*
19. Madeira, Madhusoodanan, Lee (2024) "The EMBL-EBI Job Dispatcher Sequence Analysis Tools Framework in 2024" *Nucleic Acids Research*
20. (2020) "WHO HIVResNet HIV Drug Resistance Laboratory Operational Framework"
21. Nka, Bouba, Teto (2023) "Evaluation of HIV-1 Capsid Genetic Variability and Lenacapavir (GS-6207) Drug Resistance-Associated Mutations According to Viral Clades Among Drug-Naive Individuals" *Journal of Antimicrobial Chemotherapy*
22. Fabeni, Armenia, Abbate (2024) "HIV-1 Transmitted Drug Resistance in Newly Diagnosed Individuals in Italy Over the Period 2015-21" *Journal of Antimicrobial Chemotherapy* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724256&blobtype=pdf | # Phospho-activated non-muscle myosin IIA heavy chain supports different mechanisms of virus uptake
Natalia Ansin, Kathia Guardado, Matías Ponce, María García, Camila Ladra, Irene Ferreiro, Gonzalo Moratorio, Nicolás Sarute
## Abstract
Limiting virus entry is crucial to prevent infection, since this step of the virus life cycle is a major determinant of cellular tropism, host range, and pathogenesis. We recently showed that the transmembrane protein signal regulatory protein alpha (SIRPA), a negative regulator of F-actin-dependent phagocytosis, limits endocytosis of enveloped RNA viruses from unrelated families. Mechanistically, surface engagement of SIRPA recruits the SH2 domain-containing protein tyrosine phosphatases (SHP)-1 and -2, initiating a signaling cascade that dephosphorylates different pro-phagocytic proteins, including the heavy chain (MYH9) of the molecular motor non-muscle myosin IIA (NM-IIA), to ultimately suppress phagocytosis. Here, we analyzed whether MYH9 would enhance viral infection, as it does phagocytic engulfment, and how its activity in infection is regulated. We found that MYH9 expression increases the infection of viruses exploiting different endocytic pathways/mechanisms to enter cells, including flavivirus, arenavirus, rhabdovirus, and togavirus, in human and mouse cells. Furthermore, we demonstrated that MYH9 is actively translocated to the plasma membrane upon viral infection to support post-binding steps of entry, and that phosphorylation of two key tyrosine residues in its head and tail domains is essential for its function in viral infection. By using small molecule inhibitors and MYH9 knockdown cells, we suggest that members of the Src family kinases may phosphorylate/activate MYH9 to support virus entry.IMPORTANCE Viral infections represent a major threat to global public health and pose a huge social and economic cost; thus, a better understanding of how cell-intrin sic factors modulate the outcome of infection is of great importance to better under stand virus-host interactions and to our ability to develop prophylactic measures and therapeutics. In this work, we show that MYH9 is a broad proviral host factor that enhances entry of divergent families of human pathogenic RNA viruses, which exploit different endocytic pathways to infect cells. By determining that virus infection triggers the phosphorylation of MYH9 in two tyrosine residues essential for its proviral activity, and the family of non-receptor tyrosine kinases involved in this process, we may also provide new cellular targets to develop antiviral therapies. KEYWORDS virus entry, MYH9, proviral factor, tyrosine phosphorylation V iruses can hijack a variety of cellular processes and molecules to promote entry and replication in host cells (1). Although virus entry is a major determinant of cellular tropism, host range, and pathogenesis, there are relatively few host factors acting at this step whose activity has been characterized (2). We previously uncovered that signal regulatory protein alpha (SIRPA) is an intrinsic antiviral factor that limits entry of pathogenic enveloped RNA viruses from divergent families (3). SIRPA is a receptor-type transmembrane glycoprotein highly expressed on myeloid cells, harboring two immunoreceptor tyrosine-based inhibition motifs (ITIM) in its cytoplasmic domain
. Phosphorylation of the SIRPA ITIMs provides docking sites for the recruitment and activation of the cytosolic SH2 domain-containing protein tyrosine phosphatases (SHP)-1 and -2, initiating a negative signaling cascade that ultimately inhibits F-actindependent phagocytosis and immune signaling (4,5). The phosphorylation levels of SIRPA are greatly enhanced by the engagement of the membrane protein CD47 in trans, which leads to the recruitment of the SHP phosphatases and the subsequent dephosphorylation of several prophagocytic proteins, including the heavy chain (MYH9) of non-muscle myosin IIA (NM-IIA) (6)(7)(8).
NM-IIA is a hexameric protein with motor and contractile properties involved in cell migration, adhesion, phagocytosis, and other cellular processes, which consists of homodimers of MYH9, regulatory light chains (RLC), and essential light chains (9). Each MYH9 monomer comprises an N-terminal motor domain, which regulates the production of mechanical force through magnesium-dependent ATP hydrolysis (ATPase activity) and binding to actin filaments, and a C-terminal cargo-binding tail domain (9,10). Phosphor ylation of the RLCs regulates the assembly of NM-IIA into bipolar filaments, which is essential to engage and contract the actin cytoskeleton (11,12). It was shown that SIRPA limits viral endocytosis by a mechanism that resembles its inhibitory activity on phagocytosis (3), and that F-actin-mediated phagocytosis is mostly driven by phosphoactivated MYH9, since its pharmacological inhibition prevents particle engulfment to a similar extent as the negative signaling initiated by the SIRPA-CD47 interaction (8,13). Therefore, we speculated that other phagocytic proteins could also have a role in virus infection.
Here, we investigated whether MYH9 enhances viral infection and how its activity would be regulated during this process. We found that MYH9 increases infection levels of unrelated enveloped viruses, including flavivirus, arenavirus, rhabdovirus, and togavi rus, which use different mechanisms of virus entry, and that its ATPase activity and self-oligomerization are essential for its proviral activity. Furthermore, we showed that virus infection drives the accumulation of MYH9 at the plasma membrane to support post-binding steps of virus entry and that phosphorylation of the tyrosine residues 277 and 1805 is critical for its function in infection. By analyzing different families of non-receptor tyrosine kinases (NRTKs), we identified that members of the Src family kinase would phosphorylate MYH9 upon viral infection, which could be therapeutically exploited.
## RESULTS
## MYH9 enhances viral infection in human and mouse cells
Given the significant role of MYH9 in distinct cellular processes that require the generation of mechanical force, MYH9 knockout cells are mostly non-viable and Myh9 depletion is embryonic lethal in mice (9,14). Thus, to analyze the role of MYH9 in viral infection, we did short-interfering RNA (siRNA)-mediated knockdowns in human and mouse cells, which transiently reduced MYH9 mRNA and protein expression (Fig. 1A), without compromising cell viability 48 h post-transfection (Fig. 1B). siMYH9-transfected cells were infected with replicative-competent viruses at a multiplicity of infection (MOI) of 1, and viral replication levels were measured at 24-48 h post-infection (hpi) by quantitative polymerase-chain reaction (qPCR). MYH9 knockdown resulted in signifi cantly lower infection levels of the New World arenavirus (NWA) Tacaribe (TCRV), the Old World arenavirus (OWA) Lymphocytic Choriomeningitis virus (LCMV), the flavivirus Zika (ZIKV), the rhabdovirus vesicular stomatitis virus (VSV), and the togavirus Mayaro (MAYV) (Fig. 1C), while infection levels of the retrovirus murine leukemia virus (MLV) were not reduced upon siMYH9 transfection with respect to a siRNA control (siCTRL) (Fig. 1D). As a control for this experiment, we used herpes simplex virus type-1 (HSV-1), which was first described to use MYH9 as a cellular receptor (15), observing lower HSV-1 infection levels upon MYH9 knockdown as expected (Fig. 1E). To analyze Myh9 activity in viral infection, we isolated bone marrow-derived macrophages (BMDMs) from 8-to 12-week-old C57BL/6 mice, and we transfected a mousespecific siRNA (simMYH9) in fully differentiated BMDMs for 48 h (Fig. 1A). Next, we infected simMYH9-transfected cells with TCRV and LCMV at an MOI = 1 for 48 h, which were shown to infect primary mouse cells at high levels (16), and we found that reduced expression of Myh9 also resulted in a significant decrease in viral infection levels for both viruses (Fig. 1F), indicating that MYH9 proviral activity is conserved in both human and mouse cells.
We next asked whether viral infection could upmodulate the expression levels of MYH9 to further support infection. To this end, we quantified MYH9 mRNA and endogenous protein expression in A549 cells after viral infection. We observed that MYH9 transcript levels were increased at 8 hpi, while at 24 hpi, the expression levels were comparable to those observed for MOCK-infected cells (Fig. 1G). Western blot analysis showed that MYH9 levels were significantly increased at 24 hpi (Fig. 1H), suggesting that virus infection establishes a positive feedback-like mechanism to further increase infection levels. (H) MYH9 expression was analyzed by western blot in A549 cells infected with TCRV for 24 h using a rabbit anti-MYH9 (CST) and relative MYH9 expression levels are depicted above each lane. Shown is the average ± SD of three independent experiments. An unpaired t test was used to determine significance. *, P ≤ 0.04. A mouse anti-Tubulin (Thermo) was used as a control.
## The ATPase activity and self-assembly of MYH9 promote infection of viruses exploiting different uptake mechanisms
There is now a consensus about the existence of six main endocytic pathways in mammalian cells, which can be broadly divided based on their dependence on dynamin-2 (DYN-2) activity, a large GTPase involved in pinching off endocytic vesi cles from the plasma membrane (17). DYN-2-dependent endocytic processes include clathrin-mediated (CME), fast endophilin-mediated, and clathrin-independent carrier pathways, while dynamin-independent mechanisms include caveolae, macropinocytosis, and phagocytosis (17). In Fig. 1, we showed that MYH9 enhances infection of viruses that exploit different endocytic pathways to enter cells: TCRV, ZIKV, VSV, and MAYV use CME (18)(19)(20)(21), while LCMV exploits a macropinocytosis-like (MPL) mechanism for cellular entry (22). To formally examine whether the activity of MYH9 in viral infection is independent of the virus entry route, we analyzed infection levels in MYH9 knockdown cells upon DYN-2 inhibition; for these experiments, we used the NWA TCRV and the OWA LCMV, whose entry is DYN-2-dependent (CME) and -independent (MPL), respec tively. In brief, TCRV entry is initiated by the binding of the viral glycoprotein (GP) to a yettobeidentified surface receptor(s). Unlike pathogenic NWAs, TCRV infects human cells independently of transferrin receptor 1 (TfR1) (2). In this context, we showed that L-type voltage-gated calcium channels are required for efficient TCRV entry (16). Subsequent to the interaction of TCRV GP with its receptors/entry factors on the cell surface, viral internalization occurs through CME, followed by trafficking to a pH 5 late endosome where virus-cell fusion occurs (2,18). On the other hand, LCMV GP interacts with its well-characterized bona fide receptor α-dystroglycan (α-DG) on the cell surface, promoting viral internalization via a non-classic macropinocytosis mechanism where early macropinosomes deliver the virus-receptor complex to the late endosome for subsequent fusion of the viral and endolysosomal membranes facilitated by the mucin receptor CD164 (23)(24)(25).
To first corroborate the role of DYN-2 in TCRV and LCMV infection, we pre-incubated A549 cells for 30 min with the inhibitor dynasore, followed by viral infections (MOI = 1) in the presence of the inhibitor for 1 h, and viral RNA levels were analyzed by reverse transcription quantitative PCR (RT-qPCR) at 24 hpi. As expected, only TCRV, but not LCMV, infection was decreased in cells treated with dynasore with respect to a vehicle control (dimethyl sulfoxide [DMSO]) (Fig. 2A). Next, we transfected a siMYH9 for 48 h before dynasore treatment and virus infection, and we analyzed TCRV and LCMV RNA levels. MYH9 knockdown did not further decrease TCRV infection levels in dynasore-treated cells, suggesting that MYH9 enhances DYN-2-dependent TCRV infection, whereas LCMV infection levels were reduced only in those cells transfected with a siMYH9 irrespective of the treatment with dynasore (Fig. 2B), indicating that MYH9 is supporting virus infection independently of the viral entry route used. To rule out that TCRV and LCMV could use alternative entry pathways, which may be also contributing to infection and thus be targeted by the activity of MYH9, we tested the specific inhibitors chlorproma zine (CME), nystatin (caveolae), 5-(N-ethyl-N-isopropyl)amiloride (EIPA) (macropinocyto sis), and wortmannin (phagocytosis) prior to and during viral infection, as described above. We found that only chlorpromazine significantly reduced TCRV infection levels, confirming that this virus mostly uses CME to enter cells (Fig. 2C). Interestingly, nystatinand EIPA-treated cells showed increased levels of TCRV infection, which agrees with the reported upregulation of CME when other endocytic routes are perturbed/depleted (17). On the other hand, LCMV infection levels were only decreased in EIPA-treated cells due to the inhibition of MPL, its main entry route (Fig. 2C). Collectively, these data suggest that MYH9 is a pleiotropic proviral host factor that supports infection of viruses exploiting both DYN-2-dependent and -independent uptake mechanisms.
Many of the cellular functions performed by NM-IIA require the generation of motor/ contractile force by the head domain of MYH9, which may be suppressed either by inhibiting its Mg 2+ -dependent ATPase activity or by preventing the serine/threonine phosphorylation of the RLCs (9)(10)(11)(12). Thus, to determine the molecular basis of the activity of MYH9 in viral infection, we first used the inhibitor blebbistatin, which binds to MYH9 ATPase site, hence suppressing the overall mechanical activity of NM-IIA (26). The treatment of A549 cells with blebbistatin before and during infection resulted in a significant reduction in TCRV and LCMV RNA levels at 24 hpi, suggesting that the ATPase activity is necessary for the proviral function of MYH9 (Fig. 2D).
It was also shown that phosphorylation of threonine (T) 18 and serine (S) 19 residues in the RLCs of NM-IIA by myosin light chain kinase (MLCK) and Rho-associated protein kinases (ROCK) regulates the assembly of bipolar filaments that engage and contract actin filaments (11,12). Interestingly, these kinases are simultaneously active and compete for a limiting pool of cellular NM-IIA monomers; thus, the assembly of NM-IIA filaments could be driven not only by direct phosphorylation/activation of the T18/S19 residues by a given kinase, but also by inactivating the competing kinase(s) (27,28). To first corroborate that the phosphorylation of the RLC would not be completely prevented by only inhibiting MLCK or ROCK kinases, we analyzed the phosphorylation of the T18/S19 residues by western blot, upon treatment with the inhibitors ML-7 (MLCK) and Y-27632 (ROCK) in A549 cells pre-incubated with the calcium ionophore calcimycin to promote the phosphorylation of the RLCs. The individual treatment with ML-7 (10 μM) or Y-27632 (25 μM) modestly decreased phosphorylation of T18/S19; however, phosphorylation of the RLCs upon calcimycin treatment was markedly reduced when both inhibitors were used simultaneously (Fig. 2E). To test whether inhibiting the phosphorylation of RLCs modulates the levels of viral infection, we infected A549 cells with LCMV or TCRV in the presence of both ML-7 and Y-27632. Viral infection was analyzed 24 hpi, detecting a significant decrease in infection levels in those cells treated with the inhibitors with respect to control cells (Fig. 2F). Hence, preventing the assembly of NM-IIA in bipolar filaments, which is essential for the motor/contractile activity of MYH9, also reduces viral infection.
## Viral invasion drives the translocation of MYH9 to the plasma membrane
Phagocytosis is a finely coordinated process that involves the assembly of the cytoskele ton and the rapid accumulation of prophagocytic proteins at the phagocytic synapse to drive engulfment, including F-actin, NM-IIA, paxillin, and actinin (29). Given the fundamental function of MYH9 during phagocytic uptake and its role as a surface receptor for HSV-1 (8,15), we next asked whether MYH9 might function during entry of the viruses herein analyzed. To this end, we assessed the sub-cellular localization of endogenously expressed MYH9 during viral entry by confocal microscopy. Specifically, we analyzed the expression pattern of MYH9 after TCRV binding (MOI = 50) on ice for 1 h (0′) and at different intervals during viral internalization at 37°C (5′ and 15′), using a monoclonal anti-MYH9 and a cross-reactive anti-Junín virus antibody that binds to TCRV nucleoprotein (NP). We found that virus binding to A549 cells did not alter the localization of MYH9 with respect to MOCK-infected cells, as it was uniformly distributed along the cytoplasm (Fig. 3A; MOCK 0′, 5′, 15′ and TCRV 0′). However, when TCRV-infec ted cells were shifted to 37°C, a permissive temperature for viral internalization, we observed a marked accumulation of MYH9 at the plasma membrane in a time-depend ent fashion, which was not observed in MOCK cells (Fig. 3A; TCRV 5′, 15′). Since other cytoskeletal proteins also accumulate at the plasma membrane during phagocytosis and viral endocytosis, we also analyzed whether actin would translocate to the membrane by staining for actin filaments with Phalloidin-Alexa Fluor 568. We found that actin expression was also enriched at the cell periphery and that it increasingly colocalized with MYH9 during TCRV internalization at 37°C, which again was not observed for MOCK-infected cells (Fig. 3A), suggesting that these motor proteins collectively support TCRV endocytosis. Although we could not stain the NP of LCMV particles using a commercial antibody (clone VL4), we did observe the translocation of MYH9 during LCMV internalization, but not upon binding of viral particles (Fig. 3B; LCMV 5′, 15′). Thus, MYH9 may facilitate the internalization step of TCRV and LCMV entry when it translocates to the cell periphery.
## MYH9 supports post-binding steps of virus entry
To formally determine which step(s) of virus entry is promoted by the activity of MYH9, we next performed virus binding, internalization, and fusion assays with variable expression levels of MYH9 in A549 cells. In brief, TCRV and LCMV binding were assessed by quantifying viral RNA from bound particles to siMYH9-and siCTRL-transfected cells after 1 h on ice, while internalized viral RNA was analyzed after virus binding, incubation of infected cells at 37°C for 45 min and stripping off non-internalized viral particles with Proteinase K. MYH9 knockdown did not alter TCRV nor LCMV binding to cells but significantly decreased internalization levels of both viruses (Fig. 4A andB). Lastly, to determine whether MHY9 would modulate virus-cell fusion, we analyzed syncytium formation as a surrogate assay. To this end, A549 cells were co-transfected with a siMYH9 and a FLAG-tagged LCMV GP construct, and 48 h later, the transfected cells were pulsed with sodium citrate (pH = 5 or pH = 7) to assess the number and size of cell syncytia by immunofluorescence. As depicted in Fig. 4C, there was no syncytium formation when GP-transfected cells were pulsed at pH = 7, given the requirement of an acidic environment for arenavirus GP-mediated cell fusion (30). When the cells were pulsed at pH = 5, we observed a significantly lower number of syncytia, as well as cells per syncytium in siMYH9-transfected cells, with respect to siCTRL-transfected cells (Fig. 4C andD), indicating that MYH9 is also required for an efficient process of virus-cell fusion.
## Tyrosine phosphorylation of MYH9 is essential for its role in virus entry
Although MYH9 participates in several cellular processes requiring the generation of mechanical/contractile force, to our knowledge, there are a few studies analyzing how its activity is regulated by tyrosine phosphorylation (pY). Of note, MYH9 is phosphoactivated in different tyrosine residues upon B cell antigen receptor (BCR) stimulation, bacterial invasion, and to efficiently drive F-actin-driven phagocytosis (8,31,32). To analyze whether virus invasion triggers pY of MYH9, we infected A549 cells with TCRV for 15 min, and we either immunoprecipitated tyrosine-phosphorylated proteins using a mix of anti-pY antibodies or MYH9 with a monoclonal antibody. Western blot analysis showed that pY levels of MYH9 were significantly increased in TCRV-infected cells with respect to MOCK-infected control cells (Fig. 5A), indicating that virus entry triggers pY and activation of MYH9.
Tsai et al. demonstrated that pY of 277 and 1805 residues is essential for the function of MYH9 in phagocytic engulfment, given that point mutations in any of these residues inhibited its contractile function during the process (8). To analyze whether pY of these key residues would also modulate the function of MYH9 in virus entry, we mutated 277Y and 1805Y to phenylalanine (F) individually or in combination (227F, 1805F, and 277,1805F [2F]), using a construct of MYH9 WT fused to green fluorescent protein (GFP) as template (Fig. 5B). We also analyzed the MYH9 3A mutant, which harbors alanine (A) replacements in three serine (S) residues (1914S, 1915S, and 1946S), whose phosphory lation is essential for the activity of MYH9 in focal adhesion and protrusion extension (Fig. 5B) (33). To examine the contribution of these mutant versions of MYH9 to virus entry, we first assessed and quantified the expression of GFP in transfected cells, finding that the MYH9 227F, 1805F, 2F, and 3A mutants expressed at similar levels with respect to the MYH9 WT construct (Fig. 5C andD). Interestingly, forced expression of MYH9 WT resulted in a gain-of-function in viral internalization, while we did not observe an increase in internalization levels in 227F-, 1805F-, or 2F-expressing cells, with respect to an empty vector (EV) control (Fig. 5E). Overexpression of the 3A mutant enhanced viral internalization to similar levels to the WT construct (Fig. 5E), indicating that substituting 277Y and 1805Y residues, but not 1914S, 1915S, and 1946S, abrogates the activity of MHY9 during virus entry. Since we found that the MYH9 phospho-inactive mutants cannot support virus entry although their expression levels were comparable to the WT construct, we then asked whether these mutants may have an impairment to translocate to the cell periphery to promote viral entry. To test this, we analyzed GFP expression in MYH9 WT-and 2F-trans fected cells during TCRV internalization, an entry step enhanced by the activity of MYH9, by confocal microscopy. In MOCK-infected cells, both MYH9 WT and 2F constructs were distributed in the cytoplasm when incubated at 37°C for 5′ or 15′ (Fig. 5F; MOCK 5′, 15′). However, the MYH9 2F mutant did not accumulate at the plasma membrane during viral internalization, as it was observed for the WT protein for both time points (Fig. 5F; TCRV 5′, 15′). Collectively, these results indicate that pY of the 277 and 1805 residues is required for the translocation of MYH9 to the cell periphery, and hence for its proviral activity during viral entry.
## Src family kinases likely phosphorylate MYH9 upon virus entry
To further characterize the role of pY in infection, we lastly sought to identify which kinase(s) phosphorylate MYH9 upon virus entry. There are nine families of NRTKs in mammals, known as Src, Abl, Fes, Jak, Ack, Syk, Tec, Fak, and Csk (34), which could phosphorylate/activate MYH9 to promote infection. To analyze this, we investigated whether the activity of Src, Jak, and/or Syk kinases would modulate virus entry by using selective inhibitors. Briefly, A549 cells were pre-treated with small molecules targeting Src (PP1), Syk (piceatannol), and Jak (Jak inhibitor I) families for 45 min, to later assess viral internalization in the presence of the inhibitors. PP1 treatment markedly reduced internalization levels of both viruses, piceatannol did not affect TCRV nor LCMV internalization, while Jak inhibitor I had opposite effects: it greatly enhanced TCRV internalization but decreased the process for LCMV (Fig. 6A), indicating that members of the Src family phosphorylate cellular proteins relevant for both TCRV and LCMV entry. Since Src kinases phosphorylate hundreds of proteins to modulate their function in several physiological and pathological processes (35), we coupled MYH9 knockdown with PP1 treatment to analyze whether these kinases may be activating MYH9 upon virus infection. Interestingly, TCRV and LCMV internalization levels were similarly decreased in siMYH9-and siCTRL-transfected cells pre-incubated with PP1, while piceatannol-treated cells showed decreased internalization levels only when transfected with a siMYH9 but not with a siCTRL, as was observed for the vehicle (DMSO) control (Fig. 6B). These results suggest that members of the Src family may phosphorylate MYH9 during virus entry, since PP1 treatment did not further reduce viral internalization levels in siMYH9-trans fected cells. To further analyze this, we immunoprecipitated pY upon TCRV infection (15 min) in PP1-treated and control cells, as well as in MOCK-infected cells. Western blot analysis showed that pY of MYH9 during TCRV invasion was diminished in the presence of PP1 to levels comparable to those detected in MOCK cells (Fig. 6C); as expected, we did not detect MYH9 when using an IgG control antibody (raIgG) for the immunoprecipi tation (Fig. 6C). These results support that Src kinases are likely to phosphorylate and activate MYH9 during virus entry to further promote infection.
## DISCUSSION
NM-IIA is a ubiquitous actin-binding protein with critical functions in a wide range of cellular physiological processes, including cell adhesion, migration, and phagocyto sis, and more recently it has also been implicated in modulating viral and bacterial infection (9,36). NM-IIA is a molecular motor that provides mechanical force through ATP hydrolysis; hence, its function in infection is often associated with the motor and contractile properties of MYH9 (36). MYH9 was first identified as a cellular receptor for HSV-1, and later studies showed that it also promotes infection of other herpesvi ruses (Kaposi's-sarcoma associated Herpesvirus; KSHV) and arterivirus (severe fever with thrombocytopenia syndrome virus and porcine reproductive and respiratory syndrome virus) (15,(37)(38)(39)(40). Here, we showed that MYH9 enhances infection of endocytic viruses from divergent families, including Arenaviridae, Flaviviridae, Rhabdoviridae, and Togaviridae, and that such activity is conserved in human and mouse cells (Fig. 1C andF). We also found that viral infection upregulates MYH9 endogenous expression, suggest ing a virus-driven positive regulatory mechanism (Fig. 1G andH). This observation is not without precedent, since the upregulation and/or activation of cellular proteins to support infection has been reported. For instance, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection upregulates the expression of the viral receptor ACE2 through the transcription factor GATA6, while Influenza A virus infection enhances galectin-3 expression to modulate the activity of the viral polymerase complex (41,42). Thus, MYH9 is a broad proviral host factor whose expression is positively modulated to increase viral infection levels by a mechanism currently under investigation. By analyzing TCRV and LCMV infection, which, respectively, use CME and MPL for cell entry, we showed that MYH9 supports infection of viruses exploiting both dyna min-dependent and -independent uptake processes (Fig. 2B andC). Interestingly, it was shown that NM-IIA supports entry of KSHV by CME in foreskin fibroblasts and by macropinocytosis in dermal endothelial cells, which are both dependent on the activation of the E3 ubiquitin ligase c-Cbl by tyrosine phosphorylation (37,38). KHSV interacts with and activates Ephrin A2 early during infection, which in turn associates with phospho-c-Cbl, NM-IIA, as well as clathrin and its adapter protein AP2 to pro mote CME (37). In the case of macropinocytosis, the interaction of phospho-c-Cbl with phospho-RLC in KSHV-induced blebs leads to the association of NM-IIA with actin filaments to accelerate actomyosin contraction and bleb retraction (38). Future studies will be aimed at determining whether MYH9 forms part of a conserved protein signaling complex to promote distinct viral endocytic routes.
To get further insight into the proviral activity of MYH9, we sought to analyze the effect in viral infection of decreasing the motor/contractile activity of NM-IIA or preventing its self-assembly into bipolar filaments. The small molecule blebbistatin binds to the ATPase site of MYH9, decreasing the phosphate release rate and thereby suppressing its motor activity, which is needed for efficient phagocytic engulfment (8,26,43). In blebbistatin-treated cells, TCRV and LCMV infection levels were greatly reduced (Fig. 2D). Moreover, we determined that preventing the phosphorylation of the RLCs in the S18/T19 residues, which modulates the self-assembly of NM-IIA monomers into functional bipolar filaments that interact with the actin cytoskeleton (11,28), also reduced virus infection (Fig. 2E andF). These results hence indicate that the motor/ contractile activity and the self-oligomerization of NM-IIA are essential for the proviral function of MYH9.
Although we showed that MYH9 is a host factor that enhances infection of divergent endocytic viruses, its role in viral infection was first described for HSV-1; Arii et al. proposed that MYH9 functions as a cellular receptor for HSV-1, favoring virus binding and virus-cell fusion at the plasma membrane by directly interacting with the viral gB glycoprotein (15). Later studies suggested that MYH9 would also promote cell binding of arteriviruses by interacting with surface glycoproteins (39,40), although it was not established how MYH9, a cytoplasmic protein, would become inserted or expressed at the cell surface to promote virus binding (44). We examined whether MYH9 may serve as an entry factor for TCRV and LCMV by analyzing its sub-cellular localization upon infection. We found that virus binding did not induce the accumulation of MYH9 at the cell periphery, and that the levels of viral RNA from bound particles were similar in siMYH9-and siCTRL-transfected cells (Fig. 3A, B, and4A). However, when the infected cells were shifted to a permissive temperature for virus internalization, we observed a sustained increase in the expression of membrane-associated MYH9 and a marked decrease in internalized viral RNA in MYH9 knockdown cells (Fig. 3A, B, and4A), indicating that MYH9 expression is relevant for virus uptake but not for virus-cell binding. We also detected actin enrichment and increased co-localization with MYH9 during viral internalization, suggesting that these motor proteins cooperate to enhance this process (Fig. 3A). Lastly, we used a surrogate system to determine whether MYH9 would also modulate virus-cell fusion. Arenavirus glycoproteins are class I fusion proteins that get primed at an acidic pH to mediate viral-endosomal membrane fusion (45). Our fusion assay showed that MYH9 knockdown significantly reduced the formation of cell syncytia and the number of cells per syncytium at pH = 5 (Fig. 4C andD), indicating that MYH9 is likewise important for virus-cell fusion. Collectively, our results demonstrate that MYH9 translocates to the plasma membrane upon infection to support post-binding steps of viral endocytosis.
We next investigated how the function of MYH9 in virus entry may be regulated, because it was not previously addressed (15,39,40). In fact, little is known about how post-translational modifications in MYH9 modulate the activity of NM-IIA in other cellular processes that require motor/contractile force (9, 10). To our knowledge, there are three studies analyzing the role of tyrosine phosphorylation of MYH9 in phagocyto sis, BCR stimulation, and bacterial invasion (8,31,32). Specifically, phosphorylation of the Y277/1805 residues favors the polymerization of NM-IIA dimers into bipolar filaments, which is essential for phagocytic engulfment, while Listeria monocytogenes infection triggers the phosphorylation of Y158 which affects the ability of MYH9 to bind to and/or hydrolyze ATP thus preventing bacterial dissemination (8,32). Upon demonstrating that viral invasion triggers tyrosine phosphorylation of MYH9 (Fig. 5A), we examined whether phosphorylation of Y277 and/or Y1805 might regulate its function in virus infection. We first analyzed these residues because we showed that the contractile activity of MYH9 is needed to enhance entry (Fig. 2D), and we hypothesize that phagocytic proteins would also have a role in viral endocytosis. We showed that the MYH9 phospho-inactive mutants 277F, 1805F and 2F cannot functionally contribute to virus entry, indicating that phosphorylation of both tyrosine residues is needed for the proviral activity of MYH9 (Fig. 5E). Furthermore, we also determined that MYH9 2F had an impairment to translocate to and accumulate at the plasma membrane to facilitate entry (Fig. 5F). Thus, our results indicate that viral invasion triggers tyrosine phosphorylation of MYH9, which is essential to regulate its contractile activity to promote virus entry.
Lastly, we sought to determine which tyrosine kinase(s) would phosphorylate/acti vate MYH9 to enhance virus entry. Typically, viral infection involves the activation of many surface molecules and downstream proteins/signaling pathways, including receptor and non-receptor kinases, cytoskeletal proteins, among others (1). In this regard, it was shown that binding of the OWA Lassa virus GP to its bona fide receptor α-DG induces its tyrosine phosphorylation by NRTKs, which regulates the internalization of the virus-receptor complex (46). Here, we examined whether the Jak, Syk, and/or Src families of NRTKs may phosphorylate MYH9 upon virus infection. We initially analyzed these NRTKs because of their role in endocytic processes and virus entry: Syk kinases phosphorylate C-type lectin receptors which are used as entry factors (47,48), Jak kinases modulate CME (49), and Src kinases phosphorylate different proteins to facilitate virus entry and replication (50)(51)(52). Moreover, members of the Src family phosphorylate the Y158 residue of MYH9 during bacterial invasion (32). We found that inhibiting the Src kinases with the small molecule PP1 decreased internalization of both TCRV and LCMV (Fig. 6A). Thus, we next analyzed the effect of combining PP1 treatment and MYH9 knockdown to examine whether Src kinases may phosphorylate MYH9 upon virus entry. PP1 treatment did not further decrease viral internalization in siMYH9-transfected cells with respect to control cells, and the addition of the inhibitor reduced virus-triggered tyrosine phosphorylation of MYH9 during entry (Fig. 6B andC), suggesting that MYH9 may be activated by Src kinases upon infection. However, these observations need to be further corroborated by using selective inhibitors of individual members of the Src family, Src knockout systems, and by performing in vitro kinase assays. Since MYH9 expresses at high levels in many cell types and it is involved in several cellular processes, it cannot be depleted nor inhibited for therapeutic purposes (9). Thus, identifying the kinase(s) that activate MYH9 upon virus entry may provide new druggable targets to limit the early steps of infection of a variety of RNA viruses of public health importance.
## MATERIALS AND METHODS
## Mice
C57BL/6 mice were housed according to the policies of the Animal Care Committee of the Institut Pasteur de Montevideo. BMDMs were isolated from the hind limbs of 8-to 12-week-old mice as previously described (3).
## Cell lines and viruses
Vero, A549, U2OS, BHK-21, NIH-3T3, HeLa, and 293T cells were grown in Dulbecco's modified Eagle Medium (DMEM; Gibco) supplemented with 2 mM glutamine, 10% fetal bovine serum (FBS, Invitrogen) and penicillin (100 U/mL)-streptomycin (100 µg/mL) (Invitrogen) (DMEM complete). 293T cells stably expressing mouse mCAT-1 (293T-mCAT) were also cultured with geneticin (1 µg/mL). TCRV, VSV, ZIKV (PRVABC59 strain), and HSV-1 (KOS strain) were propagated in Vero cells, while LCMV (Armstrong strain) was grown in BHK-21 cells. Briefly, cell monolayers were infected at 70%-80% confluence at an MOI of 0.01-0.03, culture medium was removed at 24 hpi, and the cells were washed with phosphatebuffered saline (1× PBS) and refed with DMEM containing 2% FBS. Culture supernatant from TCRV, VSV, ZIKV, and LCMV infected cells was collected at 4-8 dpi, passed through a 0.45 µm filter, concentrated with Vivaflow 50 system filters (Sartorius), and stored at -80°C until use. For HSV-1, cells were cultured for 2-3 dpi until 100% of cells displayed cytopathic effect (CPE). Cells were then frozen at -80°C and thawed at 37°C three times, and the cell suspension was sonicated three times for 30 s to release virus particles. Moloney MLV was harvested from the supernatants of stably infected NIH 3T3 fibroblasts. The supernatant was passed through a 0.45 µm filter and treated with 20 U/mL DNase-1 (Roche) at 37°C for 30 min. MAYV was obtained by transfecting a cDNA infectious clone (53) in HeLa cells using Lipofectamine 3000 (Thermo). Infectious MAYV particles were collected from culture supernatant when full CPE was observed.
## Virus titration
TCRV titers were determined by an infectious center assay using a mouse monoclonal anti-Junín nucleoprotein (clone IC06-BE10, BEI resources) and an anti-mouse IgG-Alexa Fluor 488 secondary antibody (Invitrogen). LCMV, ZIKV, VSV, and HSV-1 titers were established by plaque assays in Vero cells using 1% agarose overlays. MLV titers were determined by a focus-forming assay on NIH-3T3 cells by using a monoclonal anti-ENV (Ab538), as previously described (3).
## RNA interference
Silencer Select siRNAs were purchased from Thermo. Human (#4390824; ID: s534066) and mouse (#4390771; ID: s70267) MYH9 siRNAs and a negative control (siCTRL) were used following a forward-transfection protocol with Lipofectamine RNAiMax (Invitrogen). Briefly, adherent cells were transfected at 60%-70% confluence with the indicated siRNAs for 48 h, and knockdowns were verified by RT-qPCR using the primers described in Table 1. siRNA-transfected cells were then infected for 24-48 h or used for virus entry assays.
## Chemicals
Blebbistatin, ML-7, Y-27632, chlorpromazine, nystatin, 5-(N-ethyl-N-isopropyl)amiloride (EIPA), dynasore, PP1, piceatannol, wortmannin, and Jak inhibitor I were purchased from Sigma and reconstituted with DMSO or methanol following the manufacturer's instructions. Adherent cells were pretreated with the inhibitors for 30-60 min, followed by viral infections for 1 h in the presence of the inhibitors. After the infection was completed, the cells were incubated without inhibitors for an additional 24 h.
## Generation of mouse primary cells
BMDMs were isolated from the hind limbs of 8-to 12-week-old C57BL/6 mice as previously described (3). Macrophages were cultured in DMEM complete supplemen ted with 100 µg/mL of macrophage colony-stimulating factor (M-CSF; Gibco) and 0.1% sodium pyruvate (Gibco). Cells were harvested 7 days after plating and seeded in 24-well plates for virus infections.
## In vitro and ex vivo infections
A549 cells were infected with TCRV, LCMV, ZIKV, and VSV at an MOI = 1 for 24 h, while HSV-1 infections (MOI = 1) were done in U2OS cells for 48 h. Viral nucleic acids were isolated at the indicated time points to assess infection levels by RT-qPCR (TCRV, LCMV, ZIKV, and VSV) or qPCR (HSV-1). MLV infections of 293T-mCAT cells were done at an MOI = 0.1 and viral DNA was measured by qPCR at 48 hpi. BMDMs were infected with TCRV and LCMV at an MOI = 1, and the cells were harvested at 48 hpi to assess infection levels by RT-qPCR.
## Western blot
Protein extracts (50 µg) were resolved on 10% SDS-polyacrylamide gels, transferred to polyvinylidene difluoride membranes (General Electric), and blocked with 3% bovine serum albumin (BSA; Sigma) or 5% non-fat milk. The following antibodies were used to detect endogenous, overexpressed, or fused proteins: rabbit anti-MYH9 (CST; #3403), rabbit anti-GAPDH (CST; #2118), mouse anti-FLAG (Invitrogen; # MA1-91878), rabbit anti-RLC (CST; #8505), rabbit anti-pRLC (Thr18/Ser19) (CST; #3674), mouse anti-Tubulin (Thermo; #A11126).
## Immunoprecipitation of endogenous proteins
A total of 500 µg of protein lysate per condition was incubated with a mix of rabbit anti-phospho-tyrosine antibodies (CST MultiMab; #8954), a rabbit monoclonal anti-MYH9 (Sigma; #M8064), or a rabbit IgG control polyclonal antibody (Proteintech; # 30000-0-AP) and 20 µL of Pierce Protein A/G agarose beads (Thermo; #20421) on rotation and at 4°C for 14 h. The immunocomplexes were then washed off five times with 1× cell lysis buffer and loaded onto SDS-polyacrylamide gels for western blot analysis.
## Nucleic acid isolation and RT-qPCR
DNA was isolated with the DNeasy Blood & Tissue Kit (Qiagen), and total RNA was isolated using the GeneJET RNA Purification Kit (Thermo). RNA was used as a template for cDNA synthesis using the RevertAid Reverse Transcriptase (Thermo) in a reaction mixture primed with 50 ng/µL of random hexamers (Macrogen). Viral and cellular RNAs were detected by RT-qPCR using a QuantStudio 3 Real-Time PCR System (Applied Biosystems) with specific primer pairs (Table 1), and the RNA expression was normalized to GAPDH. RT-qPCR reactions were done using Power SYBR Green Master Mix (Applied Biosystems), under these amplification conditions: 50°C for 2 min, 95°C for 10 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min. The efficiency of amplification was determined for each primer set by a standard curve with 10-fold serial dilutions of DNA of known concentration. The slope values of the standard curves for the primer pair amplicons ranged from 3.5 to 3.2. For each primer pair, a non-template control was included, and each sample was run in triplicate.
## Confocal microscopy
A549 cells were infected with TCRV and LCMV at an MOI = 50 on ice for 1 h and then shifted to 37°C for different time intervals (0, 5, and 15 min). MYH9 sub-cellular localiza tion and TCRV particles during virus entry were analyzed in a Zeiss LSM800 Confocal Microscope using a rabbit anti-MYH9 (Sigma; #M8064) and an anti-Junín nucleoprotein (clone IC06-BE10; BEI resources), and Alexa Fluor-conjugated (488 or 647) secondary antibodies (Invitrogen). Actin localization was determined using Texas Red-X Phalloidin (Invitrogen; #T7471).
## Virus entry assays
## Binding assay
A549 cells were incubated on ice with TCRV or LCMV (MOI = 20) for 1 h, washed three times with 1× PBS to remove unbound particles, and total RNA was isolated to analyze genomic RNA levels of bound virions by RT-qPCR.
## Internalization assay
A549 cells were infected on ice as described, washed with 1× PBS, shifted to 37°C for 45 min, treated with 1 mg/mL of Proteinase K (Gibco) for 45 min to strip off non-inter nalized particles from the cell surface, and treated with 2 mM of phenylmethylsulfonyl fluoride (PMSF) (Sigma) to inactivate Proteinase K. RNA was next isolated to analyze internalized viral RNA by RT-qPCR.
## Fusion assay
A549 cells were co-transfected with a siMYH9 and a FLAG-tagged LCMV GP construct using Lipofectamine 3000 (Thermo) for 48 h. Next, transfected cells were pulsed with sodium citrate at pH = 5 or pH = 7 for 10 min, and the number and size of cell syncytia were determined by epifluorescence microscopy in four randomly sampled regions per condition, by analyzing the GP expression using an anti-FLAG antibody (Thermo).
## Generation of MYH9 mutant constructs
DNA constructs encoding full-length MYH9 wild-type (WT) (#11347) and the 3xA mutant (#101041) fused to GFP were purchased from Addgene. The 277F, 1805F and 2F constructs were generated by PCR-based sitespecific mutagenesis using the MYH9 WT construct as template and the Q5 Site-Directed Mutagenesis Kit (New England Biolabs).
The primers used to generate these mutants are detailed in Table 2. All introduced mutations were validated by Sanger sequencing (Macrogen).
## DNA transfection
DNA expression constructs were transfected into 80%-90% confluent cells using Lipofectamine 3000 (Thermo) for 24 h according to the manufacturer's instructions. Cells were then lysed with 1× RIPA buffer supplemented with 1× Protease and Phosphatase Inhibitors (Thermo), the supernatants were clarified by centrifugation at 13,000 rpm and 4°C for 15 min, sonicated for 15 s, and stored at -70°C until use.
## Statistical analysis
Each experiment was done with three technical replicates per experiment. The data shown is the average of at least three independent experiments, or as indicated in the figure legends. Statistical analysis was performed using the GraphPad 8.1/PRISM software. Tests used to determine significance are indicated in figure legends.
## References
1. Yamauchi, Helenius (2013) "Virus entry at a glance" *J Cell Sci*
2. Sarute, Ross (2021) "The board is set, the pieces are moving: modulation of New World arenavirus entry by host proteins" *PLoS Pathog*
3. Sarute, Cheng, Yan et al. (2021) "Signal-regulatory protein alpha is an anti-viral entry factor targeting viruses using endocytic pathways" *PLoS Pathog*
4. Barclay, Brown (2006) "The SIRP family of receptors and immune regulation" *Nat Rev Immunol*
5. Barclay, Van Den Berg (2014) "The interaction between signal regulatory protein alpha (SIRPα) and CD47: structure, function, and therapeutic target" *Annu Rev Immunol*
6. Veillette, Thibaudeau, Latour (1998) "High expression of inhibitory receptor SHPS-1 and its association with protein-tyrosine phosphatase SHP-1 in macrophages" *J Biol Chem*
7. Oldenborg, Gresham, Lindberg (2001) "CD47-signal regulatory protein α (SIRPα) regulates Fcγ and complement receptor-mediated phagocytosis" *J Exp Med*
8. Tsai, Discher (2008) "Inhibition of "self" engulfment through deactivation of myosin-II at the phagocytic synapse between human cells" *J Cell Biol*
9. Pecci, Ma, Savoia et al. (2018) "MYH9: structure, functions and role of non-muscle myosin IIA in human disease"
10. Vicente-Manzanares, Ma, Adelstein et al. (2009) "Nonmuscle myosin II takes centre stage in cell adhesion and migration" *Nat Rev Mol Cell Biol*
11. Chou, Christensen, Frische et al. (2004) "Non-muscle myosin II and myosin light chain kinase are downstream targets for vasopressin signaling in the renal collecting duct" *J Biol Chem*
12. Billington, Wang, Mao et al. (2013) "Characteriza tion of three full-length human nonmuscle myosin II paralogs" *J Biol Chem*
13. Morrissey, Kern, Vale (2020) "CD47 ligation repositions the inhibitory receptor sirpa to suppress integrin activation and phagocyto sis" *Immunity*
14. Mhatre, Li, Bhatia et al. (2007) "Generation and characterization of mice with Myh9 deficiency" *Neuromolecular Med*
15. Arii, Goto, Suenaga et al. (2010) "Non-muscle myosin IIA is a functional entry receptor for herpes simplex virus-1" *Nature*
16. Sarute, Ross (2020) "CACNA1S haploinsufficiency confers resistance to New World arenavirus infection" *Proc Natl Acad Sci*
17. Rennick, Johnston, Parton (2021) "Key principles and methods for studying the endocytosis of biological and nanoparticle therapeu tics" *Nat Nanotechnol*
18. Roldán, Martínez, Forlenza et al. (2016) "Human transferrin receptor triggers an alternative Tacaribe virus internalization pathway" *Arch Virol*
19. Agrelli, De Moura, Crovella et al. (2019) "ZIKA virus entry mechanisms in human cells" *Infect Genet Evol*
20. Zhang, Kim, Fox et al. (2018) "Mxra8 is a receptor for multiple arthritogenic alphaviruses" *Nature*
21. Sun, Yau, Briggs et al. (2005) "Role of clathrin-mediated endocytosis during vesicular stomatitis virus entry into host cells" *Virology (Auckl)*
22. (2025) *Full-Length Text Journal of Virology*
23. Rojek, Perez, Kunz (2008) "Cellular entry of lymphocytic choriomeningitis virus" *J Virol*
24. Cao, Henry, Borrow et al. (1998) "Identification of alphadystroglycan as a receptor for lymphocytic choriomeningitis virus and Lassa fever virus" *Science*
25. Iwasaki, Ngo, De La Torre (2014) "Sodium hydrogen exchangers contribute to arenavirus cell entry" *J Virol*
26. Bakkers, Moon-Walker, Herlo et al. (2022) "CD164 is a host factor for lymphocytic choriomeningitis virus entry" *Proc Natl Acad Sci*
27. Kovács, Tóth, Hetényi et al. (2004) "Mechanism of blebbistatin inhibition of myosin II" *J Biol Chem*
28. Croft, Coleman, Li et al. (2005) "Actin-myosin-based contraction is responsible for apoptotic nuclear disintegration" *J Cell Biol*
29. Beach, Bruun, Shao et al. (2017) "Actin dynamics and competition for myosin monomer govern the sequential amplification of myosin filaments" *Nat Cell Biol*
30. Diakonova, Bokoch, Swanson (2002) "Dynamics of cytoskeletal proteins during Fcγ receptor-mediated phagocytosis in macrophages" *Mol Biol Cell*
31. Borrow, Oldstone (1994) "Mechanism of lymphocytic choriomenin gitis virus entry into cells" *Virology (Auckl)*
32. Baba, Fusaki, Shinya et al. (2003) "Myosin is an in vivo substrate of the protein tyrosine phosphatase (SHP-1) after mIgM cross-linking" *Biochem Biophys Res Commun*
33. Almeida, Mesquita, Cruz et al. (2015) "Src-dependent tyrosine phosphorylation of non-muscle myosin heavy chain-IIA restricts Listeria monocytogenes cellular infection" *J Biol Chem*
34. Rai, Thomas, Beach et al. (2017) "Myosin IIA heavy chain phosphorylation mediates adhesion maturation and protrusion in three dimensions" *J Biol Chem*
35. Amatya, Lin, Andreotti (2019) "Dynamic regulatory features of the protein tyrosine kinases" *Biochem Soc Trans*
36. Kumar, Jaggi, Singh (2015) "Pharmacology of Src family kinases and therapeutic implications of their modulators" *Fundam Clin Pharmacol*
37. Tan, Yuan, Liu et al. (2019) "Non-muscle myosin II: role in microbial infection and its potential as a therapeutic target" *Front Microbiol*
38. Dutta, Chakraborty, Bandyopadhyay et al. (2013) "EphrinA2 regulates clathrin mediated KSHV endocytosis in fibroblast cells by coordinating integrin-associated signaling and c-Cbl directed polyubiquitination" *PLoS Pathog*
39. Veettil, Sadagopan, Kerur et al. (2010) "Interaction of c-Cbl with myosin IIA regulates Bleb associated macropi nocytosis of Kaposi's sarcoma-associated herpesvirus" *PLoS Pathog*
40. Sun, Qi, Liu et al. (2014) "Nonmuscle myosin heavy chain IIA is a critical factor contributing to the efficiency of early infection of severe fever with thrombocytopenia syndrome virus" *J Virol*
41. Gao, Xiao, Wang et al. (2016) "MYH9 is an essential factor for porcine reproductive and respiratory syndrome virus infection" *Sci Rep*
42. Israeli, Finkel, Yahalom-Ronen et al. (2022) "Genome-wide CRISPR screens identify GATA6 as a proviral host factor for SARS-CoV-2 via modulation of ACE2" *Nat Commun*
43. Yang, Chen, Wang et al. (2023) "Upregulation of galectin-3 in influenza A virus infection promotes viral RNA synthesis through its association with viral PA protein" *J Biomed Sci*
44. Sosale, Rouhiparkouhi, Bradshaw et al. (2015) "Cell rigidity and shape override CD47's "self"-signaling in phagocytosis by hyperactivating myosin-II" *Blood*
45. Matozo, Kogachi, De Alencar (2022) "Myosin motors on the pathway of viral infections" *Cytoskeleton (Hoboken)*
46. York, Dai, Amberg et al. (2008) "pH-induced activation of arenavirus membrane fusion is antagonized by small-molecule inhibitors" *J Virol*
47. Moraz, Pythoud, Turk et al. (2013) "Cell entry of Lassa virus induces tyrosine phosphorylation of dystroglycan" *Cell Microbiol*
48. Chen, Lin, Huang et al. (2008) "CLEC5A is critical for dengue-virusinduced lethal disease" *Nature*
49. Chen, Liu, Wu et al. (2012) "CLEC5A regulates Japanese encephalitis virusinduced neuroinflammation and lethality" *PLoS Pathog*
50. Karim, Saul, Ghita et al. (2022) "Numbassociated kinases are required for SARS-CoV-2 infection and are cellular targets for antiviral strategies" *Antiviral Res*
51. (1016)
52. Mainou, Dermody (2011) "Src kinase mediates productive endocytic sorting of reovirus during cell entry" *J Virol*
53. Delorme-Axford, Sadovsky, Coyne (2013) "Lipid raft-and SRC family kinase-dependent entry of coxsackievirus B into human placental trophoblasts" *J Virol*
54. Kumar, Agrawal, Khan et al. (2016) "Identification and characterization of the role of c-terminal Src kinase in dengue virus replication" *Sci Rep*
55. Chuong, Bates, Weger-Lucarelli (2019) "Infectious cDNA clones of two strains of Mayaro virus for studies on viral pathogenesis and vaccine development" *Virology (Auckl)*
56. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645942&blobtype=pdf | # Viral RNA pUGylation promotes antiviral immunity in C. elegans
David Lowe, Aditi Shukla, Scott Kennedy
## Abstract
RNA interference (RNAi) is a component of the innate immune systems of many eukaryotes, including C. elegans. During RNAi in C. elegans, the nucleotidyltrans ferase RDE-3 modifies the 3′ termini of mRNAs with polyUG (pUG) tails, which recruit RNA-dependent RNA polymerase (RdRP) enzymes that drive gene silencing by synthesiz ing antisense small interfering (si)RNAs. During normal growth and development, RDE-3 pUGylates transposon RNAs to silence transposons and protect genomic integrity. How C. elegans identifies specific RNAs for pUGylation and whether the pUGylation system is used for other biological purposes are not yet known. Here, we show that pUGylation contributes to antiviral immunity in C. elegans: During infection of C. elegans with Orsay virus, RDE-3 adds pUG tails to viral RNAs, which converts these RNAs into templates for RdRP-based antiviral siRNA production, thereby limiting viral replication. We present evidence that MUT-15 is critical for viral pUGylation because it interacts with RDE-3 and the NYN domain-containing endonuclease RDE-8, thus bridging the enzymes that cleave and pUGylate viral RNA, ensuring efficient antiviral immunity. We conclude that pUGylation promotes antiviral immunity in C. elegans, and we provide molecular insights into how C. elegans identifies and neutralizes its internal and external parasitic threats.IMPORTANCE Viruses are a threat to all organisms. Therefore, organisms have evolved numerous systems to recognize and neutralize viruses. Many of these systems, which are referred to as innate immune systems, function by recognizing unique molecular characteristics of the viral genetic material. One such innate immune system is RNA interference (RNAi). RNA interference uses double-stranded RNA, which is an obligatory byproduct of replication for many viruses, as a weapon to fight viruses. In this work, we provide molecular insights into how the nematode C. elegans uses RNA interference and viral double-stranded RNAs to defend itself against viral invaders. KEYWORDS small RNA, antiviral RNAi, poly(UG) RNA, Orsay virus, C. elegans, pUGasome V iruses pose a near-ubiquitous threat to all forms of life, prompting the evolution of diverse innate immune systems that detect and restrict viral replication by recognizing molecular signatures unique to viral nucleic acids. Among these defense mechanisms, RNA interference (RNAi) has emerged as a conserved antiviral strategy in plants and invertebrates (1-11). In C. elegans, RNAi is triggered by long double-stranded RNAs (dsRNAs), which are processed by the RNase III-like enzyme DICER into 20-25 nucleotide small interfering RNAs (siRNAs) (3,(12)(13)(14). These primary siRNAs are loaded onto the Argonaute protein RDE-1, which guides the complex to complementary target RNAs via Watson-Crick base-pairing (14-17). Target engagement is thought to recruit the endonuclease RDE-8, which cleaves RNAs, enabling RNA-dependent RNA polymerases (RdRPs) to generate secondary siRNAs off 19). RdRP-amplified siRNAs are loaded onto secondary Argonautes that reinforce gene silencing (17,20,21). RdRP-based siRNA amplification is a major contributor to the potency and heritability of . Recent discoveries have expanded our understanding of the RNAi pathway by identifying a post-transcriptional RNA modification-poly(UG), or pUG
tailing-that facilitates secondary siRNA biogenesis (25,26). The nucleotidyltransferase RDE-3 appends 3' pUG tails to RNA fragments generated by the coordinated action of RDE-1 and RDE-8 (25,27). pUG tails then adopt G-quadruplex-like structures that recruit RdRPs, enabling efficient secondary siRNA synthesis from pUGylated RNA templates (25,28). RDE-3 has been shown to pUGylate transposon-derived RNAs during normal development, thereby restricting these endogenous mobile elements and preserving genomic stability (25,27). However, how RDE-3 identifies its RNA substrates and whether RDE-3 functions with co-factors to achieve selective RNA targeting remain unclear.
In germ cells, RDE-3 localizes to cytoplasmic foci known as Mutator foci (25,29). The formation of Mutator foci requires the low-complexity protein MUT-16, which is thought to scaffold the assembly of small RNA biogenesis machinery within Mutator foci (29,30). Colocalization studies have shown that Mutator foci contain several essential RNAi factors, including RDE-3, the exonuclease MUT-7, the RdRP RRF-1, and MUT-15a protein of unknown domain structure with limited conservation outside nematodes (29)(30)(31). Current models posit that MUT-16 concentrates these components, as well as other factors such as RDE-8, together to facilitate production of the secondary siRNAs required for transposon silencing and germline RNAi (29,30,32,33). Although many of these proteins are also required for RNAi in somatic tissues, Mutator foci have not been detected in the soma (30), raising questions about how the pUG RNA and siRNA biogenesis machinery is organized in the absence of Mutator foci.
The Orsay virus is a naturally occurring positive-sense single-stranded RNA virus that infects C. elegans. Its genome comprises two RNA segments: RNA1 encodes the viral RNA-dependent RNA polymerase, and RNA2 encodes the capsid protein (34)(35)(36). During replication, viral dsRNA intermediates are formed, which are recognized by the RNAi machinery (1,36,37). For example, RDE-1 and RRF-1 are required to limit Orsay virus replication, establishing RNAi as a functional component of C. elegans antiviral immunity (1,36). Whether RDE-3-mediated pUGylation also contributes to the host's antiviral response is not known.
Here, we show that viral RNA pUGylation is required for Orsay viral immunity in C. elegans. We present evidence that MUT-15 promotes viral immunity by recruiting RDE-8, the enzyme responsible for cleaving viral RNAs, into proximity with the pUGylase RDE-3. We also present evidence that the role of MUT-16 in viral immunity is to bring RdRPs into proximity with RDE-3 and RDE-8, thus allowing efficient conversion of pUGylated viral RNA templates into antiviral siRNAs. The findings provide mechanistic insights into how C. elegans defends itself against viral pathogens through a unified RNA modification and silencing strategy.
## RESULTS
## RNA pUGylation contributes to innate immunity in C. elegans
The NT RDE-3 appends pUG tails to transposon RNAs in the C. elegans germline to limit transposon replication (25). We wondered if the pUG system might play additional roles in protecting C. elegans against nucleic acid parasites. To test this idea, we infected wild-type (WT) animals harboring an rde-3 deletion that removes conserved domains (RDE-3(-)), animals harboring two different RDE-3 catalytic site mutations (G366R or D189N), or RDE-3(G366R) animals reverted to wild-type (RDE-3(GR366G-reverted)) with Orsay virus and used qRT-PCR to monitor levels of Orsay viral RNAs (ORV1 and ORV2) in these animals 4 days post-infection. We observed a ~ 6-7 log2-fold increase in viral ORV1 and ORV2 RNA (henceforth, viral load) in animals that lacked functional RDE-3, indicating RDE-3 is needed to limit Orsay viral replication in C. elegans (Fig. 1A andB).
Orsay virus infects and replicates within the intestinal cells of C. elegans (36,38,39). To determine the site of RDE-3′s antiviral activity, we employed the auxin-inducible degron (AID) system to deplete RDE-3 in a tissue-specific manner (40,41). We generated animals that expressed GFP::AID::RDE-3 in all tissues (25) and the F-box Transport Inhibitor Response 1 (TIR1), which is needed for auxin-induced degradation, in specific tissues (40,41). Pan-soma or intestinal depletion, but not muscle depletion, of RDE-3 led to animals unresponsive to RNAi targeting the intestinally expressed dpy-6 mRNA, which indicates tissue-specific RDE-3 depletion was successful (Fig. S1). We performed auxin-induced degradation of RDE-3 in all somatic cells (pan-soma) or in the muscle or intestinal cells of animals infected with Orsay virus. Pan-soma or intestinal depletion, but not muscle depletion, of RDE-3 resulted in ~2-4 log2-fold increase in viral load post-Orsay virus infection (Fig. 1C andD). The increase in viral load observed after intestinal RDE-3 depletion indicates RDE-3 acts cell-autonomously to limit viral replication. The failure of Orsay virus to accumulate to the same level exhibited by rde-3(-) animals after intestinal RDE-3 knockdown may be due to incomplete knockdown of RDE-3 in these experiments, or it may indicate additional, non-autonomous roles for RDE-3 in antiviral immunity.
Because the pUGylase RDE-3 restricts viral RNA expression, we hypothesized that pUGylated viral RNAs might be produced during Orsay virus infection. To test this hypothesis, we performed RT-PCR specific for pUG-modified RNAs to detect pUGylated OrV RNA1 transcripts in WT or RDE-3(-) animals that had been infected with Orsay (Fig. S2A). The analyses revealed Orsay virus-derived pUGylated RNAs generated post-infec tion, which required RDE-3 for their biogenesis (Fig. 2A andB). Note that failure to detect viral pUG RNA in rde-3 mutants indicates that the pUG RNA assay detects viral pUG RNAs and not other types of viral RNA (Fig. S2A). An in vitro synthesized gfp pUG RNA, which was spiked-in during library preparation, was detected at similar levels in both WT and RDE-3(-) animals, indicating failure to detect Orsay pUGs in RDE-3(-) animals was not due to technical issues (Fig. S3). Additionally, we employed Nanopore-based cDNA sequencing to sequence pUGylated RNAs in RNA isolated from WT and RDE-3(-) animals infected with Orsay virus (see Methods). Nanopore sequencing detected ORV1 and ORV2 RNAs possessing 3′ non-templated pUG tails in WT animals, but not in RDE-3(-) animals (Fig. 2B). Analysis of the Nanopore sequencing data revealed that pUG tails were added to fragments of ORV1 and ORV2 RNAs and that sites of pUGylation on these RNAs were distributed across both viral RNAs (Fig. 2C). Collectively, these findings show that Orsay RNA undergoes RDE-3-dependent pUGylation.
Following infection, rde-3 mutants exhibited elevated viral loads and lacked detecta ble viral pUG RNAs, suggesting that pUGylated viral RNAs contribute to antiviral immun ity. How might pUGylated viral RNAs restrict viral replication? RDE-3 generates RNAs with pUG tails that form atypical G-quadruplex RNA structures, which interact with the RNAdependent RNA polymerases (RdRPs) RRF-1 and EGO-1 (hereafter referred to collectively as RdRP) (25,28). RdRP utilizes pUGylated RNAs as templates to synthesize antisense siRNAs, which function in trans to silence complementary RNAs (25,28). Thus, viral pUG RNAs could facilitate antiviral immunity by promoting RdRP-mediated antiviral siRNA synthesis. Supporting this model, previous work demonstrates increased Orsay viral loads in animals deficient for RdRP (37,42,43). Using custom-designed Taqman probes (Fig. S2B), we quantified levels of two antiviral siRNAs, which were identified in previous studies (37,42,43), in WT or RDE-3(-) animals infected with Orsay. The analysis showed that RDE-3(-) animals produced approximately ~4-6 log2-fold fewer antiviral siRNAs than WT animals after infection (Fig. 2D andE). These results suggest that viral pUGylated RNAs inhibit viral replication by acting as templates for antiviral siRNA biogenesis.
## In silico predictions of RDE-3 interactors
To explore the molecular basis of RDE-3-based RNA pUGylation and to identify potential RDE-3 co-factors, we leveraged immunoprecipitation-mass spectrometry (IP-MS) data that previously identified 21 proteins enriched in MUT-16 pulldowns, including the pUGylase RDE-3 (18,30,31). To predict potential RDE-3-interacting proteins, we used AlphaFold2 to assess whether any MUT-16-interacting proteins might be predicted to interact directly with RDE-3. As a specificity control, we included 100 randomly selected, conserved, germline-expressed C. elegans proteins in the analysis. We evaluated two AlphaFold2 output metrics to predict protein-protein interactions: (i) interface-predicted template modeling (ipTM) scores, which estimate interaction likelihood (44), and (ii) the number of independent structure predictions (out of five) that support the interaction, which indicates prediction robustness (45). This analysis identified two high-confidence RDE-3 interactors: MUT-15 and UAF-1 (Fig. 3A). Given prior evidence that RDE-3 and MUT-15 colocalize in germ cells and are both required for RNA interference and transpo son silencing (25,27,29,30,32), we focused further investigation on MUT-15. To assess the specificity of the predicted MUT-15-RDE-3 interaction, we used AlphaFold2 to test for predicted interactions between MUT-15 and 40 other nucleotidyltransferases (NTs) encoded in the C. elegans genome. RDE-3 was the only NT predicted to interact with MUT-15, indicating a degree of specificity in the predicted MUT-15-RDE-3 interaction (Fig. 3A). A predicted alignment error (PAE) plot for the predicted MUT-15-RDE-3 interaction is shown in Fig. 3B. A three-dimensional (3D) representation and four other independent PAE predictions of the predicted MUT-15-RDE-3 complex are shown in Fig. S4A. Given that RDE-3 and MUT-15 co-precipitate with MUT-16, we asked if Alphafold2-Multimer might predict a direct interaction between either RDE-3 and MUT-16 or MUT-15 and MUT-16. Previous studies have demonstrated that colocalization of RDE-3 and MUT-15 with MUT-16 in germ cells depends upon amino acids 89-384 of MUT-16, which is the only segment of MUT-16 predicted to adopt secondary structure (hereafter MUT-16(strd)) (29,30). AlphaFold2-Multimer predicted that MUT-15, but not RDE-3, interacted with MUT-16(strd). PAE plots of the predicted MUT-15-MUT-16 interaction are shown in Fig. S4B. Modeling of RDE-3, MUT-15, and MUT-16(strd) together resulted in the prediction of a trimeric complex, in which MUT-15 bridges RDE-3 and MUT-16(strd). PAE plots and a 3D representation of the predicted MUT-15-RDE-3-MUT-16(strd) interaction are shown in Fig. 3C; Fig. S4C,respectively. RDE-8, NYN-1, and NYN-2 are NYN (Nedd4-BP1, YacP Nucleases) domain-containing endonucleases. RDE-8 is required for RNAi, and NYN-1 and NYN-2 are redundantly required for RNAi in C. elegans (18). During RNAi, RDE-8 is thought to endonucleolytically cleave mRNAs targeted for silencing by dsRNAs and the Argonautes RDE-1 (18). NYN-1 and NYN-2 lack catalytic residues required for endonuclease activity (18), suggesting that NYN-1 and NYN-2 play redundant, non-catalytic roles in RNAi. MUT-16 coprecipitates with RDE-8, NYN-1, and NYN-2 (18), and MUT-16(strd) is required for RDE-8, NYN-1, and NYN-2 to colocalize with MUT-16 in C. elegans germ cells (30). Thus, RDE-8, NYN-1, and NYN-2 are strong candidates for proteins that could interact with MUT-15, RDE-3, or MUT-16. To probe potential interactions between RDE-8, NYN-1, or NYN-2 with MUT-15, RDE-3, or MUT-16, we again utilized AlphaFold2-Multimer modeling. The resulting predictions showed that NYN-1 and NYN-2, but not RDE-8, were predicted to interact with MUT-15 through the same interface on MUT-15, suggesting a mutually exclusive interaction between NYN-1 or NYN-2 with MUT-15. The predicted interaction surface on MUT-15 for NYN-1/2 was distinct from the predicted RDE-3 interaction surface on MUT-15. The predicted mutually exclusive interactions of NYN-1 or NYN-2 with MUT-15 are consistent with the redundant role played by NYN-1 and NYN-2 in RNAi. PAE plots of the predicted NYN-1-MUT-15 and NYN-2-MUT-15 interactions are shown in Fig. 3D; Fig. S5A, respectively. To assess the specificity of the predicted NYN-1/2-MUT-15 interaction, we examined 26 additional C. elegans endonucleases for their predicted binding to MUT-15. Only NYN-1 and NYN-2 displayed predicted interactions, indicating a degree of specificity in the NYN-1/2-MUT-15 predictions (Fig. 3A). Although a direct interaction between RDE-8 and MUT-15 was not predicted, additional modeling with just RDE-8 and NYN-1/2 revealed predicted interactions between RDE-8 and NYN-1 or NYN-2, which is consistent with the results of previous in vivo studies (18). PAE plots and 3D repre sentations of the predicted NYN-1/2-RDE-8 interactions are shown in Fig. 3E; Fig. S5B, respectively.
When RDE-8, NYN-1 or NYN-2, MUT-15, MUT-16(strd), and RDE-3 were modeled simultaneously, Alphafold2-Multimer predicted a heteropentameric complex. In this predicted complex, MUT-15 bridges the RNA pUGylase RDE-3, MUT-16, and NYN-1/2, while NYN-1/2 connects the endonuclease RDE-8 to the other components of the predicted complex (Fig. 3E; Fig. S6). When MUT-16 was not included in the prediction, the other four proteins (RDE-8, NYN-1 or NYN-2, MUT-15, and RDE-3) were predicted to form a similar complex, which lacked MUT-16 (Fig. 3F; Fig. S7). PAE plots and 3D representation of the predicted heteropentameric (with MUT-16) and heterotetrameric complex (without MUT-16) are shown in Fig. 3E andF; Fig. S6 andS7. Neither RDE-3 nor MUT-16 was predicted to interact with NYN-1, NYN-2, or RDE-8 within these complexes. Because data presented below will show that RDE-3, MUT-15, NYN-1/2, and RDE-8, but not MUT-16, are required for viral RNA pUGylation, we refer to the four-protein hetero tetrameric complex containing RDE-3, MUT-15, NYN-1/2, and RDE-8 as the predicted pUGasome.
## Co-immunoprecipitation studies support predicted MUT-15 interactions and the general architecture of the predicted pUGasome
To experimentally test the AlphaFold2-Multimer-based predictions of a pUGasome complex, we introduced epitope tags at the endogenous loci of the four core predicted components: 3XFLAG::RDE-3, MUT-15::3XHA, 3XV5::RDE-8, and 3XV5::NYN-1/NYN-2. Prior studies have established the necessity of RDE-3, RDE-8, and MUT-15 for efficient RNA interference (RNAi) in C. elegans. To confirm that the introduced tags did not compromise protein function, we performed RNAi assays targeting dpy-6, a gene whose silencing results in a characteristic Dumpy (Dpy) phenotype in wild-type animals. Consistent with expectations, animals harboring null mutations in rde-3, mut-15, or rde-8 failed to exhibit the Dpy phenotype upon exposure to dpy-6 RNAi (Fig. S8A). When animals expressing 3X FLAG::RDE-3, MUT-15::3X HA, or 3X V5::RDE-8 were exposed to dpy-6 RNAi, they became dumpy (Fig. S8B), indicating that the introduced tags did not disrupt protein function.
We conducted four co-IP experiments to test AlphaFold2-predicted interactions within the pUGasome. First, we examined the predicted interaction between RDE-3 and MUT-15 by conducting co-IP assays, which confirmed an in vivo physical interaction between RDE-3 and MUT-15 (Fig. 4A). Second, AlphaFold2 modeling suggested residues 2-44 of MUT-15 are essential for an interaction with RDE-3 (Fig. 4B). Modeling a truncated MUT-15 variant lacking residues 2-44 (henceforth referred to as MUT-15(Δ2-44)) failed to predict binding with RDE-3 in silico (Fig. S9). To assess if a.a. 2-44 of MUT-15 were required for MUT-15's interaction with RDE-3 in vivo, we generated a mut-15(Δ2-44) allele via CRISPR-Cas9 genome editing. MUT-15(Δ2-44) animals were unresponsive to dpy-6 RNAi, demonstrating a functional requirement for residues 2-44 of MUT-15 in RNAi (Fig. 4C). Additionally, while MUT-15(Δ2-44) was expressed, it failed to co-immunoprecipitate with RDE-3 (Fig. 4D). Third, we tested whether the predicted RDE-3-MUT-15 interaction requires MUT-16: AlphaFold2 predicts an interaction that is independent of MUT-16. Co-IP experiments in MUT-16(-) animals showed that RDE-3 and MUT-15 co-precipita ted in the absence of MUT-16, validating a MUT-16-independent interaction (Fig. 4A). Fourth, we assessed whether RDE-3 interacts with RDE-8 or NYN-1/2 through MUT-15, as computationally predicted. Indeed, RDE-3 co-immunoprecipitated with RDE-8 and NYN-1 in wild-type animals, but not in MUT-15(-) animals, consistent with the idea that MUT-15 bridges RDE-8 and NYN-1 to RDE-3 (Fig. 4E). Together, the co-IP analyses support the general architecture of the pUGasome, as predicted by Alphafold2.
## Components of pUGasome are required for viral pUG RNA synthesis and antiviral immunity
To investigate the in vivo significance of the predicted pUGasome, we examined the effects of disrupting each component on viral replication, antiviral pUG RNA synthesis, and antiviral siRNA synthesis. We first quantified Orsay virus RNA levels via quantitative RT-PCR (qRT-PCR) in wild-type animals and mutants deficient in individual pUGasome components: RDE-3(-), RDE-8(-), NYN-1(-); NYN-2(-), MUT-15(-), or MUT-15(Δ2-44). All five mutant strains exhibited significantly increased viral RNA loads, approximately ~4-5 log2-fold higher than wild-type controls (Fig. 5A andB; Fig. S10A). These results demon strate that each pUGasome component, including amino acids 2-44 of MUT-15, which are required for MUT-15 to interact with RDE-3, are required for suppressing viral replication. These observations align with previously reported findings for rde-8 mutants (18). To determine if the predicted components of the pUGasome are required for viral RNA pUGylation, we measured ORV1 pUG-modified RNAs by qRT-PCR in animals lacking components of the predicted pUGasome. Animals lacking any of the four predicted pUGasome components significantly diminished the level of ORV1 pUG RNAs after infection (Fig. 5C), establishing an essential role for the predicted pUGasome compo nents in mediating viral RNA pUGylation. We next quantified antiviral secondary siRNAs generated in response to Orsay virus infection using TaqMan-based assays. Absence of any pUGasome component resulted in loss of antiviral siRNA production (Fig. 5D; Fig. S10B). Collectively, the results establish a critical role for the four predicted components
## MUT-16 and RdRP act downstream of the pUGasome
Interestingly, MUT-16(-) animals behaved differently than animals lacking the compo nents of the predicted pUGasome when infected with Orsay virus. While MUT-16(-) animals exhibited elevated viral loads and were defective for synthesizing antiviral siRNAs (Fig. 6A andB; Fig. S11A andB), similar to animals lacking components of the predicted pUGasome, MUT-16(-) animals retained the ability to produce antiviral pUG RNAs (Fig. 6C andD). Thus, MUT-16 is not required for viral pUG RNA synthesis, but is required for antiviral siRNA synthesis and for suppressing viral load. Why might MUT-16 be needed for antiviral siRNA synthesis, but not antiviral pUG RNA synthesis? MUT-16 co-IPs with the components of the predicted pUGasome as well as with RdRP (18,31). Current models posit that RdRP uses pUGylated RNAs as templates for siRNA synthesis (25,28). Thus, MUT-16 could promote antiviral immunity by bringing RdRP into proximity with the pUGasome, thus enabling efficient conversion of pUG RNAs into antiviral siRNAs. Consistent with this idea, we observed that RdRP mutant animals behaved similarly to MUT-16(-) animals following viral infection: they had elevated viral loads and were defective for synthesizing antiviral siRNAs (Fig. 6A andB; Fig. S11A andB), which is consistent with previous reports (37,42,43,46). However, RdRP(-) animals retained the ability to produce viral pUG RNAs (Fig. 6C andD; Fig. S11A andB). Inciden tally, the lack of Orsay siRNA signals in animals lacking RdRP, which current models posit is the enzyme responsible for producing antiviral siRNAs, suggests the TaqMan Orsay siRNA assays are specific for viral siRNAs and not other types of viral RNA (Fig. S2B). The data hint that MUT-16 could promote antiviral immunity by recruiting RdRP into proximity with the pUGasome, thus coupling antiviral pUG synthesis by the pUGasome with antiviral siRNA synthesis by RdRP. The TUDOR-domain protein EKL-1 interacts with the RdRP RRF-1 to promote RdRP-based siRNA biogenesis (23,29,(47)(48)(49). We used AlphaFold2-Multimer to ask if RRF-1 and/or EKL-1 might be predicted to interact with each other and/or with MUT-16. Indeed, AlphaFold2-Multimer predicted that RRF-1 and EKL-1 interact with each other (Fig. S12A) and with a.a. 516-581 of MUT-16 (Fig. 6E andF; Fig. S12B through D). PAE plots for the predicted MUT-16-RRF-1-RdRP interaction are shown in Fig. 6E andF; Fig. S12B through D. A 3D representation of the predicted RdRP-EKL-1-MUT-16(484-632) complex is shown in Fig. S12E. The residues in MUT-16 predicted by Alphafold to interact with RRF-1/EKL are distinct from those predicted to interact with MUT-15 and the pUGasome. The RdRP-EKL-1-MUT-16 Alphafold2 prediction is experi mentally supported with published IP-MS results (23,29,(47)(48)(49), and with data showing that a.a. 484-569 of MUT-16 are necessary for RdRP to colocalize with MUT-16 (30). Taken together, the data suggest that MUT-16 promotes antiviral immunity by physically linking the pUGasome to RdRP, thereby enabling efficient conversion of pUGylated RNAs into antiviral siRNAs (Fig. 7).
## DISCUSSION
Here, we show that RNA pUGylation limits Orsay virus infection in C. elegans. We present evidence supporting the existence of a four-protein complex, which we term the pUGasome, that is required for viral RNA pUGylation and antiviral siRNA synthesis. Finally, we show that the pUGasome associates with MUT-16, and this association likely enables antiviral siRNA production by RdRP.
AlphaFold-Multimer simulations predicted that RDE-3, MUT-15, NYN-1/2, and RDE-8 assemble into a pUGasome. Consistent with this model, we find that all four components of the pUGasome are required for viral pUGylation. Biochemical tests of the AlphaFold-Multimer pUGasome prediction support the general architecture of the Alphafold The pUGasome model is consistent with existing co-IP data and subcellular colocali zation studies of the pUGasome components with one notable exception: MUT-15 is not required for RDE-3 to localize to Mutator foci, which are thought to be phase-separated germ granules that form in C. elegans germ cells (29,30). This is not the expected result if the MUT-15-RDE-3 interaction is the only mechanism localizing RDE-3 to the pUGasome and Mutator foci. It is possible (i) that RDE-3 interacts with other proteins or RNAs, which localizes RDE-3 independently of MUT-15, or (ii) that the logic of pUGylation or pUGasome architecture differs in the soma and germline. Our Alphafold-Multimer screen for RDE-3-interacting proteins identified MUT-15, which was the major focus of this study. The analysis also identified UAF-1, the largest subunit of the U2AF splicing complex, as a candidate RDE-3 interactor. The predicted RDE-3:UAF-1 interaction could be an erroneous prediction by Alphafold or, perhaps, indicative of a surprising connection between the pUGasome and pre-mRNA splicing. Asking if UAF-1 and RDE-3 interact in vivo will help distinguish these possibilities.
We speculate that the pUGasome assembles in intestinal cells to promote specificity, processivity, and/or fidelity of RNA pUGylation. During RNAi in C. elegans, dsRNA is processed into primary small RNAs by the RNase-III-like enzyme Dicer (14). The Argonaute protein RDE-1 binds these primary small RNAs to regulate other cellular RNAs via the recruitment of the endonuclease RDE-8, which cleaves mRNAs targeted by RDE-1 (14,15,18). RDE-3 appends pUG tails to RDE-8 cleaved RNAs, and RdRPs then amplify gene silencing responses by using pUGylated RNAs as templates for secondary siRNAs, which engage WAGO proteins to further silence homologous RNAs (25). We find that the components of the pUGasome are required for both RNAi and viral RNA pUGylation, suggesting that the logic of viral pUGylation is, at least partly, similar to that of canonical RNAi-based gene silencing in C. elegans. The data support a model in which Dicer processed primary small RNAs, derived from viral dsRNAs produced during viral replication, recruit the pUGasome to single-stranded viral RNAs to enable efficient cleavage (RDE-8) and pUGylation (RDE-3) of these RNAs. More specifically, pUGasome assembly may allow the 3′ termini of viral RNAs, generated by RDE-8-based endonu cleolytic cleavage, to be brought into close proximity with RDE-3 to enable efficient pUGylation. Indeed, the Alphafold-predicted pUGasome shows a groove of positively charged surface amino acids connecting the RDE-8 and RDE-3 active sites (Fig. S13A andB). Because positively charged protein surfaces often interact with negatively charged nucleic acids, we speculate that RNA, which is cleaved by RDE-8, is threaded along this positively charged groove to link RNA cleavage with RNA pUGylation (Fig. S13A through C). The translocation of RNA along this positively charged groove from RDE-8 to RDE-3 may be facilitated by helicases, such as RDE-12 and MUT-14 (50)(51)(52). Finally, RDE-3 is remarkable, in that it appends alternating uridine and guanosine nucleotides to the 3′ termini of RNAs (25,26). Thus, the RDE-3 active site must alternatively recognize and accommodate either uridine or guanosine nucleotides, and these interactions must somehow be coupled to the identity of the last nucleotide added to RNA by RDE-3. It is possible that MUT-15, or MUT-15 in conjunction with the larger pUGasome, could help facilitate the complex structural rearrangements needed for RDE-3 to accomplish this remarkable feat.
MUT-16 is a large (1,054 a.a.), low-complexity protein that co-IPs with the subunits of the pUGasome (18,(29)(30)(31). AlphaFold predicts a direct interaction between MUT-16 and MUT-15, which likely explains why MUT-16 co-IPs with pUGasome components. Interestingly, we find that MUT-16 is not needed for antiviral RNA pUGylation. Rather, MUT-16 is needed for antiviral siRNA production. Current models posit that the pUG tails appended to RNAs by RDE-3 interact with RdRP and that this interaction allows RdRP to use pUGylated RNAs as templates to synthesize siRNAs (25). MUT-16 also co-IPs with RdRP (18,31). Therefore, we hypothesize that a primary role of MUT-16 during antiviral immunity is to bring RdRP, and likely other factors such as MUT-7, into proximity with the pUGasome so that pUG RNAs can be efficiently converted into the secondary siRNAs, which are loaded onto Argonaute proteins to drive antiviral immunity. Consistent with this model, we find that RdRP is also not needed for pUGylation but is needed for antiviral siRNA synthesis. Previous studies showed that the C-terminus of MUT-16 is needed for MUT-16 to nucleate Mutator foci in germ cells and for in vitro coacervation (29,30,53). These observations hint that multivalent interactions between the C-termini of MUT-16 could underlie the assembly of Mutator foci in germ cells, perhaps via a liquid-liquid phase separation-like process. MUT-16-labeled foci are not observed in the intestinal cells of C. elegans, which are the site of antiviral pUG/siRNA biogenesis, suggesting that either MUT-16 oligomerization is a germline-specific function of MUT-16 or that oligomerization also occurs in the soma, but not to a level detectable by light microscopy (29,30). Asking whether the C-terminus of MUT-16 is needed for MUT-16 to link antiviral pUG synthesis to antiviral siRNA synthesis should distinguish these possibilities. Finally, the terminal uridyltransferase CDE-1 modifies Orsay viral RNA to promote exosome-mediated viral RNA decay in C. elegans (37). Future studies investigat ing potential crosstalk between the CDE-1 and RDE-3 3' viral RNA modification systems could reveal additional insights into antiviral immunity in C. elegans.
## MATERIALS AND METHODS
## Strains
All C. elegans strains were grown at 20°C, derived from the Bristol N2 strain, and grown on normal growth medium with OP50 bacteria, unless otherwise stated. Strains used in this study are listed in Table 1. All the sequences for oligonucleotides, gRNAs, and HR templates are listed in Table 2.
## RNA extraction
Animals were collected in TRIzol reagent and freeze-thawed at least three times. RNA was extracted with isoamyl alcohol and chloroform twice with DNase I digestion or with RNA clean and concentrator with on-column DNase I digestion to remove any residual DNA contaminants (Zymo Research, R1014). RNA was quantified using a NanoDrop 2000 to determine the concentration.
## CRISPR-Cas9
Gene editing preparation in C. elegans was performed as previously described (54). Homology repair templates were ordered from IDT as Alt-R HR oligos, if possible. Custom sgRNAs were designed for genes of interest using Benchling. The designed sgRNAs were validated using IDT's custom gRNA tool. CRISPR-Cas9 complex was assembled as previously described (55). Adult animals were injected in the germline, and F1s were phenotyped for the co-injection marker of an extrachromosomal array (rol-6). F2s were singled, and non-roller animals were maintained and genotyped. All the edited animals were validated with Sanger sequencing and outcrossed at least once.
## pUG-Seq library preparation
Spike-in RNA (gfp(UG)18) was synthesized using the Megascript T7 kit (Invitrogen, AM1334) and purified using the RNA Clean Concentrator (Zymo Research, R1014). RNAseOUT (Invitrogen, 10777019) was added with RNA to prevent unwanted degrada tion. A combination of 10 ug of total RNA and 100 pg of spike-in RNA was depleted of rRNA using a previously established RNAse-H based method (Gallucci et al). The rRNA-depleted RNA was purified using RNA clean and concentrator (Zymo Research, R1014). The eluted RNA was used for reverse transcription with Maximus H Minus Reverse Transcriptase (Thermo Fisher, EP0742), Strand Switching Primer II (SSP II), and a poly(AC) RT primer according to the Oxford Nanopore cDNA-PCR barcoding protocol. To remove unwanted fragments from the sequencing library, the newly generated cDNA was initially amplified with 2X Longamp Taq Polymerase (NEB, M0287L) for eight cycles to generate dsDNA. DASH was performed to remove any yrn-1 and RT primer-SSP II artifacts. Briefly, Cas9-gRNA complex (IDT) was prepared exactly as previously described (56,57) and pre-incubated for 15 min at 25°C. The incubated complex was used to digest the cleaned dsDNA library by incubating for 1 hour at 37°C. The digested samples were purified using a select-a-size concentrator (Zymo Research, D4080) to deplete fragments < 300 nts and eluted with nuclease-free water. The DASH-treated samples were barcoded according to the Oxford Nanopore cDNA-PCR barcoding protocol. The final library was analyzed with Tapestation (Agilent) before proceeding to sequencing.
## Nanopore sequencing
The library was pooled and prepared according to Nanopore protocol. Briefly, the pooled library was loaded into MinION r10 flow cells on a MinION Mk1C machine (Oxford Nanopore). Sequencing was set to run for at least 24 hours, barcoding was turned on with no trimming, and quality score filter of 9. The Guppy high accuracy base-calling model was used.
## Data preprocessing and transcript quantification
Fastq files that passed the quality score filter were in the fastq_pass folder. The fastq files were merged using cat *.fastq. Next, pychopper was used to identify and orient full-length transcripts. The PCS111_primers.fas was modified as follows: >VNP TTGCCTGTCGCTCTATCTTC >SSP TCTGTTGGTGCTGATATTGCTTT Pychopper was used with the following settings: -m edlib -p -k PCS111. The output fastq files were then merged and filtered for TG repeats using the following: grep -B1 -A2 --no-group-separator -E .GTGTGTGTGT.{1\}
Then, the filtered fastq files were used to map to the transcriptome, spike-in, and the Orsay virus using minimap2 -uf k14. The resulting sam files were converted to bam files that were indexed and sorted using samtools. Salmon was used to generate a quant.sf file. The quantification file was imported into R using tx2import. The libraries were normalized to RPM using the spike-in gfp pUG RNAs to normalize the libraries for any technical variations from the library generation. To analyze visually, reads that were generated from mapping the fastq files to the genome were loaded onto the integrated genome viewer or IGV (Broad Institute). Sam files were processed using pysam to analyze the soft clipped sequences that correspond to the pUG tails as well as identifying the location of pUG tail sites. Frequency of pUG tails across the viral genome (binwidth = 50) was generated using ggplot2 in R.
## Computational predictions
All protein sequences were identified from the UniProt database. To identify conserved germline proteins, we first started with a data set of germline-specific transcripts identified through SAGE (58), which was a total of 1,063 genes. We then used BioMart on the Wormbase ParaSite to select from the 1,063 germline-specific genes that are conserved in humans. We then selected 100 conserved germline-specific genes without overlapping paralogs when possible. In addition, we added Mutator proteins identified in MUT-16 IP-MS (29,31,50). The protein sequences of the final list of genes to test for RDE-3 interaction were extracted from UniProt; if there were multiple isoforms, then the longest isoform was chosen for the initial predictions. AlphaFold 2 was used to perform protein:protein interactions with the following parameters: 5 models and 5 cycles (paste the code). To analyze the outputs, we used a Python script to summarize the predictions (45). To generate the graphs of the in silico predictions, we used the avg_n_models per contact and average ipTM scores with ggplot2 in R. To create the protein structure prediction figures, the top ranked PDB file was loaded onto Pymol. The pLDDT scores were used to identify strongly and poorly predicted regions of the protein:protein interaction.
## Immunoprecipitation
Animals were mixed with 1× RIP Buffer (20 mM Tris-HCl pH 7.5, 200 mM NaCl, 2.5 mM MgCl2, 10% glycerol, 0.5% NP-40, 80 U mL -1 RNaseOUT, 1 mM dithiothreitol [DTT] and protease inhibitor cocktail without EDTA) and frozen dropwise with liquid nitrogen to generate frozen worm balls. The frozen worm balls were ground with a mortar and pestle before resuspending in the same buffer as collection. The ground lysate was clarified by a centrifuge running at 15,000 × g for 10 mins in 4°C. The supernatant was resuspended into a new tube. The protein lysate was measured for abundance with a BCA assay (Thermo Fisher, 23225). For co-immunoprecipitation assays, anti-FLAG M2 magnetic beads (Thermo Fisher, A36797) were used to immunoprecipitate 3X FLAG::RDE-3 from ~ 300-400 ug of the protein lysate. Co-immunoprecipitated proteins were eluted with 0.1M glycine pH 3.5 and 1M Tris-HCl pH 8.0 for neutralization and mixed with Laemmli buffer prior to boiling for loading. Roughly the same amount of proteins is loaded for input and IP, unless otherwise stated into a 10% mini-Protean TGX precast SDS-PAGE gels (Biorad, 4561033). The gels were semi-dry transferred using 0.2 uM nitrocellulose Trans-blot Turbo (Biorad, 1704157) and the Trans-blot Turbo Transfer System (Biorad). The membranes were blocked for 1 hour with the Intercept (TBS) blocking buffer (Licor, 927-60003) at room temperature on a shaker. Primary antibodies were incubated overnight at 4°C on a shaker and then washed four times with 1× TBS-T. The following primary antibodies were used: mouse anti-FLAG (Sigma, F3165, diluted 1:2000), rabbit anti-HA (Abcam, C29F4, diluted 1:2000), and rabbit anti-V5 (Cell Signaling Tech, D3H8Q, diluted 1:1000). Secondary antibodies were incubated for at least 1 hour at room temperature and then washed four times with 1× TBS-T prior to imaging on a Licor Odyssey Fx. The following secondary antibodies were used: anti-rabbit IR-680 (Licor, 1:10000) and anti-mouse IR-800 (Licor, 1:10000).
## Orsay virus filtrate preparation and infection assays
Orsay virus filtrate was prepared according to previous publications (59). Briefly, rde-1(ne219) animals infected with Orsay virus were grown for several generations before collecting starved animals with an M9 buffer. The animals were incubated in the buffer for at least 30 min. Then, the supernatant was collected and isolated from the pellet containing the starved animals. The virus-containing supernatant was plunged through a 22 gage needle and then isolated into 1 mL virus aliquots to avoid freeze/thaw cycles. The virus aliquots were diluted 1:1000 with 10× OP50 to seed infection plates. Animals were infected by egg prepping onto the Orsay virus-containing plates. Animals were collected at adulthood about 3-4 days post-infection. To infect animals with auxin-medi ated depletion of RDE-3 in various tissues, 1 mM auxin plates were seeded with virus, and various strains were egg prepped onto the infection plates for 3 days prior to collection.
## qRT-PCR assay
Total RNA was reverse transcribed with the Superscript IV kit (Thermo Fisher) according to the manufacturer's protocol. Random hexamer was used for quantifying total viral RNA levels from 1 ug of total RNA. A poly(AC)12 with 5′ template sequence encoding an amplification sequence was used to detect pUG RNAs from 5 ug of total RNA. An oligo(dT)20 was used to reverse-transcribe polyadenylated mRNAs from 1 ug of total RNA. cDNA diluted 1:10 was mixed with SYBR Green Master Mix (Biorad) and primers for qPCR according to the manufacturer's recommendation in a CFX Connect machine (Biorad). The CT values were used to calculate relative expression levels using the -2 ΔΔ method. To quantify viral RNA levels across samples, primers targeting ORV RNA1 (ORV1), ORV RNA2 (ORV2), cdc-42, and eft-2 were used. To quantify pUG RNAs, gsa-1 primers were used as control. Further clarification for pUG RNAs is as follows: The reverse-tran scription primers contain adapter sequences amenable for downstream nested PCR. To detect Orsay virus pUG RNAs, two rounds of nested PCR were performed. First, a primer mapping to the Orsay virus sequence was amplified in conjunction with a primer that maps to the sequence of the RT primer. The first PCR was treated with Exo I to remove any unused primers. Then, a second PCR was performed with a second primer mapping to the Orsay virus sequence with a different primer mapping to the RT primer for gel-based detection or with qPCR primers as above for quantification. Similar methods were used to detect the control gsa-1 genome-encoded pUG RNAs.
## TaqMan assay
Trizol-extracted RNA (1 ug) from infected animals (see above) was used for TaqMan assays. Small RNAs were reverse-transcribed into cDNA using the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems, 4366596). U18, antiviral siRNA1, and antiviral siRNA2 small RNAs were then quantified by qRT-PCR using TaqMan Universal Master Mix II, no UNG (Applied Biosystems, 4440040), and custom TaqMan small RNA assays from Applied Biosystems (assay IDs: U18: 001764, Orsay virus RNA1 siRNA1: CTU63TC, Orsay virus RNA1 siRNA2: CTWCXC9). qRT-PCRs were performed using the CFX Connect machine (Bio-Rad) and semi-skirted PCR plates (Bio-Rad, 2239441). No Orsay virus siRNAs were detected in non-infected animals using the custom TaqMan small RNA assays.
## Feeding RNAi assay
RNAi clones were from the Ahringer Library (Source Biosciences). RNAi clones were grown on the appropriate antibiotic. Before seeding, IPTG was added to the RNAi clone and incubated on a shaker at 37°C for 1 hour. The incubated RNAi clones were 2× concentrated and then seeded onto IPTG plates to express dsRNA. Animals were egg prepped onto the appropriate RNAi plates. L4440 was used as a negative control. For dpy-6 RNAi, after 3 days of incubation, the plates were inspected and counted for a detectable dumpy phenotype with a light microscope (Nikon). The counts were analyzed and displayed using GraphPad (Prism).
## Auxin-mediated degradation
Animals containing degron-tagged RDE-3 with TIR1 expression in either the soma, intestine, or muscle were egg prepped onto 1 mM auxin plates. Animals were grown for 4 days before collection for imaging. Animals were washed off plates using M9 + Triton X-100 buffer, which were washed with M9 buffer. Sodium azide (0.1%) was added before mounting onto slides with freshly made 2% agarose pad for imaging. For the RNAi assay with auxin, 1 mM auxin plates containing IPTG were seeded with L4440 or dpy-6 RNAi clones from the Ahringer Library (Source Bioscience). Animals were egg prepped onto the corresponding auxin + RNAi plates, and animals were collected after 4 days of incubation. All images were taken using Axio Observer.Z1 fluorescent microscope (Zeiss) with the Plan-Apochromat 20×/0.8 M27 objective.
## Quantifications and statistical analysis
All statistics were performed using GraphPad (Prism). The P-values and comparisons relevant to the text are shown in the figures. All descriptions of the statistical analysis are described in the figure legends.
## References
1. ":mRuby::unc-54 3'UTR + Cbr-unc-119(+)] II Muscle TIR1 expression with GFP::AID::RDE-3 YY2672 rde-3(gg693) I; ieSi57 [ges-1p::TIR1::mRuby::unc-54 3'UTR + Cbr-unc-119(+)] II Intestinal"
2. "X HA, and 3X V5::NYN-2 YY2511 rde-3(gg636); mut-15(gg1066) I nyn-1(gg1052) IV 3X FLAG::RDE-3, 3X V5::NYN-2, and MUT-15::3 X HA with inverted charge mutations YY2513 rde-3(gg636); mut-15(gg1067) I nyn-1(gg1052) IV 3X FLAG::RDE-3, MUT-15::3 X HA with R159TOP mutation"
3. "X HA with R159STOP mutation, and 3X V5::NYN-2 YY2555 rde-3(gg636); mut-15(gg1067) I rde-8(gg1051) IV 3X FLAG::RDE-3, MUT-15::3 X HA with R159STOP mutation, and 3X V5::RDE-8 YY1689 ego-1(gg685) rrf-1(pk1417) I Null alleles for both host RdRPs, which are in an operon GR1747 mut-15(tm1358) I Deletion of MUT-15 YY1803 nyn-1(tm5004) I; nyn-2(tm4844) rde-8(tm2252) IV Deletion of RDE-8, NYN-1, and NYN-2 YY1740 nyn-1(tm5004) I; nyn-2(tm4844) IV Deletion of NYN-1 and NYN-2 YY1739 rde-8(tm2252) IV Deletion of"
4. (2025) *Full-Length Text Journal of Virology*
5. Wilkins, Dishongh, Moore et al. (2005) "RNA interference is an antiviral defence mechanism in Caenorhabditis elegans" *Nature*
6. Tassetto, Kunitomi, Andino (2017) "Circulating immune cells mediate a systemic RNAi-based adaptive antiviral response in Drosophila" *Cell*
7. Guo, Zhang, Wang et al. (2013) "Homologous RIG-I-like helicase proteins direct RNAi-mediated antiviral immunity in C. elegans by distinct mechanisms" *Proc Natl Acad Sci*
8. Wang, Wu, Ito et al. (2010) "RNAi-mediated viral immunity requires amplification of virus-derived siRNAs in Arabidopsis thaliana" *Proc Natl Acad Sci*
9. Ying, Dong, Zhu et al. (2010) "RNA-dependent RNA polymerase 1 from Nicotiana tabacum suppresses RNA silencing and enhances viral infection in Nicotiana benthamiana" *Plant Cell*
10. Xie, Fan, Chen et al. (2001) "An important role of an inducible RNA-dependent RNA polymerase in plant antiviral defense" *Proc Natl Acad Sci*
11. Nakahara, Masuta, Yamada et al. (2012) "Tobacco calmodulin-like protein provides secondary defense by binding to and directing degradation of virus RNA silencing suppressors" *Proc Natl Acad Sci*
12. Maillard, Ciaudo, Marchais et al. (2013) "Antiviral RNA interference in mammalian cells" *Science*
13. Li, Lu, Han et al. (2013) "RNA interference functions as an antiviral immunity mechanism in mammals" *Science*
14. (2025) *Full-Length Text Journal of Virology*
15. Maillard, Van Der Veen, Poirier et al. (2019) "Slicing and dicing viruses: antiviral RNA interference in mammals" *EMBO J*
16. Lu, Maduro, Li et al. (2005) "Animal virus replication and RNAi-mediated antiviral silencing in Caenorhabditis elegans" *Nature*
17. Ketting, Fischer, Bernstein et al. (2001) "Dicer functions in RNA interference and in synthesis of small RNA involved in developmental timing in C. elegans" *Genes Dev*
18. Pak, Fire (2007) "Distinct populations of primary and secondary effectors during RNAi in C. elegans" *Science*
19. Tabara, Yigit, Siomi et al. (2002) "The dsRNA binding protein RDE-4 interacts with RDE-1, DCR-1, and a DExH-box helicase to direct RNAi in C. elegans" *Cell*
20. Tabara, Sarkissian, Kelly et al. (1999) "The rde-1 gene, RNA interference, and transposon silencing in C. elegans" *Cell*
21. Parrish, Fire (2001) "Distinct roles for RDE-1 and RDE-4 during RNA interference in Caenorhabditis elegans" *RNA*
22. Yigit, Batista, Bei et al. (2006) "Analysis of the C. elegans argonaute family reveals that distinct argonautes act sequentially during RNAi" *Cell*
23. Tsai, Chen, Conte D Jr et al. (2015) "A ribonuclease coordinates siRNA amplification and mRNA cleavage during RNAi" *Cell*
24. Aoki, Moriguchi, Yoshioka et al. (2007) "In vitro analyses of the production and activity of secondary small interfering RNAs in C. elegans" *EMBO J*
25. Vasale, Gu, Thivierge et al. (2010) "Sequential rounds of RNAdependent RNA transcription drive endogenous small-RNA biogenesis in the ERGO-1/Argonaute pathway" *Proc Natl Acad Sci*
26. Buckley, Burkhart, Gu et al. (2012) "A nuclear argonaute promotes multigenera tional epigenetic inheritance and germline immortality" *Nature*
27. Spracklin, Fields, Wan et al. (2017) "The RNAi inheritance machinery of Caenorhabditis elegans" *Genetics*
28. Gu, Shirayama, Conte et al. (2009) "Distinct argonaute-mediated 22G-RNA pathways direct genome surveillance in the C. elegans germline" *Mol Cell*
29. Burton, Burkhart, Kennedy (2011) "Nuclear RNAi maintains heritable gene silencing in Caenorhabditis elegans" *Proc Natl Acad Sci*
30. Shukla, Yan, Pagano et al. (2020) "Poly(UG)-tailed RNAs in genome protection and epigenetic inheritance" *Nature*
31. Preston, Porter, Chen et al. (2019) "Unbiased screen of RNA tailing activities reveals a poly(UG) polymerase" *Nat Methods*
32. Chen, Simard, Tabara et al. (2005) "A member of the polymerase β nucleotidyltransferase superfamily is required for RNA Interference in C. elegans" *Curr Biol*
33. Roschdi, Yan, Nomura et al. (2022) "An atypical RNA quadruplex marks RNAs as vectors for gene silencing" *Nat Struct Mol Biol*
34. Phillips, Montgomery, Breen et al. (2012) "MUT-16 promotes formation of perinuclear mutator foci required for RNA silencing in the C. elegans germline" *Genes Dev*
35. Uebel, Anderson, Mandarino et al. (2018) "Distinct regions of the intrinsically disordered protein MUT-16 mediate assembly of a small RNA amplification complex and promote phase separation of mutator foci" *PLoS Genet*
36. Manage, Rogers, Wallis et al. (2020) "A tudor domain protein, SIMR-1, promotes siRNA production at piRNA-targeted mRNAs in C. elegans" *Elife*
37. Zhang, Montgomery, Gabel et al. (2011) "Mut-16 and other mutator class genes modulate 22G and 26G siRNA pathways in Caenorhabditis elegans" *Proc Natl Acad Sci*
38. Sijen, Plasterk (2003) "Transposon silencing in the Caenorhabditis elegans germ line by natural RNAi" *Nature*
40. Félix, Wang (2019) "Natural viruses of Caenorhabditis nematodes" *Annu Rev Genet*
41. Jiang, Franz, Wang (2014) "Engineering recombinant orsay virus directly in the metazoan host Caenorhabditis elegans" *J Virol*
42. Félix, Ashe, Piffaretti et al. "2011a. Natural and experimental infection of Caenorhabditis nematodes by novel viruses related to nodaviruses" *PLoS Biol*
43. Pen, Jiang, Domenico et al. (2018) "Terminal uridylyltransferases target RNA viruses as part of the innate immune system" *Nat Struct Mol Biol*
45. Guo, Fan, Zhou et al. (2020) "Orsay virus cp-δ adopts a novel βbracelet structural fold and incorporates into virions as a head fiber" *J Virol*
46. Franz, Renshaw, Frezal et al. (2014) "Orsay, Santeuil and Le Blanc viruses primarily infect intestinal cells in Caenorhabditis nematodes" *Virology (Auckl)*
47. Zhang, Ward, Cheng et al. (2015) "The auxin-inducible degradation (AID) system enables versatile conditional protein depletion in C" *elegans. Development (Rome)*
48. Ashley, Duong, Levenson et al. (2021) "An expanded auxin-inducible degron toolkit for Caenorhabditis elegans" *Genetics*
49. Ashe, Bélicard, Pen et al. (2013) "A deletion polymorphism in the Caenorhabditis elegans RIG-I homolog disables viral RNA dicing and antiviral immunity" *Elife*
50. Long, Meng, Lu (2018) "Transgene-assisted genetic screen identifies rsd-6 and novel genes as key components of antiviral RNA interference in Caenorhabditis elegans" *J Virol*
51. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with AlphaFold" *Nature*
52. Lim, Tamayo-Orrego, Schmid et al. (2023) "In silico protein interaction screening uncovers DONSON's role in replication initiation" *Science*
53. Guo, Zhang, Wang et al. (2013) "Antiviral RNA silencing initiated in the absence of RDE-4, a double-stranded RNA binding protein, in Full-Length Text Journal of Virology November"
54. "Caenorhabditis elegans" *J Virol*
55. Thivierge, Makil, Flamand et al. (2011) "Tudor domain ERI-5 tethers an RNAdependent RNA polymerase to DCR-1 to potentiate endo-RNAi" *Nat Struct Mol Biol*
56. Rocheleau, Cullison, Huang et al. (2008) "The Caenorhabditis elegans ekl (enhancer of ksr-1 lethality) genes include putative components of a germline small RNA pathway" *Genetics*
57. Chen, Wang, Mufti et al. (2024) "Germ granule compartments coordinate specialized small RNA production" *Nat Commun*
58. Phillips, Montgomery, Breen et al. (2014) "MUT-14 and SMUT-1 DEAD box RNA helicases have overlap ping roles in germline RNAi and endogenous siRNA formation" *Curr Biol*
59. Shirayama, Stanney, Iii et al. (2014) "The vasa homolog RDE-12 engages target mRNA and multiple argonaute proteins to promote RNAi in C. elegans" *Curr Biol*
60. Yang, Vallandingham, Shiu et al. (2014) "The DEAD box helicase RDE-12 promotes amplification of RNAi in cytoplas mic foci in C. elegans" *Curr Biol*
61. Busetto, Pshanichnaya, Lichtenberger et al. (2024) "MUT-7 exoribonuclease activity and localization are mediated by an ancient domain" *Nucleic Acids Res*
62. Dokshin, Ghanta, Piscopo et al. (2018) "Robust genome editing with short single-stranded and long, partially single-stranded DNA donors in Caenorhabditis elegans" *Genetics*
63. Paix, Folkmann, Rasoloson et al. (2015) "High efficiency, homology-directed genome editing in Caenorhabditis elegans using CRISPR-Cas9 ribonucleoprotein complexes" *Genetics*
64. Mahat, Tippens, Waterton et al. (2024) "Single-cell nascent RNA sequencing unveils coordinated global transcription" *Nature*
65. Gu, Crawford, 'donovan et al. (2016) "Depletion of abundant sequences by hybridization (DASH): using Cas9 to remove unwanted high-abundance species in sequencing libraries and molecular counting applications" *Genome Biol*
66. Wang, Zhao, Wong et al. (2009) "Identification of genes expressed in the hermaphrodite germ line of C. elegans using SAGE" *BMC Genomics*
67. Sowa, Jiang, Somasundaram et al. (2020) "The Caenorhabditis elegans RIG-I homolog DRH-1 mediates the intracellular pathogen response upon viral infection" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC9175619&blobtype=pdf | # EDITOR IN CHIEF (2022)
Rozanne Sandri-Goldin, J.-H James Ou, Theodore Pierson, Viviana Simon, Hector Aguilar, Sabra Klein
## References
1. Board, Abendroth
2. Aguilar
3. Ahn "22) Tero Ahola"
4. Aiken
5. Albrecht
6. "Lorraine Albritton (' 22) José Alcami"
7. Amarasinghe
8. Apetrei
9. Arias
10. "James Arthos (' 24) Ann Arvin"
11. Asokan
12. Walter, Atwood
13. Baines
14. Baker
15. Banfield "22) Eric Barklis"
16. Barouch
17. Barrett
18. Barry
19. Bartenschlager
20. Barton
21. Bartz
22. Basler
23. Belov
24. Benmohamed
25. Bergelson
26. Bergmann, Berkhout
27. Bertke "24) Sonja Best"
28. Blankson
29. Blissard
30. David, Bloom
31. Boehme
32. Boris-Lawrie
33. Brenchley
34. Bresnahan
35. Brinton
36. Britt
37. Broder "24) Jay Brown"
38. Brune
39. Buchmeier
40. Christopher, Buck, Bukreyev
41. Michael, Bukrinsky, Byrnes
42. Cameron
43. Campadelli-Fiume
44. Campos
45. Cattaneo
46. Chahroudi
47. Chakrabarti, Chang
48. Chao
49. Chapman
50. Chattopadhyay
51. Chelico
52. Chen
53. Chinchar
54. Chomont
55. Clapham
56. "François Clavel (' 24)"
57. Cohen
58. Collman
59. Connor
60. Conzelmann
61. Coulson
62. Cullen
63. Damania
64. Danthi "22) Miles Davenport"
65. Davido "23) Victor DeFilippis"
66. Carlos De La
67. Mark, Denison
68. Derdeyn
69. Aravinda, De Silva
70. Desselberger
71. Dimaio "22) Shou-Wei Ding (' 23) Dirk Dittmer (' 22) Valerian Dolja"
72. Domingo, Doorbar
73. Dorfman
74. Doria
75. Philip, Dormitzer "22) Christian Drosten"
76. Dubuisson "23) Jaquelin Dudley (' 24) Breck Duerkop"
77. Dustin "22) Gregory Ebel"
78. Elder
79. Engelman
80. Enquist
81. David, Evans
82. Matthew, Evans
83. Oliver, Fackler
84. Falck-Pedersen, Bentley Fane
85. Paul, Farrell, Michael Farzan
86. Favoreel
87. Heinrich, Feldmann "23) Pinghui Feng (' 24) Zongdi Feng"
88. Fernandez-Sesma
89. "Svetlana Folimonova (' 24) James Craig Forrest (' 23) Steven Foung"
90. Freed
91. Matthew, Frieman "23) Shou-Jiang Gao (' 23) Robert Garcea (' 24) Fernando Garcia-Arenal"
92. Geballe "22) Chou-Zen Giam (' 24) Paul Goepfert"
93. Golovkina "23) Alexander Gorbalenya (' 24) Heinrich Göttlinger (' 23) Urs Greber"
94. Grose
95. Grubman
96. Grdzelishvili "22) Haitao Guo (' 24) Ju-Tao Guo"
97. Scott, Halstead
98. Harhaj
99. Harris
100. Harty "24) Biao He (' 22) Bin He"
101. Hearing
102. Heldwein "24) Scott Hensley (' 24) Alon Herschhorn"
103. Hirsch
104. Hiscott
105. Hogue
106. Holm
107. Holmes
108. Hoover
109. Stacy, Horner
110. Hughes
111. Iorio
112. Isaacs, Jacobs
113. Michael, Jarvis ; Dong-Yan
114. Johnson
115. Johnson "24) Clinton Jones (' 22) Robert Kalejta (' 24) Jeremy Kamil"
116. Karst, Kashanchi
117. Katz "22) Yoshihiro Kawaoka (' 22) Kenneth Kaye (' 24) Brandon Keele (' 24) Shannon Kenney"
118. Kent
119. Khanna "23) Alexander Khromykh (' 22) Margaret Kielian"
120. Paul, Kinchington, Knipe
121. Konan
122. Kousoulas
123. Kowalik
124. Orkide, Koyuncu
125. Krammer
126. Kristie
127. Krummenacher
128. Kuhn
129. Kuhn
130. Kulpa
131. Sebla, Kutluay "22) Michael Lai (' 24) Laimonis Laimins"
132. Lamb "24) Ke Lan (' 23) Nathaniel Landau"
133. Langlois "24) Benhur Lee"
134. Leib "23) Laura Levy (' 22) Renfeng Li"
135. Liang
136. Paul, Lieberman
137. Lindenbach, Maxine, Linial *22)*
138. Liu
139. Lloyd
140. Loeb
141. Volker Lohmann, Shan
142. Luftig
143. Guangxiang, Luo "22) Min-Hua Luo" *24)*
144. Lyles
145. Macdonald
146. Mackenzie
147. (2022) *Journal of Virology jvi.asm.org i* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12646006&blobtype=pdf | # The neddylation of the RNA-dependent RNA polymerase 3D of Coxsackievirus B3 promotes viral replication
Siwei Li, Yanyan Dong, Xuexuan Wang, Danxiang Feng, Tian Luan, Ziyuan Wang, Lexun Lin, Yang Chen, Yao Wang, Yanru Fei, Yan Wang, Zhaohua Zhong, Wenran Zhao
## Abstract
The pathogenesis of Coxsackievirus B3 (CVB3), the causative pathogen for severe diseases such as myocarditis, pancreatitis, and meningitis, remains largely unknown. Neddylation, the covalent modification of proteins through the conjugation of NEDD8, a ubiquitin-like molecule, plays critical roles in regulating cellular activities. Our previous study demonstrated that 3D pol , the RNA-dependent RNA polymerase of CVB3, was modified by ubiquitin. Since NEDD8 is highly homologous to ubiquitin in both the sequence and structure, we hypothesized that 3D pol may also undergo neddylation. Here, we demonstrated that 3D pol of CVB3 is modified by NEDD8. Proteo mics and immunoprecipitation studies identified that NEDD8 expression was upregu lated in CVB3-infected cells, and 3D pol was neddylated. Through the overexpression or knockdown of the subunit of NEDD8-activating enzyme or NEDD8, we show that neddylation promoted CVB3 replication. Furthermore, the overexpression of NEDD8-spe cific protease 1 completely blocked 3D pol neddylation and suppressed viral replication. The neddylation sites of 3D pol , K261, and K457 were identified by mutagenesis studies. We further showed that neddylation enhanced 3D pol stability. In contrast, when the neddylation sites were mutated (K261R, K457R, or K261/457R), the 3D pol degradation rate was significantly accelerated. Moreover, the growth kinetics of the recombinant viruses containing the mutated 3D pol at the neddylation sites was markedly decreased. Through analyzing the 3D pol interacting proteins with mass spectrometry, we identified that TRIM4 is the E3 ligase, which upregulates 3D pol neddylation. TRIM4 knockdown markedly suppressed viral replication and almost completely blocked 3D pol neddylation. In contrast, TRIM4 overexpression promoted CVB3 replication. Taken together, this study demonstrated that 3D pol neddylation facilitates CVB3 replication. IMPORTANCE CVB3 infection is commonly related to the inflammatory disease of the heart, which may develop to dilated cardiomyopathy and heart failure. Neddylation is a process in which the ubiquitin-like molecule NEDD8 is covalently linked to the specific lysine residues of the target proteins. Increasing evidence has shown that the neddyla tion of either host or viral proteins may alter viral replication. In this study, we demonstra ted that 3D pol , the RNA-dependent RNA polymerase of CVB3, is modified by NEDD8 at its lysine residues 261 and 457, and the neddylation process is mediated by the E3 ligase TRIM4. Neddylation enhances the stability of 3D pol and facilitates viral replication, while viruses with mutated 3D pol , which cannot be neddylated, showed decreased replication capacity. Our findings not only add novel insights for understanding the pathogenesis of CVB3 but also identify that targeting neddylation could be a potential antiviral strategy for the treatment of CVB3 infection.
S mall RNA viruses are widespread in nature, and some of them, such as enterovirus A71 (EV-A71), Coxsackievirus group A (CVA), or Coxsackievirus group B (CVB), can cause severe illnesses. CVBs are single-stranded, positive-sense, non-enveloped RNA viruses, which belong to the Enterovirus genus of Picornaviridae family (1). There are six known serotypes of CVB (CVB1 ~6), which can cause a variety of diseases ranging from the common cold to viral myocarditis, pancreatitis, and meningitis (2,3). CVB1, CVB3, and CVB5 are cardiophilic in nature (4), while CVB3 has been shown as the primary causative pathogen of myocarditis, dilated cardiomyopathy, and even heart failure (5)(6)(7).
CVB3, with a genome of 7.4 kb (8), consists of a single open reading frame (ORF) and two untranslated regions (UTRs) located at the ends of the viral genome, the 5'UTR and 3'UTR (1,9). The ORF is translated into a single polyprotein, which is cleaved by viral proteases 2A pro , 3C pro , and 3CD pro into four structural proteins (VP1-4) and seven nonstructural proteins (2A pro , 2B, 2C, 3A, 3B, 3C pro , and 3D pol ) (1,10). 3D pol is the RNAdependent RNA polymerase (RdRp), which mediates viral genomic replication (11,12). To replicate viral RNA, CVB3 uses its genomic RNA, a positive-sense, single-stranded RNA (+ssRNA), as a template to synthesize a complementary negative-stranded RNA (-ssRNA), which is then used as the template to produce the genomic RNA for progeny viruses (13).
The posttranslational modifications (PTMs) of proteins, such as phosphorylation, glycosylation, methylation, and ubiquitination, play critical roles in regulating the stability, enzymatic activity, and subcellular localization of the substrate proteins (14,15). PTMs are also exploited by viruses to facilitate viral replication or to inhibit cellular defense (15). Neddylation is a process of PTMs, in which substrate proteins are covalently modified by neural precursor cell-expressed developmentally downregulated protein 8 (NEDD8), a highly conserved ubiquitin-like protein with 59% amino acid identity to ubiquitin (Ub) (16,17). Similar to ubiquitination, neddylation also requires an enzymatic cascade containing NEDD8-activating enzyme (NAE), NEDD8-conjugating enzyme E2, and substrate-specific NEDD8 E3 ligases (18,19). There is only one NAE (NAE1/UBA3 heterodimer) (20) and two NEDD8-conjugating enzymes (UBC12/UBE2M or UBE2F) (21,22). Mature NEDD8 forms a thioester bond with the cysteine active site in UBA3 in an ATP-dependent manner catalyzed by NAE1 (20,22). Subsequently, NEDD8 is transferred to the cysteine active site of the NEDD8-conjugating enzyme (21,22). Finally, in the presence of E3 ligases, the glycine at the C-terminal of NEDD8 forms an isopeptide bond with the specific lysine (K) residue on the substrate protein (19,23). Unlike ubiquitination, there are a limited number of NEDD8 E3 ligases identified, and all of them are also E3 Ub ligases (24). In addition, like most of the cellular biochemical reactions, neddylation is a reversible process, which is mediated by deneddylases, such as COP9 signalosome complex and NEDD8-specific protease 1 (NEDP1). Through deneddylation, NEDD8 is removed from the target protein, allowing NEDD8 to participate in the next cycle of neddylation (23).
Previous studies showed that neddylation is involved in the life cycle of some viruses including hepatitis B virus (HBV), influenza A virus (IAV), and EV-A71 (25)(26)(27)(28). However, it is unknown whether or not neddylation also participates in CVB infection. Our previous study demonstrated that the 3D pol of CVB3 contains the ubiquitination site at its K220, which upregulates the degradation of 3D pol (29). Since NEDD8 shows high homology to Ub in both the sequence and structure, we hypothesized that 3D pol of CVB3 might also be modified by neddylation.
In this study, we demonstrated that 3D pol of CVB3 is neddylated at K261 and K457. The neddylation of 3D pol is mediated by the E3 ligase tripartite motif 4 (TRIM4). We further show that neddylation of 3D pol increases its stability, which in turn promotes viral replication. Our study adds novel insights for understanding the pathogenesis of CVB3 infection.
## RESULTS
## Viral protein 3D pol is neddylated
To study how CVB3 infection would impair host protein synthesis, we analyzed the protein expression profile of CVB3-infected cells by mass spectrometry (MS). We found that NEDD8 (Gene name: NEDD8-MDP1; Gene ID: 100528064) was one of the proteins that was significantly upregulated in CVB3-infected cells (Fig. 1A; Table S1). To evalu ate the correlation between CVB3 infection and neddylation, the overall neddylated proteins and NEDD8 expression were determined (Fig. 1B andC). We show that CVB3 infection enhanced neddylation and NEDD8 expression (Fig. 1B andC). To show whether individual viral protein exerts the capability to impact neddylation, we determined the expression of NEDD8 in the cells expressing viral mature proteins or viral precursors (Fig. 1D). We show that none of the viral proteins or viral precursor proteins was able to alter NEDD8 expression, suggesting that the upregulated NEDD8 and neddylation is the consequence of viral replication. We further wondered whether upregulation of neddylation is shared by other types of CVB or not. To this end, we determined the neddylation status of the cells infected with CVB5. We show that neddylation was significantly enhanced by CVB5 infection (Fig. 1E), further implicating the importance of the upregulated neddylation in CVB infection.
Our previous study demonstrated that 3D pol of CVB3 is ubiquitinated (29). Since Ub and NEDD8 are highly homologous, we hypothesized that viral 3D pol might be neddylated. To validate this hypothesis, we first observed if there were 3D pol molecules with high molecular weight (MW) when cell lysates were analyzed by immunoblotting. 3D pol of CVB3 is usually migrating at around 58 kDa in the SDS-PAGE gel. With the covalent modification of NEDD8 or Ub, the MW of 3D pol would become much higher than 58 kDa. As shown in Fig. 1F, 3D pol with an MW higher than 180 kDa appeared at 12 h of post-infection (indicated by arrows), while there was no viral 3C pro with an MW higher than 180 kDa. Proteins migrating between >70 and ~110 kDa showed similar patterns revealed by either anti-3C and anti-3D (Fig. 1F), indicating that these proteins contain both 3C pro and 3D pol , which are likely viral precursors derived from P3 (3CD, 3BCD, and 3ABCD). Since 3C pro -containing protein molecules did not appear at high MW (Fig. 1F), we proposed that it is 3D pol rather than 3C pro or the precursor of 3D pol, which is under PTM. To further reveal that 3D pol contains modification during virus infection, cells were infected with CVB3, and 3D pol was analyzed at different time points of post-infection. We found that 3D pol with high MW was accumulating as viral infection continues (Fig. 1G).
To demonstrate whether or not 3D pol is modified by NEDD8, denatured Co-IP was carried out for the cells overexpressing 3D pol (Fig. 1H andI) or infected with CVB3 (Fig. 1J andK). We show that 3D pol of CVB3, when ectopically expressed, was neddylated (Fig. 1H andI). To avoid the artificial effects of NEDD8 overexpression, the neddylation of 3D pol was determined in the natural process of virus infection (Fig. 1K). We demonstrated that 3D pol was neddylated in CVB3-infected cells (Fig. 1J andK). Furthermore, we confirmed that neither 3 CD nor 3C pro was neddylated (Fig. 1L andM). Collectively, these results demonstrate that 3D pol of CVB3 is neddylated.
## Neddylation facilitates CVB3 replication
With the identification that 3D pol is neddylated, we further determined how neddylation would impact viral replication. To this end, CVB3 replication was observed in the cells overexpressing NAE1, the subunit of the NAE, or NEDD8 (Fig. 2A through J). We show that overexpression of NAE1 or NEDD8 obviously led to the increased levels of viral RNA in a dose-dependent manner (Fig. 2A andB). Similarly, levels of viral proteins 3D pol and VP1 were also increased in the cells overexpressing NAE1 or NEDD8 (Fig. 2C through H). Moreover, NAE1 or NEDD8 overexpression significantly increased the virus yield (Fig. 2I andJ). In contrast, knockdown of NAE1 or NEDD8 resulted in markedly decreased viral RNA levels (Fig. 2K andL). Collectively, these data demonstrate that neddylation facilitates CVB3 replication.
## Deneddylation inhibits CVB3 replication
Neddylation is reversed by deneddylation, the process in which NEDD8 is removed from the neddylated substrate protein. NEDP1, a NEDD8-specific protease, mediates the removal of NEDD8 from a wide variety of substrates (30). To further verify that neddyla tion is utilized to enhance CVB3 replication, we determined how deneddylation would impact CVB3 replication. To this end, the neddylation of 3D pol of CVB3 was determined in the cells with or without NEDP1 overexpression. As shown in Fig. 3, overexpression of NEDP1 (Fig. 3A), but not the mutated NEDP1 (without enzymatic activity) (Fig. 3B), completely blocked the neddylation of 3D pol . Moreover, the overexpression of NEDP1, but not the mutated NEDP1, resulted in a significant decrease in viral protein 3D pol and VP1 (Fig. 3C through E). These results further demonstrate that 3D pol of CVB3 is neddyla ted, and neddylation enhances viral replication.
## The 3D pol of CVB3 is neddylated at K261 and K457
With the finding that neddylation of 3D pol facilitates CVB3 replication, we wondered which specific lysine residue in 3D pol is neddylated. Before answering this question, we asked if other species of enteroviruses also contain 3D pol neddylation since the amino acid sequences of 3D pol among enteroviruses share more than 50% homology. To this end, we compared the 3D pol amino acid sequence of CVB3 (GenBank: U57056.1), CV-A16 (GenBank: U05876.1), and EV-A71 (Genbank: U22521.1) (Fig. 4A), and the neddylation status of the 3D pol from these viruses was determined with Co-IP analysis (Fig. 4B). We show that neddylation was identified only in the 3D pol of CVB3, implicating that the neddylation sites of 3D pol of CVB3 are not shared by the 3D pol of either CV-A16 or EV-A71.
According to the amino acid sequence of 3D pol (Fig. 4A), we identified 12 K residues (K22, 51, 96, 133, 134, 196, 250, 255, 261, 409, 436, and 457), which exclusively occur in 3D pol of CVB3. To determine the neddylation site of 3D pol , constructs expressing 3D pol mutations were generated, in which K was mutated to arginine (R). The neddylation of the mutated 3D pol was determined by Co-IP (Fig. 4C). We show that when 3D pol was mutated at K261 or K457, 3D pol neddylation was significantly inhibited. When both K261 and K457 were mutated, 3D pol neddylation was completely blocked (Fig. 4D). These data demonstrate that NEDD8 modification of 3D pol occurs at K261 and K457.
## Neddylation enhances 3D pol stability
Our previous study has shown that the ubiquitination of 3D pol of CVB3 promotes its degradation through proteasomal activity (29). However, CVB3 replication is not disrupted in spite of the proteasomal degradation of 3D pol , suggesting that a counterac tive mechanism might be involved to maintain the stability of viral proteins. Since the conjugation of NEDD8 often induces a conformational change of its targets, and hence their biochemical properties (31), we raised the question of how neddylation would impact the stability of 3D pol . To this end, cycloheximide (CHX) chase assay was used to measure the degradation rate of 3D pol or the mutant 3D pol at K261, K457, and K261/457. As shown in Fig. 5, 3D pol degradation was significantly inhibited in the cells overexpress ing NEDD8 (Fig. 5A andB), indicating neddylation enhances 3D pol stability. In contrast, 3D pol with mutation at either K261 or K457 or both (K261/457) showed significantly accelerated degradation (Fig. 5C through H), demonstrating that neddylation promotes 3D pol stability.
## Viral replication is impaired by the mutations at the neddylation sites of 3D pol of CVB3
To further reveal that neddylation of 3D pol is used by CVB3 as a strategy to facilitate viral replication, we generated recombinant CVB3 viruses that contain mutated 3D pol at its neddylation sites. We compared the viral growth kinetics and viral replication between wild-type CVB3 and the mutated CVB3. As shown in Fig. 6, recombinant viruses were generated based on pMKS1, which contains the cDNA of the entire CVB3 genome (Fig. 6A). The constructs of the recombinant CVB3 were sequenced. Viruses were recovered by transfecting HEK293T cells with these constructs. The virus titer was determined with TCID 50 assay using HeLa cells (Fig. 6B). HeLa cells were infected with wild-type or mutant CVB3 at various MOI (0.1 and 1) for 24 h, and viral protein 3D pol and VP1 were determined (Fig. 6C). Viral growth kinetics was determined by infecting cells with 0.1 MOI of wildtype or mutant CVB3. The cell culture was harvested at various time points of postinfection, and virus titers were determined with TCID 50 assay (Fig. 6D). We show that the replication of mutated CVB3 was impaired, represented by the significantly decreased levels of 3D pol , VP1, and viral growth kinetics (Fig. 6C andD). Furthermore, overexpression of NEDD8 significantly increased the levels of 3D pol and VP1 in the cells infected with wild-type CVB3 (Fig. 6E), while NEDD8 overexpression did not alter both 3D pol and VP1 levels in the cells infected with mutated viruses (Fig. 6E), demonstrating that 3D pol neddylation promotes viral replication. Similarly, NEDD8 overexpression significantly increased the production of viral particles of wild-type CVB3 (Fig. 6F), while the viral particle production of the recombinant viruses was not significantly altered (Fig. 6G through I).
It should be noted that for the cells without NEDD8 expression, the virus yield of wildtype CVB3 was up to 10 6 (Fig. 6F), while the yield of mutant CVB3 declined to 10 4 (Fig. 6G andH), demonstrating that both neddylation sites in 3D pol are indispensable for the efficient replication of CVB3. When cells were infected with the recombinant CVB3, which lost the two neddylation sites of 3D pol , the virus yield further declined (Fig. 6I). Collec tively, these data confirmed that the neddylation of 3D pol promotes CVB3 replication.
To further show the replication difference between the wild-type CVB3 and the mutated CVB3, the cytopathic effect (CPE) of the viruses was visualized (Fig. 6J). We observed that the CPE of the wild-type CVB3 was obvious, while CVB3 with 3D pol mutation at K261 or K457 or K261/K457 showed significantly reduced CPE. Collectively, these data further demonstrated that 3D pol neddylation facilitates the replication of CVB3.
## E3 ligase TRIM4 upregulates 3D pol neddylation
To further understand how neddylation is exploited by CVB3 to promote 3D pol stability, we carried out Co-IP to obtain the proteins that are interacting with 3D pol , and these 3D pol -interacting proteins were analyzed by MS. To this end, cells were transfected with pFlag-3D for 36 h and then infected with CVB3 for 18 h, followed by Co-IP assay to obtain cellular proteins that are co-precipitated with 3D pol . These proteins were analyzed by MS. Among the 3D pol -interacting proteins identified by MS (Fig. 7A; Table S2), we paid special attention to the Ub E3 ligases since all the known NEDD8 E3 ligases are Ub E3 ligases. To validate the role of the E3 ligases identified by MS in 3D pol neddylation, Co-IP was carried out. We show that TRIM4 interacted with 3D pol and enhanced 3D pol neddylation in a dosedependent manner (Fig. 7B through D). To evaluate the functional involvement of TRIM4 in 3D pol neddylation, cells were transfected with the construct expressing the mutant TRIM4 (TRIM4-C27S), which lost its E3 ligase activity (32). We found that compared to the wild-type TRIM4, the mutant TRIM4 displayed weakened capability to enhance 3D pol neddylation (Fig. 7E), suggesting that the E3 ligase activity of TRIM4 is indispensable for mediating 3D pol neddylation. Although the observations above show that TRIM4 mediates 3D pol neddylation, we cannot exclude that there are other E3 ligases that also contribute to this process. To this end, we determined the impact of TRIM40 and RNF169 on 3D pol NEDDylation. It has been shown that TRIM40 suppresses antiviral response through mediating the ubiquitination of RIG-I (33), indicating that TRIM40 might be exploited by viruses. RING finger proteins (RNF) show functional flexibility as the E3 ligases for ubiquitination, SUMOylation, and ISGylation (34), implicating the possible involvement of RNFs in neddylation. Therefore, we tested the correlation of these E3 ligases, TRIM40 and RNF169, with 3D pol neddylation (Fig. 7F andG). In addition, E3 ligase c-CBL was selected as a negative control (Fig. 7H). Our results showed that overexpression of TRIM40, RNF169, or c-CBL did not alter the levels of neddylated Flag-3D (Fig. 7F through H). To further study the interaction of 3D pol and TRIM4, HEK293T cells were co-transfected with constructs expressing EGFP-3D and mCherry-TRIM4. The result showed that 3D pol and TRIM4 were colocalized in the cytoplasm (Fig. 7I). Overall, these observations demonstrate that TRIM4, which interacts with and promotes 3D pol neddylation, likely functions as the E3 ligase of NEDD8-conjugation.
## TRIM4 promotes CVB3 replication
It has been reported that TRIM4 is required for EV-A71 replication with an unknown mechanism (35), and our results suggest that TRIM4 also plays a role in CVB3 infection. To show how TRIM4 would influence CVB3 infection, viral replication was determined in the context of either knockdown or overexpression of TRIM4. As shown in Fig. 8, changes in TRIM4 expression were determined in CVB3-infected cells (Fig. 8A through C). We found that, compared with the mock-infected cells, virus infection induced a significant increase in TRIM4 expression (Fig. 8A through C). Furthermore, in the cells with TRIM4 knockdown, viral protein 3D pol and VP1 were markedly reduced (Fig. 8D through F), indicating that TRIM4 plays a critical role for the efficient replication of CVB3. Moreover, knockdown of TRIM4 almost completely blocked 3D pol neddylation (Fig. 8G), demonstrating that TRIM4 is essential for 3D pol neddylation.
The impact of TRIM4 on virus infection was further determined in the cells with TRIM4 overexpression. We found that overexpression of TRIM4 promoted the levels of viral proteins 3D pol and VP1 in a dose-dependent manner (Fig. 8H through J). Evidently, increased virus yield was also observed in the cells overexpressing TRIM4 (Fig. 8K).
Collectively, these data demonstrate that TRIM4 promotes CVB3 replication through mediating 3D pol neddylation.
## DISCUSSION
Viruses optimize the cellular environment for efficient replication. NEDD8, one of the ubiquitin-like molecules, shares structural similarity with Ub. Neddylation, the covalent linkage of NEDD8 with the specific lysine residue of the substrate proteins, plays an important role in regulating the stability, subcellular localization, and function of proteins (23). Our previous study showed that the 3D pol of CVB3 is ubiquitinated at K220, leading to upregulated degradation of 3D pol through proteasomal activity (29). However, whether or not 3D pol of CVB3 is modified by covalent linkage with NEDD8, which shares up to 60% sequence identity to Ub, remains unknown. This study shows that 3D pol of CVB3 is neddylated at both K261 and K457 mediated by E3 ligase TRIM4. Neddylation enhances 3D pol stability and promotes viral replication. PTM is a strategy used by viruses to facilitate the viral life cycle and to optimize the host environment. Studies have implicated that neddylation plays essential roles in the life cycle of viruses such as HBV, IAV, and EV-A71 (25)(26)(27)(28)36). For instance, the neddylated X protein (HBx) of HBV exhibited enhanced stability and chromatin localization, which in turn promoted viral replication and HBV-induced tumor growth (25). In contrast, the neddylation of PB2 and M1 protein of IAV negatively regulates viral replication (26,27). Similarly, it was reported that the capsid protein VP2 of EV-A71 is neddylated, leading to the reduced stability of VP2 and inhibited viral replication (28). Pharmacological inhibition of neddylation with MLN4924, the selective inhibitor of NAE, enhanced EV-A71 replication. However, another study obtained a contradictory conclusion that neddyla tion promoted EV-A71 replication (37). Nonetheless, according to these reports and our previous study, in which the ubiquitination of 3D pol of CVB3 has been demonstrated (29), we postulated that 3D pol might also be modified by neddylation.
During the infection of enteroviruses, multiple host proteins are involved in every step of the viral life cycle (38). We previously demonstrated that TRIM56 mediates the ubiquitination of 3D pol and promotes 3D pol proteasomal degradation (29). In spite of this, viral replication is not compromised, suggesting that a counteractive strategy might exist to ensure the stability of 3D pol and the successful replication of CVB3. Importantly, our proteomics study using liquid chromatography-tandem mass spectrometry (LC-MS) identified that NEDD8 level was increased significantly in CVB3-infected cells (Table S1). A subsequent verification study demonstrated that it is the overall impact of CVB3 replication rather than the expression of individual viral protein that upregulated NEDD8 expression. Concerning that NEDD8 is homologous to Ub in structure and sequence and that neddylation often stabilizes substrate proteins, we hypothesized that 3D pol might be neddylated. Co-IP analysis identified that 3D pol is covalently conjugated to NEDD8 in CVB3-infected cells. Furthermore, we demonstrated that overexpression of either NAE1 or NEDD8 significantly elevated the virus titer, while knockdown of NAE1 or NEDD8 inhibited viral replication. Moreover, the CVB3 growth rate was significantly decreased when 3D pol lost its neddylation sites at K261 and/or K457. These data demonstrated that 3D pol neddylation facilitates CVB3 replication.
PTM affects almost all aspects of protein function (39). Similar to ubiquitination, neddylation is the conjugation of NEDD8 with the specific lysine residue of the substrate protein via an isopeptide bond (40). Neddylation often changes the stability, intracellu lar localization, and enzymatic function of the substrate protein (23). Protein stability is crucial for the normal function of the cell, and abnormally activated proteasomal degradation often leads to diseases (41). 3D pol , as the RdRp of CVB3, determines the synthesis rate of the viral genome and subsequent translation of viral proteins. Therefore, it is not surprising that the stability of 3D pol is maintained by PTM. We previously demonstrated that TRIM56 mediates the ubiquitination and proteasomal degradation of 3D pol (29). With the identification that 3D pol of CVB3 is also neddylated, we wondered how these different modification processes interact in regulating the stability of 3D pol . To get an initial understanding of this issue, we determined the ubiquitination of 3D pol in the context of NEDD8 overexpression (Fig. S1). Our data show that the ubiquitination of 3D pol was declined along with the increased expression of NEDD8 in a dose-dependent pattern. These data are consistent with the favorable effect of the upregulated expres sion of NEDD8 on CVB3 replication, while an in-depth study is needed to reveal how neddylation is interfering with the ubiquitination of 3D pol .
It has been reported that neddylation promotes the nuclear import of the transform ing growth factor-β (TGF-β)-activated kinase 1 (TAK1) (42). In addition, it has been reported that 3D pol and its precursor 3 CD of poliovirus are capable of entering the nucleus in virus-infected cells, where the viral 3C pro , matured from 3 CD, may cleave nuclear proteins (43). Therefore, to further investigate the impact of 3D pol neddylation on the host cell as well as virus, we observed the cellular localization of the neddylated 3D pol (Fig. S2). We found that, although a small portion of 3D pol is located in the nucleus in a dotted pattern, the neddylated 3D pol is primarily localized in the cytoplasm. This finding implies that neddylation might be critical for 3D pol to function as RdRp since the genome replication of enterovirus occurs in the cytoplasm.
Due to the sequence and structural homology of 3D pol among enteroviruses (44), we initially supposed that other enteroviruses might also contain 3D pol neddylation. It was reported that the 3D pol of EV-A71 contains phosphorylation at S184, which is essential for viral replication (45). However, we found that 3D pol neddylation was absent in either CV-A16 and EV-A71, suggesting that the neddylation site is unique for CVB3 3D pol . Through comparing the amino acid sequences of these enteroviruses, we identified 12 lysine residues that are found only in CVB3. Subsequent analysis demonstrated that neddylation occurs at K261 and K457 of CVB3 3D pol . We also show that the mutated CVB3, which lost its two neddylation sites in 3D pol , exhibited much slower growth kinetics than the wild-type virus. Even with one mutated neddylation site, virus yield was significantly reduced, demonstrating that both neddylation sites of 3D pol are critical for viral replication. In addition, we found that CVB5 infection upregulates neddylation. Sequence alignment shows that there is more than 95% identity in the 3D pol sequence of different types of CVB, while K261 and K457 are well conserved (Fig. S3). These data imply that 3D pol neddylation is shared by all types of CVB.
To further consolidate that neddylation facilitates CVB replication, MLN4924 was used to show the impact of neddylation on virus replication (Fig. S4). We found that MLN4924 suppressed the replication of both wild-type and mutated CVB3 (CVB3-3D K261R or CVB3-3D K457R ), further demonstrating that neddylation of 3D pol is critical for CVB replication.
This study demonstrated that TRIM4 is the E3 ligase that mediates the neddylation of 3D pol of CVB3. To search for the E3 ligase that mediates the neddylation of 3D pol , an MS study was carried out to identify the proteins that interact with 3D pol in CVB3-infec ted cells (Table S2). Until now, all the known E3 ligases that mediate the final step of neddylation are Ub E3 ligases. Therefore, we paid special attention to the E3 ligases in the MS data. We confirmed that TRIM4 specifically interacts with and enhances 3D pol neddylation, while other E3 ligases do not. Moreover, E3 ligase-deficient TRIM4 failed to upregulate 3D pol neddylation. These data show that the promoting effect of TRIM4 on 3D pol neddylation depends on its E3 ligase activity. In addition, we demonstrated that NEDP1, the highly conserved deneddylation cysteine protease, completely blocked 3D pol neddylation, while the mutated NEDP1, which lacks the enzymatic activity, did not. However, further study is needed to characterize the intrinsic role of the NEDP1 during CVB3 infection. Overall, these observations implicate that the key enzymes involved in neddylation, including NAE, NEDD8-conjugating enzyme E2, TRIM4, and NEDP1, are critical for CVB infection.
The interaction and the colocalization of TRIM4 and 3D pol were confirmed in this study. However, which domain of TRIM4 is interacting with 3D pol was not determined. It is also unclear whether or not the interaction between 3D pol and TRIM4 is facilitated by other cellular or viral proteins. Our MS analysis identified up to hundreds of proteins that are interacting with 3D pol in CVB3-infected cells (Table S2). Except for the ribosomal proteins required for viral translation, which are almost co-existing with viral genome synthesis at the same intracellular space, other proteins identified by MS may also be involved in the stability, modification, or localization of 3D pol . Therefore, further in-depth study is needed.
Here, we identified the neddylation of 3D pol of CVB. However, it is unknown whether or not neddylation also occurs in other proteins of CVB. It has been demonstrated that VP2 of EV-A71 is neddylated at the lysine residue at position 69 (28). Through comparing the amino acid sequence of VP2 between CVB3 and EV-A71, we found that K69 is unique for the VP2 of EV-A71 (Fig. S5). However, the presence or absence of neddylation for other proteins of CVB still needs further investigation.
In summary, this study demonstrated that the 3D pol protein of CVB3 can be modified by NEDD8 at its lysine residues 261 and 457. The neddylation of 3D pol , which is medi ated by the E3 ligase TRIM4, promotes its stability and facilitates CVB3 replication. Our findings not only revealed a novel mechanism involved in CVB3 pathogenesis but also suggest that targeting the process of neddylation might be a potential antiviral strategy against CVB3 infection.
## MATERIALS AND METHODS
## Cell culture
HeLa and HEK293T cells were maintained and cultured in Dulbecco's modified Eagle medium (DMEM, Gibco) supplemented with 10% (vol/vol) fetal bovine serum (FBS, Bioindustry, Israel) and 1% antibiotics (100 µg/mL penicillin and 100 µg/mL streptomy cin) at 37°C in a 5% CO 2 incubator and passaged every 2 days. After virus inoculation, cells were maintained in the medium containing 2% FBS.
## Virus
CVB3 Woodruff strain was kindly provided by the Scrips Research Institute (San Diego, USA). CVB5 was provided by Professor Shen of Jiangsu University, Zhenjiang, China. To amplify viruses, HeLa cells were infected with CVB3 or CVB5 for 24 h. Cell cultures were subjected to three freeze-thaw cycles and centrifuged at 12,000 × g (Thermo Fisher, Waltham, MA) for 5 min to obtain the virus stock solution, which was stored at -80°C. The virus titer was determined by the 50% tissue culture infective dose (TCID 50 ) assay as described previously (46). In this study, the TCID 50 of CVB3 was 1 × 10 -6.5 /mL. TCID 50 of CVB5 was 1 × 10 -5 /mL.
## Chemical and antibodies
MLN4924 (Abmole, Shanghai, China) was dissolved in DMSO to prepare the stock solution (100 mM) and stored at -80°C. The working solution of MLN4924 was pre pared with DMEM and used only once to avoid inactivation. Antibodies against NEDD8, GAPDH, β-actin, β-tubulin, Myc, Flag, HA, EGFP, and mCherry were purchased from Proteintech (Wuhan, China). Antibodies against V5, horseradish peroxidase (HRP)-conju gated goat anti-mouse IgG and HRP-conjugated goat anti-rabbit IgG were purchased from Servicebio (Wuhan, China). Antibody against TRIM4 was purchased from CUSABIO (Wuhan, China). Polyclonal antibodies against 3D pol , 3C, and VP1 of CVB3 were prepared in our laboratory.
## Plasmids and siRNAs
The plasmids expressing Myc-NEDD8, mCherry-NEDD8, Flag-NAE1, HA-TRIM4, HA-Ub, and V5-NEDP1 were purchased from Miaoling (Wuhan, China). Plasmids pMKS1-CVB3, pEGFP-C1, pEGFP-2A, pEGFP-2B, pEGFP-2C, pEGFP-3A, pEGFP-3B, pEGFP-3C, pEGFP-3D, pEGFP-VP1, and pEGFP-VP2 ~4 were kindly provided by Professor Zhaohua Zhong, Department of Microbiology, Harbin Medical University. The plasmids expressing EGFP-3D K22R, -K51R, -K96R, -K133/134R, -K196R, -K250/255R, -K261R, -K406/409R, -K436, -K457R, -K261/457R, V5-NEDP1-C163A, and HA-TRIM4-C27S were generated by sitespecific mutagenesis with PrimeSTAR Max Premix (Takara, Beijing, China) by following the manufacturer's directions. The plasmids expressing Flag-3D of CVB3, Flag-3D of EV-A71, EGFP-3D of CV-A16, and Myc-3CD were constructed based on pcDNA3.1-EGFP. The plasmid expressing mCherry-TRIM4 was constructed based on pcDNA3.1-mCherry. NEDD8 siRNA was purchased from Ribobio Technology (Guangzhou, China). The other siRNAs were synthesized by GenePharma (Shanghai, China). The siRNA sequences are listed in Table 1.
## Transfection
HEK293T cells were cultured in 6-well plates to 50% ~ 60% confluence 24 h prior to transfection. The transfection mix was prepared with plasmids and polyethyleneimine (PEI) (Polysciences, USA) in the ratio of 1 : 2.5 dissolved in 1 mL of DMEM, and cells were incubated with the mixture for 6 h at 37°C. After transfection, cells were cultured in fresh medium for 36 to 48 h and then harvested for immunoblotting and RT-qPCR. siRNA (25 nmol) was transfected into cells with X-tremeGene siRNA Transfection Reagent, according to the instructions recommended by the manufacturer. Cells were harvested at 24 h after transfection for further analysis.
## Cytotoxicity assay
HEK293T cells were plated into 96-well plates and cultured to 80% of confluence. Cells were treated with various concentrations of MLN4924 for 24 h. The cell viability was determined with the methylthiazolydiphenyl-tetrazolium bromide (MTT) (Yeasen, Shanghai, China) according to the protocol recommended by the provider. The MTT solution was diluted with DMEM to a concentration of 10% and then added to each well of the culture plate, which was measured by a microplate reader Epoch2 (BioTek) at 570 nm.
## Proteasomal degradation assay
MG132 (Selleck, Shanghai, China) was dissolved in DMSO to prepare the stock solution (10 mM), which was stored at -80°C. The fresh working solution of MG132 (10 µM) was prepared with DMEM. Six hours before the endpoint of the culture, cells were treated with MG132. Cells were harvested for further analysis.
## Cycloheximide chase assay
Cycloheximide (CHX) (Abmole, Shanghai, China) was dissolved in DMSO to 10 mM stock solutions (stored at -80°C). The working concentration of CHX was 100 nM, which was diluted with DMEM and used only once to avoid inactivation. HEK293T cells were used to carry out CHX chase assay after transfection. The cell lysates were prepared at different time points after CHX treatment and analyzed by immunoblotting.
## RNA extraction, reverse transcription, and real-time quantitative PCR
Total RNA was extracted by TRIzol (Yeasen, Shanghai) according to the protocol recommended by the providers, and the concentration of RNA was quantified by NanoDrop 2000 (Thermo Fisher). The reverse transcription system was prepared with 1,000 ng RNA, 1 µL of gDNA remover, and 4 µL of 5 × Trans Script All-in-One Super Mix (TransGen, Beijing, China) and was made up to 20 µL with RNase-free sterile water. The mixture was carried out at 50°C for 5 min, followed by heating at 85°C for 2 min to create cDNA. Twenty microliters of quantitative PCR system consisted of 1 µL of cDNA, 0.4 µL of forward and reverse primers (10 µM), 10 µL of 2 × TransStart Top Green qPCR Super Mix (Trans Gen, Beijing, China), and 8.2 µL RNase-free sterile water.
Quantitative PCR was performed in LightCycler 96 (Roche, Basel, Switzerland) for 45 cycles. Each amplification cycle contained denaturation at 94°C for 5 s, annealing at 58°C for 15 s, and extension at 72°C for 1 min. Relative RNA quantity was calculated using the 2 -ΔΔCT method and normalized to the quantity of GAPDH. Primers were synthesized by Comate Bio (Jilin, China). The sequences of primers for PCR are listed in Table 2.
## Immunoblotting
Cells were cultured in 6 or 12-well plates to 80% confluency and lysed by using RIPA lysis buffer (Beyotime, Beijing, China) containing 1% protease inhibitor PMSF (Beyotime) on ice for 20 min and then treated with ultrasound under the power of 120W for 8 s, repeated three times. The cell lysates were centrifuged at 12,000 × g at 4°C for 15 min to collect the supernatant, and the concentration of protein samples was detected by using a BCA protein assay kit (Beyotime). Proteins were separated on 10% or 12.5% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE). After electrophoresis, proteins were transferred from gel to polyvinylidene difluoride (PVDF) membranes (Millipore, USA), which were blocked with skimmed milk for 1 h and then incubated with the indicated primary antibody at 4°C overnight. The membranes were washed three times with 0.5% Tween-20 in TBST and incubated with the secondary antibody (anti-mouse or anti-rabbit IgG) for 1 h at room temperature. Finally, the blots were visualized by Tanon 5200 Chemiluminescent Imaging System (Biotanon, Shanghai, China) and analyzed by ImageJ.
## Immunoprecipitation
To explore the interaction between two ectopic expression proteins, HEK293T cells were transfected with the indicated plasmids for 36 or 48 h and then washed twice with cold PBS. The cell lysates were collected with NP40 lysis buffer (Beyotime) containing 1% protease inhibitor PMSF (Beyotime). Soluble lysates were incubated on ice for 30 min and then centrifuged at 12,000 × g at 4°C for 15 min to collect the supernatant. After determining the protein concentration, the samples were incubated with anti-Flag or anti-Myc magnetic beads (MCE, Shanghai, China) at room temperature for 2 h with rotation. The control cell lysates were incubated with mouse IgG (Beyotime) at 4°C overnight with rotation, followed by incubation of protein A/G agarose beads (MCE).
The precipitated mixture was placed on a magnetic stand to remove the supernatant, and the protein-magnetic bead mixture was washed with IP washing buffer 10 times by gently pipetting. Finally, the beads were incubated with 100 µL elution buffer at 100°C for 10 min. The precipitated proteins were collected for immunoblotting.
To identify the linkage between NEDD8 and 3D of CVB3, HeLa cells were infected with CVB3 (MOI = 10) for 8 or 12 h and collected with NP40 lysis buffer (Beyotime) containing 1% protease inhibitor PMSF (Beyotime). The cell lysates were incubated with 4 µg of 3D polyclonal antibody or mouse IgG (Beyotime) at 4°C overnight on a rotating mixer, followed by incubation of protein A/G agarose beads (MCE). The rest of the steps are the same as above.
## Fluorescence microscopy
HEK293T cells were grown on coverslips in 6-well plates and transfected with 1 µg of plasmid (pEGFP-3D, pmCherry-TRIM4 or pmCherry-NEDD8) using PEI according to the manufacturer's protocol for 36 h. To visualize cells with fluorescence microscopy, cells were fixed with 4% paraformaldehyde for 20 min at room temperature. Nuclei were stained using 4' ,6'-diamidino-2-phenylindole (DAPI) for 15 min. Coverslips were mounted with FluorSave (Calbiochem). Images were acquired with a Cell Voyager 1000 (Yokogawa, Japan) confocal laser scanning microscope.
## Mass spectrometry
HeLa cells were infected or mock-infected with CVB3 (MOI = 1) for 24 h. The cells were washed with cold PBS three times, and the cell lysates were prepared with NP40 lysis buffer (Beyotime) supplemented with 1% protease inhibitor PMSF (Beyotime). Cell lysates were analyzed by nanoscale liquid chromatography coupled to tandem mass spectrometry (EASY-nLC 1200, Thermo Fisher).
HEK293T cells were transfected with the constructing plasmid Flag-3D for 36 h, followed by CVB3 infection (MOI = 5) or mock-infection. At 18 h post-infection, the cells were lysed in NP40 lysis buffer (containing 1% PMSF) on ice for 30 min and then centrifuged at 12,000 × g for 15 min at 4℃. The supernatants were incubated with anti-Flag magnetic beads (MCE) for 2 h on a rotating mixer, and then the protein-bead mixture was washed with washing buffer for 10 times. The precipitated proteins were removed from the agarose beads and then trypsinized. The digested peptides were analyzed by HPLC-MS/MS.
## Production of wild-type and mutant CVB3 viruses
Full-length genomic cDNA of CVB3 was synthesized and cloned in the pMKS1 vector by GenScript (Nanjing, China), yielding plasmid pMKS1-CVB3-WT. A T7 promoter was placed at the 5' end for in vitro transcription. The point mutations were introduced into this plasmid using PrimeSTAR Max Premix (Takara, Beijing, China), and the desired sequence alterations were confirmed by DNA sequencing. To recover CVB3 viruses, the plasmids containing the wild-type or mutant CVB3 genomic cDNA were transfected into HEK293T cells using PEI. The CVB3 viruses were collected 3 days post-transfection and propagated in HeLa cells. The wild-type or mutant CVB3 viruses were collected by three freeze-thaw cycles and titrated by TCID 50 assay.
## Viral growth kinetics
HeLa cells were infected with wild-type CVB3 or mutant virus CVB3-3D K261R , CVB3-3D K457R , and CVB3-3D K261/457R at an MOI of 0.1. At 1 h post-infection, the cells were cultured in fresh DMEM with 2% FBS at 37°C for 12, 24, 36, or 48 h. The supernatants were subjected to three freeze-thaw cycles, and virus titers were calculated by TCID 50 assay.
## Statistical analysis
All experiments were repeated three times. Data were analyzed by Graphpad Prism 8, and the quantitative data were presented as mean ± SD. Student's t test, one-way ANOVA, and two-way ANOVA were used for statistical analyses. P < 0.05 is considered statistically significant.
## References
1. Garmaroudi, Marchant, Hendry et al. (2015) "Coxsackievirus B3 replication and pathogenesis" *Future Microbiol*
2. Weng, Zhu, Wu et al. (2025) "Research progress and application prospects of animal models of group B Coxsackievirus infections" *Emerg Microbes Infect*
3. Kim, Hufnagel, Chapman et al. (2001) "The group B coxsackieviruses and myocarditis" *Rev Med Virol*
4. Gauntt, Paque, Trousdale et al. (1983) "Temperature-sensitive mutant of coxsackie virus B3 establishes resistance in neonatal mice that protects them during adolescence against coxsackievirus B3-induced myocarditis" *Infect Immun*
5. Yang, Yan, Song et al. (2022) "Whole-genome analysis of coxsackievirus B3 reflects its genetic diversity in China and worldwide" *Virol J*
6. Bouin, Gretteau, Wehbe et al. (2019) "Enterovirus persistence in cardiac cells of patients with idiopathic dilated cardiomyopathy is linked to 5' terminal genomic RNAdeleted viral populations with viral-encoded proteinase activities" *Circulation*
7. Gaaloul, Riabi, Harrath et al. (2014) "Coxsackievirus B detection in cases of myocarditis, myopericarditis, pericarditis and dilated cardiomyopathy in hospitalized patients" *Mol Med Rep*
8. Muckelbauer, Kremer, Minor et al. (1995) *Structure*
9. Bouin, Nguyen, Wehbe et al. (2016) "Major persistent 5' terminally deleted Coxsackievirus B3 populations in human endomyocardial tissues" *Emerg Infect Dis*
10. Huber, Gauntt, Sakkinen (1998) "Enteroviruses and myocarditis: viral pathogenesis through replication, cytokine induction, and immunopathogenicity" *Adv Virus Res*
11. Song, Luo, Shi et al. (2022) "Exploration of IRES elements within the ORF of the Coxsackievirus B3 genome" *Biomed Environ Sci*
12. Chen, Li, Han et al. (2024) "The nucleoside analog 4'-fluorouridine suppresses the replication of multiple enteroviruses by targeting 3D polymerase" *Antimicrob Agents Chemother*
13. Bedard, Semler (2004) "Regulation of picornavirus gene expres sion" *Microbes Infect*
14. Xu, Wu, Han et al. (2019) "Post-translational modification control of RNA-binding protein hnRNPK function" *Open Biol*
15. Kumar, Mehta, Mishra et al. (2020) "Role of hostmediated post-translational modifications (PTMs) in RNA virus pathogenesis" *Int J Mol Sci*
16. Van Der Veen, Ploegh (2012) "Ubiquitin-like proteins" *Annu Rev Biochem*
17. Kamitani, Kito, Nguyen et al. (1997) "Characterization of NEDD8, a developmentally down-regulated ubiquitin-like protein" *J Biol Chem*
18. (2025) *Full-Length Text Journal of Virology*
19. Kamada (2013) "Inhibitor of apoptosis proteins as E3 ligases for ubiquitin and NEDD8" *Biomol Concepts*
20. Zhou, Zhang, Sun et al. (2018) "Protein neddylation and its alterations in human cancers for targeted therapy" *Cell Signal*
21. Walden, Podgorski, Huang et al. (2003) "The structure of the APPBP1-UBA3-NEDD8-ATP complex reveals the basis for selective ubiquitin-like protein activation by an E1" *Mol Cell*
22. Huang, Paydar, Zhuang et al. (2005) "Structural basis for recruitment of Ubc12 by an E2 binding domain in NEDD8's E1" *Mol Cell*
23. Gong, Yeh (1999) "Identification of the activating and conjugating enzymes of the NEDD8 conjugation pathway" *J Biol Chem*
24. Zhao, Morgan, Sun (2014) "Targeting Neddylation pathways to inactivate cullin-RING ligases for anticancer therapy" *Antioxid Redox Signal*
25. Enchev, Schulman, Peter (2015) "Protein neddylation: beyond cullin-RING ligases" *Nat Rev Mol Cell Biol*
26. Liu, Zhang, Yang et al. (2017) "HDM2 promotes NEDDylation of hepatitis B virus HBx to enhance its stability and function" *J Virol*
28. Zhang, Ye, Yang et al. (2017) "NEDDyla tion of PB2 reduces its stability and blocks the replication of influenza A virus" *Sci Rep*
29. Li, Chai, Ye et al. (2020) "Neddylation of M1 negatively regulates the replication of influenza A virus" *J Gen Virol*
30. Wang, Zhong, Cui et al. (2022) "Neddylation of Enterovirus 71 VP2 protein reduces its stability and restricts viral replication" *J Virol*
31. Wang, Dong, Luan et al. (2024) "TRIM56 restricts Coxsackievirus B infection by mediating the ubiquitination of viral RNA-dependent RNA polymerase 3D" *PLoS Pathog*
32. Mendoza, Shen, Botting et al. (2003) "NEDP1, a highly conserved cysteine protease that deNEDDylates Cullins" *J Biol Chem*
33. Rabut, Peter (2008) "Protein modifications: beyond the usual suspects" review series" *EMBO Rep*
34. Yan, Li, Mao et al. (2014) "TRIM4 modulates type I interferon induction and cellular antiviral response by targeting RIG-I for K63-linked ubiquitination" *J Mol Cell Biol*
35. Pan, Xie, Zhang et al. (2024) "EGFR core fucosylation, induced by hepatitis C virus, promotes TRIM40-mediated-RIG-I ubiquitination and suppresses interferon-I antiviral defenses" *Nat Commun*
36. Cai, Tang, Zhai et al. (2022) "The RING finger protein family in health and disease" *Signal Transduct Target Ther*
37. Li, Jian, Yin et al. (2019) "Elucidating the host interactome of Enterovirus A71 2C reveals viral dependency factors" *Front Microbiol*
38. Zhang, Yu, Li et al. (2024) "Protein neddylation and its role in health and diseases" *Signal Transduct Target Ther*
39. Zhang, Guo, Wang et al. (2021) "Inhibition of the neddylation pathway suppresses Enterovirus replication" *Virol Sin*
40. Huan, Qu, Li (2022) "Host restrictive factors are the emerging storm troopers against Enterovirus: a mini-review"
41. Lee, Hammarén, Savitski et al. (2023) "Control of protein stability by post-translational modifications" *Nat Commun*
42. Baek, Scott, Schulman (2021) "NEDD8 and ubiquitin ligation by cullin-RING E3 ligases" *Curr Opin Struct Biol*
43. Wang, Jiang, Zhang et al. (2024) "Ubiquitin-like modification dependent proteasomal degradation and disease therapy" *Trends Mol Med*
44. Li, Fang, Cui et al. (2020) "Neddylation promotes protein translocation between the cytoplasm and nucleus" *Biochem Biophys Res Commun*
45. Sharma, Raychaudhuri, Dasgupta (2004) "Nuclear entry of poliovirus protease-polymerase precursor 3CD: implications for host cell transcription shut-off" *Virology (Auckl)*
46. Gruez, Selisko, Roberts et al. (2008) "The crystal structure of coxsackievirus B3 RNA-dependent RNA polymerase in complex with its protein primer VPg confirms the existence of a second VPg binding site on Picornaviridae polymerases" *J Virol*
47. Lin, Dong, Xiao et al. (2023) "Proteomic and phosphoproteomic analysis of responses to enterovirus A71 infection reveals novel targets for antiviral and viral replication" *Antiviral Res*
48. Wang, Zhao, Chen et al. (2020) "N-Acetyl cysteine effectively alleviates Coxsackievirus B-Induced myocarditis through suppressing viral replication and inflammatory response" *Antiviral Res*
49. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12369362&blobtype=pdf | # Learning microbiology and virology from the cinema
Oliver Schildgen, Jan Smetana, Jan Ehlers, Kathrin Dreckmann
## Abstract
In this report, an alternative approach to teach virology and microbiologyrelated topics is presented. The teaching approach is suitable to overcome the learning fatigue of students who were tired from media reporting on viruses during the COVID-19 pandemic or were less motivated for the discipline.
## PROCEDURE
The course was a mixture of short lectures by university teachers, short students' talks, movie-watching, and blended learning, i.e., for each session, movies and accompa nying texts/papers were provided according to the students' learning levels. Prerequi sites for successfully passing the course were active participation, including watching the movies/streaming series/videos, reading course literature (ranging from Wikipedia entries to students' textbooks and journal articles), and presenting of a short talk on one of the course topics. No written examination was required to successfully pass the course.
The course was held in the frame of the so-called Studium fundamentale (https:// www.uni-wh.de/en/your-studies/degree-programmes/studium-fundamentale), which is a mandatory part of any studies at the University of Witten/Herdecke and which aims to provide interdisciplinary insights into topics beyond the main studies. Consequently, the participants came from the human medicine degree program, dental medicine degree program, psychology degree program, nursing science degree program, and the business administration degree program. The level ranged from the first semester to the sixth semester, independent of the discipline. In total, 23 students participated in the course, and gender and degree balance were considered. Six out of 23 students came from the business administration study program, psychology, and nursing studies. More details, however, cannot be provided here due to reasons of anonymity requested by the ethical vote. The course was a pilot project.
The topics and movies included in the course are presented in Table 1, and the major learning objectives are pointed out in the table, too. Each lesson followed the same scheme and started with a flipped classroom. Students had to watch at least one of the movies (better both, if two movies were chosen) and read the literature provided for them according to their level of knowledge. The total time for this was about 2 h of reading plus watching the movies or TV series, the latter not longer than the movies, i.e., three to a maximum of five episodes. The students had to prepare a short presentation, no longer than 15 min, in which they included the aspect they identified as relevant. The presentations were then discussed and critically but constructively commented on by the lecturers, who in turn added a short presentation with the aspects they identified as relevant, which in turn was again discussed with all participants. Finally, key messages and major learning points were summarized and collected by all. The course was open to all levels of students, from absolute beginners to advanced students, with the most advanced students studying medicine in the eighth semester. The sessions in which the short presentations were held were scheduled every two weeks.
For the analysis of course success, a mixed-method study design with a pre-post setting was chosen. Students were asked to fill out a questionnaire before and after the course. The questionnaire included qualitative and quantitative self-assessment questions with scales from 1 (fully agree) to 6 (fully disagree), as well as topic-specific questions on the course contents that were designed to deliver a totally correct (1) to a totally incorrect (6) answer. Additionally, the questionnaires contained questions that were of a qualitative nature and that were dedicated to learning about the students' experiences with microbiology. As examples, it was asked if the students had experi enced discrimination or stigmatization because of infectious diseases themselves or by others. The students were also asked for their expectations before the course (What do you expect to learn?) and after the course (Were your expectations fulfilled? What did you miss?). Additionally, students were asked to provide suggestions, topics, and ideas for future rounds of the course. Finally, the questionnaires also contained questions addressing prejudices and clichés circulating around microbiology and virology (e.g., "French disease"). Some representative questions, including the answering results before and after the course, are presented in the Results section. The questionnaires were not designed to be examinations, and this was communicated explicitly to participants. Student participation was fully voluntary. The study, including the questionnaires, was approved by the local ethical committee of the Private University of Witten/Herdecke (vote no. S-125/2024). The pilot project was internally and externally evaluated.
Additionally, the students were asked to anonymously provide written feedback, in which they were asked about their motivation before and after the course, as well as their personal opinion about the course format.
## Results
The students were asked to watch a selection of popular movies or streaming content, including but not limited to movies/series like Outbreak, Contagion, The Andromeda Strain, Guardians of the Galaxy, The Walking Dead, Bohemian Rhapsody, or Philadelphia (Table 1). A small group of students was selected for each topic, respectively, and prepared a short presentation of the scientific content of the respective movie while also taking into account cultural and societal aspects, e.g., stigmatization as shown in Philadelphia, or the aspect of population control as discussed in Inferno. After discussing the students' presentation, further input was given by the teachers who were specialized in virology/microbiology and media and cultural studies.
Table 2 shows some representative questions and the students' replies (in percent age). Questions were separated into self-assessment questions, in which a range from fully agree to fully disagree could be marked. Knowledge-based questions are asked, for example, the definition of relevant terms such as "pandemic. " Answers to those questions were marked like school examination degrees (Table 2) from 1 to 6 (German system, with 1 being fully correct) or A to F (according to the American system). The third type of question included was to be answered with true or false (or correct, not correct).
As shown in Table 2, there are clear positive effects, such as an increase from 21% of fully correct (A-level) to 61% of fully correct answers for the definition of pandemics. This positive effect was observed for all questions and was independent of the main degree course of the students. However, not all students answered all questions in total; thus, there are some discrepancies between the percentages and the total number of students if reverse calculated. This is because of the voluntary character of the questionnaire and the small total number of participants.
However, all students who filled out the questionnaire gave relevant free-text replies on their motivation and self-assessed learning effect and motivation. All 19 of the 23 students who finished the questionnaire reported that the approach caused a higher motivation to learn virology and microbiology and that learning was easier and more successful than in conventional courses. This was supported by the objective measures from the mixed-methods pre-post analyses that have shown a significant increase in the virological and microbiological knowledge of the students. Also, the self-assessment of the students was more positive after they completed the course than in advance of the first lesson. Additionally, students reported gaining skills due to the incorporation of media and cultural science studies in the course. Some students of medicine and dental medicine have had microbiology/virology lectures and courses before. The qualitative self-assessment (i.e., free-text answers after the course) revealed that those students felt they had easier access to the topics by way of the movie-based teaching scheme than in conventional case-based teaching or frontal lectures.
As examples, knowledge related to biosafety issues and food safety was claimed to be highly useful, followed by trained soft skills during anamnestic discussions with patients. An increase of up to 90% in correctly answered topic-specific questions and self-assessment was measured independently of the study discipline.
The increase in knowledge, both on self-assessment and objectively measured, was independent of the degree course of the students. This indicates that the positive effects were independent of the degree course, but as the course itself was not mandatory, the positive effects may have also been caused by intrinsic motivation, even if the students were not explicitly STEM majors. Figure 1 shows some examples of the increased knowledge the students acquired from the course. Questions not only included virology and bacteriology-specific questions but also questions on epidemiology, genetic engineering in virology and microbiology, environmental aspects, biological warfare, and animal experimentation in virology and microbiology. In Fig. 1a, the average increase in correctly answered questions after the course compared to the initial replies before the course is shown, i.e., over all questions asked. The increase in correct answers is directly obvious, and the percentage of totally wrong answers (5 and 6 marks) is clearly reduced. For some specific questions, e.g., on zoonosis (Fig. 1b) or epidemiology (Fig. 1c) in microbiology/virol ogy, the effect is even more pronounced. Thereby, it has to be mentioned that the students who attended the course came from different scientific disciplines, i.e., they studied medicine, dental medicine, psychology, and economics/business. No study-spe cific effects were observed. A total of 19 out of 23 students replied to the questionnaires.
During the entire course, we observed an open-minded and motivated group of students who voluntarily worked on relevant topics of microbiology and virology. Except for two cases of illness at a single session, respectively, no absence periods were observed, and all participants actively contributed. Both the internal feedback from the questionnaires and the external evaluations were overwhelmingly positive for the course format.
Free-text replies from the students were critical but entirely positive. The inclusion of Media and Cultural Sciences was asked to be improved in the sense of becoming a more introductory format, as this discipline was completely new for the students. All students who provided free-text answers/comments reported that their knowledge of microbiology and virology increased subjectively and that they were motivated to learn more about the topic, independent of the discipline they studied. The stu dents also reported that they had learned where to get relevant and evidence-based information on virology/microbiology topics and were able to mention sources such as PubMed, CDC/ECDC, Robert-Koch-Institute, or scientific journals. The students also reported that they recognized that the disciplines of microbiology/virology were often not sufficiently represented in the media and movies and that objective reporting and scientific discourses appeared often to be hindered or mono-dimensional. More general comments were that the students would recommend the course to their fellow students as one of the preferred courses they have participated in so far and recommended extending the format to other movies and, if possible, other clinical entities.
Additionally, a further positive side effect of the teaching format is that students were trained to differentiate between dramaturgic exaggerations in movies and scientific reality, a skill that is increasingly important in times of a plethora of fake news on science and medicine.
## CONCLUSION
We conclude that the teaching scheme is suitable not only for students of medicine but also for other disciplines like nursing, psychology, and even business administra tion. Students from these disciplines who participated demonstrated increased learning based on self-assessment and objective measurements. Surprisingly, comments that were solely negative were not included in student feedback, but that may be due to the voluntary format of the questionnaire. While it is a weakness of this study that the number of students is rather low, the results are encouraging to roll out the model to larger student groups. As at our University, the groups are in principle smaller because the University of Witten is a small private university with a focus on small but intensive learning groups. This means that the teaching scheme will be continuously evaluated in future semesters. As a further weakness, the course was not compared to an existing course, as it was a pilot format, and the general intention was to analyze whether the concept is accepted by the students. This has to be addressed if the course should be used to replace existing teaching formats.
Finally, a statistical effect has to be discussed. The percentages shown in Table 2 and Fig. 1 do not always fully align with the number of "physical" students, as not all questions were fully answered by all students, and we decided to round up or round off the percentages to a full percent. Thus, "re-calculation" from percentage back to student number may result in "half-students. " In order to overcome this technical bias, a larger student cohort should participate, and further studies are required.
The course format thus could be a template for microbiology and virology education on several levels from high school to university studies, as already concluded by Linares and coworkers (1), and could overcome motivation gaps resulting from the COVID-19 pandemic.
## References
1. Linares, López-Ejeda, Álvarez et al. (2021) "Service-learning, movies, and infectious diseases: implementation of an active educational program in microbiology as a tool for engagement in social justice" *Front Microbiol*
2. Wester, Walsh, Arango-Caro et al. (2024) "Student reflections on emotional engagement reveal science fatigue during the COVID-19 online learning transition" *J Microbiol Biol Educ*
3. "Alteration of wrong and correct replies to a question on epidemiology. In all diagrams, it is obvious that there was an increase in correct answers and that the students experienced an increase in measurable knowledge" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12681262&blobtype=pdf | # Dengue Fever and Its Burden in Burkina Faso: An Overview
Wendimi Belem, Ibrahim Sangare, Caroline Gilbert, | Wilfried, Wenceslas Bazié, Wilfried Bazié
## Abstract
Dengue fever is an arbovirus disease caused by the dengue virus and has been diagnosed in Burkina Faso for many years. In recent decades, the disease has become a growing concern, thereby impacting the public health system. Several factors contribute to the pathogenesis of dengue fever, including the immune system and the virulence of different serotypes. Additionally, multiple complex conditions, including the spread of the Aedes mosquito vector and meteorological factors, contribute to the disease's spread. Therefore, effective disease management must be comprehensive, involving strategic combinations including community engagement, mosquito control, and public health measures. This approach has been implemented in Burkina Faso, with some success. Although several studies have focused on viral control, the isolation of virus serotypes, the prevalence and seroprevalence of the disease in specific populations, information on the overall burden of dengue fever is scarce in the country, as it presents classic symptoms similar to those of malaria and some arbovirus diseases encountered in the country. However, limited access to diagnostic tools, an inadequate surveillance system, a lack of awareness among healthcare workers, auto-medication, and ongoing conflicts in the country may lead to an underestimation of its burden and a limited understanding of its epidemiology. Here, we discuss dengue fever and the factors associated with the underestimation of its burden in Burkina Faso, drawing on government documents and published data. This review aims to describe the impact of managing this neglected tropical disease, advocating for improved surveillance and control efforts in the country.
| IntroductionDengue fever (DF) is a vector-borne viral disease caused by the dengue virus (DENV), which is transmitted to humans by mosquitoes [1]. The global burden of the disease is substantial. An estimated 50 million cases occur annually in approximately 100 countries, with the potential for further global dissemination. The incidence of DF has risen significantly worldwide in recent decades, as evidenced by the increasing number of cases reported by the World Health Organization (WHO), which climbed from 505,430 in 2000 to 5.2 million in 2019 on a global scale [2]. Remarkably, the year 2019 witnessed the highest
number of dengue cases ever documented worldwide [2]. An estimation from a modelling study suggests that there are 390 million occurrences of DFs annually, of which 96 million display clinical symptoms [3]. Another investigation on the occurrence of dengue reveals that approximately 3.9 billion individuals are susceptible to the DF [4]. The disease is now endemic in over 100 nations spanning the WHO regions of Africa, the Americas, the Eastern Mediterranean, South-East Asia, and the Western Pacific. Notably, the Americas, South-East Asia, and Western Pacific territories bear the brunt of this disease, with Asia alone accounting for approximately 70% of the overall disease burden [2,3]. Reports also suggest that people of the African population do not experience severe DF at rates comparable to Europeans or Asians [5]. However, several factors might explain these situations in Africa [6].
In Burkina Faso, DF is a growing problem that significantly impacts public health [7,8]. The country has sometimes faced epidemics in recent decades, resulting in increasingly high morbidity and mortality. While the central region is heavily affected, the spread to other parts of the country is concerning and demands strong preventive measures [8]. In a context of the endemicity of other febrile infectious diseases, the lack of awareness among healthcare workers [9], the self-medication, ongoing conflicts in Burkina Faso [10], the limited diagnostic tools [11] and an inadequate surveillance system raise questions about the exact DF burden in the country. Here, we discuss DF and the factors associated with the underestimation of its burden in Burkina Faso, drawing on government documents and published data. This review aims to describe the impact of managing this neglected tropical disease, advocating for improved surveillance and control efforts in the country.
## 1.1 | Virus and Serotypes Associated With Various
Outbreaks DENV, a member of the Flavivirus genus in the Flaviviridae family, is a positive-sense single-stranded RNA virus about 11 kilobases in length. It has a diameter of approximately 50 nm and contains three structural proteins (capsid, precursor membrane, and envelope proteins) and seven non-structural (NS) proteins named NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5 [12] Four distinct serotypes of DENV-serotypes 1 (DENV-1), 2 (DENV-2), 3 (DENV-3), and 4. (DENV-4)-have been identified globally. These serotypes exhibit unique antigenic features, distinguishing them from one another. All serotypes include multiple genotypes (four to seven), which differ in the amino acid present in the envelope protein (≤ 10%) [13]. These genotypes add to the diversity and complexity of the DENV population. [14] Understanding serotype-specific traits and genetic variation is crucial for effective monitoring, diagnosis, and control of DF as well as understanding epidemiological patterns, disease progression, transmission dynamics, and immune responses, ultimately informing public health strategies and vaccine development.
Historical data confirm the presence of all serotypes of DENV in Burkina Faso [13]. Therefore, DENV-2 [15] and DENV-3 [16] were the most prevalent serotypes involved in different outbreaks (Table 1). Co-infections involving multiple serotypes, such as DENV-1 and DENV-2, DENV-1 and DENV-3, DENV-2 and DENV-3, DENV-1 and DENV-4, and DENV-4 and DENV-3, have also been reported in previous studies [22]. However, there is limited research on the serotypes and genotypes associated with severe cases and case fatality rates of DF, underscoring the need for further investigation to understand their implications for the severity of DF in the country.
## 2 | Immunopathogenesis
The immunopathogenesis of DF involves host-specific immune responses, including immune cell activation, complement activation, the production of inflammatory mediators, and autoimmunity, as well as the virulence of the virus, host genetic factors [23], including the antibody receptor attachment molecules [24], human leucocyte antigens (HLA), and singlenucleotide polymorphisms [25]. To control the infection, a variety of immune and inflammatory mediators are involved in the innate and adaptive antiviral response to dengue [26]. Asymptomatic or mild infections are characterised by efficient viral clearance due to the balance of these immune components [12].
Several key observations highlight the immune system's role in dengue pathogenesis. First, the convergence of symptoms, vascular leakage, and viral control aligns with a 'cytokine storm' marked by an abundance of pro-inflammatory mediators [27] during primary and secondary infections. In DF immunopathogenesis, the 'cytokine storm' contributes to changes in endothelial permeability [12], which are seen in severe disease forms. Capillary permeability results in plasma leakage, which can potentially lead to haemodynamic instability, shock, organ failure [28], and bleeding [29]. For instance, previous studies have shown that TNF-α and IL-10 are associated with haemorrhagic symptoms and platelet breakdown, respectively [30]. TNF-α and IL-1β [31] or IL-2 and TNF-α [32] directly cause vascular leakage. An inadequate antiviral response, characterised by the overproduction of IL-10, could suppress the immune system, leading to an uncontrolled inflammatory response and severe disease progression [29]. Additionally, recent reports indicate that the synergistic effect of TNFα and IFN-γ leads to inflammatory cell death, triggering the cascade of the 'cytokine storm' [29]. Secondly, severe DF may occur upon subsequent infection with a different serotype following an impaired adaptive immune response [33]. Indeed, this severe DF is believed to be partly due to immune mechanisms involving B-and T-cell immunopathology. However, these events are still debated. When an individual is infected with one serotype, it results in homotypic immunity that provides lifelong protection against that serotype and a period of temporary cross-immunity to other serotypes [34]. The pathogenesis of disease in secondary infections may involve antibodydependent enhancement (ADE) [35]. However, current studies showed that viraemia levels are neither the main cause of inflammation [36] nor linked to DF severity [37], as severe symptoms mainly occur when viraemia decreases rapidly [32]. Another explanation of the severity of disease pathogenesis during secondary heterotypic DF is the original antigenic sin of T lymphocytes [38]. The phenomenon of original antigenic sin in DF occurs when pre-existing low-affinity and cross-reactive memory T lymphocyte expansion leads to ineffective viral control [38]. However, they excessively release pro-inflammatory cytokines ('cytokine storm'), causing an exacerbation of the disease [39]. Recent studies challenge the idea that cross-reactive Tcell responses are only associated with pathogenesis, as they can also induce a robust and multifunctional response that can provide protection [40]. The protective or detrimental effects of DENV-specific T cells may be associated with host factors based on studies of HLA associations with dengue [41]. Few investigations focused on the immunopathogenesis of DF in African populations. BECQUART and collaborators found that acute DENV-2 infection in Gabonese patients (African population) was associated with an early innate immune response [42]. Therefore, studies are needed to investigate the implications of some immunological events in the pathogenesis of DF to understand the mechanism of this disease and identify easily measurable biomarkers for patient management, mainly in the African context, including Burkina Faso.
## 3 | Dengue Burden
DF has been a public health concern in Burkina Faso in recent decades. Although the exact prevalence and fatality rates of severe DF cases remain unknown, more research is necessary to understand the disease better, enabling improved management [43]. The DF outbreak was first recorded in Upper Volta (now Burkina Faso) in 1925 [17]. The first DF outbreak with germ isolation was reported in 1982, documenting 30 cases (29 expatriates and one African) [19]. Since 2000, DF has become endemic in Burkina Faso [44] with some occasional epidemic peaks. In 2016 and 2017, Burkina Faso experienced its first major DF outbreaks, with 1266 probable cases (in Burkina Faso) and 15 dengue-related deaths (from all 12 districts of Ouagadougou) [15], and 8804 probable cases, with 30 dengue-related deaths reported, respectively [45]. The largest DF outbreak in Burkina Faso's history occurred in 2023, recording 160,751 probable cases and 751 deaths (0.4%) according to the country's health statistics yearbooks [46]. The epidemic continued into 2024, with 83,869 cases and 99 deaths [47]. Notably, the number of suspected dengue cases and deaths has increased over recent years from 2020 to 2024, based on data from Burkina Faso's health statistics yearbooks (Figure 1) [46][47][48][49][50]. Some reports also provided other data on disease prevalence in specific population groups and regions. The data were extracted and summarised in Table 2.
## 4 | Diagnosis of DF
Laboratory diagnosis of DF is established through direct or indirect identification of DENV and its components in the bloodstream. The diagnostic methods involve virus isolation, detection of virus antigen (Ag) or RNA, and the presence of DENV-specific antibodies [56,57] using serological tests, molecular detection (Nucleic acid tests), viral culture (research purposes), sensing technologies, and clustered regularly interspaced short palindromic repeats-based genome editing technology [58].
Detecting dengue nucleic acid by real-time reverse transcriptase PCR (rRT-PCR) appears to be the most sensitive and specific means of detecting dengue viraemia [59]. Recently, platforms based on rRT-PCR have been developed, which can differentiate each of the four serotypes of DENV (multiplex rRT-PCR) or three arboviruses, including DENV, Chikungunya, and Zika (Trioplex rRT-PCR tests) from blood samples [58] in areas endemic to these arboviruses. Unlike molecular tests, serological tests, including Enzyme-linked immunosorbent assays (ELISA) [56,57] and rapid diagnostic tests (RDTs) [60], are the most widely used techniques in detecting DF due to their low cost and operational simplicity. The ELISA procedures use tests based on IgM, IgG, IgM/IgG ratio, and NS1 [58]. Therefore, detecting NS1 antigens through immunoassays provides acceptable levels of sensitivity and specificity in detection [59]. However, ELISA diagnostics are instrument-dependent and require facilities that enable the performance of the tests [59]. Implementing RDTs based on immunochromatographic methods offers simpler point-of-care diagnostic options in many dengue-endemic settings, particularly in resource-limited
## Years of outbreak occurrence
## Detection in humans or vectors Incriminated serotypes Localisation
1925 [17] Human Unknown serotype -
## Authors and references Title Prevalence
Ouattara et al., Journal of vector borne diseases (2025) [13] Predominance of DENV-3 among patients in ouagadougou, Burkina Faso diagnostic laboratories [60]. They are valuable tools for the efficient and accurate diagnosis of DF, with several advantages, including a simple workflow, rapid turnaround time, and minimal equipment requirements [60], as well as an affordable cost [61]. WHO recommends diagnostic tests based on the level of laboratory surveillance. Although all of these methods can be used to establish an etiological diagnosis, rapid point-of-care tests for detecting antigen, antibody, or both antigen and antibody are preferred, whether they have satisfactory sensitivity and specificity [62]. Several other serological tests are used for DENV diagnosis, including haemagglutination inhibition (HI), Luminex-based serological analysis, and neutralisation assays.
The HI test is a standard test for primary and subsequent infections distinct from DENV, with the disadvantage of not detecting early disease. In contrast, neutralisation assays have the advantage of efficiently determining asymptomatic DF [58].
In addition, due to its high specificity, neutralisation assays may be used to confirm the positive ELISA tests for DENV in endemic areas of flaviviruses [61]. However, this test has disadvantages, such as time and labour [58]. Luminex-based serological analysis allows the simultaneous detection of IgG and IgM [63] and antibodies against different DENV serotypes [64]. Luminex-based serological analysis is highly specific, but they are costly. The effectiveness of each method in detecting various markers (Ag NS1, Ig M, IgG, and RNA) is influenced by the duration of the patient's illness (Figure 2).
In Burkina Faso, RDTs/Duo Dengue are recommended for outbreak response in detecting probable cases. Among the many tests available in the country, some have not been evaluated or validated at the national level [65]. RDTs were mostly available only in private laboratories and health facilities, and were not subsidised, which created a barrier to access for the majority of the population [55]. However, during the 2023 outbreak, the government announced free testing in public health facilities.
Confirmation of probable cases (NS1Ag þ or IgMþ) must be conducted in the national reference laboratory for viral haemorrhagic fevers (NRL/VHF) through ELISA and PCR, as well as multiplex techniques to differentiate DENV serotypes and the detection of DENV, Chikungunya, or Zika [66,67]. Interestingly, the health system has a pyramidal organization at the central (tertiary), intermediate (secondary), and peripheral (primary) levels [8]. The lack of clear regulatory guidelines for laboratory tests and the unavailability of tests and reagents from national drug purchasing centres hinder disease diagnosis in primary healthcare settings in Burkina Faso [65]. Therefore, access to testing is limited, and confirmatory tests are often not performed [11].
## 4.1 | Transmission, and Spread of DF
DENV can infect non-humans (sylvatic cycle), humans [68], or vertically from mosquitoes to their offspring [69], thereby contributing to the spread of the disease. The DENV to human transmission occurs primarily through the bites inflicted by infected female mosquitoes (Figure 3) (Aedes aegypti mosquito), followed by Aedes albopictus [12], and possibly Aedes polynesiensis [30]. Although studies have indicated that Aedes aegypti is the most frequently collected vector in containers across some domestic and peri-domestic areas of Burkina Faso, its abundance correlates with a higher risk of DF transmission. Sylvatic circulation of DENV [70] and its vectors, such as Aedes luteocephalus, Aedes africanus, and Aedes cumminsi [71], have also been documented in other regions of the country.
Generally, several complex factors contribute to the spread of DF, such as globalisation, the dissemination of the Aedes mosquito vector, insufficiently planned urbanisation, and the FIGURE 2 | Detection of DENV and its components in the bloodstream. Detection of DENV and its components in the bloodstream can be direct or indirect, depending on the timeline of illness. The DENV genome and NS1 antigens can be identified early in infection, within 24-48 h before symptoms appear, and up to 7 days after the start of fever using PCR and serological tests. In primary infection, anti-DENV IgM and IgG antibodies usually appear around 3-5 days and 8-10 days after fever begins, respectively. These antibodies can be detected during the specified periods by serological testing.
lack of an authorised anti-dengue therapeutics [72], ineffective vector control strategies, climate change, population growth, and high population density [73]. Meteorological conditions (e.g., temperature, humidity, and wind speed) influence the spread of DF and have a significant nonlinear effect on DENV transmission in the country [15,74]. Therefore, several studies have reported the seasonality of dengue transmission in Burkina Faso (between October and November), often with a time lag [8], especially in the largest cities, Ouagadougou and Bobo-Dioulasso [71]. For example, an optimal temperature range of 27°C-32°C has been identified for DENV transmission in Burkina Faso, corresponding with the temperature of October and November (period following the rainy season) [8,75]. Additionally, the geographic expansion of mosquito vectors and their resistance to insecticides has contributed to the resurgence of arboviral disease outbreaks in sub-Saharan Africa [76]. Aedes aegypti populations have shown resistance to multiple insecticide classes in Burkina Faso [71], which may reduce the effectiveness of chemical control methods.
## 4.2 | Surveillance System and Other Factors Associated With the Underestimation of DF
Scant or no DF surveillance activities were carried out before the 2000s, during which sporadic dengue outbreaks occurred in other parts of Africa, accompanied by the emergence of DHF in some countries [77]. Therefore, its magnitude could not be monitored by routine surveillance, which explains the uneven data. Nevertheless, the high prevalence of dengue NS1 antigen in Burkina Faso confirms the endemicity of DF and active disease transmission [78,79]. Burkina Faso has been officially adding DF to its routine national surveillance system for diseases with epidemic potential since 2016 [15]. The surveillance activities at each level of the health system have been clearly defined in the technical guide for integrated disease surveillance and response and by the job descriptions of the different actors involved in the Health Information System (HIS) [80]. Using weekly health service reports, epidemiological disease surveillance data were obtained from the National Health Management Information System of the Ministry of Health of Burkina Faso, operating on the District 2 Health Information Software platform (version 2.38) [75]. Cases were reported yearly with epidemic peaks in some areas [71]. In addition to epidemiological surveillance, routine health service reports, programme management, administration, and resource management, community-based surveillance, and periodic surveys and studies are components built around a HIS [80]. The compilation of cases from health service reports has been reported in the Health Statistics Yearbooks [8]. Thus, the data extracted from these yearbooks showed a heterogeneous spatial distribution of DF in Burkina Faso (Figure 4). The highest probable dengue cases were reported in the Central Region, followed by the Upper Bassin Region, indicating the Central Region was the focal point of the DF [46]. In addition, a key element is the implementation of the national strategy for the control of arboviruses (dengue, Zika, and chikungunya). It operates like a sentinel surveillance network, comprising seven sentinel sites distributed throughout the national territory, and has been operational since 2017 [81]. As part of virological surveillance, confirmation, serotyping, and genotyping are carried out using suspect/probable samples (sentinel sites and epidemic regions) by the NRL/VHF [82]. However, dengue vector surveillance has not yet been implemented in Burkina Faso [71]. In the absence of a national entomological surveillance system for arbovirus vectors, several investigations have reported arbovirus indicators in vectors in some regions [70,71].
Passive surveillance systems associated with underutilisation of health services [10], a lack of awareness among healthcare workers [9], and the lack of diagnostic tools [11] may lead to the underestimation of the DF burden due to misdiagnosis, misclassification, and underreporting [6] .Burkina Faso has complied with the new WHO classification of DF. Thus, cases must now be recorded and reported according to the severity of the disease, including group A, B1 (DF without warning signs), B2 (DF with warning signs), and C (Severe DF) [83,84]. DF was reported as the most frequently neglected tropical disease in medical centres and hospitals, and the fifth most common in primary health care facilities in 2020 [52]. he clinical presentation of DF is difficult to distinguish from other febrile infectious diseases common in some African countries, such as malaria, typhus, and other arboviral infections [85]. In Burkina Faso, there exists confusion among health professionals about DF and malaria. 'Malaria-dengue' is a term that refers to either malaria or DF, as well as malaria and DF coinfections, leading to complications in disease identification and reporting, especially in remote rural areas [10]. Therefore, DF is often eclipsed by the substantial malaria burden in Burkina Faso [86]. Furthermore, the underutilisation of health services, particularly in resource-limited countries, makes it difficult to estimate the true health and economic burden of DF [44]. Among the reasons for the underutilisation of health services in Burkina Faso are self-medication and ongoing conflicts [10].
## 4.3 | Prevention and Control Measures Against
## DF Outbreaks
In response to DF outbreaks, a multifaceted approach was implemented, including communication and awareness campaigns, case prevention and management, laboratory strengthening, logistical support, enhancement of epidemiological surveillance, case mapping, vector control, and management skill development [82]. Response activities were coordinated by the Health Emergency Response Operations Centre. Vector control measures mainly focused on eliminating breeding sites and applying insecticide fogging to kill adult mosquitoes during outbreaks [71]. For example, during the 2023 outbreak, vector management included physically destroying larval breeding sites, conducting spatial and indoor spraying, and using drones for areas that are difficult to access [82]. A national campaign to eliminate larval nests was also launched in Ouagadougou by the Ministry of Environment, Water, and Sanitation, while the Ministry of Agriculture and Animal Resources provided sprayers to support spraying efforts [87]. Strengthening epidemiological surveillance, case prevention, and laboratory capacity was also carried out to improve data collection through the availability of dengue surveillance tools and support, training, and the supply of necessary materials and equipment [82]. Additionally, Burkina Faso's response to DF outbreaks included conducting sentinel seroepidemiology studies to determine the prevalence of DF and identify circulating serotypes [54]. The spatiotemporal distribution of dengue cases has been studied to pinpoint high-risk areas and develop more effective preventive strategies [8]. Communities were equipped with essential knowledge about dengue prevention strategies [88] and instructions to avoid self-medicating in case of suspected symptoms of DF [89] The communication, awareness-raising, and mobilisation messages were broadcast on the radio and television [82].
## 4.4 | Future Strategies for DF Prevention
Although response efforts to the 2023 DF outbreak have improved compared to previous ones, some gaps remain and need to be addressed [82]. Indeed, future strategies (Figure 5) should focus on strengthening the disease surveillance system by implementing active surveillance for a few years, especially during high-risk periods. There is an urgent need to develop national guidelines for the comprehensive management of febrile infectious diseases and to implement an integrated febrile infectious disease surveillance strategy nationwide. Effective prevention of DENV vectors involves reducing the source, destroying larvae, early treatment before the mosquito season, using bed nets, mass trapping, eliminating adult mosquitoes in the environment, screening dwellings [73], and cooperating regionally or internationally to control vectors. Additionally, DF places significant demands on healthcare systems. Strengthening healthcare workers' knowledge, improving early diagnosis, and deploying tools for predicting the risk of severe cases are needed to facilitate rapid triage at the primary care level and daily effective clinical management of the disease. Licensure and validation of the sensitivity and specificity of diagnostic tests should be a higher priority [65] for the health ministry, followed by widespread availability of diagnostic tests. Investment in training programs, research initiatives, and regional collaboration is also essential for knowledge exchange and resource sharing to drive progress in dengue prevention and control. Using a participatory approach, community engagement and educational efforts could promote preventive behaviours. This component should include educational lectures, door-to-door outreach, and theatrical performances focused on vector knowledge and source reduction, personal protective measures, and knowledge of the disease and causative agent. In the absence of DENV-specific antivirals and the availability of two tetravalent vaccines [90], vaccination could be an alternative for prevention, especially in high-risk population areas, as recommended by the WHO [91].
## 5 | Conclusion
In a context of limited resources, the presence of conditions favourable to the spread of the disease, vectors, and factors that can lead to misdiagnosis, misclassification, and underreporting, DF remains a neglected tropical disease and a public health concern in Burkina Faso. Despite efforts to fight the DF epidemics, some gaps still exist and need to be addressed. It is therefore essential to implement comprehensive and proactive measures as well as strengthen regional collaboration to improve disease control and prevention. Burkina Faso has recently adopted a novel integrated surveillance strategy for viral haemorrhagic fevers (VHF) for 2025-2029, accompanied by a technical guide. Its implementation will offer promising prospects for improvement. It includes a 'one-health' approach aimed at enhancing multisectoral integration of VHF surveillance, including DF. In the future, evaluating this novel strategy could determine its effectiveness in combating the disease in Burkina Faso.
## References
1. Denv-2 "Eastern upper volta 1982 [19] Vectors DENV"
2. Bobo, Ouagadougou (2013) "Human DENV-3 Central region 2016 [15] Human DENV-2 predominating Central region and around ouagadougou 2017 [21] Human DENV-1, DENV-2, and DENV-3 with DENV-2 predominating All health regions 2023 [16] Human DENV-1, DENV-2, and DENV-3 with DENV-3 predominating All health regions References"
3. (2005) "Natural Cycles of Vector-borne Pahogens Biology of Disease Vectors"
4. (2023) "Dengue and Severe Dengue Fact Sheet"
5. Bhatt, Gething, Brady (2013) "The Global Distribution and Burden of Dengue" *Nature*
6. Al, Das, Nesa (2023) "Importance of Wolbachia-mediated Biocontrol to Reduce Dengue in Bangladesh and Other dengue-endemic Developing Countries" *Biosafety and Health*
7. Fagbami, Onoja (2018) "Dengue Haemorrhagic Fever: An Emerging Disease in Nigeria, West Africa" *Journal of Infection and Public Health*
8. Gainor, Harris, Labeaud (2022) "Uncovering the Burden of Dengue in Africa: Considerations on Magnitude, Misdiagnosis, and Ancestry" *Viruses*
9. Im, Balasubramanian, Ouedraogo (2020) "The Epidemiology of Dengue Outbreaks in 2016 and 2017 in Ouagadougou" *Heliyon*
10. Ouattara, Traore, Sangare et al. (2016) "Spatiotemporal Analysis of Dengue Fever in Burkina Faso From"
11. Hashimoto, Kutsuna, Maeki (2017) "A Case of Dengue Fever Imported From Burkina Faso to Japan in October 2016" *Japanese Journal of Infectious Diseases*
12. Bilgo (2024) "The Unseen Battle: Interpreting the 2023 World Malaria Report From Burkina Faso's Frontlines"
13. Zongo, Carabali, Munoz et al. (2018) "Dengue Rapid Diagnostic Tests: Health Professionals' Practices and Challenges in Burkina Faso" *SAGE Open Medicine*
14. Fernandes-Santos, De Azeredo (2022) "Innate Immune Response to Dengue Virus: Toll-Like Receptors and Antiviral Response" *Viruses*
15. Ouattara, Bello, Traoré et al. (2025) "Predominance of DENV-3 Among Patients in Ouagadougou, Burkina Faso" *Journal of Vector Borne Diseases*
16. Leitmeyer, Vaughn, Watts (1999) "Dengue Virus Structural Differences That Correlate With Pathogenesis" *Journal of Virology*
17. Lim, Seydou, Carabali (2019) "Clinical and Epidemiologic Characteristics Associated With Dengue During and Outside the 2016 Outbreak Identified in Health Facility-Based Surveillance in Ouagadougou" *PLoS Neglected Tropical Diseases*
18. Poda, Da, Somé (2025) "Deadly Dengue Epidemic Outbreak in Burkina Faso in 2023" *International Journal of Infectious Diseases*
19. Amarasinghe, Kuritsk, Letson et al. (2011) "Dengue Virus Infection in Africa" *Emerging Infectious Diseases*
20. Hervy, Legros, Roche et al. (1984) "Circulation du Virus Dengue 2 Dans Plusieurs Milieux Boisés des Savanes Soudaniennes de la Région De Bobo-Diou-lasso" *Cahiers-ORSTOM Entomologie médicale et parasitologie*
21. Gonzalez, Du, Saussay et al. (1985) "Dengue in Burkina Faso (ex-Upper Volta): Seasonal Epidemics in the Urban Area of Ouagadougou]" *Bulletin de la Societe de Pathologie Exotique*
22. Tarnagda, Congo, Sagna (2013) "Outbreak of Dengue Fever in Ouagadougou" *International Journal of Microbiology and Immunology Research*
23. Letizia, Pratt, Wiley (2022) "Retrospective Genomic Characterization of a 2017 Dengue Virus Outbreak" *Emerging Infectious Diseases*
24. Bello, Tapsoba, Zoure (2024) "Molecular Characterization of the Four Serotypes (DENV-1, DENV-2, DENV-3 and DENV-4) of Dengue Virus Circulating in Ouagadougou" *Open Journal of Epidemiology*
25. Fiestas Solorzano, Da Costa Faria, Santos (2021) "Different Profiles of Cytokines, Chemokines and Coagulation Mediators Associated With Severity in Brazilian Patients Infected With Dengue Virus" *Viruses*
26. Tsai, Chuang, Lin et al.
27. (2013) "An Emerging Role for the Anti-Inflammatory Cytokine Interleukin-10 in Dengue Virus Infection" *Journal of Biomedical Science*
28. Pare, Neupane, Eskandarian (2020) "Genetic Risk for Dengue Hemorrhagic Fever and Dengue Fever in Multiple Ancestries" *EBioMedicine*
29. Lee, Voon, Lim et al. (2022) "Innate and Adaptive Immune Evasion by Dengue Virus" *Frontiers in Cellular and Infection Microbiology*
30. Srikiatkhachorn, Mathew, Rothman (2017) "Immune-Mediated Cytokine Storm and its Role in Severe Dengue" *Seminars in Immunopathology*
31. Tantawichien, Thisayakorn (2017) *Dengue. Neglected Tropical Diseases-South Asia*
32. Bhatt, Varma, Sood (2024) "Temporal Cytokine Storm Dynamics in Dengue Infection Predicts Severity" *Virus Research*
33. Malavige, Fernando, Fernando et al. (2004) "Dengue Viral Infections" *Postgraduate Medical Journal*
34. Malavige, Jeewandara, Ogg (2020) "Dysfunctional Innate Immune Responses and Severe Dengue" *Frontiers in Cellular and Infection Microbiology*
35. Khanam, Gutiérrez-Barbosa, Lyke et al. (2022) "Immune-Mediated Pathogenesis in Dengue Virus Infection" *Viruses*
36. Rouers, Chng, Lee (2021) "Immune Cell Phenotypes Associated With Disease Severity and Long-Term Neutralizing Antibody Titers After Natural Dengue Virus Infection" *Cell Reports Medicine*
37. Sebayang, Fahlena, Anam (2021) "Modeling Dengue Immune Responses Mediated by Antibodies: A Qualitative Study" *Biology*
38. Kamath, Olakkengil (2023) "Original Antigenic Sin in Dengue -Hoskins Effect" *APIK Journal of Internal Medicine*
39. De Arruda, Bavia, Mosimann (2023) "Viremia and Inflammatory Cytokines in Dengue: Interleukin-2 as a Biomarker of Infection, and Interferon-α and -γ as Markers of Primary Versus Secondary Infection" *Pathogens*
40. Singla, Kar, Sethi (2016) "Immune Response to Dengue Virus Infection in Pediatric Patients in New Delhi, India-Association of Viremia, Inflammatory Mediators and Monocytes With Disease Severity" *PLoS Neglected Tropical Diseases*
41. Tian, Grifoni, Sette et al. (2019) "Human T Cell Response to Dengue Virus Infection" *Frontiers in Immunology*
42. Manh, Weiss, Thuong (1980) "Kinetics of CD4(þ) T Helper and CD8(þ) Effector T Cell Responses in Acute Dengue Patients" *Frontiers in Immunology*
43. Zompi, Harris (2013) "Original Antigenic Sin in Dengue Revisited" *Proceedings of the National Academy of Sciences*
44. Rothman (2011) "Immunity to Dengue Virus: A Tale of Original Antigenic Sin and Tropical Cytokine Storms" *Nature Reviews Immunology*
45. Becquart, Wauquier, Nkoghe (2010) "Acute Dengue Virus 2 Infection in Gabonese Patients Is Associated With an Early Innate Immune Response, Including Strong Interferon Alpha Production" *BMC Infectious Diseases*
46. Dagenais, Hébert, Ridde (2021) "Video as an Effective Knowledge Transfer Tool to Increase Awareness Among Health Workers and Better Manage Dengue Fever Cases" *Journal of Global Health Reports*
47. Bello, Houkpevi, Zackari (2022) "Epidemiology of Dengue in Patients With Febrile Syndrome at Saint Camille Hospital, Ouagadougou, Burkina Faso From 2020 to 2021" *African Journal of Clinical and Experimental Microbiology*
48. Tougma, Zoungrana/Yaméogo, Dahourou (2017) "Dengue Virus Infection and Pregnancy Outcomes During the" *PLoS One*
49. (2023) *Health Statistics Yearbook*
50. (2024) *Health Statistics Yearbook*
51. (2022) *Health Statistics Yearbook*
52. (2021) *Health Statistics Yearbook*
53. (2020) *Health Statistics Yearbook*
54. Ouedraogo, Ilboudo, Bado (2024) "Estimating Dengue Burden Among Family Contacts Through Cluster Investigation Around Probable Cases in 2022 and 2023 in the Central Region" *Infectious Diseases of Poverty*
55. Ouedraogo, Ilboudo, Compaore (2024) "Determinants and Prevalence of Symptomatic Dengue Fever Among Adults in the Central Region of Burkina Faso: A Hospital-based cross-sectional Study" *BMC Infectious Diseases*
56. Ouédraogo, Debe, Ilboudo (2023) "Séroprévalence et facteurs de risque démographiques de la dengue au Burkina Faso: résultats d'une enquête nationale" *Revue d'Épidémiologie et de Santé Publique*
57. Donatien, Hien, Salam (2023) "Seroepidemiological Study of Dengue Virus Infection Suspected Cases in Burkina Faso" *Journal of Biosciences and Medicines*
58. Sondo, Diendere, Meda (2015) "Severe Dengue in Adults and Children"
59. Mungrue (2014) "The Laboratory Diagnosis of Dengue Virus Infection, a Review" *Advance Laboratory Medicine International*
60. Wiwanitkit (2010) "Dengue Fever: Diagnosis and Treatment" *Expert Review of Anti-infective Therapy*
61. Kabir, Zilouchian, Younas et al. (2021) "Dengue Detection: Advances in Diagnostic Tools From Conventional Technology to Point of Care" *Biosensors (Basel)*
62. Hunsperger, Sharp, Lalita (2016) "Use of a Rapid Test for Diagnosis of Dengue During Suspected Dengue Outbreaks in Resource-Limited Regions" *Journal of Clinical Microbiology*
63. Blacksell (2012) "Commercial Dengue Rapid Diagnostic Tests for Point-of-Care Application: Recent Evaluations and Future Needs?" *Journal of Biomedicine and Biotechnology*
64. Ndiaye, Woolston, Gaye (2023) "Laboratory Evaluation and Field Testing of Dengue NS1 and IgM/IgG Rapid Diagnostic Tests in an Epidemic Context in Senegal" *Viruses*
65. Guzman, Gubler, Izquierdo et al. (2016) "Dengue Infection" *Nature Reviews Disease Primers*
66. Falconi-Agapito, Kerkhof, Merino (2022) "Peptide Biomarkers for the Diagnosis of Dengue Infection" *Frontiers in Immunology*
67. Tinto, Kaboré, Kania (2022) "Serological Evidence of Zika Virus Circulation in Burkina Faso" *Pathogens*
68. Manigart, Ouedraogo, Ouedraogo et al. (2024) "Dengue Epidemic in Burkina Faso: How can the Response Improve?" *Lancet*
69. Gomgnimbou, Belem, Some (2024) "Utilization of Novel Molecular Multiplex Methods for the Detection and, Epidemiological Surveillance of Dengue Virus Serotypes and Chikungunya Virus in Burkina Faso, West Africa" *Molecular Biology Reports*
70. Gomgnimbou, Belem, Bilgo (2024) "Potential Performance of Two New RT-PCR and RT-qPCR Methods for Multiplex Detection of Dengue Virus Serotypes 1-4 and Chikungunya Virus in Mosquitoes" *Current Issues in Molecular Biology*
71. Vasilakis, Cardosa, Hanley et al. (2011) "Fever From the Forest: Prospects for the Continued Emergence of Sylvatic Dengue Virus and its Impact on Public Health" *Nature Reviews Microbiology*
72. Saepudin, Kasjono, Martini (2022) "Detection of Dengue Virus Transovarial Transmission in Dengue Hemorrhagic Fever Endemic Areas" *KEMAS: Jurnal Kesehatan Masyarakat*
73. Hien, Sangaré, Sawadogo (2022) "Chikungunya (Togaviridae) and Dengue 2 (Flaviviridae) Viruses Detected From Aedes aegypti Mosquitoes in Burkina Faso by qRT-PCR Technique: Preliminary Results and Perspective for Molecular Characterization of Arbovirus Circulation in Vector Populations" *Frontiers in Tropical Diseases*
74. Ouattara, Sangare, Namountougou (2019) "Surveys of Arboviruses Vectors in Four Cities Stretching Along a Railway Transect of Burkina Faso: Risk Transmission and Insecticide Susceptibility Status of Potential Vectors" *Frontiers in Veterinary Science*
75. Simmons, Farrar, Nguyen et al. (2012) "Dengue" *New England Journal of Medicine*
76. Kayesh, Khalil, Kohara et al. (2023) "Increasing Dengue Burden and Severe Dengue Risk in Bangladesh: An Overview" *Tropical Medicine and Infectious Disease*
77. Ouattara, Traore, Traore et al. (2022) "Climate Factors and Dengue Fever in Burkina Faso From 2017 to 2019" *Journal of Public Health in Africa*
78. Otshudiema, Diao, Ouedraogo (2018) "Estimating Dengue Outbreak Thresholds in West Africa: A Comprehensive Analysis of Climatic Influences in Burkina Faso"
79. Nasir, Bakare, Ayodeji et al. (2018) "Prevention of Dengue Virus Infection" *Saudi Journal of Medicine & Medical Sciences*
80. Diallo, Diouf, Gaye (2022) "Dengue Vectors in Africa: A Review"
81. Hien, Thiombiano, Ilboudo (2021) "Analysis of Immune Markers and Hematological Features in Plasmodium/Dengue Virus Co-Infected Patients in Ouagadougou"
82. Saré, Pérez, Somé et al. (2018) "Community-Based Dengue Control Intervention in Ouagadougou: Intervention Theory and Implementation Fidelity" *Global Health Research and Policy*
83. Diallo, Schioler, Samuelsen et al. (2022) "formation System as Part of Epidemic Management in Burkina Faso: From Plan to Reality (Field Findings)"
84. Sanou, Dirlikov, Sondo (2017) "Building Laboratory-Based Arbovirus Sentinel Surveillance Capacity During an Ongoing Dengue Outbreak"
85. Diao, Bicaba, Sanou et al. (2024) "After Action Review: Preparedness and Response to the Dengue Epidemic in 2023 in the Centre Region" *Frontiers in Tropical Diseases*
86. (2020) "Pan American Health Organization/World Health Organization (PAHO/WHO)"
87. Potts, Rothman (2008) "Clinical and Laboratory Features That Distinguish Dengue From Other Febrile Illnesses in Endemic Populations" *Tropical Medicine and International Health*
88. Ouédraogo, Benmarhnia, Bonnet (2018) "Evaluation of Effectiveness of a Community-Based Intervention for Control of Dengue Virus Vector"
89. (2023) "World Health Organization. Regional Office for Africa" *Weekly Regional Dengue Bulletin*
90. Khan, Adil, Wang (2022) "A cross-sectional Study to Assess the Epidemiological Situation and Associated Risk Factors of Dengue Fever; Knowledge, Attitudes, and Practices About Dengue Prevention in Khyber Pakhtunkhwa Province, Pakistan" *Frontiers in Public Health*
91. Ouoba, Dori, Semdé (2024) "Dengue Epidemic in Burkina Faso: Concerns About the Informal Use of Traditional Herbal Remedies" *Pan African Medical Journal*
92. Rosales-Rosas, Goossens, Chiu (2024) "The Antiviral JNJ-A07 Significantly Reduces Dengue Virus Transmission by Aedes aegypti Mosquitoes When Delivered via blood-Feeding" *Science Advances*
93. (2023) "Conclusions and Recommendations" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12697093&blobtype=pdf | # Phages infecting Bacteroides thetaiotaomicron
Joanna Steczynska, Kelly Williams, John Dennehy
## Abstract
Bacteroides thetaiotaomicron is a prevalent human gut commensal that plays an important role in polysaccharide breakdown and is of interest as a probiotic candidate. We report isolation of lytic phages able to infect B. thetaiotaomicron, with nine unique genome sequences belonging to four congeneric species. KEYWORDS phage, gut commensal, bacteroides, bacteroides phage, Bacteroides thetaiotaomicron, anaerobes B acteroides thetaiotaomicron (Bt) is a common human gut commensal, with potential therapeutic uses (1). Here, we report isolation of nine phages for this organism that fall into four congeneric species.Anaerobic digester sewage samples were harvested from the Livermore Water Reclamation Plant (Livermore, CA, USA) wastewater facility on 20 September 2024. Samples were centrifuged, and the supernatants were filtered (0.45 µm) and enriched by adding an equal volume of fresh BHI medium with 1:100 dilution of B. thetaiotaomi cron VPI-5482. All growth was anaerobic (Don Whitley, A35 chamber, 5% H2), at 37°C. Enriched supernatant was filtered, assayed for plaques on BHI media (37 g/L Millipore), supplemented with yeast extract (5 g/L, Fisher Scientific) and 5% each of newborn calf serum and sheep serum (RmBio). Plaques were clear; 41 were plaque-purified 3 times. Genomic DNA was extracted from high-titer (>10 8 pfu/mL) stocks using the Norgen Biotek Phage DNA isolation kit, and libraries were prepared using the Illumina DNA prep kit and sequenced using the MiSeq V3 150-cycle kit in paired-end mode using default protocols.Reads for 36 phages were assembled using SPAdes (2) v3.15.5. Each assembly yielded a single (~38 kbp) high-coverage contig, except for BT24, whose seven high-coverage contigs were readily assembled based on 55 bp overlaps confirmed with ReadStepper (3). Genomic DNA libraries were re-sequenced using the kit above, in single-read mode to resolve the repetitive gene 21/22 intergentic space in combination with ReadStepper (3). The longer (150 bp) reads also enabled resolution of a problematic tandem repeat region in gene 16, except for BT03 where a tandem 9 bp repeat was truncated at its minimum supported length of 16 copies. Ultimately, all genomes formed closed circles, and the set was reduced to nine unique genome sequences (Table 1). These had close sequence relationships to the 27 genomes of the alpha cluster of previously reported Bt phages (4), to which we added 18 related genomes (names formulated as BTxPy) found at NCBI. Phylogenetic analysis of this alpha-cluster genome set (Fig. 1A) revealed four groups of new phages, represented by BT03, BT04, BT12, and BT24, each considered a separate species within the same genus using VIRIDIC (5), all determined to be tailed phages. There were no matches to any of the 8,630 reference ICTV phages, so taxonomic names were not assigned. Pharokka (6) v1.7.5 (flags: -g prodigal-dnapler) was used for genome annotation, leaving many genes without assigned function. Annotations were verified and further manually supplemented by BLASTP and HMM searches (HHPred https://toolkit.tuebingen.mpg.de/tools/hhpred and HMMER http://hmmer.org/). Default parameters were used for all software unless specified. Gene order comparison (Fig. 1B)
revealed a conserved virion gene cluster of structural genes (BT03 g6-18) and a less conserved cluster of replication genes (BT03 g25-46, itself with a conserved core g26-28), as well as lytic and terminase functions. The main variation among new phages was in the presence and type of DNA methyltransferase genes, which likely influence host range. Sequencing of these phages provides an avenue for genetic manipulation of B. thetaiotaomicron and better understanding of its viruses.
## References
1. Margaret, Imke, Elizabeth et al. (2019) "Bacteroides thetaiotao micron ameliorates colon inflammation in preclinical models of Crohn's Disease" *Inflamm Bowel Dis*
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. Hryckowian, Merrill, Porter et al. (2020) "Bacteroides thetaiotao micron-infecting bacteriophage isolates inform sequence-based host range predictions" *Cell Host Microbe*
5. Moraru, Varsani, Kropinski (2020) "VIRIDIC-A novel tool to calculate the intergenomic similarities of prokaryote-infecting viruses" *Viruses*
6. Bouras, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737738&blobtype=pdf | # Co-Circulation of Tick-Borne Bandaviruses and Orthonairoviruses Across Humans, Livestock, and Rats in Pakistan: Serologic Evidence and Public Health Implications
Nikos Vasilakis, Pat Nuttall, Muhammad Ammar, Shengyao Chen, Muhammad Saqib, Jingyuan Zhang, Awais-Ur-Rahman Sial, Asad Zia, Yaohui Fang, Muhammad Khalid Mansoor, Abulimiti Moming, Asim Shahzad, Rehman Hafeez, Aneela Javed, Ali Hassan, Ben Hu, Ali Zohaib, Shu Shen, Fei Deng
## Abstract
Tick-borne viruses (TBVs) pose significant public health and economic threats. Pakistan has endemic Crimean-Congo hemorrhagic fever virus (CCHFV), but evidence suggests broader TBV circulation. This study assessed the seroprevalence of thirteen TBVs (seven are members of the genus Orthonairovirus and six are members of the genus Bandavirus) in humans, livestock, and rats in Punjab, Pakistan. Serum samples (n = 794: 321 livestock, 253 human, and 220 rat) were collected from the Narowal, Lahore, and Faisalabad districts. Antibodies to viral NPs were detected using the luciferase immunoprecipitation system (LIPS). The overall seroprevalence was 19.14% (152/794); it was highest in livestock (27.10%), then humans (20.55%), and then rats (5.91%). The highest seroprevalence rates were 3.12% for CCHFV in livestock, 3.56% for Yezo virus (YEZV) in humans, and 0.91% for Tamdy virus (TAMV) and Tacheng tick virus 1 (TcTV-1) in rats. Neutralizing antibodies were detected against CCHFV (1 cattle, 4 humans), Bhanja virus (BHAV) (3 livestock, 1 rat), TAMV (1 cattle), Guertu virus (GTV) (1 cattle), and Dabie bandavirus (2 cattle). Sixteen samples showed antibodies to both orthonairoviruses and bandaviruses, indicating co-exposure. Further analysis showed that seropositivity was not randomly distributed. Livestock kept in commercial farming systems and people working mainly outdoors had distinctly higher
## 1. Introduction
Tick-borne viral diseases (TBVDs), caused by tick-borne viruses (TBVs), have attracted increasing attention due to the growing threats they pose to global public health and their consequent heavy economic burden on the medical and livestock industries [1,2]. For an extended period of time, the tick-borne CCHFV has been circulating in Pakistan, while serologic evidence suggests the presence of other TBVs having caused human exposure there [3][4][5].
Ticks serve as a critical reservoir of TBVs and act as the major vector for the transmission of TBVs by taking blood meals from hosts. Currently, over 40 tick species from Hyalomma, Haemaphesalis, and Rhipicephalus genera are widely distributed in Pakistan, particularly in the Punjab region, where they frequently infest livestock, including cattle, buffalo, sheep, and goats, and are established as the principal vectors of TBVs [6,7]. Surveys carried out in Punjab have also reported high levels of tick infestation in livestock across several districts in central and northeastern parts of the province, such as Narowal, Lahore and Faisalabad, which share similar ecological conditions [2,8]. Ticks feed as larvae and nymphs on a wide range of vertebrate hosts before developing into adults, which brings them into regular contact with large and small ruminants. Consequently, livestock like cattle, buffalo, goats, and sheep in Punjab could be repeatedly exposed to feeding ticks when moving through farms, grazing areas, and animal markets [9,10]. Humans are at potential risk of exposure when they care for or handle these animals, particularly during herding, milking, slaughtering, and agricultural work, where ticks can easily attach to clothing or exposed skin [11]. Although livestock are the primary hosts, rodents, including rats, can also interact with questing ticks in animal sheds, fodder piles, and grain stores in mixed farm environments, and rodents are recognized as important in maintaining and amplifying several TBV hosts [12]. All this establishes an ecological overlap between humans, livestock, and rats potentially linked by ticks, which is supported by previous field surveys showing high infestation rates of ticks on these animals across Punjab [9,10]. Furthermore, viruses persisting in tick populations throughout generations due to transstadial and transovarial transmission increase their chances of spreading to hosts during the whole life cycle [13].
In Pakistan's complex tick-borne viral landscape, bandaviruses and orthonairoviruses represent the most significant public health challenges due to their notable transmissibility and high fatality rates. The orthonairovirus CCHFV is one of the most lethal tick-borne pathogens long endemic to Pakistan. Human infection with CCHFV leads to severe hemorrhagic disease with high mortality rates and can trigger nosocomial transmission through human-to-human contact, posing an ongoing threat to livestock workers and healthcare personnel [14]. Among bandaviruses, the SFTSV exemplifies an emerging tickborne viral threat. First identified in China's Henan Province in 2011 [15], SFTSV has subsequently been reported in South Korea and Japan [16,17], demonstrating a trend of continuous geographical expansion. Seropositivity detected in the Pakistani population in 2020 further indicates the potential risk of local circulation [4]. Although SFTSV shares some clinical manifestations with CCHFV, it is more prone to underdiagnosis, and its true prevalence in Pakistan remains unclear. Research on these viruses is critically important for establishing disease early-warning systems and presents a proactive strategy for mitigating large-scale outbreaks of emerging and re-emerging infectious diseases.
To address this gap, serum samples from humans, livestock, and rats were collected from Punjab's integrated economic zones, where human-livestock-rodent contact intensifies spillover risk [4,14,18]. Serologic exposure to thirteen TBVs-including seven members of orthonairoviruses (TAMV, CCHFV, Wenzhou tick virus (WzTV), Songling virus (SGLV), Huangpi tick virus-1 (HpTV-1), TcTV-1, and YEZV) and six members of bandaviruses (SFTSV, GTV, Heartland virus (HRTV), BHAV, Lone Star virus (LSV), and Hunter Island group virus (HIGV))-were investigated, and subsequently neutralization antibodies to specific viruses were examined. Possible routes for TBV circulation among humans, livestock, and rats in Pakistan and potential risk factors were discussed based on the results of serologic tests. The findings will improve our understanding of TBV prevalence and may facilitate the prevention and control of TBV-associated diseases in Pakistan [4,14].
## 2. Materials and Methods
## 2.1. Sample Collection
Serum samples were collected in 2022 from humans, rats, and animals (n = 794: 321 livestock (90 buffalo, 200 cattle, 14 dogs, 12 goats, and 5 sheep), 253 humans (173 males and 80 females), and 220 rats (77 males and 143 females)) in Punjab's integrated economic zones, which include Narowal (32 • N, 74.8 • E), Lahore (31.5 • N, 74.3 • E), and Faisalabad (31.5 • N, 74 • E). Human sampling employed a convenience strategy due to sociopolitical constraints, limited healthcare infrastructure, and logistical challenges in obtaining farmer participation. Random selection was prioritized where feasible. All the participants gave their informed consent verbally or in writing, regardless of their reading level. Blood (4 mL) was taken from consenting study participants from the peripheral vein into the gel-clot activator containing vacutainer (Improve Medical, Guangzhou, China). Livestock were sampled from households, animal selling markets, and farms where human sampling was conducted, forming spatial correlation for a comparative investigation into exposure to tick-borne viruses. The selection of livestock was also made conveniently, taking into consideration owner permission and farm access.
Rat samples were collected from places where there is high human and animal interaction, particularly from local godowns, shops, livestock farms, and open markets, important sites for the inter-district transportation of food, animals, animal products, crops, and forages. Livestock and rat blood (4 mL) were collected by veterinarians via the jugular or tail veins using identical vacutainers. The samples were shipped to The Islamia University of Bahawalpur, Punjab, Pakistan, under continuous cold-chain maintenance. After centrifugation (5000× g, 12 min), serum was aliquoted into cryovials (Imec Medical Co., Zhangjiagang, China) and stored at -40 • C until subjected to further experimentation.
All the procedures involving human participants and animals were reviewed and approved by the Departmental Bioethics Committee, Department of Microbiology, The Islamia University of Bahawalpur (REC. BEC No. 05-2021-21/2; approval date: 2 May 2021), Pakistan, and by the ethics committees of Wuhan Institute of Virology, Chinese Academy of Sciences (WIVHF33202404; approval date: 29 September 2024). These committees also serve as the institutional animal ethics and animal care and use bodies for this project, and all livestock and rat sampling was conducted in accordance with their approved protocols and relevant animal welfare regulations.
## 2.2. Cells, Viruses, and Antibodies
Human embryonic kidney cells (CRL-11268), African green monkey kidney cells (CCL-81), and African green monkey kidney E6 cells (ATCC-1586) were acquired from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultivated in Eagle's Minimum Essential Medium (EMBM) (NZK Biotech, Nanjing, China) with 10% FBS and Dulbecco's Modified Eagle Medium (DMEM) (NZK Biotech, Nanjing, China) with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA) and 1% penicillin-streptomycin (PS; Gibco, Grand Island, NY, USA).
The SFTSV strain WCH (IVCAS6.6088; GenBank accession no. JQ341190.1), CCHFV strain YL16070 (IVCAS6.6329; GenBank accession no. KY354080.1), TAMV strain YL16082 (IVCAS6.7499; GenBank accession no. MT815991.1), HRTV strain MO-4 (IV-CAS6.6330; GenBank accession no.LC629153.1), GTV strain DXM (IVCAS6.6106; GenBank accession no.KT328591.1), HpTV-1 strain HTV1/SZYD9 (IVCAS6.9465; GenBank accession no.MW721869.1), and BHAV strain M3811 (IVCAS6.9001; GenBank accession no. JQ956378.1) were obtained from the National Virus Resource Center (NVRC), China.
Rabbit polyclonal antibodies against nucleoprotein (NP) of SFTSV [19], CCHFV [20], TAMV [21], and GTV [22] were previously prepared, and the rabbit polyclonal antibodies against NPs of HRTV, HpTV-1, and BHAV were prepared in-house following the procedure outlined in previous studies [20][21][22]. The anti-FLAG Tag antibody (Sangon Biotech, Shanghai, China) was used as the primary antibody in verifying the expression of the LIPS-related fusion protein. GAPDH mouse mAb (abclonal, Wuhan, China) was used as the inner control. The goat anti-rabbit IgG (H + L) conjugated with horseradish peroxidase (HRP) (Proteintech, Wuhan, China) was used as the secondary antibody to verify the expression of LIPS-related fusion protein. Goat anti-rabbit IgG H&L conjugated with Alexa Fluor 488 (Abcam, Shanghai, China), goat anti-human IgG H&L conjugated with horseradish peroxidase (HRP) (Abcam, Shanghai, China), and HRP-Protein A (proteintech, Wuhan, China) were used as secondary antibodies for immunofluorescence assays (IFAs) or Western blot.
## 2.3. Reagents
Hoechst 33258 (Beyotime, Shanghai, China) was used for staining cell nuclei. ClonExpress ® Ultra One-step Cloning Kit (Vazyme, Nanjing, China) for gene cloning in pREN2 plasmids in-frame with the Renilla luciferase (Rluc) and Flag tag. Lipofectamine TM 3000 (Invitrogen, Carlsbad, CA, USA) was used to transfect plasmids into 293T cells. Sea kidney luciferase reporter gene cell lysate buffer (Beyotime, Shanghai, China). Pierce Protein A/G UltraLink Resin beads (Thermo Fisher Scientific, Waltham, MA, USA), Renilla-Lumi™ Plus and Luciferase Assay Kit (Beyotime, Shanghai, China).
## 2.4. Serological Testing Using Luciferase Immunoprecipitation System (LIPS) Followed by Western Blot and Microneutralization Assays
The presence of antibodies was initially detected using the luciferase immunoprecipitation system (LIPS) assays and subsequently confirmed by Western blot as previously described [23]. To construct the plasmids expressing viral antigen, the NP genes of thirteen viruses, including TcTV-1, HpTV-1, WzTV, SGLV, YEZV, HRTV, HIGV, BHAV, and LSV, were cloned into pREN2 plasmids in-frame with the Renilla luciferase (Rluc) and Flag tag by using the ClonExpress ® Ultra One-Step Cloning Kit (Vazyme, Nanjing, China) according to the manufacturer's instructions, and were further verified by Sanger sequencing. For CCHFV, TAMV, SFTSV, and GTV, NP-based plasmid constructs and related assays had been previously established and validated by [23] and were used in this study as controls for validating viral protein expression. Subsequently, the plasmids inserted with or without the viral NP gene were transfected into HEK293T cells using Lipofectamine™ 3000 Transfection Reagent (Invitrogen, America). At 48 h post-transfection, cells were fixed to visualize viral protein expression by IFAs or lysed using sea kidney luciferase reporter gene lysis buffer (RG129M, Beyotime, Shanghai, China) and subjected to Western blot to validate expression of viral antigens. The anti-FLAG antibody was used as the primary antibody for both tests. The remaining cell lysates were either stored at -80 • C for later analysis or used immediately for LIPS assays as previously described [23]. The Luminescence Unit (LU) values for each tested sample were measured using Renilla-Lumi™ Plus Luciferase Assay Kit (Beyotime, Shanghai, China) by using a luminometer, the GloMax Multi+ Detection System (Promega, Madison, WI, USA). The cut-off value for antibody positivity was determined by calculating the average and standard deviation (STDEV) of LU values from all tested samples, which was expressed as:
Cut-off = Mean LU + 3 × STDEV Samples with LU values above this cut-off were considered seropositive. The expression of the NP genes of thirteen viruses, including CCHFV, TAMV, TcTV-1, HpTV-1, WzTV, SGLV, YEZV, SFTSV, GTV, HRTV, HIGV, BHAV, and LSV, in fusion with Flag-tag in the plasmid-transfected cells was verified by Western blot using anti-FLAG Tag antibody (Sangon Biotech, Shanghai, China) (1:1000 dilution) as the primary antibody and goat anti-rabbit IgG (H + L) conjugated with horseradish peroxidase (HRP) (Proteintech, Wuhan, China) (1:3000 dilution) as the secondary antibody. GAPDH expression was blotted using GAPDH Mouse mAb (abclonal, China) as the inner control. To confirm the seropositive samples identified by LIPS, Western blot was performed using purified respective viral particles as antigens, followed by incubation with serum samples or specific polyclonal viral antibodies as controls and the commercially provided HRP-labeled secondary antibodies specific to animals or humans [23].
To identify neutralizing antibodies, serum samples were heat-inactivated at 56 • C for 30 min to inactivate complement and other non-specific antiviral factors. Then, the presence of neutralizing antibodies against CCHFV, TAMV, HpTV-1, SFTSV, GTV, HRTV, and BHAV were detected using the method as previously described [3,21,22]. IFAs were performed to visualize virus infection in each well. For each dilution, the tests were performed in triplicate. The neutralization titer was defined as the reciprocal of the highest serum dilution that completely prevented virus infection in all three replicate wells.
## 2.5. Infection Assays and Virus Purification
We cultured cells for virus propagation using Vero cells for SFTSV, HRTV, GTV, and BHAV and Vero E6 cells for CCHFV, TAMV, and HpTV-1. Cells were seeded in T75 tissue culture flasks and maintained in Dulbecco's Modified Eagle Medium (DMEM, 2%) and Eagle's Minimum Essential Medium (EMEM, 2%), respectively. Both media were supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. Cells were incubated at 37 • C in a humidified atmosphere containing 5% CO 2-until they reached approximately 80-90% confluency. For viral infection, 100 µL of virus stock solution was added to each flask, followed by incubation for five days.
At 5 days post-infection (p.i.), supernatants were collected and centrifuged at 4000× g to remove residual cells. Viral particles were concentrated using a polyethylene glycol (PEG8000) precipitation method. A 5× PEG8000/NaCl stock solution was prepared by dissolving 50 g of PEG8000 and 8.766 g of NaCl in 200 mL of distilled water and was autoclaved at 121 • C for 30 min. Virus-containing supernatants were filtered through a 0.45 µm membrane filter, mixed with the pre-cold PEG8000/NaCl solution (w/w = 5:1), and incubated at 4 • C overnight. Viral particles were pelleted by centrifugation at 4000× g for 20 min at 4 • C. The viral pellet was harvested and applied to Western blot as antigens to examine antibody response among serum samples [24]. The laboratory-prepared rabbit anti-nucleoprotein (Np) antibodies (1:2000) to respective viruses were used as the primary antibodies. For human serum samples HRP-conjugated goat anti-human IgG (H + L) (1:5000 dilution) or HRP-Protein A (1:5000 dilution) were used as the secondary antibodies to immunoblot antibody response from human serum samples or livestock and rats, respectively.
## 2.6. Phylogenetic Analysis and Pairwise Sequence Identity
The nucleotide sequences of the nucleoprotein (NP) gene segment were aligned by ClustalW (v2.1) implemented in MEGA 11 (version 11.0.10; Mega Limited, Auckland, New Zealand), and the phylogenetic tree was constructed using the maximum likelihood method with 1000 bootstrap replicates. The Gn and NP amino acid sequences were aligned using Clustal Omega from the European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL-EBI, Hinxton, UK; https://www.ebi.ac.uk/Tools/msa/clustalo/, accessed on 15 March 2025) and then analyzed using an online Chiplot tool (https: //www.chiplot.online/, accessed on 15 March 2025) to generate the pairwise percent identity matrix. Genotypes were distinguished according to the procedure described in previous studies [25,26].
## 2.7. Statistical Analysis
Statistical analysis was performed using SPSS Statistics Version 21 (IBM Corporation, Armonk, NY, USA) and GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Chi-squared testing examined correlations between host risk factors and viral infection with a 95% confidence interval. Positive samples were also analyzed statistically for associations between age, gender, job, and animal contact for humans; sex and environment for rats; animal type (buffalo, cattle, dogs, goats, and sheep), sex, and farming system for livestock by using the χ 2 test (or Fisher's exact test) at a 95% confidence interval. For all analyses, a p-value of <0.05 was considered significant.
## 3. Results
We collected 794 serum samples from 321 livestock (buffalo, cattle, dogs, goats, and sheep), 253 humans (convenience-sampled), and 220 rats. Antibody responses to seven members of orthonairoviruses (TAMV, CCHFV, WzTV, SGLV, HpTV-1, TcTV-1, and YEZV) and six members of bandaviruses (SFTSV, GTV, HRTV, BHAV, LSV, and HIGV) were examined among these samples by LIPS, as these viruses are either known human pathogens or have potential to infect humans (Supplementary Table S1). Consequently, an expanded LIPS assay was built by including the nine additional TBVs upon the previously established methods for TAMV, CCHFV, SFTSV, and GTV [5]. NP expression of all thirteen TBVs was confirmed in the transfected cells, with those of TAMV, CCHFV, SFTSV, and GTV revalidated as controls (Supplementary Figure S1). Subsequently, the LIPS assay identified an overall seroprevalence rate of 19.14% (152/794), with livestock exhibiting the highest rate (87/321, 27.10%), followed by humans (52/253, 20.55%) and rats (13/220, 5.91%) (Table 1). Species analysis revealed CCHFV as the most prevalent virus in livestock (10/321, 3.12%), while GTV was the least (4/321, 1.25%). Humans exhibited the highest seroprevalence for Yezo virus (YEZV) (9/253, 3.56%) and the lowest for Tacheng tick virus-1 (TcTV-1) (1/253, 0.40%). Rats demonstrated a higher exposure to TAMV and TcTV-1 (2/220, 0.91%) and lower exposure to the other tested viruses (1/220, 0.45%) (Table 1, Supplementary Figure S2). Antibody responses detected by LIPS for SFTSV, GTV, BHAV, HRTV, TAMV, CCHFV, and HpTV-1 (viruses preserved by China's NVRC) were validated by Western blot using purified viral particles on 85 LIPS-positive samples. The confirmation rates were 50% for SFTSV (6/12), 40% for GTV (4/10), 36.36% for BHAV (4/11), 75% for HRTV (9/12), 100% for TAMV (11/11), 62.50% for CCHFV (10/16), and 46.15% for HpTV-1 (6/12) (Supplementary Figure S3). The variation in confirmation rates may be affected by the differences in method sensitivity or the preparations of purified viral particles.
Limited serologic cross-reaction within the viral groups was suggested by the amino acid identities of either NPs or GPs among the seven orthonairoviruses (NP:44.24-59.96%; GP: 30.91-68.21%) or among the six bandaviruses (NP: 32.27-74.90%; GP: 36.56-71.65%) (Supplementary Figure S4). However, this cross-reaction was still possible, as some samples were detected by Western blot antibody-positive to multiple viruses (Supplementary Figure S3), and the only cross-reaction was previously documented between the SFTSV and GTV bandaviruses [12]. Sixteen samples (11 livestock, 4 human, and 1 rat) showed simultaneous reactivity to both orthonairoviruses and bandaviruses, thereby indicating co-exposure events to the two viral groups (Supplementary Table S2).
We subsequently compared the seroepidemiological rates of the six bandaviruses and seven orthonairoviruses regarding the different factors among livestock, humans, and rats so as to identify critical factor(s) which could be associated with seroprevalence of specific viruses (Table S3). Overall, SFTSV exhibited seroprevalence rates significantly higher in sheep (1/5, 20.00%; 95% CI: 3.62-62.45%) than cattle (6/200, 3.00%; 95% CI: 1.38-6.39%) and in people performing outdoor jobs (4/128, 3.12%; 95% CI: 1.22-7.76%) than those performing indoor jobs (0/125, 0.00%; 95% CI: 0.00-2.98%). Seroprevalence rates of GTV were found to be significantly higher in males (5/173, 2.89%; 95% CI: 1.24-6.59%) than females (0/180, 0.00%; 95% CI: 0.00-2.09%) among humans. HRTV revealed significant variation across animal species, with sheep having the highest occurrence (1/5, 20.00%; 95% CI: 3.62-62.45%), followed by goats (1/12, 8.33%; 95% CI: 1.49-35.39%), dogs (1/14, 7.14%; 95% CI: 1.27-31.47%), cattle (4/200, 2.00%; 95% CI: 0.78-5.03%), and buffalo (1/90, 1.11%; 95% CI: 0.20-6.03%). Among humans, the rates for HRTV were significantly higher in males (3/173, 1.73%; 95% CI: 0.59-4.97%) than females (0/180, 0.00%; 95% CI: 0.00-2.09%). The highest prevalence of HIGV was recorded in sheep (1/5, 20.00%; 95% CI: 3.62-62.45%), followed by goats (1/12, 8.33%; 95% CI: 1.49-35.39%), cattle (4/200, 2.00%; 95% CI: 0.78-5.03%), and buffalo (1/90, 1.11%; 95% CI: 0.20-6.03%), with a significant difference between animal species. Similarly, BHAV also presented the highest rates in sheep (1/5, 20.00%; 95% CI: 3.62-62.45%), followed by goats (2/12, 16.67%; 95% CI: 4.70-44.80%), dogs (1/14, 7.14%; 95% CI: 1.27-31.47%), and cattle (1/200, 0.50%; 95% CI: 0.09-2.78%), with statistically significant differences. Moreover, the seroprevalence rates of BHAV in men (1/95, 1.05%; 95% CI: 0.19-5.72%) were significantly higher than in women (4/226, 1.77%; 95% CI: 0.69-4.46%). For CCHFV, the seroprevalence rates were significantly higher in men (1/95, 1.05%; 95% CI: 0.19-5.72%) than in women (9/226, 3.98%; 95% CI: 2.11-7.39%). Animal species were found significantly associated with TAMV seroprevalence, as the highest rates were recorded in sheep (1/5, 20.00%; 95% CI: 3.62-62.45%), followed by goats (1/12, 8.33%; 95% CI: 1.49-35.39%), dogs (1/14, 7.14%; 95% CI: 1.27-31.47%), buffalo (1/90, 1.11%; 95% CI: 0.20-6.03%), and cattle (2/200, 1.00%; 95% CI: 0.27-3.57%). Similar results were observed for HpTV-1, WzTV, and SGLV, which showed identical rates that were highest among sheep, followed by goats, dogs, cattle, and buffalo, with statistically significant differences. YEZV was found in higher rates in people aged ≤30 (9/121, 7.44%; 95% CI: 3.69-13.53%) than in those aged >30 (0/132, 0.00%; 95% CI: 0.00-2.83%), in men (9/173, 5.20%; 95% CI: 2.76-9.59%) than women (0/180, 0.00%; 95% CI: 0.00-2.09%), and in those working indoors (9/125, 7.20%; 95% CI: 3.83-13.12%) than those working outdoors (0/128, 0.00%; 95% CI: 0.00-2.91%).
Neutralization tests for antibodies specific to the TBVs available in the NVRC (SFTSV, GTV, BHAV, HRTV, TAMV, CCFHV, and HpTV-1) were performed; tests for HIGV, LSV, YEZV, SGLV, TcTV-1, and WzTV were not possible due to the lack of live viruses. The results demonstrated that specific infection occurred in 13 of the 85 LIPS-positive samples, including in eight livestock (8/321, 2.49%), four humans (4/253, 1.58%), and one rat (1/220, 0.45%) (Table 1; Supplementary Table S2). Of those, the livestock (cattle: L143; 1/321, 0.31%) and four people (4/253, 1.58%) exhibited neutralizing activities to CCHFV, as demonstrated by the neutralization antibody titers of 160 in the cattle and 40 in the four individuals. Three livestock (3/321, 0.93%), including two goats (L71 and L195), one sheep (L173), and one rat (R218; 1/220, 0.45%) exhibited neutralizing activities to BHAV, resulting in titers ranging from 20 to 40. One cattle (L29; 1/321, 0.31%) was neutralizing-positive for TAMV with a titer of 20, while one other was for GTV (cattle: L123; 1/321, 0.31%) with a titer of 40. Two livestock showed neutralization to SFTSV (cattle: L246 and L302; 2/321, 0.62%) with titers of 40 (Table S2). These results confirmed the presence of TAMV in Pakistan, supporting earlier findings [5], and revealed livestock reservoirs for SFTSV and GTV beyond known human infections [4,5]. Neutralization to BHAV provided the first evidence of BHAV presence in Pakistan. No cross-neutralization was observed among the 13 samples (Supplementary Table S2).
## 4. Discussion
This study provides serological evidence for the circulation of multiple TBVs among humans, livestock, and rats in the integrated zones of Narowal, Lahore, and Faisalabad in Punjab, where there is a socio-ecological network for the frequent movement of livestock, forage, and animal products that promotes virus dissemination (Figure 1A). Lahore hosts one of the province's largest livestock markets (mandis), where livestock from not only nearby districts such as Narowal and Faisalabad but also from throughout Punjab and other provinces of Pakistan are transported for the purpose of trade, slaughter, or redistri-bution [8]. Narowal, a region distinguished by its significant rice and forage production, supplies these to Lahore and Faisalabad, which function as major storage and distribution centers for the wider supply of Punjab and other provinces [8]. Faisalabad, a prominent industrial hub, receives substantial quantities of livestock, forage, and animal by-products. These are processed within the city to satisfy domestic consumption needs and to facilitate export to other regions [8]. These interactions occur in rural and peri-urban regions where rodents such as rats are frequently found in fields, warehouses, and animal shelters, and livestock are grown close to residential areas [18]. In this region, humans, animals, and rodents coexist in close proximity, thereby generating conditions conducive to the circulation of viruses among these populations [18]. where samples were collected in this study. This zone is composed of Narowal (N), Lahore (L), and Faisalabad (F). Zoomed-in district maps display sampling locations as red dots. Faisalabad processes significant livestock, forage, and by-products for domestic and export markets. Narowal supplies key rice and forage to Lahore and Faisalabad, which function as primary storage/distribution centers. Lahore hosts a major provincial livestock market, acting as a central hub for animals sourced locally and across wider regions. It is a highly integrated socio-economic unit with frequent movement of people, animals, and goods. This intense activity occurs in rural/peri-urban areas where rodents (notably rats) inhabit fields, warehouses, and shelters and livestock are reared near human residences, enabling frequent human-animal contact and rodent coexistence. (B) The circulation of TBVs among humans, livestock, and rats is based on the seroprevalence observed in this study. Viruses that were identified as being positive for neutralizing antibodies were indicated by red characters in their respective hosts. Red arrows indicate the potential circulation of CCHFV between livestock and humans, and BHAV between livestock and rats according to their identified neutralizing antibodies and possible contacts within the eco-social and ecological surroundings. The figure was created using Adobe Photoshop (Adobe Inc., San Jose, CA, USA) for illustrative purposes.
This study investigated the seroprevalence of seven members of orthonairoviruses (TAMV, CCHFV, WzTV, SGLV, HpTV-1, TcTV-1, and YEZV) and six members of bandaviruses (SFTSV, GTV, HRTV, BHAV, LSV, and HIGV) by LIPS among livestock, humans, and rats collected from the integrated zones. The LIPS assay is a highly sensitive liquidphase immunoassay method that has been widely used for high-throughput detection of antibodies against viral proteins and is characterized by its ability to preserve native antigen conformation, producing high sensitivity and low background, which improves sensitivity and specificity, unlike solid-phase assays such as ELISA [22,27]. However, due to its high sensitivity, this method may detect non-specific antibody-antigen reactions, which could be included in the resulting positive rates when multiple antigenically related viruses are tested in parallel. Nevertheless, LIPS could still be considered for the initial screening of a large number of samples from different host species, which could be followed by a different method to verify or confirm the reactions using a different antigen. Therefore, we verified the LIPS-positive samples by Western blot using purified viral particles as antigen. LIPS detects antibodies based on the conformational recognition of viral NPs, while Western blot detects antibodies against linear, denatured epitopes in viral particles. This led to the different sensitivities in the two methods. The number of viral particles used in this study and their differences in antigenicity may also have affected the Western blot analyses confirmation rates. Nevertheless, we identified antibody responses to SFTSV, GTV, BHAV, HRTV, TAMV, CCHFV, and HpTV-1 the LIPS-positive samples and subsequently confirmed the antibody responses to SFTSV, GTV, BHAV, HRTV, TAMV, CCHFV, and HpTV-1. Neutralizing antibodies specific to SFTSV, GTV, BHAV, and TAMV were further identified. Despite this, there could be false positive or negative outcomes during the initial survey of the serum samples; subsequent examination using different methods would help to confirm specific antibody reactions, particularly by the neutralization assays.
The results by LIPS showed YEZV and CCHFV had the highest seroprevalence rates of all the tested TBVs in the three investigated hosts and the highest subtotal rate of all tested viruses among livestock animals (Table 1). These findings suggest a more comprehensive seroprevalence of the tick-borne orthonairoviruses and bandaviruses among livestock, humans, and rats beyond those reported in the previous study [5]. By analyzing the variations in the positive rates of each virus among livestock, humans, and rats, we found that the livestock species may commonly play a significant role in the seroprevalence of SFTSV, HRTV, HIGV, BHAV, TAMV, TcTV-1, HpTV-1, WzTV, and SGLV, even though the highest rates were not recorded in the same livestock species. Of the livestock species, sheep exhibited the highest rates of up to 20.00% for most of the tested TBVs, probably attributable to the limited sample size of five individuals. Further surveys including more samples from sheep would reveal a more precise seroprevalence rate of these viruses. Furthermore, gender could be another critical factor that affected the seroprevalence of GTV, BHAV, CCHFV, and YEZV among humans. Men generally exhibited significantly higher rates than women, probably due to greater engagement in outdoor occupations such as livestock handling, farming, herding, and animal market work, which place them in frequent contact with tick-infested environments. Consequently, higher rates of SFTSV and YEZV were found among people performing outdoor jobs than those indoors. The prevalence rates for all tested TBVs were not significantly associated with either gender or rat habitat.
Together with the results from the identification of neutralizing antibodies, our data echoes the previous findings of CCHFV prevalence in Pakistan [3,5] and suggests a high likelihood of CCHFV circulation among humans and livestock. The results also indicated the presence and dissemination of BHAV, which is capable of infecting the central nervous system in humans [28], livestock, and rodents for the first time in Pakistan. Since BHAV was identified in ticks in India in 1954 [23], it has been necessary to conduct subsequent surveys on the prevalence of BHAV and its association with human disease in Pakistan. Furthermore, the presence of neutralizing antibodies indicated past infection of SFTSV and GTV in livestock, thereby revealing the expanding SFTSV and GTV seroprevalence not only in humans [4,5]. The presence of TAMV in Pakistan was also confirmed by further serologic evidence that supports the previous findings [5]. Taken together, the identification of neutralizing antibodies against CCHFV in humans and livestock; against TAMV, SFTSV, and GTV in livestock (including cattle, goats, and sheep); and against BHAV in livestock and rats suggests that these TBVs were comprehensively prevalent in Pakistan. Given that the samples were from the integrated zones in Punjab, where human, livestock and rats had frequent contact (Figure 1A), the potential circulating links of CCHFV between humans and livestock and BHAV between livestock and rats were proposed. Moreover, livestock species exhibited neutralizing antibodies to more TBVs, probably because they were more likely to be exposed to the parasite by ticks, resulting in an increased incidence of exposure to TBVs in addition to CCHFV (Figure 1B). These results suggest the need for further investigation of the above TBVs in an expanded region a larger sample size from hosts as well as ticks.
A key limitation of this study is the inherent sampling bias due to narrow geographic coverage and a non-random sampling approach. Samples were gathered from only three areas within Punjab's integrated socio-economic zone, which may not completely represent the larger ecological and epidemiological diversity across Pakistan. Human participants were included by convenience sampling due to geopolitical constraints and limited access to rural communities, potentially overrepresenting individuals with higher health-seeking behavior or closer linkages to livestock agricultural activities. Similarly, livestock and rat sampling depended on owner consent, farm accessibility, and the availability of trapping sites, which may have biased detection toward settings with increased human-animal interaction. These restrictions limit the generalizability of the seroprevalence estimates and may either underestimate or overstate the true burden of tick-borne viruses in the region. Future large-scale, active surveillance, including systematic tick sampling and community-based randomized host sampling, will be required to overcome these biases and more precisely determine TBV circulation in Pakistan. Despite these limitations, the circulation of the virus among them was evident as samples were taken from within an ecologic and socio-economic unit. Due to the lack of live viruses, neutralization to HIGV, LSV, YEZV, WzTV, SGLV, and TcTV-1 could not be analyzed. Future research should focus on combining cross-neutralization assays and phylogenetic analysis to strengthen serological specificity, clarify potential cross-reactivity among antigenically related viruses, and improve our understanding of virus-host relationships.
## 5. Conclusions
This study developed LIPS assays for an initial screening of antibodies against seven tick-borne orthonairoviruses and six tick-borne bandaviruses, followed by confirmation using Western blot and microneutralization tests. The results revealed substantial seroprevalence of these viruses in human, livestock, and rodent populations in Punjab, Pakistan, and identified neutralizing antibodies against CCHFV, TAMV, SFTSV, GTV, and BHAV. The findings clearly demonstrate a significant exposure risk to the tick-borne orthonairoviruses and bandaviruses in Pakistan, which highlights the necessity for implementing an integrated "One Health" monitoring strategy that encompasses ticks, livestock, rodents, and high-risk human groups in this region.
## Supplementary Materials:
The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/v17121620/s1, Figure S1. Validation of NP protein expression in HEK293T cells transfected with respective plasmids by (A) Western blot analyses and (B) immunofluorescence assays. NC, negative control. Bars, 150 µm; Figure S2. Antibody responses to thirteen tick-borne viruses among livestock (A), humans (B), and rats (C) by using the luciferase immunoprecipitation system. The levels of antibody response to each specific virus were expressed as fold changes to cut-off values; Figure S3. Western blot assay to examine antibody responses to (A) SFTSV, (B) GTV, (C) BHAV, (D) HRTV, (E) TAMV, (F) CCHFV, and (G) HpTV-1. Sample IDs labeled in red indicate the LIPS-positive samples; samples positive with antibody responses to multiple viruses marked with symbol "#"; and samples that were confirmed with neutralization are marked by an asterisk symbol *; Figure S4. Pairwise sequence identity and phylogenetic analysis of (A) Bandaviruses and (B) Orthonairoviruses. The pairwise percent identities of NP (bottom left) and Gn (top right) amino acid sequences aligned with the respective phylogenetic trees, with genotypes indicated. The sequence of Hantavirus was used as an outgroup for both trees; Table S1. Taxonomic and biological characteristics of the seven Orthonairoviruses and six Bandaviruses investigated in this study; Table S2. Background information and detailed results of the cases that possess positive serological activities against tick-borne viruses in livestock, humans, and rats. Table S3. Univariable analysis of associations between host demographic factors and thirteen tick-borne viruses seropositivity in livestock, humans, and rats.
## References
1. Omar, Albarrak (2025) "The Acaricidal and Repellent Efficacy of Essential Oils and Their Immunomodulatory Effects against Hyalomma Ticks: A Review Article. Pak" *Vet. J*
2. Ahmad, Sindhu, Zafar et al. (2025) "Risk Factors, and Molecular Characterization of Hard Ticks in Two Diverse Agro-Ecological Zones of Punjab, Pakistan. Pak" *Vet. J*
3. Zohaib, Saqib, Athar et al. (2015) "Crimean-Congo Hemorrhagic Fever Virus in Humans and Livestock" *Emerg. Infect. Dis*
4. Zohaib, Zhang, Saqib et al. (2020) "Serologic Evidence of Severe Fever with Thrombocytopenia Syndrome Virus and Related Viruses in Pakistan" *Emerg. Infect. Dis*
5. Chen, Saqib, Khan et al. (2024) "Risk of Infection with Arboviruses in a Healthy Population in Pakistan Based on Seroprevalence" *Virol. Sin*
6. Jamil, Li, Wang et al. (1276) "High Diversity and Low Coinfections of Pathogens in Ticks from Ruminants in Pakistan. Microorganisms"
7. Karim, Budachetri, Mukherjee et al. "A Study of Ticks and Tick-Borne Livestock Pathogens in Pakistan" *PLoS Negl*
8. Shah, Mahmood Malik, Arshad et al. (2016) "Dynamics of Livestock Population Diversity in Punjab Province of Pakistan" *Veterinaria*
9. Khan, Ahmed, Afzal et al. (2022) "Distribution and Identification of Ticks on Livestock in Pakistan" *Int. J. Environ. Res. Public Health*
10. Hussain, Hussain, Yu et al. "Correction: Geographical Epidemiology of Hyalomma Anatolicum and Rhipicephalus Microplus in Pakistan: A Systematic Review" *PLoS ONE*
11. Sajid, Ali, Rahman et al. (2023) "Sero-Epidemiology of Crimean Congo Hemorrhagic Fever Virus (CCHFV) in Human and Livestock Population in District Mardan" *Pakistan. Int. J. Agric. Biol*
12. Shi, Hu, Deng et al. (2018) "Tick-Borne Viruses" *Virol. Sin*
13. Hoogstraal (1979) "Review Article 1: The Epidemiology of Tick-Borne Crimean-Congo Hemorrhagic Fever in Asia, Europe, and Africa" *J. Med. Entomol*
14. Tabassum, Naeem, Khan et al. "Crimean-Congo Hemorrhagic Fever Outbreak in Pakistan, 2022: A Warning Bell amidst Unprecedented Floods and COVID 19 Pandemic" *Health Sci*
15. Yu, Liang, Zhang et al. (2011) "Fever with Thrombocytopenia Associated with a Novel Bunyavirus in China"
16. Kim, Yi, Kim et al. (2012) "Severe Fever with Thrombocytopenia Syndrome, South Korea" *Infect. Dis*
17. Takahashi, Maeda, Suzuki et al. (2014) "The First Identification and Retrospective Study of Severe Fever with Thrombocytopenia Syndrome in Japan" *J. Infect. Dis*
18. Blasdell, Morand, Laurance et al. (2022) "Rats and the City: Implications of Urbanization on Zoonotic Disease Risk in Southeast Asia" *Proc. Natl. Acad. Sci*
19. Zhang, Shen, Shi et al. (2017) "Isolation, Characterization, and Phylogenic Analysis of Three New Severe Fever with Thrombocytopenia Syndrome Bunyavirus Strains Derived from Hubei Province"
20. Guo, Shen, Zhang et al. (2017) "A New Strain of Crimean-Congo Hemorrhagic Fever Virus Isolated from Xinjiang" *Virol. Sin*
21. Moming, Shen, Fang et al. (2021) "Evidence of Human Exposure to Tamdy Virus, Northwest China" *Emerg. Infect. Dis*
22. Shen, Duan, Wang et al. (2018) "A Novel Tick-Borne Phlebovirus, Closely Related to Severe Fever with Thrombocytopenia Syndrome Virus and Heartland Virus, Is a Potential Pathogen" *Emerg. Microbes Infect*
23. Chen, Xu, Wu et al. (2022) "A New Luciferase Immunoprecipitation System Assay Provided Serological Evidence for Missed Diagnosis of Severe Fever with Thrombocytopenia Syndrome" *Virol. Sin*
24. Dai, Guo, Yu et al. (2022) "SECISBP2L-Mediated Selenoprotein Synthesis Is Essential for Autonomous Regulation of Oligodendrocyte Differentiation" *J. Neurosci*
25. Ma, Lv, Zhang et al. (2021) "Identification of a New Orthonairovirus Associated with Human Febrile Illness in China" *Nat. Med*
26. Tamura, Stecher, Kumar (2021) "MEGA11: Molecular Evolutionary Genetics Analysis Version 11" *Mol. Biol. Evol*
27. Burbelo, Lebovitz, Notkins (2015) "Luciferase Immunoprecipitation Systems for Measuring Antibodies in Autoimmune and Infectious Diseases" *Transl. Res*
28. Vilibic-Cavlek, Stevanovic, Krcmar et al. (2155) "Detection of Bhanja Bandavirus in Patients with Neuroinvasive Disease of Unknown Etiology in Croatia" *Microorganisms*
29. "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=PMC12607652&blobtype=pdf | # Evaluation of a sample-to-answer real-time PCR assay for enterovirus detection in cerebrospinal fluid
Grace Perkins, Jennifer Swink, Kevin Wade, Lori Hughes, Bijal Parikh, Neil Anderson
## Abstract
Enteroviruses frequently cause aseptic meningitis, necessitating differentia tion from bacterial meningitis to avoid unnecessary antimicrobial treatment and improve patient management. Polymerase chain reaction (PCR) remains the gold standard for enterovirus detection in cerebrospinal fluid (CSF) due to its high sensitivity and rapid turnaround time. The discontinuation of the Cepheid Xpert EV assay has created a gap in diagnostics. This study aimed to optimize two laboratory-developed real-time PCR assays using Diasorin direct amplification disc (DAD) and universal disc (UD) methods compared to the previously standard Cepheid Xpert EV assay. A total of 87 clinical CSF specimens were tested to assess sensitivity, specificity, and overall performance. There was no significant difference between the performance of the three assays compared to standard of care results. The Xpert EV and DAD methods had a 79.6% positive agreement, while the UD method had an 86.4% positive agreement. All methods achieved 100% negative agreement. Specificity testing against non-enterovi ral pathogens confirmed no cross-reactivity. The estimated limit of detection was 100 copies/mL for Xpert EV, 250 copies/mL for UD, and 1,000 copies/mL for DAD, though clinical sensitivity remained high. Blood contamination affected the DAD assay but not the UD method due to its RNA extraction step. This study demonstrates the potential of DAD and UD methods as viable alternatives for rapid enterovirus detection in CSF, essential for patient management following the discontinuation of the Xpert EV assay. Further validation and comparison are needed for broader clinical applications. IMPORTANCE Rapid and accurate detection of enteroviruses in cerebrospinal fluid is crucial for patient management in cases of aseptic meningitis, especially following the discontinuation of the Cepheid Xpert EV assay. This study evaluates two laboratorydeveloped real-time PCR assays utilizing Diasorin direct amplification disc and univer sal disc methods, demonstrating their potential as viable alternatives for enterovirus detection with high sensitivity and specificity. KEYWORDS virology, pcr, enterovirus E nteroviral infections are common and generally asymptomatic or self-limiting (1). Nevertheless, enterovirus is the leading cause of aseptic meningitis, responsible for 80%-92% of cases globally (2). Distinguishing aseptic meningitis from the more lethal bacterial meningitis based solely on symptoms remains challenging. While treatment for aseptic meningitis is primarily supportive, bacterial meningitis requires targeted antimicrobial therapy and has a mortality rate of up to 50% (3). Thus, a rapid diagnostic test for enterovirus in cerebrospinal fluid (CSF) is critical. A timely positive CSF enteroviral result can prevent needless antibiotic exposure in patients, as well as promote more swift symptomatic management and triage (2).Polymerase chain reaction (PCR) is the gold standard for the detection of enterovirus in CSF specimens due to its advantages of fast turnaround time and high sensitivity
compared to viral culture (4,5). Currently, there are few Food and Drug Administra tion (FDA)-cleared enterovirus diagnostic PCRs for CSF specimens. The BioFire FilmAr ray Meningitis/Encephalitis panel and QIAstat-DX Meningitis/Encephalitis panels are multiplex PCRs and array diagnostic methods that assess for common bacterial, viral, and fungal pathogens of the central nervous system and yield results in approximately 1 hour. The FilmArray Meningitis/Encephalitis panel for enterovirus detection has a reported sensitivity of 89% and specificity of 100% (6). The QIAstat-DX ME panel has variable sensitivity based on the enteroviral strain but a reported sensitivity of contrived specimens of 88%-99% (7,8). The Cepheid Xpert EV assay is the only FDA-cleared assay targeting solely enterovirus. The assay is a real-time RT-PCR assay that results in approximately 2.5 hours with a reported sensitivity of 96% and specificity of 99% (9). Unfortunately, this assay has been discontinued by the manufacturer as of October 2023. Given the limited diagnostic options for enteroviral meningitis, we describe the evaluation of a laboratory-developed sample-to-answer real-time PCR assay utilizing commercially available analytespecific reagents for the detection of enterovirus from CSF specimens.
## MATERIALS AND METHODS
## Specimens
This study was approved by the Washington University Institutional Review Board (ID#202312029). A total of 87 previously analyzed CSF specimens submitted for routine clinical testing at Barnes Jewish Hospital Molecular Infectious Disease Laboratory and Saint Louis Children's Hospital Microbiology Laboratory were utilized in this study. Specimens ranged in collection dates from September 2004 to July 2023 and were stored at -80°C after initial clinical analysis. Inclusion criteria required at least 1 mL of CSF specimen to allow all necessary testing to be performed. A total of 44 of the specimens were previously enterovirus-positive by standard-of-care molecular diagnostic methods, with the other 43 specimens being previously enterovirus negative by standard-of-care molecular diagnostic methods. From 2004 to 2008, the standard-of-care molecular diagnostic method at Barnes Jewish Hospital and Saint Louis Children's Hospital was a laboratory-developed test utilizing a Qiagen One-Step RT-PCR kit. The method utilized from 2008 to 2024 was the Cepheid Xpert EV RT-PCR assay.
## DAD protocol optimization
Four PCR protocols were tested to determine the best performance using the Diasorin 8-well direct amplification disc (DAD) (Diasorin, Cypress, California). The DAD is an 8-well, sample-to-answer PCR amplification disc that does not require RNA extraction prior to use. The four protocols were trialed with and without RNasin Ribonuclease Inhibitor (Catalog No: N2615; Promega), as well as with or without a filtration step to achieve maximum performance. Analytespecific reagents for this assay were obtained from Diasorin. The Diasorin Enterovirus primer pair (MOL9020) is comprised of a Scorpion primer/probe that targets a conserved 5′ non-coding region. The primer pair targets Enterovirus serotypes 68, 70, and 71, as well as other Coxsackieviruses and Echoviruses. All reactions were prepared immediately before amplification as master mixes containing all reaction components: (i) 1 µL of Enterovirus primer pair, (ii) 20 µL TA master mix (MOL9070), (iii) 1 µL Reverse Transcriptase (MOL9018), (iv) 1 µL Simplexa Extraction and Amplification Control (SEAC) RNA primer pair (MOL9200), (v) 3 µL SEAC RNA template (MOL9200), and (vi) 23 µL water. Fifty microliters of sample or control was added to the "SAMPLE" well, and 50 µL of reaction mix was added to the "R" well.
Dilutions of control material were made by diluting commercially available inactiva ted Coxsackievirus A9 (Exact Diagnostics) in CSF matrix at concentrations of 100,000, 10,000, 5,000, 1,000, 500, 250, 125, and 62.5 copies/mL, based on the value provided by the manufacturer. Fifty microliters of diluted control material was used per reaction.
RNasin was added to the PCR master mix at 1 µL per reaction. An Amicon Ultra 0.5 mL filter tube (Catalog No: UFC501096) was utilized, with filtration RPM at 14,000 g for 2 minutes, as per the package insert. PCR amplification was performed using the Diasorin LIAISON MDX instrument. All reactions were prepared immediately prior to amplifica tion. Target and internal control fluorescence thresholds were 2,000 mean fluorescence intensity units. Data collection and analysis were performed with the LIAISON MDX Studio software. The protocol definitions and results can be found in the supplemental materials as Table S1.
Protocols were evaluated for sensitivity by comparing cycle thresholds for dilutions of control material. The dilutions evaluated and the results can be seen in the supplemental materials as Table S2. Protocol 2 with RNasin was chosen given its balance of sensitivity and favorable cycling parameters (Ct values less than 40). This protocol will be referred to from here as the "DAD method" and will be compared to the 96-well format Diasorin universal disc (UD) as well as the Cepheid Xpert EV performance.
## Diasorin UD protocol
The Diasorin UD protocol utilizes the consumable 96-well UD (Diasorin) for PCR amplification, which requires an additional RNA extraction step. Before extraction, 400 µL of specimen or control was spiked with 15 µL of SEAC RNA template (MOL9200). The SEAC RNA acted as an internal control for the extraction and amplification of the reaction. RNA extraction was performed on the bioMerieux NucliSENS easyMAG (bioMerieux, Maryc l'Etoile, France). Five microliters of extracted CSF samples or controls was used. The remaining reaction components totaled 5 µL for a final reaction volume of 10 µL. All reactions were prepared immediately prior to amplification as master mixes containing all reaction components: (i) 0.2 µL Enterovirus primer pair (MOL9020), (ii) 4 µL TA master mix (MOL9070), (iii) 0.5 µL Reverse Transcriptase (MOL9018), (iv) 0.2 µL SEAC RNA primer pair (MOL9200), and (v) 0.1 µL water PCR amplification was performed using the Diasorin LIAISON MDX instrument, and data collection and analysis were performed with the LIAISON MDX Studio software.
## Cepheid Xpert EV protocol
The Cepheid Xpert EV assay is a sample-to-answer reverse-transcription PCR (RT-PCR) assay for the detection of enteroviral RNA in CSF specimens. The protocol was followed as per manufacturer instructions on the GeneXpert Dx system.
## Comparative accuracy assessment
The DAD method, the UD method, and the Xpert EV assays were performed on 87 previously tested clinical CSF specimens. The performance of each tested method was compared to the clinical standard-of-care reported result. Positive percent agree ment was calculated as the proportion of positive standard of care results that also tested positive by each of the three tested methods. Negative percent agreement was calculated as the proportion of negative standard of care results that also tested negative by each of the three methods. Overall percent agreement for each method was also calculated. Confidence intervals were calculated via bootstrapping. Cohen's Kappa was calculated and interpreted using Landis and Koch's classification of agree ment. Additionally, the result agreement between each of the three tested assays was determined to help compare performance.
## Analytical specificity assessment
The analytical specificity of both DAD and UD assays was assessed by utilizing resid ual clinical CSF samples that had previously tested positive for non-enteroviral patho gens. Additionally, control samples that contained multiple pathogens were tested. The pathogens tested were herpes simplex virus 1 (HSV-1), HSV-2, Human herpesvirus 6, Varicella zoster virus (VZV), adenovirus, cytomegalovirus, Mycoplasma pneumoniae, Parvovirus B19, and Parechovirus.
## Comparative analytical sensitivity assessment/limit of detection
Serial dilutions of control material (inactivated Coxsackievirus A9, Exact Diagnostics) at 5,000, 1,000, 500, 250, and 100 copies/mL were run in triplicate using the Xpert EV, DAD, and UD assays. The performance of the DAD assay for enterovirus D68, which the Xpert EV assay detects less efficiently due to a higher limit of detection for this serotype, was compared to that of Xpert EV (7). Dilutions of Fermon enterovirus D68 RNA at 5,000, 1,000, 500, 250, 100, 50, and 25 TCID50/mL using previously enterovirus-negative patient CSF were run in triplicate on both Xpert EV and DAD assays. The estimated limit of detection for each assay was determined as the concentration at which three of three (100%) of replicates were detected.
## Inhibition assessment
To ensure the assay performance with the most likely inhibitory substance encountered in CSF, we performed both DAD and UD assays with a dilution of CSF and EDTA blood. CSF specimens that tested previously positive for enterovirus were pooled to make a single pooled specimen. The pooled specimen was run "neat" in addition to dilutions A, B, and C. The total volume of each dilution was 500 µL. Dilution A was made with 25 µL EDTA blood in 475 µL of the pooled specimen (dilution factor 1:20). Dilution B was made with 50 µL EDTA blood in 450 µL pooled specimen (dilution factor 1:10). Dilution C was made with 100 µL EDTA blood in a 400 µL pooled specimen (dilution factor 1:5).
## RESULTS
## Demographic results
Demographic information for patients from whom clinical specimens were utilized is listed in Table 1. Due to the age of specimens, some records predate the currently used electronic medical record, and thus, certain elements were unable to be recovered (listed as "unknown"). Patient ages were stratified into three groups: 0 to <2 years, 2 to <18 years, and >18 years. The Freeman-Halton extension of the Fisher exact test was utilized to analyze this data. There was a statistically significant difference (P < 0.001) between age groups, suggesting that the positive cohort was more likely to be composed of very young children (<2 years of age). We used the Fisher's exact test to evaluate the differences in sex composition between the two groups, with statistical significance set at P < 0.05. There was not a significant difference between the positive and negative cohorts' sex composition. For the remainder of the data points, we utilized the Wilcoxon rank sum test to compare the differences between the positive and negative cohorts. The significant differences between the positive and negative cohorts are seen in the nucleated cell count and whether additional viral testing of CSF was performed. The positive cohort had higher nucleated cell counts in CSF than the negative cohort. The negative cohort was more likely to have concurrent viral testing for HSV, VZV, and Parechovirus in CSF. These findings are to be expected, as a negative enterovirus test for a viral meningitis clinical picture would prompt clinicians to perform other viral testing of CSF.
## DAD protocol optimization
The results of protocol optimization are depicted in Table S2. Protocol two with RNasin was selected as the most sensitive protocol with Ct values below 40.
## Comparative accuracy
The accuracy results for all three tested methods are shown in Table 2. Utilizing the previously obtained standard of care testing as the gold standard, there was no significant difference between the performance of the three assays. The Xpert EV and DAD assays both identified 35 of 44 previous positive specimens, with a positive percent a Key: PPA (positive percent agreement), NPA (negative percent agreement), OPA (overall percent agreement). b DAD, direct amplificiation disc. c UD, universal disc. agreement of 79.6% (95% confidence interval [CI] 67.5%-90.7%) and Cohen's Kappa of 0.794. The UD 96-well assay correctly identified 38 of 44 positive specimens, with a positive percent agreement of 86.4% (95% CI 75.6%-95.5%) and Cohen's Kappa of 0.862. All three assays identified 43 of 43 previously negative specimens as negative, equating to a negative percent agreement of 100% for all three assays. The overall percent agreement for both the Xpert EV and the DAD assays was 89.7% (95% CI 83.9%-95.4%). The overall percent agreement for the UD assay was slightly higher at 93.1% (95% CI 87.4%-97.7%). Invalid results were not observed in any of the three tested assays. The positive result agreement between each of the three tested assays is shown in Fig. 1. Twenty-nine of 44 (66.0%) positive specimens tested positive for enterovirus by all three methods. Thirteen of 44 (30.0%) positive specimens tested positive by one or multiple methods, but not by all three methods. Two of 44 (5.0%) previously positive specimens tested negative by all three methods. Utilizing a two out of three composite reference standard, the DAD showed 91.9% sensitivity and 98% specificity, the UD had 94.6% sensitivity and 94% specificity, and the Xpert EV demonstrated 91.9% sensitivity and 98% specificity. Discordant results show Ct values ranging from 32.2 to 37.9 (Xpert EV), 39.4-41.1 (DAD), and 37.1-38.9 (UD). These findings suggest that the discordance may be due to low viral load in these specimens, given higher Ct values.
## Comparative analytical specificity
None of the tested pathogens elicited a positive result on the DAD and UD assays, which suggests high specificity for enteroviral detection among other common CSF pathogens.
## Comparative analytical sensitivity assessment/ limit of detection
The results of the estimated limit of detection studies are depicted in Table 3. Of the three methods tested, the Xpert EV had the lowest estimated limit of detection at 100 copies/mL. The UD method had the second-lowest estimated limit of detection, estimated to be 250 copies/mL. The DAD method had an estimated limit of detection of 1,000 copies/mL. Although this suggests a difference in analytical sensitivity for the tested control material, the difference in clinical sensitivity is likely minimal, given the results of patient sample testing. The positive percent agreements between the standard of care Xpert EV method and the proposed DAD and UD methods were very similar, with the UD method having the greatest positive percent agreement at 86.4%. The negative percent agreement for all three methods was identical. The results of the D68 comparative analysis are depicted in Table 4. The DAD assay had a lower estimated limit of detection of 25 TCID50/mL, while the Xpert EV assay had an estimated limit of detection of 100 TCID50/mL. These results suggest that the DAD assay is more sensitive than the Xpert EV assay for the detection of this serotype.
## Inhibition assessment
Inhibition assay results are listed in Table 5. For the DAD method, qualitative inhibition was noted at a 1:5 dilution of blood in CSF. Since this method uses a direct sample without an extraction method, these results can be expected. For the UD method, qualitative inhibition was not noted for all tested levels of blood. This is likely due to the additional extraction step that precedes PCR.
## DISCUSSION
While the Cepheid Xpert EV has been the standard of care for dedicated Enterovirus testing, its discontinuation creates a clinical need for rapid enterovirus testing. Herein, we describe two alternative methodologies with similar clinical performance. When weighing these alternatives, it is important to consider the testing time and ease of use. The standard of care Xpert EV assay requires minimal hands-on time, while the proposed DAD and UD methods utilize more technical input. The RNA extraction step necessary for the UD method requires a separate instrument and a more complicated workflow, which can add at least 1 hour of additional testing time.
In comparison to published studies, the proposed DAD and UD methods show similar clinical performance to enterovirus detection methods for CSF. The Xpert EV has a reported sensitivity of 96%-100% for enterovirus detection in CSF (9,10). The FilmAr ray ME and QIAstat-DX ME panels have reported sensitivity of 89% and 88%-99% for enterovirus detection in CSF, respectively (6)(7)(8). Herein, we found direct positive percent agreement between the Xpert EV and the DAD and UD methods to be 79.6% and 86.4%, respectively. Based on this agreement and composite reference analysis, the clinical accuracy of the DAD method may be acceptable for clinical use in the microbiology laboratory.
A key strength of our study is the identification of two potential novel methods for the rapid diagnosis of enteroviral meningitis. The DAD method is a sample-to-answer assay, like the previously utilized standard of care Xpert EV method, which is an attractive option for large, busy microbiology laboratories. The UD method performs similarly, though it may have an increased analytical sensitivity. One must evaluate whether the extended RNA extraction time necessitated by the UD method justifies its application within their laboratory and for their patient population. Further clinical comparisons between these two methodologies are likely needed in future studies. CSF specimens can be contaminated with blood through the process of lumbar puncture. We show that blood contamination at a dilution of at least 1:5 can cause a lack of detection of enterovirus in CSF specimens using the DAD assay. The UD method did not show any decrease in sensitivity with possible blood contamination at the dilutions tested. This is likely due to the offline extraction step used for the UD method increasing sensitivity. Other studies utilizing the DAD have had success with diluting patient samples 1:1 in phosphate buffered saline to offset contamination, although this must be validated (11,12). Overall, the results of this study and others indicate that for specimens with high blood inhibition, dilution can be validated, or alternatively, the UD method may be a more suitable choice for analyzing bloody CSF specimens.
One of the limitations of our study was the relative scarcity of acceptable positive specimens. For this study, 1 mL of remnant CSF was required; however, CSF is typically obtained in limited quantities. Consequently, some of the specimens tested were as old as 19 years past original collection, potentially introducing a selection bias, as the positive cohort comprised older specimens than the negative cohort. Additionally, extended frozen storage of RNA virus particles may have resulted in degradation, which could account for some of the discrepant results. Furthermore, the age of the specimens contributed to certain data collection challenges, rendering the demographic statistics incomplete.
Another potential limitation of our study is the absence of serotyping for the samples. Given the reported higher limit of detection for enterovirus D68 compared to other enteroviruses, the Xpert EV method may have failed to detect low viral loads of the D68 strain in the positive cohort. Without serotyping the positive cohort, definitive conclusions about the performance of the three assays for specific enteroviral strains in clinical samples cannot be drawn. We investigated the performance of the DAD assay against the Xpert EV assay for the detection of enterovirus D68 in non-clinical specimens, and the DAD proved to be analytically sensitive for the detection of this specific serotype.
In regard to analytical specificity, we have shown that the DAD and UD assays have high specificity for enterovirus detection among other common CSF pathogens. However, specificity was not assessed for rhinoviral infections, which represents a limitation of this study given known cross-reactivity between enterovirus and rhinovirus via nucleic acid testing methodologies. Although rhinoviral infections are exceedingly rare causes of meningoencephalitis, we acknowledge that the presence of rhinoviral nucleic acid has been shown to elicit cross-reactivity in nucleic acid amplification testing due to their degree of homology (13,14). The Diasorin enteroviral primer pair is not known to have binding for rhinovirus; however, it targets the 5′ non-coding region of enteroviruses, so further validation for rhinovirus positive clinical samples is needed. A rapid diagnostic test for diagnosing enteroviral meningitis is crucial to help guide treatment decisions and length of stay. With a relative lack of commercially availa ble assays, we describe two potential alternatives for rapid diagnosis of enteroviral infections of CSF. Both the DAD and UD methods utilizing Diasorin analytespecific reagents demonstrate similar clinical sensitivity compared to the previously available FDA-approved Xpert EV assay. The sample-to-answer DAD assay meets our laboratory's validation criteria for acceptability and has been implemented for clinical use.
## References
1. Zaoutis, Klein (1998) "Enterovirus infections" *Pediatr Rev*
2. Ramers, Billman, Hartin et al. (2000) "Impact of a diagnostic cerebrospinal fluid enterovirus polymerase chain reaction test on patient management" *JAMA*
3. Akaishi, Tarasawa, Fushimi et al. (2024) "Demographic profiles and risk factors for mortality in acute meningitis: a nationwide population-based observational study" *Acute Med Surg*
4. Gorgievski-Hrisoho, Schumacher, Vilimonovic et al. (1998) "Detection by PCR of enteroviruses in cerebrospinal fluid during a summer outbreak of aseptic meningitis in Switzerland" *J Clin Microbiol*
5. (2025) *Full-Length Text Journal of Clinical Microbiology*
6. He, Kaplan, Kamboj et al. (2016) "Laboratory diagnosis of central nervous system infection" *Curr Infect Dis Rep*
7. Schnuriger, Vimont, Godmer et al. (2022) "Differential performance of the FilmArray Meningitis/Encephalitis assay to detect bacterial and viral pathogens in both pediatric and adult populations" *Microbiol Spectr*
8. Benschop, Staines, Mckloud et al. (2024) "Performance evaluation of molecular detection of enteroviruses: results of 18 years of quality control for molecular diagnostics (QCMD) external quality assessment program, 2005-2022" *J Med Virol*
9. Sundelin, Bialas, De Diego et al. (2023) "Evaluation of the QIAstat-Dx Meningitis/Encephalitis Panel, a multiplex PCR platform for the detection of community-acquired meningoencephalitis" *J Clin Microbiol*
10. Lin, Li, Lan et al. (2022) "Evaluation of GeneXpert EV assay for the rapid diagnosis of enteroviral meningitis: a systematic review and metaanalysis" *Ann Clin Microbiol Antimicrob*
11. Marlowe, Novak, Dunn et al. (2008) "Performance of the GeneXpert enterovirus assay for detection of enteroviral RNA in cerebrospinal fluid" *J Clin Virol*
12. Zhen, Berry (2021) "Herpes simplex virus-1 and -2 rapid detection in whole blood" *Mol Diagn Ther*
13. Colasante, Makari, Hummel et al. (2024) "Sample-to-answer direct real-time PCR detection of Anaplasma phagocytophilum, Ehrlichia spp., and Babesia spp. infections in whole-blood specimens" *Microbiol Spectr*
14. De Almeida, Zerbinati, Tateno et al. (2010) "Rhinovirus C and respiratory exacerbations in children with cystic fibrosis" *Emerg Infect Dis*
15. Jaramillo-Gutierrez, Benschop, Claas et al. (2010) "multi-centre study in the Netherlands examining laboratory ability to detect enterovirus 68, an emerging respiratory pathogen" *J Virol Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12786400&blobtype=pdf | # Special Issue "Influenza Viruses: Infection and Genomics"
Daniele Focosi
Of the four known influenza virus types affecting humans (A, B, C, D), influenza virus A (IAV) has threatened human health globally for centuries. Decades of genomic studies have increasingly refined the taxonomy of IAV, which now includes subtypes, clades, and genotypes. In this Special Issue of the Journal, titled "Influenza Viruses: Infection and Genomics", several intriguing primary research articles are introduced.
Savin et al., at the Siberian Branch of the Russian Academy of Sciences, used integrative bioinformatics analysis to identify the core genes (Birc5, Cdca3, Plk1, Tpx2, Prc1, Rrm2, Nusap1, Spag5, Top2a, and Mcm5) and transcription factors (E2F1, E2F4, NF-YA, NF-YB, and NF-YC) involved in persistent lung injury and regeneration processes in IAV-infected murine lung tissue. Further in vivo verification of the core signature genes confirmed their involvement not only in IAV infection but also in COVID-19 and lung neoplasm development, suggesting their potential role in abnormal epithelial proliferation and oncotransformation [Contribution 1].
Yin et al., at the Nanjing Agricultural University, reported an increased expression of Linc01615, a long intergenic non-coding RNA (LincRNA), in A549 cells upon IAV PR8 infection. The authors proved that DHX9, a protein interaction predicted by the catRAPID website, binds with Linc01615 to partake in IAV replication and that Linc01615 helps to activate the intracellular immune system, which was confirmed by knockdown experiments and by cross-linking immunoprecipitation and high-throughput sequencing (CLIP-seq) experiments [Contribution 2].
Hao et al., at the University of Texas Health Science Center, reported that upon H 1 N 1 IAV infection, Runx3 promoted enlargement of mediastinal lymph nodes (mLNs) and enhanced CD8 + and CD4 + T cell expansion in lung-draining mLNs but not in lungs. Runx3 is required for CD43 core 2 O-glycosylation on activated CD8 + T lymphocytes, and the involved Runx3 signal pathway may mediate the CD8 + T lymphocyte phenotype for generation of pulmonary CTLs [Contribution 3].
Savenkova et al., at the Russian State Research Center of Virology and Biotechnology "Vector", reported that in mice, the knockout of the tumor necrosis factor alpha (TNF-α), a well-known proinflammatory cytokine, leads to increased viral loads compared to the parental strain but similar amounts of live virus and lower interalveolar septal infiltration [ Contribution 4].
IAV(H 5 N 1 ) clade 2.3.4.4b is becoming enzootic in mammals and reinforces its position as a pandemic candidate. This Special Issue finally includes a review on 2.3.4.4b, in particular genotype B3.13, which has recently caused an outbreak in US dairy cattle. Since pandemic preparedness is largely based on the availability of prepandemic candidate vaccine viruses (CVVs), we have reviewed challenges for H 5 vaccine manufacturing and delivery [ Contribution 4]. In this regard, mRNA vaccines represent the most promising pipeline in terms of turnaround times and combination.
Much remains to be discovered about IAVs. The exact subtype and clade that will cause the next pandemic remain difficult to predict, with a long list of candidate subtypes circulating at low levels across birds and mammals and occasionally causing outbreaks in humans exposed for professional reasons. Much research is still needed to precisely identify host restriction factors. While we have gained some insights into serological crossreactivities and hence into candidate vaccine virus (CVV) efficacy, under pre-pandemic scenarios, the ability to maintain a robust veterinary genomic surveillance program will be fundamental to facilitate drug manufacturers, including those dealing with passive immunotherapies. Similarly, the ability to manufacture combined vaccinations could boost the otherwise declining vaccine compliance rates. Moreover, concerns persist regarding the real-world clinical efficacy of small-molecule antivirals and the risk of mutational escape during massive deployments. As such, preparedness should include frameworks for immediately starting well-designed randomized clinical trials.
## References
1. Savin, Sen'kova, Goncharova et al. "Novel Core Gene Signature Associated with Inflammation-to-Metaplasia Transition in Influenza A Virus-Infected Lungs" *Int. J. Mol. Sci. 2024*
2. Yin, Hu, Huang et al. "The Identification and Function of Linc01615 on Influenza Virus Infection and Antiviral Response" *Int. J. Mol. Sci. 2024*
3. Hao, Kundu, Shetty et al. (2024) "Runx3 Regulates CD8+ T Cell Local Expansion and CD43 Glycosylation in Mice by H1N1 Influenza A Virus Infection" *Int. J. Mol. Sci*
4. Savenkova, Gudymo, Korablev et al. (2024) "Knockout of the Tnfa Gene Decreases Influenza Virus-Induced Histological Reactions in Laboratory Mice" *Int. J. Mol. Sci*
5. "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=PMC12382465&blobtype=pdf | # The first discovery of the group IV GETV from Pangolin, China
Pengpeng Xiao, Yujia Hao, Min Zheng, Duo Zhang, Nan Li
## Abstract
We confirmed novel Group IV Getah virus (GETV) was detected from pangolin, which is the first discovered new host. And this is the first discovery of Group IV GETV in Guangxi province, China. This strain, named CHIKV-China/GX2020, is highly homologous to the 2017 Thailand strain. Bayesian analysis shows that GETV has a faster evolutionary rate with 3.487 (2.213-4.771) × 10 -4 substitutions/site/year. The Lollipop plot demonstrated in detail the amino acid differences between the isolated strain and the other sequences of GETV of Group IV and revealed that more mutations occur in the structural protein E2. Compared with the 2017 Thailand isolate, GETV-China/GX2020 has been discovered important amino acid mutation of E2-H207N. This mutation results in a change in the conformation of the E2 protein, which may in turn affect the infectivity or pathogenicity of the virus. Existing surveillance programs and diagnostics efficiency of potential outbreaks of GETV need strengthen.
## Introduction
Getah virus (GETV) is an emerging mosquito-borne alphavirus of Togaviridae, which, as an emerging virus, is a continuing threat to public health [1]. GETV was first identified in Malaysia in 1955 with subsequent sporadic outbreaks. Since 1955, the geographic distribution of GETV has ranged from tropical (Malaysia) to cold regions (Russia). To date, GETV has been isolated in 13 countries, including Vietnam, Thailand, Cambodia, Malaysia, the Philippines, India, Sri Lanka, Australia, Japan, Korea, Mongolia, China, and Russia. GETV has gradually spread from tropical to temperate regions and has even been found in the frigid Arctic tundra [2]. GETV was first isolated in China in 1964 and is now found in 65% (22/34) of China's provinces [2].
GETV is known to be pathogenic to horses, pigs, and cattle, resulting in huge economic losses. GETV infection in horses was first reported in Japan in 1962, and since then, seven large-scale equine GETV outbreaks have been reported worldwide [3]. Clinical manifestations of GETV infection in horses usually include fever, generalized rash, hindlimb edema, and enlarged lymph nodes. It is important to note that pigs are more severely infected than horses. Infected piglets exhibit depression, tremors, diarrhea, hind limb paralysis, and high mortality rates, and infected sows abort [4]. GETV infection in pigs was first reported in Japan in 1965 and rapidly spread throughout Asia. GETV was not proven to be able to infect cattle until 2019, when it was isolated from feverishly ill cattle in China [5]. In addition, GETV can infect wildlife such as foxes, causing fever, anorexia, depression, neurological symptoms and even death, and wild boars [6].
Additionally, GETV is constantly expanding its host range, including rabbits, monkeys, orangutans, mice, hamsters, guinea pigs, chickens, ducks, goats, kangaroos [7], and can even be detected in human serum [8]. In our study, Group IV GETV was first discovered in Guangxi province, China. Novel GETV was isolated from pangolin, which is the first discovered new host. It poses new challenges for the prevention and control of GETV, which can draw attention to the research in this field. Given the further expansion of GETV epidemiology and host species, we used whole genome sequences to explore the epidemiology and evolution of newly identified GETV.
## Materials and methods
## Swab sample collection
In July 2020, to investigate the carriage of GETV by pangolins, we randomly selected 34 healthy pangolins from the Pangolin Rescue and Breeding Base of Guangxi Academy of Forestry Sciences in Nanning, China, for swab sample collection (Figure 1A), with one anal or throat swab sample per pangolin. The species of pangaolins is Manis Javanica (M. Javanica). Twelve pangolins underwent anal swab sample collection and 24 pangolins underwent throat swab collection. After collected, swab samples were stored at -80°C for later use. The samples were divided into two groups based on the swab types. The pangolin samples counting and grouping were shown in Table S1 + 2. Our study was approved by the Research Ethics Committee of Wenzhou University. The study has adhered to the ARRIVE guidelines.
## GETV detection and sequencing
Viral RNA was first extracted from swab samples. Subsequently, GETV was specifically detected by realtime reverse transcription PCR (RT-PCR). Primers used for RT-PCR to detect GETV are shown in Table S3. Viral isolation was performed in BHK-21 and Vero cells. Virus isolation identified by CPE (Cytopathic Effect). Cells were cultured in medium (DMEM, HyClone, Logan, UT, USA) containing 10% fetal bovine serum. Virus blind passaging was performed using positive samples. Afterward, virus blind passaging cell cultures stored at -80ºC were shipped to Sangon (Sangon Biotech, Shanghai, China) for subsequent sequencing. Complete genome sequencing was accomplished by the Sanger method, which has been described in previous studies [9,10]. Sequence analysis and full genome assembly were completed using SeqMan software (DNASTAR, Madison, WI). Multiple comparisons were performed using MAFFT7, and maximum likelihood trees were constructed using MEGA7.
## Phylogenetic and Bayesian analysis
GETV complete genome sequences were downloaded from the NCBI database. We aligned sequences using MAFFT (https://mafft.cbrc.jp/alignment/software) to obtain the sequences from E2 of GETV and inferred a preliminary maximum-likelihood tree using IQ-TREE (http://www.iqtree.org). Temporal signal evaluation of sampling using TempEst v 1.5.3 (http://tree. bio.ed.ac. uk/software/tempest) regressed root-tip genetic distances on the ML tree against the exact sampling date to test whether the sampling date could be Temporal molecular clock correction for Bayesian, and the results of the analysis were visualized using the R language. The Bayesian software package BEAST version 1.10.4 (https://beast.com munity) was used to estimate a time-calibrated phylogeny employing the TN93+F+G4 nucleotide substitution model, the Uncorrelated Relaxed Clock model and the Bayesian Skyline coalescent tree prior, and the Bayesian Markov Chain Monte Carlo (MCMC) chain length was set to 60 million generations for each set, sampled every 6,000 generations. Tracer version 1.7.1 (https://beast.com munity/tracer) was used to check the Convergence (Effective sample sizes >200) of MCMC chains. We summarized maximum clade credibility (MCC) trees using TreeAnnotator (https://beast.community/treeannotator) after discarding 10% as burn-in, and the evolutionary tree was viewed using FigTree.
## Amino acid mutation analysis
Amino acid mutation analysis of each protein of GETV of Group IV was performed using BioAider software using the isolated strain in this study as the reference sequence, and the Lollipop plot was used to show the mutation results. The tertiary structures of the E2 proteins of the isolated strains in this study and the 2017 Thailand isolate LC534253.1 which has the closest relative to it were homology modeled using the i-TASSER server (https://zhanglab.ccmb.med.umich.edu/I-TASSER/), and the one with the highest C-score was selected as the best model. The structures were visualized and analyzed in ChimeraX.
## Results
## Analysis of GETV detection and sequencing
Two GETV-positive samples from each group were detected by real-time RT-PCR, suggesting that a total of four pangolins were GETV positive, and the Ct values of the positive samples ranged from 16.5 to 28.3. Virus blind passaging was performed for five generations, and the CPE results were all negative. GETV blind passaging cell culture (in Vero cells) from No. 19 Pangolin throat swab was completely sequenced (Table S2). A genome-wide linear sequence of 11,288 nt GETV, named GETV-China/GX2020, was obtained from a throat swab and submitted to GenBank (accession number OR373097).
## Phylogenetic analysis
The phylogenetic tree was constructed by comparing the whole genome sequence of GETV-China/GX2020 with that of other GETV isolates in GenBank. Initially, the genome of GETV-China/GX2020 was compared with 38 other isolates representing all GETV genotypes (Figure 1B). In addition, the GETV-China/GX2020 genome was compared with three sequences of group IV GETV (Figure 1C). The maximum likelihood tree proved that GETV-China/GX2020 belonged to group IV GETV and became the first discovered group IV GETV in Guangxi province, China. Our team discovered group III GETV in a previous study of the pangolin virome [11], and followed up with a comprehensive screening of this batch of pangolin swab samples for GETV carriage, and for the first time, the presence of group IV GETV was detected in pangolins. The comparison of GETV-China /GX2020 with the group IV GETV studied strains revealed a nucleotide divergence from 2.7% (Thailand 2017 strain) to 5.6% (Russia 2020 strain). Study on polyprotein showed that amino acid differences ranged from 0.8% (Thailand 2017 strain) to 3.1% (Russia 2020 strain). Remarkably, GETV-China/GX2020 presented 14 unique amino acids compared to the other group IV GETV (Figure 1D), of which 5 were conservative substitutions among hydrophobic (n = 3) and polar (n = 2) residues, while the remaining 9 were non-conservative substitutions, 5 located in the structural proteins (Figure 1D).
## Bayesian analysis
The temporal signal detection results showed a strong enough temporal signal (correlation coefficient = 0.4593, R 2 = 0.2109) to estimate the temporal calibration phylogeny using the molecular clock model. The MCC tree has revealed the detection of all groups and subtypes of GETV, and from this analysis, we estimated the time of the most recent common ancestor (tMRCA) that occurred in 1867.9 (1746. 4-1935.4). The estimated tMRCA was 1978. 3 (1953.9-1998.5) for Group IV. The mean evolutionary rate was estimated at 3.487 (2.213-4.771) × 10 -foot_1 substitutions/site/year (Figure 2).
## Amino acid mutation analysis
The Lollipop plot demonstrated in detail the amino acid differences between the isolated strain and the other sequences of GETV of Group IV and revealed that more mutations occur in the structural protein E2 (Figure 3). E2 is the main glycoprotein by which the virus enters the host cell and binds to the cellular receptor, and the hydrophobicity of the E2 protein of the isolated strain was changed by a mutation at amino acid position 207 from N to H as compared to the other sequences (Figure 4).
## Discussion
GETV is currently divided into four main groups: Group I~IV. Yuan-Yuan L et al. [12] analyzed the genetic evolution of the E2 gene of GETVs isolated from 1955 to 2014 and concluded that the Most Recent Common Ancestor (MRCA) of GETVs appeared about 145 y ago, after which they sequentially evolved into four distinctly different evolutionary populations. The first strain of GETV was first isolated from Culex collected from Hainan Island in China in 1964. Since then, a total of 70 strains of GETV have been isolated from a variety of mosquitoes, midges, pigs and foxes, with the major group III GETV [8,[13][14][15][16][17][18].
In our study, the GETV-China/GX2020 does not show close links with other strains in China, which strongly supports the independent introduction of GETV into border province in China. Moreover, it is the first discovery of group IV GETV in Guangxi province, China. Indeed, the GETV-China/GX2020 is distinct from other GETVs circulating in China, but has the highest homology with the Thailand 2017 strain isolated from Sus scrofa. Due to the influence of monsoon circulation, the summer monsoon brings abundant precipitation to the subtropical regions of Thailand and China during the hot season, forming a warm and humid climate, which facilitates the transmission of the virus from Thailand insect vectors biting group IV GETV-positive Sus scrofa to the coastal areas of Guangxi, China. This provides favorable conditions for the breeding and transmission of the virus by insect vectors. Due to the limited number of GETV nucleotide sequences, the specific pathway of GETV transmission between East and Southeast Asia needs further exploration.
Notably, GETV-China/GX2020 was firstly detected in pangolin. It suggests that GETV may once again break through the interspecific barrier and further increase the host range. Bayesian analysis shows that GETV has a faster evolutionary rate. The detection of GETV from healthy pangolins indicates that carriage of GETV does not result in clinical symptoms, suggesting that pangolins may be potential reservoir hosts for GETV, which needs further confirmation. New hosts may cause new adaptive mutations in GETV [19], which may further increase the risk of GETV transmission, deserving further in-depth research, and pose new potential threats to public health. Furthermore, GETV-China/GX2020 presented 14 unique amino acids when compared with group IV members. In addition, nine other substitutions deserve special attention (Figure 1D): NS1-G64R, NS1-S241P, NS2-H162P, NS3-E237A, C-Q85H, E2-K112N, E2-H207N, E1-S238P, E1-N396H. This analysis demonstrates evidence of independent amino acid evolution as a result of the previously depicted genetic drift [20]. GETV-China/GX2020 has the highest consistency with the nucleotide and amino acid of the 2017 Thailand isolate, suggesting that GETV-China/GX2020 may have evolved from the 2017 Thailand Sus scrofa isolate. Furthermore, compared with the 2017 Thailand isolate, GETV-China/GX2020 has been discovered important amino acid mutation of E2-H207N. This mutation results in a change in the conformation of the E2 protein, which may in turn affect the infectivity or pathogenicity of the virus. This needs to be verified by further viral reverse genetics experiments. The fact suggests the need to strengthen existing surveillance programs and enhance the efficiency of rapid diagnostics to detect potential outbreaks of GETV and other related Alphaviruses that may occur in Guangxi Province and throughout the country.
## References
1. Sing-Sin, Noor-Adila, Boon-Teong (2022) "Group IV getah virus in Culex mosquitoes" *Malaysia. Emerg Infect Dis*
2. Bin, Huanyu, Guodong (2022) "Getah virus (Alphavirus): an emerging, spreading zoonotic virus" *Pathogens*
3. Nemoto, Bannai, Koji (2014) "Getah virus infection among racehorses" *Emerg Infect Dis*
4. Jin, Simon, Ziqing (2023) "Early genomic surveillance and phylogeographic analysis of Getah virus, a reemerging arbovirus, in livestock in China" *J Virol*
5. Hao, Xu, Li-Xia (2019) "First isolation and characterization of getah virus from cattle in northeastern China" *BMC Vet Res*
6. Ning, Li-Xia, Rong-Guang (2017) "Highly pathogenic swine getah virus in blue foxes, eastern China" *Emerg Infect Dis*
7. Jiao, Yan, Zhai (2023) "Transcriptome screening identifies TIPARP as an antiviral host factor against the Getah virus" *J Virol*
8. Chaonan, Jianfeng, Zongji (2020) "Isolation and characterization of getah virus from pigs in Guangdong Province of China" *Transbound Emerg Dis*
9. Omer, Lyudmila "The short-and long-range RNA-RNA interactome of SARS-CoV-2"
10. (2020) *Mol Cell*
11. Nan, Chengcheng, Yuge (2023) "A new cluster of chikungunya virus West Africa genotype isolated from Aedes albopictus in China" *J Infect*
12. Duo, Ying (2022) "Multiple novel mosquito-borne zoonotic viruses revealed in pangolin virome" *Front Cell Infect Microbiol*
13. Yuan-Yuan, Hong, Shi-Hong (2017) "From discovery to spread: the evolution and phylogeny of Getah virus" *Infect Genet Evol*
14. Tongwei, Xiangling, Qingrong (2022) "Construction and characterization of a full-length infectious clone of Getah virus in vivo" *Virol Sin*
15. Noor-Adila, Sing-Sin, Boon-Teong (2022) "Genomic and in vitro phenotypic comparisons of epidemic and non-epidemic Getah virus strains" *Viruses*
16. Ningning, Xiaofeng, Xiaoling (2022) "Attenuation of getah virus by a single amino acid substitution at residue 253 of the E2 protein that might be part of a new heparan sulfate binding site on alphaviruses" *J Virol*
17. Khwankamon, Noppadol, Nattakarn "A serological survey and characterization of Getah virus in domestic pigs in Thailand"
18. (2022) *Transbound Emerg Dis*
19. Xiaozhou, Qikai, Liwei (2021) "Metagenomic sequencing reveals viral abundance and diversity in mosquitoes from the Shaanxi-Gansu-Ningxia region" *PLOS Negl Trop Dis*
20. Tongwei, Qingrong, Yuxu (2020) "Emergence and phylogenetic analysis of a Getah virus isolated in southern China" *Front Vet Sci*
21. Shi, Zhu, Qiu (2022) "Origin, genetic diversity, adaptive evolution and transmission dynamics of Getah virus" *Transbound Emerg Dis*
22. Haiwei, Yongjie, Ramunas (2017) "Excess of non-conservative amino acid changes in marine bacterioplankton lineages with reduced genomes" *Nat Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12179355&blobtype=pdf | # Virologist Opinions: An Important Component for the Governance of the Convergence of Artificial Intelligence and Dual-Use Research of Concern
Matthew Walsh, Gigi Gronvall
## Abstract
Background: The convergence of artificial intelligence (AI) and the life sciences has brought in silico research into policy conversations around dual-use research of concern and pathogens with enhanced pandemic potential research. This study considers the expert opinions of virologists on governance of AI and life sciences research. Methods: Semi-structured interviews with virologists were conducted and qualitatively analyzed to explore expert opinions about AI and virology. Interviewees were asked about the risks and benefits of AI, policy development considerations, and about evaluating the capability of AI tools in the field of virology.Results: Interviewed virologists generally expressed similar sentiments in responses to questions, including that benefits and risks of AI use in virology research are still to come, that policy and governance should be a process that includes virologist input, and that it is challenging to predict the capability of AI tools without experimental wet-lab validation. Discussion: Governance should be informed by expert opinions of practitioners, and it is important to consider how such opinions are incorporated. Expert opinions are valuable in understanding the impact of governance measures on beneficial research and development, and ensuring that governance measures are practicable and applicable. Virologists interviewed generally had similar opinions around AI and virology topics and often expressed an expectation that their opinions would develop over time. Conclusion: Given the uncertainty around the capability of AI technologies in the life sciences, it may be better to focus on developing frameworks for how governance measures will be developed, and to monitor developments, than to focus on specific interventions.
## Introduction
The impact of artificial intelligence (AI) technologies within the life sciences has recently become a focus of multiple U.S. government and advisory committee policies. A key concern is that these technologies could lower barriers to the development of biological weapons and democratize the ability to cause harm. The convergence of AI and the life sciences has thus brought in silico research into policy discussions surrounding dualuse research of concern (DURC) and pathogens with enhanced pandemic potential (PEPP) research. [1][2][3] However, poorly designed policy-based risk mitigation measures could inadvertently limit the potential benefits of AI without meaningfully mitigating biological risk. To address this, we interviewed virologists to gather their opinions about the potential applications of AI in virology research and on proposed risk mitigation measures. This effort supports the development of rational, policybased risk mitigation measures through publications like this one and informs our contributions as subject matter experts in interviews and conference and workshop discussions.
The United States released an updated policy guiding the oversight of DURC and PEPP research in May of 2024, which included limited discussion of in silico research. 4 The new policy aims to unify federal governance of such research, expand the scope of oversight, and clearly delineate the roles and responsibilities of stakeholders involved in DURC and PEPP research. The new policy seeks to keep pace with scientific advancements and encourages, but does not require, institutional oversight of in silico research and the development of AI tools that can design novel pathogens. Set to take effect in May of 2025, is the result of a multi-year process marked by tension within the biosecurity, biosafety, and life science communities.
In February 2022, the National Science Advisory Board for Biosecurity (NSABB), a Department of Health and Human Services advisory committee, was tasked by the White House with evaluating the effectiveness of the two major U.S. policies governing DURC and research with PEPPs. Over the following year, the NSABB reviewed existing policies, consulted subject matter experts, and considered public comments. The NSABB published a report with their findings and recommendations in early 2023. 5 Among its key findings and recommendations, the NSABB called for the unification of existing federal policies and expansion of the scope of research requiring federal review for potential DURC.
Many virologists, however, expressed concerns that the expanded scope recommended by the NSABB would encompass too much research where the benefits far outweigh the risks, including vaccine development, and potentially overburden those involved in the oversight process. 6 Some virologists also argued that the NSABB's recommendations were developed without adequate virology expertise. 7 In apparent response to these concerns, the Office of Science and Technology Policy (OSTP), which was responsible for the development of the updated policy, issued a Request for Information (RFI) seeking feedback on the NSABB recommendations. While the NSABB did not recommended including in silico experiments within the updated policy's scope, it did recommend continued assessment of their risks and benefits. As part of the RFI, the OSTP sought out opinions regarding whether in silico experimental approaches should require risk assessment and review. The updated policy, which incorporates many of the NSABB's recommendations, goes a step further with regards to in silico experimentation. It encourages voluntary "institutional oversight of in silico research that could result in the development of potential dual-use computational models directly enabling the design of a PEPP or a novel biological agent or toxin." Recognizing the ongoing development of AI technology, the update policy also commits to periodic review and potential updates at least every 4 years.
There are few examples of AI tools that produce outputs typically subject to oversight under traditional DURC/PEPP frameworks. For instance, researchers have developed AI tools capable of predicting SARS-CoV-2 viral escape mutants as accurately as some laboratory methods. 8 However, there are significantly more AI tools developed outside the pathogen context that could theoretically be misused to engineer a pathogen. For example, tools intended to guide protein redesign to enhance desirable characteristics or tools that predict protein structures, could potentially be misapplied to engineer pathogen proteins instead of therapeutics or biotechnology-relevant enzymes. [9][10][11] Governance of these tools will be challenging, as the concerning uses are often distinct from the intended applications. Additionally, it may be unclear how effectively these tools would function in misuse scenarios, further complicating efforts to assess and mitigate potential risks.
In addition to concerns about the misuse of biologyspecific AI tools, there have been concerns about the potential misuse of general purpose chatbots (e.g., ChatGPT) by nefarious actors to aid in the development or production of biological weapons. These concerns are reflected in public policy. President Biden's Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence describes AI tools as enabling the creation and proliferation of biological VIROLOGIST OPINIONS: AI AND DUAL-USE RESEARCH OF CONCERN weapons. 12 This Order is somewhat at odds with the way existing DURC/PEPP oversight frameworks operate. Current frameworks focus on overseeing how a technology (e.g., synthetic biology) is applied, rather than imposing limitations on the development or distribution of the technology itself. Provisions within the National Security Memorandum on Artificial Intelligence (AI NSM) look to address this discrepancy by requiring OSTP, National Security Council Staff, and the Office of Pandemic Preparedness and Response Policy to develop guidance around in silico biological research by April 2026. 13 In the coming months and years, the governance of AI technologies within the life sciences will continue to mature. However, poorly designed policy-based risk mitigation measures risk inadvertently stifling the potential benefits of these technologies without meaningfully mitigating biological risk. Relying solely on the theoretical potential for misuse as a measure of risk may overestimate the likelihood of misuse and mischaracterize the negative impacts of a technology. Instead, incorporating an understanding of AI model performance and observed usage patterns may provide a more accurate assessment of the likelihood and risks of misuse. 14 To inform the development of policy-based risk mitigation measures, we interviewed virologists to identify commonalties in their expert opinions surrounding the use of artificial intelligence in virology research and topics around policy-based risk mitigation measures.
## Materials and Methods
The Johns Hopkins Bloomberg School of Public Health Institutional Review Board Office determined that this research was not human subjects research (FWA #00000287).
## Interviews
Between May and July 2024, a series of interviews were conducted with 21 interviewees. A total of 75 potential interviewees were identified through authorship on peer-reviewed publications relating to virology policy and contacted for participation via email. Potential interviewees who were contacted held a PhD in the life sciences and typically were faculty at academic institutions. An additional nine potential participants were identified through snowball sampling and/or through review of websites of virology departments at major research universities that were not previously represented. In total, six potential participants declined to participate due to time constraints or self-determined lack of expertise in artificial intelligence. Three participants initially agreed to participate but did not respond to scheduling emails. The remaining 54 potential interviewees did not respond.
Researchers conducted semi-structured interviews via Zoom and followed an interview guide. The semistructured interview format included predetermined questions and gave the interviewer the ability to ask additional questions based on the flow of the conversation. The interview guide with predetermined questions was developed based on three areas of active discussion around the incorporation of AI in the field of virology and the researchers' personal experience and expertise related to the convergence of AI and the life sciences and associated policies. For each topic area, between 4 and 7 specific questions were included in the interview guide. Participants were introduced to the three topic areas at the beginning of the interview (Table 1), but were afforded the ability deviate from responding within the bounds of the pre-written questions. Interviewees were encouraged to respond based on whatever AI tool came to mind (e.g., ChatGPT or AlphaFold). All interviewees were provided an open-ended opportunity to share any additional thoughts at the end of the interview.
Interviews were typically around 45 minutes long and conducted on a not-for-attribution basis. Each interview was recorded with the interviewee's consent and an automated transcript was generated using Zoom. The automated transcript was then reviewed against the recording • What are the benefits of AI in the life sciences broadly and virology specifically?
• What are the risks of AI in the life sciences broadly and virology specifically?
• Are there any experiments which should be off-limits from being conducted in silico?
## Governance
• Who should be involved in developing policies relating to AI and virology?
• What are the challenges to achieving effective policy?
• How should the international community address these risks? Some foreign nations do not share the same values as the United States, how does this influence your thinking?
Risk Assessment and Evaluations
• What evidence would you need to see to convince you of the capability of an AI tool?
## 126 WALSH AND GRONVALL
and edited by a researcher to ensure the content of the transcript was accurate.
## Analysis
A qualitative approach was used to analyze the content of the interviews. One interview was conducted with an epidemiologist and infection preventionist but given their limited hands-on experience with virology research the interview excluded from analysis. Researchers reviewed the interview transcripts and identified seven areas of discussion for analysis based on their frequency. As the expertise of interviewees varied, not all questions in the interview guide were covered although each of the three major topic areas were. The areas for analysis typically aligned with specific questions from the interview guide. Each interviewee's response was then coded to identify answers to the question as well as sentiments that were expressed throughout the conversations. The responses to each of these areas were then aggregated and described below.
## Results
Of the 20 interviews analyzed, 19 were with academic virologists (a list of interviewees is provided in the acknowledgements section). Interviewees studied multiple types of viruses, most commonly human influenza viruses, coronaviruses, and cytomegaloviruses, with one being the Director of a biosafety level (BSL)-3 laboratory. Interviewees came from 17 different U.S. academic institutions and two interviewees held a second terminal degree (one MD, one DVM).
## Risks and Benefits
When describing benefits of AI in virology and the life sciences, interviewees described virology-specific and broadly applicable benefits. These benefits corresponded to two categories of AI-tools: Large language model (LLM)-based chatbots and biology-specific AI tools. 15 The virology-specific benefits typically represented enhanced or new capability. These included improved understanding of the relation between the structure and function of biomolecules, as well as broader advancements in understandings disease dynamics, spillover risk, and biological systems.
Interviewees also described generally applicable benefits of AI, which included significant resource savings, such as time and money. Multiple interviewees referenced AI's ability to assist virologists in reformatting and analyzing large datasets, as well as extracting and summarizing information from text, such as peerreviewed publications. Additionally, they highlighted AI tools' capacity to help write code and to help non-native speakers communicate their scientific research more effectively.
When discussing risks, interviewees raised a range of concerns. Some specifically mentioned the potential for AI to contribute to DURC. Interviewees also cited the misuse of AI for data falsification and its potential to mislead users by providing incorrect or inaccurate information. These concerns stemmed partly from anecdotal experiences with AI systems producing incorrect outputs and partly from the lack of transparency with how some AI systems operate. This lack of transparency creates challenges in verifying or validating the output of blackbox AI models during the peer-review process. Others expressed broader worries about the impact on public trust in science if AI-generated outputs were conveyed as scientifically proven when they were not. A few interviewees described concerns about de-skilling, noting that reliance on AI could lead to a diminished understanding of the underlying science and scientific method, and therefore reduced troubleshooting ability in the next generation of scientists.
All interviewees were asked if they believed there to be any experiments that should categorically not be conducted in silico. The responses to this question almost always indicated a belief that there should not be a defined set of experiments that are off-limits for conducting in silico. However, the rationale for these responses were varied. Some participants believed that meaningful risk was attributable to real-world experimentation and that, at the current time, any in silico or AI-based experiment would need to be accompanied by real-world lab work that is subject to various biosafety and biosecurity provisions. This would then raise the question of the value of doing experimentation in silico alone. Additionally, a few interviewees qualified their answers by saying that researchers working on certain pathogens should have no expectation of privacy while others felt that some experiments only made sense in a controlled environment where access was limited or restricted to a subset of individuals.
## Policy
The interviewees' experiences and knowledge of ongoing policy discussions varied, with very few demonstrating a significant understanding of current policy conversations around the convergence of AI and the life sciences. As a result, conversations focused broadly on policy development rather than specific policy proposals. Overwhelmingly, interviewees emphasized the importance of involving subject matter experts in virology and artificial intelligence in the policy development process. Less frequently, they indicated national security experts, publishers, policymakers, and the public should be included. Interviewees typically favored a policy development process that incorporates subject matter expertise during the drafting phase, rather than drafting policies without such input and seeking feedback from experts afterward.
During a few conversations, interviewees highlighted various challenges, including the difficulty of developing policies for rapidly advancing technologies, the need for all stakeholders involved in policy development to have a solid technical understanding, and the politicized atmosphere surrounding science in the United States. While the importance of international coordination on scientific policy was often acknowledged, achieving consensus on a single, multilateral policy to govern international research was often viewed as an insurmountable challenge. However, establishing guidelines or reaching agreement as to what should be subject to a policy was seen as a more achievable goal.
## Risk Assessment and Evaluation
Many proposed policy-based risk mitigation measures rely on anticipating whether an AI model possesses specific capabilities that have been determined to contribute to biological risk. This interview section was initially intended to cover red teaming exercises and other risk analysis and evaluation methodologies that have been reported in the literature. [16][17][18] However, most interviewees were unfamiliar with these evaluation approaches. Consequently, after the first few interviews, the scope was narrowed to focus on the evidence required to determine if an AI model could design a novel pathogen. While most participants acknowledged the difficulty of addressing the question due to the lack of a clear definition for a "novel pathogen", the majority indicated that experimental wet-lab data would be necessary to substantiate such a claim. Although not all participants were asked this specific question, none suggested that in silico results alone would be sufficient to conclude that an AI model could design a novel pathogen.
## Cross-Cutting Themes and Sentiments
Interviewees often elaborated on their answers, providing rationale or anecdotal experiences to support their responses. In doing so, they often expressed sentiments that were shared across multiple interviews. Some of these commonalities are included in Table 2.
## Discussion
Virologists generally share similar opinions about the governance of the convergence of AI and the life sciences. While they recognize AI as an emerging technology with significant promise, many expressed skepticism its current performance in contexts specific to their research. This skepticism extended to the near-term benefits and risks of biology-specific AI tools within the life sciences. When discussing frontier models, interviewees were even more doubtful about their ability to assist novices in crafting detailed experimental plans, aligning with prior evaluations of earlier versions of frontier models. 17 Many interviewees expressed uncertainty in their opinions, noting an expectation that their perspectives will develop, and possibly change, over time and as they gain familiarity with AI and as the technology matures. This uncertainty emphasizes the importance of governance approaches that are adaptable and flexible, ensuring responsiveness to advances in and changing perceptions of AI technologies.
## Table 2. Common themes and sentiments across interviews
## Major category Minor points
Cautious optimism about artificial intelligence (AI) in the life sciences
• In many cases, interviewees pushed back on the premise of a question. When this happened, it typically occurred in the context of the capability of AI tools and the interviewee expressing the opinion that AI tools would have a capability in the near future.
• A view that tools would get better in the future • Personal anecdotes about AI not performing well where AI provided an incorrect or incomplete answer Hesitation that AI tools could enable the design of "worse" pathogens
• Nature is the best at making pathogens worse, why do we think AI would be better?
• The scientific community does not have the ability to routinely make pathogens worse, why do we think AI would be able to?
Trouble justifying why someone would use AI as part of a plan to misuse biology
• Biology can already be very deadly, there are good options to misuse biology • Biology can be unpredictable, so why not cause harm another way
## Miscellaneous
• Concerns about politicizing science • Stating that they "should" be using AI more in their work
• Questioning how regulation can be applied to someone working on a computer
## 128 WALSH AND GRONVALL
Multiple federal stakeholders are responsible for the governing the convergence of AI and the life sciences. The United States and the United Kingdom independently established AI Safety Institutes (AISIs) in 2023 to help develop safety and security guidelines for AI generally, and for the use of AI within the life sciences specifically. In the United States, the AI NSM designates the U.S. AISI as the lead entity on these efforts. Additionally, the AI NSM instructs the OSTP to develop specific "guidance to promote the benefits and mitigate the risks of in silico biological and chemical research", in consultation with relevant stakeholders, of which the AISI is likely to be a key partner.
Involving subject matter experts in technical governance is crucial to ensuring policy measures are practical and to help understand and address potential implementation challenges. The opinions of virologists on some of the existing language in the updated DURC/PEPP policy may foreshadow definitional issues that OSTP will face when crafting a policy on in silico biological research. As stated in the updated DURC/PEPP policy, institutional review committees should oversee the development of AI tools that could be used to directly design novel pathogens or PEPP.
However, interviewees often took issue with a lack of specificity in the definition of a "novel pathogen." When asked what information they would require to determine if a tool could design a novel pathogen, interviewees often first noted that this lack of specificity. For example, interviewees described how "novel" could refer to something that has never been seen before but is highly similar to something that has been (e.g., a SARS-CoV-2 variant) or could describe something with a never-before-seen function. The OSTP simultaneously released a companion document providing implementation guidance for the updated DURC/ PEPP policy, which could serve as a venue for resolving definitional challenges.
When asked to set aside the definitional challenges, all interviewees who were asked about validating a model's capability agreed that some level of wet-lab experimental data would be necessary to justify claims about an AI model's ability to design novel pathogens. Some interviewees indicated that as examples of AI models with the ability to design novel pathogens are reported, they might require less wet-lab data. However, they also highlighted significant risks and limited, if any, benefits in generating what is theorized to be a novel pathogen purely to evaluate an AI tool's capability, especially if the AI tool's primary purpose is unrelated to pathogen design. Additionally, and as noted in the AI NSM, it will be important to ensure that such evaluations are not misconstrued as offensive biological weapons research.
The implementation guide for the updated DURC/ PEPP policy provides a workable definition of "reasonably anticipated" as it relates to experimental outcomes in DURC/PEPP contexts. The OSTP could adopt a similar approach to guide judgements about whether an AI model possesses a particular capability. However, interviewed virologists described how the "black box" nature of some AI systems creates a significant challenge. Unlike DURC/PEPP experimental outcomes where there is a scientific basis for determining what is reasonably anticipated, the "black box" nature of AI systems makes it difficult to understand their operations and therefore predict their behavior and potential outcomes.
Another reason to involve subject matter experts in the development of governance is to ensure a clear understanding of the impact on the beneficial use cases, allowing for a well-informed weighing of risks and benefits. This sentiment was frequently expressed in interviews, often in the context of concerns about the politicization of science. Many interviewees indicated that these concerns stemmed expressed that this concern was derived from their experiences during the COVID-19 pandemic, where they perceived that politicized fear led to calls for overly broad restrictions on their work and that of others.
There are several frameworks available to guide the development of informed governance approaches that balance diverse stakeholder interests. For instance, some frameworks focus on addressing societal concerns, while others seek to balance innovation with security. 19 The effective use of such frameworks depends on achieving broadly agreeable assessment of risks and benefits, which can themselves be evaluated using various methodologies. 14,20 However, given that most virologists interviewed are not yet routinely using AI in their daily research, the risks and benefits of AI remain largely theoretical and carry significant uncertainty. This uncertainty makes it challenging to reach consensus on what the risks and benefits entail. Therefore, it is essential to devote time and effort to discuss and deliberate on how such assessments will be conducted. Doing so can increase the likelihood that the outcomes are broadly accepted and effectively inform governance decisions.
Opportunities to advance the technical and theoretical foundations of such evaluations and assessments are forthcoming. The NIH issued a new charge to the NSABB on November 21, 2024, to provide recommendations related to in silico research in life science settings. This charge directs the NSABB to recommend strategies to identify risks associated with in silico research that could contribute to the design of a PEPP and to develop methods for assessing risks and benefits against clear criteria. While this work would benefit from the expertise of virologists who routinely use AI models, the interviews conducted for this study suggest that such experts may be relatively few in number.
## Study Limitations
While this study identified commonalities and trends within the interviewees' responses, it was not designed to explicitly elicit or evaluate areas of agreement or disagreement. Additionally, the interviews were primarily conducted with academic virologists based in the United States. The findings should not be generalized across virologists globally, to non-virologists in the life sciences, or to those working outside of academic contexts. Finally, this study was conducted shortly after the publication of the updated DURC/PEPP policy and during ongoing public policy debate about governance of AI and the life sciences. Given that this study spanned three months, participants interviewed later in the study period may have had more exposure to the updated DURC/ PEPP policy, and therefore different opportunities to inform and develop their opinions compared to those interviewed earlier.
Additionally, this study specifically solicits and characterizes the opinions and views of one group of stakeholders in the governance of AI and the life sciences: virologists. As such, the findings represent only one piece of the broader conversation. Further research is needed to explore the perspectives of other stakeholder groups, such as biosafety professionals, non-virologist researchers within the life sciences, and AI model developers, to provide a more comprehensive understanding and to inform governance discussions.
## Conclusions
Incorporating subject matter expertise into the development of technology governance is essential to ensure that governance and risk mitigation measures are practical and do not unnecessarily hinder the beneficial applications of a technology. For the governance of the convergence of AI and the life sciences, virologists are one critical group of subject matter experts who must be consulted. However, there currently appears to be limited adoption of advanced, biology-specific AI tools in the routine research activities of virologists. Many virologists interviewed anticipated their opinions would change over time as their familiarity with the technology grows. Considering this, it may be more productive for those responsible for developing governance measures (e.g., the AISIs, NSABB) to prioritize establishing robust frameworks for assessing the risks and benefits of AI technologies in the life sciences rather than pursuing specific interventions at this stage. Establishing such frameworks can provide a basis for more informed and adaptive governance as both AI technologies and their applications in the life sciences mature.
## References
1. Carter, Wheeler, Chwalek (2023) "The convergence of artificial intelligence and the life sciences"
2. Bloomfield, Pannu, Zhu (2024) "AI and biosecurity: The need for governance" *Science*
3. Vindman, Trump, Cummings (2024) "The convergence of AI and synthetic biology: The looming deluge" *arXiv*
4. (2024) "United states government policy for oversight of dual-use research of concern and pathogens with enhanced pandemic potential" *OSTP*
5. (2023) "Proposed biosecurity oversight framework for the future of science. NIH: Bethesda, MD"
6. Goodrum, Lowen, Lakdawala (2023) "Virology under the microscope-a call for rational discourse" *J Virol*
7. Rasmussen, Gronvall, Lowen (2024) "Public role in research oversight" *J Virol*
8. Thadani, Gurev, Notin (2023) "Learning from prepandemic data to forecast viral escape" *Nature*
9. Sumida, Núñez-Franco, Kalvet (2024) "Improving protein expression, stability, and function with ProteinMPNN" *J Am Chem Soc*
10. Lin, Akin, Rao (2023) "Evolutionary-scale prediction of atomic-level protein structure with a language model" *Science*
11. Abramson, Adler, Dunger (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
12. (2023) "Executive order on the safe, secure, and trustworthy development and use of artificial intelligence. The White House: Washington, DC"
13. (2024) "Memorandum on Advancing the United States' Leadership in Artificial Intelligence; Harnessing Artificial Intelligence to Fulfill National Security Objectives; and Fostering the Safety, Security, and Trustworthiness of Artificial Intelligence. The White House: Washington, DC"
14. Goldstein, Sastry (2024) "The PPOu framework: A structured approach for assessing the likelihood of malicious use of advanced AI systems" *Aies*
15. Donkor, Walsh, Titus (2024) "Computing in the life sciences: From early algorithms to modern AI" *Quantitative Biology*
16. Li, Pan, Gopal (2024) "The WMDP benchmark: Measuring and reducing malicious use with unlearning" *Computer Science*
17. Mouton, Lucas, Guest (2024) "The operational risks of AI in large-scale biological attacks: Results of a red-team study"
18. Patwardhan, Liu, Markov (2024) "Building an early warning system for LLM-Aided biological threat creation"
19. Gillum (2024) "Balancing innovation and safety: Frameworks and considerations for the governance of dual-use research of concern and potential pandemic pathogens" *Appl Biosaf*
20. Walsh (2024) "Towards risk analysis of the impact of ai on the deliberate biological threat landscape" *Computer Science* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12590810&blobtype=pdf | Takashi Tanikawa, Masashi Kitamura
## Abstract
Following the publication of the original article [1], it was noted that due to a typesetting error the figure legends were paired incorrectly. The figure legends for Figs. 2 and4 were wrongly given as captions for Fig. 4 and2 respectively.The correct figures and captions have been included in this correction, and the original article has been corrected.
Correction: Inhibitory effect of [6]-shogaol against 3CLpro activity and SARS-CoV-2 infection Takashi Tanikawa 1* , Tsuyoshi Hayashi 2 , Yuka Kiba 1 , Hitoshi Kamauchi 1,3 , Yuichi Someya 2 and Masashi Kitamura 1*
## References
1. Tanikawa, Hayashi, Kiba (1186) "Inhibitory effect of [6]-shogaol against 3CLpro activity and SARS-CoV-2 infection" *BMC Complement Med Ther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12446966&blobtype=pdf | # Corrigendum to "Effects and improvement mechanisms of ultrasonic pretreatment on the quality of fermented skim milk". [Ultrason. Sonochem. 108 (2024) 106958]
Hongsen Yu, Xinyue Cheng, Hang Li, Qiwei Du, Xiaoqun Zeng, Zhen Wu, Yuxing Guo, Daodong Pan, Ultrasonics Sonochemistry
## Abstract
The authors regret, that the "Original Figure 5a" corrigendum is the "Correct image of Figure 5a" in "Fig. 5. Scanning electron microscopy images of microstructures of NUSFM (a) and USFM (b) samples. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802171&blobtype=pdf | # Symptomatic SARS-CoV-2 breakthrough infections broaden the repertoire of Spike-reactive CD4 T cells
Emil Johansson, Yeji Lee, Vicente Fajardo-Rosas, Adam Abawi, Ashmitaa Logandha, Ramamoorthy Premlal, April Frazier, Jason Greenbaum, Pandurangan Vijayanand, Ricardo Da, Silva Antunes, Alessandro Sette
## Abstract
SARS-CoV-2 breakthrough infections (BTIs) are relatively common, but little is known in terms of differentiating responses associated with asymptomatic BTI (ABTI) versus symptomatic BTI (SBTI). Here, we investigated the impact of ABTI and SBTI on antibody and T-cell responses toward Spike. SBTI donors had significantly higher plasma anti-Spike RBD IgG titers compared to both ABTI and SARS-CoV-2 vaccinated donors with no signs of previous infection (VAX) donors. While no impact of ABTI or SBTIs was found in the magnitude of Spike-specific CD4 and CD8 T-cell responses, both ABTI and SBTI donors had significantly higher CD4 T-cell responses toward non-Spike antigens. In-depth characterization of Spike-specific CD4 and CD8 T cells at scRNA/TCRseq level revealed that ABTI and SBTI induced different alterations of the CD4 compartment. These included an IFNG high -skewed response among cytokine-producing cells and a concurrent expansion of type II IFN-responsive T H 17-like cells in SBTI. SBTI donors were further found to have an increased CD4 TCR repertoire diversity compared to VAX donors. ABTI-associ ated alterations of the Spike-specific compartment were intermediate between SBTI and VAX groups, likely reflecting lower antigen exposure and less inflammatory environment during ABTIs versus SBTIs.IMPORTANCE SARS-CoV-2 mRNA vaccines have been shown to induce robust T-cell responses, crucial for long-term protection against severe SARS-CoV-2 infection. Hybrid immunity, created by a combination of vaccination and infection, has been associated with improved protection against severe SARS-CoV-2 infection. Here, we have investi gated the impact of asymptomatic and symptomatic breakthrough infections (BTIs) on T-cell responses toward Spike, compared to donors that have only received mRNA SARS-CoV-2 vaccines. Symptomatic, and to a lesser extent asymptomatic, BTIs broad ened the repertoire of Spike-reactive CD4 T cells and induced a more pro-inflammatory Spike-reactive T-cell response capable of enhancing activation of TH17-like cells. These findings represent the first characterization of T-cell responses in ABTI in comparison to SBTI, and in comparison to vaccinated individuals (VAX) who did not experience BTIs. KEYWORDS T cell, SARS-CoV-2, vaccines, T-cell repertoire S ARS-CoV-2 breakthrough infections (BTIs) have become relatively common, driven by viral evolution that enables immune escape from B-and T-cell responses induced by previous vaccinations and infections, and waning of vaccine-or previous infectioninduced immunity over time (1-4). BTIs may function as natural boosters that broaden and renew the immune protection provided by vaccination or previous infection and are recognized as playing an important role in maintaining an "immunity wall" against severe COVID-19 and its complications (4-7).BTIs have been generally characterized in terms of magnitude and breadth of cross-neutralizing antibody responses (8-10), durability of T-cell immunity across
SARS-CoV-2 variants (11)(12)(13)(14), and the complex interplay between vaccination timing, viral load, and immune response (15)(16)(17). BTIs boost T-cell responses, broaden the range of antigens targeted by cellular responses beyond the Spike (S) antigen (12,15,18,19), and generate novel T-cell epitopes that recognize mutations found in Delta and Omicron variants (19)(20)(21).
It is commonly observed that BTIs encompass a wide spectrum of clinical presen tations, ranging from symptomatic disease (symptomatic BTIs; SBTIs) to completely asymptomatic (asymptomatic BTIs; ABTIs) (22,23). However, the immunological profiles of T cells responding to antigen stimulation as a function of SBTI versus ABTI have not been described. The different inflammatory environments associated with BTIs may also drive distinct T-cell differentiation programs, underlying differences in T-cell subset activation, functional programming, and effector capabilities that remain poorly characterized at the single-cell level (24,25). We have previously shown that ABTI was associated with an expansion of T H 17-like cells compared to vaccinated individuals without infection (VAX) (26), but it is not known how these cells are impacted by SBTI. Further research is therefore needed to elucidate whether this represents a protective immune signature that prevents progression to SBTI, or whether the expansion reflects inflammatory events occurring during ABTI that would be amplified in SBTI.
Activation-induced marker (AIM) assays have become a standard method for measuring and characterizing antigen-specific T-cell responses, including those against SARS-CoV-2, due to their HLA-agnostic and function-independent attributes (27)(28)(29). It was recently reported that certain T-cell subsets, including T H 17 and T reg cells, can undergo TCR-independent activation by cytokines during antigen stimulation, leading to non-specific upregulation of activation markers (11,30). However, "bystander" popula tions may themselves provide important biological insights into vaccine or infection in vivo responses and reactogenicity (31,32).
Here, we used single-cell RNA sequencing (scRNAseq) and T-cell receptor sequencing (scTCRseq) to characterize Spike (S)-specific T-cell responses in VAX, ABTIs, and SBTIs, to define distinct immunological signatures related to T-cell functionality and T-cell subsets responding to antigen stimulation. We found that SBTIs broadened the repertoire of S-reactive CD4 T cells and induced an IFNG high -skewed pro-inflammatory T H 1 response that enhanced the activation of S-specific T H 17-like cells.
## RESULTS
## BTI increases antibody titers and T-cell response breadth
Our goal was to investigate the impact of SARS-CoV-2 SBTI on the antibody quantity and on the quantity and quality of T-cell responses, in comparison to responses observed in ABTI and VAX donors. Accordingly, we assembled a study cohort of COVID-19 vaccinated donors (Table 1) which included 13 donors who had reported a SBTI and compared them to 15 donors who had experienced an ABTI and 18 VAX donors, both selected from a previous study (26). The VAX and ABTI donors consistently tested negative in all PCR or antigen SARS-CoV-2 diagnostic tests that they undertook and remained asymptomatic for an average of 199 and 187 days, respectively (26). Donors were defined as having experienced ABTI on the basis of having T-cell responses toward non-S antigens, as detected using a peptide "megapool" (MP) of experimentally validated SARS-CoV-2 non-S epitopes (CD4RE) with minimal homology to endemic common cold coronaviruses (26,33). Conversely, VAX donors had no detectable reactivity to the CD4RE pool using a stringent stimulation index (SI) cutoff (26,33).
Donors were matched for age, number of vaccine doses, days since last vaccination, year of sample donation, and ethnicity (Table 1). Samples were collected on average 129, 133, and 158 days post last vaccination for VAX, ABTI, and SBTI groups, respectively. For SBTI donors, samples were collected at an average of 103 days after symptom onset. All SBTI donors had fully recovered from acute COVID-19 symptoms and tested PCR-negative for SARS-CoV-2 at the time of blood collection.
As expected (19,26), SBTI donors had significantly higher anti-S RBD IgG titers compared to both VAX (P < 0.001) and ABTI (P < 0.001) groups (Fig. 1A). In line with the previously described high stability of S-specific T-cell responses (26), the magnitude of S-reactive CD4 and CD8 T-cell responses was not significantly different between the three groups (Fig. 1B), as measured by the combined use of a MP of peptides spanning the entire sequence of the S antigen and an AIMassay (34). The magnitude of CD4 T-cell responses toward non-S antigens (CD4RE) showed the highest response in the SBTI group, followed by the ABTI group, with both groups being significantly higher compared to VAX donors (P<0.001) (Fig. 1C).
Taken together, these results show that while S-reactive T-cell responses are relatively stable in terms of magnitude and are not affected by BTIs, BTI increases breadth of SARS-CoV-2-reactive CD4 T cells, and SBTI increases the anti-S RBD IgG titers. Although not significant, responses to the non-S antigens were highest in SBTI donors compared to ABTI donors, possibly reflecting stronger or longer antigenic stimulation, or a more pronounced pro-inflammatory environment.
## Definition of S-reactive CD8 T-cell subsets associated with the three study groups
We hypothesized that differences in the inflammatory environment and antigenic stimulation associated with the different vaccination and infection exposures could also alter the phenotypes and quality of S-reactive T cells. We therefore sorted AIM + CD8 T cells following S MP stimulation and performed scRNAseq and scTCRseq of the exclusive S-reactive cells (CD69 + CD137 + ). Samples were obtained from the same individuals in the cohorts described above. In total, 89,839 CD8 T cells were included in the analysis following quality control (QC) assessment of the data set.
Based on their gene expression and shared nearest neighbor (SNN)-based cluster analysis, we identified nine CD8 T-cell clusters (Fig. 2A). Fig. S1A shows cluster marker gene expression, respectively, which was further visualized using UMAP (Fig. S1B). High expression of GZMK was found in the two most abundant clusters, Clusters 0 (GZMK high ), and Cluster 1 (T CM GZMK high ) which further had high expression of central memory markers TCF7, IL7R, and KLF2. Cluster 2 (T EFF GNLY high CTSW high ) contained effector cells
## Phenotypes of S-reactive CD8 T cells are not altered by BTIs
We did not observe significant differences in the cluster frequency of CD8 T-cell subpo pulations across the three groups (Fig. S1C). Conflicting findings have been reported regarding the induction of CD8 T-cell exhaustion following repeated SARS-CoV-2 vaccination and infection (19,(39)(40)(41). Accordingly, we generated an exhaustion signa ture score based on 62 genes previously linked to CD8 T-cell exhaustion (39) (Table S1). In line with our previous findings (19,26), we did not observe increased expression of the exhaustion-associated signature following ABTI or SBTI (Fig. S1D).
We next investigated the impact of BTI on the CD8 TCR repertoire by comparing the diversity of CDR3 sequences from paired TCRα and TCRβ chains, based on Chao1, Gini-Simpson, D50 diversity metrics. Consistent with the similar CD8 cluster frequencies across the three groups, we did not observe any significant differences in TCR repertoire diversity (Fig. S1D).
Taken together, these results show that the subsets of S-reactive CD8 T cells, although phenotypically heterogeneous, are remarkably similar across study groups and are not significantly altered by BTIs. Likewise, S-reactive CD8 T cells are associated with similar TCR repertoire diversity.
## S-reactive CD4 T cells are phenotypically heterogenous
We performed scRNAseq and scTCRseq on S-reactive CD4 T cells by sorting AIM + (OX40 + CD137 + ) cells following S MP stimulation. After applying similar QC measures as for CD8 T cells, 140,640 CD4 T cells were included in the analysis. In total, we identified 11 CD4 T-cell clusters (Fig. 2B), which were annotated on the basis of their differentially expressed genes (Fig. S2A) and further visualized using UMAP (Fig. S2B). Cluster 0 (T reg ) cells showed upregulated expression of the canonical Treg markers FOXP3 and IKZF2 (Fig. S2A). Cluster 1 (T H 17-like) cells were enriched for T H 17 signature genes, including DPP4, IL4I1, and CTSH (Fig. S2A). Cluster 2 (T CM ) contained cells with enrichment for central memory signature genes TCF7, IL7R, and KLF2 (Fig. S2A). Cluster 3 (cytotoxic T H 17-like) cells had high expression of T H 17-associated genes, as well as the cytotoxic genes GNLY, CCL5, and PRF1 (Fig. S2A). A T H 1 signature was found in Clusters 3, 4, 5, and 8-10 (Fig. S2A), including TBX21 (T-bet), IL2, and TNF. Cells in Cluster 3 (T H 1/T FH ) cells had low cytokine expression but high expression of ID3, a marker of memory CD4 T cells with the capacity to give rise to both T H 1 and T FH cells (42). Cells in Cluster 5 (T H 1 IFNG high ) expressed high levels of IFNG, IL2, and TNF. Cluster 6 (CD4 CTL ) contained cells with high expression of cytotoxic genes GZMB, GZMA, NKG7, and CCL5 (Fig. S2A). Cluster 7 (T H 2) contained cells with high expression of IL4, IL5, IL13, and the Th2 lineage master regula tor GATA3 (Fig. S2A). Within clusters 9 and 10, the expression of CSF2 was highest in Cluster 9 (T H 1 CSF2 high ) cells and the expression of CCL20 in Cluster 10 (T H 1 CCL20 high ) cells (Fig. S2A).
These results demonstrate that both CD4 and CD8 T-cell subsets responding to in vitro S MP stimulation display remarkable phenotypic heterogeneity. The various subsets overlap with those reported in prior studies conducted by our laboratory and other groups (26,41,43).
## SBTI induces phenotypic changes of S-reactive CD4 T cells
Based on cluster relationship at different clustering resolutions, the 11 CD4 T-cell clusters could be combined into two primary groups each accounting for approximately 33% and 66% of the total number of cells (Fig. S2C), respectively. One containing the T H 17-like, cytotoxic T H 17-like, T CM and T reg clusters (Group 1) and the second containing the remaining clusters of T H 1, T H 2, and T CTL cells (Group 2). The Group 1 cluster (T H 17-like, cytotoxic T H 17-like, T CM , and T regs ) was significantly enriched in frequency in SBTI compared to the VAX donors (P = 0.012; Fig. 2C). Within Group 1, the frequency of Clusters 1 (T H 17-like) and 3 (Cytotoxic T H 17-like) was significantly higher in SBTI donors (P = 0.002 and P = 0.029, respectively) compared to VAX donors (Fig. 2D; Fig. S2D). In Group 2, SBTI donors had significantly lower frequencies of Cluster 4 (T H 1/T FH ) and Cluster 9 (CSF2 high ) (P=0.004 and P=0.010, respectively) compared to VAX donors (Fig. 2E). ABTI donors showed frequencies of these four clusters that were intermediate between VAX and SBTI donors although no statistical significance was observed.
Taken together, these results show that while BTIs do not alter the magnitude of S-specific CD4 T-cell responses, BTIs alter the phenotype of the S-reactive T cells. These changes were strongest among the SBTI donors, with a trend for similar alterations among the ABTI donors.
## T H 1-biased CD4 T-cell responses promote T H 17-like cell expansion
Recent studies highlighted the interplay between cells in vitro following peptide stimulation (11,32). These are mediated by cytokines and receptor-ligand interactions between conventional T cell (T CONV ) subsets such as T regs and T H 1/T FH 11 , as well as between T CONV and MAIT cells (32). As several cytokine genes were upregulated in Group 2 clusters (Fig. S2A), we hypothesized that the activation of Group 1 clusters might be enhanced by the activation of Group 2 cells. DGE analysis of the two groups revealed distinct expression patterns. Group 2 was characterized by strong cytokine expression, with the top 10 differentially expressed genes including IL2, IL3, IL13, CSF2, IL21, IFNG, and CCL20. These genes are involved in the Reactome Gene Sets "Signaling by Interleukins. " In contrast, Group 1 showed elevated expression of genes involved in the GO Biological Process "Cellular response to cytokine stimulus, " including cytokine receptors (IFNGR2, IL2RB, and IL2RG) and Interferon-induced genes (IFI6, IFIT1, and ISG15) (Table S2; Fig. S3A). Of note, these two groups resemble two previously described modules of activated CD4 T cells identified following TCR engagement (36), one with high expression of TCR signaling genes and one with high type II interferon signaling genes (30, 36) (Fig. S3B).
We next examined the frequency of the clusters within Group 2 separately to determine how the frequency of these cells was associated with the expansion of Group 1 clusters. Interestingly, we observed an inverse association between Cluster 4 (T H 1/T FH ) and Cluster 5 (T H 1 IFNG high ). Specifically, we found a trend for decreased frequency of Cluster 4 (T H 1/T FH ) (P = 0.135) and increased frequency of Cluster 5 (T H 1 IFNG high ) (P = 0.161) (Fig. 3A), and a significantly higher T H 1 IFNG high :T H 1/T FH ratio (P = 0.043) in SBTI donors compared to VAX donors (Fig. 3B). We observed a similar trend for higher T H 1 IFNG high :T H 1/T FH ratio (P = 0.141) in the ABTI donors, in line with the moderate perturbation of Group 1 and Group 2 clusters observed in Fig. 2. We further found that the frequency of the T H 17-like cluster was negatively associated with the frequency of the T H 1/T FH cluster (r = -0.647, P < 0.001) and positively associated with the frequency of the T H 1 IFNG high cluster (r = 0.427, P = 0.003) among all Group 2 clusters (Fig. 3C). No significant associations were found for T CM or T reg cells.
These results suggest that an IFNG high -T-cell response could promote the expansion of the T H 17-like cells. Supporting this observation, cells from the T H 17-like cluster in SBTI donors exhibited increased expression of genes involved in the Reactome Gene Sets "Interferon Signaling, " including BST2, IFI6, and ISG15 (Fig. S3C; Table S3), in comparison to VAX donors. Taken together, these results highlight the interplay between subsets of S-reactive CD4 T cells, suggesting that TCR-dependent IFNγ release drives enhanced activation of type II IFN-responsive CD4 T cells (36).
## TCR profiles associated with the different donor cohorts
When comparing the TCR repertoire of all CD4 T cells, we found that SBTI donors had significantly higher Chao1 (P = 0.014) diversity index (Fig. 4A), reflecting an increased richness of the repertoire, that is, number of clones, driven by an increased proportion of singleton clones. In line with this, the D50 diversity index, defined as the number of clonotypes accounting for 50% or more of the total TCR repertoire, was significantly higher in SBTI compared to VAX donors (P = 0.026). Finally, the Gini-Simpson diversity index was significantly higher in SBTI donors compared to VAX donors (P = 0.021), highlighting a more even distribution of clonal sizes across the TCR repertoire. Taken together, these results indicate a reduced dominance of robustly expanded clones and increased frequency of weakly expanded and singleton clones following SBTIs. Similar to the phenotypic changes of S-reactive T cells, the characteristics of the TCR repertoire of the ABTI donors were intermediate to the VAX and the SBTI donors.
We next assessed clonal expansion and clonal sharing at the individual cluster level. We found expanded clones in all clusters, but in varying proportions. The lowest proportion of expanded cells was found in T reg cells, while the largest proportion was found in T H 2 cells (Fig. 4B). Extensive overlap between expanded clones was found across clusters, except for the T reg cluster (Fig. 4C). Consistent with a recent publication (11), the T reg cluster had the lowest proportion of clones shared with other clusters, and these cells potentially represent those activated in a bystander fashion by cytokines released from cells reactive to the S peptide pool. Examination of TCR clone expansion in individual clusters, downsampled to equal number of cells per cluster and donor group, revealed that expanded clones in ABTI and SBTI donors were predominantly small (2-5 cells) or medium (5-20 cells) sized, while the VAX donors also displayed large (20-100 cells) and hyperexpanded clones (>100 cells) (Fig. 4D). When comparing the proportion of small expanded clones in each cluster, we found that the SBTI donors had a higher proportion of small T-cell clones in all clusters compared to the VAX donors (P = 0.10; Fig. 4E,left). Importantly, the proportion of small expanded clones among Group 2 clusters remained significantly higher in SBTI compared to VAX donors (P = 0.037; Fig. 4E,right), but not for Group 1 clusters (P = 0.118; Fig. 4E,center). Taken together, these results show that SBTI infection increases the richness of the TCR repertoire of S-reactive CD4 T cells, associated with an increased frequency of small expanded clones. This increase was most prominent among the cytokine-producing Group 2 clusters, with non-significant trends observed among Group 1 clusters. This supports the conclusion that symptomatic BTI expands the bona fide Spike-specific TCR repertoire beyond what is achieved by vaccination alone. The characteristics of the TCR repertoire of the ABTI donors were intermediate to the VAX and the SBTI donors.
## Assessment of TCR specificity
A recent study suggested that T H 17-like cells found among AIM + cells were activated in a TCR-independent bystander fashion (30). To address this, we leveraged the VDJ database to identify S-specific TCRα chains (44). In total, we identified 668 S-specific cells, 647 of which were specific for the previously described immunodominant peptide TFEYVSQPFLMDLE (45,46), present in our S MP (Table S4), representing 4.2% of all clonally expanded cells. In line with the recent publication (30), we found that these cells were primarily found among Group 2 clusters (Fisher's exact test P < 0.001), with few cells observed across Group 1 T H 17-like, cytotoxic T H 17-like, T CM , and T reg cells (Fig. 5A). To assess if the TCR clones in the Group 1 clusters represented cells activated through potential bystander activation, we next assessed if TCRα chains matched to previously described non-SARS-CoV-2 antigens in the VDJ database were accumulated in Group 1 clusters. However, we only found 168 cells with TCRα chain matches in the VDJ database (Fig. S3D; Table S5). The most common match was to Influenza A, a ubiquitous virus in the United States (47). Among all CD4 T cells, we found only 90 cells (0.1%) expressing previously defined Influenza A-specific TCRα, with no selective accumulation in Group 1 compared to Group 2 clusters (Fisher's exact test P = 0.741; Fig. 5B), although an accumulation of Influenza A-specific TCRα was identified among cytotoxic T H 17-like cells (Fisher's exact test P < 0.001).
In conclusion, these results suggest that SBTIs broaden the repertoire of S-reactive T cells and promote an IFNG high T H 1 response that more efficiently activates S-reactive T H 17-like cells.
## DISCUSSION
Here, we present the first characterization of T-cell responses in ABTI in comparison to SBTI and VAX donors. The characterization included determining the magnitude of responses and characterizing the functional T-cell subsets responding to antigen stimulation and associated TCR repertoires. The results indicate that ABTI responses are associated with characteristics that are intermediate between SBTI and VAX groups. The fact that SBTI responses are in general higher than ABTI indicates that the symptomatic nature of the BTI is not associated with an intrinsic defect of these donors to mount a T-cell response. In fact, individuals emerging from a SBTI episode, because of the higher level of immunity, might be associated (albeit temporarily) with higher levels of protection from further reinfection and symptomatic disease. Conversely, ABTIs are associated with intermediate levels of immune responses, which are lower compared to SBTIs. This might reflect lower levels of infection either in terms of viral titers or duration.
In line with previous studies (19,26,48), we found that the magnitude of S-reactive CD4 and CD8 T-cell responses was not significantly altered following BTIs, highlighting the stability of vaccine-induced T-cell responses. However, by employing scTCRseq, we found that following SBTIs, the S-reactive CD4 T cell repertoire is broadened by the emergence of smaller clones and reduced dominance of larger clones. In our previous study, we showed that BTIs induced de novo responses toward variant-specific epitopes (19), which could in part explain the observed broadening of the TCR repertoire in SBTI donors. In the case of ABTI donors, the median values of all three diversity measurements of ABTI donors were intermediate to that of VAX and SBTI donors, possibly due to reduced antigen exposure in these donors.
Further characterization of the S-reactive CD4 T cells revealed two main subsets, one containing cytokine-producing cells and a second group containing cytokine-responsive cells. SBTI donors were associated with an increased T H 1 IFNG high :T H 1/T FH ratio among the cytokine producing cells, paralleling previous observations that hybrid immunity (SARS-CoV-2 vaccination and infection) increases frequencies of IFNγ-releasing CD4 T cells (41). Further, the frequency of IFNG high cells was positively associated with the expansion of type II IFN-responsive T H 17-like cells. These results are in line with two major modules of memory CD4 T cells following TCR-mediated activation: one con taining cytokine-producing cells, and a second containing cells whose activation was modulated through type II IFN signaling (36).
Consistent with the impact of IFNγ on T H 17-like cells observed here, two recent studies showed that T-cell responses were augmented through a feedback loop where IFNγ, released by S-specific memory T cells, activated innate-like T cells (11), which in turn amplified the signal by releasing more IFNγ (32). This effect was dependent on the time between vaccination doses, with longer intervals being associated with reduced inflammatory signatures and higher recall-expansion of TCR clones observed in the previous vaccine dose (11). The increased T H 1 IFNG high :T H 1/T FH ratio in SBTI donors observed here could therefore in part be reflective of a more recent exposure to SARS-CoV-2 compared to ABTI and VAX donors, combined with the heightened inflammatory state during a BTI (49). The bidirectional cross-talk between innate-like T cells and conventional T cells plays an important role during the mounting of an immune response toward vaccines and viruses (50,51). The clear phenotypic alteration of the CD4 T cell compartment observed in SBTI suggests that the type II IFN-responsive cells also play an important role during a BTI, potentially in the orchestration of immune responses and protection from recurrent infection. However, it is still not known if these cells play a pro-inflammatory or an autoregulatory role. Here, T regs were found to be the biggest type II IFN responsive cluster, but we did not observe any significant differences between the three study groups. The second largest cluster contained T H 17-like cells with a gene signature, resembling pro-inflammatory T H subsets linked to autoimmunity (52). However, we did not observe high levels of IL-17 family of cytokine genes or IL22 in this cluster, and previous studies have not reported IL-17 to be an abundantly produced cytokine by S-reactive CD4 T cells (28,53). More research is therefore needed to elucidate the role of the type II interferon responsive T H 17-like cells observed here, and to determine if these cells represent bona fide T H 17 memory cells, reflect a BTI-imprinted T H 17-like transcriptional signature in S-reactive cells, or infection-primed S-specific cells developed during the BTI (49,54). The T H 17-like populations identified in our study (clusters 1 and 3) likely represent cells with T H 17 characteristics representing transient activation states. The selective enrichment of these populations in SBTI compared to both VAX and ABTI likely reflects differences in the inflammatory environment and antigen exposure between natural infection and vaccination. During natural SARS-CoV-2 infection, prolonged systemic antigen exposure combined with tissue damage-associ ated inflammatory mediators (IL-6, IL-1β, and IL-23) may drive T H 17 differentiation more effectively than the localized inflammation induced by mRNA vaccination (55). Whether the T H 17-like populations represent beneficial barrier immunity to reinfections or contribute to symptomatic disease and inflammation requires further functional validation.
Surprisingly, we observed T H 2 cell expansion in all three study groups, even though T H 2 cells are not typically reported as a major subset of Spike-reactive T cells (26,56). While T H 2 responses are generally considered less protective against viral infections, T H 2 expansion might represent a regulatory response to limit excessive T H 1/T H 17-mediated inflammation, and a recent study found enrichment of the expression T H 2 cytokines in long-persisting CD4 T-cell clones (57), suggesting that these cells might play a role in the long-term protection against SARS-CoV-2.
Recent reports have raised concern that subpopulations of AIM + cells, including T H 17-like and T reg cells, become activated through TCR-independent activation by cytokines (11,30). Similar to a previous study (11), we observed that the majority of expanded TCR clones found in T reg cells were found exclusively within this cluster, suggesting that these cells may indeed represent bystander-activated cells. However, the remaining clusters showed a high degree of clonal sharing between clusters. Corroborating recent findings (30), we observed that CD4 T cells specific for the S peptide TFEYVSQPFLMDLE were enriched among non-T H 17-like cells. We therefore next investigated if we could find an accumulation of TCR clones previously described to be specific for non-SARS-CoV-2 antigens in the VDJ database among T H 17-like cells, but no such enrichment of Influenza A-specific T cells was observed. However, it is important to note that this analysis is limited by the availability of TCR sequences from CD4 T cells at the VDJ database, which does not include TCR sequences specific for the ubiquitous virus EBV (58). We did not find sufficient evidence that the CD4 T cells with a cytokine-activated gene signature are Spike-specific CD4 T cells. However, as shorter antigen-exposure intervals have been shown to induce reduced expansion of recalled clones among AIM + cells (11), the TH17-like cells could thus represent de novo responses to variant-specific epitopes following BTIs. The specificity of Group 1 cells could in future studies be addressed by sorting Group 1 cells and restimulating them with Spike peptides, or engineered T cell lines with TCRs from Group 1 cells could be generated and tested for Spike reactivity. Given the modest sample sizes of our study groups, the findings presented in this study should be considered exploratory and require validation in a larger cohort with improved statistical power. This is particularly important for confirming trends observed in comparisons between ABTI and SBTI, where differences were subtle and did not always reach statistical significance.
In conclusion, our data show that ABTI and SBTI are associated with different features of T-cell responses. This likely reflects the increased antigen exposure, reduced exposure interval, and heightened inflammatory environment during an SBTI compared to an ABTI, and in turn might influence the degree and duration of protection from further BTIs.
## MATERIALS AND METHODS
## Subjects and samples
All samples were collected by the Clinical Core at the La Jolla Institute and provided informed consent (26). Sample collection was approved by the LJI Institutional Review Board under IRB Protocol #VD-214. The VAX and ABTI donors were part of a longitudinal study where they were rigorously screened to ensure they tested negative for SARS-CoV-2 both before and during the sample collection period either by antigen or PCR test (26). The VAX and ABTI donors were selected to match the SBTI donors as best as possible in regards to gender balance, ethnicity, age, time since last vaccine dose, and number of vaccine doses (Table 1). Importantly, we found no significant differences between the three study groups in these parameters.
## Peripheral blood mononuclear cells (PBMC) and plasma isolation
Blood samples were collected in heparin-coated blood bags, and PBMCs were isolated by density-gradient sedimentation with Ficoll-Paque PLUS (Cytiva), as previously described (26,59).
## SARS-CoV-2 S-RBD ELISA
Plasma titers of anti-S RBD IgG titers were determined by ELISA as described in previous studies (26,60,61).
## AIM assay
To study T-cell responses against SARS-CoV-2, two peptide MPs were prepared follow ing the MP approach, previously outlined as a comprehensive method for analyzing T-cell responses across diverse epitopes and populations (34). A MP of 15-mer peptides overlapping by 10 spanning the entire S protein sequence (253 peptides) and a MP (CD4RE) composed of 284 experimental defined epitopes from non-S (R) region of SARS-CoV-2 (Table S4) were selected as previously described (28,33).
To detect T-cell-specific responses, we employed an AIM assay methodology in combination with peptide pool stimulation using the dual activation marker expres sion of OX40 + CD137 + or CD69 + CD137 + to detect antigen-specific CD4 and CD8 T cells, respectively (34). Briefly, 1-2 × 10 6 PBMCs were plated and immediately stimulated with MPs (1 μg/mL), or phytohemagglutinin-L (PHA) (10 μg/mL; Roche, San Diego, CA, USA) and DMSO as positive and negative controls, respectively, in RPMI 1640 supplemented in 5% human serum (Gemini Bio-Products) for 18-24 h. After culture, cells were collected, washed, and stained with Live/Dead eFlour 506 (Thermo Fisher), CD3 BUV805 (UCHT1, BD Biosciences), CD8 BV650 (RPA-T8, BioLegend, San Diego, CA, USA), CD19 V500 (HIB19, BD Biosciences), CD14 V500 (M5E2, BD Biosciences), CD4 BV605 (RPA-T4, BD Bioscien ces), OX40 PE-Cy7 (Ber-ACT35, BioLegend), CD137 APC (4B4-1, BioLegend), and CD69 PE (FN50, BD Biosciences) for 30 min at 4°C. All samples were acquired on a Bio-Rad ZE5 Analyzer (Bio-Rad Laboratories, Hercules, CA, USA) and analyzed with FlowJo 10.9 software (Tree Star, Ashland, OR, USA), as previously described (26).
The data were normalized with a minimum response level set at 0.005%. The specific T-cell responses were calculated by subtracting the background (DMSO stimulation) values. For each population, the limit of detection (LOD) was determined as the upper 95% confidence interval (CI) of the DMSO values, while the limit of sensitivity (LOS) was calculated as the median plus two times the standard deviation (SD) of DMSO. The SI was calculated as the percentage of stimuli response divided by the percentage of response in the DMSO control. Positive responses were defined as responses greater than LOS, with SI > 2 for CD4 T cells or SI > 3 for CD8 T cells. Responses with SI < 2 for CD4 T cells or SI < 3 for CD8 T cells were normalized to LOD. ABTI donors were identified by the first time point at which their CD4RE MP reactivity responses surpassed a highly stringent 10-fold SI threshold (26). While this cutoff provides good confidence in identifying individuals with previous SARS-CoV-2 exposure, the VAX donor group may still include individuals with previous exposure that did not generate detectable CD4RE responses.
## AIM + T cell sorting and sample preparation
According to the previously described AIM assay protocol, PBMCs were thawed and stimulated for 24 h with the S MP. Cells were washed and stained with an antibody cocktail containing the antibodies described above for 30 min at 4°C, protected from light. TotalSeq-C oligonucleotide-conjugated antibodies (BioLegend) were also added at this step at 0.01 mg/mL final concentration (one distinct antibody per sample). After two washes in PBS, cells were resuspended in 500 μL of FACS buffer (PBS containing 2 mM EDTA (pH 8.0) and 0.5% BSA) and stored at 4°C until flow cytometry acquisition. S-specific CD4 + and CD8 + AIM + T cells were subsequently sorted using a BD FACSAria Fusion cell sorter (Becton Dickinson), as previously described (26). On average, 20,000-40,000 AIM + T cells per subject were sorted for single-cell RNA-seq assays (10x Genomics, Pleasanton). Cells were collected directly into low-retention, sterile 1.5 mL collection tubes (Thermo-Fisher) pre-chilled on ice and containing 500 μL of PBS:FBS solution (1:1, vol:vol) supplemented with RNAse inhibitor (1:100, Takara Bio).
## Cell isolation and single-cell RNA-seq and TCR-seq library preparation
A total of five to six different samples were multiplexed, each labeled with distinct TotalSeq-C DNA-oligo barcoded, totaling approximately 60,000 cells per 10× lane (equivalent to one well of the 10× chip). Samples were centrifuged at 600 × g for 10 min at 4°C, and the supernatant was carefully removed, leaving behind 10-12 μL of residual volume. Cell pellets were resuspended in 25 μL of 10x Genomics-compatible resuspension buffer (0.22 μm filtered PBS containing 0.04% ultrapure bovine serum albumin, Sigma-Aldrich). 33 μL of the resuspended cells was transferred to an 8-strip PCR tube for downstream processing as per the 10x Genomics protocol. The remaining cells were used for cell-counting QC.
Following the manufacturer's recommendations, single-cell RNA libraries were generated using the 10x Genomics standard 5′TAG v2 chemistry. Both cDNA ampli fication and library preparation were carried out using 13 PCR cycles. Barcoded cDNA products were collected, quantified, and pooled at equimolar concentrations. The libraries were sequenced on the Illumina Novaseq6000 platform with paired-end sequencing (S4 100 × 100 cycles, Illumina) configured as follows: read length: read 1, 100 cycles; read 2, 100 cycles; i7 index, 10 cycles; and i5 index 10 cycles.
Size and quantity QCs were performed throughout the procedure by capillary DNA high sensitivity electrophoresis (HS NGS Fragment Kit, 1-6000 bp, Fragment Analyzer, Agilent) and Picogreen assay (Quant-iT PicoGreen dsDNA Assay Kits and dsDNA Reagents). Sequencing depth was aimed to reach 30,000 reads per cell for gene expression 8,000 reads/cell for TCR, and antibody-feature sequencing (multiplexing analysis).
## Single-cell transcriptome analysis
The Cellranger multi pipeline (v8.0.1) was used to perform alignment to the pre-built GRCh38 human genome reference (2020-A), UMI counting, sample deconvolution, and TCR clonotype calling. The Seurat toolkit (v5.2.1) was employed in R (v4.4.1) to per form QC, unbiased clustering, dimensionality reduction, and cluster annotation on each aggregated data set. The following QC criteria were enforced to minimize doublets and eliminate low-quality transcriptomes: 1,500 ≥ unique molecular identifier (UMI) count ≤ 20,000; 800 ≥ gene count ≤ 4,400; and mitochondrial UMI percentage ≤ 7%. As we had previously found that MAIT cells are captured among AIM + cells (26), we excluded cells expressing the known MAIT TCRα chain segment TRAV1-2-TRAJ33/12/20, and the TCRβ chain segment TRBV6-1/6-4/20-1 (62). To prevent donor-specific gene expression of individual TCR genes to influence gene expression-based clustering, TCRA/B/D/G genes were pulled from the gene expression matrix and counts were aggregated into a single gene feature for each gene (63). Samples were normalized using the SCTransform function ("V2" regularization, v0.4.2) with mitochondrial content as the variable to regress and 3,000 variable features. Samples were thereafter integrated using the RunHarmony function (v1.2.3), with sequencing batch and donor ID speci fied as covariates. For CD4 T cells, 20 PCA dimensions were used as cell embedding features, while 15 PCA dimensions were used for the CD8 T cells. FindNeighbors and RunUMAP were performed using 30 dimensions, and FindClusters was performed using a resolution of 0.2 for CD4 T cells and 0.3 for CD8 T cells. The Clustree package (v0.5.1) was used to plot cluster number and cell distribution at different cluster resolutions. The Seurat AddModuleScore function was used to score the CD8 exhaustion gene signature, using 62 genes listed in Table S1, and an average score per donor was calculated.
## Single-cell differential gene expression analysis (DGEA)
The FindAllMarkers function from Seurat (64) (v3.2.3) was used to perform DGEA with MAST (v1.32.0) employed as the statistical framework for differential expression analysis (65). The function was used to identify transcripts enriched in T-cell clusters, as well as to compare the transcriptome of the same cluster between two study groups. Mitochon drial and ribosomal genes were excluded for cluster DGEA between study groups. A gene was considered as differentially expressed if Benjamini-Hochberg adjusted P value < 0.05 and either log2 fold change (LFC) ≥ 0.25 or LFC ≤ -0.25. Metascape (66) (http:// metascape.org) was used for gene set overrepresentation analyses.
## TCR repertoire analysis
The immunarch R package (v0.9.1) was used to calculate the TCR repertoire diversity of cells with paired TCRα and TCRβ chains. To identify cells with TCR sequences with known antigen-specificity, CDR3 amino acid sequence, V-segment, and J-segments from paired TCRα and TCRβ chain sequences, as well as individual TCRα and TCRβ chains, were mapped to sequences deposited at the VDJ database (Table S6). The scRepertoire (v2.2.1) package was used to visualize the frequency of expansion from paired TCRα and TCRβ clones per cluster and make circle plots showing clonal sharing among clusters.
## Quantification and statistical analysis
Statistical analyses were performed in GraphPad Prism 10.4.2. Data plotted in linear scale were expressed as median ± 95% CI, while the data plotted in logarithmic scales were expressed as geometric mean ± 95% CI. Unpaired comparisons between groups were performed using the nonparametric two-tailed Kruskal-Wallis test adjusted with Dunn's test for multiple comparisons. The Pearson correlation coefficient test was used for association analysis.
## References
1. Tan, Kwan, Rodríguez-Barraquer et al. (2023) "Infectiousness of SARS-CoV-2 breakthrough infections and reinfections during the Omicron wave" *Nat Med*
2. Kuhlmann, Mayer, Claassen et al. (2022) "Break through infections with SARS-CoV-2 omicron despite mRNA vaccine booster dose" *Lancet*
3. Servellita, Syed, Morris et al. (2022) "Neutralizing immunity in vaccine breakthrough infections from the SARS-CoV-2 Omicron and Delta variants" *Cell*
4. Feikin, Higdon, Abu-Raddad et al. (2022) "Duration of effectiveness of vaccines against SARS-CoV-2 infection and COVID-19 disease: results of a systematic review and meta-regression" *Lancet*
5. Da Silva Antunes, Grifoni, Frazier et al. (2023) "An update on studies characterizing adaptive immune responses in SARS-CoV-2 infection and COVID-19 vaccination" *Int Immunol*
6. Dan, Da, Antunes et al. (2022) "Observations and perspectives on adaptive immunity to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)" *Clin Infect Dis*
7. Altarawneh, Chemaitelly, Ayoub et al. (2023) "Effects of previous infection, vaccination, and hybrid immunity against symptomatic Alpha, Beta, and Delta SARS-CoV-2 infections: an observational study" *EBioMedicine*
8. Ahmed, Einhauser, Peiter et al. (2024) "Evolution of protective SARS-CoV-2-specific B and T cell responses upon vaccination and Omicron breakthrough infection"
9. Einhauser, Asam, Weps et al. (2024) "Longitudi nal effects of SARS-CoV-2 breakthrough infection on imprinting of neutralizing antibody responses" *EBioMedicine*
10. Lee, Tan, Reynaldi et al. (2023) "Durable reprogramming of neutralizing antibody responses following Omicron breakthrough infection" *Sci Adv*
11. Murray, Amini, Ferry et al. (2025) "Dosing interval is a major factor determining the quality of T cells induced by SARS-CoV-2 mRNA and adenoviral vector vaccines" *Sci Immunol*
12. Nesamari, Omondi, Baguma et al. (2024) "Post-pandemic memory T cell response to SARS-CoV-2 is durable, broadly targeted, and cross-reactive to the hypermuta ted BA.2.86 variant" *Cell Host Microbe*
13. Matei, Chivu-Economescu, Dragu et al. (2025) "Long-term durability and variant-specific modulation of SARS-CoV-2 humoral and cellular immunity over two years" *Int J Mol Sci*
14. Kim, Kim, Jung et al. (2024) "Omicron BA.2 breakthrough infection elicits CD8 + T cell responses recognizing the spike of later Omicron subvariants" *Sci Immunol*
15. Painter, Johnston, Lundgreen et al. (2023) "Prior vaccination promotes early activation of memory T cells and enhances immune responses during SARS-CoV-2 breakthrough infection" *Nat Immunol*
16. Koutsakos, Reynaldi, Lee et al. (2023) "SARS-CoV-2 breakthrough infection induces rapid memory and de novo T cell responses" *Immunity*
17. Lim, Tan, Bert et al. (2022) "SARS-CoV-2 breakthrough infection in vaccinees induces virus-specific nasalresident CD8+ and CD4+ T cells of broad specificity" *J Exp Med*
18. Fenn, Koycheva, Kundu et al. "INSTINCT Study Investigators. 2025. Early de novo T cell expansion following SARS-CoV-2 infection predicts favourable clinical and virological outcomes" *EBioMedicine*
19. Tarke, Ramezani-Rad, Pereira Neto et al. (2024) "SARS-CoV-2 breakthrough infections enhance T cell response magnitude, breadth, and epitope repertoire" *Cell Rep Med*
20. Dang, Anzurez, Nakayama-Hosoya et al. (2023) "Breadth and durability of SARS-CoV-2-specific T cell responses following long-term recovery from COVID-19" *Microbiol Spectr*
21. Lineburg, Crooks, Raju et al. (2023) "Breakthrough SARS-COV-2 infection induces broad anti-viral T cell immunity"
22. Lee, Woo, Kim et al. (2022) "Clinical manifestations of COVID-19 breakthrough infections: a systematic review and metaanalysis" *J Med Virol*
23. Bergwerk, Gonen, Lustig et al. (2021) "Covid-19 breakthrough infections in vaccinated health care workers" *N Engl J Med*
24. Kotliar, Curtis, Agnew et al. (2025) "Reproducible single-cell annotation of programs underlying T cell subsets, activation states and functions" *Nat Methods*
25. Fischer, Ansari, Wagner et al. (2021) "Single-cell RNA sequencing reveals ex vivo signatures of SARS-CoV-2-reactive T cells through "reverse phenotyping" *Nat Commun*
26. Da Silva Antunes, Fajardo-Rosas, Yu et al. (2025) "Evolution of SARS-CoV-2 T cell responses as a function of multiple COVID-19 boosters" *Cell Rep*
27. Dan, Arlehamn, Weiskopf et al. (2016) "A cytokine-independent approach to identify antigen-specific human germinal center T follicular helper cells and rare antigen-specific CD4+ T cells in blood" *J Immunol*
28. Grifoni, Weiskopf, Ramirez et al. (2020) "Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals" *Cell*
29. Poloni, Schonhofer, Ivison et al. (2023) "Tcell activation-induced marker assays in health and disease" *Immunol Cell Biol*
30. Zheng, Burmas, Tan et al. (2025) "Deconvoluting TCR-dependent and -independent activation is vital for reliable Ag-specific CD4 + T cell characterization by AIM assay" *Sci Adv*
31. Yosri, Dokhan, Aboagye et al. (2024) "Mechanisms governing bystander activation of T cells" *Front Immunol*
32. Amini, Garner, Shaw et al. (2025) "MAIT and other innatelike T cells integrate adaptive immune responses to modulate intervaldependent reactogenicity to mRNA vaccines" *Sci Immunol*
33. Yu, Wang, Garrigan et al. (2022) "Development of a T cell-based immunodiagnostic system to effectively distinguish SARS-CoV-2 infection and COVID-19 vaccination status" *Cell Host & Microbe*
34. Da Silva Antunes, Weiskopf, Sidney et al. (2023) "The MegaPool approach to characterize adaptive CD4+ and CD8+ T cell responses" *Curr Protoc*
35. Boutet, Benet, Guillen et al. (2021) "Memory CD8 + T cells mediate early pathogen-specific protection via localized delivery of chemokines and IFNγ to clusters of monocytes" *Sci Adv*
36. Szabo, Levitin, Miron et al. (2019) "Single-cell transcriptomics of human T cells reveals tissue and activation signatures in health and disease" *Nat Commun*
37. Wu, Yoshikawa, Inoue et al. (2023) "CD83 expression characterizes precursor exhausted T cell population" *Commun Biol*
38. Lugli, Galletti, Boi et al. (2020) "Stem, effector, and hybrid states of memory CD8+ T cells" *Trends Immunol*
39. Kusnadi, Ramírez-Suástegui, Fajardo et al. (2021) "Severely ill COVID-19 patients display impaired exhaustion features in"
40. Cells *Sci Immunol*
41. Fuentes, Mohamed, De et al. (2022) "Evidence of exhausted lymphocytes after the third anti-SARS-CoV-2 vaccine dose in cancer patients" *Front Oncol*
42. Cai, Gao, Adamo et al. (2023) "SARS-CoV-2 vaccination enhances the effector qualities of spike-specific T cells induced by COVID-19" *Sci Immunol*
43. Shaw, Deng, Omilusik et al. (2022) "Id3 expression identifies CD4(+) memory Th1 cells" *Proc Natl Acad Sci*
44. Zhang, Upadhyay, Hao et al. (2023) "Multimodal single-cell datasets characterize antigen-specific CD8 + T cells across SARS-CoV-2 vaccination and infection" *Nat Immunol*
45. Goncharov, Bagaev, Shcherbinin et al. (2022) "VDJdb in the pandemic era: a compendium of T cell receptors specific for SARS-CoV-2" *Nat Methods*
46. Mudd, Minervina, Pogorelyy et al. (2022) "SARS-CoV-2 mRNA vaccination elicits a robust and persistent T follicular helper cell response in humans" *Cell*
47. Minervina, Komech, Titov et al. (2021) "Longitudinal high-throughput TCR repertoire profiling reveals the dynamics of T-cell memory formation after mild COVID-19 infection"
48. Schmidt, Lapuente (2021) "T Cell Immunity against Influenza: the long way from animal models towards a real-life universal flu vaccine" *Viruses*
49. Uwamino, Yokoyama, Sato et al. (2025) "Breakthrough Infection after a primary series of COVID-19 vaccination induces stronger humoral immunity and equivalent cellular immunity to the spike protein compared with booster shots" *Vaccines (Basel)*
50. Sl, Solis, Chen et al. (2024) "SARS-CoV-2 inflammation durably imprints memory CD4 T cells" *Sci Immunol*
51. Amini, Klenerman, Provine (2024) "Role of mucosal-associated invariant T cells in coronavirus disease 2019 vaccine immunogenicity" *Curr Opin Virol*
52. Proulx, Wiggins, Reames et al. (2025) "Noncanonical T cell responses are associated with protection from tuberculosis in mice and humans" *J Exp Med*
53. Schnell, Littman, Kuchroo (2023) "TH17 cell heterogeneity and its role in tissue inflammation" *Nat Immunol*
54. Juno, Tan, Lee et al. (2020) "Humoral and circulating follicular helper T cell responses in recovered patients with COVID-19" *Nat Med*
55. Cano-Gamez, Soskic, Roumeliotis et al. (2020) "Single-cell transcriptom ics identifies an effectorness gradient shaping the response of CD4 + T cells to cytokines" *Nat Commun*
56. Martonik, Parfieniuk-Kowerda, Rogalska et al. (2021) "The role of Th17 response in COVID-19" *Cells*
57. Sureshchandra, Lewis, Doratt et al. (2021) "Single-cell profiling of T and B cell repertoires following SARS-CoV-2 mRNA vaccine" *JCI Insight*
58. Liu, Antoun, Fries et al. (2025) "Long-persisting SARS-CoV-2 spikespecific CD4+ T cells associated with mild disease and increased cytotoxicity post COVID-19" *Nat Commun*
59. Dunmire, Verghese, Balfour (2018) "Primary Epstein-Barr virus infection" *J Clin Virol*
60. Arlehamncsl (2021) "PBMC Isolation & Cryopreservation from Whole Blood v1"
61. Dan, Mateus, Kato et al. (2021) "Immunological memory to SARS-CoV-2 assessed for up to 8 months after infection" *Science*
62. Zhang, Mateus, Coelho et al. (2022) "Humoral and cellular immune memory to four COVID-19 vaccines" *Cell*
63. Garner, Amini, Fitzpatrick et al. (2023) "Single-cell analysis of human MAIT cell transcriptional, functional and clonal diversity" *Nat Immunol*
64. Lewis, Sutherland, Soldevila et al. (2023) "Identification of cow milk epitopes to characterize and quantify diseasespecific T cells in allergic children" *J Allergy Clin Immunol*
65. Satija, Farrell, Gennert et al. (2015) "Spatial reconstruction of single-cell gene expression data" *Nat Biotechnol*
66. Finak, Mcdavid, Yajima et al. (2015) "MAST: a flexible statistical framework for assessing transcriptional changes and characterizing heterogeneity in single-cell RNA sequencing data" *Genome Biol*
67. Zhou, Zhou, Pache et al. (2019) "Metascape provides a biologist-oriented resource for the analysis of systems-level datasets" *Nat Commun* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12671204&blobtype=pdf | # Transiently produced IgGs enable universal SARS-CoV-2 diagnosis and differentiation recent from past infections
Chuchu Li, Jingzhi Li, Chen Dong, Qian Zhen, Xiaoyu Li, Lan Yang, Jingjing Cao, Xiaoxiao Kong, Hua Tian, Lu Zhou, Fengcai Zhu, Hongwei Ma, Liguo Zhu, Alexander Bello
## Abstract
Early diagnosis and differentiating infected from vaccinated animals (DIVA) are the milestone in mitigation, control, and eradication of virus. However, the continu ous evolution of new virus strains has hindered the success of many classical techniques, and no DIVA-compatible vaccines are available for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). We here develop a universal diagnostic protein-peptide hybrid microarray (PPHM) for different SARS-CoV-2 strains infections and extend the DIVA concept to differentiate recently infected from historically vaccinated/infected hosts (DIVH) in the context of SARS-CoV-2 using in a single sample. First, we analyzed the amino acid sequence identities, which exceeded 96.6%, among the SARS-CoV-2 strains-Delta, Omicron, and New Omicron-and the wild strain. Based on this analysis, we identified a unique diagnostic combination for each of the three SARS-CoV-2 strains using the same peptide probes from the wild strain. Subsequently, we developed a diagnostic combination without regard to strain specificity, PPHM SARS-CoV-2 , suitable for all three strains, demonstrating a specificity of 99.0% (99/100) and a sensitivity of 90.2% (175/194). Finally, we successfully achieved DIVH and classified 314 individuals to the classifications of uninfected (57/314, 18.2%), newly vaccinated or infected (5/314, 1.6%), recently infected (78/314, 24.8%), and historically vaccinated or infected (174/314, 55.4%) using PPHM SARS-CoV-2 . IMPORTANCE Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused global health concerns and led to economic losses reaching trillion dollars since 2019. Many diagnostic techniques are showing great success; for instance, nucleic acid amplification showed high specificity and sensitivity in SARS-CoV-2 infections. However, the results of these techniques can be affected by the continuous evolution of SARS-CoV-2 strains, and these techniques were not compatible with differentiating infected from vaccinated animals (DIVA). To circumvent these unfavorable outcomes, a novel strategy is imperative. In this study, we made an exciting discovery: transiently produced IgGs can be used not only for diagnosis and DIVA but also for differentiating recent infections from past infections (approximately 70 days after the occurrence of vaccina tion or infection). This protein-peptide hybrid microarray technique will have a positive impact on the mitigation and eradication of coronaviruses. KEYWORDS DIVH, transiently produced IgGs, infection diagnosis, PPHM, SARS-CoV-2 strains T he COVID-19 pandemic, caused by the severe acute respiratory syndrome corona virus 2 (SARS-CoV-2), emerged in December 2019 and rapidly became a global threat (1, 2). With the incessant transmission of COVID-19, SARS-CoV-2 has mutated into a variety of variants due to numerous duplications (3). Preventing the mutation and transmission of SARS-CoV-2 seems to be a big issue. It is widely recognized that
controlling the spread of infectious diseases involves three fundamental strategies: eliminating the infection sources, cutting off the infection routes, and protecting the susceptible populations (3). Timely isolation of infected individuals has proven effective in curbing the spread of COVID-19. Therefore, developing a technology that accurately and rapidly identifies COVID-19 infections within a short timeframe is of paramount importance. In serological surveys, through this indicator, we can infer the recent infection rate of SARS-CoV-2 in the population, rather than the cumulative positive rate of this virus. Thus, we can timely assess the disease burden caused by this virus infection.
Three diagnostic assays are commonly used for detecting SARS-CoV-2 infections: (i) nucleic acid amplification test (NAAT), (ii) antigen detection, and (iii) antibody detec tion. These assays detect viral genetic material, viral proteins, and antibodies in serum, respectively (4). Currently, NAAT is considered the gold standard for confirming COVID-19 infections due to its high sensitivity and specificity. However, the accuracy of NAAT results can be influenced by factors in the pre-analytical, analytical, and post-analytical phases, including sampling, nucleic acid extraction, and analysis (5). Antigen-based detection tests are generally the most straightforward diagnostic assays but tend to have lower specificity and sensitivity compared to NAAT (6). The protein-based assays have been developed and serve as a foundation for serological tests to detect coronaviruses in human serum samples (7,8). However, positive results can indicate either a past or current infection or a vaccination (9,10), as the COVID-19 vaccines have been widely administered in the population.
The differentiating infected from vaccinated animals (DIVA) concept is crucial in veterinary medicine, as it allows for differentiating infected and vaccinated animals using a single serum sample, which is essential for the complete eradication of a virus. This concept has been successfully applied to various viruses, such as foot-and-mouth disease virus (11), and pest des petite ruminants virus (PPRV) (12). In the classical DIVA approach, vaccines lacking specific proteins are developed, and animal serum samples containing antibodies against these proteins indicate exposure to the natural infections rather than vaccination. However, developing such marker vaccines is both time-consuming and uncertain (11).
Therefore, a novel serological technology for coronavirus should be capable of rapidly, conveniently, and efficiently detecting various strains of SARS-CoV-2 and should incorporate DIVA functionality in a single serum sample. In our previous research, the peptides synthesized according to the amino acid sequences of structural proteins (e.g., the fusion protein of PPRV and capsid protein of porcine circovirus 2 (PCV2)), along with proteins, were used to create a protein-peptide hybrid microarray (PPHM). This platform detects two distinct types of IgG responses: transiently produced IgGs (TPIs), which are anti-peptide antibodies reflecting short-lived immune responses, and persistently produced IgGs (PPIs), which are anti-protein antibodies associated with longer-lasting immunity (10,13). The PPHM enabled the successful diagnosis of PPRV and PCV2 infections and supported DIVA strategy implementation (9,10). Meanwhile, this serological assay utilizes a "digital microarray index" (DMI) to diagnose viral infections, addressing issues of non-reproducible interactions (NRIs) and non-specific interactions (NSIs) (14).
In this study, we aim to establish a standardized protocol for the development of a PPHM that possesses diagnostic capabilities, and to extend the DIVA concept to differentiate recently infected from historically vaccinated/infected hosts (DIVH) within the context of SARS-CoV-2. Initially, we analyzed the amino acid sequence identities between the SARS-CoV-2 wild strain and three other strains: Delta, Omicron, and the New Omicron variant. We then used the same probes on Microarray-#1.5 to generate a unique diagnostic combination for these three SARS-CoV-2 strains. Subsequently, we sought to develop a strain-agnostic diagnostic combination, PPHM SARS-CoV-2 , applicable to all three strains. Finally, we intended to apply the PPHM SARS-CoV-2 to screen serum samples from 314 individuals undergoing physical examination to evaluate its DIVH functionality.
## MATERIALS AND METHODS
## Serum samples
At the verification phase, we used 225 serum samples, which contained 27 For evaluating DIVH functionality, 314 serum samples were collected from persons who were under physical examination in April 2023.
All serum samples were inactivated at 56°C for 30 min and stored at -20°C before testing.
## Peptides and proteins
By analyzing the amino acid sequence of the SARS-CoV-2 wild strain (MN908947), 20-mer peptides with an overlap of 10 aa residues, partially covering four structural proteins (S, N, M, and E) of SARS-CoV-2, were chemically synthesized by GenScript (Jiangsu, China), and ultimately yielded 136 peptides as peptide probes (Table S1). RBD and N proteins (GenScript, Jiangsu, China) of SARS-CoV-2 and set them as protein probes in the experiments. In a previous study, six probes (S39, S95, S37, N15, S21, and N protein) were excluded for they had excessive antigenicity in the discovery phase (15). In this study, we first used 131 peptides (Table S1, black font), and RBD protein.
SARS-CoV-2 Delta (GenBank No. XBA20213), Omicron (GenBank No. WZA33875), and New Omicron (GenBank No. XBA90971) strains were analyzed using NCBI.
## Microarrays
Microarray-#1.5 was formed with one protein (RBD protein), and 131 peptides (Table S1, black font). Briefly, 0.1 mg/mL of each peptide or protein was printed onto the iPDMS substrate membrane using the non-contact printer sciFLEXARRAYER S1 (Scienion, Berlin, Germany) to form a 12 × 12 array (one microarray with 12 arrays for serum screening). In each array, eight positive controls were printed with human IgG at a concentration of 10 µg/mL, and one negative control was printed with buffer.
PPHM SARS-CoV-2 was formed after the verification stage and used for the validation stage of serum screening, which consisted of one RBD protein and eight peptides (including N10, N24, N37, N39, S46, S58, S63, and S71). Briefly, 0.1 mg/mL of each peptide or protein was printed onto the iPDMS substrate membrane using the non-con tact printer sciFLEXARRAYER S1 (Scienion, Berlin, Germany) to form a 4 × 4 array (one microarray with 48 arrays for serum screening). In each array, three positive controls were printed with human IgG at a concentration of 10 µg/mL, and one negative control was printed with buffer.
## Serum screening with microarrays
Two hundred twenty-five serum samples used in the verification stage were screened using Microarray-#1.5. Serum was first diluted 1:100 with serum-dilution buffer (1% bovine serum albumin, 1% casein, 0.5% sucrose, 0.2% polyvinylpyrrolidone, and 0.5% Tween20 in 0.01 M phosphate-buffered saline, pH = 7.4) and 450 mL was added into each array of Microarray-#1.5, incubated for 30 min on a shaker (500 rpm, 37°C). One array incubated with serum-dilution buffer was used as a negative control. The microarray was then rinsed three times with washing buffer and incubated with 450 mL/ array of horseradish peroxidase (HRP)-conjugated goat anti-human IgG (Sigma-Aldrich) diluted 1:10,000 in Peroxidase Conjugate Stabilizer/Diluent (Thermo Scientific) for another 30 min on a shaker (500 rpm, 37°C), followed by the same washing steps as described above. Then, 100 mL of chemiluminescence substrate (Thermo Scientific) was added onto each array of the microarray, and the images were taken at a wavelength of 635 nm using Clear 4 imaging system (Suzhou Epitope, Suzhou, China). The signal of any dot was defined as signal readout of dot minus signal readout of background.
Two hundred ninety-four serum samples used in the validation stage and 314 serum samples from physical examination were screened using PPHM SARS-CoV-2 . Serum was first diluted 1:100 with serum-dilution buffer (1% bovine serum albumin, 1% casein, 0.5% sucrose, 0.2% polyvinylpyrrolidone, and 0.5% Tween20 in 0.01 M phosphate-buffered saline, pH = 7.4) and 100 mL was added into each array PPHM SARS-CoV-2 , incubated for 30 min on a shaker (500 rpm, 37°C). One array incubated with serum-dilution buffer was used as a negative control. The PPHM was then rinsed three times with washing buffer and incubated with 100 mL/array of HRP-conjugated goat anti-human IgG (Sigma-Aldrich) diluted 1:10,000 in Peroxidase Conjugate Stabilizer/Diluent (Thermo Scientific) for another 30 min on a shaker (500 rpm, 37°C), followed by the same washing steps as described above. Then, 90 mL of TMB (Thermo Scientific) was added into each array of the PPHM, incubated at 37°C for 10 min, then rinsed three times with pure water. Finally, the informative signal of IgGs against probes using a microarray imager (Suzhou Epitope). The data were processed using IBT software, which was also developed by Suzhou Epitope. The signal for each dot was calculated using the following equation: signal dot = signal readout -signal background.
## Statistical analysis
We use the formula signal = (peptide signal intensity -background intensity)/(back ground intensity) to convert the intensity to a signal value and define signal = 2 as the filter value of a single probe. The whole-microarray-level cutoff value (i.e., 2) used a DMI, which represented the overall responding peptide probes on PPHM. ROC was performed using GraphPad Prism. P value was calculated by SPSS.
Based on previous studies (10, 13), we defined: both DMI positive and anti-protein positive as recent infection; DMI negative and anti-protein positive as past infection; both DMI and anti-protein negative as uninfected; DMI positive and anti-protein negative as early infection or early seroconversion.
## RESULTS
## Study design
The antibody response is largely contingent upon the ability of antibodies elicited during a natural infection or vaccination to recognize viral antigens upon exposure to the virus. These antibodies may be circulating in the blood, or generated by memory B cells and plasma cells upon re-exposure to the viral antigens (16). In the context of coronaviruses, the majority of antibodies target the SARS-CoV-2 viral antigens, notably the spike surface glycoprotein (S) and the nucleocapsid phosphoprotein (N) (17,18). The S protein is targeted due to its critical role in viral entry (19), and the N protein is targeted due to its high expression levels during infection (20), rendering both key antigens for antibody recognition.
Minor differences in amino acid sequences can significantly influence pathogenicity and variations in antibodies induced by conformational epitope, while differences at the linear epitope level are likely minimal and predictable (21). Consequently, we posit that serological diagnostics based on peptides (i.e., linear epitopes) may effectively cover different strains. If substantial differences in linear epitopes exist or if they elicit distinct humoral immune responses, strain-specific identification may also be feasible. Under this hypothesis, we compared the amino acid sequences of the S and N proteins from four SARS-CoV-2 strains, including wild-type (Wuhan-Hu-1, Clade 19A), Delta (B.1.617.2, Clade 21A), Omicron (B.1.1.529, Clade 21K), and New Omicron (XBB.1.5, Clade 23A). These classifications were based on PANGO lineage and GISAID clade definitions. Despite variant divergence, the amino acid sequence identities of the S and N proteins remained above 96.6% (Fig. 1; Table 1), allowing us to use wild-type-derived peptide probes for diagnostic combination design across variants.
## One unique diagnostic combination for each of the three SARS-CoV-2 strains
As previously reported, a diagnosis combination of 131 peptides was identified during the discovery phase (15). Here, we presented details from the verification phase. Specifically, we used 27 serum samples collected in 2019 (Control 1: negative, pre-COVID-19 outbreak), 30 serum samples from individuals vaccinated over 200 days ago, but confirmed uninfected with SARS-CoV-2 (Control 2), and 50 serum samples from Delta-infected individuals for serological screening using Microarray-#1.5 (containing 132 probes, i.e., 131 peptides and a RBD protein). Ultimately, we identified a diagnostic combination of six peptides for confirming Delta strain infections: S46, S58, S63, S71, S82, and N24 (Combination-Delta).
For Control 1, the serum samples were theoretically negative as they were collected prior to the COVID-19 outbreak, and the results showed no response to SARS-CoV-2 probes (Fig. 2a andc-green). Regarding Control 2, as the samples were collected over 200 days post-vaccination from individuals confirmed uninfected, they were expected to be negative for anti-peptide antibodies (10,13). The results showed only three peptides responded (peptides S58, S82, and N24), and the positive rates were all below 20% (Fig. 2a). Using a cutoff of DMI ≥ 2, the positive rate for Control 2 was 0% (Fig. 2b). In contrast, the positive rates for the peptides screened with Delta-infected serum samples ranged from 30.0% to 84.0%, exhibiting over a 30% difference in response rate between control and Delta-infected groups (Fig. 2a). This Combination-Delta showed significant differences in responses between controls and Delta-infected groups (P < 0.005), achieving 100% specificities and 94.0% sensitivity (Fig. 2b andd), with an area under the ROC curve (AUC) of 0.9889 (Fig. 2e), indicating excellent diagnostic perform ance.
Protein-based assays utilizing the RBD as a detection antigen can also be developed on Microarray with a signal value cutoff ≥2. Serum samples from Control 2 tested positive for anti-RBD antibodies due to vaccination, indicating that protein-based assays cannot be used for DIVA. Serological screening revealed sensitivity for RBD protein in Delta-infec ted to be 98.0%, and specificities for Control 1 and Control 2 to be 100% and 6.7%, respectively (Fig. 2d). The responses of anti-RBD were with no significant differences between Control 2 and Delta-infected groups, and with significant differences between Control 1 and Delta-infected groups (P < 0.005) (Fig. 2c).
Employing the same methodology used for the Combination-Delta, we applied this approach during the verification phase for diagnosing the Omicron and New Omicron strains. In addition to the two Control groups, we collected 85 positive serum samples from Omicron-infected individuals and 33 from New Omicron-infected individuals to screen using Microarray-#1.5. We then identified a diagnostic combination of seven peptides for the Omicron strain: S46, S58, N10, N14, N16, N37, and N39 (Combination-Omicron) (Fig. S1a). Combination-Omicron showed significant differences in responses between controls and Omicron-infected groups (P < 0.005), achieving 100% (Control 1) and 93.3% (Control 2) specificities (Fig. S1b andd), and 92.9% sensitivity (Fig. S1b andd), with an AUC of 0.9634 (Fig. S1e). Furthermore, we developed a diagnostic combination for the New Omicron strain, comprising five peptides and an RBD protein: S46, S71, N25, N37, N39, and RBD (Combination-New Omicron) (Fig. S2a). The inclusion of the RBD protein may be necessary due to the limited recognition between the peptide and antibodies induced by infections with the new Omicron variant. Combina tion-New Omicron showed significant differences in responses between controls and New Omicron-infected groups (P < 0.005), achieving 100% (Control 1) and 96.7% (Control 2) specificities (Fig. S2b andd), and 90.9% sensitivity (Fig. S2b andd), with an AUC of 0.9983 (Fig. S2e).
The results of these three diagnostic combinations suggest that the DMI can not only be used for infectious diagnosis, but also be able to DIVA.
## A diagnostic combination suitable for all three SARS-CoV-2 strains
As previously mentioned, serum samples infected with three SARS-CoV-2 strains were initially screened using the same set of probes (i.e., Microarray-#1.5), with amino acid sequence identities reaching 96.6% between the peptide probes from the wild strain and each of the three SARS-CoV-2 strains (Fig. 1). The screening results further indicated cross-reactivity among serum samples infected with different SARS-CoV-2 strains when using the unique diagnostic combination for each strain (Table 2): (i) Combination-Delta exhibited diagnostic sensitivities of 67.1% and 81.8% for Omicron and New Omicron strains, respectively; (ii) Combination-Omicron showed sensitivities of 86.0% and 75.8% for Delta and New Omicron strains, respectively; and (iii) Combination-New Omicron displayed sensitivities of 86.0% and 80.0% for Delta and Omicron strains, respectively. antibodies or a signal value ≥2 for anti-protein antibodies, the groups displayed varying specificities and sensitivities. At a DMI cutoff of ≥2, both control groups showed 100% specificity, with a 94.0% sensitivity for the Delta-infected group. Using RBD protein as a probe with a signal value ≥2, the two control groups had specificities of 100% and 6.7%, respectively, and the Delta-infected group had a sensitivity of 98.0%. (e) Using a DMI cutoff ≥2, the Combination-Delta demonstrated a specificity of 100% and a sensitivity of 94.0% for Control 1 and Delta-infected samples, respectively. Statistical significance: ns, not significant; ****P < 0.005.
These data suggested that cross-reactivity presented a challenge in achieving strain-spe cific diagnostic combination. To facilitate screening, we developed a new combination (N10, N24, N37, N39, S46, S58, S63, and S71), along with RBD protein, named PPHM SARS- CoV-2 , intended for positive diagnosis of SARS-CoV-2 without regard to strain specificity in the verification phase (Fig. 3a). To validate the feasibility of the PPHM SARS-CoV-2 , we used 100 negative serum samples collected before 2019, and a positive group comprising 92 Delta-infected, 75 Omicron-infected, and 27 New Omicron-infected serum samples, for serological screening with PPHM SARS-CoV-2 in the validation phase. The screening results showed that the positive rates of probes in the 194 positive serum samples ranged from 8.2% to 78.9% (Fig. 3a), with the RBD protein achieving a 94.3% positive rate (Fig. 3b). Ultimately, the PPHM SARS-CoV-2 assay using DMI demonstrated a 90.2% sensitivity for positive serum samples and a 99.0% specificity for negative serum samples (Fig. 3c andd), which showed DMI can improve simultaneously sensitivity and specificity beyond what is achievable with a single peptide probe (Fig. S3).
Additionally, we evaluated the sensitivity of the PPHM SARS-CoV-2 assay for each SARS-CoV-2 strain, finding sensitivities of 89.1%, 93.3%, and 85.2% in the Delta-infec ted, Omicron-infected, and New Omicron-infected groups, respectively (Fig. S4). This suggested that the diagnostic performance of the PPHM SARS-CoV-2 assay is consistent across different strains. For further differentiation of infecting strains, we could employ NAAT and whole-genome sequencing to identify the specific strain.
## PPHM SARS-CoV-2 can DIVH
The ability to achieve DIVH is crucial for using the PPHM to determine recent infec tions and serves as the foundation for subsequent strain differentiation using NAAT. Previous studies have shown that anti-peptide antibodies are TPIs with a duration of approximately 70 days after the occurrence of vaccination or infection, while anti-protein antibodies are PPIs (10,13). Therefore, by considering the timing of sampling, and the status of anti-peptide and anti-protein antibodies in a single serum sample, we can further determine whether the serum host has a past vaccination/infection (the occurrence of vaccination or infection events exceeding 70 days) or a recent infection (Fig. 4a). Note that in the screening of PPHM SARS-CoV-2 , RBD can also be used as one of the detection antigens, which allowed for the detection of both anti-peptide and anti-protein antibodies in a single serum sample and provided the foundation of DIVH. Next, we screened 314 serum samples collected during physical examinations in April 2023 to determine the DIVH ability of PPHM SARS-CoV-2 .
The screening results identified four classifications (Table 3): (i) DMI (-[negative]) RBD (-), indicating the absence of both anti-peptide and anti-protein antibodies, sugges ted that these individuals were unvaccinated and uninfected, comprising a total of 57 samples (18.2%) (Fig. 4b); (ii) DMI (+, positive) RBD (-), signifying the presence of anti-peptide antibodies and the absence of anti-protein antibodies; theoretically, this classification may only exist in cases of occurrence of vaccination/infection, where anti-peptide antibodies appeared earlier than anti-protein antibodies, with a total of only five samples (1.6%) (Fig. 4c); (iii) DMI (+) RBD (+), indicating the presence of both anti-peptide and anti-protein antibodies, suggested recent vaccination/infection, with a total of 78 samples (24.8%) (Fig. 4d); and (iv) DMI (-) RBD (+), where anti-pep tide antibodies are negative and anti-protein antibodies are positive, indicated past vaccination/infection, as the anti-peptide antibodies induced by past vaccination/infec tion have disappeared by the time of sampling, with a total of 174 samples (55.4%) (Fig. 4e).
The data indicated that employing both anti-peptide antibodies (DMI) and anti-RBD protein antibodies in a single serum can reveal the vaccination or infection status. Among the serum samples collected in April 2023, over half (55.4%, 174 samples) demonstrated past infections [DMI (-), RBD (+)], and 24.8% indicated recent infections [DMI (+), RBD (+)], which corresponded with China's situation following the easing of COVID-19 restrictions at the end of December 2022. Furthermore, 18.2% of the population was unvaccinated and uninfected [DMI (-), RBD (-)], with this group predominantly comprising elderly individuals, aligning with the demographic profile in China.
If only anti-peptide antibodies are assessed, we can only determine whether the DMI value is negative or positive. According to the previous research, a negative DMI does not necessarily indicate that the serum sample is from an uninfected or unvaccinated individual, as anti-peptide antibodies are short-lived and disappear over time (13). If only anti-RBD protein antibodies are assessed, it is not possible to DIVH (10). Therefore, simultaneous testing of anti-peptide and anti-RBD protein antibodies is beneficial for accurately determining the vaccination or infection status of a serum sample and is advantageous for epidemiological investigations of COVID-19.
## DISCUSSION
In this study, we demonstrated that PPHM SARS-CoV-2 can effectively (i) diagnose infections of SARS-CoV-2 and (ii) achieve DIVH, that is, differentiate recently infected from historically vaccinated/infected hosts. Despite the existence of various diagnostic methods for COVID-19, including NAAT and antigen detection, their effectiveness has often been suboptimal, especially concerning DIVH. Our findings suggested that serological antibody screening with PPHM SARS-CoV-2 plays a pivotal role in diagnosing and monitoring COVID-19, thereby contributing to the containment of the virus's spread.
During the COVID-19 pandemic, the incomplete understanding of antibodies generated by SARS-CoV-2 infections, coupled with prevalent issues such as NSI and NRI in traditional serological assays, has impeded the broad application of serological antibody testing for infection monitoring. Factors like heterophile antibodies (22) and high antibody concentrations (14,23) can lead to binding with different target anti gens, exacerbating NSI/NRI issues. Our PPHM approach recognizes and mitigates these influences by employing a DMI method to counteract their adverse effects.
Conceptually, proteins are regarded as compositions of multiple linear epitopes besides conformational epitopes, offering greater sensitivity than single peptides when utilized as probes in diagnostic assays. For instance, when diagnosing infections with PPHM SARS-CoV-2 , the sensitivity of the RBD protein was 94.3% (Fig. 3b), whereas each corresponding peptide (S46, S58, S63, and S71) exhibited significantly lower sensitivities: 10.8%, 71.1%, 40.7%, and 8.2%, respectively (Fig. 3a). After applying the DMI analysis, peptides derived from the RBD protein and N protein achieved a sensitivity of 90.2% and specificity of 99.0% for diagnosing SARS-CoV-2 infections (Fig. 3d), indicating that both sensitivity and specificity were improved simultaneously, surpassing the level achieva ble with a single peptide probe (Fig. S3). The observed higher sensitivity of the RBD protein may be attributed to the implementation of the extensive COVID-19 vaccination campaign in China in 2021, leading to a robust positive signal across populations.
Recall that proteins, being mixtures of linear epitopes, not only possess higher sensitivity but also facilitate the aggregation of signals (24). Recent assays for rapid COVID-19 detection, such as label-free fiber optic surface plasmon resonance (SPR)-based biosensors (25), localized SPR-based nano-plasmonic biosensors (26), and microfluidic-based point-of-care biosensors (27), still face challenges when using proteins as probes, which can lead to aggregated signals. Our previous research highlighted the existence of different longevity antibodies, namely short-lived anti-pep tide antibodies termed TPIs, and long-lived anti-protein antibodies termed PPIs, allowing us to develop a DIVH strategy: monitoring whether TPI reappears due to infection after disappearance, without the need of developing marker vaccines (with negative tags) (10,11). In this study, we verified the presence of both TPIs and PPIs, as well as the DIVH functionality of TPIs. In the detection of serum from individuals uninfected but vaccinated for more than 200 days, the anti-peptide antibodies showed a 100% negative result, indicating that (i) the anti-peptide antibodies were indeed TPIs and (ii) the negative results of anti-peptide antibodies indicated no recent infection. The RBD protein showed a positive rate of 93.3%, also indicating that (i) the anti-protein antibodies were mainly PPIs and can remain positive more than 200 days after vaccina tion, and (ii) the anti-protein antibodies remained consistently positive and therefore cannot indicate recent infection (Fig. 2; Fig. S1 andS2). In addition, we further catego rized an individual's immune status according to the occurrence and development of anti-peptide and anti-protein antibodies: (i) when DMI (-) RBD (-) occurs, it indicates In an ideal scenario, the detection of anti-protein antibodies should correlate with viral infection, ensuring positive responses to pathogenic protein probes. To minimize NSI/NRI due to molecular mimicry, it is essential to use corresponding pathogenic proteins and their derived peptides as probes. However, RNA viruses like SARS-CoV-2, known for their error-prone replication process, are prone to mutation, thus foster ing virus evolution through replication-associated changes (28). This implies that if diagnosis is required for every SARS-CoV-2 infection strain, at least six different detection kits would need to be developed. To alleviate the corresponding workload, we have developed a versatile PPHM SARS-CoV-2 assay, capable of diagnosing infections across different strains, based on (i) the high amino acid sequence similarity (up to 96.6%) among strains, (ii) the utilization of linear peptides as probes, and (iii) the ability of DMI to overcome NSI/NRI. The PPHM SARS-CoV-2 assay has demonstrated a sensitivity of 90.2% for diagnosing SARS-CoV-2 infections.
Thus, PPHM SARS-CoV-2 holds distinct advantages in diagnosing SARS-CoV-2, partic ularly for differentiating between past vaccination or infection and recent infection. While this study did not include head-to-head comparisons with commercial ELISA or lateral flow assays, we have discussed their conceptual differences in assay design and diagnostic capability. Future benchmarking studies are warranted. However, this study also has several limitations. First, this assay cannot directly identify the infecting strain, and confirmation of the strain type still requires NAAT after infection diagnosis. Second, due to the challenge of obtaining individual continuous serum samples, we were only able to derive estimated occurrence and development curves for anti-protein and anti-peptide antibodies. Moreover, we acknowledge that longitudinal monitoring of antibody dynamics would provide deeper insights into TPIs. Third, potential crossreactivity with antibodies against other human coronaviruses was not fully evaluated and warrants further investigation. Finally, the relatively limited sample size and lack of validation in diverse populations, such as elderly or immunocompromised individuals, may affect the generalizability of our findings.
With the development of new vaccines to prevent human diseases and the imple mentation of public health measures, there is a growing demand for the assessment of vaccine efficacy, disease diagnosis, and prognosis. It is essential to develop rapid and reliable assays to meet these assessment needs, such as those for Human Papillo mavirus in cervical cancer (29), and Epstein-Barr Virus in nasopharyngeal cancer (30). Our research on diagnosing the SARS-CoV-2 infections has positive implications for the early prevention and diagnosis of cancers caused by these viruses. Moreover, the assays developed for differentiating past vaccination or infection from recent infection present a promising approach for controlling and potentially eradicating the virus.
## References
1. Wang, Horby, Hayden et al. (2020) "A novel coronavirus outbreak of global health concern" *Lancet*
2. Huang, Wang, Li et al. (2020) "Clinical features of patients infected with 2019 novel coronavirus in Wuhan" *Lancet*
3. Li, Xiong, Deng et al. (2022) "The utility of SARS-CoV-2 nucleocapsid protein in laboratory diagnosis" *J Clin Lab Anal*
4. Yüce, Filiztekin, Özkaya (2021) "COVID-19 diagnosis -A review of current methods" *Biosens Bioelectron*
5. Rahbari, Moradi, Abdi (2021) "rRT-PCR for SARS-CoV-2: analytical considerations" *Clin Chim Acta*
6. Toptan, Eckermann, Pfeiffer et al. (2021) "Evaluation of a SARS-CoV-2 rapid antigen test: potential to help reduce community spread?" *J Clin Virol*
7. Okba, Widjaja, Li et al. (2020) "Serologic detection of Middle East respiratory syndrome coronavirus functional antibodies" *Emerg Infect Dis*
8. Haynes, Miao, Harcourt et al. (2007) "Recombinant protein-based assays for detection of antibodies to severe acute respiratory syndrome coronavirus spike and nucleocapsid proteins" *Clin Vaccine Immunol*
9. Xue, Xu, Liu et al. (2020) "Epitope-containing short peptides capture distinct IgG serodynamics that enable differentiating infected from vaccinated animals for liveattenuated vaccines" *J Virol*
10. Chen, Li, Yang et al. (2022) "Immunoglobin G sero-dynamics aided host specific linear epitope identification and differentiation of infected from vaccinated hosts" *J Virol*
11. Meeusen, Walker, Peters et al. (2007) "Current status of veterinary vaccines" *Clin Microbiol Rev*
12. Murr, Hoffmann, Grund et al. (2020) "A novel recombinant newcastle disease virus vectored DIVA vaccine against peste des petits ruminants in goats" *Vaccines (Basel)*
13. Xu, Chen, Li et al. (2024) "Altering the competitive environment of B cell epitopes significantly extends the duration of antibody production" *Int Immunol*
14. Pan, Yang, Wu et al. (2022) "Previously unrecognized nonreproducible antibody-probe interactions" *Anal Chem*
15. Xu, Chen, Li et al. (2023) "Removing negative impacts from inevitable nonreproducible and nonspecific antibody-probe interac tions in viral serology" *Anal Chem*
16. Palm, Henry (2019) "Remembrance of things past: long-term B cell memory after infection and vaccination" *Front Immunol*
17. Huang, Garcia-Carreras, Hitchings et al. (2020) "A systematic review of antibody mediated immunity to coronaviruses: kinetics, correlates of protection, and association with severity" *Nat Commun*
18. Weissleder, Lee, Ko et al. (1931) "COVID-19 diagnostics in context" *Sci Transl Med*
19. Tortorici, Veesler (2019) "Structural insights into coronavirus entry" *Adv Virus Res*
20. Liu, Leng, Huang et al. (2006) "Immunological characterizations of the nucleocapsid protein based SARS vaccine candidates" *Vaccine (Auckland)*
21. Jian, Wang, Yisimayi et al. (2025) "Evolving antibody response to SARS-CoV-2 antigenic shift from XBB to JN.1" *Nature*
22. Lippi, Aloe, Meschi et al. (2013) "Interference from heterophilic antibodies in troponin testing. Case report and systematic review of the literature" *Clin Chim Acta*
23. Huang, Ma, Liu et al. (2015) "Initiator integrated poly(dime thysiloxane)-based microarray as a tool for revealing the relationship between nonspecific interactions and irreproducibility" *Anal Chem*
24. Lu, Li, Teng et al. (2015) "Chimeric peptide constructs comprising linear B-cell epitopes: application to the serodiagnosis of infectious diseases" *Sci Rep*
25. Qu, Leirs, Maes et al. (2022) "Innovative FO-SPR label-free strategy for detecting anti-RBD antibodies in COVID-19 patient serum and whole blood" *ACS Sens*
26. Bhalla, Payam, Morelli et al. (2022) "Nanoplasmonic biosensor for rapid detection of multiple viral variants in human serum" *Sens Actuators B Chem*
27. Mao, Zhang, Yang (2020) "An integrated biosensor system with mobile health and wastewater-based epidemiology (iBMW) for COVID-19 pandemic" *Biosens Bioelectron*
28. Banerjee, Mossman, Grandvaux (2021) "Molecular determinants of SARS-CoV-2 variants" *Trends Microbiol*
29. Prue, Baker, Graham et al. (2018) "It is time for universal HPV vaccination" *Lancet*
30. Li, Li, Guo et al. (2023) "Anti-Epstein-Barr virus BNLF2b for mass screening for nasopharyngeal cancer" *N Engl J Med* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548465&blobtype=pdf | # Mutation in enterovirus 71 nonstructural protein 3A increases genome replication fidelity and exhibits attenuated virulence in mice
Fang Yu, Dongyan Xiong, Yao Zhong, Qiu-Yan Zhang, Cheng-Lin Deng, Zhe-Rui Zhang, Chang Wang, Yang Qiu, Han- Qing Ye, Peng Gong, Bo Zhang
## Abstract
Increasing numbers of studies have highlighted the important implications of the replication fidelity of RNA viruses for virus replication, pathogenesis, and the development of antiviral drugs and live-attenuated vaccines. However, most of the research has focused on viral polymerase, and little information is available about the potential role of other viral replicase proteins. Here, we demonstrated that the mutations of 3A V75A and 3D V63A+M393L in enterovirus 71 (EV71) conferred high-fidelity phenotypes through deep sequencing of virus populations in cell culture and animal models of infection. The 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L high-fidelity variants were highly attenuated in immunocompetent suckling mice. The RdRP enzymology data indicated that the V63A + M393L mutation of 3D pol increased the fidelity of RNA synthesis. The addition of the purified 3AB, either WT or V75A, as previously reported, was able to stimulate the RNA synthesis; interestingly, it effectively enhanced the replication fidelity of WT 3D pol . Moreover, the RNA chaperone activity assay showed that V75A 3AB had reduced helix unwinding activity compared with WT 3AB, implying the potential correlation between chaperone activity and fidelity regulation. Together, our results reveal a novel role of the 3A protein in fidelity, providing a basis for the development of antiviral inhibitors and live-attenuated EV71 vaccines.IMPORTANCE Numerous viral polymerases in various RNA viruses, such as EV71 3D pol , have been reported to be involved in the regulation of replication fidelity, while the role of other viral replicases in this process is poorly understood. In this study, we demon strate that the 3A V75A variant of EV71 confers increased fidelity and attenuated virulence, and the addition of 3AB, either WT or V75A mutant, can enhance the replication fidelity of WT 3D pol during RNA synthesis. Collectively, this work identifies EV71 nonstructural protein 3A as a previously unrecognized fidelity determinant. KEYWORDS replication fidelity, deep sequencing, RNA synthesis, virulence, helix unwinding activity R NA viruses have high mutation rates, ranging from 10 -4 to 10 -6 mutations per round of genome replication due to the lack of the same proofreading system as DNA viruses (1, 2). Such error-prone replication confers RNA viruses the capability of rapidly adapting to changes in environments, thereby influencing virus evolution and pathogenesis (3, 4). However, it still needs to be strictly controlled at an optimal fidelity level, as too high mutation rates may lead to error catastrophe, and too low mutation rates may fail to overcome selection pressure. Manipulating fidelity levels has also been developed as a novel strategy for the rational design of live-attenuated vaccines. Thus,
it is necessary to identify important fidelity determinants and investigate the precise mechanism underlying the fidelity regulation.
As a core element for virus replication, RNA-dependent RNA polymerase (RdRP) is the key regulator of replication fidelity, especially for the positive-strand RNA viruses with genomes < 20 kilobases (kb) lacking proofreading and other postreplicative repair mechanisms (5). Multiple RdRP-related fidelity variants have been identified in many viruses, including 3D G64S poliovirus (PV; a fidelity variant with a change of Gly-64 to Ser in the polymerase 3D pol ) (6), 3D G64R /3D G64T /3D L123F /3D S264L enterovirus 71 (EV71) (7,8), 3D M296I foot-and-mouth disease virus (FMDV) (9), 3D F232L coxsackievirus B3 (CV) (10), nsP4 C483Y chikungunya virus (CHIKV) (11), PB1 V43I influenza A virus (12), and nsP12 V553I coronaviruses (CoVs) (13). Besides the RdRP, a viral proofreading exonuclease (ExoN) in CoVs, a unique viral protein in large RNA viruses (i.e., > 20 kb), was first identified to be involved in fidelity regulation by excising misincorporated nucleotides or nucleotide analog inhibitors at the 3' end of the nascent RNA (14)(15)(16)(17). More recent work demonstra ted that the nonstructural protein 2 (nsP2) in CHIKV, a viral helicase/protease, could regulate replication fidelity in concert with the viral polymerase nsP4 (18). Together, the current findings not only emphasize the importance of RdRP for replication fidelity but also underscore the necessity of exploring other viral protein components that regulate fidelity.
EV71 is the causative agent of hand, foot, and mouth disease (HFMD) in infants and children and belongs to the genus Enterovirus within the family Picornaviridae (19). The genome contains a 7.5 kb positive-sense single-stranded RNA, encoding a large polyprotein that is processed into the following three primary precursors: one structural region P1 (VP4-VP1), and two nonstructural regions P2 and P3 (2A, 2B, 2C, 3A, 3B, 3C, and 3D) (20). The P3 region includes crucial components of the replication complex, in which a functional precursor protein 3CD can be further cleaved to form 3C (viral proteinase) and 3D pol (viral RdRP) (21). 3AB, another precursor protein encoded by the P3 region, is a small, membrane-binding protein containing a core hydrophobic region in 3A that inserts into the membrane, flanked by an N-terminal (the soluble moiety of 3A) and a C-terminal (the 3B peptide) regions. It can be further hydrolyzed into 3A and 3B (also known as VPg, a protein primer for RNA replication) (22,23). 3A/3AB have multiple functions in virus replication, such as anchoring the viral polymerase 3D pol to the membranous vesicles where replication occurs (24,25); stabilizing the complex between RNA template, primer, and 3D pol to stimulate RNA polymerase activity (26); and acting as a nucleic acid chaperone to facilitate the proper RNA folding (27,28). We previ ously selected two NITD008 (an adenosine analog developed as a drug candidate for dengue virus)-resistant EV71 isolates with mutations mapped to 3A (3A V75A ) and 3D pol (3D V63A+M393L ) (29). It was demonstrated that either 3A V75A or 3D V63A+M393L mutation alone could lead to resistance to NITD008, and the combination of 3A and 3D mutations enhanced this resistance further. Given that selection of nucleoside/nucleotide analogs (NAs)-resistant variants is a commonly used approach to discover fidelity determinants, we hypothesize that these mutations may represent new fidelity determinants; in particular, 3A protein may also play an important role in the regulation of viral replication fidelity through an unknown mechanism.
In this study, we quantitatively assessed the changes in fidelity of EV71 NITD008resistant variants in cell culture systems utilizing deep sequencing across the full-length viral genome. It was demonstrated that the 3A V75A variant manifested a high-fidelity phenotype besides the 3D V63A+M393L variant. The double mutant 3A V75A -3D V63A+M393L showed a further increase in fidelity. Consistently, deep sequencing analysis of virus populations in ICR suckling mice revealed that the high-fidelity variants produced less genetically diverse populations within target organs, including the mouse brain and muscle, compared with WT EV71, along with a significant reduction in virulence. To investigate the underlying mechanisms of 3A and 3D pol proteins regulating replication fidelity, we performed an in vitro RdRP assay and assessed the effect brought by purified 3AB protein. During RNA synthesis, the V63A + M393L 3D pol exhibited a higher fidelity than WT 3D pol , and 3AB protein, irrespective of the mutation, had the ability to increase the replication fidelity of WT 3D pol . Moreover, the RNA chaperone activity assay showed that V75A 3AB had a reduced helix unwinding activity relative to WT, which may contribute to the increased fidelity. Our results demonstrate that the 3A protein in EV71 plays an important role in fidelity regulation and adds a new member to the list of fidelity regulators.
## RESULTS
The EV71 nonstructural proteins 3A and 3D collaboratively regulate replica tion fidelity in cell culture Among the fidelity assessment approaches, deep sequencing of virus population samples is far more informative, as it allows millions or billions of reads to be generated in a single experiment, then the alignment reads showing the amount of minority variants across the whole genome, and the analysis of minority variant diversity can determine fidelity from different populations (13,14,18). Thus, to test the role of mutations within 3A and 3D proteins in fidelity, the viral supernatants of P1 WT and variants (3A V75A , 3D V63A+M393L , 3A V75A -3D V63A+M393L , as well as a low-fidelity variant control, 3D F232L [10]) harvested at 72 h post-infection (hpi) were subjected to RNA extraction, followed by deep sequencing. Here, the NGS data for minority variants above 1% frequency were mined to calculate the number of minority variants. Under this parameter setting, ViVan generated more conservative mutation profiles that more closely resembled previously reported values for wild-type and low-fidelity variants obtained by molecular clone sequencing (30)(31)(32). Meanwhile, all minority variants were aggregated to calculate the RMSD value, which can be used as a metric of viral popula tion diversity (13,18). In this analysis, the reduction in the number of minority variants or RMSD implies that the virus possesses a high-fidelity phenotype. Each sample achieved a mean depth of more than 100,000 and 100% genome sequence coverage (Fig. 1A).
The inherent fidelity of the variants was first examined. Analysis of the minority variants showed that the 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L viruses produced fewer mutations than WT virus, with 1.24-fold, 1.29-fold, and 1.39-fold increases in fidelity, respectively, whereas the control 3D F232L virus resulted in a 1.1-fold decrease in fidelity compared to WT virus (Fig. 1B). Consistently, except for the 3D F232L virus with a higher RMSD value, the RMSD values of 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L viruses were significantly lower relative to WT virus, suggesting that these mutations in 3A and/or 3D protein(s) can increase the fidelity of EV71 (Fig. 1C). In parallel, we also compared the fidelity of each variant and WT virus after treatment with NITD008. Both data from the minority variants and RMSD values came to the same conclusion with the fidelity order of WT < 3A V75A < 3D V63A+M393L < 3A V75A -3D V63A+M393L (Fig. 1B andC), further confirming the high-fidelity phenotypes of these three variants. In summary, 3A and 3D proteins collaboratively regulate replication fidelity in cell culture.
Besides, we compared the multi-step growth curves and one-step growth curves of WT, 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L variants in Vero cells, and no significant differences in viral production were observed between each variant and the WT virus (Fig. 1D andE), further excluding the possibility that increased fidelity in variants resulted from compromised viral replication.
## High-fidelity variants are attenuated in immunocompetent mice
Due to restricted replication or lower fitness, previous studies have shown that highfidelity RNA variants are attenuated in animals (3,6,33). To assess whether limiting population diversity incurs a fitness cost in cell culture, the adaptability of the WT virus and high-fidelity variants was compared. As expected, the high-fidelity variants did not compete effectively with the WT virus and decreased viral fitness (Fig. 2A andB).
To further evaluate whether these high-fidelity variants were attenuated, we performed infection assays in immunocompetent neonatal ICR mice. Two-day-old ICR mice were inoculated intraperitoneally (IP) in parallel with 10 7 PFU of WT, 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L viruses. Compared with the WT virus, the survival times of three high-fidelity variants were prolonged (Fig. 2C), and the clinical symptoms, such as hind limb paralysis, were weakened (data not shown), presenting attenuated phenotypes. We then determined the viral titers in various tissues, includ ing brain, muscle, liver, lung, spleen, and small intestine, of these variants at 4 days post-infection (4 dpi) (Fig. 2D). There were no significant differences in the viral loads in most of the tissues between these variants and WT viruses, except for a little reduction observed in the brain and muscle of 3D V63A+M393L and 3A V75A -3D V63A+M393L variants (one-way ANOVA; P < 0.01) and in the liver of the 3D V63A+M393L variant (one-way ANOVA; P < 0.05), respectively. To study the intrahost evolution of EV71 high-fidelity variants, viral populations were isolated from the major target tissues (brain and muscle) at 4 dpi and analyzed by deep sequencing to assess the population genetic diversity. In brain and muscle, the diversities of high-fidelity populations were reduced relative to that of WT virus as characterized by RMSD values, although there were no statistical differences in muscle between WT virus and 3A V75A /3D V63A+M393L variants (Fig. 2E). Later, single-nucleotide polymorphisms (SNPs) with a frequency of 0.5% or greater detected in mouse were identified to characterize tissue-specific patterns of diversity, which refers to the composition of SNPs found in brain or muscle tissue (Fig. 2F). In the brain, the number of SNPs above 1% from high-fidelity strains is less than that from WT, and this difference is even greater in the number of 0.5%-1% SNPs. In muscle, the high-fidelity variants also had a reduced number of 0.5%-1% and > 1% SNPs compared with WT (Fig. 2F), and there are some non-shared SNPs different from the brain (Table 1; Fig. S1) and Vero cells (Table S1; Fig. S1). In the composition of SNPs across the genome, the diversity patterns of WT populations isolated from brain and muscle tissues were similar, consistent with those of the high-fidelity strains (Fig. 2F). These results indicated that the high-fidelity variants produced less genetically diverse populations in the mouse brain and muscle, consistent with the results of cell culture. Taken together, these results indicated that high-fidelity variants of EV71 were attenuated in the host and exhibited limited tissue-specific patterns of diversity and fitness.
## In vitro polymerase assay data indicate that the V63A and M393L mutations in 3D pol increase the fidelity of RNA synthesis
To further investigate how the V63A and M393L mutations affect fidelity, an in vitro polymerase assay was performed to evaluate the kinetics of incorporation for cognate and non-cognate nucleotides at relatively saturating nucleotide triphosphate (NTP) substrate concentrations. The T33/P10 RNA construct containing a 33-mer template strand and a 10-mer primer strand was used in a primer-dependent polymerase assay as previously described (34). Briefly, in vitro reactions were performed using equal amounts of each purified 3D pol . When only GTP and ATP are provided, the primer strand is expected to elongate by four nucleotides (nt) to produce a 14-mer product (Fig. 3A). For both EV71 WT 3D pol and V63A + M393L 3D pol , similar RNA elongation profiles of 14-mer products at each time point were shown, and the observed rate constant of correct nucleotide incorporation (k obs ), estimated by the crude accumulation rate of 14-mer under relatively saturating NTP concentrations (300 µM), showed no obvious difference (Fig. 3B and C, 0.080 min -1 vs. 0.075 min -1 ). In contrast, the k obs of F232L 3D pol was larger than those of WT 3D pol (0.140 min -1 vs. 0.080 min -1 ). These data indicate that the V63A + M393L mutation in 3D pol does not have an impact on RNA synthesis when using cognate NTPs as substrates.
To obtain a more quantitative perspective of changes in fidelity, the fidelity values (see legend of Fig. 3) of WT, V63A + M393L, and F232L 3D pol were evaluated. It is known that the substitution of 2′-OH group of NTP ribose with hydrogen (e.g., 2′-dATP and 2′-dCTP) impairs the interaction between NTP and viral RdRP, resulting in a reduction in the efficiency of nucleotide incorporation (35). We first used 2′-dATP to mimic a non-cognate nucleotide and calculated the fidelity value as the correct k obs divided by the incorrect k obs (Fig. 3D andE). The V63A + M393L 3D pol was on the order of 1.25-fold and 1.1-fold more faithful than WT 3D pol at 100 µΜ and 300 µΜ concentrations of 2′-dATP, respectively (Fig. 3F). Under the same conditions, the fidelity of the control F232L 3D pol was reduced by 2.15-fold and 1.56-fold, respectively, consistent with the previous study (10). Furthermore, the fidelity value was evaluated by monitoring the utilization of the NTP form of NITD008 (NITD008-TP; a chain terminator). By adding GTP and NITD008-TP to initiate the reaction, the incorporation of NITD008 monophosphate (NITD008-MP) can be monitored over time by the accumulation of 12-mer product (Fig. 3G). As observed in Fig. 3H andI, V63A + M393L 3D pol consistently incorporated less NITD008-TP over time, with a 2.36-fold higher fidelity value than WT (Fig. 3I), whereas the fidelity value of F232L 3D pol was 1.05-fold decrease relative to WT 3D pol , confirming the increased fidelity of V63A + M393L 3D pol .
Additionally, GTP-only incorporation was tested, and similar accumulation rates of 11-mer products were observed between WT and V63A + M393L 3D pol (0.055 min -1 vs. 0.053 min -1 ), indicating the first-step GTP incorporation had little effect on the incorpora tion rate of the subsequent NTPs or mismatching/non-cognate NTPs, including NITD008-TP in multi-step reactions (Fig. 4A andB). To further accurately determine the differences of NITD008-TP discrimination between WT and V63A + M393L 3D pol , a single-nucleotide incorporation assay was performed. As described in Materials and Methods, driven by the polymerase, a "G" is incorporated into the RNA template/primer duplex to form elongation complex (EC, with an 11-mer product), which is then purified and used to determine the incorporation rate of NITD008-TP by monitoring the generation of 12-mer products from U: NITD008-TP mismatch (Fig. 4C) (36,37). As shown in Fig. 4D andE, V63A + M393L EC yielded a 12-mer product at a slower rate than WT EC. Collectively, these data suggest that the V63A + M393L mutation increased the ability in mismatch discrimination of 3D pol during RNA synthesis.
## 3AB protein regulates the fidelity of RNA synthesis catalyzed by WT 3D pol , besides stimulating the activity of polymerase
In order to explore the regulation mechanism of 3AB protein on fidelity, we first studied the effects of 3AB on 3D pol -catalyzed U: NITD008-TP mismatch using an in vitro poly merase assay. It has been demonstrated that the 3A and 3B regions of 3AB have a synergistic effect to stimulate 3D pol -catalyzed RNA synthesis in vitro (38); hence, the 3AB protein was used in the following assays. To improve the solubility of 3AB protein, the maltose-binding protein (MBP) tag was added at the N-terminus of 3AB (MBP-3AB) to facilitate its folding. Meanwhile, the control MBP-only protein was also expressed and purified. The RNA products of different combinations between WT and mutant in 3AB and 3D pol proteins were measured according to the reaction flow chart (Fig. 5A). As previously reported (26,39), compared with 3D pol protein alone, the addition of 3AB protein significantly stimulated the yield of 14-mer RNA products to varying degrees in four combinations (Fig. 5B, lanes 2 to 7; P < 0.05, two-way ANOVA). In order to further quantify the stimulatory effect of 3AB on the RNA synthesis by 3D pol , the k obs values of 3D pol with and without 3AB in the correct nucleotide incorporation assay were analyzed. Both WT and V75A 3AB could stimulate the polymerase activity of WT and V63A + M393L 3D pol (P < 0.05 for all, Unpaired t test; Fig. 5C), and no significant difference was observed between WT and V75A 3AB, indicating that the V75A mutation did not affect the activity of 3AB to stimulate 3D pol -catalyzed RNA synthesis.
a The asterisk represents the stop codon. syn, synonymous mutations. Amino Acid Sub, Amino acid substitu tion. WT, wild-type strain. 3A, 3A V75A strain. 3D, 3D V63A+M393L strain. 3A-3D, 3A V75A -3D V63A+M393L strain. The color-coding scheme is as follows: diamonds, found in both tissues in mice; triangles, found in the brain only in mice; squares, found in the muscle only in mice; circles, found in vitro in Vero cells.
To further investigate whether the addition of 3AB protein would regulate the fidelity of 3D pol , the ATP substrate was replaced with NITD008-TP in the same reaction (Fig. 5D). As shown in Fig. 5E, the yields of 12-mer RNA product catalyzed by WT/mutant 3D pol were significantly accelerated at different times when adding either WT or V75A 3AB protein (lanes 2 to 7; P < 0.05, two-way ANOVA), although there were no significant differences for the k obs between with and without WT 3AB in the context of WT 3D pol (Fig. 5F). By comparing the fidelity values for each group, it was shown that the addition of WT and V75A 3AB greatly increased the fidelity of WT 3D pol (1.320/1.070 vs. 1.880/1.070), but had little effect on the fidelity of V63A + M393L 3D pol . Besides, the control MBP protein had no effect on the activity of polymerase (Fig. 5F). At the same time, we also followed the same strategy to study the effects of 3AB protein on the incorporation of NITD008-TP in the 11-mer EC reaction system by pre-incubating EC with 3AB (Fig. S2A), but no obvious stimulatory effect was observed (Fig. S2B andC), which confirmed the previous report that 3AB does not stimulate 3D activity on a template that is stably base paired to a primer (39). These results indicate that both WT and V75A 3AB are able to increase the fidelity of RNA synthesis catalyzed by WT 3D pol . The V75A mutation reduces the nucleic acid helix-destabilizing activity of 3AB protein
It has been reported that 3AB functions as a nucleic acid chaperone with helix-destabi lizing activity (27,28). To address whether the high-fidelity phenotype of the 3A V75A variant is correlated with the chaperone activity of the 3AB protein, an in vitro nucleic acid helix-unwinding assay was performed to investigate the effect of the V75A mutation on the RNA chaperone activity of 3AB. Different concentrations of purified 3AB proteins were added to the nucleic acid helix unwinding reaction consisting of hexachloro fluorescein (HEX)-labeled RNA substrates to initiate the reaction, and the yield of unwinding double-stranded (ds) RNA was determined after 1 h by native-PAGE gel electrophoresis followed by Typhoon imager scanning. In comparison with WT 3AB, V75A 3AB exhibited a slower helix unwinding activity, especially at a low concentration (1 µΜ) (Fig. 6A andB). Subsequently, the kinetics of dsRNA unwinding in WT and V75A 3AB-initiated reactions were measured at a concentration of 1 µΜ by calculating the percentage of unwound RNA at each time point (Fig. 6C). Compared with WT 3AB, the V75A 3AB displayed a slower helix unwinding activity (0.450 h -1 vs. 0.060 h -1 , a single exponential equation; Fig. 6D), implying a potential relationship between decreased RNA chaperone activity and high-fidelity regulation of 3AB mutant.
$$5'UTR C373G ♢ ♢ 5'UTR G385T ♢ ♢ ♢ ♢ □ ♢ ♢ 5'UTR A388C ♢ ♢ ♢ □ ♢ □ VP4 A37P ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 2A G122G (syn) ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 2A L131V (syn) ♢ ♢ △ ♢ □ ♢ 2A G133G (syn) ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 2A A135G ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 2A V137V (syn) ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 2C Q52H ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ ○ ○ ○ 2C R100G □ 3C G164G (syn) ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 3D A78G ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 3D Y120* ♢ ♢ ♢ △ ♢ ♢ ♢ 3D S121C ♢ ♢ ♢ △ ♢ ♢ ♢ 3D A122D ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 3D L123* ♢ ♢ ♢ △ ♢ ♢ ♢ 3D I125V ♢ ♢ ♢ ♢ ♢ ♢ ♢ ♢ 3D K126N ♢ □ □ ♢ 3D K127* □ □ □ 3D K127I □ 3D E366G □ □$$
## DISCUSSION
Strong evidence from cell culture and animal assays indicates that 3A, in concert with 3D pol , is an important fidelity regulator
In the presence of nucleotide analogs, the factors affecting replication fidelity can often be identified by passaging the virus population to screen for specific amino acid substitutions in resistant variants. Using the same passaging strategy, we previously selected NITD008-resistant variants of EV71 with 3A and 3D pol mutations (29). It thus prompted us to further investigate whether these mutations are involved in fidelity regulation. As demonstrated by deep sequencing analysis, all these mutations, including 3A V75A , 3D V63A+M393L and 3A V75A -3D V63A+M393L , resulted in high replication fidelity, featured with (i) 1.24-fold, 1.29-fold, and 1.39-fold lower mutation frequency in the 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L populations of cell culture (Fig. 1);
(ii) a significant reduction in the sequence diversities and tissue-specific patterns of diversity in brain and muscle tissues relative to WT virus (Fig. 2). Consistent with previous reports that alterations of fidelity always result in inefficient adaptation and virulence attenuation (3,32,33), the high-fidelity 3A V75A , 3D V63A+M393L , and 3A V75A -3D V63A+M393L variants exhibited reduced fitness and significant attenuation in immunocompetent suckling mice. These data provide strong evidence that the EV71 nonstructural proteins 3A and 3D pol collaboratively regulate replication fidelity, highlighting their potency as a vaccine candidate.
In terms of the potential mechanism underlying the observed attenuation of such mutants, as we know, there are currently the following opinions, most of them based on the studies on the well-characterized high-fidelity PV 3D G64S variant exhibiting decreased virulence: (i) the PV 3D G64S variant reduces viral fitness under a defined selective pressure, making it likely that the reduced spread in mouse tissue could be caused by the increased fidelity of the viral polymerase (3); (ii) it is also considered that the virulence of this variant is determined by the interplay between different variants within the quasispecies that may facilitate expansion and replication in mice (6); and (iii) the replicative speed is more decisive for viral virulence than genetic diversity, as demonstrated by the fact that the compensation for the growth defect of PV 3D G64S restored virulence, whereas compensation of the fidelity phenotype did not (40). In our study, different from the growth defect caused by the G64S mutation in PV 3D pol , the mutations in either EV71 3A or 3D pol protein had little effect on virus proliferation in cells (Fig. 1D andE; no significant differences relative to WT). It thus implied that the observed attenuation by these 3A and 3D pol variants in our study may not, at least not mainly, be attributed to the differences in viral replicative speed between them and the WT virus.
## Underlying mechanism of 3A and 3D pol proteins regulating fidelity
Over the past decades, a considerable amount of work has uncovered the molecular determinants of replication fidelity in RdRPs. Unexpectedly, the variation/mutation sites are widely distributed in the RdRP core. It is thus believed that there may exist a more complicated mechanism (41). Here, both V63A and M393L polymerase mutations in EV71 fall into the near sites known to regulate RNA synthesis. At the C-terminus of V63, the residue G64 plays a critical role in the hydrogen bonding network involving the N-terminus and polymerase motif A; its mutation G64S has been documented to be a known high-fidelity mutation of picornaviruses and increases resistance to ribavirin. Additionally, residue M393 is adjacent to the polymerase motif E and forms hydrophobic interactions with motif E residues K376 and R377, which in turn interact with the -1 to -3 backbone region of the product RNA (29). In vitro polymerase assay data showed that the V63A + M393L mutation increased nucleotide discrimination, reducing the rate of RNA elongation in both processive elongation and single-nucleotide assays when non-cognate NTP (2'-dATP/NITD008-TP) was provided. Recently, an analogous site of M393V in the polymerase of PV was also reported to confer the high-fidelity phenotype (42), further confirming the important role of this site in fidelity regulation. In contrast, there is currently no report regarding the 3A/3AB protein in fidelity regulation. Our in vitro polymerase assay data demonstrated that 3AB protein, regardless of being WT or V75A mutant, resulted in distinct effects on the fidelity during RNA synthesis in the context of WT and V63A + M393L 3D pol proteins (increase vs. no effect), despite both polymerases exhibiting consistent stimulation by 3AB (Fig. 5). Many studies on the stimulation of 3D pol -catalyzed RNA synthesis by 3AB have been conducted on PV. It is thought that 3AB stimulates RNA synthesis through promoting the utilization of 3′-hydroxyl termini as sites for chain elongation by 3D pol (39). During this process, it may require 3AB to interact with and stabilize these sites and/or may recruit 3D pol to the site (39,43). It has been indicated that many of the binding sites with either RNA or 3D pol are within 3B sequences, although there may exist a synergistic effect between 3A and 3B regions on the 3AB functions (44,45), and the corresponding contact surface on 3D pol lies in a hydrophobic patch near conserved motif E (24,45). Thus, it is plausible that the V75A mutation in the 3A sequence had little effect on the stimulation activity of 3AB. In terms of how 3AB, irrespective of the mutation, regulates the fidelity of WT 3D pol -catalyzed RNA synthesis, it remains unknown. On the other hand, more work is also needed to determine whether the inability of 3AB to influence the fidelity of RNA synthesis by V63A + M393L 3D pol is attributed to the impairment of 3AB-3D pol interaction caused by the M393L mutation, which resides at the contact surface on 3D pol .
It has been reported that 3AB can function as a nucleic acid chaperone with helixdestabilizing activity to facilitate RNA proper folding, although the mechanism is still unclear (28). An in vitro helix-unwinding assay was then performed to investigate the effect of the V75A mutation on the RNA chaperone activity of 3AB. Additionally, due to the absence of 3D pol protein, this assay can, in a way, dissect the role of 3A/3AB in fidelity regulation even further, providing a straightforward explanation for the high-fidelity phenotype of the 3A V75A variant. The results showed that despite the V75A mutation being located in the membrane binding region of 3A protein, beyond the reported chaperone functional region (the last 7 C-terminal amino acids of 3A plus the full 3B protein [28]), it was able to cause a reduction in helix-unwinding rate, especially at a low concentration (Fig. 6). In comparison with alanine, valine is a larger hydrophobic residue and thereby has more hydrophobic interaction with membranes. It is often found in the tightly packed hydrophobic core of membrane proteins (46,47). Analysis of EV71 3A/3AB topology (48,49) revealed that it is the case for 3A/3AB protein, as there are quite a few Val residues residing in the predicted membrane binding region, implying the essential function of Val residues (Fig. S3). Although it is still unclear regarding the interplay between membrane binding region and chaperone functional region, it seems to be a common event in many RNA viruses that RNA remodeling proteins, like RNA chaperone or helicase, play an important role in fidelity regulation, as demonstra ted that the high-fidelity G641D nsP2 of CHIKV viral helicase-protease protein) also exhibited a delayed helicase activity (18). Based on such common features shared by EV71 3AB and CHIKV nsP2, the reduction in helix-unwinding activity of V75A 3AB protein may contribute to the enhanced fidelity of 3A V75A /3A V75A -3D V63A+M393L variants. The detailed mechanism behind the slowed helix-unwinding activity of 3AB-induced high-fidelity phenotype remains to be explored in future work. Together, our findings have identified that 3A regulates the replication fidelity of EV71. This supports a model in which EV71 uses multiple nonstructural proteins to replicate RNA genomes faithfully. Besides, our data suggest that the 3AB protein with RNA helix-unwinding activity and polymerase may have dynamic interactions to coordinate the replication fidelity of EV71.
## MATERIALS AND METHODS
## Viruses, cells, and compound
Vero cells were cultured in Dulbecco modified Eagle medium (DMEM; Invitrogen) with 10% fetal bovine serum (FBS), 100 U/mL of penicillin, and 100 µg/mL of streptomycin. The WT viruses and variants (designated as 3A V75A , 3D V63A+M393L , 3A V75A -3D V63A+M393L , and 3D F232L , respectively) were prepared from the infectious cDNA clone of EV71 (29,50), stored as aliquots at 80°C. The supernatants of transfected cells were harvested at 96 h post-transfection, designated as passage 0 (P0) of the rescued viruses. The P0 was passaged to the second generation (P1), followed by viral titration on Vero cells, and the viruses were stored at 80°C for later use. All viruses (P1) were subject to nucleotide sequencing. NITD008 was synthesized as previously reported (51).
## Deep-sequencing sample preparation and analysis
For multi-step growth curves, Vero cells were infected at MOI = 0.1 and, at different time points, quantified progeny virus by plaque assay. After the P0 generation infected the cells, the viral supernatant of the P1 generation was harvested at 72 h post-infection. The viral RNA was extracted using the QIAamp Viral RNA Mini Kit (Qiagen, Hilden, Germany) and quantified by quantitative reverse transcription-PCR (qRT-PCR) using VP1 primers (Table 2). Then, the RNA was converted to double-stranded DNA prior to generating a compatible sequencing library using the Nextera XT DNA Library Preparation kit. NGS was performed on the DNBSEQ T7 sequencing platform (MGI, Shenzhen, China).
Sequencing runs were analyzed using the ViVan bioinformatics pipeline. SNP detection was performed with default parameters, except that the minimum coverage was set to 3,000, minimum variant frequency was set to 1.0%, and the ploidy was set to 1 (13,52,53). For each position throughout the viral genome, base composition and frequency were counted to calculate the number of variants for the whole genome, each protein, and (non-)synonymous mutation.
To calculate RMSD for each nucleotide position, the RMSD formula (54) was modified to compare the distribution of nucleotides between each sample by PHDtools (55), which were previously utilized for analyses of population diversity (56)(57)(58)
. P[Xi,p] and P[Yi,p] are the probabilities of nucleotide p, at position i, for the samples X and Y, respectively. The nucleotide, p, is an element of the nucleotide set [A, T, G, C, -, I], where the -and I characters represent deletion and insertion. The denominator within the square root operator, 6, is the number of symbols in the nucleotide set used.
$$: RMSD (X, Y) i = ∑ p = {A, T, G, C, -, I} P X i, p -P Y i, p2 6$$
## Construction of recombinant baculoviruses
The cDNAs for EV71 3AB from the plasmid containing full-length EV71 cDNA (29,50), and MBP fragments were amplified by polymerase chain reaction (PCR), followed by cloning into the vector pFastBac. The desired plasmids were subjected to the Bac-to-Bac baculovirus system to express the recombinant proteins with an MBP fused at the N-terminus as previously described (28).
## Protein expression and purification
The pET26b-Ub vector-based plasmid containing the EV71 3D pol (RdRP) gene (50) was used as the original cloning template to construct the mutant plasmids according to previously described methods, designated as the EV71 3D pol WT and V63A + M393L 3D pol , respectively (59). Cell growth, isopropyl-β-D-thiogalactopyranoside (IPTG) induction, cell harvesting, cell lysis, protein purification, and protein storage were performed as previously described (60). The expression and purification of MBP alone and MBP-3AB proteins were performed as previously described (61). Briefly, Spodoptera frugiperda insect cells (Sf9) were infected with the recombinant baculoviruses and harvested at 72 h postinfection. Cell pellets were resuspended, lysed by sonication, and subjected to centrifugation for 30 min at 11,000 × g to remove debris. The protein in the supernatant was purified using amylase affinity chromatography (New England BioLabs, Ipswich, MA) according to the protocol and then further purified by Superdex 75 column in the AKTA system.
## Primer-dependent polymerase assay
For the primer-dependent polymerase assay of EV71 3D pol , a 50 μL-reaction mixture containing 6 µM 3D pol or its variant, 4 µM RNA construct (T33/P10) (34,62), a certain NTP/NITD008-TP substrate combination in buffer (50 mM Tris-HCl [pH 7.0], 75 mM KCl,
## Elongation complex (EC) purification and the EC extension assay
EC assembly and purification were performed as previously described (36). Briefly, the EC was assembled using the T33/P10 and EV71 RdRP upon incorporation of a G nucleotide with GTP as the only NTP substrate, producing an 11-mer (P11)-containing EC (EC11) and purified using a Capto HiRes Q column (GE Healthcare). The final buffer condition for EC storage was 20 mM HEPES [pH 7.0], 100 mM NaCl, 2 mM MgCl 2 , and 4 mM TCEP.
The EC elongation experiments were performed in a 50 µL reaction mixture containing 4 µM purified EC, 300 µM NITD008-TP at 10°C. Aliquots (8 µL) were withdrawn and quenched with an equal volume of stop solution at the different time points. RNA bands were visualized using Stains-All (Sigma-Aldrich) staining and quantified by ImageJ software.
## Nucleic acid helix unwinding assay
The standard helix destabilizing assay was performed as previously described (61). RNA helix substrates were prepared by annealing two complementary nucleic acid strands, RNA1 and RNA2 (Table 2). Briefly, different concentrations of proteins and 0.1 pmol of helix substrate were added to a mixture containing a final concentration of 25 mM HEPES-KOH (pH 8.0), 50 mM NaCl, 1 mM MgCl 2 , 5 U RNasin (Promega), and 10 mM ATP and incubated at 37°C for different times. Mixtures were electrophoresed on 12% native-PAGE gels. Gels were scanned with a Typhoon 9200 imager (GE Healthcare). The ratio of released single strands versus the total substrates was quantified with ImageJ software.
## Animal experiment
The progenies of the pregnant ICR mice were assigned randomly to five groups, and each group had 10 newborn mice. Two-day-old ICR newborn mice (n = 10) were inoculated intraperitoneally with viruses/virus-free cell culture supernatant at a concentration of 10 7 PFU/ mouse. The suckling mice were monitored daily for body weight, clinical symptoms, and mortality for 20 days. To minimize animal suffering, the mice were euthanized if they were quadriplegic. Mice from each treatment group were euthanized on 4 dpi, and the brain, muscle, liver, lung, spleen, and small intestine were collected. Samples were homogenized by using a TissueLyser LT homogenizer (Qiagen) in DMEM. The virus titers in the supernatants of clarified homogenates (3,000 × g for 10 min at 4°C) were determined by plaque assay. The statistical analyses of virus titers were performed using one-way ANOVA, and P values of 0.05 were considered significant.
For NGS samples from brain and muscle tissues, target enrichment on equivalent quantities of viral RNA was performed by high-fidelity RT-PCR (Accuscript PfuUltra II) of four cDNA amplicons spanning the EV71 5′ to 3′ UTRs (Table 2). The follow-up procedure was the same as above.
## Direct competition fitness assay
For direct competition fitness assays, each variant was mixed with the WT virus at a ratio of 1:1 to infect Vero cells in triplicate wells at an MOI of 0.1 over three passages. Viral RNA was extracted, and the region flanking 3A amino acid 75 and 3D amino acid 63 or 393 was amplified by RT-PCR for Sanger sequencing. The abundance of each compet itor was measured as the height of the nucleotide encoding either the WT (3A V75A -GTG; 3D V63A -GTG; 3D M393L -ATG) or variants (3A V75A -GCG; 3D V63A -GCG; 3D M393L -TTG) in sequencing chromatograms and quantified by ImageJ software. The primers used in the direct competition assays are listed in Table 2.
## References
1. Sanjuán, Nebot, Chirico et al. (2010) "Viral mutation rates" *J Virol*
2. Jenkins, Rambaut, Pybus et al. (2002) "Rates of molecular evolution in RNA viruses: a quantitative phylogenetic analysis" *J Mol Evol*
3. Pfeiffer, Kirkegaard (2005) "Increased fidelity reduces poliovirus fitness and virulence under selective pressure in mice" *PLoS Pathog*
4. Vignuzzi, Wendt, Andino (2008) "Engineering attenuated virus vaccines by controlling replication fidelity" *Nat Med*
5. Steinhauer, Domingo, Holland (1992) "Lack of evidence for proofreading mechanisms associated with an RNA virus polymerase" *Gene*
6. Vignuzzi, Stone, Arnold et al. (2006) "Quasispe cies diversity determines pathogenesis through cooperative interactions in a viral population" *Nature*
7. Sadeghipour, Mcminn (2013) "A study of the virulence in mice of high copying fidelity variants of human enterovirus 71" *Virus Res*
8. Meng, Kwang (2014) "Attenuation of human enterovirus 71 highreplication-fidelity variants in AG129 mice" *J Virol*
9. Arias, Arnold, Sierra et al. (2008) "Determinants of RNA-dependent RNA polymerase (in)fidelity revealed by kinetic analysis of the polymerase encoded by a foot-andmouth disease virus mutant with reduced sensitivity to ribavirin" *J Virol*
10. Campagnola, Mcdonald, Beaucourt et al. (2015) "Structure-function relationships underlying the replication fidelity of viral RNA-dependent RNA polymerases" *J Virol*
11. Rozen-Gagnon, Stapleford, Mongelli et al. (2014) "Alphavirus mutator variants present host-specific defects and attenuation in mammalian and insect models" *PLoS Pathog*
12. Cheung, Watson, Choy et al. (2014) "Generation and characterization of influenza A viruses with altered polymerase fidelity" *Nat Commun*
13. Sexton, Smith, Blanc et al. (2016) "Homology-based identification of a mutation in the coronavirus RNA-dependent RNA polymerase that confers resistance to multiple mutagens" *J Virol*
14. Eckerle, Becker, Halpin et al. (2010) "Infidelity of SARS-CoV Nsp14-exonuclease mutant virus replication is revealed by complete genome sequencing" *PLoS Pathog*
15. Graham, Becker, Eckerle et al. (2012) "A live, impaired-fidelity coronavirus vaccine protects in an aged, immunocompromised mouse model of lethal disease" *Nat Med*
16. Denison, Graham, Donaldson et al. (2011) "Coronaviruses: an RNA proofreading machine regulates replication fidelity and diversity" *RNA Biol*
17. Smith, Case, Blanc et al. (2015) "Mutations in coronavirus nonstructural protein 10 decrease virus replication fidelity" *J Virol*
18. Stapleford, Rozen-Gagnon, Das et al. (2015) "Viral polymerase-helicase complexes regulate replication fidelity to overcome intracellular nucleotide depletion" *J Virol*
19. Wang, Zhu, Zhao et al. (2012) "Characteri zation of full-length enterovirus 71 strains from severe and mild disease patients in northeastern China" *PLoS One*
20. Mcminn (2002) "An overview of the evolution of enterovirus 71 and its clinical and public health significance" *FEMS Microbiol Rev*
21. Toyoda, Nicklin, Murray et al. (1986) "A second virus-encoded proteinase involved in proteolytic processing of poliovirus polyprotein" *Cell*
22. Suhy, Giddings, Kirkegaard (2000) "Remodeling the endoplasmic reticulum by poliovirus infection and by individual viral proteins: an autophagy-like origin for virus-induced vesicles" *J Virol*
23. Paul, Van Boom, Filippov et al. (1998) "Protein-primed RNA synthesis by purified poliovirus RNA polymerase" *Nature*
24. Lyle, Clewell, Richmond et al. (2002) "Similar structural basis for membrane localization and protein priming by an RNA-dependent RNA polymerase" *J Biol Chem*
25. Xiang, Cuconati, Hope et al. (1998) "Complete protein linkage map of poliovirus P3 proteins: interaction of polymerase Full-Length Text Journal of Virology October"
26. "Dpol with VPg and with genetic variants of 3AB" *J Virol*
27. Rodriguez-Wells, Plotch, Destefano (2001) "Primer-dependent synthesis by poliovirus RNA-dependent RNA polymerase (3D pol )" *Nucleic Acids Res*
28. Destefano, Titilope (2006) "Poliovirus protein 3AB displays nucleic acid chaperone and helix-destabilizing activities" *J Virol*
29. Tang, Xia, Wang et al. (2014) "The identification and characterization of nucleic acid chaperone activity of human enterovirus 71 nonstructural protein 3AB" *Virology (Auckl)*
30. Deng, Yeo, Ye et al. (2014) "Inhibition of enterovirus 71 by adenosine analog NITD008" *J Virol*
31. Isakov, Bordería, Golan et al. (2015) "Deep sequencing analysis of viral infection and evolution allows rapid and detailed characterization of viral mutant spectrum" *Bioinformatics*
32. Gnädig, Beaucourt, Campagnola et al. (2012) "Coxsackievirus B3 mutator strains are attenuated in vivo" *Proc Natl Acad Sci U S A*
33. Levi, Gnädig, Beaucourt et al. (2010) "Fidelity variants of RNA dependent RNA polymerases uncover an indirect, mutagenic activity of amiloride compounds" *PLoS Pathog*
34. Coffey, Beeharry, Bordería et al. (2011) "Arbovirus high fidelity variant loses fitness in mosquitoes and mice" *Proc Natl Acad Sci*
35. Wu, Lu, Zhang et al. (2015) "Perturbation in the conserved methyltransferase-polymerase interface of flavivirus NS5 differentially affects polymerase initiation and elongation" *J Virol*
36. Gohara, Arnold, Cameron (2004) "Poliovirus RNA-dependent RNA polymerase (3D pol ): kinetic, thermodynamic, and structural analysis of ribonucleotide selection" *Biochemistry*
37. Shu, Gong (2016) "Structural basis of viral RNA-dependent RNA polymerase catalysis and translocation" *Proc Natl Acad Sci*
38. Gong, Kortus, Nix et al. (2013) "Structures of coxsackievirus, rhinovirus, and poliovirus polymerase elongation complexes solved by engineering RNA mediated crystal contacts" *PLoS One*
39. Plotch, Palant (1995) "Poliovirus protein 3AB forms a complex with and stimulates the activity of the viral RNA polymerase, 3Dpol" *J Virol*
40. Richards, Ehrenfeld (1998) "Effects of poliovirus 3AB protein on 3D polymerase-catalyzed reaction" *J Biol Chem*
41. Fitzsimmons, Woods, Mccrone et al. (2018) "A speed-fidelity trade-off determines the mutation rate and virulence of an RNA virus" *PLoS Biol*
42. Liu, Shi, Gong (2018) "A unique intra-molecular fidelity-modulating mechanism identified in a viral RNA-dependent RNA polymerase" *Nucleic Acids Res*
43. Kempf, Watkins, Peersen et al. (2020) "An extended primer grip of picornavirus polymerase facilitates sexual RNA replication mechanisms" *J Virol*
44. Lama, Sanz, Rodríguez (1995) "A role for 3AB protein in poliovirus genome replication" *J Biol Chem*
45. Xiang, Cuconati, Paul et al. (1995) "Molecular dissection of the multifunctional poliovirus RNA-binding protein 3AB" *RNA*
46. Choe, Kirkegaard (2004) "Intracellular topology and epitope shielding of poliovirus 3A protein" *J Virol*
48. Pechmann, Frydman (2014) "Interplay between chaperones and protein disorder promotes the evolution of protein networks" *PLoS Comput Biol*
49. Mughram, Catalano, Herrington et al. (2023) "3D interaction homology: the hydrophobic residues alanine, isoleucine, leucine, proline and valine play different structural roles in soluble and membrane proteins" *Front Mol Biosci*
50. Horova, Lyoo, Różycki et al. (2019) "Convergent evolution in the mechanisms of ACBD3 recruitment to picornavirus replication sites" *PLoS Pathog*
51. Gangaramani, Eden, Shah et al. (2010) "The twenty-nine amino acid C-terminal cytoplasmic domain of poliovirus 3AB is critical for nucleic acid chaperone activity" *RNA Biol*
52. Shang, Deng, Ye et al. (2013) "Develop ment and characterization of a stable eGFP enterovirus 71 for antiviral screening" *Antiviral Res*
53. Yin, Chen, Schul et al. (2009) "An adenosine nucleoside inhibitor of dengue virus" *Proc Natl Acad Sci*
54. Van Slyke, Arnold, Lugo et al. (2015) "Sequence-specific fidelity alterations associated with West Nile virus attenuation in mosquitoes" *PLoS Pathog*
55. Bordería, Isakov, Moratorio et al. (2015) "Group selection and contribution of minority variants during virus adaptation determines virus fitness and phenotype" *PLoS Pathog*
56. Li, Venter, Yooseph et al. (2010) "ANDES: statistical tools for the ANalyses of DEep Sequencing" *BMC Res Notes*
57. Xiong, Zhang, Xu et al. (2024) "PHDtools: a platform for pathogen detection and multi-dimensional genetic signatures decoding to realize pathogen genomics data analyses online" *Gene*
58. Xiong, Zhang, Yu et al. (2020) "Distribution of intra-host variations and mutations in the genomes of SARS-CoV-2 and their implications on detection and therapeutics" *MedComm*
59. Zhang, Xiong, Yu et al. (2021) "Genetic polymor phism drives susceptibility between bacteria and bacteriophages" *Front Microbiol*
60. Oyejobi, Xiong, Shi et al. (2022) "Genetic signatures from adaptation of bacteria to lytic phage identify potential agents to aid phage killing of multidrug-resistant Acinetobacter baumannii" *J Bacteriol*
61. Gohara, Ha, Kumar et al. (1999) "Production of "authentic" poliovirus RNA-dependent RNA polymerase (3D pol ) by ubiquitin-protease-mediated cleavage in Escherichia coli" *Protein Expr Purif*
62. Shi, Ye, Deng et al. (2020) "A nucleobasebinding pocket in a viral RNA-dependent RNA polymerase contributes to elongation complex stability" *Nucleic Acids Res*
63. Yang, Cheng, Zhang et al. (2014) "A cypovirus VP5 displays the RNA chaperonelike activity that destabilizes RNA helices and accelerates strand annealing" *Nucleic Acids Res*
64. Wang, Wu, Wang et al. (2020) "Structural basis for RNA replication by the SARS-CoV-2 polymerase" *Cell*
65. Chen, Chen, Zhang et al. (2021) "The genome sequence archive family: toward Full-Length Text Journal of Virology October"
66. "explosive data growth and diverse data types" *Genom Proteom Bioinform*
67. Bai, Bao, Bei et al. (2024) "Database resources of the National Genomics Data Center, China National Center for Bioinformation in 2024" *Nucleic Acids Res*
68. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12798023&blobtype=pdf | # Whole genome sequence analysis and characterization of virulent Newcastle disease virus isolates from chicken and pheasants during 2020/21 outbreaks in Nepal
Meera Prajapati, Joshua Lynton-Jenkins, Aashirbad Pokharel, Shresha Rayamajhi, Madhav Acharya, Manita Aryal, Suruchi Karn, Joe James, Ashley Banyard, Craig Ross
## Abstract
Newcastle Disease (ND) remains a major threat to poultry production worldwide, particularly in regions where it is endemic, like Southern Asia. The disease is caused by virulent forms of avian paramyxovirus-1, commonly termed Newcastle Disease Virus (NDV), a highly contagious virus with significant genetic diversity and evolving pathogenicity. This study aimed to molecularly characterize NDV isolates obtained from chickens and pheasants during the 2020/21 ND outbreaks in Nepal, to understand their genetic makeup, phylogenetic relationships, and implications for control strategies. Necropsy samples, including trachea, liver, intestine, spleen, lungs, heart, and proventriculus were collected from ten birds. Isolates from five clinical samples were typed as NDV by hemagglutination and hemagglutination inhibition (HA/HI) assays and were subjected to whole genome sequencing (WGS). Full genomes of 15,192 nucleotides were recovered from each isolate. Fusion (F) gene sequence analysis revealed the presence of multi-basic cleavage site motif 112 RRQKRF 117 in all isolates, indicative of virulent strain and suggesting a potentially velogenic or mesogenic phenotype. Phylogenetic analyses consistently classified all isolates within genotype VII.2 of class II NDV. Further comparative analysis indicated a close genetic relationship between the Nepalese isolates and strains reported from India and Bangladesh, and BEAST analysis suggested Southern Asia as the likely source of introduction into Nepal. These viral genomes provide additional insight into contemporary NDV circulating in an area of endemicity.
## Introduction
Newcastle disease (ND), caused by virulent forms of avian paramyxovirus-1 (APMV-1), commonly termed Newcastle Disease Virus (NDV), is a highly contagious, devastating viral infection of domesticated and wild birds. The infection can cause severe clinical disease, high mortality, and significant economic losses due to decreased productivity, increased mortality, and associated trade restrictions. ND remains endemic in many parts of the world, particularly in resource limited settings, where outbreaks lead to high flock mortalities and threaten poultry-based livelihoods.
The causative agent, avian paramyxovirus-1, is an enveloped, single stranded negative sense RNA virus belonging to the family Paramyxoviridae, genus Orthoavulavirus [46]. The viral genome is approximately 15.2 kilobases in length and encodes six structural proteins: nucleocapsid protein (NP), phosphoprotein (P), matrix protein (M), fusion protein (F), haemagglutinin-neuraminidase (HN), and large RNA-dependent polymerase (L) in the order of 3'-N-P-M-F-HN-L-5' [9]. Two accessory proteins are encoded from the P gene following ribosomal slippage, namely V and W [51]. NDV strains are broadly classified into four pathotypes: avirulent, lentogenic (low morbidity, no mortality), mesogenic (high morbidity, low mortality), and velogenic (high morbidity and high mortality), with velogenic strains causing either viscerotropic disease (with intestinal haemorrhaging) or neurotropic disease (with neurological symptoms such as torticollis and encephalitis). A key molecular marker of virulence is the fusion (F) protein cleavage site (CS). All virulent forms have a multi-basic cleavage site from positions 113-116 and a phenylalanine at position 117, allowing cleavage of the F protein from inactive F 0 to the active form F 1 and F 2 by Furin-like proteases [37,57], ubiquitously expressed in multiple cell types, thus resulting in systemic infection, including detection in the brain [2]. Cleavage sites of low or avirulent viruses have di-basic sequence at position 113-116 and a leucine at position 117, cleaved by trypsin-like proteases which are only detected in the respiratory and intestinal tracts, and thus limits infections to these tissues [8,16]. The presence of a virulent CS is one of the determinants for declaring ND in a poultry flock, along with a score of > 0.7 from the diagnostic Intracerebral Pathogenicity Index (ICPI) assay, as decreed by the World Organisation for Animal Health (WOAH). Besides these molecular markers, antigenic epitopes on the F and HN proteins, the viral surface glycoproteins, play a crucial role in host immune recognition and vaccine efficacy. Several neutralizing epitopes have been mapped on the F protein, while the HN protein contains major antigenic sites that mediate receptor binding and neuraminidase activity, making both genes key targets for molecular characterization and comparative analysis [19,36].
Although all APMV-1 are of a single serotype, phylogenetic analysis of the F gene has resulted in the distinction of two APMV-1 classes (class I and II) [11]. Class I viruses are largely avirulent in nature and mostly found in wild birds [11] whereas class II contains twenty genotypes (I -XXI, with genotype XV unassigned) of both avirulent strains (e.g. vaccine strains La Sota, Hitchner B1, Ulster 2C), and virulent strains responsible for the majority of ND outbreaks worldwide [11,30]. The severity and outcome of ND outbreaks are largely determined by the virulence of the circulating NDV strain and its interaction with the host immune response and vaccination status. The use of vaccines is common in both countries where NDV is absent and endemic [27]. Despite global vaccination efforts, NDV continues to cause outbreaks which could be due to improper vaccine storage, handling, and administration, as well as incomplete flock coverage [11,13,23].
APMV-1 has been detected in over 241 different bird species, in 27 different orders [20]. Infections with velogenic NDV strains are most commonly reported in domesticated Galliformes, whereas lentogenic and avirulent variants are also detected in wild and aquatic birds [11,28]. Studies have indicated that Phasianidae (e.g., pheasants and partridges), are equally susceptible to virulent ND genotype VII as chickens and, where raised in backyard settings, and have the potential to serve as intermediaries between wild birds and domestic poultry [1,48]. However, where pheasants are more typically reared in commercial settings, such as in South Asia, reports of pheasant-specific outbreaks are rare.
Nepal is a resource-limited country in South Asia where poultry farming plays a vital role in food security and income generation, particularly among rural populations. The poultry industry comprises a mixture of commercial and backyard systems, both of which are vulnerable to infectious diseases such as ND. Although vaccination is widely practiced (primarily using genotype II strains), ND outbreaks remain frequent and severe, undermining disease control efforts [33,42]. NDV outbreaks have been reported in pheasants and chickens [38,42]. More recent outbreaks from 2021 were identified as genotype VII.2 and published partial sequencing of the fusion protein gene [33,42]. However, few studies have investigated the molecular characteristics of these emerging strains in Nepal.
Genotype VII of NDV was initially identified in the Far East during the mid-1980s [19,25], with the first known isolate later determined to have occurred in Indonesia in 1976 [31] and has since emerged as a major cause of ND outbreaks across Southeast, East, and South Asia, as well as parts of Africa and Europe [17]. This genotype has become endemic in many countries and is associated with large-scale poultry losses and economic disruption [7,12]. Although genotype VII strains are increasingly detected in Nepal, there is a lack of whole genome sequencing (WGS) data to characterize their genetic makeup. Without detailed molecular information, it is difficult to track virus evolution, identify markers of virulence, or transmission [47]. Following an outbreak in poultry in the Kathmandu region of Nepal in 2020/21, NDV was identified and isolated from chickens and pheasants. We performed WGS of five NDV isolates and analysed them phylogenetically to understand their genetic features, virulence determinants, and evolutionary relationships. This is the first study to apply WGS to NDV strains circulating in Nepal and is expected to generate crucial insights for guiding future vaccine development and improving national disease control strategies.
## Methodology
## Sample collection and processing
Poultry suspected of NDV infection were brought for postmortem examination (PME) at the National Animal Health Research Center (NAHRC), Khumaltar, Lalitpur. Organ samples, including trachea, liver, intestine, spleen, lungs, heart and proventriculus were collected at the time of necropsy. The lesions observed during necropsy were haemorrhages in trachea, gastrointestinal tract, pinpoint haemorrhages in proventriculus, nephritis, hepatomegaly and splenomegaly. Samples from birds which showed the lesion commonly associated with ND virus and detected positive by NDV antigen test (Bionote, Cat no: RG1503DD) were stored at -80 o C. Altogether, 18 organ samples from ten birds were collected for virus isolation from pheasants and backyard chickens (Table 1). These samples were promptly processed for virus isolation at NAHRC, Nepal and NDV positive allantoic fluids were kept at -80 °C until it was sent to Animal and Plant Health Agency, UK for whole genome sequencing.
## Virus isolation
For each bird, multiple tissue samples were collected, pooled and a 20% (w/v) tissue homogenate was prepared using phosphate buffered saline supplemented with gentamycin (10 mg/ml). The suspension was placed at 4 o C for 2 h and then clarified by centrifugation at 3000 rpm for 10 min. Approximately 0.2 ml of the supernatant from the tissue homogenate was inoculated into the allantoic cavity of 10 days old embryonated fowls eggs (EFEs) which were tested as NDV free as stated by WOAH [59]. The eggs were incubated at 37 °C until the embryos died or for a maximum period of 96 h. Embryos were candled every 24 h, and the dead embryos were stored at 4 °C prior to collection of the allantoic fluids. Embryos that died before 24 h of incubation were excluded. The harvested allantoic fluid from each egg was checked for hemagglutination activity (HA) [59]. HA positive allantoic fluids were confirmed as NDV by haemagglutination inhibition test (HI) using the ND antiserum (Ulster Newcastle Disease Virus Antiserum, UK. Lot no: 1/20) at NAHRC, Nepal and by rRT-PCR at the Animal and Plant Health Agency (APHA), UK.
## RNA Extraction, PCR and detection
Nucleic acid extraction was performed using the QIAmp viral RNA mini kit (QIAGEN, Manchester, UK) according to the manufacturer's instructions. The presence of APMV-1 viral RNA (vRNA) in tissues was determined using the L-gene rRT-PCR assay [43,53].
## Whole genome sequencing and phylogenetic analysis
WGS of five virus isolates from the incursion was carried out as described previously [43]. In brief, cDNA was generated using the SuperScript IV First-Strand Synthesis System with random hexamers (ThermoFisher), and then to double-stranded cDNA using the NEBNext Ultra II Non-Directional RNA Second Strand Synthesis Module (New England Biolabs). cDNA was purified using Agencourt AMPure XP beads (Beckman Coulter). Subsequently, 1 ng of purified dsDNA was used as template in the sequencing library generated using the Nextera XT kit (Illumina). Sequencing was performed on a NextSeq 550 (Illumina) with 2 × 150 base paired end reads. Raw sequencing reads were assembled using a custom de novo assembly approach. F gene sequences were combined with representative APMV-1 sequences obtained from GenBank ( w w w . n c b i . n l m . n i h . g o v / g e n b a n k /) and from the NDV consortium sequence database ( h t t p s : / / g i t h u b . c o m / N D V c o n s o r t i u m / N D V _ S e q u e n c e _ D a t a s e t s). Sequences were first aligned using Mafft [22], M o d e l F i n d e r [21] was then used to determine an appropriate phylogenetic model and trees were inferred using maximum-likelihood models in IQ-Tree v2.2.5 [29] with 1,000 ultrafast bootstraps [18]. Bayesian phylogenetic analysis of selected full F-gene sequences was carried out using BEAST v 1.1.10.4 [52] in combination with Beagle library [4]. We employed an uncorrelated relaxed clock and constant population size (see Table S1) using a General Time Reversible substitution model [55] with separate partitions for codon positions 1 plus 2 versus position 3. Two independent MCMC chains with a length of 200,000,000 and sampling every 20,000 iterations were carried out, with the first 10% discarded as the burn-in. Convergence was assessed using Tracer v1.7.2 [39]. The maximum clade credibility (MCC) tree was summarized using TreeAnnotator v1.10.4 [52] and visualized using R with the tidyverse, treeio and phytools packages [45,56,58]. To determine regional spread, APMV-1 samples were designated a region determined by the UN geoscheme [54] with the Nepal designated independently. To explore the pattern of spatial diffusion among the geographic regions, discrete phylogeographic analyses using location as a trait were performed [26]. We assumed an asymmetric non-reversible transition model and incorporated Bayesian stochastic search variable selection [26]. SpreaD3 was used to measure rates of transmission using Bayes Factor (BF) and was used to determine likelihood of transmission between locations [5]. The support for BF transmission was as described previously [24]. BF and representative transitions were visualized as described previously [50]. Markov jump counts were used to measure the number of viral movements along the branches of the phylogeny and estimated the Markov rewards to quantify the time the virus spent in each geographical region [32]. As an alternative method to examine evolutionary and geographic spread, analysis was carried out using the mugration model in Treetime, using default settings [49]. The analysis was carried out on a previously generated M-L tree, with molecular clock estimation, using the UN geoscheme designated regions.
For the analysis of amino acid mutations in the F and HN genes, the nucleotide sequences of these genes from all five isolates were translated into amino acid sequences using the translate function from the Biopython Seq module. Along with these, comparative sequences included the reference strain APMV 1/chicken/ NL/152,608/93 (GenBank accession AEZ00711) and four vaccine strains: LaSota (AAC28374.1), Mukteswar (EF201805.1), Ulster 2 C (PQ301174.1) and Queensland V4 (JX524203.1). All sequences were aligned using MAFFT v7.505 with default parameters. The F gene cleavage site (residues 112-117) was analysed to determine the virulence-associated motif. Epitope regions in both the F and HN genes were identified based on previously reported studies and examined for strain-specific amino acid substitutions. Comparative analysis was performed using custom Python scripts to extract and tabulate mutations at the defined positions.
## Results
## Haemagglutination and haemagglutination inhibition test
Of the ten initial sample pools analysed, five isolation attempts were found negative in HA/HI testing and five were found positive for APMV-1 in HA/HI. A second passage was performed on the positive isolate and confirmed to be APMV-1 by rt-PCR analysis with WGS carried out on each sample. The details of the samples are shown in Table 1.
## Phylogenetic and comparative genome analysis
Following WGS, the full F-gene sequence from across all twenty APMV-1 genotypes was used to determine the genotype of the five Nepalese isolates. Analysis demonstrated that all five Nepalese isolates were genotype VII.2 (Fig. 1 and Figure S1) as described by the classification system [11]. Three of the isolates clustered in one group with pairwise nucleotide identity analysis of the F gene showing them to be identical at the nucleotide level (100% identity), the remaining two isolates formed a second group also sharing 100% within-group F gene identity.
Further analysis of F-gene sequences previously designated as genotype VII.2 demonstrated that the Nepalese samples were closely related to samples observed in India and Bangladesh between 2020 and 2024 (Figure S2). For example, strains reported from Guwahati and Chhattisgarh in India showed the closest pairwise nucleotide identity with the two unique Nepalese F gene sequences reported in this study (NDV/Owl/Guwahati/01/20 (MZ546197) [15], 99.3% and 99% identity with Nepalese strains; NDV/Chicken/RPR/01/23 (OR185447) [41], 99.1% and 98.8% identity). While the next closest strains were reported from Chattogram in Bangladesh (e.g., APMV-1/NDV/Chicken/BD/37o/2024 (PV656753), 99.1% and 98.8% identity).
To further determine the incursion of virulent APMV-1 into Nepal, BEAST analysis was carried out. This determined that the most likely region of entry into Nepal was from Southern Asia (Fig. 2A, B andC) with the time to most recent common ancestor (TMRCA) to isolates from India estimated to be September 2018 (range Nov 2017 -Dec 2019, [Treetime Feb 2018, range Jan 2016 -Mar 2018] (Figure S3)). As previously noted, within Nepal, the five viruses split into two distinct clades, with the TMRCA for the Nepalese split predicted to be May 2019 (range June 2018 -Mar 2020, [Treetime-not determined]).
Spread analysis suggests that, with a high degree of posterior probability, that the ancestor was originally in South-east Asia, transitioned into Southern Asia, and subsequently moved into Nepal, when these incursions were detected (Fig. 2B and Table S2). This is backed up by the phylogenetic analysis where the last common ancestor prior to detection in Nepal was from Southern Asia (Fig. 2A and Figure S3). The TMRCA from South-east Asia into Southern Asia was estimated to have occurred in April 2016 (range Feb 2015 -May 2017, [Treetime Feb 2015, range June 2014 -Sept 2016]). Markov jump analysis suggests that the precursors transitioned from Southeast Asia to both Southern and Eastern Asia, before transitioning from Southern Asia into Nepal (Fig. 2 C). Markov jump counts suggest that the precursors to the Nepalese incursion spent the longest time in South-east Asia, before being transmission into both Southern and Eastern Asia (Fig. 2D), and subsequently into Nepal (via Southern Asia).
## F and HN protein analysis
F gene cleavage site analysis showed a motif ¹¹²RRQKR/ F¹¹⁷ in all five isolates. In contrast, the vaccine strains possessed either ¹¹²GRQGR/L¹¹⁷ or ¹¹²GKQGR/L¹¹⁷ motifs, except for the Mukteshwar strain, which showed ¹¹²RRQRR/F¹¹⁷. Mukteshwar, being mesogenic, is commonly used as a booster in ND endemic regions and was included for sequence comparison along with other vaccine strains in this study. Furthermore, amino acid analysis of the F and HN genes revealed no major substitutions or alterations in the field isolates when compared with the reference and vaccine strains (Table S3).
## Discussion
Though ND is endemic in Nepal, detailed study on the circulating virus is limited. This is the first study conducting whole genome sequencing with phylogeographic analysis of viral isolates from infected poultry, identifying genotype VII.2 of class II NDV as the causative agent of the outbreaks.
NDV strains belonging to genotype VII, which can be further categorized into genotype VII.1.1, VII.1.2 and VII.2 based on amino acid substitutions, are implicated in numerous outbreaks of virulent Newcastle disease (ND) across Asia, Africa, Europe and the Middle East since the late 1990 s [11,12,31]. Among these, genotype VII.2 has emerged as a dominant lineage, often associating with panzootic-like spread and high mortality in poultry across South-East and Southern Asia [11]. The findings from the whole genome sequencing revealed that all five samples analysed belong to genotype VII.2, concordant with previous findings [33] who also reported genotype VII.2 from NDV outbreaks in 2021. These five isolates clustered into two groups each with 100% within-group F gene identity. Two isolates forming one group were isolated from samples originating from two distinct farms (pheasants and chickens), whereas the second group of three isolates were collected from the same farm at different time points during the outbreak. The nucleotide identity between the two groups was 98.74%, indicating minor sequence differences between them. The identification of genotype VII.2 in Nepal underscores its persistence in the region [33,42]. Spread analysis and Markov jump analysis demonstrated that the initial precursor isolates were present in South-east Asia, before transitioning into Southern Asia (TMCRA estimation April 2016 [range Feb 2015 -May 2017]) before subsequently spreading throughout India, Bangladesh and Nepal. These South-east Asian isolates have also caused outbreaks in Pakistan, which were subsequently linked to incursions in Western Asia and Western Europe (Fig. 2B). This spread of Genotype VII.2 virus across Southern, South-eastern and Western regions of Asia demonstrate the difficulty of control of this virus, even where vaccination is routinely practised.
Genotype VII.2 is historically associated with velogenic strains of NDV, with high ICPI values resulting in severe systemic disease and high mortality in poultry. The conservation of the F-gene cleavage site (¹¹²RRQKR/F¹¹⁷) in all field isolates strongly suggests their virulent nature, aligning with class II virulent NDV strains known to cause outbreaks in poultry. This is consistent with previous studies reporting that multiple NDV outbreaks in Asia are driven by virulent strains with multi-basic cleavage sites [12]. Analysis of F gene revealed that the Nepalese isolates were closely related to India and Bangladesh between 2020 and 2023. Though directly not assessed in this study, this close genetic relationship may suggest potential cross-border viral movement influenced by regional poultry trade, migratory birds, or biosecurity lapses. In this study, three ND isolates were isolated from a vaccinated flock of 2-3 months old layers. Similarly, previous study reported infection of ND in vaccinated poultry [42]. Vaccination together with strict biosecurity measures is important to contain the disease. However, both biosafety and biosecurity measures were found to be generally poor in commercial chicken farms across Nepal [33]. This condition is worse in the case of backyard poultry where there is no biosecurity present. A pilot study reported that farms achieved only 42% compliance with personnel safety standards and 3% for rodent control, reflecting the widespread absence of disinfectants, footbaths, and protective clothing for staff [40]. In addition, field vaccination practices often suffer from inadequate cold chain maintenance, improper handling or reconstitution of vaccines, and lack of standardized administration schedules, all of which can reduce vaccine efficacy [6]. Moreover, feed contamination with mycotoxins such as aflatoxin B₁ has been reported in Nepalese poultry feed [3] and can impair immune responses, reducing antibody production even after proper vaccination [35]. These factors collectively contribute to reduced vaccine efficiency and recurrent NDV outbreaks despite routine vaccination.
Structural weaknesses such as substandard housing, free-ranging birds, and insecure feed storage further enabled access by wild birds and rodents, significantly elevating disease risk [10]. In such conditions, introduction and transmission of disease is easy from farm to farm contributing to the endemic nature of NDV in this region. The open border to India and proximity with Bangladesh, combined with the lack of strict veterinary checks, provides a critical pathway for the introduction of NDV into Nepal. Informal trade in live birds, hatching eggs, and poultry products from neighbouring Indian states, where repeated outbreaks of genotype VII have been documented, likely facilitated cross-border incursions of NDV [34]. In addition, Nepal's position along the Central Asian Flyway, with shared wetlands and transboundary protected areas such as Chitwan-Valmiki, allows the movement of migratory birds and raptors between these countries. Although the precise role of wild birds in genotype VII transmission remains debated, their frequent interaction with backyard poultry in border regions may have further promoted viral spillover [14]. Once introduced, the weak biosecurity practices as described above may have allowed NDV to spread rapidly within Nepal, thereby establishing endemic circulation.
## Conclusions
This study has successfully isolated ND virus that circulated among poultry population in 2021 and provided whole genome sequences of NDV isolates offering valuable insights into molecular epidemiology. Phylogenetic analysis demonstrated close links with isolates from Bangladesh and India, with proposed routes of incursion outlined. The identification of genotype VII.2 strains in vaccinated birds demonstrates the capacity of this virus to cause outbreaks despite flock vaccination, highlighting existing limitations in local biosecurity and the application of effective vaccination.
devolved administrations of Scotland and Wales through grant SV3002 and SE2228.
## References
1. Aldous, Alexander (1016) "Newcastle disease in pheasants (Phasianus colchicus): a review" *Vet J*
2. Alexander (2000) "Newcastle disease and other avian paramyxoviruses" *Rev Sci Tech*
3. Aryal, Karki (2009) "Prevalence of aflatoxin B1 and B2 in poultry feed" *Nepal Agric Res J*
4. Ayres, Darling, Zwickl et al. (2012) "BEAGLE: an application programming interface and high-performance computing library for statistical phylogenetics" *Syst Biol*
5. Bielejec, Baele, Vrancken et al. (2016) "SpreaD3: interactive visualization of spatiotemporal history and trait evolutionary processes" *Mol Biol Evol*
6. Birhane, Fesseha (2020) "Vaccine failure in poultry production and its control methods: a review" *Biomed J Sci Tech Res*
7. Charkhkar, Bashizade, Sotoudehnejad et al. (2024) "The evaluation and importance of Newcastle disease's economic loss in commercial layer poultry" *J Poult Sci Avian Dis*
8. De Bruin, Spronken, Bestebroer et al. (2023) "Conserved expression and functionality of Furin between chickens and ducks as an activating protease of highly pathogenic avian influenza virus hemagglutinins" *Microbiol Spectr*
9. De Leeuw, Peeters (1999) "Complete nucleotide sequence of Newcastle disease virus: evidence for the existence of a new genus within the subfamily Paramyxovirinae" *J Gen Virol*
10. Dhakal, Devkota, Jethara et al. (2025) "Assessment of biosecurity in poultry farms in Chitwan, Nepal" *Vet Med Sci*
11. Dimitrov, Ramey, Qiu et al. (2016) "Temporal, geographic, and host distribution of avian paramyxovirus 1 (Newcastle disease virus)" *Infect Genet Evol*
12. Dimitrov, Abolnik, Afonso et al. (2019) "Updated unified phylogenetic classification system and revised nomenclature for Newcastle disease virus" *Infect Genet Evol*
13. Dortmans, Peeters, Koch (2012) "Newcastle disease virus outbreaks: vaccine mismatch or inadequate application?" *Vet Microbiol*
14. Eid, Hussein, Hassanin et al. (2022) "Newcastle disease genotype VII prevalence in poultry and wild birds in Egypt" *Viruses*
15. Gaurav, Deka, Das et al. (1080) "Isolation of genotype VII avian orthoavulavirus serotype 1 from barn Owl from Northeast India" *Avian Pathol*
17. Glickman, Syddall, Iorio et al. (1988) "Quantitative basic residue requirements in the cleavage-activation site of the fusion glycoprotein as a determinant of virulence for Newcastle disease virus" *J Virol*
18. Hicks, Dimitrov, Afonso et al. (2019) "Global phylodynamic analysis of avian paramyxovirus-1 provides evidence of inter-host transmission and intercontinental spatial diffusion" *BMC Evol Biol*
20. Hoang, Chernomor, Haeseler et al. (2018) "UFBoot2: improving the ultrafast bootstrap approximation [Evaluation study. Research, Support Non-US Gov't"
21. Hu, He, Deng et al. (2022) "Current situation and future direction of Newcastle disease vaccines" *Vet Res*
22. Jin, Wei, Bi et al. (1186) "Identification of a potential neutralizing linear epitope of hemagglutinin-neuraminidase in Newcastle disease virus" *Virol J*
23. Kaleta, Baldauf (1988) "Newcastle disease in free-living and pet birds"
24. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: fast model selection for accurate phylogenetic estimates" *Nat Methods*
25. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
26. Kondakova, Agranovsky, Ryabchevskaya et al. (2025) "Genetic diversity of Newcastle disease virus and its implications for vaccine development" *Vet Sci*
27. Lee, Wagenmakers (2014) "Bayesian cognitive modeling: a practical course"
28. Lee, Sung, Choi et al. (1080) "Molecular epidemiology of Newcastle disease viruses isolated in South Korea using sequencing of the fusion protein cleavage site region and phylogenetic relationships" *Avian Pathol*
29. Lemey, Rambaut, Drummond et al. (2009) "Bayesian phylogeography finds its roots" *PLoS Comput Biol*
30. Mayers, Mansfield, Brown (2017) "The role of vaccination in risk mitigation and control of Newcastle disease in poultry" *Vaccine*
31. Miller, Koch (2020) "Diseases of poultry"
32. Miller, Decanini, Afonso (2010) "Newcastle disease: evolution of genotypes and the related diagnostic challenges" *Infect Genet Evol*
33. Miller, Haddas, Simanov et al. (2015) "Identification of new sub-genotypes of virulent Newcastle disease virus with potential panzootic features" *Infect Genet Evol*
34. Minh, Schmidt, Chernomor et al. (2020) "Iq-tree 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol*
36. Minin, Suchard (1512) "Fast, accurate and simulation-free stochastic mapping" *Philos Trans R Soc Lond B Biol Sci*
37. Napit, Poudel, Pradhan et al. (2023) "Newcastle disease burden in Nepal and efficacy of tablet I2 vaccine in commercial and backyard poultry production" *PLoS One*
38. Nath, Barman, Kumar (2016) "Molecular characterization of Newcastle disease virus strains isolated from different outbreaks in Northeast India during 2014-15" *Microb Pathog*
39. Olariu, Fiţ, Bouari et al. (2025) "Mycotoxins in broiler production: impacts on growth, immunity, vaccine efficacy, and food safety" *Toxins*
40. Omony, Wanyana, Mugimba et al. (2021) "Epitope peptide-based predication and other functional regions of antigenic F and HN proteins of waterfowl and poultry avian avulavirus serotype-1 isolates from Uganda. Front Vet Sci"
41. Panda, Huang, Elankumaran et al. (2004) "Role of fusion protein cleavage site in the virulence of Newcastle disease virus" *Microb Pathog*
42. Poudel, Karna, Gompo (2023) "Study on management status and disease aspects of pheasant (Phasianus colhicus) of Kathmandu Valley" *Nepal. Nepal Vet J*
43. Poudel, Sharma, Dhital et al. (2024) "Antimicrobial stewardship hindered by inadequate biosecurity and biosafety practices, and inappropriate antibiotics usage in poultry farms of Nepal-a pilot study" *PLoS One*
44. Rambaut, Drummond, Xie et al. (2018) "Posterior summarization in bayesian phylogenetics using tracer 1.7" *Syst Biol*
45. Reddy, Patil, Shah et al. (2024) "Deciphering whole genome sequence of a Newcastle disease virus genotype VII.2 isolate from a commercial poultry farm in India" *Gene Rep*
46. Regmi, Bhatta, Pal et al. (2024) "Clinicopathological and molecular investigation of Newcastle disease outbreaks in vaccinated and non-vaccinated broiler chicken flocks in Nepal" *Animals*
47. Reid, Skinner, Sutton et al. (2023) "Understanding the disease and economic impact of avirulent avian paramyxovirus type 1 (APMV-1) infection in Great Britain" *Epidemiol Infect*
48. Revell (2024) "Phytools 2.0: an updated R ecosystem for phylogenetic comparative methods (and other things)" *PeerJ*
49. Rima, Balkema-Buschmann, Dundon et al. (2019) "ICTV virus taxonomy profile: Paramyxoviridae" *J Gen Virol*
50. Rohaim, Mq, Naggar et al. (2022) "Evolutionary trajectories of avian avulaviruses and vaccines compatibilities in poultry. Vaccines"
51. Ross, Skinner, Sutton et al. (2023) "Game birds can act as intermediaries of virulent genotype VII avian orthoavulavirus-1 between wild birds and domestic poultry" *Viruses*
52. Sagulenko, Puller, Neher (2018) "TreeTime: maximum-likelihood phylodynamic analysis" *Virus Evol*
53. Shabbir, Mahmood, Ul-Rahman et al. (2024) "Genomic diversity and evolutionary insights of avian paramyxovirus-1 in avian populations in Pakistan" *Viruses*
54. Steward, Vipond, Millar et al. (1993) "RNA editing in Newcastle disease virus" *J Gen Virol*
55. Suchard, Lemey, Baele et al. (2018) "Bayesian phylogenetic and phylodynamic data integration using BEAST 1.10" *Virus Evol*
56. Sutton, Allen, Fuller et al. (2019) "Development of an avian avulavirus 1 (AAvV-1) L-gene real-time RT-PCR assay using minor groove binding probes for application as a routine diagnostic tool" *J Virol Methods*
57. (2021) "Standard country or area codes for statistical use (M49)"
58. Vinh, Haeseler (2006) "Computational molecular evolution-Ziheng Yang"
59. Wang, Yu, Huo et al. (2017) "Comprehensive analysis of amino acid sequence diversity at the F protein cleavage site of Newcastle disease virus in fusogenic activity" *PLoS One*
60. Wang, Lam, Xu et al. (2020) "Treeio: an R package for phylogenetic tree input and output with richly annotated and associated data" *Mol Biol Evol*
61. Wickham, Averick, Bryan et al. (2019) "Welcome to the Tidyverse" *J Open Source Softw*
62. Swayne, Brown (2023) "Newcastle disease (infection with Newcastle disease virus), 3.3.14 in WOAH Terrestrial Manual 2023. World Organisation for Animal Health" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817942&blobtype=pdf | # A murine coronavirus infection platform identifies proviral and proinflammatory activities of SARS-CoV-2 accessory protein 7a
Grant Hawkins, Enya Qing, Julisa Salgado, Pearl Chan, Edward Campbell, Stanley Perlman, Tom Gallagher
## Abstract
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and related sarbecoviruses encode a set of accessory proteins (3a, 3b, 6, 7a, 7b, 8, 9b, and 10) that control host responses to infection and promote virus growth. Of these accessory proteins, 7a is set apart by its intracellular localization near CoV budding sites and its incorporation into secreted virions. To investigate 7a functions during CoV infections under biosafety level 2 conditions, we constructed recombinant mouse hepatitis viruses (rMHVs) (rMHV strain A59) that express sarbecovirus 7a genes. Comparative infections revealed that 7a increased viral replication and viral output in immortalized murine cell cultures and in primary bone marrow-derived macrophages (BMDMs). This proviral effect was independent of a previously reported 7a-mediated interferon antagonizing activity. 7a is a type I transmembrane protein with a short cytoplasmic tail that oper ates in subcellular trafficking and signal transduction. To further elucidate tail func tions, we generated a set of rA59 viruses expressing substitutions in the tail di-lysine motifs. Several substitutions reduced 7a proviral activities; notably, the K119A change in rA59-7a-KRATE eliminated 7a support of virus yield. The K119A change also reduced mouse-adapted SARS-CoV-2 virus yields in infected BALB/c mice. 7a expression was proinflammatory in BMDMs, as measured by cytokine arrays. Cytoplasmic tail substitu tions tempered these proinflammatory responses, implying connections with proviral activities. SARS-CoV-2-infected macrophages have been implicated in inflammatory COVID-19, and these findings point to 7a cytoplasmic tails as potential contributors to cytokine-mediated disease.
IMPORTANCEThis study shows that SARS-CoV-2 accessory protein 7a promotes infection of a phylogenetically distinct embecovirus and, in doing so, elicits proinflamma tory and potentially disease-relevant host responses. The proviral and proinflammatory activities were traced in part to a short 7a cytoplasmic tail. The findings localize and highlight a specific proviral component in a sarbecovirus accessory protein.
frequently attenuates in vivo virus virulence (4)(5)(6)(7)13). Several accessory gene products, notably 3a and 6, have clear proviral functions that are also evident in vitro (4,14,15). Here we became drawn to protein 7a, in large part because of its particular structural features and wide range of reported functions. 7a is a virion-associated type I transmem brane protein (16). The 87-amino-acid Ig-like 7a ectodomain may bind alternative host cell-surface proteins and hypothetically expand virus spread (17)(18)(19). 7a accumulation within cells also generates several potentially proviral and proinflammatory outcomes by restricting intracellular transport and function of MHC-I (20), SERINC5 (21), and tetherin (22); by interfering with autophagy (23)(24)(25); or by inducing apoptosis (26), endoplasmic reticulum (ER) stress (26), and inflammation (27,28).
Some of these varied responses may trace to a short fiveaminoacid 7a cytoplas mic tail (29). The carboxy-terminal tail encodes a canonical ER retention motif, which potentially interacts with COPI coatomer complexes, possibly holding 7a in position for incorporation into virions (16). Intracellular retention traps both 7a and 7a-associated MHC-I molecules in the ER, reducing viral antigen presentation (20). While embedded in intracellular membranes, 7a tails bind antiapoptotic BclXL (26), affecting apoptotic responses. 7a tails interact with RNF121, an E3 ubiquitin ligase (27) which gener ates a membrane-associated polyubiquitin substrate that then accumulates secondary effectors that block STAT2 phosphorylation (30,31) and activate NF-kB, ultimately modulating proinflammatory responses (27,28,32).
Of note, nearly all of these reported mechanisms by which 7a effects functional responses have come from experiments in which the 7a proteins were overexpressed in the absence of infection and/or modified by epitope tags at carboxy-terminal tails. These experimental approaches have value but may not reflect unaltered 7a biology in the natural infection contexts that include many diverse virus-host interactions. Therefore, we aimed to evaluate 7a mechanisms operating during a complete infec tion process, using an orthologous murine coronavirus that can be investigated in biosafety level 2 laboratories. Rodentspecific embecoviruses have been proven to be useful tools for assessing functions of human pathogenic sarbecovirus accessory proteins (33,34). The embecovirus subgenus does not encode any accessory proteins homologous to SARS-CoV-2 (35)(36)(37), making it so that 7a might singularly add unique functions to embecovirus infections. For these reasons, we constructed recombinant mouse hepatitis embecoviruses encoding SARS-CoV-2 7a (rA59-7a) and then evaluated 7aspecific functions. The tests were facilitated by prior knowledge that antiviral host responses to MHV infection (38,39) can be measured ex vivo in immunologically competent cell types such as primary bone marrow-derived macrophages (BMDMs) (40)(41)(42). Using this orthologous platform to evaluate sarbecovirus accessory protein function, we found that 7a promotes viral replication and viral output while inducing proinflam matory responses that are mediated, in part, through 7a cytoplasmic tail interactions. Subsequent to these studies, we validated findings in recombinant SARS-CoV-2 infection contexts, demonstrating that 7a cytoplasmic tail modifications reduce virus growth in mouse infection models.
## RESULTS
## Construction of recombinant embecoviruses expressing sarbecovirus accessory 7a genes
To investigate potential functions of SARS-CoV-2 accessory protein 7a in the context of an orthologous and biosafe betacoronavirus infection, we used reverse genetics and circular polymerase extension reaction (CPER) methodologies (43,44) to generate a recombinant MHV (strain A59) containing SARS-CoV-2 ORF7a, designated as rA59-7a (Fig. 1a). ORF7a was inserted directly downstream of ORF E, under the control of a dedicated transcription regulatory sequence to ensure subgenomic 7a RNA synthesis (45,46) (Fig. 1b). In the rA59-7a genome, ORF4 was replaced with a nanoluciferase (Nluc) reporter gene, thereby allowing Nluc enzyme activities to serve as proxies for viral replication. Isogenic control rA59-7a-Null viruses were constructed by introducing three stop codons downstream of the ORF7a start codon (Fig. 1c). The recombinant viruses were generated within HEK293 cells transfected with CPERamplified DNAs, with secreted viruses then harvested, plaquepurified, and grown up to stock titers on mouse delayed brain tumor (DBT) astrocytoma cells.
Sequenceverified rA59-7a viruses were then used to infect 17Cl-1 mouse fibroblast cells. The infected cells were fixed and stained for immunofluorescence (IF) imaging at 8 h post-infection (hpi). Infected cells were identified by staining for replication organ elles using an anti-nsp3 antibody, and 7a expression was assessed with its respective antibody. The results showed abundant 7a expression in the rA59-7a infected cells but none in rA59-7a-Null (Fig. 1d).
## 7a does not antagonize antiviral interferon during rA59 infection
SARS-CoV-2 accessory protein 7a was previously shown to function as an interferon antagonist by preventing STAT2 phosphorylation (30)(31)(32). These findings were obtained from experiments in which the 7a proteins were expressed from plasmid DNAs in the absence of a complete coronavirus infection. Here we considered whether 7a proteins might increase the known interferon antagonizing potential that is evident in authentic MHV-infected cells (47,48). To this end, mouse DBT cells were exposed to serially diluted type I interferon 8 h prior to rA59 infection. At 12 hpi, viral Nluc accumulations were compared between rA59-7a and rA59-7a-Null (Fig. 2a). A proviral 7a effect was observed in mouse DBT cells and mouse BMDMs, but no significant difference was observed in an immortalized mouse macrophage cell line (NR-9456). Normalized data processing was used to evaluate effects of interferon (Fig. 2b). Surprisingly, interferon treatments did not suppress rA59-7a or rA59-7a-Null infections. By contrast, VSV infections were clearly dose-dependently suppressed, indicating intact antiviral interferon signaling in the DBT cells. The experiment was repeated in immortalized mouse macrophages and in IFN-sensitive primary mouse BMDMs. In these settings, interferon equally suppressed both rA59-7a and rA59-7a-Null due to interferon treatment, up to 10-fold and 100-fold in the immortalized and primary macrophages, respectively (Fig. 2c andd). These results demonstrate that 7a did not operate to antagonize the anti-MHV activities of type I interferon in murine cells, affording an opportunity to further dissect IFN-independent 7a activities.
## 7a increases virus replication and viral titers
We sought further characterization of the protein 7a-dependent increases in viral RNA replication (Fig. 2a). Experiments were initiated using the DBT cells that did not elicit interferon-induced rA59 restriction. Infected DBT cells were incubated with a live cell nanoluciferase substrate and read at hourly intervals by luminometry. rA59-7a replication began ~1 h earlier than rA59-7a-Null, with the two parallel infections then accumulating Nluc similarly (Fig. 3a). The earlier onset of rA59-7a replication was consistent with RT-qPCR measurements showing higher levels of 7a-positive viral genomes at 12 hpi (Fig. 3b). The findings also aligned with measurements of secreted rA59 viruses, revealing that 7a promoted ~10-fold increases in virus infectivity (Fig. 3c). Supporting 7a proviral activity, the rA59-7a plaques were significantly larger than rA59-7a-Null (Fig. 3d). To further assess 7a-dependent effects on infection, pelleted rA59 virus particles were resuspended, and virion proteins were detected by Western blotting. rA59-7a virion yields were clearly higher than rA59-7a-Null, as evidenced by S, N, and M band intensities (Fig. 3e). 7a incorporated into MHV virions, as might be predicted from evaluations of SARS-CoV ( 16), but did not alter the virion S:N or M:N ratios (Fig. 3e). The increased yields of secreted rA59-7a virion proteins concorded with increased virion proteins in cell lysates (Fig. 3e), suggesting that 7a promotes viral replication or translation but may not specifically augment virion assembly or secretion processes.
Similar experiments were performed using primary BMDMs. rA59 can be restricted in BMDMs (Fig. 2), and primary BMDMs are relevant ex vivo models for MHV-host cell interactions (40,41,49). Furthermore, SARS-CoV-2 infection of macrophages, while abortive (50)(51)(52)(53), elicits disease-relevant cytokine responses (54)(55)(56), making BMDMs a reasonable choice for further evaluation of 7a phenotypes. Infected BMDMs were monitored for Nluc reporter accumulations. Again, rA59-7a replication advanced earlier than 7a-Null, here about 2 h earlier (Fig. 3f). Relative to rA59-7a-Null, the rapid rA59-7a replication onset generated ~10-fold more viral genomes at 16 hpi (Fig. 3g), ~20-fold more secreted virion infectivities (Fig. 3h), and significantly more intracellular N proteins (Fig. 3i). These findings matched those obtained using DBT cells.
## Substitutions in the 7a cytoplasmic tail reduce proviral activities
The 121-aa transmembrane 7a proteins have short cytoplasmic tails (K 117 RKTE 121 ) that include canonical di-lysine ER retention motifs (KXKXX). Host-cell protein trafficking machinery, such as the COPI coatomer complex, interacts with these motifs and likely retains 7a to the ER and ER-Golgi intermediate compartment (20,29,57,58), the membrane organelles essential for viral replication (59,60) and viral assembly (61)(62)(63). Specifically, the K117 and K119 residues are considered essential for COPI coatomer binding (64).
Additionally, the K119 residue has been implicated in ubiquitination (26,27,30) and subsequent host factor binding (27), potentially augmenting interactions with host factors and other viral structural components (16). Therefore, to assess the relevance of the cytoplasmic motif, recombinant viruses were constructed in which each lysine (K117 or K119) was substituted with an alanine. The K117A and K119A recombinant viruses were designated rA59-7a-ARKTE and rA59-7a-KRATE, respectively.
Mouse BMDMs were inoculated with the mutant viruses, and replication kinetics were monitored through a single infection cycle. rA59-7a-ARKTE and rA59-7a-KRATE advanced at rates that were intermediate between the fast rA59-7a and the slow A59-7a-Null (Fig. 4a). Viral genome levels at 16 hpi were concurrently modestly decreased relative to rA59-7a (Fig. 4b). Secreted virus rA59-KRATE infectivities were expectedly low; however, there was a puzzling high rA59-ARKTE infectivity yield that was discordant with the reduced replication of this variant (Fig. 4c). rA59-7a-KRATE-infected cells contained viral nucleocapsid proteins at levels intermediate between the respective high-and lowoutput 7a and 7a-Null infected cells (Fig. 4d). To further assess the effect of 7a cytoplas mic tail substitutions on viral replication, infected BMDMs were fixed and probed for viral replication organelles using an anti-nsp3 antibody (Fig. 4e). Relative to the rA59-7a infections, there were significantly decreased numbers of BMDMs with established replication organelles in rA59-7a-KRATE-infected cultures (Fig. 4f). For the most part, these results reflect the measurements of viral genome accumulations through 16 h (Fig. 4b), although there was some discordance with the high rA59-7a-KRATE genome levels. The results suggest that lysine substitutions in 7a cytoplasmic tails reduce 7a proviral activities and that proviral effects take place at early RNA replication stages. The findings also leave open the possibility that the K119 residue, a known ubiquitin substrate (26,27,30), also facilitates virus assembly or egress (Fig. 4c).
While it has been assumed that the COPI coatomer complex interacts with the cytoplasmic ER retention motif of 7a, a direct interaction has not been shown. Therefore, 7a proteins were immunoprecipitated (IP) from infected DBT cell lysates. IP of 7a from the rA59-7a infection co-precipitated the β-COPI subunit of the coatomer complex (65,66), while IP of 7a-ARKTE and 7a-KRATE precipitated significantly less β-COPI (Fig. 4g). High-resolution IF images were taken using structured illumination microscopy (SIM) to assess potential effects of the reduced coatomer interactions on 7a localization. 7a, 7a-ARKTE, and 7a-KRATE all partially co-localized with viral S proteins and also with Golgin-97, a Golgi resident protein (Fig. 4h). Thus, surprisingly, disruption of the 7a-COPI interaction did not significantly alter 7a intracellular localization during MHV infection. This raised questions of whether 7a incorporation into virions was affected by the reduced COPI binding. Western blot analysis of secreted virion proteins revealed that the KRATE variant was present in the virion protein preparations in abundance similar to wild type 7a, while the ARKTE variant 7a showed less virion incorporation (Fig. 4i). The results demonstrated that these substitutions in the 7a cytoplasmic tail reduced interactions with coatomer complexes but had variable and somewhat marginal effects on 7a intracellular localization and virion incorporation.
## 7a mediates a proinflammatory response in BMDMs that depends on the cytoplasmic tails
Several previous studies have documented that accessory protein 7a can promote proinflammatory cytokine production (27,28,32), potentially mediated by the 7a cytoplasmic tails, specifically K119 (27), and operating through the NF-κB pathway (27,28). This suggests that 7a contributes to the elevated levels of proinflammatory cytokines, such as IL-6 and IL-12, observed in sera obtained from severe COVID-19 patients (67). We considered whether the various rA59-7a infections might generate differential proinflammatory responses and, if so, whether there are correlations between proinflammatory cytokines and virus replication and secreted virus yield. To this end, conditioned media from infected BMDMs were evaluated using a LEGENDPlex Mouse Anti-Virus Response panel kit (Fig. 5). rA59-7a and rA59-7a-ARKTE had significantly greater levels of IL-6 and IL-12 compared to rA59-7a-Null and rA59-7a-KRATE (Fig. 5a andb). These findings aligned with cytokine RNA levels, as measured by RT-qPCR (Fig. 5d and e), indicating controls at signaling and cytokine gene expression, not at the latest stages of cytokine protein secretion from BMDMs. The decreased IL-6 and IL-12 levels for rA59-7a-Null and rA59-7a-KRATE strongly correlated with their decreased viral titers, suggesting a link between proinflammatory cytokine production and virus secretion.
TNF-α cytokine and NFKBIA gene expression levels were assessed to determine if the 7a-dependent proinflammatory response observed in infected BMDMs was due to activation of the NF-κB pathway. TNF-α cytokine production and NFKBIA gene expres sion are common correlates for NF-κB activation (27,28,68). TNF-α levels were signifi cantly decreased in rA59-7a-Null infections but only slightly decreased in the parallel rA59-7a-ARKTE and rA59-7a-KRATE-infected cells (Fig. 5c). Relative to rA59-7a, NFKBIA gene expression was significantly decreased only in the rA59-7a-Null infected cells (Fig. 5f). The results suggest that the NF-κB pathway may be activated by 7a, without the essential participation of the K117 or K119 residues of the cytoplasmic tail. Yet the K residues, most notably K119, appear to have a significant role in promoting IL-6 and IL-12 proinflammatory cytokine production by infected BMDMs in an NF-κB independent manner.
## The 7a K119A substitution attenuates mouse-adapted SARS-CoV-2 infections
To determine whether findings in the rA59 virus background are similarly observed in SARS-CoV-2 infections, we introduced stop codons and K119A codon changes into a mouse-adapted (MA) SARS-CoV-2 bacterial artificial chromosomal plasmid (BACmid) using fragment assembly methods. Transfection of the resulting recombinant BACmids into Vero-ACE2-TMPRSS2 cells generated a set of three recombinant SARS-CoV-2 viruses: unaltered MA (MA-7a-WT), MA-7a-NULL, and MA-7a-KRATE. These viruses were used to infect susceptible BALB/c mice. In each of the three infections, weight losses were similar, although some recovery from MA-7a-KRATE infections was evident at 6-7 dpi (Fig. 6a). Both the elimination of the 7a expression (MA-7a-NULL) and the 7a-K119A substitution (MA-7a-KRATE) significantly attenuated the infections as measured by survival scores (Fig. 6b). This attenuation accorded with viral RNA levels in lungs at 3 dpi; mice infected with MA-7a-KRATE had two-to sixfold less RNA accumulation than MA-7a-WT (Fig. 6c). Correspondingly, MA-7a-NULL and MA-7a-KRATE viral titers were significantly lower at 3 dpi (Fig. 6d). These findings further demonstrate that residue K119 in the 7a cytoplasmic tail supports coronavirus infection and validates the results observed in an rA59 virus background.
## DISCUSSION
Severe COVID-19 is often correlated with a dysregulated release of proinflammatory cytokines. Patients have elevated levels of IL-6, IL-12, IL-1β, and other proinflammatory cytokines that contribute to infection clearance but at high levels cause severe, even fatal disease (67). SARS-CoV-2-infected macrophages have been implicated as key drivers of inflammatory cytokines and resultant severe COVID-19 (54)(55)(56). However, consistent in vitro model systems for SARS-CoV-2 macrophage infection have yet to be established (50)(51)(52)(53), as was the case for SARS-CoV (69,70). MHV productively infects BMDMs ex vivo due to their high expression of the MHV receptor murine carcinoembryonic antigen-rela ted cell adhesion molecule 1 (CEACAM1) (42,56). This BMDM susceptibility provides a premise to use recombinant MHVs as vectors to express SARS-CoV-2 genes in betacor onavirus infection contexts. We constructed recombinant MHVs expressing the SARS-CoV-2 accessory gene 7a and infected murine BMDMs. We discovered that SARS-CoV-2 protein 7a was a likely contributor to the elevated proinflammatory response that occurs during infection. Recombinant MHV A59-7a infections elicited IL-6 and IL-12 production in BMDMs. Cytokine release was linked to virus secretion through an unknown mecha nism requiring native 7a cytoplasmic tails, specifically 7a tail residue K119. It remains unclear how 7a cytoplasmic tails manipulate the host cell to promote virus secretion and a proinflammatory response, but the mechanism appears to be independent of the COPI coatomer complex and the NF-κB pathway.
Utilizing recombinant MHVs to infect BMDMs also revealed a role for SARS-CoV-2 accessory protein 7a in promoting viral replication and output. In both murine BMDMs and DBT cells, 7a proteins accelerated the kinetics of viral replication (Fig. 3a andf) and significantly increased secreted virus levels (Fig. 3c andh). These 7a-dependent proviral effects were independent of a reported 7a-mediated interferon antagonizing activity (30)(31)(32). Indeed, in our tests, 7a did not antagonize antiviral interferon responses (Fig. 2). The proviral effects of 7a were diminished by substituting cytoplasmic tail lysine residues, particularly K119, whose exchange to A reduced viral titers. The mechanisms by which the cytoplasmic tails elicit proviral effects remain unknown. Substitutions in SARS-CoV-2 7a cytoplasmic tails have been shown to relocalize host factors, particularly to affect factors influencing MHC-I localization ( 20), but we did not find that mutant 7a proteins were significantly relocalized in infected cells (Fig. 4h). Relocalized 7a may not have been detectable due to MHV infection-induced disruption of the secretory pathway; note, for example, the faint, scattered Golgin-97 signal in SIM images (Fig. 4h). 7a proteins may interfere with cellular autophagy (23,24). As coronaviruses are thought to egress through an autophagosome-lysosome pathway (71), 7a may facilitate virus flow through secretory autophagosomes. Similar proviral effects have been reported for SARS-CoV-2 accessory protein 3a, which manipulated autophagy and modulated intracellular trafficking through its cytoplasmic tail to promote viral output (14,24,72,73). Future studies are aimed at determining whether 7a promotes replication by modifying the autophagosomal organelles supporting coronavirus replication and assembled virus secretion (74,75).
This study establishes the utility of a recombinant MHV system to analyze individual SARS-CoV-2 proteins. Genes and proteins of pathogenic human coronaviruses can be evaluated at BSL-2 when placed in the MHV background. This was needed for the 7a study, as the proviral role for 7a required analyses in the complete viral infection context. The MHV system addresses functional redundancies among sarbecovirus accessory proteins; notably, the deletion of ORF7a in SARS-CoV-2 generated very minor effects on viral replication (4,8,9), possibly due to compensating SARS-CoV-2 gene products (14,24,72,73). MHV only encodes three accessory proteins, ORF2a, ORF4, and ORF5a, which lack homology with SARS-CoV-2 accessory proteins (35)(36)(37). The distinctly orthologous set of accessory proteins in MHV permits it to be used as a controlled recombinant virus platform for isolating the effects of particular SARS-CoV-2 accessory proteins.
The non-essential SARS-CoV-2 accessory proteins impact virulence by controlling innate immune responses and by promoting particular stages of virus production (1, 2, 4-6, 14, 15, 31). This study highlighted these attributes of accessory proteins by revealing a proviral effect of SARS-CoV-2 7a during infection of primary innate immune cells and mice. Specifically, 7a promoted viral output and proinflammatory responses through a process requiring native 7a cytoplasmic tails. The importance of the native 7a cytoplas mic tail was further recognized in SARS-CoV-2 infection contexts. These effects on virus production require additional investigation because they appear to be unrelated to activities previously assigned to the 7a cytoplasmic tails, including coatomer-mediated protein trafficking and suppression of interferon signaling (20,30). Additional studies will also be necessary to dissect the proviral mechanisms by which SARS-CoV-2 accessory protein 7a contributes to the cytokine storms observed in severe cases of COVID-19.
## MATERIALS AND METHODS
## Cells
HEK293T cells (obtained from Edward M. Campbell, Loyola University Chicago) were cultured in Dulbecco's modified Eagle medium (DMEM) (#10-013-CV, Corning) supple mented with 0.1 mM non-essential amino acids (Gibco), 2 mM L-glutamine (Cytiva), 10 mM HEPES (Corning), 100 U/mL penicillin (Cytiva), 100 µg/mL streptomycin (Cytiva), and 10% fetal bovine serum (FBS) (Bio-techne). Mouse DBT cells (obtained from Susan Baker, Loyola University Chicago) were cultured in minimum essential medium (MEM) Alpha Medium (#11900-024, Gibco) supplemented with 2.2 g/L NaHCO 3 (Fisher Chemical), 10% Tryptose Phosphate Broth (BD Biosciences), 2 mM L-Glutamine (Cytiva), 100 U/mL Penicillin (Cytiva), 100 µg/mL Streptomycin (Cytiva), and 5% FBS (Bio-techne). 17Cl-1 mouse fibroblast cells (obtained from Susan Baker, Loyola University Chicago) were cultured in DMEM (#10-013-CV, Corning) supplemented with 2 mM L-Glutamine (Cytiva), 5% Tryptose Phosphate Broth (BD Biosciences), 100 U/mL Penicillin (Cytiva), 100 µg/mL Streptomycin (Cytiva), and 5% FBS (Bio-techne). The mouse macrophage cell line, NR-9456, (obtained from BEI Resources) was cultured in DMEM (#10-013-CV, Corning) supplemented with 0.1 mM non-essential amino acids (Gibco), 2 mM L-Gluta mine (Cytiva), 10 mM HEPES (Corning), 100 U/mL Penicillin (Cytiva), 100 µg/mL Strep tomycin (Cytiva), and 10% FBS (Bio-techne). L929 mouse fibroblast cells (obtained from Andrew T. Ulijasz, Loyola University Chicago) were cultured in minimum essential medium Eagle (MEM) (#10-010-CV, Corning) supplemented with 2 mM L-glutamine (Cytiva), 1 mM sodium pyruvate (Corning), 10 mM HEPES (Corning), and 10% FBS (Bio-techne). All cell lines were cultured in a 5% CO 2 incubator at 37°C.
## BMDMs
Male and female black/B6 mouse bone marrow (obtained from Dorothy Sojka, Loyola University Chicago) was pooled then cultured in BMDM media, containing DMEM (#10-017-CV, Corning), 30% L929 cell supernatant, 2 mM L-glutamine (Cytiva), 1 mM sodium pyruvate (Corning), 10 mM HEPES (Corning), 100 U/mL penicillin (Cytiva), 100 µg/mL streptomycin (Cytiva), 3.5 × 10 -4 % 2-mercaptoethanol (Sigma), and 25% FBS (Bio-techne), to obtain differentiated BMDMs. Differentiated BMDMs were maintained in BMDM media (as described above).
## Recombinant virus
Recombinant MHV-A59 viruses were generated using the CPER method as described in Amarilla et al. and Torii et al. (43,44). Recombinant viral genomes were transfected (LipoD293 Transfectant Reagent, SignaGen) into HEK293T cells and co-cultured with mouse DBT cells. Upon observation of cytopathic effects (CPE), cell supernatant was collected and clarified (centrifugation at 300 × g for 10 min, followed by 3,000 × g for 10 min at 4°C) as the viral stock. Recombinant viruses were propagated and plaque purified on mouse DBT cells. The modified genomic sequences of all virus stocks were converted to cDNA via RT-PCR (SuperScript IV One-Step RT-PCR System, Invitrogen) and sequence verified by ACGT. Recombinant MHV sequences were aligned to the murine hepatitis virus strain A59 complete genome (GenBank accession no. AY700211). To obtain purified viral particles, the clarified viral supernatant was concentrated 100-fold through centrifugation by overlaying the viral supernatant onto a 20% (wt/wt) sucrose cushion and performing slow-speed pelleting (SW28; 7,500 rpm at 4°C for 24 h). The resulting pellet was resuspended in serum-free DMEM to 1/100 of the original medium volumes.
## Virus infections
Mouse DBT cells were infected with indicated viral strains at a multiplicity of infec tion (MOI) of 0.1, unless otherwise indicated, in serum-free DMEM media (#10-013-CV, Corning) for 1 h at 37°C in a 5% CO 2 incubator. After 1 h, the viral inoculum was rinsed off with two PBS rinses followed by the addition of serum-containing MEMs (as described above). Mouse BMDMs were infected with indicated viral strains at an MOI of 0.1, unless otherwise indicated, in serum-free DMEM media (#10-017-CV, Corning) for 1 h at 37°C in a 5% CO 2 incubator. After 1 h, the viral inoculum was rinsed off with two PBS rinses followed by the addition of serum-free DMEMs (#10-017-CV, Corning). Cell supernatant was collected and clarified (centrifugation at 300 × g for 10 min, followed by 3,000 × g for 10 min at 4°C) and used for plaque assay and cytometric bead array (CBA) analysis. Infected cells were utilized for Trizol (Ambion) RNA extraction or to generate cell lysates using 1% NP-40 lysis buffer (#AAJ60766AK, Thermo Scientific), containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, and 1:1,000 Protease Inhibitor Cocktail (#P1860, Sigma-Aldrich).
## Plaque assays
Virus infectivities were determined by plaque assay using mouse DBT cells as indicator cells. Cells were infected with 10-fold serial dilutions of viral samples for 1 h at 37°C, followed by overlaying with a 0.4% (wt/vol) Noble agar (BD Biosciences) and 1% FBS DMEM (#10-013-CV, Corning) mixture. Plates were incubated at 37°C for 48 h and fixed using 3.7% formaldehyde-PBS solution for 30 min. Viral plaques were visualized by staining with 0.1% crystal violet for 15 min.
## Antibodies
MHV S proteins were detected with the R3300 rabbit polyclonal antibody (PAb) that targets the S2 subunit (76). MHV S proteins were also detected using a mCEACAM-Fc construct that was previously generated (77). MHV N proteins were detected with murine monoclonal antibody (MAb) J3.1. MHV M proteins were detected with murine MAb J1.3. The J3.1 and J1.3 MAbs were a generous gift from John Fleming. SARS-CoV-2 7a proteins were detected with murine MAb clone 3C9 obtained from Invitrogen (#MA5-35944). MHV nsp3 proteins were detected using the anti-D3 rabbit PAb, which was a generous gift from the Susan Baker lab (78). The β-COPI subunits were detected with a rabbit PAb obtained from Abcam (#ab2899). Golgin-97 proteins were detected using a rabbit PAb obtained from Invitrogen (#PA5-30048). β-Actin subunits were detected using murine MAb clone AC-15 conjugated with peroxidase obtained from Sigma-Aldrich (#A3854).
## Western blots
Samples in SDS solubilizer (0.0625 M Tris-HCl [pH 6.8], 10% glycerol, 0.01% bromo phenol blue, 2% [wt/vol] SDS, and 2% 2-mercaptoethanol) were heated at 95°C for 5 min, electrophoresed through 4%-20% or 8%-16% Mini-PROTEAN TGX Precast Protein Gels (Bio-Rad), transferred to nitrocellulose membranes (Bio-Rad), blocked with TBST-M (25 mM Tris-HCl [pH 7.5], 140 mM NaCl, 27 mM KCl, 0.05% Tween 20, and 5% nonfat milk powder), and incubated with the indicated primary antibody. After incubation with the appropriate horseradish peroxidase-tagged secondary antibody and chemiluminescent substrate (Thermo Scientific), the blots were imaged and processed with a FluorChem E apparatus (ProteinSimple).
## RT-qPCR
RNA was isolated from mouse DBT cells or BMDMs using Trizol (Ambion) via phase separation. RT-qPCR was performed using Luna Universal One-Step RT-qPCR Kit (New England BioLabs) on a CFX Opus 96 Real-Time PCR System (Bio-Rad). qPCR primers are listed in Table 1. Cycle thresholds were normalized to that of the housekeeping gene β-actin by the following the equation ∆∆C T = ∆C T (gene of interest) -∆C T (β-actin -average) . ∆C T values were determined by the following equation: ∆C T = C T (gene of interest) -C T (β-actin) . All results are shown as a ratio to β-actin calculated as 2 -∆∆CT .
## Infection time course luminescence assay
Mouse DBT cells or BMDMs in their respective media (as described above), contain ing Nano-Glo Endurazine Live Cell Substrate (#N2570, Promega) as per manufacturer's instructions, were infected with the indicated viral strains at an MOI of 0.1 and incubated at 37°C in a 5% CO 2 incubator. Relative light unit readings were taken on a GloMax Explorer luminometer (Promega) at the indicated timepoints.
## Interferon treatment assay
Mouse DBT cells, NR-9456 mouse macrophages, or mouse BMDMs were pre-treated with 10-fold titrations of Universal Type I IFN (#11200, PBL Assay Science) in their respective media (as described above) for 8 h prior to infection. After the 8 h IFN treatment, cells were rinsed with PBS and infected as described above, unless otherwise indicated. Infected cells were incubated with Nano-Glo Endurazine Live Cell Substrate (#N2570, Promega), and the results were read on a GloMax Explorer luminometer (Promega) at the indicated timepoints. Results were depicted as fold change in signal relative to the 0 U/mL IFN condition.
## Immunofluorescence assay
Mouse fibroblast 17Cl-1 cells or mouse BMDMs were grown on 12 mm coverslips (#12-545-80, Fisher Scientific) and infected before being fixed with 3.7% paraformalde hyde in PBS at the indicated hpi times. After permeabilization using 0.2% Triton X-100 in PBS and subsequent blocking with 1% goat serum (Gibco) in PBS, proteins of interest were detected using the indicated primary antibodies followed by incubation with secondary antibody conjugated to Alexa 488, Alexa 594, or Alexa 647 fluorophores (Life Technologies). mCEACAM-Fc was directly conjugated with a CF488 fluorophore using the Mix-n-Stain CF488A Antibody Labeling Kit (#MX488AS20, Sigma-Aldrich). Cell mem branes were stained with wheat germ agglutinin, Alexa Fluor 647 conjugate (#W32466, Invitrogen). Nuclei were stained using Hoescht 33258 (Molecular Probes). Coverslips were mounted with Fluoro-Gel with Tris Buffer (#17985-10, Electron Microscopy Sciences) onto glass slides for imaging. Fluorescence signals were captured with a DeltaVision wide field fluorescence microscope (Applied Precision, GE) equipped with a digital camera (CoolSNAP HQ, Photometrics) and a 1.4-numerical aperture ×100 objective lens. Z-stack images were collected and deconvolved with SoftWoRx deconvolution software (Applied Precision, GE). High-resolution SIM images were captured with a Lattice SIM5 (Zeiss). Microscopy images were processed using Imaris (Bitplane).
## Immunoprecipitation assay
Mouse DBT cells were infected with indicated viral strains at an MOI of 0.1 in serum-free DMEMs (#10-013-CV, Corning) for 1 h at 37°C in a 5% CO 2 incubator. After 1 h, the viral inoculum was rinsed off with two PBS rinses followed by the addition of serum-contain ing MEMs (as described above). At 16 hpi, infected cells were lysed using 1% NP-40 lysis buffer (#AAJ60777AK, Thermo Scientific) containing 50 mM Tris-HCl (pH 7.4), 150 mM NaCl, 5 mM EDTA, and 1:1,000 Protease Inhibitor Cocktail (#P1860, Sigma-Aldrich). The cell lysates were utilized for co-immunoprecipitation assays using Protein G Dynabeads (#10003D, Invitrogen) as per manufacturer's instructions. Briefly, Protein G Dynabeads were incubated with 1-10 µg of the specified target antibody, followed by incubation of the bead-antibody complex with the indicated cell lysates. The target antigen was eluted from the bead-antibody-antigen complex using a 50 mM glycine, pH 2.8, elution buffer. The eluates were then used for Western blot analysis.
## Cytometric bead array
Infected mouse BMDM cell supernatant was collected and clarified (centrifugation at 300 × g for 10 min, followed by 3,000 × g for 10 min at 4°C) to remove cellular debris.
The CBA was performed using LEGENDplex Mouse Anti-Virus Response Panel (13-plex) with V-bottom Plate (BioLegend) according to the manufacturer's instructions. Bead fluorescence was measured with an LSR Fortessa cell analyzer (BD Biosciences) or Cytek's Aurora System (Cytek), and data were analyzed using the cloud-based LEGENDplex Data Analysis Software Suite (BioLegend).
## Recombinant SARS-CoV-2 viruses
BACmids containing the cDNA genome of mouse-adapted (MA30) SARS-CoV-2 (79) were PCRamplified into six fragments, each with 30 base-pair overlaps. Fragments were isolated by gel electrophoresis, excised and purified using PureLink PCR reagents, and assembled using Gibson assembly (#E2621L, NEBuilder HiFi DNA Assembly). Assembled products were transformed into NEB 10-beta Competent Escherichia coli (#C3019H). All bacterial cultures were grown at 20°C. Chloramphenicol-resistant colonies were selected after 2 days, further amplified in LB24 broth for 3 days, with BACmids then purified using NucleoBond Xtra Midi (TaKaRa Bio) reagents. SARS-CoV-2 recombinant MA30 BACmids with engineered mutations were verified by sequencing. The entire BACmids were fully sequenced. Amounts of each BACmid (1.5 µg) were transfected into 10 6 Vero-ACE2-TMPRSS2 (Vero-AT) cells using LipoD293 In Vitro Transfection Reagent (SignaGen). Cells were then observed for cytopathic effect. Cells were then frozen at -80°C when overt syncytial cytopathologies were evident, typically 3 days post-transfection. Secreted virus titers were determined by plaque development on Vero-AT indicator cells.
## Mouse infections
BALB/c mice (8-10 weeks old) were anesthetized with ketamine-xylazine and inoculated intranasally with 50 µL DMEM solution mixed with the indicated amounts of virus (n = 4-5 mice for each virus infection) (79). Weights and survival scores were recorded daily for seven consecutive days. In repeat infections, mice were euthanized at 3 dpi with isoflurane overexposure. Lungs were collected and homogenized in TRIzol (Invitrogen). After centrifuging the samples, total RNA extraction was performed following the manufacturer's protocol. cDNA was synthesized using M-MuLV reverse transcriptase (New England Biolabs). Then, samples were amplified by RT-qPCR using PowerUp SYBR Green Master Mix (Applied Biosystems). For data analysis, the expression of each gene was averaged from two technical replicates and normalized to Gapdh expression using the ΔΔCt method. The qPCR primer sequences are listed in the table below. The qPCR primer sequences for SARS-CoV-2 detection have been reported previously (80). For obtaining lung viral titers, at 3 dpi, mice were euthanized. Lungs were collected and homogenized in PBS. Samples were centrifuged at 1,000 rpm for 5 min. The supernatants were used for infection in plaque assays using VeroE6-TMPRSS2 indicator cells.
## Biosafety
In vitro work with infectious SARS-CoV-2 was completed at Loyola University Chicago in restricted access tissue culture rooms under negative air pressure and at BSL2+ containment. In vivo SARS-CoV-2 infection studies were completed in BSL2 laboratories at the University of Iowa.
## Statistical analysis
Statistical comparisons were made by the unpaired Student t-test, one-way ANOVA, two-way ANOVA, or log-rank test. Error bars indicate the standard error mean of the data. P values of less than 0.05 were considered statistically significant.
## References
1. Redondo, Zaldívar-López, Garrido et al. (2021) "SARS-CoV-2 accessory proteins in viral pathogenesis: knowns and unknowns" *Front Immunol*
2. Hurtado-Tamayo, Requena-Platek, Enjuanes et al. (2023) "Contribution to pathogenesis of accessory proteins of deadly human coronaviruses" *Front Cell Infect Microbiol*
3. Liu, Fung, Chong et al. (2014) "Accessory proteins of SARS-CoV and other coronaviruses" *Antiviral Res*
4. Silvas, Vasquez, Park et al. (2021) "Contribution of SARS-CoV-2 accessory proteins to viral pathogenicity in K18 human ACE2 transgenic mice" *J Virol*
5. Lin, Fu, Xiong et al. (2023) "Unconventional secretion of unglycosylated ORF8 is critical for the cytokine storm during SARS-CoV-2 infection" *PLoS Pathog*
6. Kehrer, Cupic, Ye et al. (2023) "Impact of SARS-CoV-2 ORF6 and its variant polymorphisms on host responses and viral pathogene sis" *Cell Host Microbe*
7. Nemudryi, Nemudraia, Wiegand et al. (2021) "SARS-CoV-2 genomic surveillance identifies naturally occurring truncation of ORF7a that limits immune suppression" *Cell Rep*
8. Nhu, Labroussaa, Ebert et al. (2020) "Rapid reconstruction of SARS-CoV-2 using a synthetic genomics platform" *Nature*
9. Hou, Okuda, Edwards et al. (2020) "SARS-CoV-2 reverse genetics reveals a variable infection gradient in the respiratory tract" *Cell*
10. Wong, Perlman (2022) "Immune dysregulation and immunopathol ogy induced by SARS-CoV-2 and related coronaviruses -are we our own worst enemy?" *Nat Rev Immunol*
11. Yoshimoto (2020) "The proteins of severe acute respiratory syndrome coronavirus-2 (SARS CoV-2 or n-COV19), the cause of COVID-19" *Protein J*
12. Zandi (2022) "ORF9c and ORF10 as accessory proteins of SARS-CoV-2 in immune evasion" *Nat Rev Immunol*
13. Liu, Zhang, Liu et al. (2022) "A live-attenuated SARS-CoV-2 vaccine candidate with accessory protein deletions" *Nat Commun*
14. Chen, Zheng, Sun et al. (2021) "ORF3a of SARS-CoV-2 promotes lysosomal exocytosis-mediated viral egress" *Dev Cell*
15. Miyamoto, Itoh, Suzuki et al. (2022) "SARS-CoV-2 ORF6 disrupts nucleocytoplasmic trafficking to advance viral replication" *Commun Biol*
16. Huang, Ito, Tseng et al. (2006) "Severe acute respiratory syndrome coronavirus 7a accessory protein is a viral structural protein" *J Virol*
17. Hänel, Willbold (2007) "SARS-CoV accessory protein 7a directly interacts with human LFA-1" *Biol Chem*
18. Nizamudeen, Xu, Karthik et al. (2021) "Structural assessment of SARS-CoV2 accessory protein ORF7a predicts LFA-1 and Mac-1 binding potential" *Biosci Rep*
19. Zhou, Huang, Zhou et al. (2021) "Structural insight reveals SARS-CoV-2 ORF7a as an immunomodulating factor for human CD14 + monocytes"
20. Arshad, Laurent-Rolle, Ahmed et al. (2023) "SARS-CoV-2 accessory proteins ORF7a and ORF3a use distinct mechanisms to down-regulate MHC-I surface expression" *Proc Natl Acad Sci*
21. Timilsina, Umthong, Ivey et al. (2022) "SARS-CoV-2 ORF7a potently inhibits the antiviral effect of the host factor SERINC5" *Nat Commun*
22. Martin-Sancho, Lewinski, Pache et al. (2021) "Functional landscape of SARS-CoV-2 cellular restriction" *Mol Cell*
23. Hou, Wang, Wang et al. (2023) "The ORF7a protein of SARS-CoV-2 initiates autophagy and limits autophagosome-lysosome fusion via degradation of SNAP29 to Full-Length Text Journal of Virology"
24. (1080) *Autophagy*
25. Hayn, Hirschenberger, Koepke et al. (2021) "Systematic functional analysis of SARS-CoV-2 proteins uncovers viral innate immune antagonists and remaining vulnerabilities" *Cell Rep*
26. Koepke, Hirschenberger, Hayn et al. (2021) "Manipulation of autophagy by SARS-CoV-2 proteins" *Autophagy*
27. Liu, Fu, Huang et al. (2022) "Ubiquitination of SARS-CoV-2 ORF7a prevents cell death induced by recruiting BclXL to activate ER stress" *Microbiol Spectr*
28. Nishitsuji, Iwahori, Ohmori et al. (2022) "Ubiquitination of SARS-CoV-2 NSP6 and ORF7a facilitates NF-κB activation" *mBio*
29. Su, Wang, Yoo (2021) "Activation of NF-κB and induction of proinflammatory cytokine expressions mediated by ORF7a protein of SARS-CoV-2" *Sci Rep*
30. Nelson, Pekosz, Lee et al. (2005) "Structure and intracellular targeting of the SARS-coronavirus Orf7a accessory protein" *Structure*
31. Cao, Xia, Rajsbaum et al. (2021) "Ubiquitination of SARS-CoV-2 ORF7a promotes antagonism of interferon response" *Cell Mol Immunol*
32. Xia, Cao, Xie et al. (2020) "Evasion of type I interferon by SARS-CoV-2" *Cell Rep*
33. García-García, Fernández-Rodríguez, Redondo et al. (2022) "Impairment of antiviral immune response and disruption of cellular functions by SARS-CoV-2 ORF7a and ORF7b"
34. Pewe, Zhou, Netland et al. (2005) "A severe acute respiratory syndromeassociated coronavirusspecific protein enhances virulence of an attenuated murine coronavirus" *J Virol*
35. Körner, Majjouti, Alcazar et al. (2020) "Of mice and men: the coronavirus MHV and mouse models as a translational approach to understand SARS-CoV-2" *Viruses*
36. Ontiveros, Kuo, Masters et al. (2001) "Inactivation of expression of gene 4 of mouse hepatitis virus strain JHM does not affect virulence in the murine CNS" *Virology*
37. Koetzner, Kuo, Goebel et al. (2010) "Accessory protein 5a is a major antagonist of the antiviral action of interferon against murine coronavirus" *J Virol*
38. Stokes, Chang, Chua et al. (2009) "Organspecific attenuation of murine hepatitis virus strain A59 by replacement of catalytic residues in the putative viral cyclic phosphodiesterase ns2" *J Virol*
39. Case, Li, Elliott et al. (2018) "Murine hepatitis virus nsp14 exoribonuclease activity is required for resistance to innate immunity" *J Virol*
40. Alhammad, Parthasarathy, Ghimire et al. (2023) "SARS-CoV-2 Mac1 is required for IFN antagonism and efficient virus replication in cell culture and in mice" *Proc Natl Acad Sci*
41. Deng, Hackbart, Mettelman et al. (2017) "Coronavirus nonstructural protein 15 mediates evasion of dsRNA sensors and limits apoptosis in macrophages" *Proc Natl Acad Sci*
42. Grunewald, Chen, Kuny et al. (2019) "The coronavirus macrodomain is required to prevent PARP-mediated inhibition of virus replication and enhancement of IFN expression" *PLoS Pathog*
43. Hirai, Ohtsuka, Ikeda et al. (2010) "Role of mouse hepatitis virus (MHV) receptor murine CEACAM1 in the resistance of mice to MHV infection: studies of mice with chimeric mCEACAM1a and mCEACAM1b" *J Virol*
44. Amarilla, Sng, Parry et al. (2021) "A versatile reverse genetics platform for SARS-CoV-2 and other positive-strand RNA viruses" *Nat Commun*
45. Torii, Ono, Suzuki et al. (2021) "Establishment of a reverse genetics system for SARS-CoV-2 using circular polymerase extension reaction" *Cell Rep*
46. De Haan, Van Genne, Stoop et al. (2003) "Coronaviruses as vectors: position dependence of foreign gene expression" *J Virol*
47. De Haan, Haijema, Boss et al. (2005) "Coronaviruses as vectors: stability of foreign gene expression" *J Virol*
48. Garlinghouse, Smith, Holford (1984) "The biological relationship of mouse hepatitis virus (MHV) strains and interferon: in vitro induction and sensitivities" *Arch Virol*
49. Martínez-Sobrido, Scott, García-Sastre et al. (2007) "Inhibition of the alpha/beta interferon response by mouse hepatitis virus at multiple levels" *J Virol*
50. Kindler, Gil-Cruz, Spanier et al. (2017) "Early endonuclease-mediated evasion of RNA sensing ensures efficient coronavirus replication" *PLoS Pathog*
51. Dalskov, Møhlenberg, Thyrsted et al. (2020) "SARS-CoV-2 evades immune detection in alveolar macrophages" *EMBO Rep*
52. García-Nicolás, 'kovski, Zettl et al. (2021) "No evidence for human monocyte-derived macrophage infection and antibody-mediated enhancement of SARS-CoV-2 infection" *Front Cell Infect Microbiol*
53. Munnur, Teo, Eggermont et al. (2021) "Altered ISGylation drives aberrant macrophage-dependent immune responses during SARS-CoV-2 infection" *Nat Immunol*
54. Wendisch, Dietrich, Stillfried et al. (2021) "SARS-CoV-2 infection triggers profibrotic macrophage responses and lung fibrosis" *Cell*
55. Grant, Morales-Nebreda, Markov et al. (2021) "Circuits between infected macrophages and T cells in SARS-CoV-2 pneumonia"
56. Sefik, Qu, Junqueira et al. (2022) "Inflammasome activation in infected macrophages drives COVID-19 pathology" *Nature*
57. Mazzarella, Santoro, Ravasio et al. (2023) "Inhibition of Full-Length Text Journal of Virology"
58. "the lysine demethylase LSD1 modulates the balance between inflammatory and antiviral responses against coronaviruses" *Sci Signal*
59. Fielding, Tan, Shuo et al. (2004) "Characterization of a unique groupspecific protein (U122) of the severe acute respiratory syndrome coronavirus" *J Virol*
60. Pekosz, Schaecher, Diamond et al. (2006) "Structure, expression, and intracellular localization of the SARS-CoV accessory proteins 7a and 7b" *Adv Exp Med Biol*
61. Knoops, Kikkert, Worm S Van Den et al. (2008) "SARS-coronavirus replication is supported by a reticulovesicular network of modified endoplasmic reticulum" *PLoS Biol*
62. Cortese, Lee, Cerikan et al. (2020) "Integrative imaging reveals SARS-CoV-2-induced reshaping of subcellular morphologies" *Cell Host Microbe*
63. Krijnse-Locker, Ericsson, Rottier et al. (1994) "Characteriza tion of the budding compartment of mouse hepatitis virus: evidence that transport from the RER to the Golgi complex requires only one vesicular transport step" *J Cell Biol*
64. De Haan, Rottier (2005) "Molecular interactions in the assembly of coronaviruses" *Adv Virus Res*
65. Klein, Winter, Wachsmuth-Melm et al. (2020) "SARS-CoV-2 structure and replication characterized by in situ cryo-electron tomography" *Nat Commun*
66. Dey, Singh, Khan et al. (2022) "An extended motif in the SARS-CoV-2 spike modulates binding and release of host coatomer in retrograde trafficking" *Commun Biol*
67. Cattin-Ortolá, Welch, Maslen et al. (2021) "Sequences in the cytoplasmic tail of SARS-CoV-2 Spike facilitate expression at the cell surface and syncytia formation" *Nat Commun*
68. Burman, Bourbonniere, Philie et al. (2008) "Scyl1, mutated in a recessive form of spinocerebel lar neurodegeneration, regulates COPI-mediated retrograde traffic" *J Biol Chem*
69. Diamond, Kanneganti (2022) "Innate immunity: the first line of defense against SARS-CoV-2" *Nat Immunol*
70. Nilsson-Payant, Uhl, Grimont et al. (2021) "The NF-κB transcriptional footprint is essential for SARS-CoV-2 replication" *J Virol*
71. Yilla, Harcourt, Hickman et al. (2005) "SARS-coronavirus replication in human peripheral monocytes/macrophages" *Virus Res*
72. Tseng, Perrone, Zhu et al. (2005) "Severe acute respiratory syndrome and the innate immune responses: modulation of effector cell function without productive infection" *J Immunol*
73. Ghosh, Dellibovi-Ragheb, Kerviel et al. "Altan-Bonnet N. 2020. βcoronaviruses use lysosomes for egress instead of the biosynthetic secretory pathway" *Cell*
74. Miao, Zhao, Li et al. (2021) "ORF3a of the COVID-19 virus SARS-CoV-2 blocks HOPS complexmediated assembly of the SNARE complex required for autolysosome formation" *Dev Cell*
75. Walia, Sharma, Paul et al. (2024) "SARS-CoV-2 virulence factor ORF3a blocks lysosome function by modulating TBC1D5-dependent Rab7 GTPase cycle" *Nat Commun*
76. Twu, Lee, Kim et al. (2021) "Contribution of autophagy machinery factors to HCV and SARS-CoV-2 replication organelle formation" *Cell Rep*
77. Shang, Zhuang, Zhang et al. (2021) "Inhibition of autophagy suppresses SARS-CoV-2 replication and ameliorates pneumonia in hACE2 transgenic mice and xenografted human lung tissues" *J Virol*
78. Gallagher, Parker, Buchmeier (1990) "Neutralization-resistant variants of a neurotropic coronavirus are generated by deletions within the amino-terminal half of the spike glycoprotein" *J Virol*
79. Gallagher (1997) "A role for naturally occurring variation of the murine coronavirus spike protein in stabilizing association with the cellular receptor" *J Virol*
80. Gosert, Kanjanahaluethai, Egger et al. (2002) "RNA replication of mouse hepatitis virus takes place at double-membrane vesicles" *J Virol*
81. Wong, Zheng, Wilhelmsen et al. (2022) "Eicosanoid signalling blockade protects middle-aged mice from severe COVID-19" *Nature*
82. Cheung, Yang, Honne et al. (2025) "SAMHD1 promotes SARS-CoV-2 infection by enhancing HNF1dependent ACE2 expression in lung epithelial cells" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607584&blobtype=pdf | # Nasal mucus-derived KLK13 restricts SARS-CoV-2 infection via proteolytic cleavage of spike
Wenying Cao, Ningze Zheng, Hehe Cao, Ran Chen, Jianheng Chen, Xueyi Deng, Hui Zhang, Shuofeng Yuan, Guigen Zhang
## Abstract
The epithelial cilia are the first line of defense against respiratory pathogens. For the first time, we found that Kallikrein-related peptidase 13 (KLK13), a serine protease expressed in airway ciliated epithelial cells with cell type specificity, was secreted into nasal mucus. KLK13 efficiently cleaved the spike of SARS-CoV-2, resulting in the inhibition of SARS-CoV-2 cell entry and spike protein-mediated cell-cell fusion. Recombinant KLK13 protease efficiently cleaved the spike protein as well as virus particles in vitro. Only KLK13, but not other members of the KLK family, specifically cleaved the spike proteins of SARS-CoV-2 as well as other coronaviruses. We also confirmed that endogenous KLK13 stimulated by CRISPR activation (CRISPRa) in A549 cells inhibited SARS-CoV-2 pseudovi rus entry. The mRNA level of KLK13 was stimulated by poly (I:C) in both A549 and HeLa cells, and its expression level was also increased in SARS-CoV-2-infected clinical samples. Recombinant adeno-associated virus packaged KLK13 (AAV-KLK13) reduced SARS-CoV-2 replication in a K18-ACE2 mouse model. Collectively, the nasal mucus-derived KLK13 functions as a scissor of coronaviruses and holds the potential to be further developed as a broad-spectrum antiviral against coronaviruses. IMPORTANCE Epithelial cilia directly come into contact with inhaled pathogens. The nasal mucus functions as a formidable barrier against penetration of viral particles. KLK13 is secreted into nasal mucus and efficiently cleaves the spike proteins across different coronavirus species. KLK13-mediated cleavage of spike inhibits SARS-CoV-2 entry and syncytium formation. Intranasally delivered KLK13 also restricts SARS-CoV-2 infection in vivo. The finding that KLK13 acts as a scissor of viral spike in nasal mucus paves the way for the development of new antivirals against respiratory viruses. KEYWORDS coronavirus, nasal mucus, restriction factor, KLK13 protease, spike protein E pithelial cilia are the first line of defense against respiratory viruses, such as influenza A viruses, human coronaviruses, and rhinoviruses. Respiratory viruses target the ciliated respiratory epithelial cells at the early stage of infections (1-5). Respiratory epithelial cells are covered by a gel-like layer of mucus, which contains a range of mucin glycoproteins secreted. The heavily glycosylated mucins form a formidable barrier against the penetration of respiratory virus particles. It has been reported that the components of mucus, such as mucins and defensins, restrict respiratory viral infections (6-8).Coronaviruses (CoVs) are positive-sense single-stranded RNA viruses that are classified into four genera: α-, β-, γ-, and δ-CoVs. In the past 20 years, there have already been three outbreaks caused by β-CoVs, including severe acute respiratory syndrome coronavirus (SARS-CoV) (9-11), Middle East respiratory syndrome coronavirus (MERS-CoV) (12, 13), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (14, 15). All these viruses placed a substantial burden on global public health, especially SARS-CoV-2, which has caused over seven million deaths. SARS-CoV-2 infection is initiated by
binding of the spike protein to its receptor, angiotensin-converting enzyme 2 (ACE2), on the cell surface. After attachment, additional cleavage at the S2′ site by cell surface proteases, such as transmembrane protease serine 2 (TMPRSS2), or by cathepsins in endosomes following endocytosis, releases the fusion peptide and mediates membrane fusion (16,17). Spike-mediated syncytium formation was observed both in vitro and ex vivo infected with SARS-CoV, MERS-CoV, or SARS-CoV-2 (18)(19)(20)(21)(22)(23). Syncytium formation in the lung tissues of infected patients contributes to SARS-CoV-2 pathogenesis and is a hallmark of severe COVID-19 pathology. Host restriction factors, such as IFN-induci ble lymphocyte antigen 6 complex locus E (LY6E), cholesterol 25-hydroxylase (CH25H), inhibited coronavirus infection-associated syncytia (24)(25)(26).
Kallikrein-related peptidases (KLKs), a subgroup of serine proteases with diverse physiological functions, have emerged as new regulators of viral infections in recent years. For instance, KLK1, KLK5, and KLK12 have been shown to cleave hemagglutinin proteins of influenza viruses and mediate viral activation (27)(28)(29). KLK8 proteolytically processes human papillomaviruses to facilitate viral entry into host cells (30). KLK13 was upregulated in varicella zoster virus-infected keratinocytes and HCoV-HKU1-infected human airway epithelial cells (31,32). A recent report claimed that KLK13 serves as a priming protease facilitating viral entry during HCoV-HKU1 infection (32). KLK13 is expressed in multiple human tissues and detected in several biological fluids, including cervicovaginal fluid, seminal plasma, and saliva (33,34). For the first time, we found that KLK13 was secreted into nasal mucus. KLK13 cleaves the SARS-CoV-2 spike protein and antagonizes SARS-CoV-2 infection. The discovery of KLK13 as a scissor of viral spike gave new insights for the development of novel antiviral strategies.
## RESULTS
## KLK13 specifically cleaves the spike of SARS-CoV-2
By analyzing the public data available from the Human Protein Atlas (https://www.pro teinatlas.org/humanproteome/tissue+cell+type/lung), we found that several membrane proteases were expressed in ciliated epithelial cells with high cell type specificity, including the serine proteases TMPRSS4, PRSS12, PCSK4, PLG, KLK12, and KLK13 as well as metalloprotease CLCA2 (Fig. 1A andB). The expression levels of these genes in ciliated respiratory cells were further evaluated using single-cell sequencing data (Fig. S1A). To investigate whether these proteases are involved in SARS-CoV-2 spike cleavage, each of them was cotransfected with SARS-CoV-2 spike expression vector (Wuhan-Hu-1 strain) in HEK293T cells. Among them, PCSK4 and TMPRSS4 dramatically reduced the production of the S2 subunit. In the presence of PLG or TMPRSS2, the spike protein was cleaved, and additional bands were observed, while CLCA2, PRSS12, and KLK12 (a short form) had no such effect. Besides TMPRSS2, TMPRSS4, and plasminogen (PLG) were reported to cleave the SARS-CoV-2 spike (35,36), which is consistent with our findings. The role of PCSK4 in SARS-CoV-2 spike protein processing and maturation is under investigation. Immunoblotting analysis showed that, in the presence of KLK13, two faster-migrating fragments of the spike protein were detected (Fig. 1C). One fragment was close to, but smaller than, S2, and the other one was slightly larger than 45 kDa. These results indicated that KLK13 cleaves the spike of SARS-CoV-2. Similarly, KLK13 also cleaved the spike of SARS-CoV-2 Omicron BA.1 strain (Fig. S1B). We further investigated whether the other members of the KLK family are able to cleave the spike protein. Strikingly, only KLK13, but not KLK10, KLK11, KLK12, or KLK14, can cleave the spike of SARS-CoV-2 (Fig. 1D; Fig. S1C). To evaluate whether the serine protease activity of KLK13 is necessary for the cleavage of spike, a catalytically inactive mutant KLK13 S218A was constructed (37), which was not able to cleave the SARS-CoV-2 spike protein (Fig. 1E through F). Addition ally, to analyze whether the cleavage of spike protein by KLK13 is dose-dependent, we co-transfected the spike expression vector with different amounts of KLK13 expression vector in HEK293T cells. As shown in Fig. S2A, the cleaved fragments of spike (above 75 kDa) are increased, along with the increase of KLK13, which confirms that KLK13
## KLK13 is interferon-stimulated and secreted into nasal mucus
We analyzed the single-cell sequencing data from both COVID-19 patients and healthy individuals. The expression of KLK13 in SARS-CoV-2-infected individuals was upregulated compared with that in non-infected individuals (Fig. 2A). To understand whether KLK13 is an interferon-stimulated gene (ISG), we stimulated both A549 and HeLa cells with poly(I:C), an analog of dsRNA that mimics the molecular pattern associated with viral infection. As shown in Fig. 2B andC, similar to that of IFIT3, the mRNA level of KLK13 was significantly increased in both A549 and HeLa cells upon poly(I:C) stimulation. In addition, the promoter region of KLK13 was predicted to harbor a putative IFN-γ activation site (Fig. 2D). The IFN-γ activation sites in the promoter regions of GBP, Ly-6E, and FcγR1 were shown in parallel (38). These results indicated that KLK13 is an ISG and upregulated upon SARS-CoV-2 infection. Moreover, since KLK13 was reported to be secreted into several body fluids (33,34), we therefore tested whether it exists in the nasal mucus. As shown in Fig. 2E, KLK13 was detected in the nasal mucus of both healthy and COVID-19 volunteers, in line with the previous nasal mucus proteomics studies (39)(40)(41). We further confirmed that KLK13 was detected in the supernatants of HEK293T cells transfected with KLK13 cDNA (Fig. 2F). SARS-CoV-2 NSP2 was previously reported to be secreted (42) and was thereby used here as a positive control. The results confirmed that KLK13 is interferon stimulated and can be secreted into nasal mucus.
## Recombinant KLK13 cleaves the spike protein both in vitro and in vivo
To investigate whether KLK13 directly cleaves spike protein in vitro, we incubated the purified SARS-CoV-2 (WH strain) spike protein (S1+S2 ECD) with recombinant KLK13 in vitro. As shown in Fig. 3A, recombinant KLK13 protease was able to cleave the recombinant spike protein. Since the recombinant spike protein (S1+S2 ECD) consists of 1,193 amino acids, without the transmembrane domain (TMD) and cytoplasmic tail, the cleaved fragments are slightly smaller than those in the full-length spike. We further analyzed whether the recombinant KLK13 protease is capable of cleaving the spikes on the virus particles. The replication-competent SARS-CoV-2 virus-like particles (SARS-CoV-2 GFP/ΔN trVLP) were produced and propagated in the Caco-2-N packaging cells (43). The SARS-CoV-2 virus-like particles were incubated with recombinant KLK13 protease in vitro. Indeed, KLK13 protease also cleaved the spikes on the virions (Fig. 3B).
Moreover, we checked the cleavage activity of KLK13 in the context of virus infection. HEK293T-ACE2 cells overexpressing KLK13 were infected with the authentic SARS-CoV-2 virus (Omicron BA.5 strain). Immunoblotting analysis showed that the spike protein was cleaved in SARS-CoV-2-infected cells, as evidenced by the two characteristic bands (Fig. 3C). These results confirmed that KLK13 cleaves both the recombinant spike protein and the spikes on the virions in an enzymatic activity-dependent manner. overexpressed in the cells. Among them, KLK12 was derived from mRNA transcript variant 5 (NM_001370126.1), which encoded a protein with 144 amino acids in length, a short form of KLK12. This experiment was performed in three biological replicates. (D) HEK293T cells were cotransfected with plasmids expressing SARS-CoV-2 (WH strain) spike together with HA-tagged KLK10, KLK11, KLK12, KLK13, KLK14, or empty vector. The cells were lysed, followed by immunoblotting analysis. The triangles indicate different members of the KLK family. HEK293T cells were cotransfected with SARS-CoV-2 WH strain (E) or Omicron (BA.1) (F) spike plasmid together with HA-tagged wild-type KLK13 or KLK13 S218A mutant. Forty-eight hours later, the cells were lysed and subjected to immunoblotting
## KLK13 cleaves the spikes of coronaviruses across different species
KLK13 hydrolyzed the substrates at basic residues with higher efficiency for Arg (R) and Lys (K) (44). The SARS-CoV-2 spike protein possesses a polybasic cleavage motif, RRAR, and is further processed to the S1 and S2 subunits (22). We generated a spike mutant with the RRAR motif deleted (ΔRRAR) (Fig. S2B). In the presence of KLK13, the upper cleaved fragment close to S2 was not detectable when the RRAR motif was deleted, while the shorter fragment, approximately 45 kDa, was still observed (Fig. S2C). These results indicated that deletion of RRAR affected, but did not completely abolish, KLK13-mediated cleavage of spike.
We further investigated whether KLK13 has a proteolytic effect on the spikes of other coronaviruses. We generated expression vectors for the spikes of different coronaviruses, including SARS-CoV-1, MERS-CoV, HCoV-HKU1, HCoV-229E, and HCoV-OC43. As shown in Fig. 3D, KLK13 cleaved the spike of SARS-CoV-1. The pattern of cleaved fragments is similar to that of SARS-CoV-2 spike. KLK13 efficiently cleaved the spike of MERS-CoV, resulting in two cleaved fragments, which are close to 65 kDa and 35 kDa (Fig. 3D). Strikingly, KLK13 also cleaved the spikes of HCoV-HKU1, HCoV-229E, and HCoV-OC43, although the patterns of cleaved fragments differ from each other. We then investigated whether KLK13 cleaves the glycoproteins of other enveloped viruses. As shown in Fig. 3E, KLK13 was not able to cleave the glycoprotein of vesicular stomatitis virus (VSV). These results indicated that KLK13 is able to cleave the spikes of different coronaviruses, although the cleavage sites are distinctive.
## KLK13 interacts with spike protein and inhibits spike-mediated syncytium formation
We further performed the co-immunoprecipitation assay to evaluate the protein-protein interaction between KLK13 and SARS-CoV-2 spike. As shown in Fig. 4A, the spike protein was coprecipitated with KLK13. We also performed a complementary co-immunoprecipi tation experiment by using IgG as control (Fig. S3A). Consistently, the spike protein was specifically coprecipitated with KLK13. These were also supported by an immunofluores cence assay that KLK13 was colocalized with the spike protein (Fig. 4B andC).
The interaction between the SARS-CoV-2 spike and its receptor ACE2 induces syncytium formation (45)(46)(47)(48). We therefore performed a syncytium formation assay to investigate whether KLK13 inhibits SARS-CoV-2-mediated cell-cell fusion (49). We transfected spike and KLK13 vectors into HEK293T-mCherry cells (effector cells). The effector cells were then mixed and co-cultured with HEK293T-ACE2/GFP cells (target cells) (Fig. 4D). As shown in Fig. 4E andF, the SARS-CoV-2 spike protein (WH strain) efficiently induced syncytium formation (top panel), which is consistent with the previous study (45). In the presence of KLK13, the spike-mediated syncytium formation was significantly inhibited (middle panel), while KLK13 S218A had no such inhibitory effect (bottom panel). Similarly, KLK13 also blocked SARS-CoV-2 Omicron BA.1 spike-mediated syncytium formation (Fig. S2D andE). Furthermore, we performed a second type of syncytium formation assay by transfecting SARS-CoV-2 spike protein (Wuhan strain or Omicron BA.1 strain), together with KLK13-WT or KLK13 S218A expression vectors in HEK293T-ACE2-GFP cells. Consistent with the findings shown in Fig. 4E and F and Fig. S2D and E, KLK13 WT, but not KLK13 S218A , extensively inhibited SARS-CoV-2 spike-induced syncytium formation (Fig. S3B-C). These results confirmed that KLK13 significantly inhibited spike-mediated syncytium formation. the JASPAR database; a threshold of 0.9 matrix similarity score was applied. (E) The nasal mucus collected from two healthy adult individuals and four COVID-19 patients was individually mixed with 2× SDS lysis buffer. The lysates were analyzed by immunoblotting with KLK13-, nucleocapsid-, and GAPDH-specific antibodies, while cell culture medium was used as a control. (F) HEK293T cells were transfected with KLK13-HA, SARS-CoV-2 NSP8-Strep, NSP2-Strep, or empty vector, respectively. The culture supernatants and cell lysates were harvested and prepared for immunoblotting analysis. spike was cotransfected, together with KLK13-HA, KLK13 S218A mutant, or empty vector into HEK293T-mCherry cells for 5 h. The cells were digested and mixed with HEK293T cells expressing ACE2 and GFP at a ratio of 1:1. After 36 h of coculture, the cells were fixed and stained with DAPI. Immunofluorescence microscopy was used to capture fluorescence and bright field images. The syncytium was depicted by merged red-green color and circled by dashed lines. (F) Quantification of the syncytia is shown in (E). Syncytium number was pooled from eight microscope fields for each experiment (n = 8). Student's t-test was used to perform statistical analysis. ***P < 0.001. These experiments were performed in two biological replicates.
## Endogenous and secreted KLK13 inhibit SARS-CoV-2 entry
We analyzed whether the cleavage of spike protein by KLK13 inhibits SARS-CoV-2 entry. HEK293T-ACE2 cells expressing KLK13 or vector control were transduced with pseudo typed lentiviral particles (WH strain or Omicron BA.1). As presented by the pseudovi rus-based entry assays, KLK13-expressing cells showed a lower level of infection than control cells (Fig. 5A andB). Consistently, KLK13 also inhibited SARS-CoV-1 pseudovirus entry (Fig. 5C), but had no effect on the entry of VSV pseudovirus (Fig. 5D). In addition, we investigated whether the inhibition of KLK13 on SARS-CoV-2 pseudovirus entry is dose dependent. Different amounts of KLK13 expression vector were transfected into HEK293T-ACE2 cells. The entry of SARS-CoV-2 pseudovirus decreased, along with the increase of KLK13 (Fig. 5E).
We further analyzed whether endogenous KLK13 plays an inhibitory role in SARS-CoV-2 pseudovirus entry. Since endogenous KLK13 was not detected in different cells, including BEAS-2B, HEK293T, A549, Caco-2, and Calu-3 cells (Fig. S4A andB), we exploited the CRISPR/Cas9-mediated gene activation (referred to as CRISPR activation, CRISPRa) technique to activate the expression of endogenous KLK13 in A549-hACE2-TMPRSS2 cells (50). In this experiment, a CRISPR-dCas9-VPR activation system was used, in which two specific sgRNAs were designed to target the KLK13 gene promoter region, while a scrambled sgRNA was used as a negative control. As shown in Fig. 5F, compared to scrambled sgRNA, sgRNA-2 stimulated KLK13 mRNA expression, whereas sgRNA-1 showed no such effect, which is also considered a control. We then performed the pseudovirus-based entry assay to assess the effect of endogenous KLK13 on viral entry. Consistently, endogenous KLK13 stimulated by CRISPRa inhibited SARS-CoV-2 pseudovi rus entry (Fig. 5G).
To study whether secreted KLK13 inhibits SARS-CoV-2 pseudovirus entry, recombi nant KLK13 or KLK13 S218A protein was concentrated from cell culture supernatants and incubated with pseudotyped lentiviral particles (Fig. 5H). Recombinant KLK13 protein from culture supernatants dose dependently inhibited SARS-CoV-2 pseudovirus entry, while KLK13 S218A had no such inhibitory effect (Fig. 5I andJ). To further verify if it is the secreted KLK13 that inhibits pseudovirus entry, we designed and synthesized a specific KLK13 peptide inhibitor (Biotin-VRFR-CMK) based on a previous study (51). KLK13-con taining supernatants were incubated with KLK13 inhibitor and SARS-CoV-2 pseudovi ruses (Fig. S4C). As shown in Fig. S4D, KLK13 inhibitor dose dependently reversed the suppression of KLK13 on SARS-CoV-2 pseudovirus entry. These findings confirm that secreted KLK13 specifically suppresses SARS-CoV-2 entry.
## KLK13 restricts SARS-CoV-2 infection in vivo
KLK13 was expressed in ciliated epithelial cells with cell type specificity. To verify whether KLK13 restricts SARS-CoV-2 infection in vivo, we intranasally infected K18-hACE2 transgenic mice with rAAV-KLK13 or rAAV-EGFP (Fig. 6A). Thirty days before SARS-CoV-2 infection, the K18-hACE2 mice were intranasally infected with rAAV-KLK13 or rAAV-EGFP. Each was intranasally challenged with SARS-CoV-2 (HKU-001a strain) on day 0. The animals were sacrificed at 2 dpi (Fig. 6A). The viral burden in the nasal turbinate and lung was measured by plaque assays (Fig. 6B). The results showed that the titer in the lungs of KLK13-expressing mice was reduced, as well as in the nasal turbinate. Moreover, the pathology induced by SARS-CoV-2 in the nasal turbinate and lungs was attenuated by KLK13 (Fig. 6C). In the KLK13 group, exudation and inflammatory cell infiltration in the nasal turbinate were decreased, while consolidation and immune cells were rarely seen in the lung. Histology scores of lung tissues in different groups were also shown (Fig. 6D). These results confirmed that KLK13 restricts SARS-CoV-2 infection in mice. To further investigate the role of KLK13 in viral infection in vitro, we overexpressed KLK13 in A549-hACE2-TMPRSS2 cells. Overexpression of KLK13 reduced viral RNA copies in the supernatants (Fig. 6E, left panel), although the viral RNA copies in the cell lysates were not significantly changed (Fig. 6E, right panel). We also performed a time-course infection experiment by using a transcription-and replication-competent SARS-CoV-2 virus-like particle system (SARS-CoV-2 GFP/ΔN trVLP), which enables the complete viral life cycle in HEK293T-ACE2 cells expressing the SARS-CoV-2 nucleocapsid protein (43). Consistently, KLK13-WT, but not KLK13 S218A , robustly inhibited viral RNA replication 24, 48, and 72 hpi, indicated by the levels of both SARS-CoV-2 genomic and subgenomic RNA (Fig. 6F). Here, the antiviral effect of KLK13 in A549 cells is not as strong as that in HEK293T cells, because the transfection efficiency of KLK13 expression vector in A549 cells is much lower compared to that in HEK293T cells. These results demonstrate that KLK13 exerts antiviral activity against SARS-CoV-2 both in vivo and in vitro.
## DISCUSSION
Ciliated epithelium is present in the nasal and sinus cavities, as well as the proximal and distal conducting airways. Notably, ciliated epithelial cells are the major targets of respiratory viruses at the early stage of infections. For instance, influenza viruses specifically infected ciliated cells in an in vitro model of human ciliated airway epithelia (HAE) (2,3). Human respiratory syncytial virus (RSV), which causes serious pediatric respiratory disease worldwide, preferentially infects the ciliated cells of the airway epithelium via the apical surface (1). SARS-CoV-1 only infects human HAE derived from nasal and tracheobronchial regions via the apical surface, but not undifferentiated primary epithelial cells (2). The respiratory epithelial cells are covered by a gel-like layer of mucus, which protects against inhaled pathogens. We confirmed that KLK13 was secreted in the nasal mucus of both healthy individuals and COVID-19 patients. MUC5AC and MUC5B, two major secreted airway mucins, are the major components that constitute the airway mucus layer. MUC1, MUC5AC, and MUC5B are known to protect against respiratory viral infections at the respiratory surface (7,(52)(53)(54). It seems that the secretion of antiviral components into the mucus is a common strategy to protect against respiratory viral infections. In this study, for the first time, we reported that KLK13 is a new antiviral component secreted into nasal mucus. KLK13 acts as a scissor and cleaves the spikes of different coronaviruses and protects against SARS-CoV-2 infection in vivo. Previously, it was reported that TMPRSS4 and plasminogen (PLG) can cleave the SARS-CoV-2 spike (35,36), which is consistent with our findings (Fig. 1C). Besides, we found that PCSK4 dramatically reduced the production of S1/S2 subunits (Fig. 1C), and the underlying mechanism of which is under investigation.
After cleavage by KLK13, two cleaved fragments of SARS-CoV-2 spike were observed. KLK13 also cleaves the spike of SARS-CoV-1. The pattern of cleaved fragments of SARS-CoV-1 spike is similar to that of SARS-CoV-2. There were two possible cleavage sites in the spike of SARS-CoV-1 and SARS-CoV-2. The upper fragment was not detected when the RRAR motif of spike was deleted in the presence of KLK13 (Fig. S2C). Moreover, KLK13 also cleaved the spikes of MERS-CoV, HCoV-HKU1, HCoV-229E, and HCoV-OC43, although the pattern of cleaved fragments differs from each other. The cleavage sites in these spikes need to be further investigated. It has been reported that KLK13 could cleave the spike of HCoV-HKU1 and serve as a priming protease during HCoV-HKU1 infection (32). It seems that after cleavage by KLK13, the spikes of coronaviruses can either be primed or inactivated. A recent study reported that KLK13, together with KLK12, enhanced SARS-CoV-2 replication; however, the individual role of KLK13 was not fully investigated (55). In our study, both types of syncytium formation assays analysis was conducted using a paired Student's t-test, *P < 0.05, **P < 0.01, ***P < 0.001. (n = 3). (G) The cells described in (F) were then transduced with pseudotyped lentiviral particles. The entry of pseudovirus was quantified through measuring the luciferase activity (n = 4). This experiment was performed in three biological replicates. (H) Schematic representation of pseudovirus entry assay with KLK13-containing supernatants. (I) HEK293T cells were transfected with KLK13-WT or KLK13 S218A expression vector. The supernatants were collected and concentrated using ultrafiltration tubes. Pseudotyped lentiviral particles were pre-incubated with KLK13 WT-or KLK13 S218A -containing supernatants at different concentrations at 37℃ overnight. The KLK13-treated pseudoviruses were used to perform the pseudovirus-based entry assays. The luciferase activity was measured (n = 3). (J) The concentrated supernatants used in (I) were blotted with specific antibodies. This experiment was performed in three biological replicates. confirmed that KLK13 WT, but not KLK13 S218A mutant, efficiently restricted SARS-CoV-2 spike--mediated cell-cell membrane fusion, which indicates that KLK13 could not be an activator of the SARS-CoV-2 spike protein. The roles of KLK13 in other coronaviral infections, such as MERS-CoV, HCoV-229E, and HCoV-OC43, will be further evaluated both in vitro and in vivo. The expression of KLK13 mRNA in various cell lines is very low or undetectable based on analyzing the data from the Human Protein Atlas database (Fig. S4A). KLK13 was not detected in different cells by Western blotting, including A549, Caco-2, BEAS-2B, and Calu-3 cells, although different KLK13 antibodies were used (Fig. S4B). We also performed the mRNA-seq experiment in differentiated primary human nasal epithelial cells under ALI conditions. The mRNA level of KLK13 was also not high in the differentiated epithelial cells (data not shown). Thereby, we could not perform KLK13 knockdown or knockout experiments in these cells. Instead, we exploited the CRISPR activation (CRISPRa) technique to stimulate endogenous KLK13 expression in A549-hACE2-TMPRSS2 cells. Previously, we have successfully performed the CRISPR inhibition (CRISPRi) experiments (56). Two sgRNAs targeting the KLK13 gene promoter region were used in this experiment. Scrambled sgRNA was used as a negative control.
## Research
As shown in Fig. 5G, stimulation of endogenous KLK13 by CRISPR-dCas9-VPR/sgRNA2 inhibits SARS-CoV-2 pseudovirus entry. Human KLK13 harbors a SNP (rs34089525) at amino acid position 109, resulting in a His (H) to Tyr (Y) substitution. This SNP is present at 3% frequency in the European populations and 2.4% frequency in the Latin American population, but is nearly absent in Asian populations (Fig. S5A). We introduced this H109Y variant into a KLK13 expression construct. As shown in Fig. S5B andC, both KLK13 WT and KLK13 H109Y are expressed at similar levels, while the cleavage efficiency of KLK13 H109Y toward spike is slightly reduced compared with that of KLK13 WT, suggesting that the naturally occurring variants in KLK13 may possibly affect SARS-CoV-2 infectivity. Currently, we have no epidemiological evidence indicating that the KLK13 H109Y mutation is associated with the disease severity of COVID-19 patients. Whether this missense mutation is associated with the severity of respiratory infectious diseases in these populations needs further investigation. KLK13 is not sufficient to control virus spread and transmission in the human population. However, the presence of KLK13 in the nasal mucus likely contributes to reducing disease severity via restricting virus infection. Nevertheless, we report that KLK13, a novel restriction factor, is secreted into nasal mucus and inhibits SARS-CoV-2 infection via cleavage of the spike protein (Fig. S5D).
## MATERIALS AND METHODS
## Cell culture
HEK293T cells were purchased from American Type Culture Collection (ATCC). HeLa, BEAS-2B, Caco-2, and A549 cells were obtained from the Cell Bank of Shanghai Institute The pathological changes of lung and NT were presented. In this experiment, the naïve K18-hACE2 mice were used as controls. Consolidation and immune cell infiltration in the lung were marked by asterisk or arrow, respectively. Exudation and inflammatory cell infiltration in NT were denoted by an arrow.
Representative images from five mice in each group were shown at 10× magnification (left, scale bar = 100 µm) and 20× magnification (right, scale bar = 50 µm). (D) Histology scores of lung tissues in different groups. Data are mean ± SD. **P < 0.01. (E) A549-hACE2-TMPRSS2 cells were transfected with KLK13-WT, KLK13 S218A overexpression plasmid, or empty vector, respectively. After 24 h, the cells were infected with SARS-CoV-2 (HKU-001a) at an MOI of 0.01. Forty-eight hours post-infection, the supernatants and cell lysates were harvested and subjected to qPCR assays to quantitate the viral copy numbers. Student's t-test, *P < 0.05, n.s., not significant. (F) HEK293T-ACE2 cells were cotransfected with the SARS-CoV-2 nucleocapsid plasmid and KLK13-WT or KLK13 S218A for 36 h. The cells were split and infected with SARS-CoV-2 GFP/ΔN for 3 h (MOI of 0.05). Total RNA was extracted 24, 48, and 72 h post-infection (hpi) for RT-qPCR to detect relative levels of SARS-CoV-2 genomic RNA and subgenomic RNA. Data are presented as mean ± SD. *P < 0.05, ****P < 0.0001 (n = 3).
(∆RRAR) was generated according to the protocol of the ClonExpress II One Step Cloning Kit, with wild-type Omicron spike as a template. The sequences of all cloning plasmids were confirmed by Sanger sequencing.
The expression vectors rAAV-KLK13-HA-2A-EGFP and rAAV-KLK13-HA-2A-KLK13 were constructed by BrainVTA (BrainVTA Co., Ltd., Wuhan, China). AAV6 particles encoding human KLK13 or EGFP were also packaged by BrainVTA.
HEK293T cells were seeded one day prior to transfection, and the indicated plas mids were transfected using polyethylenimine (PEI, Sigma) or Fugene (E2311, Promega) following the manufacturer's instructions.
## CRISPR activation
The plasmids Lenti-SAM-v2-puro (Addgene, 92062) and Lenti-MS2-P65-HSF1 Hygro (Addgene, 61426) were obtained from Addgene. The sgRNAs targeting the promoter regions of KLK13 were synthesized and cloned into the BsmbI site of the Lenti-SAM-v2puro plasmid. Scrambled sgRNA was used as a negative control.
HEK293T cells were cotransfected with plasmids expressing sgRNA or P65, psPAX2, and pMD2.G at a ratio of 4:3:1.2 to generate lentiviruses. The culture medium was replaced with fresh medium 6 h after transfection, and lentivirus supernatants were further collected at 48 h, filtered using a 0.45 µm filter, and stored at -80°C. A549-hACE2-TMPRSS2 cells were then transduced with both lentiviral vectors that express sgRNA and MS2-P65-HSF1 Hygro. The mRNA expression levels of KLK13 were determined by quantitative real-time PCR to detect the activation efficiency.
The following sgRNA primers were used in this study:
(1) KLK13#1: 5'-GGCCACATGGCTCCGGGATC-3';
(2) KLK13#2: 5'-GGGTGCAGTGGCGAGGTGGG-3';
(3) scrambled sgRNA: 5′-AAGATGAAAGGAAAGGCGTT-3′.
## Animal experiments
After obtaining consent from the Committee on the Use of Live Animals in Teaching and Research (CULATR) of the University of Hong Kong, all the animal experiments were carried out in biosafety level 3 animal facilities and performed in accordance with the standard operating procedures as well as the NIH Guide for Care and Use of Laboratory Animals. K18-hACE2 transgenic mice were housed and bred at the Center for Compara tive Medicine Research (CCMR), affiliated with the University of Hong Kong. The mice were maintained in BSL-2 housing with food and water freely and were delivered at 6 to 8 weeks of age. The animals were transferred to a BSL-3 animal facility for virus challenge. All the rooms housing animals were held at 25°C and 50% humidity. Four-to six-week-old male Balb/c mice were randomly divided into two groups (n = 5 per group): the rAAV-EGFP group and rAAV-KLK13 group. Thirty days before viral infection, each mouse received either 5 × 10 5 vg of rAAV-EGFP or rAAV-KLK13 through intranasal administration (20 µL per mouse). Thirty days post-inoculation, all mice were anesthetized with ketamine and xylazine, then intranasally infected with 2 × 10 3 plaque-forming units (PFU) of SARS-CoV-2 (HKU-001a) per mouse. Two days post-challenge, all mice were euthanized for tissue collection. Half of the nasal turbinate and right lung homogenate were used for viral load determination via plaque assay, while the remaining nasal turbinate and left lung were preserved for histological analysis.
## Virus titration by plaque assays
For viral titer determination, harvested nasal turbinates and lung tissues were homogen ized in 1 mL DMEM using a Tissue Lyzer II (Qiagen, Germany). Following centrifugation, supernatants were tenfold serially diluted and inoculated onto Vero E6-TMPRSS2 cell monolayers in 24-well plates. After 2 h, the cell medium was replaced with 1% low-melt ing agarose in DMEM containing 1% FBS. Forty-eight hours post-inoculation, cells were fixed with 4% paraformaldehyde and stained with 0.5% crystal violet in 25% ethanol/dis tilled water for plaque visualization. Viral titers were quantified as plaque-forming units per gram of tissue (PFU/g) for both lung and nasal turbinate (NT) samples.
## Histology and scoring
Lung tissues were collected and fixed in 10% neutralbuffered formalin. Nasal turbinates underwent decalcification in 10% formic acid for seven days before processing with a TP1020 Leica semi-enclosed benchtop tissue processor. Tissue sections were stained with Gill's hematoxylin and eosin-Y for H&E staining. Images were acquired using an Olympus BX53 light microscope.
For semi-quantitative histological scoring, blinded evaluation was performed to assess pathological changes in bronchioles, alveoli, and blood vessels using the following criteria: For bronchioles: 0 = normal structure; 1 = mild peribronchiolar infiltration; 2 = peribronchiolar infiltration plus epithelial cell death; 3 = score 2 plus intrabronchiolar wall infiltration and epithelium desquamation.
## Pseudovirus packaging and pseudovirus entry assay
To obtain pseudoviruses bearing VSV glycoprotein, SARS-CoV-1 spike, SARS-CoV-2 (WH strain) spike, or Omicron (BA.1) spike, HEK293T cells were cotransfected with plasmids encoding VSV glycoprotein, SARS-CoV-1 spike, SARS-CoV-2 spike, or Omicron spike along with psPAX2 and pHIV-Luciferase at a ratio of 15:10:10. The medium was refreshed at 6 h post-transfection. Supernatants containing viral particles were harvested 48 h later and subsequently filtered using a 0.45 µm filter and frozen at -80°C.
To perform pseudovirus entry assays, HEK293T-ACE2 cells were seeded in 48-well plates in triplicate and inoculated with pseudovirus the following day. Six hours after transduction, the medium was replaced, and the cells were cultured for another 48 h. The cells were lysed with lysis buffer and subjected to the detection of firefly luciferase signal through a luminometer.
## Immunofluorescence
HEK293T cells were seeded in 24-well plates with cell-climbing slices at a density of 10 × 10 4 and cultured overnight. The Omicron spike plasmid was transfected into HEK293T cells along with the KLK13-HA plasmid. At 36 h post-transfection, the cells were fixed in 4% paraformaldehyde for 10 min and then permeabilized in 0.2% Triton X-100 in PBS for 10 min at room temperature. Following blocking in 1% BSA diluted in PBS at 37°C for 1 h, the sample was incubated with SARS-CoV-2 spike-or HA tagspecific primary antibody at 37°C for 1 h. After washing three times with PBS to remove unbound primary antibody, goat anti-mouse FITC (015-090-050, Jackson) and goat anti-rabbit Alexa Fluor 568 secondary antibodies (A11036, Invitrogen) in 1% BSA were added and incubated at 37°C for 1 h. The sample was washed three times with PBS and then incubated with DAPI diluted in PBS at room temperature for 7 min. Afterward, the cell-climbing slice was mounted on microscope slides with antifade mounting medium (H-1000-10, Vector). Fluorescence images were captured using a Nikon C2 confocal microscope (Nikon Eclipse Ni-E), and colocalization analysis was performed using ImageJ software (ImageJ_v1.8.0) with the Plot Profile function.
## Cell-cell fusion assay
HEK293T cells stably expressing mCherry were cotransfected with SARS-CoV-2 (WH strain) spike or (BA.1) spike expression plasmids, together with plasmid of KLK13-HA, KLK13 S218A , or an empty vector, which served as effector cells. At 5 h post-transfection, the cells were detached and mixed at a 1:1 ratio with HEK293T cells stably expressing GFP and ACE2, which were used as target cells. Thirty-six hours after coculture, the mixed cells were fixed with 4% paraformaldehyde and subsequently stained with DAPI. Fluorescent and brightfield images were captured using a fluorescence microscope (Olympus). The number of syncytia, indicated by merged green-red color, was counted and analyzed from eight fields per experiment.
## RNA extraction, cDNA preparation, and quantitative real-time PCR
HeLa cells or A549 cells were seeded into 12-well plates at a density of 3 × 10 5 and cultured overnight. Both cell lines were transfected with poly(I:C) using FuGENE HD transfection reagents, with ddH 2 O as a negative control. Total RNA in cells was extracted at 18 h post-transfection using MagZol Reagent (Magen) following the manufacturer's instructions. The concentrations of purified RNA were measured using a NanoDrop spectrophotometer, and 1 µg of total RNA was used for reverse transcription with HiScript II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China). Quantitative real-time PCR was performed by using ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) to analyze the mRNA expression level of KLK13 in each sample, with IFIT3 serving as a positive control. To quantify the relative expression level of target genes, the housekeeping gene GAPDH served as a reference for normalization. Three technical replicates in each group were adopted in this experiment.
The following primers were used in this study for qPCR: KLK13-Fwd: 5′-CAGCCCCCAGGTGAATTAC-3′; KLK13-Rwd: 5′-CAGGAGACGATGCCATACAGT-3′; IFIT3-Fwd: 5′-GAAGAAATGAAAGGGCGAAGG-3′; IFIT3-Rwd: 5′-AGGACATCTGTTTGGCAAGGAG-3′; GAPDH-Fwd: 5′-TGCACCACCAACTGCTTAGC-3′; GAPDH-Rwd: 5′-GGCATGGACTGTGGTCATGAG-3′;. SARS-CoV-2 genomic RNA-Fwd: 5′-AGAAGATTGGTTAGATGATGATAGT-3′; SARS-CoV-2 genomic RNA-Rwd: 5′-TTCCATCTCTAATTGAGGTTGAACC-3′; SARS-CoV-2 subgenomic RNA-Fwd: 5′-CTTCCCTCAGTCAGCACCTC-3′; SARS-CoV-2 subgenomic RNA-Rwd: 5′-AACCAGTGTGTGCCATTTGA-3′;
## Coimmunoprecipitation
HEK293T cells were seeded in 6-well plates at a density of 3 × 10 5 and cultured overnight. The SARS-CoV-2 WH strain or Omicron (BA.1) spike plasmid was transfected into HEK293T cells along with KLK13-HA or an empty vector for 48 h. The cells were lysed in 500 µL TBST lysis buffer (20 mM Tris-HCl pH 7.4, 1 mM EDTA, 150 mM NaCl, 1% Triton X-100), containing 1× proteinase inhibitor cocktail, among which 50 µL of lysate was mixed with 4× loading buffer and then denatured at 98°C for 10 min, while 450 µL of lysate was incubated with 20 µL of anti-HA magnetic beads (Pierce, 88836) or 20 µL protein A/G agarose beads conjugated with normal IgG (as a control) or specific IgG against HA on a rolling platform at 4°C overnight. The beads were separated using a magnetic stand or centrifugation and washed six times with TBST lysis buffer. Proteins bound to beads were eluted by mixing with 4× loading buffer and boiling at 98°C for 10 min. Both cell lysate (input) and immunoprecipitated (IP) samples were subsequently used for immunoblotting analysis.
## Nasal mucus collection and analysis
At enrollment, nasal mucus of volunteers was collected in 15 mL centrifuge tubes and placed on ice immediately. The mucus was mixed with 1% SDS lysis buffer (1% SDS, 50 mM Tris-HCl pH 8.1, 10 mM EDTA pH 8, 1 mM Phenylmethylsulfonylfluoride) containing 2× proteinase inhibitor cocktail at a ratio of 1:1 and further sonicated. The cellular debris in the lysate was removed by centrifugation at 12,000 × g at 4°C for 10 min. The concentration of total protein in lysate was determined using a BCA Protein Assay Kit (23227, Thermo Scientific). In each sample, 18 µg of protein was mixed with 4× loading buffer and denatured at 98°C for 10 min and subjected to immunoblotting analysis.
## Immunoblotting
Cells were harvested and lysed in 1× SDS lysis buffer (containing 1× proteinase inhibitor cocktail), while culture media were mixed with 4× loading buffer (containing 4× proteinase inhibitor cocktail). All the samples were denatured at 98°C for 10 min. Proteins were separated by SDS-PAGE and then transferred to PVDF membranes (Merck Millipore).
After blocking in 5% skim milk for 1 h, the membranes were incubated with specific primary antibodies on a rolling platform at 4°C overnight. Following three washes with PBST, the membranes were incubated with HRP-conjugated secondary antibodies (115-035-003, Jackson) at room temperature for 1 h. After washing three times with PBS, chemiluminescent images were captured using an image developer (ChemiDoc XRS+) with chemiluminescent HRP substrate (WBKLS0500, Merck).
## SARS-CoV-2 spike cleavage experiment in vitro
The recombinant human KLK13 protein (10199-H08H, SinoBiological) is in the form of proenzyme and needed to be activated by lysyl endopeptidase (JP05061, Wako). The lyophilized powder of KLK13 protein and lysyl endopeptidase was diluted in activation buffer (0.1 M Tris, pH 8.0) to the final concentration of 100 µg/mL and 0.8 µg/mL, respectively. The activation assay was conducted by adding 1 µL lysyl endopeptidase to 40 µL of KLK13 protein solution, followed by incubation at 37℃ for 30 min. For the in vitro SARS-CoV-2 spike cleavage experiment, the SARS-CoV-2 spike protein (40589-V27B-B, SinoBiological) was firstly diluted in reaction buffer (50 mM tris, pH 7.5) to the final concentration of 25 µg/mL. Then, 10 µL of activated KLK13 protein or BSA (100 µg/mL) solution was fully mixed with 40 µL of SARS-CoV-2 spike protein, respectively, and incubated at 37℃ for 3 h. The samples were mixed with 4× loading buffer and then denatured at 98℃ for 10 min and subjected to immunoblotting analysis.
## scRNA-seq data analysis
All data sets used for expression analysis originated from public data sets. To determine the expression levels of identified serine proteases and metalloproteases across various cell types of the human airway, the human airway single-cell RNAseq data were downloaded from Synapse (https://accounts.synapse.org/) (accession code: EGAS00001004344). The single-cell RNA-seq data from healthy individuals and SARS-CoV-2-infected patients were available at a web portal (https://covid19cellatlas.org) and used for differential expression analysis of identified genes in specific human airway cell types. Data processing and plotting were conducted in R software.
## Statistical analysis
All data are presented as the mean ± SD. Comparisons were performed by using two-tailed Student's unpaired t tests. Differences between two groups were considered statistically significant when P < 0.05.
## References
1. Zhang, Peeples, Boucher et al. (2002) "Respira tory syncytial virus infection of human airway epithelial cells is polarized, specific to ciliated cells, and without obvious cytopathology" *J Virol*
2. Sims, Baric, Yount et al. (2005) "Severe acute respiratory syndrome coronavirus infection of human ciliated airway epithelia: role of ciliated cells in viral spread in the conducting airways of the lungs" *J Virol*
3. Matrosovich, Matrosovich, Gray et al. (2004) "Human and avian influenza viruses target different cell types in cultures of human airway epithelium" *Proc Natl Acad Sci U S A*
4. Ahn, Kim, Hong et al. (2021) "Nasal ciliated cells are primary targets for SARS-CoV-2 replication in the early stage of COVID-19" *J Clin Invest*
5. Wu, Lidsky, Xiao et al. (2023) "SARS-CoV-2 replication in airway epithelia requires motile cilia and microvillar reprogramming" *Cell*
6. Holly, Diaz, Smith (2017) "Defensins in viral infection and pathogenesis" *Annu Rev Virol*
7. Chatterjee, Huang, Mykytyn et al. (2023) "Glycosylated extracellular mucin domains protect against SARS-CoV-2 infection at the respiratory surface" *PLoS Pathog*
8. Chatterjee, Van Putten, Strijbis (2020) "Defensive properties of mucin glycoproteins during respiratory infections-relevance for SARS-CoV-2" *mBio*
9. Drosten, Günther, Preiser et al. (2003) "Identification of a novel coronavirus in patients with severe acute respiratory syndrome" *N Engl J Med*
10. Ksiazek, Erdman, Goldsmith et al. (2003) "A novel coronavirus associated with severe acute respiratory syndrome" *N Engl J Med*
11. Peiris, Yuen, Osterhaus et al. (2003) "The severe acute respiratory syndrome" *N Engl J Med*
12. Zaki, Van Boheemen, Bestebroer et al. (2012) "Isolation of a novel coronavirus from a man with pneumonia in Saudi Arabia" *N Engl J Med*
13. Wit, Doremalenn, Falzarano et al. (2016) "SARS and MERS: recent insights into emerging coronaviruses" *Nat Rev Microbiol*
14. Zhu, Zhang, Li et al. (2019) "China Novel Coronavirus I, Research T. 2020. A novel coronavirus from patients with pneumonia in China"
15. Zhou, Yang, Wang et al. (2020) "A pneumonia outbreak associated with a new coronavirus of probable bat origin" *Nature*
16. Bestle, Heindl, Limburg et al. (2020) "TMPRSS2 and furin are both essential for proteolytic activation of SARS-CoV-2 in human airway cells" *Life Sci Alliance*
17. Zhao, Yang, Yang et al. (2021) "Cathepsin L plays a key role in SARS-CoV-2 infection in humans and humanized mice and is a promising target for new drug development" *Signal Transduct Target Ther*
18. Franks, Chong, Chui et al. (2003) "Lung pathology of severe acute respiratory syndrome (SARS): a study of 8 autopsy cases from Singapore" *Hum Pathol (N Y)*
19. Matsuyama, Nagata, Shirato et al. (2010) "Efficient activation of the severe acute respiratory syndrome coronavirus spike protein by the transmembrane protease TMPRSS2" *J Virol*
20. Qian, Dominguez, Holmes (2013) "Role of the spike glycoprotein of human middle east respiratory syndrome coronavirus (MERS-CoV) in virus entry and syncytia formation" *PLoS One*
22. Chan, Chan, Choi et al. (2013) "Differential cell line susceptibility to the emerging novel human betacoronavirus 2c EMC/2012: implications for disease pathogenesis and clinical manifestation" *J Infect Dis*
23. Hoffmann, Kleine-Weber, Pöhlmann (2020) "A multibasic cleavage site in the spike protein of SARS-CoV-2 is essential for infection of human lung cells" *Mol Cell*
24. Xu, Shi, Wang et al. (2020) "Pathological findings of COVID-19 associated with acute respiratory distress syndrome" *Lancet Respir Med*
25. Pfaender, Mar, Michailidis et al. (2020) "LY6E impairs coronavirus fusion and confers immune control of viral disease" *Nat Microbiol*
26. Wang, Li, Hui et al. (2020) "Cholesterol 25-hydroxylase inhibits SARS-CoV-2 and other coronaviruses by depleting membrane cholesterol" *EMBO J*
27. Zang, Case, Yutuc et al. (2020) "Cholesterol 25-hydroxylase suppresses SARS-CoV-2 replication by blocking membrane fusion" *Proc Natl Acad Sci*
28. Hamilton, Whittaker (2013) "Cleavage activation of human-adapted influenza virus subtypes by kallikrein-related peptidases 5 and 12" *J Biol Chem*
29. Leu, Yang, Chung et al. (2015) "Kallistatin ameliorates influenza virus pathogenesis by inhibition of kallikrein-related peptidase 1-mediated cleavage of viral hemagglutinin" *Antimicrob Agents Chemother*
30. Magnen, Gueugnon, Guillon et al. (2017) "Kallikreinrelated peptidase 5 contributes to H3N2 influenza virus infection in human lungs" *J Virol*
31. Cerqueira, Ventayol, Vogeley et al. (2015) "Kallikrein-8 proteolytically processes human papillomaviruses in the extracellular space to facilitate entry into host cells" *J Virol*
32. Jones, Dry, Frampton et al. (2014) "RNA-seq analysis of host and viral gene expression highlights interaction between varicella zoster virus and keratinocyte differentiation" *PLoS Pathog*
33. Milewska, Falkowski, Kulczycka et al. (2020) "Kallikrein 13 serves as a priming protease during infection by the human coronavirus HKU1" *Sci Signal*
34. Shaw, Diamandis (2007) "Distribution of 15 human kallikreins in tissues and biological fluids" *Clin Chem*
35. Lilja (1985) "A kallikrein-like serine protease in prostatic fluid cleaves the predominant seminal vesicle protein" *J Clin Invest*
36. Zang, Castro, Mccune et al. (2020) "TMPRSS2 and TMPRSS4 promote SARS-CoV-2 infection of human small intestinal enterocytes" *Sci Immunol*
37. Hou, Yu, Wang et al. (2022) "Competitive cleavage of SARS-CoV-2 spike protein and epithelial sodium channel by plasmin as a potential mechanism for COVID-19 infection" *Am J Physiol-Lung Cellular Mole Physiol*
38. Yousef, Chang, Diamandis (2000) "Identification and characteri zation of KLK-L4, a new kallikrein-like gene that appears to be downregulated in breast cancer tissues" *J Biol Chem*
39. Darnell, Kerr, Stark (1994) "Jak-STAT pathways and transcrip tional activation in response to IFNs and other extracellular signaling proteins" *Science*
40. Soler, Schlosser, Mulligan et al. (2021) "Olfactory cleft mucus proteome in chronic rhinosinusitis: a case-control pilot study" *Int Forum Allergy Rhinol*
41. Workman, Nocera, Mueller et al. (2019) "Translating transcription: proteomics in chronic rhinosinusitis with nasal polyps reveals significant discordance with messenger RNA expression" *Int Forum Allergy Rhinol*
42. Yoshikawa, Wang, Jaen et al. (2018) "The human olfactory cleft mucus proteome and its age-related changes" *Sci Rep*
44. Zheng, Liu, Chen et al. (2024) "SARS-CoV-2 NSP2 as a potential delivery vehicle for proteins" *Mol Pharm*
45. Ju, Zhu, Wang et al. (2021) "A novel cell culture system modeling the SARS-CoV-2 life cycle" *PLoS Pathog*
46. Andrade, Assis, Santos et al. (2011) "Substrate specificity of kallikrein-related peptidase 13 activated by salts or glycosaminoglycans and a search for natural substrate candidates" *Biochimie*
47. Buchrieser, Dufloo, Hubert et al. (2020) "Syncytia formation by SARS-CoV-2-infected cells" *EMBO J*
48. Xia, Liu, Wang et al. (2020) "Inhibition of SARS-CoV-2 (previously 2019-nCoV) infection by a highly potent pan-coronavirus fusion inhibitor targeting its spike protein that harbors a high capacity to mediate membrane fusion" *Cell Res*
49. Braga, Ali, Secco et al. (2021) "Drugs that inhibit TMEM16 proteins block SARS-CoV-2 spike-induced syncytia" *Nature*
50. Zhang, Zheng, Niu et al. (2021) "SARS-CoV-2 spike protein dictates syncytiummediated lymphocyte elimination" *Cell Death Differ*
51. Papa, Mallery, Albecka et al. (2021) "Furin cleavage of SARS-CoV-2 spike promotes but is not essential for infection and cell-cell fusion" *PLoS Pathog*
52. Qi, Larson, Gilbert et al. (2013) "Repurposing CRISPR as an RNA-Guided Platform for Sequence Specific Control of Gene Expression" *Cell*
53. Gruba, Bielecka, Wysocka et al. (2019) "Development of chemical tools to monitor human kallikrein 13 (KLK13) activity" *Int J Mol Sci*
54. Delaveris, Webster, Banik et al. (2020) "Membrane-tethered mucin-like polypeptides sterically inhibit binding and slow fusion kinetics of influenza A virus" *Proc Natl Acad Sci U S A*
55. Iverson, Griswold, Song et al. (2022) "Membrane-tethered mucin 1 is stimulated by interferon and virus infection in multiple cell types and inhibits influenza a virus infection in human airway epithelium" *MBio*
56. Wallace, Liu, Van Kuppeveld et al. (2021) "Respiratory mucus as a virus-host range determinant" *Trends Microbiol*
57. Kim, Kang, Kim et al. (2024) "The host protease KLK5 primes and activates spike proteins to promote human betacoronavirus replication and lung inflammation" *Sci Signal*
58. Liang, Zhang, Li et al. (2021) "TRIM26 is a critical host factor for HCV replication and contributes to host tropism" *Sci Adv*
59. Pan, Chen, He et al. (2021) "Infection of wild-type mice by SARS-CoV-2 B.1.351 variant indicates a possible novel crossspecies transmission route" *Sig Transduct Target Ther* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12625748&blobtype=pdf | # A 1-year study on SARS-CoV-2 variant shifts in wastewater using dPCR: comparison with clinical and GISAID data
Mohammad Sayed, Mosavi, Patrick Acer, Patrick Andersen, Robbie Barbero, Stephanie Barksdale, Sophia Bellakbira, Dalton Bunde, Ross Dunlap, James Erickson, Daniel Goldfarb, Tara Jones-Roe, Michael Kilroy, Hien Le, Benjamin Lepene, Emily Milich, Ayan Mohamed, Denton Munns, Jared Obermeyer, Anurag Patnaik, Ganit Pricer, Marion Reven, Dalaun Richardson, Chamodya Ruhunusiri, Saswata Sahoo, Lauren Saunders, Olivia Swahn, Kalpita Vengurlekar, David White, Jeremy Davis-Turak, Aaron Stanz, Aouda Flores-Baffi, Jean Lozach, Tim Wesselman, Stephen Hilton, Siya Kashwala, Pengbo Liu, Christine Moe, Orlando Sablon, Yuke Wang, Marlene Wolfe, Dagmara Antkiewicz, Erica Camarato, Kayley Janssen, Adélaïde Roguet, Shreya Shrestha, Regan Wied, Johannah Gillespie, Jim Huang, Andrew Jones, Sarah Kane, Dolores Gonzalez, Modou Jarju, Chi-Yu Lin, Mayumi Pascual, Rachel Poretsky, Michael Secreto, Ian Bradley, Sydney Gallo, Yinyin Ye, Elizabeth Donahue, Stephanie Greenwald, Sarah Owens, Rosemarie Wilton
## Abstract
Wastewater testing can be used to monitor SARS-CoV-2 infections in communities. Data from PCR-based wastewater testing are usually available to public health authorities within 5-7 days after excreta and other body fluids enter the sewer. While PCR-based methods can accurately detect and quantify SARS-CoV-2, sequencingbased methods are usually required to distinguish between variants, delaying the results and adding cost to the process. We developed and assessed a novel, customizable digital PCR (dPCR)-based genotyping method for SARS-CoV-2 variant detection in wastewater, which is more cost-effective, faster, and more accessible than sequencing. This approach was applied to more than 1,400 wastewater samples collected from six states between April 2023 and May 2024, with results displayed on a public dashboard alongside clinical data from the same period. The wastewater dPCR-based method effectively detected emerging variants, mirroring trends observed in clinical settings and in data on the Global Initiative on Sharing All Influenza Data (GISAID) platform; this method also provided early warning signals, as variants like EG.5 and FL were identified in wastewater before clinical detection. A subset of the wastewater samples was analyzed using both dPCR genotyping and sequencing, with good agreement between the two methods. The range of concordance for four different variants was between 62% and 98%. The successful development and implementation of this dPCR-based genotyping for wastewater samples demonstrates its cost-effectiveness and scalability. With the decline in clinical testing, wastewater surveillance becomes increasingly vital for monitoring SARS-CoV-2 variants and supplementing clinical surveillance efforts. IMPORTANCE As clinical specimens are collected and analyzed less for SARS-CoV-2, variant detection in wastewater provides a readily accessible and rich source of information on SARS-CoV-2 evolution. The detection of pathogen targets in wastewater samples using PCR assays is a sensitive, cost-effective way to monitor the levels of infectious diseases, like SARS-CoV-2, in a community. Unfortunately, because PCR-based methods are typically not used to distinguish between viral variants, most wastewater testing labs must rely on more expensive, time-consuming, and resource-intensive sequencing methods for these results. Building upon recent developments for variant detection using quantitative PCR, we developed and assessed a novel, customizable digital PCR-based genotyping method for SARS-CoV-2 variant detection in wastewater,
the utility of wastewater-based genotyping as an effective way to monitor SARS-CoV-2 variants. By integrating dPCR-based genotyping and sequencing methods, we compared the prevalence of variants in wastewater samples with clinical and GISAID data sets, demonstrating a strong correlation and emphasizing the utility of wastewater testing for community-level tracking of variants.
## RESULTS
## Developing the dPCR-based wastewater genotyping method
We began by demonstrating the feasibility of qPCR-based genotyping in wastewater samples using archived nucleic acids from 10 samples in Georgia during the BA.1-to-BA.2 variant transition (February-April 2022). The results revealed a strong correlation between whole-genome sequencing (WGS) and qPCR results, with a t-test P-value of 0.9269 and 0.1983 and r-values of 0.6514 and 0.9922 for BA.1 and BA.2, respectively (Fig. S4).
Next, we applied dPCR-based genotyping to nine archived samples from Georgia collected during the BQ-to-XBB variant transition (January-March 2023). The results indicated no significant differences between the two methods and showed a strong correlation, with t-test P-values of 0.8342 and 0.8685 and correlation coefficients of 0.8256 and 0.8537 for BQ.1 and XBB, respectively (Fig. S5). Based on these findings, we developed a standard operating procedure for a 2-marker panel to monitor BQ.1 and XBB variants, which was used to test 86 samples in Georgia between 11 April and 19 June 2023. Wastewater genotyping results aligned with clinical data, with XBB being the dominant variant detected (88.23% vs 96.39%; Fig. S7).
In June 2023, we updated the variant panel to reflect the evolving shift in prevalence by adding markers for XBB, EG, FD, and FL. This updated panel was used to evaluate 73 archived samples collected between June and September 2023, confirming XBB as the most prevalent variant in Georgia during this period (72.81% in wastewater samples and 66.47% in clinical samples). Notably, the FL and EG.5 variants were detected 22 and 31 days earlier in wastewater samples than in clinical data (Fig. S9). Finally, in late 2023, as the JN variant began rising in prevalence, we validated a new assay to detect XBB, EG.5, FL, and JN in wastewater samples, enabling continued monitoring during early 2024 (Supplemental material).
## Wastewater genotyping results correlate with and complement clinical genotyping and GISAID results over a 1-year timeframe in Georgia
Using state-level data in Georgia between April 2023 and April 2024, we examined variant prevalence reported by three different methods-wastewater genotyping, clinical sample genotyping, and GISAID-for several variants: XBB, BQ, EG.1, EG.5/FL, and JN (Fig. 2). The wastewater genotyping results for XBB tracked a steady decline in variant prevalence from nearly 100% in April 2023 until it almost completely disappeared by January 2024, a result that was mirrored in the GISAID data (Fig. 2A). The clinical genotyping data similarly tracked this decline until October 2023 when it stopped reporting XBB completely.
Because the BQ and EG.1 variants were both present at low levels during the study period in Georgia, we combined these results into a single chart (Fig. 2B). For most of the study period, the wastewater genotyping results were consistent with the reported prevalences of these variants in clinical genotyping and GISAID results. Notably, however, the wastewater testing results showed an increase in the BQ variant to 27% prevalence in May 2023, which was not reported in either of the other surveillance approaches.
While the wastewater genotyping panel 2 was able to distinguish between EG.5 and FL variants, the genotyping panel that was used for clinical samples between July 2023 and October 2023 could not distinguish between these two variants. Thus, to facilitate a comparison between the wastewater genotyping and clinical genotyping in Georgia during this time period, we combined the prevalence of EG.5 and FL into one chart (Fig. 2C). The wastewater genotyping results track closely with the clinical genotyping and GISAID results. Also, the data indicate that EG.5/FL was detected in wastewater samples 22 days earlier than in clinical samples (Fig. 2C).
Fig. 2D illustrates the rapid rise and dominance of the BA.2.86*/JN* variant detected in wastewater samples and the GISAID data sets between October 2023 and January 2024. The "unknown variant" represents SARS-CoV-2 signals that did not match known variant-specific mutations in the genotyping panel, likely indicating emerging variants. The upward trend of the unknown marker in wastewater, which began in October 2023 and peaked at around 50% in January 2024, sharply declined after the transition to the new panel with the JN assay in February 2024.
## Wastewater dPCR-based genotyping shows strong agreement with wholegenome sequencing for variant detection
We compared the performance of the dPCR genotyping method with WGS using 339 data points from the ROSALIND wastewater dashboard (panel 3) using data from Wisconsin and Illinois, where corresponding sequencing data were available. Details of the sequencing methods used in this study are described in Materials and Methods.
The Wisconsin State Laboratory of Hygiene (WSLH) processed biological duplicates of 129 wastewater samples. One replicate was concentrated using Nanotrap Microbiome A Particles (Ceres Nanosciences) and extracted with a Maxwell HT Environmental TNA kit (Promega) for Illumina WGS. The other replicate was concentrated using Nanotrap Microbiome A Particles and extracted with a MagMAX wastewater extraction kit (Thermo Fisher Scientific). At the University of Illinois, 210 wastewater samples were processed following the dPCR genotyping protocol. The purified SARS-CoV-2 RNA from these samples was tested using both dPCR and WGS. We evaluated the results from 339 combined samples processed by both methods.
## Concordance analysis
To synchronize the results, we aggregated WGS sublineages into broader lineages with shared mutations, aligning with the dPCR detection panels. Both positive and negative detections were considered. Concordance between the two methods ranged from 61.9% to 98.2% (Table 1).
## Correlation analysis
The correlation analysis between dPCR and WGS results for JN, EG.5, XBB, and FL variants further supported the agreement between the findings and revealed varying degrees variant is observed across all data sets over the study period. In October 2023, the prevalence of the XBB* variant was still at 30% prevalence in both wastewater and GISAID results, but it disappeared from the clinical genotyping results. The vertical dotted line demarcates the last date that the clinical genotyping results reported XBB*. (B) The prevalence of BQ* and EG.1* variants detected in wastewater, clinical, and GISAID data sets. A notable spike in the BQ* variant is observed in wastewater samples in May 2023, reaching a prevalence of 27%, which is not reflected in clinical or GISAID data. The EG.1* variant was only detected in the wastewater data set during the fall and winter of 2023, with a prevalence of 1% in September and 3% in December, and showed no signal in the clinical and GISAID data sets. (C) The prevalence of EG.5* and FL* variants detected in wastewater, clinical, and GISAID data sets. The data show that EG.5/FL was detected in wastewater samples 22 days earlier than in clinical samples. The first vertical dotted line demarcates the date that EG.5 and FL were first reported in wastewater genotyping samples in Georgia, and the second vertical dotted line demarcates the date that EG.5 and FL were first reported in clinical genotyping samples in Georgia. (D) The prevalence of the BA.2.86*/JN* variant and an unknown marker detected in wastewater samples and the GISAID data set. The second peak of the unknown marker in the wastewater data set from October 2023 to January 2024 is associated with the BA.2.86*/JN* variant. Note that because there were very few clinical genotyping results during this time period, they were not included in the figure .
of consistency between the two methods (Fig. 3). For the JN variant, there was a strong positive correlation (r = 0.7477, P < 0.0001), indicating a high level of agreement between dPCR and WGS. Similarly, the EG.5 variant demonstrated a statistically significant positive correlation (r = 0.7067, P < 0.0001), reflecting strong alignment in detection results. In contrast, the XBB variant showed a moderate positive correlation (r = 0.3367, P < 0.0001), highlighting considerable variability and reduced agreement between the methods. For the FL variant, no significant correlation was observed (r = 0.02055, P = 0.7062), with notable differences in detection frequencies; WGS identified FL in only 3% of samples, compared to 37% in the dPCR results. These findings underscore differences in detection sensitivity, methodological resolution, and baseline detection levels between the two approaches, particularly for recombinant variants like XBB and FL.
## Lessons learned and opportunities for improvement
We encountered several challenges during this project, resulting in valuable lessons for future deployments of dPCR-based genotyping for wastewater, which we have summarized in Table 2.
## Evaluating dPCR genotyping of wastewater at a national level
To evaluate the utility of, and any challenges associated with, utilizing this dPCR-geno typing method at a national level, we worked with multiple testing laboratories starting on 31 October 2023. The wastewater genotyping panel 2 for variants XBB, EG.1, EG.5, and FL was deployed by five laboratories around the United States. These laboratories, collectively, tested samples from six states (California, Georgia, Illinois, Louisiana, New York, and Wisconsin) at a cadence of roughly 17 samples per lab per week.
Between 31 October 2023 and 16 May 2024, 1,181 samples from these six states were processed and analyzed using the dPCR genotyping approach (Fig. 4). During that same time frame, 2,323 clinical samples nationwide were analyzed for the same variants using qPCR-based genotyping, and 139,631 clinical sequencing results were deposited in GISAID. As observed in the detailed analysis of Georgia samples, the emergence of the JN.1 variant was detected in early November 2023 (Fig. 2). This variant showed a rapid rise in prevalence in combined wastewater data from these six states, increasing from a 2-week average of approximately 9% in December 2023 to around 75% in January 2024 in wastewater samples (Fig. 4). A similar pattern of rapid emergence was observed in the GISAID data.
From a broader perspective, the wastewater variant profiles in these six states during this period mirrored the patterns observed in the GISAID data set for the entire country. For example, the wastewater dPCR-based genotyping data showed a JN marker prevalence of 57.16% (dark green in Fig. 4), which closely aligns with the 61.31% observed in the GISAID sequencing-based data set. Note that the JN marker detected by the dPCR genotyping BA.2.86*/JN* assay corresponds to JN.1, JN.1.1, and JN.1.4 in the GISAID data set. Furthermore, the "Other" category in the GISAID graph includes 27.79% of BA.2.86 sublineages. This results in an overall prevalence of 61.31% for the JN marker in the GISAID data set for this period. This strong agreement highlights the a A total of 339 wastewater samples were counted based on the detection or absence of the marker by each method.
b Combined positive and negative percent agreement.
reliability of dPCR-based wastewater genotyping as a complementary tool to clinical and sequencing-based approaches for variant monitoring.
## DISCUSSION
In this study, we present a novel approach for detecting SARS-CoV-2 variants in wastewater using dPCR and mutation-specific assays. Several studies have previously utilized mutation-specific PCR assays to detect SARS-CoV-2 variants of concern in wastewater (14,16,17,21). However, these studies primarily used individual mutationspecific assays to track specific mutations, either retrospectively or in real-time, without the capability to detect emerging variants circulating in the community.
Compared to these approaches, our study employs panels of mutation-specific duplex assays with probe-specific mutation and wild-type probes, specifically designed for real-time wastewater surveillance of Omicron BA.2 sublineages, enhancing detec tion accuracy. Additionally, we introduce a novel prevalence calculation method for recombinant variants sharing mutations within our assay panel and demonstrate the ability to detect and measure the prevalence of emerging variants, making this approach valuable for near-real-time proactive surveillance. Moreover, our workflow includes an automated viral concentration and nucleic acid extraction process, improving efficiency and reproducibility.
Although the accuracy of wastewater testing can be affected by sewer shed-specific characteristics (22,23), our 1-year study in the state of Georgia shows that the prevalence of variants detected in the wastewater closely matched the variant prevalence in clinical specimens, reinforcing the reliability of this approach as reported by other studies (12, 14, 16). A detailed analysis of multiple data sets from Georgia revealed several key findings. First, the dominance and transition of SARS-CoV-2 variants, such as the shift from XBB to BA.2.86*/JN* in early 2024, were effectively captured in both wastewater and clinical samples (Fig. 2). We also noted the clinical genotyping data tracked XBB's decline until October 2023 when reporting ceased due to a sharp reduction in genotyped samples. The number of clinical specimens declined from 62 in early August to just 2 by late September 2023, and no clinical samples were genotyped by the reference labs after 8 February 2024. This decline in testing likely hindered accurate prevalence assessment in the qPCR-based clinical results, but with sufficient samples, the trendline would have likely mirrored the GISAID and wastewater genotyping results. The data from Georgia indicate that some variants were transient, like BQ* and EG.1*. The brief spike in the BQ* variant detected in wastewater in May 2023, and the absence of a corresponding spike in clinical and GISAID data, raises questions about the possible reasons for this discrepancy. It may reflect a lag in clinical testing, differences in population sampling, or the variant's low pathogenicity, leading to fewer clinical cases with specimen collection and genotyping despite widespread circulation. The detection of the EG.1* variant solely in the wastewater data during the fall and winter of 2023, with no corresponding signals in clinical or GISAID data sets, suggests that this variant did not lead to significant numbers of clinical cases.
We observed two significant spikes in the Unknown/Others category during the study: in June and October 2023. These spikes indicate ongoing viral evolution and the emergence of new variants. The June spike likely resulted from the introduction of the FL variant, which was reported in GISAID data but went undetected by the wastewater genotyping panel, as it did not include specific primers and probes for FL at that time (Fig. 2C). In October 2023, we suspected that the unknown variants detec ted were JN. Retesting archived nucleic acids from 23 wastewater samples collected
## Challenge
## Potential improvement
The primer and probe design tools we used were not optimized for the dPCR system that we were using for our testing, which caused some of the primer and probe sets to fail when implemented on the dPCR system.
Utilize primer and probe design tools developed by the manufacturer of the dPCR instrument.
The assays that we used for genotyping were not available with customized concentrations of primers and probes, which caused some of the primer and probe sets to fail when implemented on the dPCR system.
Work with an assay manufacturer that provides customized concentrations of primers and probes, allowing for optimization specific to wastewater samples.
The JN variant rapidly grew in prevalence and was the dominant variant in samples before we had a JN assay validated and deployed to the testing labs, resulting in high levels of "unknowns" detected in the wastewater samples during this time period.
New variants can arise quickly and are expected to appear in wastewater samples earlier than in clinical samples. Having verified assays ready for deployment as soon as possible is desirable. This may require designing, ordering, and testing assays on a more frequent cadence.
Using a single-target mutation assay for each variant detection is not the most cost-effective or labor-efficient way to utilize dPCR testing.
Instead of single-target mutation assays, which require running four separate dPCR plates each week in the testing lab, use multiplex assays to reduce the burden on the testing laboratories. Based on our estimated costs for assays, controls, reagents, and plates, moving to a four-plex dPCR assay could reduce the cost per sample by $66.
The data analysis pipeline failed when the dPCR instrument manufacturer published a software update on the instrument. This update modified the data output format and caused issues during data upload to the ROSALIND Tracker.
Implement a QC check on data format in the data management process to prevent data upload/transfer issues. Coordinate with the instrument manufacturer to be prepared for pending software updates and the impact those changes might have on the process.
Overall concentrations of SARS-CoV-2 RNA in the wastewater samples were not reported in this project, potentially diminishing the value of this testing method.
Including an overall SARS-CoV-2 target in the test panel can be a useful metric for monitoring community transmission. Understanding virus levels alongside variant prevalence provides a better picture of transmission dynamics. Although this study did not measure SARS-CoV-2 levels in wastewater, it is possible to do so by analyzing mutant and wild-type quantities in each assay. Further study is needed to assess the accuracy of SARS-CoV-2 levels based on these results.
a QC, quality control.
between 11 December 2023 and 1 January 2024 confirmed this, as the unknown variants disappeared after deploying the JN assay (Fig. S13 and Fig. 2D). The concurrent rise of unknown and JN variants from October 2023 to April 2024 demonstrated strong alignment between wastewater and GISAID trendlines. However, the rapid rise of JN in late 2023 posed a challenge. We were unable to validate and deploy the assay across all five testing laboratories in time to track JN in real time. As a result, a large fraction of unknown variants appeared in wastewater samples tested during January and February 2024. After assay validation, laboratories retested retained RNA samples, and the updated results were uploaded to the ROSALIND Tracker. This highlights the need for more frequent assay design and validation to keep pace with emerging variants. The concordance study between dPCR and WGS results for JN, EG.5, XBB, and FL variants in Wisconsin and Illinois demonstrates that dPCR genotyping is an accurate and reliable tool for monitoring variants in wastewater. These findings align with previous studies highlighting the higher sensitivity and lower sample input requirements of PCR-based methods (17,24,25). However, we noted varying degrees of agreement between dPCR and WGS variant prevalence results. The lower agreement for XBB and FL markers likely stems from differences in single-nucleotide polymorphism detection in dPCR and lineage classification protocols in WGS. Recombinant variants such as XBB -formed from multiple lineages-can carry mutations shared in other variants like FL (26) (https://covariants.org/variants/22F.Omicron; https://genspectrum.org/). These mutations are detected by dPCR, but they are classified under different categories in WGS, resulting in discrepancies between the two methods. Similarly, FL variant detection often requires multiple markers, introducing greater variability due to the additional measurements and percentage normalization steps. In contrast, the JN marker showed higher agreement due to the presence of unique mutations not shared with other variants, enabling consistent identification across methods. In addition, there are some limitations with sequencing-based genotyping when genome coverage is low, since errors in variant identification might occur due to only partial information on the analyzed RNA (17,24). Despite detection sensitivity, variant nomenclature, and methodological resolution, dPCR genotyping offers a reliable approach for rapid and targeted variant monitoring, complementing the higher-resolution capabilities of WGS.
## Conclusion and recommendations
PCR-based genotyping offers a potential alternative to sequencing for rapid, cost-effective variant detection in wastewater. However, this study identified key challenges in implementing dPCR-based wastewater genotyping and proposed improvements. Assay failures occurred due to primer and probe design tools not being optimized for the dPCR system, which could be mitigated by using manufacturer-recommended design tools and customized primer concentrations. Single-target mutation dPCR assays proved inefficient and costly, suggesting a shift to multiplex assays. Additionally, a software update from the dPCR instrument manufacturer disrupted the data pipeline, highlighting the importance of proactive quality control measures and coordination with manufacturers. Lastly, the study did not measure overall SARS-CoV-2 concentrations in wastewater, limiting its utility for community transmission analysis. Future efforts should include an overall SARS-CoV-2 target to enhance surveillance capabilities.
Despite challenges, the dPCR genotyping method offers significant advantages over WGS for routine variant surveillance, including lower costs, simplicity, and faster turnaround times. As of 2024, singleplex dPCR reduced reagent costs by 36% and sample-to-result times by 68% compared to Illumina sequencing (Table S7). Multiplexing in dPCR could further cut reagent costs by 84% and processing times by 90%. The streamlined dPCR workflow requires fewer steps and less expertise, and it allows rapid marker updates without extensive revalidation, making it ideal for tracking evolving variants. While WGS provides greater precision and detects recombinants, dPCR is a practical, cost-effective choice for ongoing high-throughput SARS-CoV-2 wastewa ter surveillance. However, unlike WGS, PCR-based genotyping cannot identify novel mutations or provide comprehensive genomic information, making it less suitable for exploratory surveillance. Compared to qPCR, dPCR offers high sensitivity, accuracy, and resilience to inhibitors, but the run time is longer, and it may be less suitable for detecting multiple targets in a high-throughput format due to the cost and complexity of multiplexing (27,28).
In our 1-year study in Georgia, we observed that one variant consistently exceeded 50% prevalence each quarter, with no more than three variants dominating the profile at any given time. This suggests that, for community-level surveillance programs, a broad detection tool like dPCR-based genotyping can provide sufficient resolution to monitor major shifts in viral variants and may be adequate for understanding clinical outcomes. High-resolution methods like WGS, which detect sublineages, might not always be necessary in such contexts. Furthermore, these findings emphasize the need for regularly developing and validating multiplexed wastewater panels to capture the continuously evolving variant landscape effectively. During the study, we also observed the presence and a sudden increase in the Unknown variants at different time points. Establishing a detection threshold for the Unknown marker and monitoring the rate of increase could serve as indicators to trigger new assay development and validation efforts. The ability to rapidly identify emerging variants is critical for timely public health interventions.
We compared the wastewater genotyping results from six states, representing the Northeast, South, Midwest, and West regions of the United States, to clinical qPCR-based genotyping data and sequencing results in GISAID. Despite the smaller regional coverage compared to nationwide data and the smaller sample size in the wastewater data set relative to clinical data sets, the comparison demonstrated similar trends and variant profiles. This strong agreement underscores the reliability of dPCR-based wastewater genotyping as a complementary tool to clinical specimen testing and sequencing-based approaches for monitoring SARS-CoV-2 variants.
Overall, our findings support the use of wastewater-based SARS-CoV-2 genotyping as a cost-effective and complementary approach to clinical sample genotyping and sequencing, offering a broader and more inclusive picture of variant prevalence and transmission. This approach is particularly valuable in times when clinical testing is limited, as observed during the study period. Beyond SARS-CoV-2, the dPCR-based genotyping platform described in this study holds promise for tracking other patho gens in wastewater. Its high sensitivity and specificity make it particularly suitable for monitoring low-abundance targets, such as emerging variants of other respiratory viruses (e.g., influenza and respiratory syncytial virus [RSV]), antibiotic resistance genes, or enteric pathogens like norovirus and Salmonella. As public health agencies expand their use of wastewater surveillance, dPCR can serve as a flexible and scalable tool for early detection and population-level monitoring of a wide range of infectious diseases.
## MATERIALS AND METHODS
## Wastewater collection
We processed and analyzed 1,416 wastewater samples across six states from April 2023 to May 2024. Table 3 lists the number of samples tested from each state and the collection period for those states.
Wastewater samples were collected from multiple types of sources, including wastewater treatment plants and correctional facilities, as part of regular wastewater testing routines conducted by the five testing laboratories. Wastewater samples were collected from 91 sites across six U.S. states using primarily automated composite sampling methods. Most sites employed 24-hour time-or flow-weighted composite sampling of influent wastewater from treatment plants. California sites collected effluent samples, and one site in Illinois provided grab samples. The sampling frequency ranged from once to twice weekly, and volumes collected per site ranged from 40 mL to 500 mL. Environmental monitoring data, including sample arrival temperature, total flow, and optionally pH, DO, and conductivity, were recorded where available-most comprehen sively in Wisconsin and Illinois. Full sampling specifications by state, including site type, method, and environmental parameters, are provided in Table S8.
Across the six participating states, most wastewater samples were processed promptly upon arrival or stored under appropriate conditions to preserve sample integrity prior to analysis. When processing delays occurred, they were typically due to logistical factors such as sample shipping schedules or batching strategies for efficient lab workflow. More details are provided in the Supplemental material.
## Wastewater processing
Automated SARS-CoV-2 concentration was accomplished using Nanotrap Microbiome A Particles and Enhancement Reagent 1 (ER1) on a Thermo Scientific KingFisher Apex System. In brief, 75 µL of Nanotrap Microbiome A Particles and 50 µL of Nanotrap ER1 were mixed with ~5 mL of wastewater in two replicate wells with a total of 10 mL of wastewater for each sample. Concentrated viruses were lysed in 500 µL Microbiome Lysis Buffer at 56°C. For all samples except the samples from WSLH that underwent WGS, the following nucleic acid extraction method was followed. After viral concentration, the samples were processed for RNA extraction using the Applied Biosystems MagMAX Wastewater Ultra Nucleic Acid Isolation Kit on the KingFisher Apex System. Briefly, 400 µL of lysate was mixed with 550 µL of MagMAX binding mix, and 10 µL of proteinase K was added to each sample prior to running on the KingFisher Apex. Wash 1 and wash 2 used 1 mL of MagMAX wash solution and 80% ethanol, respectively. RNA was eluted in 100 µL Microbiome Elution Solution. For the WSLH samples that underwent WGS, nucleic acids were extracted using the Maxwell HT Environmental TNA kit (Promega).
Extracted nucleic acids were quantified post-extraction and either analyzed via dPCR genotyping within 24 hours or were stored in a freezer at -20°C to -80°C until they were analyzed.
## SARS-CoV-2 dPCR-based genotyping
SARS-CoV-2 genotyping was accomplished using the Applied Biosystems TaqMan SARS-CoV-2 Mutation Panel (Cat# A49785). Table 4 summarizes the assays and targeted mutation sites used in this study.
The marker selection for SARS-CoV-2 variant detection and lineage assignment methods used in this study has been previously described (15). Sequences for positive controls (mutation and wild type) were designed in silico by ROSALIND Bio and were constructed and manufactured using the gBlocks service by Integrated DNA Technolo gies (IDT) or the GeneArt service by ThermoFisher Scientific. Assays and controls were validated by Emory University and Ceres Nanosciences, Inc., using the QIAcuity dPCR system.
The mutation detection assay was performed on the QIAGEN QIAcuity Digital PCR system. In brief, the reaction mix was made by mixing 10 µL OneStep Advanced Probe Master Mix and 0.4 µL OneStep Advanced RT Mix. The volume was brought up to 30 µL by adding RNase-free water. Ten microliters of RNA template was added to the reaction mix, and then the entire volume of 40 µL was transferred into 26K 24-well QIAGEN Nanoplate. The QIAGEN QIAcuity Digital PCR system was used to amplify and detect the signals. Amplification was accomplished according to the following steps: (i) one cycle at 50°C for 40 minutes; (ii) one cycle at 95°C for 2 minutes; (iii) 45 cycles at 95°C, for 3 seconds; (iv) 60°C for 30 seconds. Only for the FL assay, step iii was modified to 45 cycles at 95°C for 30 seconds, and step iv was modified to 57°C for 1 minute. Signal detection was obtained using default settings for exposure duration and gain in each channel. QIAcuity Software Suite (version 2.2) was used to analyze the data. A common threshold was applied across the samples to clearly separate negative partitions from positive partitions. Mutation detection results were exported in comma-separated values (CSV) format.
## Calculating variant percentages
Three genotyping panels were used to track SARS-CoV-2 variants as their prevalence shifted in the United States (Table S1 through S3). Each panel included 2-4 digital PCR assays targeting specific mutations and their wild-type counterparts. Mutation fractions were calculated as follows:
mutation fraction % = mutant concentration mutant concentration + wild type concentration × 100.
Unique mutations identified specific lineages, while shared mutations required subtraction of overlapping fractions in a pre-defined order. Variant prevalence was normalized if the total exceeded 95%, ensuring the sum approached 100%. Undetected variants were calculated by subtracting the total mutation fraction from 100%. Detailed calculations are in Table S4.
## Data analysis and presentation on the public dashboard
Details of the controls and sample analysis criteria can be found in the Supplemental material. The ROSALIND classification algorithm automated mutation fraction analysis from QIAcuity dPCR outputs, processing thousands of specimens per minute on Google Cloud's secure virtual private cloud. Metadata files with sample origin and collection dates were required for processing. Results were published on the publicly accessible ROSALIND Tracker dashboard (https://tracker.rosalind.bio/tracker/dashboard/).
## WSLH sequencing method
WSLH sequences roughly 20% of the wastewater samples it processes. Illumina WGS data are processed through the Viralrecon workflow (https://nf-co.re/viralrecon/). The bioinformatics algorithm Freyja is used to evaluate the relative proportion of the SARS-CoV-2 lineages present in wastewater samples. Data are manually curated to only display the lineages according to the World Health Organization (WHO) and Nextstrain nomenclatures. These data and visualizations are available on a dashboard accessible to the public (https://dataportal.slh.wisc.edu/sc2-ww-dashboard).
WSLH utilized a library prep method for sequencing using the following protocol. The SARS-CoV-2 libraries were prepared using the QIAseq DIRECT SARS-CoV-2 Enhancer kit using the Booster primers (Qiagen). Briefly, single-stranded viral RNA molecules were reverse transcribed into cDNA using hexaprimers. The SARS-CoV-2 genome was then specifically enriched using a SARS-CoV-2 primer panel. The panel consists of approxi mately 550 primers for creating 425 amplicons, covering the entire SARS-CoV-2 viral genome. Prior to sequencing, library quality was assessed using the QIAxcel Advanced System (Qiagen) and quantified by qPCR using the QIAseq Library Quant System kit (Qiagen). Libraries were sequenced on a MiSeq Illumina platform using MiSeq Reagent v2 (300 cycles) kits targeting a median coverage of at least 500× and at least 80% of the genome covered at 10×.
To assess variant proportions, WGS data were analyzed using Freyja v.1.3.11, a tool specifically designed to estimate SARS-CoV-2 variant proportions in deep sequence data containing mixed populations (29). BAM files, generated using viralrecon v2.5, were processed through Freyja, utilizing the Wuhan-Hu-1 reference genome (MN908947.3) to produce variant and depth files. The median estimates were obtained through Freyja's bootstrap boot function (nb = 10). All samples were processed using Freyja's UShER barcode reference database updated on 13 April 2024. This ensured the inclusion of all the most recent variant detections in samples processed during earlier periods.
Variant proportions derived from Illumina sequencing are accessible through a publicly available dashboard hosted at https://dataportal.slh.wisc.edu/sc2-ww-dash board. This dashboard showcases the proportions of major variant groups listed on https://covariants.org/, with estimations generated using Freyja's raw calculations. Additional details on the methodology can be found at https://github.com/wslh-ehd/ sc2_wastewater_data_analysis.
## University of Illinois Chicago sequencing method
The SARS-CoV-2 libraries prep, sequencing, and data analysis were similar to the WSLH method, except that the libraries were sequenced on an Illumina Nextseq 2000 using NextSeq 1000/2000 P1 Reagents (300 Cycles) kits, generating a median of 800,000 reads per sample. All samples were processed using Freyja's UShER barcode reference database and updated weekly during the testing period to ensure the inclusion of the most recent variants detected in samples in real time.
## References
1. Muneer, Arshad (2023) "Insight of pandemic COVID-19, develop ments and challenges" *Pak J Sci*
2. Muneer, Munir, Abbas et al. (2021) "Facts and figures on COVID-19 pandemic outbreak" *PJZ*
3. Abbas, Simon, Qureshi et al. (2021) "Beyond the death and infection plagues of COVID-19 on the globe: a critical analysis of its diverse effects" *Virol Mycol*
4. Abbas, Iqbal, Javid et al. (2019) "COVID-19 attack, prevention, precaution and managemental strategies" *Int J Innov Res Educ Sci*
5. Gagliano, Biondi, Roccaro (2023) "Wastewater-based epidemiology approach: the learning lessons from COVID-19 pandemic and the development of novel guidelines for future pandemics" *Chemosphere*
6. Kilaru, Hill, Anderson et al. (2023) "Wastewater surveillance for infectious disease: a systematic review" *Am J Epidemiol*
7. O'reilly, Wade, Farkas et al. (2025) "Analysis insights to support the use of wastewater and environmental surveillance data for infectious diseases and pandemic preparedness" *Epidemics*
8. (2022) "SARS-CoV-2 wastewater surveillance testing guide for public health laboratories"
9. Farkas, Hillary, Malham et al. (2020) "Wastewater and public health: the potential of wastewater surveillance for monitoring COVID-19" *Curr Opin Environ Sci Health*
10. (2023) "Wastewater-based disease surveillance for public health action"
11. Karthikeyan, Ronquillo, Belda-Ferre et al. (2021) "High-throughput wastewater SARS-CoV-2 detection enables forecasting of community infection dynamics in San Diego County. mSystems"
12. Amman, Markt, Endler et al. (2022) "Viral variant-resolved wastewater surveillance of SARS-CoV-2 at national scale" *Nat Biotechnol*
13. Markov, Ghafari, Beer et al. (2023) "The evolution of SARS-CoV-2" *Nat Rev Microbiol*
14. Wolfe, Hughes, Duong et al. (2022) "Detection of SARS-CoV-2 variants Mu, Beta, Gamma, Lambda, Delta, Alpha, and Omicron in wastewater settled solids using mutation-specific assays is associated with regional detection of variants in clinical samples" *Appl Environ Microbiol*
15. Lai, Kennedy, Lozach et al. (2022) "A method for variant agnostic detection of SARS-CoV-2, rapid monitoring of circulating variants, and early detection of emergent variants such as omicron" *J Clin Microbiol*
16. Heijnen, Elsinga, De Graaf et al. (2021) "Droplet digital RT-PCR to detect SARS-CoV-2 signature mutations of variants of concern in wastewater" *Sci Total Environ*
17. Wilhelm, Schoth, Meinert-Berning et al. (2023) "Interlaboratory comparison using inactivated SARS-CoV-2 variants as a feasible tool for quality control in COVID-19 wastewater monitoring" *Sci Total Environ*
18. Tiwari, Ahmed, Oikarinen et al. (2022) "Application of digital PCR for public health-related water quality monitoring" *Sci Total Environ*
19. Khare, Gurry, Freitas et al. (2021) "GISAID's role in pandemic response" *China CDC Wkly*
20. Swaminathan (2020) "The WHO's chief scientist on a year of loss and learning" *Nature*
21. Farkas, Pellett, Williams et al. (2023) "Rapid assessment of SARS-CoV-2 variantassociated mutations in wastewater using real-time RT-PCR" *Microbiol Spectr*
22. Kallem, Hegab, Alsafar et al. (2023) "SARS-CoV-2 detection and inactivation in water and wastewater: review on analytical methods, limitations and future research recommendations" *Emerg Microbes Infect*
23. Parkins, Lee, Acosta et al. (2024) "Wastewater-based surveillance as a tool for public health action: SARS-CoV-2 and beyond" *Clin Microbiol Rev*
24. Lou, Sapoval, Mccall et al. (2022) "Direct comparison of RT-ddPCR and targeted amplicon sequencing for SARS-CoV-2 mutation monitoring in wastewater" *Sci Total Environ*
25. Lind, Barlinn, Landaas et al. (2025) "Methods and Protocols mSystems November"
26. Holberg-Petersen (2021) "Rapid SARS-CoV-2 variant monitoring using PCR confirmed by whole genome sequencing in a high-volume diagnostic laboratory" *J Clin Virol*
27. Tamura, Ito, Uriu et al. (2023) "Virological characteristics of the SARS-CoV-2 XBB variant derived from recombination of two Omicron subvariants" *Nat Commun*
28. Ding, Xu, Deng et al. (2024) "Comparison of RT-ddPCR and RT-qPCR platforms for SARS-CoV-2 detection: Implications for future outbreaks of infectious diseases" *Environ Int*
29. Erler, Droop, Lübbert et al. (2024) "Analysing carbapenemases in hospital wastewater: Insights from intracellular and extracellular DNA using qPCR and digital PCR" *Sci Total Environ*
30. Karthikeyan, Levy, Hoff et al. (2022) "Wastewater sequencing reveals early cryptic SARS-CoV-2 variant transmission" *Nature* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817944&blobtype=pdf | # Attenuation of canine distemper virus leads to a potent antiviral innate immune response with restricted infection of alveolar macrophages
Pauline Pöpperl, Elisa Chludzinski, Melanie Stoff, Robert Geffers, Martin Ludlow, Andreas Beineke
## Abstract
Canine distemper virus (CDV, species Morbillivirus canis) is a highly contagious pathogen with a broad host range among carnivores. In common with measles virus, alveolar macrophages (AMs) are among the first target cells of infection in the respiratory tract. Therefore, in vitro infections of primary canine AMs were performed with the attenuated Onderstepoort (Ond) and field R252 strain of CDV over a period of 6 days. This showed that AMs are permissive to CDV infection and that such infections are productive with respect to the release of new virus particles. Phenotypic differences were observed over the entire course of the experiment, as higher levels of infection and virus production were observed in CDV R252-infected AMs, while infection with CDV Ond resulted in more prominent cytopathic effects, including syncytium formation. Transcriptome analyses of samples from 1 day post-infection via total RNA sequenc ing demonstrated further marked differences with respect to the pro-inflammatory response and cell death pathways. CDV Ond-infected AMs exhibited robust induction of pro-inflammatory mediators including type I interferon-related signaling pathways, whereas CDV R252-infected cells showed much weaker expression of these pathways. These transcriptomic differences were further highlighted by the detection of the highest rates of cell apoptosis and lactate dehydrogenase activity in the supernatants of CDV Ond-infected AM cultures over the entire course of the experiment. In addition, transcriptome differences indicate disturbances of homeostatic AM functions associated with CDV infection. These results provide insights into early events in the pathogenesis of CDV infection and mechanisms underlying vaccine strain attenuation. IMPORTANCE Morbilliviruses, including canine distemper virus (CDV) and human measles virus, cause severe systemic disease with respiratory distress, immunosuppres sion, and neurologic signs. While natural infection in dogs has become rare due to efficient vaccination, outbreaks in wildlife populations can be devastating, and concerns about zoonotic potential of CDV have been raised. The impact of CDV infection on the transcriptome of alveolar macrophages has not been elucidated thus far. Knowledge about early events in CDV pathogenesis and phenotypic consequences of vaccine attenuation is therefore necessary to protect endangered wildlife populations and might furthermore serve as a model for human measles. This study presents the first transcrip tomic analyses of primary AMs during the initial phase of morbillivirus infection. These results provide insights into early events in the pathogenesis of CDV infection and mechanisms serving to restrict the spread of an attenuated virus strain. KEYWORDS canine distemper virus, alveolar macrophages, vaccines, innate immunity C anine distemper virus (CDV, species Morbillivirus canis) is a member of the genus Morbillivirus within the family Paramyxoviridae (1). It causes canine distemper, a
highly contagious and severe systemic disease in which respiratory distress, immunosup pression, and neurologic signs are typical clinical signs of infection. In addition to dogs, CDV infects wild carnivores, including seals, foxes, raccoons, bears, and mustelids, as well as large felids and non-human primates, thereby representing a serious threat to endangered wildlife species (2)(3)(4)(5). Moreover, CDV outbreaks in rhesus and cynomolgus macaques in China and Japan, respectively, have raised concerns about the zoonotic potential of CDV, particularly due to decreasing human measles virus (MeV) vaccination rates in many regions (6)(7)(8). Live attenuated morbillivirus vaccines have been extremely successful in reducing morbidity and mortality levels in animals and humans in the last 60-70 years. However, our understanding of phenotypic differences between attenuated and wild-type strains in primary target cells is limited.
CDV has a non-segmented, negative sense, single-stranded RNA genome, six structural proteins, and two non-structural proteins (1). Morbillivirus non-structural V proteins interact with retinoic acid-inducible gene I (RIG-I)-like receptors, melanoma differentiation-associated protein 5 (MDA5), probable ATP-dependent RNA helicase DHX58, signal transducer and activator of transcription (STAT) 1, and STAT2, thereby inhibiting the production of type I interferons (IFNs) and tumor necrosis factor-α (TNF-α) (9)(10)(11)(12)(13).
A notable feature of morbilliviruses is their ability to infect epithelial cells of the respiratory tract, from which infectious virus is released and efficiently transmitted to other hosts via aerosols or respiratory droplets (14,15). It is suspected that the initiation of CDV infection in a susceptible host is similar to MeV, bypassing the epithelial barrier of the respiratory tract via pulmonary dendritic cells and alveolar macrophages (AMs), which express the viral entry receptor CD150 (14,(16)(17)(18). Here, AMs are suggested to sustain primary respiratory MeV infection and are the immune cell type infected at the highest level in the lung of mice expressing human CD150 receptor during the early infection phase (16,17). These infected cells subsequently transit the epithelial barrier of the respiratory tract with virus amplification occurring in lymphatic tissues of the respiratory tract, prior to the first viremic phase (16,19,20). The induction of timely and robust innate and humoral immune responses during early stages of infection in the respiratory tract could lead to a more restricted infection and effective virus elimination (21).
AMs are a distinct population of tissue resident macrophages originating from fetal progenitors (22,23) and represent the first line of defense within lung alveoli, given the continuous exposure to infectious agents (24,25). Under steady-state conditions, the functions of AMs include surfactant metabolism, phagocytosis, and clearance of cellular debris in order to maintain homeostasis within the lung microenvironment (26). Recognition of pathogen-or damage-associated molecular patterns by pattern recognition receptors and loss of their connection to epithelial cells in injured tissue can lead to a shift of AMs from a tolerogenic toward a pro-inflammatory phenotype, associated with the production of pro-inflammatory cytokines (27)(28)(29)(30)(31). AMs have been identified as early targets for MeV in transgenic mouse and cynomolgus monkey models, but there is a lack of knowledge about cellular responses upon infection (16,17,32). Their role as early target cells in CDV infection has also been suggested in ferret studies and ex vivo infection models (14,33).
Modulation of innate immune cells provides a potential target for treatment and prophylactic approaches to mitigate the impact of viral diseases. However, knowledge about pulmonary innate immunity in morbillivirus infections and its impact on disease pathogenesis is still sparse. In particular, the transcriptional and phenotypic properties of AMs in canine distemper have not yet been investigated. Elucidating the regula tory mechanisms through which pathogens regulate innate immune cell plasticity will contribute to the discovery of therapeutic targets in morbillivirus-induced diseases and thus reduce virus transmission to other hosts. In this study, we have investigated the ability of a field and attenuated strain of CDV to productively infect primary canine AMs and show that this results in differential cytopathic effects and pro-inflammatory innate immune responses.
## RESULTS
## CDV infection of primary canine AMs is associated with a restricted infection by an attenuated strain
Productive infection with the attenuated Onderstepoort (Ond) or field R252 strains was confirmed by immunofluorescence staining (Fig. 1a through c) and virus titration at 6 hours, 1 day, 3 days, and 6 days post-infection (dpi) (Fig. 1d). Noteworthy, CDV R252 infected significantly higher numbers of Iba1 + AMs compared to the vaccine strain CDV Ond at all time points post-infection (Fig. 1c). Accordingly, a concomitant increase in the release of infectious viral particles was found in CDV R252-infected AMs at all time points (Fig. 1d). While a time-dependent decrease of infection rates and virus titers was observed in CDV Ond-infected cells, the number of infected cells and virus titers remained high in CDV R252-infected AM cultures. CDV-infected primary canine AMs showed prominent syncytium formation with significantly more multinucleated cells observed following CDV Ond infection in comparison to CDV R252-infected cells at all examined time points (Fig. 2a through c). However, much larger intracellular inclusions and aggregates of N protein were observed in CDV R252-infected AM cultures (Fig. 2a andb). These results show that canine AMs are permissive for CDV infection, but that viral replication and release of infectious virus particles is strain-dependent, suggesting differential antiviral responses and thus consequently different rates of viral elimination in infected cells. Correlating to these results in vitro, CDV antigen can be detected in alveolar histiocytes of naturally infected dogs. Immunohistochemistry and immunofluorescence double labeling were used to confirm natural CDV infection of Iba1 + AMs in lungs (Fig. S1a andb).
## AMs show enhanced cytokine responses following infection with an attenu ated CDV strain
To characterize cellular responses following in vitro CDV infection, cytokine expression of infected primary canine AMs at 6 hours, 1 dpi, 3 dpi, and 6 dpi was assessed by reverse transcription quantitative PCR (RT-qPCR). The transcription of TNF-α was most prominent in CDV Ond-infected samples with significantly increased levels compared to non-infec ted controls and CDV R252-infected cells at 6 hours pi and 1 dpi. In addition, a significant increase of TNF-α mRNA in CDV Ond-infected cells compared to controls was found at 3 dpi. Significantly higher expression of TNF-α mRNA in CDV R252-infected cells compared to non-infected AMs was found at 1 dpi (Fig. 3a). In contrast, significantly decreased interleukin (IL)-1β mRNA expression was found in both CDV Ond-and CDV R252-infected cells compared to non-infected samples at 6 hours pi. In addition, IL-1β mRNA expression in CDV Ond-infected cells showed a tendency toward lower expression compared to CDV R252-infected AMs (P = 0.065; Fig. 3b). Moreover, a statistical trend toward decreased IL-1β mRNA expression was found in CDV Ond-infected cells compared to R252-infected samples at 3 dpi (P = 0.075). Highest mRNA expression levels of other investigated cytokines were also found in CDV Ond-infected cells, peaking at 6 hours pi (IL-10 and IL-12) and 1 dpi (IL-6, IL-8, TGF-β, and IFN-γ), but group differences did not reach the levels of significance (Fig. S2).
## Transcriptome analyses reveal CDV strain-specific responses of AMs
In order to investigate the transcriptomic signature of canine AMs following CDV infection in more detail, total RNA sequencing (RNA-seq) of infected and non-infec ted samples during the early infection phase (1 dpi) was performed. Key functional differences were computed via principal component analysis, which clustered samples into three distinct subsets, corresponding with the group assignment (Fig. 4a). Pairwise comparison showed fundamental changes in the transcriptome, with higher numbers of up-or downregulated differentially expressed genes (DEGs) in canine AMs observed following CDV Ond infection compared to CDV R252 infection (Fig. 4b andc). In total, 492 genes were uniquely upregulated in CDV Ond-infected cells, 227 upregulated DEGs were shared in both CDV Ond-and CDV R252-infected samples, and 61 genes were upregulated solely in CDV R252-infected cells. For downregulated DEGs, 587 genes were exclusively downregulated in CDV Ond-infected cells, 206 genes were shared, and 34 genes were downregulated solely in CDV R252-infected samples. Hierarchical clustering based on expression profiles identified six different clusters (Fig. 4d; Table S1). DEGs in cluster 4 (n = 593) were upregulated in both CDV-infected groups, with significantly higher expression values in CDV Ond-infected AMs. Cluster 6 (n = 151) genes were almost exclusively upregulated in CDV-Ond infected AMs. DEGs in cluster 1 (n = 560), cluster 2 (n = 243), and cluster 5 (n = 52) were downregulated in both CDV-infected groups. Of note, CDV R252-infected samples often showed significantly higher expression levels of genes in cluster 1 and partly in cluster 2 compared to CDV Ond-infected samples. Cluster 3 contained 63 DEGs with higher expression levels in CDV-infected AMs compared to non-infected controls. Analysis of viral gene expression (N, P, M, F, H, and L genes) at 1 dpi did not reveal significant differences between both CDV strains (Fig. S3).
## Genes related to pro-inflammatory and antiviral responses are preferentially upregulated in AMs infected with an attenuated CDV strain
The differential expression profiles were further analyzed with respect to biological context by performing Gene Ontology (GO) enrichment analysis ("biological process"; Table S2) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (Table S3). Most GO terms and KEGG pathways enriched in cluster 4 were related to innate immune processes and defense responses (Table 1).
KEGG pathway analysis revealed enrichment of several pathways associated with sensing of and reaction to pathogens in cluster 4, including NOD-like receptor signaling pathway (Fig. 5a), Toll-like receptor signaling pathway, and RIG-I-like receptor signaling pathway. Initiation of pro-inflammatory reactions to CDV infection was reflected by the enrichment of several GO terms and KEGG pathways associated with responses to virus infection, including the NF-κB signaling pathway, which is associated with the induc tion of pro-inflammatory components in immune cells and initiating antiviral defense responses (34) (Fig. 5b).
In agreement with RT-qPCR results (Fig. 3a), RNA-seq analysis revealed increased TNF-α gene expression in CDV Ond-infected canine AMs, while no upregulation of TNF-α was observed in CDV R252-infected canine AMs. Moreover, transcription of TNF-α -related genes was primarily upregulated in CDV Ond-infected cells (Fig. 5c). Similar to TNF-α mRNA expression (Fig. 3b), secretion of TNF-α by CDV Ond-infected AMs was significantly higher compared to both CDV R252-and non-infected AMs at 6 hours pi and 1 dpi, as shown by quantitative enzyme-linked immunosorbent assay (ELISA). TNF-α secretion by both CDV-infected AM cultures was significantly higher compared to the non-infected AMs at 6 hours pi, 1 dpi, and 3 dpi, with highest values found at 1 dpi (Fig. 5d). In addition, genes enriched in KEGG pathways apoptosis and necroptosis and GO terms associated with cell death were also present in cluster 4 (Fig. 5d ande). Interestingly, CASP8 and FADD, associated with the initiation of the extrinsic apoptosis pathway by TNF-α, were only significantly higher expressed in CDV Ond-infected AMs compared to controls and CDV R252-infected AMs, while there was no upregulation of expression of these genes in CDV R252-infected AMs.
CDV infection led to an upregulated transcription of IFN-α and IFN-β genes as well as their receptor and downstream transcription factor STAT1 in canine AMs. Increased transcription of the IFN-α gene IFNA5 (logFC 15) and IFN-β gene IFNB1 (logFC 11.5) was most prominent in CDV Ond-infected cells (Fig. 6a). Quantitative ELISA also revealed significantly increased IFN-α secretions by both CDV Ond-and R252-infected AMs compared to non-infected controls at all investigated time points. Significantly higher IFN-α secretion by CDV Ond-infected AMs compared to CDV R252-infected AMs was detected at 6 hours pi, 1 dpi, and 3 dpi (Fig. 6b).
Other cluster 4 genes, for which increased transcription is triggered by type I IFNs (IFN-stimulated genes [ISGs]) (35,36) are presented in Fig. 6c. These genes were all upregulated in both CDV Ond-and CDV R252-infected AMs but most prominently in CDV Ond-infected AMs. These findings were subsequently confirmed on a protein level, by performing immunofluorescence staining of two selected ISG proteins (ISG15 and MX1) in infected AM cultures (Fig. 6d ande). In accordance with transcriptome analysis, the numbers of both ISG + Iba1 + cells and MX1 + Iba1 + cells at 1 dpi were highest in CDV Ond-infected AM cultures compared to CDV R252-infected cells and non-infected controls. In addition, MX1 protein expression in CDV R252-infected AMs was significantly lower compared to non-infected cells (Fig. 6f andg).
Along with type I IFN and TNF-α signaling, several genes encoding for chemokines were upregulated as parts of clusters 4 and 6. While CDV Ond-infected AMs showed a significant upregulation of CXCL10, CCL8, CCRL2, CCL5, and CCL4 compared to non-infec ted controls, only CXCL10 was upregulated in CDV R252-infected cells.
KEGG pathway and the eponymous GO term antigen processing and presenta tion were enriched in cluster 4 and upregulated in CDV-infected cells, containing genes related to antigen processing and presentation via MHC peptides. All genes related to antigen presentation were slightly upregulated (logFC < 2.5), and for most genes, differential expression was significant only between CDV Ond-infected cells and non-infected controls.
## Attenuation of CDV is associated with increased cytopathic effects in infected AM cultures
Both the observation of syncytium formation and the enrichment of cell death pathways in cluster 4 are indicative of increased cytopathic effects in AMs following CDV infection. To further evaluate the cell death of CDV-infected primary canine AMs, lactate dehydro genase (LDH) activity was measured in supernatants collected at 6 hours pi, 1 dpi, 3 dpi, and 6 dpi (Fig. 7a). At 6 hours pi, LDH release by CDV Ond-infected AMs was significantly higher compared to non-infected AMs, indicating reduced cell viability. At 1 and 3 dpi, LDH release by CDV Ond-infected AMs peaked and was significantly higher compared to both other groups of AMs. Also, CDV R252-infected AMs released significantly more LDH compared to non-infected control AMs at 1 and 3 dpi.
Further insights into different cell death pathways were obtained by differential cell staining to distinguish between live, apoptotic, and necrotic cells at 1 dpi (Fig. 7b through d). While the number of necrotic cells was negligible (<0.5%, data not shown), apoptotic cell counts showed significant differences between the three groups. While no differences were found between CDV R252-infected cells and non-infected controls and rates of apoptosis were below 4.5%, markedly higher rates of apoptosis up to 36% were observed in CDV Ond-infected cells (Fig. 7e). Corresponding with both increased LDH activity and apoptotic cell count in CDV Ond-infected cells, live cell counts were lower than in both other groups (Fig. 7f).
To examine apoptosis induction in CDV-infected AMs over the entire course of the experiment, immunofluorescence staining of cleaved caspase-3 (CC3) was performed (Fig. 7g). Here, CDV Ond-infected AMs showed a higher proportion of CC3 + cells at 6 hours pi compared to CDV R252-infected cells and at 1, 3, and 6 dpi compared to both other groups. The number of CC3 + AMs peaked at 1 dpi. Interestingly, at 6 hours pi, 1 dpi, and 3 dpi, the percentage of CC3 + AMs was significantly lower in CDV R252-infected AMs compared to non-infected AMs. Taken together, this suggests that CDV Ond-infection is associated with reduced cell viability and enhanced apoptosis of canine AMs.
## CDV infection of AMs downregulates genes related to cell homeostasis and immune cell interaction
Analysis of cluster 2 identified genes, which were downregulated in both groups of CDV-infected cells compared to non-infected AMs, including genes related to phagocy tosis, cell adhesion, migration, and cell-cell interaction. Genes related to the GO terms regulation of leukocyte-mediated immunity, regulation of immune effector process, and T cell-mediated immunity were associated with pro-inflammatory signaling (IL1B), leukocyte adhesion (FUT7 and ITGAM) (37,38), phagocytosis of apoptotic cells (efferocytosis) (RAC2 and ITGAM) (39,40), and reactive oxygen species generation (RAC2) (41) (Fig. S4a). Furthermore, genes related to the GO term actin cytoskeletal organization were part of cluster 2 and significantly downregulated in CDV Ond-infected samples (Fig. S4b). Organization of the cell cytoskeleton is a key component of phagocytic function. In addition, cluster 1 contained CD36, a macrophage scavenger receptor involved in the recognition of bacterial pathogens and phagocytosis of apoptotic cells (42)(43)(44), which was downregulated in both infected groups. This clearly indicates that pro-inflammatory responses, which were pronounced in CDV Ond-infected cells, were accompanied by the downregulation of genes related to homeostatic functions.
## CDV infection alters the transcription of metabolic genes in AMs
The majority of enriched KEGG pathways and GO terms within cluster 1 were associ ated with cellular metabolism and downregulated in CDV-infected AMs. Promoters of fatty acid oxidation CD36 and IRF4 (45-47) showed highest expression in non-infected samples and were downregulated predominantly in CDV Ond-infected samples. The KEGG pathway fatty acid metabolism contained genes associated with peroxisomal and mitochondrial fatty acid β-oxidation (Fig. S4c). In addition, several GO terms and KEGG pathways associated with nucleotide metabolism were enriched in cluster 1 (ribonucleo tide biosynthetic process, purine nucleotide biosynthetic process, and purine metabo lism) (Fig. S4d). DEGs of the GO term ribonucleotide biosynthetic process were downregulated predominately in CDV Ond-infected AMs. Furthermore, genes enriched in the KEGG pathway purine metabolism, including genes encoding for enzymes involved in de novo synthesis of purines, were significantly downregulated in CDV-Ondinfected samples compared to non-infected controls. Collectively, transcriptome analyses from the initial infection phase revealed alterations of cellular lipid and nucleotide metabolism in AMs following CDV infection, suggestive of a shift from AM homeostatic fatty acid oxidation toward a more glycolytic energy generation, which is associated with the pro-inflammatory polarization of macrophages.
## DISCUSSION
Morbilliviruses are highly contagious due to efficient spread via the respiratory route with AMs and dendritic cells identified as the primary target cells during early stages of infection in the respiratory tract (16,17,32). Similarly, infection of Iba1 + macrophages within alveoli and associated expression of TNF-α and ISG proteins has been reported in dogs naturally infected with CDV (48). This study shows higher infection rates and virus titers in primary canine AMs following infection with the field CDV R252 strain compared to the CDV Ond vaccine strain over the entire course of the experiment. Moreover, accelerated virus elimination in CDV Ond-infected AMs is indicative of a more robust and efficient innate immune response. Accordingly, transcriptome analyses at one dpi, representing the initial infection phase, revealed increased gene expression associated with pro-inflammatory pathways and type I IFN signaling in AMs infected with the attenuated virus strain. These findings highlight the phenotypic consequences of morbillivirus strain attenuation in a primary cell model in which virus spread is restricted in comparison to a field isolate. ISGs, found to be upregulated in CDV-infected AMs, encode for multiple proteins, which have been shown to restrict virus replication in different stages of the viral life cycle. This includes transcriptional elongation (MX1) (49), inhibition of the production of viral proteins (PKR, encoded by EIF2AK2) (50), interference with viral replication (viperin, encoded by RSAD2) (51), and the degradation of viral RNA via OAS proteins that activate RNase L (encoded by RNASEL) (52). Moreover, type I IFNs have the potential to increase virus-induced apoptosis (53). Immunofluorescence staining showed that MX1 and ISG15 proteins are expressed prominently in AMs infected with the attenuated CDV Ond strain. Thus, enhanced expression of ISGs could contribute to viral elimination observed in CDV Ond-infected AM cultures, while less efficient antiviral signaling by AMs during initial CDV R252 infection might facilitate prolonged virus infection, as reflected by continu ously high infection rates and virus titers over the 6-day course of the experiment.
In viral diseases, apoptosis is a basic mechanism to limit the extent of viral replication and cell-to-cell spread (54,55). Wild-type CDV has been shown to prevent apoptosis of infected kidney epithelial cells in vitro and of infected immune cells of experimen tally infected ferrets in vivo, representing a possible mechanism of immune evasion (56). In contrast to wild-type CDV strains, infection of Vero cells with CDV Ond causes caspase-3-and caspase-8-mediated apoptosis (57). TNF-α gene expression and protein secretion were increased primarily in CDV Ond-infected AMs but only slightly in CDV R252-infected AMs, indicating an insufficient cytokine response in the latter. Enhanced TNF-α expression in CDV Ond-infected AMs might thus contribute to reduced viral replication and prevention of viral spread by its function as an inducer of cell death. This is supported by the upregulation of apoptosis-related genes during initial infection and high LDH release and increased apoptosis rates in these CDV Ond-infected AM cultures. Suppressed TNF-α expression in peripheral blood mononuclear cells from experimentally CDV-infected ferrets has been previously observed: while TNF-α mRNA expression is downregulated in blood leukocytes from virulent CDV-infected ferrets, V protein-knock out recombinant CDV induces increased TNF-α mRNA expression (9). Thus, enhanced cell death of primary canine AMs following CDV Ond-infection is probably associated with virus attenuation, leading to more rapid viral clearance. In the present study, CDV R252-infected AMs maintained high infection rates and production of infective virus particles over the entire course of the experiment, accompanied by low apoptosis rates. Thus, field CDV strains such as R252 might inhibit apoptosis of primary target cells, which could contribute to viral immune evasion and increased spread (56). The enhanced type I IFN and TNF-α signaling in CDV Ond-infected AMs contributes to the activation of death signaling cascades of infected cells and thus efficient elimination of virus-infected cells (58)(59)(60). Given the essential role of early pro-inflammatory responses of AMs in other respiratory viral infections, including influenza A virus and respiratory syncytial virus, an insufficient response and lack of cell death during aerogenic CDV infection might promote early viral propagation in AMs and spread to other cells of the respiratory tract (61,62).
Another function of IFN signaling is the initiation of antigen presentation to induce adaptive T-cell responses (63,64). To maintain tolerance to innocuous antigens within the alveolar niche, AMs are relatively poor antigen presenters and can actively suppress T-cell activation by dendritic cells under homeostatic conditions (65)(66)(67). CDV Ond-infec ted AMs showed increased expression of genes associated with the immunoproteasome and loading of MHC class I molecules and genes related to MHC class II-mediated antigen presentation, probably as a consequence of type I IFN signaling (63,64). In addition, the enhanced chemokine response in CDV Ond-infected cells corresponds well with the higher transcription of genes related to type I IFN and TNF-α signaling. Chemoattraction of inflammatory cells is necessary to create an antiviral microenvironment and initiate adaptive immunity (65,68). Therefore, the lack of chemokine induction together with a disturbed antigen presentation capacity in CDV R252-infected AMs might facilitate virulence and impair antiviral immunity.
The IL1B gene was significantly downregulated preferentially in CDV Ond-infected AMs as shown by RT-qPCR and RNA-seq analyses of the AMs during the initial infection phase. Type I IFN signaling decreases the amount of IL-1β, since type I IFN treatment of human macrophages suppresses pro-IL-1β protein availability, its caspase-1-medi ated cleavage, and the release of IL-1β (69,70). Similarly, human metapneumovirus suppresses IL-1β, associated with IFN-β signaling in human monocytes in vitro (71). Hence, type I IFN responses in CDV Ond-infected AMs and to a lesser extent in CDV R252-infected AMs may account for the downregulation of IL1B expression found in the current study.
Formation of syncytia is a common finding in lungs of CDV-infected dogs (48). CDV Ond infection was observed to cause more extensive cytopathic effect in infected AMs, as evidenced by increased numbers of syncytia. Increased syncytium formation in Vero cells is a characteristic feature of this attenuated strain and is related to properties of the H protein (72). In vitro studies on MeV-infected human epithelial and dendritic cells revealed a correlation between cell fusion with the consequent formation of multinucleated cells and increased cellular IFN-β production, as is also observed in the current study (73).
Viruses benefit from faster levels of replication, which involves active nucleotide metabolism (74). Several IFN-regulated mechanisms are known to disturb nucleotide metabolism, which has been shown to restrict lentivirus and herpesvirus infections. Genes related to nucleotide metabolism were downregulated in AMs infected with both CDV strains but more pronounced in CDV Ond-infected AMs, possibly representing another antiviral mechanism during the initial phase of CDV infection.
AMs exhibit several important regulatory functions, protecting the lung environ ment from harmful overreactions of the immune system in response to inhaled particles. Migration along the respiratory epithelium, phagocytosis of inhaled particles, scavenging microbes, clearance of apoptotic cells (efferocytosis), and maintenance of pulmonary surfactant homeostasis are all essential tasks of AMs within their anatomical niche (24,75). Under physiologic conditions, AMs have a low rate of glycolysis and a dominance of energy metabolism utilizing lipids (76). Non-infected AMs showed the highest expression of genes related to fatty acid catabolism, which is essential to remove excess surfactant and a hallmark of anti-inflammatory M2-polarized macrophages (77). Genes associated with fatty acid metabolism were downregulated early during CDV infection, which was most pronounced in CDV Ond-infected cells. This downregulation in CDV-infected cells might indicate a metabolic switch to a pro-inflammatory and glycolytic cell type, representing a downstream effect of Toll-like receptor signaling (78). Downregulation of IDH1 (encoding for isocitrate dehydrogenase) implies a metabolic switch and M1 polarization of CDV-infected AMs (79). Disruption of homeostatic AM functions during initial CDV infection could predispose infected animals to secondary bacterial pneumonia, as shown in respiratory syncytial virus and murine cytomegalovirus infection of mice (80,81).
The V protein of wild-type morbillivirus strains is able to block antiviral responses by inhibiting the translocation of transcription factors STAT1 and STAT2 to the nucleus (10,82,83). Moreover, the interaction of the V protein with RIG-I-like receptors MDA5 and LGP2 directly interferes with IFN-β transcription (11,13,(84)(85)(86)(87). Thus, the transcriptional differences observed between CDV R252-and CDV Ond-infected AMs could be due to a change in the V protein of CDV Ond affecting STAT1 translocation to the nucleus (10,88). Given that morbillivirus attenuation occurs due to mutations of several genes (89), future studies could use the AM model to better delineate the contribution of specific viral molecular determinants to observe transcriptional and phenotypic differences in CDV Ond-and CDV R252-infected AM cultures.
In summary, the present study shows that primary canine AMs can be efficiently and productively infected by CDV. Virus infection induces pro-inflammatory innate immune responses during the initial infection phase, which is dominated by type I IFN signaling. Comparison of virus strains reveals enhanced pro-inflammatory signaling and cytotoxic effects in AMs infected with the attenuated strain of CDV, suggestive of an insufficient induction of antiviral pathways by CDV R252. Despite the induction of a highly activated, antiviral transcriptional signature, CDV infection also induced changes in gene expres sion associated with homeostatic processes of AMs. Disturbance of cellular metabolism and reduced ability to clear apoptotic cells from the alveolar microenvironment might represent factors facilitating secondary bacterial infections upon natural CDV infection. Thus, this study not only identifies AMs as target cells of CDV and confirms previously suggested mechanisms of viral interference with innate immune signaling but also highlights functional disturbances of key AM functions.
A limitation of the current study is the performance of bulk RNA-seq solely at the early initial phase of infection (1 dpi). Therefore, the temporal development of antiviral responses in CDV-infected AMs could not be monitored. In addition, further experiments, including functional assays, are clearly needed to get in-depth insights into disturbed homeostatic AM functions and cell metabolism during morbillivirus infection, as indicated by transcriptomic alterations.
## MATERIALS AND METHODS
## Isolation and culture of primary canine AMs
Isolation of AMs was performed according to a protocol published by Busch and co-workers with slight modifications (90). In brief, AMs were isolated from bronchoal veolar lavage (BAL) fluid of 11 recently deceased dogs. The authors confirm that no animals were sacrificed for the purpose of this study. All dogs used in the present study were dead at the time of submission to routine necropsy service following euthanasia due to animal welfare reasons. Owners declared written consent for sample collection. Histopathologic examination of lung tissue performed by European College of Veterinary Pathologists board-certified veterinary pathologists (AB, MS) ruled out respiratory disease and evidence for viral, bacterial, mycotic, or parasitic infections of the respiratory tract. The collection of BAL was performed using pre-warmed (37°C) phosphate-buffered saline (PBS)-based buffer containing 2 mM EDTA (Biochrom) and 0.5% (vol/vol) fetal bovine serum (Capricorn Scientific). The lavage fluid was diluted, filtered through a 100 µm cell sieve, and centrifuged. Red blood cell lysis was performed by incubating cell pellets for 5 minutes with 10% lysis buffer (0.155 M NH 4 Cl, 0.01 M KHCO 3 , and 0.1 M EDTA), and cells were centrifuged and filtered through a 40 µm cell sieve twice. Cells were seeded at a maximum density of 5 × 10 7 cells per T75-flask in Roswell Park Memorial Institute (RPMI)-medium 1640 with L-alanyl-glutamine and sodium bicarbonate (Thermo Fisher Scientific), supplemented with 1% (wt/vol) sodium pyruvate (Sigma-Aldrich), 1% (vol/vol) penicillin/streptomycin (Sigma-Aldrich), and 10% (vol/vol) fetal bovine serum. After overnight incubation at 37°C with 5% CO 2 , cells were carefully collected from the flasks with a cell scraper and seeded at a density of 1 × 10 5 cells per well in a 24-well plate (Sarstedt) for use in virus infection experiments. Infection was performed 1 day after cell culture preparation in order to prevent cellular adaptation to culture conditions (91). Immunofluorescence staining for Iba1 revealed a purity of histiocytic cells of over 90% (median).
## Cell lines
DH82 cells were obtained from the European Collection of Authenticated Cell Cultures (ECACC No. 94062922) and used as a control for immunofluorescence staining. This cell line originates from a malignant histiocytosis of a dog (92). The derivation of DH82 cells persistently infected with the Ond strain of CDV was performed as previously described (93). Cells were cultured in Minimal Essential Medium with Earle's salts (Thermo Fisher Scientific) with 10% (vol/vol) fetal bovine serum, 1% (vol/vol) penicillin/streptomycin, and 1% (vol/vol) non-essential amino acids (Sigma-Aldrich). Vero cells stably expressing canine SLAM (Vero-SLAM cells) (94) were used to determine the tissue culture infec tious dose-50 (TCID 50 ) of virus present in collected supernatants. Vero-SLAM cells were cultured in Dulbecco's Modified Eagle's medium (Thermo Fisher Scientific) containing 1% (vol/vol) penicillin/streptomycin, 10% fetal bovine serum, and 0.5 mg/mL Zeocin (InvivoGen).
## Viruses
Two different strains of CDV were used for infection. CDV R252 was isolated from spleen tissue of a naturally infected dog (kindly provided by Prof. S. Krakowka, Ohio State University, Columbus, OH, USA). The strain was shown to cause lethal infection with severe pathology of the lymphoid and/or central nervous system in gnotobiotic dogs and ferrets, respectively, and to induce cytopathic effects in primary canine brain cell cultures (95)(96)(97)(98). CDV R252 was propagated in Vero cells, reaching a titer of 10 6.5 TCID 50 /mL. CDV Ond was first isolated in 1939 following an outbreak of canine distemper in North American ranched foxes and has undergone multiple passages in ferrets and eggs (99). The used strain originates from the Belfast variant of CDV Ond, harboring the Y110D mutation in the V protein (10). It was propagated in Vero cells and attained a titer of 10 6 TCID 50 /mL.
## Virus infection of primary cultures
Prior to infection, cells were washed twice in cell culture medium devoid of serum (washing medium), followed by incubation with washing medium containing virus inoculum to enable infection at a multiplicity of infection of 1. After 3 hours of incu bation at 37°C and 5% CO 2 , washing medium was replaced by culture medium and cells were incubated until harvesting at 6 hours pi, 1 dpi, 3 dpi, and 6 dpi. Cells were carefully detached from the wells using a cell scraper (Sarstedt), and an aliquot was taken for Cytospin centrifugation. A centrifugation step was performed to remove the supernatant, and both supernatants and cell pellets were rapidly frozen in liquid nitrogen and stored at -80°C until further use.
## Virus titration
To assess viral loads within supernatant, a TCID 50 assay was performed. Therefore, 3 × 10 4 Vero-SLAM cells were seeded in each well of a 96-well plate in 100 µL growth medium. Supernatants obtained from infection experiments were serially diluted in washing medium with four replicates per dilution and medium-only controls. After 6 days, the presence of cytopathic effects in each well was assessed by light microscopy, and the TCID 50 was calculated according to the method of Spearman-Kärber (100).
## Immunofluorescence labeling
Cytospin slides from the cell suspension aliquots were prepared with the Cytospin 4 Cytocentrifuge (Thermo Fisher Scientific) according to the manual. Subsequently, slides were fixed with 4% (wt/vol) paraformaldehyde for 20 minutes and kept at -80°C until use. Cytospin slides of uninfected and persistently CDV Ond-infected DH82 cells were used as negative and positive controls, respectively. Frozen slides were thawed slightly and washed in PBS containing 0.25% (vol/vol) Triton X-100 (Sigma-Aldrich). Afterward, unspecific binding of the respective secondary antibody was blocked by incubation in 20% (vol/vol) normal goat serum in PBS with 3% (wt/vol) bovine serum albumin (BSA) and 0.25% (vol/vol) Triton X-100 for 15 minutes. Primary antibodies anti-Iba1 (FUJIFILM) and anti-CDV-N (Table S4) were concurrently diluted in PBS with 3% (wt/vol) BSA and 0.25% (vol/vol) Triton followed by an overnight incubation at 4°C. Negative controls were incubated with normal rabbit serum and ascites fluid from non-immunized BALB/c mice instead of the primary antibodies. Slides were washed in PBS with 0.25% (vol/vol) Triton X-100 and incubated in the dark for 2 hours at room temperature with a secon dary polyclonal antibody at a dilution factor of 1:200. Alexa Fluor 488-conjugated goat anti-rabbit (Jackson ImmunoResearch Europe) and Cy3-conjugated goat anti-mouse (Jackson ImmunoResearch Europe) were used to visualize signals. For staining of CC3, a primary labeled antibody was used (Table S4). After thawing, slides were rinsed with PBS thrice, and unspecific binding of the secondary antibody was blocked by incubation in 20% (vol/vol) normal goat serum in PBS with 3% (wt/vol) BSA and 0.25% (vol/vol) Triton X-100 for 60 minutes. Subsequently, the diluted antibody was applied and incubated overnight. Lastly, slides were washed in PBS and mounted with fluorescence mounting medium containing DAPI (Dako), followed by storage at 4°C in the dark. Representative fluorescence pictures for the figures were taken using a Keyence BZ-X800 microscope (Keyence).
## Immunohistochemistry and immunofluorescence for the detection of CDV antigen in lung tissues derived from naturally infected dogs
In order to confirm natural CDV infection of AMs, immunohistochemistry and immu nofluorescence of lung tissue from five naturally infected dogs was performed. Lung tissue was taken at necropsy and fixed in 4% buffered formaldehyde solution, processed routinely, and embedded in paraffin. Sections were cut to a thickness of 2 µm, depar affinized using Roticlear (Carl Roth), and rehydrated through a graded alcohol series. Endogenous peroxidase activity was suppressed by treating the samples with 0.5% hydrogen peroxide in 85% ethanol. Pretreatment involved a 20-minute incubation in citrate buffer (pH 6.0) using a microwave at 800 W, followed by blocking nonspecific bindings with goat normal serum (1:5) for 30 minutes. The primary antibody, mouse anti-CDV-N (Table S4), was applied overnight at 4°C, while negative control samples were treated with ascites fluid from nonimmunized BALB/c mice. The secondary antibody, goat anti-mouse (Vector Laboratories), was applied at a 1:200 dilution in PBS and incubated for 45 minutes at room temperature. This was followed by incubation with the avidin-biotin complex (Vectastain Elite ABC kit, Vector Laboratories) for 20 minutes at room temperature. Antigen-antibody interactions were visualized using 3,3′-diamino benzidine tetrahydrochloride with 0.03% hydrogen peroxide for 5 minutes, and the slides were counterstained with Mayer's hemalum for 30 seconds.
For immunofluorescence double labeling of lung tissue slides, deparaffinization, rehydration, and pretreatment were performed as mentioned above. To minimize non-specific binding, the slides were incubated for 30 minutes with 20% goat nor mal serum in PBS containing 1% BSA and 0.1% Triton X-100 (Sigma-Aldrich). Primary antibodies against Iba1 (Invitrogen) and CDV-N (Table S4) were simultaneously diluted in PBS with 1% BSA and 0.1% Triton X-100 and incubated overnight at 4°C, while negative controls were treated as mentioned above. Signal visualization was achieved using secondary polyclonal antibodies (Alexa Fluor 488-conjugated goat anti-mouse and Cy3-conjugated goat anti-rabbit; Jackson ImmunoResearch Europe) diluted 1:200 in PBS with 1% BSA and 0.1% Triton X-100, followed by a 45-minute incubation at room temperature in the dark. After washing with distilled water, autofluorescence was reduced using the Vector TrueVIEW Autofluorescence Quenching Kit (Vector Laborato ries). Nuclei were visualized with Bisbenzimide Hoechst 33,258 (1:100 in sterile doubledistilled water; Sigma-Aldrich Chemie), and the slides were mounted using fluorescence mounting medium (Dako).
## Digital image analysis
For quantification of Iba1 + cells, infection rates, and expression of MX1, ISG15, and CC3 on Cytospin slides, digitization was performed using an Olympus VS200 Digital slide Scanner (Olympus Europe). Image analysis was performed with the open-source software QuPath (version 0.4.3) (101). Total cell count was determined by using the "cell detection" tool in the DAPI channel. Afterward, object classifiers for both Iba1 + (macro phages), CDV + , MX1 + , ISG15 + , and CC3 + cells were trained to determine the number of single or double-positive cells. To quantify formation for syncytial cells, Iba1 + cells with three or more nuclei on digitized double-labeled Cytospin slides were counted manually using the "counting" tool in QuPath.
## RNA isolation and reverse transcription
Total RNA was isolated using the RNeasy Micro Kit (Qiagen) according to the manufac turer's instructions including an on-column DNA digestion. RNA quality and concentra tion were measured with a Multiskan GO microplate spectrophotometer (µDrop plate, SkanIt software version 5.0.0.42, Thermo Fisher Scientific), and RNA was stored at -80°C until further use. For transcription of total RNA in complementary DNA (cDNA), the Sensiscript RT Kit (Qiagen) supplemented with RNaseOUT Recombinant Ribonuclease Inhibitor (Thermo Fisher Scientific) and random primers (Promega Corporation) was used following the supplier's protocol.
## Generation of standard dilutions
To amplify gene products for the generation of standard dilutions, primer sequences for glyceraldehyde-3-phosphate dehydrogenase (GAPDH), TNF-α, and IL-6 were obtained from previous studies (93,102), (Table S5). Plasmids based on the pEX-A128 vector containing under 300 bp of the respective canine cDNA genome sequences of IL-1β, IL-8, IL-10, IL-12, TGF-β, and IFN-γ (purchased from Eurofins Genomics) were used to generate standard dilutions (Table S6). In addition, cDNA which had been isolated from naturally CDV-infected canine lung tissue (GAPDH) or lymph nodes from uninfected dogs (TNF-α, IL-6) was also used to produce standard dilutions via PCR. The mastermix for PCR amplification contained Taq DNA Polymerase (Invitrogen, Thermo Fisher Scientific) with 1.5 mmol/L MgCl2, 0.2 mmol/L dNTP mix (New England Biolabs), and 300 nmol/L of each primer. A T-Gradient thermocycler (Biometra) was used with 40 cycles of an initial denaturing step of 94°C, an annealing step of 58°C (TNF-α, IL-6) or 59°C (GAPDH) for 45 seconds and elongation at 72°C for 40 seconds. Visualization of PCR was achieved by agarose gel electrophoresis. The respective band was extracted with NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel) according to the manufacturer's instructions. Absorbance at 260 nm was measured with a Multiskan GO microplate spectrophotome ter to calculate the DNA concentration. PCR products or plasmids were diluted in DNase and RNase-free water to concentrations ranging from 10 2 to 10 8 copies/µL.
## Reverse transcription quantitative PCR
For RT-qPCR detection of canine GAPDH, TNF-α, IL-6, IL-10, IL-12, TGF-β, and IFN-γ, previously published primer sequences were used (48,(103)(104)(105) (Table S7). The Primer-BLAST software tool by the National Library of Medicine ( 106) was used to design analogous primers for IL-1β and IL-8 (Eurofins Genomics) (Table S7). RT-qPCR assays were performed using the AriaMx Real-Time PCR System (Agilent Technologies; Agilent Aria software version 1.71). The standard dilution series was included in every experimental setup to determine the copy numbers using the Brilliant III Ultra-Fast SYBR Green QPCR Master Mix (Agilent Technologies) according to the manufacturer's instructions. Primers were added at a concentration of 200 nmol/L and carboxy-X-rhodamine served as a reference dye. Annealing steps were performed at 56°C (IL-1β), 60 °C (TNF-α), or 64 °C (GAPDH). Calculated copy numbers were normalized with the housekeeping gene GAPDH.
## Total RNA-seq
Quality and integrity of total RNA was assessed using an Agilent Technologies 2100 Bioanalyzer (Agilent Technologies). The RNA-seq library was generated from 50 ng total RNA using NEBNext Single Cell/Low Input RNA Library Prep Kit for Illumina (New England BioLabs) according to manufacturer's protocols. The libraries were sequenced on an Illumina NovaSeq 6000 using NovaSeq 6000 S1 Reagent Kit (100 cycles, paired-end run) with an average of 3 × 10 7 reads per RNA sample. A quality report was gener ated by FASTQC tool. Each sequence in the raw FASTQ files was trimmed on base call quality and sequencing adapter contamination using fastq-mcf (http://expressiona nalysis.github.io/ea-utils/). Reads shorter than 15 bp were removed from FASTQ files. Trimmed reads were aligned to reference genomes (dog: ROS_Cfam_1.0, ensemble database version 105; CDV strain R252: Genbank Acc. KF640687; CDV strain Ond: Genbank Acc. AF378705) using open source short read aligner STAR (107). The sub sequent analysis was performed using R version 4.2.1 (23 June 2022; https://www.Rproject.org/). Raw count table was annotated using the biomaRt package (version 2.54.1) (108). For filtering, features with annotation type "rRNA" or "pseudogene" were removed from the data set and libraries in which the size differed by more than three standard deviations. For normalization of library sizes, the calcNormFactors function was used to find a set of scaling factors for the library sizes that minimizes the log-fold changes between the samples for most genes (109). Differential expression analysis for multi factor experiments was performed utilizing the generalized linear models (glm)-based statistical method of the edgeR Bioconductor package (version 3.38.4) (110). Disper sion was estimated using the Cox-Reid profile-adjusted likelihood method. Trended dispersions were estimated prior to estimating tagwise dispersions. After dispersion estimation, negative binomial glm were fitted to the data, after which the DEGs were determined using quasi-likelihood F-test. The statistical analysis to identify differential gene expression was performed using a multivariate regression model. In the following, a heatmap plot (threshold: absolute FC > 1, false discovery rate [FDR] < 0.05) was generated using the ComplexHeatmap package (version 2.12.1) (111), and GO enrich ment analysis ("biological process") and KEGG pathway analysis were performed from DEGs with the ClusterProfiler function (version 4.6.2) (112). Heatmaps of smaller subsets of genes were generated in R studio (version 2023.12.1) with the function heatmap2 of package gplots (version 3.1.3; https://github.com/talgalili/gplots). Venn's diagrams were created using R package VennDiagram (version 1.7.3) (113).
## LDH assay
To determine cell viability, LDH activity in the supernatants was determined in triplicate using the Cytotoxicity Detection Kit (Roche) according to the manufacturer's instructions. The absorbance at 492 nm and 630 nm (reference) was measured three times using a microplate reader (Fluostar Optima, BMG Labtech). Arithmetic means were calculated for each supernatant and the mean of the medium-only control as well as the absorbance value of the reference measurement were subtracted from each value.
## Quantitative sandwich-ELISA
Sandwich-ELISAs were used to quantify TNF-α and IFN-α secretions in supernatants of AM cultures, following the manufacturer's instructions. The Quantikine ELISA Canine TNF-α Immunoassay (Bio-Techne) and the canine IFN alpha ELISA Kit (Invitrogen) were used. All standards, samples, and controls were measured in duplicates. For antigen binding, samples, standards, and controls were pipetted in 96-well plates coated with specific antibodies to canine TNF-α and IFN-α, respectively, followed by an incubation period. For measurement of TNF-α, samples were diluted 1:3. Bound antigen was labeled using biotinylated detection antibodies and streptavidin-horseradish peroxi dase. Colorimetric quantification was achieved by adding a chromogen and measur ing absorbance at 450 nm and 540 nm (background) with the SpectraMax ABS Plus reader (Molecular Devices). Using Softmax Pro software (version 7.2, Molecular Devices), four-parameter logistic standard curves were fitted to the standards, and absorbance values were plotted to determine TNF-α and IFN-α concentration, respectively, after subtraction of background absorbance. Mean values were calculated of all duplicates, and TNF-α values were multiplied with the dilution factor.
## Apoptosis/necrosis assay
To investigate cell death in CDV-infected AMs, differential staining of cells to distinguish between live, apoptotic, and necrotic cells was performed, using the Apoptosis/Necrosis Assay Kit (Abcam) according to the manufacturer's instructions with slight modifications. Briefly, 50,000 cells per well were grown in a black 96-well plate with clear bottom, and at 1 dpi, staining was performed. Two washing steps with RPMI 1640 medium without phenol red (Thermo Fisher Scientific) including centrifugation of the plate were performed and subsequently, a staining mix containing 2 µL/well Apopxin Green Indicator, 1 µL/well 7-AAD 200×, and 1 µL/well CytoCalcein Violet 450 was added. After a 45-minute incubation period in the dark, the plate was washed three times. After a final centrifugation step, five images per well were taken with the Keyence BZ-X800 microscope at 10× magnification, and the number of live, apoptotic, and necrotic cells was determined using the manual counting tool in QuPath version 0.5.1.
## Statistical analysis and graph design
R statistic program was used to perform statistical analysis, and box plots were generated with GraphPad Prism 10 (GraphPad Software). The presence of significant differences was tested using the non-parametric Kruskal-Wallis test. Subsequently, pairwise comparisons among groups were performed with multiple two-tailed Mann-Whitney U tests with the FDR adjustment of Benjamini and Hochberg for multiple group comparisons. Significance was assumed at a q value of 0.05 or smaller.
## References
1. Rima, Balkema-Buschmann, Dundon et al. (2019) "ICTV virus taxonomy profile: Paramyxoviridae" *J Gen Virol*
2. Beineke, Baumgärtner, Wohlsein (2015) "Cross-species transmis sion of canine distemper virus-an update" *One Health*
3. Martinez-Gutierrez, Ruiz-Saenz (2016) "Diversity of susceptible hosts in canine distemper virus infection: a systematic review and data synthesis" *BMC Vet Res*
4. Kennedy, Earle, Omar et al. (2019) "Canine and phocine distemper viruses: Global spread and genetic basis of jumping species barriers" *Viruses*
5. Feng, Yu, Wang et al. (2016) "Fatal canine distemper virus infection of giant pandas in China" *Sci Rep*
6. Yoshikawa, Ochikubo, Matsubara et al. (1989) "Natural infection with canine distemper virus in a Japanese monkey (Macaca fuscata)" *Vet Microbiol*
7. Qiu, Zheng, Zhang et al. (2011) "Canine distemper outbreak in rhesus monkeys" *China. Emerg Infect Dis*
8. Sakai, Nagata, Ami et al. (2013) "Lethal canine distemper virus outbreak in cynomolgus monkeys in Japan in 2008" *J Virol*
9. Von Messling, Svitek, Cattaneo (2006) "Receptor (SLAM [CD150]) recognition and the V protein sustain swift lymphocyte-based invasion of mucosal tissue and lymphatic organs by a morbillivirus" *J Virol*
10. Röthlisberger, Wiener, Schweizer et al. (2010) "Two domains of the V protein of virulent canine distemper virus selectively inhibit STAT1 and STAT2 nuclear import" *J Virol*
11. Svitek, Gerhauser, Goncalves et al. (2014) "Morbillivirus control of the interferon response: relevance of STAT2 and mda5 but not STAT1 for canine distemper virus virulence in ferrets" *J Virol*
12. Ramachandran, Parisien, Horvath (2008) "STAT2 is a primary target for measles virus V protein-mediated alpha/beta interferon signaling inhibition" *J Virol*
13. Motz, Schuhmann, Kirchhofer et al. (2013) "Paramyxovirus V proteins disrupt the fold of the RNA sensor MDA5 to inhibit antiviral signaling" *Science*
14. De Vries, Ludlow, De et al. (2017) "Delineating morbillivirus entry, dissemination and airborne transmission by studying in vivo competition of multicolor canine distemper viruses in ferrets" *PLoS Pathog*
15. Sawatsky, Cattaneo, Messling (2018) "Canine distemper virus spread and transmission to naive ferrets: selective pressure on signaling lymphocyte activation molecule-dependent entry" *J Virol*
16. Lemon, De Vries, Mesman et al. (2011) "Early target cells of measles virus after aerosol infection of non-human primates" *PLoS Pathog*
17. Ferreira, Frenzke, Leonard et al. (2010) "Measles virus infection of alveolar macrophages and dendritic cells precedes spread to lymphatic organs in transgenic mice expressing human signaling lymphocytic activation molecule (SLAM, CD150)" *J Virol*
18. Shin, Chludzinski, Wu et al. (2022) "Overcoming the barrier of the respiratory epithelium during canine distemper virus infection" *mBio*
19. Von Messling, Milosevic, Cattaneo (2004) "Tropism illuminated: lymphocyte-based pathways blazed by lethal morbillivirus through the host immune system" *Proc Natl Acad Sci*
20. Krakowka, Higgins, Koestner (1980) "Canine distemper virus: review of structural and functional modulations in lymphoid tissues" *Am J Vet Res*
21. De Swart, De Vries, Rennick et al. (2017) "Needle-free delivery of measles virus vaccine to the lower respiratory tract of non-human primates elicits optimal immunity and protection" *NPJ Vaccines*
22. Evren, Ringqvist, Doisne et al. (2022) "CD116+ fetal precursors migrate to the perinatal lung and give rise to human alveolar macrophages" *J Exp Med*
23. Yona, Kim, Wolf et al. (2013) "Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeostasis" *Immunity*
24. Aegerter, Lambrecht, Jakubzick (2022) "Biology of lung macrophages in health and disease" *Immunity*
25. Blériot, Chakarov, Ginhoux (2020) "Determinants of resident tissue macrophage identity and function" *Immunity*
26. Gautier, Shay, Miller et al. (2012) "Gene-expression profiles and transcriptional regulatory pathways that underlie the identity and diversity of mouse tissue macrophages" *Nat Immunol*
27. Bissonnette, Debley, Ziegler (2020) "Cross-talk between alveolar macrophages and lung epithelial cells is essential to maintain lung homeostasis" *Front Immunol*
28. (2026) *Full-Length Text Journal of Virology*
29. Goritzka, Makris, Kausar et al. (2015) "Alveolar macrophage-derived type I interferons orchestrate innate immunity to RSV through recruitment of antiviral monocytes" *J Exp Med*
30. Mould, Jackson, Henson et al. (2019) "Single cell RNA sequencing identifies unique inflammatory airspace macrophage subsets" *JCI Insight*
31. Kumagai, Takeuchi, Kato et al. (2007) "Alveolar macrophages are the primary interferon-alpha producer in pulmonary infection with RNA viruses" *Immunity*
32. Divangahi, King, Pernet (2015) "Alveolar macrophages and type I IFN in airway homeostasis and immunity" *Trends Immunol*
33. De Vries, Lemon, Ludlow et al. (2010) "In vivo tropism of attenuated and pathogenic measles virus expressing green fluorescent protein in macaques" *J Virol*
34. Gonzales-Viera, Woolard, Keel (2023) "Lung and lymph node explants to study the interaction between host cells and canine distemper virus" *Res Vet Sci*
35. Oeckinghaus, Ghosh (2009) "The NF-κB family of transcription factors and its regulation" *Cold Spring Harb Perspect Biol*
36. Schoggins (2019) "Interferon-stimulated genes: what do they all do?" *Annu Rev Virol*
37. Rusinova, Forster, Yu et al. (2013) "INTERFEROME v2.0: an updated database of annotated interferon-regulated genes" *Nucleic Acids Res*
38. Wagers, Lowe, Kansas (1996) "An important role for the α1,3 fucosyltransferase, FucT-VII" *Blood*
39. Discipio, Daffern, Schraufstätter et al. (1998) "Human polymorphonuclear leukocytes adhere to complement factor H through an interaction that involves α M β 2 (CD11b/CD18)" *J Immunol*
40. Hoppe, Swanson (2004) "Cdc42, Rac1, and Rac2 display distinct patterns of activation during phagocytosis" *Mol Biol Cell*
41. Cabec, Carréno, Moisand et al. (2002) "Complement receptor 3 (CD11b/CD18) mediates type I and type II phagocytosis during nonopsonic and opsonic phagocytosis, respectively" *J Immunol*
42. Hordijk (2006) "Regulation of NADPH oxidases: the role of Rac proteins" *Circ Res*
43. Baranova, Kurlander, Bocharov et al. (2008) "Role of human CD36 in bacterial recognition, phagocytosis, and pathogen-induced JNK-mediated signaling" *J Immunol*
44. Areschoug, Gordon (2009) "Scavenger receptors: role in innate immunity and microbial pathogenesis" *Cell Microbiol*
45. Fadok, Warner, Bratton et al. (1998) "CD36 is required for phagocytosis of apoptotic cells by human macrophages that use either a phosphatidylserine receptor or the vitronectin receptor" *J Immunol*
46. Nagy, Tontonoz, Alvarez et al. (1998) "Oxidized LDL regulates macrophage gene expression through ligand activation of PPARgamma" *Cell*
47. Tontonoz, Nagy, Alvarez et al. (1998) "PPARgamma promotes monocyte/macrophage differentiation and uptake of oxidized LDL" *Cell*
48. Huang, Smith, Everts et al. (2016) "Metabolic reprogramming mediated by the mTORC2-IRF4 signaling axis is essential for macrophage alternative activation" *Immunity*
49. Chludzinski, Klemens, Ciurkiewicz et al. (2022) "Phenotypic and transcriptional changes of pulmonary immune responses in dogs following canine distemper virus infection" *Int J Mol Sci*
50. Haller, Staeheli, Schwemmle et al. (2015) "Mx GTPases: dynamin-like antiviral machines of innate immunity" *Trends Microbiol*
51. Balachandran, Roberts, Brown et al. (2000) "Essential role for the dsRNA-dependent protein kinase PKR in innate immunity to viral infection" *Immunity*
52. Kurokawa, Iankov, Galanis (2019) "A key anti-viral protein, RSAD2/ VIPERIN, restricts the release of measles virus from infected cells" *Virus Res*
53. Li, Banerjee, Wang et al. (2016) "Activation of RNase L is dependent on OAS3 expression during infection with diverse human viruses" *Proc Natl Acad Sci*
54. Balachandran, Roberts, Kipperman et al. (2000) "Alpha/beta interferons potentiate virusinduced apoptosis through activation of the FADD/Caspase-8 death signaling pathway" *J Virol*
55. Benedict, Norris, Ware (2002) "To kill or be killed: viral evasion of apoptosis" *Nat Immunol*
56. Savill, Dransfield, Gregory et al. (2002) "A blast from the past: clearance of apoptotic cells regulates immune responses" *Nat Rev Immunol*
57. Pillet, Messling (2009) "Canine distemper virus selectively inhibits apoptosis progression in infected immune cells" *J Virol*
58. Kajita, Katayama, Murata et al. (2006) "Canine distemper virus induces apoptosis through caspase-3 and -8 activation in vero cells" *J Vet Med B Infect Dis Vet Public Health*
59. Wong, Goeddel (1986) "Tumour necrosis factors alpha and beta inhibit virus replication and synergize with interferons" *Nature*
60. Rath, Aggarwal (1999) "TNF-induced signaling in apoptosis" *J Clin Immunol*
61. Upton, Chan (2014) "Staying alive: cell death in antiviral immunity" *Mol Cell*
62. Schneider, Nobs, Heer et al. (2014) "Alveolar macrophages are essential for protection from respiratory failure and associated morbidity following influenza virus infection" *PLoS Pathog*
63. Kolli, Gupta, Sbrana et al. (2014) "Alveolar macrophages contribute to the pathogenesis of human metapneumovirus infection while protecting against respiratory syncytial virus infection" *Am J Respir Cell Mol Biol*
64. Shin, Seifert, Kato et al. (2006) "Virus-induced type I IFN stimulates generation of immunoproteasomes at the site of infection" *J Clin Invest*
65. Aki, Shimbara, Takashina et al. (1994) "Interferon-γ induces different subunit organizations and functional diversity of proteasomes" *J Biochem*
66. Blumenthal, Campbell, Hwang et al. (2001) "Human alveolar macrophages induce functional inactivation in antigen-specific CD4 T cells" *J Allergy Clin Immunol*
67. (2026) *Full-Length Text Journal of Virology*
68. Chelen, Fang, Freeman et al. (1995) "Human alveolar macro phages present antigen ineffectively due to defective expression of B7 costimulatory cell surface molecules" *J Clin Invest*
69. Holt, Oliver, Bilyk et al. (1993) "Downregulation of the antigen presenting cell function(s) of pulmonary dendritic cells in vivo by resident alveolar macrophages" *J Exp Med*
70. Melchjorsen, Sørensen, Paludan (2003) "Expression and function of chemokines during viral infections: from molecular mechanisms to in vivo function" *J Leukoc Biol*
71. Guarda, Braun, Staehli et al. (2011) "Type I interferon inhibits interleukin-1 production and inflammasome activation" *Immunity*
72. Díaz-Pino, Rice, Felipe et al. (2024) "Type I interferon regulates interleukin-1beta and IL-18 production and secretion in human macrophages" *Life Sci Alliance*
73. Loevenich, Montaldo, Wickenhagen et al. (2023) "Human metapneumovirus driven IFN-β production antagonizes macrophage transcriptional induction of IL1-β in response to bacterial pathogens" *Front Immunol*
74. Von Messling, Zimmer, Herrler et al. (2001) "The hemagglutinin of canine distemper virus determines tropism and cytopathogenicity" *J Virol*
75. Herschke, Plumet, Duhen et al. (2007) "Cell-cell fusion induced by measles virus amplifies the type I interferon response" *J Virol*
76. Ariav, Ch'ng, Christofk et al. (2021) "Targeting nucleotide metabolism as the nexus of viral infections, cancer, and the immune response" *Sci Adv*
77. Kulikauskaite, Wack (2020) "Teaching old dogs new tricks? The Plasticity of lung alveolar macrophage subsets" *Trends Immunol*
78. Svedberg, Brown, Krauss et al. (2019) "The lung environment controls alveolar macrophage metabolism and responsiveness in type 2 inflammation" *Nat Immunol*
79. Biswas, Mantovani (2012) "Orchestration of metabolism by macrophages" *Cell Metab*
80. Rodríguez-Prados, Través, Cuenca et al. (2010) "Substrate fate in activated macrophages: a comparison between innate, classic, and alternative activation" *J Immunol*
81. Jha, Huang, Sergushichev et al. (2015) "Network integration of parallel metabolic and transcriptional data reveals metabolic modules that regulate macrophage polarization" *Immunity*
82. Shibata, Makino, Ogata et al. (2020) "Respiratory syncytial virus infection exacerbates pneumococcal pneumonia via Gas6/Axlmediated macrophage polarization" *J Clin Invest*
83. Baasch, Giansanti, Kolter et al. (2021) "Cytomegalovirus subverts macrophage identity" *Cell*
84. Chinnakannan, Nanda, Baron (2013) "Morbillivirus v proteins exhibit multiple mechanisms to block type 1 and type 2 interferon signalling pathways" *PLoS One*
85. Palosaari, Parisien, Rodriguez et al. (2003) "STAT protein interference and suppression of cytokine signal transduction by measles virus V protein" *J Virol*
86. Childs, Stock, Ross et al. (2007) "Mda-5, but not RIG-I, is a common target for paramyxovirus V proteins" *Virology (Auckl)*
87. Andrejeva, Childs, Young et al. (2004) "The V proteins of paramyxoviruses bind the IFNinducible RNA helicase, mda-5, and inhibit its activation of the IFN-beta promoter" *Proc Natl Acad Sci*
88. Parisien, Bamming, Komuro et al. (2009) "A shared interface mediates paramyxovirus interference with antiviral RNA helicases MDA5 and LGP2" *J Virol*
89. Poole, He, Lamb et al. (2002) "The V proteins of simian virus 5 and other paramyxoviruses inhibit induction of interferon-beta" *Virology (Auckl)*
90. Wyss, Gradauskaite, Ebert et al. (2022) "Efficient recovery of attenuated canine distemper virus from cDNA" *Virus Res*
91. Liu, Wu, Li et al. (2016) "Evolutionary characteristics of morbilliviruses during serial passages in vitro: gradual attenuation of virus virulence" *Comp Immunol Microbiol Infect Dis*
92. Busch, Favret, Geirsdóttir et al. (2019) "Isolation and long-term cultivation of mouse alveolar macrophages" *Bio Protoc*
93. Subramanian, Busch, Molawi et al. (2022) "Long-term culture-expanded alveolar macrophages restore their full epigenetic identity after transfer in vivo" *Nat Immunol*
94. Wellman, Krakowka, Jacobs et al. (1988) "A macrophagemonocyte cell line from a dog with malignant histiocytosis" *In Vitro Cell Dev Biol*
95. Puff, Krudewig, Imbschweiler et al. (2009) "Influence of persistent canine distemper virus infection on expression of RECK, matrix-metalloproteinases and their inhibitors in a canine macrophage/monocytic tumour cell line (DH82)" *Vet J*
96. Von Messling, Springfeld, Devaux et al. (2003) "A ferret model of canine distemper virus virulence and immunosuppression" *J Virol*
97. Mccullough, Krakowka, Koestner (1974) "Experimental canine distemper virus-induced lymphoid depletion" *Am J Pathol*
98. Ludlow, Nguyen, Silin et al. (2012) "Recombinant canine distemper virus strain Snyder Hill expressing green or red fluorescent proteins causes meningoencephalitis in the ferret" *J Virol*
99. Orlando, Imbschweiler, Gerhauser et al. (2008) "In vitro characterization and preferential infection by canine distemper virus of glial precursors with Schwann cell characteristics from adult canine brain" *Neuropathol Appl Neurobiol*
100. Seehusen, Orlando, Wewetzer et al. (2007) "Vimentinpositive astrocytes in canine distemper: a target for canine distemper virus especially in chronic demyelinating lesions?" *Acta Neuropathol*
101. Martella, Blixenkrone-Møller, Elia et al. (2010) "Lights and shades on an historical vaccine canine distemper virus, the Rockborn strain" *Vaccine (Auckl)*
102. (2026) *Full-Length Text Journal of Virology*
103. Kärber (1931) "Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche" *Archiv f experiment Pathol u Pharmakol*
104. Bankhead, Loughrey, Fernández et al. (2017) "QuPath: open source software for digital pathology image analysis" *Sci Rep*
105. Gröne, Frisk, Baumgärtner (1998) "Cytokine mRNA expression in whole blood samples from dogs with natural canine distemper virus infection" *Vet Immunol Immunopathol*
106. Smolinski, Leverkoehne, Samson-Himmelstjerna et al. (2005) "Impact of formalin-fixation and paraffin-embedding on the ratio between mRNA copy numbers of differently expressed genes" *Histochem Cell Biol*
107. Spitzbarth, Bock, Haist et al. (2011) "Prominent microglial activation in the early proinflammatory immune response in naturally occurring canine spinal cord injury" *J Neuropathol Exp Neurol*
108. Schwartz, Puff, Stein et al. (2011) "Pathogenetic factors for excessive IgA production: Th2-dominated immune response in canine steroid-responsive meningitis-arteritis" *Vet J*
109. Ye, Coulouris, Zaretskaya et al. (2012) "Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction" *BMC Bioinformatics*
110. Dobin, Davis, Schlesinger et al. (2013) "STAR: ultrafast universal RNA-seq aligner" *Bioinformatics*
111. Durinck, Spellman, Birney et al. (2009) "Mapping identifiers for the integration of genomic datasets with the R/Bioconductor package biomaRt" *Nat Protoc*
112. Robinson, Oshlack (2010) "A scaling normalization method for differential expression analysis of RNA-seq data" *Genome Biol*
113. Chen, Lun, Smyth (2016) "From reads to genes to pathways: differential expression analysis of RNA-Seq experiments using Rsubread and the edgeR quasi-likelihood pipeline" *F1000Res*
114. Gu, Eils, Schlesner (2016) "Complex heatmaps reveal patterns and correlations in multidimensional genomic data" *Bioinformatics*
115. Wu, Hu, Xu et al. "a universal enrichment tool for interpreting omics data" *The Innovation*
116. Chen, Boutros (2011) "VennDiagram: a package for the generation of highly-customizable Venn and Euler diagrams in R" *BMC Bioinformat ics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12707402&blobtype=pdf | # Letters to the Editor
Departments Of Ophthalmology, Microbiology, Thuhin Krishna
## Prevalence of herpes simplex virus 1 and 2 in culture-negative suppurative keratitis using polymerase chain reaction -A pilot study
Dear Editor, Around 8 million people are blind due to corneal diseases globally. [1] Corneal blindness, the third leading cause of blindness in India, is most commonly due to microbial keratitis. [2,3] Corneal scraping for smear and culture is routinely done to establish etiology. However, culture positivity is seen only in 35-55% of patients with suppurative keratitis. [4,5] Since there is a significant proportion of smear/culture negativity, it is necessary to explore other potential infectious agents other than bacteria/fungi, especially viruses like the Herpes Simplex Virus (HSV) that can mimic bacterial or fungal suppurative corneal ulcers. Using newer molecular diagnostic tools like polymerase chain reaction (PCR) can enhance the identification of viral etiology, such as HSV 1 and HSV 2, in patients with culture-negative suppurative keratitis. We report the prevalence of HSV 1 and 2 in culture-negative suppurative keratitis using qualitative real-time duplex PCR. To the best of our knowledge, there is no prospective study data on the presence of HSV in culture-negative suppurative keratitis. This prospective observational case series was conducted as a pilot study at the Departments of Ophthalmology, Clinical Microbiology, and Clinical Virology in a tertiary care hospital in South India and was approved by the institutional review board (IRB Min no 9265). We included patients who presented with scrapable suppurative corneal ulcers, from March 14, 2015 to September 30, 2015, that were smear and culture-negative. Informed consent was obtained by the Principal Investigator (PI) from all the study participants. Suppurative corneal ulcer was defined as the presence of a corneal epithelial defect with an underlying stromal infiltrate measuring at least 1 mm 2 in size, with or without hypopyon. Patients were excluded if corneal scraping was not medically indicated, in the event of keratitis cases with impending perforation, or if the ulcer was smaller than 1 mm, as these conditions typically yield insufficient material for reliable testing. Children (under 15 years of age) and pregnant individuals were also excluded from the study. The time taken to heal was the time taken for complete closure of the epithelial defect and the absence of stromal inflammatory cells in the cornea. A 2-year retrospective audit (July 2012 to May 2014) examined 399 corneal scraping specimens from clinically nonviral, suppurative keratitis cases. The culture positivity rate was 44%, consistent with other studies (35-70%). Anticipating 30 culture-negative ulcers within 6 months among eligible adult patients, our pilot study with a sample size of 30 was conducted. This study aimed to determine the prevalence rate of HSV positivity with a 95% confidence interval (CI) among culture-negative suppurative keratitis in a hospital setting using PCR.
Corneal ulcer scraping was then performed, and samples were collected for smear and culture. Simultaneously, corneal scrapings were obtained in all cases for PCR analysis. Real-time PCR assay was used to detect HSV as its sensitivity is in the range of 85-100%, with a specificity of 70-75%. The samples were stored at -80°C until required for virus detection by real-time PCR assay. When no growth was noticed, bacterial cultures were usually declared negative in 5 days, while fungal cultures were usually declared negative in 2 weeks. However, in this study, we had set a waiting period of 5 days for the culture report for both bacterial and fungal growth. Empirical therapy based on clinical suspicion was started during this period. After 5 days, preserved samples of negative smear and culture reports were processed for virus isolation by real-time PCR assay in these patients to avoid delay in treatment with antivirals if found positive. The time interval between scraping and availability of PCR results was the sum of 5 days of no response and additional days from the time PCR testing was initiated to the results being available. The culture-negative samples to be tested for PCR were arranged on a PCR rack with a negative control (water) included after every third sample to detect potential cross-contamination. An external quality control (EQC) sample consisting of a low positive previously tested sample was included in every test run to ensure that the PCR conditions did not miss low positive reactive samples. DNA extraction was done as per the manufacturer's protocol of the Qiagen DNA Blood Mini kit (Qiagen, GmBH, Germany). The treatment strategy for culture and smear-negative cases in this study included an empirical approach with antibacterial (Cefazoline 5% and Fortified Gentamycin 1.4% eye drops) and/or antifungal (Natamycin 5% eye drops) agents for culture-negative corneal ulcers based on clinical presentation. If a patient tested positive for HSV 1 and 2 by PCR and showed no improvement or worsened with the initial treatment, they were additionally administered oral Acyclovir 400 mg (GlaxoSmithKline India) five times a day for 14 days and then 400 mg BD maintenance (tapering), Acyclovir ointment 3% (FDC Limited India) five times a day, tapered based on clinical response, and topical prednisolone acetate 1% (Allergan India) 4 hourly for a week and tapered depending on the response. Antifungals were stopped after starting antivirals and topical steroids. While we acknowledge the importance of other pathogens, this study was specifically designed to examine the role of HSV in culture-negative corneal ulcers.
During the study period, 23,282 patients reported to the ophthalmology department for an eye check-up. Of them, 96 patients were diagnosed clinically with suppurative (clinically nonviral) keratitis. Thus, 0.412% of all patients had suppurative keratitis. 71 patients were included in the study as per inclusion/ exclusion criteria. A detailed flow of the study is shown in Fig. 1A. Forty-three (61%) had confirmed infectious keratitis based on positive smear and/or culture results; fungi were identified in 31 (72%) and bacteria in 11 (26%). Acanthamoeba, a less common organism, was found in one patient (2% of positive cases). There was no discrepancy in the organism between the smear and culture reports. Out of 31 cases of fungal corneal ulcers, 21 were positive on both smear and culture, and 5 showed no hyphae on smear but grew on culture. Antivirals were also started empirically in 4 cases (5.6%). Some patients were on more than one medication. Among the four patients on antivirals in the current episode, 3 (75%) turned out to be positive on smear and/or culture for bacteria or fungus.
Twenty-eight smear and culture-negative samples were processed for duplex PCR (both HSV 1 and 2). Of these, 25 cases were negative for both HSV 1 and 2, while 3 cases were positive for HSV 1. None of the cases were positive for HSV 2. In this pilot study, we identified three cases that were HSV-positive, giving a prevalence of 10.7% (95% CI 2.3%-28.2%) using the Binomial (Clopper-Pearson) 'exact' method. The culture positivity rate in our study was 60%. All three HSV positive keratitis cases were clinically suppurative keratitis. None of them had any clinical features of viral keratitis, like old scarring with vascularization, significantly reduced corneal sensation, or rolled out margins. 48 (67.6%) patients received empirical antibacterial and 27 (38%) received empirical antifungal agents. None of these patients was on empirical antiviral therapy. Letters to the Editor Characteristics of the smear/culture-negative patients with suppurative keratitis are shown in Table 1. Past history of red eyes (P = 0.019), corneal ulcers (P = 0.006), and history of antiviral use were associated with HSV positivity (P = 0.001). The presence of a corneal scar was also significantly associated with HSV positivity (P = 0.001). While two-thirds of HSV-positive patients presented with red eye, a history of red eye or corneal ulcer did not significantly correlate with smear-negative suppurative keratitis (P = 0.156 and P = 0.154, respectively). Specifically, among eight patients with a history of red eye, 3 (37.5%) were smear or culture-positive. Of 6 patients with a prior corneal ulcer, 2 (33.4%) had positive smear or culture results (one bacterial, one fungal).
Although two of the three HSV-positive patients had a history of antiviral use, none were using antivirals during the current episode due to a presumed bacterial or fungal etiology. Furthermore, only one of the 28 culture-negative patients (none of whom were HSV-positive) was on antivirals, demonstrating no significant relationship between antiviral use at the current episode and HSV positivity (P = 0.724). Of the 28 culture negatives, 11 cases were suspected to be clinically bacterial, 14 were clinically fungal, and 3 could not be commented on. Of the 3 HSV-positive cases, two were presumed to be clinically bacterial and one was presumed to be fungal, as shown in Fig. 1B. All three ulcers had remained static with no improvement at the time of initiation of antiviral therapy after obtaining PCR results. In our study, the time taken to heal was 47 days, 58 days, and 51 days, respectively, in the three HSV positive patients, with an average of 52 days after starting oral and topical Acyclovir along with steroid drops. This pilot study offers valuable preliminary insights into the potential role of viral etiology of culture-negative suppurative keratitis. Herpes keratitis can present as suppurative keratitis without the classical clinical characteristics of a viral ulcer. Looking for viral etiology in all culture-negative cases may not be feasible. Past history of red eyes or corneal ulcer, past history of antiviral use, and presence of corneal scar are associated with viral etiology. On encountering patients with these presenting features in remote places where any testing may be unavailable, a high index of suspicion for viral etiology must be considered, and prompt treatment with antivirals may help. In a tertiary referral center that has testing facilities, if the smear and culture have been negative, viral etiology should be considered, PCR tested for HSV 1 if possible, and appropriate treatment should be started. Adding PCR to selected cases of culture-negative suppurative keratitis improves recovery rates for microbial agents in refractory keratitis that do not respond to conventional treatment.
The small sample size and focus on a single viral agent limit the generalizability of our findings. Further research with a larger, more comprehensive study design is necessary to confirm these findings and explore the potential role of other viral pathogens, fungi, and eubacteria. Additionally, future studies should incorporate longer follow-up periods, particularly for culture-negative and HSV-negative patients, to better understand the long-term clinical course of these cases.
Financial support and sponsorship: Nil.
## References
1. Anitha, Tandon, Shah et al. (2023) "Corneal blindness and eye banking: Current strategies and best practices" *Indian J Ophthalmol*
2. (2020) "National Blindness and Visual Impairment Survey 2015-2019. India Vis. Atlas NPCB2019"
3. Oliva, Schottman, Gulati (2012) "Turning the tide of corneal blindness" *Indian J Ophthalmol*
4. Ting, Ho, Deshmukh et al. (2021) "Infectious keratitis: An update on epidemiology, causative microorganisms, risk factors, and antimicrobial resistance" *Eye*
5. Srinivasan, Mascarenhas, Prashanth (2008) "Distinguishing infective versus noninfective keratitis" *Indian J Ophthalmol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793446&blobtype=pdf | A Migliavacca
## Abstract
10 IU/mL decline from baseline [BL]), alanine aminotransferase (ALT) normalization, change from BL in ALT, combined response (VR and ALT normalization), and change from BL in HBV DNA and liver stiffness, all assessed over 96 weeks with vs without NAs.
Background. CDC's Vessel Sanitation Program (VSP) works with the cruise industry to monitor and investigate outbreaks of acute gastroenteritis (AGE) on cruise ships. Since 2019, 81% of cruise ship AGE outbreaks posted to VSP's website were attributed to norovirus. 1,2 Previously, most norovirus outbreaks in the United States (US) and cruise travel were attributed to the GII.4 genotype, recently, GII.17 has become the predominate genotype. 3 Due to its new emergence in the US, characteristics of illness with GII.17 are not well described in adult populations. We compare case demographics and disease severity between GII.4 and GII.17 from cruise ship norovirus outbreaks. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12606699&blobtype=pdf | # Asymptomatic Shedding of Enteric Viruses in Young Children: Insights From a Year-Long Prospective Surveillance Study
Lital Hadad, | Avelson, Maya Goldberger, | Hezi Beniacar, | Ayelet, | Hazan, | David Greenberg, Dana Danino
## Abstract
Enteric viruses are a leading cause of gastroenteritis in young children. With the increasing use of multiplex PCR, these viruses are frequently detected in asymptomatic children. In this prospective study conducted from March 2024 to February 2025, rectal swabs were collected from children < 5 years with no history of diarrhea or vomiting in the preceding month and no fever in the prior week. Samples were tested for rotavirus, norovirus, adenovirus, astrovirus, and sapovirus. Parental questionnaires and follow-up interviews assessed risk factors and monitored symptom development. Enteric viruses were detected in 17.6% of asymptomatic children, most commonly rotavirus (6.2%) and norovirus (5.1%). Detection peaked during the second year of life. No differences were observed in daycare attendance, number of siblings, or underlying conditions between virus-positive and virus-negative asymptomatic children. Breastfeeding was significantly protective (OR 0.19, 95% CI 0.06-0.51), while recent antibiotic use was more common in virus-positive cases. Compared with symptomatic children tested during the same period, asymptomatic children had significantly higher cycle threshold (Ct) values, except for astrovirus. Seasonal distribution of viruses was similar across both groups. These findings suggest that asymptomatic children may contribute to viral transmission and that Ct PCR results, alongside clinical context, may aid result interpretation.
| IntroductionEnteric viruses are the leading cause of acute gastroenteritis (AGE) in children under 5 years of age [1]. Following the introduction of the rotavirus vaccine, the most commonly identified pathogens are rotavirus and norovirus, followed by adenovirus, sapovirus, and astrovirus [2][3][4][5][6].Molecular testing is currently the most effective diagnostic method due to its rapid turnaround time and high sensitivity, enabling the simultaneous detection of multiple pathogens. However, distinguishing between nonviable viral fragments and true disease-causing infection remains a challenge [7,8]. Asymptomatic viral detection may reflect either post-diarrheal shedding or subclinical infection.
McMurry et al. investigated the duration of post-diarrheal shedding in young children in a low-resource setting by testing both diarrheal and monthly non-diarrheal stool samples from children aged 0-24 months using quantitative polymerase chain reaction (PCR). They reported median shedding durations of 8.1 days [95% CI, 6.2-9.6] for rotavirus, 17.7 days [9.3-20.6] for astrovirus, 18.1 days [15.4-20.8] for norovirus, and 22.9 days [20.5-25.0] for sapovirus [9].
In addition, a study conducted in a daycare setting found that 22.1% of stool samples from children up to 3 years of age, 95% of whom were asymptomatic, tested positive for enteric viruses [10].
In our region, rotavirus vaccination was introduced into the national immunization program at the end of 2010, with coverage rates reaching 65%-85% within 2 years. Multiplex PCR testing for enteric viruses became the standard diagnostic method for AGE in 2017. Among children under 5 years of age with diarrhea, the overall viral positivity rate from all rectal swabs was 45.3%, with rotavirus remaining predominant despite widespread vaccination. Multiple virus detection occurred in 9.2% of positive cases [11,12].
The current study aimed to evaluate the positivity rate of enteric viruses detected by PCR in rectal swabs from asymptomatic children under 5 years of age and to compare these findings with data from symptomatic children. Specifically, we sought to (1) determine the prevalence and types of enteric viruses in asymptomatic children; (2) examine seasonal and demographic patterns of viral detection; and (3) explore associations between viral carriage and potential influencing factors. Understanding the background carriage of these viruses in healthy children may improve interpretation of positive multiplex-PCR results and their clinical relevance in diagnosing gastroenteritis.
## 2 | Methods
## 2.1 | Setting
The Soroka University Medical Center (SUMC) pediatric hospital has 220 admission beds, and 60,000 pediatric emergency department (PED) visits a year. It is the only hospital in the Negev region of southern Israel and serves a population of 97,500 children < 5 years old, where almost all children are born and receive emergency and inpatient services. All outpatient services in the Negev region and the sole inpatient hospital (SUMC) in the region are served by a single central laboratory that collects and processes all stool samples. The southern region of Israel (the Negev) has a heterogeneous population, consisting of ~80% Jews (representing only ~50% of all births), who live mainly in urban centers and a few rural communities, and 20% Bedouin Arabs (~50% of all births), who are in various stages of transition from semi-nomadism to settled modern-day life. The two groups are different in many aspects, with the Bedouin population characterized as being of a lower socioeconomic status, overcrowding and larger family size. Contact between the two populations is rare except in the hospital setting. Medical insurance in Israel is universal and is provided free of charge. Compared to Jewish children, Bedouin children have a higher incidence rate of hospitalizations and PED visits for diarrheal diseases [5].
Rotavirus vaccines were licensed in Israel in 2007. In December 2010 RotaTeq was added to the NIP, and since then is the sole rotavirus vaccine used in Israel given free of charge. In southern Israel, by the end of 2012, ∼95% (≥ 1 doses), ∼95% (≥ 2 doses) and ∼85% (3 doses) of the Jewish population 8-11 months received RotaTeq. The respective figures among Bedouin infants 8-11 months were ∼90% (≥ 1 doses), ∼85% (≥ 2 doses) and ∼65% (3 doses) (p < 0.001 between Jewish and Bedouin infants) [13].
## 2.2 | Study Design
This was a prospective study. Rectal swabs for multiplex RT-PCR targeting rotavirus, norovirus, adenovirus, astrovirus, and sapovirus were collected from asymptomatic children under 5 years of age between March 2024 and February 2025, following informed consent from their parents or legal guardians. Each week, 8 samples were obtained from children residing in the Negev region who presented to either the ambulatory or inpatient services of the hospital and showed no symptoms of gastroenteritis; no fever during the week before sample collection and no diarrhea (loose stools) or vomiting during the preceding month.
## 2.3 | Laboratory Methods
Rectal swabs were collected from the perianal region using synthetic fiber swabs with plastic shafts and placed into tubes containing viral transport medium. (VTM), transported to the laboratory within 24 h and tested the same day; samples arriving during weekends/holidays were kept at room temperature and tested on the next working day. Viral nucleic acids were extracted using the Seegene Nimbus automated system and tested with the Allplex™ GI-Virus Assay (Seegene Inc., Seoul, Republic of Korea), a multiplex real-time RT-PCR detecting rotavirus A, norovirus GI/GII, adenovirus F40/41, astrovirus, and sapovirus with an internal control to monitor extraction and amplification. Results were analyzed with Seegene Viewer software. A sample was considered positive at a cycle threshold (Ct) ≤ 38, per manufacturer′s recommendations.
Parents or legal guardians completed a questionnaire regarding the child′s: antibiotic use in the past month; episodes of fever in the past month (children with fever in the week before sampling were excluded); dietary habits; breastfeeding status; number of siblings; daycare attendance; known immunodeficiency; hospitalization in the past month; antibiotic treatment in the past month, chronic gastrointestinal disease; immunosuppressive therapy; and rotavirus vaccination status, including number of doses and the date of last administration.
Children were defined as fully vaccinated against rotavirus if they were under 6 months of age and had received 2 doses of the vaccine, or older than 6 months and had received 3 doses. Children who received a rotavirus vaccine within 2 weeks before sampling and tested positive for rotavirus were excluded from the analysis, as it was not possible to differentiate vaccinederived from wild-type virus strains [14][15][16].
Parents and primary care physicians were notified of positive results. Follow-up phone interviews 1 week after a positive result were conducted to assess the occurrence of fever within 5 days of sample collection and of diarrhea or vomiting within 5 days of testing. This timeframe was chosen to capture symptoms likely related to the detected infection while minimizing unrelated events in a population with frequent intercurrent illnesses.
Additionally, we retrospectively reviewed all positive enteric virus PCR results during the study period from children with gastroenteritis, based on data from the microbiology laboratory that exclusively processes all rectal swab samples from both outpatients and inpatients in our region. For these cases, we recorded the date of testing, patient age, viral identification, and Ct value. Multiple tests identifying the same virus in symptomatic children within a 2-month period were considered part of a single episode [9,17,18].
## 2.4 | Statistical Analysis
Descriptive statistics and univariate analyses were performed to summarize viral detection rates and clinical characteristics. Initial comparisons were made between asymptomatic subjects with positive and negative PCR results. This was followed by analyses comparing symptomatic and asymptomatic individuals. Variables found to be significant in the univariate analysis were included in multivariate logistic regression models to identify factors associated with viral PCR positivity. Finally, Ct values were compared between symptomatic and asymptomatic subjects to explore differences in viral load. All data preprocessing, statistical analyses, and visualizations were conducted using RStudio (version 2024.12.1, Build 563) and Microsoft Excel.
## 3 | Results
Throughout the study year, 330 rectal swabs for multiplex enteric viral PCR were collected from children under 5 years of age who had no symptoms of gastroenteritis in the month before testing and no febrile illness in the preceding week. Of these, 58 samples (17.6%) tested positive for at least one virus, with 6 (10.3%) showing multiple virus detections.
The most frequently detected virus was rotavirus (21, 6.2%), followed by norovirus (17, 5.1%), adenovirus (10, 3.0%), sapovirus (8, 2.4%), and astrovirus (8, 2.4%). Viral detection was more common among Bedouin children (43, 21.2%) than among Jewish children (21, 15.8%) and was highest during the first year of life (36, 18.4%). The distribution of viruses differed between the two ethnic groups: among Bedouin children, norovirus (14, 6.9%) and rotavirus (12, 5.9%) were the most common, while among Jewish children, rotavirus (9, 6.8%) and adenovirus (5, 3.8%) were more frequently detected (Figure 1).
The highest positivity rate (52.2%) was observed in January. Rotavirus and astrovirus were most frequently detected during the cold months (winter-spring), adenovirus during the warmer months (spring-summer), and sapovirus during the fall-winter months. Norovirus was detected throughout the year (Figure 2).
When comparing asymptomatic children with viral detection to those without, no significant differences were found in age, ethnicity, or underlying conditions (e.g., immunodeficiency, chronic gastrointestinal disease, or recent healthcare exposure). Both groups had similar daycare attendance and number of siblings. However, asymptomatic children with viral detection were more likely to be formula-fed and to have received antibiotics in the month before testing (53.1% vs. 33.1%, p < 0.01% and 18.6% vs. 9.6%, p = 0.06, respectively) (Table 1). After adjusting for age and ethnicity, breastfeeding was found to be 3 and4).
Among asymptomatic children with viral detection, only 1 (1.7%) child developed vomiting and diarrhea within 5 days, 3 (5.2%) experienced only vomiting and 1 (1.7%) only diarrhea. None developed fever during the subsequent week. During the same period, 1,715 rectal swabs were collected from symptomatic children with gastroenteritis in both outpatient and inpatient settings. Of these, 700 (40.8%) children tested positive for at least one enteric virus, among them 27 had two viruses and 2 had three viruses, resulting in a total of 731 viruses detected. The most common viruses were rotavirus (271, 37.1%) and norovirus (179, 24.5%). No significant differences were found in age, sex, or viral distribution between symptomatic children and asymptomatic children with viral detection (Table 2). However, cycle threshold (Ct) values were significantly lower in symptomatic children compared to asymptomatic ones (p < 0.01). This difference was observed overall and for each virus individually, except for astrovirus (Figure 5). Viral seasonality was similar in symptomatic and asymptomatic children for rotavirus and overall viruses (Figure 6, Supplementary Figure 1).
## 4 | Discussion
The prevalence of enteric virus detection by multiplex PCR among asymptomatic children under 5 years of age was 17.6%. The most commonly detected viruses were rotavirus and norovirus, mirroring those found in symptomatic children of the same age group in our region. Many studies have shown that enteric viruses are frequently detected in the stool of children without symptoms of acute gastroenteritis, particularly when using highly sensitive molecular methods such as PCR [3,19]. These children are often referred to as healthy controls or asymptomatic carriers.
In a pilot study of five healthy Australian infants, 40.7% of weekly stool samples collected over 2 years tested positive for at least one of six enteric viruses [20]. Geographic variation plays a significant role in the prevalence of asymptomatic viral shedding, with higher detection rates consistently observed in lowincome countries. For example, in Madagascar, 51.7% of stool samples from apparently healthy children aged 2-5 years were positive for at least one enteric virus [21], while in Finland, only 14% of asymptomatic hospitalized children tested positive [22]. These findings align with the differences in prevalence observed between the two ethnic populations in our region.
Prevalence also varies by virus type. Norovirus is detected in approximately 7% of healthy individuals worldwide [23,24]. Rotavirus is found in 1%-11% of asymptomatic individuals, with some of the higher rates attributed to vaccine virus shedding [25,26]. Adenovirus shows the widest detection range, from 1.4% to 43% [3,4]. Sapovirus is detected in 3%-5% of healthy children in the US and up to 20% in certain populations [22,25]. Astrovirus has the lowest prevalence overall, typically ranging from 0.3% to 10%, depending on the region [3,27].
Age-related differences are also notable, with asymptomatic viral detection generally higher in infants and young children [25]. This frequent shedding suggests that a positive PCR result for an enteric virus does not necessarily indicate the causative agent of a child's illness. Furthermore, asymptomatic individuals can act as reservoirs, contributing to transmission and outbreaks.
Several risk factors have been associated with asymptomatic enteric viral shedding in children. Attendance at childcare settings has consistently been linked to increased risk, as close contact facilitates transmission. Studies in both England and the US found higher detection rates of rotavirus and other pathogens among children attending daycare or school [25,26]. Household dynamics also play a role; living with a baby in diapers was associated with increased rotavirus shedding in older adults [26]. Poor dietary quality and maternal overnutrition were associated with higher viral detection, emphasizing the importance of nutrition and maternal health [21]. Interestingly, infestation with Ascaris lumbricoides appeared to be protective, suggesting complex interactions between enteric pathogens and the host immune system [21].
Although previous studies have demonstrated an association between increased contact and asymptomatic infection in young children, we did not observe such a correlation in our cohort. Formula feeding was the only risk factor significantly more common among asymptomatic children with viral detection, while other factors, such as daycare attendance, number of siblings, and underlying medical conditions, did not differ between groups. Breastfeeding and absence of recent antibiotic use were both protective against viral shedding, likely due to their beneficial effects on gut immunity. Recent studies have shown that greater gut microbial diversity enhances immunity and protects against symptomatic viral gastroenteritis.
The similar seasonal distribution observed for symptomatic and asymptomatic infections supports the hypothesis that both are transmitted via similar routes, primarily person-to-person. It is therefore likely that host immunity, rather than the route of infection, determines whether illness develops.
We found that Ct values were significantly lower in symptomatic children compared to asymptomatic ones, both across all viruses and for each virus individually, except for astrovirus. Symptomatic children with acute gastroenteritis generally have higher viral loads than asymptomatic carriers [4,22]. However, other studies, Ct values alone, particularly for viruses like norovirus and sapovirus, could not reliably differentiate between symptomatic and asymptomatic individuals [28].
The strengths of our study include its prospective design, with active sampling of asymptomatic children throughout the year, and the application of strict criteria to define asymptomatic status. Children with diarrhea, loose stools, or vomiting within the month before sampling, as well as those with fever in the preceding week, were excluded. Additionally, detailed parental questionnaires captured sociodemographic and medical history, and follow-up phone interviews were conducted for all children with positive results to monitor for subsequent symptoms.
Our study has several limitations. First, it was challenging to recruit a consistent number of children each month, particularly during colder months, due to the stringent criteria for asymptomatic status. Because most children in this age group frequently experience fever or gastrointestinal symptoms, it was difficult to identify children who had been symptom-free for at least 1 month. As a result, the number of samples collected varied over time, which may have influenced detection rates. Second, since the participants were recruited from a hospital setting, they may not fully represent the broader healthy pediatric population, particularly those who do not access hospital care. Third, we collected only a single sample per child, making it difficult to determine whether a detected virus reflected early incubation, prolonged shedding, or a persistent subclinical infection. To mitigate this, we excluded children with any recent gastrointestinal symptoms and conducted follow-up interviews to identify any delayed symptom onset. Fourth, we did not perform viral genotyping and therefore could not compare strains from asymptomatic children with those from symptomatic cases. We were also unable to distinguish between vaccine-derived and wild-type rotavirus strains. To address this, we excluded children who tested positive for rotavirus and had received the vaccine within 2 weeks before sample collection. Finally, the number of astroviruspositive cases was small in both asymptomatic and symptomatic groups, which may have limited the ability to detect meaningful differences for this virus.
In conclusion, the frequent detection of enteric viruses among asymptomatic children under 5 years of age underscores their potential role in transmission and highlights the limitations of using multiplex PCR alone for diagnosing gastroenteritis. Positive test results should always be interpreted within the full clinical and epidemiological context.
## References
1. Osborne, Montano, Robinson et al. (2015) "Viral Gastroenteritis in Children in Colorado 2006-2009" *Journal of Medical Virology*
2. Chirinda, Manjate, Garrine (2008) "Detection of Enteric Viruses in Children Under Five Years of Age Before and After Rotavirus Vaccine Introduction in Manhiça District" *Viruses*
3. Halasa, Piya, Stewart (2021) "The Changing Landscape of Pediatric Viral Enteropathogens in the Post-Rotavirus Vaccine Era" *Clinical Infectious Diseases*
4. Hassan, Kanwar, Harrison (2019) "Viral Etiology of Acute Gastroenteritis in & amp;lgt;2-Year-Old US Children in the Post-Rotavirus Vaccine Era" *Journal of the Pediatric Infectious Diseases Society*
5. Leshem, Givon-Lavi, Tate et al. (2016) "Real-World Effectiveness of Pentavalent Rotavirus Vaccine Among Bedouin and Jewish Children in Southern Israel" *Clinical Infectious Diseases*
6. Muhsen, Cohen (2017) "Rotavirus Vaccines in Israel: Uptake and Impact" *Human Vaccines & Immunotherapeutics*
7. Corcoran, Van Well, Van Loo (2014) "Diagnosis of Viral Gastroenteritis in Children: Interpretation of Real-Time PCR Results and Relation to Clinical Symptoms" *European Journal of Clinical Microbiology & Infectious Diseases*
8. De Grazia, Bonura, Bonura (2020) "Assessing the Burden of Viral Co-Infections in Acute Gastroenteritis in Children: An Eleven-Year-Long Investigation" *Journal of Clinical Virology*
9. Mcmurry, Mcquade, Liu (2021) "Duration of Postdiarrheal Enteric Pathogen Carriage in Young Children in Low-Resource Settings" *Clinical Infectious Diseases*
10. Enserink, Scholts, Bruijning-Verhagen (2014) "High Detection Rates of Enteropathogens in Asymptomatic Children Attending Day Care" *PLoS One*
11. Danino, Hazan, Mahajna (2023) "Implementing a Multiplex-PCR Test for the Diagnosis of Acute Gastroenteritis in Hospitalized Children: Are All Enteric Viruses the Same?" *Journal of Clinical Virology*
12. Hazan, Goldstein, Greenberg (2024) "Comparing Single Versus Multiple Virus Detection in Pediatric Acute Gastroenteritis Postimplementation of Routine Multiplex RT-PCR Diagnostic Testing" *Journal of Medical Virology*
13. Givon-Lavi, Ben-Shimol, Cohen et al. (2015) "Rapid Impact of Rotavirus Vaccine Introduction to The National Immunization Plan in Southern Israel: Comparison Between 2 Distinct Populations" *Vaccine*
14. Kaneko, Takanashi, Inoue (2019) "Detection of Mutations in the VP7 Gene of Vaccine-Derived Strains Shed by Monovalent Rotavirus Vaccine Recipients" *Access Microbiology*
15. Li, Cao, Gao (2018) "Faecal Shedding of Rotavirus Vaccine in Chinese Children After Vaccination With Lanzhou Lamb Rotavirus Vaccine" *Scientific Reports*
16. Markkula, Hemming, Vesikari (2015) "Detection of Vaccine-Derived Rotavirus Strains in Nonimmunocompromised Children Up to 3-6 Months After RotaTeq® Vaccination" *Pediatric Infectious Disease Journal*
17. Clark, Mckendrick (2004) "A Review of Viral Gastroenteritis" *Current Opinion in Infectious Diseases*
18. Richardson, Grimwood, Gorrell et al. (1998) "Extended Excretion of Rotavirus After Severe Diarrhoea in Young Children" *Lancet*
19. Grant, O'brien, Weatherholtz (2017) "Norovirus and Sapovirus Epidemiology and Strain Characteristics Among Navajo and Apache Infants" *PLoS One*
20. Ye, Whiley, Ware (2017) "Detection of Viruses in Weekly Stool Specimens Collected During the First 2 Years of Life: A Pilot Study of Five Healthy Australian Infants in the Rotavirus Vaccine era" *Journal of Medical Virology*
21. Razanajatovo, Andrianomiadana, Habib (2023) "Factors Associated With Carriage of Enteropathogenic and Non-Enteropathogenic Viruses: A Reanalysis of Matched Case-Control Data From the AFRIBIOTA Site in Antananarivo, Madagascar" *Pathogens*
22. Pitkänen, Markkula, Hemming-Harlo (2022) "Sapovirus, Norovirus and Rotavirus Detections in Stool Samples of Hospitalized Finnish Children With and Without Acute Gastroenteritis" *Pediatric Infectious Disease Journal*
23. Payne, Vinjé, Szilagyi (2013) "Norovirus and Medically Attended Gastroenteritis in U.S. Children" *New England Journal of Medicine*
24. Wang, Gao, Yang et al. (2023) "Global Prevalence of Asymptomatic Norovirus Infection in Outbreaks: A Systematic Review and Meta-Analysis" *BMC Infectious Diseases*
25. Harrison, Hassan, Lee (2021) "Multiplex PCR Pathogen Detection in Acute Gastroenteritis Among Hospitalized US Children Compared With Healthy Controls During 2011-2016 in the Post-Rotavirus Vaccine era" *Open Forum Infectious Diseases*
26. Phillips, Lopman, Rodrigues et al. (2010) "Asymptomatic Rotavirus Infections in England: Prevalence, Characteristics, and Risk Factors" *American Journal of Epidemiology*
27. Khumela, Kabue, Traore et al. (2021) "Human Astrovirus in Symptomatic and Asymptomatic Children: A Cross-Sectional Study on Hospitalized and Outpatients From Rural Communities of South Africa Between 2017-2021" *Pathogens*
28. Bruijnesteijn Van Coppenraet, Flipse, Wallinga (2021) "From a Case-Control Survey to a Diagnostic Viral Gastroenteritis Panel for Testing of General Practitioners' Patients" *PLoS One* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793584&blobtype=pdf | Pierluigi Salvo, Francesca Lombardi, ; Gianmaria Baldin, Valeria Campolattano, Ilenia Aversa, Antonio Abatino, Simona Giambenedetto, Camillo Palmieri, Carlo Torti
Background. Dengue is the most widespread arboviral infection, with approximately half of the world's population at risk. In 2023, an autochthonous dengue outbreak occurred in Rome, highlighting the increasing relevance of this infection in non-endemic areas. PLWH are at an increased risk of developing severe dengue, particularly in cases of secondary infection. The aim of this study was to evaluate and compare the T cell-mediated immune response following dengue infection in PLWH and HIV-negative individuals.
Methods. In this cross-sectional study, we enrolled PLWH and HIV-negative individuals with a history of dengue infection, identified through DENV RT-PCR or NS1 antigen assays in a hospital setting. IFN-γ enzyme-linked immunospot assay S1330 • OFID 2026:13 (Suppl 1) • Poster Abstracts (ELISPOT) was used to analyze T cell response, using peptide pools representing three structural proteins (prM/E/C) and non-structural proteins (NS3 and NS2A/B, NS4A/ B, NS5).
Results. We enrolled 9 PLWH and 10 HIV-negative indiduals. All but one HIV-negative subjects showed a positive T-cell response to at least one stimulating peptide pool. Overall, HIV-negative individuals exhibited stronger ELISPOT responses than PLWH across all antigens, though without statistical significance, likely due to sample size limitations. The NS3 antigen showed the most pronounced difference (Mean: 23.5 in HIV-negative vs 2.0 in PLWH, p=0.223); prM/E/C responses were also higher in the HIV-negative group (14.3 in HIV-negative vs. 5.8 in PLWH, p = 0.327), while NS4/AB (5.4 in HIV-negative vs 2.5 in PLWH, p=0.451) and NS2/A/ B (3.0 in HIV-negative vs 2.0 in PLWH, p =0.578) responses were only slightly elevated in the HIV-negative group. NS5 elicited the weakest response in both groups (0.5 in HIV-negative vs 0.0 in PLWH, p = 0.408).
Conclusion. The non-uniform antigen recognition pattern between the 2 groups suggests potential differences in immune history, genetic background, or prior exposure between groups, influencing T-cell activation levels. Overall, our findings suggest that prior dengue exposure does not guarantee uniform T-cell reactivity, and HIV status may modulate the strength and pattern of antigen-specific immunity.
Disclosures |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12596242&blobtype=pdf | # Beyond the Amazon: vector surveillance and emerging Or opouche virus in South America
Ruth Dálety, Silva Brito, Jéssica Farias, Alexander Birbra Ir, Luís Carlos De Souza Ferreira, Paloma Oliveira Vidal, Jaime Amorim
## To the E ditor,
We commend the recent study by Vieira et al. (Vieira et al. 2024) for the innovative approach of integrating xeno-monitoring and metatranscriptomic sequencing to investigate arboviral diversity. This innovative approach not only revealed diverse co-circulating arboviruses in Australia but also exemplifies how next-generation surveillance can outperform conventional methods in ecologically complex regions. Tropical areas in South and Central America, especially the Amazon basin and surrounding biomes , share similar conditions for arboviral endemicity, vector heterogeneity, and underdiagnosis due to symptom overlap following infection with different co-circulating viruses (Files et al. 2022, Bai et al. 2025). These findings highlight the potential of genomic surveillance tools in arbovirus-endemic regions.
Of particular relevance stand neglected arboviruses such as the Oropouche virus (OROV), which is currently expanding its geogr aphic and epidemiological footprint across the Americas (Gutierrez et al. 2020, Moreira et al. 2024). While historically endemic to the Amazon basin, OROV recently caused outbreaks in previously non-endemic areas of Brazil, including the northeast (e.g. Bahia State), the southeast (e.g. Espírito Santo State), and the south (e.g. Santa Catarina State), as well as in neighbouring countries, such as Peru and Panama (Moreira et al. 2024, Bai et al. 2025). These outbreaks were geographically distinct across multiple Brazilian states and neighbouring countries, with partially overlapping time windows during 2024. Recent genomic surveillance revealed that latest outbreaks are associated with a novel reassortant virus variant, whose small and large segments are mor e closely related to Iquitos virus, a related Orthobunyavirus, while the medium segment retains similarity to the OROV prototype strain (Gutierrez et al. 2020, Moreira et al. 2024). This reassortment event is accompanied by the emergence of non-synonymous mutations across different genomic segments (Fig. 1). In the M segment, substitutions V61F (in the Gn envelope glycoprotein), V526I, I958T, S1268P, and F/L521S (in the Gc envelope glycoprotein), as well as I393T in the nonstructural protein NSm, were identified. Mutations in the Gc envelope glycoprotein, which functions as the viral antireceptor, may inf luence host range and susceptibility by altering virus-receptor interactions. In the L segment, which encodes the RNA-dependent RNA polymerase (RdRp), substitutions such as I758V, I1942V, H476Y, E847G, and R346K were identified. These evolutionary and geographic events are summarized in Table 1. These amino acid changes may have phenotypic consequences, potentially affecting viral fitness, replication efficiency, and virulence (Moreira et al. 2024, Silva Júnior et al. 2024, Scachetti et al. 2025). Despite these alarming developments, OROV remains a neglected pathogen, with widespread underreporting due to limited molecular diagnostic c apacity and lack of systematic vector surveillance programs (de Souza Luna et al. 2017, Bai et al. 2025). In addition, it is important to emphasize that in most places, vector surveillance is focused on mosquitoes and not midges. This diagnostic gap hinders early detection, impairs outbreak response, a nd masks the true disease burden, particularly in rural and peri-urban settings.
Effective surveillance of OROV must consider the diversity of vectors involved in the transmission cycle. Species such as Aedes serratus, Culex quinquefasciatus, Coquillettidia venezuelensis, and Mansonia venezuelensis have been found naturally infected in the field, indicating their participation in the virus's sylvatic cycle. Laboratory studies further confirmed that mosquitoes, including A. serratus, Aedes scapularis, Aedes albopictus, C. quinquefasciatus, and Psorophora ferox, can be experimentally infected and are capable of transmitting OROV, with C. quinquefasciatus exhibiting the highest vector competence described to date. However, it is important to clarify that the actual epidemiological relevance of mosquitoes remains uncertain, and no species has yet demonstrated the consistent field-to-human transmission dynamics established for biting midges. Nevertheless, the biting midge Culicoides paraensis remains the primary vector historically associated with human outbreaks, with its competence firmly established through virus isolation in wild specimens and successful transmission to susceptible hosts under laboratory conditions (Bai et al. 2025). The possibility of expansion of vector range raises urgent questions about urban spillover and the risk Figure 1. Schematic representation of the recent evolutionary events leading to the emergence of epidemic OROV in Brazil, in which a genomic reassortment likely occurred between 2022 and early 2023, involving the replacement of the small (S) and large (L) segments of the prototype OROV genome by those of Iquitos virus (IQTV), while retaining the medium (M) segment from the classical OROV lineage, and in mid-2023, a novel epidemic strain-AM0088-was identified, carrying multiple non-synonymous substitutions in the M segment, and was associated with the unprecedented OROV epidemic in Brazil in 2024, marked by increased viral replication in mammalian cells, lar ger and earlier plaque formation, and substantial reduction in neutralization by antibodies induced by prior OROV infection (Scachetti et al. 2025). Scachetti et al. 2025 Notes: Events are aligned with genomic surveillance reports cited in the main text. Terminology harmonized with Fig. 1 for consistency. Spillover refers to reported outbreaks in previously non-endemic reg ions.
of future outbreaks in densely populated regions. In this context, metatranscriptomic surveillance applied directly to fieldcollected vector pools-as implemented in the Australian studyoffers a powerful, unbiased approach to detect emerging OROV variants, including reassortant variants and co-infections with other arboviruses. Such methods can uncover evolutionary trajectories, inter-species transmission events, and genomic signatures of adaptation that are otherwise missed by conventional virological assays (Gutierrez et al. 2020, Vieira et al. 2024). Integrating these techniques into routine vector monitoring would not only enhance early detection of OROV but also support more accurate modelling of its spatial dynamics and evolutionary potential.
The successful implementation of genomic arbovirus surveillance through xeno-monitoring and metatranscriptomics in Australia offers a timely and transferable model for Latin America, where arbovirus emer gence is shaped by ecological pressures, human mobility, and climate change (Sah et al. 2024). We advocate for the urgent adaptation of such strategies to South American contexts, particularly in Brazil, where arboviruses like OROV, mayaro virus (MAYV), yellow fever virus (YFV), a nd others demonstrate the capacity to cause unexpected outbreaks through sylvatic-urban spillover events (Moreira et al. 2024, Bai et al. 2025). Establishing regional genomic surveillance networks-integrating public health laboratories, academic institutions, and field entomology teams-would not only enhance early warning capabilities but also foster data sharing and coordinated responses across national borders. This collaborative infrastructure is essential to detect, characterize, and contain emerging viral threats before they reach epidemic scale. In addition, broader impact modelling of metatranscriptomic implementation and comparative epidemiological insights across related arboviruses represent promising directions for future research. However, the implementation of such strategies in Brazil and neighbouring countries faces practical challenges. These include the high cost of sequencing, limited access to equipped laboratories, the need for specialized training, and barriers to data integration across institutions. Recognizing and addressing these constraints is essential for the success of genomic surveillance in the region. In addition, recent studies have emphasized that the resurgence of OROV poses a diagnostic challenge in the region (Cherem and Barçante 2025), with frequent misclassification as dengue, reinforcing the need for expanded genomic surveillance and improved differential diagnosis.
Conf lict of interest: None declared.
## References
1. Bai, Denyoh, Urquhart (2025) "A comprehensive review of the neglected and emerging Oropouche virus" *Viruses*
2. Cherem, De Barçante (2025) "Addressing the escalating burden of dengue in the Americas amid global c hallenges" *J Infect Public Health*
3. De Souza Luna, Rodrigues, Santos (2017) "Oropouche virus is detected in peripheral blood leukocytes fr om patients" *J Med Virol*
4. Files, Hansen, Herrera (2022) "Baseline mapping of Oropouche virology, epidemiology, therapeutics, and vaccine research and development" *Vaccines*
5. Gutierrez, Wise, Pullan (2020) "Evolutionary dynamics of Oropouche virus in South America" *J Virol*
6. Moreira, Rr, Dutra et al. "Oropouche virus genomic surveillance in Br azil" *Lancet Infect Dis*
7. Sah, Srivastava, Kumar (2024) "Oropouche fever outbreak in Brazil: an emerging c oncern in Latin America" *Lancet Microbe*
8. Scachetti, Forato, Claro (2025) "Re-emergence of Oropouche virus between 2023 and 2024 in Brazil: an observ ational epidemiological study" *Lancet Infect Dis*
9. Júnior, Lopes, Pedroso (2024) "Unprecedented Oropouche fever outbreak in Brazil: could the M segmentencoded proteins pro vide clues to possible insights" *J Med Virol*
10. Vieira, Onn, Shivas (2024) "Long-term co-circulation of multiple arboviruses in Southeast Australia revealed by xeno-monitoring and viral whole-genome sequencing" *Virus Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12518673&blobtype=pdf | # Seasonal trends and molecular evolution of SARS-CoV-2 in acute respiratory illness patients during the fifth year of COVID-19, Thailand
Jiratchaya Puenpa, Preeyaporn Vichaiwattana, Ratchadawan Aeemjinda, Lakkhana Wongsrisang, Sumeth Korkong, Jen-Ren Wang, Nasamon Wanlapakorn, Yong Poovorawan
## Abstract
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is relentlessly evolving, with emerging variants exhibiting heightened transmissibility and immune escape capabilities. Understanding the genetic diversity and seasonal transmission dynamics of SARS-CoV-2 in Thailand is crucial for implementing effective public health interventions. This study aims to elucidate the genetic diversity and seasonal transmission patterns of SARS-CoV-2 among patients with acute respiratory illness in Thailand in 2024. Among 8,096 Acute Respiratory Infection (ARI) cases analyzed, 1,152 samples tested positive for SARS-CoV-2, resulting in a positivity rate of 14.2%. The majority of infections occurred during a significant Outbreak from late summer to early rainy season, particularly between April and June, accounting for nearly 49% of all positive cases. The highest infection rate occurred in adults aged 31-40 years, with no significant association between gender and infection status (p = 0.583). Extensive genomic sequencing identified over seven distinct SARS-CoV-2 lineages, with the JN.1 lineage dominating early in the year. Recombinant variants, notably XEC, XDV.1, and XDY, emerged as significant contributors to the evolving landscape, reaching a prevalence of 57.1% by December. Phylogenetic analyses demonstrated a consistent evolutionary rate and identified critical emergence dates for new lineages, underscoring the virus's ongoing evolution in Thailand. Our findings emphasize the necessity of continuous genomic surveillance for tracking variant transmission dynamics and managing public health responses effectively. Furthermore, the evolving genetic landscape of SARS-CoV-2 highlights the urgent need for adaptive vaccination strategies aligned with seasonal transmission patterns in Thailand.
## Results
## Trends in COVID-19 incidence and demographics of infected patients in 2024
According to the Ministry of Public Health (MoPH), a total of 46,079 COVID-19 cases and 220 fatalities were reported between January and December 2024, reflecting a 6% reduction in cases and a 20% decrease in deaths compared to the previous year (Fig. 1a andb) 27 . Regional distribution of cases revealed that the Bangkok Metropolitan Region (BMR) accounted for 30.8% of the total cases, followed by the Northeastern region at 20.0% and the Southern region at 13.8%. The Western region recorded the lowest proportion, contributing only 3.4% to the overall total. The disproportionately high number of cases in the BMR may reflect its dense population, greater mobility, and broader access to testing services, which could have contributed to both increased transmission and case detection. Notably, the highest incidence of deaths occurred in May and June, accounting for 41% of total fatalities (Fig. 1b).
In this study, 8,096 Acute Respiratory Infection (ARI) cases submitted by Thai residents to partnering hospitals were analyzed. Of these, 1,152 samples tested positive for SARS-CoV-2 via real-time RT-PCR, resulting in a positivity rate of 14.2% (Fig. 1c). Data collection was conducted across Thailand's seasonal divisions: summer (February to mid-May), rainy season (mid-May to October), and winter (November to February). The findings revealed an initial SARS-CoV-2 Outbreak wave in January, accounting for 13.3% of total positive samples (153/1,152), followed by a significant second wave from April to June, which represented 48.9% of total positives (563/1,152) during the summer months. Notably, positivity rates for SARS-CoV-2 declined below 10% from July to December, indicating a reduction in transmission during the latter part of the year.
Among the 1,152 SARS-CoV-2 infected patients, 539 (46.8%) were Male and 613 (53.2%) were female (Table 1). A Chi-square test showed no significant association between gender and infection status, χ²(1, N = 8,096) = 0.30, p = 0.583. The highest proportion of SARS-CoV-2 infections occurred in individuals aged 31-40 years (18.3%), followed by those aged 41-50 (17.7%) and 1 month-10 years (16.9%). The lowest was in the 11-20 years group (5.9%). Pairwise proportion tests with Bonferroni correction showed significantly higher infection rates in the 31-40 age group compared to the 0-10 (p < 0.001), 11-20 (p = 0.0011), and 51-60 (p < 0.001) groups. No significant differences were observed when compared to the 21-30, 41-50, or > 60 age groups (p > 0.05). These findings highlight a greater burden of infection among middle-aged adults.
## SARS-CoV-2 variants distribution in Thailand
To investigate the distribution of SARS-CoV-2 lineages in Thailand from January to December 2024, a comprehensive analysis was conducted on 689 SARS-CoV-2 positive samples. Among these, 60 samples were randomly selected for complete genome sequencing, while the remaining 629 underwent partial spike sequencing. The study identified more than seven distinct lineages, each accompanied by its descendant variants [Fig. 2]. Notable lineages included JN. In alignment with the observed variant frequencies (Fig. 2a andb), the JN.1 lineage Maintained the highest prevalence from January to April, peaking at 90.2% (55/61) in February, followed by JN.1.16 at 11.1% (4/36) in March and KP.2 at 23.2% (16/69) in April. Following this period, the highest prevalence transitioned to KP.2 during May and June, reaching a peak of 28.6% (32/112) in June, while JN.1.16 emerged as the second most prevalent variant, accounting for 23.4% (32/137) and 25.9% (29/112) in those respective months. Additionally, LB.1 was first detected in April with a prevalence of 4.3% (3/69) and subsequently peaked at 35.0% (21/60) in July, establishing itself as a dominant variant before declining and ultimately disappearing by November.
From August to September, the predominant variant shifted once again to KP.3, which accounted for 29.0% of cases in both months (11/38 in August and 9/31 in September), primarily represented by sublineages KP. KP.3.2, and KP.3.3.1 (Supplement Fig. 1). This variant maintained its status as the second most prevalent variant through the end of December.
The recombinant lineages were first identified in May, beginning with an initial prevalence of 8.8% (12/137). They subsequently became predominant during the last three months of the year, reaching 38.2% (13/34) in October and peaking at 57.1% (4/7) in December. For example, XDV.1 began to significantly contribute in mid-year, first detected in May at 5.1% (7/137) and increasing to 10.5% (4/38) in August (Supplement Fig. 1). Meanwhile, XDY peaked in September at 9.7% (3/31). The variant XEC emerged as the most prevalent within the recombinant lineages, first identified in September at 3.2% (1/31), and then rising to 15.8% (3/19) in November and 14.3% (1/7) in December. Additionally, in October, other recombinant variants detected included XEF at 2.9% (1/34), and both XEL and XEN at 8.8% each (3/34). This dynamic landscape of variant prevalence highlights the ongoing evolution and transmission patterns of SARS-CoV-2 in Thailand, underscoring the necessity for continuous genomic surveillance to adapt public health responses effectively.
## Maximum likelihood and time-scaled phylogenetic reconstruction of sars-cov-2 variants
To establish initial trees for Bayesian phylodynamic analyses, time-scaled phylogenies were constructed based on the maximum likelihood (ML) tree topologies and their corresponding collection dates (Fig. 3a andb). The ML phylogenetic tree, derived from complete genome sequences, illustrates the molecular evolution of SARS-CoV-2 isolates both from Thailand and globally. Molecular clock analysis of the dataset provided robust evidence for the time-stable, clock-like evolution of these lineages, demonstrating an R² value of 0.5 and a correlation coefficient of 0.7, along with a substitution rate of 1.38 × 10 -3 substitutions per site per year (Fig. 3c). This analysis Fig. 3. (a) A global maximum likelihood (ML) tree was constructed from a dataset of complete genome sequences (N = 423) using TreeTime, employing the oldest method for phylogenetic analysis. (b) A timescaled ML tree was generated by integrating collection dates into the ML tree from panel (a), thus serving as the primary reference for further analysis. (c) A regression analysis of root-to-tip genetic distances against sampling dates for this comprehensive dataset, estimated using TempEst, revealed a significant positive molecular clock signal.
employed the optimal rooting approach, which minimizes the mean squared residuals, to effectively investigate the relationships between genetic divergence and sampling dates.
The Bayesian phylogenetic tree, constructed from an Additional set of 60 complete SARS-CoV-2 genomes (Supplement Table 1) alongside global sequences collected between December 2023 and December 2024 (Fig. 4), indicates that the most recent common ancestor of this dataset is estimated to have emerged in August 2022. Overall, the Thai and global SARS-CoV-2 samples did Not exhibit distinct phylogenetic clustering, suggesting similar patterns of viral circulation and shared lineages during the study period. However, after September 2024, a divergence was observed, with Thai samples not clustering within the sublineage clusters KP3.1.1, KP3.2, and KP3.3.1. The absence of these newly emerging KP3 sublineages in Thailand during the study period May reflect differences in transmission dynamics, introduction events, or the temporal scope of sampling. The molecular evolutionary rate for the dataset was estimated at 0.63 × 10 -3 nucleotide substitutions per site per year (sub/site/ year), with a 95% highest posterior density interval (HPDI) ranging from 0.57 × 10 -3 to 0.70 × 10 -3 sub/site/year. Table 2 details the sequence change rates and the time to the most recent common ancestors (TMRCA) as determined by Markov chain Monte Carlo (MCMC) methods. Among all variants, XDV showed the highest substitution rate (0.89 × 10 -3 ) and the greatest nucleotide divergence (4.56 × 10 -4 ), with a tMRCA estimated around late November 2023. In contrast, JN.1 had the lowest substitution rate (0.45 × 10 -3 ) and an earlier tMRCA in early March 2023. Variants KP.1 to KP.3 and LB.1 shared comparable nucleotide divergence values (3.66 × 10 -4 and 4.09 × 10 -4 , respectively), with substitution rates of 0.56 × 10 -3 and 0.75 × 10 -3 . However, KP.1-3 emerged later, with a tMRCA in mid-November 2023, while LB.1 diverged slightly after, in early January 2024. XEC and XDY both emerged more recently, with tMRCAs in May 2024. While their substitution rates were similar (0.55 and 0.56 × 10 -3 ), XDY showed notably lower nucleotide divergence (1.72 × 10 -4 ) compared to XEC (3.33 × 10 -4 ), possibly reflecting differences in evolutionary dynamics or sampling time frames.
## Discussion
The COVID-19 Outbreak, first identified in late 2019, arrived in Thailand in early 2020, leading to a series of ongoing transmission events that prompted national concern 28 . Following nearly three years of extensive outbreaks and significant public health challenges, the Thai Ministry of Public Health reclassified COVID-19 from a 'dangerous infectious disease' to a 'disease under surveillance' and announced a revised reporting protocol in October 2022 that focuses exclusively on case counts among hospitalized patients 29 . By 2023, the severity of COVID-19 had significantly diminished, evidenced by a marked reduction in fatalities 27 . In 2024, the reported death toll decreased to just 220 cases, representing a stark contrast to the high mortality rates observed in the first three years of the pandemic, which included 21,614 deaths in 2021, 11,971 in 2022, and 848 in 2023 27 . This reduction in severity can be attributed to several factors, including increased vaccination coverage, the emergence of less virulent variants, and improved treatment modalities. As a result, COVID-19 is now recognized as a seasonal respiratory illness, similar to other common respiratory infections.
This study investigated the seasonal trends and genetic diversity of SARS-CoV-2 among patients with acute respiratory illness in Thailand during the fifth year of the COVID-19 pandemic, providing a comprehensive analysis of acute respiratory infection cases throughout 2024, which includes both inpatient and outpatient data across all seasons. The results demonstrate that COVID-19 transmission follows a distinct seasonal pattern, marked by a significant increase in cases from late summer to the early rainy season, particularly between April and June.
In the early years of COVID-19's seasonal transition, its incidence peaked earlier than other respiratory infections, such as influenza and respiratory syncytial virus 30,31 . Although Thailand's official rainy season begins in mid-May, the interplay of school reopenings and the Songkran festival in mid-April, characterized by extensive travel and gatherings, contributes significantly to an early rise in infection rates.
Following this peak, case numbers gradually declined from September to the end of the year, before experiencing another surge in early winter (January to mid-March), primarily due to lower temperatures. Unlike temperate regions where COVID-19 surges predominantly occur during winter [32][33][34] . Thailand's tropical climate lacks a distinct cold season. While there is evidence of increased cases during cooler months, the most pronounced peak reliably occurs at the onset of the rainy season each year. This highlights the unique seasonal dynamics of COVID-19 in Thailand, characterized by year-round transmission with a marked surge during early monsoon months. Such insights are crucial for informing public health strategies tailored to local seasonal patterns.
This study observed the highest prevalence of COVID-19 infection among individuals aged 31 to 40 years, with statistically significant differences compared to several other age groups, including children (0-10 years), adolescents (11-20 years), and older adults aged 51-60 years. These findings are consistent with previous research reporting a greater burden of SARS-CoV-2 infection among middle-aged adults, possibly due to increased social mobility and occupational exposure during the post-pandemic reopening phase 35 . Although a slightly higher proportion of female cases was observed, statistical analysis showed no significant association between gender and infection status. This aligns with earlier studies that reported comparable infection rates between males and females 35 . However, consistent with earlier findings, male patients have been reported to experience more severe disease outcomes and increased mortality 36 . Sequence change rates and time to the most recent common ancestors (TMRCA) using Markov chain Monte Carlo (MCMC) methods. * The SARS-CoV-2 variants, along with their descendants, have been included. † Mean value with the 95% HPD interval in parentheses. # Time before the present of the most recent common ancestor with the 95% HPD interval in parentheses. ǂ Mean pairwise P distances with standard deviation in parentheses.
## SARS-CoV
The current study revealed that the predominance of variants shifted throughout the year, with JN.1 dominating in the earlier part of the year, followed by transitions to KP.2 and LB.1. After August, KP.3 became the predominant variant, primarily represented by the sublineages KP.3.1.1, and KP.3.3.1. The present study aligns with a report from South Korea, which found that the proportions of various sub-lineages, including JN.1, KP.2, LB.1, and KP.3, exhibited the highest prevalence during the period from April to August 2024 37 . That report also indicated that the KP.3 sub-lineages with the highest proportions were identified as KP.3.3.1, KP.3.3, and KP.3.1.1. The ongoing genetic evolution of SARS-CoV-2 poses significant challenges for vaccine development, as newly emerging variants may diminish the efficacy of existing vaccines. The variants identified in Thailand closely align with those circulating globally, highlighting the interconnected nature of SARS-CoV-2 evolution and underscoring the need for continuous genomic surveillance and vaccine adaptation to effectively combat the evolving pandemic. Notably, while the overall variant dynamics in Thailand closely mirrored global trends, the absence of the newly emerging KP.3 sublineages (KP.3.1.1, KP.3.2, and KP.3.3.1) in Thai samples after September 2024 suggests potential regional differences in viral evolution and spread. This highlights the importance of timely genomic surveillance to detect emerging sublineages that may initially appear in global datasets but remain undetected locally.
The current study estimated a nucleotide evolutionary rate of approximately 0.63 × 10 -3 substitutions per site per year (sub/site/year), consistent with reported during the XBB wave in the Republic of Korea (late 2022 to late 2023), where substitution rates ranged from 0.56 × 10 -3 to 0.91 × 10 -3 sub/site/year 38 . Similarly, a study conducted in the United States reported a whole-genome evolutionary rate of 0.67 × 10 -3 (sub/site/year), further supporting the comparability of our findings 39 . When compared to other respiratory viruses, this rate is similar to that of human rhinovirus (0.66 × 10 -3 ) and human metapneumovirus (0.71 × 10 -3 ), while it is lower than that of respiratory syncytial virus (0.76 × 10 -3 ) and influenza A, which ranges from 2.21 to 3.37 × 10 -340-43 .
The present study provides important insights into the molecular epidemiology and evolutionary dynamics of SARS-CoV-2 circulating in Thailand and globally. By integrating genomic surveillance with phylogenetic analysis, the findings contribute to the early detection of emerging variants and the identification of transmission patterns. The ability to monitor viral evolution in near real-time supports early warning systems and enables the timely adjustment of public health measures. Moreover, the phylogenetic similarities observed between Thai and global samples emphasize the importance of international surveillance and data sharing to anticipate local outbreaks influenced by global viral movements. These insights are instrumental for guiding vaccine policy, resource allocation, and national preparedness planning, particularly in the context of rapidly evolving variants such as JN.1 and KP.1-KP. 3.
During the study period, vaccination strategies globally were updated to address the Omicron JN.1 subvariant, with reformulated vaccines being deployed. In Thailand, however, these JN.1-targeted vaccines were introduced later, toward the end of 2024, and their uptake was limited due to high cost and voluntary administration. This may have influenced the transmission dynamics observed in the region.
This study provides a comprehensive year-long analysis of the seasonal distribution of COVID-19 in Thailand, utilizing a substantial dataset collected over the entire year with a large number. However, several limitations must be acknowledged. Since the analysis focuses exclusively on a single year, multi-year investigations are necessary to fully understand the seasonal fluctuations of COVID-19 outbreaks. Regarding variant analysis, despite efforts to conduct sequencing throughout the year, resource constraints, including budget limitations, limited the ability to sequence all cases. The limited number and temporal distribution of fully sequenced genomes may affect the representativeness of lineage dynamics, so caution is warranted when interpreting results based on this subset. Additionally, in instances where patients exhibited low viral loads (as indicated by high Ct values), variant identification was not feasible. For evolutionary analysis, the inclusion of more robust metadata would enhance the accuracy and effectiveness of datasets used to estimate the nucleotide substitution rate and the time to the most recent common ancestor (tMRCA).
In conclusion, this study provides a comprehensive analysis of COVID-19 incidence and variant dynamics in Thailand throughout 2024. It highlights a notable seasonal trend in infection rates and emphasizes the predominance of specific SARS-CoV-2 lineages, including KP.3 and its sublineages, particularly in the latter part of the year. The findings underscore the importance of continuous genomic surveillance to monitor the evolving landscape of variants and inform public health strategies effectively. Additionally, the estimated molecular evolutionary rate and the emergence timing of key lineages contribute valuable insights to the understanding of SARS-CoV-2 evolution and its impact on the pandemic landscape in Thailand.
## Materials and methods
## Institutional review board statement
The research was carried out following the Good Clinical Practice (GCP) under the principles outlined in the Declaration of Helsinki and received approval from the Institutional Review Board of the Faculty of Medicine, Chulalongkorn University, Thailand (approval number IRB0933/67). To ensure patient confidentiality, all data and identifiers were anonymized. Since this was a retrospective study, the Institutional Review Board waived the requirement for informed consent.
## Specimens collection
In this study, 8,096 specimens were collected using throat swabs or nasopharyngeal swabs with Flock swabs (Copan Diagnostics, Murrieta, CA) or Modono sterile swabs (Modono, New Delhi, India), and were subsequently placed in viral transport media (VTM) or universal transport media (UTM). The samples were obtained from patients presenting with acute respiratory illness characterized by symptoms such as fever, sore throat, rhinorrhea, cough, or dyspnea within seven days of symptom onset. The study included both outpatients and inpatients seeking medical care at various hospitals in Bangkok as well as at Chum Phae Hospital in Khon Kaen province. Collected samples were tested for multiple respiratory viruses, including influenza viruses (A and B), SARS-CoV-2, respiratory syncytial virus (RSV), adenovirus, parainfluenza viruses, human metapneumovirus, seasonal coronaviruses, and rhinovirus. Residual specimens were then sent to the Center of Excellence in Clinical Virology at the Faculty of Medicine, Chulalongkorn University, where they were stored at -20 °C until further analysis for this study.
## RNA extraction and molecular testing
RNA extraction and molecular testing were conducted as previously outlined 23 . In brief, RNA was extracted from a 200-µL aliquot of the supernatant using the MagLEAD 12gC instrument (Precision System Science, Chiba, Japan), following the manufacturer's protocols meticulously. Real-time reverse transcription-PCR (RT-PCR) was performed using the Roche LightCycler ® 480 (LC480) instrument, employing specific primers and probes targeting the nucleocapsid gene (N1 and N2) in accordance with the guidelines established by the Centers for Disease Control and Prevention (CDC) 44 . The amplification of the housekeeping gene Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was included as an endogenous control for each sample, utilizing primers and probes as outlined in prior studies 45 . The thermocycling conditions included an initial 10-minute incubation at 45 °C, followed by a 5-minute denaturation step at 95 °C, and then 45 amplification cycles at 95 °C for 15 s and 60 °C for 30 s. Each testing run included positive and negative (nontemplate) controls, with a sample classified as positive for SARS-CoV-2 if the cycle threshold (CT) value was ≤ 38. Additionally, a selection of positive SARS-CoV-2 samples was randomly chosen for variant classification via partial spike sequencing and complete genome sequencing.
## Amplification and sequencing of the partial Spike gene
From January to December 2024, a total of 629 SARS-CoV-2 RNA samples were processed, with partial amplification of the spike gene performed on individuals in Thailand confirmed to be infected with COVID-19. The amplification utilized primers detailed in Supplementary Table 2. Briefly, RT-PCR was performed in a total volume of 25 µL, consisting of 2-3 µL of total RNA (ranging from 100 ng to 1 µg), 0.5 µM of each primer, 12.5 µL of 2X Reaction Mix (which contains 0.4 mM of each dNTP and 3.2 mM MgSO4), 1 µL of the SSIII RT/ Platinum Taq Mix, and nuclease-free water. The Superscript III One-Step RT-PCR system with Platinum Taq High Fidelity was utilized according to the manufacturer's guidelines (Invitrogen, Carlsbad, CA, USA). The PCR protocol involved an initial incubation at 45 °C for 30 min, followed by 40 cycles that included denaturation at 95 °C for 30 s, annealing at 50 °C for 30 s, and extension at 68 °C for 1 min and 45 s. A final extension step was carried Out at 68 °C for 5 min. Both forward and reverse primers were utilized concurrently for sequencing and product amplification, conducted at First BASE Laboratories Sdn Bhd (Selangor Darul Ehsan, Malaysia), ensuring comprehensive coverage and accuracy in the analysis.
## Whole-genome sequencing
Residual SARS-CoV-2 PCR-positive respiratory specimens (N = 60) were utilized for viral sequencing. Samples with a Ct value of 25 or lower were specifically selected for whole-genome sequencing, which was carried out by adapting a previously established protocol to enhance accuracy and efficiency in the genomic analysis 23 . The sequencing and identification of complete SARS-CoV-2 genomes were conducted using the Celemics Comprehensive Respiratory Virus Panel (Celemics Inc., Incheon, Republic of Korea), facilitating accurate and efficient genomic analysis. Briefly, RNA extraction involved combining 25 ng of isolated RNA with an RNA fragmentation buffer, followed by first-strand cDNA synthesis using a specialized master mix. The cDNA was converted into double-stranded form through incubation with a second-strand synthesis mix, after which it was cleaned, repaired, and modified with poly(A) tail oligomers. The A-tailed DNA was then ligated to adapters and purified using CeleMag cleanup beads before amplification to create an adapter-ligated library with CLM polymerase and unique dual-index primers. The quality of the library was assessed using automated capillary gel electrophoresis, ensuring DNA fragments were in the 200 to 400 bp range. Next-Generation Sequencing (NGS) was conducted on the Illumina NextSeq 500 system, and the resulting FASTQ data were trimmed, assembled, and analyzed through the Celemics Virus Verifier pipeline to generate consensus sequences.
## Maximum likelihood phylogenetic analysis and molecular clock assessment
In this study, a comprehensive dataset of 423 genomic sequences was compiled, which included 60 newly collected sequences from this research and 116 sequences from other provinces in Thailand (Supplement Table 3). This dataset was further enriched by integrating 247 globally representative SARS-CoV-2 genomes retrieved from the GISAID database within the timeframe of the study, from December 2023 to December 2024, thereby enhancing its robustness. The dataset was aligned utilizing MAFFT v.7 46 , and the alignments were subsequently partitioned by codon position. TreeTime 47 was used to construct a maximum likelihood tree, which provided the framework for the Bayesian phylodynamic analyses based on the tree topologies and corresponding collection dates. The resulting maximum likelihood phylogeny and time-scaled tree were visualized using the ggtree package 48 in R. v.4.4.2 49 . Additionally, a regression analysis of root-to-tip genetic distance against sampling time was conducted using TempEST v1.5.3 50 .
## Bayesian phylogenetic analysis and genetic distance estimation
A complete genome dataset was utilized to reconstruct time-scaled phylogenies through Bayesian inference, employing Markov Chain Monte Carlo (MCMC) techniques via the BEAST software (v.2.4.8) 51 . To determine the most appropriate evolutionary model, combinations of three coalescent tree priors (constant population size, exponential growth, and Bayesian skyline) and two molecular clock models (strict and uncorrelated lognormal relaxed clocks) were independently evaluated. Model fit was assessed through marginal likelihood estimation (MLE) via path sampling and stepping-stone sampling 52 . Log Bayes Factors (logBF), calculated as the difference in MLE between competing models, were interpreted following Kass and Raftery's criteria 53 , with logBF > 5 indicating very strong support. Among all tested models, the strict molecular clock with a constant population size prior exhibited the highest marginal likelihood and was decisively supported over alternative models (Supplement Table 4). Final phylogenetic inference was therefore conducted under the selected best-fit model. Two independent MCMC chains, each consisting of 200 million steps, were run and combined using the BEAGLE library 54 to enhance computational performance. Parameters and trees were sampled every 20,000 steps, with the initial 20% discarded as burn-in. Convergence and adequate sampling were confirmed in Tracer v1.7.1 55 , with effective sample sizes (ESS) exceeding 200 for all key parameters. The resulting posterior tree distributions were summarized using TreeAnnotator v1.8.4 to generate a maximum clade credibility (MCC) tree, and phylogenies were visualized with FigTree (https://github.com/rambaut/figtree/releases).
Average genetic distances were calculated using MEGA-X version [10.2.6], employing the Kimura 2-parameter model with variance estimation enabled 56 .
## Statistical analysis
Associations between the categorical variable gender and infection status were examined using Pearson's chisquare test of independence. Pairwise comparisons of infection proportions among age groups were conducted with Bonferroni correction to adjust for multiple testing. Statistical significance was set at p < 0.05. All analyses were performed using R version 4.4.2 49 .
## References
1. (2020) "Coronaviridae Study Group of the International Committee on Taxonomy of Viruses. The species severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2" *Nat. Microbiol*
2. Hu, Guo, Zhou et al. (2021) "Characteristics of SARS-CoV-2 and COVID-19" *Nat. Rev. Microbiol*
3. Malone, Urakova, Snijder et al. (2022) "Structures and functions of coronavirus replication-transcription complexes and their relevance for SARS-CoV-2 drug design" *Nat. Rev. Mol. Cell. Biol*
4. Yang, Rao (2021) "Structural biology of SARS-CoV-2 and implications for therapeutic development" *Nat. Rev. Microbiol*
5. Rashid (2022) "Roles and functions of SARS-CoV-2 proteins in host immune evasion" *Front. Immunol*
6. Mariano, Farthing, Lale-Farjat et al. (2020) "Structural characterization of SARS-CoV-2: where we are, and where we need to be" *Front. Mol. Biosci*
7. Fan (0997) "SARS-CoV-2 Omicron variant: recent progress and future perspectives" *Signal. Transduct. Target. Ther*
8. Fischer (2025) "Emergence and spread of the SARS-CoV-2 Omicron (BA.1) variant across africa: an observational study" *Lancet Glob Health*
9. Viana (2022) "Rapid epidemic expansion of the SARS-CoV-2 Omicron variant in Southern Africa" *Nature*
10. Karim, Karim, Omicron (2021) "variant: a new chapter in the COVID-19 pandemic" *Lancet*
11. Tan (2022) "Comparative neutralisation profile of SARS-CoV-2 omicron subvariants BA.2.75 and BA"
12. Cao (2022) "Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies" *Nature*
13. Dejnirattisai (2022) "SARS-CoV-2 omicron-B.1.1.529 leads to widespread escape from neutralizing antibody responses" *Cell*
14. Rambaut (2020) "Addendum: A dynamic nomenclature proposal for SARS-CoV-2 lineages to assist genomic epidemiology" *Nat. Microbiol*
15. O'toole (2021) "Assignment of epidemiological lineages in an emerging pandemic using the Pangolin tool" *Virus Evol*
16. Tamura (2023) "Virological characteristics of the SARS-CoV-2 XBB variant derived from recombination of two Omicron subvariants" *Nat. Commun*
17. Ma (2023) "Genomic surveillance for SARS-CoV-2 variants: circulation of Omicron XBB and JN.1 Lineages -United states" *MMWR Morb Mortal. Wkly. Rep*
18. Lu, Ao, He et al. (2020) "The rising SARS-CoV-2 JN.1 variant: evolution, infectivity, immune escape, and response strategies" *MedComm*
19. Ou (2024) "Evolving immune evasion and transmissibility of SARS-CoV-2: the emergence of JN.1 variant and its global impact" *Drug Discov Ther*
20. (2025) "WHO COVID-19 Dashboard"
21. Buathong (2021) "Multiple clades of SARS-CoV-2 were introduced to Thailand during the first quarter of 2020" *Microbiol. Immunol*
22. Puenpa (2023) "Investigation of the molecular epidemiology and evolution of Circulating severe acute respiratory syndrome coronavirus 2 in Thailand from 2020 to 2022 via Next-Generation sequencing" *Viruses*
23. Puenpa (2022) "Genomic epidemiology and evolutionary analysis during XBB.1.16-predominant periods of SARS-CoV-2 Omicron variant in bangkok" *Sci. Rep*
24. (2025) "Global Initiative on Sharing All Influenza Data (GISAID)"
25. Zanobini (2022) "Global patterns of seasonal influenza activity, duration of activity and virus (sub)type circulation from 2010 to 2020. Influenza Other Respir Viruses"
26. Suntronwong "Climate factors influence seasonal influenza activity in bangkok"
27. (2020)
28. (2025) "COVID-19 Situation"
29. Hinjoy (2020) "Self-assessment of the Thai department of disease control's communication for international response to COVID-19 in the early phase" *Int. J. Infect. Dis*
30. Suntronwong (2017) "Characterizing genetic and antigenic divergence from vaccine strain of influenza A and B viruses Circulating in Thailand" *Sci. Rep*
31. Thongpan, Vongpunsawad, Poovorawan (2020) "Respiratory syncytial virus infection trend is associated with meteorological factors" *Sci. Rep*
32. (2024) "COVID-19 can surge throughout the year"
33. Patterson (2024) "Characterising COVID-19 school and childcare outbreaks in Canada in 2021: a surveillance study" *BMJ Public. Health*
34. Shamsa, Shamsa, Zhang (2023) "Seasonality of COVID-19 incidence in the united States" *Front. Public. Health*
35. Zaher, Basingab, Alrahimi et al. (2023) "Gender differences in response to COVID-19 infection and vaccination" *Biomedicines*
36. Jin (2020) "Gender differences in patients with COVID-19: focus on severity and mortality" *Front. Public. Health*
37. Kim (2024) "Prevalence of recent COVID-19 variants and Cell-based Omicron KP.3 infectivity analysis" *PHWR*
38. No (2024) "Dynamics of SARS-CoV-2 variants during the XBB wave in the Republic of Korea" *Virus Res*
39. Wang (2022) "Molecular evolutionary characteristics of SARS-CoV-2 emerging in the united States" *J. Med. Virol*
40. Phyu (2019) "Evolutionary dynamics of Whole-Genome influenza A/H3N2 viruses isolated in Myanmar from 2015 to" *Viruses*
41. Bhattacharjee, Chakrabarti, Kanungo et al. (2023) "Evolutionary dynamics of influenza A/H1N1 virus Circulating in India from 2011 to 2021" *Infect. Genet. Evol*
42. Briese (2008) "Global distribution of novel rhinovirus genotype" *Emerg. Infect. Dis*
43. Yang (2009) "Genetic diversity and evolution of human metapneumovirus fusion protein over Twenty years" *Virol. J*
44. (2019) "The CDC 2019 novel coronavirus (2019 nCoV) real time Rt-PCR diagnostics"
45. Puenpa, Suwannakarn, Chansaenroj et al. (2017) "Development of single-step multiplex real-time RT-PCR assays for rapid diagnosis of enterovirus 71, coxsackievirus A6, and A16 in patients with hand, foot, and mouth disease" *J. Virol. Methods*
46. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol. Biol. Evol*
47. Sagulenko, Puller, Neher (2018) "& TreeTime Maximum-likelihood phylodynamic analysis" *Virus Evol*
48. 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*
49. Core (2014) "R: A Language and Environment for Statistical Computing"
50. Rambaut, Lam, Carvalho et al. (2016) "Exploring the Temporal structure of heterochronous sequences using tempest (formerly Path-O-Gen)" *Virus Evol*
51. Bouckaert (2014) "BEAST 2: a software platform for bayesian evolutionary analysis" *PLoS Comput. Biol*
52. Lartillot, Philippe (2006) "Computing Bayes factors using thermodynamic integration" *Syst. Biol*
53. Kass, Raftery (1995) "Bayes factors" *J. Am. Stat. Assoc*
54. Ayres (2012) "BEAGLE: an application programming interface and high-performance computing library for statistical phylogenetics" *Syst. Biol*
55. Rambaut, Drummond, Xie et al. (2018) "Posterior summarization in bayesian phylogenetics using tracer 1.7" *Syst. Biol*
56. Kumar, Stecher, Li et al. (2018) "MEGA X: molecular evolutionary genetics analysis across computing platforms" *Mol. Biol. Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793654&blobtype=pdf | # P-2185. Management of CMV Prophylaxis in Lung Transplant Patients at a Large Academic Medical Center
Kaitlyn Reasoner, Andrea Ito, Richard Merkhofer, Michael Zou, Kyle Enriquez, Emily Moore, Barbara Mora, Wonbeom Paik, Luke Pryke, Megan Uehling, Ashley Zeoli, ; Staub, ; Augusto, Dulanto Chiang, Casey Smiley
Background. Standard cytomegalovirus (CMV) prevention in transplant recipients uses antiviral prophylaxis based on risk stratification by CMV serology. The American Society of Transplantation updated guidelines for CMV management in 2025, but practice varies widely. Lung transplant prophylaxis durations at Vanderbilt University Medical Center (VUMC) tend to be shorter than other institutions with maximum duration of 9 months for high-risk serotypes (Fig 1). This project aimed to assess current practice variation and assess organization readiness for a standardized approach incorporating CMV T-cell immunity testing at Vanderbilt University Medical Center (VUMC). Results. During the study period, 165 patients collectively had 4485 CMV quantitative PCR tests, and 54 patients (32.9%) had a CMV quantitative PCR >500 IU/mL (Table 1). Twenty lung transplant clinic providers completed ORC surveys (Figures 2 &3), and 7 (46.7%) providers felt patients received appropriate CMV prophylaxis durations, but 15 (66.7%) providers felt prophylaxis duration was changed somewhat or very often due to valganciclovir toxicities or cost.
Conclusion. Our results suggest there are opportunities for improving the VUMC CMV prophylaxis process for lung transplant patients and clinic providers. VUMC plans to implement a T-cell-mediated immunity test which may improve CMV risk-stratification and optimize antiviral prophylaxis duration, thereby reducing direct and indirect effects of CMV infection and antiviral toxicities. Forthcoming data on patient-specific factors that increase reactivation risk may aid a more nuanced risk stratification and testing strategy. The Second Affiliated Hospital of Xi'an Jiaotong University, Xi'an, Shaanxi, China 2 Xi'an Jiaotong University Second Affiliated Hospital, Xi'an, Shaanxi, China 3 Second Hospital of Xi'an Jiaotong University, Xi'an, Shaanxi, China |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817895&blobtype=pdf | # SGIV envelope protein VP088 facilitated virus replication via interacting with other viral proteins and promoting p62dependent autophagic degradation of TBK1
Mengdi Yuan, Ya Zhang, Xiaolin Gao, Wenji Wang, Yin Zhao, Qiwei Qin, Xiaohong Huang, Youhua Huang
## Abstract
Singapore grouper iridovirus (SGIV), a novel member of the genus Ranavirus, family Iridoviridae, frequently causes a severe disease with high mortality in grouper aquaculture. Although previous findings have demonstrated that SGIV envelope protein VP088 was crucial for its infectivity, the underlying mechanism still remained uncertain. Here, we screened the potential viral proteins that interacted with VP088 during SGIV infection using GFP pull-down assay. Co-immunoprecipitation (Co-IP) assays verified the interactions between VP088 and VP018, VP068, or VP156 in vitro. Moreover, confocal microscopy analysis showed that VP088 markedly altered the cellular distri bution of exogenously expressed VP018 and VP068 and ultimately translocated into virus assembly sites together upon SGIV infection. Differently, VP088 mostly co-localized with exogenous VP156 in co-transfected cells and almost simultaneously translocated into the virus assembly sites, suggesting that VP088 participated in SGIV replication through interactions with other viral proteins in different ways. Interestingly, VP088 also abrogated IFN response induced by grouper (Epinephelus coioides) EccGAS-EcSTING and EcTBK1 in vitro. Co-IP assays showed that VP088 interacted with EccGAS, EcST ING, EcTBK1, or EcIRF3, while only degraded EcTBK1 via Ecp62-mediated autophagic degradation. Furthermore, VP088 decreased EcTBK1-induced EcIRF3 phosphorylation and nuclear translocation. In addition, the ectopic expression of VP088 attenuated the antiviral function of EcSTING/EcTBK1/EcIRF3 against red-spotted grouper nervous necrosis virus (RGNNV) infection. Thus, our results not only identified the association between SGIV VP088 and other viral proteins during replication, but also for the first time demonstrated that an iridoviral envelope protein could function as an immune evasion protein via abrogating EcTBK1-induced interferon response. IMPORTANCE Iridovirus infection frequently causes high levels of morbidity and mortality among commercially and ecologically important fish, crustaceans, amphibians, and reptiles. However, the molecular mechanism of iridovirus pathogenesis still remains largely unknown, and few effective countermeasures have been developed to date. Using the Singapore grouper iridovirus (SGIV) infection model in vitro, we identified the potential viral proteins that interacted with envelope protein VP088 during virus replication. Moreover, for the first time, we demonstrated that VP088 interacted with EccGAS, EcSTING, EcTBK1, and EcIRF3, but only degraded EcTBK1 via Ecp62-mediated autophagic degradation, thereby inhibiting the host IFN response. Thus, our results not only contribute to elucidating the mechanism of SGIV pathogenesis but also provide a novel molecular target for the construction of immunogenic live vaccines against iridoviral diseases in the future.
I nnate immunity is an important first-line defense against various pathogens. Upon virus invasion, host cell pattern recognition receptors (PRRs) recognize pathogen-asso ciated molecular patterns of these viruses and trigger a cascade of signal transduction to activate interferon-mediated innate immune responses (1)(2)(3). Among these PRRs, cyclic GMP-AMP synthase (cGAS) senses cytoplasmic double-stranded DNA, stimulating its own activity to form cGAMP (4), and then cGAMP acts as a second messenger to activate the adaptor stimulator of interferon genes (STING) (5). Subsequently, STING recruits TANK-binding kinase 1 (TBK1) and IFN regulatory factor 3 (IRF3), leading to their phosphorylation (6). The phosphorylated IRF3 forms dimers and translocates into the nucleus and finally induces the expression of the interferon-stimulated genes (ISGs) to resist the invading pathogens, especially DNA viruses (7).
During the co-evolution between virus and host, viruses have also developed various strategies to evade host immune attack for efficient replication (8,9). For DNA viruses, the increasing amount of evidence reveals that multiple viral proteins are able to abrogate the host IFN immune response by interfering with the normal functions of key molecules in the cGAS-STING signaling pathway (10)(11)(12). For instance, herpes simplex virus 1 (HSV-1) VP22 inhibited the enzymatic activity of cGAS (13), while UL41 targeted the degradation of cGAS to evade the immune response mediated by cGAS-STING (14). HSV-1 UL38 also inhibited the binding of cGAMP to STING through direct interaction with STING, thereby blocking the formation of STING-TBK1-IRF3 complex and antagoniz ing the IFN immune response (15). Notably, structurally related herpesvirus tegument proteins, such as VP22 and UL37, not only played roles in viral entry and viral capsid transportation but also suppressed the type I IFN signaling pathway and other innate immune responses (13,16). Moreover, the deletion of such viral immune evasion genes might provide a novel strategy for constructing safe and immunogenic live vaccines against infectious pathogens (17).
Singapore grouper iridovirus (SGIV), a novel member of the genus Ranavirus, is a highly pathogenic DNA virus isolated from the spotted-grouper (Epinephelus tauvina), causing more than 90% mortality in juvenile grouper (18)(19)(20). The high pathogenicity might be partially due to its ability to disrupt the host antiviral immune response (21). Consistently, recent studies have demonstrated that several viral proteins from SGIV abrogate host antiviral immune response for efficient replication (22)(23)(24)(25). For instance, VP131 interacted with EcSTING and degraded it through both the autophagy-lyso some pathway and the ubiquitin-proteasome pathway (24). Similarly, VP018 degraded EcSTING, EcTBK1, and EcIRF3 proteins in vitro and reduced their induction of interferon response (25). Similar to HSV-1, SGIV also encoded multiple structural proteins. Among them, VP088 was identified as an envelope protein involved in virus entry (26), and deletion of VP088 significantly reduced the infectivity of SGIV, but it showed no obvious effects on genomic stability, replication, and release of progeny viruses during SGIV infection (26,27). Whether VP088 exerted other functions during SGIV-host interactions remained uncertain.
In this study, we first screened the potential viral proteins that interacted with VP088 during SGIV infection using GFP pull-down and MS analysis and verified their interactions using co-immunoprecipitation (Co-IP) assay. Based on the results that VP088 interacted with VP018 and VP018 abrogated host interferon response during SGIV infection (25), we also clarified the regulatory roles of VP088 on EcSTING-EcTBK1-mediated antiviral IFN-I signaling. Our results will not only shed new light on understanding the roles of structural proteins in fish virus infection but also provide novel molecular targets for the development of vaccines against iridoviral diseases in the future.
## MATERIALS AND METHODS
## Cells and viruses
Grouper spleen (GS) cells were established by our laboratory and subcultured in Leibovitz's L15 medium (Gibco) supplemented with 10% fetal bovine serum (Excell) at 28°C (28). SGIV and red-spotted grouper nervous necrosis virus (RGNNV) were propaga ted in GS cells. Virus stocks were stored at -80°C until used.
## Antibodies and reagents
The anti-HA and anti-Flag antibodies were purchased from Cell Signaling. Anti-GFP and anti-β-tubulin antibodies were purchased from Abcam. Rabbit against SGIV VP018, VP068, and VP156 polyclonal antibodies and mouse against VP088 and RGNNV capsid protein (CP) antibodies were prepared by Wuhan GeneCreate Biological Engineering Co., Ltd (China) (21,28). The anti-IRF3, anti-p-IRF3, and anti-Lamin B1 antibodies were purchased from ABclonal. The anti-p62 antibody was obtained from Selleck. HRP-conju gated anti-HA and anti-Flag were purchased from AlpalifeBio. Peroxidase-conjugated goat anti-mouse or anti-rabbit IgG, Alexa Fluor 555 goat anti-mouse IgG (H+L), and Alexa Fluor 488 goat anti-rabbit IgG (H+L) were purchased from Thermo Fisher Scientific.
MG132 (ubiquitin-proteasome inhibitor), NH 4 Cl (lysosome inhibitor), and 4,6-diami dino-2-phenylindole (DAPI) were purchased from Sigma Aldrich, and 3-MA (autophagy inhibitor) was purchased from Selleck. Paraformaldehyde (PFA) was purchased from Thermo Fisher Scientific. The stock of NH 4 Cl and 3-MA was prepared using sterile water, and MG132 was diluted with dimethyl sulfoxide.
## Plasmid construction
The full-length SGIV VP088 (GenBank accession no. YP_164183.1) was amplified by PCR from SGIV genomic DNA. To investigate the subcellular localization and function of VP088, the full-length ORF of VP088 was subcloned into pcDNA3.1-HA-N (Clon tech), pCMV-3×Flag-N (Clontech), pEGFP-C1 (Clontech), and pmCherry-C1 (Clontech), respectively. The recombinant plasmids, including pcDNA3.1-HA-VP088 (pHA-VP088), pCMV-3×Flag-VP088 (pFlag-VP088), pEGFP-C1-VP088 (pGFP-VP088), and pmCherry-C1-VP088 (pmCherry-VP088), were validated by sequencing. The full-length ORF of Ecp62 was subcloned into pcDNA3.1-HA-N (Clontech) to obtain pcDNA3.1-HA-Ecp62 (pHA-Ecp62). The corresponding primers used in this study were listed in Table S1 (Table S1).
The recombinant plasmids, including pHA-VP018, pHA-VP068, pHA-VP156, pHA-EcSTING, pHA-EcTBK1, pHA-EcTBK1-∆C, pHA-EcTBK1-∆N, pHA-EcIRF3, pFlag-EccGAS, pFlag-EcSTING, pFlag-EcTBK1, pGFP-VP018, pGFP-VP068, pGFP-VP156, pGFP-EccGAS, pGFP-EcSTING, pGFP-EcTBK1, and pGFP-EcIRF3, were described previously (23,24,29).
## Cell transfection and virus infection
GS cells were seeded in 24-well plates or dishes overnight and transfected with the indicated plasmids using Lipofectamine 2000 (Thermo Fisher Scientific) according to the manufacturer's instructions. To screen the potential viral or cellular proteins that interacted with VP088, GS cells were seeded into 10 cm² dishes and then transfected with pGFP-C1 or pGFP-VP088. At 24 h post-transfection, cells were infected with SGIV and then collected for Co-IP assays and LC-MS/MS analysis (Guangzhou Fitgene Biotechnol ogy Co., Ltd., China) as described previously (30).
To evaluate the effect of VP088 on the expression of IFN-related genes induced by EccGAS/EcSTING or EcTBK1, GS cells were co-transfected with pHA-VP088 and pFlag-Ecc GAS plus pGFP-EcSTING or pGFP-EcTBK1, respectively. At 48 h post-transfection, cells were collected for quantitative PCR (qPCR) analysis.
In addition, to clarify whether VP088 affected the antiviral activities of EcSTING, EcTBK1, or EcIRF3, GS cells were co-transfected with pHA-EcSTING, pHA-EcTBK1, or pHA-EcIRF3, together with either pFlag-VP088 or the empty vector pCMV-3×Flag. At 24 h post-transfection, cells were infected with RGNNV (which lacks VP088 or its homolog). The infected cells were harvested for subsequent qPCR and immunoblotting (IB) analysis.
## Subcellular localization
To observe the localization of VP088 and its interacting viral proteins, GS cells were seeded into glass bottom cell culture dishes (35 mm) and co-transfected with 0.5 µg pmCherry-VP088 and 0.5 µg pGFP-VP018, pGFP-VP068, or pGFP-VP156, respectively. To examine the trafficking events of these viral proteins during SGIV infection, the co-transfected cells were infected with SGIV at 24 h post-transfection and then harvested after another 24 h incubation.
To detect the localization of VP088 and its interacting cellular proteins, GS cells were co-transfected with pmCherry-VP088 and pGFP-EccGAS, pGFP-EcSTING, pGFP-EcTBK1, or pGFP-EcIRF3, respectively. At 48 h post-transfection, cells were fixed and stained with DAPI for 10 min. All samples were imaged under a confocal laser scanning microscope (CLSM, Leica) equipped with an oil-immersion objective lens (63×).
## Indirect immunofluorescence assay (IFA)
To determine the intracellular distribution of viral proteins during infection, IFA was carried out using specific antibodies against different viral proteins as described previously (24). In brief, GS cells were seeded into 35 mm glass bottom cell culture dishes overnight and infected with SGIV for 12 or 24 h. Then, the cells were fixed by 4% PFA for 1 h and then permeabilized with 0.1% Triton X-100 for 15 min at room temperature. After blocking with 2% bovine serum albumin, cells were incubated with primary antibodies for 2 h. The primary antibodies used included anti-HA (1:3,000), anti-VP088 (1:1,000), anti-VP018 (1:1,000), anti-VP068 (1:1,000), and anti-VP156 (1:1,000). Then, cells were washed and incubated with Alexa Fluor 555-conjugated goat anti-mouse IgG (1:800) or Alexa Fluor 488-conjugated goat anti-rabbit IgG (1:800) for another 2 h. Finally, the nuclei were stained with DAPI, and the samples were observed under a CLSM.
## Dual-luciferase reporter assay
To evaluate the effect of SGIV VP088 on cellular IFN immune response, the IFN1, IFN3, and ISRE promoter activities were determined using dual-luciferase reporter assay as described previously (24). Briefly, GS cells were seeded into 12-well plates overnight and then co-transfected with pHA-VP088 (or the empty vector control), along with pGFP-EcTBK1 (or the combination of pFlag-EccGAS and pGFP-EcSTING), reporter gene plasmid (pIFN1-Luc, pIFN3-Luc, or pISRE-Luc), and internal control plasmid pRL-SV40 (Promega). At 24 h post-transfection, cells were harvested and lysed to detect the luciferase activities using a Dual-Luciferase Reporter System (Promega) according to the manufacturer's manual. Luciferase activities were read using a microplate reader from Tecan (Switzerland).
## RNA interference
To evaluate the effect of VP088 knockdown on IFN immune response during SGIV infection, three specific small interfering RNA (siRNA) oligonucleotides targeting VP088 (si-VP088) and stealth RNAi negative control (siRNA-NC) were designed by Sangon Biotech (Shanghai), and the specific sequences were listed in Table S1. GS cells were transfected with si-VP088-1, si-VP088-2, si-VP088-3, or siRNA-NC for 24 h and then infected with SGIV for another 24 h. The silencing efficiency of si-VP088 was determined by western blotting using anti-VP088 (1:1,000), and si-VP088-1 was screened as the optimal candidate. Subsequently, GS cells transfected with si-VP088-1 (160 nM/well) or siRNA-NC were infected with SGIV and harvested at 24 h post-infection (p.i.) for qPCR analysis.
To determine whether Ecp62 was involved in VP088-mediated degradation of EcTBK1, three specific si-RNAs targeting Ecp62 (si-Ecp62) were synthesized by Sangon Biotech (Shanghai), and the silencing efficiency was determined by IB analysis. Subsequently, GS cells were transfected with si-Ecp62-1, pHA-EcTBK1, and pFlag-VP088 for 48 h and then harvested for IB analysis.
## RNA isolation and qPCR
Total RNA was extracted from cells using the Cell Total RNA Isolation Kit (FORE GENE) according to the manufacturer's instructions and then reverse transcribed using the ReverTra Ace qPCR RT Kit (TOYOBO). Subsequently, a qPCR assay was performed using the 2× PolarSignal SYBR Green qPCR Mix (MIKX) under an Applied Biosystems Quant Studio 5 Real Time Detection System (Thermo Fisher Scientific). The qPCR conditions were as follows: 94°C for 20 s, followed by 40 cycles at 94°C for 10 s, 55°C for 10 s, and 72°C for 10 s. The relative expression levels (fold change) were calculated with the 2 -ΔΔCt method using β-actin as an internal control. The IFN-related genes, including ISG15, ISG56, Viperin, and myxovirus resistance gene (MX1), were detected, and the corresponding primers were used as described previously (24).
## Co-IP assay
The interactions between VP088 and key molecules (including VP018, VP068, VP156, EccGAS, EcSTING, EcTBK1, EcTBK1-∆C, EcTBK1-∆N, EcIRF3, and Ecp62) were carried out using Co-IP assay as described previously (24). Briefly, GS cells were seeded into 10 cm 2 dishes overnight and then co-transfected with pHA-VP088, pFlag-VP088, or pGFP-VP088 and pHA-VP018, pHA-VP068, pHA-VP156, pFlag-EccGAS, pHA-EcSTING, pHA-EcTBK1, pHA-EcTBK1-∆C, pHA-EcTBK1-∆N, pHA-EcIRF3, pHA-Ecp62, or empty vectors, respec tively. At 48 h post-transfection, cells were collected and lysed by IP lysis buffer (Thermo Fisher Scientific). The whole cell lysis buffer (WCLs) was centrifuged at 12,000 × g for 5 min, and the supernatants were collected for immunoprecipitation using the HA-tag Protein IP Assay Kit with Magnetic Beads (Beyotime), Flag-tag Protein IP Assay Kit with Magnetic Beads (Beyotime), and ChromoTek GFP-Trap Magnetic Particles Kit (Proteintech). These immunoprecipitated proteins and WCLs were further subjected to IB analysis using antibodies as indicated.
## IB analysis
For IB analysis, the immunoprecipitated proteins and WCLs were separated by 10% SDS-PAGE and then transferred onto 0.2 µm polyvinylidene difluoride membranes (Millipore). The membranes were blocked with 5% skim milk in tris-buffer saline with 0.5% Tween-20 (TBST) for 2 h at room temperature. Subsequently, the membranes were incubated with the indicated primary antibodies, including anti-GFP (1:5,000), anti-HA (1:5,000), anti-Flag (1:5,000), anti-VP088 (1:1,000), anti-p62 (1:1,000), anti-IRF3 (1:1,000), anti-p-IRF3 (1:1,000), anti-Lamin B1 (1:1,000), anti-HA (HRP) (1:1,000), anti-Flag (HRP) (1:1,000), anti-RGNNV CP (1:2,000), or anti-β-tubulin (1:5,000) for another 2 h. After washing, the membranes were incubated with secondary antibodies, such as horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit IgG (1:10,000). Finally, the membranes were visualized with the ECL chemiluminescence solution, and the intensities of protein bands were quantified using Image J software. Data were representative of three independent experiments.
## Nucleocytoplasmic separation
To clarify whether VP088 affected the nuclear translocation of EcIRF3, GS cells were transfected with pFlag-EcTBK1 alone or co-transfected with pGFP-VP088 for 48 h. The nucleocytoplasmic separation was performed using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime) as described previously (25), and then the different fractions were subjected to IB assay.
## Statistical analyses
All statistical analyses were carried out using GraphPad Prism Software. The obtained data were expressed as mean ± standard deviation. Differences between two groups were calculated using two-tailed Student t-tests. P < 0.05 was considered statistically significant (*P < 0.05).
## RESULTS
## VP088 interacted with VP018/VP068/VP156
As an envelope protein, SGIV VP088 was demonstrated to be involved in viral assem bly and function as an anchor protein during SGIV infection (31). Thus, we identified the potential proteins that interact with VP088 using GFP pull-down assay, followed by MS analysis (Table S2). The MS results showed that VP088 was associated with viral proteins including SGIV VP018, VP068, and VP156. Here, Co-IP assays verified the interactions between VP088 and VP018, VP068, or VP156 in vitro without SGIV infection (Fig. 1A, D, andG). Moreover, confocal microscopy analysis showed that VP088 altered the intracellular distribution of VP018 and VP068 in co-transfected cells. In detail, the fluorescence of VP018 alone was present throughout the cells in both the nucleus and the cytoplasm. However, the co-expression of VP088 prevented the trafficking of VP018 from the cytoplasm into the nucleus (Fig. 1B andC). Differently, VP088 co-expression only partially prevented the transport of the fluorescence of VP068 from the cytoplasm into the nucleus, and a small fraction of fluorescence in the cytoplasm was overlapped between VP088 and VP068 (Fig. 1E andF). In addition, the fluorescence of VP156 was localized in the cytoplasm and partially overlapped with that of VP088 (Fig. 1H andI).
## VP088 and its interacting proteins were involved in virus assembly
To clarify whether VP088-interacting viral proteins were involved in virus assembly, GS cells were co-transfected with pmCherry-VP088 and pGFP-VP018, pGFP-VP068, or pGFP-VP156 and then incubated with SGIV for 12 or 24 h, respectively. Extranuclear DAPI staining was used to identify the sites of viral DNA replication indicative of virus assembly sites (32). As shown in Fig. 2, VP018 and VP068 were partially translocated into the virus assembly sites before VP088 at 12 h p.i., although both of them were present there at 24 h p.i. (Fig. 2A andB). Differently, the fluorescence of VP156 and VP088 was almost simultaneously present in the virus assembly sites, and their abundances were both increased at 24 h p.i. (Fig. 2C). Using the specific antibodies, we found that the majority of fluorescence from endogenous VP018, VP068, VP156, and VP088 proteins was all present in the virus assembly sites at 24 h p.i. The minority of fluorescence was labeled on the viral particles which were present throughout the cytoplasm (Fig. 2D). Thus, our results showed that VP088 might participate in SGIV replication through the interactions with other viral proteins in different ways.
## VP088 abrogated the host IFN response
Given that VP088 interacted and altered the localization of VP018 without virus infection, and VP018 abrogated the host IFN response for efficient replication (25), we speculated that VP088 might also antagonize host innate immune defenses during virus infection. Interestingly, our findings showed that VP088 overexpression alone down-regulated the host IFN response (Fig. 3A). To assess the effect of VP088 knockdown on host response, we designated the specific siRNAs for VP088 and first detected their silencing effects on VP088 protein synthesis during SGIV infection. According to the knockdown efficacy, si-VP088-1 was chosen for the following experiment (Fig. 3B). Consistently, transfection of si-VP088-1 led to the restoration of the IFN response compared with wild-type SGIV-infected cells (Fig. 3C). Moreover, VP088 overexpression not only significantly down-regulated the promoter activities of IFN1 (Fig. 3D andH), IFN3 (Fig. 3E andI), and ISRE (Fig. 3F andJ) activated by EccGAS+EcSTING and EcTBK1 but also markedly decreased the transcription levels of ISG15, ISG56, Viperin, and MX1 induced by these proteins (Fig. 3G andK). Collectively, SGIV VP088 was speculated to abrogate EccGAS-EcSTING-EcTBK1-activated host IFN response in vitro.
## VP088 targeted EcTBK1 to regulate the IFN signaling
To clarify the critical molecules hijacked by VP088 to regulate the IFN response, we first performed Co-IP assay and western blotting analysis in vitro. GS cells were co-transfected with VP088 and EccGAS, EcSTING, EcTBK1, or EcIRF3, respectively. As shown in Fig. 4A, Flag-tagged EccGAS interacted with HA-tagged VP088 but not HA peptide when they were co-expressed in GS cells. Similarly, HA-tagged EcSTING (Fig. 4B), EcTBK1 (Fig. 4C), and EcIRF3 (Fig. 4D) all interacted with Flag-tagged VP088 but not Flag peptide when they were co-transfected in GS cells. Further, confocal microscopic observation revealed that the fluorescence signal from VP088 almost overlapped with those from EccGAS, EcSTING, EcTBK1, or EcIRF3 in co-transfected cells (Fig. 4E through H). Thus, VP088 was elucidated to interact with EccGAS, EcSTING, EcTBK1, and EcIRF3 in vitro.
Next, we examined the potential effect of VP088 on EccGAS-EcSTING in co-transfec ted cells. As shown in Fig. 5, VP088 overexpression showed no obvious effect on the stability of the exogenous EccGAS and EcSTING in transfected cells (Fig. 5A andB). Moreover, VP088 also did not affect the dimerization of EcSTING (Fig. 5C). Notably, once EcSTING and EcTBK1 were co-expressed, the formation of EcSTING-EcTBK1 complex was markedly reduced after VP088 overexpression (Fig. 5D). Thus, it was proposed that VP088 might target EcTBK1 but not EccGAS/EcSTING to regulate the IFN signaling.
## Ecp62 was essential for VP088-mediated autophagic degradation of EcTBK1
Overexpression of VP088 markedly decreased the abundance of EcTBK1 in co-transfected cells in a dose-dependent manner (Fig. 6A). To identify the key domains of EcTBK1 involved in its interaction with VP088, we constructed two EcTBK1 truncating mutants, including EcTBK1-∆C (lacking the C-terminal, amino acids 1-308) and EcTBK1-∆N (lacking the N-terminal transmembrane region, amino acids 309-723) (Fig. 6B). Interestingly, VP088 was immunoprecipitated with HA-tagged EcTBK1 and its truncated mutants, but not HA peptide when they were co-expressed in GS cells (Fig. 6C). Furthermore, confocal microscopic analysis showed that both EcTBK1-∆N and EcTBK1-∆C were partially co-localized with VP088 (Fig. 6D), suggesting that the interaction between VP088 and EcTBK1 was independent of its N-or C-terminal domain.
To further clarify the pathway involved in VP088-induced degradation of EcTBK1 protein in vitro, pFlag-VP088 and pHA-EcTBK1 were co-transfected into GS cells and treated with MG132, 3-MA, or NH 4 Cl to block different degradation pathways. Interest ingly, 3-MA and NH 4 Cl, but not MG132, markedly weakened the degradation effect of decreased EcTBK1-induced EcIRF3 phosphorylation (Fig. 7D). Consistently, western blotting analysis showed that VP088 markedly decreased the abundance of EcIRF3 in the nucleus fraction after EcTBK1 activation. Confocal microscopic observation also demon strated that the fluorescence of EcIRF3 in the nucleus was markedly reduced in VP088 and EcTBK1-EcIRF3 co-transfected cells compared with EcTBK1-EcIRF3 transfected cells (Fig. 7E andF). Thus, VP088 was speculated to affect EcIRF3 activity via decreasing EcTBK1-induced EcIRF3 phosphorylation and nuclear translocation.
## VP088 attenuated the antiviral actions of EcSTING-EcTBK1-EcIRF3 axis
Next, we also evaluated the regulatory effect of VP088 on the antiviral activities of EcSTING, EcTBK1, or EcIRF3. Overexpression of EcSTING, EcTBK1, or EcIRF3 all markedly inhibited the progression of RGNNV-induced cytopathic effect (CPE). However, their inhibitory effects were partially reversed by VP088 overexpression (Fig. 8A through C). Consistently, the transcription levels (Fig. 8D through F) and protein expression levels (Fig. 8G through I) of RGNNV CP were decreased by EcSTING, EcTBK1, or EcIRF3 overexpression but rescued by VP088 co-transfection. Thus, we speculated that VP088 antagonized the antiviral actions regulated by EcSTING-EcTBK1-EcIRF3 axis.
## DISCUSSION
During the co-evolution between viruses and their hosts, viruses have developed various strategies to evade the host immune defense. For DNA viruses, African swine fever virus (ASFV) B175L inhibited the downstream signals of IFN-mediated antiviral response by interfering with the interaction between cGAMP and STING (33). ASFV pI215L negatively regulated the cGAS-STING signaling pathway by recruiting RNF138 (34). In fish, cyprinid herpesvirus 2 (CyHV-2) KLP degraded STING by the autophagy-lysosomal pathway and further diminished the antiviral ability of STING (35). ORF67 also inhibited IFN expression by competitively obstructing STING phosphorylation (36). SGIV VP131 interacted with and degraded STING and TBK1, thereby disrupting the antiviral activity of STING or TBK1 (24). As a large DNA virus, whether more viral proteins were involved in immune evasion of SGIV remained fascinating.
Recent studies demonstrated that VP088 functioned as an endoplasmic reticulumlocalized protein and participated in virus assembly (31). As an anchor protein, VP088 (31). To clarify this hypothesis, we screened the potential interacting proteins using GFP pull-down assay, and the results showed that VP088 might be associated with VP018/VP068/VP156 during SGIV infection. Their interactions were verified in co-expressed cells by Co-IP analysis. Moreover, VP088 was able to alter the localization of VP018 and VP068 without SGIV infection. Interestingly, upon SGIV infection, the exogenous VP088 and VP156 were almost simultaneously transferred into the virus assembly sites, while VP018 and VP068 were translocated at the late stage of SGIV infection. Consistently, endogenous VP018, VP068, VP156, and VP088 were all present in the virus assembly sites during SGIV infection. Consistently, exogenous VP088 also colocalized with MCP (VP072) and translocated to the virus assembly sites together at the late stage of SGIV infection (31). Directional or temporal translocation events were found to facilitate localization-depend ent protein interactions and contribute to either host defense or virus replication (37). Therefore, we speculated that VP088 interacted with these viral proteins in different ways and participated in virus assembly during SGIV infection. Although we also identified several cellular proteins that interacted with VP088 during infection, such as Rab7 and Rab5C, their detailed roles in SGIV replication needed further investigation.
Notably, among these interacted viral proteins, VP018 was previously elucidated to exert dual roles during SGIV infection (25,38). VP018 not only played an important role in the expressions of viral late genes and virion assembly (38) but also abrogated EcSTINGinduced IFN response via disrupting the assembly of the EcSTING-EcTBK1 and EcTBK1-EcIRF3 complexes and reducing the nuclear translocation of EcIRF3 (25). Based on our findings that VP088 prevented the trafficking of VP018 into the nucleus completely, we speculated that VP088 might also function via regulating the host IFN response. As expected, the transcription levels of IFN-related genes in VP088 alone overexpressing cells were inhibited compared with control cells, while those in VP088-silenced infected cells were significantly upregulated compared with SGIV-infected cells, suggesting that VP088 was crucial for SGIV to inhibit the IFN response. Moreover, co-transfection with VP088 significantly inhibited EccGAS-EcSTING-and EcTBK1-induced IFN response, evidenced by the decrease of the IFN promoter activities and the expression levels of IFN-related genes. Similarly, exogenous DNA and RNA-mediated IFN activation were both abrogated by CyHV-2 KLP (35). SGIV VP131 alone also weakened EcSTING-, EcTBK1-, or EcMDA5-induced IFN response (24), suggesting that VP088 might be another candidate for SGIV to evade host antiviral immune response via EcSTING-EcTBK1-EcIRF3 axis.
Although numerous viral immune evasion proteins have been reported to abrogate cGAS-STING-mediated IFN signaling, their interactions with target molecules in the cGAS-STING pathway remained at a variety of levels (39,40). Here, the Co-IP assay showed that VP088 interacted with EccGAS, EcSTING, EcTBK1, and EcIRF3 in co-transfec ted cells. VP088 did not significantly alter the protein abundance of the exogenous EccGAS and EcSTING, as well as the level of EcSTING dimerization. However, VP088 reduced the assembly of the EcSTING-EcTBK1 complex. Consistently, VP088 decreased the abundance of exogenous EcTBK1 protein in a dose-dependent manner, and the degradation of EcTBK1 was via the Ecp62-mediated autophagy pathway, suggesting that EcTBK1 was a critical target for VP088 to regulate the host IFN response. Similarly, Avibirnavirus VP3 could inhibit TRAF6-mediated IFN-β production to evade host innate immunity by inducing TRAF6 autophagic degradation in a p62-dependent manner (41). Encephalomyocarditis Virus Structural Protein VP3 also triggered MAVS degradation through the p62-mediated autophagy pathway (42). Differently, Grass Carp Reovirus VP4 recruited and interacted with toll-interacting protein to degrade STING via the autoph agy-lysosome pathway (10). Thus, we speculated that Ecp62-mediated autophagy was crucial for VP088-triggered EcTBK1 degradation. Whether other autophagic receptors were involved in this process needed further investigation.
As a downstream target of EcTBK1, EcIRF3 exerted the physiological function, usually accompanied by its phosphorylation, dimerization, and nuclear translocation (43,44). The activated IRF3 subsequently caused the transcriptional activation of IFN and ISRE promoters, finally resulting in the induction of IFN-I and numerous ISGs for the establish ment of an antiviral state (7). Human cytomegalovirus US9 evaded the type I interferon response by disrupting IRF3 nuclear translocation (45). HSV-1 VP24 also abrogated the interaction between TBK1 and IRF3, hence impairing IRF3 activation including its phosphorylation and dimerization (46). In our study, although the presence of VP088 showed no obvious effect on the degradation of exogenous EcIRF3 protein, this not only hindered the assembly of EcTBK1-EcIRF3 complex but also reduced EcIRF3 dimerization. Moreover, EcTBK1-activated EcIRF3 phosphorylation and nuclear translocation were also markedly decreased in VP088 co-transfected cells, thereby interfering with the activation of IFN antiviral response. Consistently, VP088 overexpression attenuated the antiviral activities of EcSTING/EcTBK1/EcIRF3 against RGNNV infection. Thus, we speculated that the attenuated activation of IRF3 was crucial for VP088 to abrogate the host IFN response.
In conclusion, our study identified the interacting components of envelope pro tein VP088 during virus assembly. Moreover, we for the first time demonstrated that VP088 also functioned as an immune evasion protein via Ecp62-mediated autophagic degradation of EcTBK1, thereby inhibiting the host IFN response. Our results will provide a novel molecular target for the construction of immunogenic live vaccines against grouper iridoviral diseases.
## Funder Grant(s)
Author(s)
innovative team building project of Guangdong Modern Agricultural Industrial Technology System 2024CXTD27 Xiaohong Huang
## References
1. (2026) "3-MA (F), or NH 4 Cl (G), on the degradation effect of VP088 on exogenous EcTBK1. GS cells were co-transfected with pFlag-VP088 and pHA-EcTBK1, then treated without or with MG132, 3-MA, or NH 4 Cl, and collected for western blotting. (H, I) The interactions between Ecp62 and VP088 (H) or EcTBK1 (I) in vitro. GS cells were transfected with Ecp62 and VP088 or EcTBK1 and then collected for Co-IP assay and western blotting assay. (J) The silencing effect of siRNA on Ecp62 protein synthesis. GS cells were transfected with siRNA-NC, si-Ecp62-1, si-Ecp62-2, or si-Ecp62-3 and then collected for western blotting analysis. (K) Ecp62 was essential for the degradation effect of VP088 on EcTBK1. GS cells were co-transfected with si-Ecp62-1, pFlag-VP088, and pHA-EcTBK1, and then the cells were collected for western blotting analysis"
2. Osorio, Reis E Sousa (2011) "Myeloid C-type lectin receptors in pathogen recognition and host defense" *Immunity*
3. Akira, Uematsu, Takeuchi (2006) "Pathogen recognition and innate immunity" *Cell*
4. Takeuchi, Akira (2010) "Pattern recognition receptors and inflammation" *Cell*
5. 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*
6. Bridgeman, Maelfait, Davenne et al. (2015) "Viruses transfer the antiviral second messenger cGAMP between cells" *Science*
7. Ishikawa, Ma, Barber (2009) "STING regulates intracellular DNAmediated, type I interferon-dependent innate immunity" *Nature*
8. Dalskov, Gad, Hartmann (2023) "Viral recognition and the antiviral interferon response" *EMBO J*
9. Goubau, Deddouche, Reis E Sousa (2013) "Cytosolic sensing of viruses" *Immunity*
10. Beachboard, Horner (2016) "Innate immune evasion strategies of DNA and RNA viruses" *Curr Opin Microbiol*
12. Wang, Wang, Li et al. (2025) "Grass carp reovirus VP4 manipulates TOLLIP to degrade STING for inhibition of IFN production" *J Virol*
13. Shi, Jia, He et al. (2023) "Immune evasion strategy involving propionylation by the KSHV interferon regulatory factor 1 (vIRF1)" *PLoS Pathog*
14. Zhu, Zheng (2020) "The race between host antiviral innate immunity and the immune evasion strategies of herpes simplex virus 1" *Microbiol Mol Biol Rev*
15. Huang, You, Su et al. (2018) "Herpes simplex virus 1 tegument protein VP22 abrogates cGAS/STING-mediated antiviral innate immunity" *J Virol*
16. Su, Zheng (2017) "Herpes simplex virus 1 abrogates the cGAS/STINGmediated cytosolic DNA-sensing pathway via its virion host shutoff protein, UL41" *J Virol*
17. Wang, Peng, Fan et al. (2025) "Herpes simplex virus 1 encodes a STING antagonist that can be therapeutically targeted" *Cell Rep Med*
18. Zhang, Zhao, Xu et al. (2018) "Species-specific deamidation of cGAS by herpes simplex virus UL37 protein facilitates viral replication" *Cell Host Microbe*
19. Brar, Farhat, Sukhina et al. (2020) "Deletion of immune evasion genes provides an effective vaccine design for tumor-associated herpesviruses" *NPJ Vaccines*
20. Qin, Chang, Ngoh-Lim et al. (2003) "Characterization of a novel ranavirus isolated from grouper Epinephelus tauvina" *Dis Aquat Organ*
21. Qin, Lam, Sin et al. (2001) "Electron microscopic observations of a marine fish iridovirus isolated from brown-spotted grouper, Epinephelus tauvina" *J Virol Methods*
22. Liu, Zhang, Yuan et al. (2024) "Integrated multi-omics analysis reveals liver metabolic reprogramming by fish iridovirus and antiviral function of alpha-linolenic acid" *Zool Res*
23. Wang, Zhang, Guo et al. (2023) "Singapore grouper iridovirus infection counteracts poly I:C induced antiviral immune response in vitro" *Fish Shellfish Immunol*
24. Gao, Lin, Zhao et al. (2024) "SGIV evades interferon immune response via the degradation of STING-TBK1 complex by VP149" *Aquaculture*
25. Wang, Liu, Wang et al. (2024) "SGIV VP82 inhibits the interferon response by degradation of IRF3 and IRF7" *Fish & Shellfish Immunology*
26. Zhang, Gao, Yang et al. (2022) "Singapore grouper iridovirus VP131 drives degradation of STING-TBK1 pathway proteins and negatively regulates antiviral innate immunity" *J Virol*
27. Wang, Zhi, Liu et al. (2025) "Singapore grouper iridovirus VP018 abrogates the interferon response by targeting STING-TBK1-IRF3 axis" *Int J Biol Macromol*
28. Yuan, Wang, Liu et al. (2016) "Singapore grouper iridovirus protein VP088 is essential for viral infectivity" *Sci Rep*
29. Yuan, Hong (2016) "Subcellular redistribution and sequential recruitment of macromolecular components during SGIV assembly" *Protein Cell*
30. (2026) *Full-Length Text Journal of Virology*
31. Huang, Huang, Sun et al. (2009) "Characterization of two grouper Epinephelus akaara cell lines: application to studies of Singapore grouper iridovirus (SGIV) propagation and virus-host interaction" *Aquaculture*
32. Zhang, Zhang, Liao et al. (2023) "Grouper cGAS is a negative regulator of STINGmediated interferon response" *Front Immunol*
33. Zhang, Luo, Lv et al. (2023) "TRAP1 inhibits MARCH5-mediated MIC60 degradation to alleviate mitochondrial dysfunction and apoptosis of cardiomyocytes under diabetic conditions" *Cell Death Differ*
34. Zhao, Huang, Liu et al. (2023) "Near-atomic architecture of Singapore grouper iridovirus and implications for giant virus assembly" *Nat Commun*
35. Stefanovic, Windsor, Nagata et al. (2005) "Vimentin rearrangement during African swine fever virus infection involves retrograde transport along microtubules and phosphorylation of vimentin by calcium calmodulin kinase II" *J Virol*
36. Ranathunga, Dodantenna, Cha et al. (2023) "African swine fever virus B175L inhibits the type I interferon pathway by targeting STING and 2'3'-cGAMP" *J Virol*
37. Huang, Xu, Liu et al. (2021) "African swine fever virus pI215L negatively regulates cGAS-STING signaling pathway through recruiting RNF138 to inhibit K63-linked ubiquitination of TBK1" *J Immunol*
38. Lu, Li, Zhou et al. (2023) "Fish herpesvirus KLP manipulates beclin1 to selectively degrade MITA through a precise autophagic manner for immune evasion" *Water Biology and Security*
39. Cui, Zhang, Zhou et al. (2025) "Cyprinid herpesvirus 2 (CyHV-2) ORF67 inhibits IFN expression by competitively obstructing STING phosphorylation" *Fish Shellfish Immunol*
40. Cook, Cristea (2019) "Location is everything: protein translocations as a viral infection strategy" *Curr Opin Chem Biol*
41. Wang, Bi, Chen et al. (2008) "ORF018R, a highly abundant virion protein from Singapore grouper iridovirus, is involved in serine/ threonine phosphorylation and virion assembly" *J Gen Virol*
42. Cheng, Kanyema, Sun et al. (2023) "African swine fever virus L83L negatively regulates the cGAS-STINGmediated IFN-I pathway by recruiting tollip to promote STING autophagic degradation" *J Virol*
43. Christensen, Jensen, Miettinen et al. (2016) "HSV-1 ICP27 targets the TBK1-activated STING signalsome to inhibit virus-induced type I IFN expression" *EMBO J*
44. Deng, Hu, Wang et al. (2022) "TRAF6 autophagic degradation by avibirnavirus VP3 inhibits antiviral innate immunity via blocking NFKB/NF-κB activation" *Autophagy*
45. Zhao, Hou, Zhang et al. (2024) "Encephalomyocarditis virus structural protein VP3 interacts with MAVS and promotes its autophagic degradation to interfere with the type I interferon signaling pathway" *Front Biosci (Landmark Ed)*
46. Wu, Hu, Song et al. (2023) "Lysine methyltransferase SMYD2 inhibits antiviral innate immunity by promoting IRF3 dephosphorylation" *Cell Death Dis*
47. Li, Fan, Zhu et al. (2023) "Tyrosine phosphorylation of IRF3 by BLK facilitates its sufficient activation and innate antiviral response" *PLoS Pathog*
48. Choi, Park, Kang et al. (2018) "Human cytomegalovirus-encoded US9 targets MAVS and STING signaling to evade type I interferon immune responses" *Nat Commun*
49. Zhang, Su, Zheng (2016) "Herpes simplex virus 1 serine protease VP24 blocks the DNA-sensing signal pathway by abrogating activation of interferon regulatory factor 3" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12483027&blobtype=pdf | # Seoul Virus Infection Acquired at Private Pet Rat Breeding Facility, Germany, 2024
Fabian Baalmann, Mario Hönemann, Stephan Drewes, Martin Eiden, Calvin Mehl, Dennis Tappe, Birte Pantenburg, Steffi Bellmann, Ingrid Möller, Melanie Maier, Corinna Pietsch, Rainer Ulrich, Johannes Münch
## Abstract
ence range <0.07 g/g]). Renal ultrasound showed enlarged, edematous kidneys (Appendix Figure 1, panel B), mild ascites, and splenomegaly. We observed no sonographic sign of chronic kidney disease or liver pathology. Given the deteriorating kidney function (peak serum creatinine was 14.3 mg/ dL [reference range 0.67-1.18 mg/dL]), we initiated acute hemodialysis and performed a kidney biopsy.Histopathologic examination revealed severe acute tubular injury with interstitial hemorrhage, edema, and inflammation (Figure 1,panel A). Only 2 of 21 glomeruli were sclerotic, and we observed minimal interstitial fibrosis with tubular atrophy. Specific immunohistochemical staining for hantavirus antigen was positive within the tubular cells and the tubulointerstitial space (Figure 1, panel B). Concurrently, lineblot assays performed by using re-
## DISPATCHES
H antaviruses, a diverse group of zoonotic pathogens carried primarily by rodents, can cause hemorrhagic fever with renal syndrome (HFRS) and hantavirus cardiopulmonary syndrome (1). In Europe, Puumala virus and Dobrava-Belgrade virus are the most prevalent human pathogenic hantaviruses (2). Seoul virus (SEOV) (species Orthohantavirus seoulense) is the only hantavirus with a worldwide distribution. Norway rats (Rattus norvegicus), black rats (R. rattus), and related species represent the main natural reservoir; however, pet rats recently have emerged as an important source of SEOV infections (3). The clinical spectrum of SEOV infection in humans ranges from mild febrile illness to HFRS-like disease that can include fever, hepatitis, and gastroenteritis, and, less commonly, kidney involvement (4,5). In contrast to its worldwide distribution, autochthonous SEOV infections in Germany remain rare; only a few cases have been reported to date (4,6).
## The Study
In March 2024, a previously healthy 44-year-old woman without any underlying conditions sought care at the nephrology clinic at University Hospital Leipzig (Leipzig, Germany) with a 6-day history of fever, fatigue, diarrhea, and hypotension. Five weeks before symptom onset, the patient visited a private pet rat breeding facility and subsequently acquired rats for her children. Initial laboratory workup revealed thrombocytopenia, mildly elevated transaminases, and acute kidney injury (Appendix Figure 1, panel A; https://wwwnc.cdc.gov/ EID/article/31/10/25-0362-App1.pdf). Urinalysis revealed hematuria and nephrotic-range proteinuria (urine protein-creatinine ratio >6 g/g [refer-comLine HantaPlus (Mikrogen Diagnostics, https:// www.mikrogen.de) of admission serum samples demonstrated positivity for hantavirus IgG and IgM, supporting the diagnosis of hantavirus-associated nephritis. Reactive bands included Hantaan virus antigen for IgG and Puumala, Hantaan, Dobrava-Belgrade, and Seoul virus antigens for IgM (Appendix Table 1). To further delineate the infecting hantavirus, we used a panhantavirus reverse transcription PCR targeting a 412-nt region of the viral large segment (7) on blood and urine samples. The subsequent sequence analysis and comparison to GenBank entries confirmed an infection with SEOV.
After 3 hemodialysis sessions (hospital days 5-7), the patient's clinical condition improved; we observed normalization of platelet counts by day 3, transaminase activities by day 6, and serum creatinine within 3 weeks. At 2-and 8-month follow-ups, serum creatinine had returned to normal (0.75 mg/dL), and proteinuria had resolved. Of note, the patient's 2 daughters and husband were serologically negative and did not show any symptoms related to a hantavirus infection throughout the initial treatment and follow-up period. We performed those serologic tests 2 weeks after the initial diagnosis of the patient for the 2 daughters and 2 months later for the husband.
Given the unexpected finding of SEOV as causative agent, we initiated an epidemiologic investigation in collaboration with the local public health and veterinary department, which led to the identification of the pet rat breeder. The breeding facility (≈30 m 3 in size) was located inside a private apartment, and the rats were bred with the intention of being pet animals. During the epidemiologic investigation (and 3 weeks after the patient tested positive for hantavirus IgG), the breeder, her husband, and her daughter also tested positive for hantavirus IgG but could not recall having symptoms suggestive of a hantavirus infection.
To assess the source of infection, we also analyzed samples from the index patient's rats and from the breeding facility. For all 4 of the patient's rats, we could not determine evidence of SEOV infection by serologic and molecular assays. In contrast, among the 6 pet rats obtained from the breeding facility, all yielded positive results across >1 assays (Appendix Table 2). One rat (source of sample KS24/530) was seropositive but negative by molecular testing, possibly reflecting a low viral load or virus clearance. Another rat (source of sample KS24/528) was SEOV RNA-positive but seronegative, suggesting an acute phase of infection. In general, results were concordant across 4 different assays performed on liver and lung tissues. In addition, we performed a phylogenetic analysis of a 731-nt long region of the viral small segment of the strains obtained from the breeder rats and publicly available SEOV sequences. All strains clustered closely together and shared a relatedness to sequences from other SEOV outbreaks in Germany but also to breeder rats or pet rats from other countries in Europe and the United States (Figure 2). Together with the high sequence similarity of the large segment sequences of 291 nt between the virus detected in the patient and those from the rats of the breeding facility (Appendix Figure 2), those findings strongly suggest the breeding facility as the source of SEOV infection. Given that the pet rats owned by the patient had no evidence of SEOV, the infection probably was acquired through aerosols during a visit to the breeding facility before purchase of the pet rats, followed by a 5-week incubation period until onset of disease symptoms.
## Conclusions
This report underscores several important observations. First, SEOV infection can lead to major clinical disease, including severe acute kidney injury requiring hemodialysis, even in the absence of hemorrhagic fever. Second, the proportion of asymptomatic infection in exposed persons suggests that SEOV infection might be underdiagnosed. Moreover, with the increasing popularity of pet rats, breeding facilities might serve as critical sources for SEOV, warranting enhanced surveillance. Proactive measures, such as routine screening of rat colonies and comprehensive public education, are warranted.
In Germany, practical implementation of a One Health approach for SEOV must consider the current lack of regulation for private breeders and the effect of animal welfare constraints, which preclude routine invasive testing of asymptomatic animals. Instead, feasible measures should focus on strict hygiene protocols, appropriate ventilation of animal housing, quarantine of newly acquired rats, and thorough documentation of animal origin and health status to facilitate traceability in case of outbreaks (9). Given that SEOV is primarily transmitted through inhalation of virus-contaminated aerosols from rodent excreta, risk reduction strategies should include avoiding dry cleaning methods, using damp cleaning techniques, and wearing protective masks during cage maintenance. Immunocompromised persons and other vulnerable groups should be explicitly advised against keeping pet rats. In addition, raising public awareness of zoonotic disease risks through veterinarians, breeders, pet stores, and public health agencies remains essential. Clinicians should routinely inquire about rodent exposures when evaluating patients with unexplained febrile illnesses, particularly when acute kidney injury has occurred (9).
Looking ahead, the development and provision of easy-to-use, noninvasive, and animal welfarecompatible diagnostic tools, such as point-of-care tests for SEOV RNA detection from environmental, oral, or fecal swabs, would offer veterinarians and breeders a practical means of monitoring colonies without causing undue stress to the animals. Such approaches could contribute substantially to early detection and prevention efforts, thereby strengthening One Health strategies in this context.
## References
1. Jonsson, Figueiredo, Vapalahti (2010) "A global perspective on hantavirus ecology, epidemiology, and disease" *Clin Microbiol Rev*
2. Kruger, Figueiredo, Song et al. (2015) "Hantaviruses-globally emerging pathogens" *J Clin Virol*
3. Cuperus, De Vries, Hoornweg et al. (2021) "Seoul virus in pet and feeder rats in the Netherlands" *Viruses*
4. Hofmann, Ulrich, Mehl et al. (2024) "Hantavirus disease cluster caused by Seoul virus" *Germany. Emerg Infect Dis*
5. Clement, Leduc, Lloyd et al. (2019) "Wild rats, laboratory rats, pet rats: global Seoul hantavirus disease revisited" *Viruses*
6. Hofmann, Heuser, Weiss et al. (2020) "Autochthonous ratborne Seoul virus infection in woman with acute kidney injury" *Emerg Infect Dis*
7. Johansson, Yap, Low et al. (2010) "Molecular characterization of two hantavirus strains from different Rattus species in Singapore" *Virol J*
8. Heuser, Drewes, Trimpert et al. (2023) "Pet rats as the likely reservoir for human Seoul orthohantavirus infection" *Viruses*
9. Haake, Eisenberg, Heuser et al. (2024) "Cuddle with care! A current overview of zoonotic pathogens transmitted by pet rats" *Berl Munch Tierarztl Wochenschr*
10. "Address for correspondence" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12697166&blobtype=pdf | # Complete genome sequence of Jobypre, an L3 subcluster mycobacteriophage isolated from marsh soil in Charleston, South Carolina
Riley Polic, Christa Joby, James Herring, Patrick Arnsberger, Ashanti Carter, Charlotte Drainville, Della Evans, Alyson Fincke, Calvin Geisel, Terranee Hines, Ryochi Jimenez, Joshua King, Malori Lesesne, Sophia Morrison, Emma Peluso, Vineel Prathipati, Cara Samson, Mattie Tharp, Mouna Dibenedetto, Christine Byrum, John Dennehy
## Abstract
Jobypre, an L3 subcluster mycobacteriophage (Family: Vilmaviridae) extracted from marsh soil at Joe Riley Waterfront Park in Charleston, South Carolina, has siphovirus morphology and infects Mycobacterium smegmatis mc 2 155. The Jobypre genome is 75,624 bp long and contains 128 protein-coding genes, 10 transfer RNA sequences, and no transfer-messenger RNA.
using Newbler v2.9 (6) and checked for accuracy and genome termini on Consed v29.0 (7).
Genome annotation was performed using the workflow tool PECAAN (8), and files were subsequently transferred to DNA Master v5.23.2 (https://phagesdb.org/DNAMas ter). Putative genes were identified by consensus with Glimmer v3.02 (9), Starterator v1.1 (10), Genemark v3.25 (11), Phamerator Actino_prophage v606 (12), ARAGORN v1.2.38 (13), and tRNAscan-SE v3.0 (14). Gene functions and domains were predicted by consensus using BLASTp v2.8.2+ (15), HHpred v2.1 (16), TMHMM Deep v1.0.42 (17), SOSUI v1.11 (18), and the NCBI Conserved Domain Database (CDD) (19). Default settings were applied to all programs, except as stated in https://seaphages.org/forums/topic/ 5398. The Jobypre genome is 75,624 bp with 59.3% guanine and cytosine (GC) content and contains 128 protein-coding sequences (48 with identified putative functions; two orphan genes, gp67 and gp118; 68 hypothetical proteins with homologs), 10 tRNA sequences, and no tmRNAs (Table 1). This phage is in the L cluster/L3 subcluster (cluster members share >50% nucleotide sequence identity; >35% gene content [GCS] similarity) (20,21), and the bioinformatic analysis revealed 3′ sticky overhangs at the termini (5′-TCGATCAGCC-3′). The presence of an immunity repressor (gp38 is homologous to proteins preventing phages with same immunity type from infecting the host), tyrosine integrase (gp36), and excise (gp40), as well as turbid plaques, suggest a temperate lifestyle. Whole-genome BLASTn comparison reveals that genomes among members of this subcluster (Family: Vilmaviridae) are highly conserved, and that three L3 viruses, namely, Snenia (GenBank KT281794.1), MsGreen (GenBank MK878900.1), and Lumos (GenBank KT372003.1), are nearly identical to Jobypre (>99.9% identity, 100% coverage). Interestingly, of the 24 L3 subcluster bacteriophages currently in the Actinobacterio phage Database (https://phagesdb.org/) (July 2025), all come from tropical or subtropical zones (southeastern US, Texas, South Africa, and Taiwan), except Whirlwind (Pennsylva nia; GenBank KF024725.1).
## References
1. Diacon, Guerrero-Bustamante, Rosenkranz et al. (2022) "Mycobacteriophages to treat tuberculosis: dream or delusion?" *Respiration*
2. Nick, Dedrick, Gray et al. (2022) "Host and pathogen response to bacteriophage engineered against Mycobacterium abscessus lung infection" *Cell*
3. Jordan, Burnett, Carson et al. (2014) "A broadly implementable research course in phage discovery and genomics for first-year undergraduate students" *mBio*
4. Poxleitner, Pope, Jacobs-Sera et al. (2018) "Phage discovery guide"
5. Russell (2018) "Sequencing, assembling, and finishing complete bacteriophage genomes" *Methods Mol Biol*
6. Margulies, Egholm, Altman et al. (2005) "Genome sequencing in microfabricated high-density picolitre reactors" *Nature*
7. Gordon, Green (2013) "Consed: a graphical editor for next-generation sequencing" *Bioinformatics*
8. Rinehart, Gaffney, Smithjr (2016) "PECAAN: phage evidence collection and annotation network user guide"
9. Delcher, Harmon, Kasif et al. (1999) "Improved microbial gene identification with GLIMMER" *Nucleic Acids Res*
10. Pacey (2016) "Starterator guide"
11. Lukashin, Borodovsky (1998) "GeneMark.hmm: new solutions for gene finding" *Nucleic Acids Res*
12. Cresawn, Bogel, Day et al. (2011) "Phamerator: a bioinformatic tool for comparative bacteriophage genomics" *BMC Bioinformatics*
13. Laslett, Canback (2004) "ARAGORN, a program to detect tRNA genes and tmRNA genes in nucleotide sequences" *Nucleic Acids Res*
14. Lowe, Eddy (1997) "tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence" *Nucleic Acids Res*
15. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol*
16. Söding, Biegert, Lupas (2005) "The HHpred interactive server for protein homology detection and structure prediction" *Nucleic Acids Res*
17. Hallgren, Tsirigos, Pedersen et al. (2022) "DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks" *bioRxiv*
18. Hirokawa, Boon-Chieng, Mitaku (1998) "SOSUI: classification and secondary structure prediction system for membrane proteins" *Bioinformatics*
19. Marchler-Bauer, Derbyshire, Gonzales et al. (2015) "CDD: NCBI's conserved domain database" *Nucleic Acids Res*
20. Hatfull, Jacobs-Sera, Lawrence et al. (2010) "Comparative genomic analysis of 60 mycobacteriophage genomes: genome clustering, gene acquisition, and gene size" *J Mol Biol*
21. Hatfull (2020) "Actinobacteriophages: genomics, dynamics, and applications" *Annu Rev Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12779694&blobtype=pdf | # Genome surveillance of SARS-CoV-2 Omicron XBB subvariants in Wuhan in June 2023
Yi Yan, Zhiyong Pan, Liangwei Mao, Peihua Niu, Fei Lu, Yirong Li, Zhiqiang Li
## Abstract
The COVID-19 pandemic caused by SARS-CoV-2 has had a significant impact on global social and economic activities. The continuous emergence of new SARS-CoV-2 variants has made the global pandemic of COVID-19 difficult to predict. Therefore, it is of great significance to closely monitor the SARS-CoV-2 epidemic and understand the evolution and transmission characteristics of SARS-CoV-2. In this study, we examined 36 cases of SARS-CoV-2 infection in Wuhan in June 2023. Genomic surveillance revealed that this outbreak was caused by Omicron XBB variants, among which FY.3 and its descendants FY.3.1 and FY.3.2 dominated, and this trend was consistent with the prevalence of SARS-CoV-2 nationwide in June and July. Analysis of sequence variations within hosts and between populations suggested that the S and ORF8 genes had undergone positive selection, possibly due to their important role in host adaptation, suggesting that some of these variations might be harmful mutations that reduce vaccine effectiveness. These findings may provide insights that will aid in predicting the evolutionary direction of SARS-CoV-2 in terms of variant competition, informing the design of next-generation multivalent vaccines and therapeutic strategies targeting conserved or rapidly evolving regions such as the S and ORF8 genes and supporting the evaluation and adjustment of surveillance, prevention, and control strategies in the Wuhan region.
## Introduction
The coronavirus disease 19 (COVID-19) pandemic caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has become a major threat to public health [1,2]. SARS-CoV-2 has accumulated numerous mutations, which have resulted in variants with increased transmissibility, replication efficiency, and immune evasion capability [3][4][5][6]. Some of the variants that have posed a higher publichealth risk were declared "variants of concern (VOC)" by the World Health Organization (WHO). So far, the VOCs include the Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), and Omicron (B.1.1.529) variants [7,8]. The latest variant, Omicron, which was identified in South Africa on November 24, 2021, was the most prevalent SARS-CoV-2 variant at the end of 2021, representing over 60% (9,682,591/16,129,598, as of January 29, 2024) of the complete SARS-CoV-2 genome sequences in the GISAID database [9]. Studies have suggested that it has an infection rate that is 2.7-3.7.7.7 times higher than that of the Delta variant [10].
The Omicron subvariant XBB is a recombinant virus derived from two BA.2 subvariants (BA.2.10.1 and BA.2.75) Yi Yan and Zhiyong Pan contributed equally to this work.
Handling Editor: Sheela Ramamoorthy Yirong Li liyirong838@163.com Zhiqiang Li lizhiqiang@whu.edu.cn that was first identified in mid-2022. Compared with earlier Omicron subvariants such as BA.1, BA.2, and BA.5, XBB exhibits enhanced immune evasion capability due to a unique combination of spike protein mutations, including R346T, F486S/P, and N460K. These mutations enable the virus to escape neutralizing antibodies elicited by prior infection or vaccination while maintaining efficient ACE2 receptor binding [11,12]. By early 2023, XBB-related variants became dominant in several countries and regions, including the United States and parts of Asia, including China.
After nearly three years of strict pandemic containment efforts, the Chinese government relaxed its control policies at the end of 2022 to adapt to the current international epidemiological landscape. Since the resulting infection peak, the positive rate of SARS-CoV-2 has remained relatively low (< 5%) across the country, with only sporadic cases reported. However, the Chinese Center for Disease Control and Prevention (CDC) reported that the positive rate of SARS-CoV-2 rebounded and reached a new peak in May, which might have been related to the International Workers' Day holiday. During that period, there was largescale population movement across the country. Wuhan, an important transportation hub, continues to hold significant public-health significance in the ongoing monitoring and management of the COVID-19 pandemic.
Despite their global prevalence, local transmission dynamics, mutational spectrum, and evolutionary features, XBB subvariants have remained insufficiently characterized in China, particularly during regional outbreaks that have occurred since the change in epidemic control policies. Therefore, the aim of this study was to address this gap through genomic surveillance of SARS-CoV-2 variants in Wuhan in June 2023. We conducted comprehensive analysis on the epidemic SARS-CoV-2 variants, genome-wide variation hotspots, and intra-host single-nucleotide variations (iSNVs). In addition to a detailed description of the characteristics of the recent epidemic of SARS-CoV-2 in Wuhan, this study provides data that may be useful for the formulation, evaluation, and adjustment of prevention and control strategies for respiratory pathogens in Wuhan.
## Materials and methods
## Ethics statement
This study and the use of all samples were approved by the Ethics Committee of Zhongnan Hospital (no. 2023111 K). Informed consent was obtained from patients and/or their legal guardians when samples were collected, and the participants were aware of the purpose, risks, and benefits of the study. Participants' names and other Health Insurance Portability and Accountability Act (HIPAA) identifiers have been omitted from all sections of this article, including the supplementary information. This study was performed in accordance with the Declaration of Helsinki.
## Patients and specimens
Clinical throat swab samples were collected in Zhongnan Hospital in Wuhan in June 2023. All specimens were obtained from patients who tested positive for SARS-CoV-2 during routine clinical diagnosis. No selection was made based on vaccination status, recovery stage (convalescence), or specific clinical manifestations. The swabs were immediately placed into sterile tubes containing 3 mL of inactivated viral transport medium (VTM) and sent to the detection laboratory of Primbio Genes Bio-Tech. Demographic and clinical data from enrolled patients were obtained from their electronic medical records.
## RNA extraction and qRT-PCR
Total viral genomic RNA was extracted from clinical specimens, using an Ex-DNA/RNA Virus Extraction Kit (T014, Xi'an Tianlong Science and Technology Co., Ltd, China). Briefly, nucleic acids were isolated from specimens, using lysis buffer containing a strong protein denaturant and proteinase K treatment. The mixture was then incubated with magnetic beads to purify the RNA. Using a Novel Coronavirus (2019-nCoV) Nucleic Acid Diagnostic Kit (Sansure Biotech Inc., Changsha, China), viral RNA obtained from clinical samples was detected using qRT-PCR targeting open reading frame 1ab (ORF1ab) and the nucleocapsid (N) protein gene of SARS-CoV-2. All experiments were performed in accordance with the standard protocols of the kit instruction manuals.
## Next-generation sequencing
A total of 38 SARS-CoV-2-positive samples (all of the positive samples available during the study period) were analyzed by meta-transcriptome sequencing, using a TruePrep RNA Library Prep Kit for Illumina (Vazyme, Nanjing, China) with some modified steps for library construction as described previously [13]. Sequencing was carried out on an MGISEQ platform with 150-bp paired-end raw reads after converting the sequencing library style to "MGI library".
## Viral genome assembly
Raw reads generated by next-generation sequencing (NGS) were quality-filtered using FastQC v0.12.1 and fastp 0.23.4 [14,15]. Megahit v1.2.9 was used for de novo assembly, and samples with poor assembly quality (coverage ≤ 90%) were 1 3 assembled with reference to the most closely related SARS-CoV-2 isolates identified by BLAST 2.12.0+ search using STAR 2.7.11a or Bowtie2 v2.5.0 [16][17][18]. bcftools 1.14 and bam-readcount were used to generate a file containing the base ratio of the mutation site [19]. Finally, the mutation sites were screened for an allele frequency (AF) ≥ 0.6 and depth >5 to generate a consensus viral genome sequence.
## Lineage division and phylogenetic analysis
To construct a representative dataset for phylogenetic analysis, we first used BLAST to identify the top 50 most similar SARS-CoV-2 genome sequences for each sample from the GISAID databases. To ensure temporal relevance, we retained genome sequences obtained within three months prior to our sampling period (i.e., March to June 2023). In total, 67 closely related sequences were analyzed together with sequences from this study. A multiple sequence alignment was performed using MAFFT v7.505 [20]. After selection of the best nucleotide substitution model using ModelFinder, a maximum-likelihood phylogenetic tree was constructed using IQ-Tree v2.2.2.5 with 1000 ultrafast bootstrap replicates [21][22][23]. SARS-CoV-2 lineages and variants were analyzed using Nextclade CLI (database updated on October 26, 2023).
## Single-nucleotide polymorphism (SNP) analysis
A Perl script ( h t t p s : / / g i t h u b . c o m / z e r 0 l i u / b i o u t i l s / t r e e / m a s t e r / s n p) was used to identify SNPs. The earliest genome sequence of an XBB strain (GISAID ID, EPI_ISL_402124) was used as a reference. SNP types, distributions, and densities were determined using the ggplot2 package v.
$$3.4.1 in R v.4.2.2.$$
## iSNV calling
Clean reads were compared to the consensus sequence for each sample, using Bowtie2 v2.5.0 with default parameters, and files containing base proportions at mutated sites were generated using bcftools 1.14 and bam-readcount. Finally, iSNVs were screened using the parameters 0.05 ≤ AF ≤ 0.5 and sequencing depth >5. A custom script was used to distingush iSNVs in viral genomes ( h t t p s : / / g i t h u b . c o m / z e r 0 l i u / b i o u t i l s / b l o b / m a s t e r / s n p / s t a t _ v a r _ c o d o n . p l).
## Results
## Clinical epidemiological characteristics
From June 8, 2023 to June 23, 2023, a total of 38 throat swab samples from SARS-CoV-2-positive patients were collected at Zhongnan Hospital, including 15 females and 23 males (Table 1). The age of the patients ranged from 1 month to 79 years, with a median age of 51.5 years. Of these patients, 5.26% were in the age group of 0-17 years, 71.05% were adults aged 18-65 years, and 23.68% were in the age group of 66-79 years (Table 1).
In this set of patients, clinical symptoms manifested as chest pain and cough, with a diagnosis of pulmonary infection accounting for 36.84% (14 out of 38 individuals). One patient, aged 79, experienced respiratory failure. The remaining cases presented varying degrees of complications (Supplementary Table S1). Given the wide age distribution of patients, it is plausible that age-related immune landscape differences may influence selective pressures on viral populations, potentially contributing to the emergence and maintenance of specific mutations within SARS-CoV-2 lineages.
## Quality of the sequence data
A total of 161.51 Gb of raw sequence data were obtained from 38 samples, with an average yield of 4.25 Gb (0.48-10.45.48.45 Gb, median 3.54 Gb). Of the 38 samples, 36 yielded high-quality sequence data, which were used for subsequent analysis. The average genome coverage was 99.69% (97-100%, median 99.83%), the average 5x coverage was 98.8% (90.25-100.25%, median 99.37%), and the average 10x coverage was 97.24% (76.64-100.64%, median 99.26%). The average sequencing depth was 1,026.72 (5.72-23.72,535.91, median 118.51) (Supplementary Table S1).
The statistics for the qPCR cycle threshold (Ct) values in this sample set ranged from 16.5 to 26.2 for the N gene, The 36 genome sequences from this study and 67 complete SARS-CoV-2 genome sequences obtained from the GISAID database were used for phylogenetic analysis. The phylogenetic results were consistent with those predicted using Nextclade (Fig. 2A, Supplementary Fig. S2). It is noteworthy that the phylogenetic tree revealed the presence of two groups of samples with strikingly similar infection times and evolutionary distances, indicating a potential occurrence of the same SARS-CoV-2 infection event in each group or successive transmission events (Fig. 2B andC).
with a median of 23.85 (Q1-Q3 = 21.98-24.78.98.78) and from 16.9 to 26.6 for the ORF gene, with a median of 24.1 (Q1-Q3 = 22.65-25.2.65.2) (Table 1). The quality of the SARS-CoV-2 genome sequences (depth) showed a negative correlation with Ct values (Supplementary Fig. S1).
## Genotypic variety and phylogeny of SARS-CoV-2 in the current epidemic
In order to investigate the genotypes associated with this epidemic, the genotypes of the viruses from this study were determined using Nextclade CLI. The results indicated that there were 17 different genotypes among the 36 isolates (Fig. 1). The majority belonged to Pango lineage FY. were synonymous (S) (Fig. 4D). Notably, all of the SNPs identified in ORF7b, ORF8, and ORF10 were NS, suggesting possible functional adaptations, especially given the role of ORF8 in immune modulation. The NS/S ratio approached 1 in ORF1a, was <1 in the ORF1b, ORF3a, membrane (M), and N genes, and >1 in other genes, indicating variable selective pressures across the genome (Fig. 3D). The five most frequent substitution types were A>G (26.18%), C>T (24.00%), T>C (18.55%), G>T (17.64%), and G>A (13.64%), primarily occurring in ORF1a (36.72%), ORF1b (16.91%), and the S gene (25.45%) (Fig. 3E; Supplementary Table S3). Transitions were more common than transversions, with an average Ti/Tv ratio of 1.52 across the genome, peaking in the N gene
## Genomic diversity of SARS-CoV-2 variants at the population level
Using the XBB strain as a reference (GISAID ID, EPI_ ISL_402124), a total of 641 SNP sites were identified among the 36 samples (Supplementary Table S2), each with 12-26 SNPs (average, 18). All of the mutations were located in the CDS region (Fig. 3A), and the majority occurred in ORF1a, the S gene, or ORF1b (Fig. 3B). After normalizing for gene length, the SNP density was the highest in the ORF8, S, and N genes, each exceeding 1 bp/kb (Fig. 3C). Of the SNPs, 361 (56.32%) were non-synonymous (NS), and 280 (43.68%) iSNVs (average, 10 iSNVs). All mutation sites were located within the CDS region, and most were in the ORF1ab, S, and ORF3a genes (Fig. 4A). The number of iSNVs was not related to the sequencing volume of the sample (Fig. 4B). Normalization of iSNVs within the SARS-CoV-2 genome according to the length of the ORF showed that the frequency of iSNVs in the ORF3a gene was the highest (0.9477 iSNVs/kb), followed by the ORF6 gene (0.7305 (Fig. 4E), which may reflect an underlying mutational bias and host-related RNA editing mechanisms.
## Genomic diversity of SARS-CoV-2 variants at the individual level
A total of 373 iSNVs were identified in 36 samples (Supplementary Table S4), with each sample possessing 0-39 are subject to adaptive evolution during replication and transmission. The S protein, particularly the receptorbinding domain, is a major target of neutralizing antibodies. Mutations in this region can reduce antibody binding affinity, facilitating immune escape and enabling viral persistence in vaccinated or previously infected individuals [26][27][28]. In ORF8, a G27915T mutation introduced a premature stop codon at the eighth codon (G8), truncating the polypeptide. ORF8 is known to mediate immune evasion by downregulating major histocompatibility complex class I (MHC-I), thereby interfering with cytotoxic T cell recognition [29]. Previous studies showed that a similar early stop codon at Q27 increased MHC-I expression, presumably due to loss of this immune modulatory function [29]. Therefore, the G27915T mutation may also alter the host immune response, although this hypothesis requires further experimental validation. These findings underscore the mutation tolerance and functional redundancy of the ORF8 gene, consistent with its rapid evolutionary trajectory [30].
Different viral genes exhibited different types of selection pressure. For example, the NS/S ratios of the iSNVs of ORF1ab and ORF3a within the host were 4.76 and 4.8 respectively, but the NS/S ratios of SNPs between populations were 0.80 and 0.29. This is consistent with previous findings that the selection of mutations for adaptive traits within hosts and across populations may be uncorrelated or even antagonistic [31]. In contrast, the ORF1b region showed low mutation density and low NS/S ratios, with very few SNPs or iSNVs detected. ORF1b encodes enzymes that are critical for viral replication and proofreading, including RNA-dependent RNA polymerase (nsp12), helicase (nsp13), and exonuclease (nsp14). These proteins are highly conserved and essential for viral fitness, and even minor changes can impair replication efficiency [32]. The purifying selection observed in ORF1b highlights its evolutionary constraints and genomic stability, as has been reported previously [33].
This study has some limitations, including a relatively small sample size and a narrow surveillance period, which might have introduced biases in the observed epidemiological and genomic characteristics and thus might not fully represent the SARS-CoV-2 transmission dynamics in the broader Wuhan population. Moreover, the short sampling window limits our ability to observe temporal changes in the prevalence of variants and the accumulation of mutations. As all of the samples were collected at a single tertiary hospital, regional and demographic diversity is likely to be underrepresented. Nevertheless, the study provides a valuable snapshot of the post-policy-relaxation viral landscape in Wuhan and provides insights into the local circulation and evolution of Omicron XBB subvariants. Future studies should incorporate larger and more-diverse cohorts, iSNVs/kb) and the S gene (0.5729 iSNVs/kb). Among all iSNVs, 307 (82.31%) were identified as non-synonymous (NS) substitutions, and 66 (17.69%) as synonymous (S) substitutions (Fig. 4C). In these XBB variants, the ORF7a and ORF8 iSNVs were non-synonymous. The NS/S ratios in the ORF1ab, S, ORF3a, and M genes were 4.76, 5.08, 4.8, and 4 respectively, and the NS/S ratios in the ORF6 and N genes were 1.5 and 1.6, respectively. The non-synonymous iSNVs occurred mainly in the first (127, 41.37%) and third (117, 38.11%) codon positions, and synonymous iSNVs were mainly located in the second codon position (56, 84.85%) (Fig. 4D, Supplementary Table S5).
To assess whether the age of the patient was associated with the rate of intra-host viral evolution, we examined the correlation between patient age and mutation metrics including the total number of SNPs and iSNVs, as well as their respective NS/S ratios. No significant correlation was observed between age and any mutation metric, with R 2 values ranging from 7e-5 to 0.0048 (SNP: 0.0017-0.0228.0017.0228). This suggests that, in our cohort, age was not a major determinant of SARS-CoV-2 mutation burden or selective pressure.
$$Nonsynonymous 0 20 40 60 80 O R F 1 a b S O R F 3 a M O R F 6 O R F 7 a O R F 8 N O R F 1 a b S O R F 3 a M O R F 6 O R F 7 a O R F8$$
## Discussion
Currently, most Asian countries, including China, have entered a new stage of epidemic prevention and control against SASR-CoV-2, and people's lives are gradually returning to normal. However, the continual emergence of new variants still poses a substantial threat to global health. In this study, we performed genomic surveillance of 36 SARS-CoV-2 samples from Wuhan in June 2023 and found that Omicron XBB subvariants -especially FY.3 and its descendants -were the dominant circulating variants. This pattern was consistent with national surveillance data and reflects the evolutionary advantage of XBB variants under postrelaxation transmission dynamics [24]. All of these variants belong to the BA.2 subtype. Recent studies have also highlighted the transmissibility of SARS-CoV-2 strains, revealing that the original strain had a basic reproduction number (R0) of 2-3, the Delta strain had an R0 of 6-7, the Omicron BA.1 variant had an R0 of 7-8, and the BA.2 variant had an R0 as high as 9.1 [25]. In this case, long-term epidemiological surveillance is critical for early detection, decision-making, and response to the threat of Omicron variants.
It is worth noting that, at both the population level and individual level, the S gene and ORF8 exhibited consistent signals of strong selection pressure. The S gene showed a high density of SNPs and iSNVs as well as elevated NS/S ratios, while all of the observed mutations in ORF8 were non-synonymous. This pattern suggests that both genes 1 3 extend sampling over longer time periods, and, ideally, involve multi-center surveillance to assess viral dynamics and lineage shifts. Integration with clinical, immunological, and vaccination data would provide important information about the effect of individual variations.
The emergence of new SARS-CoV-2 variants and successive waves of epidemics have posed serious challenges to global public health. This study investigating the unique mutational characteristics of the XBB variant in Wuhan in June 2023 revealed strong selective pressure acting on the S and ORF8 genes. Given the high population density and mobility in Wuhan, continuous genomic surveillance focusing on these key genes may provide early warnings of fitness-enhancing mutations. These findings support the prioritization of S and ORF8 as molecular targets for future variant monitoring and risk assessment. Strengthening region-specific surveillance and integrating genomic data into real-time public-health decision-making will be critical for responding effectively to future outbreaks.
## References
1. Menon, Mohapatra (2022) "The COVID-19 pandemic: virus transmission and risk assessment" *Curr Opin Environ Sci Health*
2. Petersen, Koopmans, Go et al. (2020) "Comparing SARS-CoV-2 with SARS-CoV and influenza pandemics" *Lancet Infect Dis*
3. Chen, Zhang, Case et al. (2021) "Resistance of SARS-CoV-2 variants to neutralization by monoclonal and serum-derived polyclonal antibodies" *Nat Med*
4. Liu, Liu, Plante et al. (2022) "The N501Y spike substitution enhances SARS-CoV-2 infection and transmission" *Nature*
5. Plante, Liu, Liu et al. (2021) "Spike mutation D614G alters SARS-CoV-2 fitness" *Nature*
6. Hou, Chiba, Halfmann et al. (2020) "SARS-CoV-2 D614G variant exhibits efficient replication ex vivo and transmission in vivo" *Science*
7. Salehi-Vaziri, Fazlalipour, Khorrami et al. (2022) "The ins and outs of SARS-CoV-2 variants of concern (VOCs)" *Arch Virol*
8. Khandia, Singhal, Alqahtani et al. (2022) "Emergence of SARS-CoV-2 Omicron (B.1.1.529) variant, salient features, high global health concerns and strategies to counter it amid ongoing COVID-19 pandemic" *Environ Res*
9. Hoang, Chernomor, Haeseler et al. (2018) "UFBoot2: improving the ultrafast bootstrap approximation" *Mol Biol Evol*
10. Li, Hu, Qu et al. (2024) "Molecular epidemiology and population immunity of SARS-CoV-2 in Guangdong (2022-2023) following a pivotal shift in the pandemic" *Nat Commun*
11. Tao, Tzou, Nouhin et al. (2021) "The biological and clinical significance of emerging SARS-CoV-2 variants" *Nat Rev Genet*
12. Tye, Jinks, Haigh et al. (2022) "Mutations in SARS-CoV-2 spike protein impair epitope-specific CD4(+) T cell recognition" *Nat Immunol*
13. Starr, Greaney, Hilton et al. (2020) "Deep mutational scanning of SARS-CoV-2 receptor binding domain reveals constraints on folding and ACE2 binding" *Cell*
14. Bhattacharya, Sharma, Dhama et al. (2022) "Omicron variant (B.1.1.529) of SARS-CoV-2: understanding mutations in the genome, S-glycoprotein, and antibody-binding regions" *Geroscience*
15. Zhang, Chen, Li et al. (2021) "The ORF8 protein of SARS-CoV-2 mediates immune evasion through down-regulating MHC-I" *Proc Natl Acad Sci U S A*
16. Du, Ding, Li et al. (2020) "Genomic surveillance of COVID-19 cases in Beijing" *Nat Commun*
17. Hou, Shi, Gong et al. (2023) "Intra-vs. interhost evolution of SARS-CoV-2 driven by uncorrelated selection-the evolution thwarted" *Mol Biol Evol*
18. Brant, Tian, Majerciak et al. (2021) "SARS-CoV-2: from its discovery to genome structure, transcription, and replication" *Cell Biosci*
19. Zhao, Hall, De Cesare et al. (1987) "The mutational spectrum of SARS-CoV-2 genomic and antigenomic RNA" *Proc Biol Sci*
20. "Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations"
21. Jung, Kmiec, Koepke et al. (2022) "Omicron-what makes the latest SARS-CoV-2 variant of concern so concerning?" *J Virol*
22. Espenhain, Funk, Overvad et al. (2021) "Epidemiological characterisation of the first 785 SARS-CoV-2 Omicron variant cases in Denmark" *Euro Surveill*
23. Wang, Iketani, Li et al. (2023) "Alarming antibody evasion properties of rising SARS-CoV-2 BQ and XBB subvariants" *Cell*
24. Kurhade, Zou, Xia et al. (2023) "Low neutralization of SARS-CoV-2 Omicron BA.2.75.2, BQ.1.1 and XBB.1 by parental mRNA vaccine or a BA.5 bivalent booster" *Nat Med*
25. Lu, Yan, Dong et al. (2021) "Integrated characterization of SARS-CoV-2 genome, microbiome, antibiotic resistance and host response from single throat swabs" *Cell Discov*
26. Chen, Zhou, Chen et al. (2018) "fastp: an ultra-fast all-inone FASTQ preprocessor" *Bioinformatics*
27. Brown, Pirrung, Mccue (2017) "FQC Dashboard: integrates FastQC results into a web-based, interactive, and extensible FASTQ quality control tool" *Bioinformatics*
28. 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*
29. Ye, Mcginnis, Madden (2006) "BLAST: improvements for better sequence analysis" *Nucleic Acids Res*
30. Danecek, Bonfield, Liddle et al. (2021) "Twelve years of SAMtools and BCFtools"
31. Langdon (2015) "Performance of genetic programming optimised Bowtie2 on genome comparison and analytic testing (GCAT) benchmarks" *BioData Min*
32. Nakamura, Yamada, Tomii et al. (2018) "Parallelization of MAFFT for large-scale multiple sequence alignments" *Bioinformatics*
33. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: fast model selection for accurate phylogenetic estimates" *Nat Methods*
34. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12792179&blobtype=pdf | # P-1800. Saliva as a Promising Additional Specimen for Enhanced RSV Detection in Pediatric Patients; A New Vaccine Surveillance Network Study
Anjana Sasidharan, Montserrat Santos, Varun Chandra Boinpelly, ; Dithi Banerjee, Brian Lee, Jennifer Schuster, Dinah Dosdos, Mary Moffatt, Kirsten Weltmer, Gina Weddle, Casey Kalman, Heidi Moline, Rangaraj Selvarangan
Background. Respiratory syncytial virus (RSV) detection is essential for pediatric acute clinical care. While nasopharyngeal (NPS) and mid-turbinate swabs (MTS) remain the gold standard, they may be distressing to children. Saliva has shown promise as a diagnostic specimen in adults for RSV, but its diagnostic utility in children is unknown. In this study, we evaluated the performance of saliva compared to paired MTS samples for RSV detection in children. Results. A total of 786 paired MTS and saliva samples were collected. Among the 468 (59.5%) paired MTS and saliva specimens with sufficient volume, 154 (33%) were positive for RSV in MTS or saliva: 71 (15%) MTS specimens, 83 (18%) saliva specimens. Median (IQR) Ct values were 25.6 (21.6 -30.9) in MTS and 29.1 (26.0 -33.1) in saliva. Saliva missed 7 (1.5%) cases detected by MTS (median MTS Ct: 31.3), while detecting 19 (4.1%) additional cases (median saliva Ct: 35.2) suggesting viral persistence in oral secretions (Table 1.). Of these, 13 cases were RSV-B and 6 were RSV-A. Saliva increased diagnostic yield by 20% compared to MTS. Positive and negative percent agreement between both sample types were 90% and 95%, respectively.
Demographic and clinical symptoms were broadly comparable, although cough was less common in the saliva-only group (84.2% vs 98.4%, p=0.053). Co-infection was lower in saliva-only group (5.3% vs 18.8%, p=0.289), though not statistically different (Table 2).
Conclusion. Saliva showed greater positive percent agreement for RSV detection compared to MTS samples, particularly in cases with higher Ct values suggestive of lower viral loads. These findings support saliva as a complementary specimen, reinforced by the observed 20% increase in diagnostic yield.
Disclosures. Brian R. Lee, PhD, MPH, Merck: Grant/Research Support Rangaraj Selvarangan, PhD, Altona: Grant/Research Support|Biomerieux: Advisor/Consultant| Biomerieux: Grant/Research Support|Biomerieux: Honoraria|Cepheid: Grant/ Research Support|Hologic: Grant/Research Support|Hologic: Honoraria|Meridian: Grant/Research Support|Qiagen: Grant/Research Support |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12731289&blobtype=pdf | # A case of Escherichia coli native mitral valve endocarditis complicated by annular abscess and central retinal artery occlusion
Amgad Ghoprial, Nazary Nebeluk, Uzoamaka Eke
## Abstract
Infective endocarditis (IE) is most frequently caused by gram-positive organisms such as Staphylococci, Streptococci, and Enterococci. IE due to gram-negative organisms is uncommon and is most often associated with the HACEK group (Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, and Kingella) of organisms. IE due to other gram-negative organisms such as Escherichia coli (E. coli) is rare and less is known about its natural pathology. Valvular abscesses due to E. coli IE have been described, however these are not usually reported in a native mitral valve. Here we describe the case of a 69 year old female who initially presented with pyelonephritis and was found to have E. coli IE of the posterior mitral valve leaflet associated with a large annular abscess, which ultimately led to septic emboli causing stroke and central retinal artery occlusion (CRAO), an extremely rare sequela of IE. Definitive diagnosis and mitral valve replacement were delayed by a low initial suspicion for IE, as well as complications including pulmonary decompensation and incidental discovery of a mycotic aneurysm of the left anterior cerebral artery.
## Introduction
Infective endocarditis (IE) is an infection of the endocardium, most commonly involving a heart valve. The vast majority of cases are caused by gram-positive cocci such as Staphylococci, Streptococci, and Enterococci. Discussion of gram-negative bacterial etiologies tend to focus on the HACEK group (Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, and Kingella) which are known to cause IE albeit at an overall low epidemiological rate. Non-HACEK gram-negative organisms of significant clinical burden in other infectious syndromes such as Escherichia coli (E. coli) are rarely identified as the cause of IE although there are estimates they may in fact account for 0.5 % of all cases [1]. There have been reports of cardiac valvular abscesses in E. coli endocarditis, however these are usually associated with prosthetic valves, and most often occur in the aortic valve, even though the valve most frequently affected by E. coli endocarditis is the mitral valve in both native and prosthetic valve IE [2]. In this article, we describe a case of native mitral valve endocarditis secondary to E. coli bacteremia complicated by annular abscess development and significant embolic complications, including central retinal artery occlusion (CRAO). CRAO is an exceedingly uncommon sequela of infective endocarditis with only 10 cases noted in the literature from any pathogen, and no cases attributed to E. coli. This case highlights the need to maintain E. coli on the differential for infective endocarditis even without persistent bacteremia to prevent patient morbidity.
## Case
A 69-year-old white female with no pertinent past medical history presented to an outside emergency department with a chief complaint of fevers, chills, and lower back pain for approximately 2 weeks. She reported an abnormal urinalysis from her primary care provider, for which she completed a 5-day course of nitrofurantoin the morning of presentation without symptomatic improvement. On presentation, she was afebrile and hemodynamically stable on room air. Physical exam revealed only mild right costovertebral angle tenderness. Laboratory studies were significant for leukocytosis (51,000 white blood cells/µL) with neutrophilic predominance, creatinine elevation (1.2 mg/dL) from previous baseline (0.6 mg/dL), and moderate pyuria (11-25 white blood cells/high power field). Urine and blood cultures were collected. A computed tomography (CT) scan of the abdomen and pelvis with intravenous contrast showed a diffusely heterogenous right kidney concerning for pyelonephritis, an unremarkable left kidney, and no evidence of inflammation in the bladder or ureters.
She was started on cefepime and received one dose of intravenous vancomycin. A few hours after admission, she became hypoxic (76 % O 2 saturation) and tachycardic (146 beats per minute). Her hypoxia resolved with nasal cannula. Chest X-ray showed mildly reduced lung volumes with scattered patchy parenchymal opacities. Additional labs were drawn, showing elevated lactate (7.6 mmol/L), D-dimer (31.54 mg/L fibrinogen equivalent units), and N-terminal prohormone of brain natriuretic peptide (30.168 pg/mL). The next day, she developed altered mental status with labile blood pressures and worsening hypoxia requiring non-rebreather, and was upgraded to the intensive care unit (ICU). A non-contrast CT scan of the chest demonstrated patchy central interstitial thickening and ground glass opacities. An expanded respiratory viral pathogen multiplex polymerase chain reaction panel was negative, and expectorated sputum cultures showed no growth. Ultimately the new symptoms were attributed to nitrofurantoin related pulmonary toxicity, for which she was treated with IV methylprednisolone. Blood cultures collected in the ED returned positive for E. Coli in 4/4 bottles, and the urine culture resulted without growth. Repeat blood cultures two days later were negative. Final sensitivities returned on hospital day 4, and the patient was de-escalated to ceftriaxone.
To evaluate for a possible heart failure exacerbation, a transthoracic echocardiogram (TTE) was obtained. This demonstrated a mobile echodensity on the posterior leaflet of the mitral valve (Fig. 1) suspicious for multiple mitral annular calcifications (MAC) with trace mitral regurgitation, and a normal left ventricular ejection fraction without right ventricular strain. Transesophageal echocardiogram (TEE) was discussed but ultimately deferred due to the patient's pulmonary condition and the high risk of intubation. She was transferred out of the ICU on hospital day 15 on nasal cannula, and completed her 14-day course of ceftriaxone for pyelonephritis on hospital day 18. The patient remained admitted to wean her oxygen requirements to her baseline of room air.
On hospital day 20, a stroke alert was called for "shadowy vision that started at lunch time" per the patient. Ophthalmology evaluated the patient and diagnosed a left sided CRAO. CT scan of the head and neck with contrast was negative for acute stroke, but did incidentally note a 4 × 4 mm saccular aneurysm of the left anterior cerebral artery (A1 segment). Magnetic Resonance Imaging (MRI) of the brain showed extensive small infarcts in the posterior fossa and supratentorial region concerning for septic emboli. The infectious disease service was reengaged, and new blood cultures were obtained which did not show growth. TEE was performed on day 22, which demonstrated an independently mobile 1.5 × 1.5 cm mass on the atrial aspect of the posterior leaflet of the mitral valve suspicious for infective endocarditis versus caseous MAC (Fig. 2) as well as mild mitral regurgitation. Ceftriaxone was restarted, and two days later the patient was transferred to our tertiary care center for assessment by cardiothoracic surgery. The patient was deemed an appropriate surgical candidate because of the persistent vegetation noted on repeat TEE, evidence of septic emboli on head MRI, and progression of disease despite prior antibiotic therapy.
Neurosurgery clipped the patient's saccular aneurysm, delaying her cardiothoracic surgery by one week so that she could be safely heparinized. Finally, on hospital day 39, she underwent mitral valve replacement with a bovine bioprosthetic valve. Intraoperatively, a large posterior annular abscess and extensive involvement of the posterior leaflets with vegetations were noted and this tissue was all excised. The valve tissue culture grew E. coli 3 days after surgery, and it was sensitive to most antibiotics tested, including ceftriaxone. The patient was discharged on a 6-week course of ceftriaxone daily and aside from residual left sided vision loss did not have sequalae of disease or recurrent signs of infection at multiple subsequent outpatient visits.
## Discussion
This is a rare case of a native mitral valve IE due to E. coli pyelonephritis which progressed to the development of an annular abscess despite two weeks of appropriate antibiotic therapy and resulted in embolic sequelae including CRAO and multiple strokes albeit without significant residual motor deficits. This patient's case is consistent with what is described in the literaturethe median age of patients with IE due to E. Coli is 59.6 years, urinary tract infection as the preceding event was described in 52 % of patients and the mitral valve was the most commonly affected [2]. Approximately 68 % of patients with native valve E. coli IE did not have any underlying valvular disease, a higher proportion than in bacterial IE overall where only 43 % of patients did not have underlying disease [2,3]. Embolic complications from E. coli IE appear to be less frequent with only 24 % of patients suffering an embolic complication compared to 39.5 % of all bacterial IE [2,4]. The rate of progression to valvular abscess appears comparable between the two groups (18 % for E. coli IE versus 14.4 % for all bacterial IE) [2,4]. Up to 25 % of patients with IE experience some form of neurologic complication. In general, these patients are less likely to have had a cardiac surgical intervention, and when done, it is unclear whether early or later intervention is more beneficial to mortality [5]. However, given the concern for morbidity from embolization, recent guidelines tend to favor early surgery, but the decision should be individualized to each patient's clinical presentation and co-morbid risk factors [6][7][8].
CRAO in older patients is typically associated with vascular factors such as hypertension and hyperlipidemia, embolic factors such as atrial fibrillation, and neurologic factors such as migraine [6]. Out of these risk factors, our patient only had dyslipidemia, so it is unclear if she had any other underlying predisposition for CRAO. CRAO is an extremely rare sequela of IE, with only 10 case reports identified on literature review. Of these, four were due to Streptococcal species [9][10][11][12], two were due to Bartonella henselae [13,14], two were in the setting of culture negative endocarditis [15,16] and one was attributed to Staphylococcus aureus [17] and Erysipelothrix rhusiopathiae respectively [18]. The cases reviewed were in patients with ages ranging from 14 to 81 and did not have any significant commonalities observed. However, risk factors for embolization in IE have been described which include location on the mitral valve (as opposed to the aortic valve), and large size (>10 mm), which were both characteristics of this case [19]. This case appears to be the first reported instance of CRAO as a sequela of E. coli IE.
Prompt diagnosis of IE is crucial as early surgical intervention has been shown to have a significant mortality benefit with a recent metaanalysis showing an almost 5-fold reduction in mortality risk in patients who received surgery when compared to conservative treatment [20]. Indications for surgical management of IE include new onset heart failure, evidence of uncontrolled infection such as abscess formation, enlarging vegetation or persistent bacteremia, and established embolism or high risk of embolization due to a vegetation > 10 mm [21]. Surgical intervention also provides a morbidity benefit as it helps prevent systemic embolization from the IE. In a randomized control trial, patients assigned to an early surgery group had no embolic events at up to six months of follow up compared to 21 % of patients in the conventional treatment group [22]. This data highlights the need for early TEE to both make the diagnosis and risk stratify as the current American Heart Association/American College of Cardiology (AHA/ACC) guidelines recommend surgical valve replacement for lesions > 10 mm in patients with a history of embolic events and > 15 mm in patients without [4].
The definitive diagnosis of IE in this case was delayed due to the deferral of TEE to the outpatient setting. This decision was made due to concern for periprocedural aspiration pneumonitis further complicating the patient's pulmonary compromise and inducing an episode of acute hypoxic respiratory failure requiring intubation. The low clinical suspicion for IE, based on negative repeat blood cultures on day two of admission and underecognition of E. coli as a causative agent, also contributed to the decision. However, viable E. coli was later recovered from the operative cultures on hospitalization day 39 indicating ongoing infection that had evaded detection by standard blood cultures. In similar cases, newer diagnostic techniques, such as microbial cell free DNA (mcfDNA) metagenomic sequencing, could play a complementary role in raising an earlier clinical suspicion for IE. While mcfDNA metagenomic sequencing does not confirm IE it does detect bacterial DNA in the bloodstream, even after up to 30 days of antibiotic therapy [23]. If E. coli DNA had been recognized in the patient's bloodstream within the first week of hospitalization, despite negative repeat cultures, this could have prompted an inpatient TEE given the patient's initial TTE findings pointed to a vegetation as the most likely source. Thus, mcfDNA metagenomic sequencing could serve as a complementary non-invasive diagnostic strategy helping to identify persistent infection and guide further workup in culture negative patients. The patient's pulmonary compromise was attributed to the course of nitrofurantoin she had taken prior to admission, however respiratory toxicity is primarily associated with prolonged courses of therapy [24], so her short course seems unlikely to have been the trigger. Similarly, her endocarditis was unlikely to have caused the pulmonary symptoms as she never had symptoms of valve or heart failure that would have precipitated pulmonary edema.
Granted no additional likely etiology was identified during her hospitalization or subsequent review of her case.
## Conclusion
This case highlights a 69-year-old woman who presented for treatment of pyelonephritis and had a series of complications including acute respiratory failure, infective endocarditis, central retinal artery occlusion, and mycotic aneurysm that together led to a 51-day long hospitalization. While mortality was thankfully avoided, earlier diagnosis and intervention may have prevented the severe morbidity due to the sequelae of IE in this case. In patients with a visible vegetation on TTE and cultures positive for E. coli, clinical suspicion should be maintained, and a TEE should be pursued to evaluate for IE. In cases where TEE may not be safe or feasible, alternative diagnostics such as mcfDNA metagenomic sequencing could be considered or empiric therapy discussed with the patient in a shared decision-making context.
## CRediT authorship contribution statement
Uzoamaka Akudo Eke: Writingreview & editing, Writingoriginal draft, Supervision. Nazary Nebeluk: Writingreview & editing, Writingoriginal draft, Project administration, Conceptualization. Amgad Ghoprial: Writingreview & editing, Writingoriginal draft, Investigation.
## Consent
Written informed consent was obtained from the patient for publication of this case report and accompanying images. A copy of the written consent is available for review by the Editor-in-Chief of this journal on request.
## Ethical approval
Consent: Authors received written inform consent from patient and a copy is available for review upon request
## References
1. Chavada, Anwar, Dumont et al. (2023) "Escherichia coli endocarditis and cardiac abscess: a rare presentation in a patient with a prosthetic aortic valve" *Eur J Case Rep Intern Med*
2. Akuzawa, Kurabayashi (2018) "Native valve endocarditis due to Escherichia coli infection: a case report and review of the literature" *BMC Cardiovasc Disord*
3. Castillo, Anguita, Castillo et al. (2015) "Changes in clinical profile, epidemiology and prognosis of left-sided native-valve infective endocarditis without predisposing heart conditions" *Rev Española De Cardiolía (Engl Ed)*
4. Otto (2020) "ACC/AHA guideline for the management of patients with valvular heart disease" *J Thorac Cardiovasc Surg*
5. García-Cabrera (2013) "Neurological complications of infective endocarditis" *Circulation*
6. Yanagawa (2016) "Surgical management of infective endocarditis complicated by embolic stroke" *Circulation*
7. Barsic (2013) "Influence of the timing of cardiac surgery on the outcome of patients with infective endocarditis and stroke" *Clin Infect Dis*
8. Okita (2016) "Optimal timing of surgery for active infective endocarditis with cerebral complications: a Japanese multicentre study" *Eur J CardioThorac Surg*
9. Kato, Takeda, Matsuyama (2001) "Combined Occlusion of the Central Retinal Artery and Vein in a Pediatric Patient Secondary to Infective Endocarditis"
10. Serras-Pereira, Fernandes, Azevedo et al. (2020) "Central retinal artery occlusion from Streptococcus gallolyticus endocarditis" *BMJ Case Rep*
11. Flores, Blanco, Rivas et al. (2015) "Oclusión de la arteria central de la retina y endocarditis infecciosa: el rigor sí importa" *Arch Soc Esp Oftalmol*
12. Chawla, Goldblatt, Morgan et al. (2023) "Central retinal artery occlusion with concomitant intracranial hemorrhage secondary to streptococcus gordonii endocarditis" *Case Rep Ophthalmol Med*
13. Woo, Ahn, Song et al. (2018) "A case of retinal vessel occlusion caused by Bartonella infection" *Korean Acad Med Sci*
14. Chebolu, Wallsh, Falk et al. (2023) "Central retinal artery occlusion as presentation of bartonella endocarditis"
15. Ammous, Braham, El Amri Mezghanni et al. (2018) "Central retinal artery occlusion revealing an infectious endocarditis"
16. Ziakas, Kotsidis, Ziakas (2014) "Central retinal artery occlusion due to infective endocarditis" *Int Ophthalmol*
17. (2025) "From heart to eye: central retinal artery occlusion secondary to endocarditis: septic embolism" *JACC Case Rep*
18. Urzedo, Predabon, Filho et al. (2025) "Bilateral retinal artery occlusion as the initial presentation of infectious endocarditis: a case report" *J Med Case Rep*
19. Yang (2019) "Clinical and echocardiographic predictors of embolism in infective endocarditis: systematic review and meta-analysis" *Clin Microbiol Infect*
20. Caldonazo (2024) "Conservative versus surgical therapy in patients with infective endocarditis and surgical indication-meta-analysis of reconstructed time-to-event data" *J Am Heart Assoc*
21. Delgado (2023) "2023 ESC Guidelines for the management of endocarditis" *Eur Heart J*
22. Kang (2012) "Early Surgery versus conventional treatment for infective endocarditis" *N Engl J Med*
23. Eichenberger (2023) "Microbial cell-free DNA identifies the causative pathogen in infective endocarditis and remains detectable longer than conventional blood culture in patients with prior antibiotic therapy" *Clin Infect Dis*
24. Santos, Batech, Pelter et al. (2016) "Evaluation of the risk of nitrofurantoin lung injury and its efficacy in diminished kidney function in older adults in a large integrated healthcare system: a matched cohort study" *J Am Geriatr Soc* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12813283&blobtype=pdf | # In vitro activity of cefepime/zidebactam against sulbactam/durlobactam-susceptible and -resistant Acinetobacter baumannii clinical isolates
Carlo Tascini, Gabriele Bianco, Robert Bonomo, Paolo Gaibani, J Chemother
## Abstract
Objectives: To evaluate the in vitro activity of cefepime in association with a β-lactamase inhibitor (enmetazobactam) or β-lactam enhancer (BLE), zidebactam, against carbapenem-resistant Acinetobacter baumannii (CRAB) strains susceptible or resistant to sulbactam/durlobactam.Material and methods: Twenty-one CRAB clinical isolates were characterized by WGS and AST to cefepime/ enmetazobactam, cefepime/zidebactam and comparators was determined.Results: Resistome analysis revealed that all CRAB carried bla OXA-23 carbapenemase genes, while bla ADC-25 and bla OXA-66 β-lactamase genes were observed exclusively in sulbactam/durlobactam-resistant strains. Analysis of penicillin-binding protein genes demonstrated the presence of specific mutations within PBP3 (N392T) previously associated to the resistance to sulbactam/durlobactam. Phenotypic analysis revealed that cefepime/ enmetazobactam did not exert antibacterial activity against CRAB, while cefepime/zidebactam displayed potent bactericidal activity against both sulbactam/durlobactam-susceptible and -resistant strains.Conclusions: These results demonstrate that cefepime/zidebactam exhibited potent in vitro antibacterial activity against CRAB producing OXA carbapenemase and support its clinical use against both sulbactam/ durlobactam-susceptible or -resistant isolates.
## Introduction
Acinetobacter baumannii is an opportunistic Gram-negative pathogen and a leading cause of hospital-acquired infections, particularly in critically ill and immunocompromised patients. Historically considered 'low virulence', resistant strains are now associated with mortality rates up to 70%. 1,2 The bacterium rapidly acquires antimicrobial resistance via mobile genetic elements, upregulation of efflux pumps, β-lactamases, aminoglycoside-modifying enzymes and target gene mutations. 3 The global emergence of MDR, XDR and PDR strains represents a major clinical challenge. 3,4 Carbapenem-resistant A. baumannii (CRAB) is frequently cross-resistant to last-line agents such as colistin and tigecycline, and has been classified by the WHO as an urgent public health threat, highlighting the need for novel therapeutic options. 5 Sulbactam/durlobactam is a recently approved combination for CRAB infections. Sulbactam, a β-lactamase inhibitor with intrinsic antibacterial activity, binds penicillin-binding proteins, whereas durlobactam, a novel diazabicyclooctane (DBO) β-lactamase inhibitor similar to avibactam, protects sulbactam from degradation by serine β-lactamases, particularly OXA-type enzymes. 6,7 This combination demonstrated clinical efficacy in HABP and VABP in the ATTACK trial and is currently recommended by the Infectious Diseases Society of America Guidance as a preferred option for CRAB infections, often in combination with a carbapenem. 7,8 Resistance has been reported via PBP mutations and metallo-β-lactamase expression. 9 Among recently developed agents are two cefepime-based β-lactam/β-lactamase inhibitor combinations. Cefepime, a fourth-generation cephalosporin, is bactericidal through PBP inhibition and resistant to most AmpC β-lactamases but hydrolysed by ESBLs. Cefepime/enmetazobactam combines cefepime with enmetazobactam, a β-lactamase inhibitor similar to tazobactam that protects cefepime from ESBL-mediated degradation. 10,11 Cefepime/zidebactam is another novel combination, where zidebactam binds PBP2 and cefepime targets PBP3, producing rapid bactericidal activity, even at sub-MIC levels, independently of β-lactamase expression. 12 Zidebactam exerts its activity by dual mechanism of action by enhancing the activity of the β-lactams partner. 12 In detail, zidebactam inhibits class A, C and some D B-lactamase, and directly binds to Penicillin-Binding Protein 2 (PBP2) in Gram-negative bacteria, thus enhancing the killing of Enterobacterales and Pseudomonas aeruginosa and producing KPC or MBL when combined with cefepime. 13 However, a significant discrepancy in susceptibility was observed in carbapenem-resistant and/or carbapenemaseproducing A. baumannii, with rates of 95.7% using the provisional PK/PD breakpoint (≤64 mg/L) versus 24.9% with the CLSI cefepime breakpoint (≤8 mg/L). 12 The aim of the present study is to evaluate the in vitro activity of cefepime/enmetazobactam and cefepime/zidebactam against CRAB isolates characterized by WGS, including strains susceptible and resistant to sulbactam/durlobactam.
## Material and methods
## Bacteria characterization
The CRAB isolates were selected from two Italian strain collections from university hospitals in Northern Italy and consisted of Acinetobacter baumannii bloodstream isolates from hospitalized patients. The selection comprised all consecutive blood culture isolates collected in 2022-2023 that were resistant to sulbactam/ durlobactam. In addition, seven consecutive isolates susceptible to sulbactam/durlobactam were included. Duplicate CRAB isolates collected from the same patient were excluded. Antimicrobial susceptibility testing was performed using Vitek2 system (Biomerieux, France) and results were confirmed by SensititreTM Plate EUMDRXXF (Thermofisher, USA). MICs for cefiderocol were assayed by microdilution in iron depleted-Mueller-Hinton broth with the ComASP cefiderocol test (Liofilchem, Italy), while MICs for sulbactam/durlobactam, cefepime/enmetazobactam and cefepime/zidebactam were evaluated by MIC TestStrip (Liofilchem, Roseto degli Abruzzi, Italy). All MIC values were determined in triplicate. Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as quality control strains to validate antimicrobial susceptibility testing. Results were interpreted following EUCAST and CLSI clinical breakpoints (for cefepime/zidebactam, the provisional PK/PD susceptibility breakpoint was >64 mg/L). Comparison between MICs of different antibiotics was performed using Student's t-test analysis implemented in GraphPad Prism v.10.1.11 (San Diego, CA, USA).
## Genome-sequencing analysis
Genomic DNA were extracted from purified bacterial cultures of A. baumannii using the DNeasy Blood&Tissue Kit (Qiagen, Switzerland) and cleaned up with AMPure XP magnetic beads (Beckman Coulter). Bacterial genomes were sequenced as previously described. 14 Briefly, libraries were prepared with DNA Prep Library Preparation Kit (Illumina, USA) and sequenced by the Illumina MiSeq platform (Illumina, USA) using the MiSeq Reagent Kit v.3 with 2 × 300 paired-end reads. The quality of the reads was evaluated using FastQC v.12.1 software (hps://www. bioinformatics.babraham.ac.uk/projects/fastqc/, accessed on 1 January 2023) and assembly was performed using SPAdes v.3.10. Antimicrobial resistance determinants and MLST analysis were evaluated using an online platform (available at https://www. genomicepidemiology.org/). The phylogenetic trees based on core genomes were generated as previously described. 14 Analysis of PBPs were performed by aligning gene sequences against reference genome of ATCC17978 A. baumannii strain using ClustalW. 14 SNP detection was manually curated using Unipro UGENE v.49.1.
## Results
Genomic characteristics of the CRAB clinical isolates included in the study are shown in the Table 1. Analysis of antimicrobial determinants related to β-lactams-resistance showed that all isolates also carried bla OXA-23 , while 61.9% (13/21), 47.6% (10/21) and 14.2% (3/21) carried bla ADC-25 , bla OXA-66 and bla NDM-1 β-lactamase genes, respectively. These genetic antimicrobial determinants were observed only in the sulbactam/ durlobactam-resistant strains. Notably, 75% (6/8) of the sulbactam/durlobactam-susceptible CRAB carried bla TEM-1 .
Analysis of genes related to resistance to other antimicrobial class showed that 90.5% (19/21) carried aph(3″)-Ib, 71.4% (15/ 21) carried sul1/2 or tetB and 57.1% (12/21) carried aph(6)-Id. Analysis of PBPs sequences revealed all strains harboured a similar allelic profile (i.e. wild type, PBP1a; P112S, PBP1b; wild type, PBP2; N392S, PBP5) excluding NDM-producing CRAB isolates. At the same time, specific mutations were observed within PBP3 gene between sulbactam/durlobactam-resistant (90%, N392T) and -susceptible (100%, A515V) strains. In addition, exclusive mutations within PBP genes were observed among NDM-producing CRAB [i.e. PBP1a (T38A, A244T and T766A); PBP1b (N513H); PBP2 (P665A); PBP3 (L480I, T511S)].
To evaluate the genomic relatedness of CRAB strains included in the study, phylogenetic analysis based on core-genome SNPs was performed. The derived phylogenetic tree showed that the isolates included in the study were related to other CRAB strains isolated in Italy, while strains carrying bla NDM-1 segregated separately to other Italian strains (Figure S1, available as Supplementary data at JAC Online in the Supplementary Material).
Phenotypic characteristics of the CRAB clinical isolates included in the study are shown in the Table S1 (in the Supplementary Material). Phenotypic characteristics of the CRAB clinical isolates included in the study are shown in Table S1 (in the Supplementary Material). Phenotypic results showed that all CRAB strains included in this study exhibited high MIC values for meropenem (median >16 mg/L, IQR 16 mg/L), cefepime (median >64 mg/L, IQR 64 mg/L) and cefepime/enmetazobactam (median >64 mg/L), whereas they showed low MICs for cefiderocol (median 0.5 mg/L, IQR 0.125-1 mg/L). At the same time, the median MIC for sulbactam/durlobactam against susceptible strains was 2 mg/L (IQR 2 mg/L), while it was 64 mg/L (IQR 64 mg/L) against resistant CRAB strains.
Of note, our results showed that cefepime/zidebactam demonstrated potent antibacterial activity against all CRAB strains Tascini et al. Cefepime combined with zidebactam against CRAB included in the study (Figure 1) by showing a significant reduction in MICs (P < 0.0001) in comparison to cefepime and cefepime/ enmetazobactam. In addition, our results showed that cefepime/zidebactam exerted higher antibacterial activity against sulbactam/durlobactam-susceptible strains (median 8 IQR 6.5-8; P < 0.0001) than sulbactam/durlobactam-resistant strains (median 8 IQR 4-8; P < 0.01) (Figure S2 in the Supplementary Material).
## Discussion
Here we showed that cefepime/zidebactam displayed a high antibacterial effect against CRAB strains including sulbacatam/ durlobactam-susceptible and -resistant clinical strains. Our results are partially in agreement with data presented in the literature showing a broad range of MICs of cefepime/zidebactam against A. baumannii clinical strains. [15][16][17][18] In particular, our results reveal that the MIC 90 for cefepime/zidebactam was 8 mg/L against CRAB strains included in this study, and similar results were observed by Mushtaq and co-workers against acquired OXA carbapenemases strains and Sader et al. against imipenemsusceptible strains A. baumannii clinical isolates. 15,16 Of note, we observed that statistical differences were not observed between sulbactam/durlobactam-susceptible and -resistant strains. These findings are in accordance with genotypic results by showing the absence of mutations within the main target of zidebactam (i.e. PBP2) among CRAB strains included in the study, thus resulting in a similar antibacterial activity against these two groups independent of the sulbactam/durlobactam resistance. We also hypothesized that the differences in cefepime/zidebactam activity against sulbactam/durlobactam-susceptible and -resistant strains is probably caused by the different β-lactamase gene contents (i.e. bla ADC-25 , bla OXA-66 β-lactamase genes present in sulbactam/ durlobactam-resistant strains) in association with specific mutations within the PBP-1 and PBP-3, which are targets of cefepime. 17 A recent review highlighted resistance mechanisms to cefepime/ zidebactam in Gram-negative bacteria, suggesting that resistance may involve multiple mutations in PBPs, as well as alterations in efflux systems. 12 In this context, the combination of efflux pump alterations, β-lactamase expression and PBP mutations could explain the high MIC observed in the CRAB48 isolate.
In addition, we observed that enmetazobactam did not increased the activity of cefepime against CRAB strains independent of the reduced susceptibility to sulbactam/durlobactam. These results are in agreement with previous studies showing the absence of inhibitory activity of cefepime/enmetazobactam against A. baumannii independent of antimicrobial resistance genes carried by isolates. 18,19 Here we provided the in vitro evidence of the antibacterial activity of cefepime/zidebactam against sulbactam/durlobactamsusceptible and -resistant CRAB strains. In this context, a recent study showed that human-simulated exposure of cefepimezidebactam determined a significant bactericidal effect against carbapenem-resistant A. baumannii expressing OXA carbapenemases in the murine thigh infection model, thus suggesting in vivo efficacy of this combination. 20 In conclusion, our findings suggest that cefepime/zidebactam may have potential activity against infections caused by carbapenem-resistant A. baumannii, including both sulbactam/ durlobactam-susceptible and -resistant strains, and could represent a potential candidate for the treatment of CRAB infections. However, this study has some limitations. The generalizability of our findings is limited by the relatively small number of bacterial strains included and the fact that isolates were collected from only two Italian centres. Larger multicentre studies, ideally including global strain collections, are needed to confirm and extend these findings and to better define the clinical impact of cefepime/zidebactam for the treatment of infections caused by sulbactam/durlobactam-resistant CRAB.
## References
1. Lee, Chen, Wu (2014) "Risk factors and outcome analysis of Acinetobacter baumannii complex bacteremia in critical patients" *Crit Care Med*
2. Casale, Bianco, Bastos (1934) "Prevalence and impact on mortality of colonization and super-infection by carbapenem-resistant Gram-negative organisms in COVID-19 hospitalized patients" *Viruses*
3. Lee, Lee, Park (2017) "Biology of Acinetobacter baumannii: pathogenesis, antibiotic resistance mechanisms, and prospective treatment options" *Front Cell Infect Microbiol*
4. Magiorakos, Srinivasan, Carey (2012) "Multidrug-resistant, extensively drug-resistant and pandrug-resistant bacteria: an international expert proposal for interim standard definitions for acquired resistance" *Clin Microbiol Infect*
5. "WHO Bacterial Priority Pathogens List, 2024: bacterial pathogens of public health importance to guide research, development and strategies to prevent and control antimicrobial resistance"
6. Kubin, Garzia, Uhlemann (2025) "Acinetobacter baumannii treatment strategies: a review of therapeutic challenges and considerations" *Antimicrob Agents Chemother*
7. Kaye, Shorr, Wunderink (2023) "Efficacy and safety of sulbactam-durlobactam versus colistin for the treatment of patients with serious infections caused by Acinetobacter baumannii-calcoaceticus complex: a multicentre, randomised, active-controlled, phase 3, noninferiority clinical trial (ATTACK)" *Lancet Infect Dis*
8. Tamma, Heil, Justo (2024) "Infectious Diseases Society of America 2024 guidance on the treatment of antimicrobial-resistant Gram-negative infections" *Clin Infect Dis*
9. Principe, Bella, Conti (2022) "Acinetobacter baumannii resistance to sulbactam/durlobactam: a systematic review" *Antibiotics*
10. Papp-Wallace, Bethel, Caillon (2019) "Beyond piperacillintazobactam: cefepime and AAI101 as a potent β-lactam-β-lactamase inhibitor combination" *Antimicrob Agents Chemother*
11. Lanier, Melton, Covert (2025) "Cefepime-enmetazobactam: a drug review of a novel beta-lactam/beta-lactamase inhibitor" *Ann Pharmacother*
12. Boattini, Gaibani, Comini (2025) "In vitro activity and resistance mechanisms of novel antimicrobial agents against metallo-β-lactamase producers" *Eur J Clin Microbiol Infect Dis*
13. Papp-Wallace, Nguyen, Jacobs (2018) "Strategic approaches to overcome resistance against Gram-negative pathogens using β-lactamase inhibitors and β-lactam enhancers: activity of three novel diazabicyclooctanes WCK 5153, zidebactam (WCK 5107), and WCK 4234" *J Med Chem*
14. Caiazzo, Bianco, Boattini (2025) "Genomic characterization of carbapenem-resistant Acinetobacter baumannii clinical strains resistant to sulbactam/durlobactam from Italy" *Eur J Clin Microbiol Infect Dis*
15. Mushtaq, Garello, Vickers (2021) "Activity of cefepime/zidebactam (WCK 5222) against 'problem' antibiotic-resistant Gram-negative bacteria sent to a national reference laboratory" *J Antimicrob Chemother*
16. Sader, Rhomberg, Flamm (2017) "WCK 5222 (cefepime/zidebactam) antimicrobial activity tested against Gram-negative organisms producing clinically relevant β-lactamases" *J Antimicrob Chemother*
17. Endimiani, Perez, Bonomo (2008) "Cefepime: a reappraisal in an era of increasing antimicrobial resistance" *Expert Rev Anti Infect Ther*
18. Liu, Ko, Lee (2018) "In vitro activity of cefiderocol, cefepime/ enmetazobactam, cefepime/zidebactam, eravacycline, omadacycline, and other comparative agents against carbapenem-non-susceptible Pseudomonas aeruginosa and Acinetobacter baumannii isolates associated from bloodstream infection in Taiwan between"
19. (2022) *J Microbiol Immunol Infect*
20. Bonnin, Jeannot, Henriksen (2025) "In vitro activity of cefepime-enmetazobactam on carbapenem-resistant Gram negativesauthor's response" *Clin Microbiol Infect* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12472342&blobtype=pdf | # Characterization of a New HIV-1 Second-Generation Circulating Recombinant Form CRF173_63A6 in the Jewish Autonomous Region of Russia
Vasiliy Ekushov, Maksim Halikov, Alexei Totmenin, Mariya Antonets, Tatyana Tregubchak, Andrey Murzin, Marina Pavlova, Anastasia Troianova, Tatyana Adusheva, Svetlana Beniova, Alexandra Ermolitskaya, Irina Gorelova, Alexander Agafonov, Natalya Gashnikova
## Abstract
Studies of HIV-1 molecular epidemiology describe significant differences in HIV infection spread across geographical areas. We examined 80 HIV-1 samples from the Jewish Autonomous Region of Russia in 2024. HIV-1 genome sequences for 12 samples revealed a novel HIV-1 called CRF173_63A6. HIV-1 CRF173_63A6 was found to have arisen through recombination between a specific Russian A6 subtype and the recombinant virus CRF63_02A6, which is responsible for the PWID-associated HIV outbreak in the Siberian region of Russia. Phylogenetic analysis of pol sequences previously deposited in Genbank showed that the CRF173_63A6 samples we described are grouped into a common phylogenetic cluster that includes 54 HIV-1 samples isolated in the JAR and other areas of the Russian Far East, indicating a wide distribution of this virus genovariant. This study once again proves the significant contribution of the key PWID group not only to the development of local Russian HIV epidemics, but also to the change in the characteristics of the circulating virus population.
## 1. Introduction
Studies of HIV molecular epidemiology in Russia describe significant differences in HIV infection spread across geographical areas: while the specific Russian subtype A6 HIV-1 continues to dominate in the Central and Western parts of the country [1][2][3], the CRF63_02A6 virus predominates in the regions of Siberia, where the HIV infection rate exceeds the national average [4][5][6]. Research into the genetic diversity of HIV in the Russian Far East indicates the development of separate internal epidemics, which is associated with the remoteness and geographic isolation of a number of territories [7][8][9]. The Jewish Autonomous Region (JAR), part of the Far Eastern Federal District of Russia, is one of the most sparsely populated areas of the country, with 146,000 residents. According to the Regional AIDS Center, a total of 487 cases of HIV infection are registered in the JAR. In recent years, various forms of recombinant HIV have contributed more to the development of the local epidemic in the Russian Far East [10,11]. Based on the analysis of complete viral genomes, our study identifies and describes a new second-generation circulating recombinant form of HIV-1 that has spread in the JAR.
## 2. Materials and Methods
## 2.1. Sample Source
Plasma samples were collected from patients at the AIDS Prevention and Control Center, Birobidzhan, Jewish Autonomous Region and Regional clinical hospital No 2, AIDS Prevention and Control Center, Vladivostok, Primorsky Krai, in 2023-2024.
## 2.2. Ethics
The study was conducted in accordance with the Helsinki Declaration, and the protocol was approved by the Ethics Committee of the "JAR AIDS Prevention and Control Center" (Protocol No. 2 of 27 January 2021). All participants provided informed consent for the collection and subsequent analysis of their samples.
## 2.3. Sequence Analysis
The GenBank accession numbers for the NFLG sequences of the studied isolates are PQ523366-PQ523377 and PQ585408-PQ585414. The HIV-1 sequences we obtained were compared with reference sequences of different subtypes and recombinant forms from the international database GenBank in the MEGA11 and AliView programs [12,13]. Multiple alignment was performed using the MAFFT version 7.526 (RIMD) program with standard settings [14].
The maximum likelihood phylogenetic tree was created using the online resource IQ-TREE v1.6.12 with a bootstrap of 1000 repeats based on the GTR + I + G substitution model, and bootstrap analysis was used to evaluate the topology [15]. The phylogenetic tree was visualized using the iTOL toolkit [16].
RIP, jpHMM, and Bootscan assays were used to determine the recombination structures of viruses [17][18][19]. The following settings were used to perform the Bootscan analysis: window size 400 bp and step size 50 bp. The following reference subtypes of HIV-1 were used as comparison sequences: A6, CRF63_02A6, and a representative sequence of subtype B. Recombinant mosaic map of CRF173_63A6 was generated using the Recombinant HIV-1 Drawing Tool (https://www.hiv.lanl.gov/content/sequence/DRAW_CRF/recom_mapper. html, accessed on 26 September 2024).
The tMRCA was evaluated using the Nextstrain software package [20].
## 3. Results
Analysis of 80 HIV-1 samples, which we isolated from patients at the JAR AIDS Centre in 2024, revealed the following distribution of HIV genovariants: 51.3% belonged to subtype A6; 10.0% to subtype B; 5.0% to CRF02_AG; 6.3% to CRF63_02A6; and 1.3% each to A1, G, and URF_A6C. A group of 19 samples (23.8%) formed a separate HIV-1 phylogenetic cluster.
Near-full-length HIV-1 genome (NFLG) sequences for 12 samples from the established phylogenetic group of viruses were successfully obtained (the details of the patients' demographics and clinical characteristics can be found in Table S1). Additionally, the sequences of HIV isolated from residents of the Primorsky Territory and the JAR, related to the genetic variants A6 (14 samples) and CRF63_02A6 (5 samples), were used to analyze the genome structure of the studied viruses. Phylogenetic analysis showed that the 12 studied HIV-1 samples, which were subsequently assigned the name CRF173_63A6, grouped outside of any known HIV-1 or CRF subtype and formed a separate monophyletic branch with a bootstrap value of 100% (Figure 1), indicating their origin from a single ancestor. The 12 CRF173_63A6 genomes that we described had identical recombination profiles. Seven recombination breakpoints were identified in the genome; these were located in the gag (two breakpoints), pol (three breakpoints), vpu (one breakpoint), and env (one breakpoint) genes. All strains had these recombination breakpoints in common. Subregional phylogenetic analysis of eight genomic segments was conducted to study their probable parental lines. Phylogenetic analysis of subregions I, III, VI, and VIII confirmed their relationship with CRF63_02A6 viruses. Segments II, IV, and VII of the studied CRF173_63A6 were related to the phylogenetic group of the HIV-1 subtype A6 circulating in Russia. Fragment V does not have precise coordinates of the recombination break and is located between subtypes A6 and CRF63_02A6 on the phylogenetic tree.
Effectively, HIV-1 CRF173_63A6 resulted from recombination between CRF63_02A6 and A6 with seven breakpoints delimiting four CRF63_02A6 fragments, three fragments of different lengths of the A6 subtype, and one fragment with no clear boundaries between the subtypes. Figure 1 provides the genomic subregion and genome coordinates according to HXB2.
To understand the origin of the CRF173_63A6 variant, an additional tMRCA assessment was performed. As shown in Figure S1, CRF63_02A6 and CRF173_63A6 HIV-1 originated from the same ancestor, and the CRF173_63A6 variant we describe began to spread in the Russian Far East around 2004, 2 years after the emergence of CRF63_02A6 HIV-1 in Siberia [4].
In the group of 18 patients from our study sample in whom CRF173_63A6 HIV-1 was isolated, 6 individuals were infected sexually, and 11 by using injected drugs. With the exception of one case of vertical mother-to-child transmission in 2011, there was no epidemiological link between these patients. For nine patients, HIV was diagnosed from 2011 to 2015, and eight individuals were diagnosed from 2018 to 2024. Eleven out of eighteen patients from this group live in the Oblutchensky district in the JAR. The Oblutchensky district is the most disadvantaged area in the JAR, where 34% of all patients with HIV identified in the JAR over the entire period of the HIV spread are registered; the HIV attack rate per 100,000 population (638.4) for the Oblutchensky district is 2.2 times higher than the regional average.
Analysis of HIV-1 pol sequences previously deposited in GenBank revealed a phylogenetic cluster combining the 18 CRF173_63A6 genovariants studied by us with 36 other HIV-1 samples (Figure 2). Of the 36 HIV-1 samples, 32 samples were isolated and deposited from 2016 to 2019 in the JAR, 3 in 2019 in Blagoveshchensk, and 1 in 2021 in Khabarovsk. An analysis of epidemiological data from the JAR residents infected with CRF173_63A6 shows that this virus is actively spreading in the JAR among PWID and is transmitted through sexual contact.
## 4. Discussion
The global spread of HIV-1 in Russia since the early 1990s has been associated with an active increase in the number of people who inject drugs [21][22][23]. In many areas of the country, HIV transmission in the key group of PWID dominated until 2014-2016. After 2010, the number of people using injectable drugs in Russia began to decrease, but at the same time, the consumption of more accessible synthetic or pharmacy drugs increased sharply, and the spread of HIV began to occur in the key group of PWID and their sexual partners [24]. Currently in Russia, the transmission of HIV infection is predominantly through heterosexual contact, but unprotected sexual contact can also be provoked by the use of smoking narcotic mixtures and synthetic drugs [25].
The circulation of different genetic variants of HIV and the practice of risk behavior in relation to infection are necessary and sufficient conditions for the emergence of new recombinant forms of HIV.
Based on the phylogenetic analysis and analysis of the recombination structure complete viral genomes, our study describes a new second-generation circulating recombinant form of HIV-1 that has spread in the Far Eastern Federal District of Russia. The secondgeneration recombinant form of the HIV-1 CRF173_63A6 subtype, which developed from the Russian A6 subtype and the CRF63_02A6 recombinant virus responsible for HIV outbreaks among PWID in the Siberian region, is described. Epidemiological data of the patients with CRF173_63A6 viruses indicate that this subtype has begun to spread among injecting drug users (IDUs). Recently described by us, HIV-1 CRF157_A6C, CRF147_A6B, and CRF133_A6B have also been identified among IDUs and their sexual partners [9,26,27].
The research data once again proves the significant contribution of the key group of PWID not only to the development of local Russian HIV epidemics, but also to the change in the characteristics of the circulating virus population [9,26,27].
The increase in the registered cases of various HIV-1 URFs and the spread of the emerging recombinant HIVs emphasizes the importance of strengthening infection prevention measures, including HIV reinfection in the key groups. Monitoring the spread of new emerging HIV is essential for understanding the evolution and patterns of change in the molecular epidemiology of the virus.
## References
1. Abidi, Aibekova, Davlidova et al. "Origin and evolution of HIV-1 subtype A6" *PLoS ONE*
2. Lebedev, Lebedeva, Moskaleychik et al. (2019) "Human immunodeficiency virus-1 diversity in the Moscow region, Russia: Phylodynamics of the most common subtypes" *Front. Microbiol*
3. Van De Klundert, Antonova, Di Teodoro et al. (2022) "Molecular epidemiology of HIV-1 in Eastern Europe and Russia" *Viruses*
4. Sivay, Maksimenko, Osipova et al. "Spatiotemporal dynamics of HIV-1 CRF63_02A6 sub-epidemic"
5. Safina, Sidorina, Efendieva et al. "Molecular epidemiology of HIV-1 in Oryol oblast"
6. Rudometova, Shcherbakova, Shcherbakov et al. (2021) "Genetic Diversity and Drug Resistance Mutations in Reverse Transcriptase and Protease Genes of HIV-1 Isolates from Southwestern Siberia" *AIDS Res. Hum. Retroviruses*
7. Kotova, Trotsenko, Balakhontseva et al. (2019) "Molecular genetic characteristics of HIV-1 variants isolated in the subjects of the Russian far east" *Vopr. Virusol. (Probl. Virol. Russ. J.)*
8. Kirichenko, Kireev, Lapovok et al. (2006) "HIV-1 drug resistance among treatment-naïve patients in Russia: Analysis of the national database"
9. Halikov, Ekushov, Totmenin et al. (2024) "Identification of a novel HIV-1 circulating recombinant form CRF157_A6C in Primorsky Territory" *Russia. J. Infect*
10. Kotova, Trotsenko, Balakhontseva et al. (2019) "The importance of detection of the patterns of HIV-1 genome variability in the system of epidemiological surveillance for HIV-infection (on the example of some territories of the Eastern Federal district). Far East" *J. Infect. Pathol*
11. Kotova, Balakhontseva, Bazykina et al. (2021) "Circulating recombinant forms of HIV-1 constituent entities of the Far Eastern Federal district. Far East" *J. Infect. Pathol*
12. Koichiro, Stecher, Kumar (2021) "MEGA11: Molecular evolutionary genetics analysis version 11" *Mol. Biol. Evol*
13. Larsson (2014) "AliView: A fast and lightweight alignment viewer and editor for large datasets" *Bioinformatics*
14. Kazutaka, Rozewicki, Yamada (2019) "MAFFT online service: Multiple sequence alignment, interactive sequence choice and visualization" *Brief. Bioinform*
15. Trifinopoulos, Nguyen, Haeseler et al. (2016) "W-IQ-TREE: A fast online phylogenetic tool for maximum likelihood analysis" *Nucleic Acids Res*
16. Letunic, Bork (2024) "Interactive Tree of Life (iTOL) v6: Recent updates to the phylogenetic tree display and annotation tool" *Nucleic Acids Res*
17. Siepel, Halpern, Macken et al. (1995) "A Computer Program Designed to Screen Rapidly for HIV Type 1 Intersubtype Recombinant Sequences" *AIDS Res. Hum. Retroviruses*
18. Schultz, Zhang, Bulla et al. (2009) "Improving the reliability of recombination prediction in HIV-1" *Nucleic Acids Res*
19. Martin, Varsani, Roumagnac et al. "RDP5: A computer program for analyzing recombination in, and removing signals of recombination from, nucleotide sequence datasets"
20. Hadfield, Megill, Bell et al. (2018) "Nextstrain: Real-time tracking of pathogen evolution" *Bioinformatics*
21. Rhodes, Sarang, Bobrik et al. (2004) "HIV transmission and HIV prevention associated with injecting drug use in the Russian Federation" *Int. J. Drug Policy*
22. Rhodes, Lowndes, Judd et al. (2022) "Explosive spread and high prevalence of HIV infection among injecting drug users in Togliatti City" *Russia. Aids*
23. Hamers, Downs (2003) "HIV in central and eastern Europe" *Lancet*
24. Meylakhs, Aasland, Grønningsaeter (2017) "Until people start dying in droves, no actions will be taken": Perception and experience of HIV-preventive measures among people who inject drugs in northwestern Russia" *Harm Reduct. J*
25. Zyryanova, Astakhova, Ismailova et al. (2020) "Detection of HIV-1 resistant to antiretroviral drugs among tomsk oblast population with newly diagnosed HIV-infection" *J. Infectol*
26. Maksimenko, Sivay, Totmenin et al. (2022) "Novel HIV-1 A6/B recombinant forms (CRF133_A6B and URF_A6/B) circulating in Krasnoyarsk region" *Russia. J. Infect*
27. Maksimenko, Sivay, Antonets et al. (2024) "Expanding HIV-1 diversity in Russia: Novel circulating recombinant form between subtypes A6 and B (CRF147_A6B)" *J. Infect*
28. "The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817012&blobtype=pdf | # Onderstepoort Journal of Veterinary Research
Wegdan Ali, Intisar Saeed, Shaza Mutwakil, Muzdalifa Alamin, Abdelgader Ball, Mona Ahmed, Abubaker Sa, Rayan Asil, Omer Algezoli, Muaz Abdella, Yahia Ali
## Abstract
The present study confirmed the existence and isolation of type A AI virus from different species of wild birds as well as subtyping of its virus for the first time in Khartoum State, Sudan.
## Introduction
Influenza A viruses are grouped in the Influenza virus A genus of the Orthomyxoviridae family (Charostad et al. 2023). Avian influenza (AI) viruses are globally prevalent in domestic and non-domestic birds. The viruses are grouped into low-pathogenic (LP) and highpathogenic (HP) AI viruses by the Office International des Epizooties (OIE 2012). Most of the haemagglutinin-neuraminidase (HA-NA) combinations were identified in avian species, while highly pathogenic avian influenza (HPAI) viruses were detected only in viruses of the subtypes H5 or H7. Highly pathogenic AI virus infection may result in up to 100% losses with drastic production loss in affected domestic poultry and non-domestic birds.
Asymptomatic infections can occur in wild birds; HPAI viruses are seldom detected (Charostad et al. 2023). Waterfowl have been reported to have the highest AI virus prevalence rates, with the greatest subtype variety (Diskin et al. 2020). Wild birds are considered a natural reservoir and may carry LP AI strains. However, viral mutations may lead to the existence of highly pathogenic ones, commonly in H5 and H7 genotypes (Ghersi et al. 2011). All AI type-A subtypes are found in different populations of water birds, which shed large quantities of the virus for long durations in respiratory secretions and/or faeces (Runstadler et al. 2007). The main wild bird species involved in AI are from the orders Anseriformes (geese, swans and ducks) as well as Charadriiformes (especially, waders, gulls and terns) (Olsen et al. 2006).
Despite extensive global knowledge on AI in wild birds, local data from Sudan remain limited. In Sudan, previously a high titre (86.4%) of AI virus type A antibodies was detected in chickens (Ali, Kheir & Ballal 2007), and 28.4% positivity was recorded by competitive enzyme-linked immunosorbent assay (c-ELISA) in Khartoum State (Elamin 2000). Further, sera positive for AI type A were subtyped, 45.3% subtype H9 and 3.5% subtype H7 were Avian influenza (AI) is a significant disease affecting chickens and other avian species. Wild birds are thought to contribute to the virus transmission. The present study intends to explore the existence of AI type A virus in wild birds at the Six April Zoo, Khartoum State, Sudan. A total of 42 cloacal and tracheal swabs were collected from clinically healthy individuals belonging to five different wild bird species. The selected wild bird species were Common crane Grus grus, Sudan crowned crane Balearica pavonina, Helmeted guinea fowl Numida meleagris, Duck sp. Anatidae and Chestnut-billed sand grouse Pterocles exustus. Swabs were examined for AI virus antigen using the agar gel immunodiffusion (AGID) test, and all tested swabs produced positive results. The swab samples were inoculated into embryonated chicken eggs. The isolated virus was identified by AGID test and polymerase chain reaction. The virus was isolated from swabs collected from Grus grus, Balearica pavonina, Numida meleagris, Duck sp. Anatidae and Pterocles exustus. Subtyping of the isolated viruses was performed using reverse transcriptase-polymerase chain reaction, which identified the H5 subtype.
## Isolation and subtyping of avian influenza A virus from wild birds in Khartoum, Sudan
Read online:
## Detection, isolation and identification of the virus
The collected samples were examined for AI type A virus antigen using agar gel immunodiffusion (AGID) test. Virus isolation was adopted according to the OIE procedure (OIE 2012). Briefly, the prepared inocula were inoculated into the allantoic cavity of 9-10-day-old embryonated chicken eggs and run passages 4-6 times.
## Determination of the isolated virus
## Serological identification of the isolated virus
Haemagglutination test: The existence of haemagglutinating viruses in harvested allantoic fluid was determined using the HA test technique with 1% chicken red blood cell (RBC) suspension. Newcastle disease (ND) virus was excluded by the haemagglutination inhibition (HAI) test against reference ND virus antiserum.
## Agar gel immune diffusion test:
The AI virus identification for the harvested allantoic fluid was examined by AGID test. Allantoic fluids negative for ND virus were tested for avian influenza virus (AIV) with reference AI type A antiserum using AGID test as described (OIE 2012).
## Molecular identification
To identify the isolated virus, the harvested allantoic fluid was examined for AI type A and subtype H5 using reverse transcriptase-polymerase chain reaction (RT-PCR).
Reverse transcriptase-polymerase chain reaction: Reverse transcriptase-polymerase chain reaction, targeting M and HA genes, was adopted for the identification and subtyping of influenza A virus.
## Ribonucleic acid extraction:
The nucleic acid of the virus collected randomly from the five wild bird species investigated was purified from samples using QIAamp ® Viral Ribonucleic Acid (RNA) Mini Kit (Qiagen, Venlo, the Netherlands) as in the manufacturer's protocol instructions.
The extracted RNA was frozen at -80 °C till used.
## Positive and negative controls
Avian influenza genome of domestic poultry locally isolated strain (Ali & Kheir 2007) was kindly provided by the CVRL, Khartoum. Serial dilutions were used to estimate the optimum deoxyribonucleic acid (DNA) concentration and double-distilled water (DDW) was used as negative control.
## Synthesis of complementary deoxyribonucleic acid
Synthesis of complementary deoxyribonucleic acid (cDNA) was adopted by transcriptor first-strand cDNA synthesis kit (Roche, Inc., Rotkreuz, Switzerland). Briefly, 1 µL random hexanucleotide primers (600 pmole/µL), 4 µL 5x reaction buffer (8 mM MgCl 2 ), 0.5 µL of ribonuclease (RNase) inhibitor (40 U/µL) and reverse transcriptase (20 U/µL), 2 µL 10 mM deoxyribonucleotides (dNTP) mix and 7 µL water free of nuclease were added to 5 µL of deoxyribonucleotides (DNA). The reaction was then completed in accordance with the procedure provided.
## Amplification of M gene
The reaction was performed using primers designed by Capua and Alexander (2009) for AI identification (Table 1). Polymerase chain reaction (PCR) mix includes 5 µL of buffer, 2.5 mM MgCl 2 , 1 mM dNTPs, 100 mM dithiothreitol (DTT), 0.3 µM primers (Table 1), 10 U RNase inhibitor, 2.5 U Taq DNA polymerase (QIAGEN). The PCR condition set was as follows: 42 °C for 20 min, then followed by 95 °C for 5 min; denaturation at 94 °C for 60 s, annealing at 55 °C for 60 s and then 40 cycles of extension at 72 °C for 60 s, with a final elongation at 72 °C for 10 min. Amplicons were detected in ethidium bromide-stained agarose gel (2%).
## Haemagglutinin gene amplification
Amplification of the HA gene was achieved using primers derived from the published sequences and according to the procedure described by Slomka et al. (2007) (Table 1).
The reaction was completed using One-Step enzyme mix (Qiagen One-Step RT-PCR Kit), including primers and 8 U of RNase inhibitor (Promega). The thermal cycle or heating conditions were as follows: for 30 min at 50 °C; 94 °C for 15 min; 40 cycles at 94 °C for 30 s, 58 °C for 1 min and then 68 °C for 2 min and then with a final extension at 68 °C for 7 min. Amplicons were visualised on a 2% agarose gel.
## Ethical considerations
Ethical clearance to conduct this study was obtained from the Central Veterinary Research Laboratory, Department of Virology, Khartoum, Sudan. The material used is only cloacal and tracheal swab specimens collected from clinically healthy wild bird species. The samples collected during the study are only cloacal and tracheal swabs; no administration of any hazardous materials was done to any of the sampled birds.
## Results
## Detection, isolation and identification of the virus
Using AGID, AI type A virus antigen was detected in all tested cloacal and tracheal of wild birds (Grus grus, Balearica pavonina, Numida meleagris, Duck sp. Anatidae and Pterocles exustus). The detected AI virus was isolated from the embryonated chicken eggs.
## Serological identification of the isolated virus
## Haemagglutination test
The harvested allantoic fluid of the inoculated embryos was found to agglutinate 1% chicken RBCs using HA test.
## Agar gel immunodiffusion test
Existence of AI virus in the harvested fluids was confirmed by using AGID test.
## Molecular identification of the isolated virus
## Reverse transcriptase-polymerase chain reaction
All samples under test gave positive results. Clear bands were detected in the ethidium bromide-stained gel that correspond to the probable band size for M gene (244 base pairs [bp]) and HA gene (320 bp). Control positive gave positive results, while no products were amplified when DDW was used as template (Figure 1 and Figure 2).
## Discussion
Wild birds are reported as a natural reservoir for AI virus subtypes (Capua & Alexander 2009;Graziosi et al. 2024;Swayne 2008). Low-pathogenic viruses are the main circulating AI viruses for domestic galliforms with no obvious clinical signs in infected birds. However, these viruses can spread to other birds, where some strains may be adapted (Brown, Poulson & Stallknecht 2014). Mostly, after the establishment of AI virus in a new host, wild birds' involvement in the viral transmission and maintenance is not significant because of the host adaptation changes, for that, determination of HPAI viruses in wild birds is rare (Brown et al. 2014). (Bevins et al. 2016).
In the present study, subtype H5 was detected for the first time in all subtyped samples collected from the five examined wild bird species. This agrees with reported subtypes in Central, Eastern, Southern and West African countries, where H5 was the most detected subtype (79%) in both domesticated and wild birds (Kalonda et al. 2020;Kirunda et al. 2014). The same results were obtained in wild birds in Korea (Lee et al. 2017) and Japan (Abao et al. 2013). Subtype H5 was found as well in wild birds in China (Cui et al. 2020), Germany (King et al. 2021), the Netherlands (Engelsma et al. 2022), Egypt (El-Shesheny et al. 2021;Mahmoud et al. 2024), the US (Bevins et al. 2016) and in water habitats of birds (Kenmoe et al. 2024).
## Limitation of the study
The study was focused and limited to a single area with a limited number of samples; it was intended only to shed light on the existence of AI infection in wild birds and to open the floor for further epidemiological and molecular studies in the future.
## Conclusion
The present study confirmed the isolation of the virus from five different wild bird species, along with the subtyping of the identified virus. Further routine surveillance and research studies are required to be carried out to determine the virus's virulence, its molecular characterisation, host susceptibility and the prevalence of viral infection in each species, to aid in the effective control of this viral infectious disease.
## References
1. Abao, Jamsransuren, Bui et al. (2009) "Surveillance and characterization of avian influenza viruses from migratory water birds in eastern Hokkaido, the northern part of Japan" *Virus Genes*
2. Ali, Mansour (2015) "Serological survey of avian influenza type A and subtype A antibodies in chickens sera in Sudan using AGID and HAI tests" *International Journal of Preventive Medicine*
3. Ali, Zakia (2015) "Molecular detection and histopathological evaluation of naturally occurring HPAI virus in chicken in Sudan" *International Journal of Animal Research*
4. Ali, Kheir (2007) "Isolation of highly pathogenic avian influenza virus (H5N1) from poultry in Sudan in 2006" *Journal of Animal and Veterinary Advances*
5. Ali, Kheir, Ballal (2007) "Serological survey of type A avian influenza antibody in chicken sera in Sudan using indirect ELISA" *Research Journal of Veterinary Sciences*
6. Bevins, Dusek, White et al. (2016) "Widespread detection of highly pathogenic H5 influenza viruses in wild birds from the Pacific Flyway of the United States" *Scientific Reports*
7. Brown, Poulson, Stallknecht (2014) "Wild bird surveillance for avian influenza virus" *Animal Influenza Virus. Methods in Molecular Biology*
8. Capua, Alexander (2009) "Avian influenza and Newcastle disease: A field and laboratory manual"
9. Charostad, Rukerd, Mahmoudvand et al. (2023) "A comprehensive review of highly pathogenic avian influenza (HPAI) H5N1: An imminent threat at doorstep" *Travel Medicine and Infectious Disease*
10. Cui, Li, Li et al. (2020) "Evolution and extensive reassortment of H5 influenza viruses isolated from wild birds in China over the past decade" *Emerging Microbes & Infections*
11. Diskin, Friedman, Krauss et al. (2020) "Subtype diversity of influenza A virus in North American waterfowl: A multidecade study" *Journal of Virology*
12. El-Shesheny, Kandeil, Mostafa et al. (2021) "H5 influenza viruses in Egypt" *Cold Spring Harbor Perspectives in Medicine*
13. Elamin (2000) "Studies on avian influenza in Khartoum"
14. Elnasri, Abdel Rahim (2009) "Sero-epidemiological survey of anti-avian influenza virus antibodies in five avian species in the Sudan" *Sudan Journal of Veterinary Research*
15. Engelsma, Heutink, Harders et al. (2020) "Multiple introductions of reassorted highly pathogenic avian influenza H5Nx viruses clade 2.3.4.4b causing outbreaks in wild birds and poultry in the Netherlands" *Microbiology Spectrum*
16. Ghersi, Sovero, Icochea et al. (2011) "Isolation of low-pathogenic H7N3 avian influenza from wild birds in Peru" *Journal of Wildlife Diseases*
17. Graziosi, Lupini, Catelli et al. (2024) "Highly pathogenic avian influenza (HPAI) H5 clade 2.3.4.4b virus infection in birds and mammals" *Animals*
18. Kalonda, Saasa, Nkhoma et al. (2020) "Avian influenza viruses detected in birds in sub-Saharan Africa: A systematic review" *Viruses*
19. Kenmoe, Takuissu, Ebogo-Belobo et al. (2024) "A systematic review of influenza virus in water environments across human, poultry, and wild bird habitats" *Water Research*
20. Khatun, Tasnim, Hossain et al. (2024) "Molecular epidemiology of avian influenza viruses and avian coronaviruses in environmental samples from migratory bird inhabitants in Bangladesh" *Frontiers in Veterinary Science*
21. King, Harder, Conraths et al. (2006) "The genetics of highly pathogenic avian influenza viruses of subtype H5 in Germany" *Transboundary and Emerging Diseases*
22. Kirunda, Erima, Tumushabe et al. (2014) "Prevalence of influenza A viruses in livestock and free-living waterfowl in Uganda" *BMC Veterinary Research*
23. Lee, Kang, Song et al. (2017) "Surveillance of avian influenza viruses in South Korea between 2012 and 2014" *Virology Journal*
24. Lestari, Lubis, Dharmawan et al. (2020) "Co-circulation and characterization of HPAI-H5N1 and LPAI-H9N2 recovered from a duck farm" *Transboundary and Emerging Diseases*
25. Mahmoud, Gomaa, El Taweel et al. (2024) "Transmission dynamics of avian influenza viruses in Egyptian poultry markets" *Viruses*
26. Mutisari, Muflihanah, Wibawa et al. (2021) "Phylogenetic analysis of HPAI H5N1 virus from duck swab specimens in Indonesia" *Journal of Advanced Veterinary Research*
27. Oie (2012) "Manual of Diagnostic Test and Vaccines for Terrestrial Animals"
28. Olsen, Munster, Wallensten et al. (2006) "Global patterns of influenza A virus in wild birds" *Science*
29. Parvin, Kabiraj, Mumu et al. (2020) "Active virological surveillance in backyard ducks in Bangladesh: Detection of avian influenza and gammacoronaviruses" *Avian Pathology*
30. Pinon, Vialette (2019) "Survival of viruses in water" *Intervirology*
31. Runstadler, Happ, Slemons et al. (2005) "Using RRT-PCR analysis and virus isolation to determine the prevalence of avian influenza virus infections in ducks at Minto Flats State Game Refuge" *Archives of Virology*
32. Slomka, Coward, Banks et al. (2007) "Identification of sensitive and specific avian influenza polymerase chain reaction methods through blind ring trials organized in the European Union" *Avian Diseases*
33. Stallknecht, Luttrell, Poulson et al. (2012) "Detection of avian influenza viruses from shorebirds: Evaluation of surveillance and testing approaches" *Journal of Wildlife Diseases*
34. Swayne (2008) "Epidemiology of avian influenza in agricultural and other manmade systems" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12172444&blobtype=pdf | # A veterinary virapalooza: a summary of the 2024 American Society for Virology (ASV) Veterinary/Zoonotic Virology Satellite Symposium and online H5N1 panel discussion
| Virology, | Minireview, Andrew Broadbent, Nicola Stonehouse
## Abstract
The year 2024 saw veterinary/zoonotic virology take center stage once more as the American Society for Virology (ASV) hosted a satellite symposium on the subject in June and an online panel discussion in December. The symposium comprised six talks from experts on viruses of economic importance to agriculture and of public health importance. The viruses in question spanned foot and mouth disease virus (FMDV), African swine fever virus (ASFV), Marek's disease virus (MDV), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and influenza A viruses (IAVs), and topics covered fundamental virology, applied virology, epidemiology, and surveillance. The goal was to emphasize that improving the control of animal viral diseases requires an integra ted, holistic approach involving academia, government, and industry labs undertaking research on basic virology, vaccinology, epidemiology, and surveillance. Moreover, the symposium aimed to highlight career opportunities in the agricultural/veterinary sector for those with virology training. Six months following the symposium, the ASV held an online panel discussion on the ongoing H5N1 IAV situation in poultry, cattle, and people to provide more up-to-date information to its membership. A summary of the talks and discussions is presented here.
KEYWORDS veterinary virology, zoonoses, virology, agricultural viruses, H5N1T he American Society for Virology (ASV) is committed to representing and supporting its members who specialize in viral diseases of animals. To this end, the 2024 annual meeting held in June at the Greater Columbus Convention Center in Columbus, Ohio, and hosted by The Ohio State University featured veterinary and zoonotic virology in the form of workshops, plenary speakers, state-of-the-art talks, and a satellite sympo sium. The veterinary virology community selected a broad theme for the symposium to encompass researchers working on One Health, non-zoonotic animal viral diseases, and vaccines and requested speakers from the industry and government in addition to the academia. As a result, the 2024 symposium was titled "A Veterinary Virapalooza: Improving the Control of Viral Diseases of Animals, " and the goal was to highlight that improving the control of animal viral diseases requires an integrated, holistic approach involving academia, government, and industry labs undertaking research on basic virology, vaccinology, epidemiology, and surveillance. To this end, the symposium was structured into sessions on fundamental virology, applied virology, and epidemiology and surveillance, with the latter also being a One Health theme tackling zoonoses. Moreover, a key aim of the symposium was to demonstrate career opportunities for virologists in the agricultural/veterinary sector.Since the last veterinary virology satellite symposium in 2021, major animal virus outbreaks have continued, including African swine fever virus (ASFV) in Asia, Russia, and Europe (1), severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in wild animals
replication without affecting host gene expression. This technique has achieved 57% survival of ASFV-infected pigs compared to 0% survival with no treatment, and it is hoped this could be used in an outbreak setting. His work illustrates how fundamental virology informs the development of control strategies.
## Applied virology
Dr. Claudia Osorio [Associate Advisor for US Poultry Elanco Animal Health, Inc., and President of the American Association of Avian Pathologists (AAAP), 2024-2025] delivered a presentation on the critical role of applied virology in poultry production. The US poultry industry is a global leader, producing 9.17 billion broilers (chickens sold for meat) in 2022, totaling to $76.9bn in sales (17). Given the scale of the industry, effective disease prevention measures, such as vaccination, are essential to prevent production losses and ensure flock health. Dr. Osorio highlighted the use of vaccines in commercial poultry settings, where up to 20 commercial vaccines are administered, depending on the breed, age, and specific disease risk. In addition to commercial vaccines, autoge nous vaccines are used to target farm-specific pathogens. She emphasized the various methods used to administer vaccines, including inoculation into the embryonated egg (in ovo), which allows for high automation, delivery to the mucosal surface by spray or drinking water, and administration by subcutaneous or intramuscular injection. One such success story from the industry was the development of vaccines against Marek's disease virus (MDV), a herpesvirus responsible for tumors in chickens. Before the introduction of the vaccine, MDV caused mortality rates between 30 and 80% in affected flocks. Today, thanks to vaccination, the disease is no longer clinically in the US. Moreover, the herpesvirus of turkeys (HVT) is a useful vector for the delivery of antigens to protect chickens against other diseases and is becoming popular in vaccine schedules (18). One main take-home from Dr. Osorio's talk was that there is a pressing need for scientists with virology and immunology expertise to help make the next generation of poultry vaccines. These efforts are crucial for maintaining the health and productivity of poultry flocks and for preventing economic losses within the industry. Dr. Osorio highlighted that many virologists already have positions within the poultry industry in a variety of roles. Furthermore, Dr. Osorio encouraged members of ASV to explore career opportunities in the poultry industry and to get involved with the AAAP to strengthen links between the two societies.
## Epidemiology and surveillance
Addressing the topic of One Health, Dr. Andrew Bowman (Professor, College of Veteri nary Medicine, The Ohio State University) gave a talk on SARS-CoV-2 in white-tailed deer. In January-March 2021, Dr. Bowman's team processed nasal swabs from 360 deer from nine urban metro parks in Ohio and found that 129 (35.8%) were positive for SARS-CoV-2. The virus was found at all sites, and male deer had higher rates than female. Whole-genome sequencing was performed, and the entire genome sequences of 14 strains were obtained from six sites, and three lineages were discovered. Moreover, the lab was able to isolate two of the virus samples using Vero cells (19). Following on from this study, 83/88 counties in Ohio were tested, and nearly 60% had positive deer, with 163/1522 (10.7%) animals being antigen-positive and 274/1164 (23.5%) seropositive. The lab identified six independent human-to-deer transmission events, indicating that deer are highly susceptible to infection from people, and meaning that SARS-CoV-2 is a reverse zoonosis, as well as a zoonosis. Viruses were sequenced during the delta wave, yet two clusters were found to be the alpha variant, suggesting months of persistence in the deer population, which is different from the situation in humans. Interestingly, despite this persistence, the rate of evolution was threefold higher in deer compared to people (20). These data demonstrate how viruses can evolve differently in animal populations compared to humans, emphasizing the importance of continued surveil lance.
Addressing companion animals, Dr. Colin Parrish (Professor, College of Veterinary Medicine, Cornell University) discussed the epidemiology of H3N8 and H3N2 influenza A viruses (IAVs) in dogs and horses. The H3N8 IAV first emerged in horses in South America in the 1960s before spilling over into dogs in 1999. Canine H3N8 was first recognized in greyhounds in a training facility in Florida and initially spread with greyhound movement in the US to many regions in the South and Midwest but then quickly began to die out, with phylogenetic analyses revealing only sustained transmission in geographically constrained areas around Denver and Colorado Springs and in the North East USA, being maintained in dog shelters in New York City. The virus eventually became extinct in the dog population in 2016 ( 21) but continues to circulate in horses.
In contrast, the H3N2 IAV resulted from an avian virus transferring to dogs around 2004 and was first reported from South Korea and China (22). Canine H3N2 then formed a Chinese lineage and a South Korean lineage. In 2015, canine H3N2 was first transmitted to North America in rescued dogs from South Korea that were brought to the US. Initially, it was believed that this virus subsequently died out similar to the H3N8 viruses (23), but the virus re-emerged in Florida in 2021, and that virus spread to the Los Angeles area soon after where it likely infected over 50,000 dogs. The evidence points to this being a separate reintroduction into the US, likely due to the importation of large numbers of dogs during the coronavirus disease 2019 pandemic, when so-called "pandemic pups" were becoming popular because of social distancing at the time.
Phylogenetic and Bayesian analyses suggest that canine H3N2 viruses died out in South Korea around 2017 but became established in China, after which they appear to have seeded recent North American outbreaks, with two or three introductions in the past 3 years (22). Genomic epidemiology conducted by Dr. Parrish's team confirms that within North America, the H3N2 virus spreads very rapidly among dogs in kennels and shelters in different regions but then dies out locally, likely once the animals have become infected and immune. Household dogs in North America appear not to be connected enough to maintain the virus in circulation, so sustaining the epidemic requires the movement of the virus to more distant dog populations with dense populations where the virus can spread (22). Moreover, the strain is evolving at a constant rate, consistent with other influenza viruses, and has not yet gained properties that would keep the virus in prolonged circulation among dogs (22).
Sticking with the topic of influenza, Dr. Erica Spackman (Distinguished Senior Research Scientist, Exotic & Emerging Avian Viral Diseases Research, US National Poultry Research Center) gave a talk on the H5N1 IAV situation. Typically, low-pathogenicity avian influenza viruses can spill over from ducks to poultry populations. If poultry become infected with H5 or H7 subtypes, the disease may become highly pathogenic avian influenza (HPAI) for reasons that are poorly understood. HPAI can cause 100% mortality in infected poultry flocks, which can be rapid and accompanied by clinical signs, such as lethargy, neurological signs, and hemorrhagic lesions in chickens in wattles, combs, legs, and internal organs (24). Since H5N1 was first detected in wild birds in the US at the end of 2021, many spillover events have occurred from wild waterfowl into poultry, causing significant losses to the poultry industry (at the time of writing, 168.26 million poultry in 1,674 flocks have been affected by H5N1 in the USA since the outbreak in poultry began on 8 February 2022 [25]). Then, in early 2024, a reassortant strain emerged in wild birds and transmitted directly to cattle in the US and has since spread (26) (at the time of writing, 996 dairy herds in 17 states have been affected by H5N1 in the USA since the outbreak in cattle began on 25 March 2024 [27]). While the wild bird H5N1 strains mostly belong to genotype D1.1, the vast majority of cattle have been affected by genotype B3.13. Now, there is a breadth of species infected; the geographic range is wider; and the load of virus is high in resident birds maintaining it locally, so the landscape has changed (at the time of writing, there have been three separate spillover events of H5N1 into cattle, one of genotype B3.13 and two of genotype D1.1 [28]). Moreover, the H5 virus has also transmitted from cattle to poultry, and now poultry continues to be infected from the bovine B3.13 genotype in addition to aquatic waterfowl D1.1 genotype (29). Ducks shed higher titers of the virus for a longer time compared to gallinaceous avian species. In one study, chickens shed virus for 2-3 days post-exposure (24), whereas in another study, ducks shed virus for the duration of the experiment (14 days) (30). Ducks also have a wide geographic distribution, which could increase the exposure of other species, enhancing the potential for spillover into a broader range of hosts. Currently, the focus of control efforts has been to preserve the US export of poultry and dairy products, but open questions remain. For example, do we eradicate, do we depopulate, or do we vaccinate? The goals can be different depending on the industry, and it is important to note that vaccines only work well when there is good biosecurity in place. Several H5 vaccines are licensed in the US: an inactivated vaccine based on a reverse-genetics virus with a backbone comprising lab-strain PR8, an RNA particle vaccine comprising a non-replicating alphavirus vector, two recombinant HVT vectored vaccines, and a fowlpox vectored vaccine, which is not produced in the US anymore (31). However, trade restrictions exist for vaccinated production animals, and vaccinating can be labor-intensive, particularly if given intramuscularly or subcuta neously. Valuable birds have been vaccinated; for example, 140 condors were vaccinated with an inactivated H5 vaccine, which is an example of a vaccine success (32). Whether we vaccinate poultry depends on several factors, including the cost benefit, and whether a test can be developed to ensure differentiation of infected from vaccinated animals (at the time of writing, an inactivated H5 vaccine has been conditionally approved by the USDA for use in chickens [33], but a decision to vaccinate US flocks has not yet been made).
## H5N1 ONLINE DISCUSSION 12/11/24
Six months after the symposium, ASV hosted an online discussion on the ongoing H5N1 outbreak. This was the first online content delivered for its membership. Dr. Andrew Bowman (Professor, The Ohio State University) and Dr. Kay Russo (Veterinarian, RSM Consulting) gave updates. Dr. Russo gave a brief overview of the outbreak: in early March 2024, she received a call from dairy veterinarians who had discovered that some cattle had respiratory signs and fever, and that rumination had abruptly ceased. Furthermore, many animals had mastitis that was not associated with bacterial infection. After discovering that wild birds around the dairies were dying from H5N1, Dr. Russo suggested that the cattle also be tested for the virus, which subsequently came back as positive, and those were the first-ever cases of H5N1 diagnosed in dairy animals. In the field, clinical presentation and mortality percentages vary, with some herds having asymptomatic infection, whereas in other herds, up to 20% of the animals have clinical signs (26) for reasons that remain unknown. The virus is shed in the milk to very high levels (26) and is efficiently moving across the country with cattle movement. However, we still do not know how the virus is moving from farm to farm, or even cow to cow. Poultry populations are now at risk from strains circulating in bovine populations, as well as migratory bird populations. Unfortunately, it is still HPAI in poultry, and so as spillover into poultry continues, so too does the depopulation, and as a result, eggs are now expensive and hard to come by.
Dr. Bowman elaborated on the challenge studies his group has done in cows: intranasal challenge led to minimal replication in the respiratory tract and mammary gland and no evidence of transmission to co-housed chickens. In contrast, inter-mam mary challenge was easily accomplished, and within a couple of days, the cows were euthanized for humane reasons. Other in vivo challenge studies have yielded similar results (34,35). Furthermore, it has not been possible to reproduce transmission via contaminated milking equipment under experimental conditions, even though it is one of the leading hypotheses of how the virus is spreading on farms (36). Potentially, it is not a highly efficient route of transmission, and so with the small n numbers seen in experimental studies, it is hard to recapitulate consistently. Consequently, it is not possible to model potential interventions to put in place.
Dr. Russo explained current control strategies: in poultry, if an operation tests positive, the birds are depopulated as humanely as possible, as otherwise they would die from the infection. Next, a control zone is instituted around the operation within a 10 km radius, and everything in and out is tightly controlled. In cattle, there is not the same degree of mortality, so the approach to control it is different. However, as we have a poor understanding of how the virus is moving cow-to-cow, it is difficult to provide farmlevel recommendations for producers. Compounding the issue, there is a lot of cattle movement in the US, creating a perfect situation for spreading the disease. Currently, there are efforts to quarantine farms, but there is still a lot of movement of animals and people on and off these farms, for example, non-lactating animals going to slaughter or dairy workers picking up shifts at poultry barns, etc. Dr. Russo emphasized that this is the first major infection event that the dairy industry has dealt with for decades, and biosecurity programs and protocols still need to be established and optimized. Vaccination of the animals needs to be heavily considered. However, the reason it is not yet implemented is in large part because of the implications on trade, and trade agreements would need renegotiating.
Regarding the situation in people, the majority of cases from dairy farms have been described as mild to moderate, with many patients presenting with conjunctivitis. However, there has been significant underreporting in farm workers. To date, those spillover events have been contained. At the time of writing, there have been 70 human cases of H5N1 in the US, including one death (41 from exposure to dairy herds, 24 from exposure to poultry, two from exposure to other animals, and three where the exposure source is unknown) (37). There has also been concern about people being infected with H5N1 from consuming milk. The majority of American homes have milk in the refrigera tor, so protecting the milk supply is of paramount importance. Abnormal milk, or milk from sick animals, should not be going into the milk supply, so at present, that is waste milk, and, thankfully, several groups have demonstrated that pasteurization is effective in inactivating infectious virus in milk (38)(39)(40)(41). However, despite these interventions, RNA from H5N1 is found in milk bought from the grocery store, and there is a risk of infection from drinking unpasteurized milk. One other way to control infection in people is personal protective equipment (PPE) for farm workers. However, they can make doing the job more challenging, which leads to low adoption. So, there needs to be a middle ground of what really is the most effective PPE and where to focus control efforts.
The panelists then took questions from the audience, with the first being whether cows could act like a "mixing vessel" for the emergence of reassortment strains. Dr. Bowman answered that this remains unknown, but as cattle have not typically been endemically infected with IAVs, there has not been anything to reassort with. However, as H5N1 has been detected in pigs (42), the veterinary public health community is concerned about more pig populations becoming infected with H5N1 and reassortment with endemic swine strains, so the connections between the dairy and swine industries should be further examined. Concerns were also raised from the audience regarding contamination of farm waste. Dr. Russo explained that dumping waste milk into manure lagoons has happened, and it is important to not spread that manure near poultry houses. Another question regarded whether cats were still becoming infected on dairy farms, and whether there has been any evidence of spread to rodents. Dr. Bowman replied that cats seem to be very susceptible and have been found with severe disease, including neurological signs (43,44). Additionally, cats will sometimes disappear and may be found dead later in neighboring fields by farm workers picking vegetables, etc., so raising local awareness when the virus burden is high in the area is important. Dr. Russo expanded, describing how sometimes wildlife services will come in and euthanize different creatures surrounding an operation to test them, and they have found mice positive for H5N1, although how those contribute to the movement of the virus remains unclear, and the cats may be becoming infected by eating the mice or by drinking contaminated milk. One question on immunity referred to whether cattle were protected from reinfection after they recovered from a primary infection. Dr. Bowman stressed that there was no robust data on this yet. Dr. Russo expanded, describing how dairy farmers regularly replace approximately 30% of the herd on average yearly, and whether the continual replenishment of naïve animals into a herd helps perpetuate the virus still needs to be determined. The final question asked whether there have been any challenge studies done in cows to examine the transmission of the virus from cows to other mammalian hosts. Dr. Bowman replied that it was difficult to get cattle into containment to do those kinds of studies, but potentially they could be done in the future.
## CONCLUDING REMARKS
In summary, the symposium united experts across virus families, species, and career stages from academia, government, and industry, thus fostering cross-discipli nary discussion and networking, and the online discussion further connected veteri nary clinicians with researchers. Strengthening these scientific networks benefits US agricultural research, which in turn improves the long-term sustainability of US food systems.
## References
1. Li, Zheng (2025) "Insights and progress on epidemic characteristics, pathogenesis, and preventive measures of African swine fever virus: a review" *Virulence*
2. Goldberg, Langwig, Brown et al. (2024) "Widespread exposure to SARS-CoV-2 in wildlife communities" *Nat Commun*
3. Usda (2025) "First outbreak of foot-and-mouth-disease in Germany since 1988 GAIN UFAS"
4. Peacock, Moncla, Dudas et al. (2025) "The global H5N1 influenza panzootic in mammals" *Nature*
5. Webby, Uyeki (2024) "An update on highly pathogenic avian influenza A(H5N1) virus, clade 2.3.4.4b" *J Infect Dis*
6. Arzt, Sanderson, Stenfeldt (2024) "Foot-and-mouth disease" *Vet Clin North Am Food Anim Pract*
7. Ward, Lasecka-Dykes, Neil et al. (2022) "The RNA pseudoknots in foot-andmouth disease virus are dispensable for genome replication, but Minireview Journal of Virology" *PLoS Pathog*
8. Neil, Newman, Stonehouse et al. (2024) "The pseudoknot region and poly-(C) tract comprise an essential RNA packaging signal for assembly of foot-and-mouth disease virus" *PLoS Pathog*
9. Ward, Lasecka-Dykes, Dobson et al. (2024) "The dual role of a highly structured RNA (the S fragment) in the replication of foot-andmouth disease virus" *FASEB J*
10. Williams, Mettenleiter, Blome (2024) "African swine fever: advances and challenges" *Rev Sci Tech Special Edition*
11. Gladue, Borca (2022) "Recombinant ASF live attenuated virus strains as experimental vaccine candidates" *Viruses*
12. O'donnell, Holinka, Gladue et al. (2015) "African swine fever virus Georgia isolate harboring deletions of MGF360 and MGF505 genes is attenuated in swine and confers protection against challenge with virulent parental virus" *J Virol*
13. Borca, Rai, Espinoza et al. (2023) "African swine fever vaccine candidate ASFV-G-ΔI177L produced in the swine macrophage-derived cell line IPKM remains genetically stable and protective against homologous virulent challenge" *Viruses*
14. Espinoza, Spinard, Rai et al. (2024) "Lyophilization of ASFV vaccine candidate ASFV-G-ΔI177L offers long term stability" *Sci Rep*
15. Dinhobl, Spinard, Birtlery et al. (2025) "African swine fever virus biotype identification tool" *Microbiol Resour Announc*
16. Dinhobl, Spinard, Tesler et al. (2023) "Reclassification of ASFV into 7 biotypes using unsupervised machine learning" *Viruses*
17. Usda (2025) "Poulty & eggs -sector at a glance"
18. Liao, Bajwa, Reddy et al. (2021) "Methods for the manipula tion of herpesvirus genome and the application to Marek's disease virus research" *Microorganisms*
19. Hale, Dennis, Mcbride et al. (2022) "SARS-CoV-2 infection in free-ranging white-tailed deer" *Nature*
20. Mcbride, Garushyants, Franks et al. (2023) "Accelerated evolution of SARS-CoV-2 in free-ranging white-tailed deer" *Nat Commun*
21. Wasik, Rothschild, Voorhees et al. (2023) "Understanding the divergent evolution and epidemiology of H3N8 influenza viruses in dogs and horses" *Virus Evol*
22. Wasik, Damodaran, Maltepes et al. (2025) "The evolution and epidemiology of H3N2 canine influenza virus after 20 years in dogs" *Epidemiol Infect*
23. Voorhees, Dalziel, Glaser et al. (2018) "Multiple incursions and recurrent epidemic fade-out of H3N2 canine influenza A virus in the United States" *J Virol*
24. Spackman, Leyson, Youk et al. (2023) "Pathogenicity in chickens and turkeys of a 2021 United States H5N1 highly pathogenic avian influenza clade 2"
25. Usda (2025) "Confirmations of highly pathogenic avian influenza in commercial and backyard flocks"
26. Caserta, Frye, Butt et al. (2024) "Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle" *Nature*
27. Cdc (2025) "Current situation: bird flu in dairy cows"
28. Usda (2025) "APHIS identifies third HPAI spillover in dairy cattle"
29. Puryear, Runstadler (2024) "High-pathogenicity avian influenza in wildlife: a changing disease dynamic that is expanding in wild birds and having an increasing impact on a growing number of mammals" *J Am Vet Med Assoc*
30. Spackman, Mj, Lee et al. (2023) "The pathogenesis of a 2022 North American highly pathogenic clade 2.3.4.4b H5N1 avian influenza virus in mallards (Anas platyrhynchos)" *Avian Pathol*
31. Spackman, Suarez, Lee et al. (2023) "Efficacy of inactivated and RNA particle vaccines against a North American clade 2.3.4.4b H5 highly pathogenic avian influenza virus in chickens" *Vaccine (Auckl)*
32. Mills (2023) "Protecting the USA's rarest wild birds" *Veterinary Record*
33. Cohen (2025) "U.S. conditionally approves vaccine to protect poultry from avian flu" *Science*
34. Baker, Arruda, Palmer et al. (2025) "Dairy cows inoculated with highly pathogenic avian influenza virus H5N1" *Nature*
35. Halwe, Cool, Breithaupt et al. "4b dynamics in experimentally infected calves and cows" *Nature*
36. Anderer (2024) "Bird flu is primarily transmitted among dairy cattle through milking, study suggests" *JAMA*
37. Cdc (2025) "H5 bird flu: current situation"
38. Alkie, Nasheri, Romero-Barrios et al. (2025) "Effectiveness of pasteurization for the inactivation of H5N1 influenza virus in raw whole milk" *Food Microbiol*
39. Schafers, Warren, Yang et al. (1173) "Pasteurisation temperatures effectively inactivate influenza A viruses in milk" *Nat Commun*
40. Spackman, Anderson, Walker et al. (2024) "Inactivation of highly pathogenic avian influenza virus with high-temperature short time continuous flow pasteurization and virus detection in bulk milk tanks" *J Food Prot*
41. Spackman, Jones, Mccoig et al. (2024) "Characterization of highly pathogenic avian influenza virus in retail dairy products in the US" *J Virol*
42. Cdc (2024) "CDC A(H5N1) bird flu response update"
43. Chothe, Srinivas, Misra et al. (2025) "Distribution of lesions and detection of influenza A(H5N1) virus, clade 2.3.4.4b, in ante-and postmortem samples from naturally infected domestic cats on U.S. dairy farms" *Emerg Microbes Infect* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645935&blobtype=pdf | # N-glycosylation at the receptor binding site drives differences in receptor binding specificity between influenza B virus lineages
Caroline Page, M Mubassir, Pradeep Chopra, Lindsey Gay, Ginger Geiger, Sean Ray, Justin Shepard, Rose Miller, Daniel Perez, Justin Bahl, Josephus Gerardus, Boons, Stephen Tompkins
## Abstract
Receptor specificity plays a critical role in influenza virus host tropism and pathogenesis. Influenza A and influenza B viruses (FLUAVs and FLUBVs, respectively) utilize N-glycans with terminal sialic acids on glycoproteins decorating the surface of mucosal epithelial cells as receptors for the virus hemagglutinin (HA) protein. For FLUAVs, the specificity of HA binding to distinct sialic acid linkages on host glycans is a major determinant of species specificity. Amino acid motifs and N-linked glycosylation sites influencing FLUAV HA receptor specificity are well defined. In contrast, considerably less is known regarding determinants of FLUBV receptor specificity, despite its significant contribution to the global influenza disease burden and unique restriction for human hosts. To address this knowledge gap, we utilize microarrays populated with glycans resembling structures found in the respiratory tract to comprehensively define the receptor binding profiles of FLUBVs from different decades and lineages and confirm these results with complementary virus-glycan binding assays. Using wild-type and reverse genetics FLUBVs having singular mutations in the HA receptor binding site (RBS), as well as structural models, we identify an N-glycosylation site at amino acid 196 within the RBS that determines the breadth of HA binding to terminal sialic acids. The definition of the presence of an N-linked glycan as the primary determinant for FLUBV receptor specificity provides a clear mechanism for lineage-specific differences in HA receptor binding. This may help explain the distinct tropism observed between Victoria and Yamagata lineage FLUBVs and provide insights into the disappearance of the Yamagata lineage.IMPORTANCE Influenza B viruses (FLUBVs) are a major cause of human respiratory disease, but the molecular determinants influencing receptor specificity for the hemagglutinin protein remain largely undefined. We defined the receptor specificity of a panel of Early, Victoria, and Yamagata lineage viruses spanning over 50 years and showed that Victoria lineage viruses can have expanded receptor specificity, compared to Yamagata lineage viruses. We identified a critical N-glycosylation site within the hemagglutinin that regulates hemagglutinin binding to sialic acid receptors on host cells. Recent successful subclades of the Victoria lineage viruses lost this glycosylation site, enabling binding to both human-type and avian-type sialic acid receptors, which may influence respiratory tract tropism. These viruses also showed increased endemic activity. The expanded receptor tropism of Victoria lineage viruses may have fitness benefits, helping to explain epidemiologic features of the lineage, perhaps contributing to the lineage's recent success, while the Yamagata lineage appears to have become extinct.
I nfluenza B viruses (FLUBVs) have circulated exclusively in humans for over 80 years, contributing substantially to the influenza disease burden, particularly in children and the elderly (1)(2)(3)(4). FLUBVs are classified into two antigenically distinct lineages, Victoria and Yamagata, which diverged in the 1970s (5)(6)(7)(8)(9). While the evolutionary dynamics of FLUBVs are complex, their separation is ultimately defined by differences in their hemagglutinin (HA) protein (7). Unlike influenza A viruses (FLUAVs), which infect a broad range of avian and mammalian species and pose a pandemic risk through zoonotic transmission, FLUBVs circulate exclusively in humans, causing only epidemics (10)(11)(12). While they lack the antigenic and host diversity of FLUAVs, FLUBVs nonetheless undergo antigenic variation through reassortment with co-circulating viruses and the accumulation of mutations, which together drive antigenic drift (13,14). These processes contribute to the divergence of the lineages and have been associated with increased epidemic activity (14). Notably, despite co-circulating for nearly four decades, Yamagata viruses have not been detected since March 2020, ultimately leading the World Health Organization to recommend their removal from seasonal influenza vaccine formulations (15,16).
A key determinant of FLUAV and FLUBV receptor specificity is the ability of the HA protein to bind N-acetylneuraminic acid (Neu5Ac) in the receptor binding pocket. Glycoproteins on the cells lining the upper and lower respiratory tracts of host species have poly-N-acetyllactosamine structures terminating with either α2,6 or α2,3 linked sialic acids (Sia), which FLUAVs and FLUBVs variably utilize for attachment and infection (12). The receptor specificity of FLUAVs has been extensively studied to help define factors contributing to host range and the risk of viruses crossing species barriers. In contrast, relatively few studies have defined the receptor specificity and tropism of FLUBVs (17). Wang et al. analyzed glycan-binding properties of FLUBVs using clinical isolates from Taiwan collected between 2001 and 2007. They found that Yamagata viruses exclusively bound to α2,6-linked Neu5Ac, while some Victoria viruses bound both α2,6 and α2,3-linked Neu5Ac, and others bound only α2,6-linked Neu5Ac (18). A correlation between the expanded glycan binding profile of Victoria lineage viruses and year of isolation suggested viral evolution might influence receptor specificity. Despite these observations, most studies defining FLUBV receptor binding focused on earlier viral isolates, often from limited geographic regions, leaving the receptor binding characteristics of recent, globally diverse FLUBV strains largely undefined.
The FLUBV HA protein contains a receptor-binding site (RBS) that includes parts of the 190-helix, 240-loop, and 140-loop (19). Four key residues, namely Phe-95, Trp-158, His-191, and Tyr-202, form the base of the RBS and are conserved across all FLUBVs (20,21). Structural comparisons between the HA of B/Hong Kong/8/73 and influenza A H3 HA (X31) highlighted conserved residues, including Thr-139, Ser-140, and Gly-141, within the 190-helix, and Pro-238 and Ser-240 within the 240-loop, which interact with Neu5Ac (7). Single amino acid mutations within the globular head of the HA can alter receptor specificity and the avidity of the HA-Sia interaction, particularly if the mutation affects a glycosylation site defined as N-X-S/T (22,23). Glycosylation of influenza HA plays an important role in the virus's evolution, enabling immune evasion by shielding antigenic sites, altering receptor binding, and affecting host specificity (24)(25)(26)(27). Experimentally passaging FLUBVs in embryonated chicken eggs induces mutations at residues 194-196, which alter an N-glycosylation site and increase the virus's affinity for α2,3-linked Neu5Ac (28)(29)(30)(31)(32). While this augmentation is artificial, natural differences in glycosylation between the lineages exist, and it remains unknown whether mutations separating the lineages contribute to their differences in sialic acid receptor binding.
Extensive research has characterized how amino acid mutations within the HA, changes in glycosylation sites, and other molecular variations drive shifts in receptor specificity for FLUAVs; however, significant gaps remain in our understanding of these factors for FLUBV receptor specificity (23,33). While structural and predictive models provide valuable insights into the molecular factors affecting FLUBV receptor specificity, experimental data to substantiate these models are limited (19,34,35). To address this knowledge gap, we utilized microarrays populated with glycans that mimic structures found in respiratory tissue to comprehensively define the receptor binding profiles of FLUBVs from different decades, lineages, and clades. Importantly, all the FLUBVs tested were isolated and cultured exclusively in mammalian cell substrates, avoiding the potential for egg-adaptive mutations influencing the results. Using reverse genetics viruses with single-point mutations and molecular dynamics, we examined the role of the 196 N-glycosylation site as a molecular determinant of FLUBV receptor specificity. Our findings show that this site is critical in determining the breadth of Victoria and Yamagata lineage binding to terminal sialic acid conformations. This data enhances our understanding of FLUBV glycan-host interactions and provides a clear mechanism for lineage-specific differences in HA receptor binding.
## RESULTS
## Genetic variation affecting the receptor-binding domain of influenza B viruses
The phylogenetic analysis of original or cell-passaged FLUBV HA sequences isolated between 1960 and 2024 revealed distinct evolutionary N-glycosylation patterns at residue 196 between the Early, Victoria, and Yamagata lineage viruses (Fig. 1A). While asparagine is the dominant residue at position 196 for both Victoria and Yamagata lineages, mutations at this site are more frequently observed in Victoria lineage sequences (Fig. S1). Mutations within the RBS at position 196 and its flanking residues can abolish the N-glycosylation sequon (N-X-S/T), thereby preventing the addition of an N-glycan (Fig. 1B). The appearance of specific variants (mutations N196D and N196E) is clustered closely with the emergence of successful contemporary subclades, such as V1A.3a.1 and V1A.3a.2, for the Victoria lineage (Fig. 1A). These subclades have been associated with increased endemic activity and have been isolated globally, including in East Asia, Southeast Asia, and North America (14). In contrast, substitutions at this position in the Yamagata lineage were observed less frequently, with no single variant rising to dominance.
Our data set of viruses used for glycan microarray analysis, indicated by circles in Fig. 1A, represents an antigenically diverse subset of FLUBVs, including Early viruses that circulated before the lineage split, as well as non-contemporary and contemporary Victoria and Yamagata lineage isolates (Table S1). Sequence analysis of the mammalian cell-propagated FLUBVs selected for microarray analysis revealed patterns consistent with the lineage differences in N-glycosylation at position 196 (Fig. S2), as observed in our phylogenetic analysis of the full data set (Fig. 1A). Our data showed that 25% (1/4) of Early viruses and 47% (7/15) of Victoria wild-type (WT) viruses had mutations at position 196, while all (10/10) Yamagata viruses retained the N-glycosylation motif (Fig. 1C; Fig. S2). Taken together, these data highlight distinct evolutionary glycosylation patterns at residue 196 for WT viruses from each lineage, independent of egg adaptations. Muta tions at this site, potentially selected for due to their impact on viral fitness, may be contributing to the evolutionary success of recently emerged Victoria lineage subclades.
## Receptor binding profiles of FLUBVs using glycan array technology
Most data regarding FLUBV receptor-binding profiles originated from isolates limited to small geographic regions, with minimal representation across evolutionary clades, leaving gaps in the comprehensive evolutionary characterization of the receptor-binding profiles of FLUBVs. Additionally, recent studies highlighted that, for FLUAV, it is not only the terminal sialic acid but also the structural elements of the glycan that influence binding (36)(37)(38). However, the significance of these factors for FLUBV remains undeter mined.
To address this gap, we collected a globally diverse set of FLUBVs spanning 57 years, exclusively propagated in mammalian cell culture, and assessed their receptor-binding profiles using a glycan microarray designed to replicate structures found in the human respiratory tract (39). The microarray included symmetrical and asymmetrical bianten nary N-glycans with multiple N-acetyl lactosamine (LacNAc) repeating units, capped by either α2,3-or α2,6-sialosides. Glycans are categorized by terminal modifications: α2,3sialosides (1-12, light blue bars), α2,6-sialosides (13-23, dark blue bars), or unmodified galactose-terminating structures (24-26, tan bars) (Fig. 2A).
Cell-cultured virus isolates were applied to the microarray and detected using a human FLUBV HA stem monoclonal antibody (5E04) and a goat anti-human IgG antibody labeled with AlexaFluor-647 (40). The assay was performed in the presence of oseltamivir (OC), a neuraminidase (NA) inhibitor, to avoid interactions between the enzymatic protein and the glycans. The major binding differences observed between Early, Victoria, and Yamagata viruses lie in their ability to accommodate binding to α2,3-sialosides. Notably, only select Early and Victoria isolates exhibited this binding capability, while none of the Yamagata viruses did. Our earliest FLUBV, B/Netherlands/1000/62, along with an early Victoria strain, B/Netherlands/1000/1977, and more contemporary Victoria strains, B/Santiago/51375/2018 and B/Texas/43/2019, exhibited promiscuous binding to α2,3-sialosides and α2,6-sialosides, with varying mono-, di-, and tri-LacNAc repeating units (Fig. 2B andC; Fig. S3A). Other α2,3-sialoside-binding viruses, such as B/ Maryland/15/2016, selectively bound compound 9, which is a symmetric bi-sialylated mono-LacNAc moiety, whereas the less contemporary B/Netherlands/1000/1990 preferentially bound compounds 1 and 2, which are asymmetric mono-sialylated α1,3antenna mono-and di-LacNAc moieties, respectively. Viruses isolated after 1962 consistently bound to α2,6-sialosides, with the strongest responsiveness to the extended symmetric bi-antennary compounds 22 and 23, which represent tri-and tetra-LacNAc moieties at the α1,3-and α1,6-antenna (bottom and top arms, respectively). Across lineages and clades, for α2,6-sialosides, we observed preferential binding to asymmetric di-and tri-LacNAc moieties compared to mono-Lac NAc moieties (compounds 14, 15 and 18, 19 vs 13 and 16). Interestingly, reduced binding was observed when the Neu5Ac was presented at the α1,6-antenna compared to the α1,3-antenna. For example, viruses that bound to the asymmetric mono-LacNAc moiety, such as B/Bolivia/111/2018, lost recognition for this structure if the Neu5Ac was presented at the α1,6-antenna instead of the α1,3-antenna (13 vs 16). A similar pattern was observed with longer di-and tri-LacNAc isomeric structures, where viruses showed stronger responsiveness to compound 14 vs 17 and 15 vs 18, reinforcing the preference for the Neu5Ac presented on the α1,3-antenna. Of the asymmetric α2,6-sialo sides, FLUBV's showed the strongest binding to compounds 15 and 19, which represent a mono-sialylated α1,3-antenna tri-LacNAc and a bi-sialylated α1,3-antenna tri-LacNAc, respectively, highlighting the overall binding preference for longer glycan structures.
Sequence analysis revealed that viruses with dual binding capabilities frequently had mutations at or near the glycosylation site at position 196 (Table 1). No other shared mutations were identified that could be associated with the expansion of receptor specificity (Fig. S2). This suggested that the N-glycosylation site at position 196 may be a critical molecular determinant of receptor specificity and could drive lineage-specific binding differences in FLUBVs.
## Loss of glycosylation site broadens receptor specificity
After characterizing FLUBV receptor-binding patterns and identifying a potential role for the glycosylation site at position 196 (Table 1), we sought to investigate its impact on receptor binding in more detail. Due to its proximity to the RBS, the loss of this specific glycosylation motif may have exposed the receptor-binding pocket, enabling interac tions with α2,3-sialosides in addition to α2,6-sialosides (34). Sequence analysis and binding assays following serial passage of WT FLUBVs in eggs suggested an expanded binding profile associated with mutations at this site (28,29,31,32,41). However, the direct contribution of HA residue 196 to receptor-binding specificity in WT viruses remained to be experimentally validated.
We first examined the receptor-binding profiles of WT viruses (Fig. 3A through C). B/Hawaii/01/2018 (Victoria WT N196) naturally retained the 196 N-glycosylation site and preferentially bound α2,6-sialosides, with minimal α2,3 binding (Fig. 3A). In contrast, B/Santiago/51375/2018 (Victoria WT N196K) contained a naturally occurring mutation at position 196 and bound promiscuously to both α2,3-and α2,6-sialosides of varying LacNAc lengths and structures (Fig. 3B). Due to the rarity of mutations at the 196-glycosy lation site in WT Yamagata viruses, B/Oklahoma/10/2018 (Yamagata WT N196) was the only WT virus used to represent this lineage. This isolate preferentially bound exten ded symmetric bi-antennary compounds 22 and 23 with responsiveness to asymmetric compounds 15 and 19 (Fig. 3C). These WT data demonstrated that natural variation at position 196 can correlate with distinct receptor-binding profiles, but other differences in the remaining six gene segments could also contribute.
To isolate the specific effect of HA residue 196, we generated reverse genetics (RG) viruses on a B/Brisbane/60/2008 backbone, including Victoria RG N196, Victoria RG N196K, Yamagata RG N196, and Yamagata RG N196D (mutation-induced variant not found in WT viruses) (42,43). Figure 3D through G showed that the RG viruses repro duced the binding differences observed in the corresponding WT viruses, confirming that HA residue 196 is the primary determinant of receptor-binding specificity. The RG approach allowed for isolation of the specific role of mutations at the 196 N-glycosyla tion site in determining receptor specificity by eliminating influence from other gene segments, such as NA, which has an impact on receptor binding of some FLUAVs (44).
We further assessed real-time virus-glycan interactions using biolayer interferometry (BLI), a label-free technology. Streptavidin-coated BLI sensors were loaded with biotinyla ted polyacrylamide (PAA) polymers capped with α2,3-or α2,6-sialyl-LacNAc or lactose. Loaded sensors were then dipped into a buffer containing the virus of interest to allow HA binding. Oseltamivir was included to block NA activity. Our Victoria RG N196 virus bound strongly to α2,6-sialyl-LacNAc, with minimal α2,3 binding (Fig. 3H), while mutat ing the glycosylation site led to equal binding with both α2,3 and α2,6 polymers (Fig. 3I). Similarly, Yamagata RG N196 showed strong α2,6 binding, with α2,3 binding comparable to the negative lactose (Fig. 3J). Notably, the Yamagata RG N196D virus exhibited increased binding to α2,3-sialyl-LacNAc, compared to the Yamagata RG N196 virus, while maintaining the α2,6 binding (Fig. 3K). These findings highlight the significant role of single amino acid mutations that disrupt the 196 N-glycosylation motifs in influencing influenza B glycan-virus interactions and shaping receptor-binding properties.
## Molecular dynamics simulations reveal molecular origin of glycan-mediated receptor specificity
Building on experimental data, we performed molecular dynamics (MD) simulations to further explore the role of the 196 N-glycan in receptor binding of FLUBVs with α2,3-and α2,6-sialosides. The HA trimer structures of FLUBV B/Hawaii/01/2018 and B/Oklahoma/10/2018 were modeled using AlphaFold3 (45), with point mutations at residue 196 (N196K and N196D, respectively) introduced using the Rosetta software suite (46). Sialoside receptor analog comprising a linear pentasaccharide with a terminal α2,3-linked Neu5Ac was modeled using the GLYCAM-Web server (47), and the 2,6-linked Full-Length Text conformation was derived from the H3N2 crystal structure (PDB: 6AOV). The confidence of AlphaFold3-predicted structures was assessed using the per-residue confidence metric (pLDDT). Both B/Hawaii/01/2018 (Victoria) and B/Oklahoma/10/2018 (Yamagata) HAs displayed high pLDDT scores, with most residues scoring above 90 for the RBS (Fig. S4). MD simulations were performed on Victoria (B/Hawaii/01/2018) and Yamagata (B/Oklahoma/10/2018) HA trimers in complex with α2,3-and α2,6-linked sialosides, with an N-glycan attached at residue 196. Proximity analysis of sialoside atoms and N-glycan atoms at position N196 revealed distinct interaction patterns. The α2,3-linked sialosides consistently approached the N-glycan at closer distances compared to α2,6-linked sialosides, often entering the hydrogen-bond (H-bond) interaction zone, suggesting stronger interactions between α2,3-linked sialosides and N-glycan for both the Yamagata and Victoria simulations (Fig. 4A andB). Analysis of the H-bond interactions between sialosides and the N-glycan revealed striking differences between α2,3-and α2,6-linked sialosides. For both Victoria and Yamagata lineage viruses, α2,6-linked sialosides rarely formed H-bonds with the N-glycan, whereas α2,3-linked sialosides formed multiple H-bonds consistently with N-glycan over the 500 ns simulation period (Fig. 4C andD). This phenomenon was further supported by representative MD snapshots (Fig. 5A through D). In the Victoria α2,3 simulation (Fig. 5A), the sialoside mostly occupied conformations within close distances to the N-glycan, indicating persistent favorable interactions with the N-glycan. A similar trend emerged for the Yamagata α2,3 simulation (Fig. 5C), where the sialoside remained in close contact with the N-glycan. By contrast, α2,6-linked sialosides in Victoria (Fig. 5B) and Yamagata (Fig. 5D) simulations generally maintained a greater distance than α2,3-linked sialosides.
## Structural interactions between sialosides and HA altered by the presence of N-glycan
To obtain structural insight into how sialoside linkage type modulates glycan binding in Victoria and Yamagata lineage viruses, we extracted a representative snapshot from 1,000 molecular dynamics frames by principal component analysis (PCA) (Fig. S5). In the Victoria-α2,3 complex, 140-loop residues Thr140, Ser141, Gly142, together with 240-loop residues Gln241 and Gly243, form hydrogen bonds to Neu5Ac. This replicates the binding pattern of B/HK HA-LSTa crystal (2RFT), where Neu5Ac made multiple hydrogen bonds with these 140-loop residues (Thr140, Ser141, Gly142) (19). The proximal Gal also hydrogen-bonds to the backbone carbonyl near Pro240, and proximal GlcNAc forms hydrogen-bonds with Ser242, which showed the same exact interactions that were observed for B/HK HA-LSTa crystal complex (2RFT) (19). The Yamagata-α2,3 complex with the intact N-glycan at Asn196 showed almost similar interaction patterns as the Victoria-α2,3 complex, where Neu5Ac is stabilized by Thr139, Ser140, Gly141, and Asp195, while the distal GlcNAc remains sterically blocked by the N-glycan. Notably, in Yamagata, Trp158-Neu5Ac hydrophobic packing is evident. For the human α2,6 sialopentasaccharide (LSTc), our models from representative 500-ns MD frame show that Sia-1 is anchored by multiple hydrogen bonds to the 140-loopbackbone N-H and C=O of Gly142, the side-chain hydroxyl of Ser141, and the backbone carbonyl of Thr140, with additional contacts from the Neu5Ac to Asp195, consistent with the B/HK HA-LSTc crystal (19). Although the crystal 2RFU exhibits fragmented density and limited asialo contacts, we observe stable distal interactions in the complexes, including a recurrent distal-Gal and Lys149 hydrogen bond (Yamagata) and a reproduci ble Gal and backbone-carbonyl interaction at residue Pro240 (Victoria), in line with the weak Gal contact noted in the HA-LSTc crystal complex 2RFU (19). For all the sialopenta saccharides, our models and the previous co-crystals (2RFT, 2RFU) indicate engagement beyond Sia-1 from the nearer Gal, the adjacent GlcNAc, and the distal Glc.
Hydrogen-bond network analysis showed that, in the glycosylated state (N196 present), α2,6-linked sialosides form a markedly denser and more extensive hydrogenbond network than α2,3-linked sialosides in both Victoria and Yamagata HAs. Disruption of the N196 glycosylation site (Victoria N196K; Yamagata N196D) abolishes this disparity, yielding convergent, similarly distributed hydrogen-bonding patterns for α2,3 and α2,6 sialosides (Fig. S6).
Root-mean-square deviations (RMSD) calculation for backbone atoms showed that all systems reached equilibrium at early stages and remained stable over 500 ns, with backbone RMSD values ranging within a few Å (Fig. S7). Root-mean-square fluctuation (RMSF) analysis of HA backbone atoms revealed no significant differences in flexibility between α2,3-and α2,6-linked sialosides (Fig. S8), including the RBS residues, suggesting that the divergent binding preferences are driven by local interaction effects, such as N-glycan-mediated interactions with α2,3 sialosides, rather than global changes in HA dynamics.
## DISCUSSION
In this study, we comprehensively characterized receptor binding across Early, Victoria, and Yamagata lineage FLUBVs, including contemporary strains, using glycan arrays that mimic structures found in the human respiratory tract. Historically, influenza receptor binding was thought to be determined primarily by the terminal sialic acid; however, recent studies have shown that the structural topology of the glycan itself also plays a critical role in receptor binding (37,38,48,49). Our binding studies with symmetri cal and asymmetrical N-glycans revealed that, in addition to the length of the Lac NAc structure, the presentation of Neu5Ac on a specific antenna influenced binding. FLUBV's across lineages and clades preferentially bind extended symmetric bi-antennary compounds sialylated at the α1,3 and α1,6-antennas. These viruses also showed greater responsiveness to α2,6-linked sialosides presented on extended LacNac structures at the α1,3-antenna. In contrast, shorter mono-and di-LacNac structures abolished binding when the epitope was presented on the α1,6-antenna, showing a preference for the α1,3-antenna. This preference aligns with the activity of human sialyltransferase ST6Gal1, which preferentially modifies the α1,3-antenna over the α1,6-antenna on N-linked glycans, indicating an adaptation of FLUBVs to human hosts (40,50). The primary difference between lineages was the expansion of binding profiles in some Early and Victoria lineage viruses. Our findings supported previous research showing that Yamagata viruses primarily bound α2,6-linked sialosides, while Victoria viruses preferentially bound α2,6 but can also adapt to bind α2,3-sialosides. Additionally, early FLUBVs share similar binding preferences to Victoria viruses, with some viruses able to expand from binding only α2,6-sialosides to also bind α2,3-sialosides. Given the closer evolutionary relationship between Early and Victoria viruses, it follows that they exhibit similar binding patterns.
For FLUAVs, the importance of receptor specificity is well-defined, with human viruses binding α2,6-sialosides and avian viruses binding α2,3-sialosides. Mutations within the RBS have been shown to broaden the binding capacity of avian viruses, enabling them to recognize α2,6-sialosides and thereby facilitating interspecies transmission (33,51,52). The importance of receptor specificity for FLUBVs is more nuanced, as some Early and Victoria lineage viruses can bind both α2,6-and α2,3-sialosides, yet these viruses remain restricted to human hosts. This limitation is partly due to FLUBVs' long-standing adaptation to human hosts, which requires specific temperature and pH conditions that are found in the human upper respiratory tract (53). For FLUBVs, receptor specificity may contribute to the observed differences in age-related host susceptibility between the lineages: Victoria viruses typically infect younger children, who have a higher prevalence of α2,3-linked sialosides in their upper respiratory tract, while Yamagata viruses predominantly infect older populations with increased expression of α2,6-linked sialosides in their upper respiratory tract (1,34,(54)(55)(56)(57). Additionally, dual sialoside-bind ing FLUBV viruses have been associated with increased disease severity by causing gastrointestinal illness and lower lung infections, likely due to the higher proportion of α2,3-linked sialosides in these mucosal tissues, enabling disseminated infection (18,(56)(57)(58). While we do not suggest that the broadened receptor specificity of Early and Victoria viruses facilitates cross-species transmission, it is important to consider the broader host implications.
Molecular determinants influencing receptor specificity for FLUBVs remain largely undefined. Sequence analysis of the cell-cultured FLUBVs used in the glycan array studies revealed mutations at the 196 N-glycosylation site in Early and Victoria viru ses, which resulted in expanded binding profiles from α2,6-sialosides to also include α2,3-sialosides (Table 1). These findings suggested that this mutation may be a key determinant of receptor specificity in FLUBVs. Previous structural modeling (19) based on FLUBV HA sequences suggested that N-glycosylation within the RBS may influence receptor binding by physically obstructing the binding pocket (34). Our experimental data support this hypothesis: the loss of the 196 N-glycosylation site broadens recep tor specificity, as shown using a glycan microarray (Table 1; Fig. 2; Fig. S3). This expan ded specificity was further validated using the FLUBV RG system, where RG-generated viruses from both lineages with the mutated N-glycosylation site bound both α2,3and α2,6-sialosides, whereas RG viruses retaining the N-X-S/T glycosylation motif bound exclusively to α2,6-sialosides. These results underscore the critical role of this N-glyco sylation site in modulating FLUBV receptor binding, enabling interaction with both α2,3-and α2,6-linked sialic acids independent of lineage (Fig. 3). While other distinct glycosylation patterns exist, such as the functional glycosylation site at position 233 found only in Victoria lineage viruses (34), the distal location of this glycan from the RBS makes it less likely to play a role in modulating receptor specificity differences between the lineages. For FLUAV, the balance between receptor-binding HA and the enzymati cally active NA is known to significantly impact receptor binding (59). Although we did not explicitly investigate HA:NA balance in FLUBVs, our observation that RG viruses maintained consistent binding profiles regardless of gene segment swaps suggests that HA:NA balance may play a less critical role in the more conserved FLUBVs. Further studies are needed to confirm this hypothesis and explore its implications.
Our results highlighted distinct evolutionary mechanisms between the FLUBV lineages, with Victoria viruses exhibiting greater variability at the 196-glycosylation site, allowing for mutations that abolish the glycosylation motif. In contrast, such mutations are rarely observed in Yamagata viruses (34). We suggest that Victoria viruses have selectively favored these mutations, as shown by our phylogenetic analysis, which indicated that variants with N196 mutations and therefore, broadened sialoside binding, are clustered with successful contemporary subclades, such as V1A.3a.1 and V1A.3a.2. These subclades dominated the post-2021 influenza seasons and were subsequently included in the annual influenza vaccine formulation. In contrast, while isolated N196 mutations occasionally occurred in Yamagata viruses, no single variant has achieved dominance, underscoring the different evolutionary strategies of the two lineages. The limited acquisition of N196 mutations in Yamagata, combined with widespread COVID-19 mitigation measures, such as masking, social distancing, and reduced travel, may have contributed to the disappearance of this lineage from global circulation. Interestingly, our findings showed that while wild-type Yamagata viruses infrequently acquire this mutation, they can accommodate the change, and like the Victoria lineage, it expanded their ability to bind α2,3-linked sialosides. This adaptation could have implications for the tropism and fitness of Yamagata viruses, potentially increas ing susceptibility in younger populations or contributing to more severe symptoms associated with infection.
Our study also provides insights into the structural role of the 196 N-glycan on receptor specificity by integrating molecular modeling, site-directed mutagenesis, and MD simulations. The presence of this N-glycan creates a pronounced difference in interactions that allows for binding of α2,3-linked sialosides with N glycan, while disfavoring binding of α2,6-linked sialosides in both Victoria and Yamagata lineages (Fig. 5). Notably, when interacting with FLUBV HA, the α2,3-linked sialosides adopt a trans conformation, similar to observations with other influenza viruses like avian H5 and avian H3 (19,33,(60)(61)(62). This restricted their flexibility and directed the α2,3-linked sialosides to interact with the N-glycan, which results in stable hydrogen-bond formation with the N-glycan instead of the RBS alpha helix region. In contrast, α2,6-linked sialosides predominantly assumed a cis conformation with enhanced conformational freedom, allowing them to escape interaction with N-glycan and achieve extended engagement with RBS. This Neu5Ac conformation-dependent differential interaction emerged as a major determinant of receptor specificity for FLUBVs (19). The hydrogen-bond network analysis identified interactions between the Neu5Ac and conserved Thr-140, Ser-141, and Gly-142, along with various other residues, corroborating our model while build ing off previous structural analysis. Our 500-ns MD for the α2,6-linked LSTc aligns with crystal complex 2RFU: Sia-1 is anchored by the 140-loop (Thr140/Ser141/Gly142), matching the crystal geometry. For the α2,3-linked LSTa, our complexes align the HA-LSTa crystal complex 2RFT where besides Neu5Ac both proximal and distal Gal and GlcNAc make interactions with receptor binding site residues. These findings underscore the regulatory influence of glycosylation at N196 on receptor specificity, with potential ramifications for viral host age range and pathogenicity. Moreover, our MD simulations revealed that these local receptor-binding perturbations exert minimal impact on the global conformational landscape of the HA protein.
Despite the insights provided by our integrative modeling and simulation approach, several limitations should be noted. Our structural analyses relied on AlphaFold-predic ted HA models with Rosetta-introduced mutations, which, while supported by high local confidence scores, do not replace experimentally determined FLUBV HA-sialoside co-crystal structures. Similarly, the glycan conformations used in docking and MD simulations were derived from modeled or heterologous templates and may not capture the full conformational diversity of host receptors. Although 500 ns production MD simulations showed stable trajectories, longer timescales may be needed to sample slower rearrangements or rare binding events. All computational analyses depend on force fields, parameterization, and modeling assumptions, introducing additional uncertainty. Additionally, while our receptor binding analysis represents a significant improvement over what is currently available, it is heavily biased toward viruses obtained from the Netherlands. This geographic bias may limit the generalizability of our findings. Therefore, our conclusions from structural models, MD simulations, and the data set should be viewed as complementary to, rather than a substitute for, direct experimental validation and broader sampling.
In conclusion, this study provided crucial insights into the receptor specificity of FLUBVs, emphasizing the role of the N-glycosylation site at residue 196 in modulating binding profiles. Our findings highlighted how mutations at this site can expand receptor specificity, allowing for the recognition of both α2,6-and α2,3-linked sialosides. These molecular determinants offer a clear mechanism for lineage-specific differences in HA receptor binding, contributing to a deeper understanding of FLUBV evolution.
## MATERIALS AND METHODS
## Viruses
A collection of 16 influenza B viruses, exclusively passaged in mammalian cells, was provided by Dr. Ron Foucher's lab (Table S1). Additional parental isolates were obtained from the International Reagent Resource (IRR), with passage history and IRR numbers listed in Table S1. Viral stocks were generated in Madin-Darby Canine Kidney (MDCK-ATL; FR-926) cells cultured in Dulbecco's Modified Eagle Medium (DMEM) with 5% fetal bovine serum (FBS). Cells were inoculated with viral dilutions in DMEM without FBS and 1 µg/mL tosylsulfonyl phenylalanyl chloromethyl ketone (TPCK)-treated trypsin and incubated at 35°C with 5% CO₂ for 72 h. Supernatants were collected, aliquoted, and stored at -80°C. All HA sequences were confirmed by next-generation sequencing and are available in the GISAID EpiFlu database (Table S1). Reverse genetics was used to generate four influenza B viruses with an isogenic internal gene and NA background derived from B/Brisbane/60/2008, differing only in their HA (63). HA sequences were derived from B/Oklahoma/10/2018 (wild type), B/Oklahoma/10/2018 (N196D), B/Hawaii/01/2018, and B/Santiago/51375/2018. HA plasmids were synthesized by Twist Bioscience (San Francisco, CA) and cloned into the bidirectional pDP2002 vector (64). Virus generation was performed using a co-culture of HEK293T and MDCK cells seeded at a ratio of 6:1 and transfected with 1 µg of each of the eight plasmids (total 8 µg) using 18 µL of TransIT-LT1 transfection reagent (Mirus Bio LLC, Madison, WI). The transfection mixture was incubated for 45 min and then added to the cells. After overnight incubation, the transfection mixture was replaced with Opti-MEM media containing 1% antibiotic-anti mycotic solution (Life Technologies, Carlsbad, CA). At 24 h post-transfection, the media were supplemented with 1 µg/mL of (TPCK)-treated trypsin (Worthington Biochemicals, Lakewood, NJ). Viral sequences were confirmed by Sanger sequencing, and viruses were propagated in MDCK cells to prepare viral stocks.
## Phylogenetic analysis
All hemagglutinin (HA) sequences and corresponding metadata were obtained from GISAID (65). HA residues are numbered according to the mature protein sequence of B/Brisbane/60/2008 (PDB: 6FYW, Victoria lineage) after cleavage of the signal peptide; position 196 corresponds to the major N-glycosylation site. Sequences spanning the years 1960-2024 were downloaded with the following inclusion criteria: human host, complete HA gene, and original cell-passaged sequences only. For sequences from 1960 to 2000, due to the unavailability of original-passaged complete HA sequences, we included cell-passaged sequences. Sequence alignment was performed using MAFFT (v7) (66), which was implemented in the Seqtron tool (67). Duplicate amino acid sequences were filtered, and the corresponding DNA sequences for the unique amino acid sequences were retained for downstream analyses (6,357 sequences). Maximum likelihood phylogenetic trees were generated using IQ-TREE (v2.2.2) (68) with the best-fit model TVM+F+R4. The tree was then used as input for TreeTime (v0.11) (69) to calibrate the phylogeny to a time scale. We used the Baltic Python package to map and visualize mutation information on the fixed maximum likelihood tree generated by IQ-TREE. We also analyzed all publicly available IBV sequences deposited in GISAID through 31 December 2024 (n = 78,052) and identified the presence or absence of the 196 N-glycosylation site.
## Microarray printing and binding analysis
All N-glycans bear an α-amine at the reducing end asparagine moiety and were printed on amine reactive, NHS-ester activated glass slides (NEXTERION Slide H, Schott Inc.) using a Scienion sciFLEXARRAYER S3 non-contact microarray equipped with a Scienion PDC80 nozzle (Scienion Inc.). Individual samples were dissolved in sodium phosphate buffer (50 µL, 0.225 M, pH 8.5) at a concentration of 100 µM and were printed in replicates of 6 with a spot volume of ~400 pl at 20°C and 50% humidity. Each slide has 24 subarrays in a 3 × 8 layout. After printing, slides were incubated in a humidity chamber for 8 h and then blocked for 0.5 h in a Tris buffer (pH 9.0, 50 mM) containing 5 mM ethanolamine at 50°C. Blocked slides were rinsed with DI water, spun dry, and kept in a desiccator at room temperature for future use.
Binding analysis was performed by incubating the slides with virus isolates at optimized dilution in TSM binding buffer ([TSM-BB]); Tris-HCl 20 mM pH 7.4, NaCl 150 mM, CaCl 2 2 mM, MgCl 2 2 mM, containing 1% BSA and 0.05% Tween-20) in the presence of a neuraminidase inhibitor (Oseltamivir carboxylate, OC, 10 µM) for 2 h at room temperature. Then, slides were washed by sequentially dipping in TSM wash buffer (2 min, containing 0.05% Tween 20), TSM buffer (2 min), and water (2 × 2 min), followed by centrifugation. Next, slides were incubated with a premixed solution of a FLUBV monoclonal antibody 321.05.05.PB.5E04 (0.1 µg/mL) and goat anti-human IgG antibody Alexa Fluor 647 (1 mg/mL) (Jackson ImmunoResearch) in TSM-BB in the presence of OC (10 µM) for 1 h, followed by washing and drying.
The slides were scanned using a GenePix 4000B microarray scanner (Molecular Devices) at the appropriate excitation wavelength with a resolution of 5 µm. Optimum gains and PMT values were employed for the scanning, ensuring that all signals were within the linear range of the scanner's detector and that there was no saturation of signals. The images were analyzed using GenePix Pro 7 software (version 7.2.29.2, Molecular Devices). The data were analyzed with a home-written Excel macro. The highest and the lowest values of the total fluorescence intensity of the replicate spots were removed, and the remaining values were used to provide the mean value and standard deviation. The fluorescence values were plotted using Prism 10 software (GraphPad Software, Inc.); bars represent the mean ± SD for each treatment.
## Biolayer interferometry
BLI analysis was performed using an Octet RED384 system at 30°C. Streptavidin (SA) biosensors were loaded with biotinylated synthetic PAA polymers capped with either α2,3-or α2,6-sialyl-LacNAc or lactose at a concentration of 5 µg/mL for 600 s. WT and RG viruses were diluted to equivalent hemagglutinating units of 128 for Victoria lineage viruses and 32 for Yamagata lineage viruses in TSM-BB containing 0.1% BSA, 0.05% Tween-20, and 10 µM OC. Real-time virus association was monitored for 3,600 s by transferring the loaded sensors into 80 µL dilutions of virus. Dissociation was recorded by immediately moving the sensors into TSM-BB for 1,800 s. Data were analyzed using Octet Data Analysis software, and binding curves were generated in GraphPad Prism.
## Structural modeling and mutagenesis
Although IBV HA co-crystal structures exist (e.g., PDB 2RFT, 2RFU), our Victoria (B/ Hawaii/01/2018) and Yamagata (B/Oklahoma/10/2018) sequences differ substantially from those templates, so we predicted the exact lineage HA trimers using Alpha Fold3 (70). Point mutations at residue 196 were introduced using Rosetta (71): N196K for B/Hawaii/01/2018 and N196D for B/Oklahoma/10/2018. Each mutated structure underwent Cartesian relaxation in Rosetta to refine sidechain packing and backbone geometry. For receptor analogs, we focused on sialosides with five terminal sugar units: DNeup5Aca2-3DGalpb1-4DGlcpNAcb1-3DGalpb1-4DGlcpNAca1-OH for 2,3-linked sialic acid, and DNeup5Aca2-6DGalpb1-4DGlcpNAcb1-3DGalpb1-4DGlcpNAca1-OH for 2,6-linked sialic acid. Our assays used receptor analogs matching host receptors 3′SLN-β1-4 (α2-3) and 6′SLN-β1-4 (α2-6), but no IBV HA co-crystals contain these full five-sugar ligands (IBV crystal structures largely feature shorter α2-6 or α2-3 with β1-3 internal linkages). Accordingly, we extracted α2-6 sialoside from the H3 LSTc complex (PDB 6AOV), and for α2-3, we built the α2-3 3′SLN-β1-4 model with GLYCAM-Web server, then superposed the sialic-acid/Gal anchor onto our Alphafold3 IBV models and applied local minimization before MD (47). For HAs retaining the native N196 residue, a high-mannose type N-glycan (DManpa1-3[DManpa1-6]DManpb1-4DGlcpNAcb1-4DGlcpNAcb1-OH) was appended at the glycosylation site using the glycoprotein builder tool of the GLYCAM server (47). All sialic acid placements were guided by structural superposition of both the HA head receptor-binding domain (residues 71-271) and sialosides to the reference crystal structures (PDBs: 2RFT, 4YYA, and 6AOV) using PyMOL (72). Then those systems containing both protein and sialosides were exported and used as the initial structure for molecular dynamics simulation.
## Molecular dynamics simulations and analysis
Using AmberTools24 (73), a total of eight systems were prepared for MD simulations: four with native N196 glycans with N-glycan attached to the site, and four with N196K or N196D mutations, each bound to either 2,3-or 2,6-linked sialic acid. All the systems were then solvated in an octahedral box of TIP5P water models (74) and were neutral ized with counter-ions. To ensure stable dynamics, a 10-step MD preparation protocol described by Roe and Brooks (75) was adopted. This protocol includes sequential energy minimizations and short MD relaxation steps with gradually reduced restraints, followed by a final NPT simulation until density stabilization. Using Amber24 (76), production runs of 500 ns at 300 K and constant pressure were conducted. For glycans and sialic acid, GLYCAM_06j-1 (77) and for HA protein ff19SB (78) force fields were used. Trajecto ries were analyzed using CPPTRAJ (79) for RMSD, RMSF, and hydrogen-bond networks. Besides, to identify the most representative conformation of the complex, PCA was done on 1,000 frames of the simulation trajectory, and the frame corresponding to the highest density in the projection was extracted for visualization and was examined with the Protein-Ligand Interaction Profiler server (80).
## References
1. Bhat (2020) "Influenza B infections in children: a review" *World J Clin Pediatr*
2. Peltola, Ziegler, Ruuskanen (2003) "Influenza A and B virus infections in children" *Clin Infect Dis*
3. Nerome, Hiromoto, Sugita et al. (1998) "Evolutionary characteristics of influenza B virus since its first isolation in 1940: dynamic circulation of deletion and insertion mechanism" *Arch Virol*
4. Paul Glezen, Schmier, Kuehn et al. (2013) "The burden of influenza B: a structured literature review" *Am J Public Health*
5. Kanegae, Sugita, Endo et al. (1990) "Evolutionary pattern of the hemagglutinin gene of influenza B viruses isolated in Japan: cocirculating lineages in the same epidemic season" *J Virol*
6. Rota, Wallis, Harmon et al. (1983) "Cocirculation of two distinct evolutionary lineages of influenza type B virus since" *Virology (Auckl)*
7. Wang, Cheng, Lu et al. (2008) "Crystal structure of unliganded influenza B virus hemagglutinin" *J Virol*
8. Lindstrom, Hiromoto, Nishimura et al. (1999) "Comparative analysis of evolutionary mechanisms of the hemagglutinin and three internal protein genes of influenza B virus: multiple cocirculating lineages and frequent reassortment of the NP, M, and NS genes" *J Virol*
9. Chen, Guo, Wu et al. (2007) "Exploration of the emergence of the Victoria lineage of influenza B virus" *Arch Virol*
10. Cauldwell, Long, Moncorgé et al. (2014) "Viral determinants of influenza A virus host range" *J Gen Virol*
11. Wilson, Akin, Zhou et al. (1956) "The influenza B virus Victoria and Yamagata lineages display distinct cell tropism and infection-induced host gene expression in human nasal epithelial cell cultures" *Viruses*
12. Long, Mistry, Haslam et al. (2019) "Host and viral determinants of influenza A virus species specificity" *Nat Rev Microbiol*
13. Rosu, Lexmond, Bestebroer et al. (2022) "Substitutions near the HA receptor binding site explain the origin and major antigenic change of the B/Victoria and B/ Yamagata lineages" *Proc Natl Acad Sci*
14. Virk, Jayakumar, Mendenhall et al. (2020) "Divergent evolutionary trajectories of influenza B viruses underlie their contemporaneous epidemic activity" *Proc Natl Acad Sci*
15. Vajo, Torzsa (2022) "Extinction of the influenza B Yamagata line during the COVID pandemic-implications for vaccine composition" *Viruses*
16. Caini, Meijer, Nunes et al. (2023) "Is influenza B/Yamagata extinct and what public health implications could this have? An updated literature review and comprehensive assessment of global surveillance databases"
17. Page, Tompkins (2024) "Influenza B virus receptor specificity: closing the gap between binding and tropism" *Viruses*
18. Wang, Chang, Wang et al. (2012) "Characterization of glycan binding specificities of influenza B viruses with correlation with hemagglutinin genotypes and clinical features" *J Med Virol*
19. Wang, Tian, Chen et al. (2007) "Structural basis for receptor specificity of influenza B virus hemagglutinin" *Proc Natl Acad Sci*
20. Krystal, Elliott, Benz et al. (1982) "Evolution of influenza A and B viruses: conservation of structural features in the hemagglutinin genes" *Proc Natl Acad Sci*
21. Krystal, Young, Palese et al. (1983) "Sequential mutations in hemagglutinins of influenza B virus isolates: definition of antigenic domains" *Proc Natl Acad Sci*
22. Vines, Wells, Matrosovich et al. (1998) "The role of influenza A virus hemagglutinin residues 226 and 228 in receptor specificity and host range restriction" *J Virol*
23. Obadan, Santos, Ferreri et al. (2019) "Flexibility in vitro of amino acid 226 in the receptor-binding site of an H9 subtype influenza A virus and its effect in vivo on virus replication, tropism, and transmission" *J Virol*
24. Gambaryan, Marinina, Tuzikov et al. (1998) "Effects of host-dependent glycosylation of hemagglutinin on receptor-binding properties on H1N1 human influenza A virus grown in MDCK cells and in embryonated eggs" *Virology (Auckl)*
25. Ohuchi, Ohuchi, Feldmann et al. (1997) "Regulation of receptor binding affinity of influenza virus hemagglutinin by its Full-Length Text Journal of Virology November"
26. *J Virol*
28. Munk, Pritzer, Kretzschmar et al. (1992) "Carbohydrate masking of an antigenic epitope of influenza virus haemagglutinin independent of oligosaccharide size" *Glycobiology*
29. Schulze (1997) "Effects of glycosylation on the properties and functions of influenza virus hemagglutinin" *J Infect Dis*
30. Robertson, Naeve, Webster et al. (1985) "Alterations in the hemagglutinin associated with adaptation of influenza B virus to growth in eggs" *Virology (Auckl)*
31. Gambaryan, Robertson, Matrosovich (1999) "Effects of eggadaptation on the receptor-binding properties of human influenza A and B viruses" *Virology (Auckl)*
32. Gambaryan, Tuzikov, Piskarev et al. (1997) "Specification of receptor-binding phenotypes of influenza virus isolates from different hosts using synthetic sialylglycopolymers: non-egg-adapted human H1 and H3 influenza A and influenza B viruses share a common high binding affinity for 6'-sialyl(N-acetyllactosamine)" *Virology (Auckl)*
33. Lugovtsev, Smith, Weir (2009) "Changes of the receptor-binding properties of influenza B virus B/Victoria/504/2000 during adaptation in chicken eggs" *Virology (Auckl)*
34. Saito, Nakaya, Suzuki et al. (2004) "Antigenic alteration of influenza B virus associated with loss of a glycosylation site due to host-cell adaptation" *J Med Virol*
35. Lin, Zhu, Wang et al. (2024) "A single mutation in bovine influenza H5N1 hemagglutinin switches specificity to human receptors" *Science*
36. Vijaykrishna, Holmes, Joseph et al. (2015) "The contrasting phylody namics of human influenza B viruses. eLife 4:e05055"
37. Ni, Mbawuike, Kondrashkina et al. (2014) "The roles of hemagglutinin Phe-95 in receptor binding and pathogenicity of influenza B virus" *Virology (Auckl)*
38. Jia, Byrd-Leotis, Matsumoto et al. (2020) "The human lung glycome reveals novel glycan ligands for influenza A virus" *Sci Rep*
39. Chandrasekaran, Srinivasan, Raman et al. (2008) "Glycan topology determines human adaptation of avian H5N1 virus hemagglutinin" *Nat Biotechnol*
40. Gambaryan, Tuzikov, Pazynina et al. (2008) "6-sulfo sialyl Lewis X is the common receptor determinant recognized by H5, H6, H7 and H9 influenza viruses of terrestrial poultry" *Virol J*
41. Ma, Liu, Eggink et al. (2024) "Asymmetrical biantennary glycans prepared by a stop-and-go strategy reveal receptor binding evolution of human influenza A viruses" *JACS Au*
42. Broszeit, Van Beek, Unione et al. (2021) "Glycan remodeled erythrocytes facilitate antigenic characterization of recent A/H3N2 influenza viruses" *Nat Commun*
43. Chen, Aspelund (2008) "Stabilizing the glycosylation pattern of influenza B hemagglutinin following adaptation to growth in eggs" *Vaccine (Auckl)*
44. Cardenas-Garcia, Caceres, Rajao et al. (2020) "Reverse genetics for influenza B viruses and recent advances in vaccine development" *Curr Opin Virol*
45. Nogales, Perez, Santos et al. (2017) "Reverse genetics of RNA viruses: methods and protocols"
46. Byrd-Leotis, Cummings, Steinhauer (2017) "The Interplay between the host receptor and influenza virus hemagglutinin and neuraminidase" *Int J Mol Sci*
47. Abramson, Adler, Dunger et al. (2024) "Accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
48. Leman, Weitzner, Lewis et al. (2020) "Macromolecular modeling and design in Rosetta: recent methods and frameworks" *Nat Methods*
49. Woods (2018) "Predicting the structures of glycans, glycoproteins, and their complexes" *Chem Rev*
50. Yang, Carney, Chang et al. (2015) "Structure and receptor binding preferences of recombinant human A(H3N2) virus hemagglutinins" *Virology (Auckl)*
51. Peng, De Vries, Grant et al. (2017) "Recent H3N2 viruses have evolved specificity for extended, branched human-type receptors, conferring potential for increased avidity" *Cell Host Microbe*
52. Joziasse, Schiphorst, Van Den Eijnden et al. (1985) "Branch specificity of bovine colostrum CMP-sialic acid: N-acetyllactosaminide alpha 2--6-sialyltransferase. Interaction with biantennary oligosaccharides and glycopeptides of Nglycosylproteins" *J Biol Chem*
53. Good, Fernández-Quintero, Rodriguez et al. (2024) "A single mutation in dairy cow-associated H5N1 viruses increases receptor binding breadth" *Nat Commun*
54. De Graaf, Fouchier (2014) "Role of receptor binding specificity in influenza A virus transmission and pathogenesis" *EMBO J*
55. Pekarek, Weaver (2023) "Existing evidence for influenza B virus adaptations to drive replication in humans as the primary host" *Viruses*
56. Caini, Kusznierz, Garate et al. (2019) "The epidemiological signature of influenza B virus and its B/Victoria and B/ Yamagata lineages in the 21st century" *PLoS One*
57. Sočan, Prosenc, Učakar et al. (2014) "A comparison of the demographic and clinical characteristics of laboratory-confirmed influenza B Yamagata and Victoria lineage infection" *J Clin Virol*
58. Nicholls, Bourne, Chen et al. (2007) "Sialic acid receptor detection in the human respiratory tract: evidence for widespread distribution of potential binding sites for human and avian influenza viruses" *Respir Res*
59. Walther, Karamanska, Chan et al. (2013) "Glycomic analysis of human respiratory tract tissues and correlation with influenza virus infection" *PLoS Pathog*
60. Bell, Severi, Owen et al. (2023) "Biochemical and structural basis of sialic acid utilization by gut microbes" *J Biol Chem*
61. Guo, Rabouw, Slomp et al. (2018) "Kinetic analysis of the influenza A virus HA/NA balance reveals contribution of NA to virus-receptor binding and NA-dependent rolling on receptor-containing surfaces" *PLoS Pathog*
62. Eisen, Sabesan, Skehel et al. (1997) "Binding of the influenza A virus to cell-surface receptors: structures of five hemagglutinin-Full-Length Text Journal of Virology November"
63. "sialyloligosaccharide complexes determined by X-ray crystallography" *Virology (Auckl)*
64. Gamblin, Vachieri, Xiong et al. (2021) "Hemagglutinin structure and activities" *Cold Spring Harb Perspect Med*
65. Shi, Wu, Zhang et al. (2014) "Enabling the "host jump": structural determinants of receptor-binding specificity in influenza A viruses" *Nat Rev Microbiol*
66. Santos, Finch, Sutton et al. (2017) "Development of an alternative modified live influenza B virus vaccine" *J Virol*
67. Nogales, Perez, Santos et al. (2017) "Reverse genetics of influenza B viruses" *Methods Mol Biol*
68. Shu, Mccauley (2017) "GISAID: Global initiative on sharing all influenza data -from vision to reality" *Euro Surveill*
69. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
70. Fourment, Holmes (2016) "Seqotron: a user-friendly sequence editor for Mac OS X" *BMC Res Notes*
71. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol*
72. Sagulenko, Puller, Neher (2018) "TreeTime: maximum-likelihood phylodynamic analysis" *Virus Evol*
73. Abramson, Adler, Dunger et al. (2024) "Addendum: accurate structure prediction of biomolecular interactions with AlphaFold 3" *Nature*
74. Das, Baker (2008) "Macromolecular modeling with rosetta" *Annu Rev Biochem*
75. Delano (2002) "Pymol: an open-source molecular graphics tool" *CCP4 Newsl Protein Crystallogr*
76. Case, Aktulga, Belfon et al. (2023) *AmberTools. J Chem Inf Model*
77. Khalak, Baumeier, Karttunen (2018) "Improved general-purpose five-point model for water: TIP5P/2018" *J Chem Phys*
78. Roe, Brooks (2020) "A protocol for preparing explicitly solvated systems for stable molecular dynamics simulations" *J Chem Phys*
79. Case, Cheatham, Iii et al. (2005) "The Amber biomolecular simulation programs" *J Comput Chem*
80. Kirschner, Yongye, Tschampel et al. (2008) "GLYCAM06: a generalizable biomolecular force field. Carbohydrates" *J Comput Chem*
82. Tian, Kasavajhala, Belfon et al. (2020) "ff19SB: amino-acidspecific protein backbone parameters trained against quantum mechanics energy surfaces in solution" *J Chem Theory Comput*
83. Roe, Cheatham, Iii (2013) "PTRAJ and CPPTRAJ: software for processing and analysis of molecular dynamics trajectory data" *J Chem Theory Comput*
84. Salentin, Schreiber, Haupt et al. (2015) "PLIP: fully automated protein-ligand interaction profiler" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12473128&blobtype=pdf | # The Omicron Wave in Tunisia: Dynamic, Diversity, and Phylogenetic Analyses
Yasmine Chaari, Sondes Haddad-Boubaker, Haifa Khemiri, Wasfi Fares, Anissa Chouikha, Cesare Camma, Alessio Lorusso, Hanen Smaoui, Khaoula Meftah, Ouafa Kallala, Abdelhalim Trabelsi, Amel Chtourou, Awatef Taktak, Olfa Bahri, Manel Hamdoun, Yassine Chaabouni, Henda Touzi, Amel Sadraoui, Zina Meddeb, Nissaf Alaya, Mariem Gdoura, Henda Triki
## Abstract
The SARS-CoV-2 Omicron variant has exhibited a rapid progression around the world, but its molecular insights in North Africa remain understudied. This study characterizes the genetic diversity, dynamics, and evolutionary trends of the Omicron variant in Tunisia over a 33-month period (December 2021-August 2024). In total, 928 high-quality wholegenome sequences were considered in this study, of which 559 were retrieved from the GISAID database and 369 were generated in our laboratory. Phylogenetic analysis of the dominant subvariants (BA.1, BA.2, and BA.5) was performed using IQ-TREE. BA.2 was the predominant subvariant (38%), followed by BA.1 (24.0%), Omicron recombinants (19%), and BA.5 (18%). BA.2 diversified into JN, KP, and BN sub-lineages. Recombinants were dominated by XBB (98.8%), primarily including EG.4, XBB.1.5, and XBB.2.3.11, with rare detection of XDK and XDQ. Phylogenetic analysis revealed local clusters in BA.1, BA.2, and BA.5 alongside imported strains. Tunisia's Omicron wave was mainly driven by BA.2 and its recombinants, with evidence of localized viral evolution and sporadic introductions. The detection of rare recombinants underlines the importance of integrating regional genomic surveillance with epidemiological data in order to help guide future public health strategies.
## 1. Introduction
After its emergence in late 2019 in Wuhan, China, Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) spread quickly around the globe; it took only a few months for a pandemic state to be issued by the World Health Organization (WHO) on 11 March 2020 [1]. As of 22 September 2024, a total of 776,386,491 confirmed Coronavirus Disease (COVID- 19) cases have been reported globally, with over 7 million deaths worldwide [2].
SARS-CoV-2 belongs to the Coronaviridae family, Orthocoronavirinae subfamily, Betacoronavirus genus, and Sarbecovirus subgenus. It is an enveloped single-stranded, positive-sense RNA virus with a spherical shape that contains spike-like projections on its surface, giving it a crown-like appearance [3]. Throughout the pandemic, the SARS-CoV-2 virus has been evolving continuously with a mean substitution rate of 0.6-1.6 10 -3 substitutions per site per year, depending on the variant [4]. SARS-CoV-2 variants were classified by the WHO into three groups: Variants of Concern (VOCs), Variants of Interest (VOIs), and Variants Under Monitoring (VUMs) [5]. To track global SARS-CoV-2 lineage transmission, a nomenclature system known as the Pango lineage nomenclature was developed by Rambaut (2020) [6]. The major VOCs identified so far are Alpha (B.1.1.7), Beta (B.1.351), Gamma (P.1), Delta (B.1.617.2), and Omicron (B.1.1.529), each with distinct protein sequences and varying biological characteristics [7].
The Omicron variant was first reported on 24 November 2021, in Gauteng, South Africa, and was classified as a VOC just two days later [7]. By 16 December 2021, it had spread to 87 countries and was associated with a sharp surge in COVID-19 cases [7]. Compared to earlier VOCs, Omicron accumulated a significantly higher number of mutations [8]. Notably, the triple mutations in the furin-like cleavage site, H655Y, N679K, and P681H, have been linked to its enhanced transmissibility [9], contributing to a faster and more widespread transmission. Additionally, the Omicron variant exhibits a 2-to-2.5-fold greater binding affinity to the angiotensin-converting enzyme 2 (ACE2) receptor than the original SARS-CoV-2 strain, largely due to the T478K, Q493K, and Q498R mutations within its receptor-binding domain (RBD) [10]. Furthermore, Omicron has demonstrated a remarkable ability to evade most virus-neutralizing antibodies, whether induced by vaccination or prior infection with other variants [11].
According to the report published by WHO in April 2022, the five major Omicron subvariants are BA.1, BA.2, BA.3, BA.4, and BA.5 [7].
During the pandemic, the ongoing emergence of these SARS-CoV-2 variants has facilitated co-infection, increasing the likelihood of genetic recombination, a phenomenon achieved when two different strains infect the same cell. This led to the emergence of recombinant subvariants, not only between distinct lineages, such as the Deltacron (XD, XF, and XE), but also within the same lineage [12]. A prominent example is the Omicron XBB recombinant, which stemmed from two second-generation Omicron BA.2 sub-lineages [12].
Despite the global interest in Omicron's evolution, data in Tunisia and North Africa remain limited. Chouikha et al. and Haddad-Boubaker et al. offered a genetic analysis of SARS-CoV-2 evolution in Tunisia over the first 17 months of the pandemic and the Delta wave in Tunisia, respectively [13,14]. In a subsequent study of the Tunisian pediatric population, Khemiri et al. reported Delta as the predominant variant (39.8%) between April 2020 and February 2022, followed by Omicron (24.2%) and Alpha (13.9%) [15].
The present study aims to analyze the Omicron wave epidemiology and its genetic features in Tunisia, to identify predominant subvariants, and to assess their phylogenetic relationships with Omicron strains reported from other regions of the world.
## 2. Materials and Methods
## 2.1. Ethical Statement
This study was approved by the Bio-Medical Ethics Committee of the Pasteur Institute of Tunis, Tunisia (2020/14/I/LR16IPT/V1), on 24 November 2020. It was performed under ethical standards according to the 1964 Declaration of Helsinki and its later amendments. All samples were investigated after de-identification with respect to patient anonymity.
## 2.2. Samples and Viral Genome Sequencing
A total of 369 sequences were obtained and submitted at the laboratory of Clinical Virology of the Pasteur Institute of Tunis. These sequences originated from SARS-CoV-2 positive nasopharyngeal swabs collected between 1 January 2022 and 13 February 2023, corresponding to the Omicron circulation period in Tunisia. Samples were gathered at the Pasteur Institute of Tunis along with various laboratories across Tunisia. All contributors are detailed in Supplementary Table S1. They were transported under refrigerated conditions and processed within 24 h with genome extraction and real-time PCR detection as previously described [14].
Full-genome sequencing was performed using NGS platforms at two centers: the Pasteur Institute of Tunis (Illumina COVID Seq 1000 (San Diego, CA, USA); 339 sequences) and the collaborating Istituto Zooprofilattico Sperimentale dell'Abruzzo e del Molise, Teramo, Italy (NextSeq 500; n = 30). Viral RNA extraction, RT-PCR genotyping, amplification, and sequencing were performed using either commercial kits or in-house protocols as previously described [14,15]. Consensus genome sequences in FASTA format were uploaded to Pangolin (v4.3) for lineage assignment (https://cov-lineages.org (accessed on 20 March 2025)). Only sequences covering at least 85% of the genome were selected for sub-lineage classification. Nextclade v3.16.0 (https://clades.nextstrain.org/ (accessed on 19 March 2025)) was used to assign clades and assess diversity, amino acid changes, and mutation profiles. The identification of SARS-CoV-2 lineages and sub-lineages was performed on the consensus sequences in FASTA format using Pangolin (version 4.3) (https://cov-lineages.org/pangolin.html (accessed on 10 March 2025)).
## 2.3. Viral Sequences
A total of 1194 were retrieved from the Global Initiative on Sharing All Influenza Data (GISAID) database [16] and from GenBank (NCBI) [17]. They included Omicron Tunisian genomic sequences (n = 559) and other worldwide sequences (n = 635).
Concerning the Omicron Tunisian genomic sequences, they were derived from samples collected between December 2021 and August 2024; the selection criteria included sequences from "Africa/Tunisia" and the variant criteria were "Former VOC Omicron GRA (B.1.1.529 + BA.*)". Samples originated from 21 out of 24 Tunisian governorates as detailed in Supplementary Table S1. As for the worldwide sequences, they were randomly distributed as follows: Africa (n = 12), Asia (n = 121), Europe (n = 242), North America (n = 198), Oceania (n = 9), and South America (n = 53), as provided in Supplementary Table S2. To ensure the accuracy and reliability of the genomic analysis, only sequences with less than 10% ambiguous nucleotide positions were considered.
## 2.4. Phylogenetic Analyses
Phylogenetic analysis concerned major Omicron subvariants: BA.1, BA.2, and BA.5. The selected dataset comprised 657 high-quality Omicron sequences (BA.1, BA.2, and BA.5) from Tunisia and 635 worldwide sequences. Sequence alignment was conducted using the MAFFT online platform (version 7) (https://mafft.cbrc.jp/alignment/software/ (accessed on 30 March 2025) [18] with default settings. Maximum Likelihood phylogenetic trees were constructed with IQ-TREE multicore software (v1.6.12) (http://iqtree.cibiv.univie.ac.at/ (accessed on 7 April 2025) [19] employing 1000 bootstrap replicates to evaluate the robustness of the tree topology. The resulting phylogenies were visualized and annotated using FigTree (version 1.4.4) (http://tree.bio.ed.ac.uk/software/figtree/ (accessed on 8 April 2025) [20].
## 3. Results
## 3.1. Epidemiological Features of Collected Samples
This study included 928 Tunisian sequences obtained from 414 males and 514 females, with a sex ratio of 1.24. The ages of the individuals ranged from 18 days to 98 years, with a mean age of 42.82 years and a median age of 44 years. The highest proportion of cases was observed among individuals aged 18-44 years (30.8%), followed by those aged 45-64 years (28.5%) and ≥65 years (18.0%). Pediatric cases were less common, with children aged 1 month to 9 years representing 9.3% of the cohort and adolescents aged 10-17 years accounting for 7.5%. Considering second-generation subvariants, a total of 107 were identified (Figure 2). Within BA.1, the detected lineages included BA.
## 3.2. Variant Assignment
## 3.3. Distribution Timeline of the Omicron SARS-CoV-2 Subvariants
The first Omicron sequence in Tunisia was reported on 2 December 2021. The distribution over time of the main Omicron subvariants is described in Figure 3
## 3.5. Major Omicron Recombinant Mutations
The XBB recombinant issued from reassortment between BA.
## 3.6. Phylogenetic Analysis
The phylogenetic analyses were performed on the most predominant Omicron subvariants in Tunisia: BA.1, BA.2, and BA.5.
## 3.6.1. BA.1
The phylogenetic tree of Omicron BA.1 was constructed using 223 Tunisian sequences belonging to this subvariant along with 226 global ones. The tree revealed multiple mixed clusters comprising both Tunisian and global sequences, along with several collections containing only Tunisian sequences; these collections appeared genetically independent from the global sequences (Figure 7).
## 3.6.2. BA.2
The phylogenetic tree of the Omicron BA.2 subvariant was constructed using 312 Tunisian sequences along with 304 global ones. The tree showed limited intermixing of Tunisian sequences within global clusters and displayed several groups composed exclusively of Tunisian sequences. Tunisian clusters appeared genetically independent from global sequences (Figure 8). The phylogenetic tree of the Omicron BA.5 subvariant was constructed using 122 Tunisian and 105 global ones. The tree showed broadly gathered Tunisian sequences within well-defined clades. The remaining sequences were distributed showing intermixing with global sequences. Notably, the tree also displayed a few elongated branches, suggesting the presence of a higher genetic divergence within some sequences (Figure 9).
## 4. Discussion
The Omicron variant caused the most recent wave of COVID-19 infections and played a key role in the ongoing evolution of the SARS-CoV-2 virus [21]. According to prior research, Omicron harbors a significantly higher number of mutations compared to previous variants [8,11]. Furthermore, Omicron's major subvariants, namely, BA.1, BA.2, BA.3, BA.4, and BA.5, showed an increased ability to escape the neutralization efficiency induced by prior vaccination or infections. Moreover, several Omicron recombinants, XBB, XBD, and XBF, have emerged and shaped this variant landscape [7]. Although Omicron has been well studied in several regions of the world, data from Tunisia were not exhaustive and covered only a few aspects of the variant epidemiology, such as its circulation among the pediatric population [14] or a case report related to a specific sub-lineage like JN.1 [22]. The present retrospective study aims to better analyze the Omicron wave and to gain a broader understanding of its dynamic in the Tunisian general population.
This study covers a 33-month period starting from 2 December 2021, the date of the first Omicron case detection in Tunisia. A diverse array of Omicron subvariants was found: BA.2, BA.1, recombinant lineages, BA.5, BA.4, and the ancestral B.1.1.529 strain, along with their respective derivatives. Notably, BA.3 was not detected, and no BA.3 sequence from Tunisia was available in GISAID up to date, supporting its total absence in the country [15]. Globally and in the USA, Muthusamy et al. reported B.1.1.529, followed by BA.2 and XBB.1, as major subvariants, while in Italy, Bergana et al. reported BA.1 as initially dominant, followed by BA.2 and BA.5 [23,24].
In early 2022, BA.1 became dominant among Tunisian isolates. This lineage contains numerous mutations, such as G339D, S371L, and N501Y, which have been associated with increased transmissibility and immune evasion [25]. The BA.1 phylogenetic analysis showed exclusive Tunisian clustering indicating local transmission, while other clusters displayed intermixing of Tunisian and worldwide sequences. This supports the occurrence of multiple importation events alongside autochthonous spread. Our findings align with previous reports in North Africa that pointed to multiple introductions of the Omicron variant likely originating from England, Scotland, and the United States [26]. By March 2022, BA.1 was supplanted by BA.2, which turned out later to be the most prevalent variant in this study. The BA.2 peak was observed during the boost immunization period (PIP3: 1 December 2021 to 3 March 2022), which is consistent with the hypothesis of increased immune evasion [27]. Additionally, prior studies have linked BA.2 evolution to host immune evasion and adaptation to epidemiological conditions [11,26]. Chatterjee et al. related BA.2's transmissibility to the H78Y mutation, enabling its global dominance [7]. Furthermore, our study found various derivatives of BA.2, such as JN, BN, and KP. Notably, BA.2 derivatives re-emerged long after BA.2's initial peak, and JN subvariant (BA.2.86.1 descendant) frequency began rising in January 2024, despite a significant decline in COVID-19 cases. A Brazilian study identified different JN circulation patterns: its emergence following low Severe Acute Respiratory Infection incidence, co-circulation with XBB.1, and delayed waves post-XBB.1 peaks [28]. Notably, JN.1's S:L455S mutation, previously linked to enhanced viral fitness, may explain JN outcompeting XBB transmissibility in our study [29]. As for the BA.2 phylogenetic tree, it revealed large Tunisian clusters, attributable to the establishment of a localized evolution and a possible adaptation of the circulating strains. The independent Tunisian clades, observed in both BA.1 and BA.2 phylogenetic trees, likely reflect their heavy circulation in early 2022, a period marked by an increased economic activity, school reopening, and relaxed preventive measures. Similar exclusive genetic clusters were previously observed in the Tunisian population during the Delta wave [14].
Another major subvariant in this work was BA.5, along with its derivatives BQ and BE. In contrast, BA.4 was far less frequent, confirming its minor role in the Omicron wave in Tunisia. This aligns with global patterns where BA.5 consistently outcompeted BA.4 due to its higher effective reproduction number (R e ) [30]. Previous studies have attributed BA.5's strong immune evasion to mutations like L452R and F486V in addition to enhanced binding affinity and viral fitness [31]. In our study, BA.5 started peaking around May 2022, and its derivatives, BQ and BE, took over later that year. BA.5's peak coincided with PIP4 (3 March to 1 December 2022), a period characterized by reduced population immunity and relaxed non-pharmaceutical interventions which may have facilitated sustained viral circulation and local diversification of the BA.5 lineage [27]. Phylogenetic analysis of BA.5 showed well-defined clusters, strongly suggesting in-country evolution rather than through importation events. Overall, a similar Omicron subvariant dynamic has been reported in the literature. In Japan, BA.2 was dominant from 25 April to 26 June 2022, followed by BA.5 from 18 July to 25 September 2022 [32], and, in India, second-generation BA.2 lineages circulated extensively during summer 2022 but failed to spread widely in regions where BA.5 was dominant [30].
Another prominent group identified in the present study was the Omicron recombinants group, nearly all belonging to the XBB sub-lineage. Focosi and Maggi documented the identification of over 75 recombinants so far among SARS-CoV-2, ranging from XA to XY [33]. Based on the literature, the high genetic divergence during the Omicron era, along with broad diversity and regional variations, were favorable for co-infection with distinct viral strains, a prerequisite for recombination [34]. The XBB subvariant, as evidenced previously, arose from BA.2.10.1 (BJ.1) and BA.2.75 (BM.1.1.1) and acquired critical mutations like Y144del, V83A, N460K, and F486S. Tamura et al. linked those mutations to an increased viral fitness and immune evasion. They also described XBB as the first SARS-CoV-2 variant to increase its fitness through recombination rather than substitution [35]. In Tunisia, recombinants started emerging in late 2022, similarly to what was reported in Singapore [36], and went on to account for nearly 20% of all Omicron cases. Subsequently, two successive XBB-driven waves were observed during spring and summer 2023. The shift towards BQ and XBB at the time was attributed to their higher R e compared to predecessor BA.5 [35], and this period was referred to as the 'variant soup,' as it was marked by the co-circulation of multiple subvariants [37]. Global trends, where XBB outcompeted earlier Omicron subvariants, supported recombination as a key driver of the virus's evolution [12]. In the present study, the EG lineage (XBB.1.9.2 descendant) was the predominant recombinant, including EG.4, EG.4.5, and EG.13. While in the US, EG.5 ('Eris') caused 20% of the infections in August 2023, its detection was rare in our cohort [38]. Other frequent recombinants were detected in our study: XBB.2.3.11, XBB.1.5, XBB.1.9, and XBB.1.16, along with XBB-derived lineages like FL, GS, and JG. FL.1.5.1's detection is interesting given it was highly reported in the US [39]. In addition to XBB, rare detection of JN.1-derived recombinants (XDK and XDQ) was noted. These recombinants carried the 17MPLF insertion, previously documented to enhance spike compactness and ACE2 binding [40].
It is worth noting that random sampling of the global sequences in the phylogenetic analysis may not capture full subvariant diversity, limiting comparative analyses, and that the absence of clinical data hinders correlation with epidemiological trends.
## 5. Conclusions
The Omicron era in Tunisia was marked by a high diversity of subvariants, shaped by importation events as well as localized evolution. BA.2 was the most prevalent subvariant, and its recombinant descendant, XBB, played a significant role in defining the Omicron wave in Tunisia. The gradual shift toward BA.2 derivatives and recombinants in the wave's later stages highlights the virus's continuous adaptation to host immunity. These findings emphasize the critical importance of sustained genomic surveillance integrated with epidemiological data to anticipate and mitigate the impact of future recombinantdriven outbreaks.
## Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/microorganisms13092162/s1, Table S1: Accession ID and virus name and related details of complete genome SARS-CoV-2 Tunisian sequences analyzed in this study and retrieved from GISAID. Table S2: Accession ID, submitters, and geographical origin of complete genome SARS-CoV-2 sequences used to build phylogenetic trees and retrieved from GenBank.
## Informed Consent Statement:
The parents or legal tutors of children provided informed and written consent to collect samples and data specifically for this study.
## References
1. Cucinotta, Vanelli (2020) "WHO Declares COVID-19 a Pandemic" *Acta Biomed*
2. (2024) "COVID-19 Cases|WHO COVID-19 Dashboard"
3. Wang, Horby, Hayden et al. (2020) "A Novel Coronavirus Outbreak of Global Health Concern" *Lancet*
4. Rouzine (2025) "Evolutionary mechanisms of the emergence of the variants of concern of SARS-CoV-2. Viruses"
5. (2024) "Tracking SARS-CoV-2 Variants"
6. Rambaut, Holmes, O'toole et al. (2020) "A Dynamic Nomenclature Proposal for SARS-CoV-2 Lineages to Assist Genomic Epidemiology" *Nat. Microbiol*
7. Chatterjee, Bhattacharya, Nag et al. (2023) "A Detailed Overview of SARS-CoV-2 Omicron: Its Sub-Variants, Mutations and Pathophysiology, Clinical Characteristics, Immunological Landscape, Immune Escape, and Therapies. Viruses"
8. Das, Samanta, Banerjee et al. (2022) "Is Omicron the End of Pandemic or Start of a New Innings" *Travel Med. Infect. Dis*
9. He, Hong, Pan et al. "SARS-CoV-2 Omicron Variant: Characteristics and Prevention" *MedComm*
10. Shah, Woo "Omicron: A Heavily Mutated SARS-CoV-2 Variant Exhibits Stronger Binding to ACE2 and Potently Escapes Approved COVID-19 Therapeutic Antibodies" *Front. Immunol. 2022*
11. Wang, Møhlenberg, Wang et al. (2023) "Immune Evasion of Neutralizing Antibodies by SARS-CoV-2 Omicron" *Cytokine Growth Factor Rev*
12. Thakur, Thakur, Kumar et al. (2022) "Emergence of Novel Omicron Hybrid Variants: BA(x), XE, XD, XF More Than Just Alphabets" *Int. J. Surg*
13. Chouikha, Fares, Laamari et al. "Molecular epidemiology of SARS-CoV-2 in Tunisia (North Africa) through several successive waves of COVID-19" *Viruses*
14. Haddad-Boubaker, Arbi, Souiai et al. (2023) "The Delta Variant Wave in Tunisia: Genetic Diversity, Spatio-Temporal Distribution, and Evidence of the Spread of a Divergent AY" *Sub-Lineage. Front. Public Health*
15. Khemiri, Mangone, Gdoura et al. (2020) "Dynamic of SARS-CoV-2 Variants Circulation in Tunisian Pediatric Population, During Successive Waves" *Virus Res*
16. Virusbethesda, Md (2004) "National Library of Medicine (US), National Center for Biotechnology Information"
17. Katoh, Rozewicki, Yamada et al. (2019) "Multiple Sequence Alignment, Interactive Sequence Choice and Visualization" *Brief. Bioinform*
18. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Mol. Biol. Evol*
19. Rambaut (2010)
20. González-Candelas, Shaw, Phan et al. (2021) "One Year into the Pandemic: Short-Term Evolution of SARS-CoV-2 and Emergence of New Lineages" *Infect. Genet. Evol*
21. Hamzaoui, Ferjani, Kanzari et al. (2024) "Unveiling the Emergence of SARS-CoV-2 JN.1 Sub-Variant: Insights from the First Cases at Charles Nicolle Hospital" *Tunisia. Acta Microbiol. Immunol. Hung*
22. Muthusami, Saritha (2023) "Exploratory Analysis of SARS-CoV-2 Omicron Variant and Its Subvariant Propagation: Global Predominance of BA.1, BA.2, BA.5, BE.1, and BQ.1" *Proc. Indian Natl. Sci. Acad. Part A Phys. Sci*
23. Bergna, Lai, Sagradi et al. "Genomic Epidemiology of the Main SARS-CoV-2 Variants Circulating in Italy During the Omicron Era" *J. Med*
24. Kumar, Thambiraja, Karuppanan et al. (2022) "Omicron and Delta Variant of SARS-CoV-2: A Comparative Computational Study of Spike Protein" *J. Med. Virol*
25. Menasria, Aguilera "Genomic Diversity of SARS-CoV-2 in Algeria and North African Countries: What We Know So Far and What We Expect? Microorganisms 2022"
26. Abroug, Ouanes-Besbes, Dachraoui et al. (2024) "Impact of Pharmaceutical and Non-Pharmaceutical Interventions on COVID-19 in Tunisia"
27. Tort, Naveca, Nascimento et al. "SARS-CoV-2 Omicron XBB Infections Boost Cross-Variant Neutralizing Antibodies, Potentially Explaining the Observed Delay of the JN.1 Wave in Some Brazilian Regions"
28. Li, Zhang, Lu et al. (2024) "Molecular Epidemiology and Population Immunity of SARS-CoV-2 in Guangdong (2022-2023) Following a Pivotal Shift in the Pandemic" *Nat. Commun*
29. Yajima, Ito, Ueno et al. (2024) "Molecular and Structural Insights into SARS-CoV-2 Evolution: From BA.2 to XBB Subvariants. mBio"
30. Cao, Wang, Jian et al. (2022) "BA.2.12.1, BA.4 and BA.5 Escape Antibodies Elicited by Omicron Infection" *Nature*
31. Nakakubo, Kishida, Okuda et al. (2023) "Associations of COVID-19 Symptoms with Omicron Subvariants BA.2 and BA.5, Host Status, and Clinical Outcomes: A registry-based observational study in Sapporo" *Lancet Infect. Dis*
32. Focosi, Maggi (1239) "Recombination in Coronaviruses, with a Focus on SARS-CoV-2. Viruses"
33. Wang, Iketani, Li et al. (2023) "Alarming antibody evasion properties of rising SARS-CoV-2 BQ and XBB subvariants" *Cell*
34. Tamura, Ito, Uriu et al. (2023) "Virological characteristics of the SARS-CoV-2 XBB variant derived from recombination of two Omicron subvariants" *Nat. Commun*
35. Chia, Young, Chia (2025) "The Omicron-Transformer: Rise of the Subvariants in the Age of Vaccines" *Annals Singapore*
36. Focosi, Quiroga, Mcconnell et al. (2023) "Convergent evolution in SARS-CoV-2 spike creates a variant soup from which new COVID-19 waves emerge" *Int. J. Mol. Sci*
37. Sil, Gautam, Saxena et al. (2025) "Comprehensive Analysis of Omicron Subvariants: EG.5 Rise, Vaccination Strategies, and Global Impact" *EurekaSelect. Available online*
38. Şimşek-Yavuz (2023) "COVID-19: An Update on Epidemiology, Prevention and Treatment" *Infect. Dis. Clin. Microbiol*
39. Chakraborty "Rapid Worldwide Spread of 17MPLF Spike Insertion Mutants (JN.1-JN.1.25, KP.1, KP.2, KQ.1, KR.1, XDD, XDP, XDK, XDQ Subvariants) of Omicron Coronaviruses and Spike Gene 5 ′ -End Sequencing Problem" *SciTe.ai. ResearchSquare*
40. "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=PMC12817905&blobtype=pdf | # Reactive oxygen species promotion drives auranofin's antiviral activity against hepatitis E virus
Kateland Tiller, S Williams, Bo Wang, Debin Tian, Xiang-Jin Meng, James Weger-Lucarelli
## Abstract
Hepatitis E virus (HEV) causes roughly 20 million yearly global infections and is associated with chronic hepatitis, neurological sequelae, and pregnancy-related adverse outcomes that require antiviral intervention. While there are no approved HEV-specific therapeutics, ribavirin and pegylated interferon, prescribed off-label, remain the current standard of care. However, ribavirin resistance and toxicity highlight the unmet clinical need to identify safer, HEV-specific antivirals. Here, we identify reactive oxygen species (ROS) promotion as a previously unrecognized host-directed antiviral mechanism against HEV, revealed through the activity of the FDA-approved drug auranofin. Auranofin, which is known to elevate intracellular ROS, displays antiviral activity against several viruses. We revealed here that auranofin exhibits robust, dose-dependent antiviral activity against two clinically relevant HEV genotypes and a ribavirin treatment failure-associated mutant. ROS inhibition reversed auranofin-mediated ROS promotion and antiviral activity, establishing a mechanistic link between ROS promotion and antiviral activity. Treatment with D-amino acid oxidase, which breaks down D-amino acids producing the ROS H 2 O 2 , exerted dose-dependent anti-HEV activity. This effect was reversed by ROS inhibition, demonstrating that ROS accumu lation alone is sufficient for antiviral activity. We also revealed that ROS promotion by auranofin drives activation of antioxidant, ER stress, and interferon-stimulated gene expressions, further supporting induction of ROS-dependent antiviral signaling. Lastly, we demonstrated that combined treatment with auranofin and ribavirin exhibits synergistic antiviral activity in vitro. These findings highlight the promotion of ROS as a previously underappreciated host-directed antiviral mechanism and support the repurposing of auranofin-alone or in combination with ribavirin-as a therapeutic strategy against HEV.IMPORTANCE Hepatitis E virus (HEV) lacks approved virus-specific antiviral therapies, and off-label treatments with ribavirin and pegylated interferon are limited by toxicity and emerging resistance mutants. This study identifies reactive oxygen species (ROS) promotion mediated by the FDA-approved drug auranofin and D-amino acid oxidase as an effective antiviral strategy against multiple genotypes of HEV, including two globally relevant human-associated genotypes and a ribavirin treatment failure-associated HEV mutant. The observed synergistic anti-HEV activity in vitro for combined treatment with both auranofin and ribavirin suggests a potential clinically effective combinational therapeutic approach. ROS promotion through auranofin or other means represents an underexplored antiviral strategy with potential for broad-spectrum activity against a range of viral diseases.
H epatitis E virus (HEV) is a globally distributed virus that has been estimated by the World Health Organization (WHO) to cause approximately 20 million infections every year, with 3.3 million symptomatic infections and approximately 44,000 deaths, making HEV a major cause of acute and chronic viral hepatitis worldwide (1)(2)(3). Most HEV infections in developing countries are acquired through virus-contaminated water during large-scale outbreaks. In contrast, in industrialized countries with good sanitary infrastructure, cases often result from zoonotic transmission through consumption of undercooked or raw animal meat products (4,5). Based on these multiple routes of transmission, it is suggested that the actual global disease burden caused by HEV is greatly underestimated (6,7).
HEV is a single-stranded positive-sense RNA virus belonging to the family Hepeviridae, which consists of two subfamilies: Orthohepevirinae and Parahepevirinae (8). The major HEV genotypes known to infect humans belong to the species balayani in the genus Paslahepevirus. Genotypes 1 and 2 (HEV-1 and HEV-2) exclusively infect humans and are responsible for large-scale outbreaks in developing countries. In contrast, HEV-3 and HEV-4 infect both humans and other animals, causing sporadic cases of zoonotic transmission (9). HEV is a non-enveloped, spherical virus with particles of approximately 30-35 nm in stool samples, although virions circulating in the blood of infected individuals and those produced in cell culture exist as quasi-enveloped particles (10)(11)(12)(13). The HEV genome is approximately 7.2 kb in length, containing three partially overlap ping open reading frames (ORFs): ORF1 encodes non-structural proteins responsible for replication, ORF2 encodes the structural capsid protein, and ORF3 encodes a small protein involved in virus replication and assembly (14)(15)(16)(17).
HEV poses a significant global health burden due to its varied clinical manifesta tions and potential for severe disease in vulnerable populations. While most patients experience a self-limiting, acute infection, HEV infection can also result in chronic and/or deadly outcomes in at-risk groups, including immunocompromised people, pregnant women, or those who have pre-existing liver disease (18)(19)(20). Immunocompromised individuals, such as those who are HIV-positive, solid organ transplant recipients, and patients undergoing chemotherapy treatments, are more likely to develop chronic HEV infections, which can become deadly if liver fibrosis or cirrhosis develops (21)(22)(23). Severe disease outcomes associated with HEV infection disproportionately affect pregnant women and their developing fetuses. HEV infection in the second or third trimesters of pregnancy greatly increases the risk of developing fulminant hepatic failure (FHF) and death (19,24). Vertical transmission has also been reported, resulting in adverse fetal outcomes of preterm delivery, fetal distress, and/or low birth weight (25), although others failed to transmit HEV vertically under experimental conditions (26,27). A significant proportion of HEV-infected individuals develop various neurological sequelae, including Guillain-Barré syndrome and neuralgic amyotrophy (28)(29)(30). Recently, HEV has been recognized as the third leading cause of foodborne viral illness (31), and pork is a leading source of foodborne HEV infections (32)(33)(34). Chronic hepatitis E, HEV-asso ciated neurological complications, foodborne hepatitis E, and the severity of disease outcomes in certain at-risk populations highlight the urgent need for effective HEV-spe cific antivirals.
Currently, no therapeutics are approved for treating hepatitis E, and the only hepatitis E vaccines approved for use are in China (35) and Pakistan (36). Ribavirin and pegylated interferon are used as off-label treatments and represent the current standard of care for treating HEV infection. However, ribavirin resistance has been reported (37,38), and pegylated interferon is associated with significant side effects (39)(40)(41), highlighting the need for safer and more effective virus-specific antivirals to treat hepatitis E. Reflecting this importance, in 2024, the WHO listed HEV-3 as a prototype pathogen for the Hepeviridae family (42). Prototype pathogens are designated to accelerate pandemic preparedness by serving as representative viruses from high-risk families, through which broadly applicable antivirals, vaccines, and diagnostic tools can be developed and later adapted to novel emerging viral threats. Therefore, antiviral testing against HEV is significant not only for identifying urgently needed treatment options for HEV infections, but also for contributing to global preparedness efforts against future viral outbreaks.
Recently, there has been an increasing interest in repurposing FDA-approved drugs as antivirals. This process is faster, safer, and more cost-effective than identifying and developing novel antivirals (43,44). One such FDA-approved drug is auranofin, a gold-based compound used in patients with rheumatoid arthritis. Auranofin has shown promise in treating a wide range of ailments, including viral, bacterial, and fungal infections as well as cancer (45). Notably, it has demonstrated antiviral activity against chikungunya virus (46), human immunodeficiency virus (47,48), and SARS-CoV-2 (49,50). However, auranofin's antiviral mechanism remains poorly defined, and its efficacy against HEV has not been examined. We tested the antiviral activity of auranofin against clinically relevant genotypes of HEV using a human hepatocyte cell line. We found that auranofin displayed dose-dependent antiviral activity against HEV at non-toxic concentrations. Further studies into the antiviral mechanism of action revealed that reactive oxygen species (ROS) mediate auranofin's antiviral activity and that promoting ROS alone displayed robust antiviral activity against HEV. Transcriptional analyses also revealed ROS-promoted modulation of antioxidant, ER stress, and interferon-stimulated pathways, supporting ROS-driven antiviral activity. Furthermore, we also demonstrated that combined treatment with ribavirin and auranofin yields a synergistic antiviral effect, underscoring the potential for more effective combinational therapies for chronic hepatitis E and HEV-associated neurological sequelae. Collectively, our results suggest that ROS promotion is a promising host-directed antiviral strategy against HEV infection.
## RESULTS
## Auranofin displays dose-dependent antiviral activity against HEV-1 Sar55 Gluc replicon and HEV-1 infectious reporter virus Sar55(Hib)
To establish a range of non-toxic concentrations to test auranofin's antiviral activity, cell viability was assessed 72 h post-compound application on Huh7-S10-3 cells via an MTS assay. At concentrations below 2 μM, auranofin displayed low toxicity in Huh7-S10-3 cells. Cell viability decreased at concentrations above 2 μM, resulting in a CC 50 of 2.53 μM (Fig. 1A). Based on these results, we performed subsequent experiments using auranofin at concentrations of 2 μM or below.
To assess auranofin's antiviral activity against HEV, we first tested non-toxic doses against the HEV-1 Sar55 Gluc replicon and the HEV-1 Sar55(Hib) HiBiT-tagged infectious reporter virus that enables luciferase-based detection of extracellular viral particles (51). The Gluc expression (72 h post-inoculation) and HiBiT expression (7 days post-inocula tion) were measured to indicate HEV-1 replication. Auranofin exerts robust dosedependent antiviral activity against the HEV-1 Sar55 Gluc replicon (Fig. 1B) and the HEV-1 Sar55(Hib) HiBiT-tagged infectious virus (Fig. 1C), demonstrating auranofin's antiviral activity against a replicon and infectious HEV-1 system.
## Auranofin also displays antiviral activity against HEV-3 P6 Gluc replicon, HEV-3 infectious reporter virus P6(Hib), and a ribavirin treatment failureassociated HEV-3 P6 G1634R mutant
To determine if auranofin displays antiviral activity against other clinically relevant HEV genotypes, we assessed its activity against a genotype 3 HEV (HEV-3), a zoonotic genotype. As with HEV-1, auranofin displayed dose-dependent antiviral activity against the HEV-3 P6 Gluc replicon at non-toxic concentrations (72 h post-inoculation), resulting in an EC 50 of 0.54 μM (Fig. 1D; Fig. S1). To assess auranofin's efficacy against an HEV-3 P6 infectious virus, we constructed an HiBiT-expressing P6 infectious clone, HEV-3 P6(Hib), using a bacterial-free cloning approach (51)(52)(53). We showed that auranofin also displays dose-dependent antiviral activity against the HEV-3 P6(Hib) infectious virus (Fig. 1E). To determine if auranofin is effective against a ribavirin treatment failure-associated HEV mutant, antiviral testing was performed against an HEV-3 P6 Gluc mutant containing the G1634R mutation, which is associated with ribavirin treatment failure in vivo (37,54). Similar to the WT HEV-3 P6 Gluc and HEV-3 P6 (Hib) infectious virus, the HEV-3 G1634R P6 Gluc mutant displayed susceptibility to auranofin treatment (Fig. 1F). Altogether, these results demonstrated that auranofin is an effective antiviral against replicon and infection systems of two different HEV genotypes, as well as a mutant associated with ribavirin treatment failure.
## ROS inhibitors, N-acetylcysteine (NAC) and dithiothreitol (DTT), reverse auranofin's antiviral activity
Auranofin has been reported to inhibit glutathione peroxidase (GPx) and thioredoxin reductase (TrxR), resulting in increased intracellular ROS levels (55)(56)(57)(58). ROS are chemically active byproducts of cellular metabolism that can play important roles in cell signaling pathways and immune activation (59). Thus, we hypothesized that ROS production may contribute to auranofin's antiviral activity against HEV. To test this hypothesis, we treated cells with common ROS inhibitors, NAC and DTT, in the pres ence of auranofin. NAC reduces ROS by promoting glutathione production, directly scavenging ROS, and modulating redox-signaling pathways (60,61). DTT is thought to neutralize ROS and other free radicals, in addition to protecting against mitochondrial oxidative damage and regenerating glutathione from oxidized glutathione (GSSG) (62,63), suggestive of differential anti-ROS mechanisms for NAC and DTT.
Co-treatment with 10 mM NAC and 2 μM auranofin completely reversed the antiviral activity of 2 μM auranofin against the HEV-3 P6 Gluc replicon (Fig. 2A) and the HEV-3 P6 HiBiT infectious virus (Fig. 2B). To validate the impact on intracellular ROS levels, we used the fluorescent probe H 2 DCFDA to measure general intracellular ROS levels. When compounds and the H 2 DCFDA probe were applied to Huh7-S10-3 cells, we observed an increase in ROS in the presence of 2 μM auranofin (Fig. 2C). This increase was reversed with the combined treatment of 10 mM NAC, mirroring the reversal of antiviral activity. NAC treatment resulted in a robust reversal of baseline ROS, which was normalized to the DMSO control, resulting in negative values. Importantly, DMSO increases ROS, likely accounting for the values below the baseline (64). To test the contribution of ROS in an orthogonal manner, we tested 500 μM DTT in the presence of 2 μM auranofin. Again, we observed a reversal of antiviral activity against the HEV-3 P6 Gluc replicon (Fig. 2D) and the HEV-3 P6 HiBiT infectious virus (Fig. 2E). ROS promotion was also reversed in the presence of DTT (Fig. 2F). Together, these results suggest that ROS promotion is necessary for auranofin's antiviral effects against HEV.
## The ROS promoter D-amino acid oxidase (DAAO) displays antiviral activity against HEV
To directly test whether ROS production alone is sufficient to induce antiviral activity, we tested the ROS promoter DAAO. DAAO specifically promotes hydrogen peroxide (H 2 O 2 ) production through the breakdown of D-amino acids, a ROS-promoting mechanism distinct from auranofin and independent of GPx and TrxR inhibition (65). DAAO displayed strong antiviral activity against HEV-3 P6 Gluc and no cytotoxicity at the doses tested, with a CC 50 of >1,000 µg/mL and an EC 50 of 388.1 μg/mL (Fig. 3A). The antiviral activity (Fig. 3B) and ROS promotion (Fig. 3C) of 200 μg/mL DAAO were reversed by co-treatment with 30 mM NAC. These results identify that ROS promotion alone is sufficient for anti-HEV activity and identify a novel antiviral mechanism against HEV replication.
## Auranofin treatment induces ROS-dependent upregulation of antioxidant, ER stress, and antiviral transcripts
Auranofin treatment can modulate downstream signaling pathways related to oxidative stress, innate immune responses, and inflammation (66)(67)(68)(69). To examine the downstream impacts of ROS promotion via auranofin treatment, we performed reverse transcriptionquantitative polymerase chain reactions (RT-qPCR) on RNA extracted from HEV-3 P6 Gluc-transfected Huh7-S10-3 cells treated with auranofin and/or NAC for 24 h. At this time point, the antiviral activity of auranofin was robust and was reversed by NAC treatment (Fig. 4A). We first tested the expression of antioxidant defense genes associ ated with ROS-induced nuclear factor E2-related factor 2 (Nrf2) oxidative stress defense responses (69,70). NAD(P)H quinone oxidoreductase 1 (NQO1) (Fig. 4B) and heme oxygenase-1 (HMOX1) (Fig. 4C) were significantly upregulated by auranofin treatment, an effect that was reversed by NAC. The upregulation of NQO1 and HMOX1 is consistent with a functional role for ROS activation of the Nrf2 pathway by auranofin. In parallel, guanylate-binding protein 5 (GBP5) (Fig. 4D) and interferon-stimulated gene 15 (ISG15) (Fig. 4E) were also upregulated by auranofin treatment and reversed by NAC treatment. Both ISG15 and GBP5 are established ISGs and antiviral effectors against multiple viruses (71)(72)(73)(74)(75), demonstrating that ROS induced by auranofin activates antiviral signaling.
We also examined transcripts associated with endoplasmic reticulum (ER) stress, as ROS promotion is known to disrupt the tightly regulated redox environment of the ER that is required for proper disulfide bond formation and protein folding, leading to the accumulation of misfolded proteins that can induce ER stress (76,77). We found that ER stress sensors, including glucose-regulated protein 78 (GRP78), also known as BiP (binding immunoglobulin protein) (Fig. 4F), and X-box binding protein 1 (Xbp1) (spliced) (Fig. 4G), were upregulated by auranofin treatment and reversed by NAC. The upregula tion of GRP78 and Xbp1 (spliced) is indicative of the activation of the inositol-requiring enzyme 1 alpha (IRE1α) branch of ER stress response, which attempts to resolve ER stress by reducing the ER folding load and decreasing the amount of misfolded proteins (77). We also examined the upregulation of transcripts associated with the ER-associated degradation (ERAD) response toward misfolded proteins. Transcripts of homocysteineresponsive endoplasmic reticulum-resident ubiquitin-like domain member 1 (HERPUD1) or HERP (Fig. 4H) and ER degradation-enhancing alpha-mannosidase-like 1 (EDEM1) (Fig. 4I) were upregulated by auranofin treatment and reversed by NAC. HERPUD1 and EDEM1 are downstream effectors of ER stress that facilitate ERAD and help clear misfolded proteins or target them for degradation (78)(79)(80). Taken together, these transcriptional perturbations highlight the downstream functional consequences of auranofinmediated ROS promotion and provide evidence for potential antiviral mechanisms of action of ROS promotion.
## Auranofin and ribavirin combined treatment exhibits synergistic antiviral effects
Since the antiviral activity of auranofin and ribavirin is likely through different mecha nisms, we hypothesized that a combined treatment would exert synergistic antiviral activity (81). This is an important consideration because multiple ribavirin-resistant mutants have been identified in patients, and a combined antiviral treatment could prevent the development and further expansion of ribavirin-resistant HEV strains (82), representing a more effective treatment option. We first assessed the cytotoxicity and antiviral activity of ribavirin to determine the optimal concentration range for combined treatment testing. Ribavirin displays stable cell viability in Huh7-S10-3 cells, with a CC 50 > 500 μM, and shows dose-dependent antiviral activity with an EC 50 of 15.41 μM (Fig. 5A).
To determine the potential synergistic antiviral activity, auranofin and ribavirin were tested in combination against the HEV-3 P6 Gluc replicon at concentrations ranging from 0 to 1.5 μM and 0 to 25 μM, respectively. Antiviral data were uploaded to the Synergy Finder web application to identify synergy, defined as a synergy score greater than 10. The highest single agent (HSA) model was used to determine whether the combined effect of the two drugs exceeds the sum of their individual effects. This provides a stronger weight toward lower concentrations, which are less likely to display toxicity (83). We observed a dose-response to combined auranofin and ribavirin treatments against HEV-3 P6 Gluc for each tested concentration combination (Fig. 5B), with average inhibition scores shown. Based on the HSA model, three of the concentration combina tions tested in the study display a synergy score over 10, which can be interpreted as 10% of response beyond the individual compound effects (83) (Table S1), resulting in an overall HSA synergy score of 4.034 ± 2.72 (Fig. 5C). The three combinations that display synergy are 0.5 μM auranofin + 5 μM ribavirin with a synergy score of 13.03 (Fig. 5D), 0.5 μM auranofin + 10 μM ribavirin with a synergy score of 10.24 (Fig. 5E), and 1 μM auranofin + 20 μM ribavirin with a synergy score of 11.26 (Fig. 5F). Overall, this data demonstrates that auranofin and ribavirin display synergistic antiviral activity at select concentrations in vitro.
## DISCUSSION
HEV infection is associated with serious clinical conditions, including chronic infection, fulminant hepatic failure, and neurological sequelae, which require effective antiviral intervention. Due to the emergence of HEV strains resistant to ribavirin (37,38), the current off-label standard of care, it is essential to identify more effective antivirals that act through different mechanisms. Since auranofin is an FDA-approved, ROS-promoting compound shown to be effective against several other viruses (46, 48, 49), we investi gated its potential as an antiviral against HEV while seeking to define its mechanism of action. We demonstrated that auranofin exhibits dose-dependent antiviral activity against HEV replicons and fully infectious reporter virus systems of two different HEV genotypes (HEV-1 and HEV-3), as well as an HEV-3 mutant associated with ribavirin treatment failure. We further showed that auranofin and ribavirin displayed synergis tic antiviral activity, supporting a combined therapeutic strategy to enhance antiviral efficacy and prevent antiviral resistance. We identified that ROS promotion was necessary for the antiviral activity of auranofin, as reversing ROS through two distinct mechanisms reversed its antiviral effects. We then established ROS as sufficient for driving anti-HEV activity by showing that DAAO, an enzyme that produces H 2 O 2 , had robust anti-HEV activity with no toxicity at the doses tested. These findings were functionally suppor ted by upregulated transcripts associated with ROS promotion, Nrf2 activation, ISG induction, ER stress, and ER-associated degradation of misfolded proteins. Altogether, this work has revealed the promotion of ROS as a novel, host-directed antiviral strategy against HEV and highlights the potential for combinational treatment of auranofin with ribavirin to induce synergistic antiviral effects via differential mechanisms. These findings also redefine redox modulation as a deliberate antiviral strategy, establishing a foundation for ROS-promoted, host-targeted antiviral design. Auranofin is well-studied and has both anti-cancer and broad-spectrum anti-patho gen effects (45). Auranofin's anti-cancer activity has been linked to the accumulation of ROS via the inhibition of the redox enzymes, GPx and TrxR, which play important roles in the glutathione and thioredoxin redox systems, respectively (55,84,85). These disulfide reductase systems are antioxidant in nature and function to maintain ROS homeostasis within cells (86). Numerous studies present the connection between the inhibition of these enzymes, the upregulation of ROS, and auranofin's anti-cancer effects (57,87,88). This connection is supported by the finding that NAC, a ROS inhibitor, and auranofin co-treatment reverse the toxicity of auranofin (58,87,88). Here, we found that NAC also reversed auranofin's antiviral activity, a novel finding in the investigation of auranofin's antiviral mechanism of action.
Because auranofin targets host cells to modulate redox pathways, the level of auranofin-dependent ROS production varies across cell lines (89). Levels of ROS promotion are also dependent upon the method of perturbation. For example, auranofin and DAAO induce ROS through different mechanisms, which likely contribute to differences in ROS levels. DAAO breaks down D-amino acids into H 2 O 2 (65), while auranofin has been shown to lead to the production of H 2 O 2 (55,90) and superoxide (O₂•⁻) (58,87,91). This highlights a limitation of this study, as the H 2 DCFDA ROS probe used to measure ROS levels is non-specific. While this probe is commonly used and accepted in auranofin literature (58,87,92,93), the future use of more specific ROS probes, such as HyPer7 (94) or Amplex UltraRed (55), will help to flesh out the type of ROS responsible for auranofin's antiviral activity. To mitigate this limitation, we employed an orthogonal approach to investigate ROS-induced transcriptional pathways, providing evidence for ROS-promoted antiviral activity against HEV. The finding that HMOX1 and NQO1 are upregulated during auranofin treatment (Fig. 4B andC) and reversed by NAC treatment indicates that ROS-induced signaling pathways related to oxidative stress defense are likely activa ted. HMOX1 and NQO1 are antioxidant defense genes that are induced following the activation of Nrf2, a transcription factor that binds to antioxidant response elements (AREs), resulting in increased antioxidant activity in response to oxidative stress (95). It is also well established that auranofin activates Nrf2 and ARE pathways (66,96,97), and several viruses have been shown to alter Nrf2 activity (98)(99)(100). Thus, modulation of Nrf2 signaling pathways has been suggested as a potential antiviral strategy. In parallel with these findings, genes associated with IFN-induced antiviral activity, GBP5, and ISG15 were also upregulated by auranofin treatment. ISG15 is a ubiquitin-like protein that is strongly induced by type I IFN and correlates with antiviral activity against various viruses, including HEV (71,72,101,102). GBP5 is an interferon-induced GTPase that plays a crucial role in defense against various pathogens (73,(103)(104)(105). The induction of ISGs provides further support that auranofin-mediated ROS promotion activates antiviral signaling pathways that may contribute to its activity against HEV.
We also examined the modulation of transcriptional pathways associated with ER stress, as ROS promotion is known to induce stress on the ER through the accumulation of misfolded proteins (106). ROS accumulation in the ER can break disulfide bonds, induce oxidative damage, and cause non-native disulfide bonds to form, leading to misfolding of proteins that accumulate and induce ER stress (107). Our results indicate that genes in the IREα branch of the unfolded protein response (UPR), GRP78, and Xbp1 (spliced) were upregulated, suggesting activation through auranofin-mediated ROS promotion. GRP78, also known as BiP, is an ER chaperone that associates with and inhibits ER stress sensor proteins, such as IRE1α, under redox-balanced conditions (108,109). However, during ER stress and the accumulation of misfolded proteins, GRP78 dissociates from IRE1α, which then splices Xbp1, a transcription factor that regulates genes associated with maintaining ER homeostasis (110,111). The UPR also consists of the ER-associated degradation (ERAD) of misfolded proteins. The upregulation of HERPUD1 and EDEM1 suggests that ERAD is engaged as HERPUD1 helps to recruit and stabilize the ubiquitin-proteasome degradation system, while EDEM1 recognizes and marks misfolded proteins for ERAD (78,80,112,113). The finding that ER stress, UPR, and ERAD-associated transcripts are reversed by NAC suggests that they are activated by ROS, which may contribute to the antiviral activity of ROS. Altogether, these transcriptional modulations support a model in which auranofin-mediated ROS promotion shapes downstream transcriptional responses, including oxidative stress, IFN signaling, and ER stress-related pathways that are associated with antiviral activity.
Many viruses alter ROS levels and ROS-related signaling pathways, making modula tion of ROS a promising antiviral strategy (114,115). Several studies have previously reported auranofin's antiviral activity against other viruses, although the suggested mechanisms vary. For example, auranofin was shown to be effective against chikun gunya virus in vivo. The authors speculated that oxidative folding pathways were involved, but mechanistic studies were not performed to corroborate this (46). Auranofin also impacts human immunodeficiency virus type 1 (HIV-1) reservoirs by inducing a pro-apoptotic effect via a burst of ROS in HIV-infected CD4+ T cells, which are typi cally unaffected by antiretroviral therapy (ART) (48). During a randomized clinical trial, auranofin in conjunction with ART decreased total integrated HIV-1 DNA compared to ART alone, suggesting its efficacy against HIV-1 infection in humans (47). Auranofin also inhibits SARS-CoV-2 replication and reduces cytokine production, supporting both antiviral and immunomodulatory effects (49). Auranofin and similar analogs have also been linked to anti-protease activity against SARS-CoV-2 (50). Collectively, these studies suggest that auranofin can inhibit diverse viruses, although ROS involvement has not previously been demonstrated during acute infection. There is also evidence that ROS themselves can exert antiviral effects against viruses. For instance, treatment with H 2 O 2 inhibits hepatitis C virus replication, an effect reversed by NAC (116). The viral components of phages, including proteins, nucleic acids, and lipids, can also be directly damaged by different types of ROS, resulting in their inactivation (117). Consistent with this, our findings show that DAAO, which preferentially promotes H 2 O 2 , mediates anti-HEV effects, supporting the concept that ROS promotion may be a viable broad-spectrum antiviral strategy.
Further research is needed to investigate the antiviral effects of auranofin and ROS against HEV. While we validated the antiviral effects of auranofin against two important genotypes of HEV (HEV-1 and HEV-3) in vitro, future studies should validate auranofin's efficacy against HEV using an animal model. This represents a challenge because, although immunocompetent small animal models of HEV infection are available, working with them requires overcoming obstacles related to genotype specificity, the lack of recapitulated disease, and high costs associated with animal care and housing (118). Additional work with auranofin could also explore its efficacy against chronic HEV infections, as ribavirin resistance arises in individuals with chronic HEV infections and liver failure (119). Host-targeting antivirals, such as auranofin, are expected to retain efficacy and prevent the emergence of viral resistance (120).
In conclusion, we have identified auranofin's dose-dependent antiviral activity in vitro against two different genotypes of HEV and an HEV mutant associated with ribavirin treatment failure. These antiviral effects were mediated by ROS, as ROS inhibitors reversed the antiviral effects and ROS promotion caused by auranofin treatment. We demonstrated that ROS promotion alone was sufficient for antiviral activity using the ROS promoter DAAO, which displayed strong antiviral activity with no toxicity at the tested doses. This finding further highlights ROS promotion as a novel and promising antiviral strategy against HEV. The ROS-dependent antiviral activity of auranofin was also supported via promotion of transcripts associated with oxidative defense, IFN-stimulated antiviral pathways, and ER stress. Finally, a combined treatment of auranofin and ribavirin demonstrated that synergistic antiviral activity can be achieved at specific concentra tions, providing combinational drug therapeutic potential to enhance antiviral activity and minimize ribavirin resistance. Altogether, these data support the conclusion that auranofin inhibits HEV replication by promoting the production of ROS. While various studies have shown auranofin's antiviral activity, this is the first study to mechanistically link its antiviral activity to its promotion of ROS. Overall, this work lays the foundation for further studies to explore the mechanisms of ROS-promoted antiviral activity, which may represent a broad-spectrum antiviral strategy for treating emerging and re-emerging viral infections. Importantly, HEV-3 has been identified by the WHO as a prototype pathogen, meaning that insights gained from antiviral research against HEV can inform strategies to combat similar and emerging pathogens. Thus, our findings not only advance therapeutic options for treating HEV but also contribute to the broader goal of identifying broad-spectrum antiviral strategies for future outbreaks of novel viruses.
## MATERIALS AND METHODS
## Cell culture
Huh7-S10-3 cells, a sub-clone of human hepatocyte cellular carcinoma cells (Huh7) (121), kindly provided by Suzanne U. Emerson (NIAID, NIH, Bethesda, MD), were cultured in Dulbecco's modified Eagle's medium (DMEM) (Corning) with high glucose, L-glutamine, and sodium pyruvate supplemented with 10% fetal bovine serum (FBS), 1% non-essen tial amino acids, 0.1% gentamicin sulfate, and 25 mM HEPES (herein called DMEM-10). The cells were incubated at 37°C with 5% CO 2 .
## Compounds
Auranofin, kindly provided by Veronica Ghini (Resonance Magnetic Center, Italy), was prepared in dimethyl sulfoxide (DMSO) (Sigma Aldrich) to a stock concentration of 10 mM. Additional auranofin stocks (Selleckchem) were also prepared to 10 mM in DMSO. Ribavirin (Thermo Fisher Scientific) was prepared in molecular grade water to a concentration of 40.95 mM. N-acetylcysteine (NAC) (Thermo Fisher Scientific) was prepared fresh in molecular-grade water to varying concentrations at pH 8, with the addition of sodium hydroxide (NaOH). Dithiothreitol (DTT) (VWR) was purchased at a concentration of 1 M. D-amino acid oxidase (DAAO) (Millipore-Sigma) was prepared in water to a concentration of 22 mg/mL.
## HEV indicator replicons and infectious clones
The HEV-1 Sar55 Gaussia luciferase (Gluc) indicator replicon was constructed using the HEV-1 Sar55 strain backbone (GenBank accession no. AF444002) (122), in which a portion of ORF2 was replaced with the Gaussia luciferase gene (123). Similarly, the HEV-3 P6 Gluc indicator replicon was developed from the HEV-3 strain Kernow-C1/P6 (designated as P6) backbone (GenBank accession no. JQ679013) that was serially passaged six times in cell culture (124,125). Additionally, a G1634R mutation was introduced into the HEV-3 P6 Gluc construct, based on prior findings showing that this mutation significantly enhances HEV-3 replication in vitro and promotes ribavirin resistance in vivo (54). This mutant was constructed using a site-directed mutagenesis system, as previously described ( 126). An HEV-1 infectious luminescence reporter virus Sar55 (Hib), which has a HiBiT tag in the C-terminal of ORF2 of the wild-type Sar55 strain, was generated as previously described (51). The HEV-3 P6 Gluc indicator replicon was kindly provided by Dr. Sue Emerson of the National Institute of Allergy and Infectious Diseases at NIH. The HEV-1 Sar55 Gluc indicator replicon was generously provided by Dr. Alexander Ploss of Princeton University, Princeton, NJ.
## Construction and assembly of HEV-3 P6 HiBiT-expressing infectious clone
An HEV-3 infectious luminescence reporter virus P6(Hib) was generated using the wild-type (WT) HEV-3 P6 infectious clone backbone via a bacterial-free cloning approach to incorporate a HiBiT tag into the C-terminal of the ORF2 (53). Briefly, a plasmid containing a glycine-serine linker (24 nt), HiBiT (33 nt), two stop codons (6 nt), and the last 60 nucleotides of ORF2 was synthesized by Twist Bioscience, as previously described by Nagashima et al. and Tian et al. (51,52). This synthesized sequence also contained overlapping regions in ORF2 and the C-terminus of WT HEV-3 P6 genome that are important for assembly. The insert sequence was amplified via repliQa HiFi ToughMix (Quantabio), and the WT HEV-3 P6 genome was amplified in two fragments using Platinum SuperFi II (Invitrogen). The fragments were gel purified using the NucleoSpin Gel and PCR Clean-up Kit (Macherey-Nagel) and quantified by Qubit. Fragments were assembled in equimolar ratios using the NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs), followed by exonuclease and DpnI treatment to remove unassembled components and residual plasmid template.
## Replicon and infectious clone amplification, in vitro transcription, and transfection
Replicon, infectious clone, or HiFi assembled plasmids were amplified via rolling circle amplification (RCA) as previously described (127). The RCA product was then linearized with the restriction enzyme MluI (for P6-based) or BglII (for Sar55-based) (New England Biolabs). The linearized product was purified using SPARQ PureMag Beads (Quantabio), and in vitro transcription was performed with the mMESSAGE mMACHINE T7 Transcrip tion Kit (Thermofisher Scientific) to produce capped, infectious RNA. Viral RNA was transfected via the jetMESSENGER mRNA transfection reagent (Polyplus) into Huh7-S10-3 cells and incubated with 5% CO 2 at 37°C.
## Cytotoxicity assay
Huh7-S10-3 cells were plated at 5,000 cells/well in a 96-well plate and incubated at 37°C with 5% CO 2 . Compound dilutions were prepared in DMEM-10. The vehicle (DMSO or H 2 O) was also prepared to corresponding concentrations to serve as a control. Growth media was removed from the cells and replaced with the same volume of each compound dilution. Cells were incubated for 72 h before CellTiter 96 AQueous One Solution Reagent was added to each well as recommended by the manufacturer's protocol (Promega CellTiter 96 AQueous One Solution Cell Proliferation Assay). Plates were incubated at 37°C with 5% CO 2 for 2-4 h. After incubation, the absorbance was measured at 490 nm using an Infinite M plate reader. Viability was calculated by subtracting media-only absorbance values and normalizing to the appropriate vehicle control. Cytotoxic concentration 50% (CC 50 ) values were generated via the ED50 Plus v1.0 software (128).
## HEV antiviral assays
Huh7-S10-3 cells were plated at 5,000 cells/well in a 96-well plate. Cells were transfected with HEV replicon-or infectious clone-derived RNA as described above. After a 4-hour incubation, compound dilutions were prepared in DMEM-10. The vehicle (DMSO or H 2 O) was also prepared to a corresponding concentration to serve as a control. Growth media was removed from the cells and replaced with the same volume of each compound dilution. For the HEV-3 P6 Gluc replicon and infectious reporter virus P6(Hib) testing, cells were incubated with the compounds for 72 h. For HEV-1 Sar55 Gluc replicon and infectious reporter virus Sar55(Hib) testing, cells were incubated with the compounds for 7 days. Gaussia luciferase expression was quantified from the supernatant via the Pierce Gaussia luciferase glow assay kit (Thermofisher Scientific). HiBiT activity in the supernatant of HEV-1 Sar55(Hib) and HEV-3 P6(Hib) transfected cells was quantified using the Nano-Glo HiBiT Extracellular Detection System (Promega). Luminescence was measured on an Infinite M Plex multimode microplate reader for both Gluc and HiBiT activities. Antiviral activity was calculated by subtracting the negative transfection-only luminescence values and normalizing to the appropriate vehicle control values. The half-maximal effective concentration (EC 50 ) values were generated via ED50 Plus v1.0 software (128) for select experiments.
## ROS quantification
Huh7-S10-3 cells were plated in 24-well plates to 50,000 cells/well. Compounds were prepared in DMEM-10 to the desired concentrations. The 2' ,7'-dichlorodihydrofluorescein diacetate (H 2 DCFDA) (Fisher Scientific) probe was also added to each compound dilution at a final concentration of 10 μM. Compound dilutions were then applied to the cells, and the plates were incubated for 30 min at 37°C with 5% CO 2 . The cells were then trypsinized, and DMEM-10 was added to create a single-cell suspension. Cells from two wells were pooled for each sample and pelleted at 300 × g for 5 min at 4°C. Then, the cells were washed with PBS, re-pelleted, and finally resuspended in 100 μL of PBS. Samples were analyzed by flow cytometry using the FACSAria Fusion Flow cytometer (BD Biosciences), and the median fluorescent peak was recorded. The data were then analyzed by normalization to the appropriate vehicle control: DMSO for auranofin and media alone for DAAO. Since DMSO increases ROS levels (64), normalization was expected to differ between auranofin and DAAO treatments.
## RNA extraction and reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
24 h at 37°C with 5% CO 2 . After incubation, 500 μL of TRIzol reagent (ThermoFisher) was added to the wells. RNA was extracted following the manufacturer's protocol and stored at -80°C until use. RT-qPCR was performed on the extracted RNA using the Luna Universal One-Step RT-qPCR Kit with SYBR Green (New England Biolabs). Primers were ordered from Integrated DNA Technologies (IDT) and are listed in Table S2 (129)(130)(131)(132). RT-qPCR reactions were performed using a QuantStudio Real-Time PCR System with the following reaction conditions: reverse transcription at 55°C for 20 min, reverse transcription deactivation at 95°C for 5 min, followed by 40 cycles of denaturation at 95°C for 10 s and annealing/extension at 60°C for 60 s. A melt curve was generated at 60°C to 95°C with 0.5°C increments and an interval of 0.05 s. Relative gene expression was calculated by normalizing the Ct values of genes of interest to the housekeeping gene, GAPDH.
## Synergy analysis
Huh7-S10-3 cells were plated at 5,000 cells/well in a 96-well plate. Cells were transfected with HEV-3 P6 Gluc RNA, and compounds were prepared in DMEM-10 to the desired concentrations. Auranofin was tested at concentrations of 0, 0.5, 1, and 1.5 μM, and ribavirin was tested at concentrations of 0, 5, 10, 15, 20, and 25 μM. Each concentration was tested alone and in combination with auranofin and ribavirin, with all possible concentration combinations being tested. Compounds were applied, and 72 h later, Gluc expression was quantified from the supernatant. Antiviral activity was calculated by subtracting the negative transfection only luminescence values and normalizing to the appropriate vehicle control values. Data was uploaded to the SynergyFinder web application (version 3.0) (83). The synergy between auranofin and ribavirin was determined using the HSA model. This model determines if the combined effect of two drugs exceeds the sum of their individual effects; if so, then a synergistic effect exists. Combinations with a synergy score greater than 10 were graphed independently for statistical analysis.
## Statistical analysis
Statistical analyses were performed using Graphpad Prism 10. For initial cytotoxicity and antiviral testing, tested compound concentrations were converted to a log scale, and non-linear regression ([log inhibitor] vs normalized response) was performed. To determine EC 50 and CC 50 values, data were generated via ED50 Plus v1.0 software (128). For all other antiviral and ROS testing, data were analyzed via one-way analysis of variance (ANOVA) using Dunnett's test for multiple comparisons. Combined testing of auranofin and ribavirin was analyzed via the SynergyFinder web application (version 3.0) (83). Select compound combinations were analyzed via one-way ANOVA as previously mentioned.
## References
1. (2015) "Hepatitis E vaccine: WHO position paper" *Weekly Epidemiological Record*
2. Rein, Stevens, Theaker et al. (2012) "The global burden of hepatitis E virus genotypes 1 and 2 in 2005" *Hepatology*
3. Webb, Hr (2019) "Hepatitis E: an underestimated emerging threat" *Ther Adv Infect Dis*
4. Dalton, Kamar, Izopet (2014) "Hepatitis E in developed countries: current status and future perspectives" *Future Microbiol*
5. Meng (2013) "Zoonotic and foodborne transmission of hepatitis E virus" *Semin Liver Dis*
6. Hakim, Wang, Bramer et al. (2017) "The global burden of hepatitis E outbreaks: a systematic review" *Liver Int*
7. Dalton, Kamar, Baylis et al. (2018) "EASL Clinical Practice Guidelines on hepatitis E virus infection" *J Hepatol*
8. Purdy, Drexler, Meng et al. (2022) "ICTV virus Taxonomy profile: Hepeviridae 2022" *J Gen Virol*
9. Wang, Meng (2021) "Hepatitis E virus: host tropism and zoonotic infection" *Curr Opin Microbiol*
10. Balayan, Andjaparidze, Savinskaya et al. (1983) "Evidence for a virus in non-A, non-B hepatitis transmitted via the fecal-oral route" *Intervirology*
11. Nagashima, Takahashi, Kobayashi et al. (2017) "Characterization of the quasi-enveloped hepatitis E virus particles released by the cellular exosomal pathway" *J Virol*
12. Takahashi, Yamada, Hoshino et al. (2008) "Monoclonal antibodies raised against the ORF3 protein of hepatitis E virus (HEV) can capture HEV particles in culture supernatant and serum but not those in feces" *Arch Virol*
13. Yin, Ambardekar, Lu et al. (2016) "Distinct entry mechanisms for nonenveloped and quasi-enveloped hepatitis E viruses" *J Virol*
14. Tam, Smith, Guerra et al. (1991) "Hepatitis E virus (HEV): molecular cloning and sequencing of the full-length viral genome" *Virology (Auckl)*
15. Kenney, Meng (2019) "Hepatitis E virus genome structure and replication strategy" *Cold Spring Harb Perspect Med*
16. Yin, Ying, Lhomme et al. (2018) "Origin, antigenicity, and function of a secreted form of ORF2 in hepatitis E virus infection" *Proc Natl Acad Sci*
17. Ding, Heller, Capuccino et al. (2017) "Hepatitis E virus ORF3 is a functional ion channel required for release of infectious particles" *Proc Natl Acad Sci*
18. (2026) *Full-Length Text Journal of Virology*
19. Kumar Acharya, Sharma, Singh et al. (2007) "Hepatitis E virus (HEV) infection in patients with cirrhosis is associated with rapid decompensation and death" *J Hepatol*
20. Kumar, Beniwal, Kar et al. (2004) "Hepatitis E in pregnancy" *Intl J Gynecology Obste*
21. Buescher, Ozga, Lorenz et al. (2021) "Hepatitis E seroprevalence and viremia rate in immunocompro mised patients: a systematic review and meta-analysis" *Liver Int*
22. Dalton, Bendall, Keane et al. (2009) "Persistent carriage of hepatitis E virus in patients with HIV infection" *N Engl J Med*
23. Kamar, Mansuy, Cointault et al. (2008) "Hepatitis E virus-related cirrhosis in kidney and kidney-pancreas-transplant recipients" *Am J Transplant*
24. Geng, Zhang, Huang et al. (2014) "Persistent hepatitis E virus genotype 4 infection in a child with acute lymphoblastic leukemia" *Hepat Mon*
25. Jilani, Das, Husain et al. (2007) "Hepatitis E virus infection and fulminant hepatic failure during pregnancy: hepatitis E virus and fulminant hepatic failure" *J Gastroenterol Hepatol*
26. Qian, Li, Zhang et al. (2023) "Prevalence of hepatitis E virus and its association with adverse pregnancy outcomes in pregnant women in China" *J Clin Virol*
27. Tsarev, Tsareva, Emerson et al. (1995) "Experimental hepatitis E in pregnant rhesus monkeys: failure to transmit hepatitis E virus (HEV) to offspring and evidence of naturally acquired antibodies to HEV" *J Infect Dis*
28. Guo, Zhou, Sun et al. (2007) "Egg whites from eggs of chickens infected experimentally with avian hepatitis E virus contain infectious virus, but evidence of complete vertical transmission is lacking" *J Gen Virol*
29. Ripellino, Pasi, Melli et al. (2020) "Neurologic complications of acute hepatitis E virus infection" *Neurol Neuroimmunol Neuroinflamm*
30. Kamar, Bendall, Peron et al. (2011) "Hepatitis E virus and neurologic disorders"
31. Van Den Berg, Van Der Eijk, Pas et al. (2014) "Guillain-Barré syndrome associated with preceding hepatitis E virus infection" *Neurology (ECronicon)*
32. (2023) "Joint FAO/WHO Expert Meeting on microbiological risk assessment of viruses in food Part 1: food attribution, analytical methods, and indicators"
33. Ji, Li, Jia et al. (2021) "Estimating the burden and modeling mitigation strategies of porkrelated hepatitis E virus foodborne transmission in representative European countries" *One Health*
34. García, Hernández, Gutierrez-Boada et al. (2017) "Occurrence of hepatitis E virus in pigs and pork cuts and organs at the time of slaughter" *Front Microbiol*
35. Renou, Roque-Afonso, Pavio (2014) "Foodborne transmission of hepatitis E virus from raw pork liver sausage" *France. Emerg Infect Dis*
36. Zhu, Zhang, Zhang et al. (2010) "Efficacy and safety of a recombinant hepatitis E vaccine in healthy adults: a large-scale, randomised, double-blind placebo-controlled, phase 3 trial" *Lancet*
37. Hartley, Wasuwanich, Van et al. (2024) "Hepatitis E vaccines updates" *Vaccines (Basel)*
38. Debing, Ramière, Dallmeier et al. (2016) "Hepatitis E virus mutations associated with ribavirin treatment failure result in altered viral fitness and ribavirin sensitivity" *J Hepatol*
39. Wang, Mahsoub, Li et al. (2023) "Ribavirin treatment failure-associated mutation, Y1320H, in the RNA-dependent RNA polymerase of genotype 3 hepatitis E virus (HEV) enhances virus replication in a rabbit HEV infection model" *mBio*
40. Puoti, Babudieri, Rezza et al. (2004) "Use of pegylated interferons is associated with an increased incidence of infections during combination treatment of chronic hepatitis C: a side effect of pegylation?" *Antivir Ther (Lond)*
41. Kowdley (2005) "Hematologic side effects of interferon and ribavirin therapy" *J Clin Gastroenterol*
42. Sulkowski, Cooper, Hunyady et al. (2011) "Management of adverse effects of Peg-IFN and ribavirin therapy for hepatitis C" *Nat Rev Gastroenterol Hepatol*
43. (2024) "Pathogens prioritization: a scientific framework for epidemic and pandemic research preparedness"
44. Martinez (2022) "Efficacy of repurposed antiviral drugs: lessons from COVID-19" *Drug Discov Today*
45. Trivedi, Mohan, Byrareddy (2020) "Drug repurposing approaches to combating viral infections" *J Clin Med*
46. Shen, Shen, Luo et al. (2023) "Molecular mechanisms and clinical implications of the gold drug auranofin" *Coord Chem Rev*
47. Langsjoen, Auguste, Rossi et al. (2017) "Host oxidative folding pathways offer novel anti-chikungunya virus drug targets with broad spectrum potential" *Antiviral Res*
48. Diaz, Shytaj, Giron et al. (2019) "Potential impact of the antirheumatic agent auranofin on proviral HIV-1 DNA in individuals under intensified antiretroviral therapy: results from a randomised clinical trial" *Int J Antimicrob Agents*
49. Chirullo, Sgarbanti, Limongi et al. (2013) "A candidate anti-HIV reservoir compound, auranofin, exerts a selective "anti-memory" effect by exploiting the baseline oxidative status of lymphocytes" *Cell Death Dis*
50. Rothan, Stone, Natekar et al. (2020) "The FDA-approved gold drug auranofin inhibits novel coronavirus (SARS-COV-2) replication and attenuates inflammation in human cells" *Virology (Auckl)*
51. Massai, Grifagni, Santis et al. (2022) "Gold-based metal drugs as inhibitors of coronavirus proteins: the inhibition of SARS-CoV-2 main protease by auranofin and Full-Length Text Journal of Virology"
52. *its analogs. Biomolecules*
53. Tian, Li, Heffron et al. (2025) "Antiviral resistance and barrier integrity at the maternal-fetal interface restrict hepatitis E virus from crossing the placental barrier" *Proc Natl Acad Sci*
54. Nagashima, Primadharsini, Nishiyama et al. (2023) "Development of a HiBiT-tagged reporter hepatitis E virus and its utility as an antiviral drug screening platform" *J Virol*
55. Marano, Cereghino, Finkielstein et al. (2023) "An in vitro workflow to create and modify infectious clones using replication cycle reaction" *Virology (Auckl)*
56. Debing, Gisa, Dallmeier et al. (2014) "A mutation in the hepatitis E virus RNA polymerase promotes its replication and associates with ribavirin treatment failure in organ transplant recipients" *Gastroenterol ogy*
57. Radenkovic, Holland, Vanderlelie et al. (2017) "Selective inhibition of endogenous antioxidants with Auranofin causes mitochondrial oxidative stress which can be countered by selenium supplementation" *Biochem Pharmacol*
58. Wang, Bouzakoura, De Mey et al. (2017) "Auranofin radiosensitizes tumor cells through targeting thioredoxin reductase and resulting overproduction of reactive oxygen species" *Oncotarget*
59. Park, Kim (2005) "The role of p38 MAPK activation in auranofininduced apoptosis of human promyelocytic leukaemia HL-60 cells" *Br J Pharmacol*
60. Cui, Park, Park (2022) "Anti-cancer effects of auranofin in human lung cancer cells by increasing intracellular ROS levels and depleting GSH levels" *Molecules*
61. Sies, Belousov, Chandel et al. (2022) "Defining roles of specific reactive oxygen species (ROS) in cell biology and physiology" *Nat Rev Mol Cell Biol*
62. Aldini, Altomare, Baron et al. (2018) "N-Acetylcysteine as an antioxidant and disulphide breaking agent: the reasons why" *Free Radic Res*
63. Zafarullah, Li, Sylvester et al. (2003) "Molecular mecha nisms of N-acetylcysteine actions" *Cell Mol Life Sci*
64. Rodrigues, De Oliveira, Garcia et al. (2024) "Dithiothreitol reduces oxidative stress and necrosis caused by ultraviolet A radiation in L929 fibroblasts" *Photochem Photobiol Sci*
65. Ma, Li, Tao et al. (2020) "Reductive stress-induced mitochondrial dysfunction and cardiomyopathy" *Oxid Med Cell Longev*
66. Dludla, Jack, Viraragavan et al. (2018) "A dose-dependent effect of dimethyl sulfoxide on lipid content, cell viability and oxidative stress in 3T3-L1 adipocytes" *Toxicol Rep*
67. Spyropoulos, Michel (2024) "D-Amino acid oxidase-derived chemogenetic oxidative stress: unraveling the multi-omic responses to in vivo redox stress" *Curr Opin Chem Biol*
68. Kim, Oh, Park et al. (2010) "Auranofin, a gold(I)-containing antirheumatic compound, activates Keap1/Nrf2 signaling via Rac1/ iNOS signal and mitogen-activated protein kinase activation" *J Pharmacol Sci*
69. Youn, Lee, Saitoh et al. (2006) "Auranofin, as an anti-rheumatic gold compound, suppresses LPS-induced homodimeri zation of TLR4" *Biochem Biophys Res Commun*
70. Hwangbo, Kim, Kim et al. (2021) "Anti-inflammatory effect of auranofin on palmitic acid and LPS-induced inflammatory response by modulating TLR4 and NOX4-mediated NF-κB signaling pathway in RAW264.7 macrophages" *Int J Mol Sci*
71. Fiskus, Saba, Shen et al. (2014) "Auranofin induces lethal oxidative and endoplasmic reticulum stress and exerts potent preclinical activity against chronic lymphocytic leukemia" *Cancer Res*
72. Dunigan, Li, Li et al. (2018) "The thioredoxin reductase inhibitor auranofin induces heme oxygenase-1 in lung epithelial cells via Nrf2-dependent mechanisms" *Am J Physiol Lung Cell Mol Physiol*
73. Lenschow, Lai, Frias-Staheli et al. (2007) "IFN-stimulated gene 15 functions as a critical antiviral molecule against influenza, herpes, and Sindbis viruses" *Proc Natl Acad Sci*
74. Perng, Lenschow (2018) "ISG15 in antiviral immunity and beyond" *Nat Rev Microbiol*
75. Krapp, Hotter, Gawanbacht et al. (2016) "Guanylate binding protein (GBP) 5 is an interferoninducible inhibitor of HIV-1 infectivity"
76. Li, Qu, Liu et al. (2020) "GBP5 is an interferon-induced inhibitor of respiratory syncytial virus" *J Virol*
77. Braun, Hotter, Koepke et al. (2019) "Guanylate-binding proteins 2 and 5 exert broad antiviral activity by inhibiting furin-mediated processing of viral envelope proteins" *Cell Rep*
78. Bhattarai, Riaz, Kim et al. (2021) "The aftermath of the interplay between the endoplasmic reticulum stress response and redox signaling" *Exp Mol Med*
79. Walter (2011) "The unfolded protein response: from stress pathway to homeostatic regulation" *Science*
80. Huang, Chu, Ye et al. (2014) "Role of HERP and a HERPrelated protein in HRD1-dependent protein degradation at the endoplasmic reticulum" *J Biol Chem*
81. Kny, Standera, Hartmann-Petersen et al. (2011) "Herp regulates Hrd1-mediated ubiquitylation in a ubiquitin-like domain-dependent manner" *J Biol Chem*
82. Lamriben, Oster, Tamura et al. (2018) "EDEM1's mannosidase-like domain binds ERAD client proteins in a redox-sensitive manner and possesses catalytic activity" *J Biol Chem*
83. Debing, Emerson, Wang et al. (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*
84. Marimani, Ahmad, Duse (2020) "Combination therapy as an effective tool for treatment of drug-resistant viral infections"
85. Ianevski, Giri, Aittokallio (2022) "SynergyFinder 3.0: an interactive analysis and consensus interpretation of multi-drug synergies across multiple samples" *Nucleic Acids Res*
86. Kudin, Augustynek, Lehmann et al. (2012) "The contribution of thioredoxin-2 reductase and glutathione peroxidase to Full-Length Text Journal of Virology"
87. "2 O 2 detoxification of rat brain mitochondria" *Biochim Biophys Acta*
88. Marzano, Gandin, Folda et al. (2007) "Inhibition of thioredoxin reductase by auranofin induces apoptosis in cisplatin-resistant human ovarian cancer cells" *Free Radic Biol Med*
89. Ren, Zou, Zhang et al. (2017) "Redox signaling mediated by thioredoxin and glutathione systems in the central nervous system" *Antioxid Redox Signal*
90. You, Park (2016) "Auranofin induces mesothelioma cell death through oxidative stress and GSH depletion" *Oncol Rep*
91. Abdalbari, Martinez-Jaramillo, Forgie et al. (2023) "Auranofin induces lethality driven by reactive oxygen species in high-grade serous ovarian cancer cells" *Cancers (Basel)*
92. Panieri, Santoro (2016) "ROS homeostasis and metabolism: a dangerous liason in cancer cells" *Cell Death Dis*
93. Miri, Ouf, Çimen et al. (2024) "Visualizing the H2O2-Nrf2 relationship using an oxygen-independent Nrf2 biosensor under controlled oxygen conditions"
94. Yang, Liu, Li et al. (2024) "The rheumatoid arthritis drug auranofin exerts potent anti-lymphoma effect by stimulating TXNRD-mediated ROS generation and inhibition of energy metabolism" *Redox Biol*
95. Chen, Tzekov, Su et al. (2016) "Auranofin inhibits retinal pigment epithelium cell survival through reactive oxygen species-dependent epidermal growth factor receptor/ mitogen-activated protein kinase signaling pathway" *PLoS One*
96. Yang, Zhu, Zhou et al. (2025) "The rheumatoid arthritis drug Auranofin targets peroxiredoxin 1 and peroxiredoxin 2 to trigger ROS-endoplasmic reticulum stress axis-mediated cell death and cytoprotective autophagy" *Free Radic Biol Med*
97. Pak, Ezeriņa, Lyublinskaya et al. (2020) "Ultrasensitive genetically encoded indicator for hydrogen peroxide identifies roles for the oxidant in cell migration and mitochondrial function" *Cell Metab*
98. Ma (2013) "Role of Nrf2 in oxidative stress and toxicity" *Annu Rev Pharmacol Toxicol*
99. Lee, Koh, Jun et al. (2022) "Auranofin attenuates hepatic steatosis and fibrosis in nonalcoholic fatty liver disease via NRF2 and NF-κB signaling pathways" *Clin Mol Hepatol*
100. Falchetti, Delgobo, Zancanaro et al. (2023) "Omics-based identification of an NRF2-related auranofin resistance signature in cancer: Insights into drug repurposing" *Comput Biol Med*
101. Cho, Imani, Miller-Degraff et al. (2009) "Antiviral activity of Nrf2 in a murine model of respiratory syncytial virus disease" *Am J Respir Crit Care Med*
102. Rahban, Habibi-Rezaei, Mazaheri et al. (2020) "Anti-viral potential and modulation of Nrf2 by curcumin: pharmacological implications" *Antioxidants (Basel)*
103. Lee (2018) "Therapeutic modulation of virus-induced oxidative stress via the Nrf2-dependent antioxidative pathway" *Oxid Med Cell Longev*
104. Loeb, Haas (1992) "The interferon-inducible 15-kDa ubiquitin homolog conjugates to intracellular proteins" *J Biol Chem*
105. Sooryanarain, Rogers, Cao et al. (2017) "ISG15 modulates type I interferon signaling and the antiviral response during hepatitis E virus replication" *J Virol*
106. Lu (2025) "Guanylate-binding protein 5: a promising biomarker and therapeutic target" *Infect Immun*
107. Sauter, Kirchhoff (2024) "Antiviral mechanisms of guanylate-binding protein 5: versatile inhibition of multiple viral glycoproteins"
108. Feng, Cao, Wang et al. (2017) "Inducible GBP5 mediates the antiviral response via interferonrelated pathways during influenza A virus infection" *J Innate Immun*
109. Haynes, Titus, Cooper (2004) "Degradation of misfolded proteins prevents ER-derived oxidative stress and cell death" *Mol Cell*
110. Zhang, Zhang, Zhou et al. (2019) "Redox signaling and unfolded protein response coordinate cell fate decisions under ER stress" *Redox Biol*
111. Lee (2005) "The ER chaperone and signaling regulator GRP78/BiP as a monitor of endoplasmic reticulum stress" *Methods*
112. Rutkowski, Kaufman (2004) "A trip to the ER: coping with stress" *Trends Cell Biol*
113. Acosta-Alvear, Zhou, Blais et al. (2007) "XBP1 controls diverse cell type-and condition-specific transcriptional regulatory networks" *Mol Cell*
114. Lee, Iwakoshi, Glimcher (2003) "XBP-1 regulates a subset of endoplasmic reticulum resident chaperone genes in the unfolded protein response" *Mol Cell Biol*
115. Leitman, Shenkman, Gofman et al. (2014) "Herp coordinates compartmentalization and recruitment of HRD1 and misfolded proteins for ERAD" *Mol Biol Cell*
116. Kitzmüller, Caprini, Moore et al. (2003) "Processing of N-linked glycans during endoplasmic-reticulum-associated degradation of a short-lived variant of ribophorin I" *Biochem J*
117. Sander, Fourie, Sabiu et al. (2022) "Reactive oxygen species as potential antiviral targets" *Rev Med Virol*
118. Nencioni, Sgarbanti, Amatore et al. (2011) "Intracellular redox signaling as therapeutic target for novel antiviral strategy" *Curr Pharm Des*
119. Choi, Lee, Zheng et al. (2004) "Reactive oxygen species suppress hepatitis C virus RNA replication in human hepatoma cells" *Hepatology*
120. Fu, Yu, Wang et al. (2025) "Heterogeneity in susceptibility of viruses with different structures to various reactive oxygen species: kinetics and biological mechanisms" *Eco Environ Health*
121. Xiang, He, Zhu et al. "Chinese Consortium for the Study of Hepatitis E (CCSHE). 2024. Animal models of hepatitis E infection: advances and challenges" *Hepatobiliary Pancreat Dis Int*
122. Kamar, Izopet, Pavio et al. (2017) "Hepatitis E virus infection" *Nat Rev Dis Primers*
123. Kumar, Sharma, Kumar et al. (2020) "Host-directed antiviral therapy" *Clin Microbiol Rev*
124. Emerson, Nguyen, Torian et al. (2006) "ORF3 protein of hepatitis E virus is not required for replication, virion assembly, or Full-Length Text Journal of Virology"
125. "infection of hepatoma cells in vitro" *J Virol*
126. Nguyen, Shukla, Torian et al. (2014) "Hepatitis E virus genotype 1 infection of swine kidney cells in vitro is inhibited at multiple levels" *J Virol*
127. Ding, Nimgaonkar, Archer et al. (2018) "Identification of the intragenomic promoter controlling hepatitis E virus subgenomic RNA transcription" *mBio*
128. Shukla, Nguyen, Torian et al. (2011) "Cross-species infections of cultured cells by hepatitis E virus and discovery of an infectious virushost recombinant" *Proc Natl Acad Sci*
129. Shukla, Nguyen, Faulk et al. (2012) "Adaptation of a genotype 3 hepatitis E virus to efficient growth in cell culture depends on an inserted human gene segment acquired by recombination" *J Virol*
130. Wang, Tian, Sooryanarain et al. (2022) "Two mutations in the ORF1 of genotype 1 hepatitis E virus enhance virus replication and may associate with fulminant hepatic failure" *Proc Natl Acad Sci*
131. Marano, Chuong, Weger-Lucarelli (2020) "Rolling circle amplification: a high fidelity and efficient alternative to plasmid preparation for the rescue of infectious clones" *Virology (Auckl)*
132. Vargas (2000) "ED50plus"
133. Shi, Su (2024) "HMOX1 participates in pre-eclampsia by regulating the proliferation, apoptosis, and angiogenesis modulation potential of mesenchymal stem cells via VEGF" *Biochem Genet*
134. Zhang, Wang, Li et al. (2017) "Antiviral effects of IFIT1 in human cytomegalovirus-infected fetal astrocytes: antivirus effects of IFIT1" *J Med Virol*
135. Qin, Lai, Liu et al. (2017) "Guanylate-binding protein 1 (GBP1) contributes to the immunity of human mesenchymal stromal cells against Toxoplasma gondii" *Proc Natl Acad Sci*
136. Van Schadewijk, Van't Wout, Stolk et al. (2012) "A quantitative method for detection of spliced X-box binding protein-1 (XBP1) mRNA as a measure of endoplasmic reticulum (ER) stress" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12674793&blobtype=pdf | # P18 Antibiotic duration exceeds time to clinical stability in hospital acquired pneumonia
Elliott Lewis-Douglas, Jon Spedding, Sandhya Prem, Jamie Lupton, Daniel Weiand, Iain Mccullagh, Tom Hellyer
Background: Hospital-acquired pneumonia (HAP) is a common indication for broad-spectrum antibiotics. 1 Shorter course durations are becoming more widely accepted due to increasing evidence of non-inferior clinical effectiveness and safety compared to longer course lengths. 2,3 Guidance for duration in HAP is often extrapolated from CAP and VAP evidence, and HAP lacks validated scoring systems to guide decision making.foot_0,5 Time to clinical stability (TCS) criteria (heart rate <100, systolic blood pressure >90 mmHg, temperature <37.8°C, respiratory rate <25 and oxygen saturations >90%) have been shown to be a safe metric to guide duration decision making in CAP 2 and effective in indicating risk of complications early in disease. 3 Objectives: To determine the TCS in relation to duration of antibiotics to identify whether there could be scope to further shorten antibiotic course length if guided by TCS criteria.
Methods: We automated data-collection, screening the trust wide electronic patient record, during two non-consecutive weeks. Free-text antibiotic indication data were analysed to identify inpatients with chest-infection related terminology and prescribed antibiotics for more than 48 h. Manual searches excluded non-HAP respiratory infections. Further data collection included patient demographics; time to clinical stability; chest X-ray appearance; antibiotic prescriptions; Charlston comorbidity index; radiograph findings; and pathogens.
Results: A total of 188 patients were identified in automated screening. After exclusions, 68 patients were included. The median (IQR) age was 76.5 years (62.3-82.5); 36.8% were female; and the median (IQR) Charlson Comorbidity Index was 6 (4-7.8). 34.4% of patients were surgical (23). The median (IQR) antibiotic duration was 5 (5-7) days whilst median time to clinical stability was 2 (0-3) days. For 92% of patients, clinical stability was maintained for >24 h. Twenty-five percent of patients had a change from initial antibiotic, which included equally escalations for deteriorations and de-escalations to oral antibiotics. Local prescribing guidelines were followed in 73% of patients. Piperacillin/tazobactam or a carbapenem was prescribed in 31% and 3% of patients, respectively as the initial choice of antibiotic. 68.8% of patients had chest radiograph findings recorded as positive. While an organism was identified in only 14% of patients, in 3% of patients Pseudomonas aeruginosa. was isolated. These two patients reached clinical stability at day 12 and 0.
Conclusions: In this cohort receiving predominantly broad-spectrum antibiotics, TCS were reached much earlier than the prescribed antibiotic duration. We included patients based on recorded indication and treatment with antibiotics. The study is limited by diagnostic uncertainty as organisms were rarely identified and not all patients had radiological evidence of HAP. However, the frequent use of antibiotics without supporting diagnostic evidence, highlights the need for improved antibiotic stewardship in patients with suspected HAP. This study suggests that if antibiotics were tailored to achieving TCS, rather than the commonly used fixed duration approach, antibiotic course lengths could potentially be shortened.
## References
1. (2019) *Hospital Acquired Pneumonia: Antimicrobial Prescribing. NICE NG*
2. Mo, Booraphun, Li (2024) "REGARD-VAP investigators. Individualised, shortcourse antibiotic treatment versus usual long-course treatment for ventilatorassociated pneumonia (REGARD-VAP): a multicentre, individually randomised, open-label, non-inferiority trial" *Lancet Respir Med*
3. Kuijpers, Buis, Ziesemer (2025) "The evidence base for the optimal antibiotic treatment duration of upper and lower respiratory tract infections: an umbrella review" *Lancet Infect Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12838271&blobtype=pdf | # High-pressure processing of pork liver reduces the infectivity of the hepatitis E virus
Marie Pellerin, Jean-Luc Martin, Lauranne Harlet, Virginie Doceul, Nicole Pavio, Carole Feurer
## Abstract
Hepatitis E virus (HEV) causes acute and chronic hepatitis in humans. The zoonotic HEV genotype 3 (HEV-3) is present in various animal species, including pigs, wild boars, and other game animals. Foodborne transmission with the consumption of raw or undercooked pork products is the major transmission route of HEV-3. HEV RNA has been detected in various types of food, but particularly in pork liver-based food products. High hydrostatic pressure processing (HPP) can be used for the inactivation of pathogens in food. In the present study, the impact of HPP treatments was evaluated on HEV-3 infectivity in raw pork liver. Different pressure/time combinations (500 MPa for 1 or 5 min, 600 MPa for 1, 5, or 10 min) were applied to raw pork livers, artificially contaminated with HEV-3 (8.3 log 10 HEV ge/g). Residual HEV infectivity was evaluated using the HepaRG cell culture model in p-24 well plates. The results obtained have shown the absence of residual infectious HEV particles in pork liver after a treatment of 600 MPa, during 1 min in a refrigerated room at +8°C. Then, liver sausages were prepared with pork liver treated at 600 MPa for 1 min. Technological measurements showed that the treatment had a significant impact on brightness, firmness, red hue, and cohesiveness. Nevertheless, these differences have not been perceived after food testing, which highlighted no major difference in taste or color. Thus, inactivation of HEV-3 in raw pork liver by HPP is a possible treatment to limit the risk of HEV exposure through food consumption.
IMPORTANCEThe hepatitis E virus (HEV) is the leading cause of enterically transmitted acute hepatitis worldwide. It can have a zoonotic origin through the consumption of infected meat. Pigs are the main reservoir of zoonotic HEV, and pork livers are frequently contaminated by HEV. In the present study, we investigated the use of high-pressure processing (HPP) to inactivate HEV-3 in pork liver. This study is the first to identify HPP treatment parameters that can be applied to pork liver to reduce HEV infectivity. Additionally, it is the first study to demonstrate the feasibility of processing HPP-treated pork liver into food products, such as dry liver sausage.KEYWORDS hepatitis E virus, zoonosis, high-pressure processing, pork liver sausage, food safety H epatitis E virus (HEV) is a quasi-enveloped single-stranded RNA virus classified in the Hepeviridae family, subfamily Orthohepeviridae which belongs to the species Paslahepevirus balayani (1). HEV is responsible for enterically acquired hepatitis in humans that is often acute and self-limiting but can evolve to fulminant hepatitis or cause chronic hepatitis, particularly in immunocompromised patients (2). A recent study including several hundred viruses with zoonotic potential has classified HEV among the 10 viruses with the highest risk of spillover from wild fauna (3).Four major HEV genotypes have been identified in humans. Genotypes 1 and 2 (HEV-1 and -2) strictly infect humans, while genotypes 3 and 4 (HEV-3 and -4) are
zoonotic and largely present in pigs and wild boar reservoirs (4,5). HEV-3 and -4 infections in humans have a foodborne origin. Although other transmission routes exist, foodborne transmission is considered the primary mode of HEV contamination with the consumption of meat and meat products prepared from infected animals (6). Notably, HEV-3 subtypes 3c and 3f are circulating predominantly in Europe in both animal and human populations (7).
In recent years, the incidence rate of HEV-3 infection has increased in industrial ized countries (8). Several studies have demonstrated a high prevalence of anti-HEV antibodies in the human populations of various European countries, with studies carried out in France and Germany showing seroprevalences over 20% (9,10).
Evidence indicates that infection with HEV-3 is common among domestic swine. High prevalences of antibodies have been detected in European swine populations (11,12).
Studies carried out in pigs at the slaughterhouse have shown that a significant proportion of the animals and their organs, particularly the liver, are contaminated with HEV. For example, in France, 2%-4% of pork livers collected at slaughterhouses were found to be positive for HEV RNA (11,13).
HEV RNA was detected in various types of food, but particularly in pork liver-based food products (14). The prevalence of contaminated pork products varies from 0.03% to 50% depending on the geographical region and tested meat products (15)(16)(17)(18)(19)(20).
In human populations, higher prevalence of HEV IgG has been observed in regions where products made from uncooked pork liver are consumed, for example, in several regions of France and Italy (9,21).
Because of the high prevalence of HEV RNA in food products containing raw pork liver, efficient strategies to inactivate HEV are necessary. HEV can survive for a long time in food products with only a slight decrease in infectivity (22,23). Certain traditional preservation methods, such as drying, are not significantly effective for HEV inactivation (24). High concentration of nitrite salt, often used to cure sausage, seems to be tolerated by HEV, as well as pH variations (25).
The easiest method for reducing HEV contamination is heating, and its efficiency has been shown in several studies (26,27). However, not all types of food products can be heated, and in particular, dried or fresh liver sausages, one of the main sources of HEV exposure (15). Thus, alternative methods for HEV inactivation should be developed for specific food products.
High pressure processing (HPP) is a "non-thermal pasteurization" technique that has already demonstrated hundred years ago its efficiency for inactivating microorgan isms, particularly in milk (28). Currently, HPP is used as a gentle alternative to thermal treatment in various food products such as juice, vegetables, and seafood products (29). The inactivation effect is multifactorial, due to pH change, protein denaturation, permeabilization, and/or modification in fluidity of cell membranes (30,31).
HPP treatment has also been shown to be useful for inactivating several viruses, with minimal influence on the physicochemical and sensory properties of foods (32). The HPP treatment effect on viruses is dependent on the food type and its components. The inactivation behavior can be more or less important according to the protective role of the food matrix. Thus, for feline calicivirus and murine norovirus, a higher reduction is measured in swine liver rather than in ham after the same HPP treatment (400 MPa/10 min) (31). Different HPP inactivation effects have also been shown between chicken meat and eggs for bursal disease virus (33). Even if HPP does not impact covalent bonds, quality properties of food can be affected, depending on the HPP conditions, treatment, and food type treated (34,35). Thus, testing needs to be done for each specific food matrix.
Recent studies have explored the effect of HPP on HEV inactivation in different matrices. An initial study was carried out in culture medium and liver pâté (36). A 2-log decrease in infectivity of HEV was observed in culture medium at 400 or 600 MPa for 1 or 5 min, and a 0.5-log decrease in liver pâté. A second study was carried out on the inactivation of HEV in phosphate-buffer suspension (PBS) (37). A gradual decrease in infectivity was observed by application of 100 to 600 MPa for 2 min. The virus was nearly completely inactivated at 600 MPa. At last, one study analyzed the effect of HPP on HEV inactivation in human milk (38). Their results demonstrated that HEV was not affected after moderate HPP treatments, and the milk matrix did not fully protect HEV from inactivation at 600 MPa.
Given the varying results obtained depending on the nature of the matrices, it is important to further establish an effective HPP treatment against HEV in raw pork liver.
To this end, we investigated the impact of different pressure/time combinations on HEV infectivity in raw pork livers artificially contaminated with HEV-3. This was achieved using the cell culture HepaRG cells and the adapted genotype 3f strain FR-HuHEVF3f. Then, industrial partners have assessed the technological and sensory properties of liver-dried sausages prepared with PPH-treated pork liver.
## MATERIALS AND METHODS
## Cells and virus
Human HepaRG cells were grown as previously described (39). Cells were seeded into 24-well plates (5 × 10 4 cells/well). They were maintained in growth medium for 2 weeks, and medium with 1.2% DMSO was replaced for two extra weeks for cell differentiation into hepatocytes. After 4 weeks, cells were infected overnight with virus inoculum under a maximum volume of 250 μL. The viral suspension was then removed, and cells were washed three times in PBS before adding 0.5 mL of growth medium. Fresh medium was renewed every 2-3 days. Supernatants from infected cells were collected once a week for HEV RNA detection by RT-qPCR.
The HepaRG cell model can be used to detect infectious HEV from naturally contaminated liver samples (40).
The HEV genotype 3f strain FR-HuHEVF3f (GenBank JN906974), originally derived from a French patient suffering from acute autochthonous hepatitis E, was used for all spikings and infections. Six consecutive passages of the virus were previously carried out in HepaRG (39).
## Virus production and sample inoculation
To obtain a high viral load and sufficient viral stock for all infections, HepaRG cells were infected (passage 6) in T75 flasks with 3.5 mL (8.7 log 10 HEV ge/mL) overnight. Then, the inoculum was removed, cells were washed three times with PBS, and fresh medium was added. Supernatants were collected three times per week for 50 days, from day 65 to day 115, when the medium was renewed. All virus supernatants were pooled (around 960 mL with 8.5 log 10 HEV genome equivalent in RNA copy numbers (HEV ge/mL). To keep a homogeneous viral population of quasi-enveloped HEV particles, intracellular particles were not collected.
In order to concentrate the virus, the pooled supernatants were centrifuged on a Vivaspin20® centrifugal concentrator (100 kDa). A viral stock suspension of 70 mL was obtained, with a concentration of 9.4 log 10 HEV ge/mL.
Liver from a specific pathogen-free pig, not infected by HEV, was cut into 1 cm 3 pieces and divided into 36 plastic bags, with approximately 15 pieces of liver per bag (20 g). In 30 bags, liver samples were artificially contaminated by intra-tissular injection with needles (Microlance3 Becton Dickinson, 0.3 × 13 mm) of viral suspension (2 mL), corresponding to 100-150 µL inoculum per piece, equivalent to 8.3 log 10 HEV ge/g of liver. The other six liver samples were injected with culture medium as a negative control.
Each sample was vacuum-packed and placed in a sealed plastic bag. It was then placed in a third plastic bag containing a solution of peracetic acid to comply with the hygiene and biosafety conditions imposed by the infectious nature of HEV.
## High-pressure treatment
## Treatment of artificially contaminated samples
For each HPP condition tested, samples included five HEV artificially contaminated livers, one control liver not contaminated with HEV, and three viral suspensions corresponding to the viral stock diluted 1:10 in 2 mL PBS.
The HPP treatment was carried out on a 100-L horizontal high-pressure equipment (Avure AV-10) located on an industrial platform, in a refrigerated room at +8°C. The pressure-transmitting fluid used was water, whose temperature ranged from 7.1°C to 9.0°C before treatment. The compression rate was 4.26 MPa•s -1 , and decompression was nearly instantaneous. During HPP treatment, the cylinder temperature increased because of the adiabatic heating, without exceeding 26°C.
The HPP scales tested were chosen based on the literature and, in particular, in relation to scales enabling hepatitis A virus inactivation (41,42). Samples were exposed to different pressure levels as follows: 500 MPa for 1 and 5 min, 600 MPa for 1, 5, and 10 min. The scale of 600 MPa for 10 min corresponded to a positive control of inactivation. After treatment, the pressurized samples were placed at +4°C in a cold room before direct shipping at +4°C for infectivity analysis. Shipment time lasted no longer than 1 day (24 h).
## Treatment of pork liver for liver sausage processing
The HPP treatment, allowing total inactivation of the HEV virus, was applied to 30 kg of fresh vacuum-packed pork livers. High-pressure-treated livers were used to make liver sausages on an industrial scale by two partner companies, A and B. The formulation of the liver sausages slightly differed between companies and was as follows: company A: pork meat, pork liver (30%), salt, lactose, dextrose, spices, flavor enhancer, stabilizer, preservatives, flavors, smoke flavors, starter cultures, and natural pork casing; company B: pork meat and fat, pork liver (30%), salt, lactose, dextrose, milk powder, spices, preserva tives, starter cultures, and natural pork casing. The sausages were dried for 15 days. On the finished product, the fat content was 29.4 g/100 g for company A and 35 g/100 g for company B. The organoleptic quality of the products was assessed by both companies by an internal expert panel composed of the General Manager, the Quality Manager, the Production Manager, and the Production Technician. All of them are routinely involved in evaluating newly developed in-house liver sausage formulations, as this activity forms an integral part of their professional duties. In this specific context, an Institutional Review Board is not required. The evaluation was performed according to the following criteria: appearance, color, taste, and texture. At the same time, technological measurements (see below) were carried out on both productions of liver sausage with the casing removed. The results were compared with a control production made under the same conditions with livers not treated by HPP, in each case.
## Technological measurements
## Color measurements
Color measurements were performed using a KONICA MINOLTA CM600-d spectroco lorimeter, for an 8 mm diameter measuring area, including specular mode, in CIELAB color space. The measured values were the brightness (L*) and red hue angle, in degrees (H*). Thirty independent measurements per partner company were performed on five independent liver sausages at a rate of 6 measurements per sausage.
## Texture measurements
Texture measurements were carried out using a TA-XTPlus texturometer equipped with a round-section compression tool. They were carried out on a sample of 15 mm in diameter and 10 mm in height, after removal of the casing. Six measurements were carried out on five sausages treated (n = 30) or not (n = 30) per company with the TPA method: double compression at constant speed on 7.5 cm (75%) in the center of the slice.
The values considered were firmness (N) and cohesiveness, the best representative indices of coarse-grained products.
## Data processing
Data were processed by variance analysis using R 4.3.2 software. The variances' homogeneity was confirmed using the Bartlett test with a threshold of 0.5%. The normality of the data dispersion was confirmed using the Shapiro-Wilk test with a threshold of 0.5%. The effects of company and treatment factors were determined with a threshold of 5%. Significantly different groups were determined using the Tukey-C test. Measurement dispersion was highlighted using ggplot2 graphs.
## HEV infectivity assay
After HPP treatment and shipping, all samples were stored at -80°C until analyses. Three independent experimental infections were carried out in triplicate.
Livers were homogenized twice with 40 mL PBS, using a blender. Two successive centrifugations were performed to eliminate the food matrix debris as much as possible. Supernatants were diluted in cell medium (1/20) and passed through a 0.45 µM and then a 0.22 µM pore size filter.
After the final step of inoculum preparation (homogenization, centrifugation, dilution, and filtration), the titers of the suspensions were 6.3 log 10 ₀ HEV ge/mL instead of the expected 6.7 log 10 HEV ge/mL. Thus, half of the viral load was recovered.
For each liver suspension, three independent infections were carried out on DMSOdifferentiated HepaRG cells. Infections were maintained for up to 42 days, and cell culture medium was changed three times a week. Supernatant samples were collected once a week to quantify HEV RNA production.
The limit of infectivity of HEV culture in vitro (5.3 log 10 HEV ge/mL) was defined by carrying out serial dilutions of viral suspensions and optimizing the culture conditions of HepaRG cells (See Fig. S1 at https://doi.org/10.57745/MNRNYA). For each series of infections, a positive culture control was performed with untreated virus in suspension corresponding to 5.3 log 10 HEV ge/mL. Statistical analysis was performed using GraphPad Prism v.9.0 (GraphPad Software) and the Kruskal-Wallis test method.
## HEV RNA isolation and quantification
RNA extraction from liver homogenate, PBS suspension, or cell culture supernatant was performed using a magnetic bead-based separation technology and the KingFisher Duo (Thermo Fisher Scientific, Courtaboeuf, France). Total RNAs from 200 µL supernatant were extracted using the MagMAX core nucleic acid purification kit (Thermo Fisher Scientific, Courtaboeuf, France), according to the manufacturer's instructions and the KingFisher instrument guide (40). Elution was performed in 90 μL of RNAse-free water.
HEV RNA detection and quantification were performed using a real-time quanti tative RT-PCR as previously adapted from a method described by reference 43 in reference 26.
A standard quantification curve was obtained using in vitro transcribed RNAs from the plasmid pCDNA 3.1 ORF2-3 HEV. The limit of detection of the applied system is 3.3 log 10 HEV ge/mL and 4.8 log 10 HEV ge/g, corresponding to five copies of HEV RNA in 2 μL of total RNA extract.
## RESULTS
The impact of different pressure/time combinations on HEV-3 infectivity was investiga ted in pork livers artificially contaminated with FR-HuHEVF3f and in viral PBS suspen sions, treated at 500 MPa for 1 and 5 min, and at 600 MPa for 1, 5, and 10 min.
## Effect of HPP treatment on the molecular quantification of HEV RNA by RT-qPCR
First, the impact of high-pressure treatments on the molecular detection of HEV was evaluated by quantifying HEV RNA genome equivalent in copy numbers (HEV ge) by RT-qPCR in HEV-contaminated samples (liver homogenate or PBS suspension) subjected or not to various HPP treatments. Initial HEV RNA titers in pork liver and PBS solutions were, respectively, 6.3 log 10 and 8.3 log 10 HEV ge/mL.
The results showed that the various high-pressure treatments tested did not have any impact on the molecular quantification of HEV ge/mL by RT-qPCR. HEV ge/mL was equivalent in treated and untreated samples, both in livers and in viral PBS suspensions (Table 1).
## Evaluation of HEV residual infectivity in pork liver matrix subjected to HPP using HepaRG cell cultures
In order to determine if high-pressure treatments have an impact on HEV infectivity, homogenates of liver samples subjected to HPP or not were cultured in HepaRG cells. This model allows detecting the presence of infectious HEV in samples by monitoring viral RNA production in the supernatant of infected cell cultures for 6 weeks (42 days), using RT-qPCR (40). Three independent experiments were carried out (independent liver bags treated), with three cell culture wells infected per condition for each experiment (Fig. 1).
In order to avoid the measurement of residual inoculum, cell supernatants were collected and analyzed during 6 weeks post-infection (42 days). HEV replication in cell culture was considered positive when an increase in RNA HEV level was observed during this period of time. Without treatment, the quantity of HEV RNA increased in cell supernatant between 21 and 42 days post-infection (dpi), demonstrating HEV replication (Fig. 1a). The HEV RNA measures (ge/mL) in the supernatants of the three independent experiments show variability in the levels of HEV produced of unknown origin, confirming the necessity of performing several replicates. HEV RNA was detected after 42 dpi in the supernatant of 1 to 3 cultures of HepaRG cells infected with livers treated at 500 MPa for 1 or 5 min, showing the presence of infectious viruses in these samples. No HEV RNA was detected after 42 dpi in the supernatant of HepaRG cells infected with livers treated at 600 MPa for 1, 5, and 10 min, suggesting the decrease of infectious HEV-3, under the limit of detection in these samples (Table 2). Considering that the cell culture system used allows the detection of 5.3 log 10 HEV ge/mL (See Fig. S1 at https://doi.org/ 10.57745/MNRNYA), at least a one-log reduction was reached with 600 MPa HPP from 1 min.
In the HepaRG cell culture model of HEV, it is possible to link the quantity of virus produced in the supernatants after 42 days of culture, with the quantity of infectious virus initially present (See Fig. S2 at https://doi.org/10.57745/MNRNYA). At 500 MPa, the quantity of HEV RNA increased in cell supernatant between day 21 and day 42, but the quantity of HEV RNA measured in culture supernatants at 42 days was lower in samples treated at 500 MPa than in untreated samples (by up to 2 log) (Fig. 2). This decrease was
## Evaluation of HEV residual infectivity in PBS suspension subjected to HPP using HepaRG cell cultures
In order to evaluate a possible effect of the liver matrix on HEV inactivation, the same HPP pressure/time combinations were applied to viral suspensions diluted in PBS buffer. As above, HepaRG cells were infected with viral suspension diluted in PBS, treated or not with the same HPP conditions. As for liver sample infections, HEV RNA was ge/mL of HEV-3, treated at 500 MPa (1 and 5 min), HEV RNA replication was detected in all cell cultures (n = 6). However, no HEV replication was observed in cultures infected from suspension treated at 600 MPa, whatever the treatment time (1, 5, or 10 min) (n = 6) (Table 3; See Fig. S3 at https://doi.org/10.57745/MNRNYA). A suggestion that a log reduction of >3 was reached in PBS suspension. Data with the 7.3 log 10 HEV ge/mL viral suspensions are shown in Fig. S3 at https:// doi.org/10.57745/MNRNYA. The amount of virus produced after infection with suspen sions treated at 500 MPa for 1 min was lower, with 2 log reduction and 6 HEV positive cultures out of 6, than that of untreated suspensions, indicating a lower amount of residual virus after treatment. The inactivation effect was enhanced when the treatment was increased from 1 to 5 min, with no detection with the 500 MPa/5 min treatment (n = 6), suggesting a >3 log reduction.
For infections, with 6.3 log 10 HEV ge/mL in PBS, no viral replication could be detected in any culture infected with HPP-treated suspensions in all conditions (Fig. 3). In the liver infected with 6.3 log 10 HEV ge/mL, infectious virus could be detected after treatment at 500 MPa for 1 and 5 min. Thus, for an equivalent initial viral load, more infectious HEV virus remained when the treatments were applied to the liver matrix than in the PBS suspension (Fig. 3).
## Technological measurements
The impact of the HPP treatment inactivating HEV at 600 MPa for 1 min was assessed on the technological and sensory properties of dried liver sausages produced with HPP-treated pork liver by industrial partners. Technological measurements (dispersion of color and texture measurements) were carried out on pilot productions, and the organoleptic quality of the products was assessed by both companies.
## Dispersion of color measurements
For each company, brightness and red hue measurements of not treated versus treated sausages are presented in Fig. 4. Despite the heterogeneity of the sausages composed of large grains of meat and fat, the dispersion of the 30 color measurements is not significant regardless of the sausage product (treated or not) or the company. Indeed, the coefficient of variation ranged from 7% to 9% for brightness and from 6% to 8% for red hue.
The analysis of variance showed that there was no significant effect of the factor "company" on the brightness and red hue (P > 1). The factor ''company'' includes all manufacturing parameters such as raw materials, ingredients, additives, and the technological process. However, the factor "treatment" had a significant effect on both measures (P ≤ 0.001). For both companies, treated sausages were brighter and less red than untreated ones.
Figure 4c clearly shows that the colorimetric indices of untreated sausages from both companies were very close, and that they were essentially differentiated by the treatment applied. The limit deviation of the human eye's perception of brightness and red hue is generally 5 units. In this case, the difference in brightness between the two treatment levels should be considered as not perceptible to the naked eye. In our case, the difference in red hue is just at the limit of visual perception. The treatment neither greatly improved the brightness of the sausages nor greatly deteriorated their red hue.
## Dispersion of texture measurements
For each company, firmness and cohesiveness measurements of not treated versus treated sausages are presented in Fig. 5. The dispersion of the 30 measurements is greater for firmness (coefficient of variation ranging from 18% to 28%) than for cohesiveness (coefficient of variation ranging from 9% to 11%). Despite the greater dispersion of values for firmness, the Bartlett test confirmed the homogeneity of variances for both firmness and cohesiveness.
There was no significant difference in firmness between both companies regarding untreated sausages, whereas treated sausages from company A were significantly firmer than treated sausages from company B. The characteristics of fresh sausages had a significant effect on the difference in firmness of treated sausages.
For both companies, a decrease in cohesiveness was observed after treatment. Moreover, untreated sausages from company B were significantly more cohesive than untreated sausages from company A. The same difference was observed on treated sausages, highlighting that the treatment alone had a significant effect on the difference in cohesiveness of treated sausages from companies A and B. Figure 5c shows that for texture indices, differences are linked not only to HPP treatment, but also to the product formulation of the two companies. As far as texture measurements are concerned, we have no reference for the threshold of human perception of firmness and cohesion.
In Table 4, the global effects of the high-pressure treatment on the four parameters of dried liver sausages of both companies combined are presented. The global effects for the four parameters for each company are presented in Table S1 at https://doi.org/ 10.57745/MNRNYA.
Overall, the high-pressure treatment had a significant effect, which was identical for both product references, on all the parameters studied. The treated sausages were brighter, firmer, less red, and less cohesive than untreated ones.
## Organoleptic evaluation by companies
To assess the impact of HPP treatment, each company called on an in-house panel of four tasters accustomed to consuming liver sausages. The treatment had a visual impact on the liver color, which showed a dull brown color compared to the shiny brown/red control.
Both companies indicated that the mixture before stuffing was paler, but once stuffed and during drying, the sausages regained a color comparable to that of control liver sausages. No major difference in taste in terms of aroma or acidity was observed compared to the control. The slice color was acceptable. Firmness and cohesion were not found to be different compared to the control product.
## DISCUSSION
HPP is used to inactivate microorganisms, including viruses, with minimal influence on the physicochemical and organoleptic properties of food products (29). The inactivation efficiency of HPP on viruses differs according to the viral species, the parameters applied (pressure, time, temperature...), and the composition of the food matrix (32).
As the risk of human exposure to HEV is mainly associated with the consumption of products containing infected pork liver (18)(19)(20), the efficacy of HPP treatments was investigated in raw pork liver. For this purpose, pork liver, artificially contaminated with HEV-3, subtype 3f, was submitted to different HPP treatments. The presence of residual HEV particles was evaluated using the HepaRG cell culture system for HEV. The results have shown, first, that the quantification of HEV genome, in the contaminated liver matrix after HPP treatments, was not affected. It confirms that RNA quantification does not reflect the presence of infectious particles. Second, a partial reduction of HEV replication was detected in HepaRG cells with HPP treatments of 500 MPa for 1 min (1 log) or 5 min (2 log). An absence of HEV replication was observed in cell cultures inoculated with pork liver treated at 600 MPa for 1 min.
In comparison to other enteric viruses and surrogate models, HEV showed a high stability to HPP treatments. Indeed, Hepatitis A virus (HAV), murine norovirus 1, or feline calicivirus (FCV) is inactivated, or partially inactivated with treatment from 400 to 500 MPa in various matrices (31,41,42).
Then, the effect of the matrix on HEV inactivation was assessed by comparing HEV inactivation in PBS suspensions. In this case, no HEV replication was detected regardless of the HPP treatment applied. Thus, it suggests that the raw pork liver matrix protects HEV from inactivation. HPP induces protein denaturation, membrane permeabilization, and/or alterations in membrane fluidity. Within the hepatocyte/liver microenvironment, HEV particles may interact with cellular factors or the extracellular matrix, potentially attenuating the efficacy of HPP treatment. These results are consistent with other studies, as the matrix nature has been identified as a critical parameter for the design of HPP inactivation strategies (30). For example, FCV and bovine enterovirus, a surrogate for HAV, showed a higher resistance to HPP treatments in mussels and oysters, as compared to seawater and culture medium (44). Few other studies have investigated the effect of HPP treatments on HEV infectivity in various matrices. In all of them, a high stability of HEV was shown. For example, in liver pâté matrix, a 0.5-log decrease was obtained after a treatment at 400 or 600 MPa for 5 min (36). Another study has analyzed HEV sensitivity to HPP in PBS suspension (37). The authors showed a nearly complete inactivation of HEV at 600 MPa for 2 min. A study on HPP inactivation of HEV in human milk has shown that HEV was inactivated at 600 MPa for 5 min (38).
Overall, HEV inactivation requires high-intensity HPP treatments of 600 MPa for 1 to 5 min. The reduction of infectious HEV measured may vary, depending on the nature of the food matrix (liver pâté, milk, pork liver), the HPP device, the sensitivity of the cell culture system, the method used to detect HEV, or the viral strain used (HEV subtype). Indeed, different strains of HEV may have different inactivation sensitivities to HPP treatments, as it has been shown for six different strains of human rotavirus (45).
It is known that HPP can affect the quality properties of foods (32). Hence, the influence of the HPP treatment on the technological properties of dried liver sausages was analyzed with several parameters. For this purpose, dried liver sausages were prepared with pork liver treated at 600 MPa for 1 min by two independent companies. Technological measurements showed that the treatment had a significant impact on brightness, firmness, red hue, and cohesiveness. Nevertheless, these differences have not been perceived by the companies' panels, which highlighted no major difference in taste or color compared to the control liver sausage after drying. Moreover, the impact of the treatment on the product's texture was more likely due to the company's manufacturing process rather than the HPP treatment itself.
Even if the pilot test has been validated by both companies, the feasibility needs to be confirmed on an industrial scale. Moreover, before HPP treatment of liver becomes part of the production process of dry liver sausages, several bottlenecks must be overcome. For instance, raw liver has to be vacuum-packed before treatment. Ideally, this should be done on an automatic cutting line at the slaughterhouse, which requires a major reorganization of the process. Secondly, transport time between slaughterhouses and companies must be considered as raw liver is a product with a short shelf life that must be quickly handled by food processors. Finally, the additional cost of vacuum packing, HPP treatment, and transport needs to be assessed. Excessive overcost will not be viable for companies.
Our study has certain limitations. First, raw pork liver has been artificially contamina ted, and the use of naturally infected pork liver would have been more relevant, although more difficult to collect at the slaughterhouse with an exploitable/sufficient viral load. Secondly, the detection limit of our cell culture model may not allow the detection of small residual amounts of non-inactivated virus. However, it would be important to know the minimal infectious dose for humans following food consumption. Finally, the technological properties of dried liver sausages made with HPP-treated raw pork liver have been analyzed only by two companies. This represents a small number of manufacturers that need to be confirmed on a larger scale.
## Conclusions
In conclusion, this study is the first to establish HPP parameters (600 MPa for 1 min) effective in reducing the presence of infectious HEV in raw pork liver. Moreover, it is the first to demonstrate the feasibility of incorporating HPP-treated pork liver into food products, thereby supporting novel food safety applications.
## References
1. Purdy, Drexler, Meng et al. (2022) "ICTV virus taxonomy profile: Hepeviridae 2022" *J Gen Virol*
2. Aslan, Balaban (2020) "Hepatitis E virus: epidemiology, diagnosis, clinical manifestations, and treatment" *World J Gastroenterol*
3. Grange, Goldstein, Johnson et al. "Expert Panel, PREDICT Consortium, University of Edinburgh Epigroup members those who wish to remain anonymous. 2021. Ranking the risk of animal-to-human spillover for newly discovered viruses" *Proc Natl Acad Sci*
4. Pavio, Doceul, Bagdassarian et al. (2017) "Recent knowledge on hepatitis E virus in Suidae reservoirs and transmission routes to human" *Vet Res*
5. Sooryanarain, Meng (2020) "Swine hepatitis E virus: cross-species infection, pork safety and chronic infection" *Virus Res*
6. Van Cauteren, Strat, Sommen et al. (2008) "Estimated annual numbers of foodborne pathogen-associated illnesses, hospitalizations, and deaths" *Emerg Infect Dis*
7. Abravanel, Dimeglio, Castanier et al. (2020) "Does HEV-3 subtype play a role in the severity of acute hepatitis E?" *Liver Int*
8. Van Der Poel, Hr, Johne et al. (2018) "Knowledge gaps and research priorities in the prevention and control of hepatitis E virus infection" *Transbound Emerg Dis*
9. Mansuy, Gallian, Dimeglio et al. (2016) "A nationwide survey of hepatitis E viral infection in French blood donors" *Hepatology*
10. Faber, Willrich, Schemmerer et al. (2018) "Hepatitis E virus seroprevalence, seroincidence and serorever sion in the German adult population" *J Viral Hepat*
11. Rose, Lunazzi, Dorenlor et al. (2011) "High prevalence of Hepatitis E virus in French domestic pigs" *Comp Immunol Microbiol Infect Dis*
12. Boxman, Verhoef, Dop et al. (2022) "High prevalence of acute hepatitis E virus infection in pigs in Dutch slaughterhouses" *Int J Food Microbiol*
13. Feurer, Roux, Rossel et al. (2018) "High load of hepatitis E viral RNA in pork livers but absence in pork muscle at French slaughterhouses" *Int J Food Microbiol*
14. Ferri, Vergara (2021) "Hepatitis E virus in the food of animal origin: a review" *Foodborne Pathog Dis*
15. Pavio, Merbah, Thébault (2014) "Frequent hepatitis E virus contamination in food containing raw pork liver" *France. Emerg Infect Dis*
16. Boxman, Jansen, Hägele et al. (2019) "Monitoring of pork liver and meat products on the Dutch market for the presence of HEV RNA" *Int J Food Microbiol*
17. Harrison, Ramos, Wu et al. (2021) "Presence of hepatitis E virus in commercially available pork products" *Int J Food Microbiol*
18. Locus, Lambrecht, Peeters et al. (2023) "Hepatitis E virus in pork meat products and exposure assessment in Belgium" *Int J Food Microbiol*
19. Chatonnat, Julien, Jubinville et al. (2023) "High occurrence of hepatitis E virus in raw pork liver and pork liver pâté produced in the Canadian province of Quebec" *Front Sustain Food Syst*
20. Bennett, Coughlan, Hunt et al. (2024) "Detection of hepatitis E RNA in pork products at point of retail in Ireland -are consumers at risk?" *Int J Food Microbiol*
21. Ricci, Allende, Bolton et al. (2017) "Public health risks associated with hepatitis E virus (HEV) as a food-borne pathogen" *EFS*
22. Hinrichs, Kreitlow, Siekmann et al. (2024) "Changes in hepatitis E virus contamination during the Full-Length Text Applied and Environmental Microbiology"
23. "production of liver sausage from naturally contaminated pig liver and the potential of individual production parameters to reduce hepatitis E virus contamination in the processing chain" *Pathogens*
24. Schilling-Loeffler, Meyer, Wolff et al. (2025) "Determination of hepatitis E virus inactivation during manufacturing of spreadable pork liver sausage and salami-like raw pork sausage" *Int J Food Microbiol*
25. Wolff, Günther, Johne (2022) "Stability of hepatitis E virus after drying on different surfaces" *Food Environ Virol*
26. Wolff, Günther, Albert et al. (2020) "Effect of Sodium chloride, Sodium nitrite and Sodium nitrate on the infectivity of hepatitis E virus" *Food Environ Virol*
27. Barnaud, Rogée, Garry et al. (2012) "Thermal inactivation of infectious hepatitis E virus in experimentally contaminated food" *Appl Environ Microbiol*
28. Johne, Trojnar, Filter et al. (2016) "Thermal stability of hepatitis E virus as estimated by a cell culture method" *Appl Environ Microbiol*
29. Farkas, Hoover (2000) "High pressure processing" *J Food Sci*
30. Huang, Hsu, Wang (2020) "Healthy expectations of high hydrostatic pressure treatment in food processing industry" *J Food Drug Anal*
31. Kingsley (2013) "High pressure processing and its application to the challenge of virus-contaminated foods" *Food Environ Virol*
32. Emmoth, Rovira, Rajkovic et al. (2017) "Inactivation of viruses and bacteriophages as models for swine hepatitis E virus in food matrices" *Food Environ Virol*
33. Govaris, Pexara (2021) "Inactivation of foodborne viruses by High-Pressure Processing (HPP)" *Foods*
34. Buckow, Bingham, Daglas et al. (2017) "High pressure inactivation of selected avian viral pathogens in chicken meat homogenate" *Food Control*
35. Evrendilek (2024) "Effects of high pressure processing on bioavaliabil ity of food components" *J Nutr Food Sci*
36. Bak, Bolumar, Karlsson et al. (2019) "Effect of high pressure treatment on the color of fresh and processed meats: a review" *Crit Rev Food Sci Nutr*
37. Nasheri, Doctor, Chen et al. (2020) "Evaluation of highpressure processing in inactivation of the hepatitis E virus" *Front Microbiol*
38. Johne, Wolff, Gadicherla et al. (2021) "Stability of hepatitis E virus at high hydrostatic pressure processing" *Int J Food Microbiol*
39. Bouquet, Alexandre, Lamballerie et al. (2023) "Effect of high hydrostatic pressure processing and holder pasteurization of human milk on inactivation of human coronavirus 229E and hepatitis E virus" *Viruses*
40. Pellerin, Hirchaud, Blanchard et al. (2021) "Characteri zation of a cell culture system of persistent hepatitis E virus infection in the human HepaRG hepatic cell line" *Viruses*
41. Pellerin, Trabucco, Capai et al. (2022) "Low prevalence of hepatitis E virus in the liver of Corsican pigs slaughtered after 12 months despite high antibody seroprevalence" *Transbound Emerg Dis*
42. Grove, Lee, Stewart et al. (2009) "Development of a high pressure processing inactivation model for hepatitis A virus" *J Food Prot*
43. Kingsley, Chen (2009) "Influence of pH, salt, and temperature on pressure inactivation of hepatitis A virus" *Int J Food Microbiol*
44. 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*
45. Murchie, Kelly, Wiley et al. (2007) "Inactivation of a calicivirus and enterovirus in shellfish by high pressure" *Innov Food Sci Emerg Technol*
46. Araud, Dicaprio, Yang et al. (2015) "High-pressure inactivation of rotaviruses: role of treatment temperature and strain diversity in virus inactivation" *Appl Environ Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12482455&blobtype=pdf | # BMC Infectious Diseases
Rohitha Muthugala, Erandi Ekanayake, Mihidum Govinna, Shashini Wakkumbura, Yashodha Samarajeewa, Nipuni Arachchige, Achini Weerathunga, Lakmali Rajamanthri, Gaya Ranawaka, Thulani Pattiyakumbura, Dhanushka Dasanayake
## Abstract
Background Acute febrile illnesses, including dengue fever, are common causes of hospitalization in Sri Lanka. However, a significant proportion of clinically suspected dengue cases tested negative for dengue-specific markers, raising concerns about alternative infectious causes such as hantavirus. This study aimed to detect hantavirus infections among patients clinically suspected of dengue who tested negative for dengue NS1 and IgM antibodies and to analyze the epidemiology of hantavirus infections in febrile patients.Methodology A descriptive cross-sectional study was conducted at the Department of Virology, National Hospital Kandy, from January to August 2023. A total of 415 serum samples from clinically suspected dengue patients who tested negative for dengue NS1 and IgM antibodies were selected. Hantavirus detection was performed using real-time RT-PCR and IgM antibody testing. Additional tests for flavivirus, alphavirus, and leptospirosis were also conducted. Demographic, clinical, and laboratory data were collected from hantavirus-positive cases.
ResultsAmong the 415 samples, 21 (5.1%) tested positive for hantavirus IgM antibodies. None of the samples tested positive for hantavirus RNA. The demographic analysis revealed no significant difference in age or gender between hantavirus-positive and negative patients. Co-infection with leptospirosis was observed in one patient. In addition to fever, clinical features of hantavirus infection included dry cough (41.2%), vomiting (35.3%), and shortness of breath (23.5%). Laboratory findings showed elevated liver enzymes (AST: 61.5%, ALT: 53.8%), elevated creatinine (50%), and elevated CRP (72.7%).
ConclusionThe putative 5.1% hantavirus seropositivity rate among clinically suspected dengue patients suggests that hantavirus should be considered in differential diagnoses. Although no hantavirus RNA was detected, the presence of IgM antibodies and negativity to other potential infections indicates a potential recent infection or probable cases. The study highlights the challenge of diagnosing hantavirus due to its similar presentation to dengue fever and the need for improved diagnostic testing. Early diagnosis and tailored management can improve
## Background
Acute febrile illnesses are a leading cause of hospitalization in Sri Lanka [1]. Among these, dengue fever is frequently suspected in patients presenting with fever and thrombocytopenia [2]. Routine diagnostic tests, such as NS1 antigen detection and IgM/IgG antibody assays, are commonly used to confirm the diagnosis. However, laboratory data indicate that a significant proportion of suspected dengue cases test negative, raising concerns about alternative infectious causes, including leptospirosis, chikungunya, typhus, and hantavirus infections, which share similar clinical features [3].
The Department of Virology at the National Hospital Kandy receives approximately 300 samples each month from clinically suspected dengue patients, yet around 70% of these test negative for dengue. Despite this high percentage, these cases are not routinely screened for other potential pathogens, including hantavirus. In the South Asian region, a considerable number of clinically suspected dengue patients were negative for laboratory testing for dengue [4][5][6].
Hantavirus is an emerging zoonotic infection transmitted by rodents and has become a growing public health concern in various parts of the world, including South Asia [7]. The virus causes two distinct clinical syndromes: Hemorrhagic Fever with Renal Syndrome (HFRS) and Hantavirus Pulmonary Syndrome (HPS) [8]. In addition to these classical forms, hantavirus infection can also present with mild or atypical symptoms [9][10][11]. Clinical manifestations such as fever, thrombocytopenia, severe headaches, nausea, and vascular permeability changes often resemble dengue fever, leading to frequent misdiagnosis [12]. Notably, early-stage HFRS often presents as an undifferentiated febrile illness, further complicating clinical differentiation [13]. Importantly, the aggressive fluid management recommended for dengue is not suitable for hantavirus infection, highlighting the critical need for accurate diagnosis and tailored clinical management.
Though early identification of hantavirus cases is essential for improving patient outcomes and guiding public health interventions, hantavirus infection remains underdiagnosed in Sri Lanka due to limited routine diagnostic availability [10,14].
This study aimed to detect hantavirus infections among clinically suspected dengue patients who tested negative for dengue NS1 antigen and dengue-specific IgM antibodies. By analyzing the demographic, clinical, and laboratory characteristics of hantavirus-positive cases, this research sought to address a critical gap in the understanding of hantavirus epidemiology in Sri Lanka. Early detection and accurate diagnosis of hantavirus infection can improve clinical management, reduce mortality, and develop more effective preventive strategies. Additionally, this study's findings are expected to raise awareness among healthcare professionals and support the expansion of diagnostic facilities for hantavirus infections in Sri Lanka.
## Methodology
This descriptive cross-sectional study was conducted at the Department of Virology, National Hospital Kandy. The study design was chosen for its simplicity and effectiveness in determining disease prevalence within a defined period.
The study population comprised clinically suspected dengue patients admitted to the National Hospital Kandy between January and August 2023. A total of 415 samples were selected through systematic sampling from patients who tested negative for dengue NSI and/or dengue-specific IgM by an immunochromatographic assay (STAN-DARD™ Q Dengue Duo, SD Diagnostics, South Korea). Patients included in the study were those presenting with fever and thrombocytopenia and clinically diagnosed with dengue fever by the treating physician according to the WHO case definition. Patients who tested positive for dengue NS1 antigen and/or dengue-specific IgM antibodies were excluded first and, then every. third sample with negative dengue serology was selected and samples with insufficient volume or clinical history was not included. Additionally, those with a confirmed diagnosis of another illness at the time of sample collection were also excluded.
After routine dengue NS1 and dengue-specific IgM testing, the remaining serum samples from suspected dengue cases were used for hantavirus detection through in-house real-time RT-PCR assay [15] and hantavirusspecific IgM antibody testing using a commercial immunoblot assay (Mikrogen recomLine HantaPlus IgM, Mikrogen, Germany). To rule out other viral and bacterial infections with similar clinical presentations, additional tests were conducted, including in-house RT-PCR assays for flavivirus RNA detection [16] and alphavirus RNA detection RT-PCR [17]. Furthermore, leptospirosis was excluded using real-time PCR following the World Organization for Animal Health (WOAH) guidelines and anti-leptospirosis IgM testing via a commercial ELISA kit (Panbio, Leptospira IgM ELISA, Abbott). For molecular assays, pathogen DNA/RNA was extracted and purified patient outcomes and guide public health responses. Further studies and expanded diagnostic capabilities are recommended to better understand hantavirus prevalence and to improve clinical care for affected patients.
from blood samples by using commercial pathogen DNA/RNA extraction kit (Spin Star, ADT BioTec, Malaysia) according to manufacturer's instruction.
Patient demographic information was retrieved from laboratory request forms, while clinical history and laboratory findings were extracted from hospital records for cases that tested positive for hantavirus by real-time RT-PCR or IgM detection. Data for each positive patient were recorded in an Excel sheet using laboratory reference numbers to maintain anonymity and confidentiality.
Ethical approval for the study was obtained from the Ethics Review Committees of the National Hospital Kandy (NHK/ERC/13/2023), Faculty of Allied Health Sciences, University of Peradeniya (AHS/ERC/2023/062) and Medical Research Institute (ERC/14/2022). The Director of the hospital granted permission to conduct the study. Since no personal identification or direct patient interaction was involved in the study, informed consent was waived by the Ethics Review Committees.
## Results
From January 1st to August 15th, 2023, a total of 1,762 dengue detection samples were processed, of which 361 (20.5%) tested positive for dengue NS1 antigen and/or dengue-specific IgM antibodies. The remaining 1,401 samples (79.5%) were considered negative for dengue. Among them, 415 serum samples selected systematically for hantavirus testing, 21 individuals (5.1%) were found to be positive for anti-hantavirus IgM antibodies (Table 1). All IgM positive samples gave positive band for Hantaan (HaN) virus nucleocapsid antigen coated region and 19 of them gave bands on Dobrava antigen (DobN) and Seoul virus antigen (SeoN) coated region indicating infection could be due to hantavirus which produce IgM antibodies cross reacting with Hantaan, Dobrava and Seoul virus antigens. However, hantavirus RNA was not detected in any of the samples.
## Demographic characteristics of patients with hantavirus IgM
Of the 415 samples tested, 58.8% (244) were males and 41.2% (171) were females. Among the 21 (5.1%) positive cases, 9 (42.9%) were males and 12 (57.1%) were females (Table 2). Pearson Chi-Square analysis did not reveal a significant relationship between gender and hantavirus positivity (p = 0.128).
The age of patients ranged from 2 to 85 years (Fig. 1), with a mean age of 32.26. The mean age of hantaviruspositive cases was 34.71 years. There was no statistically significant difference in the mean age between positive and negative cases (p = 0.55).
All 415 samples were subjected to RT-PCR testing for Alphavirus and Flavivirus. None of the samples were positive for Alphavirus RNA, while six samples (1.4%) tested
## Clinical and laboratory characteristics of probable hantavirus infections
Among the 17 patients with available data, clinical and laboratory findings were analyzed. Fever was reported in 100% of the patients, with headache present in 23.5% (4 patients). Vomiting and coughing were noted in 35.3% and 41.2% of patients, respectively. Abdominal pain was rare, reported in only 5.9% of patients. Arthralgia and myalgia were found in 35.3% of patients, while diarrhoea occurred in 17.6%. Back pain was present in 23.5%, and shortness of breath (SOB) was observed in 23.5%. Wheezing was noted in just one patient (5.9%). Anaemia was reported in one patient (5.9%).
Elevated WBC counts were observed in 23.5% of patients, while 29.4% had low total WBC counts. Among the elevated WBC, 53% were lymphocytes and 47% were neutrophils. Hematocrit (HCT) was decreased in 64.7% of patients, and 47.1% of the patients had a decreased platelet count. Elevated AST levels were found in 61.5% (8/13) of patients, while 53.8% (7/13) had elevated ALT levels. Elevated creatinine was seen in 50% of patients, and CRP levels were elevated in eight patients (72.7%). Chest X-ray changes indicative of pulmonary involvement were seen in 3 patients, along with crepitus and dry cough in over 40% of the cases. Additionally, two patients were diagnosed with renal failure (9.5%). These findings highlight the diverse clinical presentations of hantavirus infections, including respiratory and renal involvement, as well as laboratory abnormalities commonly associated with the disease (Tables 4 and5).
During the follow-up period, two patients developed classical hemorrhagic fever with renal syndrome (HFRS). Of these, one patient experienced severe complications, requiring intensive care unit (ICU) admission and dialysis for renal support. Patient required intensive care unit (ICU) admission and dialysis unfortunately succumbed to death due to sepsis following secondary bacterial infection.
## Discussion
Hantavirus infection is emerging as a global health concern, but the exact prevalence of the disease in Sri Lanka remains unclear. This study aimed to identify early-stage and mild/atypical forms of hantavirus infection among clinically suspected dengue patients who tested negative for dengue. Among the tested patients, 79.5% tested negative for dengue, highlighting the potential for misdiagnoses and delays in accurate diagnosis. These challenges
## Prevalence and demographic profile of probable hantavirus infections
The study found a 5.1% positivity rate for hantavirus among clinically suspected dengue patients, which is consistent with similar studies in the region [18,19]. For instance, a study in Indonesia reported a 4.23% prevalence of hantavirus among dengue-like patients, while a Cambodia study showed a 4% prevalence [20,21]. In-house real-time RT-PCR assay used for detection of hantavirus RNA was replication of method originally described by Mohamed et al., in 2013. It is SYBER green based real-time RT-PCR which has ability to detect L segment of the hantavirus genome which is the most conserved part of the viral genome in different type of hantavirus clade with analytical sensitive of less than 100 RNA copies/mL (15). Assay was validated locally by using synthetic genes of 12 different types of hantaviruses kindly provided by the authors described this assay originally. The commercial immunoblot assay (Mikrogen recom-Line HantaPlus IgM, Mikrogen, Germany) has sensitivity of 98.3% and specificity of 94.1%, based on product broacher. There is no recorded cross reactivity with other virus or bacterial antibodies except antibodies against Sand fly fever virus (SSFV) and anti-malaria antibodies. Assay itself has test line for SSFV. And there was no evidence of other Bunyavirus infections in the country other than hantaviruses and malaria was eliminated since 2016. This test assay was used in previously published serological studies on hantaviruses in Europe [22,23].
Interestingly, none of the study samples tested positive for hantavirus RNA, which could be due to factors such as low-level viremia or a short duration of viremia [24]. Hantavirus infection may present with a transient or low viral load, which may not be detectable during the sampling period [25]. In addition to that timing of sample collection and suboptimal sample handling before sending in to laboratory may have contributed for negative PCR results. Instead, Hantavirus IgM antibodies, which can remain detectable for 3-4 months after infection, may indicate recent past exposure to the virus [26]. Demonstration of four-fold rise of hantavirus IgG is confirmatory; however, it was not performed due to logistical reasons. Despite the presence of hantavirus IgM antibodies, these patients tested negative for other common febrile illnesses, including dengue, alphavirus infections, and leptospirosis, helping to narrow the potential causes of their symptoms. Therefore, these cases can be considered as probable or suspected hantavirus cases. All hantavirus IgM positive samples gave positive band on Hantaan virus nucleocapsid antigen coated region and majority has given bands on Hantaan, Dobrava and Seoul virus antigens coated region indicating infection could be due to hantavirus which antigenically related to Hantaan, Dobrava and Soul viruses. Serological evidence of Hantaan virus or Hantaan-like virus has detected in Sri Lanka previously and serological cross reactivity was noted [10,14].
Additionally, six samples tested positive for flavivirus PCR, which confirmed these cases as dengue infections. Initially, these cases were negative for the NS1 antigen but were later confirmed as dengue through PCR. This discrepancy may be attributed to the low sensitivity of NS1 detection, particularly in secondary dengue cases, where the immune response might not produce detectable levels of the NS1 antigen [27]. This highlights the need for multiple diagnostic methods to accurately identify and differentiate between similar febrile illnesses.
Regarding demographic factors, gender was not significantly associated with hantavirus seropositivity. Among the 21 positive cases, 42.9% were male and 57.1% were female, which contrasts with most studies, where a higher proportion of males are affected [14]. A significant difference was found in age groups, with most positive cases occurring in individuals under 10 years and over 61 years. While hantavirus infections in most regions, including the Americas, Asia, and Europe, primarily affect adults, studies from Iran and Barbados have highlighted the occurrence of pediatric cases, suggesting varying age distributions of hantavirus infections across different regions [28].
## Clinical and laboratory factors in probable hantavirus patients
Most hantavirus-positive patients (76.2%) did not require special care, and only 14.3% required intensive care. Two deaths were reported, corresponding to a mortality rate of 9.5%, which is slightly lower than that observed in similar studies. Fever was present in all positive cases, while other symptoms such as cough, myalgia, and arthralgia were reported in a subset of patients. The absence of severe symptoms in most cases suggests that the majority of infections were mild. Chest X-rays revealed pulmonary involvement in three patients, and renal failure was diagnosed in two patients, with symptoms consistent with both Hemorrhagic Fever with Renal Syndrome (HFRS) and Hantavirus Pulmonary Syndrome (HPS). These findings align with recent studies conducted in Sri Lanka, indicating that Hantavirus can present with a range of clinical manifestations [11]. Without specific efficient antiviral therapy, these patients were managed symptomatically with supportive care.
Regarding laboratory results, thrombocytopenia was observed in 8 out of 17 positive cases, although it was not a common finding across all patients. Leukocytosis and elevated hematocrit were observed in a few patients; however, none of the cases showed a significant increase in hematocrit. These laboratory abnormalities, which are typically seen in the later stages of the disease, may have been missed because testing was conducted during the early or mild stages of infection [29]. Additionally, elevated ALT, AST, and CRP levels were noted in most positive cases, consistent with findings from a similar study conducted in Indonesia. These markers are indicative of liver involvement and systemic inflammation, reinforcing the need for comprehensive testing during the clinical evaluation of patients [30,31].
## Conclusion
This study determined that the positivity rate of probable hantavirus infection cases among dengue-suspected patients at the National Hospital Kandy was 5.1%. The majority of these patients experienced mild febrile illness, with many of the dengue-negative samples testing negative for other common pathogens, suggesting that the fevers in these patients were likely caused by illnesses other than the tested pathogens. There was no significant association between demographic, clinical, or laboratory factors and hantavirus positivity.
## Limitations of the study
Based on IgM results alone are not sufficient to diagnose hantavirus infection with certainty as hantavirus IgM can persist 2-3 months following infection. None of the patients were positive for hanta virus RNA detection PCR. Demonstration of specific IgG was not available due to logistical reasons. Therefore, uncertainty of the confirmatory diagnosis is the key limitation in this study. The other limitation of this study was the incomplete documentation of laboratory-related factors. Some investigations were not ordered, and several patients 'clinical records were misplaced, which may have affected the completeness of the data. Renal functions and liver functions tests routinely not performed in all dengue-like patients in the hospital especially in mild cases due to rational utilization of limited resources.
## Implications and recommendations
Hantavirus infection should be considered as part of the differential diagnosis for dengue-like illnesses in Sri Lanka. The findings from this study provide valuable insights into the clinical spectrum of hantavirus infections, with the first demonstration of mild disease forms in the country, characterized by a lack of severe renal or pulmonary involvement. The study provides the foundation for future research and can help raise awareness within the medical community, potentially influencing preventive measures and improvements in hantavirus detection in laboratories.
Given the limited research on hantavirus in Sri Lanka, further studies are essential to determine its national prevalence and guide public health policies and preventive actions. Lack of adequate diagnostic facilities for detect hantavirus infection is one of the major issue to understand disease burden in the country. In addition to that, accessibility for certain confirmatory diagnostic assays like commercial hanta IgM/IgG Immuno fluorescence assays (IFA) and real-time RT-PCR kits complicate confirmatory diagnosis. The data generated from this study contribute to the growing body of knowledge about hantavirus infections in Sri Lanka and highlight the need for better diagnostics, continued surveillance and research to better understand the disease and its impact on public health.
## References
1. Prasad, Murdoch, Reyburn et al. (2015) "Etiology of severe febrile illness in low-and middle-income countries: a systematic review" *PLoS ONE*
2. Malavige, Jeewandara, Ghouse et al. (2021) "Changing epidemiology of dengue in Sri Lanka-challenges for the future" *PLoS Negl Trop Dis*
3. Singhi, Chaudhary, Varghese et al. (2014) "Tropical fevers: management guidelines" *Indian J Crit Care Med*
5. Sharada, Raghunath, Suma "Detection of NS1 antigen, IgM and IgG antibodies using a commercial dengue rapid test kit for the diagnosis of dengue infection in patients with acute febrile illness" *Indian J Microbiol Res*
6. Rehman, Anwar, Tayyab et al. (2022) "Incidence of dengue fever, serotypes, clinical features, and laboratory markers: a case study of 2019 outbreak at district shangla, KP, Pakistan" *Afr Health Sci*
8. Kodikara, Jayathilake, Kumarasinghe et al. (2016) "Prevalence of NS-1 Status of Clinically Suspected Dengue Patients in a Selected Out" *Patient Setting*
9. Jonsson, Figueiredo, Vapalahti (2010) "A global perspective on hantavirus ecology, epidemiology, and disease" *Clin Microbiol Rev*
10. Manigold, Vial (2014) "Human hantavirus infections: epidemiology, clinical features, pathogenesis and immunology" *Swiss Med Wkly*
11. Ehelepola, Basnayake, Sathkumara et al. (2018) "Two Atypical Cases of Hantavirus Infections from Sri Lanka, Case Rep" *Infect. Dis*
12. Muthugala "Hantavirus hemorrhagic fever with renal syndrome -suspected cases in Sri Lanka: clinical picture and epidemiology from"
13. (2022) *Jpn J Infect Dis*
14. Pattiyakumbura, Pathirathne, Muthugala (2024) "Hantavirus infection in central Sri Lanka -an unusual clinical presentation: a case report" *Access Microbiol*
15. (2022) "Clinical Manifestation | Hantavirus | DHCPP | CDC"
16. Cosgriff (1991) "Mechanisms of disease in hantavirus infection: pathophysiology of hemorrhagic fever with renal syndrome" *Rev Infect Dis*
18. Muthugala (2021) "Hantavirus infection with pulmonary symptoms in North central part of Sri Lanka" *Journal of Clinical Virology Plus*
19. Mohomed (2013) "Development and evaluation of a broad reacting SYBRgreen based quantitative real-time PCR for the detection of different hantaviruses" *J Clin Virol Off Publ Pan Am Soc Clin Virol*
20. (2001)
21. Tanaka (1993) "Rapid identification of flavivirus using the polymerase chain reaction" *J Virol Methods*
22. Pfeffer, Proebster, Kinney et al. (1997) "Genus-specific detection of alphaviruses by a semi-nested reverse transcription-polymerase chain reaction" *Am J Trop Med Hyg*
23. Chandy, Yoshimatsu, Ulrich et al. (2007) "Seroepidemiological study on hantavirus infections in India" *Trans R Soc Trop Med Hyg*
24. Tortosa (2024) "Seroprevalence of hantavirus infection in non-epidemic settings over four decades: a systematic review and meta-analysis"
25. (1924)
26. Lukman, Kosasih, Ibrahim et al. (2019) "A review of hantavirus research in Indonesia: prevalence in humans and rodents, and the discovery of Serang virus" *Viruses*
28. Nouhin (2023) "Updates of Hantavirus infection risk at rodent-human" *Int J Infect Dis*
29. Hofmann, Ulrich, Mehl (2024) "Hantavirus disease cluster caused by Seoul virus" *Ger EID*
30. Salimović-Bešić, Hrvo, Zahirović et al. (2023) "Expansion of hantavirus infection during the SARS-CoV-2 pandemic in Bosnia and Herzegovina, 2021" *J Med Microbiol*
32. Douglas, Samuels, Iheozor-Ejiofor et al. (2008) "Serological evidence of human orthohantavirus infections in Barbados" *Pathogens*
33. Afzal "Hantavirus: an overview and advancements in therapeutic approaches for infection"
34. Mattar, Guzmán, Figueiredo (2015) "Diagnosis of hantavirus infection in humans" *Expert Rev Anti-Infect Ther*
35. Watson, Sargianou, Papa et al. (2013) "Epidemiology of hantavirus infections in humans: a comprehensive, global overview" *Crit Rev Microbiol*
36. Fry (2011) "The diagnostic sensitivity of dengue rapid test assays is significantly enhanced by using a combined antigen and antibody testing approach" *PLoS Negl Trop Dis*
37. (2025) "Global Perspective | SpringerLink"
38. (2025) "Manual Professional MSD"
39. Lie (2018) "Case report: two confirmed cases of human Seoul virus infections in Indonesia" *BMC Infect Dis* |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.