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=PMC12506246&blobtype=pdf | # Reproducible generation of Nipah virus pseudovirions with uniform incorporation of F and G surface glycoproteins for highthroughput neutralization assays
Easwaran Sreekumar, Geetu Varghese, Vivek Vijay, Sreeja Sreedevi, Santhik Lupitha, Priya Prabhakaran, Sushama Aswathyraj, Anitha Moorkoth, Niyas Pulloor
## Abstract
Nipah virus (NiV) is a pathogen to be handled in BSL-4 facilities. Multiple surrogate systems such as virus-like particles (VLPs) and pseudoviruses enable carrying out NiV neutralization assays and study of virus entry pathways in biosafety level-2 (BSL-2) facilities. These are dual protein expression and surface display systems comprising NiV structural glycoproteins F and G as the key components. During generation of NiV VLPs or pseudovirions, ensuring batch to batch uniformity is a major concern due to the lack of proportionate incorporation of these proteins in the producer cells as well as their consistent incorporation in the particles. We established HEK293 pseudovirion producer cells that stably co-incorporated NiV F and G proteins to address the issue. Fluorescence-activated cell sorting (FACS) analysis of clonally selected cells for high and uniform level F and G protein co-incorporation; and their further expansion were carried out to further refine the system. High titer vesicular stomatitis virus (VSV)-based pseudoviruses which showed consistent incorporation of both the glycoprotein were reproducibly generated from these producer cells. In functional assays, these Nipah pseudovirions (NiV) exhibited a dosedependent neutralization by commercial anti-NiV F and G antibodies as well as by convalescent serum from Nipah recovered patients. A pseudovirus neutralization test (PVNT) with a secreted alkaline phosphatase (SEAP) as the reporter was established in the study. The assay supports high-throughput adaptability with a quick turnaround time. It will aid large-scale human and animal serosurveillance studies in Nipah endemic regions as well as screening of virus entry inhibitors and monoclonal antibodies.
## Introduction
Nipah virus (NiV) is a highly lethal zoonotic paramyxovirus causing fatal encephalitis in humans [1]. Fruit bats of the Pteropus genus serve as a natural reservoir of the virus [2]. The initial Nipah cases identified in 1998 in Malaysia and Singapore were caused by spill over infections from pigs to humans [3][4][5]. On the other hand, a human-to-human transmission was evident in the reemerging outbreaks of the disease in Bangladesh and neighbouring regions in India [6,7]. In southern peninsular India the first outbreak occurred in 2018 in Kerala with 18 laboratory-confirmed cases [8]. Phylogenetic analysis of the NiV strain causing this outbreak revealed a 96.15% similarity to the Bangladesh strain [9]. Since the first report from India in 2001, the state of Kerala witnessed multiple Nipah outbreaks since 2018 [10,11]. Sequential outbreaks of the disease in the Indian subcontinent have raised concerns of a re-emerging trend of this highly transmissible virus posing a major public health threat. It is one among the WHO priority pathogens that require urgent research attention for effective containment measures.
The high fatality rate (around 90% in recent outbreaks) [12] as well as the lack of effective antiviral treatment and preventive vaccines impose any studies on infectious NiV strictly restricted to BSL-4 bio-containment facilities. The absence of easily employable surveillance systems poses hindrance to prevalence studies in both humans and animals, and also in the effective and early detection of subclinical cases that may help in prevention of disease outbreaks. To circumvent this difficulty and facilitate research, several surrogate systems employing viruslike particles and pseudovirions that support studies in a BSL-2 facility have been developed.
The 18.2 kb negative-sense RNA genome of NiV codes six structural proteins and three non-structural proteins [13,14]. The structural proteins G and F are the key envelope glycoproteins facilitating NiV binding to its cellular receptor. The glycoprotein G binds to the Ephrin B2 or B3 receptor triggering a conformational change in the fusion protein F, supporting the viral envelope fusion with the host cell membrane [15,16]. Earlier studies with surrogate viral systems involved the use of NiV-like particles (VLPs) and pseudotyped viruses using lentivirus or Vesicular Stomatitis virus (VSV)-based systems [13,17,18]. The NiV VLPs that co-express the surface glycoproteins G, F, and the matrix protein M have been used in neutralization assays [3]. A pseudovirus production system for henipavirus using the MuLV packaging cell line expressing the LACZ reporter gene has been reported [13]. Exploiting the broad host range and robust replication properties of VSV, a recombinant VSV (rVSV-G*ΔG)-based system with glycoprotein G coding region deleted has been used to generate pseudotypes of several heterologous viruses including NiV [19][20][21][22]. Antibody neutralization assay using the pseudotyped NiV showed a good correlation with the gold standard PRNT assay that uses infectious NiV indicating their functional resemblance [23]. In addition to the replication defective VLP and pseudovirus-based systems, a recombinant, replication competent NiV F and G chimeric Cedar virus platform has also been generated recently as a surrogate neutralization system [24].
During their generation, all these surrogate systems need multiple transfections of plasmids encoding NiV F/G proteins or matrix protein M into the producer cells. There is a critical need for optimization of plasmid concentrations used in the transfection in order to avoid inconsistency in the surface incorporation of these proteins in the producer cells. This has remained challenging [25]. As a consequence, there can be variable incorporation rates of these proteins in the VLPs or pseudovirions generated and batch-to-batch variability in the infectious virus titre. There have been no major attempts made to address this issue and generate systems that can produce high titer NiV VLPs or pseudovirions. In most of the currently available surrogate systems, the reporters used as indicators of infection are fluorescence or luminescencebased, making them less amenable to high throughput automation for large scale use in surveillance or screening studies. In the present study, we address these gaps by reproducibly generating high titer VSV-based NiV pseudoviruses from F and G stably-transfected and clonally selected producer cells. A secreted alkaline phosphatase (SEAP) was incorporated as a reporter for easy readout. We also show the use of these pseudovirions in neutralization assays employing commercial antibodies as well as convalescent serum from Nipah recovered individuals.
## Materials and methods
## Generation of NiV F and G expression vectors
Complete coding region sequences of NiV F (1640 bp; position 6650-8290) and G (1808 bp; position 8939-10747) genes derived from the National Center for Biotechnology Information (NCBI) database (GenBank Accession No.MH523642; NiV strain from the outbreak in Kerala in 2018 [9]) were used to synthesize gene constructs (BR Biochem, India). Plasmid vector backbones with a CMV promoter, pLXSN and pcDNA3.1 Hygro, were derived from the mammalian expression plasmids pLXSN-Axl (Addgene # 65222) and pHMC122 (a kind gift from Dr. Erica Ollman Saphire, La Jolla Institute of Immunology, USA). NiV F coding region was cloned to the EcoRI-BamHI linearized pLXSN vector backbone and the NiV G coding region was cloned to the KpnI-HindIII linearized pcDNA3.1 Hygro vector backbone using standard cloning procedures. Chemically competent Escherichia coli (E. coli) DH5α strains were transformed with the ligated products and the plasmid DNA isolated from NiV F and G positive clones were used for transfection of mammalian cells.
## Cell lines
BHK-21 baby hamster kidney cells (ATCC; CCL-10) and HEK293 human embryonic kidney cells (ATCC; CRL-1573), were originally purchased from the American Type Culture Collection (ATCC) and were cultured in high glucose DMEM (Gibco, USA).The culture media were supplemented with 2% or 10% Foetal Bovine Serum (FBS; Gibco, USA) and 1X antimycotic antibiotic solution (Cat No. 15240062; Sigma, USA) and cells were maintained at 37 °C in a 5% CO 2 atmosphere.
## Plasmid transfections and generation of NiV F & G stable HEK293 producer cells
HEK293 cells (3 × 10 5 ) used for the production of pseudovirions were seeded in 6-well plates for 18 h and co-transfected with the protein expression plasmids pLXSN-NiV F and pcDNA-NiVG (2 µg/ml each) using polyethyleneimine (PEI Branched; Cat No.408727; MW 25000; Sigma, USA) (8 µg/ml). The selection of transfected cells stably expressing both proteins was carried out by growing the cells in the presence of geneticin (625 µg/ml; for the pLXSN-NiVF) and hygromycin (50 µg/ml; for the pcDNA-NiVG) for 60 days.
## Clonal selection of NiV F and G co-expressing HEK293 producer cells
HEK293 producer cells stably expressing NiV F and NiV G proteins were subjected to limiting dilution to separate single cells in individual wells of a 96-well plate to allow clonal expansion. After two rounds of selection, individual colonies were expanded to large-scale cultures and cryopreserved. Three, well-growing clones were selected (293 FG-5F6, 293 FG-7C5, and 293 FG-8C7) for further characterization and pseudovirion production.
## Flow cytometry analysis of membrane expression of F and G proteins in monoclonal HEK293 producer cells
293 FG-5F6, 293 FG-7C5, and 293 FG-8C7 cells were trypsinized from T25 flasks and 1 × 10 6 cells were stained with anti-NiV F (12B2; Cat. No Ab02792-1, Absolute antibody; USA) and anti-NiV G (48D3; Cat. No Ab2865-23.0, Absolute antibody; USA) antibodies in flow cytometry staining buffer (2% FBS in Phosphate Buffered Saline (PBS), pH 7.2) for 1 h in ice. After incubation, the cells were centrifuged at 1500 rpm for 5 min followed by washing the pellet in FACS washing buffer (0.5% BSA in PBS) to remove the unbound antibodies. The cell pellet was then treated with mouse (Anti-mouse IgG Fab 2 AlexaFluor; CST # 4408 S; Cell Signalling Technology (CST), MA, USA) and rabbit (Anti-rabbit IgG Fab 2 Alexa Fluor ® 594; CST # 8889 S; Cell Signalling Technology, (CST), MA, USA) secondary antibodies to anti-F and anti-G antibodies conjugated with Alexafluor 488 and AlexaFluor 597 respectively for 30 min in ice. Following washing, the cell pellet was resuspended in FACS staining buffer and the percentage of incorporation of F and G proteins was analyzed in a flow cytometer (BD FACS Aria Fusion, BD Biosciences, USA).
## Immunofluorescence detection of F & G proteins in monoclonal HEK293 producer cells
3 × 10 4 cells from expanded 293FG-5F6 monoclonal cells were cultured in 24-well plates overnight, washed with 1X phosphate-buffered saline (PBS), and fixed with paraformaldehyde or pre-chilled acetone-methanol (1:1) at -20 °C for 5 min. The fixative was removed by washing the cells twice with 1X PBS and the cells were blocked using 5% BSA in PBS, with 0.05% Tween-20 for 30 min at room temperature. Subsequently, the cells were stained with anti-NiV F (12B2; Cat. No Ab02792-1, Absolute antibody; USA) and anti-NiV G (48D3; Cat. No Ab2865-23.0, Absolute antibody; USA) primary antibodies (1:500 dilution) overnight at 4°C. Followed by washing twice with 1X PBS, the cells were incubated with Alexafluor 488 conjugated anti-mouse (CST, Cat. # 4408 S, CST, USA) and AlexaFluor 597 conjugated antirabbit secondary antibodies (CST, Cat. # 8889 S, CST, USA) respectively at 1:1000 dilution. DAPI (Cell Signalling Technology, Cat. # 4083 S, CST, MA, USA) (1 µg/ml) was used as the nuclear stain in all experiments.
## Production of NiV pseudovirions
NiV pseudovirus production was carried out as described earlier [26] using the Delta-G-VSV Pseudotyping system (Kerafast Inc. CA, USA). All assays were conducted as per the Institutional Biosafety Committee (IBSC) approved protocols in a BSL-2 laboratory. Initial rVSV G*∆G-SEAP stocks were generated in the laboratory by plasmid transfections as per manufacturer's protocol. These stock viruses were titrated by a spot forming unit Assay As described below, and were found to have a titre of approx. 2.5 × 10 6 SFU per ml. These stocks were diluted in OptiMEM medium to infect the producer HEK293 cells (NiV F and G protein expressing polyclonal or the 293 FG-5F6 monoclonal stable cells) at an MOI 1 for 2 h at 37 °C. After infection, the inoculum was removed, and the cells were washed with PBS at least 3 to 4 times to ensure the removal of trace quantities of any residual VSV-∆G-SEAP virus. The cells were cultured in complete culture medium for further 18-24 h before the pseudovirus harvest. On the day of harvest, the culture supernatants containing the pseudotyped virions were collected, centrifuged at 3000 rpm for 20 min and stored at -80 °C until use. The pseudovirions produced from NiV F and G protein expressing polyclonal HEK293 cells were labelled as N-(PV)FG and those produced from the monoclone 293FG-5F6 cells were labelled as N-(PV) FG-5F6.
## Titration of the pseudovirion stocks
Quantification of the viral particles in rVSV G*∆G-SEAP or NiV pseudovirion stocks were carried out using a colorimetric spot forming unit assay. For this an alkaline phosphatase staining method was established using BCIP/NBT reagent (Cat.No.N1113; Tokyo Chemical Industry, TCI, Japan) to count the actual number of cells infected with pseudovirus and expressing the thermostable SEAP. Single-use aliquots of stock viruses were used to avoid inconsistencies in titer that may result from repeated freeze-thawing. Serial dilutions of the pseudovirus stocks were used to infect the target BHK-21 cells in 24-well plates for 2 h at 37 °C; followed by the removal of the inoculum and washing of the monolayer with 1X PBS. After 24 h incubation in fresh DMEM with 2% FCS containing culture medium, the cell culture plates were heated by placing in a water-bath at 65 °C to inactivate the endogenous alkaline phosphatase. For colorimetric spot counting of the infected cells, supernatants were removed and the monolayers were washed twice with 1X PBS followed by addition of 200 µl/well of BCIP/NBT reagent (1X solution in 20mM Tris-HCl pH 9.5, 2mM MgCl2 and 100mM NaCl) for staining the cells. Chromogenic development in the cells were monitored at room temperature and the reaction was stopped after 2 h by rinsing the wells twice with 1X PBS. After the removal of PBS, the plates were scanned by a ELISpot reader (ImmunoSpot by CTL., USA) and was subjected to automated focus counting by BioSpot 5.0 Professional software. Average number of spots from two corresponding dilutions in three independent experiments were used to calculate the pseudovirus virus titre (SFU per ml) of the inoculum stock.
## Target cell infection and SEAP assay in culture supernatants
BHK21 target cells cultured in 96-well plates were infected with the N-(PV)FG or N-(PV)FG-5F6 pseudovirions by incubating the cells with diluted pseudovirus stocks containing the required number of spot forming units (SFU) for 2 h at 37 °C, As described above. 24 h post-infection, the plates were placed in a water bath at 65 °C and were incubated for 15 min to inactivate the endogenous alkaline phosphatase activity. The culture supernatants were analyzed for SEAP activity by mixing 50 µl of the supernatant with 150 µl of pNPP substrate solution (TCI, Japan; Cat. # N1109). After incubation for 10 min, absorbance at 405 nm (OD 405 ) was measured using GloMax® Discover Microplate Reader (Promega Inc; USA).
## Detection of F & G protein incorporation in NiV pseudovirions by immunostaining
BHK-21 target cells were cultured in 24-well plates and the cell monolayer was infected with 500 SFU N-PV(FG)-5F6-SEAP pseudovirions. BHK-21 cells, infected with N-PV (FG) or rVSV G*∆G-SEAP virus, were used As controls. Pseudovirions were allowed to adsorb to cell surface for 30 min at 25 °C (for fixation with acetone-methanol (1:1)) or for 30 min at 4 °C (for fixation with 4% paraformaldehyde in 1XPBS). After the duration of infection, the cell monolayer was washed twice with PBS and subjected to fixation. Immunostaining was done with specific antibodies As described above. The incorporated F and G proteins in the pseudovirions were microscopically visualized by imaging at 60X magnification in a Leica Stellaris5 confocal microscope (Leica Microsystems, Germany).
## NiV pseudovirus-based neutralization assays
BHK-21 target cells (1 × 10 4 ) were seeded in 96-well plates in DMEM containing 10% FBS and 1X antimycotic antibiotic solution and incubated at 37 °C in a CO 2 incubator. N-PV(FG)-5F6-SEAP pseudovirions (120 SFU) were mixed with serial dilutions of commercial anti-NiV F and G antibodies or convalescent human sera from Nipah recovered patients [27] and incubated at 25 °C for 1 h. Anti-rabies human monoclonal antibody 8889 S (17C7); Rabishield-100 (40IU/ml), a NiV non-neutralizing antibody, purchased from the Serum Institute of India Pvt Ltd or serum from two healthy individuals were used As the negative controls in the neutralization Assays. Serum samples were heat inactivated at 56 °C for 30 min before using in the assays to inactivate the complement. The pseudovirus-sera/antibody mixture was added to the monolayer of BHK-21 cells. After 2 h infection at 37 °C, the mixture was removed from the cells, replenished with DMEM containing 2% FBS and 1X antimycotic antibiotic solution, and incubation was continued for 24 h. The cell culture supernatants were harvested; and incubated at 65 °C for 15 min to inactivate the endogenous alkaline phosphatase activity and the secreted alkaline phosphatase (SEAP) was measured as described above. Percentage neutralization was calculated with respect to non-neutralized pseudovirus infected controls using the formula:
$$% Neutralization = [ 1 - (OD T est Sample-OD Cell control) (OD V irus control-OD Cell control) × 100 ]$$
## Statistical analysis
Statistical analyses were performed using Graphpad prism 8.0 software. Pearson's correlation analysis was used to quantify the pixel-to-pixel proportionality in the signal levels of the two channels in immunofluorescence experiments and to analyze the correlation between Spot Forming Units and SEAP absorbances. In experiments with control and test groups, Student's t-test (unpaired) or Wilcoxon signed-rank test was used to compare the significance between the groups. P value < 0.05 was considered statistically significant.
## Results
The general scheme of experiments followed to generate and characterize the NiV pseudovirions is shown in Fig. 1.
## Generation of pseudovirions from HEK293 producer cells stably incorporating NiV F and G protein
HEK293 producer cells were co-transfected with the pLXSN-NiVF and pcDNA3.1-NiVG clone plasmids and selected with the antibiotics hygromycin and geneticin (G418) to generate stably transfected polyclonal cells. The co-incorporation of F and G proteins in the transfected cells was confirmed by immunofluorescence analysis using specific antibodies (Fig. 2A). The formation of syncytium cells with multiple nuclei due to the fusion of adjacent cells were visible in the transiently transfected HEK293 cells expressing F and G proteins. However, in cells selected for stable expression of these proteins, there were predominantly aggregation of cells with partial fusions of the membrane, rather than formation of larger syncytial cells (Supplementary Fig. 1 A andB).
Infection of these cells with 2.5 × 10 6 SFU/ml of the rVSV-G*∆G-SEAP virus stock generated N-PV(FG)-SEAP pseudovirions 24 h post-transfection. This was evidenced by enhanced SEAP activity in the supernatants of BHK21 target cells infected with this virus compared to mock infection control (Fig. 2B).
## Clonal selection and characterization of NiV F and G co-expressing monoclonal HEK293 producer cells
Stably co-transfected HEK293 cells with NiVF and G expressing plasmids (i.e. polyclonal producer cells that may contain clones with varying levels of NiV F and G expression) were further subjected to two successive rounds of selection under the selection pressure of hygromycin and geneticin. These cells were subjected to limiting dilution for the selection of single cell clones. Three well-defined and rapidly growing monoclones selected from 96-well plates, named as 293FG-5F6, 293FG-7C5, and 293FG-8C7, were expanded in 24-well plates and subjected to FACS analysis to determine the individual As well As the co-incorporation levels of F and G proteins. These three monoclones had respectively 90.5%, 88.9%, and 29.6% cells with equal expression of NiV F and G proteins (Fig. 2C). The best among them, 293FG-5F6 monoclonal cells, when subjected to immunofluorescence analysis, indicated uniform incorporation levels and membrane co-localization of the NiV F and G proteins confirming the observation in FACS analysis (Fig. 2D). Hence, 293FG-5F6 clone was further expanded and selected as producer cells for large-scale N-PV(FG)-5F6 pseudovirus production.
## Reproducible generation of high-titre NiV N-PV(FG)-5F6 pseudovirions
BHK-21 monolayer cells infected with two-fold serial dilutions of N-PV(FG)-5F6 pseudovirions were subjected to alkaline phosphatase staining as described in the methods section. The infected cells revealed purple coloured spots indicating the production of SEAP with their counts decreasing in proportion to the dilutions (Fig. 3A). The virus titre, calculated based on the spot forming units (SFU), was found to be 1 × 10 4 SFU/ml (Fig. 3B). Using the same approach, we also estimated the lower limit of pseudovirion detection by SEAP assay. Serial dilutions of the virus inoculum with varying SFUs were used to infect BHK-21 cells and the SEAP activity was measured after 24 h. The SEAP activity correlated well with the virus quantity determined by the Spot forming assay (Fig. 3C). A perceptible difference in OD 405 was observed in cells infected with pseudovirions as Low as 4-8 SFU; and an inoculum with 120 SFU of virus was found to give an approximate OD 405 value of 1.0 which was selected for subsequent assays.
N-PV(FG) pseudovirions from the polyclonal producer cells and N-PV(FG)-5F6 pseudovirions from the 293FG-5F6 monoclonal cells were generated by infecting these cells with 2.5 × 10 6 SFU/ml of rVSV-G*∆G-SEAP virus stock. Equal volumes (100 µl) of each of the produced pseudovirions were used to infect the BHK-21 target cells to compare the SEAP activity in the culture supernatants. The results indicated a significantly high titer pseudovirion production by the monoclonal cells (Fig. 3D). Kinetic analysis of the SEAP activity in culture supernatants from these cells revealed that the activity reaches maximum by 24 h post-infection and thereafter remains stable (Fig. 3E). So this time-point was chosen as the endpoint in subsequent biological assays.
In order to check the reproducibility of pseudovirion production, three batches of N-PV(FG)-5F6-SEAP viruses were generated from 293FG-5F6 clonal cells under identical conditions of rVSV-G*∆G-SEAP infection. These three stock viruses were found to have consistent infectivity and similar virus titre as indicated by the (See figure on previous page.) Fig. 2 Analysis of co-expression of NiV surface glycoproteins F and G in HEK293 polyclonal producer cells and clonally expanded stable HEK293 cell lines.A Confocal microscopy image of immunofluorescence analysis of polyclonal HEK293 producer cells stably expressing NiV F and G proteins. Cells were stained with primary antibodies to Nipah G (48D3) and Nipah F (12B2) proteins, respectively (Magnification, 60X; Scale bar, 20 μm). B Testing pseudovirion production by transfected HEK293 polyclonal producer cells. N-PV(FG) pseudovirions were generated by infecting the NiV F & G stably expressing producer HEK293cells with rVSV-G*∆G-SEAP virus stock. The harvested virus was used to infect BHK21 target cells and the SEAP activity measurements were done in the culture supernatants after 24 h of infection. Supernatants from un-transfected HEK293 producer cells were used for control infection of target BHK21 cells; and the SEAP activity was compared. Experiments were done in duplicates and data presented as Mean ± SD of three independent experiments (N = 6); '*' indicates P < 0.05. Un-paired Student t-test was used for statistical analysis comparing the two groups. C NiV F and G protein expression in transfected, clonally selected HEK293 cells. Suspensions of cells from expanded monoclones 293FG-5F6, 293FG-7C5 and 293FG-8C7 were stained with NiV anti-G and anti-F antibodies As described in the methods; and analysed by FACS. The percentage incorporation of G and F proteins was found to be 90.5%, 88.9% and 29.6%, respectively. D Immunofluorescence analysis of NiV G and F protein co-expression in 293FG-5F6 monoclonal cells. The fixed cells were stained with primary antibodies 48D3 and 12B2; and with secondary antibodies conjugated with Alexafluor 597 and Alexafluor 488, respectively for G and F proteins. The membrane expression of the proteins were viewed under a confocal microscope. Un-transfected HEK293 cells served as mock control. DAPI was used as the nuclear stain (Magnification, 60X; Scale bar, 20 μm) SEAP activity in the target BHK-21 cell culture supernatants infected at different dilutions (Fig. 3F).
## Detection of uniform incorporation NiV F and G proteins on N-PV(FG)-5F6 pseudovirions by immunostaining with specific antibodies
To identify whether NiV F and NiV G proteins co-incorporated on the N-PV(FG)-5F6 pseudovirions, immunostaining with specific antibodies was carried out on BHK-21 target cells post-infection. Two approaches were tested in the experiments. First, a 30 min infection at 25 °C followed by fixation with acetone-methanol which permitted membrane permeabilization and staining of even pseudovirions that were already internalized. Second, a 30 min infection at 4 °C followed by fixation with 4% paraformaldehyde, which permitted staining of only the cell surface-adsorbed pseudovirions. In acetonemethanol fixed cells, analysis by confocal microscopy at 60X magnification clearly visualized the immunostained viral particles, predominantly in the cytoplasm (Fig. 4A). Superimposition of the green and red fluorescence on them clearly indicated a uniform co-distribution of F and G proteins on the N-PV(FG)-5F6 pseudovirions. This was well apparent in the yellow colour staining in the overlay images of N-PV(FG)-5F6 pseudovirion infected cells (Fig. 4A). However, on the N-PV(FG) pseudovirions, as shown in the figure, the incorporation of the G protein was significantly low compared to the incorporation of the F protein. Also, in the control cells infected with rVSV-G*∆G-SEAP virus, there was no fluorescent signal indicating the specificity of the detection of the F and G proteins by the anti-NiVF and anti-NiV G antibodies (Fig. 4A).
Scatter plot analysis of the F and G fluorescence detection on pseudovirions revealed that the pixel points had a distribution around a central straight line passing through the origin in the case of N-PV(FG)-5F6 infected cells, as against a skewed distribution on N-PV (FG) infected cells. This also indicated that these proteins had a uniform incorporation on N-PV(FG)-5F6 pseudovirions (Fig. 4B&C). Quantification of the mean fluorescence intensity of F & G proteins re-confirmed the equal levels of incorporation of both proteins in N-PV(FG)-5F6 pseudovirions, compared to that in N-PV(FG) (Fig. 4D&E). F and G protein detection levels in N-PV(FG)-5F6 pseudovirions had a Pearson's coefficient of 0.93 compared to 0.83 in N-PV(FG), indicating a significantly uniform incorporation of both the proteins in the former.
## Comparative evaluation of NiV pseudovirion production from F and G plasmids transiently transfected cells and (FG)-5F6 cells
We compared the batch to batch consistency as well as F & G protein incorporation in the pseudovirions produced from HEK293 cells transiently transfected with NiV F & G plasmids as well as from the clone (FG)-5F6 cells. The schematic of the experimental method used is shown in Fig. 5A. Four batches of pseudoviruses, produced under identical conditions, were evaluated.
As shown in Fig. 5B, SEAP evaluation in the culture supernatant from target cell infection indicated that there is a significant variability in the production levels of the NiV pseudovirions among the batches from transiently transfected cells whereas the levels were consistent when stably transfected cells were used. Also, with equal volume of virus used for target cell infection, the amount SEAP activity was almost three-times higher (mean OD 405 of 1.31 vs. mean OD 405 of 3.96) for pseudovirions produced from F&G stably expressing (FG)-5F6 cells, indicating a significantly higher amount of virus production.
(See figure on previous page.) Fig. 3 Production of N-PV(FG)-5F6 pseudovirions from monoclonal HEK293 cells A. Titration of N-PV(FG)-5F6 pseudovirions. A colorimetric spot forming assay was used to stain the thermostable-SEAP producing, pseudovirus-infected cells as described in the methods. Bright purple coloured spots indicate the infected BHK-21 target cells, detected by alkaline phosphatase staining. Representative magnified fields from the ELISPOT reader scanned image of an assay plate infected with serial dilutions of the stock virus are shown. B Quantification of the number of spots forming units (SFU) per well. Spots were counted by automated focus counting using BioSpot 5.0 Professional software in the ELISPOT reader. Data is Mean ± SD of values of two virus dilutions from three independent experiments (N = 6). C Determination of lower limit of virus detection by SEAP assay. Virus inoculums with known amount of viral particles (SFU) were used to infect target BHK-21 cells and SEAP activity was measured for each dilution after 24 h. The values are Mean ± SD of two independent experiments done in triplicates (N = 6). Linear regression analysis was used to compute the R square values. D Comparison of SEAP activity in culture supernatants of target BHK-21 target cells infected with equal volumes of N-PV(FG) and N-PV(FG)-5F6 pseudovirions. The pseudovirion harvests were obtained from polyclonal and monoclonal HEK293 producer cells, respectively. The producer cells were initially infected with identical amounts of rVSV G*(△G-SEAP) for pseudovirion production. Supernatant from un-transfected HEK293 cells, infected with rVSV G*(△G-SEAP) was used for infecting the control cells. The Mean ± SD of duplicate values from three experiments are shown. '*' indicates P < 0.05; '****' indicates P < 0.0001; Unpaired Student t-test was used for statistical analysis for comparison between the two groups E. Kinetics of SEAP production in BHK-21 target cell supernatants post-N-PV(FG)-5F6 infection. Data points indicate the time points 4, 8, 16, 24, 32, 48, 56 and 72 h post-infection. The 24 h time-point fixed for subsequent experiments is indicated by the dotted line. Each value is a Mean ± SD of two independent experiments each done in triplicates. F Validation of the reproducibility of N-PV(FG)-5F6 pseudovirion production under identical conditions. Three batches of pseudovirions harvested from 293(FG)-5F6 monoclonal producer cells were used to infect BHK-21 target cells in various dilutions. SEAP activity in the culture supernatant was measured at 405 nm and plotted. Each value indicates Mean ± SD of triplicate values from three independent experiments (N = 9). Wilcoxon signed-rank test was used for statistical analysis for comparing the three batches of pseudovirions harvested (P < 0.001); and the values among the three batches did not show any statistical significance in any of the dilutions
We also immunostained the F & G proteins incorporated onto pseudovirions after allowing them to adsorb onto BHK21 target cells and imaged them under a confocal microscope. In the target cells fixed with 4% paraformaldehyde, the membrane integrity was well preserved; and upon immunostaining, the pseudovirion particles were clearly visualized as distinct dots on the cell surface. As shown in Fig. 5C, the colocalization of the red and green fluorescent signals yielding the yellow coloured speckles and dots in the overlay image confirmed the uniform incorporation of both the proteins. There was significantly more incorporation of the F& G proteins in the NiV pseudovirions produced from (FG)-5F6 stably transfected cells than those produced after transient transfection. As with the low titre of virus, the number of fluorescent spots indicating the infectious particles were also low in the virus preparation from transiently transfected cells. These experiments clearly confirmed that the use of (FG)-5F6 clone has a distinct advantage on NiV pseudovirion production.
## Pseudovirus neutralization assays using commercial anti-NiV antibodies and convalescent serum from Nipah infection recovered individuals
The usefulness of the N-PV(FG)-5F6 pseudovirions in functional Assays were confirmed by neutralization Assays. Commercial anti-NiV F and G antibodies gave 100% neutralization of the pseudovirions (Fig. 6A). The neutralization titer decreased with decreasing antibody concentration, confirming a dose-dependent neutralizing activity and its specificity. The 17C7 non-neutralizing antibody used as negative control did not neutralize the N-PV(FG)-5F6 pseudovirions, further confirming the specificity of the observations.
The susceptibility of the N-PV(FG)-5F6 pseudovirions to neutralization by human antibodies were evaluated using convalescent serum samples from three Nipah recovered subjects (collected during the Nipah outbreak that occurred in the Kozhikode district of Kerala, India in the year 2023). Serum samples (NiV patient 01, 02, and 03) were diluted to 1:10 followed by 2-fold serial dilutions; and used in the Assays. All the three samples showed neutralizing activity against NiV pseudovirions. While NiV serum 01 and 02 could completely neutralize the pseudoviruses, NiV03 gave more than 80% neutralization at a serum dilution of 1:10 (Fig. 6B). The percentage of neutralization showed a declining trend with increasing serum dilutions. Sera from two normal individuals used as controls in the assay did not neutralize the N-PV(FG)-5F6 pseudovirions (< 20%). From these results, it could be concluded that the established high throughput system is sensitive and specific and could be used for serosurveillance in Nipah endemic regions.
## Discussion
NiV spill over and human disease outbreaks are sporadic and unpredictable. The lack of commercial rapid diagnostic or serological tests prevent early detection and large-scale epidemiological investigations. Like for other BSL-4 pathogens such as Ebola virus (EBOV), surrogate systems, such as virus-like particles and pseudoviruses, have been developed for NiV. However, unlike EBOV and similar viruses that have only one surface glycoprotein (GP) that facilitate infection and to be included in these systems [28], NiV has two surface glycoproteins F and G that are to be incorporated. In all the previous reports on NiV VLP-based or pseudovirus-based systems, the authors attempted transient co-transfections of F and G expressing plasmids with multiple optimization attempts of outer membrane and backbone plasmid concentration ratios to obtain their co-expression. This resulted in the unreliable incorporation of these proteins in the progeny particles, making the approach challenging and less reproducible [17,18,29,30]. In a recent study, a cell line stably expressing the NiV F protein was infected with a recombinant VSV expressing NiV G to produce hightitre pseudovirions [31]. Though the study had confirmed expression of the F & G proteins in the transfected and rVSV infected cell lines, it did not analyse the efficiency of incorporation of both the proteins in the generated pseudovirions.
In the present study, to overcome the above difficulties, we generated producer cells that stably incorporated the F and G proteins after transfection. Initial cloning of the F and G genes was carried out in mammalian expression vectors that have different antibiotic selection (See figure on previous page.) Fig. 4 Evaluation of uniform incorporation of F &G proteins on NiV pseudovirions.A Localization of N-PV(FG)-5F6 pseudovirions in target cells postinfection by immunostaining with anti-NiV F & G antibodies and confocal microscopy. Target BHK-21 cells were infected with 500 SFU pseudovirions for 30 min at 25 °C. Cells were fixed in acetone: methanol (1:1) and stained with primary antibodies 48D3 and 12B2; and secondary antibodies Alexafluor 597 and Alexafluor 488. The pseudovirion infection was confirmed by the colocalization of green and red fluorescence as yellow fluorescence on the surface and within the cells B & C. Scatterplot of green and red pixel intensities corresponding to NiV F and G staining respectively on N-PV(FG) and N-PV(FG)-5F6 pseudovirions. In N-PV(FG) infection, the points of the scatterplot were distributed more towards the green fluorescence probe. On the other hand, in N-PV(FG)-5F6 infection, a proportional co-distribution, where the points of scatterplot clustered around a central straight line, was observed D& E. Quantification of F& G protein co-localization using Pearson's correlation coefficient (PCC). A PCC value of 0.83 in N-PV(FG) showed that the probes did not overlap in a fixed proportion whereas a high PCC value of 0.93 for N-PV(FG)-5F6 probes reflected proportional co-distribution. The values are the Mean ± SD of 14 region-of-interests (ROIs) fields from two independent experiments. **** P < 0.0001; ns-Not significant; Unpaired Student t-test was used for statistical analysis comparing the two groups markers, G418 and hygromycin; and stable polyclonal cells were generated after multiple rounds of selection for two months. As shown in Fig. 2A, immunofluorescence analysis confirmed that the NiV F and G proteins are incorporated on these producer HEK293 cells. Several previous studies have described the formation of large multinucleated syncytial cells in multiple cell lines, including HEK293 cells, when infected with NiV or when the viral glycoproteins F and G are co-expressed [32,33]. In our studies also, during initial transient transfection with plasmids encoding the NiV F and G proteins, many syncytial cells could be observed (Supplementary Fig. 1 A). However, when these cells were selected under antibiotics for stable expression, the cells did not form large syncytium, but remained as clumps of cells with partial fusions (Supplementary Fig. 1B). As observed in previous studies, during viral infections or transient transfections, the cell fusion events happen between a donor cell expressing the viral glycoproteins and the adjacent target cells expressing the receptors for these glycoproteins. This involves cell signaling events, e.g. Rho GTPase pathway, in some viruses such as respiratory syncytial virus (RSV) facilitating the fusion between the cells [34]. However, in the HEK293 cells selected for prolonged stable expression of F and G proteins, all the cells in culture uniformly express both the viral glycoproteins as well as the Ephrin B2/B3 receptor proteins. We presume that this may lead to a two-way cross talk and reciprocal signaling among adjacent cells preventing larger cell fusion events. However, this hypothesis needs further detailed experimental validation, especially in the context that some studies have already shown the dispensability of cytoplasmic regions of Ephrin B2 receptors for facilitating NiV induced cell fusion events [35].
When pseudovirions generated from these cells were used to infect the target BHK-21 cells and the viral particles were visualized by immunostaining, both F and G proteins could be detected. However, when these images were overlayed and analysed, only a poor protein colocalization could be observed (Fig. 2A). This indicated that the N-PV(FG) pseudoviral particles generated from these polyclonal cells were still heterogenous even after prolonged selection for stable expression of both the F and G proteins.
To further improve the pseudovirion generation and make our assay more reproducible, the stable producer cells were subjected to clonal selection by limiting dilution to ensure single cell per well, which were further expanded and analyzed. Analysis of three clones (293FG-5F6, 293FG-7C5, and 293FG-8C7) from the clonal selection demonstrated a significant variation in the rate of expression of both the proteins; and only the clone 293FG-5F6 that showed more than 90% expression of both the proteins were chosen for the establishment of Nipah pseudovirus system. These clonal cells had an almost equal levels of F and G protein membrane incorporation as seen in the immunostaining (Fig. 2D). Pseudovirion particles from these producer cells had a conspicuous co-localization of both the proteins (Fig. 4A&F) and these cells showed consistency in production of pseudovirions with uniform infectivity (Fig. 3F). Co-localization of the fluorescence probes of NiV F and G have been seldom quantified in any of the previous studies, thus making this assay system novel and reliable.
Pseudovirus-based systems are finding increased use in antibody neutralization assays and antiviral screening assays for identifying virus entry inhibitors [36]; and simpler and high throughput-adaptable assays are a prerequisite for their large-scale use. The present study used secreted alkaline phosphatase (SEAP) as a reporter in the NiV pseudovirions generated. Earlier studies have used SEAP-reporter systems for the development of pseudovirus systems for other viruses such as human papillomavirus [37,38]. Except for one report that used SEAP [20], all other previous studies on NiV surrogate systems, including the recent study [31] used either fluorescence or luminescence-based reporter Assays for monitoring infection. SEAP-based Assays are non-destructive and provide about 10-fold higher sensitivity than conventional luciferase-based systems [20,39]. In our Assays, we could see that pseudovirions even As little as 4-8 SFU give a detectable, though small, expression of SEAP activity making the system sufficiently sensitive for neutralization assays with minimum residual undetected (See figure on previous page.) Fig. 5 Comparative evaluation of NiV pseudovirion production from F and G plasmids transiently transfected cells and (FG)-5F6 cellsA. Scheme of the experimental protocol followed. B Batch to batch reproducibility in NiV pseudovirion production from (A) HEK293 cells transiently transfected with NiV F & G plasmids and (B) HEK293 (FG)-5F6 clones stably expressing these proteins. As indicated in the scheme, the target BHK21 cells hexaplicate wells of a 96-well plate were infected with equal volume of the supernatants containing the NiV pseudovirions from these producer cells. The SEAP activity was measured in the BHK21 cell culture supernatants after 24 h as described in the protocols; and mean±SD was plotted. The statistical analysis was carried out with paired t-test. * P < 0.05; ** P < 0.0, *** P < 0.001, **** P < 0.0001; ns-Not significant. C Immunostaining of F and G proteins of NiV pseudovirions on the surface of target BHK-21 cells. Cells were infected for 30 min at 4 °C with 2.5 × 10 5 SFU pseudovirions and were fixed for immunofluorescence with 4% paraformaldehyde, without membrane permeabilization. The viral particles adsorbed on the surface were immunostained with primary antibodies (48D3 and 12B2) followed by secondary antibodies conjugated with Alexafluor 597 and Alexafluor 488, respectively. The G and F proteins on the pseudovirions attached to the target cell membrane were visualised at 60X magnification in a confocal microscope. The yellow speckles represent colocalized green and red fluorescence of the NiV F and G proteins on the surface on the infected target cell membrane virus (Fig. 3C). It also offers high throughput adaptability As well As easy monitoring of the pseudovirion infection using conventional ELISA readers providing easier implementation in large-scale surveillance programs. We found that the SEAP activity becomes detectable by 18 h post-infection and reaches a peak by 24 h post-infection in NiV pseudovirion infected target cells (Fig. 3E), facilitating an early readout in biological assays using this system. As surrogate systems may not necessarily mimic the structural organization and functional properties of the surface glycoproteins of natural viruses in all aspects, it is imperative to test the functionality of the pseudovirus particles generated. Neutralization by specific antibodies is a key parameter that indicates their functionality and is a useful property that helps to assess virus-specific immunity in a population. The use of pseudoviruses for evaluating antibody-mediated neutralization has been described in previous studies [40]; and also, earlier reports have indicated a good correlation of the results from these assays with conventional PRNT assays using natural, infectious viruses [23]. However, the use of too excess pseudoviruses can give false negative results in neutralization Assays and there is a need to use optimal virus concentration to obtain reliable and reproducible results. We found that a dilution of pseudovirion stock that gives a titer of approx. 120 SFU gave a dynamic working range OD 405 of 1.00-0.1 in our PVNT assays starting with virus controls to negative/background controls. Accordingly, this dilution of the stock N-PV(FG)-5F6 pseudovirions was used in all the neutralization assays.
Antibodies generated in a mouse system (commercial anti-NiVF and G mAbs) as well as human convalescent serum were used to evaluate the pseudovirion functionality in neutralization assays. Both antibodies showed a dose-dependent neutralization of the NiV pseudovirions while the negative control antibodies did not neutralize the virus as expected (Fig. 6A &B). The Assay detect the neutralizing antibodies in 1:10 diluted serum samples indicating its comparable performance with conventional virus neutralization Assays. NiV pseudovirions were designed using the sequences of the Bangladesh strain of virus that caused the Nipah outbreak in Kerala in 2018 [9]; and recent studies have pointed out that the 2023 Nipah outbreak, from where the serum samples were collected, was also caused by the same strain of NiV. There were differences in the level of neutralization of the three patient samples used pointing out that the assay detects quantitative differences in the specific antibody titers in each of these patients. Further validation of the PVNT assay using serum samples against other Nipah virus variants need to be done to evaluate the broader utility of the VSV-based NiV pseudovirions developed in this study.
## Conclusions
The present study developed a robust and reproducible system for generating VSV-based NiV pseudovirions. Compared to the transient transfection-based systems, the 293(FG)-5F6 cell line stably expressing NiV F & G proteins gave consistent and enhanced production of NiV pseudovirions (Fig. 5). These pseudovirions uniformly incorporated F and G proteins and can be effectively used in virus neutralization assays. One limitation of the current study is that it has used only a smaller number of clinical samples. We had only limited sample access as Nipah infections are not frequent in the state and all recent outbreaks are effectively contained. Further validation of the assay in a larger set of serum samples will make this rapid and high throughput-adaptable assay for cost-effective use in large scale epidemiological screening in simple laboratory settings. The assay will be particularly useful in serosurveillance studies as a generic assay when species-specific reagents are not available. This assay will also help in therapeutic monoclonal antibody and vaccine response evaluation as well as in drug discovery studies against Nipah. carried out overall supervision, arranged funding and edited and finalised the manuscript.
## Funding
The study received funding to ES from the Government of Kerala through intramural funding of IAV Flagship program as well as from the Department of Biotechnology (DBT), Government of India through the DBT-SAHAJ program (BT/INF/22/SP53419/2024). Funders had no role in the conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript.
## References
1. Epstein, Field, Luby et al. (2006) "Nipah virus: impact, origins, and causes of emergence" *Curr Infect Dis Rep*
2. Wit, Munster (2016) "Animal models of disease shed light on Nipah virus pathogenesis and transmission" *J Pathol*
3. Walpita, Barr, Sherman et al. (2011) "Vaccine potential of Nipah virus-like particles" *PLoS ONE*
4. Chua (2003) "Nipah virus outbreak in Malaysia" *J Clin Virol*
5. Paton, Leo, Zaki et al. (1999) "Outbreak of Nipah-virus infection among abattoir workers in Singapore" *Lancet*
6. Chan, Rollin, Ksiazek et al. (2002) "A survey of Nipah virus infection among various risk groups in Singapore" *Epidemiol Infect*
7. Lo, Luis, Hummel et al. (2008) "Characterization of Nipah virus from outbreaks in bangladesh" *Emerg Infect Dis*
8. Arunkumar, Chandni, Mourya et al. (2018) "Group NIPaHS : Outbreak investigation of Nipah virus disease in Kerala, India" *J Infect Dis*
9. Yadav, Shete, Kumar et al. (2018) "Nipah virus sequences from humans and bats during Nipah outbreak" *Emerg Infect Dis*
10. Who (2023) "Disease Outbreak News; Nipah Virus Infection in India"
11. Anish, Aravind, Radhakrishnan et al. (2024) "Pandemic potential of the Nipah virus and public health strategies adopted during outbreaks: lessons from Kerala"
12. Vasudevan, Subash, Mehta et al. (2024) "Global and regional mortality statistics of Nipah virus from 1994 to 2023: a comprehensive systematic review and meta-analysis"
13. Bae, Kim, Moon et al. (2019) "Construction of the safe neutralizing assay system using pseudotyped Nipah virus and G protein-specific monoclonal antibody" *Biochem Biophys Res Commun*
14. Sun, Jia, Liang et al. (2018) "Phylogeography, transmission, and viral proteins of Nipah virus" *Virol Sin*
15. Jardetzky, Lamb (2014) "Activation of paramyxovirus membrane fusion and virus entry" *Curr Opin Virol*
16. Bossart, Fusco (2013) "Broder CC. Paramyxovirus entry" *Adv Exp Med Biol*
17. Witting, Vallanda, Gamble (2013) "Characterization of a 3rd generation lentiviral vector pseudotyped with Nipah virus envelope proteins for endothelial cell transduction" *Gene Ther*
18. Luo, Wang, Huang et al. (2023) "Establishment of a neutralization assay for Nipah virus using a high-titer pseudovirus system" *Biotechnol Lett*
19. Whitt (2010) "Generation of VSV pseudotypes using recombinant ∆G-VSV for studies on virus entry, identification of entry inhibitors, and immune responses to vaccines" *J Virol Methods*
20. Kaku, Noguchi, Marsh et al. (2012) "Second generation of pseudotype-based serum neutralization assay for Nipah virus antibodies: sensitive and high-throughput analysis utilizing secreted alkaline phosphatase" *J Virol Methods*
21. Logan, Monagle, Drew et al. (2016) "Efficient generation of vesicular stomatitis virus (VSV)-pseudotypes bearing morbilliviral glycoproteins and their use in quantifying virus neutralising antibodies" *Vaccine*
22. Capcha, Lambert, Dykxhoorn et al. (2021) "Generation of SARS-CoV-2 spike pseudotyped virus for viral entry and neutralization assays: a 1-week protocol" *Front Cardiovasc Med*
23. Tamin, Harcourt, Lo et al. (2009) "Development of a neutralization assay for Nipah virus using pseudotype particles" *J Virol Methods*
24. Amaya, Yin, Yan et al. (2023) "A recombinant chimeric cedar virus-based surrogate neutralization assay platform for pathogenic henipaviruses" *Viruses*
25. Nie, Liu, Wang et al. (2019) "Nipah pseudovirus system enables evaluation of vaccines in vitro and in vivo using non-BSL-4 facilities" *Emerg Microbes Infect*
26. Lupitha, Varghese, Das et al. (2025) "Assessing anti-rabies vaccine response in humans: a rapid and high-throughput adaptable, pseudovirus-based neutralization assay as an alternative to rapid fluorescent focus Inhibition test (RFFIT)" *PLoS Negl Trop Dis*
27. As, Sahay, Radhakrishnan et al. (2024) "Clinicoepidemiological presentations and management of Nipah virus infection during the outbreak in Kozhikode district, Kerala state, India 2023" *J Med Virol*
28. Steeds, Hall, Slack et al. (1038) "Pseudotyping of VSV with Ebola virus glycoprotein is superior to HIV-1 for the assessment of neutralising antibodies. Sci Rep"
30. Pramila, Barr, Sherman et al. (2011) "Vaccine potential of Nipah Virus-Like particles" *PLoS ONE*
31. Rajan, Naira, Pillai et al. (2024) "Highly sensitive and quantitative HiBiT-tagged Nipah virus-like particles: A platform for rapid antibody neutralization studies" *Heliyon*
32. Jain, Lo, Kainulainen et al. (2023) "Development of a neutralization assay using a vesicular stomatitis virus expressing Nipah virus glycoprotein and a fluorescent protein" *Virology*
33. Negrete, Levroney, Aguilar et al. (2005) "EphrinB2 is the entry receptor for Nipah virus, an emergent deadly paramyxovirus" *Nature*
34. Gamble, Yeo, Butler et al. (2021) "Drivers and distribution of Henipavirus-Induced syncytia: what do we know??" *Viruses*
35. Gower, Pastey, Peeples et al. (2005) "RhoA signaling is required for respiratory syncytial virus-induced syncytium formation and filamentous virion morphology" *J Virol*
36. Thiel, Diederich, Erba et al. (2008) "Ephrin-B2 expression critically influences Nipah virus infection independent of its cytoplasmic tail" *Virol J*
37. Trischitta, Tamburello, Venuti et al. (2024) "Pseudovirus-based systems for screening natural antiviral agents: a comprehensive review" *Int J Mol Sci*
38. Sehr, Rubio, Seitz et al. (2013) "Highthroughput pseudovirion-based neutralization assay for analysis of natural and vaccine-induced antibodies against human papillomaviruses" *PLoS ONE*
39. Buck, Pastrana, Lowy et al. (2005) "Generation of HPV pseudovirions using transfection and their use in neutralization assays" *Methods Mol Med*
40. Yang, Sinai, Kitts (1997) "Quantification of gene expression with a secreted alkaline phosphatase reporter system" *Biotechniques*
41. Cai, Kalkeri, Zhu et al. (2024) "A Pseudovirus-Based neutralization assay for SARS-CoV-2 variants: A rapid, Cost-Effective, BSL-2-Based High-Throughput assay useful for vaccine immunogenicity evaluation" *Microorganisms* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12705370&blobtype=pdf | # EDITED AND REVIEWED BY
Matteo Becatti, Valentyn Oksenych, Oleksandr Kamyshnyi, Rostyslav Bilyy
Editorial on the Research Topic Emerging trends in cancer research: diagnostic and therapeutic breakthroughs s The Research Topic "Emerging Trends in Cancer Research: Diagnostic and Therapeutic Breakthroughs" synthesizes recent advances in oncology, highlighting strategies for early detection, improved prognostic tools, analysis of the complex tumor microenvironment (TME), and development of precision therapies. Despite extensive efforts, cancer remains a major global health challenge, encompassing diverse pathologies with different cellular origins. Hepatocellular carcinoma (HCC), for example, ranks as the third leading cause of cancer-related mortality worldwide, causing over 800,000 deaths annually (Alemayehu et al. and Sung et al., 2021), while cutaneous melanoma (CM) is a highly aggressive skin malignancy, with incidence rates rising dramatically in recent decades (Dudin et al. and Siegel et al., 2022). Addressing these persistent challenges requires breakthroughs in molecular science and integrated clinical strategies. The papers compiled here reflect this imperative, emphasizing minimally invasive molecular diagnostics, immunotherapy targeting the tumor microenvironment (Wan et al., 2025), and advanced genomic approaches to guide personalized treatment. Collectively, they aim to overcome key limitations in late-stage diagnosis and therapy resistance, advancing more effective and precise interventions for cancer management (Alemayehu et al.).
Significant efforts are focused on improving diagnostic and prognostic precision through molecular profiling and liquid biopsy technologies. The value of circulating biomarkers is well illustrated, exemplified by a systematic review and metaanalysis demonstrating that circulating microRNAs hold promise as non-invasive diagnostic biomarkers for hepatocellular carcinoma (Alemayehu et al.). For colorectal cancer (CRC), where early-stage symptoms are often lacking (Bray et al., 2024), novel markers like cystatin S (CST4) are emerging. CST4 demonstrates stability post-chemotherapy and significantly enhances diagnostic sensitivity for malignant 10.3389/fmolb.2025.1750771 colorectal lesions when combined with conventional tumor markers (CEA, CA125, CA724), representing a 28.4% increase in sensitivity over CST4 alone (Han et al.).
The increasing need for detailed genetic information drives the application of advanced sequencing methods, including Next-Generation Sequencing and Third-Generation Sequencing (TGS), to unravel circulating tumor DNA (ctDNA) mutations in liquid biopsies (Da Silva et al.). The requirement for molecular precision is underscored by cases illustrating diagnostic pitfalls, where malignancies like pulmonary Ewing sarcoma (ES) can be misdiagnosed as more common entities such as Small Cell Lung Cancer (SCLC) based solely on histopathology, necessitating early molecular testing for accurate diagnosis and management selection (Waary et al.). Moreover, prognostic models are being refined using routinely available markers, such as assessing uric acid (UR) and the Neutrophil/Lymphocyte Ratio (NLR) in CRC. Elevated UR and NLR levels were found to be independent risk factors for bone metastasis, reflecting the role of systemic inflammation in disease progression (Chen et al.).
Omelianenko et al. investigated the tumor immune microenvironment (TIME) in thyroid adenoma (TA) and carcinoma (TC) to explore its potential diagnostic value in cytopathology. In a pilot study of 72 cases (36 TA, 36 TC) with histological confirmation and preoperative Bethesda III-V cytology, the authors quantified CD8 + , CD68 + , and CD163 + immune cells and assessed STAT6 and SMAD4 expression. TC exhibited a highly immunogenic profile with abundant CD8 + lymphocytes and macrophages, whereas TA showed low immune infiltration. Immune cell counts in cytology specimens correlated strongly with histological findings. These results suggest that immune cell density in thyroid cytology may serve as an additional criterion for differentiating benign and malignant lesions.
Case report of Yang et al. focuses on a metastatic squamous cell carcinoma of unknown primary (SCCUP) in a 70-year-old female presenting with elevated CA 19-9 and a diaphragmatic mass. Despite extensive evaluation, including PET-CT and a 90-gene expression assay, the primary tumor remained unidentified. The patient underwent surgical resection followed by systemic therapy, achieving 14 months of disease-free survival. This case highlights diagnostic limitations, the potential of multimodal therapy, molecular-clinical discordance, and the need for international collaboration and comprehensive genomic profiling in CUP management.
In therapeutic breakthroughs, a key focus is the multifaceted interplay within the tumor microenvironment (TME) (Anderson and Simon, 2020;Ragunathan et al., 2020). A prominent area involves targeting Tumor-Associated Macrophages (TAMs), which exhibit plasticity, polarizing into M1 (anti-tumor) or M2 (protumor/metastasis promoting) phenotypes (Bai et al.). Therapeutic strategies aim to reprogram M2-TAMs toward the anti-tumor M1 phenotype, utilizing agents like CSF1R inhibitors or blockers targeting the CD47/SIRPα axis, or through common drugs like Metformin, which disrupts M2 polarization by activating the AMPactivated protein kinase (AMPK) pathway (Bai et al.). This emphasis on immunomodulation was heavily featured at the 8th Cancer Immunotherapy and Immunomonitoring (CITIM) conference, which highlighted the critical roles of chronic inflammation, Myeloid-Derived Suppressor Cells (MDSCs), and the emerging understanding of the neuro-metabolic-immune regulation of cancer (Rostyslav Bilyy).
Cutting-edge strategies include novel chemical agents designed to induce immunogenic cell death and prolonged immune stimulation (Arkhypov et al., 2025) and delivery systems critical for overcoming treatment obstacles, particularly in challenging diseases like Triple-Negative Breast Cancer (TNBC) (Tiwari et al.). Extracellular vesicles (EVs) are identified both as essential players in promoting TNBC progression and drug resistance (e.g., through carrying EGFR or lncRNA XIST) and as promising drug carriers for targeted therapies due to their low toxicity and ability to traverse biological barriers, Tiwari et al.
Finally, innovation in clinical safety includes the development of a novel sealant combining an absorbable gelatin sponge (mechanical occlusion) with Agkistrodon acutus-derived hemocoagulase (local coagulation, Chen et al.). This dual mechanical and pharmacological barrier significantly reduced the rate of intervention-requiring pneumothorax in pulmonary biopsies (from a literature rate of 22.1% to 2.38%) and introduced a precision-stratified safety protocol based on D-dimer levels to manage hemorrhage risk, Chen et al.
Collectively, these findings underscore a concerted movement toward integrative cancer management, where precise molecular information and sophisticated TME modulation techniques are combined to deliver more effective, personalized, and safer patient care, Alemayehu et al.
## References
1. Anderson, Simon (2020) "The tumor microenvironment" *Curr. Biol*
2. Arkhypov, Klemt, Bila et al. (2025) "Targeting Lysosomal thiols for immunogenic cancer cell death" *Angew. Chem. Int. Ed. Engl*
3. Bray, Laversanne, Sung et al. (2024) "Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries" *CA Cancer J. Clin*
4. Ragunathan, Upfold, Oksenych (2020) "Interaction between fibroblasts and immune cells following DNA damage induced by ionizing radiation" *Int. J. Mol. Sci*
5. Siegel, Miller, Fuchs et al. (2022) "Cancer statistics, 2022" *CA Cancer J. Clin*
6. Sung, Ferlay, Siegel et al. (2021) "Global cancer statistics 2020: GLOBOCAN estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries" *CA Cancer J. Clin*
7. Wan, Ren, Yang et al. (2025) "Tumor evolution and immune microenvironment dynamics in primary and relapsed mantle cell lymphoma" *Cell Rep. Med* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645919&blobtype=pdf | # Biomimetic nanovaccines with self-adjuvant effects induced broad-spectrum neutralizing antibodies against SARS-CoV-2 infection in rodents
Weiqi Wang, Pengye Du, Yongkun Zhao, Yuan Liang, Cheng Zhang, Hongjie Zhang, Xianzhu Xia, Bo Liu, Pengpeng Lei, Feihu Yan
## Abstract
The emergence of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) variants seriously threatens the efficacy of current coronavirus disease 2019 (COVID-19) vaccines. Therefore, there is an urgent need to develop next-genera tion vaccine platforms capable of counteracting current and prospective viral variants. In this study, we employed synthetic biology technology to develop self-adjuvanteffect biomimetic nanovaccines with broad-spectrum capabilities for the prevention of SARS-CoV-2 infection. The biomimetic nanovaccines prepared by loading the SARS-CoV-2 RBD protein into dendritic mesoporous organosilicon nanoparticles (DMOSN) significantly promote the recruitment of dendritic cells (DCs) to secondary lymphoid organs, thereby initiating antibody-dependent humoral immune responses mediated by follicle helper T (Tfh) cells, germinal center (GC) B cells, and plasma cells. Moreover, DMOSN@RBD induced not only potent humoral immunity but also Th2-biased cellular immunity along with robust Th1-type cellular immune responses, which are pivotal in restricting SARS-CoV-2 infection. Our work provides a simple and environmentally friendly strategy for synthesizing nanovaccines with significant immunostimulatory potential, offering novel insights for the future development of durable and effective antiviral broad-spectrum nanovaccines.
IMPORTANCEThe persistent evolution of severe acute respiratory syndrome coronavi rus-2 (SARS-CoV-2) underscores the critical need to continually assess vaccine immu nogenicity and protective efficacy against emerging variants in preclinical animal models. Our study demonstrates that biomimetic nanoparticle vaccines elicit more durable antibody responses and enhanced T cell responses compared to conven tional aluminum hydroxide-adjuvanted formulations. Notably, RBD antigen-decorated dendritic mesoporous organosilica nanoparticles (DMOSN@RBD) exhibit broad-spectrum neutralization potential against multiple SARS-CoV-2 variants of concern (VOCs). These findings establish engineered mesoporous silica nanoparticles as a potent immunos timulatory platform capable of simultaneously enhancing both humoral and cellular immunity in subunit vaccine design, particularly through the induction of robust T cell responses typically challenging to achieve with protein-based vaccines.
dendritic mesoporous organosilicon nanoparticles (DMOSN@RBD). The breakthrough infections are primarily driven by the emergence of SARS-CoV-2 variants of concern (VOCs), which have the ability to evade the immune response and compromise vaccine efficacy. Of particular concern is the Omicron variant, which exhibits increased trans missibility and significant evasion of neutralizing antibodies, substantially complicat ing pandemic containment strategies (2,3). The emergence of these variants has underscored the importance of reassessing vaccine development strategies to address declining antibody levels and the impact of viral mutations.
Rational adjuvant design integrating antigen presentation with immunomodulatory components represents a critical pathway for next-generation vaccine development (4). Nanoparticle-vaccine platforms, such as lumazine synthase, ferritin, or I53-50 protein assemblies to display SARS-CoV-2 spike or RBD antigens, have shown promise in eliciting potent neutralizing antibody responses (5)(6)(7)(8). However, many existing nanovaccines still require co-administration with potent adjuvants to achieve optimal immunogenicity (9)(10)(11)(12). Synthetic nanoparticles offer unique advantages through antigen encapsulation or adsorption that enhance antigen stability, cellular internalization, and immunos timulatory potential, positioning them as promising nanoadjuvant candidates (13). DMOSN has garnered significant attention due to its ultrahigh specific surface area, easy surface modification, tunable particle dimensions, excellent biocompatibility, and intrinsic adjuvant properties (14). The strategic incorporation of organic functional groups into inorganic silica (-Si-O-Si-) frameworks enables precise engineering of organosilica nanoparticles for enhanced antigen-presenting cell (APC) activation and effectively improved immune responses (15,16). However, the application of DMOSNbased systems as nanoadjuvants for SARS-CoV-2 vaccines remains largely unexplored.
In this study, we loaded the SARS-CoV-2 RBD protein into DMOSN to generate a novel biomimetic nanovaccine with self-adjuvant effects. DMOSN@RBD demonstra ted the ability to elicit robust cross-neutralizing antibodies in mice and golden hamsters, conferring effective protection against SARS-CoV-2-induced clinical manifes tations. Furthermore, the DMOSN-delivered RBD vaccine promoted early dendritic cell (DC) recruitment or activation and induced robust germinal center responses. Additionally, DMOSN@RBD stimulated T-cell-mediated cellular immunity characterized by enhanced cytokine production, while simultaneously maintaining potent humoral immune responses. Overall, this study presents a promising antigen delivery platform for the prevention of epidemics and promotes the application of synthetic biology technology in vaccine production.
## RESULTS
## Construction and characterization of DMOSN@RBD
The structure of the SARS-CoV-2 RBD protein is shown in Fig. 1A. Subsequently, we loaded RBD proteins into DMOSN pre-prepared by a facile anion-assisted method. Transmission electron microscopy (TEM) characterization revealed monodisperse DMOSN particles with a uniform hydrodynamic diameter of 239.5 nm (Fig. 1B; Fig. S1A andB). Energy-dispersive X-ray spectroscopy suggested that the elements Si (1.75 eV) and O (0.52 eV) were present in the prepared carriers (Fig. S1C). In addition, the X-ray photoelectron spectra (XPS) and high-resolution XPS of the corresponding elements are shown in Fig. S1D, with the peaks attributed to Si 1 s (103.4 eV) and O 1 s (533.0 eV). Furthermore, two nanoparticles were randomly selected for elemental mapping imaging analysis, as displayed in Fig. S1E. Si and O are evenly distributed throughout the particles. Figure S1F showed the nitrogen adsorption-desorption isotherms of DMOSN. The Brunauer-Emmett-Teller surface area was measured to be as high as ~422 m 2 /g. The pore size is estimated to be ~15.3 nm, confirming the mesoporous architecture suitable for vaccine delivery applications.
Therefore, DMOSN and RBD were combined to construct a silicon adjuvant-assisted vaccine carrier composite, DMOSN@RBD. The dynamic light scattering (DLS) analysis of DMOSN and DMOSN@RBD revealed that the hydrated particle sizes were 239.9 and 241.9 nm, respectively, indicating that the RBD did not obviously affect the particle size of DMOSN (Fig. 1C). Western blot analysis showed that the size of RBD monomer was consistent with the size of RBD in the nanoparticles (Fig. 1D). The ultraviolet-visible (UVvis) absorption spectra confirmed the successful loading of the RBD (Fig. S1G). Fourier transform infrared spectroscopy revealed the characteristic stretching vibrational absorption of "C=O, " which is an amide bond in proteins (Fig. 1E). The change in zeta potential further illustrated the successful construction of the composite structure, DMOSN@RBD (Fig. 1F). In addition, the good sustained-release ability of the nanovac cines for the RBD was also verified (Fig. S1H).
## Biosafety assessment
Cytotoxicity tests and hemolysis assays were performed to assess the biocompatibility of DMOSN@RBD. Mouse fibroblast L929 cells were chosen as the model of normal cells to verify the biosafety of DMOSN@RBD. The cell viability was greater than 95%, even at every concentration (Fig. 1G). The relative intensity of the 576 nm peak is considered the quantitative standard for the percent hemolysis, which is less than 1% at every concentration (Fig. 1H). To verify the long-term toxicity of DMOSN@RBD, one group of healthy mice was injected with DMOSN@RBD, and another group was injected with an equal amount of PBS as a control group. The major organs of the control and injected administration groups were collected after 28 days. As shown in Fig. 1I, there was no significant difference between the control group and the injection group in any organ (heart, liver, spleen, lung, or kidney), indicating the excellent in vivo biosafety. The above results demonstrate the low toxicity of DMOSN@RBD.
## DMOSN@RBD vaccination protects mice against lethal challenge by the SARS-CoV-2 mouse-adapted strain
Aging is a high-risk factor for severe COVID-19 (17). The protective effects of DMOSN@RBD vaccination were assessed in 9-month-old female BALB/c mice with a lethal mouse-adapted SARS-CoV-2 strain C57MA14 (18). Nine-month-old female BALB/c mice were divided into five groups receiving intramuscular administration of DMOSN@RBD, aluminum(III) hydroxide (AlumOH)+RBD, the RBD monomer, DMOSN, or PBS, following a prime-boost regimen on days 0, 7, and 21 (Fig. 2A). Neutral izing antibody assay results showed that complete neutralizing antibody titers elici ted by DMOSN@RBD were significantly higher than those in the AlumOH+RBD or monomeric RBD immunization groups (Fig. 2B). Measurement of IgG titers and IgG subtypes against the SARS-CoV-2 S protein in serum samples using enzyme-linked immunosorbent assay (ELISA) revealed significant increases following DMOSN@RBD and AlumOH+RBD immunization (Fig. 2C through E). Notably, while antibody subtypes elicited by DMOSN@RBD were dominated by IgG1, significantly higher IgG2a levels were observed compared with the AlumOH+RBD or monomeric RBD immunization groups, indicating a stronger Th1 and Th2 response induced by DMOSN@RBD (Fig. 2D andE). In contrast, immunization with the DMOSN or PBS showed negligible immunogenicity across all parameters (Fig. 2B through E). At 42 days post-vaccination (dpv), the SARS-CoV-2 mouse-adaptation strain C57MA14 was intranasally inoculated with 10 5 TCID 50 , body weight changes and survival rates were monitored for the subsequent 13 days, and blood biochemical analyses, as well as viral load determinations in the lungs and nasal turbinates, were performed at 3 days post-infection (dpi). The results demonstrated that mice inoculated with DMOSN@RBD exhibited a 100% survival rate and maintained stable body weight throughout the observation period (Fig. 2F andG). In contrast, all animals in the other immunization groups experienced rapid weight loss and succumbed to infection within 8 dpi (Fig. 2F andG). Next, we examined changes in blood cell counts, growth kinetics, and viral loads in the five groups of mice at 3 dpi. DMOSN, PBS, or monomer RBD-immunized mice exhibited increased white blood cells (WBCs), decreased platelets (PLTs), a significant decrease in the percentage of lymphocytes (LYM%), a significant increase in the percentage of neutrophils (Neu%) and monocytes (Mon%) DMOSN@RBD-immunized group were significantly superior to those in the Alu mOH+RBD group (Fig. 2H through L). At 3 dpi, three mice from each group were euthanized, and lung and nasal turbinate tissue samples were collected for viral copy analysis and virus titer determination. Animals vaccinated with three doses of DMOSN@RBD demonstrated significant inhibition of SARS-CoV-2 replication compared to the high-level viral load observed in the RBD monomer, DMOSN, or PBS immuni zation group (Fig. 2M through P). Viral loads in tissues of DMOSN@RBD-immunized mice were slightly lower than those in the AlumOH+RBD group, with notably reduced titers observed in lung and nasal turbinate samples. Specifically, two animals in the DMOSN@RBD group showed viral titers below the detection limit (Fig. 2O andP). These findings demonstrate the high efficacy of DMOSN@RBD in controlling SARS-CoV-2 morbidity, lethality, and virus replication in a mouse model of infection.
## DMOSN@RBD induced cross-neutralizing antibodies and protected against challenges in golden hamsters
Gold hamsters are susceptible to SARS-CoV-2 infection and develop the disease (20). To evaluate the broad protective efficacy of DMOSN@RBD, we utilized this animal model. Following an immune strategy similar to that used for BALB/c mice, we meas ured neutralizing antibodies and spike protein-binding antibodies at 35 dpv (Fig. 3A). Consistent with the results in mice, compared to the AlumOH+RBD immunization group, DMOSN@RBD elicited higher levels of cross-neutralizing antibodies and S protein-bind ing antibodies in golden hamsters (Fig. 3B andC). At 42 dpv, golden hamsters were infected with SARS-CoV-2 Wuhan-Hu-1 (wild type, WT) and its variants (Beta, Delta, BA.1, and BA.2) (10 7 TCID 50 ) to evaluate the broad protective effect of DMOSN@RBD. Following the challenge, AlumOh+RBD-, RBD monomer-, DMOSN-, or PBS-immunized animals developed clinical disease with progressive body weight loss. In contrast, DMOSN@RBDimmunized animals were shielded from clinical disease and body weight loss (Fig. 3D through H).
Furthermore, DMOSN@RBD immunization resulted in a significant reduction in the viral load in both the upper and lower respiratory tract (Fig. 3I through R; Fig. S2). A comparison of the titers of infectious viruses in the nasal turbinates and lungs revealed that immunization of golden hamsters with DMOSN@RBD had remarkable protective effects (Fig. 3I through R). Compared with golden hamsters vaccinated with Alu mOH+RBD or the RBD monomer, those vaccinated with DMOSN@RBD demonstrated more than 10-fold reduction in viral load in their nasal turbinates and lungs (Fig. 3I through R). Notably, the DMOSN@RBD immune group showed no detection of live virus in the nasal turbinates after challenge with Omicron BA.1 or BA.2 (Fig. 3O, P andR). These findings are consistent with the broad neutralization data, demonstrating that immuno gens based on the SARS-CoV-2 WT strain have similarly strong neutralizing effects against the beta, delta, and omicron BA.1 and BA.2 variants. This indicates that DMOSN, in conjunction with an adjuvant for RBD proteins, can induce a strong, broad immune response against SARS-CoV-2 WT and its variants.
## DMOSN@RBD enhances lymph node targeting and elicits potent immune activation
Efficient and rapid antigen uptake and activation of antigen-presenting cells are critical for vaccine-induced adaptive immune responses. To clarify whether DMOSN accelerates antigen uptake by APCs, we labeled the RBD protein with fluorescein isothiocyanate (FITC) and assessed the internalization efficiency of DMOSN@RBD in DC2.4 and RAW264.7 cells. The results showed that encapsulating the RBD protein within DMOSN significantly enhanced the internalization of RBD protein in both DC2.4 and RAW264.7 cells (Fig. S3A andB). Immunofluorescence staining also showed that RBD proteins were internalized by macrophages more rapidly than the RBD monomers (Fig. 4A).
Next, we examined the activation of DMOSN@RBD on DCs in lymph nodes in vivo. As a benchmark, we vaccinated animals with RBD protein emulsified in aluminum(III) hydroxide (alumOH), which is the most widely used adjuvant (21). DC activation in dLNs was detected at 24 h post-injection, and flow cytometry results revealed that DMOSN@RBD significantly increased the expression of the costimulatory molecules CD86 and CD40 (Fig. S3C andD), as well as that of MHC II, which is essential for T-cell recognition (Fig. S3E), on the surface of DCs. RBD+alumOH only induced elevated MHC II expression, with no significant changes in CD86 and CD40, and its activation efficiency was considerably lower than that in the DMOSN@RBD-immunized group (Fig. S3C through E). These findings clearly demonstrate that utilizing the DMOSN platform for antigen packaging successfully improves antigen presentation efficiency in mice. Considering the immunostimulatory effects observed following DMOSN@RBD injection, we conducted a multicytokine analysis. The results revealed an enrichment of IL-12p70, MCP1, and IL-1β in DMOSN@RBD-immunized groups (Fig. 4B through I; Fig. S3F andG). These results emphasize that DMOSN@RBD activates the immune response at an early stage.
## DMOSN@RBD elicits potent humoral immune responses
To evaluate the role of DMOSN in activating immune responses against SARS-CoV-2, we investigated the humoral and cellular immune responses induced by DMOSN@RBD. Six-to eight-weeks-old BALB/c mice were immunized with an equal amount of RBD protein in the presence of DMOSN or alumOH as adjuvants, while control animals received PBS immunization (Fig. 5A). The DMOSN@RBD group exhibited significantly higher IgG titers than the alumOH adjuvant group (Fig. 5B). Moreover, consistent with the trend observed in anti-S specific IgG detection, the DMOSN@RBD group consistently exhibited higher levels of neutralizing antibodies than the alumOH adjuvant group from 14 to 120 dpv (Fig. 5C). In terms of antibody duration, at 120 dpv, the mice of the DMOSN@RBD group maintained neutralizing antibodies at 2.1 Log, whereas the alumOH adjuvant group exhibited a decrease to less than 20 (Fig. 5C). Furthermore, we evaluated serum neutralizing antibody activity against several pseudotyped viruses from VOCs and a series of Omicron mutants. Our results demonstrated that DMOSN@RBD induced neutralizing antibodies not only against the WT pseudotyped viruses but also against the other five primary VOCs (Fig. 5D). All groups showed a slight reduction in the levels of neutralizing antibodies against B.1.1.7 (Alpha), B.1.351 (Beta), and P.1 (Gamma), whereas the reduction was more significant against B.617.2 (Delta) and BA.4 (Omicron) (Fig. 5D). Notably, the decrease in neutralizing antibodies induced by DMOSN@RBD against each mutant was significantly lower than that observed in the alumOH adjuvant group, especially against Delta and BA.4 (Fig. 5D). Since germinal centers (GCs) in lymph nodes are central to the humoral immune response (22), we further investigated whether DMOSN@RBD could elicit robust GC responses. As expected, at 35 dpv, DMOSN@RBD induced the greatest populations of follicle helper T (Tfh) cells, GCB cells, and plasma cells (Fig. 5E through G; Fig. S4), which was consistent with the detection of high-titer neutralizing antibodies. Given the rapid mutation of SARS-CoV-2 and the problem of rapid decline in neutralizing antibodies against mutant strains with existing vaccines, the significance of the limited decrease in neutralizing antibodies induced by DMOSN@RBD is clearly evident.
## DMOSN as an adjuvant induces strong cellular immune responses
Next, we evaluated systemic T-cell activation induced by the vaccines. Splenocytes were stimulated with SARS-CoV-2 S1 and S2 peptides, and flow cytometry analysis was performed after 48 h. The results showed that both the DMOSN@RBD-and alu mOH+RBD-immunized groups elicited CD4 + IL4 + TH2-biased immune responses (Fig. S5A andB), which were consistent with previous studies on subunit vaccines. Interestingly, after the third vaccination with DMOSN@RBD, there was a significant increase in the frequency of antigen-specific CD4 + TH1 cells, whereas no increase in IFN-γ + CD4 + T cells was detected in the alumOH+RBD-immunized group (Fig. 6A). Moreover, DMOSN@RBD significantly enhanced the secretion of IFN-γ by CD8 + T cells (Fig. 6B). The results of the enzyme-linked immunospot (ELISpot) assay were consistent with the flow cytometry results; it demonstrated that DMOSN@RBD induced higher levels of IL-4 and IFN-γ compared to the alumOH adjuvant (Fig. 6C; Fig. S5C). To evaluate the functionality and polarization of antigen-specific T cells, we examined the cytokines secreted by spleno cytes in response to S1 and S2 polypeptide stimulation. In DMOSN@RBD-immunized mice, both TH1 (TNF-α, IFN-γ, and IL-2) and TH2 (IL-4 and IL-5)-associated cytokines were significantly elevated; however, TH1-associated cytokines were not increased slightly in alumOH+RBD-immunized mice (Fig. 6D through H). These findings suggest that DMOSN@RBD vaccination induces a more balanced TH1 and TH2 immune response. Additionally, we characterized the long-term memory T cells induced by the DMOSN@RBD vaccine. Our results showed that immunization with DMOSN@RBD vaccine significantly induced the generation of TEM (CD44 + CD62L -) in CD8 + T cells (Fig. 6I; Fig. S6A), but there was no significant advantage of DMOSN@RBD in CD4 TEM (Fig. S6B andC). These results highlight that DMOSN, as an adjuvant, effectively delivers the antigen while stimulating a robust humoral immune response and cellular immune response. It compensates for the limitations of subunit vaccines, which may lead to an imbalance between cellular immune responses, thus assisting the body in combating viral infec tions.
## DISCUSSION
DMOSN have been widely used as delivery vehicles for small-molecule nucleic acid and protein drugs. Recently, they have also gained attention as vaccine delivery systems (14).
In this study, we successfully developed a self-adjuvanting SARS-CoV-2 biomimetic nanovaccine employing DMOSN as a delivery system. DMOSN@RBD demonstrated potent broad-spectrum protection against multiple SARS-CoV-2 mutant strains. Further more, DMOSN enhanced the internalization of the antigen by macrophages and DCs. Specifically, within 24 h after immunization with DMOSN@RBD, it led to the recruitment and activation of DCs, which in turn resulted in the amplification of antigen-specific humoral and T-cell immune responses. In summary, the preparation process of the DMOSN@RBD biomimetic nanovaccines is straightforward and environmentally friendly. It also exhibits remarkable capabilities in activating both humoral and cellular immunity.
The DMOSN-based delivery system holds great potential for delivering various antigens and constructing universal nanoadjuvant vaccines, thus presenting a novel approach for vaccine design. Subunit vaccines are weakly immunogenic, which necessitates the addition of adjuvants to achieve higher immunogenicity and reduce the required antigen dosage (23). Therefore, the development of an efficient antigen delivery system that can effectively integrate humoral and cellular immunity is crucial for adequate immune protection. Among the various virus-inspired nanocarriers, silica-based nanoparticles have been widely used in vivo studies, owing to their unique advantages such as excellent biocompatibility, versatile surface chemistry, and structural stability (24)(25)(26). Additionally, silica nanomaterials are biodegradable, capable of slowly dissolving in aqueous solutions to release nontoxic silicic acid, and can be rapidly cleared in vivo. This characteristic significantly enhances their biosafety in practical applications (27,28). In our study, we employed the SARS-CoV-2 RBD protein as an immunogen and formed DMOSN on the surface. This design ensured efficient encapsulation of the antigens, thereby improving their delivery to the host. Notably, compared with traditional alumadjuvanted vaccines, DMSON@RBD induced higher levels of neutralizing antibodies. This enhanced immune response can be attributed to the fact that the coated silica created a relatively enclosed environment for the RBD protein, reducing its interaction with the aqueous solution and improving the immobilization and reinforcement of the RBD structure (29). Conversely, when the RBD protein was adsorbed by the aluminum adjuvant, there was a significant decrease in the α-helix content, which was compensa ted by the increase in the β-sheet content. It is hypothesized that the strong ligand exchange between the aluminum adjuvant and the RBD protein may destabilize the protein structure, causing the RBD protein to aggregate on the surface of the aluminum adjuvant and increasing the content of β-sheets. Ultimately, this leads to a reduction in antigen immunogenicity (30). DCs are the most powerful and only type of APCs capable of activating naïve T cells and initiating the primary immune responses (31,32). Virus-like mesoporous silica nanoparticles possess a greater capacity for antigen adsorption or internalization compared to silica nanoparticles with a smooth surface or spherical (24,28). In vivo and in vitro studies have evidenced the superior ability of DMOSN@RBD to target lymph nodes and recruit and activate DCs in draining lymph nodes. Notably, the activation of DCs represents a pivotal step in triggering T-cell-mediated immune responses induced by DMOSN@RBD. Antigen-specific T cells play an essential role in eliminating infections during the early stages of pathogen transmission and replication (23,33,34). On the other hand, although aluminum-based adjuvants are widely used as key components of subunit vaccines (35), alum is not efficient enough in vaccine formulation due to the biased humoral immune response (36). This is particularly evident in scenarios where balanced humoral and T-cell-mediated immune responses are necessary for the preven tion of infections and clearance of infected cells (37). These findings indicate that DMOSN has both antigen delivery ability and adjuvant potential, indicating the potential clinical value of DMOSN in vaccine administration.
In conclusion, we demonstrated the potential of DMOSN as a multifunctional nanoadjuvant for recombinant subunit vaccines. Additionally, we have further illustrated that SARS-CoV-2 biomimetic nanovaccines based on DMOSN provide strong, broad-spec trum protection. Specifically, the advantages of DMOSN as a nanoadjuvant for subu nit vaccines lie in triggering an effective and longer-lasting antibody response while inducing a potent cellular immune response that confers resistance against SARS-CoV-2 infections. Overall, as a mesoporous nanoparticle, DMOSN holds significant promise for further optimization to enhance vaccination efficacy. As a nanoadjuvant, DMOSN can serve as a versatile platform for vaccine design targeting pandemics and emerging infectious diseases.
## MATERIALS AND METHODS
## Preparation of dendritic mesoporous organosilica nanoparticles
In a typical synthesis, the cationic surfactants CTAB and NaSal were used as structuredirecting agents, TEOS and BTEE were used as silica sources, and TEA was a catalyst. First, 450 mg of CTAB and 200 mg of NaSal were added to 30 mL of water and stirred gently at 50°C in an oil bath under magnetic stirring for 2 h. Next, 0.082 g of TEA was added to the above solution and kept stirred for 2 h at 80°C. Then, a mixture of 2.5 mL of TEOS and 2 mL of BTEE was added to the mixture with gentle stirring for 12 h. The products were collected by high-speed centrifugation and washed several times with ethanol to remove the residual reactants. Furthermore, the collected products were refluxed in acetone at 60°C for 6 h five times to remove the template, followed by drying under vacuum at 60°C overnight.
## Characterization of DMOSN
The morphology and composition were studied employing a field-emission SEM (S-4800, Hitachi) instrument equipped with an energy dispersive X-ray spectrometer. TEM images and elemental mapping images were obtained using an FEI Tecnai G2S-Twin instrument with a field-emission gun operating at 200 kV. XPS measurements were obtained using a Thermo Scientific K-Alpha spectrometer. UV-vis absorbance spectra were recorded with a Shimadzu UV-3600 instrument. Infrared spectra were collected with a ThermoFisher Nicolet 6700 infrared spectrophotometer. Determination of zeta potential and hydrated particle size of nanoparticles by Malvern Nanoparticle Size and Potentiometry. The nitrogen adsorption-desorption isotherms were measured by a specific surface area physical adsorption instrument (Micromeritics, ASAP2020M). Cell culture was performed with a Thermo Fisher Incubator.
## RBD loading and release
For the protein loading, the SARS-CoV-2 Wuhan-Hu-1 RBD protein was produced using the Expi293 expression system and dissolved in 0.1 M PBS (pH 7.4) to obtain a protein stock solution with a concentration of 0.2 mg/mL. The RBD solution (5 mL) was added to 10.0 mg of the nanoparticles in 10 mL capped vials. The resulting mixture was shaken at 400 rpm at room temperature for 12 h, followed by centrifugation. The obtained DMOSN@RBD was washed once with PBS and subsequently lyophilized.
For the protein release test, 3.0 mg of nanoparticles after RBD loading were immersed in 1.5 mL of PBS solution and stirred at 200 rpm at 37°C. At predetermined time points, the solution was centrifuged, and the supernatant was removed and replaced with the same amount of fresh PBS solution and used BCA detection kit for protein concentration determination.
## Preparation of FITC-labeled DMOSN@RBD
First, 10 mg of DMOSN@RBD and 0.5 mg of FITC were dissolved in 5 mL of PBS. The mixture was stirred in the dark for 4 h (room temperature, 400 rpm), followed by centrifugation to collect the product, washing twice with PBS, and lyophilization afterward.
## Western blot analysis
The proteins were collected and electrophoresed on 10% SDS-PAGE polyacrylamide gels at 150 V. The proteins were transferred to nitrocellulose membranes for western blotting. The SARS-CoV-2 RBD protein antibody (Sino Biological, China) was used at a 1:3,000 dilution, and goat anti-rabbit IgG (Bioworld, USA) was used at a 1:25,000 ratio. The bands were visualized with SuperSignal West Atto (Thermo Scientific, USA).
## Animal immunization and challenge
Female BALB/c mice aged 6-8 weeks (n = 3) were inoculated with DMOSN@RBD via intravenous injection, and various tissues were collected 28 days post-inoculation for pathological analysis.
Female BALB/c mice (6-8 weeks old, n = 10) received primary and booster immu nizations via intramuscular injection on days 0, 7, and 21. Mice were randomized to receive DMOSN@RBD, AlumOH+RBD, soluble RBD, DMOSN alone, or PBS. Dosing was standardized to 10 µg of RBD per mouse in 100 µL PBS for DMOSN@RBD, AlumOH+RBD, and monomeric RBD groups, with the DMOSN@RBD formulation containing 1 mg of DMOSN carrier. The DMOSN blank control group received an identical DMOSN dosage (1 mg) without RBD conjugation. Biological samples were collected at predefined time points for downstream analyses.
For the challenge of SARS-CoV-2 in 9-month-old BALB/c mice with SARS-CoV-2 at 21 dpv, the mice were inoculated intranasally with a lethal dose (10 5 TCID 50 ) of the mouse-adapted strain C57MA14. All experiments were performed in the BLS-3 labora tory, and clinical signs (body weight, respiratory distress, tremor, and limb paralysis) were observed daily. At 3 dpi, three mice were randomly selected for blood biochemical analysis, followed by the collection of lung, turbinate bone, and serum samples. Animals that survived the infection were euthanized at 13 dpi, and serum was collected.
For immunization and challenge, at 21 days after the booster vaccination, golden hamsters were challenged with 10 5 TCID 50 of SARS-CoV-2, including the WT, B.1.351, B.1.617.2, Omicron BA.1, and BA.2 strains. The clinical symptoms and weights of the stimulated animals were continuously recorded for 1 week after infection. At 3 dpi, nasal turbinates and lung tissue were collected for TCID 50 titration and RNA copy detection.
## Serum neutralization assay
Neutralizing antibody titers against SARS-CoV-2 were quantified via a serum neutraliza tion test with eGFP-expressing vesicular stomatitis virus (VSV) pseudotypes harboring the SARS-CoV-2 S gene. First, heat-inactivated mouse serum was incubated at 56°C for 30 min. Subsequently, the serially diluted serum samples were incubated at 37°C for 1 h with an equal volume of S-pseudotyped VSV particles and inoculated into Vero E6 cells at 37°C with 5% CO 2 cultivation for 48 h. The neutralizing antibody titer was defined as the reciprocal of the serum dilution required to eliminate all eGFP
## Enzyme-linked immunosorbent assay
Detection of SARS-CoV-2 anti-S-specific IgG and its subtypes by ELISA. In summary, the SARS-CoV-2 S protein was prepared in our laboratory and then encapsulated in 96-well polystyrene plates overnight at 4°C. After three washes with PBST, 1% bovine serum albumin (BSA, Sigma, Germany) was added to each well for 2 h at 37°C. Next, the mouse or golden hamster serum was added at multiple dilutions and incubated for 1 h at 37°C. After washing three times with PBST, HRP-labeled goat anti-mouse IgG (Bioworld, USA), IgG1, and IgG2a antibodies (Southern Biotech, USA) were added and incubated for 45 min at 37°C. After washing three times with PBST, 3, 3′, 5, 5′-tetramethylbenzi dine (TMB, Sigma, Germany) was added for color development, and the reaction was terminated by adding 2 mol/L H 2 SO 4 at the appropriate time and measuring the OD at 450 nm (Bio-Rad, USA).
## Histopathology and immunohistochemistry
Tissue was fixed in 4% paraformaldehyde, and 3-5 μm paraffin sections were prepared and stained with hematoxylin and eosin for histopathological examination.
## Quantitative reverse transcription-PCR
Tissue or viral RNA was extracted for quantitative reverse transcription-PCR (qRT-PCR) detection as previously described. The collected lung tissue and nasal turbinate were ground with a tissue homogenizer, and the supernatant was collected. Viral RNA was extracted using a Tiangen virus RNA kit (Tiangen, China), and viral RNA was quantified using qRT-PCR targeting the SARS-CoV-2 N gene.
## Viral loads
The supernatants of the nasal turbinate and lung tissue homogenates were serially diluted in DMEM, and then Vero E6 cells were added to 96-well plates. After 72 h of culture at 37°C with 5% CO 2 , the TCID 50 was detected for cytopathic effects.
## Complete blood cell counts
To determine the complete blood cell counts, samples were analyzed using an auto hematology analyzer (BC-5000vet, Mindray, China) according to the manufacturer's instructions.
## Mesoscale discovery
For serum collection, blood samples were harvested 7 days after primary immunization. Whole blood was centrifuged at 3,000 × g for 30 min at 4°C, and the supernatant serum was collected and stored for subsequent analysis. For splenocyte supernatant preparation, spleens were isolated 35 days post-immu nization, and splenic lymphocytes were separated. Cells were stimulated with the SARS-CoV-2 S1/S2 peptide library for 48 h. Cultured supernatants were then centrifuged at 3,000 × g for 20 min at 4°C, and the cell-free supernatants were collected for down stream assays.
The mesoscale discovery assay (Univ#K15048D-X, China) was carried out according to the manufacturer's instructions. The plates were analyzed on a Sector Imager 2400 system, and cytokine concentrations were calculated based on the standard curve generated in the Discovery Workbench 4.0.12 software with a 4-parameter logistic nonlinear regression analysis.
## Flow cytometry assay
Flow cytometry was performed to detect immune cells in the inguinal lymph nodes. Briefly, at 35 dpv, after the tissue was collected and ground, it was dispersed into individual cells through a 70 mm nylon filter, washed with PBS containing 0.2% BSA, and then treated with a sealing solution containing CD16/32 (Thermo Scientific, USA) for 20 min. A total of 10 (6) cells were stained with fluorescently labeled antibodies, incubated for 30 min, and then washed again. Finally, the stained cells were analyzed using a FACSVerse (Beckman Coulter, USA) instrument.
At 35 dpv, mouse spleen cells were extracted, and the cell concentrations were adjusted to 10 6 /mL with RPMI 1640 (Gibco, USA) culture medium. A lymphocyte proliferation assay was performed with polypeptide (the SARS-CoV-2 S1 [Sino Biological, Cat: PP003-A] and S2 [Sino Biological, Cat: PP003-B] peptide libraries) as the specific stimulator for 12 h, and ConA (500×; Thermo Scientific, USA) as the positive stimulator, as previously described. BFA (1000×; Thermo Scientific, USA) was added to the extracted mouse spleen lymphocytes to block the secretion of intracellular cytokines for 4 h. The cells were then stained with anti-CD3, anti-CD4, and anti-CD8 antibodies, fixed and permeabilized in Fix/Perm buffer, and stained with anti-IL-4 and IFN-γ in Fix/Perm buffer, and flow cytometry was used to detect changes in the cells.
The mouse antibodies used, including CD11c (FITC, N418), CD86 (PE, GL1), CD40 (APC, 1C10), GL-7 (PE, GL-7 [GL7]), PD-1 (PE, J43), CD38 (PerCP-eFluor 710, 90), IFN-γ (APC, 4S.B3), IL-4 (PE, 11B11), CD44 (APC, MEM-263), and CD62L (APC-eFluor 780, MEL-14), were obtained from Thermo Scientific. The antibodies against MHC II (APC/Cyanine7, M5/114.15.2), CD45R (FITC, RA3-6B2), CD4 (FITC, GK1.5), CXCR5 (APC, L138D7), CD8 (PE/ Cyanine7, 53-6.7), CD19 (APC/Cyanine7, 1D3/CD19), CD138 (PE, 281-2), and CD3 (APC/ Cyanine7, 145-2C11) were obtained from Biolegend.
In vitro cellular uptake RAW264.7 and DC2.4 cells were seeded at 1 × 10 6 cells per well in 6-well plates, and uptake experiments were performed until the cells were 80% confluent. The cells were incubated for 4 h in 1 mL of serum-free RPMI 1640 medium containing 5 µg of FITC-RBD or 20 µg of DMOSN loaded with 5 µg of FITC-RBD. Then, the cells were collected, centrifuged, and washed three times with PBS, and antigen uptake was detected using a flow cytometer or indirect immunofluorescence.
## ELISpot
An ELISpot detection kit was used to detect IFN-γ and IL-4 levels (Mabtech, Sweden). According to the manufacturer's recommended method for testing, a total of 5 × 10 5 splenic cells were added to each well and treated with 10 µg/mL peptide protein stimulation. The plate was incubated at 37°C with 5% CO 2 for 36 h. Concanavalin (Thermo Scientific, USA) was used as a positive control. The tablet was scanned on an ImmunoSpot reader. ImmunoSSpot software was used to count specific spots. The number of specific spots/pores must be twice the average value found in each negative control well, and then the background value is subtracted.
## References
1. Su, Chen, Yuan et al. (2020) "Multi-omics resolves a sharp diseasestate shift between mild and moderate COVID-19" *Cell*
2. Aggarwal, Akerman, Milogiannakis et al. (2022) "SARS-CoV-2 Omicron BA.5: evolving tropism and evasion of potent humoral responses and resistance to clinical immunotherapeutics relative to viral variants of concern" *EBioMedicine*
3. Cao, Wang, Jian et al. (2022) "Omicron escapes the majority of existing SARS-CoV-2 neutralizing antibodies" *Nature*
4. Reed, Orr, Fox (2013) "Key roles of adjuvants in modern vaccines" *Nat Med*
5. Pati, Shevtsov, Sonawane (2018) "Nanoparticle vaccines against infectious diseases" *Front Immunol*
6. Saunders, Lee, Parks et al. (2021) "Neutralizing antibody vaccine for pandemic and pre-emergent coronaviruses" *Nature*
7. Arunachalam, Feng, Ashraf et al. (2022) "Durable protection against the SARS-CoV-2 Omicron variant is induced by an adjuvanted subunit vaccine" *Sci Transl Med*
8. (2025) *Full-Length Text Journal of Virology*
9. Joyce (2021) "SARS-CoV-2 ferritin nanoparticle vaccines elicit broad SARS coronavirus immunogenicity" *Acta Crystallogr A Found Adv*
10. Wuertz, Barkei, Chen et al. (2021) "A SARS-CoV-2 spike ferritin nanoparticle vaccine protects hamsters against Alpha and Beta virus variant challenge" *NPJ Vaccines*
11. King, Joyce, Lakhal-Naouar et al. (2021) "Efficacy and breadth of adjuvanted SARS-CoV-2 receptor-binding domain nanoparti cle vaccine in macaques" *Proc Natl Acad Sci*
12. Chen, Zhang, Yuan et al. (2022) "Development of receptor binding domain (RBD)-conjugated nanoparticle vaccines with broad neutralization against SARS-CoV-2 Delta and other variants" *Adv Sci (Weinh)*
13. Geng, Tai, Baxter et al. (2021) "Novel virus-like nanoparticle vaccine effectively protects animal model from SARS-CoV-2 infection" *PLoS Pathog*
14. Diaz, Care, Sunna (2018) "Bioengineering strategies for proteinbased nanoparticles" *Genes (Basel)*
15. Hong, Zhong, Du et al. (2020) "The pore size of mesoporous silica nanoparticles regulates their antigen delivery efficiency" *Sci Adv*
16. Yang, Lu, Abbaraju et al. (2017) "Multishelled dendritic mesoporous organosilica hollow spheres: roles of composition and architecture in cancer immunotherapy" *Angew Chem Int Ed*
17. Lu, Yang, Gu et al. (2018) "Glutathionedepletion mesoporous organosilica nanoparticles as a self-adjuvant and Co-delivery platform for enhanced cancer immunotherapy" *Biomaterials*
18. Deng, Li, Sun et al. (2022) "Rational development of a polysacchar ide-protein-conjugated nanoparticle vaccine against SARS-CoV-2 variants and Streptococcus pneumoniae" *Adv Mater*
19. Yan, Li, Wang et al. (2022) "Characterization of two heterogene ous lethal mouse-adapted SARS-CoV-2 variants recapitulating representative aspects of human COVID-19" *Front Immunol*
20. Wang, Deng, Gou et al. (2020) "Preliminary study to identify severe from moderate cases of COVID-19 using combined hematology parameters" *Ann Transl Med*
21. Sia, Yan, Chin et al. (2020) "Pathogene sis and transmission of SARS-CoV-2 in golden hamsters" *Nature*
22. Pulendran, Arunachalam, Hagan (2021) "Emerging concepts in the science of vaccine adjuvants" *Nat Rev Drug Discov*
23. Lu, Shih, Qi (2017) "Ephrin B1-mediated repulsion and signaling control germinal center T cell territoriality and function" *Science*
24. Shi, Zhu, Xia et al. (2019) "Vaccine adjuvants: understanding the structure and mechanism of adjuvanticity" *Vaccine (Auckl)*
25. Wang, Wang, Tang et al. (2017) "Facile synthesis of uniform virus-like mesoporous silica nanoparticles for enhanced cellular internalization" *ACS Cent Sci*
26. Ngamcherdtrakul, Reda, Nelson et al. (2021) "In situ tumor vaccination with nanoparticle co-delivering CpG and STAT3 siRNA to effectively induce whole-body antitumor immune response" *Adv Mater*
27. Song, Yu, Lu et al. (2017) "Plasmid DNA delivery: nanotopography matters" *J Am Chem Soc*
28. Song, Nor, Yu et al. (2016) "Silica nanopollens enhance adhesion for long-term bacterial inhibition" *J Am Chem Soc*
29. Lin, Revia, Zhang (2021) "Inorganic nanomaterial-mediated gene therapy in combination with other antitumor treatment modalities" *Adv Funct Materials*
30. Wang, Liu, Xiao et al. (2018) "Biomineralization state of viruses and their biological potential" *Chemistry A European J*
31. Jesus, Fragal, Rubira et al. (2018) "The inclusion of chitosan in poly-ε-caprolactone nanoparticles: impact on the delivery system characteristics and on the adsorbed ovalbumin secondary structure" *AAPS PharmSciTech*
32. Joffre, Segura, Savina et al. (2012) "Cross-presentation by dendritic cells" *Nat Rev Immunol*
33. Segura, Amigorena (2015) "Cross-presentation in mouse and human dendritic cells" *Adv Immunol*
34. Sun, Xia (2016) "Nanomaterial-based vaccine adjuvants" *J Mater Chem B*
35. Liang, Wang, Yu et al. (2022) "Mechanistic understanding of the aspect ratio-dependent adjuvanticity of engineered aluminum oxyhydroxide nanorods in prophylactic vaccines" *Nano Today*
36. Hogenesch, Hagan, Fox (2018) "Optimizing the utilization of aluminum adjuvants in vaccines: you might just get what you want" *NPJ Vaccines*
37. Sun, Ji, Liao et al. (2017) "Enhanced immune adjuvant activity of aluminum oxyhydroxide nanorods through cationic surface functionalization" *ACS Appl Mater Interfaces*
38. Moderbacher, Ramirez, Dan et al. (2020) "Antigen-specific adaptive immunity to SARS-CoV-2 in acute COVID-19 and associations with age and disease severity" *Cell* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12855637&blobtype=pdf | # Scientific Reviewing-Editors' Memo to Emerging Reviewers -2
Shahul Ebrahim, Ziad Memish
## Abstract
Peer reviewers-the Linchpin RoleAs the linchpin of scientific publishing, reviewers are essential and usually anonymous. Reviewers are largely uncompensated custodians of the process. Review styles differ, and no two reviewers reveal the same review-but may capture major drawbacks. Following the "Swiss-cheese" analogy, multiple reviewers cover the gaps inherent in solo review. If faced with review fatigue, decline rather than offering an inappropriate review.• Editors' "Third eye".Reviewers help assess the paper's relevance to the journal's aims; the rationale (why the work was conducted); the quality of the science; its broader value (for example, whether a narrow sub-analysis adds little beyond existing work such as a sub analysis of chronic disease DALYs among 70 to 75 years old); potential harms (e.g., stigma, social isolation of a population group, or causing public harm); and the need for specialized review (e.g., statistics, drug trial experts, ethicists) or transfer to a more suitable journal.• Authors' sounding board.Reviewers have an unwritten mandate as mentors. Pair critiques with concrete suggestions (e.g., "Statement X does not capture the findings; consider revising to '…'"). Help authors surface overlooked insights and temper overstatements. Strive for objectivity and clarity. Acknowledge strengths before listing challenges. Overly harsh or unconstructive feedback should be avoided irrespective of the quality of the paper.• Equity (not only equality) enhancers.Reviewers curate scientific thought-finding needles in haystacks. Applying identical review metrics to every paper (equality) may not reveal the best science; equitable, Shahul H. Ebrahim
## 1 Introduction
"A journal is only as good as its reviewers"-wrote a journal in an annual thank you note to its reviewers. Peer review is the quality-control backbone of scientific writing. Peer review, likely began with the birth of scientific journals (1665, Philosophical Transactions, Royal Society) as editorial judgment combined with occasional consultation, streamlined in 1752 via the "Committee of Papers," and was used with increasing frequency in the late 1800s by The Lancet (1823) and Nature (1869) [1][2][3]. The post-World War II boom in science and the emergence of U.S. grant systems required institutionalization of peer review [1]. Journals adopted peer review widely in the 20th century to meet demands for objectivity amid an increasingly rising volume of papers [2]. The Committee on Publication Ethics (COPE) (https://publicationethics.org), founded in 1997, helped articulate guidance on peer review [4]. The International Committee of Medical Journal Editors (www.icmje. org) and major publishers provide further guidance.
Increases in both journals and submissions, and associated 'reviewer fatigue 'increase the demand for reviewers [5]. Article retractions increases scrutiny of reviews. Though review approaches vary across reviewers, general guidance can inform emerging reviewers. We summarize our independent observations from three decades as journal editors, external peer reviewers, and institutional reviewers in national and multinational health agencies. tailored guidance can elevate valuable work. For example, an Ebola outbreak report from Congolese researchers (writing in a second language) may be less polished than a single Ebola (imported) case report from Europe or the U.S., yet the outbreak report may offer greater public-health value. Reviewers can help polish it. The North-South divide in scientific communication remains a challenge even today [6], and reviewers can play a role through the equity lens.
## 3 Review Formats
While formats vary, the aims are consistent: improve rigor, enhance transparency, ensure utility for readers, and safeguard trust.
## Reference credibility and balance
Not every statement needs a reference. Recency and relevance of citations not dominated by self-citations and supportnot substitute for-reasoning. Useful suggestions to readers for further exploration of the key domains and statistics.
## Review comments to the authors and editors
## Respectful persuasive commenting
Editors' and authors' expectations vary. Help the Editor see the merits and demerits whereas authors would benefit from pointed suggestions to improve the marketability and readability of the work. Refrain from harsh comments and highlight the value first. If it is a perfect paper, say it.
Review of the material presented to you can reveal what is not revealed. Some authors explicitly state why certain elements are not included.
• Where to start-"the 360-degree view".
A well written piece can be spotted easily [7]. Begin with layout and structure (IMRAD: Introduction, Methods, Results, and Discussion), author list, abstract, references, disclosures, visuals, and word count. This overview should help you flag excess tables/figures, imbalanced sections, or misalignment with journal guidance. Even when IMRAD headings aren't explicit, the flow should follow IMRAD. Note observations using tracked changes or numbered comments.
• Author /institution (if not blinded) alignment with scope.
While solo authorship is fine in principle, large surveys, complex analyses (e.g., analysis of NHANES [8] on cardiometabolic indicators requiring statistical, epidemiologic, laboratory, and policy expertise), multi-country studies, or other large projects merit scrutiny if conducted by a lone author. Solo vs. multiauthor dynamics can be gleaned from language, style, structural, acknowledgment cues ("I" vs. "We"; voice consistency, usage drift ["dataset" vs. "data set", "color" vs. "colour"], different levels of jargon, terminology misalignment ["participants" vs "subjects"] asymmetry in depth and visuals). In multiauthor papers, the lead and senior authors balance styles, depths, and content symmetry. Although science is not monopoly of any discipline or specialty, products from adjacent or unusual disciplines may require some scrutiny. If it is a blind review, these will be an Editor's task.
• Title-to-abstract concordance.
Most journals require a structured abstract. Check synchrony between title and abstract. A well-synthesized abstract (that conveys the essence of the often signals a well-written paper, though AI-assisted prose can obscure weaknesses.
• The title is Window; the abstract does the advertising.
Check if the title is sufficient to prompt readers' interest and helps with inclusion in literature search. Evaluate whether the abstract accurately reflects aims, methods (quantitative vs. qualitative), recency of data, key results (with statistics where appropriate), and conclusions (without repeating the results verbatim). Conclusions are not mere repletion of findings; rather synthesized summary linked to policy or suggestive of other studies. Abstracts that relegate findings to the main body ("Findings are discussed") are inadequate. Even if the abstract is weak, review the full paper and provide guidance to strengthen it.
• Relative independence of abstract, visuals, and main text.
Abstracts should standalone-meaning the readers do not need to read the entire article to discern key findings. Similarly, tables/figures should have full, self-contained titles and define abbreviations in footnotes. The main text must let readers understand the work without chasing details elsewhere; it should also synthesize the key points from the visuals (not restate every cell).
• Read the main body; review visuals and references
The main body of the paper tells the story. As the main body of a paper is a continuous part of the paper, the sections should be devoid of verbatim repetition. Check the quality of science and assess potential reputational risk (to the journal, to the reviewer, for science).
• Introduction (≈ 10-15% of words): Asses the flow: big- picture Rationale → Contextual Gap → Objective.
• Methods (≈ 20-25%) should follow logically from intro- duction/background, enough detail to replicate; specify data sources and analytical framework; cite instruments and standards, methodological biases mitigated, confirmative analysis by second analyst if needed for highly consequential findings and presented and interpreted appropriately.
• Results (≈ 25-35%): check for objective presentation logically following from the introduction/background/ methods; check arithmetic, denominators, consistency (e.g., row vs. column percents), and statistical interpretations of data. In tables and figures, assess visual clutter (e.g., > 5 columns can hinder readability) and graphic quality. Keep totals of tables and figures combined to ≲ 5 for a ~ 3,000-word paper, place extras in supplements. Check if data interpretation is tied to reality and contemporary relevance to public health. For example, few cases of a new strains of influenza may not be statistically significant as compared to circulating strains, but they are important pointers on disease emergence patterns. Such interpretations require both a good grasp of multiple disciplines including statistics, public health, and virology. Check for the desire to interpret correlation as causation, especially in public health interventions, and should be avoided. Help the authors to frame correlations accordingly.
• Discussion (≈ 30-35%) usually begins with clear take- home points ("We found…, This study reveals…. "), relate back to aims, interpret implications, acknowledge limitations, and suggest concrete next steps-without repeating results verbatim. Look for grandiose or subjective statements without providing magnitude or evidence.
• Determine the need for specialized review (e.g.: statis- tics, virology, policy) as relevant.
• Assess balance in references, verify they truly support the claims (don't rely on titles alone) and check exaggerations ("eg: Studies show X" when there is only one study). Rather than secondary references, primary sources, review articles, government and institutional documents are better. Check every reference for authenticity.
Identify reading "speed-breakers" including abbreviation and jargon overload. Confirm ethical approval (as relevant: institutional review, consent statements, animal welfare), conflicts of interest, funding sources (major funding sources have systems in place to assure good practices and can
• "Salami slicing", substandard, and papermill products.
A paper can be well-written [7], yet not written by the submitting author, or not authentic. Be alert to Artificial Intelligence-generated or ghost-authored papers which can include superficial re-analyses of public datasets (also known as "Salami slicing": e.g., Global Burden of Diseases data, UK Biobank, NHANES), and social-media based studies that imply generalizability without robust sampling.
## 4.3 by Artificial Intelligence (AI)
Though AI is increasingly becoming a task master, journals including Journal of Epidemiology and Global Health currently disallow AI -generated reviews. Tools [9] to detect AI reviews are emerging. Large language models based on predetermined algorithms can miss "needles in the haystack" the very aspect a human reviewer is entrusted to find. Even AI-related specialties such as machine learning are challenged with AI-generated reviews (about 21% of abstracts for an upcoming conference) [9]. They flagged hallucinated citations, long and vague feedback, and request for unusual statistical analysis [9]. Uploading unpublished manuscripts (to AI platforms) can leak data and feed model training using unpublished scientific work that can impact future use of AI [5]. Uploading unpublished manuscripts into AI during peer review goes against Journal of Epidemiology and Global Health's editorial policies and suspected cases are investigated ( h t t p s : / / w w w . b i o m e d c e n t r a l . c o m / g e t p u b l i s h e d / e d i t o r i a l -p o l i c i e s). Other challenges [5] include undermining author trust in peer-review process, quality assurance burden due to increasing editor/reviewer verification workload, erosion of human critical thinking, conservatism and monoculture promotion by Large language models, and author optimization of manuscripts for reviewers using AI. Also, if a review is entirely AI-driven, there is no benefit for the Editor to solicit external reviews.
Despite the above challenges, there is value in automating the routine parts of peer review to boost speed and quality, such as identifying required elements in manuscripts [5]. If this is a desired path (more guidance will likely emerge), we need to keep human editors firmly "in the loop" to judge impact, novelty, and clinical relevance-and to retain full oversight, accountability, and final decision-making.
## 4.4 Write the Reviewer Report
Start with the work's relevance (e.g., "Given the emerging relevance of vaccine hesitancy, authors' efforts to assess vaccine uptake in X is valuable…"). Separate major (theme, methods, results, interpretive coherence, be reassuring), and data-use permissions. Though mostly benign, missing statements are red flags, as you are not sure if they are purposeful omissions. Therefore, trust but verify.
• Remember-you are not the author! Focus the review on the paper being reviewed, not what you wish the author had written or conducted. Accept that there are different ways of approaching a problem. Your suggestion for further data extrapolation should be within the context of the work in review.
• Help prune verbosity, package messages, and improve clarity and utility.
After a complete read, suggest sample rewrites that tighten prose, foreground magnitudes, and replace vague claims with specific, concise statements. Mention if comments apply across the entire paper so that you don't need to repeat the comment. A repetitive or sub-analysis paper may find utility as a letter to the editor on a previously published piece. The utility of a study that lacks population representation may be improved by specifying the 'selective' nature of the study beginning in the title (e.g.: social media-based study may lack representativeness of the entire population but can be model for future studies of the population of social media users) and may get traction as a brief report.
• Minimize bias.
Bias is human and reviewers view papers through their own concepts and experience. Bias awareness is critical for reviewers. We (reviewers) can train ourselves to minimize them by strengthening our objectivity and equity lens. Try reading without focusing attention on the authorship and institutions associated with the work. No reviewer would want to be associated with a retracted manuscript.
## 4.2 Discerning the 'Between the Lines' Part
## 4.5 Skill-Continuity and Skill-Multiplication
Reviewers aren't cloned-they evolve. Co-reviewing with a mentor and comparing notes can accelerate development. Review acumen grows with practice, topical expertise, field experience, publication track record, and knowledge of current developments in the field. Opportunities to enhance review experience include Springer Nature's new initiative, Reviewer Communities ( h t t p s : / / c o m m u n i t i e s . s p r i n g e r n a t u r e . c o m / p o s t s / j o i n -t h e -s p r i n g e r -n a t u r e -r e v i e w e r -c o m m u n i t y). Researchers can apply to review with Srpinger Nature journals, gain training on research integrity checks and how to provide a high-quality review, and access an online platform for relevant resources.
## 5 Summary
Reviewing a well-written paper is rewarding, but many are challenging. Peer review helps assess validity, originality, and clarity before research enters the scholarly record. Although imperfect-subject to bias, uneven standards, and delays -peer review remains the most widely accepted filter of scientific work. Editorial policies and journal specific policies evolve over time. ( h t t p s : / / w w w . b i o m e d c e n t r a l . c o m
## References
1. Kronick (1990) "Peer review in 18th-century scientific journalism" *JAMA*
2. Burnham (1990) "The evolution of editorial peer review" *JAMA*
3. Garner (0472) "Celebrating 350 years of Philosophical Transactions: physical sciences papers" *Philos Trans A Math Phys Eng Sci*
4. Wager (2012) "The committee on publication ethics (COPE): objectives and achievements 1997-2012" *La Presse Médicale*
5. Perlis, Christakis, Bressler (2025) *Artif Intell Peer Rev JAMA*
6. Blicharska, Smithers, Kuchler (2017) "Steps to overcome the North-South divide in research relevant to climate change policy and practice" *Nat Clim Change*
7. Ebrahim (2025) "Scientific Writing-an editor's memo to emerging Authors-1" *J Epidemiol Global Health*
8. Mcquillan, Mclean, Chiappa (1988) *National Health and Nutrition Examination Survey Biospecimen Program: NHANES*
9. Naddaf (2025) "AI is transforming peer review-and many scientists are worried" *Nature*
10. "Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12585557&blobtype=pdf | # Structural Basis of Activity of and Resistance to HIV Integrase Inhibitors
Dmitry Lyumkis
## Abstract
The Human Immunodeficiency Virus Type 1 (HIV) currently infects ~40 million people worldwide. In the absence of a cure, antiretroviral therapy represents the primary treatment option, because it slows disease progression and reduces new infections. Integrase (IN) Strand Transfer Inhibitors (INSTIs) are a class of antiretroviral therapeutics that block integration of viral DNA into host chromosomes. This process is mediated by the viral IN enzyme, which assembles into oligomeric nucleoprotein complexes on the ends of viral DNA, termed "intasomes". INSTIs selectively target intasomes and represent first-line therapies in the clinic. However, the emergence of IN variants resistant to INSTIs is becoming a greater clinical problem. We are using interdisciplinary approaches that include structural biology via cryogenic electron microscopy, biochemistry, cellular virology, and computation to provide a mechanistic understanding of both how and why select drug resistant mutations arise in response to leading clinically used drugs, identify and analyze novel pathways of drug resistance, and devise strategies to predict how resistant mutations emerge. We are also working with chemists who are developing more potent therapeutics that aim to be effective against drug resistant viruses. Our work aims to improve our understanding of an important class of drugs used to treat people living with HIV, identify mechanisms, pathways, and patterns of clinically relevant resistance to INSTIs, and provide specific guidelines for their rational improvement, under a broader umbrella of "personalized medicine". |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12772238&blobtype=pdf | # Generation, characterization, and application of caprine herpesvirus 1 secreted glycoprotein D
Sergio Minesso, Amienwanlen Odigie, Valentina Franceschi, Simone Taddei, Vittorio Madia, Maria Tempesta, Gaetano Donofrio
## Abstract
Caprine herpesvirus 1 (CpHV-1), a member of the Herpesvirales order, Herpesviridae family, Alphaherpesvirinae subfamily, and Simplexvirus genus, is classically associated with two distinct clinical syndromes. In kids, CpHV-1 induces severe systemic disease with high morbidity and mortality, whereas in adult goats, the infection leads to genital lesions such as vulvovaginitis or balanoposthitis, with abortions occurring mainly in the second half of gestation. CpHV-1 shares several biological characteristics with human herpesvirus 2, including molecular features, tropism for vaginal epithelium, genital lesion nature, and latency in the sacral ganglia. These features make CpHV-1infected goats a reliable animal model for studying human herpesvirus-induced genital disease, relevant for pathogenic research, as well as the development of new vaccines and antiviral agents. Recent full sequencing of the CpHV-1 genome has identified at least 10 genes encoding glycoproteins. Among these, glycoprotein D (gD) has been characterized but not yet exploited for immunogenic or diagnostic purposes. In this study, the structural features of CpHV-1 gD were predicted using in silico analysis. A truncated version of gD lacking the transmembrane domain (secreted glycoprotein D [Sec-gD]) was subsequently generated and expressed in mammalian cells, enabling its secretion into the culture medium. Despite the structural modifications, Sec-gD retained a conserved glycosylation pattern, as confirmed by N-glycosylation mutants generation and peptide-N-glycosidase F treatment. Furthermore, the antigenic properties of Sec-gD were preserved, as demonstrated by reverse serum neutralization assays. Notably, the culture supernatant containing Sec-gD was directly usable in diagnostic enzyme-linked immunosorbent assays, supporting its potential as a valuable tool for both diagnostic and immunization strategies. IMPORTANCE Caprine herpesvirus 1 (CpHV-1)-infected goats represent a large animal model for studying human herpesvirus-induced genital disease and could be utilized for pathogenic research, as well as for the development of new vaccines and antivi ral agents. CpHV-1 glycoprotein D can be efficiently produced and rescued from the supernatant of transfected mammalian cells, retaining its immunogenic properties, and could be used for immunogenic and diagnostic purposes. KEYWORDS experimental infection, ELISA, animal model for HSV2, caprine herpesvirus 1 C aprine herpesvirus 1 (CpHV-1) is a member of the Herpesvirales order, within the Herpesviridae family, Alphaherpesvirinae subfamily, and Simplexvirus genus (1). This virus is associated with two distinct clinical syndromes in goats: a fatal systemic illness in kids (2) and a genital disease in adults, which can manifest as balanoposthitis (3), vulvovaginitis (4), and abortion (5). Although only one complete genome sequence of CpHV-1 has been determined so far (https://www.ncbi.nlm.nih.gov/nuccore/NC_076509;
accession number: NC_076509), restriction site maps have been developed using double digestion and cross-hybridization of individual restriction fragments. The molecular weight of CpHV-1 DNA, estimated by summing the weights of fragments generated by various endonucleases, is approximately 137 kbp (6,7). From a pathogenic perspec tive, CpHV-1 infection typically begins at the respiratory or genital mucosa. The virus then disseminates systemically via a mononuclear cell-associated viremia, potentially leading to abortion in pregnant animals. CpHV-1 is shed through ocular, nasal, and genital secretions, with the genital tract considered the primary site for viral entry and persistence within herds (4). In kids, CpHV-1 causes a severe systemic disease marked by high morbidity and mortality, with ulcerative and necrotic lesions through out the gastrointestinal tract. In adult goats, the infection results in genital lesions such as vulvovaginitis or balanoposthitis. Abortions typically occur in the second half of gestation and can be experimentally induced through intranasal or intravenous inoculation of pregnant goats (8). Following intravaginal infection, the virus estab lishes latency in the sacral ganglia. Reactivation may occur under physiological stress, particularly during the breeding season, and may be influenced by hormonal changes during estrus. However, experimental reactivation is challenging and generally requires high doses of dexamethasone (9). Notably, CpHV-1 shares several biological characteris tics with human herpesvirus 2 (HSV-2) and bovine herpesvirus 1 (BoHV-1), including molecular features, tropism for vaginal epithelium, the nature of genital lesions, and latency in the sacral ganglia (10). CpHV-1 genomes have been recently fully sequenced (https://www.ncbi.nlm.nih.gov/nuccore/NC_076509; accession number: NC_076509) and exhibit the characteristic structure of a class D herpesvirus genome. This genome class is composed of a linear double-stranded DNA molecule, organized into a unique long segment and a unique short (US) segment. The US segment is flanked by internal and terminal inverted repeat sequences (IR and TR, respectively). The complete genome sequence of CpHV-1 has revealed the presence of at least 10 genes encoding glycopro teins. These glycoproteins have been identified and named based on their homology with those of herpes simplex virus type 1 (HSV-1) (11). Among these glycoproteins, glycoprotein D has been characterized (12); however, it has not been exploited as an immunogen or in terms of diagnostics. In this work, we successfully expressed and characterized glycoprotein D (gD) in a secreted form (Sec-gD) in the medium of a mammalian cell, and the medium containing Sec-gD could be directly employed for the development of a diagnostic enzyme-linked immunosorbent assay or immunization purposes.
## MATERIALS AND METHODS
## In silico analysis of CpHV-1 glycoprotein D
The amino acid sequence of CpHV-1 gD was retrieved from NCBI (https:// www.ncbi.nlm.nih.gov/protein/; accession number: AAZ66865.1) in FASTA format and used as input for the downstream analyses. PSIPRED 4.0 (13) and GOR IV web server (14) were employed for secondary structure prediction. Prediction of transmembrane helices and signal peptide was performed with Phobius web server (15), and visualiza tion of protein topology was carried out with Protter version. 1.0 (16). Prediction of glycosylation sites was performed with NetGlyc 1.0 (17). Template-based 3D modeling of CpHV-1 gD was performed using the SWISS-MODEL web server (18). A total of 172 templates matched the CpHV-1 gD sequence. Due to the unavailability of CpHV-1 gD templates, the prediction was based on crystal structures of the glycoprotein D from related herpesviruses (File S3). From these, the 10 most suitable templates were selected for further modeling. The resulting best-fitting model was selected based on sequence coverage, identity, Global Model Quality Estimate (GMQE), and the QMEANDisCo Global value. GalaxyRefine (19) was then employed to enhance the prediction accuracy. Both crude and refined models were evaluated with the SWISS-Model structure assessment tool. The final model was deposited in ModelArchive (Model ID: ma-3wwka; https:// modelarchive.org/).
## Cells
HEK 293T (human embryo kidney cells; ATCC: CRL-11268), MDBK (Mardin Darby Bovine Kidney; ATCC: CRL-6071), and BEK (bovine embryo kidney; Istituto Zooprofilattico Sperimentale, Brescia, Italy; BS CL-94) were cultured in complete Eagle's minimal essential medium (cEMEM). cEMEM was supplemented with 2 mM of L-glutamine, 1 mM of sodium pyruvate, 100 IU/mL of penicillin, 100 µg/mL of streptomycin, 0.25 µg/mL of amphotericin B, and 10% FBS. Cells were cultured in a humidified incubator at 37°C/5% CO 2 . All the supplements for the culture medium were purchased from Gibco (Segrate, MI, Italy).
## Construct generation
CpHV-1 gD secreted fragment (Sec-gD), lacking the transmembrane domain, was obtained by PCR amplification from pINT2-CMV-gDcpgD 106 (20) with NheI-Cap-gD sense (5′-CCCGCTAGCATGTGGGCCCTCGTGCTCGCAGCGCTAAGC-3′) and BamHI-HA-Cap-gD antisense (5′-GGGGGATCCTTAGGCGTAATCGGGCACGTCGTAGGGGTACGGCGCGGCGG GCGGGAGGGTAGGC-3′) primers. This pair of primers introduced an NheI restriction site at the 5′ terminal and an HA tag of the open reading frame (ORF), followed by a BamHI site at the 3′ terminal. The PCR amplification reaction was implemented in a final volume of 50 µL, containing 20 mM Tris-hydrochloride pH 8.8, 2 mM MgSO 4 , 10 mM KCl, 10 mM (NH 4 ) 2 SO 4 , 0.1 mg/mL bovine serum albumin (BSA), 0.1% (vol/vol) Triton X-100, 5% dimethyl sulfoxide (DMSO), 0.2 mM deoxynucleotide triphosphate, and 0.25 µM of each primer. One unit of Pfu recombinant DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) was used to amplify 100 ng of template DNA over 35 repeated cycles, including 1 min of denaturation at 94°C, 1 min of annealing at 60°C, and 1 min of elongation at 72°C. The resulting Sec-gD-HA amplicon was restriction digested with NheI/BamHI and subcloned in pEGFP-C1 (Clontech, San Jose, CA, USA), previously digested with the same enzymes, to generate the pCMV-Sec-gD construct.
## Transient transfection
HEK 293T cells were plated into 175 cm 2 flasks (5 × 10 6 cells/flask) and incubated at 37°C with 5% CO 2 . At sub-confluent density, the culture medium was removed, and the cells were transfected with pCMV-Sec-gD or pEGFP-C1 (as a mock control; Clontech, San Jose, CA, USA) using polyethylenimine (PEI) transfection reagent (Polysciences, Inc., Warrington, PA, USA). Briefly, DNA was mixed with PEI in a ratio of 1:2.5 (DNA:PEI) in 3.500 mL of serum-free Dulbecco's modified essential medium (DMEM) with high glucose (Euroclone, Pero, Italy) and incubated for 15 min at room temperature. Next, 4× volumes of serum-free medium were added, and the transfection solution was transferred onto the cells monolayer and left for 6 h at 37°C with 5% CO 2 , in a humidified incubator. The transfection mixture was then replaced with 21 mL of DMEM/F12 (Ham's F12 Nutrient Mixture; Euroclone Pero, Italy) (1:1) and incubated for 48 h. The cell supernatants, containing Sec-gD protein, were then harvested, clarified at 2,500 rpm at 4°C and stored at -80°C.
## Immunoblotting
Different amounts of Sec-gD protein supernatant samples were electrophoresed on 10% SDS-PAGE after total protein quantification with a BCA Protein Assay Kit (Pierce, Thermo Scientific, Waltham, MA, USA) and then transferred to PVDF membranes by electroblotting (Millipore, Merck, Rahway, NJ, USA). The membrane was subsequently blocked in 5% skim milk (Becton Dickinson, San Jose, CA, USA), incubated for 1 h at RT with a primary mouse monoclonal antibody anti-HA tag (G036, Abm Inc., New York, NY, USA), diluted at 1:15,000, and then probed with horseradish peroxidase-labeled anti-mouse immunoglobulin (A9044, Sigma-Aldrich [Merck], Tokyo, Japan), diluted at 1:15,000, and visualized using enhanced chemiluminescence (Clarity Max Western ECL substrate, Bio-Rad, Hercules, CA, USA).
## Peptide-N-glycosidase F digestion
Peptide-N-glycosidase F (PNGase F) (New England BioLabs, Ipswich, MA, USA) was employed following the manufacturer's instructions. Sec-gD protein-containing supernatants were collected from pCMV-Sec-gD transiently transfected HEK cells after 48 h of secretion. The samples were then treated with PNGase F, which cleaves between the innermost N-acetylglucosamine (GlcNAc) and asparagine residues from N-linked glycoproteins. PNGase F-treated samples were subsequently analyzed by Western immunoblotting as described above.
## Site-directed mutagenesis
N-glycosylation-mutated Sec-gD variants were generated by site-directed mutagenesis using PCR with primers carrying the desired point mutations that removed N-glycosyla tion sites (File S5). Briefly, PCR was carried out in a final volume of 25 µL containing 20 mM Tris-HCl (pH 8.8), 2 mM MgSO 4 , 10 mM KCl, 10 mM (NH 4 ) 2 SO 4 , 0.1 mg/mL BSA, 0.1% (vol/vol) Triton X-100, 10% (vol/vol) DMSO, 0.2 mM dNTPs, and 0.25 µM of each primer. One unit of recombinant Pfu DNA polymerase (Thermo Fisher Scientific, Waltham, MA, USA) was used to amplify 100 ng of template DNA over 35 cycles of 1 min denaturation at 94°C, 1 min annealing at 60°C, and 3 min extension at 72°C. PCR amplicons were separated by agarose gel electrophoresis and purified using the GeneJET Gel Extraction Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. The purified DNA was mixed at a 1:2 (vol/vol) ratio with Gibson GeneArt HiFi Master Mix (Thermo Fisher Scientific, Waltham, MA, USA), incubated for 40 min at 50°C, then cooled and transformed into ElectroMAX DH10B electrocompetent Escherichia coli cells. Positive colonies were identified by plasmid DNA mini-preparation and restriction enzyme digestion, followed by transient transfection into HEK 293T cells and Western immunoblotting as described above.
## Virus titration
BA-1 strain of CpHV-1, isolated from a latently infected goat, was cultured and titrated in MDBK cells (4). The stock viral titer of 10 6.25 50% tissue culture infectious doses (TCID 50 )/50 µL was stored at -80°C and used for the experiments. Briefly, the stock virus was serially 10-fold diluted and inoculated in quadruplicate onto MDBK cells in 96-well microtiter plates and incubated at 37°C in a 5% CO 2 atmosphere environment. The result was read after 3 days of incubation, and viral titers were expressed as logarithmic units calculated by the Reed-Muench method (21).
## Animals and experimental infection
The experimental protocol for goat infection was duly authorized (code 48E68, min aut. 869/15.11.2021) and conducted at the authorized University of Bari experimental animal facility (authorization no. 06/2023-UT).
Ten 5-year-old female goats without neutralizing antibodies to CpHV-1, as demonstra ted by seroneutralization assay (SNA), were used in this study. Prior to experimentation, the goats were held under controlled environmental conditions and examined daily for clinical evidence of disease. Nasal and vaginal swabs from each goat were collected at the time the goats arrived at the laboratory, and just before inoculation to ensure the absence of ongoing infection.
Goats were intravaginally infected with the BA-1 strain of CpHV-1, each receiving 3.0 mL (2 × 10 6 TCID 50 /mL) of the virus preparation and were daily examined for clinical evidence of infection. Observed clinical signs such as hyperemia, edema, lesions, and pain were scored as 0 (absent), 1 (mild), 2 (moderate), and 3 (severe), respectively.
Temperature elevations above normal (38.2°C-38.6°C) were graded as 1 (>0.5°C to 1°C), 2 (1.1°C-1.5°C), and 3 (>1.5°C). The score for each animal was monitored and reported daily. Blood samples were taken at day 0 and at 42 days post-infection to evaluate antibody response to CpHV-1 by means of SNA. Vaginal swabs were also obtained daily for 11 days post-infection to evaluate virus shedding.
## Serum neutralization test
The serum neutralization (SN) assay, yielding the highest degree of specificity of all the serological tests, was used to assess seroconversion following the procedure described elsewhere (22). In brief, sera were obtained from goats by venipuncture in EDTA-free vacutainer and heat inactivated at 56°C for 30 min. Subsequently, serial twofold dilutions of each serum from 1:2 up to 1:32 were mixed with 100 TCID 50 of the BA-1 strain of CpHV-1 in 96-well microtiter plates. The plates were held for 45 min at room tempera ture, and then, 20,000 MDBK cells in a volume of 0.05 mL of DMEM were added to each well, and the plates were then incubated for 3-5 days at 37°C in a 5% CO 2 humidified chamber environment. Readings were made when CPEs were complete in the virus control cultures, and the titer of each serum was expressed as the highest dilution neutralizing the virus in the well.
## Samples collection and ELISA procedure
Ninety-six-well microplates (Microlon High Binding, Greiner Bio-One, Kremsmünster, Austria) were coated overnight at 4°C with 50 ng/well of Sec-gD protein supernatant diluted in 0.1 M carbonate/bicarbonate buffer at pH 9.6. After blocking with 1% BSA (Sigma Aldrich by Merck, Rome, Italy), goat serum samples at different twofold dilutions (1/10, 1/20, 1/40, 1/80, 1/160, 1/320, 1/640, and 1/1,280) were incubated for 1 h at room temperature. Serum samples were diluted in DMEM/F12 without serum, collected from HEK 293T grown for 48 h. After three washing steps in phosphate buffer saline, 50 µL of donkey anti-goat IgG-HRP (Santa Cruz Biotechnology, Heidelberg, Germany) diluted 1:5,000 was added to each well, and the plate was incubated as above. Following the final washing step, the reaction was developed with 3,3′,5,5′-tetramethylbenzidine (Merck, Rome, Italy), stopped with 0.2 M H 2 SO 4 and read at 450 nm.
## Reverse serum neutralization test
Three heat-inactivated caprine sera previously confirmed to be positive in SNAs against CpHV-1 were selected. Twenty-five microliters of each CpHV-1 neutralizing serum sample were added to the first row of 96-well plates. An equal volume of cMEM (without FBS) was added to each well, and for each serum tested, serial twofold dilutions were performed. Next, 25 µL of medium containing Sec-gD protein was added to each well. Each serum was tested in the presence (+Sec-gD) or in the absence of Sec-gD (-Sec-gD). Positive and negative virus controls were also included. After 1 h of incubation at room temperature, 25 µL of virus suspension containing 100 TCID 50 (50% tissue cell infectious doses) of CpHV-1 strain BA-1 (23) was added to each well. After 1 h of incubation at 37°C, 50 µL of a BEK cell suspension (2 × 10 5 cells/mL) was added to each well, and the plates were incubated for 2 days at 37°C/5% CO 2 . Expression of viral infectivity and serum neutralizing activity through CPE was detected by microscopy and/or by classical crystal violet staining of the cell monolayer. The neutralization antibody titer was expressed as the reciprocal (log 2 ) of the final dilution of serum that completely inhibited viral infectivity.
## Receiver operating characteristic analysis
Receiver operating characteristic (ROC) analysis was carried out using SPSS for Windows (version 29.0.1.0, SPSS Inc., Chicago, USA), and the results were plotted using GraphPad Prism (version 8.0.1, GraphPad Software Inc., Boston, USA).
## RESULTS
## Generation and expression of CpHV-1 gD as a secreted peptide
Based on its amino acid sequence and predictions from Phobius/Protter (http:// phobius.sbc.su.se/), which are consistent with findings by Keuser et al. (12), the CpHV-1 gD ORF is 1,224 nucleotides long and encodes a 407-amino acid protein. This protein has a predicted molecular mass of 42.6 kDa and includes a signal peptide (amino acids 1-17), a hydrophobic transmembrane domain (amino acids 360-376) near the C-terminus, and a 31-amino acid cytoplasmic tail (Fig. 1A; File S1). The full-length gD can be expressed in eukaryotic systems as a membrane-bound protein (20). Based on this structure, removal of the transmembrane domain was predicted to yield a secreted form of gD (Fig. 1B; File S1). For achieving this, the extracellular domain of gD was amplified by PCR using primers, with the antisense primer incorporating an HA tag sequence in-frame with the amino-terminal rest of the protein (File S1). The resulting ORF was cloned into an expression vector, generating pCMV-Sec-gD. This construct includes the CMV promoter, the gD ORF lacking the transmembrane domain fused to an HA tag, and the bovine growth hormone polyadenylation signal. Upon transfection into HEK 293T cells, Sec-gD was efficiently secreted into the culture supernatant (Fig. 2C). Although the predicted molecular weight of Sec-gD is 40 kDa, it appears between 55 and 60 kDa in Western blot analysis, likely due to post-translational modifications such as glycosylation, as previously reported (12).
## Sec-gD structure analysis
While structural data for gD proteins from other alpha herpesviruses are available, no such information currently exists for CpHV-1. To address this gap, the secondary structure of CpHV-1 gD was predicted using two computational tools. GOR IV (accessed on 28 May 2025) estimated the protein to consist of 61.43% random coil, 20.15% alpha-helix, and 18.43% extended strand. Similarly, PSIPRED 4.0 (accessed on 28 May 2025) predicted 71.79% random coil, 19.27% alpha-helix, and 8.94% extended strand (File S2). To model the 3D structure of CpHV-1 gD, template-based modeling was performed using the SWISS-MODEL web server (accessed on 28 May 2025). Among 172 matching templates, 10 were selected based on sequence coverage (0.53-0.68), identity (43.32%-70.27%), and GMQE scores (0.39-0.51), all derived from previously resolved alpha herpesvirus gD structures (File S3). The final model was built using the bovine herpesvirus 1 gD structure (PDB ID: 6lsa.1.B) (24), which shares 65.11% sequence identity and 68% coverage with CpHV-1 gD. The resulting model spans residues 23 to 266, encompassing the majority of the extracellular domain (Fig. 2A). Model quality assessment yielded a GMQE of 0.51, a QMEANDisCo Global score of 0.81 ± 0.05, and a QMEAN Z-score of -1.94 (Fig. 2C), indicating that the predicted structure is reasonably accurate and falls within the range of experimentally determined proteins of similar size (Fig. 2D). Structural visualization and local quality estimation (Fig. 2E andF) showed high confidence in most surface-exposed regions, with lower reliability observed in loop regions around residues ~55-65 and ~200-210. To further refine the model, GalaxyRefine was employed. Both the initial and refined models were evaluated using the SWISS-MODEL structure assessment tool (accessed on 28 May 2025). Ramachandran plot analysis confirmed the structural validity, with the refined model showing improved accuracy: 98.35% of residues were in favored regions with 0.00% outliers, compared to 95.04% favored and 0.83% outliers in the crude model (File S4).
## Sec-gD exhibits conserved antigenic traits
The gD glycoprotein plays a crucial role in the attachment and entry of CpHV-1 into host cells (12). Additionally, CpHV-1 gD is essential for eliciting neutralizing antibodies and conferring protection against experimental CpHV-1 infection in goats (20,22). Protein glycosylation can significantly influence antibody function, particularly affecting antigen recognition and binding affinity. This is especially relevant for conformational antibodies, which depend on the three-dimensional structure of the antigen and are therefore sensitive to glycosylation-induced changes (25). The gD is a glycosylated protein (12), including its secreted form (Sec-gD), which contains two predicted N-glycosylation sites at amino acid positions 40 and 101 (Fig. 3A andB). These sites were experimentally confirmed through PNGase treatment and Western blot analysis (Fig. 3C). Replacing asparagine (N; red) with glutamine (Q; red) at the predicted glycosylation sites (NYTE and NATV) (Fig. S5) caused a progressive shift in the protein's molecular size, as observable on the protein band front electrophoretic mobility, relative to the unmutated form (fully glycosylated larger size) and the PNGase-digested form (fully deglycosylated smaller size) (Fig. 3D), thereby supporting the accuracy of the prediction.
Since structural prediction and analysis defined a certain degree of authenticity respect to other alpha herpesvirus glycoprotein D, it was of interest to ensure that Sec-gD retained its antigenic properties despite modifications such as removal of the transmembrane domain and addition of a 9-amino acid tag. Therefore, a reverse serum neutralization assay was conducted (Fig. 3E). As shown in Fig. 3F, preincubation of neutralizing sera with Sec-gD abolished their neutralizing activity, allowing CpHV-1 to infect and destroy the cell monolayer.
## Sec-gD allows the development of an indirect ELISA for the detection of CpHV-1-infected goats
Sec-gD was used to coat 96-well ELISA plates, and ELISA assays were conducted using serial dilutions of individual serum samples. These included 10 sera from animals that tested negative by SN prior to experimental infection, and 10 sera collected from the same animals 42 days post-CpHV-1 infection, which tested SN-positive, were clinically symptomatic, and virologically shedding as monitored by clinical score from day 0 to day 11 post-infection (data not shown). The resulting dilution curves (Fig. 4A) and the corresponding average areas under the curves (Fig. 4B) allowed for clear discrimina tion between positive and negative sera. The serum dilution yielding the highest ratio between the average signal of positive and negative sera (P/N) was determined to be 1/40 (Fig. 4C). ROC analysis (Fig. 4D) confirmed that 1/40 could represent the optimal serum dilution. Indeed, as shown by the area under the curve (AUC) values reported in Table 1, both the 1/40 and 1/80 dilutions gave an AUC value of 1. However, the maximum Youden index was obtained with different cut-off points: 0.199 for 1/40 and 0.127 for 1/80 (File S6). Therefore, the dilution with the higher cut-off was chosen, since for a diagnostic purpose, a higher threshold is generally more robust against inaccurate readings at low absorbance levels. This helps to prevent false positives resulting from background noise or potential weak cross-reactivity.
## DISCUSSION
The gD of several alpha herpesviruses plays a key role in the viral entry process, which typically occurs in two stages: initial attachment, where viral attachment proteins interact with receptors on the host cell surface, followed by membrane fusion that enables viral penetration (26). Antibodies generated against gD can neutralize the virus, and the presence of these neutralizing antibodies in vaccinated animals and humans is considered a more reliable indicator of vaccine efficacy compared to cell-mediated immunity (27,28). Despite truncated gD being used in human herpesviruses vaccine and 3) containing neutralizing antibodies against CpHV-1 were tested at the dilutions of 1/10 in the presence of Sec-gD (+Sec-gD) and in the absence of Sec-gD (-Sec-gD). Virus control was established in the absence of sera and Sec-gD (virus control) and a cells control with cells without virus, sera, and Sec-gD (cell control). Crystal violet staining allows macroscopic (not shown) and microscopic evaluation of the integrity (violet wells) of the cell monolayer.
candidates with mixed results (29), no studies have yet explored the potential use of CpHV-1 gD or its secreted form as a viable antigen for diagnostics and vaccine development. In this study, Sec-gD was successfully obtained and characterized. The in silico-derived structure of Sec-gD displays features consistent with those of gD proteins from other alpha herpesviruses, including its glycosylation pattern. Biochemical analyses confirmed that Sec-gD is properly glycosylated. SDS-PAGE and Western blot analyses revealed that Sec-gD exhibits a higher molecular weight than predicted based on its amino acid sequence, a discrepancy that was resolved upon treatment with PNGase F and site-directed mutagenesis of the glycosylation sites, indicating the presence of N-linked glycans. Glycans often dominate the surface of viral glycoproteins, making the viral glycome a key factor in shaping the antigenicity and immunogenicity. At one end of the spectrum, glycans can form essential components of epitopes recog nized by neutralizing antibodies, thus playing a critical role in the design of effective immunogens. Conversely, the success of peptide-based and bacterially expressed protein vaccines demonstrates that viral glycosylation is not always essential. Never theless, native-like glycosylation can reflect proper protein folding and the presence of conformational epitopes. Moreover, strategic modifications beyond native glycan mimicry-such as altering glycosylation site occupancy or glycan processing-may enhance the immunogenicity and protective efficacy of vaccine antigens (30). Given Statistical analyses were performed using an unpaired two-tailed Student's t-test in GraphPad Prism (P < 0.0001). (C) Optimization of sera dilution, defined as the dilution yielding the highest ratio between the average signal of positive and negative sera (P/N). Differences were statistically significant for each dilution.
(D) ROC analysis to determine the cut-off value of the test.
that proper protein folding and glycosylation can indicate the presence of conforma tional epitopes involved in the induction and recognition of neutralizing antibodies, this was validated through a reverse neutralization assay (30). In this test, Sec-gD effectively inhibited the CpHV-1 neutralizing activity of sera from CpHV-1-infected animals, confirming its role in antibody recognition. For corroborating the antigenic properties of CpHV-1 Sec-gD expressed in mammalian cells, an indirect ELISA test was developed. The ELISA test demonstrated complete accuracy in distinguishing sera from infected and non-infected animals.
This study utilized a secreted form of the gD protein (Sec-gD), expressed in mam malian cells. The Sec-gD protein demonstrated excellent immunogenicity, highlighting its potential not only for diagnostic applications but also as a subunit vaccine capa ble of inducing strong neutralizing antibody responses. Its performance underscores the importance of developing effective immunogens for successful vaccine strategies. Previous research identified the CpHV-1 gD protein as a promising vaccine candidate due to its critical role in viral attachment and entry into host cells. Recent findings confirm that this glycoprotein elicits potent neutralizing antibody responses, reinforcing its value as a key target for vaccine development. Animal studies further support this, showing that gD-based vaccines can confer protective immunity. A recombinant BoHV-4-based vector vaccine expressing the full-length CpHV-1 gD has recently been developed and shown to provide robust protection against lethal CpHV-1 infection in goats. The Sec-gD platform represents an evolution of this approach, offering greater flexibility. It can be adapted into various formats, such as a ferritin-based self-assembling nanopar ticle displaying the Sec-gD head domain (Sec-gD-ferritin) or as nucleic acid vaccines (DNA or mRNA). These formats offer several advantages, including rapid development, low-cost manufacturing, scalability, stability at lower temperatures, and suitability for deployment in outbreak-prone regions. Notably, DNA or mRNA-based Sec-gD vaccines are expected to elicit high levels of IFN-γ from T cells and induce strong gD-specific IgG antibody responses. Moreover, Sec-gD can be engineered into different structural configurations-monomeric, multimeric, or chimeric subunits-and stabilized in either fusion or post-fusion forms, enhancing its versatility as a vaccine component.
In conclusion, the promising results and proof-of-concept for using Sec-gD presented in this study lay the groundwork for developing and testing additional derivatives as prototype diagnostic tools and vaccines. These could target other significant animal pathogens-and potentially even human pathogens-given that CpHV-1 serves as a valuable large animal model for studying varicella-zoster virus in humans.
## ACKNOWLEDGMENTS
This study and the APC were funded by "Finanziamento dell'Unione Europea-Next generation EU-PRIN-2022/PNRR, grant number 2022M2HP84. " The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
## References
1. Davison (2010) "Herpesvirus systematics" *Vet Microbiol*
2. Van Der Lugt, Randles (1993) "Systemic herpesvirus infection in neonatal goats" *J S Afr Vet Assoc*
3. Tarigan, Webb, Kirkland (1987) "Caprine herpesvirus from balanoposthitis" *Aust Vet J*
4. Tempesta, Pratelli, Corrente et al. (1999) "A preliminary study on the pathogenicity of a strain of caprine herpesvirus-1" *Comp Immunol Microbiol Infect Dis*
5. Keuser, Gogev, Schynts et al. (2002) "Demonstration of generalized infection with caprine herpesvirus 1 diagnosed in an aborted caprine fetus by PCR" *Vet Res Commun*
6. Engels, Loepfe, Wild et al. (1987) "The genome of caprine herpesvirus 1: genome structure and relatedness to bovine herpesvirus 1" *J Gen Virol*
7. Hao, Mao, Li et al. (2020) "Epidemiological investigation and genomic characterization of Caprine herpesvirus 1 from goats in China" *Infect Genet Evol*
8. Uzal, Woods, Stillian et al. (2004) "Abortion and ulcerative posthitis associated with caprine herpesvirus-1 infection in goats in California" *J Vet Diagn Invest*
9. Tempesta, Pratelli, Greco et al. (1999) "Detection of caprine herpesvirus 1 in sacral ganglia of latently infected goats by PCR" *J Clin Microbiol*
10. Camero, Crescenzo, Marinaro et al. (2010) "Cidofovir does not prevent caprine herpesvirus type-1 neural latency in goats" *Antivir Ther*
11. Schwyzer, Ackermann (1996) "Molecular virology of ruminant herpesviruses" *Vet Microbiol*
12. Véronique, Detry, Thiry et al. (2006) "Characterization of caprine herpesvi rus 1 glycoprotein D gene and its translation product" *Virus Res*
13. Mcguffin, Bryson, Jones (2000) "The PSIPRED protein structure prediction server" *Bioinformatics*
14. Garnier, Gibrat, Robson (1996) "GOR method for predicting protein secondary structure from amino acid sequence" *Methods Enzymol*
15. Käll, Krogh, Sonnhammer (2007) "Advantages of combined transmembrane topology and signal peptide prediction--the Phobius web server" *Nucleic Acids Res*
16. Omasits, Ahrens, Müller et al. (2014) "Protter: interactive protein feature visualization and integration with experimental proteomic data" *Bioinformatics*
17. Gupta, Brunak (2002) "Prediction of glycosylation across the human proteome and the correlation to protein function" *Pac Symp Biocomput*
18. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-MODEL: homology modelling of protein structures and complexes" *Nucleic Acids Res*
19. Ko, Park, Heo et al. (2012) "GalaxyWEB server for protein structure prediction and refinement" *Nucleic Acids Res*
20. Donofrio, Franceschi, Lovero et al. (2013) "Clinical protection of goats against CpHV-1 induced genital disease with a BoHV-4-based vector expressing CpHV-1 gD" *PLoS One*
21. Maria, Camero, Bellacicco et al. (2007) "Cidofovir is effective against caprine herpesvirus 1 infection in goats" *Antiviral Res*
22. Tempesta, Pratelli, Normanno et al. (2000) "Experimental intravaginal infection of goats with caprine herpesvirus 1" *J Vet Med B Infect Dis Vet Public Health*
23. Buonavoglia, Tempesta, Cavalli et al. (1996) "Reactivation of caprine herpesvirus 1 in latently infected goats" *Comp Immunol Microbiol Infect Dis*
24. Yue, Chen, Yang et al. (2020) "Crystal structure of bovine herpesvirus 1 glycoprotein D bound to nectin-1 reveals the basis for its low-affinity binding to the receptor" *Sci Adv*
25. Tremain, Wallace, Lorentz et al. (2023) "Synthetically glycosylated antigens for the antigenspecific suppression of established immune responses" *Nat Biomed Eng*
26. Zhong, Zhang, Krummenacher et al. (2023) "Targeting herpesvirus entry complex and fusogen glycoproteins with prophylactic and therapeutic agents" *Trends Microbiol*
27. Dummer, Leivas Leite, Van Drunen Littel-Van Den Hurk (2014) "Bovine herpesvirus glycoprotein D: a review of its structural characteristics and applications in vaccinology" *Vet Res*
28. Georgopoulos, James (2024) "Immunogenetic profiles of 9 human herpes virus envelope glycoproteins" *Sci Rep*
30. Stanberry, Spruance, Cunningham et al. (2002) "Glycopro tein-D-adjuvant vaccine to prevent genital herpes" *N Engl J Med*
31. Newby, Allen, Crispin (2024) "Influence of glycosylation on the immunogenicity and antigenicity of viral immunogens" *Biotechnol Adv* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12673416&blobtype=pdf | # Comparison of T Cell Immune Responses Against SARS-CoV-2 Using QuantiFERON SARS-CoV-2 Assay Versus Peripheral Blood Mononuclear Cells Analysis by Flow Cytometry
Sourav Sen, Madhuri Thakar, Prashant Shivgunde, Shubhangi Bichare, Madhuri Suryawanshi, Nirmalkumar Rawandale, Yogesh Raut, Sheela Godbole, Madhuri Kanitkar
## Abstract
IntroductionCOVID-19 continues to have a global impact with the emergence of the latest variant of interest: JN.01, the commonest worldwide. While most studies on immune responses against SARS-CoV-2 are based on the evaluation of the humoral arm, data regarding cell-mediated immune responses are sparse and still emerging.
Materials and methodsA sample of 51 study participants was subjected to an interferon gamma release assay (QuantiFERON SARS-CoV-2). T cell (CD4+ or CD8+) responses to SARS-CoV-2 were characterized in a subset of 10 participants using multicolor flow cytometry.
ResultsThirty-seven participants had positive results for the anti-SARS-CoV-2 IgG antibody, and QuantiFERON SARS-CoV-2 assay antigen 1 and 2 positivity was observed in two participants only. In comparison, upon analyzing peripheral blood mononuclear cells using flow cytometry in a subset of 10 participants negative for the QuantiFERON SARS-CoV-2 assay, five had positive T cell immune responses.
ConclusionOur study demonstrates higher positivity for T cell response among 10 adult individuals when tested using flow cytometry for peripheral blood mononuclear cell analysis as compared to the interferon gamma release assay. Such data on immune responses to SARS-CoV-2 will be useful in developing predictive B-and T cellbased immune correlates and algorithms for personalized risk assessment and clinical response evaluation.
## Introduction
As per the World Health Organization weekly epidemiological update on COVID-19, as of 10 November 2024, global cumulative confirmed cases were 7,76,841,264, with 36%, 27%, 25%, 8%, 3%, and 1% contributions from Europe, the Western Pacific, the Americas, South-East Asia, the Eastern Mediterranean, and Africa, respectively [1]. However, it is important to realize that these trends do not represent the actual figures since wastewater surveillance data suggest that clinical detection of COVID-19 cases underestimates the real burden from two-to 19-fold [1]. By the end of October 2024, JN.01 was reported to represent 99.5% of all circulating variants of SARS-CoV-2 worldwide, with KP.3.1.1 being the most prevalent JN.01 descendant [1,2].
In 2022, Moss, based on a review of several studies, reinforced the critical role of T cell response for protection against SARS-CoV-2 infection [3]. It has been observed that such virus-specific T cell responses lead to viral clearance, aid in infection prevention without seroconversion, and facilitate the development of cellular memory and viral variant recognition. Similar T cell responses are observed after vaccination. While most of the available literature regarding immune correlates of protection is based on the measurement of spike-specific antibody response or neutralizing titers, data on the measurement and profiling of cellular immune response are limited. With a continuous rise in the prevalence of confirmed cases, a concurrent COVID-19 vaccination drive, and disparate host responses, it is necessary to assess individual risk and clinical response using a "personalized approach" when conducting studies involving prospective cohorts, followed by the development of predictive B-and T cell-based immune correlates and algorithms [3].
## Materials And Methods
## Study design and population
The design for the study was exploratory and observational. The study sample included 51 participants above 18 years of age from the urban area of Malegaon city in Nashik District of Maharashtra, India, who had previously been tested for SARS-CoV-2 neutralizing antibody status in May 2022, and were tested with an interferon gamma release assay (IGRA; QuantiFERON SARS-CoV-2). Of those who tested negative for the SARS-CoV-2 neutralizing antibody and were willing to undergo additional specimen collection, a subset was subjected to flow cytometry analysis of peripheral blood mononuclear cells (PBMCs). This subset was limited to 10 participants due to resource constraints.
## Specimen collection, transportation, and storage
Serum and whole blood specimens were collected in sterile and lithium heparin tubes (BD, USA) after obtaining informed consent in October 2022. Institutional ethical clearance was granted before initiation of the study (Maharashtra University of Health Sciences, Nashik, IEC No. MUHS/EC/30/2022).
## SARS-CoV-2 serology
IgG antibody levels were detected with the COVID-19 Neutralizing Antibody Micro-ELISA Kit (J. Mitra & Co., India). Specimens with >30% inhibition on testing were labeled as antibody positive.
QuantiFERON SARS-CoV-2 (research use only) assay (QIAGEN, Germany)
## Flow cytometry analysis
The T cell responses to SARS-CoV-2 were characterized in a subset of 10 vaccinated individuals with multicolor flow cytometry. The SARS-CoV-2-specific T cell response to SARS-CoV-2 peptide was assessed by intracellular cytokine staining. Briefly, the frozen PBMCs isolated from the participants were revived, rested overnight, and stimulated with SARS-CoV-2 peptides (spike PepTivator: pool of lyophilized peptides) in the presence of anti-CD107a (degranulation marker), followed by incubation with Brefeldin A and GolgiStop (monensin) at 37°C in 5% CO2. The next day, the cells were stained with a fluorescently labeled antibody cocktail (anti-CD3, anti-CD4, anti-CD8, anti-HLADR, anti-CD38, anti-CD45RA, and anti-CCR7). The cells were then washed, permeabilized, and stained with anti-IFN-γ, anti-TNF-α, and anti-IL-2 antibodies. Unstimulated PBMCs were used as a negative control, and those stimulated with staphylococcal enterotoxin B as a positive control. Cells were acquired on the FACSAria Fusion Flow Cytometer (BD Biosciences, United States) and analyzed using FlowJo software (Ashland, Oregon, United States). T cells from the stimulated PBMCs were differentiated based on CD4 and CD8 expression and assessed for activation status, memory status, and multiple cytokine (IFN-γ/IL-2/TNF-α) secretion. The Lymphocytes were identified by forward and side scatter. Lymphocytes were further drilled down to identify live cells. Live cells were then drilled down to gate CD3+ CD8(OR CD4+) T cells. CD8+ T cells were gated further to see expression of cytokines like CD107a, IL-2, TNFα and IFNγ and memory cells (Figure 1). The positivity criterion for a T cell response was a threshold of 1.0% after background subtraction.
## Statistical analyses
Frequency distributions and descriptive statistics, such as measures of central tendency and dispersion, have been used to present the results of the study. The results were calculated with the help of statistical software SPSS Statistics for Windows, version 17.0 (SPSS Inc., Chicago, Ill, USA).
## Results
The study population consisted of 51 participants (27 male and 24 female) with a median age of 40 years (range: 18-69). Only one participant reported a history of past COVID-19 infection. Additionally, 48 participants mentioned that they have been vaccinated against COVID-19 at least once.
Table 1 shows the details of anti-SARS-CoV-2 IgG antibody and QuantiFERON SARS-CoV-2 assay Ag1 and Ag2 levels for the study participants (N = 51). Of the total sample, 37 had positive results for the anti-SARS-CoV-2 IgG antibody (>30%). In terms of detectable IFN-gamma secretion to Ag1 and Ag2, one and three participants had positive results, respectively. Details of study participants subjected to PBMC analysis using flow cytometry (n = 10) can be found in Table 2. All 10 participants had received the Covishield (Serum Institute of India Pvt. Ltd., Pune, India) vaccine, except for MMP 5, for whom the type of vaccine received is unknown. The age range of the participants was 22-69 years, and there were five male and five female participants. None of the participants had received a booster dose of the vaccine at the time of the study. Regarding the second dose, dates were known for three participants; for four, the status was "vaccine taken/date not known"; and we could not confirm the vaccine status for the remaining three. and IL-2 (median: 17.65%, IQR: 8.42%-24.33%) were higher than those of CD4+ T cells secreting IFN-γ (median: 6.60%, IQR: 3.16%-17.46%) and IL-2 (median: 11.85%, IQR: 4.77%-21.40%). Regarding SARS-CoV-2-specific CD4+ and CD8+ memory T cell response, central memory was observed in 50% and 30%, respectively, while it was 10% and 30%, respectively, for effector memory response. . This longitudinal study reported a sustained T cell immune response over nine months, with no significant differences between the vaccines.
## Particulars
## Unique
Fernández-González et al. assessed the clinical performance of a specific quantitative SARS-CoV-2 IGRA assay (Euroimmun, Germany) in 239 participants, comprising 152 convalescent, 54 vaccinated, and 33 uninfected and unvaccinated persons [6]. The study compared the presence of SARS-CoV-2-specific IgG, neutralizing antibodies, and IFN-γ responses in individuals who had recovered from COVID-19 and those who had been vaccinated. The IGRA method was found to modestly increase the detection of immunity overall, with a more pronounced contribution noted in convalescent patients with mild disease, where it increased the yield of serology by 13%. The authors indicated that IGRA is a consistent approach for assessing anti-SARS-CoV-2 T cell responses after either natural infection or vaccination. However, it is important to remember that these values for quantifying T cell response are dependent on disease severity and time lapse since the primary infection and/or vaccination.
In comparison, Aiello et al. reported a lower number of T cell immune responders with the QuantiFERON SARS-CoV-2 assay when compared to measurement by a homemade IGRA-SPIKE test [7]. Such discrepant results may have resulted from differences in the nature of the spike Ag used and the variance in concentrations in the two assays.
Dourdouna et al. studied humoral immunity and CMI using the QuantiFERON assay [8]. The 41 study subjects included unvaccinated convalescent children and adults, and vaccinated uninfected or vaccinated convalescent adults. All unvaccinated and a significant number of vaccinated participants had negative QuantiFERON assay results, possibly due to either a lowering of immunity or low sensitivity of this assay [8].
Johnson et al. demonstrated 100% sensitivity and specificity of QuantiFERON SARS-CoV-2 assay for detecting SARS-CoV-2 T cell responses in acute infection (12-21 days post positive PCR) [9]. The sensitivity to this test dropped to 12.5% in those with a history of past infection (172-444 days post-positive test). Thus, the QuantiFERON SARS-CoV-2 assay had a lower sensitivity for assessing long-term T cell responses [9].
Tormo et al. evaluated the performance of an in-house-developed flow cytometry assay for intracellular cytokine staining (FC-ICS) and the QuantiFERON SARS-CoV-2 assay (QF) for detection and quantification of T cell responses after COVID-19 vaccination [10]. A significant discordance was observed between the two methods, with the discrepant results mostly being FC-ICS positive/QF negative specimens, thereby suggesting greater sensitivity of the FC-ICS assay when compared to the QF assay [10].
In our study, participants were predominantly uninfected adults with a history of COVID-19 vaccination. We studied a subset of 10 participants who tested negative for SARS-CoV-2-neutralizing antibodies to evaluate T cell response, using both the QuantiFERON SARS-CoV-2 assay and PBMC analysis by flow cytometry. We found discrepant results, wherein all participants had negative results for the former test, while five exhibited a positive response when tested by the latter. Of the five participants with a positive flow cytometry result, four had a history of vaccination, while for one, vaccination status was unknown. Our data highlights the comparatively lower sensitivity of the QuantiFERON SARS-CoV-2 assay for measuring CMI response.
Our study has certain limitations. These include a small sample size of 10 study participants assessed for CMI using IGRA and flow cytometry concomitantly. There was heterogeneity among the participants regarding those with a history of natural infection, those vaccinated, or both. A control group of participants with no history of natural infection or vaccination was not assessed.
## Conclusions
The role of T cell responses is critical in determining the outcome of SARS-CoV-2 infection, viral clearance, disease severity, and long-term immunity. It is, therefore, important to monitor these responses to gauge immune protection levels in various population groups and identify those with weak T cell responses. Evaluation of cell-mediated immune responses will be essential in guiding future vaccination designs and strategies, especially regarding the development of variant-resistant immune protection.
While IGRA and flow cytometry are the two key methodologies available for measuring T cell responses, the latter has the advantage of providing granularity for a detailed analysis of cellular subsets. In the present study, flow cytometry-based PBMC analysis was able to detect CMI, in contrast to the QuantiFERON SARS-CoV-2 assay, among adults with a history of COVID-19 vaccination. IGRAs offer the advantages of being scalable and suitable for low-resource settings, but are associated with low sensitivity, especially for the measurement of long T cell responses to SARS-CoV-2 infection and vaccination. In comparison, flow cytometry-based PBMC analysis is more sensitive and quantifies both CD4+ and CD8+ subsets independently with detailed immune profiling and is better suited for research purposes. However, larger studies are essential to elaborate on cell-mediated immune response, along with providing a concurrent evaluation of the humoral immune response, so as to provide a composite picture of immune responses among the vaccinated population.
## References
1. (2024) "COVID-19 epidemiological update"
2. (2024) "COVID-19 global risk assessment"
3. Moss (2022) "The T cell immune response against SARS-CoV-2" *Nat Immunol*
4. Jaganathan, Stieber, Rao (2021) "Preliminary evaluation of QuantiFERON SARS-CoV-2 and QIAreach anti-SARS-CoV-2 total test in recently vaccinated individuals" *Infect Dis Ther*
5. Stieber, Allen, Carpenter (2023) "Durability of COVID-19 vaccine induced T-cell mediated immune responses measured using the QuantiFERON SARS-CoV-2 assay" *Pulmonology*
6. Fernández-González, Agulló, Padilla (2022) "Clinical performance of a standardized severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) interferon-γ release assay for simple detection of T-cell responses after infection or vaccination" *Clin Infect Dis*
7. Aiello, Coppola, Vanini (2022) "Accuracy of QuantiFERON SARS-CoV-2 research use only assay and characterization of the CD4(+) and CD8(+) T cell-SARS-CoV-2 response: comparison with a homemade interferon-γ release assay" *Int J Infect Dis*
8. Dourdouna, Tatsi, Syriopoulou et al. (2023) "Evaluation of T cell responses with the QuantiFERON SARS-CoV-2 assay in individuals with 3 doses of BNT162b2 vaccine, SARS-CoV-2 infection, or hybrid immunity" *Diagn Microbiol Infect Dis*
9. Johnson, Phillips (2023) "Evaluation of QuantiFERON SARS-CoV-2 interferon-γ release assay following SARS-CoV-2 infection and vaccination" *Clin Exp Immunol*
10. Tormo, Giménez, Martínez-Navarro (2022) "Performance comparison of a flow cytometry immunoassay for intracellular cytokine staining and the QuantiFERON® SARS-CoV-2 test for detection and quantification of SARS-CoV-2-Spike-reactive-IFN-γ-producing T cells after COVID-19 vaccination" *Eur J Clin Microbiol Infect Dis*
11. Sen *Cureus* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724336&blobtype=pdf | # RNA splicing patterns contribute to burst size variation among HIV-1-infected Jurkat cell clones
Kamya Gopal, Cleo Burnett, Siarhei Kharytonchyk, Ann Emery, Edmond Atindaana, Ronald Swanstrom, Jeffrey Kidd, Alice Telesnitsky
## Abstract
Accurately quantifying virus release from HIV-1-infected cells is central to predicting infection outcomes and evaluating treatment strategies. Recent stud ies suggest that viral shedding can vary strikingly among infected cells. To identify predictors of virus release levels, a previously developed high-throughput molecular barcoding system was used to track the expression properties of individual infected clones within a polyclonal population. Consistent with previous reports, virus release spanned four orders of magnitude among clonal integrants. While reporter gene expression correlated poorly with virus release, intracellular viral RNA levels correlated well, and levels of unspliced HIV-1 RNA correlated most closely. Comparing clones with different virus release levels showed that they varied not only in total intracellular HIV-1 RNA levels but also in levels of viral RNA splicing. Remobilizing proviruses revealed that splicing differences were largely due to cell-intrinsic properties, although splicing differences were due to heritable features of the parent provirus for at least one clone. This clone displayed high levels of reporter gene expression from an HIV-1spliced RNA, but low levels of unspliced viral RNA and virus release. This over-splicing clone, which contained a non-synonymous substitution in rev, did not respond to treatment with the latency-reversing agent JQ1 but displayed increased virus release in response to treatment with pladienolide B, a splicing inhibitor. These findings demonstrate that splicing differences can contribute to virus production levels and suggest that future studies on proviral populations that include expanded clones may benefit from assessing clone-specific splicing properties. IMPORTANCE Many models of HIV-1 infection rely on the assumption that actively infected cells release similar amounts of virus, despite recent reports that suggest shedding differs drastically among infected cells. In this study, the expression pheno types of hundreds of integrant clones were analyzed to identify factors contributing to burst size variation. In agreement with previous reports, virus release spanned over four orders of magnitude within the infected pool. Both proviral expression levels and variation in splicing contributed to these burst size differences. While cell-intrinsic factors appeared to be the primary contributors to heterogeneous shedding patterns, viral point mutations were also observed and, in at least one case, contributed to particle release levels. Together, these findings demonstrate that expression variation among proviruses is both large and multifaceted and suggest that clone-specific differences in HIV-1 expression properties may contribute to unpredicted responses to treatment interventions.
H IV-1-infected cells are estimated to release on average between ~10 2 and 10 4 viral particles over their short lifetimes (1)(2)(3)(4). Virus release levels can influence clinical outcomes associated with exponential HIV-1 outgrowth, such as HIV-1 expansion during acute infection (5), viral rebound after antiretroviral therapy (ART) interruption (6), and viral spread following reactivation from latency (7). Many current therapeutic strategies focus on perturbing the reservoir of latently infected cells that persist during ART (8)(9)(10), and the ability to accurately measure cell HIV-1 expression and virus release levels would be valuable in determining the efficacy of treatment interventions (11)(12)(13)(14)(15).
Recent efforts to characterize the expression profiles of individual patient-derived proviruses have revealed that both HIV-1 transcription levels and viral bursts are highly variable across integrants in cells of the same type (16)(17)(18). However, despite known burst size variability and recognition that persistent provirus populations are often dominated by a limited number of cell clones, persistent cell population sizes are often calculated by measuring viremia and assuming averaged virus release levels (19)(20)(21). Estimates of how many cells are contributing to virus release are central to predicting infection progression and evaluating therapeutic strategies and would be improved by a better understanding of proviral expression properties and their regulation.
A significant hurdle in estimating virus release among infected cells is that they sustain highly diverse levels of HIV-1 expression due to both host-and virus-mediated variables (22)(23)(24). Intracellular markers, such as HIV-1 RNA levels (12,15,25,26), viral protein expression (27), and reporter gene expression (28)(29)(30), are typically used as measures for viral infection and replication, although consensus on an optimal predictor of virus release has yet to be achieved. Many virus expression models measure aggre gate viral activity and fail to capture the inherent variability present across different proviruses, further complicating the identification of predictors of virus release.
Our lab has previously developed a high-throughput sequencing (HTS) system that enables tracking of HIV-1 integrants within a polyclonal pool of infected cells through molecular barcodes introduced into proviral genomes (31,32). This experimen tal system is particularly well-suited to identifying correlates of virus release because multiple markers of infection, including viral RNAs and reporter gene expression, can be correlated to virus release for each of hundreds to thousands of individual proviruses within a polyclonal population. Additionally, individual integrant clones can be isolated from populations to identify mechanisms underlying provirus-specific differences in late replication steps.
In this study, HTS was used to investigate the spectrum of viral expression in a pool of Jurkat T cells infected with barcoded HIV-1 vectors. It was observed that virus release varied over 10,000-fold among integrant clones. Additional work showed that while multiple measures of reporter gene expression correlated poorly with virus release, intracellular levels of unspliced viral RNA were the strongest tested predictor of virus shedding.
## RESULTS
## Virus release levels span several orders of magnitude among integrant clones
To study differences in HIV-1 particle release, a pool of proviruses was established in the Jurkat T cell line using an HIV-1 NL4-3 strain-based library of vpr-, env-, and nefvectors called HIV-GPV-, to reflect the presence of gag and pol but the absence of vpr (31, 32) (Fig. 1A). Each vector was engineered to contain a unique 20 base barcode in U3 to aid in tracking the expression properties of individual proviral lineages via HTS. Cells were infected at a multiplicity of infection less than 0.0001 to ensure that infec ted cells contained single proviruses (Fig. 1B). Provirus-containing cells were selected independently of HIV-1 LTR activity using a constitutive puromycin-resistance cassette. An enhanced green fluorescent protein (eGFP) gene in the nef open reading frame (ORF) facilitated the identification of cells in which the HIV-1 LTR promoter was transcriptionally active. Culturing in selection medium yielded a polyclonal pool of infected cells, which was expanded and then split into two parallel cultures that were analyzed separately to assess reproducibility.
First, variation in virus shedding was examined by determining the barcode content of extracellular viral particles and infected cell subpools. Barcodes were amplified from cDNA templated by virion RNA and from cell chromosomal DNA, then subjected to HTS. The fractional abundance of each barcode in the pool of extracellular viral particles released from the unsorted pool was first determined, and then virus release was normalized to clone size as determined by barcode abundance in cellular chromosomal DNA. Per cell virus release was plotted against clone size for the 500 most abundant infected cell clones (Fig. 1C). Consistent with previous reports (16,32), virus shedding ranged over 4 orders of magnitude among the 500 analyzed clones (Fig. 1C). Later in this study, six individual clones were isolated for study from the infected pool via limiting dilution. The colorized data points shown in the graphs displaying population-level observations correspond to these isolated clones, which will be discussed below.
## Bimodal expression profiles and reporter gene expression levels correlate weakly with virus release
Having established that virus release levels differed 10,000-fold among proviruses in the infected pool, several measures of viral gene expression were then assessed to determine which best correlated with release levels. One expression parameter for at least some HIV-1 integrant clones is bimodal expression, which manifests as silent proviruses in a subset of daughter cells despite retention of LTR activity in sibling daughter cells (31)(32)(33). Indeed, phenotypic mixtures of both LTR-active and LTR-inactive cells have been observed within individual proviral clones both in infected patients and in experimen tal systems (31)(32)(33). Earlier studies on bimodal expression patterns revealed that the fractions of LTR-active and LTR-inactive cells in many integrant clones remain stable over at least 2 weeks (31).
Here, the relationship between LTR-active proportion and virus release was exam ined. As noted above, the initial polyclonal pool of integrant clones was split into two prior to sorting, and these were cultured in parallel. Cells from each of these parallel cultures were sorted into eGFP + (LTR-active) and eGFP -(LTR-inactive) fractions using fluorescence-activated cell sorting. Barcodes were then amplified from the unsorted cells as well as from the eGFP + and eGFP -cell fractions. Bimodal expression was determined by calculating the fraction of cells with a given barcode sorted into the eGFP + pool and normalizing to the total abundance of the barcode within both sorted pools. Although median burst size for <5% eGFP + clones (28 ng p24/mL) was almost 100-fold less than that for >95% eGFP + clones (2.4 µg p24/mL), when comparing bimodal expression patterns and per cell virus release, only a weak positive correlation (r 2 = 0.08) was observed, suggesting that clonal LTR-active proportions were a poor predictor of burst size (Fig. 2A). Comparing bimodal expression patterns between the two subpools for the 500 most abundant clones revealed a strong correlation, thus demonstrating reproduci bility (Fig. 2B).
Although bimodal expression was a poor predictor of clonal burst size, the possi bility remained that eGFP expression levels might correlate more closely with virus release. It has been previously observed that LTR promoter activity can vary based on integration site (34), epigenetic modifications (35,36), and transcriptional noise (37), among other factors. Therefore, a clone with a large fraction of LTR-active cells, each of which displayed low LTR activity, might express lower levels of eGFP than a clone with a smaller fraction of LTR-active cells, each displaying higher levels of LTR activity. To investigate the relationship between per cell eGFP expression levels and virus release, cells from the polyclonal pool were sorted based on mean fluorescence intensity (MFI) into low, intermediate, or high MFI subpools. To assign an approximate MFI value to each integrant clone, the fractions of its cells in the low, intermediate, and high MFI subpools were multiplied by the average MFI value for each subpool, as measured by flow cytometry. These calculations resulted in clonal MFI values, which are approximations and not absolute numbers. Plotting each clone's calculated MFI against virus release per cell again revealed a weak positive correlation (r 2 = 0.11) (Fig. 2C). Thus, per cell eGFP expression levels correlated more closely with virus release than bimodal expression patterns, but eGFP expression remained a poor predictor of virus release. the relationship between intracellular HIV-1 RNA levels and virus release per cell for the 500 most abundant clones (r 2 = 0.42, P < 0.0001). (E) A scatter plot comparing the relationship between unspliced HIV-1 RNA levels and virus release per cell for the 500 most abundant clones (r 2 = 0.62, P < 0.0001). Note that the colorized clones are the same as those in Fig. 1 and will be further addressed in Fig. 3. Intracellular HIV-1 RNA levels, and particularly levels of unspliced RNA, correlate strongly with virus release
Next, the relationship between HIV-1 RNA levels and virus release was investigated. It was reasoned that measuring total intracellular viral RNA levels, unlike reporter gene expression, would account for all HIV-1 messages contributing to viral replication and might correlate more closely with virus release. To address this, barcode abundance in total cell RNA from the unsorted pools was normalized to barcode abundance in unsorted cell DNA. When plotting HIV-1 transcript levels per cell against virus release per cell for each clone, a stronger positive correlation emerged (r 2 = 0.42), suggesting that HIV-1 RNA levels are a more accurate predictor of burst size than bimodal expression patterns or MFI (Fig. 2D).
Although total intracellular HIV-1 RNA correlated well with virus release, for a given amount of intracellular RNA (X-axis in Fig. 2D), virus release levels often ranged across two orders of magnitude (Y-axis in Fig. 2D), suggesting that additional parameters may contribute to burst size variation. One possible explanation was that RNA splicing product ratios were skewed. Transcription of integrated proviruses yields unspliced primary transcripts that can undergo alternative splicing and has been reported to Full-Length Text generate more than 100 partially or completely spliced isoforms (38)(39)(40). It has been widely observed that unspliced HIV-1 RNA comprises the majority of viral transcripts in an infected cell, and that HIV-1 is more tolerant to a deficit of spliced transcripts than to unspliced transcripts (39). While maintaining both unspliced and spliced viral transcripts is essential for productive infection, unspliced RNA serves as mRNA for the structural Gag and Gag-Pol polyproteins and also as genomic RNA for progeny viral particles (39,40). If splicing ratios were skewed, this might explain why reporter gene expression, which relied on one particular spliced RNA, did not map closely to virus release.
To address this, the relationship between unspliced HIV-1 RNA levels and virus release per cell was addressed. Unspliced viral RNA was selectively enriched from total cell RNA by affinity capture using biotinylated oligonucleotides complementary to a region of gag downstream of the 5′ major splice donor site, which is removed from all spliced HIV-1 RNAs (39,41). Barcode abundance was determined by HTS, and the amount of each barcode in cell unspliced RNA was plotted against its prevalence in extracellular virion RNA (Fig. 2E). The results demonstrated that unspliced HIV-1 RNA levels were a superior predictor of virus release compared to total viral RNA levels (r 2 = 0.62).
## Isolated clones also display expression variation
Individual integrant clones were then isolated via limiting dilution to validate popula tion-level observations. Six clones were selected for further study: five of which exhibited similar high eGFP + bimodal expression patterns but displayed dramatically different levels of virus release, and a sixth clone with fewer eGFP + cells. The limited number of clones studied here was biased toward clones that displayed high %eGFP + proportions because this allowed us to focus on clones with similar active cell populations but vastly different virus release levels. The lower %eGFP + clone was included as a comparator. The colored dots in Fig. 2 designate the clones studied in isolation in Fig. 3. Virus release per 10 6 cells (Fig. 3A), bimodal expression patterns (Fig. 3B), and eGFP expression (Fig. 3C andD) were characterized for these six isolated clones by quantifying RT activity and by flow cytometry. Some variations were observed between values determined for isolated clones and values for those same clones when assessed within the polyclonal population: notably, a twofold lower eGFP + proportion for Clone 4 in the population context than when studied individually. Although the cause of this discrepancy was not fully characterized, it is noteworthy that Clone 4 LTR-active cells displayed very low eGFP levels (Fig. 3D), and thus some of the differences in bimodal proportion calculations for this clone may have reflected gating during population flow cytometry. However, for the most part, the expression trends determined for each clone were similar when studied within populations using HTS or when determined individually for the clones after isolation, further validating the approaches used in this study. Reflective of findings in the polyclonal population, the isolated clones had viral burst sizes that spanned approximately 3 orders of magnitude, with Clone 3 shedding the least virus and Clone 5 shedding the most (Fig. 3A). Notably, despite their large differences in virus release, Clones 1-5 displayed strikingly similar bimodal expression patterns, with >85% of daughter cells from each isolated clone exhibiting LTR activity (Fig. 3B).
The possibility that eGFP expression levels might correlate more closely with virus release was examined next using flow cytometry. Measuring eGFP expression by MFI revealed highly variable amounts of reporter gene expression among the clones (Fig. 3C andD). The MFI values from five of the six clones corresponded well with virus release levels, with one notable exception, Clone 3, which displayed very high MFI but very low virus release.
## Isolated clones displayed concordant differences in splicing levels and virus release
Next, relationships between cellular RNA levels and burst size were examined for the isolated clones. RNase protection assays (RPAs) were performed with total cellular RNA from each of the six clones (Fig. 4A andB) using a probe that spanned the major 5′ splice site and thus yielded distinct protected fragments for unspliced and spliced viral RNAs. Quantification of the levels of unspliced HIV-1 RNA per total viral transcripts (%unspliced HIV-1 RNA) revealed that Clone 3, the clone with the lowest virus release among those studied, had the lowest %unspliced HIV-1 RNA. Clones 2 and 5 had the highest %unspliced HIV-1 RNA and the highest virus release levels. The HIV-1 splice product trends revealed by RPA were consistent with those determined by an alternate approach to determining splicing levels, in which total cell RNA was subjected to an HTS-based assay (Fig. 4B and see Materials and Methods). Data from both RPA and HTS experiments were consistent with earlier reports that "over-splicing" negatively impacts virus production (42,43).
Both Clone 4 and Clone 6 released relatively little virus despite having similar %unspliced HIV-1 RNA values to Clones 1, 3, and 5, which released several-fold more viral particles than Clones 4 and 6. To determine whether intracellular HIV-1 transcript levels could account for the discrepancy between %unspliced HIV-1 RNA and virus release, a quantitative PCR (qPCR) assay was performed. Relative RNA levels of GAPDH, a cellular housekeeping gene, and total HIV-1 RNA were determined using a probe set targeting GAPDH and the HIV-1 U5 region (Fig. 4C). The results indicated that Clones 4 and 6 expressed approximately 10-and 50-fold less HIV-1 RNA, respectively, than Clones 1, 3, and 5 when normalized to GAPDH, thereby reconciling the discordance between their virus release and %unspliced RNA levels. The relative amounts of specific splice classes were then determined. Using a previously described HTS-based assay on cell RNA isolated from the six clones, clone-specific differences were observed in amounts of intracellular RNAs containing splice junctions associated with tat, rev, env, and nef expression (Fig. 4D).
As demonstrated in this and earlier reports, over-splicing is deleterious to viral replication (42,43). It remained unclear, however, whether suppressing over-splicing would reverse replication defects. To determine whether the over-splicing Clone 3 was otherwise competent to release virus, clones were treated with pladienolide B (44) (Plad B), which decreases splicing by inhibiting the U2 snRNP. In parallel, the clones were treated with the latency-reversing agent (LRA) JQ1 (45), a BET-family inhibitor (Fig. 4E). Virus release increased for most clones upon treatment with JQ1 but not Plad B, particularly for Clones 4 and 6, which expressed lower levels of HIV-1 RNA than all other clones. In contrast, Clone 3 did not respond to LRA treatment but exhibited an almost 250-fold increase in virus release following treatment with the splicing inhibitor Plad B.
## Cell-intrinsic differences predominate as determinants of interclonal variation in burst size and splicing, but heritable differences among provi ruses also play a role
Having measured viral gene expression and particle release for isolated integrant clones, determinants of splicing differences among these integrants were then examined. Several factors known to contribute to differences in viral transcription, such as integration site, vary from one infected cell to another. It follows then that remobilizing virus from each integrant and infecting a fresh pool of cells would result in the averaging out of pronounced particle release and splicing phenotypes if these were characteris tic of the integration sites of the parent proviruses or other cell-intrinsic differences, whereas differences would persist in progeny if phenotypes reflected viral mutations. Thus, to assess whether burst size determinants mapped to within proviral sequen ces or if cell-specific factors contributed to the distinct properties of infected clones, the integrated proviruses from each isolated clone were remobilized by transfecting integrant clones with a VSV-G expression plasmid, with Clone 3 additionally being treated with Plad B to increase virus release levels, and the properties of pooled progeny were studied.
After selection and expansion of polyclonal progeny provirus-containing cells, virus release was measured as previously (Fig. 5A). Interestingly, while virus release of the parent integrants spanned nearly three orders of magnitude (Fig. 3A), virus release of the newly infected cell pools differed by at most approximately 2-fold among progeny of Clones 1, 2, 4, 5, and 6. Cells infected with remobilized virus from Clone 3, however, had similarly low virus release levels compared to the parental clone.
Cells infected with remobilized virus were additionally subjected to eGFP expression analysis, which revealed changes in MFIs for all remobilization pools compared to the parent proviruses (Fig. 5B). Interestingly, MFI was higher in the remobilized Clone 3 pool compared to the MFIs of remobilized provirus from other pools. Additionally, when %unspliced RNA levels were determined for each remobilized clone, Clone 3 appeared to inherit the splicing patterns of its parent provirus and expressed approximately three times less unspliced HIV-1 RNA than the other remobilized clones (Fig. 5C andD).
Taken together, these data suggest that the variation in splicing and virus release observed above for Clones 4, 5, and 6 was not passed on through proviral sequen ces, while for Clones 1 and 2, the similarities in virus release between progeny pools and original clones do not clearly discriminate between hereditary and non-hereditary effects. On the other hand, splicing and virus release defects appeared to be heritable in the case of Clone 3, suggesting that proviral sequence differences may have contributed to its characteristic replication phenotypes.
## A rev mutation observed in the oversplicing Clone 3 provirus is associated with reduced intracellular unspliced RNA levels
Sequencing the proviruses in each of the parental clones revealed mutations had arisen in HIV-1 ORFs in all but one of the clones, presumably during the single cycle of reverse transcription involved in provirus genesis (Fig. 6A). The remobilization data above suggested that most of these mutations did not cause heritable changes in viral gene expression (Fig. 5A through D). However, because the progeny of Clone 3 retained the altered splice product ratios of the parent, any mutations in this provirus would be candidate modulators of splice product levels.
Sequencing the provirus in Clone 3 revealed that it contained a synonymous mutation in tat that was non-synonymous in the overlapping rev ORF, resulting in a leucine to proline substitution in Rev residue 18 (Fig. 6A). Leu-18 resides in an α helix that has previously been implicated in facilitating Rev multimerization (46). As such, a proline substitution could conceivably affect Rev function. It has been observed that defects in Rev can result in abnormal viral splicing product levels (47), which might explain the unusual splicing phenotype of Clone 3.
To test whether or not Clone 3 splice product defects could be attributed to the L18P substitution, intracellular spliced and unspliced product ratios for HIV-GPP, an NL4-3 strain-based vpr-, env-, and nefvector with intact rev, were compared to those of HIV-GPP rev L18P , a modified version of HIV-GPP that contained the Clone 3 rev mutation, using an RPA (Fig. 6B). HIV-GPP is a vector similar to HIV-GPV-that does not express Vif or eGFP (48). HIV-GPP is not barcoded and was used to simplify procedures for this experiment, which did not examine expression at the level of individual proviruses. Whereas the RPA results on extracts of cells transfected with HIV-GPP confirmed the presence of high levels of unspliced RNA, similar to those observed for most of the remobilized vectors above, extracts from the cells transfected with HIV-GPP rev L18P displayed the strong oversplicing phenotype characteristic of Clone 3. Co-transfecting cells with equimolar HIV-GPP rev L18P and pRev, a vector expressing intact wild-type Rev under the CMV promoter, led to an intermediate phenotype. Together, these results indicate that the presence of the L18P rev mutation was sufficient to explain the heritable defect in splice product ratios observed for Clone 3.
## DISCUSSION
Here, an HTS analysis pipeline was used to investigate correlates of virus release levels in a Jurkat T cell-based HIV-1 infection model, initially using a polyclonal proviral population and subsequently validating results using individual integrant clones. While reporter gene expression correlated poorly with virus release, intracellular unspliced HIV-1 RNA levels correlated well. Consistent with the presence of outliers in the polyclonal population, isolation of individual integrant clones from the polyclonal population revealed that clone-specific low levels of viral shedding occurred despite high reporter gene expression for one isolate. This phenotype was determined to be due to over-splicing, wherein this clone retained LTR activity but produced insufficient levels of unspliced HIV-1 RNA to support robust virus production. Inhibiting splicing increased virus release in the over-splicing clone, while treatment with an LRA that did not directly inhibit splicing had no effect. Wide variation in virus release among integrants has been described in previous reports by our and other groups (16,32). The present work demonstrated that proviruses can differ in the amounts of viral transcript classes produced and that these differences can contribute to heterogeneity in viral shedding among integrants. Modest splicing differences were observed among several clones (Fig. 4B). However, provirus-specific splicing and virus release levels of the isolated clones were largely lost following proviral remobilization (Fig. 5D), suggesting that cell-specific differences, and not proviral mutations, primarily contributed to these expression patterns. In contrast, the expression properties of the over-splicing clone were heritable, and sequencing analysis suggested that a non-synonymous point mutation in rev caused this unique phenotype. Engineering the implicated rev mutation into a proviral clone and comparing its splice product phenotype to that of the parental clone confirmed that this single substitution in Rev was sufficient to generate the observed over-splicing phenotype.
Interestingly, the Rev residue that was altered in this mutant, Leu18, is only about 70% conserved among natural subtype B isolates (49). Previous work comparing multimeri zation and functioning of wild-type Rev to that of Rev mutants has shown that L18T and L18Q produce dimeric instead of higher-order complexes with cognate RNAs and also reduce Gag expression levels (50)(51)(52). Although less disruptive than the proline substitution examined here, naturally arising Leu18 substitutions clearly modulate Rev function and are common enough that it has been suggested that they may confer a selective advantage, perhaps by delaying the onset of AIDS (50).
Note, however, that the oversplicing phenotype of this rev mutant does not imply that Rev functions in splicing. Many years ago, a report claimed that the viral protein Rev suppresses HIV-1 transcript splicing by inhibiting spliceosome formation (53), which at that time appeared consistent with an earlier report that demonstrated that mutations in rev can result in increased levels of fully spliced viral RNAs (47). However, subsequent work on the functioning of Rev in unspliced RNA nuclear export suggested that the low levels of unspliced RNA observed in the absence of Rev reflect the degradation of RNA that is retained in the nucleus (54). Although some subsequent work has described possible interactions of Rev with the host splicing machinery, the prevailing model for the role of Rev in modulating RNA isoform ratios remains the promotion of unspliced HIV-1 RNA stability as a result of its nuclear export (55).
The work here does not resolve whether provirus sequence-independent differences in HIV-1 gene expression were due to integration site features or to genetic or epigenetic differences among cell clones. Jurkat T cells, an immortalized cancer line, were used in this study and in many others to investigate facets of HIV-1 infection (23,(56)(57)(58). Although Jurkat cells are a valuable model system, an important consideration is that HIV-1 may behave differently in primary cells and in people living with HIV-1 (PLWH). Notably, Jurkat cells are hypotetraploid (59,60). Although the library used here was established in Jurkat cells that underwent only limited passage after their purchase from ATCC, these cells are highly unstable, and thus it is possible that the host genomes of integrant clones may differ from one another in functionally significant ways. However, results generated using infected cells isolated from PLWH suggest that these cells also support a wide range of unspliced and spliced viral RNA levels (17), which indicates that differential splicing of HIV-1 transcripts may be meaningful in a physiologically relevant context as well.
This report presents several observations that may be relevant to the interpretation of previous studies. Constructs in which reporter genes are expressed from spliced transcripts are commonly used as a read-out of HIV-1 expression (22,30,32,61). We have demonstrated that in Jurkat cells, expression of a reporter gene from an HIV-1 spliced transcript and virus release levels correlate poorly for many integrants. Additionally, this work provided evidence that cell-intrinsic factors and mutations accrued over the course of even a single round of reverse transcription may contribute to widely observed differences in the response of infected cells to LRA treatment (62,63).
The work here advances understanding of the complex effects of HIV-1 splicing and its modulation. As has been demonstrated for 5′ splice site-defective proviruses in certain PLWH with non-suppressible viremia (64), proviruses with some types of splicing abnormalities may not be cleared by immune effectors and may be a source of ongoing viremia despite defects in virus replication. In contrast to the reduced splic ing observed in those mutants, the current study demonstrates that some proviruses can display an over-splicing phenotype and suggests that such proviruses may not respond to traditional LRA treatment. Various reports have demonstrated that inhibit ing the production of spliced transcripts significantly decreases levels of extracellular virus and that targeting splicing factors may serve as a therapeutic approach for HIV-1 infection (65)(66)(67). Although several reports show that over-splicing of the viral genome is detrimental to viral replication (39), promoting over-splicing has been less widely studied as an HIV-1 treatment. Intriguingly, while Plad B has previously been used as a latencypromoting agent (67), it may serve as a stimulator of virus release in circumstances wherein a provirus undergoes over-splicing, as seen in this study.
In summary, work here tracked facets of HIV-1 replication via virus barcodes and confirmed earlier findings that virus release levels per cell can vary by integrant over several orders of magnitude. Predictably, some component of HIV-infected Jurkat T cell burst size differences appeared due to clone-specific proviral transcription levels, since the low intracellular transcript levels of some clones were boosted by a transcriptionenhancing LRA, and transcription levels were restored in remobilized progeny provirus pools. However, some of the differences in virus release, both in a polyclonal infected pool and in isolated clones, were found to correlate with differences in intracellular unspliced HIV-1 transcript levels. For some clones, their splicing differences were not maintained when proviruses were remobilized, while for another-most notably, a rev point mutant-aberrant phenotypes were maintained upon virus passage. Thus, both cell-intrinsic factors and mutations in proviral genomes contributed to differential splicing among infected Jurkat cell clones. Although the current study describes proviruses only in transformed tissue culture cells, cells comprising the latent reservoir in PLWH are also diverse, and some component of their differences may include parameters that could affect the different classes of HIV-1 spliced RNAs. For example, there is some evidence for the presence of a virus with an attenuated Rev in a subset of long-surviving untreated PLWH (68). Speculatively, such proviruses or other splicing-defective subsets of the latent reservoir may contribute to continued inflammation by producing spliced transcript-encoded proteins, while not releasing virus or responding to LRA treatment.
## MATERIALS AND METHODS
## Cells and plasmids
Human embryonic kidney (HEK) 293T cells and Jurkat T cells were obtained from ATCC and grown at 37°C and 5% CO 2 . 293T cells were cultured in Dulbecco's Modified Eagle Medium supplemented with 10% fetal bovine serum (FBS), 50 µg/mL gentamicin, and 0.33 µg/mL amphotericin B, whereas Jurkat cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 Medium containing the same. HIV-GPP rev L18P was generated for this study. This construct was generated by subcloning an amplicon derived from the chromosomal DNA of Clone 3 containing the rev substitution into HIV-GPP.
Published plasmids used in this study include the following: HIV-GPV -(31), a vector derived from HIV-GKO (69) with functional gag-pol, tat, rev, vif, and vpu, an eGFP gene in the nef ORF, and a puromycin-resistance gene under the EF1a promoter; HIV-GPP, a vpr-, env-, and nefvector derived from HIV-1 NL4-3 (70); pRev, a rev expression plasmid (71); and pHEF-VSVG (72), which expresses vesicular stomatitis virus G protein.
## Generation of barcoded libraries and infected polyclonal pools, and transfec tion of 293T cells
Barcoded libraries were constructed as previously described (32). Briefly, HIV-GPV - was digested with ClaI and MluI, and the resulting ~11.4 kb fragment was gel-puri fied. Barcoded inserts were generated via PCR amplification of the U3 region of HIV-GKO, the parent of HIV-GPV -, using the following primers: 5′-GACAAGATATCCTTGATCT GNNNNNNNNNNNNNNNNNNNNGCCATCGATGTGGATCTACCACACACAAGGC-3′ (forward) and 5′-CGGTGCCTGATTAATTAAACGCGTGCTCGAGACCTGGAAAAAC-3′ (reverse), and the resulting ~300 bp PCR product was gel purified. The vector library was generated via Gibson assembly of the HIV-GPV -vector backbone and the PCR-generated insert pool at a 1:5 molar ratio. The resulting barcoded library was pseudotyped with pHEF-VSVG and packaged in 293T cells.
Jurkat cells were incubated with barcoded virus-containing media and 0.5 µg/mL polybrene at 37°C and 5% CO 2 for 5 hours. The infection mixture was removed, and cells were plated in fresh media as previously described (32). At 48 hours post-infection, infected cells were selected using 0.5 µg/mL puromycin for 4 days, then allowed to expand in selection-free media, yielding the polyclonal population of integrants used in this study. Where indicated, single clones were isolated from the polyclonal pool via limiting dilution.
To infect cells with remobilized provirus from single integrant clones, cells from each clone were pseudotyped with pHEF-VSVG using the Neon Electroporation System, pelleted, and resuspended in fresh culture media. The over-splicing clone was cultured in media supplemented with 200 nM Plad B to stimulate higher virus release. At 24 hours after media replacement, cells were pelleted, and virus-containing supernatants were passed through 0.22 micron filters. Virus-containing media were then used to infect Jurkat T cells as described above. For the experiments in Fig. 6B, 293T cells were transfected or co-transfected as previously described (41), with 5 µg HIV-GPP, 5 µg HIV-GPP rev L18P , or 5 µg of HIV-GPP rev L18P and 5 µg pRev.
## Flow cytometry and cell sorting
Jurkat cells were suspended in phosphate-buffered saline supplemented with 1% FBS prior to all flow cytometry experiments. Sorting was performed through the University of Michigan flow cytometry core on the Bigfoot Spectral Cell Sorter using the fluorescein isothiocyanate (FITC) channel. To analyze bimodal expression within the infected polyclonal population, cells from the infected pool were sorted into eGFP + and eGFP - subpopulations. Similarly, to analyze average MFIs, cell populations were sorted into low MFI, intermediate MFI, and high MFI subpopulations. eGFP expression levels (%eGFP + and MFI) were measured using the FITC channel on a BD LSR Fortessa. All flow cytometry data were analyzed using FlowJo, version 10.10.0.
## Nucleic acid isolation and preparation for sequencing
RNA was isolated from cells and viral particles using Trizol LS Reagent and Trizol Reagent, respectively. Genomic DNA was isolated from cells using the Dneasy Blood & Tissue Kit. All isolations were performed as per the manufacturer's protocol. Unspliced HIV-1 RNA was isolated from total cellular RNA fractions using a biotinylated oligo comple mentary to the pol region of the viral genome (5′-TTGGCCTTGCCCCTGCTTCTGTATTTCT GC/3BIO/-3′) as previously described (41). To prepare samples for HTS, barcodes were PCR-amplified directly from cellular DNA samples. For RNA, cDNA was generated using the SuperScript First-Strand Synthesis System, then barcodes were PCR-amplified from the cDNA. All PCRs to amplify barcodes were conducted using Phusion polymerase and the following primers with Illumina partial adapters: 5′-ACACTCTTTCCCTACACGAC GCTCTTCCGATCTGCCTGGCTAGAAGCACAAGA (forward) and 5′-GACTGGAGTTCAGACGT GTGCTCTTCCGATCTTGCCAATCAGGGAAGTAGCC (reverse). PCRs were purified using NEB Monarch PCR Purification Kits and submitted to Azenta for next-generation sequencing.
## HTS sequencing and analysis
Barcodes were analyzed using previously developed tools implemented in Python (31,32). After barcode clustering, the 500 most abundant barcodes present in the DNA of both replicate infected pools were used for further analysis.
To calculate virus release for each clone, barcode abundance in virion RNA was divided by barcode abundance in cellular DNA. RT assays were performed on media samples isolated from the infected unsorted pool, and a standard curve was used to determine p24 concentration in the unsorted pool (73). Relative per cell virus release was multiplied by the p24 concentration of the unsorted pool to calculate a virus release value for clones in the infected pool.
To calculate bimodal expression patterns (%eGFP + ) for each clone, the following formula was used:
, where E i is the %eGFP + of the barcode i, G i is the abundance of barcode i in the eGFP + sorted pool, P is the fraction of cells sorted into the eGFP + pool, W i is the abundance of barcode i in the eGFP -sorted pool, and Q is the fraction of cells sorted into the eGFP -pool. To calculate clonal average MFIs, the following formula was used:
, where F i is the average MFI of barcode i, L i is abundance of barcode i in the low MFI sorted subpopulation, A is the fraction of cells sorted into the low MFI subpopulation, X is the average MFI of cells in the low MFI sorted subpopulation as determined by flow cytometry after sorting, M i is the abundance of barcode i in the intermediate MFI sorted subpopulation, B is the fraction of cells sorted into the intermediate MFI subpopulation, Y is the average MFI of cells in the intermediate MFI sorted subpopulation as determined by flow cytometry after sorting, H i is the abundance of barcode i in the high MFI sorted subpopulation, C is the fraction of cells sorted into the high MFI subpopulation, and Z is the average MFI of cells in the high MFI subpopulation as determined by flow cytometry after sorting. To calculate the per cell total HIV-1 RNA for each clone, a barcode's abundance in total cell RNA is divided by its abundance in the cell DNA fraction of the unsorted pool. To calculate per cell unspliced HIV-1 RNA, a barcode's abundance in unspliced HIV-1 RNA was divided by its abundance in cellular DNA.
For the six isolated clones, splicing levels and final splice acceptors were quantified using HTS as previously described (74). The Illumina MiSeq 300 was used to perform sequencing, and the bcl2fastq pipeline (v.2.20.0) was used for data pre-processing. Splice site usage was analyzed using tools developed by the Swanstrom lab.
$$E i = {(G i × P) / [(G i × P) + (W i × Q)]} ×100$$
$$F i = [(L i × A × X) + (M i × B × Y) + (H i × C × Z)] / [(L i × A) + (M i × B) + (H i × C)]$$
## References
1. Haase, Henry, Zupancic et al. (1996) "Quantitative image analysis of HIV-1 infection in lymphoid tissue" *Science*
2. Hockett, Kilby, Derdeyn et al. (1999) "Constant mean viral copy number per infected cell in tissues regardless of high, low, or undetectable plasma HIV RNA" *J Exp Med*
3. Markowitz, Louie, Hurley et al. (2003) "A novel antiviral intervention results in more accurate assessment of human immunodeficiency virus type 1 replication dynamics and Tcell decay in vivo" *J Virol*
4. Chen, Mascio, Perelson et al. (2007) "Determina tion of virus burst size in vivo using a single-cycle SIV in rhesus macaques" *Proc Natl Acad Sci*
5. Ribeiro, Qin, Chavez et al. (2010) "Estimation of the initial viral growth rate and basic reproductive number during acute HIV-1 infection" *J Virol*
6. Davey Rt, Bhat, Yoder et al. (1999) "HIV-1 and T cell dynamics after interruption of highly active antiretroviral therapy (HAART) in patients with a history of sustained viral suppression" *Proc Natl Acad Sci*
7. Hataye, Casazza, Best et al. (2019) "Principles governing establishment versus collapse of HIV-1 cellular spread" *Cell Host Microbe*
8. (2025) *Full-Length Text Journal of Virology*
9. Archin, Liberty, Kashuba et al. (2012) "Administration of vorinostat disrupts HIV-1 latency in patients on antiretroviral therapy" *Nature*
10. Deeks (2012) "HIV: shock and kill" *Nature*
11. Darcis, Das (2018) "Tackling HIV persistence: pharmaco logical versus CRISPR-based shock strategies" *Viruses*
12. Pasternak, Jurriaans, Bakker et al. (2009) "Cellular levels of HIV unspliced RNA from patients on combination antiretroviral therapy with undetectable plasma viremia predict the therapy outcome" *PLoS One*
13. Elliott, Wightman, Solomon et al. (2014) "Activation of HIV transcription with short-course vorinostat in HIV-infected patients on suppressive antiretroviral therapy" *PLoS Pathog*
14. Wiegand, Spindler, Hong et al. (2017) "Single-cell analysis of HIV-1 transcriptional activity reveals expression of proviruses in expanded clones during ART" *Proc Natl Acad Sci*
15. Puertas, Bayón-Gil, Mc et al. (2021) "VIP-SPOT: an innovative assay to quantify the productive HIV-1 reservoir in the monitoring of cure strategies"
16. Zerbato, Khoury, Zhao et al. (2021) "Multiply spliced HIV RNA is a predictive measure of virus production ex vivo and in vivo following reversal of HIV latency" *EBioMedicine*
17. Bui, Mellors, Cillo (2016) "HIV-1 virion production from single inducible proviruses following T-cell activation ex vivo" *J Virol*
18. Sannier, Dubé, Dufour et al. (2021) "Combined single-cell transcriptional, translational, and genomic profiling reveals HIV-1 reservoir diversity" *Cell Rep*
19. Einkauf, Osborn, Gao et al. (2022) "Parallel analysis of transcription, integration, and sequence of single HIV-1 proviruses" *Cell*
20. Simonetti, Sobolewski, Fyne et al. (2016) "Clonally expanded CD4+ T cells can produce infectious HIV-1 in vivo" *Proc Natl Acad Sci*
21. Wang, Gurule, Brennan et al. (2018) "Expanded cellular clones carrying replicationcompetent HIV-1 persist, wax, and wane" *Proc Natl Acad Sci*
22. White, Kufera, Bachmann et al. (2022) "Measuring the latent reservoir for HIV-1: quantification bias in near full-length genome sequencing methods" *PLoS Pathog*
23. Demaster, Liu, Vanbelzen et al. (2016) "A subset of CD4/CD8 double-negative T cells expresses HIV proteins in patients on antiretroviral therapy" *J Virol*
24. Chen, Martinez, Zorita et al. (2017) "Position effects influence HIV latency reversal" *Nat Struct Mol Biol*
25. Valdebenito, Castellano, Ajasin et al. (2021) "Astrocytes are HIV reservoirs in the brain: A cell type with poor HIV infectivity and replication but efficient cell-to-cell viral transfer" *J Neurochem*
26. Pasternak, Lukashov, Berkhout (2013) "Cell-associated HIV RNA: a dynamic biomarker of viral persistence" *Retrovirology (Auckl)*
27. Gandhi, Mcmahon, Bosch et al. (2017) "Levels of HIV-1 persistence on antiretroviral therapy are not associated with markers of inflammation or activation" *PLoS Pathog*
28. Sowd, Aiken (2021) "Inositol phosphates promote HIV-1 assembly and maturation to facilitate viral spread in human CD4+ T cells" *PLoS Pathog*
29. Hashemi, Barreto, Bernhard et al. (2016) "HIV provirus stably reproduces parental latent and induced transcription phenotypes regardless of the chromosomal integration site" *J Virol*
30. Bruggemans, Vansant, Balakrishnan et al. (2023) "GS-9822, a preclinical LEDGIN candidate, displays a block-and-lock phenotype in cell culture" *Antimicrob Agents Chemother*
31. Cai, Gao, Zhao et al. (2021) "Infection with a newly designed dual fluorescent reporter HIV-1 effectively identifies latently infected CD4 + T cells" *Elife*
32. Read, Atindaana, Pyaram et al. (2019) "Stable integrant-specific differences in bimodal HIV-1 expression patterns revealed by high-throughput analysis" *PLoS Pathog*
33. Atindaana, Kissi-Twum, Emery et al. (2022) "Bimodal expression patterns, and not viral burst sizes, predict the effects of Vpr on HIV-1 proviral populations in Jurkat cells" *mBio*
34. Weinberger, Burnett, Toettcher et al. (2005) "Stochastic gene expression in a lentiviral positive-feedback loop: HIV-1 Tat fluctuations drive phenotypic diversity" *Cell*
35. Jordan, Defechereux, Verdin (2001) "The site of HIV-1 integration in the human genome determines basal transcriptional activity and response to Tat transactivation" *EMBO J*
37. Van Lint, Emiliani, Ott et al. (1996) "Transcriptional activation and chromatin remodeling of the HIV-1 promoter in response to histone acetylation" *EMBO J*
38. Lusic, Cereseto, Giacca (2003) "Regulation of HIV-1 gene expression by histone acetylation and factor recruitment at the LTR promoter" *EMBO J*
39. Damour, Slaninova, Radulescu et al. (1969) "Transcriptional stochasticity as a key aspect of HIV-1 latency"
40. Ocwieja, Sherrill-Mix, Mukherjee et al. (2012) "Dynamic regulation of HIV-1 mRNA populations analyzed by singlemolecule enrichment and long-read sequencing" *Nucleic Acids Res*
41. Emery, Swanstrom (2021) "HIV-1: to splice or not to splice, that is the question" *Viruses*
42. Sertznig, Hillebrand, Erkelenz et al. (2018) "Behind the scenes of HIV-1 replication: alternative splicing as the dependency factor on the quiet" *Virology (Auckl)*
43. Esquiaqui, Kharytonchyk, Drucker et al. (2020) "HIV-1 spliced RNAs display transcription start site bias" *RNA*
44. Madsen, Stoltzfus (2005) "An exonic splicing silencer downstream of the 3' splice site A2 is required for efficient human immunodeficiency virus type 1 replication" *J Virol*
45. (2025) *Full-Length Text Journal of Virology*
46. Madsen, Stoltzfus (2006) "A suboptimal 5' splice site downstream of HIV-1 splice site A1 is required for unspliced viral mRNA accumulation and efficient virus replication" *Retrovirology (Auckl)*
47. Kotake, Sagane, Owa et al. (2007) "Splicing factor SF3b as a target of the antitumor natural product pladienolide" *Nat Chem Biol*
48. Li, Guo, Wu et al. (2013) "The BET bromodomain inhibitor JQ1 activates HIV latency through antagonizing Brd4 inhibition of Tattransactivation" *Nucleic Acids Res*
49. Jain, Belasco (2001) "Structural model for the cooperative assembly of HIV-1 Rev multimers on the RRE as deduced from analysis of assembly-defective mutants" *Mol Cell*
50. Feinberg, Jarrett, Aldovini et al. (1986) "HTLV-III expression and production involve complex regulation at the levels of splicing and translation of viral" *Cell*
51. Lu, Heng, Garyu et al. (2011) "NMR detection of structures in the HIV-1 5′-leader RNA that regulate genome packaging" *Science*
52. Lebedev, Kim, Ozhmegova et al. (2024) "Rev protein diversity in HIV-1 group M clades" *Viruses*
53. Edgcomb, Aschrafi, Kompfner et al. (2008) "Protein structure and oligomerization are important for the formation of export-competent HIV-1 Rev-RRE complexes" *Protein Sci*
54. Daugherty, Booth, Jayaraman et al. (2010) "HIV Rev response element (RRE) directs assembly of the Rev homooligomer into discrete asymmetric complexes" *Proc Natl Acad Sci*
55. Edgcomb, Carmel, Naji et al. (2012) "DDX1 is an RNA-dependent ATPase involved in HIV-1 Rev function and virus replication" *J Mol Biol*
56. Kjems, Frankel, Sharp (1991) "Specific regulation of mRNA splicing in vitro by a peptide from HIV-1 Rev" *Cell*
57. Malim, Cullen (1993) "Rev and the fate of pre-mRNA in the nucleus: implications for the regulation of RNA processing in eukaryotes" *Mol Cell Biol*
58. Truman, Järvelin, Davis et al. (2020) *HIV Rev-isited: HIV Revisited. Open Biol*
59. Yeung, Houzet, Yedavalli et al. (2009) "A genome-wide short hairpin RNA screening of jurkat T-cells for human proteins contributing to productive HIV-1 replication" *J Biol Chem*
60. Rathore, Iketani, Wang et al. (2020) "CRISPR-based gene knockout screens reveal deubiquitinases involved in HIV-1 latency in two" *Jurkat cell models. Sci Rep*
61. Kumari, Dash, Kumari et al. (2022) "HIV-1 Nef hijacks both exocytic and endocytic pathways of host intracellular trafficking through differential regulation of Rab GTPases" *Biol Cell*
62. Snow, Judd (1987) "Heterogeneity of a human T-lymphoblastoid cell line" *Exp Cell Res*
63. Gioia, Siddique, Head et al. (2018) "A genome-wide survey of mutations in the Jurkat cell line" *BMC Genomics*
64. Bruggemans, Vansant, Van De Velde et al. (2023) "The HIV-2 OGH double reporter virus shows that HIV-2 is less cytotoxic and less sensitive to reactivation from latency than HIV-1 in cell culture" *J Virus Erad*
65. Pardons, Fromentin, Pagliuzza et al. (2019) "Latency-reversing agents induce differential responses in distinct memory CD4 T cell subsets in individuals on antiretroviral therapy" *Cell Rep*
66. Telwatte, Kim, Chen et al. (2020) "Mechanistic differences underlying HIV latency in the gut and blood contribute to differential responses to latencyreversing agents" *AIDS*
67. White, Wu, Yasin et al. (2023) "Clonally expanded HIV-1 proviruses with 5'-leader defects can give rise to nonsuppressible residual viremia" *J Clin Invest*
68. Daum, Engels, Mag et al. (1992) "Antisense oligodeoxynucleotide: inhibitor of splicing of mRNA of human immunodeficiency virus" *Intervirology*
69. Yeh, Jenike, Calvi et al. (2020) "Filgotinib suppresses HIV-1-driven gene transcription by inhibiting HIV-1 splicing and T cell activation" *J Clin Invest*
70. Janssens, Kim, Kim et al. (2024) "Mechanisms and efficacy of small molecule latency-promoting agents to inhibit HIV reactivation ex vivo" *JCI Insight*
71. Churchill, Chiavaroli, Wesselingh et al. (2007) "Persistence of attenuated HIV-1 rev alleles in an epidemiologically linked cohort of long-term survivors infected with nef-deleted virus" *Retrovirology (Auckl)*
72. Battivelli, Dahabieh, Abdel-Mohsen et al. (2018) "Distinct chromatin functional states correlate with HIV latency reactivation in infected primary CD4 + T cells"
73. Ding, Kharytonchyk, Kuo et al. "2021. 5′-Cap sequestra tion is an essential determinant of HIV-1 genome packaging" *Proc Natl Acad Sci*
74. Lewis, Williams, Rekosh et al. (1990) "Identification of a cis-acting element in human immunodeficiency virus type 2 (HIV-2) that is responsive to the HIV-1 rev and human T-cell leukemia virus types I and II rex proteins" *J Virol*
75. Chang, Urlacher, Iwakuma et al. (1999) "Efficacy and safety analyses of a recombinant human immunodeficiency virus type 1 derived vector system" *Gene Ther*
76. Vermeire, Naessens, Vanderstraeten (2012) "Quantification of reverse transcriptase activity by real-time PCR as a fast and accurate method for titration of HIV, lenti-and retroviral vectors" *PLoS One*
77. Emery, Zhou, Pollom et al. (2017) "Characterizing HIV-1 splicing by using next-generation sequencing" *J Virol*
78. Flynn, Telesnitsky (2006) "Two distinct Moloney murine leukemia virus RNAs produced from a single locus dimerize at random" *Virology (Auckl)*
79. Kharytonchyk, Monti, Smaldino et al. (2016) "Transcriptional start site heterogeneity modulates the structure and function of the HIV-1 genome" *Proc Natl Acad Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12706163&blobtype=pdf | # Latent reservoirs in memory T cell subsets are linked to poor immune recovery in people living with HIV
Selwyn Selva Kumar, John Demosthenes, Abi Manesh, Saravanabhavan Thangavel, Luke Hanna, Rajesh Kannangai, George Varghese, Prasun Datta, Sakthivel Govindaraj, Carmen Gasca-Capote
## Abstract
Highly active antiretroviral therapy (HAART) suppresses viral loads in 71% of people living with HIV globally, while failing to bring adequate immune reconstitution in nearly one-third of them. We hypothesize that the persistence of latent HIV reservoirs in specific memory T cell subsets contributes to impaired immune recovery. We conducted a case-control study to estimate differences in the HIV-1 proviral DNA across memory CD4-positive T cell subsets between participants with CD4 counts of over 500 cells/µL (immune responders or IRs) and those with counts of less than 350 cells/µL (immune nonresponders or INRs) with sustained viral suppression. Latent HIV reservoirs (LRs) were detected in at least one memory T cell subset in 48.33% of total participants. Latent reservoirs were more frequent among INRs than among IRs (65.38% versus 35.29%, P = 0.02), particularly in the effector memory T cell subset (34.6% in INRs versus 8.8% in IRs, P = 0.02). Thus, despite long-term viral suppression with HAART, the persistence of latent reservoirs in memory T cells is associated with poor CD4-positive T cell recovery. Emerging classes of antiretroviral agents that target latent viral pools may enhance immune restoration and bring us closer to finding an HIV cure.
## 1 Introduction
Though highly active antiretroviral therapy (HAART) has had remarkable success in achieving viral suppression in approximately 71% of people living with HIV (PLWH) globally, it fails to achieve adequate immune reconstitution in a significant proportion (nearly one-third) (1,2). Individuals on HAART with impaired immune recovery, termed immune non-responders (INRs), maintain low CD4-positive T cell counts despite the sustained suppression of viral load. The resultant subclinical immunodeficiency heightens the INRs' risk of HIV-related and HIV-unrelated complications, accelerates the progression of acquired immunodeficiency syndrome (AIDS), and increases mortality (3,4).
The underlying mechanism for poor recovery of CD4-positive T cells in INRs remains incompletely understood. Proposed explanations include persistent immune activation, immune exhaustion, virological factors, host genetic variability, and suboptimal thymic output (2). However, none of these factors fully explains the INR phenotype.
A defining feature of HIV pathogenesis is the establishment of latent reservoirs within resting memory CD4-positive T-cells, where the virus persists in a transcriptionally silent state. These reservoirs evade clearance by HAART and remain the principal barrier to curing HIV. Notably, latent HIV has been detected even after allogeneic hematopoietic stem cell transplantation from CCR5 wild-type donors, underscoring the resilience of these viral sanctuaries (5).
We hypothesize that the persistence of latent HIV reservoirs in specific memory CD4 -positive T cell subsets may be associated with impaired immune recovery in immune non-responders. Elucidating this link could provide critical insights into the pathogenesis of immune non-responsiveness and inform the development of targeted therapies aimed at both immune reconstitution and reservoir eradication.
## 2 Materials and methods
We conducted a case-control study to estimate differences in the HIV-1 proviral DNA across memory CD4-positive T cell subsets between IRs and INRs. INRs were defined as people living with HIV-1 who had been on HAART for at least two years, demonstrated sustained viral suppression, and had an absolute CD4positive T-cell count of less than 350 cells/µL for at least two years prior to recruitment. IRs were defined as individuals with CD4-positive T cell counts consistently greater than 500 cells/µL for at least two years prior to recruitment. All participants who met the inclusion criteria and visited the clinic during the study period were randomly enrolled in the study. Exclusion criteria included individuals who were below 18 years of age, had HIV-2 infection, had a documented episode of detectable viremia during the past two years, or had difficult venipuncture.
Ethical approval was obtained from the Institutional Review Board of Christian Medical College, Vellore (IRB Min. No. 13143). Sixty participants were recruited between March 2021 and August 2021 from the National AIDS Control Organization (NACO) ART Centre and the Infectious Diseases outpatient department of Christian Medical College and Hospital, Vellore. After obtaining informed written consent, 10 mL of peripheral blood was collected from each participant.
We collected data on epidemiological factors, co-morbidities, duration of HIV-1 infection, duration and type of HAART regimen, history of opportunistic infections, and metabolic complications. Peripheral blood mononuclear cells (PBMCs) were isolated from blood samples, and CD4-positive T cells were sorted based on established surface markers into central memory (TCM), transitional memory (TTM), and effector memory (TEM) subsets using flow cytometry.
## 2.1 PBMC separation
Peripheral blood mononuclear cells were isolated from leukapheresis samples using Ficoll density gradient centrifugation and stored at -80 °C for 2.5 years. For analysis, the samples were thawed at 37 °C in a water bath, washed, and assessed for viability using the trypan blue dye exclusion method. Only samples with a viability greater than 60% were included for further analysis.
## 2.2 Cell surface staining
To identify and sort CD4 + T-cell subsets, cells were stained with the fluorochrome-conjugated monoclonal antibodies listed in Table 1 (all from BD Biosciences).
For staining, 4 × 10 6 to 10 x 10 6 cells were incubated with the antibody cocktail in staining buffer (PBS + 2% FBS) for 30 min at 4 °C in the dark. After staining, cells were washed twice with staining buffer and resuspended in PBS containing 2% FBS for flow cytometric acquisition and sorting.
Prior to acquisition, cells were incubated with Fixable Viability Stain (APC-H7) according to the manufacturer's protocol to exclude non-viable cells. Live (APC-H7 -) cells were gated for further analysis and sorting.
## 2.3 Flow cytometric sorting using FACS Aria III
Cell sorting was performed on a BD FACS Aria III instrument. Compensation controls were prepared using single-stained beads or
## Marker Fluorochrome Clones
Fixable Viability Stain APC-H7 -
cells to adjust for spectral overlap. The gating strategy was as follows: Singlet cells were selected using FSC-A vs. FSC-H. Live cells were identified as APC-H7-negative. T cells were gated as CD3 + . CD4 + helper T cells were gated within the CD3 + population. Memory subsets were identified using CD45RO, CCR7, and CD27 expression: Central memory (CD45RO + CCR7 + CD27 + ), Transitional memory (CD45RO + CCR7 -CD27 + ), and Effector memory (CD45RO + CCR7 -CD27 -) (Figure 1) (6). The desired memory CD4 + T cell population was sorted into tubes containing complete RPMI-1640 medium (10% FBS).
$$CD3 Brilliant Violet 510 (BV510) UCHT1 CD4 APC-R700 SK3 CD45RO APC UCHL1 CCR7 PE-Cy7 2-L 1-A CD27 Brilliant Violet 421 (BV421) M-T271$$
## 2.4 Post-sorting quality control
A small aliquot of sorted cells was re-analyzed on the FACS Aria III to confirm post-sort purity, which consistently exceeded 95%.
## 2.5 HIV-1 pro-viral DNA quantification
A 292 bp fragment encompassing a segment of the HIV-1 LTR and gag region was amplified from the DNA of an HIV-1 infected individual using nested PCR with primers sourced from previous literature (7). HIV-1 amplicon was cloned into an expression vector (PROMEGA, 2800 MADISON, WI USA) and the standards for the in-house quantitative assay were prepared.
DNA was extracted from sorted memory T cell subsets using the QIAcube system (Qiagen, Hilden, Germany) with the QIAcube HT Kit, according to the manufacturer's instructions. The eluted DNA was used for quantification of HIV-1 pro-viral DNA. HIV-1 proviral DNA quantification was performed by amplifying a 158-bp fragment spanning portions of the long terminal repeat (LTR) and gag regions using primers F (ATCTCTAGCAGTGGCGCCCGA) and R (CCTTCTAGCCTCCGCTAGTCA). The amplified product was subsequently detected using the probe (ACGCAGGA CTCGGCTTGCTG) (7). The cycling conditions were as follows, 95 °C for 10 minutes, followed by 95 °C for 30sec, 57 °C for 30 seconds, for 50 cycles.
An in-house ERV-3 (endogenous retrovirus) quantitative real time PCR was carried out to know the copy number in memory T cells (8). This assay is based on TaqMan chemistry, and the primers and probe were taken from previously published literature (9). The thermal cycling conditions used for this assay were 95 °C for 15 min, 95 °C for 45 sec, and 60 °C for 75 sec for 50 cycles The ERV-3 quantitation was done for the samples and the copy number calculated for the HIV-1pro-viral based on the ERV-3 copies per 10000 cells (10).
## 2.6 Statistical analysis
Baseline characteristics and clinical variables were summarized using frequencies and proportions for categorical data, and means with standard deviations or medians with interquartile ranges (IQR) for continuous data. Comparisons between immune responders and non-responders were made using the independent t-test or Mann-Whitney U test for continuous variables, and Pearson's chi-square test or Fisher's exact test for categorical variables. A two-tailed p-value of less than 0.05 was considered statistically significant. The proviral DNA copies among the subsets of memory CD4-positive T-cells were log-transformed to normalize distributions and facilitate visualization. We used STATA ® version 16 (StataCorp. 2019. Stata Statistical Software: Release 16. College Station, TX: StataCorp LLC) for statistical analyses, and the figures were generated using GraphPad Prism.
## 3 Results
Among the 60 participants enrolled, 26 were classified as immunological non-responders and 34 as immunological responders. The majority were middle-aged (mean: 43.56 ± 10.22 years). The median duration of HIV infection was 10 years (IQR: 7-13), and the median duration on HAART was 7 years (IQR: 4-11). Zidovudine and tenofovir disoproxil fumarate (TDF) were the most frequently used drugs in baseline HAART regimens (43.34% and 41.67%, respectively). None of the participants had received integrase strand transfer inhibitors (INSTIs). Notably, an abacavir-based regimen was used exclusively in the INR group (30.77%) (Table 2).
Significant differences were observed between groups with respect to male gender, baseline CD4 count, and exposure to TDF-efavirenz (EFV)-based regimens. There were more male participants in the INR group than in the IR group (P = 0.021). The baseline CD4 counts and the CD4 counts during the study were significantly lower (P = 0.04 and P = <0.001) in the INR group (145.5 cells/µL with IQR of 105-261 and 299.5 cells/µL with IQR of 251-321, respectively) than in the IR group (248 cells/µL with IQR of 140-443 and 688 cells/µL with IQR of 574-778, respectively).
Exposure to TDF-EFV-based regimens was higher in the IR group than in the INR group (23% versus 13%, P = 0.003). One participant from the INR group died due to complications related to tuberculosis.
Latent HIV reservoirs (LRs) were detected in at least one memory T cell subset in 48.33% of participants. Immunological non-responders demonstrated a significantly higher frequency of detectable LRs compared to IRs (65.38% versus 35.29%, P = 0.02) (Table 3). While latent reservoirs were observed across all memory subsets, the most common sites were central memory and effector memory T-cells. A statistically significant difference between groups was noted in the TEM subset, 34.6% of which had latent LRs in INRs, while 8.8% of the TEM subset had latent LRs in IRs (P = 0.02) (Figure 2). No statistically significant difference in LR was observed when stratified by baseline CD4 count, duration of HIV infection, or duration of HAART.
In TCM cells, the median (IQR) proviral DNA copies per microliter were 1,318 (47-12,342) in INRs versus 32 in IRs (P = 0.198). For TTM cells, the medians were 12,262 (311-23,644) in INRs and 883 (392-145,738) in IRs (P = 1.000). In TEM cells, median values were 2,132 (64-4,774) in INRs and 323 324) in IRs (P = 0.866). Although these differences did not reach statistical significance, the overall trend pointed toward higher proviral DNA levels in INRs across all memory subsets.
Additionally, participants receiving zidovudine-and nevirapine-based regimens exhibited a higher frequency of detectable LRs in memory T-cell subsets when compared to those on TDF-based regimens (Table 2).
## 4 Discussion
Our study demonstrated that INRs harbored a higher burden of HIV-1 proviral DNA within circulating memory CD4-positive Tcell subsets than IRs did. This provides novel evidence that the persistence of latent reservoirs in memory T cells is associated with poor CD4-positive T cell recovery despite long-term viral suppression with HAART. While other immune cells, such as regulatory T cells (Tregs), natural killer cells, Th17 cells, and polymorphonuclear myeloid-derived suppressor cells, have been implicated in immunologic failure in PLWH, memory CD4+ T cells could be the primary reservoir of latent HIV (11). Recent studies have demonstrated that peripheral blood monocytes in INRs carry a greater burden of HIV-1 RNA and DNA compared with IRs (12). Tissue-resident memory T cells (TRMs) have been shown to serve as significant HIV reservoirs and express higher levels of HIV susceptibility markers, such as CCR5 (13). Additionally, albeit smaller, contributions come from circulating mononuclear macrophages and B cells within lymphoid structures (14).
Among the memory CD4-positive subsets, central memory T cells are known to harbor the highest levels of HIV-1 proviral DNA, followed by transitional memory T cells, effector memory T cells, stem memory T cells, and naïve memory T cells (15). In our study, proviral DNA was detected across TCM, TTM, and TEM subsets, with a higher frequency of latent reservoirs in the TEM subset among INRs than among IRs. This observation may reflect the differing distributions and trafficking patterns of memory T cell subsets between peripheral blood and tissue compartments (16). Circulating CD4-positive T cells represent only 2-2.5% of the total T cell pool in the body, and over 90% of proviruses in these cells are replication-defective (17). Therefore, our findings may underestimate the true burden of replication-competent latent reservoirs. A more accurate assessment would require a
Duration Of ART (in years) (median, IQR)
Delay in ART (in years) (median, IQR) quantitative viral outgrowth assay, which specifically measures inducible, replication-competent virus (18). Immune non-responders, despite having virologic suppression, face a threefold increased risk of both AIDS-defining and non-AIDS-defining complications (19,20). Although our study was underpowered to detect such statistical differences, herpes zoster, herpes labialis, and oral candidiasis occurred exclusively among INRs post HAART. Additionally, one INR developed diffuse large B-cell lymphoma, and another INR succumbed to extrapulmonary tuberculosis.
The impact of HAART on latent reservoirs remains an important area of investigation. Early initiation of antiretroviral therapy has been shown to reduce the reservoir size and may facilitate long-term viral control (21). Furthermore, certain antiretroviral classes, such as protease inhibitors and INSTIs, have demonstrated superior outcomes in terms of rapid viral suppression, better CD4 recovery, and lower rates of virological failure (22)(23)(24)(25)(26).
In our study, participants on zidovudine-and nevirapinecontaining regimens had a higher frequency of latent reservoirs compared to those on TDF-based regimens. This may reflect the differing pharmacokinetic profiles and reservoir penetration capacities of various drug classes. For example, INSTIs may influence viral decay kinetics in resting memory T cells (27). This could not be evaluated in our study as none of our study participants had been exposed to INSTIs during or prior to the study period. Future pharmacokinetic modeling studies are warranted to better understand how newer HAART agents influence latent reservoir dynamics.
Overall, our findings highlight a significant burden of latent HIV in circulating memory CD4-positive T cells, with 48.33% of the study population harboring proviral DNA despite having prolonged viral suppression. This emphasizes the resilience of the latent reservoir and its continued role as a barrier to HIV eradication.
We acknowledge several limitations in this study. First, the observational design and small sample size limit generalizability and statistical power. Second, prolonged cryopreservation of PBMCs prior to sorting could have led to cell loss and the underestimation of reservoir size. Third, quantification of HIV-1 proviral DNA in circulating memory T cells provides only a partial view of the overall reservoir landscape. Assessing additional circulating subsets, such as circulating T follicular helper (cTfh) cells, may offer a comprehensive understanding of latent reservoirs.
## 5 Conclusion
Latent HIV reservoirs within memory CD4-positive T cells may significantly contribute to suboptimal immune recovery in immunological non-responders. Emerging classes of antiretroviral agents that target these latent viral pools hold promise for enhancing immune restoration and bringing us closer to the goal of a functional or sterilizing HIV cure.
## References
1. (2023) "HIV statistics, globally and by WHO region"
2. Espineira, Flores-Piñas, Chafino et al. (2023) "Multi-omics in HIV: searching insights to understand immunological nonresponse in PLHIV" *Front Immunol*
3. Gazzola, Tincati, Bellistrégm et al. (2009) "The absence of CD4+ T cell count recovery despite receipt of virologically suppressive highly active antiretroviral therapy: clinical risk, immunological gaps, and therapeutic options" *Clin Infect Dis*
4. Gutieŕrez, Padilla, Masiám et al. (2008) "Patients' characteristics and clinical implications of suboptimal CD4 T-cell gains after 1 year of successful antiretroviral therapy" *Curr HIV Res*
5. Cummins, Rizza, Litzow et al. (2017) "Extensive virologic and immunologic characterization in an HIV-infected individual following allogeneic stem cell transplant and analytic cessation of antiretroviral therapy: A case study" *PloS Med*
6. Kwon, Timmons, Sengupta et al. (2020) "Different human resting memory CD4+ T cell subsets show similar low inducibility of latent HIV-1 proviruses" *Sci Transl Med*
7. Macneil, Sankaléjl, Meloni et al. (2006) "Genomic sites of human immunodeficiency virus type 2 (HIV-2) integration: similarities to HIV-1 in vitro and possible differences in vivo" *J Virol*
8. Yuan, Miley, Waters (2001) "A quantification of human cells using an ERV-3 real time PCR assay" *J Virol Methods*
9. Andersson, Yun, Sperber et al. (2005) "ERV3 and related sequences in humans: structure and RNA expression" *J Virol*
10. Sachithanandham, Ramalingam, Raja et al. (2016) "Expression of cytokine-mRNA in peripheral blood mononuclear cell of human immunodeficiency virus-1 subtype C infected individuals with opportunistic viral infections from India (South)" *Indian J Med Microbiol*
11. Grassi, Notari, Cicalini et al. (2024) "Brief report: in cART-treated HIV-infected patients, immunologic failure is associated with a high myeloid-derived suppressor cell frequency" *JAIDS J Acquired Immune Deficiency Syndromes*
12. Muñoz-Muela, Trujillo-Rodrıǵuez, Serna-Gallego et al. (2019) "HIV-1-DNA/RNA and immunometabolism in monocytes: contribution to the chronic immune activation and inflammation in people with HIV-1. eBioMedicine" *Nat Commun*
13. Chen, Zhou, Zhang et al. (2022) "The reservoir of latent HIV" *Front Cell Infect Microbiol*
14. Chomont, El-Far, Ancuta et al. (2009) "HIV reservoir size and persistence are driven by T cell survival and homeostatic proliferation" *Nat Med*
15. Sathaliyawala, Kubota, Yudanin et al. (2013) "Distribution and compartmentalization of human circulating and tissue-resident memory T cell subsets" *Immunity*
16. Ganusov, Boer (2007) "Do most lymphocytes in humans really reside in the gut?" *Trends Immunol*
17. Chun, Carruth, Finzi et al. (1997) "Quantification of latent tissue reservoirs and total body viral load in HIV-1 infection" *Nature*
18. Wen, Lu, Lin et al. (2023) "Effect of immunological non-response on incidence of Non-AIDS events in people living with HIV: A retrospective multicenter cohort study in Taiwan" *J Microbiol Immunol Infect*
19. Zoufaly, Der Heiden, Kollan et al. (2011) "Clinical outcome of HIV-infected patients with discordant virological and immunological response to antiretroviral therapy" *J Infect Dis*
20. Massanella, Ignacio, Lama et al. (2021) "Long-term effects of early antiretroviral initiation on HIV reservoir markers: a longitudinal analysis of the MERLIN clinical study" *Lancet Microbe*
21. Riddler, Haubrich, Dirienzo et al. (2008) "Class-sparing regimens for initial treatment of HIV-1 infection" *New Engl J Med*
22. Jacobson, Ogbuagu (2018) "Integrase inhibitor-based regimens result in more rapid virologic suppression rates among treatment-naïve human immunodeficiency virusinfected patients compared to non-nucleoside and protease inhibitor-based regimens in a real-world clinical setting: A retrospective cohort study" *Med (Baltimore)*
23. Sangarémn, Baril, De Pokomandy et al. (2023) "CD4/CD8 ratio outcome according to the class of the third active drug in antiretroviral therapy regimens: results from the quebec human immunodeficiency virus cohort study" *Clin Infect Dis*
24. Mills, Schulman, Fusco et al. (2021) "Virologic outcomes among people living with human immunodeficiency virus with high pretherapy viral load burden initiating on common core agents" *Open Forum Infect Dis*
25. Zhu, Rozada, David et al. (2019) "The potential impact of initiating antiretroviral therapy with integrase inhibitors on HIV transmission risk in British Columbia" *Canada. EClinicalMedicine*
26. Cardozo, Andrade, Mellors et al. (2017) "Treatment with integrase inhibitor suggests a new interpretation of HIV RNA decay curves that reveals a subset of cells with slow integration" *PloS Pathog* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12846360&blobtype=pdf | # Epizootiology of African Swine Fever in the Croatian Wild Boar Population and the Estimation of the Surviving Dynamics (2023-2024)
Magda Kamber Taslaman, Jelena Prpić, Margarita Božiković, Marica Lolić, Ljubo Barbić, Carmina Gallardo, Raquel Nieto, Lorena Jemeršić
## Abstract
This study integrates data on the prevalence, infection dynamics and risks associated with African swine fever virus (ASFV) outbreaks in Croatian wild boar during 2023-2024. Although the overall ASFV DNA prevalence in Croatia was 0.24%, the highest prevalence (2.29% in 2023 and 4.69% in 2024) was recorded in Vukovar-Srijem County. Genetic typing identified ASFV genotype II, subgroup 19, consistent with strains isolated from domestic pigs in Croatia and circulating in neighboring countries. Anti-ASFV specific antibodies were detected in 10.34% of wild boar tested in counties with previously reported DNA findings. In Vukovar-Srijem County, 4.60% of wild boar were positive for both, ASFV DNA and antibodies, suggesting ongoing virus infection, whereas the proportion of boar positive only for antibodies was 5.75%, indicating survival of acute infection. Statistical analysis revealed an increase in ASFV DNA detection from 2023 to 2024 (p = 0.043), with a higher prevalence in carcasses than in hunted animals (p = 0.001), highlighting the need for passive monitoring. While gender showed no statistical significance, a higher infection rate was observed in older animals (p = 0.001). The identified course of infection involved spillover events between domestic pigs and wild boar, with a significant anthropogenic influence.
## 1. Introduction
African swine fever (ASF) is a predominantly fatal viral disease of members of the Suidae family that causes significant socio-economic losses in affected regions. The causative agent is a large (~200 nm) DNA virus, the only member of the family Asfarviridae, genus Asfivirus [1]. To date, 24 main viral genotypes have been identified, all present in swine from Africa [2]. ASF has been spreading within sub-Saharan Africa since 1921, however, its transcontinental spread was recorded in 1957, when genotype I reached Europe, Cuba and South America. The current panzootic originated in East Africa and is caused by genotype II. The transcontinental outbreak began in 2007 in the Caucasus region (Georgia) [3] and spread primarily via wild boar across Europe, entered the European Union in 2014 and later reached Asia, Oceania and the Americas, demonstrating its presence on five continents [3,4].
In Africa, wild suids such as the African warthog (Phacochoerus africanus) and the giant forest hog (Hylochoerus meinertzhageni) act as natural virus reservoirs, showing minimal clinical signs and limited horizontal transmission [5]. In contrast, Eurasian wild boar (Sus scrofa) are highly susceptible, often developing acute, fatal infections similar to domestic pigs [6][7][8][9]. Nonetheless, some wild boar can survive the acute phase, enter a convalescent phase and potentially contribute to virus maintenance in the environment, a phenomenon referred to as the "wild boar-habitat" cycle, which is crucial for understanding the epizootiology of ASF in Europe [10,11]. Clinical outcomes of ASF vary according to viral virulence, infectious dose and route of infection, with mortality in peracute or acute forms approaching 100%, while subacute or mild courses result in 0-60% mortality [12,13]. Virus excretion in recovered animals may last up to six weeks or longer, supporting local virus circulation without constituting chronic infection [12,13].
Croatia's diverse landscape, increasing wild boar density [14] and proximity to ASFaffected countries, including Hungary (2018), Serbia (2019), Italy (2022) and Bosnia and Herzegovina (2023), created favorable conditions for ASF introduction and spread [15][16][17]. To mitigate ASF risk, preventive control measures, including active and passive surveillance, have been implemented since 2019. However, the first ASF outbreak in domestic pigs was reported in June 2023, followed by the first wild boar case in July, indicating interspecies transmission and highlighting the need for coordinated control measures [18]. Despite these efforts, ASFV persisted in the wild boar population, raising questions about their role as reservoirs and contributors to environmental viral persistence.
The aim of this study is therefore to gain a better undestanding of ASF outbreaks in the Croatian wild boar population through integrated active and passive surveillance and environmental testing over a two-year period. The results of this research will expand our understanding of viral spread and the role of wild boar as potential virus carriers by detecting both acutely infected and convalescent animals, and will also reveal the virulence characteristics of the ASFV strain circulating in Croatia under natural conditions. The results should specify targeted wildlife management and improve ASF control strategies, ultimately contributing to enhanced biosecurity in domestic pig production.
## 2. Materials and Methods
The research is carried out at the Laboratory for Classical Swine Fever Diagnostics (CSF), Molecular Virology and Genetics of the Croatian Veterinary Institute (CVI) in Zagreb, which is accredited according to the HRN EN ISO/IEC 17025:2007 standard and has been appointed by the Ministry of Agriculture (MoA) as the National Reference Laboratory of the Republic of Croatia for CSF and ASF diagnostics. Testing was also conducted at the Official Laboratory for ASF in the Vinkovci Department of CVI (since August 2024). Additionally, genetic typing was performed at the European Union Reference Laboratory (EURL) for ASF (CISA-INIA/CSIC, Madrid, Spain).
As part of the ASF active and passive surveillance programs in Croatia, established from 2019 by MoA, based on the Regulation on Control Measures for the Control of ASF in Croatia (OG 147/23), Regulation 2016/429, Delegated Regulation 2020/687, Implementing Regulation 2023/594 and the National Animal Health Act (OG 152/22, 154/22), samples for the detection of ASF were collected by hunters from the hunting grounds of twenty one Croatian counties.
## 2.1. Samples Included in This Study
In 2023 and 2024, a total of 21,934 wild boar samples (sera, spleen and bone marrow) were collected as part of the active and passive surveillance programs supported by MoA. All samples were taken and sent by official hunters and were tested for the detection of ASFV DNA. In addition, 174 samples were selected for serological testing, focusing on counties with confirmed qPCR-positive cases to investigate ASFV survival and seroprevalence in wild boar following outbreaks. The number of samples is based on a 95% confidence level within the total population of 21,934 wild boar.
Samples were categorized on the basis of active (regularly hunted wild boar with no signs of disease) and passive surveillance (wild boar carcasses and remains) and classified into age groups of up to six months, six months to one year, one to two years and over two years old, on the basis of tooth eruption and/or replacement. Both male and female wild boar were represented in the data set (Table 1). In addition, five soil and two fecal samples were collected in Vukovar-Srijem County in the Spačva hunting ground (45 • 03 ′ 38.1 ′′ N 18 • 53 ′ 21.5 ′′ E), due to reports of wild boar movement and carcasses found in the area.
## 2.2. ASFV DNA Extraction and Real-Time Polymerase Chain Reaction
Prior to analysis, sera samples were prepared by centrifugation at 1500 rpm/10 min and stored at 4 • C to 8 • C until testing. Spleen and bone marrow samples were prepared by manual homogenization (1 g of tissue resuspended in 10 mL of sterile phosphate buffered saline, PBS) vortexed for one minute and centrifuged at 3000 rpm for five minutes, then stored at -20 • C until testing. Soil and fecal samples were collected (1-5 g) and PBS was added in a ratio of 1:10 in sterile Falcon tubes. Before testing, samples were mixed using vortex for one minute and centrifuged at 3000 rpm for ten minutes. They were stored at -20 • C until testing.
DNA was extracted by an automated isolation procedure performed by a King Fisher Flex device (Thermo Fisher Scientific, Waltham, MA, USA) using the IndiMag Pathogen Kit (Indical Bioscience GmbH, Leipzig, Germany) according to the manufacturer's instructions. The exogenous positive control (Non-Target Positive Control, NTPC-ASF) was added to monitor the presence of potential PCR inhibitors. In addition, a negative ASF wild boar serum sample was included in the extraction procedure to ensure that there was no crosscontamination. A weak positive reference standard positive serum (RSPS) of wild boar prepared as a validation parameter was also included in the protocol. Isolates that were not immediately subjected to real-time polymerase chain reaction (qPCR) were stored at -20 • C until use.
QPCR was performed by the CFX96 instrument (Biorad, Hercules, CA, USA) using ID Gene™ African Swine Fever Triplex Kit (ID vet, Grabels, France) according to the manufacturer's instructions. Each reaction included an internal positive control (IPC), i.e., a positive amplification control (PAC-ASF), RSPS and a negative control. A fluorescent signal with a cycle threshold (Ct) value below 35 was considered positive. This cutoff is established based on the kit's producer's validation, ensuring reliable detection of ASFV genetic material. The Universal Probe Library (UPL) method, which is recommended by the EURL for ASF and described in the World Organization for Animal Health (WOAH) manual [19], was performed to confirm positive ID Gene ASF Triplex results. The procedure was performed using the ORA™ qPCR Probe Mix 2X (HighQu GmbH, Kraichtal, Germany), targeting an ASFV DNA fragment within the VP72 coding genomic region [20]. According to the guidelines provided by the EURL (CISA-INIA, Madrid, Spain) for ASF diagnostics and WOAH, a fluorescent signal with a Ct value below 40 was considered positive. This cutoff value is set based on the validated performance parameters of the assay to ensure that the amplification signal corresponds to the presence of ASFV DNA and not to background noise or non-specific reactions. A Ct value below 40 indicates that sufficient viral genomic material is present in the sample to allow reliable detection within the defined sensitivity of the method.
## 2.3. Indirect Immunoperoxidase Test
ASFV-specific antibodies were detected using an indirect immunoperoxidase test (IPT) validated by the EURL for ASF and performed according to its standard operating procedure (SOP) [21]. Sera, spleen and bone marrow samples were tested using antigencoated plates containing fixed, ASFV-infected VERO cells provided by the EURL.
## 2.4. Genetic Typing
Twelve representative samples, eight from domestic pigs and four from wild boar, were selected for genotyping. Initial screening was performed by partial Sanger sequencing of a fragment of the B646L/p72 gene as previously described [22]. Further molecular characterization followed the multi-gene approach described by Gallardo et al. [2], targeting six genomic regions (CVR within B602L, IGR I73R/I329L, complete O174L, partial K145R, MGF 505-9R/505-10R, and ECO2 I329L/I215L). PCRs were performed with published primers using AmpliTaq Gold DNA Polymerase (Thermo Fisher Scientific, Waltham, MA, USA). Amplicons were purified with VWR ® ExoCleanUp FAST (VWR International GmbH, Darmstadt, Germany), sequenced bidirectionally by Sanger sequencing with the BrilliantDye v1.1 Kit (NimaGen BV, Nijmegen, Netherlands), purified with the Optima DTR™ 96-Well Plate kit (Qiagen, Hilden, Germany) and analyzed on a 3730 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). Sequence quality was checked with Chromas (www.technelysium.com.au, accessed on 13 October 2025), aligned in MEGA v11 software [23], using the ClustalW algorithm with adjusted gap penalties to preserve epidemiologically informative indels (pairwise gap opening = 3, extension = 0.1; multiple alignment gap opening = 5, extension = 0.2). Following inspection and trimming of each alignment to homologous length, the six region-specific alignments were concatenated into a single composite alignment, producing one concatenated sequence per isolate.
The final concatenated alignment included 85 sequences, comprising the twelve new isolates and at least two representative isolates per genetic subgroup and/or country to capture diversity of genotype II. Phylogenetic reconstruction was performed in MEGA v11 using the Maximum Likelihood (ML) method under the General Time Reversible (GTR + G) model, selected by model testing. Bootstrap resampling (1000 replicates) was applied to assess node support. Gaps and missing data were handled by partial deletion with a 95% site-coverage cutoff, thus retaining informative indel positions.
All nucleotide sequences generated in this study were deposited in the EURL ASF sequence databank (available at https://asf-referencelab.info/sequence-database-info/, accessed on 13 October 2025).
## 2.5. Statistical Analysis
Non-parametric Chi-square tests (Pearson's Chi-square and Goodness of Fit) were performed to assess associations between ASF prevalence in 2023 and 2024 and to compare prevalence rates between carcasses and hunted wild boar, gender and age groups with 95% prevalence confidence intervals (CI). Additionally, odds ratios (OR) with 95% CI were calculated using logistic regression to quantify the association between ASF infection and categorical variables such as geographical distribution, gender and age groups. A p-value < 0.05 was considered statistically significant.
## 3. Results
## 3.1. Detection of ASFV-DNA by qPCR
The overall national prevalence of ASFV DNA positive cases was 0.15% in 2023 and increased to 0.30% in 2024 on the basis of 21,934 samples tested.
In 2023, 13 cases of ASFV DNA positive wild boar were detected in Croatia. The first case was confirmed on the 5 July 2023 in hunting ground XVI/108 Mašanj in Vukovar-Srijem County. In total, there were nine positive wild boar in Vukovar-Srijem, two in Sisak-Moslavina, one in Zadar and one in Karlovac County.
In 2024, 39 wild boar samples tested positive, 38 of which were from hunting ground XVI/11 Spačva in Vukovar-Srijem County, approximately one kilometre from the locations of outbreaks in domestic pigs and near the borders with Serbia and Bosnia and Herzegovina (up to 20 km). In November 2024, there was an additional case of ASF in hunting ground XVI/129 Vučedol, also in Vukovar-Srijem County.
ASFV DNA positive cases in wild boar are presented in Table 2 according to year of detection, location and prevalence by county. ground in Vukovar-Srijem County. The outbreak in this region peaked in March 2024, with 20 reported positive cases, followed by a sharp decline in April (Figure 1). None of the environmental soil and feces samples were positive for ASFV DNA.
## 3.2. Detection of Anti-ASFV Antibodies by IPT
IPT was performed on 36 sera, 93 spleen and 45 bone marrow samples. Seropositive cases are presented by location and prevalence by county in Table 3. As shown in Figure 3, the distribution of ASF infection stages among wild boar varied geographically, with acute, subacute and chronic infections occurring differently across counties. The majority of ASFV-positive wild boar were classified as acutely infected, particularly in Vukovar-Srijem County (XVI), where 48 acutely infected cases were recorded. Subacute infections were exclusively found in Vukovar-Srijem, with eight cases, while chronic infections were observed across three counties: Vukovar-Srijem (XVI, six cases), Karlovac (IV, two cases) and Sisak-Moslavina (III, one case). Notably, no subacute or chronic infections were detected in Zadar County (XIII).
## 3.3. Genetic Characterization of ASFV Isolates Using a Multigene Approach
A total of 16 ASFV-positive samples were analyzed, including 12 from domestic pigs and 4 from wild boar, collected in Croatia between June 2023 and July 2024 from different https://doi.org/10.3390/v18010015 outbreak locations (Table 5). Samples were selected to represent host and temporal diversity while ensuring sufficient viral load for sequencing. Twelve ASF partial B646L/p72 sequences were aligned with reference strains of ASFV genotypes I-XXIV. All Croatian samples clustered within genotype II and showed very high nucleotide identity (99.7-100%) with sequences of genotype II previously reported from Eurasia.
Molecular characterization followed the multigene approach proposed by Gallardo et al. [13], targeting six genomic regions (CVR within B602L, IGR I73R/I329L , complete O174L, partial K145R, intergenic region MGF 505-9R/505-10R , and ECO2 I329L/I215L ). All samples showed 100% identity across the six loci sequenced. The CVR region corresponded to variant I with 10 amino acid TRS, identical to the Georgia 2007/1 strain (FR682468.2). The IGR locus was consistently variant II, the most frequent subtype described in over 92% of Eurasian strains [24]. No variation was detected in O174L and K145R, both classified as variant I. Analysis of the MGF region revealed two sets of tandem repeats, ABBCD in the first block and EFGHH in the second, resulting in a total of 11 TRS. This configuration corresponds to MGF variant I and was identical to Georgia 2007/1 [2,24]. Finally, all isolates grouped with ECO2 variant II, defined by a SNP in the I215L gene leading to an amino acid substitution (Glu192Gly) as previously described [2].
The combination of these markers (CVR-I, IGR-II, O174L-I, K145R-I, MGF-I, ECO2-II) allowed the assignment of all Croatian isolates to genetic subgroup 19, according to the classification of Gallardo et al. [2]. Table 5 summarizes the ASFV-positive samples, specifying isolate ID, host species, location, sampling date and genetic profiles based on multiple markers (CVR, IGR, O174L, K145R, MGF, ECO2), as well as their assigned genetic subgroup. Phylogenetic analysis of the concatenated alignment (85 sequences, including references) corroborated this assignment (Figure 4): Croatian sequences formed a single, well-supported cluster (bootstrap = 100) together with contemporary genotype II isolates from Serbia, Greece, Bulgaria, Romania, Italy and North Macedonia [2,24].
## 3.4. Statistical Findings
Using the Pearson Chi-Square test, a statistically significant difference was found in the number of ASFV DNA-positive wild boar cases between 2023 and 2024 (p = 0.043), confirming an increasing trend in the detection of infection. ASFV DNA was detected significantly more frequently in wild boar carcasses compared to hunted animals (p = 0.001). Seasonality was also evaluated as a potential risk factor for ASFV DNA detection, with a significantly higher prevalence observed in winter compared to summer (OR = 13.89; 95% CI: 5.23-36.89; p = 0.00000011).
A statistically significant difference between age groups was detected using the Chi-Square Goodness-of-Fit test (p = 0.001), with the highest prevalence found in animals older than one year and the lowest in those younger than six months. No statistically significant difference was found in ASFV incidence between male and female wild boar (p = 0.6242). Likewise, there was no significant difference in the proportion of seropositive animals between carcasses and hunted wild boar (p = 0.6315), indicating comparable seroprevalence in both groups. However, the age-stratified analysis of seropositive animals revealed a significant difference (p = 0.001), with higher seropositivity in older individuals. No significant gender-specific differences were found among seropositive wild boar (p = 0.8065).
Karlovac County (IV) was selected as the reference group for odds ratio (OR) calculations due to its exceptionally low ASF prevalence (one positive case among 1786 wild boar tested; 0.06%), providing a stable baseline for comparison. By comparing ORs in Sisak-Moslavina (III), Vukovar-Srijem (XVI) and Zadar (XIII) Counties with Karlovac, the relative risk associated with geographical location was quantified. Males were the reference for gender-specific comparisons and the youngest age group (<6 months) was the reference for age-related analysis. This approach enabled evaluation of changes in infection probability with increasing age (Tables 6 and7).
Counties XVI and XIII showed a markedly higher ASF prevalence compared to the reference County IV. These counties exhibited very high odds ratios (72.90 and 17.16, respectively) with statistically significant p-values (<0.001 and 0.045), indicating a strong geographical influence on ASF occurrence. Gender was not significantly associated with ASF detection (p = 0.200). ASF prevalence increased with age, though none of the agerelated ORs were statistically significant. No statistically significant correlation was found for any examined risk factors. County XVI had the highest seroprevalence (20%), followed by County IV (11.11%) and County III (2.17%). The OR for County XVI is 2.00 (95% CI: 0.52-7.74), suggesting a possible trend, but the wide confidence interval and high p-value limit confidence in this effect. Gender and age groups showed small differences in seroprevalence, but none reached statistical significance.
## 4. Discussion
This study provides the first comprehensive evidence that ASF in Croatian wild boar exhibits acute, subacute and past infection courses. Active viral circulation was documented in Vukovar-Srijem County, with notable subacute cases of 4.60%. Serological evidence of past exposure was detected in 5.75% of wild boar in Vukovar-Srijem, Sisak-Moslavina and Karlovac Counties.
Following the first ASF outbreak in domestic pigs in Posavski Podgajci (Vukovar-Srijem County) on the 23 June 2023, the first ASF-positive wild boar was detected shortly afterwards on the 5 July 2023 in Mašanj, approximately 10 km from the affected farm. In total, during 2023, Croatia reported 1124 ASF outbreaks in domestic pigs, mostly in small-scale holdings (<100 pigs) with low biosecurity measures. The close temporal and spatial association between positive pigs and the first cases of ASF in wild boar highlights the possibility of virus spillover into the wild boar population. Inadequate fencing and proximity to wild boar habitats have certainly facilitated virus transmission. While alternative anthropogenic pathways, such as illegal hunting and improper carcass disposal, cannot be excluded, the regional epizootiological context supports domestic-to-wild boar transmission as a plausible contributor to these first wild boar cases, contrasting with the typical European pattern in which wild boar usually represent the primary source of infection [10]. Besides Vukovar-Srijem County, sporadic ASFV-positive wild boar cases were detected in 2023 in Sisak-Moslavina, Zadar and Karlovac counties, though these cases had minimal impact on the overall disease spread.
In 2024, the number of outbreaks in domestic pigs dropped markedly to six, reflecting successful containment measures. In neighboring Serbia, 310 outbreaks were recorded in 2024, also mainly in small-scale holdings, while other bordering countries reported fewer cases [25]. These data indicate that ASFV circulation in domestic pigs persisted in the region, maintaining the potential for further transmission, although at a reduced level. However, the situation changed in 2024 when the highest number of ASF-positive cases was detected in the Spačva hunting ground in Vukovar-Srijem County. Spačva is an oak forest and ecologically rich lowland covering over 25,000 hectares, some of which border Serbia and Bosnia and Herzegovina. It proved to be the outbreak hotspot, with 38 out of 39 ASFV DNA-positive wild boar cases, detected from January to the end of April 2024. The temporal peak of the outbreak was reached in March, followed by a significant decline in April. The statistically significant increase in ASFV DNA positive cases in 2024 compared to 2023 (p = 0.043) highlights the intensification of virus circulation within the wild boar population over time.
This study also highlights the possibility that wild boar can survive acute infection, rather than succumb to it. In Vukovar-Srijem County, 4.60% of tested wild boar were positive for both ASFV DNA and anti-ASFV antibodies, indicating prolonged viral presence and potential infectiousness beyond previously reported durations [12]. These animals may contribute to virus maintenance and spread even without visible symptoms, emphasizing the importance of active surveillance programs. Serological testing also revealed seropositive wild boar in three counties (Vukovar-Srijem, Karlovac and Sisak-Moslavina) but no detectable ASFV DNA, indicating past exposure and recovery. The presence of seropositive animals in areas without prior outbreaks, such as Sisak-Moslavina and Karlovac, suggests medium-risk zones and indicate possible silent virus circulation. These findings underscore the need for integrated surveillance combining active and passive approaches to detect both, currently infected and previously exposed wild boar, ensuring a more comprehensive understanding of ASF epizootiology.
The flooding of the Sava River in August 2023, following heavy rains, likely facilitated ASFV spread in Vukovar-Srijem County by reducing the effectiveness of containment measures and physically moving wild boar carcasses, some of which were observed floating in the floodwaters. To evaluate the potential for environmental transmission, soil and fecal samples were collected in October 2024, six months after the Spačva outbreak, at sites where ASFV-positive carcasses had previously been found. All samples tested negative for ASFV DNA, suggesting that the environment was unlikely to serve as a source of infection. However, the number of collected samples was limited due to restricted access to additional sites after control measures were implemented. Environmental persistence of ASFV is influenced by temperature, UV radiation and microbial activity, which generally promote viral inactivation [26]. Nevertheless, in November 2024, a new ASF case was detected in the Vučedol hunting ground, approximately 50 km from the Spačva outbreak. These findings suggest that local environmental sources were not responsible for this new case; instead, the virus was likely introduced through the movement of recently infected or convalescent wild boar. In contrast to environmentally related pathways, human-mediated factors appear to play a more significant role in sustaining ASF transmission in Croatia, especially among domestic pigs, but also indirectly in wild boar. Surveys of hunting practices and farm management indicate that these pathways are particularly relevant in Vukovar-Srijem County and neighboring regions.
Phylogenetic analysis of the B646L/p72 gene confirmed that all Croatian ASFV isolates, regardless of their origin, collected between June 2023 and July 2024 belong to genotype II, with nucleotide identities exceeding 99.7% when compared to reference strains. The application of the multigene approach [2] (CVR, IGR, O174L, K145R, MGF and ECO2) further demonstrated that all 16 samples analyzed were genetically homogeneous, showing complete identity across the six loci examined. Based on the updated classification of ASFV genotype II viruses circulating in Europe, which now distinguishes 28 genetic subgroups [2], all Croatian isolates were assigned to genetic subgroup 19. Unlike the situation reported in Italy, where several genetic subgroups and novel variants (e.g., subgroup 25 and 26) have been identified, the Croatian dataset did not reveal any new marker patterns, single nucleotide polymorphisms (SNPs), or alternative tandem repeat sequence (TRS) structures. This homogeneity supports a relatively recent introduction of ASFV into Croatia, with limited time for genetic diversification. The detection of subgroup 19 in Croatia is consistent https://doi.org/10.3390/v18010015 with its wide distribution in Southeastern and Eastern Europe and is predominant in Romania, Bulgaria, Serbia, Greece, and North Macedonia [2] as well as in Italy, where it has been associated with both wild boar and domestic pig outbreaks [24]. The finding that all Croatian isolates, from both wild boar and domestic pigs, belong to subgroup 19 indicates that the outbreak is part of the broader regional circulation of this lineage, without evidence of multiple introductions or diversification events, to date.
Consistent with observations from other European ASF affected countries such as Poland [27], Estonia [9] and Lithuania [28,29], this study demonstrated that ASFV DNA positive wild boar in Croatia were detected significantly more frequently among animals found dead than among those hunted, showing a statistically significant difference (p = 0.001). In 2023, 61.8% of ASF-positive detections originated from carcasses, a proportion that increased to 95% in 2024. This finding reinforces the epizootiological value of passive surveillance, which is more likely to detect animals that have died from acute ASF infection and therefore provides a more accurate reflection of active virus circulation. In contrast, active surveillance is the prefered tool when targeting surviving animals and potential reservoirs of the virus. Moreover, active surveillance can contribute to the recognition of disease or re-introduction of the virus before its manifestation or after the recognition of an outbreak. This is imperative and of crucial importance in combating ASF. Hence, integrating robust passive surveillance and active surveillance into ASF monitoring strategies can ensure effective outbreak detection and improve control interventions.
The seasonal pattern of ASF cases observed in our study is consistent with findings from other European countries, including Bulgaria, Estonia, Germany, Latvia, Lithuania and Romania [30]. Our data indicate a significantly higher transmission rate in the winter months, while in summer, new ASF cases in wild boar occur sporadically and are primarily the result of spillover events from infected domestic pigs. Lower environmental temperatures in winter slow down the decomposition of wild boar carcasses, thereby prolonging the viability of ASFV in the environment and increasing the risk of indirect transmission through contact with infectious remains [31][32][33].
Geographical location was a significant risk factor for ASFV DNA positivity in Croatian wild boar, with Vukovar-Srijem County showing the highest prevalence (3.92%) and infection probability (p < 0.0001) compared to Karlovac, while Zadar County also had an elevated risk. However, small sample sizes and wide confidence intervals warrant caution. Age trends suggested higher ASFV positivity in older animals, likely reflecting cumulative exposure and wider habitat use, but logistic regression p-values did not confirm independent age effects. This pattern aligns with findings from Lithuania [28,29] but contrasts with Estonia [34], where younger animals were more affected. Ecological factors such as predation of piglets may contribute to underrepresentation of the youngest age group. No significant differences were observed between genders, indicating that gender is not a major determinant of ASFV exposure, although males showed a slightly higher seroprevalence rate.
Although this study addressed important gaps in understanding ASF epizootiology, several limitations should be considered. First, the detection rate of ASFV DNA may have been affected by under-detection of carcasses in densely forested habitats, where limited visibility and difficult access reduce the likelihood of recovering samples. Second, the serological analysis was based on a relatively small subset of samples (n = 174), limiting statistical power and the generalizability of seroprevalence estimates. The samples were selected as representative according to their quality (absence of degradation), origin, sampling location, gender and age, as most (95%) were collected from wild boar carcasses found in counties with confirmed qPCR-positive cases. Third, the assessment of environmental transmission was limited. Sampling was restricted to soil and fecal material at a few accessible sites, precluding a comprehensive evaluation of ASFV persistence in the environment. Additionally, the role of scavengers in virus transmission and maintenance remains poorly understood, as this study did not include systematic observations of their presence or activity, particularly in the case of jackals that may share habitats with wild boar in Croatia. Fourth, small sample sizes in certain strata, particularly County XIII (n = 105) and younger age groups, may have limited the statistical power of the analyses, as indicated by wide confidence intervals and the inability to detect significant associations in logistic regression models. Although the elevated odds ratio for County XIII suggests a potentially increased risk, the wide confidence interval indicates considerable uncertainty. These results should therefore be interpreted with caution, and future studies with larger sample sizes are needed to provide more robust risk estimates.
Finally, long-term ecological monitoring through longitudinal studies is essential to assess how ASFV persists and spreads. However, our data make a considerable contribution to a better understanding of the epizootiology of ASF in the wild boar population.
## 5. Conclusions
This study reveals the epizootiological characteristics of ASF among the wild boar population in Croatia, with Vukovar-Srijem County identified as a primary hotspot marked by ongoing ASFV circulation. The detection of ASFV in Vukovar-Srijem County in both 2023 and 2024, along with evidence of subacute and post-infection stages, including seropositive cases in Karlovac and Sisak-Moslavina Counties, suggests local viral persistence and the potential role of surviving wild boar in maintaining viral spread. These findings highlight the importance of continued, integrated passive and active surveillance, particularly in highrisk areas such as the Spačva hunting ground in Vukovar-Srijem County, where indications of local endemicity are emerging. Sustained molecular and serological monitoring, together with coordinated regional control measures, will be essential to mitigate the ongoing risk of spillback from wild boar populations to domestic pigs. Given the demonstrated spillover and spillback circulation of ASFV among domestic pigs and wild boar, the potential for the development of local endemicity of ASF, particularly in North East Croatia, is a pressing epizootiological risk.
## References
1. Dixon, Escribano, Martins et al. (2005) "Virus Taxonomy, VIIIth Report of the ICTV; Fauquet"
2. Gallardo, Casado, Soler et al. (2007) "A Multi-Gene-Approach Genotyping Method Identifies 24 Genetic Clusters within the Genotype II-European African Swine Fever Viruses Circulating from" *Front. Vet. Sci*
3. Rowlands, Michaud, Heath et al. (2007) "African swine fever virus isolate, Georgia" *Emerg. Infect. Dis*
4. Usda (2022) "African Swine Fever Situation Update; United States Department of Agriculture"
5. Guberti, Khomenko, Masiulis et al. (2022) "African Swine Fever in Wild Boar-Ecology and Biosecurity, 2nd ed.; FAO Animal Production and Health Manual No"
6. Pietschmann, Guinat, Beer et al. (2015) "Course and transmission characteristics of oral low-dose infection of domestic pigs and European wild boar with a Caucasian African swine fever virus isolate" *Arch. Virol*
7. Zani, Forth, Forth et al. (2018) "Deletion at the 5 ′ -end of Estonian ASFV strains associated with an attenuated phenotype" *Sci. Rep*
8. (2026) *Viruses*
9. Jemeršić (2019) "Afrička svinjska kuga; Hrvatska akademija znanosti i umjetnosti"
10. Schulz, Staubach, Blome et al. "How to demonstrate freedom from African swine fever in wild boar-Estonia as an example"
11. Sauter-Louis, Conraths, Probst et al. (1717) "African swine fever in wild boar in Europe-A review" *Viruses*
12. Allepuz, Hovari, Masiulis et al. (2022) "Targeting the search of African swine fever-infected wild boar carcasses: A tool for early detection" *Transbound. Emerg. Dis*
13. Sánchez-Vizcaíno, Mur, Gomez-Villamandos et al. (2015) "An update on the epidemiology and pathology of African swine fever" *J. Comp. Pathol*
14. Galindo-Cardiel, Ballester, Solanes et al. (2013) "Standardization of pathological investigations in the framework of experimental ASFV infections" *Virus Res*
15. (2018) "EFSA Panel on Animal Health and Welfare. Epidemiological analyses of African swine fever in the European Union" *EFSA J*
16. Prodanov-Radulović, Mirčeta, Djurdjević et al. (2023) "African Swine Fever Outbreak in an Enclosed Wild Boar Hunting Ground in Serbia"
17. Vitale, Barzanti, Crescio et al. (2024) "Measuring transboundary disease spread-ASF in wild boars straddling Piedmont and Liguria" *Microb. Risk Anal*
18. "EFSA Panel on Animal Health and Welfare. Epidemiological analysis of African swine fever in the European Union during 2023"
19. (2025) "World Organisation for Animal Health (WOAH)"
20. Fernández-Pinero, Gallardo, Elizalde et al. (2013) *Molecular Diagnosis of African Swine Fever by a New Real-Time*
21. (2025) "Standard Operating Procedure for the Detection of Antibodies Against African Swine Fever by Indirect Immunoperoxidase Technique"
22. Bastos, Penrith, Crucière et al. (2003) "Genotyping field strains of African swine fever virus by partial p72 gene characterisation" *Arch. Virol*
23. Tamura, Stecher, Kumar (2021) "MEGA11: Molecular Evolutionary Genetics Analysis Version 11" *Mol. Biol. Evol*
24. Giammarioli, Torresi, Biccheri et al. (1185) "Genetic Characterization of African Swine Fever Italian Clusters in the 2022-2023 Epidemic Wave by a Multi-Gene Approach" *Viruses*
25. Ståhl, Boklund, Podgórski et al.
26. Taylor, Podgórski, Simons et al. (2021) "Predicting Spread and Effective Control Measures for African Swine Fever-Should We Blame the Boars?" *Transbound. Emerg. Dis*
27. Frant, Gal-Ciso Ń, Bocian et al. (1170) "African Swine Fever (ASF) Trend Analysis in Wild Boar in Poland"
28. Schulz, Masiulis, Staubach et al. (1276) "African Swine Fever and Its Epidemiological Course in Lithuanian Wild Boar" *Viruses*
29. Pautienius, Grigas, Pileviciene et al. (2014) "Stankevičius, A. Prevalence and Spatiotemporal Distribution of African Swine Fever in Lithuania"
30. Rogoll, Güttner, Schulz et al. (1955) "Seasonal Occurrence of African Swine Fever in Wild Boar and Domestic Pigs in EU Member States" *Viruses*
31. Chenais, Ståhl, Guberti et al. (2018) "Identification of Wild Boar-Habitat Epidemiologic Cycle in African Swine Fever Epizootic" *Viruses*
32. (2026) *Viruses*
33. Guberti, Khomenko, Masiulis et al. (2018) "Handbook on African Swine Fever in Wild Boar and Biosecurity During Hunting"
34. Fischer, Hühr, Blome et al. (2014) "Stability of African Swine Fever Virus in Carcasses of Domestic Pigs and Wild Boar Experimentally Infected with the ASFV "Estonia"
35. Schulz, Staubach, Blome et al. (2019) "Analysis of Estonian Surveillance in Wild Boar Suggests a Decline in the Incidence of African Swine Fever" *Sci. Rep*
36. "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=PMC12778031&blobtype=pdf | # Effects of pulsed and continuous ultrasound therapy on olfactory disorders in COVID-19 patients
Narjes Feizabadi, Abdolrahman Rostamian, Noureddin Ansari, Ehsan Moghimi, Mehdi Norouzi
## Abstract
Background and Objectives: Olfactory dysfunction is common in COVID-19 patients, with a pooled prevalence of up to 50%. This study investigated the efficacy of pulsed and continuous ultrasound treatment on olfactory disorders of these patients. Materials and Methods: Three groups of COVID-19 patients having anosmia were studied, each including 15 patients. Pulsed ultrasound and continuous ultrasound were used to evaluate their efficacy on anosmia recovery in two groups of patients. The patients were subjected to pulsed or continuous ultrasound intervention 10 times during two weeks (5 days per week). The control group received no intervention. The SIT (Smell Identification Test) was used to assess the severity of olfactory dysfunctions of all patients on days 0 and 14. Data analysis was done using MANCOVA test. Results: Totally 20 (44.4%) and 25 (55.6%) patients were affected by Delta and Omicron variants of COVID-19 virus. The SIT test results showed a significant improvement in olfactory recovery of all 30 patients except one after ultrasound treatment (p < 0.05), but this was not observed in the control group. Pulsed and continuous ultrasound treatment showed an almost equal effect on olfaction status. Conclusion: Although there was no difference in olfactory test results in the control group during intervention period, pulsed and continuous ultrasound interventions were significantly effective in improving patients' olfaction. Pulsed and continuous therapeutic ultrasound improved the COVID-19 related olfactory dysfunction and can be considered as a promising technique for postinfectious olfaction.
## INTRODUCTION
COVID-19, an infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), emerged as one of the most significant global health challenges in 2020 (1). Fever, cough, fatigue, slight dyspnea, sore throat, headache, conjunctivitis, and gastrointestinal issues are the most common symptoms of COVID-19. However, olfactory dysfunction has been increasingly reported in individuals infected with SARS-CoV-2 (2). Recent reviews estimate that up to 50% of COVID-19 patients experience olfactory dysfunction, potentially due to localized airflow impairment or sensorineural damage (3,4).
Olfactory dysfunction significantly impacts the quality of life by impairing the ability to detect food odors, environmental hazards (e.g., gas leaks, smoke), and contributing to malnutrition and psychological distress. Various treatment modalities, including olfactory training (OT), corticosteroids, and antibiotics, have been attempted to manage olfactory dysfunction. However, the uncertainty surrounding its pathogenesis complicates treatment selection (2).
Therapeutic ultrasound, an electrophysical modality that utilizes high-frequency sound waves, has been used to treat conditions such as rhinosinusitis, which can contribute to olfactory dysfunction. Previous studies suggest it may improve olfactory function in patients with chronic rhinosinusitis (5). Therefore, given the lack of effective treatments for COVID-19-related olfactory dysfunction and the potential for ultrasound to influence nasal and olfactory structures, this study aims to investigate the effects of pulsed and continuous ultrasound on olfactory recovery in COVID-19 patients.
## MATERIALS AND METHODS
Ethics approval and consent to participate. This study was approved by the Research Ethics Committee of School of Medicine -Tehran University of Medical sciences (approval ID: IR.TUMS.MEDICINE. REC.1400.1520) Study design. This study was conducted on 45 COVID-19 patients between January 2022 and December 2023. Patients were eligible if they had a positive PCR (polymerase chain reaction) test for COVID-19 and reported olfactory dysfunction. They were randomly assigned to one of three groups: pulsed ultrasound intervention (n = 15), continuous ultrasound intervention (n = 15), and a control group (n = 15). The control group received no intervention (any placebo or other therapeutic ways). All participants underwent olfactory function assessment using RT-PCR. RNA was extracted from swab samples using Qiagen DSP viral RNA kit (PishtazTeb, Tehran, Iran) and then one-step RT-PCR was performed to confirm SARS-CoV-2 infection. Each 25 μL reaction contained 5 μL of template RNA, 1 μL of 10 μM forward primer, 1 μL of 10 μM reverse primer, 0.1 mM TaqMan probe, 5 μL of distilled water, and qPCR Master Mix. RT-PCR was run as follows: 50°C for 20 min (reverse transcription), 95°C for 3 min (initial denaturation), and 45 cycles of 95°C for 10 seconds (denaturation) and 55°C for 40 s (annealing/extension). All the samples were run in duplicate.
Olfactory test. The Smell Identification Test (SIT) kit (Saba Tajhiz Sabalan, Tehran, Iran), which includes a variety of scents, was used to assess olfactory dysfunction in all patients. According to the Iran SIT kit guidelines, olfactory dysfunction is categorized as follows: anosmia (0-9 points), severe microsmia (10-13 points), mild microsmia (14-18 points), and normal olfaction (19-24 points) (Table 1). All patients in the study were classified as having anosmia or severe microsmia at baseline. Olfactory testing was performed on days 0 and 14, before and after pulsed/continuous ultrasound interventions. The control group, which did not receive any intervention, was also tested at the same time points. The SIT was administered according to standardized protocols by trained professionals at Amir Kabir University in 2016, ensuring consistency and reliability of the results.
Therapeutic ultrasound. Pulsed and continuous ultrasound were administered to two groups, each comprising 15 patients with olfactory dysfunction. Patients in the intervention groups received 10 sessions of pulsed or continuous ultrasound therapy, targeting the maxillary sinus bilaterally. The ultrasound was applied with a probe velocity of 4 cm/s, an intensity of 1 W/cm², and a duration of 8 minutes per side (6,7). Table 1. Smell identification test (SIT) results for the control, pulsed ultrasound, and continuous ultrasound groups.
SIT, administered by trained evaluators. The intervention groups received (ultrasound parameters described in section of therapeutic ultrasound). All patients provided written informed consent. The study
## After
Data analysis. Data analysis was performed using MANCOVA to compare pre-and post-treatment olfactory test scores between the intervention groups and the control group. The model adjusted for potential covariates, including baseline olfactory scores, age, and gender. Assumptions of normality and homogeneity of variances were tested and found to be met. Statistical significance was defined as a p-value of <0.05, and effect size (partial eta squared) was calculated to assess the magnitude of the treatment effects. All analyses were conducted using SPSS version 26.0.
## RESULTS
Table 1 presents the olfactory test results for the control, pulsed US, and continuous US groups. The mean olfactory test score for the pulsed US group improved significantly from 8 ± 2 (anosmia) to 15 ± 3 (mild microsmia) after the intervention. Similarly, the continuous US group showed an improvement from 5 ± 1.8 (anosmia) to 13 ± 2.5 (severe microsmia). The control group showed no significant changes, with mean scores of 7 and 8 on days 0 and 14, respectively (p = X, not significant). Statistical analysis confirmed that both pulsed and continuous US significantly improved olfactory function compared to the control group (p < 0.05). However, the difference between pulsed and continuous US treatments was not statistically significant (p <0.05). The effect sizes for pulsed and continuous US interventions were 0.5 and 0.3, respectively, indicating a moderate to strong clinical impact.
## COVID-19 variants and ultrasound treatment.
All patients were grouped according to COVID-19 variants, including Delta (n=20) and Omicron (n=25). Regardless of the virus variant, all patients demonstrated improved olfaction after pulsed / continuous ultrasound treatment, except for one (an Omicron affected patient with severe microsmia before and after continuous ultrasound treatment). No changes in olfactory status were observed in the control group (Table 2).
## DISCUSSION
The prevalence of COVID-19 related olfactory dysfunction is considerable, ranging from 41% to 62% according to two recent reviews (8,9). This may be a self-limited condition or persists for several weeks or months with partial improvement or no improvement, warranting clinical intervention. However, there is no well-defined treatment for persistent COVID-19-related olfactory dysfunction and the efficacy of available treatments remains uncertain (10).
Olfactory training (deliberate sniffing of odorants) is a treatment option considered for patients with persistent COVID-19 related olfactory dysfunction. Evidence exists for improved postinfectious olfactory function following olfactory training (11)(12)(13). Although this therapy has low cost and negligible side effects, it requires long-term commitment of at least 3 months with repeated sniffing of different odorants (14). Konstantinidis et al. demonstrated long-term olfactory training (56 weeks) produced greater improvement than short-term olfactory training (16 weeks) in patients with postinfectious olfactory loss (12).
Administration of systemic corticosteroids is another option proposed as a treatment for COVID-19 related olfactory dysfunction. Despite the debatable efficacy of this treatment, it is not currently recommended for routine management of patients with COVID-19 related olfactory dysfunction due to potential risk of harm and safety concerns (15).
Other medications have also been proposed for postinfectious olfactory dysfunction, such as intranasal sodium citrate, intranasal vitamin A, and systemic omega-3. However, there is no definitive evidence that these therapies are significantly effective in COVID-19 patients with olfactory dysfunction, and they may serve as adjuvant therapy in conjunction with olfactory training (16)(17)(18).
The effect of therapeutic ultrasound on chronic rhinosinusitis (CRS)-related olfaction dysfunction was first demonstrated by Nakhostin-Ansari A et al. (5). They showed that olfactory dysfunction and symptoms associated with chronic rhinosinusitis were substantially improved by 10 treatment sessions. Likewise, satisfactory outcomes were achieved when pulsed or continuous ultrasound was used to treat olfactory dysfunction related to COVID-19. According to this study's findings, both pulsed and continuous therapeutic ultrasound significantly improved olfactory dysfunction in COVID-19 patients, regardless of Delta or Omicron variants. No patients had anosmia after pulsed or continuous ultrasound treatment, and Mild microsmia (7) Mild microsmia (7) Omicron ( 4) Anosmia ( 1) Anosmia (1)
Mild microsmia (3) Mild microsmia (3) their olfactory status improved to severe microsmia, mild microsmia, or normal olfaction. Therapeutic ultrasound demonstrated potential benefits as a short-term intervention for postinfectious olfactory dysfunction, with clinically meaningful improvements in olfactory function. Compared to other medications, pulsed or continuous therapeutic ultrasound showed promising results as a short-term intervention for patients' recovery from postinfectious olfactory dysfunction and should be considered for further investigation..
## References
1. Moraschini, Reis, Sacco et al. (2022) "Prevalence of anosmia and ageusia symptoms among long-term effects of COVID-19" *Oral Dis*
2. Zhang, Mei, Chen et al. (2021) "Smell disorders in COVID-19 patients: role of olfactory training: a protocol for systematic review and meta-analysis" *Medicine (Baltimore)*
3. Sheen, Tan, Haldar et al. (2020) "Evaluating the onset, severity, and recovery of changes to smell and taste associated with COVID-19 infection in a Singaporean population (the CO-VOSMIA-19 trial): protocol for a prospective case-control study" *JMIR Res Protoc*
4. Mastrangelo, Bonato, Cinque (2021) "Smell and taste disorders in COVID-19: From pathogenesis to clinical features and outcomes" *Neurosci Lett*
5. Nakhostin-Ansari, Nazem, Ansari et al. (2021) "Effects of pulsed ultrasound on olfactory dysfunction in patients with chronic rhinosinusitis: A pilot study" *Complement Ther Clin Pract*
6. Feizabadi, Sarrafzadeh, Fathali et al. (2018) "Quantitative analysis of Staphylococcus aureus in patients with chronic rhinosinusitis under continuous ultrasound treatment" *Iran J Microbiol*
7. Feizabadi, Sarrafzadeh, Fathali et al. (2019) "The pulsed ultrasound strategy effectively decreases the S. aureus population of chronic rhinosinusitis patients" *BMC Res Notes*
8. Agyeman, Chin, Landersdorfer et al. (2020) "Smell and taste dysfunction in patients with COVID-19: a systematic review and meta-analysis" *Mayo Clin Proc*
9. Rocke, Hopkins, Philpott et al. (2020) "Is loss of sense of smell a diagnostic marker in COVID-19: a systematic review and meta-analysis" *Clin Otolaryngol*
10. Whitcroft, Hummel (2020) "Olfactory dysfunction in COVID-19: diagnosis and management" *JAMA*
11. Oleszkiewicz, Hanf, Whitcroft et al. (2018) "Examination of olfactory training effectiveness in relation to its complexity and the cause of olfactory loss" *Laryngoscope*
12. Konstantinidis, Tsakiropoulou, Constantinidis (2016) "Long term effects of olfactory training in patients with post-infectious olfactory loss" *Rhinology*
13. Altundag, Cayonu, Kayabasoglu et al. (2015) "Modified olfactory training in patients with postinfectious olfactory loss" *Laryngoscope*
14. Whitcroft, Hummel (2019) "Clinical diagnosis and current management strategies for olfactory dysfunction: a review" *JAMA Otolaryngol Head Neck Surg*
15. Jiang, Wu, Liang et al. (2010) "Steroid treatment of posttraumatic anosmia" *Eur Arch Otorhinolaryngol*
16. Whitcroft, Ezzat, Cuevas et al. (2017) "Hummel T. The effect of intranasal sodium citrate on olfaction in post-infectious loss: results from a prospective, placebo-controlled trial in 49 patients" *Clin Otolaryngol*
17. Hummel, Whitcroft, Rueter et al. (2017) "Intranasal vitamin A is beneficial in post-infectious olfactory loss" *Eur Arch Otorhinolaryngol*
18. Reden, Lill, Zahnert et al. (2012) "Olfactory function in patients with postinfectious and posttraumatic smell disorders before and after treatment with vitamin A: a double-blind, placebo-controlled, randomized clinical trial" *Laryngoscope* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12265398&blobtype=pdf | # Diagnostic Performance of a Combined Rapid Antigen Test for Detecting SARS-CoV-2, Influenza Virus, and Respiratory Syncytial Virus in Symptomatic Patients in Tertiary Care
Jakob Meyer, Rainer Gosert, | Roland Bingisser, Christian Nickel, Sarah Tschudin-Sutter, Karoline Leuzinger
## Abstract
Rapid antigen diagnostic tests (RDTs) can rapidly detect respiratory pathogens, allowing for the prompt initiation of infection control measures and the prevention of nosocomial spread within hospital settings. In this study, we prospectively evaluated the diagnostic performance of a combined RDT from AllTest Biotech for the simultaneous detection of SARS-CoV-2, influenza virus (IV-A/B), and respiratory syncytial virus (RSV). We compared its diagnostic performance to the Xpert-Xpress-SARS-CoV-2/Flu/ RSV molecular test using 100 naso-oropharyngeal swabs (Ct-values ≤ 35), collected from symptomatic patients with acute respiratory tract infections (RTIs) at our tertiary care hospital. The RDT showed a sensitivity of 60% (95%CI: 43.4%-74.7%) for SARS-CoV-2, with lower sensitivities for RSV at 60.0% (95%CI: 38.9%-78.2%) and IV-A/B at 54.3% (95%CI: 36.9%-70.8%). Higher sensitivities of 100% were achieved for all three viruses in respiratory samples with higher viral loads (Ct-values ≤ 25). The RDT demonstrated high specificity of > 99% for SARS-CoV-2, IV-A/B, and RSV. In conclusion, the Alltest-SARS-CoV-2/IV-A + B/ RSV RDT is effective for detecting SARS-CoV-2, IV-A/B, and RSV in samples with high viral loads, but its sensitivity significantly declines at Ct-values above 25. Therefore, negative RDT results should be confirmed with nucleic acid testing in symptomatic patients with RTIs to prevent severe consequences for clinical management.
| IntroductionRapid detection of SARS-CoV-2, influenza virus A and B (IV-A/B), and respiratory syncytial virus (RSV), causing acute respiratory tract infections (RTIs) with overlapping clinical symptoms, is essential in health care settings to implement infection control measures and guide clinical management. Nucleic acid testing (NAT) remains the gold standard due to its high sensitivity but requires specialized infrastructure and is typically associated with longer turnaround times. In contrast, rapid antigen diagnostic tests (RDTs) offer simpler workflows, faster results (within 10-15 min), and lower per-test costs. These features make RDTs well-suited for tertiary care settings, where rapid diagnosis is critical for patient triage and timely admission to inpatient care. However, ensuring the diagnostic reliability of RDTs outside centralized laboratory settings requires structured workflows,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.
robust documentation with automated result reporting, properly trained personnel, and strict adherence to quality assurance protocols. In previous studies, we reported good RDT performance for SARS-CoV-2 detection in patients with high viral loads (Ctvalue ≤ 25), with sensitivities of 91%-98% [1,2]. Given the seasonal co-circulation of SARS-CoV-2, IV-A/B, and RSV, combined RDTs have gained relevance in clinical practice. A recent study evaluating three combined SARS-CoV-2/Flu/RSV RDTs demonstrated their potential to support rapid identification of cocirculating respiratory viruses in clinical settings [3]. However, prior research has shown that RDT sensitivity can be affected by several factors, including variation in the viral antigens targeted by RDTs and patient-related characteristics such as age, immune status, and vaccination history [4,5]. Understanding these factors is essential in tertiary care settings to ensure diagnostic accuracy, support clinical decision-making, and guide effective patient management.
In this prospective study, we assessed the diagnostic performance of a combined SARS-CoV-2, IV-A/B and RSV RDT in comparison to the Xpert-Xpress-SARS-CoV-2/Flu/RSV using 100 respiratory swabs from symptomatic patients presenting with RTIs at our tertiary care hospital.
## 2 | Materials and Methods
## 2.1 | Clinical Specimens and NAT Testing
Naso-oropharyngeal swabs were collected from symptomatic patients presenting with acute RTIs, defined by the presence of at least one respiratory symptom (e.g., nasal congestion, rhinorrhea, sore throat, or cough) and one systemic symptom (e.g., fever, fatigue, headache, chills, or myalgia), in accordance with established criteria [6]. Swabs were placed in universal transport medium (UTM; Copan, Brescia, Italy), and immediately analyzed using the Xpert-Xpress-SARS-CoV-2/Flu/RSV plus test (Cepheid, CA, USA) at the Clinical Virology unit, University Hospital Basel (UHB), Switzerland, as previously described [7]. Ct-values were automatically generated by the Xpert-Xpress-SARS-CoV-2/Flu/ RSV for each pathogen based on predefined multi-gene targets. For IV-A, the reported Ct-values correspond to the matrix gene target. These values were used as semi-quantitative measures of viral RNA levels but were not calibrated to absolute viral loads.
## 2.2 | Combined SARS-CoV-2, IV-A/B and RSV RDT
From March to June 2024, 100 consecutive symptomatic patients were enrolled in the study cohort; no asymptomatic or presymptomatic patients were included. The diagnostic performance of the SARS-CoV-2/IV-A + B/RSV Antigen Combo Rapid Test (AllTest Biotech, Hangzhou, China) was assessed using 50 SARS-CoV-2-positive, 25 influenza A/B-positive, and 25 RSV-positive samples, all with Ct-values ≤ 35 on the Xpert-Xpress-SARS-CoV-2/Flu/RSV. The AllTest-SARS-CoV-2/ IV-A + B/RSV is CE-marked and conforms to EU IVD regulatory requirements. RDTs were performed according to the manufacturer's instructions within 24 h of sample collection, using naso-oropharyngeal swabs stored in UTM at 4°C. Test results were read after 10 min. Readings were performed independently by two laboratory technicians, blinded to the Xpert-Xpress-SARS-CoV-2/Flu/RSV results. All RDTs yielded valid results, with no invalid or indeterminate results.
## 2.3 | Statistical Methods
Sensitivity and specificity of the RDT was calculated according to the results of the Xpert-Xpress-SARS-CoV-2/Flu/RSV. Receiver operating characteristic analysis was done using the sensitivities of each RDT stratified by viral load. All statistical data analysis was done in R (version 4.3.3; https://cran.r-project.org), using Prism (version 10; Graphpad Software, CA, USA) for data visualization. Mann-Whitney-U test was used as indicated.
## 3 | Results
## 3.1 | Detection of Respiratory Viruses by the
Xpert-Xpress-SARS-CoV-2/Flu/RSV From July 2023 to June 2024, detection rates for RSV and IV-A/B were seasonal, with peaks in December and January, respectively, while SARS-CoV-2 activity peaked in December but remained detectable year-round with a positivity rate ranging from 5% to 10% (Figure 1A). Upon hospital presentation, patients showed median Ct-values of 24 for SARS-CoV-2 (range: 11-44), 23 for IV-A/B (range: 10-44), and 23 for RSV (range: 15-44; Figure 1B). Respiratory swabs were submitted from 6488 (90.5%) symptomatic adult and 683 (9.5%) pediatric patients (Figure 1C). The median age of SARS-CoV-2-positive patients was 71 years, while IV-A/B-and RSV-positive patients were significantly younger at 51 and 31 years (p < 0.001; Figure 1C). The study cohort used to evaluate RDT performance comprised a subset of 100 patients from this overall population.
## 3.2 | Comparison of Respiratory Virus Detection
Between the Xpert-Xpress-SARS-CoV-2/Flu/RSV and the Alltest-SARS-CoV-2/IV-A + B/RSV RDT We prospectively evaluated 100 consecutive naso-oropharyngeal swab samples and compared the results to the Xpert-Xpress-SARS-CoV-2/Flu/RSV. All virus-negative samples were nonreactive in the Alltest-SARS-CoV-2/IV-A + B/RSV, yielding a specificity of 100% for SARS-CoV-2, IV-A/B, and RSV. Among the 40 SARS-CoV-2-positive samples, 24 were correctly detected by the RDT, corresponding to an overall agreement of 84% and a moderate Cohen's kappa coefficient of 0.64 (Supporting Information Table 1A). For IV-A/B, the Alltest-SARS-CoV-2/IV-A + B/ RSV identified 19 of the 25 IV-A/B-positive samples, corresponded to an overall agreement of 84% and a moderate Cohen's kappa coefficient of 0.61 (Supporting Information Table 1B). For RSV, the Alltest-SARS-CoV-2/IV-A + B/RSV detected 15 of the 25 RSV-positive samples, resulting in an overall good agreement of 90%, and a Cohen's kappa coefficient of 0.69 (Supporting Information Table 1C). However, discrepancy rates between 40% and 46% were observed across all three respiratory viruses when using the Alltest-SARS-CoV-2/IV-A + B/RSV (Supporting Information Table 1A-C). Comparison of viral loads between concordant and discordant samples revealed significantly lower Ct-values in concordant positives than in discordant cases (p < 0.001; Figure 2A). These findings suggest that the discordant results likely occurred near the detection limit of the Alltest-SARS-CoV-2/IV-A + B/RSV.
## 3.3 | Diagnostic Performance of the Alltest-SARS-CoV-2/Iv-A + B/RSV RDT Based on Respiratory Virus Load
To further evaluate the diagnostic performance of the Alltest-SARS-CoV-2/IV-A + B/RSV, RDT sensitivity was assessed in relation to respiratory virus load. Across the full data set of all available samples with Ct-values ≤ 35, overall RDT sensitivity was 60.0% (95%CI: 43.4%-74.7%) for SARS-CoV-2, 54.3% (95% CI: 36.9%-70.8%) for IV-A/B, and 60.0% (95%CI: 38.9%-78.2%) for RSV. In samples with Ct-values ≤ 30, sensitivity improved to 81.4% (95%CI: 61.3%-93.0%) for SARS-CoV-2, 59.4% (95%CI: 40.8%-75.8%) for IV-A/B, and 68.2% (95%CI: 45.1%-85.3%) for RSV (Table 1, Figure 2B). Sensitivity increased further with higher viral loads, reaching 100% for all three viruses in samples with Ct-values ≤ 25 (Figure 2B).
When assessing viral loads measured by the Xpert-Xpress-SARS-CoV-2/Flu/RSV from June 2023 to July 2024, we found that 1478 out of 2039 (72.5%) samples had Ct-values ≤ 30 (Figure 3A). When applying sensitivity thresholds based on viral loads, the Alltest-SARS-CoV-2/IV-A + B/RSV would have detected 652 of 695 (93.8%) SARS-CoV-2-positive cases, 486 of 585 (83.1%) IV-A/B-positive cases, and 177 of 198 (89.3%) RSVpositive cases with Ct-values ≤ 30 (Figure 3B,C).
## 4 | Discussion
Rapid detection of SARS-CoV-2, IV-A/B, and RSV is critical for informed clinical decision-making in tertiary care. It enables timely implementation of isolation protocols and infection control measures to reduce the risk of nosocomial transmission. In addition, it supports efficient resource management by guiding the allocation of ICU beds, medical staff, and treatments to the patients who need them most. NAT enables rapid and highly sensitive detection of respiratory viruses from clinical specimens, often delivering results in under an hour [8][9][10]. However, NAT requires specialized equipment, trained personnel, and is associated with higher per-test costs. RDTs provide results within 10-15 min, require minimal laboratory infrastructure, and have lower operational costs, making them suitable for broader implementation across various healthcare settings. Still, the accuracy of RDTs depends on well-trained personnel, and strict adherence to quality assurance protocols. Furthermore, RDTs must be supported by structured workflows, robust documentation practices, and seamless integration with laboratory information systems to maintain traceability and ensure timely clinical access to test results.
Our study evaluated the Alltest-SARS-CoV-2/IV-A + B/RSV, a combined RDT for detecting SARS-CoV-2, IV-A/B, and RSV. The test demonstrated high specificity ( ≥ 99%) for all three viruses, consistent with previous reports on RDT performance [ 1,2,11,12]. Sensitivities were 60.0% for SARS-CoV-2, 54.3% for IV-A/B, and 60.0% for RSV, aligning with published sensitivity ranges [1][2][3][11][12][13][14]. Our findings also confirm that RDT sensitivity is highest in samples with high viral loads (Ctvalue ≤ 25), consistent with prior reports showing reduced sensitivity at Ct-values above this threshold [1][2][3]13]. Our findings also align with a recent large-scale evaluation of three multiplex SARS-CoV-2/Flu/RSV RDTs in over 1600 symptomatic patients at a tertiary care clinic. Sensitivities exceeded 80% and specificities were generally > 97% for SARS-CoV-2, IV-A, and RSV detection, though one RDT showed reduced specificity for IV-A [3]. These findings support the potential of combined RDTs for rapid detection of co-circulating respiratory viruses while underscoring the need for independent validation in clinical settings.
It is important to note that in our study, RDTs were performed under optimal laboratory conditions with trained personnel, high-quality swabs in UTM, and clear-positive samples with Ct-values ≤ 35 on the Xpert-Xpress-SARS-CoV-2/Flu/RSV. Thus, the reported sensitivity and specificity reflect diagnostic performance under best-case conditions in terms of sample quality and test execution, and may not fully represent test performance in routine clinical practice, point-of-care settings, or unsupervised self-testing at home, especially with selfcollected respiratory samples. Recent studies have shown that nasopharyngeal and nasal self-swabbing provide more reliable diagnostic results, whereas buccal swabs tend to yield lower viral RNA levels, especially following food intake [15,16]. Therefore, sample quality and operational factors are critical to RDT reliability, highlighting the need for robust quality assurance protocols and ongoing staff training.
Furthermore, patient characteristics also influence RDT performance. Our study included only symptomatic individuals, aligning with the intended clinical use of RDTs in emergency and outpatient triage settings. Since sensitivity is significantly lower in asymptomatic or presymptomatic individuals [17], our findings may not be fully generalizable to broader screening contexts. Immune status also significantly influences test performance, with reported sensitivities as low as 32% among vaccinated individuals. Patient age is another important factor affecting test reliability, with RSV sensitivity shown to decrease from 52.3% in infants to 33.3% in children > 5 years [5,13]. In our study, 80% of RSV-positive samples were from children < 5 years, likely contributing to the observed sensitivity of the RDT. Finally, the timing of diagnostic testing is also critical. At our tertiary care center, symptomatic patients presented with a wide range of viral loads. RDT performance is typically highest in the first days after symptom onset, when viral RNA levels peak [9,10,14,18], with reported sensitivity ranges from 35.7% to 71.4%, increasing to 90.6% by day four [19]. Although viral detection beyond 10 days after symptom onset is generally not considered to represent infectious virions [20], some samples with Ct-values up to 35 yielded culturable virus [18]. This indicates that patients with lower viral loads, and corresponding negative RDT results, may still be infectious, especially in the early stages of infection. In contrast, NATs may remain positive for weeks after infection, detecting residual RNA even when infectious virus is no longer present [14]. Notably, immunocompromised individuals and transplant recipients, may continue to shed infectious virus at low viral loads for extended periods [20]. These findings underscore that while high viral loads typically correlate with the presence of culturable virus, neither NAT nor antigen-based RDTs can reliably distinguish between infectious virus and noninfectious viral remnants.
Only virus isolation in cell culture can confirm infectivity, which is not feasible in routine diagnostics. Thus, the relationship between viral load and contagiousness is complex and modulated by host-and virus-specific factors such as immune status, timing of sample collection, and viral replication kinetics. Consequently, both NAT and RDT results must be interpreted with caution and in clinical context, when guiding infection control measures and patient management.
These considerations are particularly relevant when evaluating the clinical utility of RDTs in healthcare settings. Although the evaluated RDT meets the World Health Organization's performance criteria ( ≥ 80% sensitivity and ≥ 97% specificity), its clinical applicability remains limited. While the high specificity observed in our study supports accurate identification of uninfected individuals, even highly specific RDTs may produce a high proportion of false positives during periods of low respiratory virus circulation. Therefore, their use should be guided by local epidemiological trends and supported by confirmatory molecular testing when clinical suspicion is low or testing occurs outside peak virus seasons. Additionally, the moderate sensitivity observed for SARS-CoV-2, IV-A/B, and RSV reduces the reliability of RDTs for comprehensive respiratory virus screening. Collectively, these findings indicate that the clinical utility of RDTs remains limited in high-risk settings such as tertiary care facilities. Some limitations should be considered when interpreting the findings of this study. First, we did not include patients who tested negative for all three respiratory viruses. Although this limited the ability to fully assess specificity in a virus-negative population, our use of mono-positive samples allowed for the practical evaluation of cross-reactivity. All samples negative for a given virus, tested negative in the corresponding RDT lane, supporting the test's high analytical specificity. Second, our sample size (n = 100) was relatively small, limiting the statistical power for subgroup analyzes. However, the inclusion of well-characterized, symptomatic patients with a range of viral loads enabled the identification of meaningful trends in diagnostic performance. These results provide a foundation for future studies involving larger and more diverse cohorts to further evaluate the performance of multiplex RDTs.
## References
1. Greub, Caruana, Schweitzer (2021) "Multicenter Technical Validation of 30 Rapid Antigen Tests for the Detection of SARS-CoV-2 (VALIDATE)" *Microorganisms*
2. Leuzinger, Roloff, Egli et al. (2022) "Impact of SARS-CoV-2 Omicron on Rapid Antigen Testing Developed for Early-Pandemic SARS-CoV-2 Variants" *Microbiology Spectrum*
3. Murphy, Mak, Cheng (2024) "Diagnostic Performance of Multiplex Lateral Flow Tests in Ambulatory Patients With Acute Respiratory Illness" *Diagnostic Microbiology and Infectious Disease*
4. Wagenhäuser, Knies, Pscheidl (2024) "SARS-CoV-2 Antigen Rapid Detection Tests: Test Performance During the COVID-19 Pandemic and the Impact of COVID-19 Vaccination" *EBioMedicine*
5. Krone, Wagenhäuser, Knies (2023) "Clinical Accuracy of SARS-CoV-2 Rapid Antigen Testing in Screening Children and Adolescents" *Journal of Infection*
6. Ison, Hirsch (2019) "Community-Acquired Respiratory Viruses in Transplant Patients: Diversity, Impact, Unmet Clinical Needs" *Clinical Microbiology Reviews*
7. Goldenberger, Leuzinger, Sogaard (2020) "Brief Validation of the Novel Genexpert Xpress SARS-CoV-2 PCR Assay" *Journal of Virological Methods*
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. Leuzinger, Gosert, Søgaard (2021) "Epidemiology and Precision of SARS-CoV-2 Detection Following Lockdown and Relaxation Measures" *Journal of Medical Virology*
10. Leuzinger, Roloff, Gosert (2020) "Epidemiology of Severe Acute Respiratory Syndrome Coronavirus 2 Emergence Amidst Community-Acquired Respiratory Viruses" *Journal of Infectious Diseases*
11. Dinnes, Sharma, Berhane (2022) "Rapid, Point-of-Care Antigen Tests for Diagnosis of SARS-CoV-2 Infection"
12. Bruning, Leeflang, Vos (2017) "Rapid Tests for Influenza, Respiratory Syncytial Virus, and Other Respiratory Viruses: A Systematic Review and Meta-Analysis" *Clinical Infectious Diseases*
13. Savolainen, Peltola, Hilla et al. (2025) "Clinical Performance of Two Commercially Available Rapid Antigen Tests for Influenza, RSV, and SARS-CoV-2 Diagnostics" *Microbiology Spectrum*
14. Korenkov, Poopalasingam, Madler (2021) "Evaluation of a Rapid Antigen Test to Detect SARS-CoV-2 Infection and Identify Potentially Infectious Individuals" *Journal of Clinical Microbiology*
15. Dräger, Bruni, Bernasconi (2024) "Impact of Swabbing Location, Self-Swabbing, and Food Intake on SARS-CoV-2 RNA Detection" *Microorganisms*
16. Puyskens, Michel, Stoliaroff-Pepin (2023) "Direct Comparison of Clinical Diagnostic Sensitivity of Saliva From Buccal Swabs Versus Combined Oro-/Nasopharyngeal Swabs in the Detection of SARS-CoV-2 B. 1.1. 529 Omicron" *Journal of Clinical Virology*
17. Veroniki, Tricco, Watt (2023) "Rapid Antigen-Based and Rapid Molecular Tests for the Detection of SARS-CoV-2: A Rapid Review With Network Meta-Analysis of Diagnostic Test Accuracy Studies" *BMC Medicine*
18. Yamada, Fukushi, Kinoshita (2021) "Assessment of SARS-CoV-2 Infectivity of Upper Respiratory Specimens From COVID-19 Patients by Virus Isolation Using VeroE6/TMPRSS2 Cells" *BMJ Open Respiratory Research*
19. Frediani, Parsons, Mclendon (2024) "The New Normal: Delayed Peak SARS-CoV-2 Viral Loads Relative to Symptom Onset and Implications for COVID-19 Testing Programs" *Clinical Infectious Diseases*
20. Puhach, Meyer, Eckerle (2023) "SARS-CoV-2 Viral Load and Shedding Kinetics" *Nature Reviews Microbiology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12561905&blobtype=pdf | # Here we go again: More diseases dubiously attributed to pegivirus infection
Adam Bailey
## Abstract
Pegiviruses (PgVs) are a genus of +ssRNA viruses currently classified within the Flaviviridae family [1,2]. Pegivirus infections are widespread in mammals, and two divergent PgVs have been discovered in humans: HPgV-1 (aka. Pegivirus hominis), which infects ~15% of the global human population; and the much rarer HPgV-2 (aka. Pegivirus columbiaense) [1,2]. These viruses cause high-titer viremia that can persist for years to decades, with the phenomenon of viral clearance being well-documented but poorly understood. Due to the high prevalence of HPgV infection in humans and the abundance of viral RNA in infected individuals, metagenomic studies on human samples routinely detect HPgV infections. Iatrogenic HPgV infections, via procedures like blood transfusion, are presumed to be common but HPgV diagnostics are not used clinically.PgVs are phylogenetically related to several important human pathogens with hepatitis C virus (HCV) being the closest medically relevant virus, though these viruses share only ~50% nucleotide and ~30% amino acid sequence identity. However, unlike HCV, HPgV has never been shown to play a causal role in any disease. This is not for lack of effort: since the discovery of HPgV, the HPgV field has lurched from one disease association to the next. When HPgV was first identified in 1995 in patients with hepatitis, two different research groups gave it the names "Hepatitis G virus (HGV)" and "GB virus C (GBV-C)" with the assumption that this virus was hepatotropic and responsible for chronic hepatitis [3,4]. Indeed, due to its high prevalence in the human population and its shared modes of transmission with other hepatitis viruses, early studies consistently identified HGV in cohorts of patients with unexplained hepatitis. It took years of careful epidemiological studies comparing healthy and diseased populations to definitively rule out HGV/GBV-C as a cause of hepatitis (summarized in [1,5]), which was further confirmed by prospective longitudinal studies in chimpanzees [6]. Nevertheless, this misconception persists to this day [7]. In the 2000s, several large clinical studies associated HPgV (officially renamed from GBV-C/HGV in 2013) with improved outcomes in HIV+ individuals [8]. Since this time, the majority of research on HPgV has focused on understanding the interactions between HPgV and HIV/AIDS disease progression (summarized in [9]). Although some evidence suggests that HPgV may directly inhibit HIV replication (e.g., by reducing HIV receptor/co-receptor expression in CD4 T cells), the mechanism(s)
underlying the observed clinical associations remain far from clear. Later, in the 2010s, several studies identified a weak association between HPgV infection and the development of various lymphoid malignancies (see [10] for a meta-analysis), though again, the mechanistic underpinnings of this association remain unclear.
Recently, several case reports and small case-control studies (including one published in the New England Journal of Medicine) have attempted to establish a link between HPgV infection and a spectrum of neurological disorders including encephalitis [11][12][13][14][15][16], osloclonus-myoclonus [17], optic neuritis and myelitis [18,19], and Parkinson's disease [20]. Given the prevalence of HPgV, its identification in any population-including patients with neurological diseases-is unsurprising [21]. Nevertheless, a link has been pursued. Imaging and histologic analyses in these studies have revealed nonspecific abnormalities that could be due to any number of nonHPgV-related processes [11,13,15,18,20]. Detection of HPgV replication intermediates or proteins within neural tissues have relied upon techniques with a high potential for generating false-positive results (i.e., strand-specific PCR and/or immunohistochemistry), the results of which have been published without sufficient data on assay optimization, critical controls, quantification, or consideration for typical patterns of viral staining (e.g., cytoplasmic localization of viral RNA) [11,13,20]. Unconvincing arguments have also been built upon phylogenetic analyses that purport "compartmentalization" of HPgV replication within the central nervous system [11][12][13]. Small sample sizes and a lack of HPgV+ control cases without disease also limit the power of these studies. Altogether, the data show nothing more than what would be expected from a coincidental HPgV infection [11,12].
It remains possible that HPgV infection is associated with a range of diseases, and/or that HPgV may even play a causal role in the development of some diseases. However, the overwhelming evidence collected to date suggests that HPgV is nonpathogenic; thus, changing this consensus requires substantial evidence to the contrary-evidence that, in this author's opinion, is currently lacking. While some of the recent articles are appropriately reserved when interpreting their findings [14,21,22], notable exceptions include causality-implying statements such as "neurological involvement due to Pegivirus" [19]; the suggestion that HPgV has "neurotropism and a causative role in pegivirus-associated encephalomyelitis" [11]; or the implication that HPgV plays an active role in "alter[ing] brain and and blood immune and transcriptomic profiles" [20]. The creation of terms like "pegivirus-associated encephalomyelitis" [11] now enshrining this tenuous connection via a self-perpetuating snowball effect that is all too common in medical literature.
As physicians and scientists, we must temper our urge to attribute infections to diseases without definitive evidence. This urge is particularly strong for viruses which, by definition, must infect the cells of its host to survive. Premature attribution of disease can have long-lasting harmful effects on a field. Ironically, the HPgV field is already a prime example of this, with rafts of speculative low-quality studies littering the HPgV literature over the past several decades. Ultimately, this has created an "unserious" reputation that discourages innovation and funding. The current disease-oriented culture of scientific funding is partially to blame, which has forced
## References
1. Stapleton, Foung, Muerhoff et al. (2011) "The GB viruses: a review and proposed classification of GBV-A, GBV-C (HGV), and GBV-D in genus Pegivirus within the family Flaviviridae" *J Gen Virol*
2. Yu, Wang, Yang et al. (2022) "Review of human pegivirus: prevalence, transmission, pathogenesis, and clinical implication" *Virulence*
3. Simons, Leary, Dawson et al. (1995) "Isolation of novel virus-like sequences associated with human hepatitis" *Nat Med*
4. Linnen, Wages, Jr et al. (1996) "Molecular cloning and disease association of hepatitis G virus: a transfusion-transmissible agent" *Science*
5. Alter (1997) "G-pers creepers, where'd you get those papers? A reassessment of the literature on the hepatitis G virus" *Transfusion*
6. Bukh, Kim, Govindarajan et al. (1998) "Experimental infection of chimpanzees with hepatitis G virus and genetic analysis of the virus" *J Infect Dis*
7. Vazzana, Mularoni, Vaiana et al. (2025) "Shotgun metagenomics detects the human pegivirus complete genome in a pediatric patient with acute hepatitis of unknown etiology: a case report" *Front Genet*
8. Xiang, Wünschmann, Diekema et al. (2001) "Effect of coinfection with GB virus C on survival among patients with HIV infection" *N Engl J Med*
9. Bhattarai, Stapleton (2012) "GB virus C: the good boy virus?" *Trends Microbiol*
10. Fama, Larson, Link et al. (2020) "Human pegivirus infection and lymphoma risk: a systematic review and meta-analysis" *Clin Infect Dis*
11. Scheibe, Melchert, Radbruch et al. (2025) "Pegivirus-associated encephalomyelitis in immunosuppressed patients" *N Engl J Med*
12. Bukowska-Ośko, Perlejewski, Pawełczyk et al. (2018) "Human pegivirus in patients with encephalitis of unclear etiology" *Poland. Emerg Infect Dis*
13. Balcom, Doan, Branton et al. (2018) "Human pegivirus-1 associated leukoencephalitis: Clinical and molecular features" *Ann Neurol*
14. Tuddenham, Eden, Gilbey et al. (2020) "Human pegivirus in brain tissue of a patient with encephalitis" *Diagn Microbiol Infect Dis*
15. He, Yang, Niu (2025) "Human pegivirus detected in patient with reversible severe encephalitis and axillary lymphadenopathy: a case report" *Diagn Pathol*
16. Fridholm, Sørensen, Rosenstierne et al. (2016) "Human pegivirus detected in a patient with severe encephalitis using a metagenomic pan-virus array" *J Clin Virol*
17. Taga, Kolchinski, Reed et al. (2025) "Bridging the gap: opsoclonus-myoclonus syndrome: human pegivirus encephalomyelitis diagnosed through metagenomic next-generation sequencing" *Neurology*
18. Valyraki, Maillart, Pourcher et al. (2023) "Human pegivirus identified in severe myelitis and optic neuritis in immunocompromised patients: a pathogenic role for a forgotten virus" *Rev Neurol*
19. Pourcher, De, Motte et al. (2025) "Lack of efficacy of sofosbuvir in human pegivirus associated neurological disorders" *Rev Neurol*
20. Hanson, Dang, Jamshidi et al. (2025) "Human pegivirus alters brain and blood immune and transcriptomic profiles of patients with Parkinson's disease" *JCI Insight*
21. Hardie, Smuts (2017) "Human pegivirus-1 in the CSF of patients with HIV-associated neurocognitive disorder (HAND) may be derived from blood in highly viraemic patients" *J Clin Virol*
22. Carmona R De, Cc, Cilli et al. (2023) "Pegivirus detection in cerebrospinal fluid from patients with central nervous system infections of unknown etiology in Brazil by viral metagenomics" *Microorganisms*
23. Manns, Maasoumy (2022) "Breakthroughs in hepatitis C research: from discovery to cure" *Nat Rev Gastroenterol Hepatol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12230535&blobtype=pdf | # Correction to "Hantaan Virus (HTNV) Human Infection on Jeju Island, South Korea: Unique Phylogeny and Epidemiology of HTNV" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12680079&blobtype=pdf | # Genoprevalence of cutavirus in benign and malignant intestinal and breast tissues
Irini Assimakopoulou, Ushanandini Mohanraj, Taina Sipponen, Anna Lepistö, Dalia Kadry, Rana Hamdy, Mahmoud Kamel, Heba El-Batal, Ahmed Abdel-Moneim, Maria Söderlund-Venermo
## Abstract
Cutavirus (CuV), the newest human protoparvovirus (PPV), has gained attention due to its significant association with cutaneous T-cell lymphoma (CTCL) and its precursor, parapsoriasis, whereas the other human PPVs, bufavirus and tusavirus, show no such link. Given this association, it is important to investigate the prevalence of CuV DNA in other tissues, particularly those affected by malignancy or inflammation. This study assessed, by multiplex quantitative PCR, the genoprevalences of all three PPVs in 427 fresh-frozen intestinal biopsies from inflammatory bowel disease (IBD), colorectal cancer, adenomas or healthy mucosa of 185 individuals, as well as in 94 formalin-fixed paraffin-embedded (FFPE) biopsies from malignant and non-malignant breast conditions of 85 patients. The study also compared the DNA prevalences of human herpesvirus (HHV)-6A, -6B and -7 in the breast tissues. CuV mRNA was assayed with reverse-transcription PCR, and corresponding FFPE sections underwent in situ hybridization. CuV DNA was detected in intestinal IBD or healthy mucosa from 6/185 (3.2%) subjects, but no CuV mRNA or in situ signals were detected. In breast biopsies, HHV-6B and HHV-7 DNAs were present in 20.3 and 5.1%, respectively, while all PPVs and HHV-6A were absent. Overall, CuV was absent in all 70 cancer tissues, underscoring its association with CTCL. The low CuV DNA loads and prevalences in intestinal and breast morbidities, and lack of activity, suggest that CuV is unlikely to play a role in these malignancies or inflammatory conditions. In contrast, HHV-6B may be more relevant to breast pathology, even though it is also widely detected in healthy tissues. Nevertheless, our study provides insight into persistent DNA viruses implicated in cancer and highlights their occurrence across various disease manifestations, laying a foundation for future studies.The Microbiology Society is a membership charity and not-for-profit publisher.Your submissions to our titles support the community -ensuring that we continue to provide events, grants and professional development for microbiologists at all career stages.Find out more and submit your article at microbiologyresearch.org
## INTRODUCTION
In 2012-2016, three new parvoviruses of the Protoparvovirus genus, named bufavirus (BuV), tusavirus (TuV) and cutavirus (CuV), were by metagenomics discovered in stools [1][2][3]. CuV has since been associated with cutaneous T-cell lymphoma (CTCL) and its precursor, parapsoriasis en plaques (PP), a chronic inflammatory condition in skin [3][4][5][6]. Recent studies have shown that in individuals with CTCL or PP, CuV remains persistent in skin for years, actively expresses mRNA, spreads via circulating immune cells within the body and is excreted to the environment in stools as potentially infectious virions [4][5][6]. CuV-specific seropositivity does not seem to hamper the active persistence of this virus.
Inflammatory bowel disease (IBD) is a long-term inflammatory condition that affects the digestive tract and is categorized as either Crohn's disease or ulcerative colitis (UC) [7]. The cause of IBD is unknown, but it is considered the result of an inappropriate immune response against environmental factors, including luminal and microbial antigens in genetically susceptible hosts [8]. Associations between other persistent parvovirus infections such as parvovirus B19 (B19V) or human bocavirus (HBoV) and IBD or gastrointestinal tumours have been suggested, but their causative role has not been verified [9][10][11][12][13]. Further, several persistent DNA viruses have been implicated in cancer, including human herpesviruses (HHVs), papillomaviruses and polyomaviruses [14]. The nine HHVs are globally ubiquitous pathogens that persist lifelong and can reactivate, sometimes causing severe diseases. However, whether HHV-6 or -7 has direct roles -or can act as contributory factors -in tumourigenesis is less clear; their prevalence and role in breast cancer have rarely been studied [15,16].
Hence, the aim of our current study was to elucidate the DNA prevalence of recently discovered human protoparvoviruses (PPVs), BuV, TuV and CuV, in intestinal and breast biopsy specimens from patients with IBD, colorectal cancer or malignant or non-malignant breast conditions. Additionally, as all PPV PCRs were negative in the breast samples, we also investigated the DNA prevalence of the more common HHV-6A, -6B and -7, since they are rarely examined in breast biopsy specimens.
## METHODS
## Patients and clinical specimens
## Gut cohort
In total, 427 fresh-frozen intestinal specimens from 185 individuals (aged 20-86 years; mean 51; median 52.3) were biopsied at the Helsinki University Hospital (Helsinki, Finland) and included in our study. These samples had been previously studied for parvovirus B19 and HBoV infections [13]. Samples were immediately immersed in RNAlater at 4 °C and subsequently preserved at -20 °C until nucleic-acid extraction. Of the 185 individuals presenting with malignancy (N=16), active (N=42) or inactive (N=33) UC, or adenoma (N=39), 129 provided paired mucosa of the disease-affected colon and adjacent healthy colon or ileum segments or both. The remaining 55 individuals were by histopathological examination considered healthy, providing only healthy colon or ileum biopsies.
Of these biopsied individuals, only a total of 13 serum samples were available, from 4 patients with inactive UC, 7 with active UC and 2 with adenoma. For in situ hybridization, 11 formalin-fixed paraffin-embedded (FFPE) samples were additionally obtained from Helsinki Biobank from 5 CuV DNA-positive and, as negative controls, 2 CuV PCR-negative individuals, all from the same intestinal regions as the fresh-frozen RNAlater-stored samples.
## Breast cohort
In total, 94 FFPE tumour tissue specimens from 85 patients (aged 27-82 years; mean 52; median 50.5) were collected from the Baheya Centre for Early Detection and Treatment of Breast Cancer and National Research Centre (Cairo, Egypt) between 2022 and 2023. Patients were diagnosed with breast cancer (n=53), benign breast conditions (n=28), other benign conditions (n=3) or prostate cancer (n=1), as detailed in Table S3, available in the online Supplementary Material. Tissue slices were cut from FFPE tissue blocks and stored as rolls in microtubes at room temperature until DNA extraction.
## Ethics
The study was approved by the Ethics Committee of the HUS Helsinki University Hospital (Decision number 553/E6/01, amendments 15 July 2014 and 2322/2024) and the Baheya Research Ethics Committee (Decision number 202302200006), and informed consent was obtained from all individuals.
## Methods
All methods mentioned here, in short, are further described in File S1.
## DNA and RNA extraction, PCR and EIA
Total DNA and RNA were separately extracted, PPV and HHV DNA detected and quantified by multiplex quantitative PCRs (qPCRs), spliced CuV mRNA detected and quantified by reverse-transcription (RT) PCR (Table S2) and PPV IgG and IgM measured by in-house enzyme immunoassays (EIAs), as previously published [4,6,[17][18][19][20]. In addition, the CuV DNA-positive samples, identified by PPV multiplex-qPCR screening (targeting a short 91 bp VP2 region), were subsequently selected for amplification of also longer (>500 bp) VP1 and VP2 fragments for phylogenetic analysis; all primers and probes are described in Table S2, and the resulting sequences were submitted to GenBank (PQ553262-PQ553268) [2]. All samples also underwent human RNaseP qPCR or RPII RT-PCR to control for cell DNA or RNA integrity, respectively.
## RNAscope in situ hybridization
RNAscope ISH (RISH) technology (ACD, Newark, CA) was applied on the seven gut FFPE biopsies with Probe-V-CuV-NS1, with positive and negative controls [6].
## Statistical analysis
The Fisher's exact test was used for comparison of virus prevalence in different disease groups. P values<0.05 were considered statistically significant.
## RESULTS
## PPV DNA prevalence in intestinal biopsy specimens of the gut cohort
Only 7 of 427 (1.6%) intestinal specimens from 6/185 (3.2%) individuals harboured CuV DNA: 1/42 (2.4%) with active UC, 1/33 (3%) with inactive UC, 1/39 (2.6%) with adenoma and 3/55 (5.5%) from healthy subjects, in various intestinal sites (Table 1). Moreover, only one of these six individuals exhibited CuV DNA in both biopsies, while otherwise mostly in the healthy segment. Only one CuV DNA-positive specimen (IBD 45) showed histopathological signs of disease: a hyperplastic polyp. All CuV DNA loads were low, <10 5 CuV copies per one million cells (cpm) (Table 1). Interestingly, CuV DNA was not detected in any of the 16 malignant gut tissues. The prevalence of CuV DNA in patients of each disease group separately, or together, was somewhat lower than that among the healthy individuals, but without reaching statistical significance. All specimens were human RNaseP-PCR positive but BuV-and TuV-DNA negative. The amplification plot and quantification cycles of the CuV plasmid standards are shown in Fig. S1A and Table S1. *Copies per one million cells, based on the single-gene RNase P copies; mean of duplicate wells. BuV and TuV PCRs were negative. Patients aged 25-79 years, mean 54.7, median 57.7. na, not enough sample available; nd, not done. In all samples, RNaseP DNA copies per µl ranged between 3.6×10 4 and 1.5×10 5 and RPII RNA copies between 4.9×10 1 and 5.4×10 2 , exhibiting a melting temperature of 84.5 °C (Fig. S1B).
## CuV mRNA and RISH of intestinal biopsies
To look for potential virus activity, we searched for mRNA by RT-PCR in the four available CuV DNA-positive fresh-frozen RNAlater-submerged intestinal biopsies, but despite all being human RPII mRNA positive (Fig. S1B), they were CuV mRNA negative (Table 1). To look for viral cell tropism, we applied RISH to all five available CuV PCR-positive FFPE intestinal biopsies, but none showed CuV-specific in situ signals, likely due to the low viral loads (Fig. S2).
## CuV DNA sequence analysis
The seven 91 nt CuV VP2-amplicon sequences from six patients were similar but differed between individuals (Fig. 1). For four patients, longer, ~500 nt, CuV VP1 and VP2 sequences were successfully obtained and aligned with sequences in GenBank. Based on phylogeny, three longer VP1 sequences clustered together, whereas the corresponding three VP2 sequences clustered less closely (Fig. 2). The fourth VP2 sequence of IBD 45 was closer to the Japanese LC sequences. The BuV2-VP2 outgroup is situated closer to the CuV sequences, while the CuV-VP1 sequences formed a separate clade. We were not able to sequence longer fragments. The tree with the highest log likelihood for VP1 (-3845.91) is shown in Fig. 2(a) and for VP2 (-2432.05) in Fig. 2(b).
## Serology of the gut cohort
Only 1/13 (7.7%) serum samples was CuV IgG positive with OD 2.4 (IBD 45 with inactive UC), who also harboured CuV DNA in gut tissue. This individual was further both CuV IgM and CuV DNA negative in serum, indicating prior immunity. We found two additional individuals with elevated CuV and BuV2 IgG, with low absorbances, but the results were regarded as inconclusive based on the competitive blocking assay [21]. All sera were IgG negative for BuV1, 3 and TuV. No sera were available from the breast cohort.
## PPV and HHV DNA in malignant and benign tissue specimens of the breast cohort
No PPV DNA was detected in 94 FFPE breast and prostate tumour tissue specimens from 85 patients. Of the 79/85 patients with breast specimens still available for herpesvirus PCR, 16 (20.3%) harboured HHV-6B DNA and 4 (5.1%) HHV-7 DNA, while none were positive for HHV-6A DNA. In total, HHVs were found in tissues of 20 out of 79 (25.3%) patients (Table 2).
The HHV-6B DNA loads ranged between 2.1×10 2 and 6.7×10 6 cpm, while the HHV-7 DNA loads were 1.1×10 2 to 1.6×10 3 cpm. The occurrence of HHV DNAs, separately or together, in the malignant and benign disease groups was compared with Fisher's exact test, but the differences did not reach statistical significance in any combinations. All tissue samples contained human DNA as extraction control.
## DISCUSSION
Investigating the prevalence of newly discovered viruses across different tissue specimens is crucial to determine their association with disease and where they persist [22]. Given the association of CuV with CTCL and its precursor PP [3][4][5][6][23][24][25], it is necessary to investigate its prevalence, persistence, activity and possible tissue-distribution sites in other malignant, inflamed or other tissue types. This is the first study to elucidate the presence of CuV, BuV and TuV DNA in intestinal and breast tissues. The overall intestinal CuV DNA prevalence among all 185 patients in the gut cohort was low (3.2%), and no CuV DNA was detected in breast tissues of the 85 patients in the breast cohort. Our genoprevalences are comparable with those previously reported in skin biopsies of transplant patients (2.9%), melanomas (1.1%), head-and-neck tumours (4%) and healthy individuals (0%), which correlate to the previously reported low CuV IgG seroprevalence among healthy adults (0-7%) [4,21,[26][27][28][29], and much lower than those in skin from patients with CTCL (especially subtype mycosis fungoides; MF) (8.5-38%) and its precursor PP (38-66.7%) [3][4][5][6][23][24][25].
Notably, we did not detect CuV DNA in any malignant tumours from colorectal, breast or prostate cancer patients, even if some parallel healthy segments did harbour the virus, emphasizing its specific and significant association with the skin lymphoma.
The 91 bp CuV sequences obtained from six patients were not identical, ruling against plasmid or cross-contamination. Phylogenetic analysis of the longer CuV DNA sequences from four isolates in the VP1 region showed close clustering with CuV strains from PP patients in Finland (PP001441.1 and PP001443.1) [6]. In contrast, the VP2 sequences displayed a slightly different pattern, with one isolate clustering with strains from France and Botswana (KT868815.1 and KT868813.1, respectively) and another isolate with strains from Japan [3,25,30].
Our previous findings indicate that persistent CuV is actively transcribing in CTCL and PP tissues, as spliced viral mRNA was detected in the skin of these patients [6]. Expression of mRNA indicates active viral gene transcription in the nucleus, potentially impacting the cellular and tumour microenvironment. However, in this study, CuV mRNA was not detected in the four available CuV DNA-positive fresh-frozen tissues, raising a hypothesis that CuV activity may be specific to CTCL-MF and PP tissues. Our results thus suggest that CuV, in non-CTCL-MF tissues, persists in a dormant state, a characteristic commonly observed with other human parvoviruses [13,22]. Further studies involving diverse tissue types are necessary to confirm these observations.
No CuV-specific signals were detected by RISH in the corresponding FFPE biopsies from the five patients in the gut cohort with CuV PCR-positive fresh-frozen intestinal biopsies, possibly due to the lower sensitivity of the RISH assay compared to that of the qPCR assay. The hospital does not store serum samples from endoscopy patients; however, we were able to obtain sera from 13 patients, one of whom happened to carry CuV DNA in his intestinal mucosa. In all, 1/13 (7.7%) exhibited CuV IgG, whereas all were IgM negative, indicating a lack of acute infections. Given the limited number of serum samples, chance may have contributed to the observed seroprevalence. In previous studies, the CuV IgG prevalence has been observed to be similar among healthy adults (0-7%) compared to the slightly higher seroprevalence in CTCL and PP patients (9.5-33.3%), <65 years of age [4,6,19,21,27]. Of note, differences in age and ethnicity may affect the seroprevalence.
In an earlier study of the gut cohort, parvovirus B19 DNA persisted frequently in the intestinal mucosa of patients with colorectal cancer (50%), UC (47%), adenoma (31%) and healthy controls (27%) [13]. In our present study of the same cohort, the frequency of CuV DNA persistence was much lower than that of B19V, revealing respective prevalence rates of 0, 2.7, 2.6 and 5.5%, which are closer to the genoprevalence of HBoV DNAs in the same cohort [13]. Interestingly, similarly to HBoV DNA that was detected in only histopathologically healthy areas of the mucosa, CuV DNA was also found in the healthy or inactive mucosal regions. B19V DNA, however, persisted in both healthy and diseased tissues, but it was significantly more prevalent in the healthy colonic segments of UC patients than in the diseased ones [13]. In both the previous and the current studies, the differences in virus occurrence between each disease group and the healthy control patients did not reach statistical significance. Of the seven CuV DNA-positive samples, two (IBD 45; sigmoid colon, and IBD 155; ileum) also carried B19V DNA.
The B19V genoprevalence across various tissues ranges from 20 to 60% [31][32][33][34], whereas CuV shows a broader variability, ranging from 0 to 67%, depending on the underlying disease, although some estimates are based on relatively small sampling sizes [3-6, 23, 24, 26, 28-30, 35-37]. *Precise clinical diagnoses are found in Table S3. RNaseP copies per µl ranged between 4×10 1 and 2.2×10 4 .
Assimakopoulou et al., Journal of General 2025;106:002184
In addition to PPV DNA, the breast cohort was also tested for HHV-6 and -7, which have, similar to B19V, been shown to persist in tissues [38]. The results showed a higher prevalence of HHV-6B at 20.3%, compared to HHV-7 at 5.1% and HHV-6A at 0%. This is the first large-scale study to search for HHV-6 and HHV-7 in breast specimens, including both malignant and benign cases. Only a few studies have explored the relationship between breast cancer and HHV-6, and none of HHV-7 [16,39]; one examining if inherited chromosomally integrated HHV-6 is a breast cancer risk factor but without any supporting evidence, and the other comprising ten breast cancer control specimens, all negative for HHV-6.
In our study, the prevalence rates of both HHV-6B and HHV-7 DNA in breast tissues were slightly lower in breast cancer patients compared to those of benign breast conditions, ruling against oncogenicity, but the differences were not significant. HHV-6 and HHV-7 are pathogens that persist for life in our tissues and can reactivate. That HHV-6A and HHV-6B have been detected in various cancers does not provide direct evidence of them being oncogenic; instead, they are suggested to indirectly promote tumour cell growth alongside other viruses [16]. We did not find any tissue samples with both CuV and HHV DNAs in the same tissue.
The breast cohort comprised only FFPE samples, which may have led to increased DNA fragmentation and, consequently, lower PCR sensitivity and lower human RNase P loads compared to the fresh gut samples; however, it is comparable to the FFPE samples of CTCL and PP skin [4,5]. Moreover, although RISH has been successful in our past CTCL and PP studies [5,6], the CuV DNA loads were too low for any signals to be detected in this study. Despite these factors, the patient materials from both cohorts are valuable, particularly as we report novel findings on CuV, a lymphoma-linked virus, in malignant or inflamed and other benign intestinal and breast tissues, as well as on the pathogenic herpesviruses HHV-6 and HHV-7 in large-scale breast specimens. While the links between protoparvo-and herpesviruses and diseases remain inconclusive, screening for these DNA viruses -which persist and are implicated in cancer -provides valuable insights into the prevalence and persistence of them, laying the foundation for further studies. The human bocavirus is associated with some lung and colorectal cancers and persists in solid tumors. PLoS One 2013;8:e68020.
## References
1. Phan, Vo, Bonkoungou et al. (2012) "Acute diarrhea in West African children: diverse enteric viruses and a novel parvovirus genus" *J Virol*
2. Phan, Sdiri-Loulizi, Aouni et al. (2014) "New parvovirus in child with unexplained diarrhea" *Tunisia. Emerg Infect Dis*
3. Phan, Dreno, Da Costa et al. (2016) "A new protoparvovirus in human fecal samples and cutaneous T cell lymphomas (mycosis fungoides)" *Virology*
4. Väisänen, Fu, Koskenmies et al. (2019) "Cutavirus DNA in malignant and nonmalignant skin of cutaneous T-cell lymphoma and organ transplant patients but not of healthy adults" *Clin Infect Dis*
5. Mohanraj, Konttinen, Salava et al. (2023) "Significant association of cutavirus with parapsoriasis en plaques: high prevalence both in skin swab and biopsy samples" *Clin Infect Dis*
6. Mohanraj, Väkevä, Ranki et al. (2024) "Prevalence, tropism, and activity of cutavirus in circulating blood lymphocytes, stool, and skin biopsy specimens of patients with cutaneous T-cell lymphoma and parapsoriasis en plaques" *J Med Virol*
7. Jairath, Feagan (2020) "Global burden of inflammatory bowel disease" *Lancet Gastroenterol Hepatol*
8. De Souza, Fiocchi, De Souza (2016) "Immunopathogenesis of IBD: current state of the art" *Nat Rev Gastroenterol Hepatol*
9. Li, Wang, Zhu et al. (2007) "Detection of parvovirus B19 nucleic acids and expression of viral VP1/VP2 antigen in human colon carcinoma" *Am J Gastroenterol*
10. Pironi, Bonvicini, Gionchetti et al. (2016) "Screening of human bocavirus in surgically excised cancer specimens" *Abdel-Moneim AS*
11. Xu, Leskinen, Gritti et al. (2022) "Prevalence, cell tropism, and clinical impact of human parvovirus persistence in adenomatous, cancerous, inflamed, and healthy intestinal mucosa" *Front Microbiol*
12. Damania (2007) "DNA tumor viruses and human cancer" *Trends Microbiol*
13. Eliassen, Lum, Pritchett et al. (2018) "Human herpesvirus 6 and malignancy: a review" *Front Oncol*
14. Arivananthan, Yadav, Kumar (1997) "Detection of HHV-6 genotypes by in situ hybridization with variant-specific oligonucleotide probes" *J Virol Methods*
15. Sadeghi, Aaltonen, Hedman et al. (2014) "Detection of TS polyomavirus DNA in tonsillar tissues of children and adults: evidence for site of viral latency" *J Clin Virol*
16. Xu, Arku, Jartti et al. (2017) "Comparative diagnosis of human bocavirus 1 respiratory infection with messenger rna reverse-transcription polymerase chain reaction (PCR), DNA quantitative PCR, and serology" *J Infect Dis*
17. Chesnut, Mohanraj, Thapa et al. (2025) "In search of human protoparvovirus acute infections" *Virology*
18. Väisänen, Paloniemi, Kuisma et al. (2016) "Epidemiology of two human protoparvoviruses, bufavirus and tusavirus" *Sci Rep*
19. Väisänen, Mohanraj, Kinnunen et al. (2018) "Global distribution of human protoparvoviruses" *Emerg Infect Dis*
20. Söderlund-Venermo (2019) "Emerging human parvoviruses: the rocky road to fame" *Annu Rev Virol*
21. Kreuter, Nasserani, Tigges et al. (2018) "Cutavirus infection in primary cutaneous B-and T-cell lymphoma" *JAMA Dermatol*
22. Hashida, Nakajima, Higuchi et al. (2023) "Involvement of cutavirus in a subset of patients with cutaneous T-cell lymphoma with an unfavorable outcome" *J Clin Virol*
23. Hashida, Nakajima, Higuchi et al. (2024) "Cutavirus infection in large-plaque parapsoriasis, a premalignant condition of mycosis fungoides" *J Infect Dis*
24. Wieland, Silling, Hufbauer et al. (2019) "No Evidence for role of cutavirus in malignant melanoma" *Emerg Infect Dis*
25. Mohanraj, Jokinen, Thapa et al. (2021) "Human protoparvovirus DNA and IgG in children and adults with and without respiratory or gastrointestinal infections" *Viruses*
26. Jauhiainen, Xu, Pyöriä et al. (2021) "The presence of herpesviruses in malignant but not in benign or recurrent pleomorphic adenomas" *Tumour Biol*
27. Jauhiainen, Mohanraj, Perdomo et al. (2024) "Presence of herpesviruses, parvoviruses, and polyomaviruses in sinonasal lymphoma" *Eur Arch Otorhinolaryngol*
28. Hashida, Higuchi, Daibata (2023) "Cutavirus on the skin in an Asian cohort: identification of a novel geographically related genotype" *Virol J*
29. Söderlund-Venermo, Hokynar, Nieminen et al. (2002) "Persistence of human parvovirus B19 in human tissues" *Pathol Biol*
30. Norja, Hokynar, Aaltonen et al. (2006) "Bioportfolio: lifelong persistence of variant and prototypic erythrovirus DNA genomes in human tissue" *Proc Natl Acad Sci*
31. Adamson-Small, Ignatovich, Laemmerhirt et al. (2014) "Persistent parvovirus B19 infection in non-erythroid tissues: possible role in the inflammatory and disease process" *Virus Res*
32. Pyöriä, Toppinen, Mäntylä et al. (2017) "Extinct type of human parvovirus B19 persists in tonsillar B cells" *Nat Commun*
33. Mollerup, Fridholm, Vinner et al. (2017) "Cutavirus in cutaneous malignant melanoma" *Emerg Infect Dis*
34. Zanella, Laubscher, Docquier et al. (2021) "Unmasking viral sequences by metagenomic next-generation sequencing in adult human blood samples during steroid-refractory/ dependent graft-versus-host disease" *Microbiome*
35. Li, Zheng, He et al. (2023) "First detection of cutavirus DNA in stools of patients with rheumatic diseases in Guangzhou" *Virol Sin*
36. Pyöriä, Pratas, Toppinen et al. (2023) "Unmasking the tissue-resident eukaryotic DNA virome in humans" *Nucleic Acids Res*
37. Gravel, Dubuc, Brooks-Wilson et al. (2017) "Inherited chromosomally integrated human herpesvirus 6 and breast cancer" *Cancer Epidemiol Biomarkers Prev* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12345212&blobtype=pdf | # CDC dengue typing kit fails to detect dengue virus 2 sylvatic genotype
Diamilatou Balde, Mignane Ndiaye, Agathe Efire, Ousmane Faye, Amadou Sall, Manfred Weidmann, Oumar Faye, Idrissa Dieng
## Abstract
AUTHOR AFFILIATIONS See affiliation list on p. 3. KEYWORDS CDC dengue typing kit, DENV-2/GVI, RT-qPCR, target failure D engue is considered as one of the most prevalent arboviral threats worldwide (1). The virus exists in four (DENV-1, DENV-2, DENV-3, and DENV-4) antigenically and phylogenetically distinct forms, namely viral serotypes (2). In Africa, the virus was thought to be rare since it was first reported from the continent in the 19th century (3). The virus epidemiology is marked by circulation in two transmission cycles: the epidemic (urban) cycle and the sylvatic cycle described in West Africa and southern Asia (4).Worldwide, the epidemic cycle has been the main driver of DENV circulation and transmission in recent decades. For example, all reported dengue circulation from 2010 to 2020 in Africa was linked to urban epidemics DENV 1-4 (5), and DENV-2 has been the main serotype reported in Africa for years (6). Recently, new NGS data helped to characterize six DENV-2 genotypes at the subserotype level, including VDEN-2/GVI, namely sylvatic DENV-2 (7). Its circulation and association with outbreaks in 2000 were reported from southeastern Senegal (Kedougou), but this particular genotype has circulated in Senegal since 1970 (8). In 2020, sylvatic DENV re-emerged in an outbreak with 59 recorded cases (8, 9). Previously, the most recent case of a dengue virus linked to a sylvatic strain in Africa was reported from one DHF (Dengue Hemorraghic Fever) patient from Guinea-Bissau returning to Spain in 2009 (10).For rapid assignment of DENV serotype associated with dengue cases, many RT-qPCR systems are used globally (9, 11). In Senegal, systematic molecular serotyping of circulating DENV serotypes has been implemented since 2017 (12). This allows us to assess real-time mapping and track the changing patterns of circulating viral serotypes, which is key to public health mitigation of the dengue disease burden (5).In November 2021, while performing molecular serotyping, we noticed serotype assignation failure using the original protocol of the Center for Disease Control and Prevention (CDC) dengue typing kit (11). For this patient sample confirmed to have dengue infection from the Saré Yoba health district, located in the Kolda region (southern Senegal), amplification curves were only observed in positive controls of the CDC dengue typing kit (13). Interestingly, the patient sample was panDENV kit positive (assay targeting DENV 3′-UTR region) (14) with a Cq (cycle threshold or crossing point) value of 26.04, indicating a detectable dengue virus RNA titer in the patient serum sample and a potential gene target failure of the CDC DENV typing kit oligos. Whole genome sequencing revealed that the virus strain unable to be assigned to a serotype by the CDC kit was a DENV-2/GVI strain (13).To further investigate these preliminary findings and assess the link of this failure to serotype DENV-2/GVI, we retrospectively sampled six viral strains belonging to DENV-2/GVI and urban epidemic DENV 1-3 strains (n = 6) from the WHO collaborating center for arboviruses and hemorrhagic fever viruses at IPD (Table 1). Extracted RNA from these selected isolates (Table 1) was tested using the panDENV assay and subjected to
molecular serotyping using the CDC dengue typing kit. None of the tested DENV-2/GVI RNA samples were assigned to a serotype by the CDC dengue typing kit.
Additionally, we retrieved the genomic data corresponding to all strains tested by RT-qPCR available in GenBank. In combination with previously characterized contempo rary DENV-2/GVI sequences obtained in the Kédougou region in 2020 (Table S1), we performed an in silico analysis of DENV-2 CDC dengue typing kit oligos to evaluate signature erosion (mismatches between viral oligos sequences against binding sites of used DENV-2 strains). Accession numbers of all used sequences during in silico analysis are listed in Table S1.
Obtained results unveil hotspots of mismatches of DENV-2 CDC dengue typing kit oligos against DENV-2/GVI strains in both forward primer, reverse primer, and probe binding regions (Fig. 1). While only one mismatch was observed in the forward primer binding region, and two mismatches in the reverse primer were noticed for urban dengue sequences (Fig. 1), in DENV-2/GVI sequences, three, five, and one mismatches were observed in forward, probe, and reverse primers binding sites, respectively. These mismatches are located at critical positions known to have an adverse impact on RT-qPCR reaction efficiency (15)(16)(17). The most prominent mismatch is on the 3′ end of the forward oligonucleotide, which is known to upset efficient elongation (18).
Mismatches near the 3′ end of primers are well-documented to significantly affect qPCR efficiency and often lead to higher cycle threshold (Cq) values (19,20). The high number of mismatches (n = 5) observed in the probe target sequence likely affects the hybridization efficiency of the probe.
Altogether, these findings highlight that the CDC dengue typing kit oligos cannot identify re-emerging DENV-2/GVI. Previous studies highlighted the risk of sylvatic DENV-2 to spillover and the potential to cause epidemics. While the significance of sylvatic DENV-related disease in humans has been largely dismissed, we contend that such a conclusion is premature given the limited data available on sylvatic DENV infections in humans (21,22). It has become increasingly evident that among all viruses with the potential to jump from animal reservoirs to humans, those carried by our closest relatives, such as non-human primates, are the most likely to make this transition, b and sylvatic DENV is a prime example of such a virus (8,23). Therefore, comprehen sive prospective epidemiological and ecological studies in enzootic locations of Asia and West Africa are clearly a research priority (21). To allow real-time monitoring of DENV-2/GVI emergence and circulation, reliable diagnostic tools are needed. Since the target failure is limited to DENV-2/GVI strains, we propose a simple and cost-effective algorithm (Fig. S1) to assess the presence of DENV-2/GVI. This approach might be more cost-effective than whole genome sequencing to monitor DENV-2/GVI prevalence and cryptic circulation in dengue endemic areas (Fig. S1). Real-time monitoring of this particular DENV-2 variant is important in public health in West Africa, where extensive urban dengue epidemics are increasingly being reported (12).
## References
1. Bhatt, Gething, Brady et al. (2013) "The global distribution and burden of dengue" *Nature*
2. Katzelnick, Fonville, Gromowski et al. (2015) "Dengue viruses cluster antigenically but not as discrete serotypes" *Science*
3. Amarasinghe, Kuritsk, Letson et al. (2011) "Dengue virus infection in Africa" *Emerg Infect Dis*
4. Wang, Ni, Xu et al. (2000) "Evolutionary relationships of endemic/epidemic and sylvatic dengue viruses" *J Virol*
5. Dieng, Talla, Barry et al. (2024) "The spatiotemporal distribution and molecular characteri zation of circulating dengue virus serotypes/genotypes in Senegal from 2019 to 2023" *Trop Med Infect Dis*
6. Mwanyika, Mboera, Rugarabamu et al. (2021) "Dengue virus infection and associated risk factors in Africa: a systematic review and meta-analysis" *Viruses*
7. Harapan, Michie, Sasmono et al. (2020) "Dengue: a minireview" *Viruses*
8. Diallo, Ba, Sall et al. (2003) "Amplification of the sylvatic cycle of dengue virus type 2, Senegal, 1999-2000: entomologic findings and epidemiologic considerations" *Emerg Infect Dis*
9. Vogels, Hill, Breban et al. (2024) "DengueSeq: a panserotype whole genome amplicon sequencing protocol for dengue virus" *BMC Genomics*
10. Franco, Caro, Carletti et al. (2010) "Recent expansion of dengue virus serotype 3 in West Africa" *Euro Surveill*
11. Santiago, Vergne, Quiles et al. (2013) "Analytical and clinical performance of the CDC real time RT-PCR assay for detection and typing of dengue virus" *PLoS Negl Trop Dis*
12. Dieng, Ndione, Fall et al. (2017) "Multifoci and multiserotypes circulation of dengue virus in Senegal between" *BMC Infect Dis*
13. Dieng, Sagne, Ndiaye et al. (2021) "2022 Detection of human case of dengue virus 2 belonging to sylvatic genotype during routine surveillance of fever in Senegal" *Front Virol*
14. Wagner, De With, Huzly et al. (2004) "Nosocomial acquisition of dengue" *Emerg Infect Dis*
15. Süss, Flekna, Wagner et al. (2009) "Studying the effect of single mismatches in primer and probe binding regions on amplification curves and quantification in real-time PCR" *J Microbiol Methods*
16. Rana, Pokhrel (2020) "Sequence mismatch in PCR probes may mask the COVID-19 detection in Nepal" *Mol Cell Probes*
17. Koo, Kaur, Teh et al. (2016) "Genetic variability in probe binding regions explains false negative results of a molecular assay for the detection of dengue virus" *Vector Borne Zoonotic Dis*
18. Newton, Graham, Heptinstall et al. (1989) "Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS)" *Nucleic Acids Res*
19. Bru, Martin-Laurent, Philippot (2008) "Quantification of the detrimental effect of a single primer-template mismatch by real-time PCR using the 16S rRNA gene as an example" *Appl Environ Microbiol*
20. Ayyadevara, Thaden, Reis (2000) "Discrimination of primer 3'-nucleotide mismatch by taq DNA polymerase during polymerase chain reaction" *Anal Biochem*
21. 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" *Nat Rev Microbiol*
22. Vasilakis, Shell, Fokam et al. (2007) "Potential of ancestral sylvatic dengue-2 viruses to re-emerge" *Virology (Auckl)*
23. Vasilakis, Holmes, Fokam et al. (2007) "Evolutionary processes among sylvatic dengue type 2 viruses" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12230527&blobtype=pdf | # Correction to "Whole Genome Sequencing and Genetic Diversity of Respiratory Viruses Detected in Children With Acute Respiratory Infections: A One-Year Cross-Sectional Study in Senegal"
## Abstract
In the list of authors, the last name of the second author, "Cortaderona", is incorrect. The correct spelling is: "Cortaredona."We apologize for this error. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548419&blobtype=pdf | # An orally available 4'-fluorouridine prodrug inhibits SFTSV and LCMV infection
Xiaoqin Jian, Tianwen Hu, Huan Xu, Yuxi Wen, Yumin Zhang, Mengwei Xu, Xiaming Jiang, Junyuan Cao, Li Xiang, Jingshan Shen, Guanghui Tian, Gengfu Xiao, Leike Zhang
## Abstract
Bunyaviruses, a subset of segmented negative-sense RNA viruses, include pathogenic species capable of zoonotic transmission to humans via arthropod vectors and rodent hosts. Pathogenic bunyavirus infections can cause severe hemorrhagic fevers and other life-threatening diseases, posing threats to human health and social stability; however, therapeutic strategies for treating bunyavirus infections remain limited. Here, we report that VV251 hydrochloride salt (VV251), an optimized oral prodrug derivative of 4′-fluorouridine (EIDD-2794), exhibits potent efficacy against severe fever with thrombo cytopenia syndrome virus (SFTSV) and lymphocytic choriomeningitis virus (LCMV) both in vitro and in vivo. In various cell lines, VV251 inhibits SFTSV and LCMV with EC 50 values in the nanomolar to micromolar range. In lethal rodent models, once-daily oral administration of VV251 at low doses (10 mg/kg for SFTSV; 1 mg/kg for LCMV) achieves complete protection (100% survival), matching the efficacy of T-705 at 300 mg/kg. Additional pharmacokinetic analysis indicates that VV251 has favorable absorption and exposure profiles in both Sprague-Dawley rat and cynomolgus monkey models. This study evaluates the antiviral profile of VV251 and supports its further development as a promising therapeutic candidate. IMPORTANCE Bunyaviruses encompass numerous highly pathogenic agents that pose significant threats to human health, including the causative agents of Crimean-Congo hemorrhagic fever, Lassa fever, and Rift Valley fever. The World Health Organization has identified Lassa fever as a priority pathogen requiring urgent research and development efforts in emergency contexts, underscoring the critical need for effective oral antiviral therapies to enhance pandemic preparedness. Here, we report that VV251 hydrochloride salt (VV251), an optimized oral prodrug derivative of 4′-fluorouridine (4′-FlU, EIDD-2794), shows significant efficacy against severe fever with thrombocytopenia syndrome virus and lymphocytic choriomeningitis virus infections, with inhibitory activity in cell culture and protective effects in lethal animal models. Building on the established broad-spec trum antiviral activity of 4′-FlU against multiple high-consequence pathogens (including severe acute respiratory syndrome coronavirus 2, respiratory syncytial virus, Lassa virus, and Junin virus), VV251 emerges as a promising next-generation oral antiviral candidate, offering an orally available therapeutic option to combat these formidable pathogens. KEYWORDS severe fever with thrombocytopenia syndrome virus, lymphocytic choriomeningitis virus, 4'-fluorouridine, nucleoside analog, antiviral agent S egmented negative-sense RNA viruses include bunyaviruses, which are classified within the order Bunyavirales and exhibit a broad geographical distribution (1, 2). While most bunyaviruses are transmitted through arthropods and rodents, patho genic bunyaviruses, such as Crimean-Congo hemorrhagic fever virus, severe fever with thrombocytopenia syndrome virus (SFTSV), Hantaviruses, Lassa virus (LASV), and
lymphocytic choriomeningitis virus (LCMV), can cross species barriers to infect humans, causing clinical manifestations like leukopenia, high fever, and neurological symptoms, posing serious threats to human health and social stability (1,3,4).
Severe fever with thrombocytopenia syndrome (SFTS) is a severe tick-borne zoonotic disease caused by SFTSV infection (5). The first case was reported in China in 2009, with subsequent cases documented in Japan, South Korea, Vietnam, and Thailand (3,(6)(7)(8)(9). SFTS manifests as a potentially fatal hemorrhagic fever in humans, with a case fatality rate (CFR) reaching up to 30% (10). The expanding range of the tick reservoir Haemaphy salis longicornis facilitates the potential spread of SFTSV infection, increasing the risk of a global SFTS pandemic (11). In 2017, the World Health Organization prioritized SFTS for research and development, underscoring the urgent need to identify effective antivirals and prepare for its potential impact (12).
Currently, treatment options for SFTSV infection remain limited due to the absence of Food and Drug Administration (FDA)-approved drugs or vaccines. Ribavirin, a potential antiviral against SFTSV with half-maximal effective concentration (EC 50 ) values ranging from 3.69 to 8.72 µg/mL (13), has been reported to be insufficient for improving the outcomes of SFTS patients and shows no effect on reducing CFR in patients with a viral load exceeding 10 6 copies/mL (tested at hospital admission) in several retrospec tive studies (14,15). T-705 (favipiravir), an anti-influenza drug approved for human use in Japan and China (16), has demonstrated efficacy against SFTSV and LCMV and improves clinical outcomes in SFTS patients; however, the CFR shows no statistically significant improvement, and the risk of teratogenicity is not negligible due to its distinctive mode of action (4,17,18). Benidipine hydrochloride and nifedipine, two calcium channel blockers, were found to be effective against SFTSV infection in SFTS patients, but comprehensive clinical trials evaluating their safety and efficacy are lacking (14). Additional small molecules, including 2'-fluoro-2'-deoxycytidine (19), amodiaquine (20), hexachlorophene, and the NF-κB inhibitor SC75741, have also been reported to be effective against SFTSV in vitro (21,22), but further validation of these potential antivirals is needed.
Owing to its crucial role and structural conservation in viral replication, RNA-depend ent RNA polymerase (RdRp) is an appealing antiviral target (23). Consequently, nucleo side analogs represent the most extensive class of small molecule-based antivirals, as they inhibit viral polymerases and exert antiviral functions (24). In recent years, nucleoside analogs such as zidovudine for acquired immune deficiency syndrome, entecavir and sofosbuvir for hepatitis B and hepatitis C, respectively, and acyclovir for herpes simplex have demonstrated efficacy in treatment (25)(26)(27)(28). During the COVID-19 pandemic, remdesivir became the first nucleoside analog approved by the FDA for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) treatment (29). The aforementioned examples affirm that RdRp is a reliable viral target for drug develop ment.
4′-Fluorouridine (4′-FlU, EIDD-2794), a promising uridine analog, has shown efficacy against multiple viruses, including respiratory syncytial virus (RSV) (30), SARS-CoV-2, avian influenza virus (31), chikungunya virus (32), enteroviruses (33), and SFTSV (12). A recent study also reported its efficacy against highly pathogenic arenaviruses LASV and Junin virus (JUNV) (34), suggesting the broad-spectrum utility of 4′-FlU as an orally available antiviral drug. However, the chemical instability of 4′-FlU hinders its further development, as determined by high-performance liquid chromatography analyses, which revealed the degradation of 4′-FlU in acidic buffers at 30°C within 7 days, potentially limiting its therapeutic utility (35)(36)(37). Building on our previous screening of 4′-FlU prodrug candidates (38), we conducted a systematic evaluation of VV251 hydrochloride (VV251), a novel double prodrug derivative of 4′-FlU, for its antiviral activity against both SFTSV and LCMV. VV251 features three acetyl modifications on the ribose moiety and a nicotinoyloxymethyl group at the N3 position of the uracil base, structural optimizations specifically designed to enhance oral bioavailability while maintaining potent antiviral efficacy. Comparative analysis revealed that VV251 exhibits stronger in vitro antiviral activity against SFTSV and LCMV infections than T-705, as evidenced by lower EC 50 values under identical conditions. In addition to cell-based assays, VV251 exhibits potent antiviral activity and safety in lethal rodent models of SFTSV and LCMV infection. This study establishes the antiviral profile of VV251, support ing it as a promising antiviral candidate with demonstrated oral efficacy against lethal SFTSV and LCMV infections in rodent models, which may expand limited therapeutic options for bunyaviruses and reduce the disease burden they cause.
## RESULTS
## In vitro antiviral activities of VV251 against SFTSV and LCMV
To investigate the effectiveness of VV251 at the cellular level, we first evaluated the inhibitory activity of VV251 against SFTSV and LCMV in different cell lines, with parallel comparisons to its parent compound 4′-FlU and T-705, the latter of which exhibits documented antiviral activity against both viruses (4,17).
Cells were pretreated with different concentrations of the test compounds for 1 hour, followed by infection with SFTSV or LCMV at various multiplicities of infection (MOI) based on the growth properties of each cell line. At 48 hours post-infection, viral RNA copies in the cellular supernatant were measured using quantitative real-time PCR (qRT-PCR) to determine the half-maximal effective concentration values.
The results revealed distinct antiviral efficacy profiles among the tested compounds. VV251 demonstrated potent, broad-spectrum inhibition of both SFTSV and LCMV replication in all cell lines evaluated, with EC 50 values closely matching those of its parent compound 4′-FlU (range: 0.65-5.16 μM vs 2.03-2.37 μM for SFTSV; 0.08-0.16 μM vs 0.04-0.15 μM for LCMV), consistent with their shared active metabolite 4'-fluorouridine triphosphate (4′-FlU-TP) (Fig. 1A through F). Notably, VV251 exhibited better broad-spec trum antiviral activity compared to T-705, demonstrating lower EC 50 values across all tested cell lines (Fig. 1A through F). Against SFTSV, VV251 showed 54-fold (A549: 5.16 vs 279.82 µM), 4.8-fold (Huh-7: 0.65 vs 3.12 µM), and 7.0-fold (Vero: 2.96 vs 20.63 µM) greater potency than T-705. The potency advantage was even more pronounced against LCMV, with 849.6-fold (A549: 0.08 vs 67.97 µM) and 632.9-fold (Vero: 0.14 vs 88.60 µM) lower EC 50 values for VV251 (Fig. 1G). These results position VV251 as a substantially more potent antiviral candidate against both bunyaviruses. The efficacy test of T-705 was not performed in BHK-21 cells since it cannot be converted into its ribonucleoside 5'-mono phosphate by hypoxanthine guanine phosphoribosyl transferase; thus, its ribonucleoside 5'-diphosphate and effective active ribonucleoside 5'-triphosphate metabolites cannot be formed (39,40).
Cytotoxicity testing revealed that the half-maximal cytotoxic concentration (CC 50 ) values of VV251, 4′-FlU, and T-705 on different cell lines exceeded 500 µM (Fig. S1). Based on EC 50 values measured by viral RNA copies, these drugs exhibited a wide range of selectivity index (SI, CC 50 /EC 50 ). For SFTSV, SIs of VV251, 4′-FlU, and T-705 ranged from 96.94 to 384.25, from 210.67 to 245.98, and from 1.79 to 160.32, respectively. For LCMV, SIs of VV251, 4′-FlU, and T-705 ranged from 3177.56 to 6559.96, from 3389.68 to 12697.20, and from 5.64 to 7.36, respectively (Table S1). The cytotoxicity test of T-705 was not performed in BHK-21 cells for the same reason mentioned for the efficacy test. Taken together, these findings demonstrate VV251's favorable safety profile, supporting its further investigation as a promising therapeutic candidate.
## VV251 acts as a pyrimidine analog inhibiting SFTSV and LCMV at the postinfection stage
To preliminarily investigate the mechanism of action of VV251, we performed a nucleo side competition assay. As the results showed, the antiviral effects of VV251 against both SFTSV and LCMV were competitively reversed by supplementation with exogenous uridine and cytidine (Fig. 2A andB), supporting that VV251 functions as a pyrimidine analog.
To further characterize its antiviral mechanism, we conducted time-of-addition experiments in which SFTSV-or LCMV-infected cells were treated with VV251 according to the schematic (Fig. 2C). T-705 was included as a positive control, given its wellcharacterized activity as a viral RNA polymerase inhibitor targeting the replication stage (41). Consistent with the positive control T-705, VV251 exhibited potent antiviral activity against both SFTSV and LCMV under post-infection and full-time treatment conditions (Fig. 2D andE). In contrast, pretreatment with VV251 showed minimal to no protective effect, suggesting that its mechanism of action is primarily effective during active viral replication rather than through preemptive viral entry blockade.
## Anabolism and pharmacokinetic profiling of VV251
As an ester prodrug, VV251 is rapidly converted to its parent nucleoside 4′-FlU in the intestine and liver by various esterases, which then distributes into the bloodstream and cells and is ultimately metabolized into 4′-FlU-TP via catalysis by a series of kinases (Fig. 3A). Prior to evaluating the in vivo antiviral efficacy of VV251, we performed pharmacokinetic (PK) studies in Sprague-Dawley (SD) rats and cynomolgus monkeys following oral administration to compare its systemic exposure with equimolar 4′-FlU and establish appropriate dosing regimens for subsequent efficacy studies. Initial PK studies were conducted in SD rats, with the oral dosing regimen optimized based on our previous experimental findings (38). The PK analysis revealed that oral administration of VV251 (21 mg/kg) achieved comparable systemic exposure to equimolar 4′-FlU, with mean plasma concentrations (C max ) reaching 1,598 ng/mL and demonstrating a plasma exposure (area under the curve [AUC 0-t ]) of 9,714 ng•h/mL. These parameters closely matched those observed for 4′-FlU itself (C max : 1,369 ng/mL; AUC 0-t : 7,277 ng•h/mL; Fig. 3B). To further validate these findings in a more clinically predictive model, we extended the investigation to cynomolgus monkeys, which represent the gold standard for nonclinical safety assessment due to their superior pharmacological and metabolic relevance to humans (42). Consistent with the data from SD rats, VV251 (10.7 mg/kg) maintained similar PK characteristics to 4′-FlU (C max : 2,358 vs 1,505 ng/mL; AUC 0-t : 14,922 vs 10,687 ng•h/mL; Fig. 3C), confirming its favorable absorption and exposure profiles.
The improved absorption of VV251 compared to 4′-FlU is attributed to its optimized molecular structure. In a previous study, we found that modifying 4′-FlU by attaching three isobutyryl groups to the ribose moiety and a nicotinoyloxymethyl group linked to the imide-nitrogen on the base moiety could improve its chemical stability and enhance its membrane permeability and metabolic stability. Meanwhile, VV251 in its free base Full-Length Text form was observed to exhibit favorable stability and PK properties (38). In this study, the hydrochloride salt form of VV251 was selected for PK evaluation, given its favorable aqueous solubility, which may enhance oral bioavailability.
## VV251 is orally efficacious against SFTSV in the A129 model
To evaluate the antiviral potential of VV251 under physiologically relevant conditions, we conducted systematic in vivo efficacy studies against SFTSV infection by using IFNAR1 knockout mice (A129) model, which is well known for its hyper-susceptibility to SFTSV infection (12). T-705 was chosen as the positive control due to its reported anti-SFTSV activity in a clinical setting (17).
In the efficacy studies, the mice were intraperitoneally infected with 1 × 10 3 PFU of SFTSV or an equivalent volume of Dulbecco's modified Eagle's medium (DMEM) as mock infection. Subsequently, vehicle, T-705 (300 mg/kg), and VV251 (at doses of 1, 5, and 10 mg/kg) were orally administered via gavage for seven consecutive days in a once-daily regimen starting 1 hour post-infection. The mice were anesthetized and sacrificed 2 days post-infection to obtain organ samples (liver, spleen, lung, and kidney) and blood samples. Additionally, another set of equally treated mice was monitored for changes in body weight and survival throughout the experiment (Fig. 4A).
All animals treated with VV251 at the 10 mg/kg dose and T-705 at the 300 mg/kg dose survived. In the VV251 (10 mg/kg) group, the body weight of the mice remained stable, whereas those in the T-705 group exhibited slight initial weight loss followed by weight recovery at 5 days post-infection (Fig. 4B). Notably, four mice treated with the vehicle and one administered with the intermediate VV251 dose (5 mg/kg) died 3 days postinfection, and the remaining mice in the vehicle-treated group died at 4 days postinfection (Fig. 4C). Moreover, the mice receiving the lowest VV251 dose (1 mg/kg) exhibited prolonged survival compared with those in the vehicle group, but five of them died from infection by day 5, and a significant reduction in body weight was observed, with only one mouse surviving until the end of the study (Fig. 4B andC). Overall, compared with T-705, once-daily treatment with 10 mg/kg of VV251 proved highly efficacious in protecting mice from SFTSV infection, along with maintaining relatively stable body weight.
Compared with vehicle treatment, all doses of VV251 and T-705 led to significant reductions in the virus load in the liver, spleen, lung, and kidney, as assessed via qRT-PCR (Fig. 4D) and immunoplaque assays (Fig. 4E), respectively. Notably, while qRT-PCR analysis revealed no statistically significant reduction of viral RNA copies in the spleen following T-705 treatment compared to vehicle control, immunoplaque assays demonstrated marked decreases in infectious viral titers. This apparent discrepancy likely stems from T-705's unique mechanism of action as a viral mutagen. Unlike conven tional antivirals that directly inhibit viral replication, T-705 induces lethal mutagenesis in viral genomes. This distinct pharmacological action preferentially reduces infectivity rather than viral RNA quantity, explaining the observed dissociation between qRT-PCR measurements (reflecting total viral genetic material) and immunoplaque assay results (measuring functional viral particles) (17).
To gain better insight into the impact of VV251 on mitigating pathogenesis, histopathological analysis of the tissues was conducted. The vehicle-treated group developed pronounced splenomegaly, which was significantly ameliorated by all tested doses of both VV251 and T-705 (Fig. 4F andG). This improvement correlated with substantially reduced histopathology scores (Fig. 4G) and a dose-dependent decrease in SFTSV antigens distribution within the spleen (Fig. 4H). Hematoxylin and eosin (H&E) staining further demonstrated that VV251 and T-705 treatments effectively attenuated SFTSV-induced splenic damage, including loss of white-red pulp demarcation, perivascu lar leukocyte infiltration, and alveolar septal thickening (Fig. 4I). This protective effect extended to pulmonary tissues, where VV251 and T-705 treatments showed improved histoarchitecture and significantly lower pathology scores (Fig. S2A andB).
Furthermore, ethylenediaminetetraacetic acid (EDTA)-treated whole blood was extracted to assess hematological parameters and the plasma cytokine profile, as hemorrhagic fever can lead to a dysregulated inflammatory response and cytokine storms (43). The viremia caused by infection led to severe leukopenia, lymphocytopenia, and a reduction in the level of platelet count, which were all relieved by T-705 and VV251 treatment (Fig. S3A through D). Notably, the concentrations of interferon-γ (IFN-γ), tumor necrosis factor-α, interleukin-1β, and interleukin-6 rapidly increased in the plasma after viral infection in the vehicle-treated group, whereas treatment with all dosages of VV251 and T-705 significantly reduced the levels of these cytokines, potentially effectively preventing the formation of cytokine storms (Fig. S3E through H).
The results validated the effectiveness of VV251 against SFTSV, as demonstrated by decreased viral load, improved histopathology, reduced viremia, and attenuated proinflammatory responses. Furthermore, once-daily oral administration of VV251 at a dosage of 10 mg/kg in mice showed comparable efficacy and safety to that of T-705 at a dosage of 300 mg/kg. VV251 is effective against LCMV in the C57BL/6-Prf1 tm1Sdz /J model
To establish the broad-spectrum antiviral potency of VV251, we evaluated its efficacy against LCMV in a C57BL/6-Prf1 tm1Sdz /J mice model, which is known for effective LCMV infection establishment (44). The mice were either infected with 2 × 10 5 PFU LCMV or received an equal volume of DMEM as mock infection. Following this, they underwent gavage administration of either the vehicle, T-705 (300 mg/kg), or VV251 (at doses of 1, 5, and 10 mg/kg) starting 1 hour post-infection, once daily for seven consecutive doses. At 5 days post-infection, organs (liver, spleen, lung, and kidney) and blood samples were extracted from equally treated sets of animals for detection (Fig. 5A).
Compared with those in the mock group, the body weights of the mice treated with all doses of VV251 remained stable (Fig. 5B). Notably, even the lowest dose of VV251 (1 mg/kg) ensured the survival of the mice. In contrast, all of the mice in the vehicle group succumbed to the infection within 9 days post-infection (Fig. 5C). Compared with VV251, T-705 at 300 mg/kg provided less protection against LCMV infection, as evi denced by the parallel body weight with that of the vehicle-treated group, the early mortality starting at 7 days post-infection, and no survivors remaining by the end of the study (Fig. 5C).
Viral RNA copies and titers in the liver, spleen, lung, and kidney obtained on day 5 post-infection were assessed via qRT-PCR (Fig. 5D) and immunoplaque assay (Fig. 5E), respectively. The findings revealed that VV251 treatment significantly decreased both the number of viral RNA copies and the titer in the liver, spleen, lung, and kidney, and the number of viral RNA copies was more than one order of magnitude lower than that in the vehicle group (Fig. 5D andE). In contrast, T-705 administration reduced the viral titer in the tissues, whereas no reduction in the number of viral RNA copies was observed except in the lungs, probably due to the T-705 unique mechanism of action (Fig. 5D andE).
Additionally, VV251 treatment ameliorated spleen damage efficiently, as evidenced by the normal size of the spleen and lower pathology score (Fig. 5F andG). H&E staining also revealed a clear demarcation between the white and red pulps and the preservation of the marginal zone in the spleen in the VV251-treated groups, whereas the vehicle-and T-705-treated groups presented marked pathological lesions (Fig. 5H).
The analysis of hematological parameters in EDTA-treated whole blood indicated that the oral administration of VV251 ameliorated viremia induced by LCMV infection. In contrast, the white blood cell, platelet, and lymphocyte count of the T-705-treated group were similar to those of the vehicle-treated group (Fig. S4A through D). Additionally, we conducted an analysis of plasma cytokine levels, considering the potential role of cytokine storms in LCMV pathogenicity. Compared with vehicle treatment, treatment with VV251 at all dosages led to a statistically significant reduction in the expression of IFN-γ and granulocyte colony-stimulating factor, whereas T-705 treatment at 300 mg/kg had no significant effect (Fig. S4E andF). Therefore, the reduced cytokine response may be a crucial factor contributing to the protective effect of VV251 treatment against lethal LCMV infection.
On the basis of the preceding results, VV251 at the lowest dose of 1 mg/kg has more potent anti-LCMV effects than T-705 at 300 mg/kg does, as it can protect C57BL/6-Prf1 tm1Sdz /J model mice from lethal LCMV infection while maintaining their body weight and survival rate. VV251 treatment not only reduced the viral load after infection but also induced altered histopathological changes in multiple organs.
## DISCUSSION
Due to the lack of FDA-approved vaccines or specific drugs, clinical treatments for bunyavirus infections are limited, threatening socioeconomic stability and human health. Effective oral antiviral agents are needed to treat outpatients with risk factors for developing severe symptoms. Recently, the uracil analog 4′-FlU has been reported to effectively inhibit Heartland virus (45), SFTSV (12), LASV, and JUNV (34), which all belong to the order Bunyavirales. However, further studies demonstrated that 4′-FlU exhibits poor chemical stability in aqueous solutions, especially under acidic conditions (pH <4) (37, 46), likely due to the glycosidic bond's increased susceptibility to cleavage in such acidic environments, thus hindering its further development as a viable drug candidate.
To address these limitations, we developed VV251, an orally bioavailable prodrug and structurally optimized pyrimidine nucleoside analog of 4′-FlU, designed to enhance systemic delivery of the parent nucleoside following oral administration. Our preliminary investigations revealed that VV251's free base form demonstrated better stability and pharmacokinetic profiles compared to 4′-FlU, whereas crystal lization difficulties and formulation stability concerns precluded its advancement as a lead candidate (38). Drawing upon the successful structural optimization approach employed in VV116's development (47), we subsequently prepared VV251 hydrochloride salt, which exhibited improved solid-state characteristics and aqueous solubility. These favorable physicochemical properties qualified the hydrochloride salt as an alternative development candidate, prompting further evaluation of its pharmacodynamic properties.
In this study, we used VV251 hydrochloride salt to comprehensively evaluate its antiviral profile, demonstrating its broad-spectrum activity against both SFTSV and LCMV in diverse cell culture systems. PK characterization in SD rats and cynomolgus monkeys revealed VV251's favorable absorption profile and systemic exposure compared to equimolar 4′-FlU, establishing an optimized dosing regimen for efficacy studies. Using T-705 as a benchmark antiviral agent with established efficacy against both viruses (4,17), we evaluated VV251's in vivo therapeutic potential. To our surprise, once-daily oral administration of VV251 at lower doses (10 mg/kg for SFTSV; 1 mg/kg for LCMV) achieved complete protection (100% survival), matching the efficacy of T-705 at 300 mg/kg. Notably, VV251 treatment conferred better protective effects against LCMV infection compared to T-705, as evidenced by better maintenance of body weight and attenuated histopathological damage at lower doses.
For optimal pandemic preparedness, antiviral drugs targeting highly conserved viral proteins, particularly RNA-dependent RNA polymerase, offer distinct advantages due to their potential efficacy against emerging viral threats. The strategy of employing nucleoside analogs to inhibit viral RdRp has been successfully utilized for over three decades, with well-documented clinical success (48)(49)(50). Our findings demonstrate that VV251, building upon the well-documented antiviral efficacy of its parent compound 4′-FlU, shows promising potential for treating bunyavirus infections but may also possess broad-spectrum antiviral activity against other viruses, while dose optimization may be required when extending its application beyond bunyaviruses.
Despite the observed efficacy of VV251 in both in vitro and in vivo settings, our study still possessed limitations. On the one hand, the breadth of protection was only tested in mice, and additional animal models are needed for overall evaluation. On the other hand, the specific mechanism of action of VV251 requires further investigation since nucleoside analogs can target the viral RdRp to impede viral replication through three primary mechanisms: delayed chain termination (e.g., remdesivir), obligate chain termination (e.g., sofosbuvir), and mutagenicity (e.g., molnupiravir) (24,51,52). Previous studies have reported that the incorporation of triphosphorylated 4′-FlU by the RdRps of RSV and SARS-CoV-2 results in sequence-modulated transcriptional stalling (30), and further investigation is needed to explore the mechanism of action of VV251 against SFTSV and LCMV. Furthermore, to provide valuable insights into the molecular basis of VV251, the resistance profile of VV251 against bunyaviruses needs further investigation; however, no evidence of breakthrough variants of JUNV was observed after more than 18 passages under 4′-FlU selective pressure (34), which indicates that VV251 may possess the potential for a high genetic barrier of viral resistance since it functions in the form of a 4′-FlU-TP.
Effective oral antiviral therapies that target conserved viral proteins and have a high barrier to resistance are crucial for maximizing therapeutic efficacy against emerging viruses. Our findings establish VV251 as a potential orally efficacious inhibitor of SFTSV and LCMV, positioning it as a promising candidate for the treatment of bunyavirus infections. Building on the success of 4′-FlU against various highly pathogenic viruses, VV251 holds promise as a broad-spectrum antiviral drug available orally to combat these pathogens.
## MATERIALS AND METHODS
## Cells and viruses
A549, BHK-21, Vero, and Huh-7 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (FBS) and 1% penicil lin-streptomycin antibiotics at 37°C in a humidified 5% CO 2 incubator. SFTSV (strain HBMC16_human_2015) was obtained from the China Center for General Virus Culture Collection. LCMV (strain Armstrong) was rescued using a T7 polymerase system as previously described (53,54).
## Compound sourcing
VV251, 4′-FlU, and T-705 were obtained from Vigonvita Shanghai Co., Ltd. For in vitro studies, all compounds were initially dissolved in dimethyl sulfoxide (DMSO) and subsequently diluted in maintenance medium. For in vivo studies, VV251 and 4′-FIU were formulated in a solution of 5% DMSO, 5% ethanol, 40% PEG400, and 50% saline. T-705 was formulated in 0.4% carboxymethyl cellulose (CMC).
## Determination of antiviral activity in vitro
To perform antiviral assays, monolayer cells (A549, Huh-7, and Vero cells for SFTSV infection; A549, BHK-21, and Vero cells for LCMV infection) were plated into 48-well plates overnight. After 16 hours, the medium was replaced with 200 µL of DMEM (2% FBS) per well containing compounds at specific concentrations within several gradients, and the mixture was incubated for 1 hour. Next, the corresponding virus (MOI = 0.1 for SFTSV; MOI = 0.01 for LCMV) was added. At 48 hours post-infection, viral RNA copy numbers in the cellular supernatant were detected to determine the EC 50 values by using quantitative real-time PCR. The inhibition rates of the compounds were calculated on the basis of the number of viral copies normalized to that of vehicle-treated (DMSO) controls as the following formula:
Inhibition rate = [1 -(compound-treated/DMSO-treated viral RNA copies)] × 100%
The EC 50 values were calculated using nonlinear regression in Prism 9.5.1 software. The bar indicates the SD from at least three independent experiments.
## RNA extraction and quantitative real-time PCR
RNA was extracted from the cell supernatants with a DNA/RNA Extraction Kit (Vazyme, China), according to the manufacturer's instructions. qRT-PCR was performed according to the instructions of the HiScript II One Step qRT-PCR SYBR Green Kit (Vazyme, China) to determine the viral RNA copy number.
## Immunoplaque assay
To determine the viral titers of SFTSV or LCMV, Vero and BHK-21 cells were incubated in 24-well plates separately. After 16 hours, the cells were incubated with 200 µL of 10-fold serial dilutions of viral supernatants in serum-free DMEM for 1 hour, after which the supernatants were discarded, and the cells were washed with phosphate buffer saline (PBS) three times. Next, 1 mL of DMEM containing 1.1% sodium CMC and 2% FBS was added to each well. After incubation for 3-4 days, the cells were fixed with 4% paraformaldehyde in PBS overnight at 4°C, permeabilized, and blocked in 5% nonfat milk containing 0.2% Triton X-100 (diluted in PBS) for 1 hour. The cells were subsequently incubated with anti-SFTSV-nucleocapsid protein (NP) serum (1:16,000 dilution) or anti-LCMV-nucleocapsid protein serum (1:300 dilution) for 2 hours at room temperature. The plates were subsequently washed in PBS containing 0.05% Tween 20 three times, followed by incubation with a horseradish peroxidase (HRP)-conjugated anti-rabbit antibody (1:1,000 dilution) for 1 hour at room temperature. The viral plaques were finally stained using an enhanced HRP-DAB Chromogenic Substrate Kit (TIANGEN, PA110), and the viral titers were analyzed.
## Cytotoxicity assay
For cell viability experiment, A549, BHK-21, Vero, and Huh-7 cells were plated in 96-well plates, with each concentration tested in triplicate. Compounds were diluted twofold across eight gradients starting at 500 µM in maintenance medium (DMEM + 2% FBS). After 48 hours of incubation, the supernatants were removed, and 100 µL of diluted Cell Counting Kit-8 (Beyotime) in maintenance medium was added to the cells. Plates were incubated at 37°C for 2 hours, and absorbance was measured at 450 nm using a spectrophotometer (BioTek). Cell viability was then calculated. All compounds were diluted in maintenance medium containing 2% FBS.
## Nucleotide competition experiment
Vero and BHK-21 cells infected with SFTSV or LCMV were exposed to 10 µM VV251 in combination with 0.5-1,000 µM exogenous nucleosides (MedChemExpress) at the onset of infection. An equal volume of DMSO served as a vehicle control. At 48 hours post-infection, RNA was extracted from the cell supernatants, and the number of viral RNA copies was measured using qRT-PCR. The values are expressed relative to the values for the vehicle-treated samples. Three independent experiments were performed.
## Time-of-addition experiment
A time-of-addition assay was performed, using different treatment patterns, which were divided into three groups. Pretreatment of the cells: the cells were pretreated with VV251 or T-705 for 1 hour and incubated with the virus for 1 hour. Then, the virus was removed, and the medium was replaced with fresh medium without compounds. Post-infection treatment: the cells were infected with the virus for 1 hour, and fresh medium containing VV251 or T-705 was added after the removal of the virus. Full-time treatment: the cells were pretreated with VV251 or T-705 for 1 hour, the virus was added for infection for 1 hour, and the medium was replaced with fresh medium containing VV251 or T-705. All of the cell supernatants were collected 47 hours later for detection.
## Pharmacokinetic studies of VV251 in SD rats and cynomolgus monkeys
Pharmacokinetic studies of VV251 and 4′-FlU in SD rats and cynomolgus monkeys were conducted at Function Biomedical Co., Ltd. Six SD rats (n = 3 for each group) were randomly allocated into two groups and fasted for 12 hours before receiving an oral dose of VV251 at 21 mg/kg and 4′-FlU at 10 mg/kg. The vehicle for oral administration of the test compounds consisted of 5% DMSO, 5% ethanol, 40% PEG400, and 50% saline. Blood samples were collected at 0.25, 0.5, 1, 2, 4, 6, 8, and 24 hours post-dosing. The samples were collected in EDTA-K2 tubes, gently mixed, and then centrifuged at 4°C and 2,000 × g for 10 minutes. After centrifugation, the separated plasma samples were frozen at -70°C until analysis, which was conducted via LC-MS/MS. The method of pharmacokinetic testing in cynomolgus monkeys was similar to that in SD rats, with oral doses of VV251 and 4′-FlU at 10.7 and 5 mg/kg, respectively.
## Studies of the anti-SFTSV efficacy of VV251 in A129 mice
Six-to 8-week-old A129 mice were randomized into six groups: the mock group, the vehicle group, the group receiving 300 mg/kg T-705, and the groups receiving VV251 at 1, 5, and 10 mg/kg. After being anesthetized by isoflurane inhalation, the mice were infected with SFTSV (1 × 10 3 PFU) diluted in 200 µL of DMEM via intraperitoneal injection, whereas the mock group was injected with an equal volume of DMEM. At 2 days post-infection, tissues (liver, spleen, lung, and kidney) and blood samples were harvested for efficacy determination (n = 6 per group), and another set of equally treated mice was monitored for changes in body weight and survival throughout the experiment (n = 6 per group). The tissues were divided into two parts, which were either placed in tissue homogenization tubes for viral copy number and titer detection or fixed in 4% parafor maldehyde for H&E and immunofluorescence staining for the detection of the SFTSV antigen (anti-SFTSV-NP serum). The image information was collected using a Pannoramic MIDI system (3DHISTECH, Budapest) and FV1200 confocal microscopy (Olympus). At the end of the study (14 days post-infection or once the mice lost 20% of their body weight), the remaining mice in each group were sacrificed after anesthetization.
## Studies of the anti-LCMV efficacy of VV251 in C57BL/6-Prf1 tm1Sdz /J mice
Six-to 10-week-old C57BL/6-Prf1 tm1Sdz /J mice were randomized into six groups: the mock group, the vehicle group, the group receiving T-705 at 300 mg/kg, and the groups receiving VV251 at 1, 5, and 10 mg/kg. After being anesthetized by isoflurane inhalation, the mice were infected with LCMV (2 × 10 5 PFU) diluted in 200 µL of DMEM via intraperi toneal injection, whereas the mock group was injected with an equal volume of DMEM. On day 5, tissues (liver, spleen, lung, and kidney) and blood samples were harvested for efficacy determination (n = 6 per group), and another set of equally treated mice was monitored for changes in body weight and survival throughout the experiment (n = 6 per group). The tissues were divided into two parts, which were either placed in tissue homogenization tubes for viral copy number and titer detection or fixed in 4% paraformaldehyde for H&E staining. At the end of the study (14 days post-infection or once the mice lost 20% of their body weight), the remaining mice in each group were sacrificed after anesthetization.
## Titration of virus from excised organs
Organs for viral copy number and titer detection were weighed and homogenized in tissue homogenization tubes with 1 mL of DMEM. Next, the homogenates were centrifuged at 5,000 rpm for 15 minutes, and the viral supernatants were collected. Viral copy number and titer detection were performed via qRT-PCR and immunoplaque assay, respectively.
## Hematology and cytokine level analysis
Hematological parameters were analyzed using EDTA-treated whole blood with an automatic blood cell analyzer for animal use (Tecom). The cytokine levels in the plasma of the mice in each group were detected via an ABplex Mouse 6-Plex Custom Panel (Abclonal) according to the manufacturer's instructions.
## Hematoxylin and eosin staining and pathology score
Tissues fixed with 4% paraformaldehyde were embedded in paraffin according to standard procedures. Then, the embedded tissues were sectioned at 4 µm for staining with hematoxylin and eosin. The pathology score of the lungs is based on the infiltration of inflammatory cells in the alveoli and interstitium, thickening of the alveolar wall, alveolar and interstitial bleeding, and fibrin exudation in the alveolar cavity. A pathology score range of 0-3 was used to evaluate the extent of spleen damage, based on the area of splenic necrosis, the number of hemosiderin-containing cells in the field of view, and the dispersion of splenic nodules; while a range of 0-4 was used for lung damage, based on alveolar and interstitial inflammatory cell infiltration, alveolar septal thickening, alveolar and interstitial hemorrhage, and fibrinous exudate in alveolar spaces. Scores for each animal represent the sum of damage degrees in each spleen or lung and are presented as mean ± SD.
## Statistical analysis
Data from the cell and animal studies were compiled in Microsoft Excel and analyzed using GraphPad Prism 9.5.1 software. The data are presented as the mean ± SD. Significance was determined via one-way ANOVA compared with the vehicle-treated group, ****P < 0.0001; ***P < 0.001; **P < 0.01; *P < 0.05; ns, not significant.
## References
1. Boshra (2022) "An overview of the infectious cycle of bunyaviruses" *Viruses*
2. Aparicio, López, Felt (2022) "A virus is a community: Diversity within negative-sense RNA virus populations" *Microbiol Mol Biol Rev*
3. Cui, Liu, Lu et al. (2015) "Severe fever with thrombocytopenia syndrome bunyavirus-related human encephalitis" *J Infect*
4. Wen, Xu, Shang et al. (2022) "Visualizing lymphocytic choriomeningitis virus infection in cells and living mice"
5. Yu, Liang, Zhang et al. (2011) "Fever with thrombocytopenia associated with a novel bunyavirus in China" *N Engl J Med*
6. Matsumoto, Shinohara, Furuta et al. (2018) "Investigation of antibody to severe fever with thrombocytopenia syndrome virus (SFTSV) in blood samples donated in a SFTS-endemic area in Japan" *Vox Sang*
7. Park, Cho (2014) "National infectious diseases surveillance data of South Korea" *Epidemiol Health*
8. Rattanakomol, Khongwichit, Linsuwanon et al. (2019) "Severe fever with thrombocytopenia syndrome virus infection" *Emerg Infect Dis*
9. Tran, Yun, Van An et al. (2019) "Endemic severe fever with thrombocytopenia syndrome" *Vietnam. Emerg Infect Dis*
10. Li, Lu, Xing et al. (2018) "Epidemiological and clinical features of laboratory-diagnosed severe fever with thrombocytopenia syndrome in China, 2011-17: a prospective observational study" *Lancet Infect Dis*
11. Liu, He, Huang et al. (2014) "Severe fever with thrombocytopenia syndrome, an emerging tick-borne zoonosis" *Lancet Infect Dis*
12. Xu, Jian, Wen et al. (2024) "A nanoluciferase SFTSV for rapid screening antivirals and real-time visualization of virus infection in mice" *EBioMedicine*
13. Lee, Kim, Yi et al. (2017) "In vitro antiviral activity of ribavirin against severe fever with thrombocy topenia syndrome virus" *Korean J Intern Med*
14. Li, Zhang, Li et al. (2019) "Calcium channel blockers reduce severe fever with thrombocytopenia syndrome virus (SFTSV) related fatality" *Cell Res*
15. Liu, Lu, Cui et al. (2013) "Case-fatality ratio and effectiveness of ribavirin therapy among hospitalized patients in china who had severe Full-Length Text Journal of Virology October"
16. "fever with thrombocytopenia syndrome" *Clin Infect Dis*
17. (2024) "Study Details | Phase 3 Efficacy and Safety Study of Favipiravir for Treatment of Uncomplicated Influenza in Adults | ClinicalTrials"
18. Li, Jiang, Cui et al. (2021) "Clinical effect and antiviral mechanism of T-705 in treating severe fever with thrombocytopenia syndrome" *Sig Transduct Target Ther*
19. Sangawa, Komeno, Nishikawa et al. (2013) "Mechanism of action of T-705 ribosyl triphosphate against influenza virus RNA polymerase" *Antimicrob Agents Chemother*
20. Smee, Jung, Westover et al. "2018. 2'-Fluoro-2'deoxycytidine is a broad-spectrum inhibitor of bunyaviruses in vitro and in phleboviral disease mouse models" *Antiviral Res*
21. Baba, Toyama, Sakakibara et al. (2017) "Establishment of an antiviral assay system and identification of severe fever with thrombocytopenia syndrome virus inhibitors" *Antivir Chem Chemother*
22. Yuan, Chan, Ye et al. (2019) "Screening of an FDA-approved drug library with a two-tier system identifies an entry inhibitor of severe fever with thrombocytopenia syndrome virus" *Viruses*
23. Mendoza, Yamaoka, Tsuda et al. (2021) "The NF-κB inhibitor, SC75741, is a novel antiviral against emerging tick-borne bandaviruses" *Antiviral Res*
24. Barik (2022) "Inhibition of viral RNA-dependent RNA polymerases by nucleoside inhibitors: an illustration of the unity and diversity of mechanisms" *Int J Mol Sci*
25. Geraghty, Aliota, Bonnac (2021) "Broad-spectrum antiviral strategies and nucleoside analogues" *Viruses*
26. Mcintyre, Molenaars, Schomakers et al. (2023) "Anti-retroviral treatment with zidovudine alters pyrimidine metabolism, reduces translation, and extends healthy longevity via ATF-4" *Cell Rep*
27. Li, Wang, Yu et al. (2023) "Tenofovir vs entecavir among patients with HBV-related HCC after resection" *JAMA Netw Open*
28. Grebely, Dalgard, Conway et al. (2018) "Sofosbuvir and velpatasvir for hepatitis C virus infection in people with recent injection drug use (SIMPLIFY): an open-label, single-arm, phase 4, multicentre trial" *Lancet Gastroenterol Hepatol*
29. Corey, Benedetti, Critchlow et al. (1983) "Treatment of primary firstepisode genital herpes simplex virus infections with acyclovir: results of topical, intravenous and oral therapy" *J Antimicrob Chemother*
30. Zhang, Sun, Xie et al. (2023) "A viral RNA-dependent RNA polymerase inhibitor VV116 broadly inhibits human coronaviruses and has synergistic potency with 3CLpro inhibitor nirmatrelvir" *Sig Transduct Target Ther*
31. Sourimant, Lieber, Aggarwal et al. "2022. 4'-Fluorouridine is an oral antiviral that blocks respiratory syncytial virus and SARS-CoV-2 replication" *Science*
32. Lieber, Aggarwal, Yoon et al. "2023. 4'-Fluorouridine mitigates lethal infection with pandemic human and highly pathogenic avian influenza viruses" *PLoS Pathog*
33. Yin, May, Lello et al. (2024) "Fluorouridine inhibits alphavirus replication and infection in vitro and in vivo. MBio:e0042024"
34. Chen, Li, Han et al. (2024) "The nucleoside analog 4'-fluorouridine suppresses the replication of multiple enteroviruses by targeting 3D polymerase" *Antimicrob Agents Chemother*
35. Welch, Spengler, Westover et al. (2024) "Delayed low-dose oral administration of 4'-fluorouridine inhibits pathogenic arenaviruses in animal models of lethal disease" *Sci Transl Med*
36. Abas, Tallei, Fatimawali et al. "2023. 4'-fluorouridine as a potential COVID-19 oral drug?: a review" *F1000Res*
37. Li, Chen, Trajkovski et al. "2020. 4'-fluorinated RNA: synthesis, structure, and applications as a sensitive 19F NMR probe of RNA structure and function" *J Am Chem Soc*
38. Xiang, Hu, Xue et al. (2023) "Synthesis and evaluation of NHC derivatives and 4'-fluorouridine prodrugs" *Org Biomol Chem*
39. Cheng, Zheng, Dong et al. (2025) "Design and development of a novel oral 4'-fluorouridine double prodrug VV261 against SFTSV" *J Med Chem*
40. Huchting, Vanderlinden, Van Berwaer et al. (2019) "Cell line-dependent activation and antiviral activity of T-1105, the nonfluorinated analogue of T-705 (favipiravir)" *Antiviral Res*
41. Cambray, Roberts, Carver et al. (1987) "A comparison of the radiation-induced mutation at the HGPRT locus in V79 and BHK cells" *Mutat Res*
42. Furuta, Gowen, Takahashi et al. (2013) "Favipiravir (T-705), a novel viral RNA polymerase inhibitor" *Antiviral Res*
43. Li, Santos, Bernal et al. (2023) "Biology and postnatal development of organ systems of cynomolgus monkeys (Macaca fascicularis)" *J Med Primatol*
44. Hawman, Feldmann (2023) "Crimean-Congo haemorrhagic fever virus" *Nat Rev Microbiol*
45. Jordan, Hildeman, Kappler et al. (2004) "An animal model of hemophagocytic lymphohistiocytosis (HLH): CD8+ T cells and interferon gamma are essential for the disorder" *Blood*
46. Westover, Jung, Alkan et al. (2024) "Modeling Heartland virus disease in mice and therapeutic intervention with 4'-fluorouridine" *J Virol*
48. Lee, Uttamapinant, Verdine (2007) "A concise synthesis of 4'-fluoro nucleosides" *Org Lett*
49. Xie, Yin, Zhang et al. (2021) "Design and development of an oral remdesivir derivative VV116 against SARS-CoV-2" *Cell Res*
50. Vivet-Boudou, Isel, Safadi et al. (2015) "Evaluation of anti-HIV-1 mutagenic nucleoside analogues" *J Biol Chem*
51. (2025) *Full-Length Text Journal of Virology*
52. Wang, Xie, Zheng et al. (2020) "Efficacy of entecavir in treating hepatitis B virus-associated membranous nephropathy" *Rev Esp Enferm Dig*
53. Aydin, Tanir, Sel et al. (2019) "Acyclovir unresponsive herpes simplex encephalitis in a child successfully treated with the addition of foscarnet: case report" *Arch Argent Pediatr*
54. Deval (2009) "Antimicrobial strategies: inhibition of viral polymerases by 3'-hydroxyl nucleosides" *Drugs (Abingdon Engl)*
55. Kabinger, Stiller, Schmitzová et al. (2021) "Mechanism of molnupiravir-induced SARS-CoV-2 mutagenesis" *Nat Struct Mol Biol*
56. Zhu, Zhang, Shang et al. (2019) "Comprehensive interactome analysis reveals that STT3B is required for N-glycosylation of Lassa virus glycoprotein" *J Virol*
57. Wan, Zhu, Li et al. (2021) "High-throughput screening of an FDA-approved drug library identifies inhibitors against arenaviruses and SARS-CoV-2" *ACS Infect Dis*
58. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282363&blobtype=pdf | # HIV-1 Drug Resistance in Children and Implications for Pediatric Treatment Strategies: A Systematic Review and Meta-analysis
Joseph Fokam, Christelle Aude, Ka'e, Bouba Yagai, Maria Santoro, Judith Otieno, Natella Rakhmanina, Collins Chenwi, Alex Nka, Ezechiel Ngoufack, Jagni Semengue, Davy-Hyacinthe Gouissi, Willy Leroi, Pabo Togna, Nelly Kamgaing, Tetang Suzie, Desire Takou, Georges Teto, Tatiana Tekoh, Jeremiah Efakika Gabisa, Audrey Mundo, Lum Forgwei, Naomi-Karell Etame, Aurelie Minelle, Kengni Ngueko, Michel Carlos, Tommo Tchouaket, Boris Tchounga, Patrice Tchendjou, Joelle Nounouce, Bouba Pamen, Rogers Awoh, Gregory-Edie Halle-Ekane, Giulia Cappelli, Alexis Ndjolo, Francesca Ceccherini-Silberstein, Vittorio Colizzi, Jean Kaseya, Nicaise Ndembi, Carlo Perno
## Abstract
Introduction.Failure in the prevention of mother-to-child HIV transmission (PMTCT) and pediatric treatment challenges led to pretreatment drug resistance (PDR) and acquired drug resistance (ADR) in children with HIV (CWHIV).Method. Interventional and observational data published between 2010 and 2024 on PDR and ADR in CWHIV were included and analyzed by random effects models.Results. Overall, 72 studies encompassing 9973 children were included. The prevalence (95% CI) of PDR was 32.48% (26.08-39.21), and high among those who failed PMTCT prophylaxis (43.23% [32.94-53.82]) versus those without PMTCT-intervention (P < .01) and driven by nonnucleoside reverse transcriptase inhibitors (NNRTI) mutations (28.38% [18.74-39.08]; P = .013). The prevalence of ADR was 61. 43% (49.82-72.45), driven by ]; P < .001). INSTI-ADR was low (5.53% [2.49-9.53]) but emerging.
Conclusion.There are high burdens of PDR and ADR among CWHIV, suggesting the need to phase out pediatric NNRTIs used for either PMTCT or treatment. Emerging INSTI resistance among CWHIV highlights the relevance of drug-resistance surveillance strategies.Prospero registration No. CRD42023470034.
The risk of perinatal HIV infection without any intervention varies from 15% to 45%, with about one-fourth of exposed newborns acquiring the virus during childbirth and one-fifth during pregnancy and breastfeeding [1]. This risk of vertical transmission has reduced substantially over time with the successful implementation of HIV prevention of mother-to-child transmission (PMTCT) through screening and management of both mothers and infants [2,3]. In the current era, successful PMTCT programs are designed to achieve elimination targets by reducing HIV vertical transmission to less than 2% at 6-8 weeks and less than 5% after breastfeeding cessation throughout the PMTCT cascade care in low-and middle-income countries (LMICs) [4]. Of note, PMTCT has significantly contributed to the concept of treatment as prevention, owing to the impact of antiretrovirals in preventing HIV transmission to exposed children [5]. However, despite efforts to scale up PMTCT interventions globally, the burden of pediatric HIV infection is still of concern among exposed infants who have experienced PMTCT failure or without any PMTCT interventions. Suboptimal exposure to antiretroviral (ARV)-based prophylaxis during PMTCT increases the risk of pretreatment drug resistance (PDR) especially in LMICs like sub-Saharan Africa (SSA) where low genetic barrier drugs are still used. In SSA, about 159 000 children are newly infected with HIV every year and most of those identified are initiated on antiretroviral therapy (ART) following the test and treat strategy recommended by the World Health Organization (WHO) since 2016 [5][6][7].
Despite the pediatric ART rollout, children with HIV (CWHIV) are faced with limited treatment options and coverage, with only 57% of children aged 0-14 years accessing treatment in 2022 [8]. In July 2019, the WHO revised its recommendations regarding first-and second-line ARTs, following new evidence on the effectiveness and drug safety of preferred and alternative ART regimens [9]. More importantly, UNAIDS set a global 95-95-95 target aimed at ensuring that 95% of people with HIV know their status, 95% of those who know their status are receiving ART, and 95% of those receiving ART have achieved viral suppression by 2025 in all subpopulations, geographical settings and age groups including children and adolescents living with HIV [10]. By the end of 2023, these targets were at 86%, 89%, and 93% respectively, indicating hopes in achieving HIV elimination and sustained epidemic control beyond 2030 [10].
Achieving HIV elimination requires equitable access to, and effective use of efficacious and better tolerated ARVs [11,12]. These lifesaving treatments have normalized the life expectancy of CWHIV, allowing them to grow toward adolescence and adulthood while maintaining lifetime ART [13]. This is particularly true following the rollout of dolutegravir (DTG), with rates of viral suppression reaching more than 90% of the general population receiving ART [14]. Nevertheless, viral suppression is not uniform across diverse regions and populations [15]. Besides the benefits of ART, the lifelong therapeutic exposure also raises concerns regarding toxicities, drug interactions, suboptimal adherence to ART (more frequent in pediatric populations, chiefly driven by nondisclosure and orphan hood), and the emergence of HIV drug resistance (HIVDR) [16]. Considering programmatic challenges in LMICs, inappropriate dispensing practices, recurrent drug stock-outs, poor retention in care, limited therapeutic options, socioeconomic disparities, and limited access to specialized medical care or reference therapeutic monitoring (viral load, CD4, genotypic resistance testing, adherence biomarkers, and drug safety assessment) in LMICs, the emergence of HIVDR is becoming even more concerning and outweighs gains at global level toward pandemic control [17,18].
Suboptimal drug levels in plasma drive by HIVDR, inadequate dosing guidelines for specific ARVs and incomplete compliance, which are all more alarming in pediatric populations. For instance, previous studies have reported higher rates of HIVDR in children as compared to adults living with HIV [19,20]. Both PDR and acquired drug resistance (ADR) represent major threats in children [21][22][23][24][25][26], based on the clinical significance of PDR on response to first-line ART. At the same time, ADR is known to compromise ART efficacy, which likely leads to viral rebound, HIV-associated morbidity/mortality, and increased risk of HIV transmission at population level [27,28]. Hence, the added value of HIVDR genotyping cannot be overemphasized, and clinical HIVDR testing has become a standard of care in managing HIV infection, especially for CWHIV who need lifesaving treatments to safeguard their life expectancy.
With the goal to contribute to shaping treatment guidelines and monitoring strategies in pediatric populations at the global level, the present study aims to assess the global pooled prevalence of PDR and ADR as well as associated factors in CWHIV over the past decade.
## METHODS
## Design and Registration
This systematic review and meta-analysis was performed following the guidelines of Preferred Reporting Items for Systematic Review and Meta-Analyses [29] and was registered in the Prospective Register of Systematic Reviews (CRD42023470034).
## Design and Setting of the Study
## Exclusion criteria
Publications such as case reports, letters, comments, reviews, systematic reviews, and meta-analysis and editorials studies were excluded.
## Search Strategy
A systematic search was performed using PubMed/MEDLINE, Google Scholar, ScienceDirect, African journals online, and gray literature databases (conference abstracts) using the keywords: HIV-1, "drug resistance," "acquired drug resistance," "pretreatment drug resistance," infants, children linked by the following Boolean operators: "OR" and "AND" (Supplementary File 1 shows the detailed search strategy for PubMed). A filter was performed by year (2010-2024). Additional studies were screened manually from the references of included studies.
## Selection of Studies for Inclusion in the Review
Records from the various sources were combined in an excel spreadsheet. Duplicates were identified and removed. Relevant studies were selected by 2 study authors (A.C.K. and A.D.N.) after an independent examination of the titles and abstracts of the eligible studies. Before data extraction, discordant ideas between investigators regarding the selection of the studies were resolved by discussion, consensus, or intervention of a third person (B.Y. or E.N.J.S.) when necessary.
## Data Extraction and Management
Four study authors (A.C.K., A.D.N., J.E.G., and T.T.) used a Google form questionnaire to extract and verify data from the included studies. The extracted data were: the name of the first author, the year of publication, the study design, the inclusion criteria, sampling method, sampling period, age, gender, sample size, sample type, the rate of HIVDR, the rate of HIVDR by classes, HIV-1 subtypes, PMTCT exposure (no PMTCT exposure vs PMTCT exposure), ART (drug class, ART regimen), WHO clinical staging (I, II, III, IV), and geographical location (country/continent) whenever available. Disagreements observed by different data extractors during data extraction were resolved by discussion and/or consensus.
## Data Analysis
I 2 and H statistics were used to estimate the heterogeneity among studies [31]. The I 2 value indicates the degree of heterogeneity, with values of 0%, 18%, 45%, and 75% designating no, low, moderate, and high heterogeneity [32], respectively. Regarding the H statistics, lack of evidence on heterogeneity among studies was designated by obtaining an H value close to 1. Otherwise, these values were inversely proportional with the degree of heterogeneity. The pooled prevalence of PDR and ADR and 95% confidence intervals (95% CI) were estimated by random effect models [33]. Random effects metaregression was used for subgroup analysis. PDR prevalence was categorized as low (<5%), moderate (5%-15%), or high (>15%) per WHO guidelines [34]. The R version 3.6.0 software (packages "meta," "metafor," and "ggplot 2") was used to perform all meta-analysis, through the R studio interface [35,36]. The dynamic curves of HIVDR were performed according to sampling years and trend proportion test was used to evaluate the significance of the evolution overtime.
## Risk of Bias Assessment
The quality of each study was independently assessed by 3 study authors (A.C.K., A.D.N., and J.E.G.) using a dedicated scale for prevalence studies that is based on 10 components divided into 2 groups: internal and external validity of the study (Supplementary File 2) [37]. A score of 0 or 1 was assigned to each question in the assessment tool for a total score of 10 per study. The scores of 0-3, 4-6, and 7-10 represented a high, moderate, and low risk of bias, respectively. For nonrandomized studies, the risk of bias was evaluated using ROBINS-1 [38], whereas ROBIS [RoB 2.0] was used for randomized controlled trials [39] (Supplementary File 3). Importantly, divergence in risk of bias assessment among the review authors was solved through discussion and consensus or by arbitration of a third review author.
## RESULTS
## Population Characteristics
A total of 325 studies were identified through the electronic search strategy. Duplicates (n = 12), irrelevant studies based on titles, and abstracts were removed (n = 159); 152 studies were assessed for full text eligibility and 72 studies finally met all inclusion criteria. Figure 1 shows the study selection process.
Overall, studies in this review encompassed 9973 children aged from 0 to 15 years with 52.5% being male and 47.5% female. Most (38.46%) of studies were cross-sectional, followed by clinical trials (21.7%), cohort studies (17.9%), surveys (2.56%), surveillance (1.28%), and chart review (1.28%). About 16.7% of studies included were without study design reported. Characteristics of the included studies are available on Table 1.
## Rates of Pretreatment and Acquired Drug Resistance
Data on PDR [21-26, 40, 43, 45, 47, 50, 57-75, 93-97, 99-101, 103-106] were available for 5884 participants. Regarding PMTCT exposure, information was available for 3836 participants with 66.36% being exposed to PMTCT prophylaxis. Globally, the pooled prevalence [95% CI] of PDR was 31.94% [25.50-38.72]; Figure 2A. The prevalence among participants exposed to PMTCT prophylaxis was high, about 43.23% [32.94-53.82] compared to 19.40% [14.45-24.84] among those without any PMTCT exposure, P < .01. With respect to regional distribution, burden of pooled prevalence was high in SSA with a prevalence of 39.13% [31.36-47.17 [21,25,47,58,59,62,107]. NNRTI resistance-associated mutations mostly reported were K101E, K103S, E138A, E138K, E138Q, H221Y, M230L, K103N, V106A, V106M, V108I, Y181C, Y181I, Y188C, Y188H, Y188L, G190A, G190S, and P225H [21,25,40,47,58,59,62,107]. PI resistance-associated mutations reported were D30N, [72] Cross-sectional Multicenter South Africa Low risk of bias Tadesse et al, 2019 [24] Cross-sectional study Multicenter Ethiopia Low risk of bias Bennett et al, 2020 [73] Not reported/ unclear Monocenter Zambia Moderate risk of bias Dambaya et al, 2020 [22] Cross-sectional study Multicenter Cameroon Low risk of bias Boerma et al, 2016 [74] Cohort Monocenter Nigeria Low risk of bias Dow et al, 2017 [75] Cross-sectional study Multicenter Tanzania Low risk of bias Namayanja et al, 2023 [76] Cross-sectional survey Multicenter Uganda Low risk of bias Boyce et al, 2023 [48] Cohort Multicenter Botswana, Kenya, South Africa, Tanzania, Thailand, Uganda, USA, Zimbabwe Low risk of bias Hackett et al, 2023 [77] Cohort Monocenter Uganda Low risk of bias Bratholm et al, 2012 [78] Cross-sectional study Monocenter Tanzania Moderate risk of bias Mohamad et al, 2012 [79] Cross-sectional study Monocenter Malaysia Moderate risk of bias Tolle et al, 2012 [80] Cross-sectional study Monocenter Botswana Moderate risk of bias Gomila et al, 2013 [81] Cross-sectional study Monocenter Botswana Moderate risk of bias Salou et al, 2016 [82] Cross-sectional study
## Nationally representative
Togo Low risk of bias Mossoro-Kpinde et al, 2017 [83] Cross-sectional study Monocenter Central African Republic Moderate risk of bias Muri et al, 2017 [84] Cohort Monocenter Tanzania Low risk of bias Fofana et al, 2018 [85] Cross-sectional study Monocenter Benin Low risk of bias Tadesse et al, 2018 [86] Cross-sectional study Monocenter South Africa Moderate risk of bias Vaz et al, 2018 [87] Cross-sectional study Multicenter Mozambique Low risk of bias Cissé et al, 2019 [88] Cross [40,48] and accessory mutations described were G163R, G140E, and G140K [40]. Additionally, E157Q, which is known to affect the susceptibility to raltegravir/elvitegravir rather than to dolutegravir, was also present [71]. PDR was associated with high pre-ART viral load and WHO clinical stage II (compared to WHO clinical stage I) [21].
## Patterns of Acquired Drug Resistance
As concerns the distribution of ADR by drug classes, NNRTI-resistance was the most predominant (OR [95% CI]: 107] and PI resistance mutations were F53L, I47V, M46I, V77I, D60E, I54V, V82A/F, V11I, I62V, I15V, I63P, G16E, L10I/M/V, I13V, K20I/R, L89I/M/V, H69 K/R/Q/Y, and M36I/L [41,42,83,86,91,92,107]. Of note, dolutegravirrelated mutations were found, such as E138K, G118R, G140A, G148K, R263K, and T66A [41,76,77,98,108]. Secondary integrase substitution E157Q and L74I were also reported [41].
## Dynamic of Pretreatment and Acquired Drug Resistance
As shown in Figure 4A, significant increase of PDR was observed between 2007 and 2018 (from 11.5% to 50.7%); the burden remains relatively the same between 2018 and 2021, when the DTG was rolled out. Nonetheless, a decrease was noted between 2021 and 2022 (from 52.6% to 33.3%), the period when DTG was fully expanded. However, during this same period, increase rates of PI and INSTI-PDR were observed (from 4B).
## DISCUSSION
In the global landscape of pediatric HIVDR prevention and control, recent systematic review and meta-analysis on HIVDR among children focused only on PDR solely in LMICs [109] and in SSA [110]. Another review without meta-analysis from Sanchez et al. evaluated HIVDR in pediatric populations but was published almost 10 years ago. To the best of our knowledge, the present study therefore seems to be the first systematic review and meta-analysis including more recent studies to describe the profile of both PDR and ADR worldwide in CWHIV over the past decade. As concerns data found, this systematic review has shown a lack of information on HIVDR in many countries from different continents, even in SSA countries with high burden of pediatric HIV infection. This study shows that the pooled prevalence of PDR is very high, more than 30% and was fitted in a trend of increasing [110]. In fact, the prevalence of PDR in SSA regions was high as compared to other regions and might be explained by the previously extended use of low genetic-barrier drugs such as nevirapine prophylaxis during PMTCT intervention among infants after childbirth or efavirenz-containing regimens used to treat pregnant and breastfeeding women [72,75]. More importantly, the prevalence of PDR was almost 3 times higher among infants with PMTCT exposure in comparison to those with no PMTCT intervention. This high prevalence (>40%) of PDR after PMTCT exposure is explained previously by the use of low genetic barrier drugs as prophylaxis during PMTCT intervention or maternal subtherapeutic regimens during pregnancy or breastfeeding [110], as supported by the WHO report showing that approximately half of CWHIV harbor PDR among selected surveys [111]. Moreover, despite the rollout of DTG since 2019, the observed persistent rate of NNRTI-PDR might be explained by the viral fitness of virus harboring NNRTI mutations (including Y181C and K103N) on the one hand and the Overall rates of ADR in this pediatric population were also high and driven by NNRTI resistance, with higher levels in SSA (>65%) as compared to other regions. As factors associated with the development of ADR might include regimen-related factors, patient factors, viral factors, and program-related factors, the phasing out of low genetic barrier (nevirapine and Efavirenz) should be accompanied by efforts to widen the use of pediatric DTG-containing regimens, strengthen the supply chain management to prevent drug stock outs, ensure timely treatment monitoring with viral load, and better position the use of other important regimens such as ritonavir-boosted darunavir (so far with rare availability) or anticipate with injectable ARVs such as CARLA (cabotegravir-rilpivirine long acting) and lenacapavir (capsid inhibitor), based on the absence or low prevalence of ADR (∼5%) to these newer drugs [112]. Regarding viral factors, resistance mutation patterns may differ across subtypes, replicative fitness varies by viral clades, and it is demonstrated that the risk to ADR after exposure to single-dose nevirapine is high in HIV-1 subtype D as compared to HIV-1 subtype A, as well as the risk of pathogenesis [113]. Moreover, another example is that natural polymorphisms in reverse transcriptase of subtype C increase the propensity of the virus to select K65R in comparison to subtype B, resulting to tenofovir loss of efficacy even before the use of this drug [114]. Indeed, SSA harbors wide HIV genetic diversity, which may influence treatment outcomes and possibly the selection of HIVDR. This raises the need to also establish personalized treatment for cases of multi-resistance or heavily treated children/adolescents. Finally, ongoing vertical transmission and the limited pediatric ART is a call for developing newer ARVs with high safety during pregnancy and infancy. Such efforts, coupled to adequate maternal adherence counseling, male partner involvement, will contribute to achieve the elimination of pediatric AIDS [111][112][113]. It is, therefore, pivotal to strengthen health systems in LMICs, set up national public health institutes, and develop a thorough sustainability roadmap and strategic plan to achieve and maintain optimal response for PMTCT and to pediatric HIV management after epidemic control beyond 2030.
## References
1. (2019) "Adolescent-friendly health services for adolescents living with HIV: from theory to practice"
2. Ka'e Ac, Nka, Yagai (2023) "The mother-to-child transmission of HIV-1 and profile of viral reservoirs in pediatric population: a systematic review with metaanalysis of the Cameroonian studies" *PLoS One*
3. Muyunda, Musonda, Todd et al. (2020) "Effectiveness of lifelong ART (Option B+) in the prevention of mother-to-child transmission of HIV programme in Zambia: observations based on routinely collected health data" *Front Public Heal*
4. Hussen, Zenebe, Mamo et al. (2022) "Determinants of HIV infection among children born from mothers on prevention of mother to child transmission programme of HIV in southern Ethiopia: a case-control study" *BMJ Open*
5. (2015) "Guideline on when to start antiretroviral therapy and on pre-exposure prophylaxis for HIV"
6. Idele, Hayashi, Porth et al. (2017) "Prevention of mother-to-child transmission of HIV and paediatric HIV care and treatment monitoring: from measuring process to impact and elimination of mother-to-child transmission of HIV" *AIDS Behav*
7. Ladner, Besson, Rodrigues (2013) "Prevention of mother-to-child HIV transmission in resource-limited settings: assessment of 99 Viramune Donation Programmes in 34 countries, 2000-2011" *BMC Public Health*
9. (2023) "The Joint United Nations Programme on HIV/AIDS (UNAIDS). Global HIV statistics"
10. (2020) "Paediatric antiretroviral medicines: market & supply update"
11. (2023) "Understanding measures of progress towards the 95-95-95 HIV testing, treatment and viral suppression targets"
12. Moore, Bartlett (2011) "Dramatic decline in the HIV-1 RNA level over calendar time in a large urban HIV practice" *Clin Infect Dis*
13. Tseng, Seet, Phillips (2015) "The evolution of three decades of antiretroviral therapy: challenges, triumphs and the promise of the future" *Br J Clin Pharmacol*
14. (2025) "How long can untreated HIV-positive children live"
15. Abudiore, Ikechukwu, Campbell (2023) "High acceptability and viral suppression rate for first-line patients on a dolutegravir-based regimen : an early adopter study in Nigeria" *PLoS One*
16. Havlir, Lockman, Ayles (2020) "What do the universal test and treat trials tell us about the path to HIV epidemic control?" *J Int AIDS Soc*
17. Fokam, Bala, Santoro (2022) "Archiving of mutations in HIV-1 cellular reservoirs among vertically infected adolescents is contingent with clinical stages and plasma viral load: evidence from the EDCTP-READY study" *HIV Med*
18. Melku, Gesesew, Ward (2022) "Magnitude and predictors of HIV-drug resistance in Africa: a protocol for systematic review and meta-analysis" *PLoS One*
19. (2025) "Children and families at risk"
20. Pursuing, Treatment, Ii et al. (2011) "Risk of triple-class virological failure in children with HIV" *Lancet*
21. Kamya, Mayanja-Kizza, Kambugu (2007) "Predictors of long-term viral failure among Ugandan children and adults treated with antiretroviral therapy" *J Acquir Immune Defic Syndr*
22. Kityo, Boerma, Sigaloff (2017) "Pretreatment HIV drug resistance results in virological failure and accumulation of additional resistance mutations in Ugandan children" *J Antimicrob Chemother*
23. Dambaya, Fokam, Ngoufack (2020) "HIV-1 drug resistance and genetic diversity among vertically infected Cameroonian children and adolescents" *Explor Res Hypothesis Med*
24. Salou, Dagnra, Butel (2016) "Short report high rates of virological failure and drug resistance in perinatally HIV-1-infected children and adolescents receiving lifelong antiretroviral therapy in routine clinics in Togo" *J Int AIDS Soc*
25. Tadesse, Tsai, Chala (2019) "Prevalence and correlates of pre-treatment HIV drug resistance among HIV-infected children in Ethiopia" *Viruses*
26. Almeida, Rodrigues, Zaparoli (2012) "Prevalence of transmitted HIV-1 drug resistance mutations in Sao Paulo, Brazil" *Pediatr Infect Dis J*
27. Karunaianantham, Kumar, Gopalan (2022) "Molecular characterization of the pol gene of vertically transmitted HIV-1 strains in children with virological failure" *AIDS Res Hum Retroviruses*
28. Hamers, Schuurman, Sigaloff (2012) "Effect of pretreatment HIV-1 drug resistance on immunological, virological, and drug-resistance outcomes of firstline antiretroviral treatment in sub-Saharan Africa: a multicentre cohort study" *Lancet Infect Dis*
29. Phillips, Stover, Cambiano (2017) "Impact of HIV drug resistance on HIV/ AIDS-associated mortality, new infections, and antiretroviral therapy program costs in sub-Saharan Africa" *J Infect Dis*
30. Kamioka (2019) "Preferred reporting items for systematic review and meta-analysis protocols (PRISMA-p) 2015 statement" *Japanese Pharmacol Ther*
31. (1999) "Countries or areas/geographical regions"
32. Higgins, Thompson, Deeks et al. (2006) "Measuring inconsistency in knowledgebases" *BMJ Open*
33. Veroniki, Jackson, Viechtbauer (2016) "Methods to estimate the between-study variance and its uncertainty in meta-analysis" *Res Synth Methods*
34. Barendregt, Doi, Lee et al. (2013) "Meta-analysis of prevalence" *J Epidemiol Community Health*
35. Who (2012) "HIV Drug Resistance Report"
36. Schwarzer (2007) "meta: An R Package for Meta-Analysis" *Germany*
37. Team R Development Core (2018) "A language and environment for statistical computing" *R Found Stat Comput*
38. Hoy, Brooks, Woolf (2012) "Assessing risk of bias in prevalence studies: modification of an existing tool and evidence of interrater agreement" *J Clin Epidemiol*
39. Sterne, Hernán, Reeves (2016) "ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions" *BMJ*
40. Whiting, Savović, Higgins (2016) "ROBIS: a new tool to assess risk of bias in systematic reviews was developed" *J Clin Epidemiol*
41. Moraka, Garcia-Broncano, Hu (2021) "Patterns of pretreatment drug resistance mutations of very early diagnosed and treated infants in Botswana" *AIDS*
42. Vavro, Ruel, Wiznia (2022) "Emergence of resistance in HIV-1 integrase with dolutegravir treatment in a pediatric population from the IMPAACT P1093 study" *Antimicrob Agents Chemother*
43. Violari, Bologna, Kumarasamy (2015) "Safety and efficacy of darunavir/ritonavir in treatment-experienced pediatric patients" *Pediatr Infect Dis J*
44. Fogel, Mwatha, Richardson et al. (2013) "Impact of maternal and infant antiretroviral drug regimens on drug resistance in HIV-infected breastfeeding infants" *Pediatr Infect Dis J*
45. Harrison, Melvin, Fiscus (2015) "HIV-1 drug resistance and second-line treatment in children randomized to switch at low versus higher RNA thresholds" *J Acquir Immune Defic Syndr*
46. Nacro, Zoure, Hien (2011) "Pharmacology and immuno-virologic efficacy of once-a-day HAART in African HIV-infected children: ANRS 12103 phase II trial" *Bull World Health Organ*
47. Beck, Jang, Mckernan-Mullin (2012) "Monitoring of HIV type 1 DNA load and drug resistance in peripheral blood mononuclear cells during suppressive antiretroviral therapy does not predict virologic failure" *AIDS Res Hum Retroviruses*
48. Inzaule, Weidle, Yang (2016) "Prevalence and dynamics of the K65R drug resistance mutation in HIV-1-infected infants exposed to maternal therapy with lamivudine, zidovudine and either nevirapine or nelfinavir in breast milk" *J Antimicrob Chemother*
49. Boyce, Smith, Nixon (2023) "Genotypic and phenotypic dolutegravir resistance in pregnant women and children living with HIV-1. International work on HIV drug resistance treatment strategies"
50. Tambuyzer, Thys, Hoogstoel (2016) "Original article assessment of etravirine resistance in HIV-1-infected paediatric patients using population and deep sequencing: final results of the PIANO study" *Antivir Ther*
51. Fokam, Salpini, Santoro (2011) "Drug resistance among drug-naive and firstline antiretroviral treatment-failing children in Cameroon" *Pediatr HIV*
52. Musiime, Kasirye, Naidoo-James (2016) "Once vs twice-daily abacavir and lamivudine in African children" *AIDS*
53. Barro, Some, Foulongne "Short-term virological efficacy, immune reconstitution, tolerance, and adherence of once-daily dosing of didanosine, lamivudine, and efavirenz in HIV-1-infected African"
54. Burkiname (2011) *J Acquir Immune Defic Syndr*
55. Huibers, Kityo, Boerma (2019) "Long-term virological outcomes, failure and acquired resistance in a large cohort of Ugandan children" *J Antimicrob Chemother*
56. Kuhn, Coovadia, Strehlau (2012) "Switching children previously exposed to nevirapine to nevirapine-based treatment after initial suppression with a protease-inhibitor-based regimen : long-term follow-up of a randomised, openlabel trial" *Lancet Infect Dis*
57. Butler, Inshaw, Ford (2016) "Health technology assessment" *Health Technol Assess*
58. Dahourou, Amorissani-Folquet, Malateste (2017) "Efavirenz-based simplification after successful early lopinavir-boosted-ritonavir-based therapy in HIV-infected children in Burkina Faso and Côte d 'Ivoire: the MONOD ANRS 12206 non-inferiority randomised trial" *BMC Med*
59. Olusola, Olusola, Oladokun et al. (2021) "Surveillance of pretreatment drug resistance among HIV-infected children in Ibadan" *AIDS Res Hum Retroviruses*
60. Children, Parham, Rivera et al. (2001) "Short communication: high prevalence of drug resistance in HIV type 1-infected children born in Honduras and Belize" *AIDS Res Hum Retroviruses*
61. Ferreira, Pinto, Kakehasi (2010) "Prevalence of primary drug resistance-associated mutations among HIV type 1 vertically infected children in Belo Horizonte, Brazil" *AIDS Res Hum Retroviruses*
62. Charpentier, Gody, Tisserand et al. (2011) "Surveillance of antiretroviral drug resistance mutations in untreated young children living in the Central African Republic" *Antivir Ther*
63. De Mulder, Yebra, Martin (2011) "Drug resistance prevalence and HIV-1 variant characterization in the naive and pretreated HIV-1-infected paediatric population in Madrid" *J Antimicrob Chemother*
64. Mulder, Yebra, Navas (2012) "Trends in drug resistance prevalence in HIV-1-infected children in Madrid: 1993 to 2010 analysis" *Pediatr Infect Dis J*
65. Neogi, Sahoo, Costa et al. (2012) "High viremia and low level of transmitted drug resistance in anti-retroviral therapy-naïve perinatally-infected children and adolescents with HIV-1 subtype C infection" *BMC Infect Dis*
66. Shao, Kifaro, Chilumba (2014) "HIV-1 drug mutations in children from northern Tanzania" *J Antimicrob Chemother*
67. Andrade, Sabid (2017) "Drug resistance in antiretroviral-naive children newly diagnosed with HIV-1 in Manaus, Amazonas" *J Antimicrob Chemother*
68. Jordan, Penazzato, Cournil (2017) "HIV drug resistance in African infants and young children newly diagnosed with HIV: a multicounty analysis" *Infect Dis Soc Am*
69. Kanthula, Rossouw, Feucht (2017) "Persistence of HIV drug resistance among South African children given nevirapine to prevent mother-to-childtransmission" *AIDS*
70. Poppe, Liyoka, Kwon (2017) "HIV drug resistance in infants increases with changing PMTCT regimens running" *AIDS*
71. Yeganeh, Kerin, Ank (2017) "HIV antiretroviral resistance and transmission in mother-infant pairs enrolled in a large perinatal study" *Clin Infect Dis*
72. Aulicino, Zapiola, Kademian (2018) "Pre-treatment drug resistance and HIV-1 subtypes in infants from Argentina with and without exposure to antiretroviral drugs for prevention of mother-to-child transmission" *J Antimicrob Chemother*
73. Frange, Veber, Chaix (2018) "Prevalence of drug resistance in children recently diagnosed with HIV-1 infection in France (2006-17): impact on susceptibility to first-line strategies" *J Antimicrob Chemother*
74. Hunt, Ledwaba, Salimo (2010) "Prevalence of HIV-1 drug resistance amongst newly diagnosed HIV-infected infants age 4-8 weeks, enrolled in three nationally representative PMTCT effectiveness surveys" *BMC Infect Dis*
75. Bennett, Chunda-Liyoka, Poppe (2020) "High nonnucleoside reverse transcriptase inhibitor resistance levels in HIV-1-infected Zambian mother-infant pairs" *AIDS*
76. Boerma, Boender, Sigaloff (2016) "High levels of pre-treatment HIV drug resistance and treatment failure in Nigerian children" *J Int AIDS Soc*
77. Dow, Schimana, Nyombi (2017) "HIV resistance and prevention of mother-to-child transmission regimen in HIV-infected infants in northern Tanzania" *AIDS Res Hum Retroviruses*
78. Namayanja, Watera, Pals (2023) "Cyclical acquired HIV drug resistance to dolutegravir among people living with HIV in Uganda-"CADRE" National Survey 2022. International work on HIV drug resistance treatment strategies"
79. Hackett (2023) "Low levels of clinically significant drug resistance mutations to dolutegravir amongst children living with HIV aged 0-15 years failing on first-line ART in Uganda"
80. Bratholm, Johannessen, Naman (2012) "Drug resistance is widespread among children who receive long-term antiretroviral treatment at a rural Tanzanian hospital 2010:1996-2000" *Braz J Infect Dis*
81. Mohamad, Deris, Yusoff et al. (2012) "Assessing subtypes and drug resistance mutations among HIV-1 infected children who failed antiretroviral therapy in Kelantan, Malaysia" *Braz J Infect Dis*
82. Tolle, Howard, Kirk et al. (2012) "Reverse transcriptase genotypes in pediatric patients failing initial antiretroviral therapy in Gaborone" *Botswana. J Int Assoc Physicians AIDS Care*
83. Gomila, Kirk, Marape et al. (2013) "Protease genotypes in patients failing protease inhibitor-based antiretroviral therapy at a pediatric" *Pediatr Infect Dis J*
84. Salou, Butel, Konou (2016) "High rates of drug resistance among newly diagnosed HIV-infected children in the national prevention of mother-to-child transmission program in Togo Monero" *Pediatr Infect Dis J*
85. Cd, Gody, Bouassa (2017) "High levels of virological failure with major genotypic resistance mutations in HIV-1-infected children after 5 years of care according to WHO-recommended 1st-line and 2nd-line antiretroviral regimens in the Central African Republic A cross-sectional STU" *Medicine*
86. Muri, Gamell, Ntamatungiro (2017) "Development of HIV drug resistance and therapeutic failure in children and adolescents in rural Tanzania: an emerging public health concern" *AIDS*
87. Fofana, Almeida, Lambert-Niclot (2018) "Resistance profile and treatment outcomes in HIV-infected children at virological failure in Benin, West Africa" *J Antimicrob Chemother*
88. Tadesse, Kinloch, Baraki (2018) "High levels of dual-class drug resistance in HIV-infected children failing first-line antiretroviral therapy in southern Ethiopia" *Viruses*
89. Vaz, Buck, Bhatt (2020) "Compromise of second-line antiretroviral therapy due to high rates of human immunodeficiency virus drug resistance in Mozambican treatment-experienced children with virologic failure" *J Pediatric Infect Dis Soc*
90. Cissé, Laborde-Balen (2019) "High level of treatment failure and drug resistance to first-line antiretroviral therapies among HIV-infected children receiving decentralized care in Senegal" *BMC Pediatr*
91. Bouassa, Cd, Gody et al. (2019) "High predictive efficacy of integrase strand transfer inhibitors in perinatally HIV-1-infected African children in therapeutic failure of first-and second-line antiretroviral drug regimens recommended by the WHO" *J Antimicrob Chemother*
92. Soumah, Veber, Moshous et al. (2019) "High rates of antiretroviral coverage and virological suppression in HIV-1-infected children and adolescents" *Med Mal Infect*
93. Hackett, Teasdale, Pals (2021) "Drug resistance mutations among South African children living with HIV on WHO-recommended ART regimens" *Clin Infect Dis*
94. Pa, Penda, Tchatchueng (2021) "Virological failure and antiretroviral resistance among HIV-infected children after five years follow-up in the ANRS 12225-PEDIACAM cohort in Cameroon" *PLoS One*
95. Soeria-Atmadja, Id, Nav (2020) "Pretreatment HIV drug resistance predicts accumulation of new mutations in ART-naïve Ugandan children" *Acta Paediatrica*
96. Chaplin, Akanmu, Inzaule (2018) "Association between HIV-1 subtype and drug resistance in Nigerian infants North-Central" *J Antimicrob Chemother*
97. Rozenszajn, Arazi-Caillaud, Taicz et al. (2022) "HIV-1 pretreatment drug resistance in vertically infected children is associated with poor virological response to protease inhibitor (PI) -based fi RST-line antiretroviral therapy (ART): results from a cohort study in Argentina" *J Antimicrob Chemother*
98. Jordan, Bikinesi, Ashipala (2022) "Pretreatment human immunodeficiency virus (HIV) drug resistance among treatment-naive infants newly diagnosed with HIV in 2016 in Namibia: results of a nationally representative study" *Open Forum Infect Dis*
99. Ebonyi, Okpokwu, Rawizza (2024) "Pretreatment and acquired drug resistance in children with human immunodeficiency virus type 1 in Jos" *Nigeria. Open Forum Infect Dis*
100. Kamori, Barabona, Rugemalila (2023) "Emerging integrase strand transfer inhibitor drug resistance mutations among children and adults on ART in Tanzania : findings from a national representative HIV drug resistance survey" *J Antimicrob Chemother*
101. Fisher, Smith, Murrell (2015) "Next generation sequencing improves detection of drug resistance mutations in infants after PMTCT failure" *J Clin Virol*
102. Holguín, Eraso, Escobar (2011) "Drug resistance prevalence in human immunodeficiency virus type 1 infected pediatric populations in Honduras and El Salvador during 1989-2009" *Pediatr Infect Dis J*
103. Antunes, Zindoga, Gomes (2015) "Development of nevirapine resistance in children exposed to the prevention of mother-to-child HIV-1 transmission programme in Maputo" *PLoS One*
104. Patel, Oyaro, Thomas (2023) "Impact of point-of-care HIV viral load and targeted drug resistance mutation testing on viral suppression among Kenyan children on antiretroviral therapy: results from an open-label" *J Int AIDS Soc*
105. Kébé, Bélec, Ndiaye (2014) "The case for addressing primary resistance mutations to non-nucleoside reverse transcriptase inhibitors to treat children born from mothers living with HIV in sub-Saharan Africa" *J Int AIDS Soc*
106. Inzaule, Osi, Akinbiyi (2018) "High prevalence of HIV drug resistance among newly diagnosed infants aged < 18 months: results from a nationwide surveillance in Nigeria" *J Acquir Immune Defic Syndr*
107. Louis, Segaren, Desinor (2013) "High levels of HIV-1 drug resistance in children who acquired HIV infection through mother to child transmission in the era of option B+" *Pediatr Infect Dis J*
108. Vaz, Giovanetti, Ferreira et al. (2015) "Molecular characterization of the human and their vertically infected children" *AIDS Res Hum Retroviruses*
109. Dambaya, Fokam, Ngoufack "HIV-1 drug resistance and genetic diversity among vertically infected Cameroonian children and adolescents" *Explor Res Hypothesis Med*
110. Frange (2019) "Bictegravir/emtricitabine/tenofovir alafenamide in paediatrics: reallife experience from a French cohort" *HIV Med*
111. Inzaule, Hamers, Bertagnolio (2019) "Pretreatment HIV drug resistance in low-and middle-income countries" *Future Virol*
112. Boerma, Sigaloff, Akanmu et al. (2017) "Alarming increase in pretreatment HIV drug resistance in children living in sub-Saharan Africa: a systematic review and meta-analysis" *J Antimicrob Chemother*
113. Bertagnolio, Luca, Vitoria (2012) "Determinants of HIV drug resistance and public health implications in low-and middle-income countries" *Antivir Ther*
114. Who (2019) "Update of recommendations on first-and second-line antiretroviral regimens"
115. Hauser, Mugenyi, Kabasinguzi et al. (2011) "Emergence and persistence of minor drug-resistant HIV-1 variants in Ugandan women after nevirapine single-dose prophylaxis" *PLoS One*
116. Brenner, Oliveira, Doualla-Bell (2006) "HIV-1 subtype C viruses rapidly develop K65R resistance to tenofovir in cell culture" *AIDS* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12395677&blobtype=pdf | # Respiratory viral infections in hospitalized adults: a comparative clinico-laboratory study of RSV, HMPV, and influenza
Virology Journal, Santhosha Devadiga, Anup Jayaram, Prasad Varamballi, Ujwal Shetty, Chiranjay Mukhopadhyay
## Abstract
Background Respiratory viral infections, including Human Metapneumovirus (HMPV), Influenza (Flu), and human Respiratory Syncytial Virus (hRSV), are major global health concerns. While their impact on vulnerable groups is known, their characteristics in healthy adults (18-65 years) are less clear. This study aimed to determine the incidence and clinical-laboratory features of RSV and HMPV in this population and compared them with those of Influenza A(H1N1) and influenza A(H3N2) for improved epidemiological and diagnostic understanding.Methodology A retrospective analysis was conducted on data from an Acute Febrile Illness surveillance (2016)(2017)(2018) in Manipal, India. The study included 96 HMPV, 68 hRSV, 75 Influenza A(H1N1), and 76 Influenza A(H3N2) positive hospitalized adults with fever (≥ 38 °C) and respiratory illness confirmed by RT-PCR. Clinical and laboratory data collected within the first 8 days of illness were statistically analyzed.
ResultsThe annual incidence rates of hRSV (0.33%-1.59%) and HMPV (0.14%-1.79%) varied. Coryza was common, but cough was most frequent in HMPV (97.9%). HMPV also resulted in increased rates of shortness of breath and chest pain. Leucopenia was most common in Influenza A(H1N1) patients, and thrombocytopenia was most common in hRSV patients. Significantly elevated leukocyte and platelet counts were observed in HMPV patients. Liver enzyme abnormalities are relatively common in hRSV and Influenza A(H1N1) patients. Symptom progression and laboratory trends revealed distinct patterns across the viruses.
ConclusionDespite overlapping initial symptoms, HMPV, hRSV, and influenza resulted in different clinical and laboratory profiles in adults. HMPV was associated with more prominent lower respiratory symptoms and a stronger inflammatory response. These distinctions can aid in the clinical differentiation and management of these common respiratory viruses in adults, highlighting the importance of timely diagnosis for improved patient care and public health strategies.
## Introduction
Respiratory viral infections are a significant global health concern, affecting millions of people annually and contributing to substantial morbidity and healthcare burdens. Human Metapneumovirus (HMPV), Influenza (Flu), and human Respiratory Syncytial Virus (hRSV) are the leading viral pathogens responsible for acute respiratory tract infections across different age groups [1]. These infections present with a wide spectrum of clinical manifestations, ranging from mild upper respiratory symptoms to severe pneumonia, acute respiratory distress syndrome (ARDS), and even mortality, particularly in vulnerable populations such as elderly and immunocompromised individuals [2]. Despite the seasonal and pandemic nature of these infections, their impact on otherwise healthy adults aged 18-65 years remains underexplored.
Understanding the incidence, clinical presentations, and laboratory findings of respiratory viral infections in adults is essential for understanding the nature of infection and circulation patterns, improving diagnostic algorithms and guiding therapeutic interventions. Clinical-laboratory features such as leukopenia, thrombocytopenia, elevated inflammatory markers (C-reactive protein, erythrocyte sedimentation rate), and liver enzyme abnormalities can aid in differentiating viral etiologies and predicting disease severity [3]. Moreover, variations in symptoms such as fever duration, respiratory distress, and the systemic inflammatory response among different viral pathogens can provide insights into disease progression and management strategies.
Given the emerging threats of seasonal and novel respiratory viruses, including the potential for coinfections and secondary complications, continuous surveillance and characterization of their clinical and laboratory profiles are crucial [4,5]. This study aimed to assess the incidence and clinical-laboratory characteristics of influenza, hRSV and HMPV infections in adults aged 18-65 years, providing valuable epidemiological and diagnostic insights. These data can support evidence-based clinical decision-making and inform public health strategies to mitigate the impact of respiratory viral infections.
## Methodology
Study design: A retrospective observational study was conducted to analyze clinical data from patients diagnosed with HMPV, hRSV and Influenza. A total of 68 hRSV and 96 HMPV-positive patients were included in the study. Respiratory specimens collected as part of Acute Febrile Illness (AFI) surveillance study conducted by Manipal Institute of Virology (MIV), with a case definition of patients admitted to the hospital with fever (≥ 38 °C) and acute respiratory illness, were included, and samples were tested for various viral pathogens, including hRSV, HMPV and Influenza virus, via molecular and serological methods from 2016 to 2018 [6].
hRSV-and HMPV-positive individuals were defined as those individuals whose results were positive according
## Graphical Abstract
Keywords Human respiratory syncytial virus, Human metapneumovirus, Influenza, Clinical, Laboratory, Comparative study to reverse transcription-polymerase chain reaction (RT-PCR) analysis of respiratory samples (i.e., nasal/ nasopharyngeal swabs). Briefly, total nucleic acid was isolated from nasal/throat swabs in viral transport medium (VTM) via a QIAmp viral RNA Mini Kit (#52906, Qiagen, Hilden, Germany) according to the manufacturer's instructions. Multiplex real-time RT-PCR was performed with primers and probes targeting HMPV and hRSV via a Respiratory Pathogens 21 Kit (#11373972, Fast Track Diagnostics -FTD, Luxembourg). RT-PCR was performed via a AgPath-ID™ One-step RT-PCR Kit (#4387391, Applied Biosystems, Foster City, USA). The reaction was performed in a QuantStudio™ 5 PCR system (#A28569, Applied Biosystems, Foster City, USA). The results are expressed as cycle threshold (Ct) values. Laboratory-confirmed cases were classified as upper respiratory tract illness (URTI) or lower respiratory tract illness (LRTI) on the basis of clinical symptoms defined by the World Health Organization (WHO) [7]. Age group matched 75 Influenza A(H1N1) positive, and 76 Influenza A(H3N2) positive cases were included for the comparison by random sampling method. Demographic, clinical and laboratory data of hRSV, HMPV and Influenza A positive patients were obtained from the AFI study database after ethical clearance was obtained and used for analysis. The current study was reviewed and approved by the Institutional Ethical Committee, Manipal Academy of Higher Education (IEC No: UEC/32/2013-14, MUEC/Renewal-08/2017, MIV/ AFI-IEC/2023).
## Data analysis
We assessed clinical and laboratory features during the first 8 days of post onset of illness. Upper, lower respiratory and gastrointestinal symptoms between the group and during the course of illness were analyzed. Hematological and biochemical parameters of the hRSV, HMPV and Influenza A virus positive cases were obtained from the AFI study database and analyzed using EZR version 1.68 (Easy R) [8]. For analysis of continuous variables, one-way ANOVA test and analysis of categorical variables, the chi-square test was used. A p-value of < 0.05 was set as the level of statistical significance.
## Results
The annual incidences of hRSV for 2016, 2017 and 2018 were 1.19% (34/2855), 0.33% (16/4847) and 1.59% (18/1126), respectively, and those of HMPV were 0.14% (4/2855), 1.79% (87/4847), and 0.44% (5/1126), respectively. The frequency of symptoms in patients infected with human Respiratory Syncytial Virus (hRSV), Human Metapneumovirus (HMPV), Influenza A (H1N1) and Influenza A(H3N2) are categorized into upper respiratory tract infections (URTIs), lower respiratory tract infections (LRTIs), and gastrointestinal symptoms. Coryza (runny nose) was most common in HMPV (85.4%), followed by Influenza A(H3N2) (72.3%), Influenza A(H1N1) (74.6%), and hRSV (70.6%). Sore throat was reported more frequently in Influenza A(H1N1) (56%) and Influenza A(H3N2) (55.2%) patients than in hRSV (50%) patients. LRTI symptoms such as cough were common across all viruses, with the highest frequencies of HMPV (97.9%) and Influenza A(H1N1) (97.3%). Shortness of breath was most reported in HMPV (22.9%), whereas hRSV, Influenza A(H1N1), and Influenza A(H3N2) showed similar frequencies (16-19%). The rate of chest pain was highest for HMPV (23.9%), followed by Influenza A(H3N2) (22.3%), while Influenza A(H1N1) (10.6%) had the lowest rate. Gastrointestinal symptoms such as nausea and vomiting were most common in Influenza A(H3N2) patients (46.1% and 67.1%, respectively), followed by Influenza A(H1N1) patients (37.3% and 32%). Chills and myalgia (muscle aches), headache and weakness were very common across all infections, ranging from 84.2 to 91.6% (Table 1).
The mean age of the hRSV patients was 40.26 years, and that of the HMPV patients was 34.4 years. Leucopenia was most frequently observed in Influenza A(H1N1) patients (31.8%), whereas thrombocytopenia was most common in hRSV patients (28%). Neutrophilia (> 70%) was more common in Influenza A(H1N1) (52.1%) and Inflammatory markers, such as C-reactive protein (CRP), were elevated in nearly half of the patients across all infections (Table 2). Notably, the total leukocyte count, platelet count, alkaline phosphatase (ALP) level and total protein level significantly differed among the hRSV and HMPV groups. The platelet count is significantly greater in HMPV compared to Influenza A(H1N1) and hRSV (Fig. 1). The total leukocyte count is also significantly greater in HMPV compared to Influenza A(H1N1), hRSV and Influenza A(H3N2). ALP levels are lower in hRSV than in the other groups, and total protein levels are significantly lower in Influenza A(H1N1) patients. Additionally, body temperature is significantly elevated in individuals with Influenza A(H1N1). Other parameters, such as the CRP level, erythrocyte sedimentation rate (ESR), respiratory rate, and diastolic blood pressure, did not significantly differ among the groups.
Coryza and sore throat are consistently highly prevalent across all infections, particularly in the early days. Shortness of breath is initially low but increases notably in HMPV cases by days 6-8. Chest pain, nausea, vomiting, and abdominal pain exhibit fluctuating patterns, with some spikes observed later in the illness. Chills and myalgia remain prevalent throughout, especially in Influenza A(H3N2) and Influenza A(H1N1) cases, with myalgia showing the highest overall consistency. Joint pain and headache also followed variable trends, but headache remained notably high in all groups. Overall, while some symptoms, such as cough, chills, and myalgia, are common and persistent, others, such as shortness of breath and gastrointestinal symptoms, are more variable, potentially indicating differing disease severity and symptom progression among viral infections (Fig. 2).
The total leukocyte count remains relatively stable, with a slight decline in some cases, whereas the neutrophil percentage decreases over time, particularly in hRSV cases, with a corresponding increase in the lymphocyte percentage. Platelet levels fluctuate, with noticeable dips in hRSV and Influenza A(H1N1) cases. Liver function markers, including AST and ALT, show intermittent peaks, particularly in hRSV and Influenza A(H1N1) infections, whereas ALP levels exhibit variable spikes. Inflammatory markers such as ESR and CRP fluctuate, with CRP peaking prominently for around 3-5 days in hRSV and HMPV patients. Urea and creatinine levels remain relatively stable, with slight elevations in later illness stages, particularly in HMPV and Influenza A(H3N2) patients. Body temperature remains at approximately 99-100 °F across all infections, with minor fluctuations (Fig. 3).
## Discussion
Human Respiratory syncytial virus (hRSV) and human metapneumovirus (HMPV) are significant viral pathogens that cause acute respiratory infections (ARIs) in children and elderly adults [9,10]. While hRSV has been extensively studied in infants and elderly individuals, its impact on otherwise healthy adults remains underappreciated. HMPV, a paramyxovirus closely related to hRSV, has also been identified as an important contributor to respiratory disease in adults. Studies have shown that the annual incidence of hRSV in adults ranges between 2% and 7%, whereas HMPV infections are observed in 1-5% of cases [3,10]. Clinically, hRSV, HMPV and influenza present with overlapping symptoms, including cough, fever, coryza, and gastrointestinal involvement. Lower respiratory tract symptoms such as cough and shortness of breath are frequently observed across infections, with HMPV showing the highest rates of cough (97.9%), corroborating previous reports on its significant involvement in bronchiolitis and pneumonia. However, hRSV tends to cause more severe lower respiratory tract involvement, particularly in individuals with comorbidities such as asthma, chronic obstructive pulmonary disease (COPD), and cardiovascular disease [11]. HMPV infections, while similar in presentation, are more frequently associated with bronchiolitis-like illness in adults. Hospitalized Influenza A(H1N1) patients exhibit leukopenia (31.8%), with thrombocytopenia being most common in hRSV patients (28%). Liver function abnormalities, particularly elevated AST and ALT levels in hRSV and Influenza A(H1N1) patients, are in agreement with prior studies showing hepatic involvement in severe viral respiratory infections. Elevated inflammatory markers, such as C-reactive protein (CRP) and the erythrocyte sedimentation rate (ESR), along with liver function abnormalities, have been reported in some studies, particularly in patients with severe infections [12]. Additionally, hypoxia and hypotension are observed in severe hRSV patients requiring intensive care. The findings of this study highlight distinct hematological and biochemical profiles associated with hRSV, HMPV, Influenza A(H1N1) and Influenza A(H3N2) infections. Elevated leukocyte and platelet counts in HMPV suggest a stronger inflammatory response, whereas lower total protein levels in Influenza A(H1N1) may indicate disease severity. Significant variations in ALP further distinguish these infections. These differences could aid in the clinical differentiation and management of viral respiratory infections.
Symptom progression and laboratory trends provide further insight into disease severity and trajectory across infections. The increase in shortness of breath in HMPV cases from days 6-8 parallels the literature highlighting prolonged lower respiratory involvement in HMPV than in hRSV and influenza. The persistence of chills, myalgia, and headache across all infections, particularly in influenza cases, aligns with reports that systemic symptoms are more pronounced in influenza patients because of their greater cytokine response [13]. Additionally, fluctuating platelet levels in hRSV and Influenza A(H1N1), along with intermittent peaks in AST and ALT, mirror previously documented cases where these infections result in transient hematological and hepatic dysfunction. The relatively stable renal function markers with minor late-stage elevations in HMPV and Influenza A(H3N2) are consistent with reports indicating minimal direct renal involvement in these infections.
The laboratory diagnosis of hRSV and HMPV is achieved primarily through molecular techniques such as real-time polymerase chain reaction (RT-PCR), which is more sensitive than traditional viral culture or antigen This study presents a comprehensive comparison of the clinical symptoms, demographic characteristics, laboratory findings, and temporal trends associated with hRSV, HMPV, Influenza A(H1N1), and Influenza A(H3N2) infections in adults over multiple years. While most existing studies on respiratory viral infections have focused on children or elderly individuals, detailed incidence rates of hRSV and HMPV in adults have rarely been reported. The observed seasonal variation in incidence, such as the increase in HMPV from 0.14 to 1.79%, is not well described and is a notable epidemiological finding. Furthermore, this study provides dynamic insights into disease progression and the kinetics of both clinical symptoms and laboratory parameters.
This study has several limitations: the sample size and assessment of symptoms and laboratory markers of disease progression were relatively limited and confined to the first 8 days of illness, and the data were derived on the basis of onset of symptoms and the day of admission. This may not fully capture the clinical course, especially in patients with prolonged or biphasic symptoms. Furthermore, coinfections with bacterial pathogens were not evaluated, which could have influenced both the clinical presentation and laboratory findings.
## Conclusion
Overall, hRSV, HMPV, and influenza A subtypes H1N1 and H3N2 share overlapping clinical features, and their progression, severity, and systemic impact differ, potentially guiding targeted clinical management strategies. Highlights the increased awareness among medical practitioners of this disease, and timely testing and diagnosis This study highlights the necessity for virus-specific diagnostic, therapeutic, and surveillance strategies in managing acute viral respiratory infections. These findings offer valuable clinical and public health insights into the distinct symptomatology, laboratory profiles, and disease progression associated with hRSV, HMPV, and influenza A subtypes H1N1 and H3N2. The consistently high prevalence of coryza and cough across all infections underscores their utility as initial screening symptoms. However, the presence of specific features such as chest pain and dyspnea in HMPV and gastrointestinal manifestations in influenza A infections can aid in early differential diagnosis. Laboratory abnormalities, including AST, ALT, total leukocyte count, and platelet counts in hRSV, HMPV, and influenza cases, suggest hepatic involvement, supporting the routine inclusion of liver function tests in the evaluation of febrile respiratory illness. Temporal shifts in lymphocyte and neutrophil counts, particularly in hRSV, may reflect stages of viral clearance or the onset of secondary bacterial infections, thereby informing antimicrobial stewardship practices. Finally, the substantial interannual variability in the incidence of hRSV and HMPV underscores the importance of sustained, adaptive surveillance systems to monitor epidemiologic trends and inform public health interventions.
This study provides valuable insights into the incidence and clinical-laboratory characteristics of influenza, hRSV, and HMPV infections in adults aged 18-65 years. While these viruses share overlapping initial symptoms, our findings highlight distinct patterns in their clinical presentation, laboratory profiles, and disease progression. Notably, HMPV was associated with a greater frequency of lower respiratory tract symptoms and a more pronounced inflammatory response than were hRSV and influenza. Furthermore, specific hematological and biochemical markers, such as elevated leukocyte and platelet counts in HMPV and liver enzyme abnormalities in hRSV and influenza, may aid in clinical differentiation. These findings underscore the importance of considering these distinctions for improved diagnostic algorithms, targeted clinical management strategies, and enhanced public health surveillance to mitigate the impact of these common respiratory viral infections in the adult population.
## References
1. Shi, Denouel, Tietjen et al. (2020) "Global Disease Burden Estimates of Respiratory Syncytial Virus-Associated Acute Respiratory Infection in Older Adults in 2015: A Systematic Review and Meta-Analysis" *J Infect Dis*
2. Ar F, Ee W (2005) "Respiratory syncytial virus infection in elderly and high-risk adults" *N Engl J Med*
3. Shirreff, Chaves, Coudeville et al. (2010) "Seasonality and Co-Detection of Respiratory Viral Infections Among Hospitalised Patients Admitted With Acute Respiratory Illness-Valencia Region"
4. Macias, Mcelhaney, Chaves et al. (2021) "The disease burden of influenza beyond respiratory illness" *Vaccine*
5. Govindakarnavar (2017) "Annual Report of Hospital Based Surveillance of Acute Febrile Illness in India"
6. Gill, Kaziev, Mtaweh et al. (2025) "Performance of the World Health Organization (WHO) severe acute respiratory infection (SARI) case definitions in hospitalized children and youth: cross-sectional study"
7. Kanda (2013) "Investigation of the freely available easy-to-use software EZR for medical statistics" *Bone Marrow Transplant*
8. Akhras, Weinberg, Newton (2010) "Human metapneumovirus and respiratory syncytial virus: subtle differences but comparable severity" *Infect Dis Rep*
9. Widmer, Griffin, Zhu et al. (2014) "Respiratory syncytial virus-and human metapneumovirus-associated emergency department and hospital burden in adults" *Influenza Other Respi Viruses*
10. Falsey, Walsh, Esser et al. (2019) "Respiratory syncytial virus-associated illness in adults with advanced chronic obstructive pulmonary disease and/or congestive heart failure" *J Med Virol*
11. Falsey, Mcelhaney, Beran et al. (2014) "Respiratory Syncytial Virus and Other Respiratory Viral Infections in Older Adults With Moderate to Severe Influenza-like Illness" *J Infect Dis*
13. García-Azorín, Santana-López, Lozano-Alonso et al. (2024) "Factors associated to the presence of headache in patients with influenza infection and its consequences: a 2010-2020 surveillance-based study" *J Headache Pain*
14. Shi, Mcallister, Brien et al. (2017) "Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in young children in 2015: a systematic review and modelling study" *Lancet* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12587420&blobtype=pdf | # Whole-Genome Sequencing of Hepatitis B Virus Genotypes E and A in Zambia Reveals Limited Viral Diversity in HIV Coinfection
Michael Vinikoor, Andreas Walker, Bright Nsokolo, Taonga Musonda, Guy Muula, Eleftherios Michailidis, Gilles Wandeler, Nadia Alatrakchi, Paul Kelly, Maximillian Damagnez, Duyen Le, Anja Voges, Nadine Lübke, Annie Kanunga, Samuel Bosomprah, Debika Bhattacharya, Carolyn Chibundi, Given Bwalya, Kalo Musukuma-Chifulo, Aleksei Suslov, Martin Feuerherd, Markus Heim, Robert Schwartz, Raymond Chung, Georg Lauer, Edford Sinkala, Jörg Timm
## Abstract
Background. The molecular characteristics of hepatitis B virus (HBV) in Africa, including the impact of HIV coinfection, are poorly understood.Methods. We performed whole-genome sequencing (WGS) on biospecimens collected before antiviral therapy in a wellcharacterized cohort of adults with HBV in Zambia, enriched for HIV coinfection (HBV/HIV). We assessed the frequency of basal core promoter (BCP) and precore variants, substitution frequencies, and the ratio of nonsynonymous to synonymous substitutions (dN/dS ratios), a surrogate for selection pressure.Results. Among 215 participants (median age, 33 years; 36% e antigen [HBeAg] positive, 35% with HBV/HIV), 114 (53.0%) had viral genotype E (gtE), and 101 (47.0%) had genotype A (gtA), subgenotype 1. BCP and precore variants, associated with HBeAg negativity, were more common with increased age, in the absence of HIV, and with gtE. Distinct from gtA, gtE had dN/dS ratios that were increased in the core vs polymerase region. Low dN/dS ratios were observed in HBV/HIV, especially at the lowest CD4 T-cell frequencies. Sequences from acute HBV infection as well as from 5 participants with chronic HBV/HIV who cleared hepatitis B surface antigen early during tenofovir-based antiretroviral therapy showed remarkably low dN/dS ratios.Conclusions. HBV gtE exhibited distinct substitution patterns compared with gtA, and HBV/HIV was associated with reduced HBV sequence diversity, consistent with impaired immune pressure.
The Africa region should be a major focus of global hepatitis elimination because of its high prevalence and incidence hepatitis B virus (HBV) infection and significant gaps in deploying evidence-based strategies to control viral transmission, mortality, and morbidity. To support viral elimination, a better understanding of HBV virology is needed in Africa and in general [1]. HBV in Africa may have important viral characteristics that are distinct from other regions. For example, HBeAg seroconversion from positive to negative, signaling a major transition in virus-host interactions, occurs at a significantly younger age in Africa compared with Asia [2][3][4]. HBV-related HCC has been reported to occur at a younger age in Africa [5], possibly underpinned by variants of genotype A (gtA) [6,7]. Unfortunately, the major HBV genotype in Africa, genotype E (gtE), is particularly neglected, with conflicting reports on whether its viral characteristics drive high mother-to-child transmission in the region [8]. While rarely seen in non-African sequencing studies [9], HBV gtE has been characterized by lower sequence diversity compared with other HBV genotypes, possibly as a consequence of its more recent emergence in West Africa ∼150-200 years ago [10]. Previous analyses of gtE often used partial viral sequencing and rarely included comparable populations (eg, shared ethnic background and environment) with another genotype.
HBV in Africa is also characterized by a high burden of coinfections, including ∼2 million individuals with HIV-1 (HBV/HIV), which is associated with a higher rate of HBV chronicity and faster liver disease progression [11,12], yet comparisons of HBV sequence features in HBV/HIV and HBV alone are scarce. Compared with HBV alone, HBV/HIV has been reported to have different HBV genotype distribution and mutational frequencies, including those associated with altered HBeAg production, and less HBsAg clearance [13][14][15]. Surprisingly, during the initial phase of treatment of HBV/HIV with tenofovir-based antiretroviral therapy (ART), HBsAg seroclearance appears to be higher than expected [16], a phenomenon that is poorly understood. High rates of HBsAg loss make HBV/HIV a potentially instructive clinical stage for understanding the host and viral mechanisms of HBV cure.
In this paper, we use whole-genome sequencing (WGS) to analyze HBV in a large and well-characterized cohort of adults in Zambia, where both HBV gtE and gtA circulate [17]. Zambia has a 6% adult prevalence of chronic HBV [18]. Compared with sequencing a limited number of genes, WGS provides higherresolution assessment of viral evolution across different genomic regions. This facilitates understanding of the genomic elements of selection pressure whether from nucleos(t)ide analogs (NAs), other therapies, natural immunity, or vaccines. The Zambian cohort uniquely includes people with chronic HBV alone, HBV/HIV, and acute resolving HBV infection. Furthermore, the cohort routinely ascertains HBsAg seroclearance among people with HBV/HIV taking tenofovir-based therapy, at a rate of 9% at 2 years and 15% at 5 years of treatment [19]. In this paper, we describe viral genotypes, molecular epidemiology, drug resistance, and viral diversity in the context of HBV/HIV, in acute HBV, and among people with chronic HBV/HIV that evolved to HBsAg loss during treatment. We hypothesize that gtE has distinct viral genetics from gtA and that viruses captured during acute and HIV coinfection, particularly in those who subsequently clear sAg, have lower viral evolution due to less immune selection pressure.
## METHODS
## HBV Cohort in Zambia
We applied WGS to plasma and serum samples from a longstanding HBV cohort in Zambia. At inception in 2013, cohort eligibility included newly diagnosed adults (age 18+ years) with HBV/HIV, defined as HIV-1/2 antibody and HBsAg positive in blood, without ART or who had received up to 1 month of antiviral therapy, and were not known to have hepatitis C. From 2016, eligibility expanded to include HBV alone (HBsAg-positive, HIV-negative) irrespective of treatment history. At enrollment, in addition to demographic characteristics and district of birth, we assessed HBV DNA (Roche cobas or Cepheid Xpert), liver enzymes (alanine aminotransferase [ALT] and aspartate aminotransferase [AST]), HBeAg (Diapro or Diasorin), hepatitis delta antibodies (Diapro or Diasorin), and we stored aliquots. In those with HBV/HIV, we also determined CD4 T-cell count and plasma HIV RNA concentration. Acute HBV was defined by hepatitis B core immunoglobulin M positivity in a participant with a typical clinical presentation. For this analysis, we excluded individuals with treatment of HBV alone before enrollment, unknown or undetectable (<10 IU/ml) HBV DNA at enrollment, insufficient stored samples, and/or enrollment HBV DNA levels of 10-500 IU/mL.
## Patient Consent
The study was approved by the ethics committees of University of Zambia and University of Alabama at Birmingham. All participants provided written informed consent.
## Amplification and Sequencing of HBV and HIV
Viral nucleic acid from 300μL of serum or plasma was extracted automatically using the Maxwell RSC Blood DNA Kit on a Maxwell RSC 48 Instrument (both Promega), and the complete HBV genome was amplified in 2 fragments as previously described [20]. In brief, 2-step nested polymerase chain reactions were performed for the core region (nt 1683-nt 2399; 717 bp according to the reference genome NC_003977.2) and the polymerase region (nt 2299-nt 1798; 2682 bp according to the reference genome NC_003977.2). Amplicons were pooled, and a library prep for Oxford Nanopore sequencing was done. Libraries were sequenced on R.10.4.1 flowcells using the R10.4.1 chemistry. Basecalled HBV reads were filtered by the ARTIC pipeline (version 1.2.4; https://github.com/articnetwork/fieldbioinformatics) according to expected amplicon lengths, gaining 2 subsets of reads corresponding to the core and polymerase regions. The length-filtered reads were separately mapped to a file containing references of all HBV genotypes using minimap2 (version 2.28) [21]. References starting with the authentic EcoRI-start or with the preCore were used for mapping core or polymerase, respectively. The numbers of primary alignments mapping to the core or polymerase region were counted for each reference. The reference with the highest number of primary alignments was determined for each region and, if identical for both regions, used for consensus sequence generation with the ARTIC pipeline. Length-filtered reads were aligned to the reference, and primers were trimmed. Variants were called with medaka, and a consensus sequence was generated for each region. Generated consensus sequences of the core and polymerase regions were merged if the overlaps were identical. HIV sequencing was performed with a protocol for routine resistance genotyping [22].
## Statistical and Bioinformatic Analysis
We compared the demographic and clinical characteristics of participants with HBV alone and HBV/HIV using chi-square and Wilcoxon rank-sum tests. We compared HBeAg by HBV genotype using multivariable logistic regression adjusted for confounders. HBV consensus sequences were aligned with the software Geneious 10.2.6 (RRID:SCR_010519) using MAFFT (PMID: 23329690). For phylogenetic analysis, a tree based on the complete HBV sequence, with references from Genebank, was calculated with the Mr. Bayes plugin [23] using the ngphylogeny pipeline (https://ngphylogeny.fr/). For visualization, the output was exported as a Newick file with support values and visualized with iTol [24]. For calculation of the ratio of nonsynonymous to synonymous substitutions (dN/dS ratio), HBV consensus sequences were separated by genotype, and separate alignments for core and polymerase were generated. These alignments were used for calculation of dN/dS ratio utilizing the SNAP tool from the HIV sequence database (www. hiv.lanl.gov). Resistance-associated mutations for HBV and HIV (if applicable) and immune escape mutations in HBsAg were analyzed with Geno2pheno. We evaluated for recombinants using SimPlot graphs and the phi(w) statistical test [25]. All sequence data are publicly available at Zenodo under https://doi.org/10.5281/zenodo.17396629.
## RESULTS
During 2013-2024, 840 HBsAg-positive adults enrolled in the cohort. We excluded from analysis 40 for prior treatment of HBV alone, 178 for unknown or undetectable (<10 IU/mL) HBV DNA levels pretreatment, 274 for insufficient stored samples, and 95 for low HBV DNA levels (between 10 and 500 IU/mL). Aliquots of plasma/serum from the remaining 253 were processed for WGS, and 215 (85.0%) high-quality sequences were generated. In addition to having higher HBV DNA levels, participants whose sequences were included in the final analysis had similar demographics (age and sex, both P > .05) to the overall cohort but were less likely to be HIV negative (32.7% vs 60.3%) and to have higher ALT (median, 31 vs 26; P < .001). The median HBV DNA in the analysis cohort (interquartile range [IQR]) was 4.94 (3.54-7.05) log 10 IU/mL, and 77 (39.7%) were HBeAg positive by serology. Seventy-five (34.9%) participants had HBV/HIV, and in that subgroup, the median CD4 count (IQR) was 133 (65-244) cells/mm [3]. None of the participants tested positive for hepatitis delta antibodies. While age and sex were similar, people with HBV/HIV had higher HBV DNA than those with HBV alone (6.5 vs 4.1 log 10 IU/mL; P < .001) (Table 1).
Within the analysis cohort, 12 individuals had acute HBV (including 2 with HBV/HIV) that resolved within 1 year of follow-up. Among the 75 participants with HBV/HIV, 5 evolved to clear HBsAg within 2 years of ART. WGS of HBV revealed 12 sequences with putative immune escape mutations in the HBsAg (6 in gtA and 6 in gtE) and 2 (0.9%) participants with lamivudine drug resistance mutations (Supplementary Table 1). Among the 75 participants with HBV/HIV, 53 had sufficient material to also undertake HIV sequencing. Sequencing of the polymerase/protease was successful for 42 (79.2%) samples, and sequencing of the integrase was successful in 45 (84.9%) samples. All sequenced HIV viruses were clade C. Among these, 16 participants had substitutions in the polymerase and 2 had substitutions in the integrase, which have been associated with reduced susceptibility to antiretroviral treatment (Supplementary Table 1). No resistance-associated substitutions were detected in the protease region. HBV recombinants were absent from the data set based on examination of SimPlots (Supplementary Figure 1B-D) and in the statistical evaluation of both gtE (P = .997) and gtA (P = .571).
## Phylogenetic Analysis of HBV Genome Sequences
Figure 1 presents a detailed phylogenetic tree illustrating HBV genotypes, HIV coinfections, acute HBV infections, and cases of HBsAg loss following ART initiation. Two HBV genotypes were detected with similar prevalence: 114 sequences (53.0%) were gtE, while 101 (47.0%) were gtA, all belonging to subgenotype A1. The distribution of HBV genotypes was not statistically different between individuals with and without HIV. Among the 75 people with HBV/HIV, 45 (60%) had gtE and 30 (40%) had gtA. There was no apparent phylogenetic clustering of HBV sequences from individuals with HBV/HIV to suggest different HBV transmission features from counterparts with HBV alone. Interestingly, 11 of 12 (91.7%) acute infections analyzed were gtE. With 2 exceptions, sequences from acute infection were dispersed throughout the phylogenetic tree. In 2 cases, viruses from acute infections clustered closely together. Among the 5 individuals with HBV/HIV who achieved HBsAg loss following antiviral therapy initiation, 3 had gtE and 2 had gtA.
## HBeAg Status by Genotype, Age, and HIV Status
HBeAg serological testing results were available for 194 (90.2%) participants. HBeAg positivity was more frequent in participants with gtA (41.6%) than in those with gtE (30.7%), though the difference was not statistically significant (P = .0778) (Figure 2A). HIV had a significant influence on HBeAg status. In gtA, HBeAg positivity was observed in 24 of 63 (38.1%) people with HBV alone, compared with 18 of 27 (66.7%) in counterparts with HBV/HIV (P = .0204). In gtE, 10 of 60 (16.7%) with HBV alone were HBeAg positive, compared with 25 of 44 (56.8%) with HBV/HIV (P < .0001) (Figure 2B). In HBV alone, HBeAg-positive individuals were significantly younger than their HBeAg-negative counterparts (P = .0362) (Figure 2C). However, this was not seen in HBV/HIV, where people with HBeAg positivity were significantly older than those with HBV alone (P = .0194) (Figure 2D). In HBV/HIV, there was a nonsignificant trend toward lower CD4 T-cell counts in HBeAg-positive individuals (P = .0913) (Figure 2E).
## Precore Variants in HBV gtE vs gtA
We next analyzed the frequencies of substitutions in the BCP and precore region that impair HBeAg production [27]. This analysis focused on 2 key BCP substitutions (A1762T + G1764A), a variant precore start codon (ATG to any variant), and a substitution introducing a stop codon at precore codon 28 (G1896A) (Figure 3A). BCP/precore variants were detected in 40% of all gtA sequences and 51% of all gtE sequences.
Although the overall prevalence of BCP variants (A1762T + G1764A) was comparable between gtA and gtE, significant differences were observed in the frequencies of other variants. The variant precore start codon (ATG variant) was exclusive to gtA, detected in 12.2% of cases alone, but was absent in gtE. In contrast, the precore stop codon (G1896A) was the most frequent variant in gtE, occurring in 22.1% of cases alone and in 8.7% alongside BCP mutations, but it was only detected in 1 patient with gtA. Of note, this patient also had the compensatory C1858T mutation that is required for the G1896A [28].
Consistent with findings on HBeAg serostatus, HBV/HIV was associated with significantly higher frequency of isolates with the prototypic PC/BCP sequence than HBV alone in both gtA (P = .0105) (Figure 3B) and gtE (P = .0052) (Figure 3C).
## Distinct Substitution Frequencies in gtA and gtE With HBV/HIV
The median genetic distance for full-length sequences was higher for gtA than gtE (0.012 vs 0.008; P < .001). For the precore/core region ("core"; nt1814-2458), this was 0.014 for gtA vs 0.013 for gtE. For the polymerase ("pol"; nt2307-1623) region, it was 0.012 for gtA and 0.005 for gtE (Figure 4A). The Phylogenetic tree with HBV sequences from Zambia. Sequences from whole HBV genomes were aligned together with genotype reference sequences and the reported genotype A/E isolate (AB194949) [26] using MAFFT. The phylogenetic tree was calculated with Mr. Bayes using the ng-phylogeny pipeline (https://ngphylogeny.fr/). Genotypes are color-coded as indicated, and infection status is marked in squares outside the sequence name. Abbreviation: HBV, hepatitis B virus.
dN/dS ratio was significantly higher in the core region of gtE compared with its pol region (P < .0001) (Figure 4B). Moreover, the dN/dS ratio in the gtE core region was also significantly higher than in the gtA core region (P < .0001) (Figure 4B). Although dN/dS ratios were generally lower in gtA than in gtE, they were significantly reduced in the core region for both genotypes in individuals with HBV/HIV compared with those with HBV alone (gtA: P = .0059; gtE: P = .0286) (Figure 4C,D). Notably, no significant differences were observed in dN/dS ratios within the polymerase (pol) region between HBV alone and HIV/HBV (Figure 4E,F). While no significant correlation was found between CD4 counts and core dN/dS ratios (Figure 4G, H), a trend was seen for gtE viruses (Figure 4H). Higher HBV DNA levels were associated with lower dN/dS ratios for both genotypes (Figure 4I,J). With 1 exception, all individuals with acute infection had gtE. In most cases, sequences from acute infections exhibited low dN/dS ratios (Figure 4C,D). Notably, individuals with HBV/HIV who achieved HBsAg loss (Figure 4D, F) also displayed low dN/dS ratios.
## DISCUSSION
In a sentinel HBV cohort in Zambia, we used WGS to analyze HBV sequence diversity and substitution frequencies in the context of chronic and acute infection with and without HIV coinfection-induced immune suppression. HBV gtE had a significantly different profile from gtA, including a higher rate of variants leading to HBeAg-negative infection. gtE was also predominant in acute infection. We also found dN/dS ratios that were higher in the core region in gtE compared with gtA, providing evidence for stronger selection pressure on this region. HBV/HIV was associated with a much higher proportion of prototype sequences in the BCP and precore region vs HBV alone. In line with impairment of immune selection pressure, dN/dS ratios were lower in HBV/HIV compared with HBV alone. Interestingly, low dN/dS ratios were also observed in the context of acute resolving HBV and HBsAg loss during tenofovir-treated chronic HBV/HIV. Together these data shed new light on HBV gtE, a prevalent but neglected genotype in Africa, further evidence that immune pressure drives HBV viral evolution, and highlight the value of WGS in viral elimination.
In this study, HBV gtE exhibited lower overall genetic distance in phylogenetic analysis, consistent with its relatively recent emergence in humans and corroborated by findings from smaller studies [8,29,30]. We built on past understanding through comparative analysis of dN/dS ratios in gtA vs gtE, revealing 2 key findings: (1) dN/dS ratios were overall higher in gtE than in gtA, indicating that gtE was under greater selection pressure; and (2) among gtE viruses, dN/dS ratios were higher in the core region than in pol, suggesting stronger selective pressure on the core. This suggested a more dynamic evolutionary process and faster adaptation kinetics in gtE than in gtA, potentially due to its relatively recent emergence and subsequent adaptation to the human population. Our observations also aligned with the hypothesis that HBV gtE core adapts more rapidly than pol to selection pressure in humans [31].
We also described HBV sequences in people with HBV/HIV, building on our past description of higher HBV DNA levels and more HBeAg positivity in this group [32]. In this study, viral WGS revealed similar distribution of genotypes in HBV/HIV and HBV alone, which conflicted with other studies where people with HIV had unique HBV transmission patterns [33]. HBV/ HIV was also associated with strikingly lower dN/dS ratios in the HBV core region, suggesting altered HBV substitution dynamics in the context of immune suppression. Collectively, higher rates of HBeAg seropositivity, lower prevalence of BCP and precore variants, and reduced dN/dS ratios provided evidence of impaired immune pressure on HBV in HBV/HIV. HBV may revert to a prototype virus in the context of HIV. We also described the frequency of common mutations associated with HBeAg seroconversion, which is thought to occur earlier in life in Africa than Asia. gtE had multiple variants leading to HBeAg-negative infection, including the G1896A, which was previously reported to be rare in gtE compared with other genotypes [8]. The same G1896A was associated with reduced HBsAg seroclearance in Côte d'Ivoire [15]. We also described in Zambia that compared with gtA, gtE had lower prevalence of HBeAg positivity, seen in only ∼15% of people with HBV alone. This conflicted with a review by Kramvis suggesting that early life loss of HBeAg in Africa was more attributed to gtA than gtE [34] and a report from West Africa suggesting that gtE had lower genetic variability, including HBeAg loss, compared with gtA [30,33]. Supporting our findings was a report by the Hepatitis B in Africa Collaborative Network, where only 5.2% of adults with chronic HBV alone were HBeAg positive in West Africa, the region where gtE is most prevalent, compared with 17.8% in Southern Africa, where gtA is more prevalent [35 ]. Our results may differ from past reports because of the larger sample size, direct comparison of gtE and gtA in a population with a similar genetic background, and/or due to our use of WGS.
This report also supported the dN/dS ratio as a predictor of clinical outcomes including HBsAg seroclearance, which is the focus of the HBV cure research agenda. Decreasing HBV DNA levels were associated with higher dN/dS ratios, which may suggest that nonsynonymous substitutions have fitness costs to viral replication or are a sign of stronger immune pressure. Previous studies reported that the HBV core protein has the highest degree of adaptation to HLA class I-associated immune pressure by CD8 T cells [31]. This may drive high dN/dS ratios in the core region, something that is reduced by HIV-mediated immunodeficiency.
We also linked low dN/dS ratios to HBsAg loss in acute resolving infection and in people with chronic HBV/HIV who achieved this outcome during ART. As in HBV/HIV, participants with acute HBV had low viral diversity, which was previously described at the nucleotide level [36]. Low viral diversity may reflect the short duration in which the virus was exposed to immune pressure or that variants with multiple substitutions are less likely to be transmitted. The Zambian cohort has reported a relatively high rate of HBsAg seroclearance in chronic HBV/HIV [19], and pretreatment viruses from 5 of these individuals had low dN/dS ratios and low genetic distance from the consensus sequences. This finding is supported by a previous study where low viral diversity at the quasispecies level, based on low haplotype number, predicted HBsAg seroclearance during tenofovir treatment of HBeAg-positive genotype A and genotype D infections [37]. BCP and precore variants have been associated with reduced HBsAg seroclearance [38]. While the phenomenon of HBsAg seroclearance during ART-treated HBV/HIV is thought to be mediated by restoration of immunity (due to HIV RNA suppression), less is known about HBV viral aspects. When tenofovir-based ART is initiated, "unadapted" HBV with no or few substitutions may be more susceptible to the combined mechanisms of direct inhibition of reverse transcriptase by tenofovir and immune reconstitution, which can sometimes trigger ALT flares. In relation to acute resolving HBV, our findings of low dN/dS ratios are in line with past analyses, where deep sequencing of the surface region [39] and WGS analysis of variants from the quasispecies [40] revealed lower diversity in acute than in chronic HBV.
This work also shed light on HBV transmission in Africa, finding that nearly every acute HBV infection was gtE, unlike among chronic infections. Although the number of acute infections was too low for solid conclusions, gtE may be more common in recent adult infections. In India, genotypic differences between acute and chronic HBV were also described, with gtD being more associated with acute sexual transmission [41]. Unlike the comparison between acute and chronic infection, we did not see a statistical difference in HBV genotype between people with and without HIV. This conflicts with a report from Cameroon, where HBV gtA was more common than gtE in those with HIV [30]. Our data suggest that HBV and HIV are transmitted independently in Zambia; most people with and without HIV likely acquired HBV at a similar time (ie, primarily at birth or early childhood).
These data are important to the field for several reasons. First, they reveal that gtE in Africa may have unique transmission and highlight clinical aspects that warrant further understanding in the context of HBV elimination. Better understanding is needed, for example, on whether gtE is more easily transmitted from mother to child or sexually in the adult population, compared with other genotypes. High rates of precore variants in gtE may explain why CHB among adults in Africa is predominantly HBeAg negative. We did not find evidence that HBV transmission differs in people with HIV compared with those without it. Higher prevalence of HBV in people with HIV in Zambia might be instead due to viral reactivation. We also linked the dN/dS ratio to resolution of infection, raising this as a possible viral biomarker, which could play a role in clinical staging of CHB and should be further evaluated as a predictor of functional cure. Viral sequencing capacity has grown substantially since the COVID pandemic, and this analysis supports an expanded role for WGS. Albeit rare, HBV drug resistance and immune escape mutations were also detected in the Zambia cohort. It is important for the Africa region to have surveillance of mutations like these that could threaten viral elimination in the future.
This analysis had significant strengths, including access to a unique and relatively large cohort in Africa and the use of WGS, but also several limitations. One limitation was that some cohort participants did not end up in the final analysis due to insufficient samples or low levels of viremia. The minimum of 500 IU/mL HBV-DNA was selected to allow efficient sequencing and generation of high-quality WGS data. It is possible that specific HBV genotypes associate with distinct immune control and low viral load. In this case, focusing on samples with HBV-DNA concentrations above a certain threshold could have introduced a bias into the analysis of genotype distribution. Moreover, our analysis of genetic distance was limited by its cross-sectional nature and reliance on genotype-specific reference consensus sequences to estimate substitution rates. Determining precise substitution rates over time and identifying residues under selection pressure would require longitudinal viral sequencing, which is not feasible after antiviral therapy initiation. Additionally, host HLA types were unavailable, preventing direct linkage of specific substitutions to CD8 and CD4 T-cell immunity. Finally, the data presented in this paper come from a single-country, single-cohort population, limiting generalizability.
In summary, HBV WGS within a unique Zambian HBV cohort yielded new information on circulating HBV genotypes, transmission, and viral evolution under immune pressure.
## References
1. Mcnaughton, Arienzo, Ansari (2019) "Insights from deep sequencing of the HBV genome-unique, tiny, and misunderstood" *Gastroenterology*
2. Chu, Liaw (2007) "Chronic hepatitis B virus infection acquired in childhood: special emphasis on prognostic and therapeutic implication of delayed HBeAg seroconversion" *J Viral Hepat*
3. Shimakawa, Lemoine, Njai (2016) "Natural history of chronic HBV infection in West Africa: a longitudinal population-based study from The Gambia" *Gut*
4. Kenfack-Momo, Kenmoe, Takuissu (2022) "Epidemiology of hepatitis B virus and/or hepatitis C virus infections among people living with human immunodeficiency virus in Africa: a systematic review and meta-analysis" *PLoS One*
5. Yang, Aziz (2017) "Characteristics, management, and outcomes of patients with hepatocellular carcinoma in Africa: a multicountry observational study from the Africa Liver Cancer Consortium" *Lancet Gastroenterol Hepatol*
6. Cohen, Ghosh, Shimakawa (2020) "Hepatitis B virus preS2Δ38-55 variants: a newly identified risk factor for hepatocellular carcinoma" *JHEP Rep*
7. Kramvis, Kew (2007) "Molecular characterization of subgenotype A1 (subgroup Aa) of hepatitis B virus" *Hepatol Res*
8. Ingasia, Kinge, Kramvis et al. (2021) "The neglected genotype of hepatitis B virus" *World J Hepatol*
9. Andernach, Hunewald, Muller (2013) "Bayesian inference of the evolution of HBV/E" *PLoS One*
10. Ingasia, Kostaki, Paraskevis et al. (2020) "Global and regional dispersal patterns of hepatitis B virus genotype E from and in Africa: a full-genome molecular analysis" *PLoS One*
11. Falade-Nwulia, Seaberg, Snider (2021) "Outcomes of acute hepatitis B virus (HBV) in HIV infection with and without HBV-active antiretroviral therapy" *AIDS*
12. Audsley, Littlejohn, Yuen (2010) "HBV mutations in untreated HIV-HBV coinfection using genomic length sequencing" *Virology*
13. Nie, Deng, Lan et al. (2020) "Pre-S deletions are predominant quasispecies in HIV/HBV infection: quasispecies perspective" *Infect Drug Resist*
14. Boyd, Moh, Maylin (2018) "Precore G1896A mutation is associated with reduced rates of HBsAg seroclearance in treated HIV hepatitis B virus co-infected patients from Western Africa" *J Viral Hepat*
15. Audsley, Avihingsanon, Littlejohn (2020) "Long-term TDF-inclusive ART and progressive rates of HBsAg loss in HIV-HBV coinfection-lessons for functional HBV cure?" *JAIDS J Acquir Immune Defic Syndr*
16. Nsokolo, Kanunga, Sinkala (2017) "Stage of disease in hepatitis B virus infection in Zambian adults is associated with large cell change but not well defined using classic biomarkers" *Trans R Soc Trop Med Hyg*
17. (2016) "Zambia Population-Based HIV Impact Assessment (ZamPHIA): Final Report"
18. Vinikoor, Hamusonde, Muula (2024) "Long-term hepatitis B and liver outcomes among adults taking tenofovir-containing antiretroviral therapy for HBV/ HIV coinfection in Zambia" *Clin Infect Dis*
19. Walker, Schwarz, Brinkmann-Paulukat (2022) "Immune escape pathways from the HBV core18-27 CD8 T cell response are driven by individual HLA class I alleles" *Front Immunol*
20. Li (2018) "Minimap2: pairwise alignment for nucleotide sequences" *Bioinformatics*
21. Lübke, Jensen, Hüttig (2019) "Failure of dolutegravir first-line ART with selection of virus carrying R263K and G118R" *N Engl J Med*
22. Huelsenbeck, Ronquist (2005) "Bayesian analysis of molecular evolution using MrBayes" *Statistical Methods in Molecular Evolution. Statistics for Biology and Health*
23. Letunic, Bork (2016) "Interactive Tree of Life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees" *Nucleic Acids Res*
24. Bruen, Philippe, Bryant (2006) "A simple and robust statistical test for detecting the presence of recombination" *Genetics*
25. Kurbanov, Tanaka, Fujiwara (2005) "A new subtype (subgenotype) Ac (A3) of hepatitis B virus and recombination between genotypes A and E in Cameroon" *J Gen Virol*
26. Buckwold, Xu, Chen et al. (1996) "Effects of a naturally occurring mutation in the hepatitis B virus basal core promoter on precore gene expression and viral replication" *J Virol*
27. Kramvis (2016) "The clinical implications of hepatitis B virus genotypes and HBeAg in pediatrics" *Rev Med Virol*
28. Mulders, Venard, Njayou (2004) "Low genetic diversity despite hyperendemicity of hepatitis B virus genotype E throughout West Africa" *J Infect Dis*
29. Olinger, Venard, Njayou (2006) "Phylogenetic analysis of the precore/core gene of hepatitis B virus genotypes E and A in West Africa: new subtypes, mixed infections and recombinations" *J Gen Virol*
30. Schwarz, Ptok, Damagnez (2025) "HBV shows different levels of adaptation to HLA class I-associated selection pressure correlating with markers of replication" *J Hepatol*
31. Muula, Bosomprah, Sinkala (2023) "HBV viral replication markers and hepatic fibrosis in untreated chronic HBV infection with and without HIV coinfection in Zambia" *AIDS*
32. Yousif, Mudawi, Bakhiet et al. (2013) "Molecular characterization of hepatitis B virus in liver disease patients and asymptomatic carriers of the virus in Sudan" *BMC Infect Dis*
33. Kramvis (2018) "Molecular characteristics and clinical relevance of African genotypes and subgenotypes of hepatitis B virus" *S Afr Med J*
34. Riches, Vinikoor, Guingane (2023) "Hepatitis B in Africa Collaborative Network: cohort profile and analysis of baseline data" *Epidemiol Infect*
35. Alestig, Söderström, Norkrans et al. (2013) "Genetic diversity of genotype D3 in acute hepatitis B" *J Med Virol*
36. Wagner, Yuen, Littlejohn (2023) "Hepatitis B virus haplotype number at baseline is a predictive marker of functional cure during antiviral therapy for patients with genotypes A and D HBeAg-positive chronic hepatitis B" *Aliment Pharmacol Ther*
37. Bayliss, Yuen, Rosenberg (2017) "Deep sequencing shows that HBV basal core promoter and precore variants reduce the likelihood of HBsAg loss following tenofovir disoproxil fumarate therapy in HBeAg-positive chronic hepatitis B" *Gut*
38. Cremer, Hofstraat, Van Heiningen et al. (2018) "Genetic variation of hepatitis B surface antigen among acute and chronic hepatitis B virus infections in The Netherlands" *J Med Virol*
39. Yang, Huang, Chen (2015) "Characterization of full-length genomes of hepatitis B virus quasispecies in sera of patients at different phases of infection" *J Clin Microbiol*
40. Sarkar, Pal, Das (2015) "Virological characteristics of acute hepatitis B in Eastern India: critical differences with chronic infection" *PLoS One* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793657&blobtype=pdf | # P-2178. The Impact of the Respiratory Syncytial Virus (RSV) Vaccination Program: Uptake and Hospitalization Trends in Older Adults in England
William Bentley, Madeleine Smith, Femke Ahlers, Tristan Curteis
## Abstract
Background. Respiratory syncytial virus (RSV) is a common infectious disease which generally presents with mild symptoms; however, severe infections in children, older adults and immunocompromised people cause a substantial number of hospitalizations in England every year. The recent approval and rollout of RSV vaccines for older adults aimed to reduce this disease burden. Here, we examine the rollout of the RSV vaccine among older adults in England from September 2024, investigating vaccine uptake and its impact on hospitalization rates.
Methods. RSV vaccine uptake for people in England aged 75-79 was calculated for the first 34 weeks of the vaccine rollout period (1 st September 2024 to 20 th April 2025) using data from the Federated Data Platform of the National Health Service and 2023 Office for National Statistics population data. Hospital admission rates for RSV positive cases for people aged 75 and above were accessed through the United Kingdom Health Security Agency's Severe Acute Respiratory Infection Watch sentinel surveillance, data from this source was compared for the 2023-2024 to the 2024-2025 season.
Results. Cumulative RSV vaccine uptake in adults aged 75-79 reached 65.6% by April 2025, which was lower than influenza vaccine uptake levels in people aged 65 or above in England (79.7% by March 2025; Figure 1). Hospital admission rates for RSV positive cases for people aged 75 or above in the 2024-2025 season were comparable to the previous year, when an RSV vaccine was not available for the older adult population (Figure 2).
Conclusion. While RSV vaccine uptake this season was below that of the influenza vaccine, uptake would be expected to increase in future seasons with greater awareness of the immunization program. Although an immediate reduction in RSV hospitalizations was not observed in 2024-2025 compared to 2023-2024, continued efforts to increase vaccine uptake rates through improved program visibility and integration with other vaccination programs could maximize benefits. Ongoing surveillance of vaccine uptake and epidemiological monitoring of the disease is essential to understand the long-term impact of the recently rolled out RSV vaccines in England and to improve future vaccination strategies.
Disclosures. William Bentley, MBiochem, Costello Medical: Employee Madeleine L. Smith, PhD, Costello Medical: Employee Femke M. Ahlers, DPhil, Costello Medical: Employee
Department of Pediatrics , Niigata University Graduate School of Medical and Dental Sciences, Niigata, Japan, Niigata, Niigata, Japan 2 Niigata University Graduate School of Medical and Dental Sciences, Niigata, Niigata, Japan 3 Niigata Univeisty, Niigata, Niigata, Japan 4 Niigata University, Niigata, Niigata, Japan 5 The University of Tokyo, Tokyo, Tokyo, Japan
$$1 2 3 3 3 4 5 4 1$$ |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793371&blobtype=pdf | # P-2182. HPV Genotyping Prevalence and Risk Factors Among Women from Two Rural Communities of Peru
Lauren Jernstadt, Ruben Briceno, Neysa Miller, Olivia Slewa, Oms Ii, Giovanna Russano, Elizabeth Lossada-Soto
## Abstract
The efficacy of RSV antivirals used for post-exposure prophylaxis (PEP) is unknown. 10 2 on Day (D)0. RSV RT-PCR was performed on nasal washes collected twice daily on D2-12. If PCR-confirmed RSV infection had not occurred by D5am after RSV exposure, participants were randomized to receive daily oral EDP-323 high dose (600mg), low dose (200mg with 600mg loading dose) or PBO for 5D. PEP efficacy was evaluated in this pre-specified population using Fisher's Exact test (two-tailed) of RSV-uninfected vs infected (pre-defined as PCR-positive on 2 consecutive specimens).
Methods. A cross-sectional study was conducted in 2024. We collected cervical cell samples from La Libertad and Loreto, with different ethnicities in two different regions. A total of 229 (La Libertad: 115,Loreto: 114) women between the ages of 13 and 60 years undergoing cervical screening were enrolled in the study. All samples were analyzed by polymerase chain reaction, and HPV presence and genotyping were obtained. Demographic, sexual history, and risk factors data were collected using a validated questionnaire. Multivariate logistic regression was performed for the statistical analysis.
Results. HPV overall prevalence was 31.0%. Loreto 36.0%; La Libertad 26.1% (p=0.012). The High Risk (HR) genotypes 16, 18, and 45 were found in 34%, 22%, and 18%, respectively. Early onset of sexual intercourse has a strong prediction for HPV infection. (95% CI 1.98-5.18, p< 0.001), STI diagnosis (aOR=2.85, 95% CI: 1.90-4.28). consistency of condom use (aOR= 2.10, 95% CI: 1.45-3.05).
Conclusion. HPV genotypes and prevalence were very high in our study compared to the national statistics in Peru. Loreto has a high incidence (31%) compared to La Libertad (26.1%) and appears to follow the prevalence trend observed in North America, with HPV type 16 accounting for cases. In regression analysis, early sexual intercourse onset (< 14 years) was the most significant risk factor for HPV infection. Our findings evidence the need for a national program with targeted screening in high-risk population.
Disclosures. All Authors: No reported disclosures
Poster Abstracts • OFID 2026:13 (Suppl 1) • S1325 |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584716&blobtype=pdf | # Identification and biochemical characterization of a novel porcine circovirus associated with porcine respiratory and diarrheal diseases
Xianhui Liu, Lin Wang, Xinming Zhang, Yilong Liu, Yixuan Li, Zan Li, Zhi Geng, Leyi Zhang, Yanling Liu, Pengshuai Liang, Yuhui Dong, Zheng Xu, Heng Zhang, Changxu Song
## Abstract
Eukaryotic circular Rep encoding single-stranded DNA (CRESS DNA) viruses exhibit high genomic diversity. Four species of circoviruses within the genus Circovirus have been identified in pigs, including the non-pathogenic porcine circovirus type 1 (PCV1), and the pathogenic PCV2, PCV3, and PCV4. In this study, a novel single circular DNA virus was identified in pigs suffering from respiratory, diarrheal, and reproductive failure diseases. The viral Rep and Cap showed low amino acid sequence identities to the four PCV species, with the highest identity of 26.2% to PCV2 Rep and 20.8% to PCV4 Cap. The classification of CRESS DNA viruses primarily relies on Rep, and biochemical analyses of the viral Rep confirm its classification as a CRESS DNA virus. Thus, it was tentatively named porcine circovirus type 5 (PCV5). This tentative designation reflects its host and genomic organization; however, phylogenetic analysis indicates that it clusters outside the family Circoviridae and is most closely related to the fur seal faces-associated circular DNA virus (FSfaCV). The positive detection rate for PCV5 in tested samples was 54.4% (294 out of 540), with 24.3% (17 out of 70) of pig farms in China testing positive for the virus. A close correlation between the copy number of PCV5 and the severity of infectious diseases was observed. Moreover, a PCV5 virus-like particles (VLPs)-based enzyme-linked immunosorbent assay, which was developed, also elucidated that PCV5 has been circulating. We subsequently established an in vitro culture system for PCV5, successfully purifying the virus. Morphological identification revealed that the purified PCV5 has a diameter of approximately 20 nm, and the recombinant PCV5 Cap could assemble into VLPs similar to purified PCV5. These results indicate that PCV5, associated with porcine circovirus-associated disease, commonly circulates within Chinese swine.
IMPORTANCEThe single-stranded circular DNA virus, tentatively named porcine circovirus type 5 (PCV5), belongs to the circular Rep encoding single-stranded DNA (CRESS DNA) viruses. Notably, PCV5 is not classified within the same family as the previously reported PCV1-PCV4. PCV5, which was associated with porcine circovirus associated-disease, is widely prevalent in Southern China. PCV5 is characterized by a larger genome (2,904 nt), along with longer Cap (1,182 nt) and Rep (1,056 nt) regions, as well as an extended stem-loop structure (stem: 9 nt and loop: 17 nt). Importantly, we subsequently established an in vitro culture system for PCV5, successfully purifying the virus. Morphological identification revealed that the purified PCV5 has a diameter of approximately 20 nm. Additionally, a variant of the PCV5 Cap, lacking the N-terminal 63 residues, was expressed in Escherichia coli, and purified Cap could assemble into virus-like particles similar to purified PCV5. These results hold significant implications for the study of CRESS DNA viruses. The emergence of PCV5 warrants further study.
E ukaryotic circular Rep encoding single-stranded DNA (CRESS DNA) viruses, which encode replication-related proteins (Rep) and infect eukaryotic organisms, possess the characteristics of small viral genomes, diverse host ranges, high prevalence, and strong rolling circle replication affinity. CRESS DNA viruses infect a wide array of different eukaryotes, such as plants, animals, humans, and fungi, and possess far-reach ing and important implications in virology (1,2). Those viruses are small non-enveloped DNA viruses with a single-stranded circular genome, primarily encoding two proteins: capsid-related protein (Cap) and replication-associated protein (Rep) (1,2). The Cap is the sole protein component of the virus particles and plays a crucial role in the entire virus replication cycle, participating in virus attachment, cell entry, genome uncoating, and the packaging of newly formed virus particles (1,3). The Rep protein facilitates the recognition of replication initiation within the viral genome sequence and pos sesses endonuclease activity, which cleaves the circular DNA and initiates rolling circle replication (1,3,4).
According to the International Committee on Taxonomy of Viruses, CRESS DNA viruses, the classification of which is based on the Rep amino acid, currently include seven family members, namely Circoviridae, Nanoviridae, Smacoviridae, Genomoviridae, Bacilladnaviridae, Geminiviridae, and Kirkoviridae (1,2). With the rapid advancement and application of new technologies such as metagenomic sequencing, novel types of CRESS DNA viruses are continuously being discovered and classified. Among the extensively studied CRESS DNA viruses, the family Circoviridae includes two genera: Cyclovirus and Circovirus (3), which are the pathogens causing porcine circovirus-associated diseases (PCVADs) (5) and beak and feather disease in birds (6), respectively. Notably, porcine circovirus type 2 (PCV2) and beak and feather disease virus (BFDV) are the most thoroughly researched CRESS DNA viruses to date (3,5,7). These viruses identified in pigs include PCV1 (5), PCV2 (5), PCV3 (8), PCV4 (9), and porcine circovirus-like virus (PCLV) (10,11). PCV2 has been associated with clinical diseases in pig farms known as PCVAD, causing substantial economic losses. At present, PCV2 and PCV3 are prevalent in the global pig industry (5,12).
In recent years, several new human CRESS DNA viruses have been discovered and studied. Researchers used metagenomic next-generation sequencing to identify an unknown species of circovirus, designated human circovirus 1 (HCirV-1), from a liver biopsy sample (13). Additionally, the researchers identified a novel circovirus, humanassociated circovirus 2 (HuCV2) from the blood of two intravenous drug users in China (14). Furthermore, the researchers reported a new Circovirus in a patient in France who had acute hepatitis of unknown origin using means of routine shotgun metagenomics (15).
Infections of unknown origin must be diagnosed promptly to ensure that appropriate measures are implemented in a timely manner to prevent the spread of potentially harmful pathogens and to facilitate effective treatment (15). Significant advancements have been made in next-generation sequencing metagenomic sequencing of viral communities, which has been extensively utilized for monitoring unknown diseases in humans, animals, plants, and other organisms (1,13,15). Additionally, designing universal primers based on uploaded metagenomic sequences in GenBank, along with existing research sequences, provides an efficient and cost-effective method for monitoring homologous viruses within a specific family or genus.
In this study, we identified a novel CRESS DNA virus, tentatively named porcine circovirus type 5 (PCV5). This tentative designation reflects its host and genomic organization; however, phylogenetic analysis indicates that it clusters outside the family Circoviridae and is most closely related to the fur seal faces-associated circular DNA virus (FSfaCV). PCV5 is associated with respiratory and diarrheal diseases in piglets, as well as reproductive failure diseases.
## RESULTS
## Epidemiologic investigation of a novel single-stranded circular DNA virus
A novel single-stranded circular DNA virus, tentatively named PCV5, was identified by designing universal primers based on uploaded metagenomic sequences in GenBank, along with existing research sequences. From October 2021 to June 2023, we identi fied 17 swine farms with pigs exhibiting respiratory, diarrheal, and reproductive failure diseases that were infected with PCV5. Pigs were classified as having PCV5 infection if PCV5 DNA was detected in clinical samples. This study found that 24.3% (17/70) of the pig farms tested positive for PCV5, indicating that PCV5 is widely prevalent in south China (Fig. 1A). However, this does not provide information on the transmission dynamics and epidemiology of PCV5. Further research on the prevalence of PCV5 in different areas and seasons is necessary.
## Sequencing and genetic analysis of the PCV5
Following PCR and Sanger sequencing of the resulting amplicons, we assembled a 2,904 nt circular genome from the intestinal homogenate of a piglet (Fig. 1B). The PCV5 genome analysis identified two open reading frames (ORFs) that encode proteins exceeding 300 amino acids (aa), with two ORFs showing significant homology to Rep and Cap of circoviruses by BLASTP, structural prediction, and comparison. The Cap is 1,182 nt, the Rep is 1,056 nt, and a stem-loop is 26 nt in length (stem: 9 nt and loop: 17 nt), which is significantly longer than PCV1, PCV2, PCV3, PCV4, and several human circoviruses (16) (Fig. 1B; Table S1 ).
## Genetic relationship of PCV5 strains to other CRESS DNA viruses
The amino acid sequence of PCV5 Rep was significantly different from those of PCV1, PCV2, PCV3, PCV4, PCLV, and several human CRESS DNA virus Rep, and the similarity was less than 26.2% (Fig. 2A). The similarity of Rep amino acid sequence between the 12 PCV5 strains ranged from 87.7% to 99.7%, and the similarity between these strains and the FSfaCV (accession KF246569.1) strain ranged from 85.8% to 92.6% (Fig. 2A). The novel circovirus identified in this paper is highly homologous to FSfaCV. The study of FSfaCV is limited, and its prevalence is unknown. This virus is closely related to PCVAD and is widely prevalent in pigs, which indicates that it may be an important pathogen, and for the convenience of subsequent studies, it is temporarily named PCV5. The amino acid of PCV5 Cap sequence was significantly different from those of PCV1, PCV2, PCV3, PCV4, PCLV, and several human CRESS DNA virus Cap, with a similarity of less than 20.8% (Fig. 2B). And the amino acid similarity of Cap between these 12 PCV5 strains and the FSfaCV strain ranges from 89.1% to 99.2% (Fig. 2B). These results indicated that PCV5 is a novel single circular DNA virus.
## Biochemical characteristics of PCV5 Rep
The classification of CRESS DNA viruses is primarily based on Rep. We subsequently investigated the biochemical functions of PCV5 Rep. It has been mentioned above that the amino acid sequence of the Rep and Cap of PCV5 is significantly different from those of PCV1, PCV2, PCV3, PCV4, PCLV, and several human CRESS DNA viruses. Subsequently, we predicted the Rep and Cap structures of PCV5 using Alpha-Fold2. The structure of PCV5 Rep is highly similar to the structure of PCV2 Rep (PDB: 5XOR, 7IAR); both possess the classic three domains: N-terminal endonuclease domain, oligomerization domain, and superfamily helicase/ATPase domain (Fig. 3A). The above results indicated that even with low sequence identities, the Rep structures among CRESS DNA viruses are highly conserved.
Then, we attempt to obtain the full-length recombination PCV5 Rep. However, the full-length Rep was expressed in Escherichia coli as an inclusion body form (Fig. 3B). According to the AF2 model, we truncated the N-terminal endonuclease domain of PCV5 Rep (termed as PCV5 Rep NΔ118), and the PCV5 Rep NΔ118 was soluble (Fig. 3C). Sizeexclusion chromatography showed that PCV5 Rep NΔ118 exists as an oligomeric form in solution (Fig. 3D). In order to investigate whether PCV5 Rep NΔ118 exhibits ATPase activity, the enzymatic kinetics of purified PCV5 Rep NΔ118 were studied using the malachite green phosphate colorimetric assay. The enzymatic kinetics data were well fitted to the Michaelis-Menten equation (Fig. 3E). Additionally, PCV5 Rep was mainly located in the cytoplasm, although a small part of Rep could enter the nucleus (Fig. S2A). Thus, the expression of purified PCV5 Rep NΔ118 exhibits ATPase activity. These bio chemical results further indicate that PCV5 Rep was similar to the classical Rep of CRESS DNA viruses, and PCV5 can be classified into CRESS DNA viruses.
## Phylogenetic analysis based on Rep
To further investigate the evolutionary relationship of PCV5 to other CRESS DNA viruses, we analyzed genome sequences from 49 representatives of the different family and 12 amino acid sequences of PCV5 Rep. The phylogenetic analysis (Fig. 4) indicates that PCV5 is most closely related to the FSfaCV (accession KF246569.1). The phylogenetic tree (Fig. 4) also suggests that PCV5 is genetically distinct from PCV1, PCV2, PCV3, PCV4, PCLV, and several human CRESS DNA viruses, indicating that PCV5 does not belong to the known family. Furthermore, PCV5 exhibits significant differences in genome length compared to PCV1, PCV2, PCV3, PCV4, PCLV, and several human CRESS DNA viruses. These results indicated that PCV5 is a novel CRESS DNA virus.
## Histological lesions associated with the presence of PCV5
To confirm the presence of PCV5 in porcine samples, a qPCR assay was developed to detect the PCV5 Rep gene (Fig. S2 to S4), which exhibited good specificity and repeata bility. The tissue homogenate samples from 12 piglets in Guangxi were strongly positive for PCV5, with cycle threshold (CT) values between 11.1 and 34 (Table 1). The virus CT values of 12 pigs were closely correlated with their clinical symptoms. When viral DNA levels were low, higher concentrations of PCV5 DNA were found in intestinal tissue, intestinal lymph nodes, and feces. However, as clinical symptoms of piglets become more severe, elevated levels of viral DNA were also detected in other organs. It was particularly notable to find that the virus can also cross the blood-brain barrier, as high levels of viral DNA were identified in brain samples. In addition, PCV5 DNA was detected in stillbirth piglets from other farms with outbreaks of reproductive disorders (data are not shown). These results demonstrate that PCV5 was detected in multiple organs and its association with PCVAD.
Then, four piglets, numbered 01, 06, 12, and 07, were selected and stained with hematoxylin and eosin (HE). The PCV5 CT value of the above four piglets gradually decreased (Table 1). The pathological examination of the four piglets revealed that the organ lesions were associated with the copies of PCV5 DNA (Fig. 5A), and the patholog ical injury of the ileum was serious in the four piglets. These results suggest a close correlation between the copy number of PCV5 and the severity of infectious diseases. Western blotting showed that the prepared polyclonal antibody can detect the viral protein in the colon sample of porcine (number 07) infected with PCV5 (Fig. 5B). These results also further confirm the association between PCV5 and histological lesions of piglets.
## Virus isolation and visualization
The PCV5 isolation from the intestinal tissue of piglets was performed in Marek's disease lymphoma cell line (MDCC-MSB1). After serial generations of the blind passages, the PCV5 isolate was obtained in MDCC-MSB1 cells (Fig. 6A). MDCC-MSB1 cells infected with PCV5 showed no cytopathic effects (Fig. S5A). For the discontinuous sucrose gradient centrifugation, the cell supernatant containing PCV5 was purified with a discontinuous 30%-50% (wt/vol in phosphate-buffered saline [PBS]) sucrose gradient (Fig. 6B).
Then, the purified sample was examined. SDS PAGE indicates that PCV5 exists in the band 1 (Fig. 6C), and western blot also indicates that PCV5 was distributed in the band 1 using PCV5 Cap polyclonal antibody (Fig. 6D). Subsequently, the sample containing PCV5 was visualized by negative-stained electron microscopy. Many non-enveloped icosahe dral particles of approximately 20 nm in diameter, morphologically with the characteris tic features of the family Circoviridae, were observed (Fig. 6E andF). In addition, we also observed virions under electron microscopy, after the sample was purified using continuous 10%-50% (wt/vol in PBS) linear sucrose gradient centrifugation (Fig. S5B through E). These results confirmed that we successfully established an in vitro culture system for PCV5, and PCV5 was successfully purified. The morphological identification of the purified PCV5 revealed that PCV5 is similar to PCV2.
## Self-assembly of PCV5 virus-like particles and their application in PCV5 seroprevalence
In order to obtain a substantial quantity of recombination Cap protein, we attempted to express PCV5 Cap in E. coli. However, we found that full-length Cap could not express in E. coli. We found the Cap structure of PCV5 is highly similar to the Cap structure of the published CRESS DNA virus (17)(18)(19)(20)(21)(22), which is a classic jelly roll (Fig. 7A). The jelly roll domain consists of two β-sheets, each of which contains four β-strands connected by loop (23)(24)(25). The above results indicated that even with low sequence identities, the Cap structures among CRESS DNA viruses are highly conserved. Additionally, full-length PCV5 Cap exists in the cell nucleus (Fig. S2B). PCV5 Cap could be expressed in large quantities after the deletion of 63 amino acids at the N-terminal (Fig. 7B). Size-exclusion chromatog raphy indicated that the truncated Cap exists as an oligomeric form in solution, and the molecular sieve peak was 8-9 mL (Fig. 7C). The purified PCV5 Cap was capable of self-assembly into virus-like particles (VLPs) (size: 17-22 nm), which was observed by transmission electron microscopy (Fig. 7D andE). Subsequently, PCV5 Cap virus-like particles with adjuvant were prepared into a vaccine and injected into rabbits to prepare polyclonal antibodies. The prevalence of anti-PCV5 Cap antibodies in porcine serum samples was assessed using an enzyme-linked immunosorbent assay (ELISA) with PCV5 Cap VLPs. The externally purified PCV5 Cap VLPs exhibit characteristics such as uniformity and stability. This study established an indirect ELISA serological detection method utilizing PCV5 VLPs, which exhibited good specificity and repeatability (Tables S2 andS3). Anti-PCV5 Cap antibodies were detected in 524 (66.84%) of 784 pig serum samples collected from multiple regions (Table 2). Among the positive samples, 298 originated from Guangdong, 56 were from Guangxi, 78 were from Hunan, and 92 were from Yunnan (Table 3). The results of PCV5 seroprevalence revealed that PCV5 has been circulating in Southern China.
## DISCUSSION
In this work, a novel single circular DNA virus identified from piglets is highly homolo gous to FSfaCV. The classification of CRESS DNA viruses primarily relies on Rep. The biochemical results about Rep of the virus further indicate that it can be classified into CRESS DNA viruses. Sequence alignment analysis revealed significant homology between PCV5 and FSfaCV. However, given the substantial difference in their primary host species, the study of FSfaCV is limited, and its prevalence is unknown, and the virus is closely related to PCVAD and is widely prevalent in pigs, which indicates that it may be an important pathogen; thus, the circovirus was named as PCV5. This nomenclature reflects the potential for distinct viral tropism between these two viruses. Furthermore, we cannot exclude the possibility of a cross-species transmission event as the origin of PCV5 (Fig. 8). Future studies will aim to provide further experimental evidence to elucidate the evolutionary origin of PCV5. We observed a close relationship between PCV5 and the severity of infectious disease. We then established an in vitro culture system for PCV5. Subsequently, PCV5 was successfully purified through sucrose gradient centrifugation, and morphological identification of the purified PCV5 revealed virus particles with a diameter of approximately 20 nm. We tried to establish an animal disease model in mice challenged with PCV5 (Fig. S6). However, the mice did not show any signs of infection, and PCV5 was not detected in multiple organs of the mice (Table S4), and all mice inoculated with PCV5 were alive during the observation period. These results indicated that PCV5 is not suitable for establishing a disease model in mice.
While much of the viral community on Earth remains largely unknown, there may be connections between some unknown human or animal diseases and CRESS DNA viruses. It is uncertain how close virologists are to uncovering the full extent of the diversity of CRESS DNA viruses. However, recent global efforts have more than doubled the number of CRESS DNA viruses in GenBank over the past decade (1, 2). There is no need to add new strain sequences to our understanding of the diversity of CRESS DNA viruses. These abundant sequences of novel CRESS DNA viruses come from a wide range of hosts and environments. The CRESS DNA viruses, in particular, have several characteristics that support their efficient transmission and spread, such as the ability to cross the placental barrier and blood-brain barrier, high viral copy numbers, and the stability of their circular DNA genomes (5,6,8,12). The researchers have better defined the diversity and prevalence of CRESS DNA viruses. The three longest-established families of CRESS DNA viruses include well-studied pathogens of animals and plants, although not all members of these families can cause disease in their hosts. The viruses of Geminiviridae and Nanoviridae infect plants, while the viruses of Circoviridae infect vertebrates (birds and mammals) and invertebrates (1). However, for other types of CRESS DNA viruses, in vitro propagation systems have not yet been established. The disease model and commercial vaccines for PCV2 have been widely accepted (5). However, other CRESS DNA viruses still face the challenge of Koch's postulates. Therefore, many studies focus on the level of gene molecular analysis, concluding that a certain CRESS DNA virus is closely associated with a particular disease rather than establishing a definitive cause-and-effect relationship. This ongoing discussion has led to the debate surrounding the pathogenicity of CRESS DNA viruses. Of course, many researchers have also detected viral antigens in samples from diseased tissues, further confirming the close connection to the disease (8,(13)(14)(15).
The logic of viral evolution is defined by the key biological features of viruses, namely their obligate intracellular parasitic lifestyle (26). This lifestyle provides viruses with ample opportunities to directly exploit and manipulate the host's cellular machinery, but it also necessitates overcoming the host's defense systems (27). Viral evolution appears to be seeded by a set of core proteins involved in genome replication (Cap and Rep), which seem to have ancestral origins dating back to pre-cellular stages of evolution (26)(27)(28). Generally, the larger the viral genome of the PCV5, the more auxiliary functions it encodes. The growth of viral genomes is a result of acquiring various auxiliary genes that enhance the adaptability or operational autonomy of the virus, although they are largely redundant with the functions of the host cell (27). The evolution of larger genomes is determined by improved infection efficiency, broader host range, potentially higher attachment success rates, and reduced decay rates-characteristics that are particularly important in resource-limited environments with low host population densities (27). CRESS DNA viruses serve as excellent models for studying biological efficiency, viral replication, and assembly in the natural process of evolution (3). The CRESS DNA virus is often composed of a single Cap, making virus particles (18)(19)(20)(21)(22)(23). The Rep protein, on the other hand, has an N-terminal endonuclease domain and a C-terminal ATPase and helicase domain (3,16,29). Although PCV1 is classified as a non-pathogenic agent, PCV2, PCV3, and PCV4 significantly affect the immune system of pigs and induce severe diseases (5,8,9,30). Some clinical symptoms caused by these viruses are collectively referred to as PCVAD (5). Over the past 20 years, PCV2 has been recognized as the primary cause of PCVAD, as its infection in pigs severely compromises the immune system, leading to immu nosuppression and lymphatic system failure (5). PCV2 primarily proliferates in lymph nodes, inducing apoptosis in immune cells, which further weakens the immunity of infected pigs (5). Evidence indicates that PCV3 can lead to some pathological symptoms, including post-weaning multisystemic wasting syndrome (PDNS), diarrhea, reproductive disorders, respiratory, systemic inflammatory diseases, and central nervous system signs (31,32).
In this article, we highlight that PCV5 is widespread in Southern China. To investigate the prevalence of PCV5 in healthy swine populations, surveillance studies were also conducted. We found PCV5 infection in clinically normal pigs, albeit with low viral titers and a low prevalence rate (data not shown). These findings suggest that PCV5 can establish subclinical infections in immunocompetent herds. Notably, during disease outbreaks on farms, PCV5 prevalence significantly increased. This pattern is consistent with potential latent infection, where viral reactivation may be facilitated by co-infection with other pathogens. Such synergistic interactions likely enhance PCV5 replication and pathogenesis, though experimental validation of this co-infection hypothesis requires further investigation. Regarding potential co-infection between PCV5 and PCV2/PCV3, a systematic analysis was not performed in this study. This knowledge gap stems from methodological differences: PCV5 was primarily detected in fecal samples, whereas PCV2/PCV3 screening relies predominantly on blood-based assays. Future studies will implement concurrent sampling of matched fecal and blood specimens to elucidate co-infection dynamics.
It is generally considered pathogenic and is associated with a variety of patholog ical symptoms similar to those caused by PCV2, including porcine diarrheal disease, respiratory disorders, and reproductive failure. Therefore, this novel CRESS DNA virus merits further investigation to clarify its significance and role in PCVAD.
## MATERIALS AND METHODS
## Samples and clinical background
In early 2022, the South China Agricultural University National Engineering Center for Swine Breeding Industry received clinical samples of different organs from approximately 12 piglets (5-20 kg) with varying degrees of respiratory and diarrheal disease. These piglets come from different litters. This farm has about 1,000 sows, and there is only a high mortality rate for farrowing piglets and nursery pigs. And the health status indicators of pigs are poor, always accompanied by respiratory diseases and diarrheal diseases. Pig farmers have not found the cause, and there are no good measures to change this phenomenon. Only PCV5 was positive, and quantitative real-time PCR (qRT-PCR) was developed to detect the loading titer of PCV5. Subsequently, we initiated a surveillance program on the epidemiology and pathogenicity of PCV5. The clinical samples, including respiratory, diarrheal, and reproductive failure diseases of pigs, were obtained from 70 swine farms in different regions of China from October 2021 to June 2023 and stored at -80°C.
## DNA/RNA extraction of viruses
Clinical samples were grounded using PBS at 4°C, and samples were repeatedly freezethawed three times. Subsequently, the viral DNA/RNA of the sample is extracted with RaPure Viral RNA/DNA Kit (Magen, R4410-3, China).
## PCR array
Microorganisms and viruses related to intestinal diseases, respiratory diseases, and reproductive disorders of pigs were detected by laboratory-preserved test methods (10,33). Multiple publicly available CRESS-DNA virus sequences were retrieved from genetic databases and subjected to sequence alignment. This analysis revealed conserved motifs within the Rep gene, which were subsequently used to design primers for amplification. Following PCR amplification using these conserved-region primers, the resulting nucleic acid products were sequenced and subjected to phylogenetic tree construction. Primers used to obtain PCV5 genome sequences were designed (Table S5).
## Real-time PCR array
To further investigate tissue tropism of PCV5 in piglets suffering respiratory and diarrheal disease, a SYBR green qRT-PCR targeting the conservative regions of PCV5 was devel oped based on the obtained PCV5 virus strains. And the designed primers were tested and showed good sensitivity, specificity, and reproducibility. Detailed information about gene primers used in the PCR and qRT-PCR is listed in Table S5.
## Phylogenetic analysis
The complete gene sequences of PCV5 viruses obtained in this article have been uploaded to GenBank with the accession numbers (Table S1). The virus genome was assembled using SnapGene. All arrangements were further aligned using the ClustalW alignment method in MegAlign (Lasergene). The phylogenetic tree was built using the maximum likelihood method and 1,000 bootstrap replicates with MEGA software. The amino acid sequence of the gene was compared by DNAman and SDT V1.2 software.
## Alpha-Fold2 predicts the structure of open reading frames
The complete three-dimensional structure of open reading frames obtained after whole-genome analysis was predicted using the artificial intelligence prediction software Alpha-Fold2. Alpha-Fold2 software, developed by DeepMind, was trained using artificial intelligence algorithms based on a data set of over 190,000 protein structures from the Protein Data Bank (PDB) (https://www.rcsb.org). This algorithm accurately predicts the three-dimensional structure of a protein based on its input amino acid sequence. On our local computing server, we input the complete amino acid sequence of the open reading frame and execute the Alpha-Fold2 command, which outputs five predicted structures. Based on the probability distribution, we selected the highest probability ranked_0.pdb file for further structural analysis. The analysis and domain delineation of the structure were performed using the PyMOL software.
## Cells and reagents
Dulbecco's modified Eagle medium (Gibco-BRL) containing 10% fetal bovine serum (Sorfa Life Science) and 1% penicillin-streptomycin (Gibco, Thermo Scientific) was used for maintaining MDCC-MSB1 cells, PK15 cells, and HeLa cells (ATCC CRL-11268), which were incubated at 37°C in 5% CO 2 .
## Virus purification
The cell medium was clarified by low-speed centrifugation (4,000 rpm) for 10 min at 4°C to remove the cell debris. For the discontinuous sucrose gradient centrifugation, the supernatant was purified with a discontinuous 30%-50% (wt/vol in PBS) sucrose gradient and centrifuged with an SW32.1 rotor (Beckman) at 100,000 × g for 3 h at 4°C. Fractions with PCV5 particles were collected. For the continuous sucrose gradient centrifugation, the continuous 10% to 50% (w/v in PBS) linear sucrose gradient was made using a Gradient Master (BioComp Instruments Inc., Canada). The concentrated virus was then placed on a 10%-50% linear sucrose gradient and centrifuged with an SW32.1 rotor (Beckman) at 140,000 × g for 14 h at 4°C. Fractions with PCV5 particles were collected. The purified PCV5 particles were then examined by western blot and imaged with negative staining EM.
## Negative stain
For examination by negative staining, an aliquot of 4 μL of purified virus was applied to freshly glow-discharged carbon-coated copper grids (Zhong Jing Ke Yi Corp., China). After 1 min, the excess liquid was removed using a filter paper. The grid was then stained with 2% uranyl acetate for 40 s and removed using a filter paper. All samples were examined on a Tecnai T12 electron microscope (FEI) operated at an acceleration voltage of 120 kV. Images were recorded using a CCD camera (Eagle, FEI).
## Construction of prokaryotic and eukaryotic plasmids
The full-length Cap and different truncated Cap constructs were cloned into a prokary otic expression plasmid. However, only the Cap protein truncated by 63 amino acids could be efficiently expressed and purified in pET-28a, and it was able to assemble into virus-like particles.
The full-length Cap was fused to the C-terminus of pCMV plasmid with a 3×FLAG tag and to the C-terminus of pCMV plasmid with a Twin-STREP tag. The full-length Rep was fused to the N-terminus of pCMV plasmid with a 3×HA tag. Detailed information on the gene primers used can be found in Table S6.
## Expression and purification of recombinant proteins
Transform the successfully constructed plasmids into BL21(DE3) competent cells. Inoculate a single positive colony into 5 mL of Luria-Bertani broth (LB) medium containing the appropriate antibiotic and incubate overnight at 37°C with shaking. Then, transfer 5 mL of the overnight culture into 1,000 mL of LB medium supplemented with antibiotics and continue shaking at 37°C for 4-6 h until the OD 600 reaches 0.6-0.8. Reduce the temperature to 16°C and add isopropyl β-D-1-thiogalactopyranoside (IPTG) to a final concentration of 0.2 mM, then continue shaking for 16-18 h to induce protein expression.
The total protein from the obtained bacterial culture was subjected to initial affinity nickel column purification, followed by ion exchange chromatography and size-exclusion chromatography, to obtain the desired protein with high purity and uniform oligomeric state.
## Indirect immunofluorescence assay
Indirect immunofluorescence assay allows for the visual observation of the expression and distribution of the target protein in cells. It involves the use of specific antibod ies (primary antibodies) that bind to the target protein, followed by the binding of fluorescently labeled secondary antibodies to the primary antibodies, generating fluorescence. The samples are then observed under a fluorescence microscope.
## Preparation of polyclonal antibodies against Cap protein and Western blotting
The recombinant Cap virus-like particle protein is administered to immunize rabbits four times with an interval of approximately 3 weeks, with each immunization consisting of 1 mg of Cap protein along with an adjuvant. After the immunizations, the serum from the rabbits is collected, which contains polyclonal antibodies against Cap protein. For the Western blotting assay, clinical samples are lysed, and the obtained rabbit serum is used as the primary antibody to detect the presence of viral proteins in the tissues.
## Development of a recombinant PCV5 Cap VLPs ELISA
The 30 serum samples obtained from a specific pathogen-free herd that tested PCV5 negative by qPCR were used as negative controls and had an average absorbance of 0.167. The cutoff value differentiating negative and positive serum samples was determined as 3 standard deviations above the mean of the negative controls (0.359). The protocol of ELISA was similar to assays previously described (34), by using 0.25 µg/mL purified PCV5 Cap VLPs to coat the wells.
## References
1. Zhao, Rosario, Breitbart et al. (2019) "Eukaryotic circular repencoding single-stranded DNA (CRESS DNA) viruses: ubiquitous viruses with small genomes and a diverse host range" *Adv Virus Res*
2. Desingu, Nagarajan (2022) "Genetic diversity and characterization of circular replication (Rep)-encoding single-stranded (CRESS) DNA viruses" *Microbiol Spectr*
3. Nath, Das, Roby et al. (2021) "Structural perspectives of beak and feather disease virus and porcine circovirus proteins" *Viral Immunol*
4. Smiley, Tompkins, Pawlak et al. (2023) "Watson-crick base-pairing requirements for ssDNA recognition and processing in replication-initiating HUH endonucleases"
5. Meng (2013) "Porcine circovirus type 2 (PCV2): pathogenesis and interaction with the immune system" *Annu Rev Anim Biosci*
6. Todd (2004) "Avian circovirus diseases: lessons for the study of PMWS" *Vet Microbiol*
7. Nauwynck, Sanchez, Meerts et al. (2012) "Cell tropism and entry of porcine circovirus 2" *Virus Res*
8. Palinski, Piñeyro, Shang et al. (2017) "A novel porcine circovirus distantly related to known circoviruses is associated with porcine dermatitis and nephropathy syndrome and reproductive failure" *J Virol*
9. Zhang, Hu, Li et al. (2020) "Novel circovirus species identified in farmed pigs designated as porcine circovirus 4, Hunan province" *Transbound Emerg Dis*
10. Liu, Zhang, Xu et al. (2021) "Emergence of porcine circovirus-like viruses associated with porcine diarrheal disease in China" *Transbound Emerg Dis*
11. Shan, Li, Simmonds et al. (2011) "The fecal virome of pigs on a high-density farm" *J Virol*
12. Opriessnig, Karuppannan, Castro et al. (2020) "Porcine circoviruses: current status, knowledge gaps and challenges" *Virus Res*
13. Pérot, Fourgeaud, Rouzaud et al. (2023) "Circovirus hepatitis infection in heart-lung transplant patient" *France. Emerg Infect Dis*
14. Li, Zhang, Ye et al. (2023) "Novel circovirus in blood from intravenous drug users"
15. Rodriguez, Boizeau, Soulier et al. (2022) "Unknown circovirus in immunosup pressed patient with hepatitis" *Emerg Infect Dis*
16. Tompkins, Houtti, Litzau et al. (2021) "Molecular underpinnings of ssDNA specificity by Rep HUH-endonucleases and implications for HUH-tag multiplexing and engineering" *Nucleic Acids Res*
17. Mo, Li, Yin et al. (2019) "Structural roles of PCV2 capsid protein N-terminus in PCV2 particle assembly and identification of PCV2 type-specific neutralizing epitope" *PLoS Pathog*
18. Hesketh, Saunders, Fisher et al. (2018) "The 3.3 Å structure of a plant geminivirus using cryo-EM" *Nat Commun*
19. Hipp, Grimm, Jeske et al. (2017) "Near-atomic resolution structure of a plant geminivirus determined by electron cryomicroscopy" *Structure*
20. Sarker, Terrón, Khandokar et al. (2016) "Structural insights into the assembly and regulation of distinct viral capsid complexes" *Nat Commun*
21. Liu, Guo, Wang et al. (2016) "9 Å resolution Cryo-EM 3D reconstruction of close-packed virus particles" *Structure*
22. Khayat, Brunn, Speir et al. (2011) "The 2.3-angstrom structure of porcine circovirus 2" *J Virol*
23. Trapani, Bhat, Yvon et al. (2023) "Structureguided mutagenesis of the capsid protein indicates that a nanovirus requires assembled viral particles for systemic infection" *PLoS Pathog*
24. Santos-Pérez, Charro, Gil-Carton et al. (2019) "Structural basis for assembly of vertical single β-barrel viruses" *Nat Commun*
25. Gil-Carton, Jaakkola, Charro et al. (2015) "Insight into the assembly of viruses with vertical single β-barrel major capsid proteins" *Structure*
26. Mart, Dolja, Koonin (2019) "Origin of viruses: primordial replicators recruiting capsids from hosts" *Nat Rev Microbiol*
27. Koonin, Dolja, Krupovic (2022) "The logic of virus evolution" *Cell Host & Microbe*
28. Krupovic (2013) "Networks of evolutionary interactions underlying the polyphyletic origin of ssDNA viruses" *Curr Opin Virol*
29. Tarasova, Dhindwal, Popp et al. (2021) "Mechanism of DNA interaction and translocation by the replicase of a circular repencoding single-stranded DNA virus"
30. Yan, Sun (2024) "Genotypic diversity and immunological implications of porcine circovirus: Inspiration from PCV1 to PCV4" *Microb Pathog*
31. Da Silva, Da Silva, Da Silva et al. (2023) "Porcine circovirus 3: a new challenge to explore" *Front Vet Sci*
32. Chen, Zhang, Xu (2023) "Pathogenicity and immune modulation of porcine circovirus 3" *Front Vet Sci*
33. Shen, Liu, Zhang et al. (2018) "Identification and characterization of atypical porcine pestivirus genomes in newborn piglets with congenital tremor in China" *J Vet Sci*
34. Wang, Li, Zeng et al. (2024) "Development of a fully automated chemiluminescent immunoas say for the quantitative and qualitative detection of antibodies against African swine fever virus p72" *Microbiol Spectr* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12702348&blobtype=pdf | # A novel monoclonal antibody targeting a conserved inner region of the hepatitis B virus envelope enables broad detection of immune escape variants
Yansong Chen, Xinyu Zhang, Asha Ashuo, Zhong Fang, Wuhui Song, Jiangxia Liu, Jieliang Chen, Yaming Li, Zhenghong Yuan
## Abstract
The detection of hepatitis B surface antigen (HBsAg) is fundamental for the diagnosis of chronic hepatitis B (CHB). However, current diagnostic assays that rely on antibodies targeting the conformational "a" determinant of HBsAg are frequently compromised by mutations in this region, leading to undetected immune escape variants. In this study, we generated a novel monoclonal antibody, designated S1705, by immunizing mice with CHO-derived HBsAg. This antibody was identified to recognize a conserved linear epitope located within the internal loop region of the HBsAg particle. Western blot analysis under denaturing conditions confirmed that S1705 robustly detects HBsAg from multiple genotypes (A-D) and common clinical mutants, including the prevalent G145R variant. Moreover, S1705 demonstrated effective utility in applications such as flow cytometry and immunofluorescence. Notably, it exhibited superior performance compared to commercial antibodies targeting conformational epitopes in detecting HBsAg escape variants. We conclude that S1705, by targeting a conserved linear epitope, enables broad and reliable detection of both wild-type and mutant HBsAg. Critically, our findings demonstrate that immunodominant antigenic regions exist beyond the conventional boundaries of the 'a' determinant. This antibody thus holds significant promise for enhancing the diagnostic coverage of HBV variants and supporting future antiviral research and clinical monitoring.
## 1. Introduction
Chronic hepatitis B virus (CHB) infection remains a major global health challenge, affecting over 250 million people worldwide and leading to a high risk of cirrhosis and hepatocellular carcinoma (Polaris Observatory Collaborators, 2023). The current diagnosis and monitoring of hepatitis B infection still rely heavily on the detection of hepatitis B surface antigen (HBsAg), which is a key serological marker reflecting active viral replication and carrier status (Tsukuda and Watashi, 2020).
HBsAg is the envelope protein of HBV, and can be detected in the bloodstream in spherical and filamentous forms (non-infectious subviral particles) or attached to viral particles (infectious virus) (Chang et al., 2021). The major antigenic region of HBsAg, known as the "a" determinant (amino acids 124-147), is the target of nearly all commercial diagnostic antibodies from vaccine-induced immune responses (Lazarevic et al., 2019). However, this region is highly prone to mutation under immune pressure, particularly in individuals with chronic infection or vaccine breakthrough. In addition, a limited number of mutation sites have been reported outside the MHR, such as positions 8, 34, 41, 44, 48, 96, 97, 175, and 176. However, the clinical significance of these mutations remains to be fully elucidated (Desmond et al., 2008;Salpini et al., 2015;Wu et al., 2012). Mutations within the "a" determinant, such as G145R, can significantly alter the antigenic conformation of HBsAg, leading to diagnostic failure and immune escape (Carman et al., 1990;Mokaya et al., 2018). These HBsAg immune escape variants pose significant challenges for accurate diagnosis, blood screening, and clinical management.
The majority of currently available HBsAg detection assays rely on antibodies targeting conformational epitopes. These epitopes are susceptible to disruption by mutations, thereby limiting their applicability in the detection of HBsAg variants and potentially leading to falsenegative results in clinical settings (Wang et al., 2020). Moreover, apart from a few studies (Matsumoto et al., 2017;Ozeki et al., 2018;Yang et al., 2016), very few antibodies are available that recognize linear epitopes located within conserved and mutation-resistant regions of HBsAg. Consequently, there remains an unmet need for broadly reactive antibodies capable of reliably detecting HBsAg across diverse genotypes and common immune-escape mutants.
In this study, we report the development and characterization of a novel monoclonal antibody, S1705, which specifically recognizes a conserved linear epitope located between amino acids 25-50 of HBsAg. This epitope resides within the inner region of the HBV envelope and is highly conserved across HBV genotypes and clinical isolates. We demonstrated that S1705 can detect multiple genotypes and various mutant HBsAg proteins including G145R in Western blot experiments. Additionally, flow cytometry and immunofluorescence experiments also proved that this antibody can be applied to various detection platforms based on hematology or histology for experimental detection. These findings indicate that S1705 represents a promising new tool for robust HBsAg detection, particularly in the context of immune escape variants and diagnostic failure.
## 2. Materials and methods
## 2.1. Generation of monoclonal antibody S1705
BALB/c mice (female, 6-8 weeks old) were purchased from Gem-Pharmatech Co., Ltd. (Jiangsu, China), and immunized intraperitoneally three times at two-week intervals with 100 μg of recombinant hepatitis B surface antigen (HBsAg, adr, genotype C) protein expressed in Chinese Hamster Ovary (CHO) cells (Luoyang Baitaike Biotechnology Co., Ltd, Cat. A93-02), emulsified in an equal volume of Freund's adjuvant (complete Freund's adjuvant for the first immunization; incomplete Freund's adjuvant for subsequent immunizations). Two weeks after the final boost, splenocytes were harvested and fused with SP2/0 myeloma cells using polyethylene glycol (PEG 4000, Sigma-Aldrich) according to standard hybridoma techniques. The fused cells were cultured in HAT (hypoxanthine-aminopterin-thymidine) selection medium in 96-well plates. Approximately 14 days later, culture supernatants were screened for HBsAg-binding antibodies under denaturing conditions via Western blot. Positive clones were subcloned by limited dilution to establish monoclonal hybridoma cell lines. A clone designated S1705 was selected for further characterization.
## 2.2. Expression and purification of S1705
Hybridoma S1705 cells were cultured in RPMI-1640 medium containing 10 % fetal bovine serum (FBS) at 37 • C in a 5 % CO₂ atmosphere. Supernatants were harvested when cells reached confluence, centrifuged at 8000 rpm for 15 min, and filtered through a 0.22 μm membrane. Immunoglobulins were purified using Protein A/G agarose affinity chromatography (Thermo Fisher Scientific). Bound antibodies were eluted with 0.1 M glycine buffer (pH 2.7) and immediately neutralized with 1 M Tris-HCl (pH 9.0). The purified antibody buffer was exchanged into phosphate-buffered saline (PBS) using a 30 kDa centrifugal filter device (Millipore) and stored at 4 • C until use.
## 2.3. Plasmids and cells
The 1.3 mer HBV-DNA constructs for genotypes A2, B5, D3 and the 3.8 mer pHBV C2 were preserved in our laboratory (Shen et al., 2018).
The pVAX1-HBsAg plasmid was available in our laboratory and served as the template for generating HBsAg mutant plasmids using a commercial site-directed mutagenesis kit (TOYOBO, Code No. SMK-101). Huh7 and HepAD38 cell lines were maintained in our lab.
## 2.4. Epitope mapping
A set of overlapping synthetic peptides covering the small HBs protein (S-HBs) sequence was synthesized (GL Biochem, Shanghai). Each peptide (15-20 amino acids, 5-amino acid overlap) was dissolved in DMSO at a concentration of 1 mg/mL. Peptides were coated onto 96well ELISA plates at 10 μg/mL and incubated overnight at 4 • C. After washing with PBST (PBS with 0.05 % Tween-20), wells were blocked with 2 % bovine serum albumin (BSA) for 2 h at room temperature. S1705 was added at a 1:1000 dilution and incubated for 1 h. After washing, horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG secondary antibody (1:5000, Jackson ImmunoResearch) was added. Color development was performed using TMB substrate (Thermo Fisher) and measured at 450 nm. The binding site was determined based on the peptide reactivity profile.
## 2.5. Ethical statement for human and animal studies
All experimental procedures involving animals and human tissues were approved by the Ethics Committee of Shanghai Medical College, Fudan University (Approval No. 2022-C001) and were performed in accordance with the relevant guidelines and regulations.
## 2.6. Mouse model of HBV
Male C57BL/6 mice (5-6 weeks old) were obtained from Gem-Pharmatech Co., Ltd. (Jiangsu, China). All mice were kept under specific pathogen-free conditions in the Enhanced Biosafety Level 2 (BSL-2+) Laboratory. The AAV-HBV mouse model (Yang et al., 2014) was constructed by injection 1 × 10 10 AAV8-HBV 1.3-wildtype/G145R (Fubio (Suzhou) Biotechnology Co., Ltd., China) via tail vein. One month later, blood was collected from the suborbital venous plexus of the mice, with a volume of 200 μl each time. The blood was incubated at 37 • C for 2 h and then centrifuged at 2000 g at room temperature for 10 min to separate the serum. The serum samples were temporarily stored at 4 • C. All procedures were in accordance with the approved guidelines of the National Institute of Biological Sciences Guide.
## 2.7. Western blot analysis
Western blot was performed using CHO-derived HBsAg (Luoyang Baitaike Biotechnology Co., Ltd, Cat. A93-02), HBsAg from different HBV genotypes (A, B, C, D), various HBsAg mutants, and mouse serum samples. The human hepatoma cell line Huh7 was cultured in Dulbecco's Modified Eagle Medium (DMEM) supplemented with 10 % fetal bovine serum (FBS). Cells on 12-well plates were transiently transfected using Lipofectamine 3000 reagent (Thermo Fisher Scientific) according to the manufacturer's instructions. Lysates of the cells and culture supernatant were collected three days post-transfection for subsequent analysis. 1/6 of the lysate was loaded per lane and were separated by SDS-PAGE and transferred to nitrocellulose membranes (Millipore). Membranes were blocked with 5 % skim milk in PBST for 1 h at room temperature, followed by incubated with S1705 antibody (1:1000 dilution) or the commercial goat polyclonal anti-HBs antibody (1:1000 dilution, bs-1557 G, Bioss) overnight at 4 • C. After washing, membranes were incubated with HRP-conjugated goat anti-mouse IgG (1:5000 dilution, AB97040, Abcam) or rabbit anti-goat IgG (1:5000 dilution, M212123S, Abmart) secondary antibodies for 1 h at room temperature. Immunoreactive bands were visualized using enhanced chemiluminescence (ECL) reagents (Bio-Rad) and imaged with a chemiluminescence imaging system (Tanon).
## 2.8. Enzyme-Linked immunosorbent assay (ELISA)
Secreted HBsAg or HBeAg in culture supernatants or mouse serum samples were detected using ELISA kits (KHB, Shanghai, China) according to the manufacturer's instructions. All samples were assayed in duplicate, and absorbance was measured at 450 nm.
## 2.9. T cell epitope prediction
Potential T cell epitopes within the HBsAg sequence were predicted using the Immune Epitope Database and Analysis Resource (IEDB, http://www.iedb.org). MHC class I binding prediction was performed for HLA-A*02 human allele. Default parameters provided by the IEDB tools were used. Predicted epitopes with high binding affinity scores (top 15) were selected.
## 2.10. Flow cytometry
HBV-replicating HepAD38 cells were harvested and resuspended in FACS buffer (PBS containing 2 % FBS). Cells were incubated with S1705 (1:500) for 2 h at 4 • C, washed three times, and then stained with FITCconjugated goat anti-mouse IgG secondary antibody (1:1000, Thermo Fisher Scientific) for 1 h at 4 • C in the dark. After final washing, samples were analyzed using an Attune NxT flow cytometer (Thermo Fisher Scientific), and data were processed with FlowJo.
## 2.11. Immunofluorescence staining
Formalin-fixed paraffin-embedded (FFPE) liver tissue sections from chronic HBV patients and HBV-associated hepatocellular carcinoma specimens (Yuanxi Biology) were deparaffinized, rehydrated, and subjected to antigen retrieval. Sections were blocked with 5 % BSA for 1 h at room temperature, followed by incubation with S1705 (1:500) for 2 h at room temperature. Sections were stained with Alexa Fluor 488-conjugated secondary antibody for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI. Fluorescence images were acquired using a Leica SP8 confocal microscope under standardized acquisition settings.
## 2.12. Statistics and data analysis
Data are presented as mean ± SD unless otherwise stated. Statistical analyses were performed using GraphPad Prism 9.0 (GraphPad Software, Boston, MA, USA).
## 3. Results
## 3.1. Generation and epitope identification of monoclonal antibody S1705
BALB/c mice were immunized with CHO-derived HBsAg. A mixture of 100 μg HBsAg and an equal volume of Freund's adjuvant was injected intraperitoneally at multiple sites every two weeks for a total of three immunizations. Two weeks after the last immunization, mouse splenocytes were isolated and fused with the myeloma cell line SP2/0. Various types of hybridoma cells were obtained. The hybridoma cells were subjected to limiting dilution cloning. Culture supernatants were collected and screened by Western blot analysis against CHO-derived HBsAg. One clone exhibiting reactivity with HBsAg under denaturing conditions was identified (Fig. 1A). In contrast, none of the other antibodies showed detectable binding to HBsAg in Western blot assays. A hybridoma clone stably secreting a monoclonal antibody against HBsAg was successfully isolated and designated as S1705. Protein A/G affinity purification yielded sufficient antibody for downstream analysis.
To determine the epitope targeted by S1705, a series of overlapping synthetic peptides spanning the small HBs protein were synthesized and analyzed by peptide-based ELISA. The antibody showed specific reactivity with the peptide corresponding to amino acids 25-50 (Fig. 1B). This region is located within the second internal loop domain of HBsAg (Fig. 1C), distinct from the commonly targeted "a" determinant (aa124-147). Further structural analysis confirmed that this region is linear, transmembrane-proximal, and highly conserved among HBV genotypes.
## 3.2. S1705 antibody recognizes HBsAg from different HBV genotypes
HBV isolates worldwide can be classified into 10 genotypes, with nucleotide sequence differences of 8 % or greater between genotypes (Abechi et al., 2025). Genotypes B and C are dominant in Asia, while genotype A has a wide global distribution, with a definite geographical predominance in Northern and Northwestern Europe, South Africa, and Brazil. Genotype D is prevalent in the Mediterranean, Northern Africa, and Russia (Chen et al., 2023). To assess the sequence conservation of the potential epitope region recognized by S1705, we performed a multiple sequence alignment of the major surface protein (amino acids 25-55) across globally relevant genotypes, including genotype A-H (Fig. 2A). The alignment revealed a remarkably high degree of conservation across all eight genotypes. The core region from positions 25 to 44 is entirely invariant. Variations are primarily concentrated in a short segment between residues 45 and 50. This analysis confirms that the S1705 target region is largely conserved, supporting its potential as a broad-spectrum antibody. Furthermore, we transfected Huh7 cells with plasmids carrying replication-competent HBV DNA from A-D genotypes, which allows production of viral antigens in vitro. Cell supernatants were analyzed by a commercial qualitative ELISA kit (KHB, Shanghai, China) (Fig. 2B) and Western blot (Fig. 2C). ELISA detection showed that all four genotypes can produce secreted HBsAg and HBeAg, S1705 can well detect the secreted HBsAg of all four genotypes, while the commercial antibody has a poor effect in detecting the secreted HBsAg of genotype D. Cell lysates were detected using S1705 and a commercial goat polyclonal anti-HBs antibody (bs-1557G, Bioss) as primary antibodies, with CHO-derived HBsAg (0.1 μg) as a positive control (Fig. 2D).
As shown, when detecinting HBsAg in cell lysates, S1705 was able to recognize all four genotypes, while the commercial antibody exhibited relatively weak detection of intracellular HBsAg of genotype B and D. Additionally, we observed that S1705 demonstrated significantly superior capability in detecting L-HBs (39/42 kDa) and M-HBs (33/36 kDa) compared to bs-1557G in Western blot analysis (Heermann et al., 1987(Heermann et al., , 1984)). This indicates that the epitope recognized by S1705 is well-conserved among major HBV genotypes, confirming the utility of S1705 in detecting HBsAg across different genotypes.
## 3.3. S1705 recognizes a broad range of immune escape mutant HBsAgs
Amino acids 99-169 form the major hydrophilic region (MHR) molecular structure of HBsAg. Within this, amino acids 124-147 constitute the "a" determinant of HBsAg, located on the outer surface of the MHR and involved in binding to anti-HBs (Lazarevic et al., 2019). Besides, the MHR encompasses critical residues governing HBV serological subtypes. Specifically, the d/y subtype is determined by the polymorphism at position 122 (d/y) and 160 (w/r). The combinations of these residues are responsible for defining the major serological subtypes of HBV (e.g., ayw, adr) (Chen et al., 2023). The MHR is a highly mutable region; for example, in genotype B, single mutations T123A, P142L, G145A, and combined mutations M103I-K122R were associated with decreased intracellular HBsAg levels; in genotype C, single mutation P120T was associated with reduced intracellular HBsAg levels (Song et al., 2025). Previously, we constructed 22 HBsAg mutant plasmids, including single and double mutation sites. The mutant and wild-type plasmids were separately transfected into Huh7 cells. Cell lysates and culture supernatants were harvested on the third day post-transfection for Western blot analysis. Huh7 cells were transfected with the wild-type and mutant plasmids, respectively. Both cell lysates and culture supernatants harvested on the third day post-transfection were subjected to subsequent analyses. As shown in Fig. 3A, a commercial ELISA kit detected only certain mutants, including several single mutants such as K141E, M103I, P142L, and T118K, while double mutants were largely undetectable. In contrast, Western blot analysis of culture supernatants using the S1705 antibody revealed that S1705 recognized the majority of immune escape mutant HBsAg proteins, including single mutants such as C137Y, C138Y, and C124R, as well as double mutants including C137Y-D144V and T123N-T143S (Fig. 3B).
Furthermore, Western blot analysis of cell lysates probed separately with S1705 and the commercial antibody bs-1557G demonstrated that S1705 detected nearly all mutants tested, including single mutants such as C137Y, C138Y, K141E, and C124R, and double mutants such as C137Y-D144V and T123N-T143S (Fig. 3C andD). Conversely, the commercial antibody failed to detect most mutants, including K141E, C124R, P142L, T116N, and the double mutant T123N-T143S. This finding indicates that the epitope recognized by S1705 is unaffected by commonly mutated residues, enabling efficient detection of HBV variants carrying immune-escape mutations in clinical settings. Consequently, S1705 shows potential to overcome current diagnostic limitations associated with immune evasion.
## 3.4. S1705 antibody efficiently detects G145R HBsAg in mouse sera
In this study, mouse models capable of stably producing wild-type HBsAg and G145R mutant HBsAg in serum were established by
## (caption on next page)
intravenous tail injection of AAV8-HBV1.3 or AAV8-HBV 1.3-G145R, respectively. Sera were collected one month post-injection, diluted 30fold, and subjected to ELISA. As shown in Fig. 4A, HBsAg was detected in the serum of both mouse models by ELISA.
For Western blot analysis, serum samples were diluted 4-fold and incubated with S1705 or the commercial anti-HBs antibody bs-1557G as primary antibodies. Naive C57BL/6 mice and CHO-derived HBsAg (0.1 μg) were used as negative and positive controls, respectively. The results (Fig. 4B) demonstrated that S1705 could effectively detect HBsAg in both wild-type and G145R mutant mouse models, consistent with the ELISA findings. In contrast, the commercial antibody bs-1557G showed barely detectable signals for serum HBsAg. These results indicate that S1705 enables sensitive detection of both wild-type and G145R mutant HBsAg in serum. Y. Chen et al. Virus Research 362 (2025) 199669
## 3.5. Application of S1705 in flow cytometry and immunofluorescence
Since S1705 recognizes a linear epitope within the membraneproximal region of HBsAg (amino acids 25-50), we hope to evaluate whether this region could be presented on surface of hepatocytes. Thus, we performed HLA-A2 restricted CD8+ T cell epitope prediction using the Immune Epitope Database (IEDB, http://www.iedb.org). Interestingly, among the top 15 predicted high-affinity T cell epitopes, three were located within the linear amino acid sequence that also constitutes the S1705 antibody-binding region (Fig. 5A). This spatial overlap raises the possibility that robust T helper cell responses, specific to peptides derived from this region, could provide effective help for B cells targeting the conformation-dependent B cell epitope within the same vicinity, potentially enhancing the overall anti-HBs immune response.
To assess the binding of S1705 to native HBsAg on the cell surface, we performed flow cytometry using HepAD38 cells. As shown in Fig. 5B, S1705 specifically labeled surface HBsAg, evidenced by a pronounced increase in fluorescence intensity compared to the isotype control. This specific binding was quantified and confirmed to be statistically significant (**p < 0.01) by comparing the mean fluorescence intensity (MFI) values. Furthermore, paraffin-embedded tissue sections from HLA-A2 positive chronic hepatitis B patients, liver cancer patients, and adjacent tissues of liver cancer patients were collected for immunofluorescence staining. Anti-total HLA-A antibody labeled with green fluorescence and S1705 labeled with red fluorescence were used to bind their target antigens, respectively, and DAPI was used to stain nuclei. Target antigen localization was observed under a confocal microscope (Fig. 5C). The results showed that S1705 could detect HBsAg signals in different tissues, corresponding to the distribution pattern of HBsAg in the respective tissues, demonstrating S1705's strong capability for intracellular HBsAg signal detection and proving its applicability in histological and blood flow-based assays.
Overall, these results indicate that S1705 is a versatile monoclonal antibody capable of detecting HBsAg from multiple genotypes and immune escape mutants under both denaturing and native conditions. Its unique linear epitope, located within the conserved inner region of HBV envelope (aa25-50), offers significant diagnostic advantages over traditional antibodies targeting variable conformational epitopes.
## 4. Discussion
Currently, in clinical practice, both HBV DNA and HBsAg serve as important virological biomarkers for monitoring the progression of chronic hepatitis B. Among them, HBsAg levels generally reflect the transcriptional activity of cccDNA and the capacity for viral protein production-closely associated with the phase of HBV infection (Kushch, 2025;Li et al., 2025;Zeng et al., 2022). HBsAg comprises L, M and S surface antigens, and the smallest surface antigen S protein (24 kDa) is 226 amino acids long (Li et al., 2020;Yuen et al., 2018) HBsAg is important for monitoring the natural infection process of HBV, assessing treatment response and predicting the stage of disease progression. However, the continuous emergence of immune escape variants, particularly those carrying mutations within the "a" determinant region, poses significant diagnostic challenges.
The G145R mutation that occurs in the "a" determinant is the earliest mutation site that was discovered and widely recognized as causing latent HBV infection (Carman et al., 1990). In addition to G145R, several other mutation sites leading to the loss of antigenicity of HBsAg have been identified in recent decades (Antoni et al., 1994;Jongerius et al., 1998;Kreutz, 2002;Lee et al., 1992;Tian et al., 2007). Although detection methods for HBsAg have been continuously updated and improved, certain mutations may still be overlooked in clinical or laboratory testing. For instance, Salpini et al. demonstrated in vivo that the T131N-M133T mutation can generate a novel glycosylation site on HBsAg, which significantly alters its antigenicity and affects its recognition by antibodies used in HBsAg quantitative assays (Salpini et al., 2020). Several studies have also indicated that different anti-HBs monoclonal antibodies (mAbs) exhibit varying binding patterns to mutant forms, and various substitutions at different positions within the "a" determinant may impair the antigenicity of HBsAg and reduce its reactivity with these antibodies (Golsaz-Shirazi et al., 2016). For example, research has shown that in seven patients undergoing maintenance hemodialysis and tested by enzyme immunoassay (EIA), mutations such as Q129R, T131N, M133S, F134L, and D144E were associated with occult hepatitis B virus infection (Tang et al., 2020). Other cases of occult infection have been linked to M133I and G145R, where HBsAg was undetectable by chemiluminescent microparticle immunoassay (CMIA) (Bubonja-Šonje et al., 2024). Furthermore, in chronic HBV carriers, substitutions including C124Y, T131N, M133T, and G145A, identified through sequencing of the HBV S region, were also undetectable by electrochemiluminescence immunoassay (ECLIA) (Liu et al., 2023). The impact of these substitutions confirms that the "a" determinant is a critical region for the recognition of the S protein by specific antibodies (Pondé and Amorim, 2024). In our study, compared to commercial antibodies, S1705 demonstrated enhanced sensitivity in detecting single mutations such as T115N, C124R, K141E, P142L, K122I, and D144E, as well as double mutations including C137Y-D144V and T123N-T143S. Relative to commercial ELISA kits, S1705 was capable of detecting single mutations including C137Y, C138Y, D144E, G145E, and G145R, in addition to double mutations such as C137Y-D144V and T123N-T143S.
Unlike most commercial antibodies that recognize conformational epitopes within the immunodominant "a" determinant, S1705 recognizes an epitope located in the inner region of the HBV envelope, which is highly conserved across HBV genotypes. Multiple functional assays, including Western blot, flow cytometry, and immunofluorescence, confirmed the ability of S1705 to recognize HBsAg in denatured states. The antibody exhibited strong reactivity against HBsAg derived from HBV genotypes A, B, C, and D. These genotypes represent the most prevalent global lineages and are associated with regional variations in transmission, disease progression, and treatment response (Chen et al., 2023). The broad genotypic coverage of S1705 suggests its potential as a universal reagent for HBV detection across diverse geographical settings.
Importantly, the linear epitope targeted by S1705 offers practical advantages for assay development. Unlike conformational epitopes, which may be disrupted during protein denaturation, linear epitopes are more stable and suitable for diagnostic platforms based on synthetic peptides, including sandwich ELISA, lateral flow tests, and even the development of multiplex antigen panels. Furthermore, the ability of S1705 to bind a conserved epitope and effectively detect HBsAg across diverse mutations suggests that regions beyond the conventional 'a' determinant may possess immunogenic potential and could represent novel targets for neutralization, a possibility that warrants further investigation.
While the 'a' determinant of HBsAg is well-established as the immunodominant epitope forming the basis of diagnostic assays and vaccines, our findings, together with previous studies, suggest that regions extending beyond its canonical limits also hold considerable biological significance. For instance, mutations in the N-terminal (approximately aa 1-98) and C-terminal (approximately aa 170-226) regions of the S domain have been implicated in altering T-cell epitopes and may serve as potential immunomodulatory targets, even if their full clinical impact remains to be fully elucidated (Desmond et al., 2008;Lazarevic et al., 2019). Moreover, the C-terminal domain plays an essential role in HBsAg secretion and virion assembly (Lazarevic et al., 2019). These adjacent regions likely contribute to antibody recognition, immune escape, and diagnostic sensitivity. Highlighting this peculiarity underscores the importance of considering not only the 'a' determinant itself but also structurally or functionally linked regions in the development of next-generation therapeutics and immunization strategies.
Although the results are promising, this study has several limitations.
It should be noted that although S1705 has been confirmed to recognize multiple mutant forms and genotypes, its reactivity against certain specific mutants and genotypes that are less prevalent in China (such as genotype E-H) has not yet been fully evaluated. In addition, the lower limit of detection for hepatitis B surface antigen in serum remains to be determined. Therefore, further investigation is warranted. Moreover, future studies should aim to evaluate the diagnostic performance of the antibody in clinical samples (like patients' sera), including those with known escape mutations. Comparative assessment of its analytical sensitivity and specificity against approved commercial assays will be essential to determine its readiness for clinical translation.
## 5. Conclusion
In this study, we developed and evaluated a novel monoclonal antibody, S1705, which targets a conserved linear epitope within the inner region of the HBV envelope. S1705 retained reactivity across major HBV genotypes and against a range of clinically relevant immuneescape variants. The antibody demonstrated robust performance across multiple detection platforms, such as Western blot, flow cytometry, and immunofluorescence. These findings provide a new strategy for improving HBV diagnosis, particularly in the context of antigenic variants and diagnostic failures, and may support the development of nextgeneration serological assays as well as tools for antiviral research.
Fig. 5. The S1705 antibody can be used in flow cytometry and immunofluorescence staining. (A). T cell epitope prediction map. Utilizing the Immune Epitope Database for T cell epitope prediction, it was found that among the top 15 high-affinity T cell epitopes predicted, three overlap with the S1705 antibody binding region. (B). Flow cytometric analysis of S1705 binding to cell surface HBsAg. HepAD38 cells were stained with either an FITC-conjugated mouse IgG isotype control or FITC-conjugated S1705. The zebra plot shows FITC fluorescence (Comp-FITC-A) after gating on single, live cells based on their side scatter (SSC-A) properties. In the bottom left, the X-axis (BL1-H/FITC-H) represents the fluorescence intensity of FITC-labeled antibodies, and theY-axis (normalized to mode) represents the relative cell frequency, with the histogram peak normalized to 1 to facilitate comparison across samples. Data are presented in mean fluorescence intensity (MFI) values with standard error of mean. **p<0.01. (C). By measuring the average fluorescence intensity, S1705 can effectively mark the hepatitis B surface antigen in liver cancer tissues and adjacent tissues. Nucleus indicates the cell nucleus stained only with DAPI (in blue). The cells were stained with the target antigens HLA-A and HBsAg respectively, showing green or red colors. Multicolor fluorescence overlap analysis is used to examine the co-localization of antigens.
## References
1. Abechi, George, Adejumobi et al. (2025) "Evolutionary insights of hepatitis B virus genotypes and profiles of mutations in surface and basal core promoter/pre-core genes among HBsAg-positive patients in north-central and southwestern Nigeria" *Viruses*
2. Antoni, Rodríguez-Crespo, Gómez-Gutiérrez et al. (1994) "Site-directed mutagenesis of cysteine residues of hepatitis B surface antigen. Analysis of two single mutants and the double mutant" *Eur. J. Biochem*
3. Bubonja-Šonje, Peruč, Abram et al. (2024) "Prevalence of occult hepatitis B virus infection and characterisation of hepatitis B surface antigen mutants among adults in western Croatia" *Ann. Hepatol*
4. Carman, Zanetti, Karayiannis et al. (1990) "Vaccine-induced escape mutant of hepatitis B virus" *Lancet*
5. Chang, Chou, Shih (2021) "A nuanced role of the small loop of hepatitis B virus small envelope protein in virion morphogenesis and secretion" *J. Biomed. Sci*
6. Chen, Li, Yin et al. (2023) "A review of epidemiology and clinical relevance of Hepatitis B virus genotypes and subgenotypes" *Clin. Res. Hepatol. Gastroenterol*
7. Desmond, Bartholomeusz, Gaudieri et al. (2008) "A systematic review of T-cell epitopes in hepatitis B virus: identification, genotypic variation and relevance to antiviral therapeutics" *Antivir. Ther*
8. Golsaz-Shirazi, Mohammadi, Amiri et al. (2016) "Localization of immunodominant epitopes within the "a" determinant of hepatitis B surface antigen using monoclonal antibodies" *Arch. Virol*
9. Heermann, Goldmann, Schwartz et al. (1984) "Large surface proteins of hepatitis B virus containing the pre-s sequence" *J. Virol*
10. Heermann, Kruse, Seifer et al. (1987) "Immunogenicity of the gene S and pre-S domains in hepatitis B virions and HBsAg filaments" *Intervirology*
11. Jongerius, Wester, Cuypers et al. (1998) "New hepatitis B virus mutant form in a blood donor that is undetectable in several hepatitis B surface antigen screening assays" *Transfusion*
12. Kreutz (2002) "Molecular, immunological and clinical properties of mutated hepatitis B viruses" *J. Cell Mol. Med*
13. Kushch (2025) "Occult hepatitis B: prevalence and clinical significance" *Vopr. Virusol*
14. Lazarevic, Banko, Miljanovic et al. (2019) "Immune-escape Hepatitis B virus mutations associated with viral reactivation upon immunosuppression" *Viruses*
15. Lee, Paglieroni, Holland et al. (1992) "Chronic hepatitis B virus infection in an anti-HBc-nonreactive blood donor: variant virus or defective immune response?" *Hepatology*
16. Li, Yan, Shi et al. (2020) "Hepatitis B virus infection: overview" *Adv. Exp. Med. Biol*
17. Li, Li, Sun et al. (2025) "Association of serum hepatitis B surface antigen with hepatitis B virus DNA and hepatic function in patients with chronic hepatitis B" *World J. Gastrointest. Surg*
18. Liu, Chen, Liu et al. (2023) "Host immunity and HBV S gene mutation in HBsAg-negative HBV-infected patients" *Front. Immunol*
19. Matsumoto, Imaizumi, Tanaka et al. (2017) "Novel and highly sensitive immunoassay for total hepatitis B surface antigen, including that complexed with hepatitis B surface antibody" *J. Gastroenterol*
20. Mokaya, Mcnaughton, Hadley et al. (2018) "A systematic review of hepatitis B virus (HBV) drug and vaccine escape mutations in Africa: a call for urgent action" *PLoS Negl. Trop. Dis*
21. Ozeki, Nakajima, Suii et al. (2018) "Analysis of hepatitis B surface antigen (HBsAg) using high-sensitivity HBsAg assays in hepatitis B virus carriers in whom HBsAg seroclearance was confirmed by conventional assays" *Hepatol. Res*
22. (2023) "Global prevalence, cascade of care, and prophylaxis coverage of hepatitis B in 2022: a modelling study" *Lancet Gastroenterol. Hepatol*
23. Pondé, De, Amorim et al. (2024) "Exchanges in the "a" determinant of the hepatitis B virus surface antigen revisited" *Virology*
24. Salpini, Colagrossi, Bellocchi et al. (2015) "Hepatitis B surface antigen genetic elements critical for immune escape correlate with hepatitis B virus reactivation upon immunosuppression" *Hepatology*
25. Salpini, Piermatteo, Battisti et al. (2020) "A hyper-glycosylation of HBV surface antigen correlates with HBsAg-negativity at immunosuppression-driven HBV reactivation in Vivo and hinders HBsAg recognition in Vitro" *Viruses*
26. Shen, Li, Wang et al. (2018) "Hepatitis B virus sensitivity to interferon-α in hepatocytes is more associated with cellular interferon response than with viral genotype" *Hepatology*
27. Song, Su, Yan et al. (2025) "Identification and characteristics of mutations promoting occult HBV infection by ultrasensitive HBsAg assay" *J. Clin. Microbiol*
28. Tang, Liu, Lu et al. (2020) "Occult Hepatitis B virus infection in maintenance hemodialysis patients: prevalence and mutations in "a" determinant" *Int. J. Med. Sci*
29. Tian, Xu, Zhang et al. (2007) "The amino acid residues at positions 120 to 123 are crucial for the antigenicity of hepatitis B surface antigen" *J. Clin. Microbiol*
30. Tsukuda, Watashi (2020) "Hepatitis B virus biology and life cycle" *Antiviral Res*
31. Wang, Wang, Huang et al. (2020) "Novel hepatitis B virus surface antigen mutations associated with occult genotype B hepatitis B virus infection affect HBsAg detection" *J. Viral. Hepat*
32. Wu, Shi, Wang et al. (2012) "A case of hepatitis B reactivation due to the hepatitis B virus escape mutant in a patient undergoing chemotherapy" *Virol. Sin*
33. Yang, Liu, Zhu et al. (2014) "A mouse model for HBV immunotolerance and immunotherapy" *Cell Mol. Immunol*
34. Yang, Song, Guan et al. (2016) "The Lumipulse G HBsAgquant assay for screening and quantification of the hepatitis B surface antigen" *J. Virol. Methods*
35. Yuen, Chen, Dusheiko et al. (2018) "Hepatitis B virus infection" *Nat. Rev. Dis. Primers*
36. Zeng, Liu, Cao et al. (2022) "Study on pathological and clinical characteristics of chronic HBV infected patients with HBsAg positive, HBV DNA negative, HBeAg negative" *Front. Immunol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12450560&blobtype=pdf | # Letter to the editor Commentary on "Complex left main bifurcation PCI in a patient with dual positivity: A unique interventional approach -A case report" by Setty et al., 2025
Mohammad Momeni, Ehsan Khozeimeh, Nezarali Moulaei, Hamidreza Zivarifar
## Abstract
We read with great interest the article by Setty et al., describing a complex left main bifurcation percutaneous coronary intervention (PCI) in a patient with dual positivity [1]. The case is indeed intriguing, highlighting innovative procedural strategies and the potential for individualized interventional approaches in high-risk coronary anatomy. The report underscores the importance of meticulous planning, intraprocedural imaging, and tailored stenting techniques in optimizing outcomes for challenging bifurcation lesions. However, despite the novelty, several considerations warrant further discussion:
## 1. Patient Selection and Risk Stratification
While the authors briefly mention dual positivity, the rationale for proceeding with PCI rather than surgical revascularization is not fully elaborated [2]. Further clarification regarding risk-benefit assessment and criteria for choosing PCI in such high-risk patients would strengthen the report.
## 2. Technical Details and Procedural Planning
The report provides limited information on lesion preparation, intravascular imaging guidance (IVUS or OCT), and bifurcation stenting strategy selection (provisional versus two-stent technique) [3]. Detailed procedural insights are crucial for reproducibility and for guiding interventionalists facing similar challenges.
## 3. Adjunctive Pharmacotherapy and Periprocedural Management
The case would benefit from elaboration on antiplatelet regimen, anticoagulation strategy, and post-procedural monitoring in the context of dual positivity, particularly if immunological or prothrombotic factors contributed to the patient's presentation [4].
## 4. Long-Term Outcomes and Follow-Up
While immediate procedural success is reported, long-term followup data including restenosis rates, major adverse cardiovascular events, and functional assessment would provide valuable evidence for the durability and safety of the described approach.
In conclusion, Setty et al. present a fascinating case that contributes valuable insights into complex left main bifurcation PCI. Addressing the above gaps through more detailed procedural reporting, risk stratification, and follow-up would enhance the educational value of the case and support the translation of such innovative strategies into routine practice.
## References
1. Setty, Patil, Kumar et al. (2025) "Complex left main bifurcation PCI in a patient with dual positivity: a unique interventional approach-a case report" *Int. J. Surg. Case Rep*
2. Fezzi, Ding, Mahfoud et al. (2024) "Illusion of revascularization: does anyone achieve optimal revascularization during percutaneous coronary intervention?" *Nat. Rev. Cardiol*
3. Fujisaki, Kuno, Numasawa et al. (2022) "Provisional or 2-stent technique for bifurcation lesions in the second-generation drug-eluting stent era" *J. Soc. Cardiovasc. Angiogr. Interv*
4. Schirmer, Bulsara, Al-Mufti et al. (2023) "Antiplatelets and antithrombotics in neurointerventional procedures: guideline update" *J. NeuroIntervent. Surg*
5. *Contents lists available at ScienceDirect International Journal of Surgery Case Reports* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12517413&blobtype=pdf | # Water-insoluble exopolysaccharide synthesized by glucosyltransferases mediates the antibacterial activity of ClyR against Streptococcus mutans
Qizhao Ma, Xiaowan Wang, Mai Xu, Ziyi Yang, Dian Zhang, Jiamin Chen, Tao Gong, Hang Yang, Yuqing Li
## Abstract
Background: Dental caries is a widespread global health issue strongly associated with Streptococcus mutans. Bacteriophage-derived lytic enzymes such as ClyR hold considerable promise as antibacterial potential, but the molecular mechanisms underlying their activity against S. mutans remain unclear.Objective: This study aimed to determine the role of water-insoluble exopolysaccharides (EPS) in mediating the antibacterial activity of ClyR against S. mutans. Design: We compared the antibacterial effects of ClyR on S. mutans UA159 and its ΔgtfB mutant, which is characterized by reduced synthesis of water-insoluble EPS. Biofilm architecture and susceptibility were assessed using scanning electron microscopy, confocal laser scanning microscopy, and biomass quantification. Adsorption assays were conducted to evaluate the interaction between ClyR and water-insoluble EPS. Results: The ΔgtfB mutant exhibited significantly higher resistance to ClyR than S. mutans UA159, with reduced biofilm disruption and bacterial loss after treatment. In vitro assays confirmed that water-insoluble EPS specifically adsorbed ClyR, with binding localized to its catalytic PlyCAC domain. Conclusions: Water-insoluble EPS synthesized by S. mutans glucosyltransferases plays a critical role in modulating bacterial susceptibility to ClyR. These findings reveal a novel mechanism underlying bacteriophage lysin activity and highlight EPS as a potential target for enhancing ClyR efficacy against cariogenic biofilms.
## Introduction
Dental caries, characterized by dental enamel demineralization, remains a major global oral health challenge. Streptococcus mutans is central to the etiology of dental caries due to its virulence factors, including acid tolerance, adhesive ability, and the synthesis of extracellular polysaccharides [1][2][3]. Advanced molecular techniques, such as fluorescence in situ hybridization (FISH) applied to plaque samples from carious sites, have unequivocally confirmed the central role of S. mutans in cariogenic biofilm formation [4], underlining its pivotal role in the caries process.
The therapeutic potential of bacteriophage lytic enzymes, or endolysins, has attracted attention due to their specificity and efficacy in lysing bacterial cell walls [5]. These enzymes, characterized by a catalytic domain (CD) and a cell binding domain (CBD), are integral to bacteriophage-mediated bacterial lysis at the end of the phage replication cycle [6,7]. Early research by Freimer [8] on the application of lytic enzymes to treat streptococcal infections set the stage for their broader therapeutic use. Protein engineering has significantly expanded this potential, allowing the synthesis of chimeric lysozymes that combine lytic domains from diverse sources [9]. These engineered constructs possess enhanced lytic breadth, activity, and stability [10,11], marking a promising frontier in the battle against infectious diseases.
ClyR is a bioengineered lysin composed of the CHAP (Cysteine, Histidine-Dependent Amidohy-Drolase/Peptidase) catalytic domain from PlyC (PlyCAC), a lysin produced by a Streptococcus pyogenes phage with a broad host range against Streptococcus species, and the cell wall binding domain from PlySs2 (PlySb), which is derived from a Streptococcus suis phage and also exhibits broad-spectrum lytic activity. By integrating these domains, ClyR was designed to enhance lytic potency and spectrum against pathogenic streptococci [12][13][14][15]. This discovery heralded ClyR as an innovative bacteriophage chimeric lysozyme with activity against S. mutans, the primary bacterial species implicated in dental caries. However, variability in the susceptibility of S. mutans clinical isolates to ClyR poses a challenge to its clinical application, underscoring the need to investigate the factors that influence bacterial resistance or sensitivity to this promising antimicrobial agent [16]. The present study is a research endeavor to unravel the mechanisms underlying the varied susceptibility of S. mutans to ClyR, deepen the understanding of ClyR's antibacterial action, and provide insights that may inform the development of future strategies for the prevention and management of dental caries, thereby contributing valuable knowledge to the field of microbial pathogenesis and therapeutic interventions.
## Materials and methods
## Bacterial strains and growth conditions
The S. mutans UA159 strain was obtained from the American Type Culture Collection (ATCC) (Manassas, VA, USA), and its derivative was provided by the State Key Laboratory of Oral Diseases at Sichuan University. S. mutans UA159 and its derivative were routinely cultured in brain heart infusion (BHI) broth (BD, USA). Monoclonal clones (single-colony isolates) were obtained by streaking S. mutans strains onto brain heart infusion agar (BHIA) plates and selecting individual well-isolated colonies. All strains were incubated under anaerobic conditions (80% N₂, 10% CO₂, 10% H₂) to simulate the oxygenlimited environment characteristic of mature dental biofilms. For biofilm assays, BHI supplemented with 1% sucrose (Sigma; w/v) was used as BHIS for culturing biofilms. Bacterial growth was monitored by measuring the optical density at 600 nm (OD 600 ). For all bactericidal assays, including both planktonic and biofilm experiments, ClyR was used at a final concentration of 300 µg/mL. Planktonic cultures were incubated with ClyR for 1 or 7 h to capture the kinetics of bacterial killing, while biofilm samples were treated with ClyR for 24 h to ensure effective exposure within the biofilm matrix and allow for sufficient antimicrobial activity against sessile cells.
## Expression and purification of ClyR, PlyCAC and ClyF proteins
ClyR, PlyCAC, and ClyF protein expression strains were constructed previously [12,17]. Escherichia coli cells were cultured to an OD 600 of 0.6-0.8, induced by 0.2-mM isopropyl-β-D-thiogalactoside, and grown for 12 h at 16 °C. The bacterial cells were lysed by a high-pressure cell cracker in PBS at pH = 7.4. The supernatant was filtered through a 0.45-μm filter and passed through a GSTSep glutathione agarose resin column pre-equilibrated with PBS. After washing with cleavage buffer [50-mM Tris-HCl (pH = 7.0), 0.15-M NaCl, 1-mM EDTA, 1-mM DTT], 2 U PreScission protease was added per 100-μg GST-tag protein and incubated at 4 °C for 12 h and eluted with cleavage buffer. Collected proteins were dialyzed against PBS and finally stored at 4 °C after filtration. Protein concentration was determined by bicinchoninic acid assay (Pierce Rockford, IL, USA), using BSA as the standard.
## Construction of S. mutans ΔgtfB mutant
The in-frame deletion mutant ΔgtfB of S. mutans UA159 was generated using a two-step transformation protocol as described previously [18]. The homologous sequences of approximately 1 kb upstream and downstream of the gtfB open reading frame were amplified using specific primers. Subsequently, the IFDC2 cassette, a laboratory-generated non-polar erythromycin resistance and p-Cl-Phe sensitivity cassette, was PCR-amplified using IFDC2F/IFDC2R primers [18]. The resulting three PCR amplicons were assembled by overlap extension PCR, yielding a 4.5-kb fragment that was subsequently transformed into S. mutans UA159 cells. Transformants were selected on BHI plates supplemented with erythromycin at a concentration of 12 μg/mL to facilitate the identification of positive clones. In the second transformation step, PCR amplification was used to generate upstream and downstream fragments corresponding to the gtfB open reading frame, which were subsequently overlapped to create an upstream-downstream amplicon. This DNA construct was then introduced into the strain obtained from the first step via transformation using BHI selection plates containing p-Cl-Phe (Sigma) at a 4-mg/mL concentration. The constructed deletion mutant was further confirmed through PCR analysis and sequencing. All primers used in this study are listed in Supplementary Table S1.
## Bactericidal assays
A series of experiments were conducted to investigate the lytic conditions of ClyR against S. mutans. Overnight cultures of S. mutans UA159 were diluted 1:10 into fresh BHI medium and further diluted 10-, 50-, and 100-fold when OD 600 reached 0.5. In a 96-well plate, 100 µL of different concentrations (range: 50-500 µg/mL) of ClyR solution in PBS and an equal volume of multiple dilutions of bacterial solution were added, followed by incubation at 37 °C for different durations (10 min, 20 min, 30 min, etc.). At the end of the incubation period, 10 µL of the mixture was dropped onto BHIA plates after a 10-fold gradient dilution. Based on colony growth, the plates were then incubated at 37 °C for 48 h for colony-forming unit (CFU) counting.
## SEM and CLSM analyses of biofilms
S. mutans biofilms were observed under a scanning electron microscope (SEM) (FEI, Hillsboro, OR, USA). Bacteria diluted after overnight growth were further diluted 1:100 into BHIS at OD = 0.5 and then added to a 24-well plate with a sterile glass sheet placed at the bottom for 24 h of anaerobic incubation. The experimental group was treated with ClyR, while the control group was treated with an equivalent dose of PBS. The treated biofilms were rinsed three times with sterile PBS and fixed overnight with 2.5% (v/v) glutaraldehyde, followed by storage at 4ºC for 16 h. Subsequently, they were washed again with PBS and dehydrated using a gradient concentration of alcohol. The samples were observed by SEM (×20,000 magnification).
The biofilm architecture was observed using a confocal laser scanning microscope (CLSM) (Nikon, N-SIM). To label the extracellular polymeric substances, 1-µM Alexa Fluor 647 dextran conjugate (Life Technologies, Grand Island, NY, USA) was added to BHIS before inoculation. The bacterial solution in the mid-logarithmic growth phase was diluted 1:100 into BHIS and then added to the bottom of a glass dish for incubation for 24 h. After incubation, the biofilms were rinsed with 0.9% NaCl three times and stained with 2.5-µM SYTO 9 (Life Technologies) for 15 min to label bacterial cells. Biofilm images were captured and collected using a ×40 objective lens at a wavelength range of 655-690 nm for Alexa Fluor 647 and 495-515 nm for SYTO 9 fluorescence signals. Three randomly selected areas within each biofilm were scanned. ImageJ software (National Institutes of Health, USA) was used to calculate biofilm thickness and biomass.
## Zymogram assays
The cells of S. mutans UA159 and its derivatives were pelleted in the mid-logarithmic phase by centrifugation (4000 g, 4 °C, 10 min). The resulting supernatant was mixed with approximately onethird of its volume of absolute ethanol and immediately frozen at -80 °C for 30 min. After another round of centrifugation (25000 rpm, 4 °C, 15 min), the pellet was resuspended in 100 μL of PBS to obtain Gtfs. To visualize the expression level of Gtfs, proteins were separated using a 6% SDS-PAGE gel after quantification with a BCA protein quantification kit (Beyotime, China) to ensure equal total protein amounts across samples. The Gtfs proteins were renatured using the 0.2-M sodium phosphate buffer containing 2.5% Triton X-100 to assess their activity. Subsequently, the gel was transferred into 0.2-M sodium phosphate buffer containing 0.2% (w/v) T-70 glucan and 5% (w/v), followed by overnight incubation at 37 °C to observe water-insoluble EPS content on the gel.
## Synthesis of EPS in vitro
We synthesized water-insoluble EPS in vitro to investigate the synergistic effect of water-insoluble EPS and ClyR. The reaction solution was prepared by adding 0.2% (w/v) T-70 glucan and 5% (w/v) sucrose to the 0.2-M sodium phosphate buffer. Gtfs (the method of obtaining was the same as above) and the reaction solution were added to a 24-well plate at a ratio of 1:200 for water-insoluble EPS synthesis at room temperature. After incubating for 24 h, the solution was centrifuged to collect insoluble EPS, which was washed with PBS three times.
## Adsorption assay of water-insoluble EPS with ClyR, PlyCAC, and ClyF
Adsorptive assays were conducted to investigate the correlation between water-insoluble EPS and ClyR. The collected water-insoluble EPS was resuspended in PBS, thoroughly vortexed, and then sonicated in a water bath for 5 min to ensure homogeneous dispersion before gradient dilution and use in subsequent assays. For control experiments, water-insoluble EPS suspensions were subjected to heat treatment (95 °C for 10 min) followed by three PBS washes to remove denatured proteins. Heated water-insoluble EPS samples were then used in adsorption assay in the same manner as untreated water-insoluble EPS. For adsorption assay, 100 µL of diluted water-insoluble EPS solution and 100 µL of ClyR were mixed thoroughly in 1.5-mL centrifuge tubes and incubated at 37 °C for approximately 7 h. After centrifugation (4000 g, 4 °C, 10 min), the supernatant and precipitate were separated by SDS-PAGE (12%) to isolate protein components. The control group consisted of EPS alone to assess background adsorptive levels. BSA was used as a control under identical conditions to determine adsorptive specificity. Similarly, waterinsoluble EPS, ClyR, and BSA (the working concentration was 300 μg/mL) were individually added to separate 1.5-mL centrifuge tubes with all other parameters kept constant. The experimental protocols used to investigate the interaction between water-insoluble EPS and PlyCAC or ClyF (the working concentration was 200 μg/mL) were consistently maintained. The adsorptive ratio (%) of proteins by water-insoluble EPS was calculated by quantifying the SDS-PAGE band intensities of the protein (ClyR, PlyCAC, or ClyF) in the supernatant before and after incubation with EPS, using ImageJ software (National Institutes of Health, USA). The ratio was calculated as: [(Intensity of protein in supernatant without EPS -Intensity of protein in supernatant with EPS)/Intensity of protein in supernatant without EPS] × 100%.
## Statistical analysis
Statistical analyses were performed using GraphPad Prism 8.0.2 (GraphPad Software Inc, San Diego, CA, USA). All the experiments were conducted in triplicate and repeated at least three times. Statistical significance between the two groups was determined using Student's t-test with a two-tailed p < 0.05.
## Results
## Impact of Gtf on ClyR antibacterial activity against S. mutans
Glucosyltransferases (Gtfs) play a crucial role in forming biofilms and synthesizing water-insoluble EPS in S. mutans. To study the effect of different EPS amounts on the lytic activity of ClyR, we utilized the mutant strain lacking GtfB (∆gtfB) [19]. This enzyme primarily synthesizes water-insoluble EPS in S. mutans. The Coomassie staining and zymogram assay were used to validate the diminished Gtfs content and consequently reduced water-insoluble EPS production in the ∆gtfB mutant (Figure 1A andB).
We evaluated the susceptibility of S. mutans UA159 and its derivative ∆gtfB to ClyR at a concentration of 300 μg/mL. After ClyR exposure, the ∆gtfB mutant exhibited a significantly higher survival rate than the UA159 strain (Figure 1C). Specifically, the ∆gtfB mutant showed a survival rate of 41.64% after 1 h of ClyR treatment, compared to a 12.99% survival rate for UA159. After 7 h of ClyR treatment, survival rates for ∆gtfB and UA159 were 28.55 and 7.23%, respectively (Figure 1D), indicating the enhanced ClyR tolerance in the ∆gtfB mutant. These findings reveal an intrinsic link between Gtf-mediated EPS production and the defensive capacity of S. mutans against ClyR, indicating that Gtfs might be pivotal determinants of bacteriophage lysin susceptibility.
## Impact of Gtf-mediated EPS production on ClyR antibacterial activity against S. mutans biofilm
To clarify the role of GtfB-mediated EPS in susceptibility to ClyR, we compared the structural and quantitative changes in S. mutans UA159 and ΔgtfB biofilms following ClyR treatment using SEM, CLSM, and biomass analysis. As shown in Figure 2A, the ΔgtfB mutant formed biofilms with a thinner and less cohesive extracellular matrix than the wild-type strain. Upon ClyR exposure, wild-type biofilms exhibited substantial disruption and loss of bacterial architecture, while the ΔgtfB mutant retained more intact biofilm structure. Quantitative analysis revealed that S. mutans UA159 biofilms exhibited significantly greater loss of biofilm thickness and water-insoluble EPS after ClyR treatment compared to the ΔgtfB mutant (Figure 2B, left and middle panels). Consistently, bacterial reduction after ClyR treatment (Figure 2B, right panel) was also more pronounced in the wild-type strain than in the ΔgtfB mutant. These results indicate the essential role of GtfB-dependent EPS production in modulating ClyR susceptibility in S. mutans.
## Specific interaction between S. mutans water-insoluble EPS and ClyR
We synthesized water-insoluble EPS using S. mutans Gtfs extracts in vitro to further investigate the potential interactions between water-insoluble EPS and ClyR. A total of 192 μg of dried EPS was resuspended in PBS by vortexing followed by mild sonication to achieve a uniform suspension. Subsequently, dilutions were prepared using a twofold gradient, resulting in final reaction concentrations of 96, 48, 24, 12, and 6 μg/mL, respectively. As shown in Figure 3A-C, increasing concentrations of waterinsoluble EPS correlated with a higher percentage of ClyR adsorption from the reaction solution, with adsorption rates of 68.04, 72.05, 78.45, 85.50, and 90.96% for each concentration, respectively, further confirmed by analyzing the precipitates formed after centrifugation, which showed ClyR accumulation. BSA was used as a control protein to investigate the specificity of water-insoluble EPS binding towards ClyR. The results indicated that BSA was not adsorbed by EPS. Similarly, when both BSA and ClyR were mixed with water-insoluble EPS simultaneously, only the amount of ClyR decreased while BSA remained unaffected in the supernatant (Figure 3D). To further exclude the possibility of residual protein contaminants, we performed adsorption assays using heat-treated water-insoluble EPS. As shown in Figure S1, increasing concentrations of heated water-insoluble EPS also resulted in a dose-dependent adsorption of ClyR, consistent with untreated water-insoluble EPS. These experimental results suggest that the waterinsoluble EPS produced by S. mutans Gtfs extracts could specifically interact with ClyR.
## Specific binding of S. mutans water-insoluble EPS to the catalytic domain of ClyR
Previous studies have established that PlyCAC functions as the CD of ClyR, while ClyF shares the same CBD but possesses a different CD compared to ClyR (Figure 4A) [12,17]. We conducted the following adsorption experiments to further explore the specific domain of ClyR that water-insoluble EPS interacts with. The results demonstrated a significant reduction of PlyCAC in the supernatant and its accumulation in the precipitates as water-insoluble EPS concentration increased (Figure 4B). Conversely, ClyF was not effectively adsorbed by water-insoluble EPS (Figure 4C). Quantitative analysis showed the high adsorption efficiency of PlyCAC, with average adsorption ratios of 98.80, 95.77, and 78.50%, respectively (Figure 4D), indicating a strong adsorptive capability of water-insoluble EPS to PlyCAC. Additionally, BSA was used as a control, and the results demonstrated that upon simultaneous addition of PlyCAC and BSA to waterinsoluble EPS, only PlyCAC was adsorbed by water-insoluble EPS, indicating a specific interaction between water-insoluble EPS and PlyCAC (Figure 4E). These findings suggest that the water-insoluble EPS produced by S. mutans Gtfs extracts could specifically interact with the catalytic domain of ClyR.
## Discussion
This study investigated the antibacterial activities of ClyR against S. mutans and its ΔgtfB derivative and explored the underlying factors contributing to the variability in the susceptibility of S. mutans. Interestingly, water-insoluble EPS synthesized by Gtfs mediated the antibacterial activity of ClyR against S. mutans. These findings provide novel insights into the interactions between ClyR and water-insoluble EPS, advancing our understanding of potential therapeutic strategies against infectious diseases.
The lyase secreted by bacteriophages enzymatically degrades the bacterial cell wall, facilitating the release of fully assembled phage particles. Delisle ingeniously incorporated the lyase derived from M102, e10 and f1, three virulent S. mutans bacteriophages that were isolated and identified in 1993 and belong to the Longtail family, into toothpaste and mouthrinse formulations as a preventive measure against dental caries and other oral diseases [20,21]. ClyR, identified by Yang through bioengineering techniques in vitro, is the first phage lyase reported to exhibit antagonistic activity against all known S. mutans serotypes [12]. However, the lytic activity of ClyR against various clinical strains of S. mutans is significantly inconsistent [16], posing a challenge to the clinical application of ClyR and necessitating a deeper understanding of the factors influencing bacterial resistance or sensitivity to these antimicrobial agents.
Bacterial biofilm is a unique mode of existence of microorganisms in nature. Compared with planktonic cells, bacterial biofilms are more adaptable to complex living environments [22]. In the classical four-factor theory of dental caries, microorganisms are the primary factor, and dental plaque biofilm is the carrier and form of microbial cariogenicity. S. mutans is the main cariogenic bacteria in biofilm formation [2,23]. Glucosyltransferase use sucrose to synthesize glucans [24]. The production of EPS, particularly waterinsoluble EPS, is a crucial cariogenic virulence factor of S. mutans [25], which facilitates bacterial adhesion, binding site formation, and biofilm matrix development [26]. Similar trends were also observed in the ∆gtfB mutant, which was engineered in our previous study [19], characterized by a single-site mutation with less Gtfs and water-insoluble EPS content. These findings suggest that the Gtfs-mediated water-insoluble EPS content within biofilms might be crucial in mediating the strain's resistance to ClyR cleavage.
The water-insoluble EPS, a polysaccharide synthesized by Gtfs from S. mutans, was synthesized in vitro using native Gtfs extracts and sucrose for functional assessment of its interaction with ClyR in this study. The water-insoluble EPS was then collected, co-incubated with ClyR, and proteins remaining in the supernatant were analyzed using 12% SDS-PAGE electrophoresis. There was a gradual decrease in ClyR concentration in the supernatant with an increasing concentration of water-insoluble EPS, accompanied by a significant accumulation of ClyR in the precipitates. The results demonstrated the ability of waterinsoluble EPS to adsorb ClyR, leading to its accumulation in deposits. Many studies have shown that phage tail silk proteins can specifically bind to and degrade capsular polysaccharides, EPS, and lipopolysaccharides on bacterial surfaces [27], possibly explaining the precise structural specificity exhibited by ClyR in its recognition of water-insoluble EPS derived from S. mutans. This suggests that the production of waterinsoluble EPS by S. mutans may be crucial in mediating its response to ClyR treatment. In the biofilm phenotype assays, both S. mutans UA159 and the ΔgtfB mutant were grown in BHI supplemented with 1% sucrose to promote EPS production. Our results show that the ΔgtfB mutant forms biofilms with a significantly reduced EPS matrix and displays greater resistance to ClyR-mediated lysis compared to the WT strain. After ClyR treatment, the reduction in bacterial biomass and the disruption of biofilm structure were much less pronounced in the ΔgtfB mutant than in the WT, supporting the conclusion that GtfBmediated EPS production plays a key role in modulating ClyR susceptibility. We also acknowledge that planktonic bactericidal assays performed without sucrose supplementation may partially reflect glucanindependent effects of gtfB deletion, which warrants further investigation.
Furthermore, the specific interaction between S. mutans water-insoluble EPS and the catalytic domain of ClyR provides a novel insight into the molecular mechanisms underlying the bactericidal action of phage lytic enzymes. The preferential interaction of water-insoluble EPS with the catalytic domain of ClyR highlights the critical role of this domain in mediating interactions with the biofilm matrix and facilitating the lytic activity of ClyR. PlyCAC, the catalytic domain of ClyR, corresponds to the CHAP domain of the PlyC lysin (amino acids 314-465). The CHAP domain, which is the enzymatic catalytic domain of PlyC lysin [28], is common to many amidases, including some peptidoglycan hydrolases [29], and mediates nucleophilic attack by using catalytic cysteine residues [30]. Crucial CHAP binding sites for waterinsoluble EPS might exist to facilitate the subsequent adsorption of ClyR, warranting further investigation.
The insights gained from this study have profound implications for developing novel therapeutic strategies for dental caries. Given that insoluble EPS produced by S. mutans can coat not only itself but also neighboring species within dental biofilms, it is plausible that the presence of glucans could increase the local concentration and retention of ClyR in the biofilm matrix, potentially exposing other bacteria embedded within the EPS to its lytic activity. Whether this would lead to effective lysis of other species may depend on the compatibility of their cell wall structures with the catalytic specificity of ClyR. In addition, targeting the biofilm's water-insoluble EPS and enhancing the lytic activity of bacteriophage enzymes might help overcome the limitations posed by the variable susceptibility of S. mutans. Future research should explore the molecular basis of water-insoluble EPS-lysozyme interactions to develop engineered lytic enzymes with improved efficacy against cariogenic biofilms. Additionally, the potential of combining bacteriophage lysozymes with other antimicrobial agents or treatment modalities warrants further exploration to maximize therapeutic outcomes.
Our study highlights the therapeutic potential of bacteriophage chimeric lysozyme ClyR, in combating dental caries by interacting with water-insoluble EPS of S. mutans. The water-insoluble EPS in S. mutans biofilms not only acts as a structural barrier but may also play a facilitating role in localizing ClyR and enhancing its lytic activity. Our data suggest that the presence of abundant EPS, which specifically binds to the catalytic domain of ClyR, may increase the local retention of the enzyme within the biofilm, thereby promoting more efficient lysis of wild-type S. mutans compared to the ΔgtfB mutant. This may explain why the ΔgtfB mutant, despite having less EPS for ClyR binding, is more resistant to lysis, as the sparse matrix is less effective at concentrating the enzyme near the bacterial cells. Collectively, this research contributes valuable knowledge towards developing novel antimicrobial strategies against dental caries and other biofilm-associated infections by providing new insights into the mechanisms underlying bacterial resistance and sensitivity to phage-derived lytic enzymes.
## References
1. Abranches, Zeng, Kajfasz et al. (2018) "Biology of Oral Streptococci" *Microbiol Spectr*
2. Lemos, Palmer, Zeng (2019) "The biology of Streptococcus mutans" *Microbiol Spectr*
3. Ren, Cui, Zeng (2016) "Molecule targeting glucosyltransferase inhibits Streptococcus mutans biofilm formation and virulence" *Antimicrob Agents Chemother*
4. Kim, Barraza, Arthur (2020) "Spatial mapping of polymicrobial communities reveals a precise biogeography associated with human dental caries" *Proc Natl Acad Sci U S A*
5. Rodríguez-Rubio, Gutiérrez, Donovan (2016) "Phage lytic proteins: biotechnological applications beyond clinical antimicrobials" *Crit Rev Biotechnol*
6. Loessner (2005) "Bacteriophage endolysins -current state of research and applications" *Curr Opin Microbiol*
7. Roach, Donovan (2015) "Antimicrobial bacteriophage-derived proteins and therapeutic applications" *Bacteriophage*
8. Freimer, Krause, Mc (1959) "Studies of L forms and protoplasts of group A streptococci. I. Isolation, growth, and bacteriologic characteristics" *J Exp Med*
9. Yang, Yu, Wei (2014) "Engineered bacteriophage lysins as novel anti-infectives" *Front Microbiol*
10. Huang, Luo, Gondil (2020) "Construction and characterization of a chimeric lysin ClyV with improved bactericidal activity against Streptococcus agalactiae in vitro and in vivo" *Appl Microbiol Biotechnol*
11. Oechslin, Daraspe, Giddey (2013) "In vitro characterization of PlySK1249, a novel phage lysin, and assessment of its antibacterial activity in a mouse model of Streptococcus agalactiae bacteremia" *Antimicrob Agents Chemother*
12. Yang, Linden, Wang (2015) "A chimeolysin with extended-spectrum streptococcal host range found by an induced lysis-based rapid screening method" *Sci Rep*
13. Mcgowan, Buckle, Mitchell (2012) "X-ray crystal structure of the streptococcal specific phage lysin PlyC" *Proc Natl Acad Sci U S A*
14. Gilmer, Schmitz, Euler (2013) "Novel bacteriophage lysin with broad lytic activity protects against mixed infection by Streptococcus pyogenes and methicillin-resistant Staphylococcus aureus" *Antimicrob Agents Chemother*
15. Yang, Bi, Shang (2016) "Antibiofilm activities of a novel chimeolysin against Streptococcus mutans under physiological and cariogenic conditions" *Antimicrob Agents Chemother*
16. Xu, Yang, Bi (2018) "Activity of the chimeric lysin ClyR against common gram-positive oral microbes and its anticaries efficacy in rat models" *Viruses*
17. Yang, Zhang, Wang (2017) "A novel chimeric lysin with robust antibacterial activity against planktonic and biofilm methicillin-resistant Staphylococcus aureus" *Sci Rep*
18. Xie, Okinaga, Qi (2011) "Cloning-independent and counterselectable markerless mutagenesis system in Streptococcus mutans" *Appl Environ Microbiol*
19. Gong, Tang, Zhou (2018) "Genome editing in Streptococcus mutans through self-targeting CRISPR arrays" *Mol Oral Microbiol*
20. Delisle, Guo, Chalmers (2012) "Biology and genome sequence of Streptococcus mutans phage M102AD" *Appl Environ Microbiol*
21. Delisle, Rostkowski (1993) "Lytic bacteriophages of Streptococcus mutans" *Curr Microbiol*
22. Whitchurch, Tolker-Nielsen, Ragas (2002) "Extracellular DNA required for bacterial biofilm formation" *Sci*
23. Bowen, Burne, Wu (2018) "Oral biofilms: pathogens, matrix, and polymicrobial interactions in microenvironments" *Trends Microbiol*
24. Hoshino, Fujiwara (2022) "The findings of glucosyltransferase enzymes derived from oral streptococci" *Jpn Dent Sci Rev*
25. Krzyściak, Jurczak, Kościelniak (2014) "The virulence of Streptococcus mutans and the ability to form biofilms" *Eur J Clin Microbiol Infect Dis*
26. Bowen, Koo (2011) "Biology of Streptococcus mutans-derived glucosyltransferases: role in extracellular matrix formation of cariogenic biofilms" *Caries Res*
27. Lee, Tu, Yang (2017) "Structural basis for fragmenting the exopolysaccharide of Acinetobacter baumannii by bacteriophage ΦAB6 tailspike protein" *Sci Rep*
28. Nelson, Schuch, Chahales (2006) "PlyC: a multimeric bacteriophage lysin" *Proc Natl Acad Sci*
29. Pritchard, Dong, Baker (2004) "The bifunctional peptidoglycan lysin of Streptococcus agalactiae bacteriophage B30" *Microbiology*
30. Rigden, Jedrzejas, Galperin (2003) "Amidase domains from bacterial and phage autolysins define a family of gamma-D,L-glutamate-specific amidohydrolases" *Trends Biochem Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12830181&blobtype=pdf | # Research note: The chicken gut virome: Spatiotemporal dynamics and divergent responses to antibiotic versus phytogenic supplementation
Li Yang, Jinlong Ru, Shun Guo, Xueqin Yang, Pengying Li, Li Deng, Xia Wang
## Abstract
This study employed metagenomic sequencing data to comprehensively investigate the gut virome, with a focus on the bacteriophage communities (the phageome), across intestinal regions and developmental stages in 360 chickens. We characterized the spatiotemporal dynamics of phage communities and assessed the impact of chlortetracycline (CTC), an antibiotic, and Macleaya cordata extract (MCE), a phytogenic supplement. Our analysis revealed that phage community assembly was highly structured, exhibiting distinct successional patterns across age and between foregut and hindgut segments. A key finding was the identification of a potential antibiotic-phage synergy, mediated by phage-encoded auxiliary metabolic genes (AMGs) involved in bacterial immune evasion, suggesting a novel mechanism for enhanced infectivity under antibiotic pressure. In contrast, phytogenic supplementation promoted gut ecosystem homeostasis by fostering significantly richer and more diverse phage communities. Our results delineate the fundamental ecology of the chicken gut virome and provide mechanistic insights into how different growth promoters exert contrasting effects on viral populations, supporting the use of phytogenics as sustainable alternatives for animal husbandry.
## Introduction
The gastrointestinal tract (GIT) of poultry is a complex ecosystem harboring a diverse community of microorganisms, including bacteria, archaea, viruses, and fungi, which play a crucial role in host health, nutrient digestion, and immune system development. Among these microbial inhabitants, bacteriophages (or phages), viruses that specifically infect bacteria, are the most abundant biological entities. As key players in the gut microbiome, phages drive bacterial community dynamics through predator-prey relationships and horizontal gene transfer, influencing overall microbial diversity and stability (Naureen et al., 2020).
While the bacterial component of the chicken gut microbiome has been extensively studied, the viral fraction, particularly the phage community, remains relatively unexplored. A critical gap in our understanding lies in the temporal and spatial distribution of these phages along the digestive tract. The chicken GIT presents distinct physicochemical and biological environments from the crop to the ceca, which likely shape unique phage populations in each compartment (Bindari and Gerber, 2022). Furthermore, how these viral communities establish and evolve as the chicken matures is not well-characterized. Elucidating this spatial variation and temporal succession is fundamental to understanding the intricate phage-bacteria interactions that underpin gut homeostasis.
The balance of this gut ecosystem is highly susceptible to external modulators, with dietary and pharmaceutical interventions being primary factors. Antibiotics, historically used as growth promoters and for disease prevention in poultry farming, are known to cause profound disruptions in the gut bacterial microbiota (Hasan and Yang, 2019). However, their specific impact on the phageome remains largely unexplored. This represents a critical knowledge gap, as antibiotic-induced stress on bacteria is known to trigger profound shifts in phage life cycles (lytic vs. lysogenic). Such shifts can not only rapidly destabilize the viral community structure, but may also accelerate the dissemination of antibiotic resistance genes (Torres-Barceló, 2018). Given the global push to reduce antibiotic use in animal agriculture, phytogenics or Chinese herbal medicines have emerged as promising natural alternatives. These compounds are believed to possess antimicrobial, anti-inflammatory, and immunity-enhancing properties (Wang et al., 2024). Their influence on the gut bacterial community is an active area of research, but their specific effects on the phage community are virtually unknown. It is plausible that these herbal compounds selectively inhibit certain bacterial hosts, thereby indirectly shaping the phage populations that prey upon them.
Therefore, this study aims to comprehensively investigate the phage community in the intestines of chickens of different ages. We will first define the spatial and temporal dynamics of the phage community in different intestinal segments. Subsequently, we will evaluate the comparative impact of antibiotic administration and Chinese herbal extract supplementation on the structure and composition of this viral community. By deciphering the interactions between these modulators and the gut phageome, this research seeks to provide novel insights into managing gut health in poultry production, potentially leading to strategies for manipulating the phageome to foster a more robust and resilient microbial ecosystem.
## Materials and methods
## Metagenomic data source and selection
This study utilized publicly available metagenomic sequencing data from the chicken intestinal microbiome project PRJNA417359. Original animal procedures were approved as detailed in (Huang et al., 2018). The original data were generated using Illumina HiSeq 2500 and HiSeq X Ten platforms with a paired-end 150 bp (PE150) sequencing strategy, yielding an average of 12 million reads per sample. From the original dataset (n=495), a subset of 360 metagenomic samples was selected for analysis based on two stringent criteria: 1) completeness of metadata across all key experimental variables (developmental stage, intestinal segment, and treatment group) to enable robust multi-factorial comparisons; and 2) detectability of viral signals, as samples failing to yield identifiable viral contigs or sufficient viral reads during the initial screening were excluded to ensure analytical reliability. The selected samples were derived from Arbor Acres broilers (AA, n=180) and Local yellow-feather chickens (LY, n=180), both breeds having comprehensive spatiotemporal data.
## Experimental design and animal groups
The experimental design encompassed multiple age points and dietary treatments. AA broilers were sampled at five ages (1, 7, 14, 28, and 42 days), while LY chickens included an additional 56-day time point. At each age (except for day 1), chickens were randomly assigned to one of five dietary treatment groups. (1) BLANK: basal diet; (2) CTC: basal diet supplemented with 50 mg/kg Citifac® (chlortetracycline 20% w/w premix); (3) MCE-L, MCE-M, MCE-H: basal diet with 15, 50, 150 mg/kg Sangrovit® (Macleaya cordata extract 3.75% w/w premix), and in these groups' names, L stands for low, M for medium, and H for high. Only the BLANK group was included for the 1-day-old chicks.
## Metagenomic virome analysis: phage identification, classification, lifestyle prediction and host relationship study and AMG identification
To profile the gut virome, we employed our recently developed ViroProfiler pipeline (Ru et al., 2023), which integrates the key analytical stages from viral contig identification and quality control to taxonomic classification, functional annotation, and viral abundance estimation. Briefly, assembled contigs were initially screened with geNomad (version 1.8.0, -min-score 0.75). A refined identification was then performed: CheckV (version 1.0.1) removed host contamination and assessed completeness, and a consensus of VirSorter2 (version 2.2.3) and VIBRANT (version 1.2.1) results was used to retain viral contigs supported by at least two tools for downstream analysis. Putative bacterial hosts were predicted using iPHoP (version 1.3.3), with a high-confidence score threshold (score ≥ 90). For functional characterization, AMGs were identified and annotated using VIBRANT (version 1.2.1) and DRAM-v (version 1.3) with their default settings. Quality-filtered reads were aligned to the non-redundant viral catalog using Bowtie2, and viral abundance was quantified using CoverM (version 0.7.0).
## Phage composition analysis
For microbial diversity analysis, ACE, Chao1, Pielou's evenness, Observed Richness, Shannon, and Simpson indices were used. The overall differences in phage community structures were evaluated by non-metric multidimensional scaling (NMDS) based on Bray-Curtis dissimilarity values.
## Statistical analyses
All statistical analyses were performed in R (v4.4.2). Data manipulation and visualization utilized the tidyverse ecosystem (v1.3.0). Student's t-test and Wilcoxon rank sum test were employed for group comparisons.
## Results and discussion
The spatiotemporal heterogeneity of phage distribution in the gut Metagenomic analysis of 360 chickens yielded 12,400 viral contigs, which were clustered into 2,118 viral operational taxonomic units (vOTUs) after quality filtering (571 of medium-to-high quality; Fig. 1A,B). Taxonomic classification assigned the majority to the classes Caudoviricetes, Malgrandaviricetes, and Arfiviricetes (Fig. 1C,D). Host prediction via iPHoP revealed 9,061 virus-host linkages, with lytic phages (5,831 vOTUs) showed a strong preference for Bacillota_A, while temperate phages (3,230 vOTUs) showed a broader host range across 10-11 bacterial phyla (Fig. 1E).
Analysis of the sample groups revealed significant spatiotemporal dynamics in the phage communities. NMDS showed clear separation between foregut and hindgut communities (AA: R² = 0.09, P < 0.001; LY: R² = 0.16, P < 0.001; Fig. 1F,H) and across developmental stages (AA: R² = 0.25, P < 0.001; LY: R² = 0.47, P < 0.001; Fig. 1G,I). Notably, the most substantial community shift occurred between 1 and 7 days of age. Furthermore, the primary site of age-related succession differed by breed: the hindgut was more dynamic in AA broilers, whereas the foregut changed more markedly in LY chickens.
Our results establish that the chicken gut virobiota is organized along two fundamental ecological gradients: space (intestinal segment) and time (age), which were consistent with the previous studies in mice, indicating that the intestinal viral compartment has a strong compartmentalization feature (Kim and Bae, 2016). The significant segregation between foregut and hindgut communities (Bray-Curtis distance; P < 0.001) likely originates from the distinct physiological environments of the small intestine (e.g., duodenum, jejunum) and the large intestine (e. g., cecum, colorectum). The small intestine is primarily responsible for the rapid digestion and absorption of nutrients, characterized by a relatively oxygen-rich, lower pH environment with a constant flux of nutrients. In contrast, the large intestine, particularly the cecum, functions as an anaerobic fermentation chamber with a near-neutral pH, rich in complex dietary fibers. This creates a unique ecological niche for strict anaerobes and their associated phages. The functional gradient from "digestion and absorption" to "fermentation" shapes distinct bacterial host communities, which in turn governs the structure of the phage communities (Zhao et al., 2025). Furthermore, the successional patterns of these phage communities were breed-specific. In AA broilers, optimized for rapid growth, the most dramatic changes occurred in the hindgut, the major site of fermentation and energy harvest. Conversely, in LY chickens, the foregut exhibited greater dynamism. This breed-level divergence may reflect genetic differences in niche adaptation along the intestinal tract. Variations in foregut physiology (e.g., enzyme secretion, motility) between LY and AA chickens could alter the ecological dynamics of both the microbiota and its phages. This implies that host genetics indirectly sculpt the virome by fine-tuning the physiological landscape of the gut. The most pronounced shift across all groups happened between day 1 and day 7, a period of intense post-hatching development where digestive enzyme activity surges and intestinal villi mature. This synchrony indicates that phage community assembly is an integral component of host gut development, rather than a passive process. The initial colonization and succession of phages appear to be orchestrated by the changing landscape of the gut itself, much like the ecological succession seen in a newly formed habitat.
## Differential effects of antibiotic and herbal additives on phage communities
Having established the baseline spatiotemporal dynamics, we investigated the impact of the CTC and the MCE on the gut phage communities. NMDS revealed no significant overall clustering among the Blank, CTC, and MCE groups (P > 0.05; Fig. 2A). However, alpha diversity analysis uncovered distinct, segment-specific responses (Fig. 2B).
In the foregut of both chicken breeds, MCE treatments, particularly at medium concentration (MCE-M), consistently resulted in significantly higher richness (ACE, Chao1) and diversity (Shannon) compared to the CTC group. In AA chickens, only MCE-M significantly increased richness indices above the untreated control, while CTC supplementation significantly reduced community evenness (Simpson index). In the hindgut, responses were more variable and breed-specific, with MCE-M reducing estimated richness (Chao1) in AA chickens and MCE-L decreasing observed richness in LY chickens. Our findings demonstrate that CTC and MCE exert fundamentally distinct influences on the gut virobiota, moving beyond a purely community-level view to reveal hostspecific and molecular consequences. At the ecological level, MCE, especially at medium concentration, acted as an ecological stabilizer in the metabolically active foregut, fostering a richer and more diverse phage community than CTC. This observation is consistent with the hypothesis that MCE promotes a more diverse and beneficial bacteriome, thereby expanding the host ecological niche space for phages. In contrast, CTC's reduction of community evenness signifies a selective pressure that favors a few dominant, pre-adapted phage taxa, likely through the reduction of bacterial diversity and niche availability.
We next analyzed phage abundance based on their predicted bacterial hosts. In the foregut, the abundances of phages infecting Escherichia, Enterococcus (and its relatives Enterococcus_F, Enterococcus_E, Enter-ococcus_B), and Ligilactobacillus were significantly altered (Fig. 2C). Phages associated with Escherichia and Enterococcus decreased in the CTC and MCE-L groups but were stable in MCE-M/H groups. Conversely, Ligilactobacillus phages increased in abundance in the CTC group. In the hindgut, phages targeting Desulfovibrio were significantly reduced specifically in the MCE-L group. This principle of host-dependent effects is further crystallized by examining specific phage-host pairs. The concurrent suppression of Escherichia and Enterococcus phages by CTC and MCE-L suggests a disruption of a potential pathogenic synergy between these genera. The stability of these populations under MCE-M/H, however, points to a more nuanced, "top-down" ecological management that suppresses virulence without collapsing the underlying network. The targeted reduction of Desulfovibrio phages (and by proxy, their host) in the hindgut by MCE-L is a positive signal, indicating an ability to mitigate a known contributor to gut dysbiosis even at low concentrations.
Functional profiling identified 43 AMGs within the vOTUs, involved in nucleotide, carbohydrate, amino acid, and vitamin B12 metabolism (Fig. 2D). The most prevalent AMG was a DNA methyltransferase (DNMT1). Differential analysis revealed that most AMGs were downregulated in MCE groups (Fig. 2E). Strikingly, the DNMT1 gene exhibited opposing regulation: it was downregulated in the foregut of MCE-treated AA chickens but upregulated in CTC-treated LY chickens, suggesting a compartmentalized and treatment-specific effect (Fig. 2E).
Functionally, these AMG abundance patterns point to a potential epigenetic dimension in the phage-additive interaction. The upregulation of DNMT1 in CTC-treated groups implies an enrichment for phage populations equipped to evade bacterial restriction-modification systems via molecular mimicry (Takahashi et al., 2025), thereby escalating the co-evolutionary arms race. Conversely, its downregulation under MCE treatment indicates a de-escalation of this conflict. We propose that by dampening overall bacterial virulence and activity, MCE lowers the selective pressure for sophisticated phage counter-defenses, thereby favoring stable coexistence over antagonism. The membrane-disrupting action of MCE's key alkaloid, sanguinarine (Qu et al., 2025), may synergize with this ecological effect by physically facilitating phage infection. Despite the lack of direct expression data, these distinct abundance patterns robustly highlight that MCE and antibiotics impose fundamentally different ecological selection pressures on the viral community. This virome-mediated homeostasis, characterized by preserved diversity and de-escalated antagonistic co-evolution, is consistent with the phenotypic improvements previously recorded in the same cohort, where MCE supplementation significantly enhanced body weight gain and feed conversion ratios (Huang et al., 2018). This concordance suggests that the viral ecological stability identified here is a key mechanistic underpinning of the zootechnical benefits observed in phytogenic-fed poultry.
We also acknowledge several limitations in this study. First, as mentioned above, this is a bioinformatics analysis based on metagenomic DNA. Therefore, our findings regarding AMG functions and phage-host interactions are predictions that reflect genetic potential rather than active gene expression, and need to be verified by future experiments (such as phage isolation and infection assays). Second, viral abundance was quantified via metagenomic read mapping, which detects viral DNA but does not distinguish between infectious virions and free DNA or inactive particles. Additionally, despite the use of state-ofthe-art tools, host prediction remains a challenge in viral ecology, potentially influencing the precision of the inferred association networks. Finally, while our dataset encompassed key developmental stages and two commercial breeds, future studies incorporating broader genetic backgrounds and diverse rearing environments are needed to verify the generalizability of these spatiotemporal patterns.
In summary, this study systematically characterizes the spatiotemporal developmental landscape of the broiler gut phageome and reveals that phytogenics (MCE) and antibiotics (CTC) drive divergent ecological trajectories within this viral community. Unlike antibiotics, which impose selective pressure that reduces community evenness, MCE fosters a more diverse and stable virobiota, likely by shaping a healthier bacterial host environment. These findings offer a novel, viro-centric perspective on the microecological effects of antibiotic alternatives, suggesting that phytogenics contribute to gut homeostasis by modulating interactions from the ecological down to the potential epigenetic level.
## Ethics statement
This study was conducted exclusively using publicly available metagenomic data downloaded from the NCBI Sequence Read Archive (SRA) under BioProject accession number PRJNA417359. As this research involved only the bioinformatic analysis of pre-existing datasets and did not involve any direct handling, intervention, or euthanasia of live animals by the authors, specific Institutional Animal Care and Use Committee (IACUC) approval or descriptions of euthanasia methods are not applicable to this study. The original animal experiments were performed in accordance with the ethical guidelines as described in the primary publication associated with the dataset.
## References
1. Bindari, Gerber (2022) "Centennial review: factors affecting the chicken gastrointestinal microbial composition and their association with gut health and productive performance" *Poult. Sci*
2. Hasan, Yang (2019) "Factors affecting the composition of the gut microbiota, and its modulation" *PeerJ*
3. Huang, Zhang, Xiao et al. (2018) "The chicken gut metagenome and the modulatory effects of plantderived benzylisoquinoline alkaloids"
4. Kim, Bae (2016) "Spatial disturbances in altered mucosal and luminal gut viromes of diet-induced obese mice" *Environ. Microbiol*
5. Naureen, Dautaj, Anpilogov et al. (2020) "Bacteriophages presence in nature and their role in the natural selection of bacterial populations" *Acta Biomed*
6. Qu, Huang, Zhu et al. (2025) "Targeting membrane integrity and imidazoleglycerol-phosphate dehydratase: sanguinarine multifaceted approach against Staphylococcus aureus biofilms" *Phytomedicine*
7. Ru, Khan Mirzaei, Xue et al. (2023) "ViroProfiler: a containerized bioinformatics pipeline for viral metagenomic data analysis" *Gut Microbes*
8. Takahashi, Hiraoka, Matsumoto et al. (2025) "Host-encoded DNA methyltransferases modify the epigenome and host tropism of invading phages"
9. Torres-Barceló (2018) "The disparate effects of bacteriophages on antibiotic-resistant bacteria" *Emerg. Microbes Infect*
10. Wang, Deng, Chen et al. (2024) "Phytogenic feed additives as natural antibiotic alternatives in animal health and production: a review of the literature of the last decade" *Anim. Nutr*
11. Zhao, Ni, Nan et al. (2025) "Expanding the chicken gut virome: uncovering viral diversity, host interactions" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12737361&blobtype=pdf | # Study of the Activity of the Staphylococcus aureus Phage vB_SaS_GE1 Against MRSA Clinical Isolates and Its Impact on the Formation of Dual-Species Biofilms with P. aeruginosa
Nino Grdzelishvili, Davit Lazviashvili, Aleksandra Kurowska, Krzysztof Pawlik, Łukasz Łaczmanski, Elene Kakabadze, Elene Zhuravliova, Nina Chanishvili, Nata Bakuradze
## Abstract
Bacteriophage therapy is regarded as a promising alternative for treating and preventing antibiotic-resistant bacterial infections. Methicillin-resistant Staphylococcus aureus (MRSA) is one of the most prevalent and difficult-to-treat pathogens. S. aureus also contributes to the formation of both single-and mixed-species biofilms. Treating biofilms remains a major challenge for antibiotic-based eradication of pathogens, as the biofilm matrix provides a protective barrier for bacteria. The selection of highly active phages targeting S. aureus is therefore crucial for medical applications, given the high prevalence and drug resistance of this pathogen. In this study, S. aureus phage vB_SaS_GE1 (GE1) was isolated and characterized as a potential therapeutic agent. The phage was isolated and propagated, and its host range was determined using standard methods. Whole-genome sequencing and annotation of the phage DNA were performed. A time-kill assay and evaluation of the anti-biofilm activity of the Staphylococcus phage, both alone and in combination with Pseudomonas phage GEC_PNG3 (PNG3) on mixed-species biofilms, were conducted. The results indicated that GE1 is a lytic phage that does not carry virulence-determining genes. The time-kill assay demonstrated sustained lytic activity of GE1 without the emergence of phage-resistant mutants in the tested MRSA strains. Although phage treatment increased biofilm matrix production compared to the control, the viable cell count within the biofilms was reduced. Overall, the characteristics assessed indicate that vB_SaS_GE1 is safe and exhibits strong antibacterial activity against MRSA strains.
## 1. Introduction
Bacteriophages, viruses that infect bacteria, are often recommended for treating bacterial infections that do not respond to conventional antibiotics. Depending on resistance properties and infection composition, phages may be administered individually, in mixtures known as phage cocktails, or in combination with antibiotics [1,2]. The demand for bacteriophage therapy has risen due to the global spread of antibiotic-resistant pathogens. Among these, Staphylococcus aureus stands out as a major cause of severe and potentially lethal infections [1], including pneumonia, wound infections, sepsis, osteomyelitis and other conditions that often require hospital admission and long-term treatment. Both MRSA and methicillin-susceptible Staphylococcus aureus (MSSA) have contributed to an increasing epidemiological and economic burden in many countries [3].
In addition, Staphylococcus aureus is frequently detected as a co-infecting agent in mixed infections, such as chronic wound or respiratory infections, and it can be often underrepresented in biofilms formed by Pseudomonas aeruginosa. The formation of mixed-species biofilms by P. aeruginosa and S. aureus may render these communities up to 1000 times more resistant to antibiotics [2]. The complex nature of such infections slows down the treatment and therefore requires the development of new therapeutic approaches that would offer alternative solutions for eliminating resistant bacteria. Bacteriophage therapy has a long history of clinical use, particularly in countries of the post-Soviet Union, and it has recently been reintroduced worldwide as a promising solution for difficult-to-treat bacterial infections. Recent retrospective analyses of hundred clinical cases globally highlight the successful outcomes of phage therapy [4]. Growing evidence shows that phages can act synergistically with antibiotics-often reversing bacterial antibiotic resistance [5] or enhancing the antibacterial activity of the drugs [6]. Phages also contribute to biofilm degradation, both directly or by promoting antibiotic efficacy. Phage-derived hydrolytic enzymes such as depolymerases and lysins degrade the exopolysaccharide structure of biofilm, improving the penetration of the phage virions as well as antibiotics to reach the bacterial cells [3].
Despite the growing interest in therapeutic phages targeting S. aureus, only a limited number of well-characterized therapeutic candidates have been evaluated against MRSA in the context of mixed-species biofilms relevant to chronic wound infections. Existing phages such as Sb-1 and ISP have demonstrated promising activity [5,6], but there remains a need for additional strictly lytic, genomically safe phages with broad host range and documented performance in dual-species communities involving P. aeruginosa. In this study, we isolated and characterized the S. aureus phage vB_SaS_GE1 (GE1), assessed its genomic safety and taxonomic placement and evaluated its activity against MRSA isolates in both singleand mixed-species biofilms with P. aeruginosa. Our aim was to determine whether GE1 represents a suitable candidate for inclusion in phage-based strategies targeting biofilmassociated MRSA infections in chronic wounds.
## 2. Materials and Methods
## 2.1. Bacterial Strains
Staphylococcus aureus and Pseudomonas aeruginosa bacterial host strains used for the isolation and further propagation of bacteriophages were obtained from the collection of Eliava Institute of Bacteriophages, Microbiology and Virology (EIBMV). In total, 66 S. aureus clinical isolates were screened for susceptibility to phage GE1, of which 6 were identified as methicillin-resistant S. aureus (MRSA) and selected for detailed characterization, and one biofilm-forming P. aeruginosa clinical isolate (1147) was used in mixed-culture and biofilm experiments.
## 2.2. Antibiotic Susceptibility of Bacterial Strains
Staphylococcus aureus isolates from the Eliava Institute collection were tested for antibiotic resistance. A Kirby-Bauer antibiotic susceptibility test was used to select the methicillinresistant S. aureus (MRSA) strains based on a cefoxitin susceptibility test. Briefly, freshly grown bacteria were grown on Muller-Hinton solid agar. Antibiotic disks were distributed over the bacterial lawn. After overnight incubation, inhibition zones were measured [7]. Interpretation of the results was based on the latest EUCAST manual (version 2024).
## 2.3. Bacterial DNA Extraction and MecA Gene Detection in MRSA Isolates
Methicillin-resistant S. aureus isolates were selected to be evaluated for the presence of the mecA gene. Freshly grown colonies grown on TSA agar were harvested and diluted in a lysis buffer. Bacterial DNA isolation was performed using an UltraClean ® Microbial DNA Isolation Kit (MO BIO, Carlsbad, CA, USA) following the manufacturer's instructions. For the mecA gene detection, primers (F: AAAATCGATGGTAAAGGTTGGC R: AGTTCT-GCAGTACCGGATTTGC) were selected based on the publication by J.H.Lee [8], and the related PCR steps were performed, respectively. To prepare 20 µL, mastermix AmpliTaq Gold ® Fast Master Mix (×1) at 10 µL, forward primer at 0.4-1 µL with final concentration of 0.2-0.5 µL/mL and reverse primer at 0.4-1 µL with final concentration of 0.2-0.5 µL/mL were used. The volume of 20 µL was filled with sterile DNA-se free water. The PCR amplification was performed with a hot-start, and primary denaturation was set at 95 • C for 10 min, which was followed by 40 cycles of denaturation at 94 • C for 30 s. Annealing was performed at 55 • C for 30 s and elongation at 72 • C for 1 min with a final elongation step at 72 • C for 5 min. For the detection of amplified gene fragments, gel electrophoresis was performed. A 1.5% agarose gel was prepared with 2.5 µL of ethidium bromide in 80 mL of agarose gel. Gel dye (6×) was diluted with PCR samples and run for 40 min at 80 v to detect 533 bp amplified gene fragments [8].
## 2.4. Isolation of the Bacteriophages Using the Enrichment Method
Isolation of the species-specific phages was performed using various sources, such as wastewater and clinical waste samples. The enrichment method was based on mixing 9 mL of potential phage source with 1 mL of 10× concentrated TSA broth and 1 mL of the bacterial culture. The mixture was incubated for 24 h aerobically at 37 • C. After incubation, the mixture was centrifuged at 6000 rpm for 30 min for pelleting bacterial debris. The supernatant was filtered through 0.20 µm pore filters [9].
## 2.5. Detection of Phages Active Against Target Pathogen and Their Propagation
Presence of species-specific phage virions in the obtained filtrates was checked using "the streak method". The bacterial cultures were streaked on the TSA agar with a 10 µL loop and air-dried. A total of 10 µL of each filtrate was dropped on the surface of the streaks. The Petri dishes were incubated for 18-24 h at 37 • C to reveal the lytic zones on the grown streaks of the bacterial cultures. Clear lytic zones were cut out with a sterile loop and incubated in 2 ml of TSA broth for 2 h at 37 • C, to which 20 µL of chloroform was then added, thoroughly vortexed and stored at 4 • C at least for 30 min until further usage. For the propagation of the phages, the suspension was titrated using the double-layer agar (DLA) method. The bacterial isolates showing phage susceptibility were selected for further analysis of phage host ranges. Ten-fold serial dilution of the phage suspension was performed up to 10 -5 , and then 1 mL of each dilution was mixed with 100 µL of susceptible bacterial culture in the exponential growth phase, and a semisolid (0.6%) TSA agar warmed to 40 • C was added. The mixture was vortexed for a few seconds and applied on top of the solid TSA agar. The plates were air-dried and incubated for 18-24 h at 37 • C. After incubation, phage plaques were formed, and the lytic zones were evaluated according to their sizes and transparency indexes. To concentrate phages, several plates with the corresponding dilution showing a meshwork of plaques were prepared. The top layer from these plates were scraped off and collected in a 45 mL centrifuge tubes and centrifuged at 6000 rpm for 30 min. Afterwards, the supernatant was filtered to remove the bacterial debris, and the DLA method was performed to determine the concentration of the phage particles in the filtrate. Countable plates were used, and the number of plaques was multiplied by the dilution number and divided by the volume applied [9].
## 2.6. Study of the Phage Morphology
Transmission electron micrographs of the phages were obtained using JEOL-JEM-1400. Phage concentrates with a titer of >2-5 × 10 10 pfu/mL were used for the transmission electron microscopy (TEM) to study the morphology of the isolated phage virion, and the sample was prepared as described previously [10].
The size of the phages was calculated using the following formula:
Size [Angstrom (Å)] = size of the image [mm] × 10 7 magnification (×220,000).
## 2.7. Phage Host Range and Efficiency of Plating (EOP)
Sixty-six bacterial isolates related to Staphylococcus aureus were used to study the phage host range. For this purpose, the streak method was applied. To evaluate the efficiency of plating, the phage was titrated and grown with each susceptible bacterial isolate using the DLA method. The efficiency of plating was measured by dividing the number of plaque-forming units (PFU) on target bacteria by its PFU on the host bacteria and was classified as given in Table 1 [11,12]. The isolates demonstrating high production (≥0.5) were selected as the host for further research.
## The Average Efficiency of Plating Value
## Classification Characteristics
## ≥0.5 High production
The productiveness of infection of the target bacterial strain results in a PFU equal to 50.0% or more compared to the PFU found for the host.
0.1-0.5
## Medium production
The productiveness of infection of the target bacterial strain is estimated between 10% and 50% of the PFU found for the host.
## 0.001-0.1 Low production
The productiveness of infection of the target bacterial strain is estimated between 1% and 0.01% of the PFU found for the host.
## ≤0.001
Inefficient production Infection productiveness on the target strain is estimated below 0.001% of the PFU found for the primary bacterial host.
## 2.8. Lytic Stability Assay
To assess the stability of GE1-mediated lysis, time-kill experiments were performed at different multiplicities of infection (MOI). Freshly grown MRSA cultures were adjusted 0.5 McFarland standard in TSA broth, to which phage GE1 was added to achieve MOI values of 1, 0.1 and 0.01, calculated as the ratio of plaque-forming units (PFU) to bacterial colony-forming units (CFU) at inoculation. The mixtures were incubated for 24 h to visualize the ability of the phage GE1 to suppress the growth of the target bacteria within a certain period of time. The incubation of the mixtures was performed for 24 h, and the bacterial growth was monitored at 3 h, 6 h, 18 h and 24 h [13].
## 2.9. Isolation of DNA for Whole-Genome Sequencing
The concentrate of the phage GE1, with a titer of ~3 × 10 10 PFU/mL, was ultracentrifuged at 21,000 rpm for 50 min to pellet the viral particles. The pellet was then resuspended in PBS and used for viral DNA isolation. To ensure samples remained uncontaminated with bacterial DNA and/or RNA in the subsequent experiments, we conducted the removal of bacterial genomic material. Phage lysates (450 µL) were treated with DNase I (1 U), RNase A (10 mg/mL) and 10× DNase buffer (50 µL) for 1.5 h at 37 • C to digest bacterial nucleic acids while preserving encapsulated phage genomes. Enzymes were inactivated with EDTA (20 mM), followed by capsid digestion using Proteinase K (1.25 µL; 20 mg/mL) for 1.5 h at 56 • C, without shaking [14]. To extract bacteriophage DNA, we used a DNA isolation kit (Qiagen DNeasy Blood & Tissue Kit, Hilden, Germany). Genome concentration was determined using a QuantiFluor ® dsDNA System on a Quantus™ Fluorometer following the manufacturer's instructions.
Isolation and sequencing of the phage GEC_PNG3 (PNG3) DNA is described in a previous publication [15].
## 2.10. Whole-Genome Sequencing
Illumina sequencing libraries were prepared with a Nextera XT DNA Library Prep Kit following the manufacturer's instructions [16]. Briefly, 150 ng of genomic DNA (5 ng/µL in 30 µL) was tagmented using Bead-Linked Transposomes at 55 • C for 15 min, the reaction was stopped with Tagment Stop Buffer at 37 • C for 15 min and the samples were washed three times with Tagment Wash Buffer on a magnetic stand. The tagmented DNA was then amplified using Enhanced PCR Mix with unique i7/i5 index adapters (5 µL each) under the following cycling: 68 • C for 3 min; 98 • C for 3 min; 5 cycles of 98 • C for 45 s, 62 • C for 30 s, 68 • C for 2 min; final extension at 68 • C for 1 min. Libraries were purified using Sample Purification Beads, quantified with a QuantiFluor ® dsDNA assay and assessed on a TapeStation 2200. Equal volumes of libraries were pooled, denatured (0.2 N NaOH), diluted to 6 pM, spiked with 1% PhiX and sequenced on a MiSeq Nano v2 300-cycle kit.
## 2.11. Genome Assembly and Annotation of the Phage GE1
Raw reads were quality-checked with FastQC v0.11.9, adapter-and quality-trimmed using Trimmomatic v0.32 (ILLUMINACLIP: NexteraPE-PE.fa:2:30:5, SLIDINGWINDOW: 4:25, MINLEN: 36) and assembled de novo using SPAdes v3.5.0 with default parameters. Annotation was performed using the PHROG database via the PHANOATE program. The visualization was prepared using the PHAROKKA program. The phage genomic DNA sequence was compared with other phage genomes using BLASTn +2.10 against the nucleotide database. Transfer RNA (tRNA)-encoding genes were identified using online tools tRNAscan-SE v2.0 (available at http://trna.ucsc.edu/tRNAscan-SE Access date: 10 June 2020) and ARAGORN v1.2.38. Virulence factors and antimicrobial resistance genes were screened with the EDGE v1.5 Gene Family module with default settings. Phage lifestyle was predicted with the DeepPL program. The presence of sequences associated with CRISPR-Cas systems was evaluated using CRISPRCas Finder (https://crisprcas.i2 bc.paris-saclay.fr/. Access date: 20 October 2025) [17]. Potential anti-CRISPR protein (Arc) sequences were searched using the ArcHub server (https://pacrispr.erc.monash. edu/AcrHub/index.jsp Access date: 20 October 2025), which integrates three predictors: PaCRISPR, AcRankeri and an HMM-based predictor. Gene-family profiling (read-and contig-based) was performed against the Antibiotic Resistance Database, Resfams antibiotic resistance functions and Virulence Factors of Pathogenic Bacteria Database as well as the Comprehensive Antibiotic Resistance Database. Taxonomy of the phage GE1 was determined, and proteomic tree analysis was accomplished using ViPTree version 4.0.
## 2.12. Nucleotide Sequence Accession Number
The genome sequence for the phage GE1 was deposited into GenBank under the accession number OM030343.1.
## 2.13. Biofilm Formation Assay and Evaluation on Viable Cell Count
Multidrug-resistant Staphylococcus and Pseudomonas isolates were tested for biofilm production. Bacterial cultures freshly grown in TSA broth were diluted with 1% glucose containing TSA broth at a 1:1 ratio [18]. A total of 100 µL of each bacterial suspension was distributed to a 96-well plate in triplicate. The plate was incubated from 4 to 24 h at 37 • C temperature. After incubation, for phenotypical evaluation of the formed biofilms, a staining procedure was performed. First, the bacterial suspension was discarded to remove the planktonic cells, and the wells were gently washed twice with PBS. Afterwards, 125 µL of 0.1% crystal violet was added to the wells to stain the biofilms and incubated for 10-15 min at the room temperature. For the next step, the stain was washed with PBS 3 to 4 times and dried for several hours [18,19]. The optical density (OD) of the stained biofilms was measured with a Thermo Scientific™ Multiskan SkyHigh Microplate spectrophotometric reader at 570 nm. For the evaluation of the number of viable bacterial cells in the biofilms, after initial washing of the wells twice with PBS, 100 µL of PBS was used to remove the formed biofilms as a homogenized suspension and serially dilute them. A total of 10 µL of each dilution was distributed on TSA solid agar for the bacterial cell count to be performed [20].
## 2.14. Mono-and Combined Phage Activity Against Single and Mixed-Agent Biofilms
A total of 2-3 colonies of bacterial culture of each species freshly grown for 18 h were separately diluted in sterile saline solution and adjusted to 1.0 McFarland turbidity standard [21]. A total of 100 µL of each bacterial suspension was added separately to 100 µL of TSA broth containing 1% glucose for stimulating better biofilm formation by each bacterial culture [18]. To produce mixed biofilms, bacterial suspensions were combined with a ratio of 1:3 v/v of P. aeruginosa to S. aureus, and 100 µL of the mixed suspension was then added to 100 µL of 1% glucose TSA broth. A total of 200 µL of the received mixture was transferred in triplicate into 96-well plates and incubated for 24 h at 37 • C under static conditions. After incubation, wells with the grown biofilm were categorized into following experimental groups: (a) Staphylococcus aureus biofilms treated with GE1, (b) Pseudomonas aeruginosa biofilm treated with phage PNG3, (c) mixed biofilms treated with the phage cocktail (GE1 + PNG3) and (d) Sterile TSA broth (control). Old media was discarded from all wells, and the wells were gently washed twice with saline solution to remove planktonic bacterial cells [20]. Then, 200 µL of the phages GE1 and PNG3 diluted in TSA broth to a final titer of 3-5 × 10 6 was added separately to their target bacterial biofilm groups, as well as in combination at a ratio of 1:1 v/v with the mixed-biofilm group, to receive MOI 0.1. An equal amount of sterile TSA broth was added to the control group. The plate was incubated for 18-24 h at 37 • C under static conditions. After incubation, antibiofilm activity was tested as described in a previous method by staining the wells with crystal violet and measuring the OD with a spectrophotometric reader at 570 nm. To measure the viable cell number, wells were gently washed twice with 200 µL of saline solution. The biofilm-entrapped cells were carefully collected by pipetting 200 µL of TSA broth and by simultaneously scraping off the biofilm with the tip of a pipette. From each well, 200 µL of biofilm was removed and placed in 1800 µL of saline solution, vortexed for homogenization and serially diluted to perform a viable bacterial count assay [22]. For counting the cells of each bacterial isolate, the following selective media were used: Pseudomonas-selective agar, and mannitol salt agar for cultivating the S. aureus cells [23].
Experiments assessing the impact of phages on biofilm formation were conducted in triplicate. The arithmetic mean of the triplicate measurements was calculated and used for data analysis. To confirm the robustness and reproducibility of the findings, independent experimental replicates (n ≥ 3) were performed.
## 2.15. Time-Kill Assay
The essential part of the study was to evaluate the ability of the phage GE1 to degrade single-and mixed-species infections. A clinical isolate from the bacterial strain collection P. aeroginosa 1147, on which the phage PNG3 demonstrated high lytic activity, was selected, along with two MRSA isolates (643 and 8497), to perform the time-kill assay. Bacterial distribution in a 96-well plate was performed similarly to the phage anti-biofilm activity assay described above. The phage solutions were applied in the same concentrations. A spectrophotometric reader was used to monitor the antibacterial activity of the phage solutions within 16 h at 600 nm [24,25].
## 3. Results
## 3.1. Antibiotic Susceptibility of Bacterial Strains
First, the antibiotic susceptibility profiles and methicillin resistance status of the S. aureus isolates were determined to define the MRSA panel used for phage testing. Six methicillin-resistant S. aureus (MRSA) strains were selected out of 66 S. aureus isolates based on a Kirby-Bauer antibiotic susceptibility assay and mecA gene amplification tests. They were identified as MRSA based on phenotypical evaluation, showing resistance to cefoxitin (Table 2). The PCR revealed a 533 bp long fragment of the mecA gene (Figure 1) (Table 3).
## 3.2. Bacteriophage Isolation and Morphological Characterization
Next, bacteriophage S. aureus vB_SaS_GE1 was isolated from the clinical waste of nasopharyngeal washing using an MSSA isolate as a host (S. aureus 152). The isolated phage produced 1 mm clear plaques on the agar plates. Electron microscopy of the GE1 phage exhibited typical characteristics of phages belonging to the Herelleviridae family, displaying an icosahedral head (72.7 nm) as well as a contractile tail (163 nm) and a baseplate attached to the tail (Figure 2). These observations are consistent with classification of GE1 as a lytic S. aureus phage and confirm its suitability for further genomic and functional analyses.
## 3.3. Study of the Phage Host Range and Efficiency of Plating
Having established the morphological features of GE1, we proceeded to evaluate its host range and efficiency of plating on S. aureus clinical isolates. A spot test assay was performed on 66 S. aureus strains, out of which 57 (86.3%) showed susceptibility according to the presence of the lytic plaques after phage application. A high productive infection (EOP ≥ 0.5) of the GE1 phage was evident for only 32.0% of the tested strains that previously have shown susceptibility results in the spot test. However, high productive infection together with the medium infection (EOP ≥ 0.1) of GE1 was shown for 70.0% out of susceptible S. aureus strains. Collectively, these data show that GE1 infects a high proportion of S. aureus isolates and achieves high or medium productivity in most susceptible strains, supporting its use as a broad anti-Staphylococcus aureus candidate.
## 3.4. Bacteriophage Genome Sequencing, Assembly and Annotation
The complete genome of bacteriophage GE1 was sequenced and annotated to assess its taxonomic position and to screen for virulence, lysogeny and antimicrobial resistance determinants. The complete genome of phage GE1 is 138,106 bp in size, has a GC content of 30.2% and contains 219 coding sequences (CDSs), of which 67 have assigned functional annotations, 41 are hypothetical proteins and 4 are tRNAs (Figure 3a, Table 4). No hits to known problematic virulence genes were detected at either the read or contig level. For comparative analysis, the GE1 genome sequence was queried by BLASTn, revealing 99.92% and 99.95% nucleotide identity to phages Sb-1 (NC_023009.1) and ISP (NC_047720.1), respectively, both of which are well-studied therapeutic phages [26,27]. Based on the BLASTn results, phage GE1 was assigned to the genus Kayvirus, within the subfamily Twortvirinae of the Herelleviridae family, representing a syphovirus morphotype. DeepPL analysis indicated that GE1 is a strictly lytic phage and lacks genes associated with lysogeny. CRISPRCasFinder identified only a single spacer and two direct repeat sequences; however, neither a leader sequence nor a cas gene cluster were detected. Out of the three algorithms used in the AcrHub ensemble, only PaCRISPR detected one putative anti-CRISPR protein, which overall does not indicate the presence of functional Arc proteins. Together, these genomic features indicate that GE1 is a strictly lytic Kayvirus lacking known virulence and resistance genes, which is desirable for therapeutic applications.
## 3.5. Antibacterial and Biofilm-Degrading Ability of the vB_SaS_GE1 Phage
Having confirmed that GE1 is a genomically safe, strictly lytic phage with broad activity against S. aureus, the antibacterial and anti-biofilm effects in single and mixed cultures of phage GE1 were evaluated. The activity of GE1 in combination with the PNG3 phage was studied in an 18 h time-kill assay on a single and mixed biofilms. Two of the six MRSA strains (S. aureus 9 and S. aureus 8497) showed no susceptibility to the GE1 phage, whether grown alone or in combination with P. aeruginosa 1147. The remaining MRSA strains showed suppressed growth in both single and mixed cultures. After 11 to 12 h, an increase in absorbance was observed, indicating renewed bacterial growth (Figure 4). A viable cell count assay on selective media demonstrated that this growth increase was due to the replication of PNG3-resistant Pseudomonas mutant forms. Overall, the time-kill assays showed that GE1 effectively suppressed the growth of GE1-susceptible MRSA strains in both single and mixed cultures, whereas the emergence of PNG3-resistant P. aeruginosa variants limited long-term control of the Pseudomonas population. Based on the spot test and time-kill assay results, two MRSA strains were selected to evaluate the stability of GE1's lytic activity over time: one fully susceptible strain (S. aureus 643) and one weakly susceptible strain (S. aureus 8497). S. aureus 643 showed no growth for 24 h at all tested MOIs. In contrast, S. aureus 8497's growth was suppressed only at an MOI of 1 and only during the first 6 h.
These two MRSA strains were then combined with the biofilm-producing strain P. aeroginosa 1147. Viable cell counts and biofilm OD were measured to determine the antibiofilm activity of phage GE1. In mixed biofilms formed by S. aureus 8497 and P. aeruginosa 1147, OD values did not change after phage treatment, but the viable counts of both species decreased by approximately 1 log (Figures 5 and6). No changes in OD or viable counts were observed when S. aureus 8497 was grown alone, confirming its resistance to GE1. In contrast, when a mixed biofilm formed by S. aureus 643 and P. aeruginosa was treated with the phage solution, no viable S. aureus cells were detected, and P. aeruginosa counts decreased by 1 log. Notably, phage treatment increased biofilm formation in this experimental group. For biofilms formed solely by S. aureus GE1 treatment did not alter OD, but the number of viable cells still decreased by 1 log compared with the control. In mixed biofilms formed by S. aureus 643 and Pseudomonas, biofilm density did not differ between treated and untreated groups, while viable counts of both S. aureus and P. aeruginosa were significantly reduced (Figures 5 and6).
In summary, phage treatment frequently reduced viable counts of MRSA and P. aeruginosa within immature biofilms, even when total biofilm biomass remained unchanged or increased, highlighting a dissociation between matrix production and cell viability under phage pressure.
## 4. Discussion
Bacteriophage therapy is emerging as a promising strategy against multidrug-resistant and biofilm-associated infections, where conventional antibiotics often fail. In this context, strictly lytic, genomically safe phages with broad activity against Staphylococcus aureus, including MRSA, and documented effects in mixed-species biofilms are of particular interest. This study positions GE1 as such a candidate, combining a favorable genomic profile with activity against MRSA in dual-species biofilms with Pseudomonas aeruginosa.
Lytic phages that lack virulence, lysogeny and antimicrobial resistance genes are widely regarded as the most suitable for therapeutic development [28,29]. GE1 fulfills these criteria and, in addition, is closely related to clinically explored phages Sb-1 and ISP at the genomic level [5,30], supporting its inclusion in the growing group of S. aureus phages with translational potential. The close relationship to Sb-1, together with shared overall genome organization and high nucleotide identity, suggests that GE1 may have comparable functional properties while still representing a distinct phage that can expand the available therapeutic repertoire.
An important implication of this work is that GE1 retains robust lytic activity against MRSA not only in planktonic cultures but also within early-stage mixed biofilms. The consistent reduction in viable MRSA counts, even when the total biofilm biomass remained unchanged or increased, underscores a mechanistic dissociation between matrix production and cell viability under phage pressure. This pattern aligns with the idea that phages can select for biofilm phenotypes while still effectively killing embedded cells, which has direct relevance for interpreting outcomes in biofilm-associated infections.
The observations in mixed cultures with P. aeruginosa highlight the complexity of polymicrobial infections. The emergence of PNG3-resistant P. aeruginosa variants, together with the modest reduction in a GE1-resistant S. aureus strain in mixed biofilms, suggests that interspecies interactions can modulate phage impact indirectly. The known ability of P. aeruginosa to alter S. aureus physiology and membrane composition may partially explain the increased vulnerability of otherwise phage-or drug-tolerant S. aureus subpopulations, pointing to a potentially exploitable synergy between community effects and targeted phage therapy [31].
The apparent increase in biofilm biomass after GE1 treatment, particularly with S. aureus 643, provides another mechanistic insight. Genomic predictions of peptidoglycan hydrolases and tail-associated proteins with potential depolymerase activity are consistent with phage-mediated remodeling of the biofilm environment, yet the net effect observed here was enhanced matrix accumulation alongside reduced viable counts. This supports the concept that, at least in immature biofilms, phage exposure can trigger a stress response in S. aureus that favors matrix overgrowth as a defense strategy, without fully preventing phage-mediated killing.
Several limitations of this work should be acknowledged. All experiments were performed in vitro using immature (24 h) biofilms and a limited number of MRSA isolates and a single P. aeruginosa strain, which may not fully capture the diversity and complexity of clinical infections. In addition, the study did not include in vivo efficacy or safety assessments, nor did it systematically evaluate combinations of GE1 with antibiotics commonly used to treat MRSA infections.
Despite these constraints, the data presented here demonstrate that GE1 is a strictly lytic Kayvirus with a favorable genomic safety profile, broad activity against S. aureus including MRSA and demonstrable efficacy in reducing viable bacterial counts in singleand dual-species biofilms. These properties support its potential as a component of phagebased strategies targeting biofilm-associated S. aureus infections, especially in polymicrobial contexts involving P. aeruginosa. Future work should extend these findings to mature biofilms and in vivo models, explore GE1 in combination with antibiotics and other phages and further dissect how interspecies interactions and phage-encoded enzymes shape biofilm architecture and treatment outcomes.
## References
1. Naghavi, Vollset, Ikuta et al. "Global burden of bacterial antimicrobial resistance 1990-2021: A systematic analysis with forecasts to 2050" *Lancet*
2. Yung, Sircombe, Pletzer (2021) "Friends or enemies? The complicated relationship between Pseudomonas aeruginosa and Staphylococcus aureus" *Mol. Microbiol*
3. Mobarezi, Esfandiari, Abolbashari et al. (2025) "Efficacy of phage therapy in controlling staphylococcal biofilms: A systematic review" *Eur. J. Med. Res*
4. Pirnay, Djebara, Steurs et al. "Personalized bacteriophage therapy outcomes for 100 consecutive cases: A multicentre, multinational, retrospective observational study" *Nat. Microbiol*
5. Wang, Tkhilaishvili, Trampuz et al. (2020) "Evaluation of staphylococcal bacteriophage Sb-1 as an adjunctive agent to antibiotics against rifampin-resistant Staphylococcus aureus biofilms" *Front. Microbiol*
6. Verheul, Mulder, Van Dun et al. "Bacteriophage ISP eliminates Staphylococcus aureus in planktonic phase"
7. Hudzicki, Kirby (2009) "Bauer disk diffusion susceptibility test protocol" *Am. Soc. Microbiol*
8. Lee (2003) "Methicillin (oxacillin)-resistant Staphylococcus aureus strains isolated from major food animals and their potential transmission to humans" *Appl. Environ. Microbiol*
9. Vandersteegen, Kropinski, Nash et al. (2013) "Romulus and Remus, two phage isolates representing a distinct clade within the Twortlikevirus genus, display suitable properties for phage therapy applications" *J. Virol*
10. Bakuradze, Merabishvili, Kusradze et al. (1042) "Characterization of a bacteriophage GEC_vB_Bfr_UZM3 active against Bacteroides fragilis" *Viruses*
11. Viazis, Akhtar, Feirtag et al. (2011) "Isolation and characterization of lytic bacteriophages against enterohaemorrhagic Escherichia coli" *J. Appl. Microbiol*
12. Khan Mirzaei, Nilsson (2015) "Isolation of phages for phage therapy: A comparison of spot tests and efficiency of plating analyses for determination of host range and efficacy" *PLoS ONE*
13. Glonti, Pirnay (1490) "In Vitro Techniques and Measurements of Phage Characteristics That Are Important for Phage Therapy Success" *Viruses*
14. Jakoči Ūn Ė, Moodley (2018) "A rapid bacteriophage DNA extraction method" *Methods Protoc*
15. Almeida, Ravantti, Grdzelishvili et al. "Relevance of the bacteriophage adherence to mucus model for Pseudomonas aeruginosa phages"
16. Nextera, Prep (2015) "Tips and Troubleshooting. Illumina"
17. Couvin, Bernheim, Toffano-Nioche et al. (2018) "an update of CRISRFinder, includes a portable version, enhanced performance and integrates search for Cas proteins" *Nucleic Acids Res*
18. Lade, Park, Chung et al. (1853) "Biofilm Formation by Staphylococcus aureus Clinical Isolates is Differentially Affected by Glucose and Sodium Chloride Supplemented Culture Media" *J. Clin. Med*
19. O'toole (2011) "Microtiter Bulaşık Biyofilm Oluşumu Testi" *J. Vis. Exp. (JoVE)*
20. Pires, Sillankorva, Faustino et al. (2011) "Use of newly isolated phages for control of Pseudomonas aeruginosa PAO1 and ATCC 10145 biofilms" *Res. Microbiol*
21. Kifelew, Warner, Morales et al. (2020) "Efficacy of lytic phage cocktails on Staphylococcus aureus and Pseudomonas aeruginosa in mixed-species planktonic cultures and biofilms" *Viruses*
22. Mendes, Leandro, Corte-Real et al. (2013) "Wound healing potential of topical bacteriophage therapy on diabetic cutaneous wounds" *Wound Repair Regen*
23. Trizna, Yarullina, Baidamshina et al. (2020) "Bidirectional alterations in antibiotics susceptibility in Staphylococcus aureus-Pseudomonas aeruginosa dual-species biofilm" *Sci. Rep*
24. Kebriaei, Lehman, Shah et al. (2023) "Optimization of Phage-Antibiotic Combinations against Staphylococcus aureus" *Biofilms. Microbiol. Spectr*
25. Barber, Shammout, Smith et al. (2021) "Biofilm time-kill curves to assess the bactericidal activity of daptomycin combinations against biofilm-producing vancomycin-resistant Enterococcus faecium and faecalis" *Antibiotics*
26. Kvachadze, Balarjishvili, Meskhi et al. (2011) "Evaluation of lytic activity of staphylococcal bacteriophage Sb-1 against freshly isolated clinical pathogens" *Microb Biotechnol*
27. Vandersteegen, Mattheus, Ceyssens et al. (2011) "Microbiological and molecular assessment of bacteriophage ISP for the control of Staphylococcus aureus" *PLoS ONE*
28. Mattila, Ruotsalainen, Jalasvuori (1271) "On-demand isolation of bacteriophages against drug-resistant bacteria for personalized phage therapy" *Front. Microbiol*
29. Latz, Wahida, Arif et al. (2016) "Preliminary survey of local bacteriophages with lytic activity against multi-drug resistant bacteria" *J. Basic Microbiol*
30. Merabishvili, Vervaet, Pirnay et al. (2013) "Stability of Staphylococcus aureus phage ISP after freeze-drying (lyophilization)" *PLoS ONE*
31. Orazi, Ruoff, O'toole (2019) "Pseudomonas aeruginosa Increases the Sensitivity of Biofilm-Grown Staphylococcus aureus to Membrane-Targeting Antiseptics and Antibiotics"
32. "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=PMC12671175&blobtype=pdf | # Adult K18-hACE2 mice are suitable for studying intranasal SARS-CoV-2 infection but not direct-contact transmission
Jiseon Kim, Sung-Hee Kim, Donghun Jeon, Haengdueng Jeong, Chanyang Uhm, Heeju Oh, Kyungrae Cho, Yejin Cho, Sumin Hur, In Park, Jooyeon Oh, Jeong Kim, Jun-Young Seo, Jeon-Soo Shin, Je Seong, Ki Taek
## Abstract
The rapid global spread of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) since 2019 emphasizes the need to understand its transmission routes, which mainly comprise airborne and contact transmission. Contact transmission, where the virus spreads through direct or indirect contact, is key to the disease epidemiology. Therefore, investigating contact transmission in animal models is crucial for understand ing SARS-CoV-2 behavior and developing effective preventive measures. Although ferrets, cats, and hamsters have been established as models for studying contact transmission, the susceptibility of mice (the most commonly used experimental animal model) to SARS-CoV-2 contact infection remains uncertain. In this study, we investigated whether SARS-CoV-2 can spread via contact transmission in adult K18-hACE2 mice with different genetic backgrounds, including those with mitomycin C-induced immunodefi ciency. We conducted contact-transmission experiments by co-housing K18-hACE2 mice intranasally infected with SARS-CoV-2 S type (isolated in Korea) alongside uninfected adult K18-hACE2 mice. Mice with genetically different backgrounds subjected to contact infection exhibited no changes in clinical signs or histopathological changes in the respiratory tract and extrapulmonary organs. Additionally, neither SARS-CoV-2 nor neutralizing antibodies were detected in any of the tested samples. Their immune responses remained unchanged, and contact transmission was not observed, even in immunodeficient mice. Collectively, these findings suggest that adult K18-hACE2 mice are not susceptible to contact infection with SARS-CoV-2, highlighting the role of immune mechanisms in viral spread and the limitations of this model for study ing human transmission pathways. Our results underscore the importance of utilizing appropriate animal models to accurately elucidate transmission dynamics. IMPORTANCE Understanding the mechanisms of severe acute respiratory syndrome coronavirus-2 infection and transmission is essential for preventing and treating coronavirus disease 2019. Varying opinions exist regarding the occurrence of contact infection in mice. Here, we aimed to induce contact infection under various conditions in K18-hACE2 mice. By measuring clinical symptoms, viral loads, and neutralizing-antibody titers and conducting pathological analyses, we demonstrated that contact infection did not occur in K18-hACE2 mice. These findings underscore the importance of selecting appropriate experimental animal models to guide future studies on viral infections. KEYWORDS SARS-CoV-2, contact transmission, hCoV-19/Korea/KCDC03/2020_WA-1, K18-hACE2 mice model, histopathological change, mouse background, immune response S evere acute respiratory syndrome coronavirus-2 (SARS-CoV-2) has rapidly spread worldwide since its initial emergence in 2019, resulting in a pandemic. The rapid spread of SARS-CoV-2 has underscored the importance of understanding its primary
transmission routes, which predominantly include airborne dissemination and direct contact (1)(2)(3)(4)(5)(6). The canonical receptor for SARS-CoV-2, known as angiotensin-convert ing enzyme 2 (ACE2), is expressed not only in respiratory organs (such as the oral and nasal mucosa and lungs) but also in digestive organs (like the stomach, small intestine, and colon). ACE2 is also expressed in immune system organs (like the lymph nodes, thymus, bone marrow, and spleen) and in the brain, thus enabling systemic infection (7,8). In addition to ACE2, alternative receptors might bind with SARS-CoV-2, including neuropilin-1, CD147, CD209L, asialoglycoprotein receptor 1, kringle contain ing transmembrane protein 1, and glucose-regulated protein 78 (9)(10)(11)(12)(13). In humans, SARS-CoV-2 detection in feces or urine has highlighted the potential for indirect transmission, augmenting concerns about escalating infection rates (14)(15)(16).
Animal models are invaluable tools for exploring complex transmission dynamics (17). Representative contact-transmission models include ferrets, hamsters, and cats (18). In ferrets, infection has been confirmed to spread through direct contact (19). Furthermore, some data have shown that viruses can be detected in the saliva, urine, and feces of infected ferrets, suggesting that noncontact transmission can also occur (20). In addition, direct-contact infection has been reported in cats co-housed with other SARS-CoV-2infected cats, where the virus was detected in the nasal and rectal swabs of the cohoused cats (21,22). Previously, hamsters showed induction of upper respiratory-tract lesions similar to those in humans and demonstrated SARS-CoV-2 contact transmission (23)(24)(25). In those studies, the virus was detected in the skin and feces of the infected hamsters, and clinical signs of infection, such as weight loss, were observed. Additionally, transmission occurred when uninfected hamsters were housed with hamsters infected with SARS-CoV-2, and those animals subsequently died (26).
As one of the most widely used experimental animal models, mice are frequently employed in studies of respiratory infections and disease pathogenesis. Techniques such as oral or direct intranasal administration and inhalation are commonly used to introduce pathogens, enabling controlled infection and the examination of disease progression (27,28). Numerous studies on SARS-CoV-2 have also utilized mice to investigate viral-infection mechanisms, immune responses, and potential therapeutic strategies, but SARS-CoV-2 signs do not develop in wild-type mice (29)(30)(31). Although some data have shown that infection can occur in severe combined immunodeficiency (SCID) mice infected with a beta variant of SARS-CoV-2, the mice did not exhibit signs of viral replication in their lungs after infection, and their lung lesions gradually improved over time (32). Therefore, experiments have been conducted using genetically modified mice engineered to express receptors capable of binding to SARS-CoV-2 or the receptorbinding domain of the SARS-CoV-2 spike (S) protein variant modified to bind mouse ACE2 with high affinity, thus rendering mice highly susceptible to SARS-CoV-2 infection (16,(33)(34)(35)(36)(37). In genetically modified mice, lesions were observed in the lungs and other organs, similar to those seen in humans (15,38,39).
While direct infection through the diet or nasal routes has been confirmed in genetically modified mice, contact transmission has been debated (37,(40)(41)(42). One research group reported rare viral detection in mice after several days of contact with infected human angiotensin-converting enzyme 2 (hACE2)-transgenic mice, although viral detection in the lungs (the main target organ) was not confirmed (41). Moreover, in keratin 18 (K18) promoter-derived, hACE2-transgenic mice, researchers observed that contact infection occurred readily during the neonatal stage. However, in adult mice, the virus was detected in only one mouse, with no virus detected in the others (43). In another study, the researchers introduced the N501Y mutation into the receptor-binding domain of SARS-CoV-2 and used it to induce SARS-CoV-2 contact infection in wild-type mice. They reported that contact infection occurred only with the beta variant (B.1.351), which was isolated from a South African traveler (37,44). However, another report demonstrated that when the same viral strain was used to study contact infection in wild-type mice, no transmission occurred (42). These results highlight the uncertainty surrounding the induction of contact infection in mice.
In this study, we aimed to evaluate SARS-CoV-2 contact infection in adult K18-hACE2 mice with two different backgrounds. We used K18-hACE2 mice with the commonly used C57BL/6 background and K18-hACE2 mice with the FVB/NJ background (developed in Korea) to assess contact transmission (45). We housed naïve K18-hACE2 mice with K18-hACE2 mice infected with S-type SARS-CoV-2 (isolated in Korea) at both the early and late stages of infection. We conducted a comprehensive analysis on the mice subjected to contact transmission, including clinical-symptom evaluation, pathologi cal analysis, and virus detection in respiratory organs. Our pathological analysis was extended to non-respiratory organs to accurately assess whether contact transmission occurred. Lastly, we studied contact infection in K18-hACE2 mice with mitomycin C (MMC)-induced immunodeficiency, which confirmed the necessity of immune mecha nisms for contact infection.
## RESULTS
## Clinical features and pathogenesis after intranasal SARS-CoV-2 infection in K18-hACE2 mice
We used K18-hACE2 mice to establish an animal model of SARS-CoV-2 infection. In situ hybridization confirmed that hACE2 was expressed in the alveoli, bronchi, and vessels of K18-hACE2 mouse lungs and in the trachea (Fig. S1A through C). To assess clinical signs and lung pathogenesis, SARS-CoV-2 was intranasally administered to K18-hACE2 mice, and observations were made at different days post-infection (dpi). The body temperature and weights of the SARS-CoV-2-infected K18-hACE2 mice decreased after 2 dpi. By 7 dpi, the infected mice exhibited significant reductions, with a >10°C decrease in body temperature and a 20% decrease in body weight. In contrast, mock-infected K18-hACE2 mice did not show any changes (Fig. 1A andB). SARS-CoV-2-infected mice began dying at 5 dpi, with only 20% surviving at 7 dpi (Fig. 1C), as reported previously (46,47). Next, we compared the compositions of white blood cells in the peripheral blood. The neutrophil:lymphocyte ratio (a diagnostic and prognostic marker of clinical outcomes) increased significantly, depending on the days post-infection (Fig. 1D) (48,49).
Plaque assays were performed to compare viral titers in lungs from SARS-CoV-2infected mice at multiple time points with those from mock-infected mice. SARS-CoV-2 PFUs were highest at 2 dpi and decreased by 7 dpi (Fig. 1E; Fig. S2A). Pathological differences occurred in the SARS-CoV-2-infected mice over time. In the lungs at 2 dpi, immune cells infiltrated the alveolar region through blood vessels, and vascular edema and capillary dilation were confirmed. At 7 dpi, the lesions worsened, and the pathologi cal scores increased significantly (Fig. 1F andG). We performed immunohistochemical (IHC) staining for the SARS-CoV-2 nucleocapsid (N) protein to verify viral distributions in the lungs of SARS-CoV-2-infected mice. Over 70% of the lung areas were positive for the N protein at 2 dpi, but by 7 dpi, the percentage of infected areas had significantly decreased (Fig. 1H andI), which correlated with the PFU data for the lungs. The absence of S gene detection in the lungs and trachea of the recipient group through in situ hybridization further confirmed that contact infection did not occur in the respiratory system of adult K18-hACE2 mice (Fig. S3A andB).
We also performed a pathological analysis of extrapulmonary organs. In the spleen, white blood cell apoptosis (associated with the occurrence of a cytokine storm) was observed at 2 dpi, with extensive damage at 7 dpi (Fig. S4A andB) (38). In the small intestine, goblet cell hyperplasia, villous atrophy, and villous necrosis were only observed at 7 dpi (Fig. S4D andE). Similarly, multifocal perivascular cuffing in the brain was only detected at 7 dpi (Fig. S4G andH) (15,50). Consistent with previous research findings, our data confirmed that K18-hACE2 mice intranasally infected with SARS-CoV-2 exhibited pathological signs in their pulmonary and extrapulmonary organs, leading to systemic infection and eventual mortality (35).
## K18-hACE2 mice were not infected with SARS-CoV-2 via contact transmission
To investigate the occurrence of SARS-CoV-2 contact transmission in mice (which can occur in humans), K18-hACE2 mice were divided into recipient groups that were co-housed with SARS-CoV-2 intranasally infected donor K18-hACE2 mice at 2 dpi (2DC) or 6 dpi (6DC) for 48 h (Fig. 2A). A 48 h co-housing period was implemented to evaluate SARS-CoV-2 transmissibility at two distinct donor infection stages: an early phase, marked by localized infection in the respiratory tract and oral cavity, and a late phase with systemic viral dissemination. Our aim was to determine whether contact transmission could be initiated during early localized viral shedding vs more advanced systemic infection. First, clinical signs were confirmed; neither the 2DC-nor 6DC-group mice showed significant changes in body temperature or weight (Fig. 2B andC). These mice subjected to contact transmission did not die during the experiment (Fig. 2D). Analysis of the composition of white blood cells in the peripheral blood revealed no changes in monocytes, eosinophils, basophils, or the neutrophil: lymphocyte ratio (Fig. 2E). In addition, neutralizing antibodies were not detected in serum samples from the co-housed mice (Fig. 2F). These data indicate that mice in contact with SARS-CoV-2-infec ted mice did not show any clinical signs.
Subsequently, we performed histopathological analyses of both the pulmonary and extrapulmonary organs. No histopathological changes were observed in the lungs of mice in the 2DC and 6DC groups (Fig. 3A through C), and PFU measurements with isolated lungs did not detect SARS-CoV-2 (Fig. 3D). Likewise, the N protein was not expressed in the nasal conchae (the initial route of respiratory infections) in the 2DC and 6DC groups at any time point (Fig. 3E through J). To enable quantitative assessment, RT-qPCR was performed on RNA extracted from lung tissues of both donor and contact groups. SARS-CoV-2 RNA targeting the surface glycoprotein gene was detected in donor mice with Ct values below 25, confirming active infection. In contrast, all contact and negative control mice exhibited Ct values of 38 or higher, or undetectable levels, indicating the absence of viral RNA (Fig. S5A).
A pathological analysis of the extrapulmonary organs, including the spleen, small intestine, and brain, was also conducted. Unlike the pathological changes observed in the donor group, no lesions were found in any of the extrapulmonary organs of the recipient group (Fig. S4A through I).
Taken together, our data revealed the absence of clinical and pathological changes in the pulmonary and extrapulmonary organs of K18-hACE2 mice in contact with SARS-CoV-2-infected mice.
## Immune responses did not change in mice subjected to potential SARS-CoV-2 contact transmission
Following SARS-CoV-2 infection, a dynamic immune response occurs in the lungs, including the activation of innate immunity through neutrophils and the subsequent activation of adaptive immunity (51,52). Therefore, IHC staining was performed to confirm the changes in each immune cell type in the lungs of both the donor and recipient groups. Initially, we focused on key innate immune cells, specifically macro phages and neutrophils. In the donor group, the abundances of F4/80-positive macro phages and Ly6-G/Ly6-C-positive neutrophils increased significantly by 8% and 5%, respectively, when compared with the corresponding abundances in mock-infected mice (Fig. 4A, C, D andF). By 7 dpi, the numbers of macrophages and neutrophils had increased by 22% and 16%, respectively (Fig. 4A, C, D andF). However, the numbers of macrophages and neutrophils did not differ significantly between the donor and negative-control group at any day post-contact (Fig. 4B, C, E andF).
In terms of the adaptive immune system, donor mice at 2 dpi showed a 3% increase in CD3b-positive T cells (Fig. 4G andI), whereas the number of PTPRC-positive B cells did not increase (Fig. 4J andL). The numbers of T and B cells were 14% and 6% higher, respectively, in infected mice at 7 dpi than in the negative-control mice (Fig. 4G, I, J andL). In contrast, recipient mice showed no changes in T and B cells in either the 2DC or 6DC group (Fig. 4H, I, K andL), similar to innate immune response data. Our data represent immunological changes that occurred in the lungs of donor and recipient mice. In donor mice, the innate immune system was activated at an early stage, and this activation had increased further at 7 dpi. The adaptive immune system was also activa ted at 7 dpi, with slightly higher activation found at 2 dpi. In recipient mice, no significant changes were observed in any immune cell population, similar to observations made with control mice. To further evaluate the innate immune response, expression levels of Cxcl10, Ifnb1, and Il6 were quantified. These genes showed significant upregulation in donor mice, whereas expression in contact mice remained comparable to uninfected controls (Fig. S6A through C). Overall, considering the pathological data, our findings indicate that contact with SARS-CoV-2-infected K18-hACE2 mice did not lead to SARS-CoV-2 infection.
## Contact infection does not occur even under conditions of immunodeficiency
Previous findings confirmed that infections in immunodeficient mice were more contagious and lasted longer (32,54). The data led us to hypothesize that the absence of contact infection in previous experiments might have been related to immunity. To test this hypothesis, immunodeficiency was induced in K18-hACE2 mice with an FVB background through drug treatment, and the occurrence of contact transmission was measured (55,56). The mice were administered MMC to induce immunodeficiency 3 days before contact infection, and a lower MMC concentration was administered following the contact period (Fig. 5A). Fewer white blood cells were observed in the peripheral blood of the MMC-treated mice (Fig. S7A). However, after contact infection, the body weights or temperatures were not different from those in the control group (Fig. S7B andC). Similar to C57BL/6-background K18-hACE2 mice, K18-hACE2 mice with an FVB background died 4 dpi following intranasal SARS-CoV-2 infection, with only one mouse surviving by 6 dpi. However, none of the mice, whether MMC-treated or not, died following contact infection (Fig. S7D). Neutralizing antibodies were not detected in the serum of the MMCtreated recipient group. In contrast to the FVB-background K18-hACE2 donor mice (which showed detectable viral PFUs in their lung tissues), no viral PFUs and SARS-CoV-2 RNA targeting the surface glycoprotein gene were detected in the MMC-treated recipient group (Fig. 5B andC; Fig. S5B). Pathological analysis of the lungs revealed no lesions in the group treated with MMC alone (Fig. S7E). Additionally, the donor group exhibited severe pneumonia, vascular edema, and pulmonary capillary dilatation, similar to C57BL/6-background K18-hACE2 donor mice. However, in the recipient groups, no lung lesions were observed at 2-or 7-days post-contact (dpc) in either the MMC-treated or untreated groups. (Fig. 5D andE). SARS-CoV-2 was not detected in the lungs of the MMCtreated recipient group at 2 and 7 dpc using in situ hybridization staining for the S gene (Fig. 5F; Fig. S7F). Similarly, the N protein was not detected in the nasal concha, the initial respiratory route (Fig. S7G). The expression levels of Cxcl10, Ifnb1, and Il6 in lung tissues did not differ between the MMC-treated recipient group and uninfected controls (Fig. S6D through F). These data suggest that contact infection was not induced, even under conditions of heightened susceptibility to infection due to immunodeficiency.
## DISCUSSION
In humans, SARS-CoV-2 infection is caused by airborne transmission or direct contact with an infected individual (14,41), and animal experiments are being conducted to model these infection routes (18). Among animals with confirmed contact infections, ferrets, cats, and hamsters can be infected through direct contact, whereas mice exhibit various responses to contact infection after SARS-CoV-2 infection. In a study by Rodri guez-Rodriguez et al., transmission experiments were conducted using adult K18-hACE2 mice and the WA-1 variant (A.1 lineage) (43). They reported only one case of contact infection. However, no decrease in body weight was observed in the control mice. Similarly, when mice were infected with the beta variant (B. 1.351; which can cause contact infection), no decrease in body weight occurred, and the virus was detected in the trachea but not in the lungs (37). These data indicate the uncertainty regarding contact infections in mouse models.
In this study, we assessed the effectiveness of contact infection in K18-hACE2 mice under various conditions. Before inducing contact infection, we intranasally infected K18-hACE2 mice with SARS-CoV-2 to observe their overall responses to the virus. The responses were similar to those of previous studies, confirming that these mice could serve as effective donors for contact-infection experiments (35,45).
We used the results from 2 dpi, where high viral titers were observed in the respira tory organs, but no severe lesions appeared in other organs or the lungs. We also used the results from 7 dpi, when the mice exhibited weight loss, reduced body temperature, and death, showing severe lesions in both respiratory and non-respiratory organs. Based on these observations, we hypothesized that contact during the early stages of infection might differ from contact during the later stages. To test this hypothesis, we induced contact infection in both environments. However, under both donor conditions, no symptoms of infection were observed in the co-housed mice. These data suggest that contact infection was not induced in K18-hACE2 mice, regardless of the timing of the infection.
We also induced contact infection using FVB-background K18-hACE2 mice in addition to C57BL/6-background mice. Previous reports have indicated that FVB-background mice exhibit varying levels of cytokine secretion in the context of infection or dis ease development compared to C57BL/6-background mice. This difference in cytokine response can lead to variations in susceptibility to diseases between these mouse strains (45,(57)(58)(59).
Finally, given prior research suggesting that the susceptibility to SARS-CoV-2 infection can increase under immunodeficient conditions, we induced immunodeficiency by administering mice MMC intraperitoneally (32). However, contact infection still did not occur after inducing immunodeficiency. These findings suggest that K18-hACE2 transgenic mice are insufficient as a model for studying transmission through contact infection, regardless of their background or immune status.
We attempted to induce contact infection in K18-hACE2 mice under various conditions, but no signs of infection were observed, strengthening the credibility of the claim that contact infection does not occur in K18-hACE2 mice. However, a limitation of our study is that we did not identify the specific transmission pathways that were impaired to explain why contact infection failed in K18-hACE2 mice. We focused solely on assessing the contact transmissibility of the Beta variant of SARS-CoV-2, without evaluating the contact infection potentials of other variants. Previous reports indicate that responses to infection can vary depending on the viral strain, and some evidence suggests that transmission may not occur with certain strains (37,39,44). However, considering the continuously evolving nature of SARS-CoV-2, the possibility of contact transmission in mice with other emerging variants should not be ruled out. Further studies are necessary to explore this potential. Previous reports have demonstrated the presence of detectable virus in the oral cavities or feces of SARS-CoV-2-infected mice. In this study, we did not measure SARS-CoV-2 titers in the nasal or oral mucosal areas of donor and recipient mice during contact infection. However, we confirmed that viral RNAs and proteins were expressed in the nasal concha following infection through in situ analysis and immunostaining. Additionally, clinical signs, serum neutralizing-antibody levels, and pathological analysis all suggested that minimal contact transmission occurred. Although previous research has shown that airborne SARS-CoV-2 transmission can lead to infection in K18-hACE2 mice, limited data exist on the viral concentrations necessary to induce infection through contact transmission (53). Also, our study used the K18-hACE2 mouse model, in which hACE2 expression is driven by the human keratin-18 promoter. This promoter induces robust expression in the lower respiratory tract, particularly in alveolar epithelial cells, but only limited expression in the upper airway epithelium, including the nasal and bronchial regions. Consequently, SARS-CoV-2 infection in K18-hACE2 mice primarily targets the lungs, with minimal replication in the upper airways. Supporting this, previous studies have shown that by 3 dpi, viral RNA levels in the lungs of K18-hACE2 mice can reach approximately 10 10 copies/g, whereas levels in the nasal epithelium are up to 10 5 -fold lower (46). Similarly, Winkler et al. reported that infectious viral titers in the nasal turbinate were approximately 100-fold lower than those in the lungs at 2 dpi (52). These findings indicate that although SARS-CoV-2 can reach the upper airways in K18-hACE2 mice, the replication levels there are insufficient to sustain efficient contact transmission. The rapid disease progression and high mortality in K18-hACE2 mice may further limit the window of viral shedding, reducing transmission opportunities. In such highly susceptible hosts, the disease course is skewed toward severe lower respiratory tract involvement rather than prolonged upper respiratory tract infection, which is typically necessary for efficient spread.
In contrast, the transgenic hACE2 mouse model used by Bao et al. was generated using the endogenous mouse ACE2 promoter, which drives hACE2 expression in a pattern more consistent with native ACE2 distribution, including robust expression in the nasal and airway epithelium (41). This difference in tissue tropism likely facilitated higher viral replication in the upper respiratory tract, prolonged viral shedding, and consequently, successful contact transmission.
Taken together, our findings support the use of adult K18-hACE2 mice for studying intranasal SARS-CoV-2 infection, while also highlighting their limitations as a model for direct-contact transmission. This limitation likely reflects differences in hACE2 transgene regulation and expression patterns compared with other hACE2 models, which can profoundly influence both disease pathogenesis and transmission potential. Improved mouse models are needed to better study alternative transmission routes, such as contact transmission. Further research is required to explore the underlying factors contributing to the resistance of K18-hACE2 mice to contact infection. These insights are critical for refining animal models and deepening the understanding of SARS-CoV-2 transmission, which will ultimately aid in developing more effective prevention and control measures during the ongoing pandemic.
## MATERIALS AND METHODS
## Animals
## Virus
For virus production, we obtained the Vero African green monkey kidney cell line (KCLB 10081) from the Korean Cell Line Bank and cultured it in Dulbecco's minimum Eagle's medium supplemented with 2 mM L-glutamine, 100 units/mL penicillin, 100 µg/mL streptomycin, and 5% fetal bovine serum. The cells were maintained at 37°C in a humidified incubator with 5% CO 2 . SARS-CoV-2 was obtained from the National Culture Collection for Pathogens of Osong, Korea (NCCP 43326, S-type variant). Virus titers were measured through plaque assays, as previously described (60). Briefly, supernatants from infected cells or homogenates from infected tissue samples were serially diluted. The supernatants were added to Vero cells that had been seeded in a six-well plate and incubated at 37°C for 1 h with gentle agitation every 15 min. The cells were then overlaid with Dulbecco's minimum Eagle's medium, 1% SeaPlaque agarose (Lonza), 2% fetal bovine serum, 100 units/mL penicillin, and 100 mg/mL streptomycin. After 3 days of incubation, the overlays were removed, and the cells were fixed with 4% paraformal dehyde and stained with a 0.5% crystal violet in a 20% methanol solution. The plaques were counted and multiplied by the dilution factor to quantify the viral titers.
## Contact-transmission model
To study SARS-CoV-2 contact transmission in mice, 12 9-week-old male K18-hACE2 mice were placed in six cages, with two mice in each cage. The mice were anesthetized with a Zoletil-Rompun mixture (4:1) and intranasally inoculated with 1 × 10 5 PFUs of SARS-CoV-2 to generate a donor group. In the recipient group, uninfected K18-hACE2 mice were co-housed with donor mice (2:4 ratio) for 48 h and then separated from the donor mice. The mice were divided into the following three groups (12 mice/group): a control group and recipient mice, co-housed with infected intranasally donor mice at 2 dpi (2DC group) or 6 dpi (6DC group). For the mock (control) infection, the donor group was administered an equal volume of phosphatebuffered saline (PBS). Mouse body weights and temperatures were monitored daily using an electronic scale and an implantable programmable temperature transponder (IP55-300; Bio Medic Data Systems, USA). The mice were euthanized 2, 7, and 14 dpc for further analysis (4 mice/time point).
## MMC treatment
Immunosuppression was induced in FVB/NJ-background K18-hACE2 mice via intraperito neal daily injections of MMC (0.02 mg/mouse), starting 3 days before infection, followed by 0.001 mg/mouse for 3 days after infection.
## Histopathological analysis
For histopathological analysis, we fixed lung, spleen, small intestine, and brain tissues in 10% neutralbuffered formalin (F5554, Sigma, St. Louis, MO, USA) for 24 h and embed ded them in paraffin. The fixed samples were sliced into 4 μm-thick tissue sections for hematoxylin and eosin (H&E) and IHC staining using a microtome (Leica, Wetzlar, Germany). For H&E staining, sections were deparaffinized through immersion three times in xylene and rehydrated sequentially in 100%, 95%, and 70% ethanol. The slides were stained with 0.1% Mayer's hematoxylin (3309, Agilent, Santa Clara, CA, USA) for 10 min and then dipped into a 0.5% Eosin Y (F5554, Sigma) solution. The slides were then washed in distilled water until the eosin stopped forming streaks and dehydrated in an ascending serial gradient of ethanol (50%, 70%, 95%, and 100%) for 1 min/dehydration step. The slides were covered with a mounting solution (6769007, Thermo Scientific, Waltham, MA, USA) and analyzed under a light microscope (BX43, Olympus, Tokyo, Japan). Histopathological analyses were performed by an animal pathologist (K.T.N.).
Histopathological severity was scored on a scale from 0 to 5 (0, none; 1, weak; 2, mild; 3, moderate; 4, severe; 5, markedly severe).
## IHC staining
For IHC staining, paraffinembedded samples were sliced into 4 μm-thick sections. The sections were deparaffinized via immersion thrice in xylene, followed by rehydration using a descending graded series of ethanol. Using pH 6.0 antigen-retrieval solution (S1699, Agilent), antigens were retrieved by pressing and boiling the sections in a high-pressure cooker for 15 min. Subsequently, the sections were cooled on ice for 1 h, washed twice with Dulbecco's PBS, and incubated in 3% H 2 O 2 for 30 min to block endogenous peroxidase activity. The sections were then washed twice with PBS and incubated in a protein-blocking solution (X0909, Agilent) for 2 h at 22°C in a humidi fied chamber. When using mouse or rat primary antibodies, the M.O.M. Kit (BMK-2202, Vector Laboratories, Newark, CA, USA) was employed before protein blocking. The slides were incubated overnight at 4°C with primary antibodies against the SARS-CoV-2 N protein (40143-MM08, Sino Biological, 1:1,000), F4/80 (70076, Cell Signaling Technol ogy, 1:1,000), CD3b (ab5690, Abcam, 1:1,000), anti-PTPRC (ab64100, Abcam, 1:1,000), and neutrophils (ab2557, Abcam, 1:2,000). The sections were then incubated with a horseradish peroxidase-conjugated secondary antibody (K4001, Agilent) for 15 min or with biotinylated anti-rat IgG (Vector) for 30 min, followed by incubation with the ABC reagent (Vector) for 30 min at room temperature. To develop the horseradish peroxidase-labeled antibodies on the sections, 3,3′-diaminobenzidine in chromogen solution (K3468, Agilent) was diluted in 20 µL to 1 mL of imidazole-HCl buffer, containing hydrogen peroxide (K3468, Agilent) and applied to the sections for 15-30 seconds to detect the signals. Mayer's hematoxylin (3309, Agilent) was used to counterstain the nuclei. After counterstaining, washing and dehydration steps were performed, and the slides were covered with mounting solution (6769007, Thermo Scientific).
## In situ hybridization
For in situ hybridization staining, the human ACE2 RNA probe and RNAscope 2.5 HD Red Assay were purchased from Advanced Cell Diagnostics ([ACD], Bio-Techne, MN, USA). In situ hybridization was performed according to the manufacturer's instructions. Briefly, paraffinembedded sections were deparaffinized in xylene and dehydrated twice with 100% ethanol. After air-drying, the slides were treated with H 2 O 2 , boiled in buffer to retrieve the targets, and treated with protease K for 30 min. The sections were incubated for 2 h with the RNA probe, after which the RNA signals were amplified using the amplifying reagent (322360, ACD) and detected with the Fast Red reagent (322360, ACD).
## Hematological analysis
For hematological analysis, peripheral blood samples were collected from mouse hearts following euthanasia using a 1 mL syringe. The collected blood samples were transfer red into 1.5 mL microtubes containing 20 µL 0.5 M ethylenediaminetetraacetic acid to prevent blood clotting. Complete blood counts were performed using a hematology analyzer (BC-5000, Mindray Global, Nanshan, China).
## Statistical analysis
Statistical analyses were performed using the GraphPad Prism software (v9.0). Statistical significance was determined using one-way analysis of variance with Šidák's multiplecomparison test. All data are presented as the mean ± SD. A value of P < 0.05 was considered to represent a statistically significant difference.
## References
1. "Writing -original draft | Yejin Cho, Conceptualization, Data curation, Investigation, Supervision, Validation, Writing -review and editing | Sumin Hur, Conceptualization, Supervision, Validation, Writing -original draft | In Ho Park, Conceptualization, Data curation, Investigation, Resources, Visualization, Writing -original draft | Jooyeon Oh, Data curation, Investigation, Resources, Visualization, Writing -original draft | Jeong Jin Kim, Data curation, Formal analysis, Resources, Validation, Writing -original draft | Jun-Young Seo, Data curation, Resources, Supervision, Writing -review and editing | Jeon-Soo Shin, Data curation, Resources, Supervision, Writing -review and editing | Je Kyung Seong, Data curation, Supervision, Writing -review and editing | Ki Taek Nam, Conceptualization, Data curation"
2. Chan, Yuan, Kok et al. (2020) "A familial cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family cluster" *Lancet*
3. Dong, Du, Gardner (2020) "An interactive web-based dashboard to track COVID-19 in real time" *Lancet Infect Dis*
4. Ghinai, Mcpherson, Hunter et al. (2020) "First known person-to-person transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in the USA" *Lancet*
5. Manathunga, Abeyagunawardena, Dharmaratne (2023) "A comparison of transmissibility of SARS-CoV-2 variants of concern" *Virol J*
6. Park, Choi, Lee et al. (2023) "Widespread household transmission of SARS-CoV-2 B.1.1.529 (Omicron) variant from children, South Korea, 2022" *Yonsei Med J*
7. Li, Burm, Hong et al. (2021) "A comprehensive review of coronavirus disease 2019: epidemiology, transmission, risk factors, and international responses" *Yonsei Med J*
8. Harmer, Gilbert, Borman et al. (2002) "Quantitative mRNA expression profiling of ACE 2, a novel homologue of angiotensin converting enzyme" *FEBS Lett*
9. 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*
10. Wang, Chen, Zhang et al. (2020) "CD147-spike protein is a novel route for SARS-CoV-2 infection to host cells" *Sig Transduct Target Ther*
11. Amraei, Yin, Napoleon et al. (2021) "CD209L/L-SIGN and CD209/DC-SIGN act as receptors for SARS-CoV-2"
12. Ibrahim, Abdelmalek, Elshahat et al. (2020) "COVID-19 spike-host cell receptor GRP78 binding site prediction" *J Infect*
13. Daly, Simonetti, Klein et al. (2020) "Neuropilin-1 is a host factor for SARS-CoV-2 infection" *Science*
14. Gu, Cao, Zhang et al. (2022) "Receptome profiling identifies KREMEN1 and ASGR1 as alternative functional receptors of SARS-CoV-2" *Cell Res*
15. Jones, Baluja, Graham et al. (2020) "Shedding of SARS-CoV-2 in feces and urine and its potential role in person-to-person transmission and the environment-based spread of COVID-19" *Sci Total Environ*
16. Gallo, Locatello, Mazzoni et al. (2021) "The central role of the nasal microenvironment in the transmission, modulation, and clinical progression of SARS-CoV-2 infection" *Mucosal Immunol*
17. Meyerowitz, Richterman, Gandhi et al. (2021) "Transmission of SARS-CoV-2: a review of viral, host, and environmental factors" *Ann Intern Med*
18. Mullin, Wyk, Asher et al. (2022) "Modeling pandemic to endemic patterns of SARS-CoV-2 transmission using parameters estimated from animal model data"
19. Shou, Liu, Yang et al. (2021) "Animal models for COVID-19: hamsters, mouse, ferret, mink, tree shrew, and non-human primates" *Front Microbiol*
20. Richard, Kok, De Meulder et al. (2020) "SARS-CoV-2 is transmitted via contact and via the air between ferrets" *Nat Commun*
21. Kim, Kim, Kim et al. (2020) "Infection and rapid transmission of SARS-CoV-2 in ferrets" *Cell Host Microbe*
22. Gerhards, Gonzales, Vreman et al. (2023) "Efficient direct and limited environmental transmission of SARS-CoV-2 lineage B.1.22 in domestic cats" *Microbiol Spectr*
23. Barrs, Peiris, Tam et al. (2020) "SARS-CoV-2 in quarantined domestic cats from COVID-19 households or close contacts" *Emerg Infect Dis*
24. Jeong, Lee, Park et al. (2022) "Comparison of the pathogenesis of SARS-CoV-2 infection in K18-hACE2 mouse and Syrian golden hamster models" *Dis Model Mech*
25. Sia, Yan, Chin et al. (2020) "Pathogenesis and transmission of SARS-CoV-2 in golden hamsters" *Nature*
26. Port, Yinda, Owusu et al. (2021) "SARS-CoV-2 disease severity and transmission efficiency is increased for airborne compared to fomite exposure in Syrian hamsters" *Nat Commun*
27. Iwatsuki-Horimoto, Ueki, Ito et al. (2023) "SARS-CoV-2 transmission from virus-infected dead hamsters"
28. Jin, Kim, Roh et al. (2023) "Characterisation of changes in global genes expression in the lung of ICR mice in response to the inflammation and fibrosis induced by polystyrene nanoplastics inhalation" *Toxicol Res*
29. Sekijima, Oshima, Ueji et al. (2023) "Toxicologic pathological mechanism of acute lung injury induced by oral adminis tration of benzalkonium chloride in mice" *Toxicol Res*
30. Dinnon, Iii, Leist et al. (2020) "A mouse-adapted model of SARS-CoV-2 to test COVID-19 countermeas ures" *Nature*
31. Letko, Marzi, Munster (2020) "Functional assessment of cell entry and receptor usage for SARS-CoV-2 and other lineage B betacoronavi ruses" *Nat Microbiol*
32. Shuai, Chan, Yuen et al. (2021) "Emerging SARS-CoV-2 variants expand species tropism to murines" *EBioMedicine*
33. Abdelnabi, Foo, Kaptein et al. (2022) "A SCID mouse model to evaluate the efficacy of antivirals against SARS-CoV-2 infection" *J Virol*
34. Chow, Brodovich, Plumb et al. (1997) "Development of an epitheliumspecific expression cassette with human DNA regulatory elements for transgene expression in lung airways" *Proc Natl Acad Sci*
35. Yang, Deng, Tong et al. (2007) "Mice transgenic for human angiotensin-converting enzyme 2 provide a model for SARS coronavirus infection" *Comp Med*
36. Kim, Kim, Jang et al. (2022) "Mouse models of lungspecific SARS-CoV-2 infection with moderate pathological traits" *Front Immunol*
37. Jiang, Liu, Chen et al. (2020) "Pathogenesis of SARS-CoV-2 in transgenic mice expressing human angiotensin-converting enzyme 2" *Cell*
38. 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*
39. Ping, Zhang, Wang et al. (2021) "Cell death and pathological findings of the spleen in COVID-19 patients" *Pathol Res Pract*
40. Lee, Lee, Hong et al. (2023) "SARS-CoV-2 Omicron variant causes brain infection with lymphoid depletion in a mouse COVID-19 model" *Lab Anim Res*
41. Bao, Deng, Huang et al. (2020) "The pathogenicity of SARS-CoV-2 in hACE2 transgenic mice" *Nature*
42. Bao, Gao, Deng et al. (2020) "Transmission of severe acute respiratory syndrome coronavirus 2 via close contact and respiratory droplets among human angiotensin-converting enzyme 2 mice" *J Infect Dis*
43. Zhang, Cui, Li et al. (2022) "The SARS-CoV-2 B.1.351 variant can transmit in rats but not in mice" *Front Immunol*
44. Rodriguez-Rodriguez, Ciabattoni, Duerr et al. (2023) "A neonatal mouse model characterizes transmissibility of SARS-CoV-2 variants and reveals a role for ORF8" *Nat Commun*
45. Montagutelli, Prot, Levillayer et al. (2021) "Variants with the N501Y mutation extend SARS-CoV-2 host range to mice, with contact transmission" *bioRxiv*
46. Seo, Son, Lee et al. (2022) "Development of transgenic models susceptible and resistant to SARS-CoV-2 infection in FVB background mice" *PLoS One*
47. Oladunni, Park, Pino et al. (2020) "Lethality of SARS-CoV-2 infection in K18 human angiotensin-converting enzyme 2 transgenic mice" *Nat Commun*
48. Mccray, Pewe, Wohlford-Lenane et al. (2007) "Lethal infection of K18-hACE2 mice infected with severe acute respiratory syndrome coronavirus" *J Virol*
49. Prozan, Shusterman, Ablin et al. (2021) "Prognostic value of neutrophil-to-lympho cyte ratio in COVID-19 compared with Influenza and respiratory syncytial virus infection" *Sci Rep*
50. Gelzo, Cacciapuoti, Pinchera et al. (2021) "Prognostic role of neutrophil to lymphocyte ratio in COVID-19 patients: still valid in patients that had started therapy?"
51. Kumari, Rothan, Natekar et al. (2021) "Neuroinvasion and encephalitis following intranasal inoculation of SARS-CoV-2 in K18-hACE2 mice" *Viruses*
52. Zuani, Lazničková, Tomašková et al. (2022) "High CD4-to-CD8 ratio identifies an atrisk population susceptible to lethal COVID-19" *Scand J Immunol*
53. Winkler, Bailey, Kafai et al. (2020) "SARS-CoV-2 infection of human ACE2transgenic mice causes severe lung inflammation and impaired function" *Nat Immunol*
54. Jeon, Kim, Kim et al. (2024) "Discovery of a new long COVID mouse model via systemic histopathological comparison of SARS-CoV-2 intranasal and inhalation infection" *Biochim Biophys Acta Mol Basis Dis*
55. Choi, Park, Choi et al. (2023) "Prevention of severe lung immunopathology associated with influenza infection through adeno-associated virus vector administration" *Lab Anim Res*
56. Jin, Liang, Ning et al. (2012) "Pathogenesis of emerging severe fever with thrombocytopenia syndrome virus in C57/BL6 mouse model" *Proc Natl Acad Sci*
57. Molyneux, Gibson, Ec et al. (2005) "The haemotoxicity of mitomycin in a repeat dose study in the female CD-1 mouse" *Int J Exp Pathol*
58. Chen, Wu, Kao et al. (2012) "Persistent hepatitis B viral replication in a FVB/N mouse model: impact of host and viral factors" *PLoS One*
59. Whitehead, Walker, Berman et al. (2002) "Allergen-induced airway disease is mouse strain dependent" *Am J Physiol Lung Cell Mol Physiol*
60. Azoulay, Brahic, Bureau (1994) "FVB mice transgenic for the H-2Db gene become resistant to persistent infection by Theiler's virus" *J Virol*
61. Couto, Gonçalves, Lamas et al. (2023) "Protocol for infecting and monitoring susceptible k18-hACE2 mice with SARS-CoV-2" *STAR Protoc* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793493&blobtype=pdf | # P-2197. Impact of ventilation on transmission risk and reproduction number of viruses in South India: implications for climate change and pandemic preparedness
Palak Shah, Brady Sack, Abdul Basith, Madolyn Dauphinais, Komal Jain, ; Maria, Florencia Martins, Sierra Wallace, Subitha Lakshminarayanan, Chelsie Cintron, Sadhana Subramanian, Apratim Sahay, Kobto Koura, Ralph Brooks, Sheela Shenoi, Pranay Sinha, Palanivel Chinnakali, Lauren Pischel
## Abstract
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.
Methods. We measured ventilation in air changes per hour in homes and healthcare offices using a carbon-dioxide decay technique. We applied the Wells-Riley equation to estimate the transmission risk and R 0 of SARS-CoV-2 and influenza in these settings, as well as under different ventilation conditions and viral shedding levels.
Results. We conducted 45 ventilation measurements across 13 homes and 7 offices; four of five air conditioning (AC) measurements were in offices. In the closed condition (doors/windows closed, fans off), mean SARS-CoV-2 transmission risk was high for high virus shedders (62.6%, SD 25.2%) and lower for low shedders (28.4%, SD 21.5%) (Figure 1A). Risk decreased significantly in the open condition (doors/windows open, fan on) for both high (29.3%, SD 15.1%; p< 0.001) and low shedders (6.3%, SD 4.4%; p< 0.001). Under AC, transmission risk remained similar to the closed condition for low shedders but was highest for high shedders (74.1%, SD 4.9%). For high shedders, R₀ equaled or exceeded 2 in both the closed (1.9, SD 0.76) and AC (2.2, SD 0.13) conditions, but stayed below 1 for low shedders across all ventilation scenarios (Figure 1B).
Similarly, for influenza, transmission risk was high in the closed condition for high virus shedders (76.8%, SD 20.4%) and decreased significantly in the open condition (47.5%, SD 18.7%; p< 0.001) (Figure 2A). For low shedders, transmission risk remained low in both the closed (24.9%, SD 19.9%) and AC (25.2%, SD 5.0%) conditions. The R₀ for influenza exceeded 2 in the closed and AC conditions for high shedders (Figure 2B).
Conclusion. The transmission risk of respiratory viruses in homes and healthcare spaces is high, particularly with the use of AC. Under-ventilation increases R 0 above 1 among high virus shedders. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12606376&blobtype=pdf | # Serotypes and antibiotic resistance patterns of group B streptococci isolated from pregnant women at Urmia University Hospital, Iran
Lida Lotfollahi, Zahra Shahabi, Zahra Mousarezai, Shabnam Kimyai, Azar Hemmati, Ehsan Shojadel
## Abstract
Background and Objectives: Group B Streptococcus (GBS) is a common bacterium found in the gastrointestinal tract and genitalia of both humans and animals. GBS infections can lead to a range of conditions, including meningitis, pneumonia, and sepsis. The present study aimed to analyze the colonization rate, antibiotic susceptibility, and serotypes of GBS in pregnant women in Urmia, Iran. Materials and Methods: Following GBS isolation from pregnant women and confirming its presence through PCR, antibiotic susceptibility testing was conducted to assess resistance patterns, followed by amplification of resistance genes (mefA, ermB, ermTR, linB) and molecular serotyping to determine the genetic characteristics of the strains. Results: Out of 400 samples, 31 (7.75%) were positive for GBS, with 22 (70.97%) showing multidrug resistance. Clindamycin had the highest resistance rate (80.65%), while penicillin showed the lowest (3.23%). Serotypes II and V were the most common (38.71% each), followed by Ia (19.35%) and III (3.23%). The ermB gene was detected in 4 strains, while mefA, ermTR, and linB were not found. Conclusion: Optimal management of GBS infections in pregnant women necessitates ongoing surveillance and antibiotic stewardship, considering penicillin resistance and observed resistance patterns.
## INTRODUCTION
The human gastrointestinal and genitourinary tracts are often colonized by Group B Streptococcus (GBS), which acts as a commensal bacterium (1). This opportunistic pathogen can lead to several infections like bacteremia, urinary tract infection (UTI), and pneumonia. GBS is a significant concern during pregnancy (2), affecting approximately 10-30% of pregnant women worldwide (3). Maternal colonization with this bacterium poses risks to both mothers and the newborns, emphasizing the impor-tance of early detection and appropriate intervention. Transmission of GBS from the mother to the infant commonly occurs during childbirth (4).
Screening pregnant women for GBS is crucial to identify carriers and take preventive measures. The most common approach involves obtaining vaginal and rectal swabs (5). Undetected or untreated GBS infection during pregnancy can cause various complications, including chorioamnionitis, endometritis, and postpartum sepsis (6). In newborns, GBS infection can result in early-onset sepsis, pneumonia, meningitis, and other life-threatening conditions (7). Commonly used antibiotics, including penicillin and ampicillin, play a key role in both managing maternal GBS infections and preventing neonatal exposure during delivery. In addition to antibiotics, implementing strategies like intrapartum antibiotic prophylaxis and maternal education can greatly decrease the risk of this bacterium transmission to the newborn (8). The use of erythromycin or clindamycin as second-line treatments is usually designated for women with penicillin allergy or those who cannot take the first-line treatment options. Although GBS is typically susceptible to penicillin, some isolates have been reported to exhibit resistance due to moderate susceptibility or a reduced minimum inhibitory concentration (MIC) (9). Macrolide resistance has also been reported in GBS in different regions around the world. The emergence of erythromycin and clindamycin resistance is linked to the vaginal swabs were taken from women attending the gynecological clinic in Kosar Hospital, Urmia. Sample collection was carried out at 35 to 37 weeks of gestation, after obtaining their informed consent. This study received ethical approval from the Urmia University of Medical Sciences Ethics Committee (IR. UMSU.REC.1395.27). Women who had symptoms of UTI, had used antibiotics, or had experienced bleeding in the two weeks prior to sampling were excluded from this study.
The vaginal-rectal swabs taken from each woman were transferred to Todd-Hewitt-broth (THB) (Hi Media, India) supplemented with gentamicin (8 μg/ mL) along with nalidixic acid (15 μg/mL) (MAST, UK). All samples were obtained aseptically and immediately sent to the lab. In the lab, the swabs were inoculated onto sheep blood (5%) agar medium (SBA) (Merck, Germany), subsequently, incubation was Macrolide-Lincosamide-Streptogramin B resistance MLSB phenotype, driven by the acquisition and ex-done at 35°C for 24 h in a 5% CO atmosphere. pression of erm genes (10).
Ten serotypes of GBS are identified by differences in capsular polysaccharide antigens encoded within the cps gene cluster. Among these, serotypes V, III, II, and Ia are most frequently associated with infections in pregnant women and newborns (11,12).
GBS colonization and resistance have been studied in various regions worldwide, however, in Iran, research has been primarily focused on the central and southern areas, leaving a notable gap in data from the northwestern regions, including Urmia.
To address this deficiency, the present study was designated to investigate the carriage rate of GBS in pregnant women admitted to Urmia University Hospital, located in northwestern Iran. Additionally, the research aimed to estimate the prevalence of macrolide-resistant strains and characterize antimicrobial resistance patterns. It also sought to identify the presence of resistance genes and characterize the distribution of molecular serotypes. This study's outcomes are anticipated to contribute to the expansion of region-specific antibiotic stewardship policies and to enhance evidence-based strategies for intrapartum prophylaxis, particularly for patients with presumed β-lactam allergies.
## MATERIALS AND METHODS
## Study design and bacterial isolation.
During this cross-sectional study, a total of 400 rectal and high GBS identification and confirmation. Colonies with expected hemolytic and morphological patterns (the whitish-grey with β-hemolytic colonies) on SBA plates were picked and assessed by using other methods such as Gram staining, bacitracin and trimethoprim plus sulfamethoxazole (SXT) susceptibility, Christie-Atkins-Munch-Petersen (CAMP) reaction, hippurate hydrolysis, and catalase. Confirmation was performed by polymerase chain reaction (PCR) targeting the dltS gene using primers described by (13). Genomic DNA from S. agalactiae ATCC 12386 was used as the positive control, whereas nuclease-free water, devoid of DNA, served as the negative control.
Analysis of antimicrobial susceptibility. The confirmed GBS isolates were investigated for susceptibility to antibiotics (MAST, UK), including tetracycline (T, 30 μg), erythromycin (E, 15 µg), quinupristin-dalfopristin (Synercid, 15 µg), penicillin G (PG, 10 units), ofloxacin (OFX, 5 μg), chloramphenicol (C, 30 μg), clindamycin (CD, 2 μg), azithromycin (AZM, 15 μg), and ciprofloxacin (CIP, 5 μg) through Kirby-Bauer method using the CLSI criteria. The control strain used was Streptococcus pneumoniae (ATCC 49619). Eventually, the diameter of the inhibition zone around antibiotic disks was measured.
## Disk induction test.
According to the CLSI guidelines, a bacterial culture suspension made to the level of 0.5 McFarland's standard was used to create a lawn B B culture on a MHA plate supplemented with 5% sheep blood. After placing for CD (2 μg) and E (15 μg) discs on this plate 12 millimeters apart from one another edge-to-edge, the plates were incubated for 24 hours at 37°C. Using the disk induction test (D-test), four phenotypes were identified (14). Diminishing of the CD inhibition zone proximal to the E disk was recognized as an inducible Macrolide-lincosamide streptogramin B (iMLS ) phenotype. Resistance to both showed evaluate the relationships between the variables. A significant threshold of p <0.05 was established.
## RESULTS
Among the 400 samples analyzed, 31 (7.75%) tested positive for GBS based on phenotypic methods. PCR analysis confirmed all phenotypically identiconstitutive MLS methylation (cMLSB) phenotype.
fied isolates, demonstrating full agreement between phenotypic and genotypic identification techniques. Detection of resistance genes. Primarily, Bacterial DNA extraction was conducted using the Geno Plus™ Genomic DNA Extraction Mini Prep System Kit (VIOGENE, Taiwan) following the manufacturer's protocols. Using the PCR technique, the S. agalactiae isolates were examined for the presence of resistance genes such as ermB, mefA, ermTR, and linB. The particular primers displayed in Table 1 were synthesized at Takapouzist, Tehran, Iran. For PCR, 5 µL of extracted DNA was used as a template and the steps were as mentioned in (15). PCR products were investigated by gel electrophoresis in agarose (approximately 1%) in the 1X TAE buffer, stained with
The antibiotic susceptibility test results are shown in Table 2. Among the 31 isolates, 22 (70.97%) were multidrug-resistant. MDR is defined as resistance to at least one antimicrobial drug in three or more antimicrobial categories. The highest resistance rate was reported to clindamycin and quinupristin-dalfopristin with 80.65% and 77.42%, respectively. Furthermore, according to our observations, resistance to penicillin G was also seen (3.23%). Of note, all chloramphenicol-resistant isolates exhibited co-resistance to clindamycin, another antibiotic used in the management of bacterial infections. All 31 GBS isolates were phenotypically evaluated for resistance. Of these, 2 (6.45%) exhibited the concerning. In our country, unfortunately, excessive use of antibiotics has become one of the most important reasons for the increasing resistance rate. Given iMLS phenotype and 9 (29.03%) showed the cMLS the increasing resistance to the first and second-line phenotype. We analyzed the isolates for resistance genes by the PCR method. Among the 31 examined isolates, only 4 harbored the ermB gene, while no other resistance genes (such as mefA, ermTR, or linB) were detected. The serotypes distribution was as follows: II (38.71%), V (38.71%), Ia (19.35%), and III (3.23%). Other serotypes were not seen and all of the isolates were typeable. Importantly, serotype V isolates exhibited resistance to all tested antibiotics. antibiotics, it seems necessary to perform antibiotic sensitivity testing to select the appropriate drug.
Our study showed that 22 isolates were MDR. The high rate of MDR in our study (70.97%) and the study conducted in Vietnam (60.66%) (24) shows the need to determine antibiotic sensitivity because this increase in MDR strains can lead to a widespread concern.
Most of the resistance mechanisms in our study are similar to previous studies which were related to cM-LS (29.03%) (21,27). Additionally, iMLS pheno-DISCUSSION As mentioned earlier, GBS screening in pregnant women is essential, however, unfortunately in our country especially in our city (Urmia) little is known about it. The GBS colonization rate in Urmia (7.75%) was similar to that reported in Kashan (6.7%) (18). This relatively low prevalence is consistent with findings from other countries such as Argentina (19) and India (20). However, it is notably lower than the rates reported by Zakerifar et al. and Rostami et al. in other regions of Iran (21,22). The reason for these different frequencies can be explained based on the differences in the sampling procedure, identification type in our study had rates of 6.45%. The only resistance gene identified in this study was ermB, which encodes 23S rRNA methylases and modifies the antibiotic target site. Similar studies, such as those by Zakerifar et al. (21), have also found this mechanism to be the primary one. In our phenotypic analysis, 22 isolates showed reduced susceptibility to erythromycin, 12 were intermediate and 10 were resistant, but only four of them carried the resistance gene. This is probably because the gene coding for rRNA methylases has been mutated. This reason also applies to other isolates that were phenotypically resistant, but the resistance gene was not observed. Out of four strains that had the ermB gene, three of them showed methods, and geographical variations. Importantly, cMLS phenotype and the other one showed iMLS all 31 isolates detected by phenotypic methods were true positives and confirmed with PCR, which was also seen in a study by HajiAhmadi et al. (23). Therefore, both phenotypic and genotypic methods seem to be appropriate approaches for GBS screening in pregnant women.
Penicillin is the first line of treatment for GBS, except in cases of hypersensitivity and allergy. Today, the level of resistance to penicillin has become concerning. In this study, the sensitivity rate to penicillin was 80.65%, which is in agreement with the findings of Zakerifar et al., who reported a rate of 78.3% (21). Surprisingly, HajiAhmadi et al. also reported that all their detected isolates (n = 36) were resistant to penicillin (23). However, 100% sensitivity has also been reported in other studies (24)(25)(26). Clindamycin and erythromycin are commonly considered the second line of treatment for people who are allergic to penicillin. The sensitivity to E and CD was also estimated as 29.03% and 16.13%, respectively, which is phenotype. In the present study, contrary to what was described by Santana et al., two other genes namely mefA and ermTR were not observed. Similarly to our study, the linB gene was absent (25). Serotyping revealed four distinct serotypes among the isolates: II and V (38.71% each), Ia (19.35%), and III (3.23%). Other serotypes were not detected. In the present study, in agreement with Savoia et al., serotype V had a high frequency (28). On the other hand, there were other studies in which this serotype had a low frequency (13,29).
Serotype II was similarly prevalent in this study, a pattern consistent with another research (26). Also, in a meta-analysis study done in Africa, the rate of this serotype was low, and more serotype V was seen there (30). These geographical variations in serotype distribution highlight the influence of regional factors and underscore the need for localized surveillance. As it is evident from Table 3, serotype V shows resistance to all antibiotics, while this situa- (31). Another study also reported that serotype V had the highest resistance rates to azithromycin, clindamycin, and erythromycin (32). Serotype V is particularly important not only because of its high prevalence in our study but also due to its resistance to multiple antibiotics. Because of these concerns, ongoing surveillance and the inclusion of serotype V in vaccine strategies are important for effective control and prevention.
## CONCLUSION
Effective management of Group B Streptococcus infections in pregnant women necessitates continual surveillance and prudent use of antibiotics. This involves careful consideration of penicillin resistance and the prevailing resistance patterns. Vigilant monitoring of GBS infections allows for timely interventions and tailored antibiotic regimens, optimizing maternal and neonatal outcomes. Additionally, implementing antibiotic stewardship programs ensures wise antibiotic usage, minimizing the risk of antimicrobial resistance emergence and preserving the effectiveness of these crucial medications for future generations. By integrating surveillance efforts and antibiotic stewardship practices, healthcare providers can mitigate the impact of GBS infections on pregnant women and their babies while combating the threat of antimicrobial resistance.
## References
1. Iran, Microbiol (2025)
2. Choi, Han, Chong et al. (2022) "Updates on group B streptococcus infection in the field of obstetrics and gynecology" *Microorganisms*
3. Nguyen, Omage, Noble et al. (2021) "Group B streptococcal infection of the genitourinary tract in pregnant and non-pregnant patients with diabetes mellitus: an immunocompromised host or something more?" *Am J Reprod Immunol*
4. Jaworowski, Huras, Zembala-Szczerba et al. (2016) "The level of implementation of the prevention program of group B streptococcal infection during pregnancy and in newborns in accordance to the recommendations of the Polish Gynaecological Society -questionnaire survey" *Przegl Lek*
5. Steer, Russell, Kochhar et al. (2020) "Group B streptococcal disease in the mother and newborn-a review" *Eur J Obstet Gynecol Reprod Biol*
6. Numanović, Smajlović, Gegić et al. (2017) "Presence and resistance of Streptococcus agalactiae in vaginal specimens of pregnant and adult non-pregnant women and association with other aerobic bacteria" *Med Glas (Zenica)*
7. Tevdorashvili, Tevdorashvili, Andghuladze et al. (2015) "Prevention and treatment strategy in pregnant women with group B streptococcal infection" *Georgian Med News*
8. Szymusik, Kosińska-Kaczyińska, Pietrzak et al. (2014) "Do we need a different approach to GBS screening?" *Ginekol Pol*
9. Ohlsson, Shah (2014) "Intrapartum antibiotics for known maternal group B streptococcal colonization" *Cochrane Database Syst Rev*
10. Verani, Mcgee, Schrag (2010) "Division of Bacterial Diseases, National Center for Immunization and Respiratory Diseases"
11. Berbel, González-Díaz, De Egea et al. (2022) "An overview of macrolide resistance in streptococci: prevalence, mobile elements and dynamics" *Microorganisms*
12. Madrid, Seale, Kohli-Lynch et al. (2017) "Infant group B streptococcal disease incidence and serotypes worldwide: systematic review and meta-analyses" *Clin Infect Dis*
13. Russell, Seale, Driscoll et al. (2017) "Maternal colonization with group B streptococcus and serotype distribution worldwide: systematic review and meta-analyses" *Clin Infect Dis*
14. Sadeh, Firouzi, Derakhshandeh et al. (2016) "Molecular characterization of Streptococcus agalactiae isolates from pregnant and non-pregnant women at Yazd University Hospital, Iran" *Jundishapur J Microbiol*
15. (2018) "Performance standards for antimicrobial susceptibility testing. 28 th ed. CLSI supplement M100"
16. Bolukaoto, Monyama, Chukwu et al. (2015) "Antibiotic resistance of Streptococcus agalactiae isolated from pregnant women in Garankuwa, South Africa" *BMC Res Notes*
17. Mousavi, Nasaj, Hosseini et al. (2016) "Survey of strain distribution and antibiotic resistance pattern of group B streptococci (Streptococcus agalactiae) isolated from clinical specimens" *GMS Hyg Infect Control*
18. Poyart, Tazi, Réglier-Poupet et al. (2007) "Multiplex PCR assay for rapid and accurate capsular typing of group B streptococci" *J Clin Microbiol*
19. Dashtizade, Zolfaghari, Yousefi et al. (2020) "Antibiotic susceptibility patterns and prevalence of Streptococcus agalactiae rectovaginal colonization among pregnant women in Iran" *Rev Bras Ginecol Obstet*
20. Bobadilla, Novosak, Cortese et al. (2021) "Prevalence, serotypes and virulence genes of Streptococcus agalactiae isolated from pregnant women with 35-37 weeks of gestation" *BMC Infect Dis*
21. Ashary, Singh, Chhabria et al. (2022) "Meta-analysis on prevalence of vaginal group B streptococcus colonization and preterm births in India" *J Matern Fetal Neonatal Med*
22. Zakerifar, Kaboosi, Goli et al. (2023) "Antibiotic resistance genes and molecular typing of Streptococcus agalactiae isolated from pregnant women" *BMC Pregnancy Childbirth*
23. Rostami, Moeineddini, Ghandehari et al. (2021) "Macrolide-resistance, capsular genotyping and associated factors of group B streptococci colonized pregnant women in Isfahan, Iran" *Iran J Microbiol*
24. Hajiahmadi, Momtaz, Tajbakhsh (2025) "Molecular characterization of Streptococcus agalactiae strains isolated from pregnant women" *Sci Rep*
25. Van Du, Dung, Toan et al. (2021) "Antimicrobial resistance in colonizing group B streptococcus among pregnant women from a hospital in Vietnam" *Sci Rep*
26. Santana, De Oliveira, Filho et al. (2020) "Streptococcus agalactiae: identification methods, antimicrobial susceptibility, and resistance genes in pregnant women" *World J Clin Cases*
27. Ali, Woldeamanuel, Asrat et al. (2020) "Features of Streptococcus agalactiae strains recovered from pregnant women and newborns attending different hospitals in Ethiopia" *BMC Infect Dis*
28. Liu, Jiang, Li et al. (2023) "Molecular characteristics and antibiotic resistance mechanisms of clindamycin-resistant Streptococcus agalactiae isolates in China" *Front Microbiol*
29. Savoia, Gottimer, Crocilla et al. (2008) "Streptococcus agalactiae in pregnant women: phenotypic and genotypic characters" *J Infect*
30. Eskandarian, Ismail, Neela et al. (2015) "Amin Nordin S. Antimicrobial susceptibility profiles, serotype distribution and virulence determinants among invasive, non-invasive and colonizing Streptococcus agalactiae (group B streptococcus) from Malaysian patients" *Eur J Clin Microbiol Infect Dis*
31. Gizachew, Tiruneh, Moges et al. (2019) "Streptococcus agalactiae maternal colonization, antibiotic resistance and serotype profiles in Africa: a meta-analysis" *Ann Clin Microbiol Antimicrob*
32. Wang, Tong, Ma et al. (2015) "Serotypes, antibiotic susceptibilities, and multi-locus sequence type profiles of Streptococcus agalactiae isolates circulating in Beijing" *PLoS One*
33. Tsai, Su, Wang et al. (2022) "Alterations in genes rib, scpB and pilus island decrease the prevalence of predominant serotype V, not III and VI, of Streptococcus agalactiae from 2008 to 2012" *Pathogens* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12859394&blobtype=pdf | # TORCH agents in women of fertile age: towards prevention of congenital infections, Italy, 2019 to 2020
Claudia Pavia, Maurizio Zavattoni¹, Luigia Scudeller³, Ambra Vola⁴, Pierangelo Clerici¹, Massimo De Paschale, Francesca Genco, Cristina Giraldi¹, Francesca Greco, Valeria Meroni, Tiziana Lazzarotto, Pavia Claudia, Zavattoni Maurizio, Scudeller Luigia, Vola Ambra, Clerici Pierangelo, Paschale De, Genco Massimo, Giraldi Francesca, Greco Cristina, Francesca, Lisa Chenal, Claudio Giacomazzi, Alessandra Sacchi, Cristina Costa, Rita Scarso, Flavia Lillo, Silvia Tonolo, Maria Straface, Margherita Longo, Gloria Gagliardi, Elisabetta Cesana, Anna Borrelli, Marco Boi, Federica Miano, Carlo Bonfanti, Cinzia Lamorgese, Alice Bonetti, Eva Robatscher, Elena Moroder, Giacomina Pavan, Cristina Canova, Anna Belgrano, Lorenzo Zandonà, Liliana Gabrielli, Simona Semprini, Michela Fantini, Roberta Schiavo, Giuliana Cascio, Giovanna Moscato, Maria Colao, Silvia Bozza, Barbara Camilloni, Monica Ferreri, Katia Marinelli, Chiara Nonne, Ombretta Turriziani, Stefania Ranno, Luana Coltella, Sandro Grelli, Pierpaolo Paba, Luigi Clerico, Claudio Palmerini, Paola Salvatore, Giuseppe Portella, Agata Calvario, Caterina Colonna, Antonietta Sinno38, Anna Giandomenico, Guido Scalia, Concetta Palermo, Simona De Grazia, Emilia Palazzotto, Giorgio Melis, Pietro Porcu, Maura Fiamma, Anna Rita
## Abstract
Background: Microbiological surveillance during pregnancy is important for better neonatal outcomes. Aim: We aimed to assess IgG and IgM antibodies against Toxoplasma gondii, parvovirus B19, Treponema pallidum, rubella virus and cytomegalovirus in women of fertile age (16-45 years) in Italy and investigate factors associated with the presence of antibodies. Methods: We collected data from clinical microbiology laboratories on test results for IgG and IgM antibodies against T. gondii, parvovirus B19, T. pallidum, rubella virus and cytomegalovirus between 1 July 2019 and 30 June 2020. Serological tests, like IgG avidity for T. gondii and cytomegalovirus, non-treponemal tests for T. pallidum and molecular tests for parvovirus B19 and rubella virus were considered as confirmatory tests for acute infections. We investigated associations between presence of antibodies with age, nationality and geographic area of residence. Results: Thirty-two laboratories submitted test results on 342,095 women.
## Introduction
The acronym TORCH is used to indicate infections caused by Toxoplasma gondii (T), other pathogens (O) referring for instance to the causative agent of syphilis, Treponema pallidum, and to parvovirus B19, rubella virus (R), cytomegalovirus (C) and herpes simplex virus (H) [1]. In healthy individuals, these infections are generally mild, but during pregnancy, these pathogens can cross the placenta and, when transmitted to the fetus, may lead to severe consequences (e.g. spontaneous abortion, stillbirth, blindness, deafness, neurodevelopment deficit, postnatal sequelae) [2]. Exposure to TORCH pathogens is affected by environmental conditions such as climate, season, vaccination coverage, cultural dietary practices, as well as by socioeconomic factors, maternal age, rate of preschool attendance by siblings and vaccine adherence [3][4][5][6]. Vaccination is currently available only against rubella virus, and in Italy it is offered within the child immunisation programme [7][8][9].
Toxoplasmosis is a zoonosis caused by the parasite Toxoplasma gondii, usually acquired via ingestion of contaminated food or water and rarely via transplanted organs [2]. Prenatal serological screening is essential for identifying women at risk and for appropriate actions (e.g. monthly screening, dietary counselling and treatment) to prevent transmission to the fetus [10].
Parvovirus B19 circulates in all areas of the world with seasonal outbreaks occurring every 3-5 years [2]. In European countries, seroprevalence in pregnant women varies from 55% to 74%. The incidence of parvovirus infection in pregnancy is approximately 1-2%, varying by season or during outbreaks. Fetal anaemia and hydrops are observed after maternal infection during pregnancy [11].
Syphilis causes more than 350,000 adverse pregnancy outcomes each year worldwide (over half as stillbirths or neonatal deaths) [7]. The recommended preventive actions include (i) strengthening surveillance, with programme monitoring and progress evaluation; (ii) prevention of sexually transmitted infections (STI); (iii) early diagnosis of STIs; (iv) patient and partner management and (v) approaches to reach the most vulnerable population groups. Targets and milestones of the global health strategies on STIs for 2022-2030 by World Health Organization (WHO) include 90% reduction of the global incidence of syphilis from 2018 and ≤ 50 cases of congenital syphilis per 100,000 live births in 80% of countries [12].
The acute illness caused by rubella virus infection is usually mild and characterised by fever and rash. If contracted in early pregnancy, it can result in congenital rubella syndrome of the fetus, with severe consequences such as cataracts, sensorineural hearing impairment, congenital heart disease, jaundice, purpura, hepatosplenomegaly and microcephaly [2].
Cytomegalovirus (CMV) is a major cause of congenital infections globally [2]. Infections in pregnancy can be primary (contracted during pregnancy) or nonprimary (reinfection with other viral strains or, more frequently, reactivation). Primary maternal cytomegalovirus infection has a high risk of in-utero transmission. Congenital infection can present with petechiae, hepatosplenomegaly, jaundice, microcephaly, thrombocytopenia and hearing loss. In some states of the United States (US) for example, targeted screening of newborns for CMV is implemented [13].
During pregnancy, early recognition of TORCH infections for which treatment is available is important to prevent fetal infection and adverse fetal and neonatal outcomes. Also, where no specific treatment is available, a correct fetal monitoring, amniocentesis and fetal management plan should be offered. Therefore, national maternal and child healthcare programmes require accurate data on the incidence and prevalence of these infections.
Presence of IgG and absence of IgM antibodies against a specific pathogen are considered a sign of a previous infection or vaccination, while negative IgG and IgM results indicate a non-immune, susceptible person [14]. Presence of IgM and IgG antibodies indicate a recent or acute infection. Additional serological or molecular tests are necessary to better evaluate the risk of transmission. Thus, IgG and IgM serosurveys can be used to measure immunity and vaccination levels.
A correct microbiological surveillance, through the evaluation of serological status, monitoring and counselling of seronegative women of fertile age and during pregnancy, results in better neonatal outcomes, and knowledge of prevalence of these infections in the population is a prerequisite for designing and maintaining serological screening programmes [2]. Currently, only local seroprevalence data about TORCH pathogens are available in Italy [15].
## What did you want to address in this study and why?
Infections during pregnancy can lead to severe consequences to the fetus and newborn. We studied protection against so called TORCH pathogens including Toxoplasma gondii, parvovirus B19, Treponema pallidum, rubella virus and cytomegalovirus but excluding herpes simplex virus, among women of childbearing age (16-45 years) in Italy and the estimated number of new infections in one year, by age, geographic area and nationality.
## What have we learnt from this study?
Immunity to these infections varied from 1% for T. pallidum, the causative agent of syphilis, to 87% for rubella virus, the latter is the only one of these microbes for which a vaccine is available. Our results also showed regional and demographic differences in seropositivity rates, underscoring the need for tailored screening and prevention initiatives.
## What are the implications of your findings for public health?
Our study results stress the importance of maternal surveillance programmes in the prevention of neonatal infections. The results highlight that fertile women might benefit from specific advice on preventive measures and closer clinical follow-up.
## KEY PUBLIC HEALTH MESSAGE
We aimed to estimate the presence and distribution of antibodies against T. gondii, parvovirus B, T. pallidum, rubella virus and cytomegalovirus in fertile women to help guide public health initiatives aimed at reducing the overall burden of maternal-fetal infectious diseases. We did not include herpes infections as no serological screening is performed in Italy and infections are diagnosed only in symptomatic cases by molecular tests.
## Methods
## Microbiological surveillance of pregnant women
In Italy, serological screening in first trimester and monthly re-testing of non-immune pregnant women is not mandatory but recommended and offered free of charge for T. gondii, rubella virus and T. pallidum, and widely performed [10,16]. At the time of data collection of our study, testing for CMV and parvovirus B19 during pregnancy was not recommended but parvovirus B19 is tested when the pregnant person or the fetus shows signs of parvovirus infection.
## Study design
We invited all Italian clinical microbiology laboratories included on the Italian Association of Clinical Microbiologist (AMCLI) mailing list to participate in a multicentre study on a voluntary basis. The laboratories were asked to submit pooled results of routine serological testing of all women aged 16-45 years, irrespective of pregnancy status, for T. gondii, parvovirus B19, T. pallidum, rubella virus and CMV between 1 July 2019 and 30 June 2020. Data were pseudonymised and then pooled by the laboratory prior to submission.
## Data collection
The laboratories were asked to submit test results by age (16-25, 26-35, 36-45 years) and nationality (Italian or other nationality). The geographic region the laboratory covered was grouped into three areas: northern Italy (Emilia Romagna, Friuli Venezia Giulia, Liguria, Lombardy, Piedmont, P.A. Bolzano, Aosta Valley, Veneto), central Italy (Abruzzo, Lazio, Marche, Tuscany, Umbria) and southern Italy (Basilicata, Calabria, Campania, Molise, Apulia, Sardinia, Sicily). The data collection form was developed in Microsoft Excel. A high IgG avidity index suggested a past infection [13,17].
For parvovirus B19, acute infections were defined as presence of IgM antibodies confirmed by detection of parvovirus DNA by commercial PCR tests.
For rubella, acute primary infections were defined as combined presence of IgM and ascertained IgG seroconversion and/or positive RT-PCR or low or intermediate IgG avidity index as recorded by the laboratories.
## Statistical analysis
Descriptive statistics were produced for each variable. Numbers and percentages are presented for categorical variables.
Meta-analytic techniques were used to obtain estimates; the pooled observed proportions per laboratory in each category were weighted according to the total number of samples processed via inverse variance weighting; binomial standard errors were calculated. Given the high heterogeneity in the analyses, restricted maximum likelihood was used for estimation. Heterogeneity was estimated with the I 2 , and Q statistics was used for homogeneity testing. Metaregression was used to investigate factors associated with higher prevalence (or incidence). Laboratories with missing data on a pathogen and serological tests were given zero weight in the corresponding analysis (equivalent to not including them).
## Toxoplasma gondii
Of the 114,678 women tested for antibodies against T. gondii, 13,700 (13%; 95% CI: 12-14) of 111,580 had IgG antibodies and 1,805 (3%; 95% CI: 1-4) of 114,678 had IgM antibodies (Table 1). Of the 1,805 women with IgM antibodies, 530 had acute infections (4/1,000 women tested; 95% CI: 3-4). Women with other nationalities had a higher proportion of IgG antibodies than those with Italian nationality (p < 0.001). No significant differences were observed between the included variables and IgM antibodies. Women with other nationalities had fewer acute infections than those with Italian nationality (p < 0.001) but women living in southern Italy had more acute infections than women in the other parts of the country (p = 0.039).
## Parvovirus B19
Of the 5,138 women tested for antibodies against parvovirus B19, 3,298 (65%; 95% CI: 60-69) had IgG antibodies and 205 (0%; 95% CI: 0-1) of 5,059 had IgM antibodies (Table 2). Thirty-three (16%) of the 205 women with IgM antibodies tested positive with PCR, confirming an acute infection (viral load > 10 4 copies/ mL). Twenty-seven of these women were Italian, six had other nationalities, four were aged 16-25 years, 17 were 26-35 years and 12 were 36-45 years. No significant differences were observed between Italian and other nationalities, ages or geographic area and presence of antibodies.
## Treponema pallidum
Of the 81,401 women tested for antibodies against T. pallidum, 889 (1%; 95% CI: 1-1) had IgG antibodies and 329 (0%; 95% CI: 0-1) had treponemal and nontreponemal antibodies (Table 3). The proportion of women with treponemal and non-treponemal antibodies was significantly higher in southern Italy than in the other parts of the country (p < 0.001).
## Rubella virus
Of the 69,865 women tested for antibodies against rubella virus, 63,828 (87%; 95% CI: 85-88) had IgG antibodies and 2,998 (2%; 95% CI: 1-3) of 55,327 had IgM antibodies (Table 4). The proportion of women with IgG antibodies was lower in southern Italy than in the other parts of the country (p < 0.001). No acute rubella cases were diagnosed.
## Cytomegalovirus
Of the 71,013 women tested for antibodies against CMV, 45,558 (66%; 95% CI: 64-68) had IgG antibodies and 3,452 (3%; 95% CI: 2-3) of 69,968 had IgM antibodies (Table 5). We estimated 449 acute infections (4/1,000; 95% CI: 3-5). Women with other nationalities had a higher proportion of IgG antibodies but a lower proportion for IgM antibodies and fewer acute infections than those with Italian nationalities (p < 0.001). Women living in southern Italy had higher proportions of IgG and IgM antibodies than women living in the other parts of the country (p < 0.001).
## Discussion
In Europe, data on the prevalence of TORCH agents in women of fertile age are sparse. We estimated presence of antibodies against five pathogens in women aged 16-45 years in Italy. Furthermore, we investigated associations between proportions of antibodies and age and geographic area. As we received data during a 12-month period from 32 laboratories from all Italian regions but one, including all the major referral centres for antenatal screening of TORCH infections, the study results can be considered representative of the status in women of fertile age accessing healthcare as part of routine practice. The results highlight that fertile women might benefit from specific advice on preventive measures and closer clinical follow-up.
The proportion of IgG antibodies against T. gondii was 13%, similar to some other studies. Seroprevalence of T. gondii antibodies has decreased in Italy [18] and some other countries [19,20]. As T. gondii is often transmitted via food, the decrease could be due to changes in alimentary habits, farming systems and improved hygienic practices [19]. The presence of IgG antibodies was significantly higher among women with other nationalities than the Italian. Seroprevalence increase with increasing age has been previously reported [21,22] and could be due to multiple exposures to T. gondii throughout life. Presence of IgM antibodies against T. gondii was less common (3%). However, testing for IgM antibodies is not sufficient to define an acute infection [23]. In our study, 29.3% of women with IgM antibodies (4/1,000; 95% CI: 3-4) had acute infections as assessed by IgG and IgM seroconversion and low or intermediate avidity. Toxoplasmosis can have a high disease burden and treatment of gestational toxoplasmosis improves the prognosis. Thus, also with a low seroprevalence, screening will be costeffective [24,25]. As T. gondii is endemic in wild and domestic animals in Italy [16], it is important to maintain the screening programme as pointed out in the Italian Guidelines for Physiological Pregnancy [10].
Of the 5,138 women tested, 65% had IgG antibodies against parvovirus B19, with no statistically significant differences between nationality, age and geographic area, contrary to a recent study in Italy [11]. Of the 5,059 women tested, 205 (2%) had IgM antibodies and 33 had acute infections. Even if serological screening is not recommended in pregnancy, during parvovirus B19 outbreaks when seroconversion rates are higher, assessment for maternal immunity is relevant to identify women at risk of acquiring B19 infection in pregnancy.
In 2016, WHO launched the 2016-2021strategy of global curable and incurable STIs and included T. pallidum among infections that need immediate action. In 2019, T. pallidum was included in testing of women of fertile age in Italy. In our study, 889 women had anti-T. pallidum antibodies and 329 were positive in treponemal and non-treponemal tests. The proportion of IgG antibodies was higher among women in southern Italy, confirming results from a previous study [15]. The Italian guidelines recommend serological screening of all pregnant women for syphilis in the first and the third trimester, which we support. To confirm the diagnosis, seropositive women need to be tested with TPHA and RPR or VDRL test [10].
According to data from the Italian Ministry of Health, in 2020, vaccination coverage against rubella virus was around 90% [15]. In our study, the overall proportion of IgG antibody positivity was 87% and significantly lower in southern Italy (77%). In previous studies, the seroprevalence was around 90% in northern Italy, 85% in central Italy and 81% in southern Italy [15,26,27]. This could reflect a lower vaccination coverage in southern Italy. We observed a significantly higher proportion of IgG antibody positivity among women aged 36-45 years. Differences across age groups have been seen in other studies [27]. Measurement of IgG antibodies against rubella virus with some of the assays used in our study may have led to false-negative results, potentially triggering unnecessary booster vaccinations. However, the number of notified rubella cases has not increased, and no congenital rubella syndrome cases have been recently reported in Italy [28]. Of the tested women in our study, 2% had IgM positive results for rubella antibodies but no acute infections were observed. Test results for IgM antibodies can be false-positive or caused by persistent IgM reactivity after infection or vaccination [27]. Women should be informed that rubella-specific IgM positive results without clinical symptoms in a previously vaccinated person are most likely not an indication of an acute infection: further testing may not be warranted and termination of pregnancy unnecessary.
The current Italian guidelines do not recommend testing for rubella antibodies in pregnancy [10]. However, European Centre for Disease Prevention and Control (ECDC) reported 227 cases of rubella in 2024. As the estimated proportion of IgG seropositivity in southern Italy was below 85% in our study, in the absence of measles-mumps-rubella (MMR) vaccination certification, it would seem prudent to test women of fertile age for rubella IgG antibodies to assess immunity to the virus and take actions to prevent congenital infection [28,29].
Of the tested women, 66% had IgG antibodies against CMV. The lower the seroprevalence of CMV, the higher the risk of acute infection during pregnancy [30]. A total of 3,452 (3%) women had IgM antibodies against CMV, and IgM antibodies can be associated with both primary and secondary CMV infections. We identified 449 acute infections (4/1,000; CI 95%: 3-5), with no differences between the three age groups. Presence of antibodies was lower among women with Italian nationality than among other nationalities. Exposure to CMV may be associated with education level, social status and lifestyle, population density, number of children per family and child-rearing practices [31]. Interventions to reduce the risk of maternal CMV infection are limited to behavioural practices. Valacyclovir has been recently approved to prevent vertical transmission. Thus, our comprehensive study on CMV infection in women of reproductive age is relevant towards our understanding of CMV and associated disease, which can guide public health strategies. Our study results support and confirm the national strategy of conducting screening for CMV in pregnancy [10].
We acknowledge some limitations. Only aggregated data were used in the analyses; therefore, we have been able to study only some factors, such as age group, nationality and geographic area. Also, the participating laboratories used a variety of commercially available tests, and we could not take into account the potential variability in test performances. Finally, the concomitant COVID 19 pandemic affected participation of the laboratories, collection and analysis of datasets. In fertile age, the identification of IgG seronegative women and an intervention based on vaccination for rubella virus before pregnancy and hygiene recommendations for CMV in pregnancy have the potential to prevent maternal and congenital infection. Our study can provide the rationale to improve existing prevention strategies in Italy.
## License, supplementary material and copyright
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC BY 4.0) Licence. You may share and adapt the material, but must give appropriate credit to the source, provide a link to the licence and indicate if changes were made.
Any supplementary material referenced in the article can be found in the online version.
This article is copyright of the authors or their affiliated institutions, 2026.
## References
1. Coyne, Lazear (2016) "Zika virus -reigniting the TORCH" *Nat Rev Microbiol*
2. Neu, Duchon (2014) "TORCH infections" *Clin Perinatol*
3. Chen, Liu, Shi et al. (2019) "Seasonal influence on TORCH infection and analysis of multi-positive samples with indirect immunofluorescence assay" *J Clin Lab Anal*
4. Akyar (2011) "Seroprevalence and coinfections of Toxoplasma gondii in childbearing age women in Turkey" *Iran J Public Health*
5. Abu-Madi, Behnke, Dabritz (2010) "Toxoplasma gondii seropositivity and co-infection with TORCH pathogens in highrisk patients from Qatar" *Am J Trop Med Hyg*
6. Guimarães, Carneiro, Carvalho-Costa (2015) "Increasing incidence of pertussis in Brazil: a retrospective study using surveillance data" *BMC Infect Dis*
7. Boyer, Boyer (2004) "Update on TORCH infections in the newborn infant" *Newborn Infant Nurs Rev*
8. De Jong, Vossen, Walther et al. (2013) "How to use... neonatal TORCH testing" *Arch Dis Child Educ Pract Ed*
9. Kishore, Misra, Paisal et al. (2011) "Adverse reproductive outcome induced by Parvovirus B19 and TORCH infections in women with high-risk pregnancy" *J Infect Dev Ctries*
10. Superiore, Sanità (2025) "Gravidanza fisiologica" *Italian*
11. De Paschale, Cerulli, Cagnin et al. (2022) "Prevalence of anti-parvovirus B19 IgG and IgM and parvovirus B19 viremia in pregnant women in an urban area of Northern Italy" *J Med Virol*
12. (2022) "Global health sector strategies on, respectively, HIV, viral hepatitis and sexually transmitted infections for the period 2022-2030"
13. Leruez-Ville, Foulon, Pass et al. (2020) "Cytomegalovirus infection during pregnancy: state of the science" *Am J Obstet Gynecol*
14. Cutts, Seroepidemiology (2016) "an underused tool for designing and monitoring vaccination programmes in low-and middle-income countries" *Trop Med Int Health*
15. Palazzotto, Bonura, Calà et al. (2023) "Serological status for TORCH in women of childbearing age: a decade-long surveillance (2012-2022) in Italy" *J Med Microbiol*
16. Dini, Morselli, Marangoni et al. (2023) "Spread of Toxoplasma gondii among animals and humans in Northern Italy: A retrospective analysis in a One-Health framework" *Food Waterborne Parasitol*
17. Garnaud, Fricker-Hidalgo, Evengård et al. (2020) "Toxoplasma gondiispecific IgG avidity testing in pregnant women" *Clin Microbiol Infect*
18. Fanigliulo, Marchi, Montomoli et al. (2020) "Toxoplasma gondii in women of childbearing age and during pregnancy: seroprevalence study in Central and Southern Italy from 2013 to 2017" *Parasite*
19. Bigna, Tochie, Tounouga et al. (2020) "Global, regional, and country seroprevalence of Toxoplasma gondii in pregnant women: a systematic review, modelling and meta-analysis" *Sci Rep*
20. Van Den Berg, Stanoeva, Zonneveld et al. (2023) "Seroprevalence of Toxoplasma gondii and associated risk factors for infection in the Netherlands: third cross-sectional national study" *Epidemiol Infect*
21. Qin, Zhang, Liu et al. (2021) "Seroepidemiology of TORCH Infections among 1.7 million women of childbearing age in rural China: a population-based cross-sectional study" *Am J Trop Med Hyg*
22. Robinson, De Valk, Villena et al. (2021) "National perinatal survey demonstrates a decreasing seroprevalence of Toxoplasma gondii infection among pregnant women in France, 1995 to 2016: impact for screening policy" *Euro Surveill*
23. Gras, Gilbert, Wallon et al. (2004) "Duration of the IgM response in women acquiring Toxoplasma gondii during pregnancy: implications for clinical practice and cross-sectional incidence studies" *Epidemiol Infect*
24. Ferrari Strang, Ferrar, Falavigna-Guilherme (2023) "Gestational toxoplasmosis treatment changes the child's prognosis: A cohort study in southern Brazil" *PLoS Negl Trop Dis*
25. Sawers, Wallon, Mandelbrot et al. (2022) "Prevention of congenital toxoplasmosis in France using prenatal screening: A decision-analytic economic model" *PLoS One*
26. De Paschale, Manco, Paganini et al. (2012) "Rubella antibody screening during pregnancy in an urban area of Northern Italy" *Infect Dis Rep*
27. Marchi, Viviani, Montomoli et al. (2019) "Elimination of congenital rubella: a seroprevalence study of pregnant women and women of childbearing age in Italy" *Hum Vaccin Immunother*
28. (2025) "Measles and Rubella monthly report"
29. Gallone, Gallone, Larocca et al. (2017) "Lack of immunity against rubella among Italian young adults" *BMC Infect Dis*
30. De Vries, Van Zwet, Dekker et al. (2013) "The apparent paradox of maternal seropositivity as a risk factor for congenital cytomegalovirus infection: a population-based prediction model" *Rev Med Virol*
31. Revello, Tibaldi, Masuelli et al. (2015) "Prevention of primary cytomegalovirus infection in pregnancy" *EBioMedicine* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12826957&blobtype=pdf | # Specific and visual detection of EBOV based on a one-pot RT-RAA-CRISPR/ Cas12a assay
Zanheng Huang, Pei Huang, Zengguo Cao, Tianyi Zhang, Kaikai Jin, Meihui Liu, Yujie Bai, Zhiyuan Gong, Xuemeng Li, Yuanyuan Li, Haili Zhang, Hualei Wang, Virologica Sinica
The frequent emergence of life-threatening infectious diseases has posed a constant global threat in recent decades. Since the first identification of Ebola Virus Disease (EVD) in the Democratic Republic of Congo (then Zaire) in 1976, there have been approximately 40 outbreaks. About 30 of these were caused by the Ebola virus (EBOV), resulting in roughly 34,000 reported cases and 15,000 deaths. The overall case fatality rate for Ebola is estimated at 44%, with EBOV being the most prevalent of the virus species within the genus Ebolavirus (WHO; Jacob et al., 2020). Early and rapid diagnosis of these diseases is important to treat infected individuals and to control pandemics (Chams et al., 2020). Evidence suggests that the survival rate for patients with EBOV infection reaches 90% when the disease is diagnosed and treated early in the course of infection, compared to only 30% for those treated at a later stage (Mulangu et al., 2019).
Reverse transcription polymerase chain reaction (RT-PCR), encompassing both conventional and real-time approaches, is the primary method for the laboratory diagnosis of EVD. Of these approaches, the GeneXpert® Ebola assay stands out as a widely used commercial kit (Kaushik et al., 2016). However, RT-PCR has several limitations, including the need for time-consuming thermal cycling on specific instruments, and for skilled laboratory technicians (Broadhurst et al., 2016). Conventional RT-PCR testing therefore requires mobile laboratories when deployed in the field, posing significant challenges in conflict zones (Mukadi-Bamuleka et al., 2023). Developing a safe, rapid, and effective detection strategy for this high-biosecurity-level pathogen in resource-limited settings remains a formidable challenge (Broadhurst et al., 2016). Hence, there is an urgent need for the development of a diagnostic method that can be implemented without reliance on complex instrumentation, thereby reducing the operational burden on testing personnel.
Accordingly, we have developed a simple, rapid and specific one-pot assay for the detection of EBOV VP40 gene using an all-in-one RT-RAA-CRISPR/Cas12a method. All components for reverse transcript recombinase-aided amplification (RT-RAA) and CRISPR-based methods can be integrated into a single reaction system. This allows for the transfer and complete mixing of amplification products without opening the lid. Ultimately, this assay is capable of detecting 3.6 copies/μL linearized plasmid within 40 min.
First, the target is amplified using RT-RAA, and the amplified product is then specifically recognized by Cas12a and activated for transcleavage. This enables the cleavage of the single-stranded fluorescent probe present in the system. Results are interpreted based on the presence or absence of a fluorescent signal. Traditional two-step methods often cause aerosol contamination when the reaction tube is opened. Furthermore, Cas12a can weaken RAA amplification efficiency by cutting system components. To address these issues, we designed a novel closure device with a tube-sleeve-tube structure. The RT-RAA system is prepared following standard protocols, and the RT-RAA reaction mixture is placed at the bottom of a 1.5 mL eppendorf tube. The Cas12a reaction mixture is then loaded into an uncapped PCR tube. After adding the sample to be examined into the RT-RAA system, the PCR tube is inserted into a 1.5 mL eppendorf tube, sealed and put into a thermal cycler or water bath at 39 • C for 20 min for amplification. After amplification, the CRISPR/Cas12a reaction mixture and RT-RAA reaction mixture are mixed by centrifugation and incubated at 37 • C for 20 min. Finally, the reaction tube is placed under ultraviolet excitation light to observe the occurrence of green fluorescence (Fig. 1A).
We first expressed and purified the Cas12a protein with a SUMO tag and confirmed that it exhibits good cleavage activity (Supplementary Fig. S1). To identify a highly specific and broadly conserved CRISPR/ Cas12a target across EBOV strains, we compared 35 EBOV sequences from the GenBank database, representing diverse geographical regions and time periods, using MAFFT version 7 for multiple sequence alignment. Based on the restriction of the PAM site (TTTV), we then selected three different highly conserved sequences of 20-23 bp in the EBOV VP40 gene as targets. Details of the EBOV strains are provided in Supplementary Fig. S2A. Additionally, the sequences of the target regions in EBOV were found to be distinct from those of other viruses, including Bundibugyo virus (BDBV), Taï Forest virus (TAFV), Reston virus (RESTV), Bombali virus (BOMV), Sudan virus (SUDV), Marburg virus (MARV), Dengue virus 2 (DENV2) and Nipah virus (NiV), demonstrating that the target regions in EBOV were distinct from those in the other viruses, and indicating that the reaction would have good specificity (Supplementary Fig. S2A). The corresponding crRNAs were designed according to the conserved targets. All three crRNAs were able to specifically recognize the target sequence, but were found to exhibit different degrees of fluorescence intensity. All three crRNAs produced strong fluorescence signals within 10 min and reached their maximum fluorescence signals within 20 min. crRNA2 not only generated the highest fluorescence signals across all time points but also reached the detection threshold in the shortest time (Supplementary Fig. S2B). Therefore, crRNA2 was selected for subsequent related experiments.
To ensure that the CRISPR/Cas12a reaction system had optimal cleavage efficiency, we optimized the reaction components related to the CRISPR/Cas12a system. First, we evaluated the effect of the concentration of crRNA. We found that within a certain range, increasing the crRNA concentration also increased the Cas12a cutting efficiency. The optimal reaction efficiency and the strongest fluorescence signal were observed at a concentration of 100 nM, but at concentrations greater than 100 nM, the efficiency declined (Fig. 1B).
Next, we optimized the Cas12a concentration. We found that the effect of Cas12a was less pronounced than that of crRNA when adjusted within a certain concentration range. However, similar to crRNA, the efficiency increased with the protein concentration, and the efficiency was optimal at a protein concentration of 100 nM, after which the efficiency enhancement was insignificant with further increases in protein concentration (Fig. 1C). Finally, to examine the effect of reaction buffers on the cleavage activity, we established four distinct CRISPR/Cas12a reaction systems with different reaction buffers. We found that the effect of different buffer compositions on the efficiency of Cas12a cleavage was pronounced, with the optimal reaction efficiency seen in the Bio-lifesci buffer, which might be attributed to its higher magnesium ion concentration and optimal pH (Supplementary Fig. S2C).
To mitigate the risk of false negatives caused by workspace contamination during the opening of the RT-RAA reaction, we developed a one-pot assay combining the RT-RAA amplification and the CRISPR/Cas12a reaction. In this assay, the RT-RAA amplification system and CRISPR/Cas12a system were preloaded into two separate reaction tubes, and the samples to be detected were added to the RT-RAA system for amplification for 20 min. After amplification, the CRISPR/Cas12a and RT-RAA systems were combined via brief centrifugation, and the CRISPR/Cas12a was used for the specific identification of the amplified target. After 10-20 min of CRISPR/Cas12a reaction, the presence of fluorescent signals was assessed visually under UV light.
After optimizing the CRISPR/Cas12a system, we next focused on developing a one-pot assay. We designed three sets of RT-RAA primers on both sides of the optimal crRNA2, and found that the primer combinations of Target2-F and Target2-R1 demonstrated superior amplification efficiency and specificity. Subsequently, we optimized the reaction temperature within the typical RT-RAA range (37-42 • C) to maximize the overall assay performance. The assay detected 3.6 copies/ μL linearized plasmid within 40 min at both 39 • C and 42 • C (Fig. 1D), but Cas12a activity was affected by the RT-RAA reaction temperature. We therefore optimized the reaction volume ratios at 39 • C. To determine the optimal volume ratio that minimizes this inhibitory effect, we compared several combinations. We found that when both systems were adjusted to 25 μL, the assay achieved a limit of detection of 3.6 copies/ μL for the linearized plasmid within 40 min, while the other two groups detected only 3.6 × 10 1 copies/μL linearized plasmid (Fig. 1E). To determine if extending the RT-RAA amplification time could improve assay sensitivity, we tested various time points. While longer amplification times resulted in stronger fluorescence signals, they did not improve the limit of detection, which remained at 3.6 copies/μL (Fig. 1F).
The sensitivity of the one-pot assay was evaluated using a serial dilution series of the pCAGGS-EBOV-linearized plasmid, with concentrations from ranging from 3.6 × 10 4 to 10 0 copies/μL. Amplification was performed according to the previously optimized conditions for the RT-RAA amplification and CRISPR/Cas12a reactions. The results showed that the assay could detect 3.6 copies/μL linearized plasmid (Fig. 1G).
Due to the high mortality rate of EBOV relative to other ebolaviruses and the fact that early infection does not have specific symptoms, it is extremely important to differentiate EBOV from other hemorrhagic fever viruses in a timely manner. In order to verify the specificity of the RT-RAA-CRISPR/Cas12a assay, we selected six viral RNA samples to assess assay specificity. To verify the specificity of the RT-RAA-CRISPR/ Cas12a assay, we assessed it against six viral RNA samples, including different subtypes of ebolaviruses (EBOV, BDBV, SUDV, TAFV) and other related viruses with similar clinical symptoms (DENV2 and NiV). To ensure the reliability of the specificity results, we set up a positive control using the pCAGGS-EBOV-VP40 linearized plasmid as a template and a negative control using nucleic acid-free water. The results showed that only the positive control and the EBOV RNA sample yielded a positive fluorescence signal, while no signal was observed in any of the remaining groups, including the negative control (Fig. 1H).
To validate the applicability of the one-pot RT-RAA-CRISPR/ Cas12a assay to real samples and to assess the clinical detection capability, we extracted high-purity (OD 260 /OD 280 = 1.8-2.0) total RNA from HEK 293T cells. Varying concentrations of EBOV RNA were spiked into the total RNA, and these samples were used to evaluate the one-pot RT-RAA-CRISPR/Cas12a assay. Based on the sensitivity results, 40 min was chosen as the reaction time for the one-pot RT-RAA-CRISPR/Cas12a assay to avoid false negatives. Fluorescence image acquisition was performed at the end of the reaction and was analyzed using ImageJ, and 57.1% (40/70) of the EBOV-positive samples tested positive (Fig. 1I). At the same time, a real-time RT-PCR method was performed to assess the same spiked RNA samples targeting the EBOV N gene, with Ct values between 13 and 31 for positive samples. Of these, 57.1% (40/70) of the EBOV-positive samples tested positive (Fig. 1J). The real-time RT-PCR method corroborated the results of our RT-RAA-CRISPR/Cas12a assay, identifying the same number of positive samples. The real-time RT-PCR has been recommended by the WHO as the standard diagnostic method for EBOV. Therefore, we selected real-time RT-PCR as the reference method for comparison. The results showed that the two methods had good consistency.
Most current isothermal amplification technologies for EBOV, as well as existing one-pot methods, are limited to a detection limit of 10 1 copies/μL (Bonney et al., 2020). Furthermore, many of these one-pot approaches suffer from relatively long detection times, even where they achieve consistent sensitivity (Xiong et al., 2022). However, the method described here achieves a detection limit of 3.6 copies/μL linearized plasmid. A major advantage of this method is its high specificity. RT-RAA, due to its low amplification temperature and longer primer sequences, is highly prone to non-specific amplification. In contrast, this method further enhances specificity by performing gRNA-specific recognition of the PAM site after RT-RAA amplification. This method minimizes the reaction volume of RT-RAA and Cas12a, significantly reducing the amount of RT-RAA reagents used. The detection time of this method is slightly longer than that of the pure RT-RAA detection method, but the entire process is completed within 40 min, making it suitable for home testing or field deployment.
While placing the Cas12a reaction system in a PCR tube enables onepot detection by simplifying the setup, it introduces a risk of premature mixing with the RAA system if the tube is accidently inverted. This can cause false-negative results, but can be avoided by carefully ensuring the liquid does not spill before centrifugation and mixing. Furthermore, visualizing low-copy samples can be difficult due to their weak fluorescence signals, which requires a dark environment. Therefore, we are working on a new device. This device will integrate a light source and a filter into a single, compact unit to provide the necessary dark environment for on-site visualization.
In conclusion, we have developed an efficient, reliable, and visual one-pot RT-RAA-CRISPR/Cas12a assay. This method effectively addresses the aerosol contamination and operational complexities commonly associated with RT-RAA amplification. The assay can complete detection within 40 min with a sensitivity of a single copy and shows excellent concordance with the gold-standard real-time RT-PCR for both positive and negative results. Therefore, it holds great promise for meeting the on-site detection needs for EBOV RNA. We expect that this technique will be used for the detection of EBOV and can be expanded to the diagnosis of other RNA/DNA viral infections.
## References
1. Broadhurst, Brooks, Pollock (2016) "Diagnosis of Ebola virus disease: past, present, and future" *Clin. Microbiol. Rev*
2. Bonney, Watson, Slack et al. (2020) "A flexible format LAMP assay for rapid detection of Ebola virus" *PLoS. Negl. Trop. Dis*
3. Chams, Chams, Badran et al. (2020) "COVID-19: a multidisciplinary review" *Front. Public Health*
4. Jacob, Crozier, Fischer et al. (2020) "Ebola virus disease" *Nat. Rev. Dis. Primers*
5. Kaushik, Tiwari, Dev Jayant et al. (2016) "Towards detection and diagnosis of Ebola virus disease at point-of-care" *Biosens. Bioelectron*
6. Mukadi-Bamuleka, Mambu-Mbika, De Weggheleire et al. (2018) "Efficiency of field laboratories for Ebola virus disease outbreak during chronic insecurity, Eastern democratic Republic of the Congo" *Emerg. Infect. Dis* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12807731&blobtype=pdf | # Vir us taxonom y: the database of the Inter national Committee on Taxonomy of Viruses
Eden Black, C Powell, Donald Dempsey, R Hendrickson, Logan Mims, Elliot Lefkowitz
## Abstract
Taxonomic classification underlies all biological science and is the basis for comparative analysis of biological organisms and therefore our understanding of life. The International Committee on Tax onom y of Viruses (ICTV) de v elops the official tax onom y f or all viruses. To ensure that the taxonomic data, associated metadata, and analytical tools used to search and visualize those data are easily accessible, the ICTV maintains a comprehensive database and website that provides these resources to the scientific community and the interested public. This report describes the e xtensiv e enhancements made to these resources since our first Nucleic Acids Research Database Issue publication in 2018. These enhancements have focused on improvements to the computational infrastr uct ure supporting the database, website, and tools; expanding the information available on the taxonomy and the viruses classified by that taxonomy; enhancing existing and developing new tools to access, search, and displa y tax onomic data; e xpanding the a v ailable methods and links used to access the taxonomic and associated data; and providing outreach and training opportunities to our users to ensure that these resources are useful and used. The data and tools provided through this effort are available from the ICTV website at https://ictv.global .
## Introduction
The classification of biological organisms requires a logical framework that can be superimposed upon the natural world [ 1 , 2 ]. The science of taxonomy provides such a framework through the classification and naming of biological entities [ 3 ]. Taxonomic classification results in organisms with similar properties organized into groups and placed into hierarchies that define different levels of shared similarities. These hierarchies are the taxonomic ranks (species, genus, family, order, class, phylum, kingdom) defined by the classification process [ 4 ]. The virus taxonomy adds an additional rank, realm, as the highest level in the taxonomy. The biological descriptors that define the taxonomic rank within a particular taxonomic lineage form a unique set of characteristics that identify the organisms classified into that rank and serve, in essence, as a barcode that can be used to identify other potential members of that taxon [ 5 -8 ]. Critically, the process of taxonomic classification provides a starting point-an initial reference-guiding all subsequent research designed to enhance our understanding of the organisms of interest. By comparing the properties D 777 of the organisms classified into one taxon (e.g. species) with those of another, we can begin to delineate the important similarities and differences between the two taxa. For example, by classifying a newly isolated virus into an existing species, we might be able to infer additional properties of the new virus by using the properties of the viruses already classified into that species as a starting point for additional research [ 9 , 10 ]. In this manner, a reiterative process of classification, inference, testing, and potentially reclassification is established that helps to expand our knowledge of the role of the newly isolated viruses in the ecosystem of interest.
Importantly, taxonomic classification is not a static endeavor. Organisms are reclassified as we learn more about their properties and the properties of related organisms, and newly discovered organisms also need to be classified [ 11 ]. Therefore, the organizations that are responsible for overseeing the taxonomy of different biological domains operate under a set of guidelines that determine the rules, policies, procedures, nomenclature, and timelines under which their taxonomy is updated. The organizations that have these responsibilities include the International Commission on Zoological Nomenclature (ICZN; https://www.iczn.org ) [ 12 ] and the International Committee on Systematics of Prokaryotes (ICSP; https://www.the-icsp.org ) [ 13 ], which are responsible for the nomenclature of animal and bacterial taxa, respectively.
The organization responsible for the taxonomy of viruses was established in 1966 and named the International Committee on Nomenclature of Viruses [ 14 -16 ]. It was renamed the International Committee on Taxonomy of Viruses (ICTV) in 1975 [ 17 ]. In contrast to other taxonomic organizations such as the ICZN and ICSP, the ICTV is responsible for both classification and nomenclature. The ICTV is charged by the Virology Division of the International Union of Microbiological Societies (IUMS) with developing, refining, and maintaining the official, universal virus taxonomy [ 18 ]. The ICTV, through its Executive Committee (EC), Subcommittees, and Study Groups, develops both the guidelines for the classification of viruses as well as the guidelines for naming of the resultant taxa. The rules governing the ICTV and its operations are defined by the ICTV Statutes ( https: // ictv.global/ about/ statutes ), while the rules for creating and naming virus taxa are provided in the International Code of Virus Classification and Nomenclature ( https://ictv.global/ about/code ). The Statutes and Code were last updated in March 2025.
Taxonomic classification and naming within the ICTV are overseen by the EC, which is composed of a president, vice president, 3 secretaries, 7 subcommittee (SC) chairs, and 11 elected members ( https:// ictv.global/ members/ ec-members ). For classification, viruses are grouped according to the type of host they infect and the molecular composition of the virus genome, with each group being the responsibility of one of the SCs ( https:// ictv.global/ about/ organization ). Each SC has Study Groups (SGs), usually one for each taxonomic family, whose responsibilities are to determine the unique set of properties that can be used to distinguish the viruses belonging to one species, genus, family, or higher taxonomic rank from those of the other viruses classified into a corresponding rank. These properties comprise the demarcation criteria that are the basis for taxonomic classification. The SGs then prepare and evaluate proposals for making changes to existing taxa and creating new taxa. There are presently 138 SGs covering all currently classified viruses ( https:// ictv.global/ study-groups ).
The ICTV also welcomes taxonomic proposals from any interested individual. Proposal template files and instructions are available at https:// ictv.global/ taxonomy/ templates . Once the relevant files are completed, they should be emailed to the appropriate SC chair. Proposals are then reviewed by the relevant SG and the EC, and any comments or questions are sent to the authors for a response.
Taxonomic proposals for viruses without an established SG are submitted to the SC chair, who oversees the group of viruses covered by the proposal (see https:// ictv.global/ studygroups ). The SC chair will then submit the proposal to the SG with the most relevant expertise. For example, the family Arteriviridae is handled by the Nidovirales SG. Any questions regarding the preparation or submission of proposals, including the demarcation criteria used to support classification, should be sent to the relevant SC chair.
## Current issues in virology
The field of virology has undergone significant changes in the past few decades, including several challenges that impact taxonomic classification. These changes and challenges must be addressed by the ICTV so that the taxonomy remains current and relevant. For example, there have been advancements in metagenomics through the bulk sequencing of mixed populations of viruses and other organisms collected from environmental or host samples [ 19 -21 ]. These sequence data can then be assembled into apparently complete virus genomes that comprise a large number of previously unidentified and therefore unclassified isolates [ 22 -24 ]. At the same time, there has been a shift by the ICTV to use properties directly derived from the complete virus genome sequences (e.g. nucleotide or amino acid sequence similarity, phylogenetic analysis, open reading frames, gene content, and synteny) as the primary basis for classification rather than physical properties of the virion or other phenotypic characteristics [ 25 ]. This approach to classification obviates the need to isolate, grow, and study a physical virus particle and use its properties as the basis for classification. Therefore, metagenomic sequencing and the use of those sequences by the ICTV to support classification represent a profound shift in the requirements necessary to define new taxa. This shift has required that the ICTV update existing demarcation criteria, adopt new techniques and new tools to process the sequence data, and establish robust and scalable databases and query tools to handle the large datasets generated by these processes [ 26 , 27 ]. SGs establish and update, as necessary, the demarcation criteria for the virus taxa under their purview. The bioinformatics tools used for classification are also described as part of the demarcation criteria and are specific to each family. The necessity for creating new or changing existing demarcation criteria will usually occur when new viruses are discovered, expanding the properties of the viruses to be classified. These criteria may also be updated when new tools become available that have been shown to do a better job classifying a family's viruses into distinct taxa.
Other challenges derive from the emergence and reemergence of viral diseases and the potential for the rapid spread of new viral pathogens as they are introduced into new hosts, evolving in real-time to better adapt to those hosts. Through the procedures outlined above that support taxonomic clas-
sification, the ICTV can assist in the worldwide response to new outbreaks of virus disease by helping to determine the classification of novel pathogens, their relationship to known viruses, and what inferences about form, function, and action might reasonably be made between known and unknown viruses based on these comparisons.
$$D 778 Black et al .$$
## Recent developments
To respond to these needs, we submitted and received a grant award from the U.S. National Institute of Allergy and Infectious Diseases of the National Institutes of Health to modernize and extend the robustness and functionality of the ICTV data infrastructure. These enhancements to the ICTV database and website have proceeded on the basis of five goals: (i) modernize the information technology infrastructure used to provide all services, moving to an open-source, cloud-based platform; (ii) expand the information provided to understand the properties of the classified viruses and the process used to support classification; (iii) develop easy-to-use software tools to search the taxonomy and associated metadata database and visualize the results; (iv) provide methods to make all of the data programmatically accessible through a variety of computational systems and data repositories; and (v) provide outreach and training assistance to ensure that users can easily access and understand the available information. In our previous report, published in the NAR Database Issues [ 28 ], we described the data, tools, and other resources developed prior to receipt of the NIH award. In this updated report, we describe the significant enhancements that have been made over the subsequent eight years.
## Infrastructure
The ICTV will celebrate its 60th anniversary in 2026, and the current ICTV database and website had their beginnings 18 years ago [ 16 ]. These long-term commitments to virus taxonomy emphasize the need to ensure a stable and reliable, secure, scalable, and sustainable infrastructure that supports the objective of communicating the products of ICTV activities to interested individuals and groups around the world. Previously, the ICTV information technology (IT) infrastructure was locally housed, proprietary, and expensive to purchase and maintain. Our goal was to develop a modern, easily maintained system, built on an open-source foundation, that would meet the needs of the ICTV for the foreseeable future. These goals have been accomplished, and the components of the upgraded systems are provided in Table 1 .
## Information
Our mission, as stated in the ICTV Statutes, is to communicate the decisions reached concerning the classification of viruses and nomenclature of virus taxa to virologists (and other interested groups and individuals) and to maintain an official index of approved names for virus taxa. Therefore, the provision of information on virus taxonomy represents our primary goal. This is accomplished by storing all data and metadata in a relational database and establishing a web-based communication infrastructure. The resources described below have been developed to provide our users with the tools required to access, search, and utilize that information.
## Taxonom y bro wsing
Once taxonomic decisions are ratified by the ICTV membership ( https:// ictv.global/ members ; "About" menu > "Membership"), they become official, and it is then the responsibility of the ICTV to provide these decisions to the scientific community. This is accomplished through the ICTV database, which stores the taxonomy structure, taxon names, and associated metadata. This information is then available online through the ICTV website as a searchable and expandable table through the taxonomy browser ( https: // ictv.global/ taxonomy ; "Taxonomy" menu > "Taxonomy Browser") as well as a downloadable spreadsheet, the Master Species List (MSL; https:// ictv.global/ msl , https:// doi.org/ 10.5281/zenodo.15042255 ; "Taxonomy" menu > "Master Species Lists"). A second spreadsheet, the Virus Metadata Resource (VMR; https:// ictv.global/ vmr , https:// doi.org/ 10. 5281/zenodo.15042309 ; "Taxonomy" menu > "Virus Metadata Resource"), containing a list of virus exemplars for each virus species, is also available. These exemplars provide an example of a well-characterized, sequenced virus isolate for each species. Entries include the virus name, isolate designation, suggested abbreviation, GenBank accession number(s), segment names, genome composition, and host or sample source.
Proposals for new or updated taxonomy that have been submitted and approved are available from the website ( https: // ictv.global/ files/ proposals/ approved ; "Taxonomy" menu > "Proposals" > "Approved Proposals"). Pending proposals submitted, but not yet officially approved, are also available ( https:// ictv.global/ files/ proposals/ pending ; "Taxonomy" menu > "Proposals" > "Pending Proposals"). Proposal template files, along with instructions for filling out and submitting new taxonomic proposals, are available for download ( https:// ictv.global/ taxonomy/ templates ; "Taxonomy" menu > "Proposal Template Files").
## Taxon history
Links from individual taxa in the taxonomy browser, as well as in the MSL and VMR spreadsheets, connect to a taxon details page that provides detailed information on each taxon, including a complete history of all updates to the taxon over time (Fig. 1 ). The history includes all changes to the taxon since it was first approved and links to the approved proposal documents explaining and justifying each change. The following changes are allowed: new, abolished, renamed, moved, lineage updated, merged, split, promoted, or demoted. Information on the etymological origins of the taxon name is also provided for all ranks at the family level and higher. A separate table listing name etymologies for a user-selected list of taxa is also available ( https:// ictv.global/ taxonomy/ etymology ; "Taxonomy" menu > "Taxon Name Etymology"; Fig. 2 ).
## ICTV Report
In-depth information on the viruses classified into a particular taxonomic family, as well as the family's constituent genera and species, is provided by the ICTV Report on Virus Classification and Nomenclature ( https:// ictv.global/ report ; "Report" menu > "Current Report Chapters"). These online report chapters replace the printed hard copy versions of the Report that had previously been published every several years and consist of nine volumes that were released from 1971 to 2011 ( https:// ictv.global/ report/ about ; "Report" menu > "About the ICTV Report") [ 14 , 17 , 18 , 29 -34 ]. The layout The past and current information technology systems used to provide all database, website, and application resources for the ICTV. and content of the current report were substantially upgraded as a part of the infrastructure migration project, as described in Table 1 . Figure 3 provides screenshots from the Coronaviridae Report chapter ( https:// ictv.global/ report/ coronaviridae ; "Report" menu > "Current Report Chapters"), showing elements that are typically present in most chapters, including descriptive information, a table of virus properties, virion diagrams, genome maps, phylogenetic trees, and a table of species with links to the genomic sequence accession number(s) of the exemplar virus(es) for each species. Descriptive information is also provided for the subfamily and genus members of the family.
The report also provides data that can be used for direct comparative analysis of the properties of the classified viruses. These data include the Virus Properties table that displays physical and biological properties of the exemplar viruses within each family, with filterable and sortable options for taxon, genome composition and topology, envelope, virion shape, and host ( https:// ictv.global/ virus-properties ; "Report" menu > "Virus Properties"; Fig. 4 ). In addition, diagrams of the exemplar virions for a family are also provided ( https: // ictv.global/ virion-diagrams ; "Report" menu > "Virion Diagrams"; Fig. 5 ). These diagrams are courtesy of ViralZone ( https:// viralzone.expasy.org/ ) [ 35 ].
## Tools
A variety of tools are provided on the ICTV website to enhance a user's ability to search for a taxon of interest and visualize its taxonomic classification. New tools, recently made available, are the Visual Taxonomy Browser, Find the Species search tool, and TaxaBLAST.
## Visual Taxonomy Browser
The visual browser ( https:// ictv.global/ visual-browser ; "Taxonomy" menu > "Visual Taxonomy Browser"; Fig. 6 ) provides a graphical representation of the taxonomy. In contrast to the original browser, which displays the taxonomy in a tabular format with expanding rows, the visual browser displays the taxonomy in an animated, branching tree. This format, although presenting the same information as the original browser, allows for better visualization of the relationships be-tween different taxonomic lineages. Other features of the visual browser include a drop-down list from which one may select a historical MSL release as the basis for a search, font size and zoom controls, and the ability to export images as PNG, SVG, or PDF files for use in publications.
## Find the SPECIES
One of the most frequently requested features for the ICTV website has been a tool that would allow users to input different names that have been associated with a virus and return the current, official species name. Find the Species is a new tool that provides this functionality ( https://ictv.global/ find-the-species ; "Search" menu > "Find the Species"; Fig. 7 ). The official ICTV species name for a virus will be returned from a query that matches all or part of a virus or virus isolate name, a past species name, or a disease name. Name synonyms, common names, acronyms, abbreviations, etc. can also be used. Selectable search parameters include finding exact matches, matches to all words or any words entered (in any order), or substring matches. Find the Species uses current and past databases from the ICTV (MSL and VMR lists), National Center for Biotechnology Information (NCBI), and the Disease Ontology to make the connection between the name entered as the search query and an ICTV taxon name. Results are dependent on identifying a match in one of these databases and determining the most recent ICTV virus taxon based on that match. Since its release, Find the Species has become the third-most accessed page of the ICTV website.
## TaxaBLAST
When a new virus is isolated, one of the first questions that arises is which already described virus is most closely related to the newly discovered virus. In essence, the discoverer of the new virus wants to know the taxonomic placement of their virus. The virus could be classified as a member of an existing species, in which case its complete taxonomic hier- archy is already defined. Or it could represent a new species or a new higher-level taxonomic rank. If this is the case, a proposal should be submitted to the ICTV proposing that a new taxon or taxonomic hierarchy be created. The properties that are used to distinguish one taxon from another are the demarcation criteria that are defined by each ICTV SG for each taxonomic rank within the family covered by the SG. The demarcation criteria can be complex, and the process of determining taxonomic placement can require the use of multiple, not always easy to use, bioinformatics tools. To provide an approximate answer to the question of taxonomic place-ment, we developed TaxaBLAST ( https:// ictv.global/ search/ T axaBLAST ; "Search" menu > "T axaBLAST") that provides a quick assessment of the taxonomic placement for a newly isolated virus based on its partial or whole genome sequence. It should not be used as justification for classification in a taxonomy proposal, but it can provide the user with useful information to guide subsequent analysis.
TaxaBLAST uses a BLAST [ 36 , 37 ] database that is created from the genomic sequences of all virus exemplars listed in the VMR. When a user enters their query nucleotide sequence into the TaxaBLAST interface, a blastn search using default D 785 Figure 7. Find the Species. A tool that will provide the current taxon name (species or higher taxonomic rank) for a virus when a full, partial, common, or disease name is entered into the search box ( https:// ictv.global/ search/ find _ the _ species ).
parameters is run comparing this sequence with the exemplar database (Fig. 8 ). The application returns a summary of the BLAST hits along with an HTML file containing the sequence alignments and a CSV file containing the BLAST hit statistics. By examining the list of hits returned and their BLAST bit scores, a user is provided with an indication of the species (and higher taxon) with the highest sequence similarity to their virus sequence.
## Accessibility
ICTV data are made available in a variety of formats for multiple audiences: interactive web visualizations and downloadable spreadsheets and documents (the MSL and VMR) for the typical user, and raw, software-consumable data via web services and database exports for the technical user. Registration and logging in to the website are not required to access any of the available data or resources. Registration will add a user to the ICTV email list for receipt of newsletters and advanced announcements of webinars and other events. The Drupal theme that provides the template for accessing all data and tools from the website uses responsive web design, so that all pages are accessible and legible on phone and tablet screens.
We have partially implemented [ 38 ] and maintain stable, unique IDs for all taxa, as well as release-specific IDs for historical versions of taxonomic records. DOIs for data releases are generated by uploading the MSL and VMR spreadsheets to the data repository Zenodo, with each release assigned a DOI to provide a stable access point for these data ( https:// zenodo.org/ communities/ ictv/ ). We also maintain stable identifiers for the exemplar virus isolate assigned to each species (in the VMR), which are linked to the NCBI accession number of each genomic sequence (and sequence segment for viruses with segmented genomes). All of these ICTV IDs provide stable links to access taxon-specific data (Table 2 ). The text and numerical identifiers that specify ICTV taxonomy releases and IDs used to identify current and historical taxa.
The application code, database schema, and data dumps of all database tables that make the data programmatically accessible are available from public GitHub repositories ( https: // github.com/ ICTV-Virus-Knowledgebase ).
## Outreach and training
One of the goals of the ICTV is to establish outreach and training programs that introduce and explain the resources of the ICTV website to current and potential users. This is accomplished through help resources provided on the website, webinars, and presentations at scientific meetings.
Instructions for using individual tools and explaining available datasets are provided on the web page of each tool and dataset. Videos explaining how to use the tools and how to find information are available from the Help menu. These how-to videos cover topics such as how to use the tabular Taxonomy Browser, the Visual Taxonomy Browser, and the Find the Species tools. They also cover how to use the MSL and VMR, and the ICTV Report. Information on the history and organization of the ICTV is available from the About menu. User questions, suggestions, and bug reports can be sent via email to the address info@ictv.global . Emails sent to this address are received by the ICTV Data Secretary, who usually provides a response within 24 h.
To provide additional user training, we host webinars focused on using the ICTV website. These webinars have been used to present an overview of the tools and resources available on the website and to demonstrate how to use tools and resources to find information about viruses and their taxonomy. Recordings of all webinars, along with questions from the attendees and answers to those questions, can be accessed under the Help menu of the website. New webinars focused on particular topics related to virus taxonomy (such as submitting taxonomic proposals) or providing instruction on the use of new website or tool features will be offered on a regular basis.
Additional material is available from the Information menu of the website, including EC meeting reports, plenary session reports from the triennial IUMS Virology Congresses, and newsletters.
Finally, to extend its outreach, the ICTV has hosted an exhibit booth at the American Society for Virology meeting over the past several years. At these meetings, all the data, tools, and other resources provided by the website are demonstrated, questions answered, and suggestions for future improvements solicited.
## Futur e dir ections
Future enhancements to the ICTV database and web infrastructure will focus on providing additional support for taxonomic classification of viruses, enhancing the user interface and tools for data access and discovery, enhancing the current application programming interface (API), and ensuring the sustainability of the resource into the future. A few specific goals are as follows:
## Support for taxonomic classification
As described above, classification of viruses into taxonomic ranks depends on using taxon-specific demarcation criteria as the basis for determining the correct classification [ 8 , 39 ]. Currently, these demarcation criteria are defined and published in various locations, including Report chapters, taxonomic proposals, and the published literature. This makes these criteria difficult to find and to determine which source provides the most current information. Therefore, we will be creating a demarcation criteria database to collect and store the pertinent properties and tools used by each individual SG to classify the viruses belonging to the family under their purview. This database will be of significant assistance to the classification of viruses assembled from sequenced metagenomic datasets generated by programs such as the NIH Common Fund-supported Human Vi- rome Program ( https:// commonfund.nih.gov/ humanvirome ) and other metagenomic virus discovery sequencing projects [ 19 , 26 , 40 -43 ].
## Data access
While the data access features of the ICTV website provide extensive search and display capabilities to support the needs of the typical site user, support for the technical user or support for computational access needs to be improved. Therefore, we will continue to improve, enhance, and extend our API to enable more customized and programmatic access to the ICTV database and tools by data consolidators, public data repositories, and advanced users. This effort will also result in broader alignment with FAIR (findability , accessibility , interoperability, and reusability) principles [ 38 ].
## Sustainability
Sustainability of the product of ICTV activities has been a primary goal since the inception of the ICTV in 1966. This is critical given that, as defined by the ICTV Statutes, the membership of the EC changes every three years, and there are strict time limits on how long any individual can serve in each position. Since 1966, the overall membership of the EC has changed 19 times. This provides opportunities for new generations of virologists to participate in and determine the processes used for the creation of the next taxonomy release. But it presents challenges in maintaining the systems and methods used to store and communicate the taxonomy. Therefore, sustainability must be an integral property built into the activities and products of the ICTV and the requirement to communicate its decisions regarding classification and nomenclature and maintain an index (database) of the resulting taxonomy. The resources described in this manuscript represent our initial attempt to fulfill these requirements. The effort expended to maintain a complete history of the virus taxonomy, along with the ability to search and visualize this taxonomic history, also contributes to information sustainability. However, as the size of the taxonomy database grows substantially in the future, and as the resources developed become greater in number and in complexity, the efforts to sustain these resources become even more complex and critical. Therefore, a major goal of our work in the future will be to focus on automation, developing computational pipelines to make it easier to update each individual resource, both when a new or updated taxonomy is published, and as new metadata describing the newly classified viruses is made available. Only in this manner, we will be able to ensure that the 60-year legacy we inherited will be passed on once again to those virologists who will extend it into the future.
## References
1. Van Regenmortel (2006) "Virologists, taxonomy and the demands of logic" *Arch Virol*
2. Van Regenmortel (2003) "Viruses are real, virus species are man-made, taxonomic constructions" *Arch Virol*
3. Mayr (1953) "Concepts of classification and nomenclature in higher organisms and microorganisms" *Ann NY Acad Sci*
4. (2020) "International Committee on Taxonomy of Viruses Executive Committee. The new scope of virus taxonomy: partitioning the ICTV database D 789 virosphere into 15 hierarchical ranks" *Nat Microbiol*
5. Bao, Amarasinghe, Basler (2017) "Implementation of Objective PASC-Derived Taxon Demarcation Criteria for Official Classification of Filoviruses" *Viruses*
6. Buchen-Osmond (1997) "Further progress in ICTVdB, a universal virus database" *Arch Virol*
7. Buchen-Osmond, Dallwitz (1996) "Towards a universal virus database-progress in the ICTVdB" *Arch Virol*
8. Lefkowitz (2015) "Manual of Clinical Microbiology . 11th edn"
9. (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 S AR S-CoV-2" *Nat Microbiol*
10. Gorbalenya, Siddell (2021) "Recognizing species as a new focus of virus research" *PLoS Pathog*
11. Zerbini, Crane, Kuhn (2025) "Summary of taxonomy changes ratified by the International Committee on Taxonomy of Viruses (ICTV)-general taxonomy proposals" *J Gen Virol*
12. Melville (1995) "Towards Stability in the Names of Animals : a History of the International Commission on Zoological Nomenclature"
13. Oren, Arahal, Goker (2023) "International Code of Nomenclature of Prokaryotes. Prokaryotic Code (2022 Revision)" *Int J Syst Evol Microbiol*
14. Wildy (1971) "Classification and Nomenclature of Viruses: First Report of the International Committee on Nomenclature of Viruses"
15. Fenner (1971) "The nomenclature and classification of viruses the International Committee on Nomenclature of V iruses" *irology*
16. Wildy, Ginsberg, Brandes (1967) "Virus-classification, nomenclature and the International Committee on Nomenclature of Viruses" *Prog Med Virol*
17. (1976) "Classification and Nomenclature of Viruses: Second Report of the International Committee on Taxonomy of Viruses"
18. King, Adams (2012) "Virus Taxonomy: Classification and Nomenclature of Viruses: Ninth Report of the International Committee on Taxonomy of Viruses"
19. Dutilh, Varsani, Tong (2021) "Perspective on taxonomic classification of uncultivated viruses" *Curr Opin Virol*
20. Simmonds, Adams, Benko (2017) "Consensus statement: virus taxonomy in the age of metagenomics" *Nat Rev Micro*
21. Zhang, Shi, Holmes (2018) "Using metagenomics to characterize an expanding virosphere" *Cell*
22. Williamson, Allen, Lorenzi (2012) "Metagenomic exploration of viruses throughout the Indian Ocean" *PLoS One*
23. Rose, Constantinides, Tapinos (2016) "Challenges in the analysis of viral metagenomes" *Virus Evol*
24. Nurk, Meleshko, Korobeynikov (2017) "metaSPAdes: a new versatile metagenomic assembler" *Genome Res*
25. Adriaenssens, Roux, Brister (2023) "Guidelines for public database submission of uncultivated virus genome sequences for taxonomic classification" *Nat Biotechnol*
26. Simmonds, Adriaenssens, Zerbini (2023) "Four principles to establish a universal virus taxonomy" *PLoS Biol*
27. Siddell, Smith, Adriaenssens (2023) "Virus taxonomy and the role of the International Committee on Taxonomy of Viruses (ICTV)" *J Gen Virol*
28. Lefkowitz, Dempsey, Hendrickson (2018) "Virus taxonomy: the database of the International Committee on Taxonomy of Viruses (ICTV)" *Nucleic Acids Res*
29. Matthews (1979) "Third report of the International Committee on Taxonomy of Viruses. Classification and nomenclature of viruses" *Intervirology*
30. Matthews (1982) "Classification and nomenclature of viruses. Fourth report of the International Committee on Taxonomy of Viruses" *Intervirology*
31. (1991) "Virus Taxonomy: Classification and Nomenclature of Viruses: Fifth Report of the International Committee on Taxonomy of Viruses"
32. Murphy, Fauquet (1995) "Virus Taxonomy: Classification and Nomenclature of Viruses: Sixth Report of the International Committee on Taxonomy of Viruses"
33. (2000) "Virus Taxonomy: Classification and Nomenclature of Viruses: Seventh Report of the International Committee on Taxonomy of Viruses"
34. Fauquet, Mayo (2005) "Virus Taxonomy: VIIIth Report of the International Committee on Taxonomy of Viruses"
35. Hulo, De Castro, Masson (2010) "ViralZone: a knowledge resource to understand virus diversity" *Nucleic Acids Res*
36. Altschul, Gish, Miller (1990) "Basic local alignment search tool" *J Mol Biol*
37. Altschul, Madden, Schaffer (1997) "Gapped BLAST and PSI-BLAST: a new generation of protein database search programs" *Nucleic Acids Res*
38. Wilkinson, Dumontier, Aalbersberg (2016) "The FAIR Guiding Principles for scientific data management and stewardship" *Sci Data*
39. Van Regenmortel, Bishop, Fauquet (1997) "Guidelines to the demarcation of virus species" *Arch Virol*
40. Zolfo, Silverj, Blanco-Miguez (2024) "Discovering and exploring the hidden diversity of human gut viruses using highly enriched virome samples" *bioRxiv*
41. Matthijnssens, Adriaenssens (2022) "Editorial overview: the virome in health and disease" *Curr Opin Virol*
42. Espinoza, Dupont (2022) "VEBA: a modular end-to-end suite for in silico recovery, clustering, and analysis of prokaryotic, microeukaryotic, and viral genomes from metagenomes" *BMC Bioinf*
43. Liang, Bushman (2021) "The human virome: assembly, composition and host interactions" *Nat Rev Micro* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282451&blobtype=pdf | # A mouse model of Crimean-Congo hemorrhagic fever virus-induced coagulopathy
Hui Zhang, Ziyang Jiang, Haidang Liao, Jiang Li, Manli Wang, Yiwu Zhou, Zhihong Hu, Jia Liu, Virologica Sinica
## Abstract
Crimean-Congo hemorrhagic fever (CCHF), caused by the CCHF virus (CCHFV), is a severe tick-borne illness with a wide geographical distribution, posing a significant threat with case fatality rates ranging from 5% to 70% (Hawman and Feldmann, 2023). Due to the lack of approved vaccines and therapeutics, the World Health Organization (WHO) has listed CCHF as one of the priority diseases (Semper et al., 2024). CCHF initially presents as a nonspecific febrile illness, characterized by fever, malaise, myalgia, and nausea, which can rapidly progress to hemorrhagic disease. The hemorrhagic stage is particularly pronounced in severe cases, with rapid progression to disseminated intravascular coagulation (DIC), overt bleeding, kidney or liver failure, and shock (Frank et al., 2024). Up to date, there is an absence of a suitable animal model that can accurately mimic the coagulopathy and bleeding associated with CCHFV infection. Consequently, our understanding of the pathogenic mechanisms underlying these conditions remains limited (Rodriguez et al., 2022).In this study, we employed mice deficient in the signal transducer and activator of transcription 1 (STAT1) protein, a pivotal component in all three interferon signaling pathways, to establish a mouse model of CCHFV-induced coagulopathy. Previously, STAT1 À/À mice have been shown to be susceptible to CCHFV infection and the infected mice could develop leukopenia and thrombocytopenia. However, whether they exhibit coagulopathy remains unknown (Bente et al., 2010). We first determined the 50 % lethal dose (LD 50 ) of CCHFV in STAT1 À/À mice. Mice were intraperitoneally inoculated with different doses (0.01-1000 TCID 50 ) of CCHFV strain YL16070 (Guo et al., 2017), and were monitored daily. Mice exhibiting a weight loss exceeding 10 %, along with two or more symptoms such as ruffled fur, hunched posture, lethargy, curling up, or shaking, were humanely euthanized for sample collection. Mice infected with high doses (1000, 100, and 10 TCID 50 ) of CCHFV showed symptoms at 3 days post-infection (d.p.i.), including weight loss (Fig.
and behavioral changes such as ruffled fur or lethargy. By 5 d.p.i., body weight continued to decrease, with a terminal mean weight loss of approximately 15 %, and these mice experienced 100 % mortality (Fig. 1A andB). Mice infected with low doses (1, 0.1, and 0.01 TCID 50 ) of CCHFV began to lose weight at 4 d.p.i. and experienced dose-dependent mortality rates. The LD 50 of CCHFV in STAT1 À/À mice was determined to be 0.07 TCID 50 , which is significantly lower than the previously reported LD 50 (4 PFU) of CCHFV strain IbAr 10,200 in STAT1 À/À mice (Bente et al., 2010), this may be related to differences in virus strains and methods of titer determination. The 1000 TCID 50 dose was used for the following experiments.
Certain clinical markers such as thrombocytopenia, leukopenia, and elevated levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and lactate dehydrogenase (LDH) are hallmarks and predictors of fatality of severe CCHF in humans (Bozkurt and Esen, 2021;Frank et al., 2024). Consequently, blood samples were collected at 5 d.p.i. from mice infected with 1000 TCID 50 of CCHFV for routine blood tests and biochemical analyses. The results revealed a significant decrease in platelet and lymphocyte counts in the peripheral blood of infected mice (Fig. 1C). Moreover, the serum levels of ALT and AST increased dramatically in CCHFV-infected mice compared to those in control mice, indicating severe liver dysfunction induced by CCHFV (Fig. 1C).
Tail bleeding assay was performed at 4 d.p.i. with mice infected with 1000 TCID 50 of CCHFV and control mice. Infected mice exhibited significantly prolonged tail bleeding times (all exceeding 10 minutes) compared to control mice, suggesting coagulopathy in the infected mice (Fig. 1D). Physiological hemostasis is a complex process involving blood vessels, platelets, coagulation system, anticoagulation system, and fibrinolytic system. Coagulation tests were performed and the results revealed that both prothrombin time (PT) and activated partial thromboplastin time (APTT) were significantly reduced by CCHFV infection at 4 d.p.i., while thrombin time (TT) remained unaffected (Fig. 1E). We further quantified the euglobulin clot lysis time (ELT), which measures the time required to dissolve a clot formed in vitro in the euglobulin fraction of plasma in the absence of plasmin inhibitors. As depicted in Fig. 1F, the ELT was significantly diminished in infected mice. Collectively, these data suggest severe coagulopathy in CCHFV infected mice.
Gross inspection of organs from 1000 TCID 50 CCHFV-infected STAT1 À/À mice at 5 d.p.i. revealed discolored liver and spleen, as well as intestinal hyperemia, while organs from mock-infected controls appeared normal (data not shown). Histopathological analysis using hematoxylin and eosin (H&E) staining showed severe pathology in CCHFV-infected mice (Fig. 1G). The liver exhibited extensive coagulation necrosis of hepatocytes (white asterisk), and the spleen showed white pulp atrophy with boundary disappearance. Additionally, small thrombi in blood vessels (yellow arrows) were observed in both the infected liver and spleen. In the intestine, mucosal epithelial cells showed nuclear contraction. Some infected mice also displayed intestinal hemorrhages (blue arrows) (Fig. 1G). Quantitative RT-PCR also revealed CCHFV could efficiently replicate in the spleen and liver, and viremia was also observed at 5 d.p.i. (Fig. 1H).
In our STAT1 À/À mouse model of CCHFV infection, mice exhibited prolonged tail bleeding time, indicating a coagulopathy. In prior investigations, it was observed that 30% of CCHF patients exhibited DIC characterized by elevated PT and APTT times (Ozturk et al., 2012;Kaya et al., 2014;Hasanoglu et al., 2016). In studies involving CCHFV-infected mice with interferon α/β receptor knockouts, a notable increase in APTT was documented without a corresponding effect on PT (Zivcec et al., 2013). However, in our STAT1 À/À mouse model of CCHFV infection, dying mice displayed shortened PT and APTT, suggesting thrombin activation, which was confirmed with evidence of thrombosis in the liver and the spleen. Concurrently, the significantly prolonged ELT suggests that these mice were also undergoing hyperfibrinolysis. The dual manifestation of thrombin activation and hyperfibrinolysis closely resembles the fibrinolytic DIC observed in cases of severe trauma (Thrombosis and Hemostasis Group, 2017;Song et al., 2022). It seems that although PT and APTT were shortened, the prevailing factors-decreased platelet counts and hyperfibrinolysis-ultimately resulted in a bleeding tendency. This was evidenced by prolonged tail bleeding time and intestinal hemorrhages in some mice. This highlights the complex and dynamic nature of the coagulopathy induced by CCHFV infection. In the current study, we collected samples at 4 or 5 d.p.i. from mice infected with a high dose of CCHFV (1000 TCID 50 ) for analyses. Further research is necessary to determine whether infected mice progress to a consumptive hypocoagulable state, which should involve examining different infection doses and time points.
In conclusion, we have developed and characterized a mouse model for CCHFV-induced coagulopathy. Our results showed that high doses of CCHFV infection in STAT1 À/À mice led to reduced platelet counts, delayed tail bleeding time, shortened PT and APTT, and prolonged ELT. Additionally, histopathology showed thrombus formation in the liver and the spleen of the infected mice, as well as intestinal hemorrhages in some mice. Although the underlying mechanisms remain to be elucidated, this model provides a validated small-animal platform that recapitulates key features of fatal human CCHF, offering a valuable tool for investigating the pathogenesis of CCHFV-induced coagulopathy and hemorrhage.
## FOOTNOTES
Fig. 1. STAT1 À/À mice infected with CCHFV exhibited severe pathology. A, B Male STAT1 À/À C57BL/6J mice of 8 ~12-week-old were intraperitoneally inoculated with 0.01-1000 TCID 50 of CCHFV or mock-infected with PBS. Body weight and behavioral changes of mice were monitored daily. Body weight changes (A) and survival rates (B) of infected and uninfected mice were monitored daily. C Blood samples were collected from control and 1000 TCID 50 of CCHFV infected mice at 5 d.p.i. The platelet and lymphocyte counts were deteced. The serum levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were analyzed. D Tail bleeding time and images of bleeding patterns on filter paper of 1000 TCID 50 of CCHFV infected mice and control mice were determined at 4 d.p.i. E Prothrombin time (PT), activated partial thromboplastin time (APTT), and thrombin time (TT) of control and 1000 TCID 50 of CCHFV infected mice were measured at 4 d.p.i. F Euglobulin clot lysis time (ELT) of control and 1000 TCID 50 of CCHFV infected mice at 4 d.p.i. G Hematoxylin and eosin (H&E) analysis of pathology in the liver, spleen and intestine of control and 1000 TCID 50 of CCHFV infected mice at 5 d.p.i. White asterisk indicates extensive coagulation necrosis of hepatocytes. Yellow arrows showed small thrombi in blood vessels in the liver and the spleen. Blue arrows showed hemorrhages in intestine. H CCHFV S segment copy number in the liver, spleen, and blood of infected mice were detected at 5 d.p.i. The data are presented as the mean AE SD. Statistical significance was determined using student's t-test. ns indicates P > 0.05. *, **, *** and **** indicate P < 0.05, P < 0.01, P < 0.001, and P < 0.0001, respectively.
## References
1. Bente, Alimonti, Shieh et al. (2010) "Pathogenesis and immune response of Crimean-Congo hemorrhagic fever virus in a STAT-1 knockout mouse model" *J. Virol*
2. Bozkurt, Esen (2021) "Association between severity grading score and acute phase reactants in patients with crimean Congo hemorrhagic fever" *Pathog. Glob. Health*
3. Frank, Weaver, Raabe (2024) "Crimean-Congo hemorrhagic fever virus for clinicians-epidemiology, clinical manifestations, and prevention" *Emerg. Infect. Dis*
4. Guo, Shen, Zhang et al. (2017) "A new strain of Crimean-Congo hemorrhagic fever virus isolated from Xinjiang" *Virol. Sin*
5. Hasanoglu, Guner, Carhan et al. (2016) "Crucial parameter of the outcome in Crimean Congo hemorrhagic fever: viral load" *J. Clin. Virol*
6. Hawman, Feldmann (2023) "Crimean-Congo haemorrhagic fever virus" *Nat. Rev. Microbiol*
7. Kaya, Elaldi, Kubar et al. (2014) "Sequential determination of serum viral titers, virus-specific IgG antibodies, and TNFα, IL-6, IL-10, and IFN-γ levels in patients with Crimean-Congo hemorrhagic fever" *BMC Infect. Dis*
8. Ozturk, Tutuncu, Kuscu et al. (2012) "Evaluation of factors predictive of the prognosis in Crimean-Congo hemorrhagic fever: new suggestions" *Int. J. Infect. Dis*
9. Rodriguez, Hawman, Sorvillo et al. (2022) "Immunobiology of Crimean-Congo hemorrhagic fever" *Antivir. Res*
10. Semper, Olver, Warner et al. (2024) "Research and product development for Crimean-Congo haemorrhagic fever: priorities for 2024-30" *Lancet Infect. Dis*
11. Song, Zhang, Zhang et al. (2017) "Consensus of Chinese experts on standardized evaluation of coagulation dysfunction in severe patients" *Hematology Society of Chinese Medical Association*
12. Zivcec, Safronetz, Scott et al. (2013) "Lethal Crimean-Congo hemorrhagic fever virus infection in interferon α/β receptor knockout mice is associated with high viral loads, proinflammatory responses, and coagulopathy" *J. Infect. Dis*
13. Zhang (2025) *Virologica Sinica* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12680965&blobtype=pdf | # Chronic Hepatitis B Virus Infection Acquired Under Cytostatic Treatment in Childhood -Clinical, Virological and Immunological Long-Term Follow-Up
Thomas Baumgarten, | Lehmann, Tina Senff, | Heiko Slanina, Christ Schüttler, Andreas Walker, Wolfram Gerlich, Reinald Repp, Jörg Timm, Dieter Glebe, Markus Metzler
## Abstract
Oncology patients receiving cytostatic therapy used to be at high risk of HBV infection when HBV screening measures were less reliable. Infections acquired under these conditions often persist, like those acquired perinatally or during early infancy. We studied the long-term clinical outcomes, viral characteristics, and virus-specific T-cell immunity of chronic HBV infection acquired during chemotherapy. We examined 16 chronically HBV-infected former paediatric oncology patients who were infected during cytostatic treatment in the 1980s. Patients underwent physical examination, laboratory liver function testing, non-invasive measurement of liver stiffness, and determination of HBV serology and DNA levels. If the material was sufficient, HBV sub-genotype, drug resistance and immune escape mutations, and mutations associated with HBeAg negativity were analysed. The frequency of HBV core-specific CD8+ T cells was measured after in vitro antigen-specific expansion. All but one patient were chronically infected with detectable HBsAg but were HBeAg-negative, mostly with low viraemia. Four patients were under ongoing effective antiviral therapy, and four required treatment initiation due to high viraemia or advanced liver disease. Hepatic effects were predominantly observed in highly viraemic patients. No drug resistance or immune escape mutations were observed. In two highly viraemic patients, basal core promoter and precore region mutations reducing HBeAg expression were identified. HBV core-specific CD8+ T cells were detected in all patients, but their frequency was low. In conclusion, more than 30 years after primary HBV infection was acquired during chemotherapy, the course of infection still resembles that of perinatally acquired infections.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.
With over 300 million people infected worldwide [1], chronic HBV (hepatitis B virus) infection is a persistent global health burden that causes substantial morbidity and mortality through associated complications, particularly liver cirrhosis and hepatocellular carcinoma. Chronic HBV infection is defined as the detectability of HBsAg (hepatitis B surface antigen) for more than six months after infection, indicating failure of immunologic control and clearance of the virus. In contrast to the predominant acute, self-limiting course of HBV infection in adolescents and adults, chronic HBV infection is mainly observed after neonatal infection or in early infancy: Chronicity rates are about 90% in perinatally infected babies, 25%-50% in children infected at 1-5 years of age, and subsequently decline towards < 5% in adolescents and adults [2][3][4][5].
The natural history of chronic HBV infection, acquired perinatally or in early childhood, is characterised by an initial phase of high replication, low inflammation, and HBeAg (hepatitis B e antigen) positivity, which usually lasts until early adulthood. Loss of HBeAg occurs in approximately half of the cases [6], sometimes accompanied by hepatitis flares. Subsequently, HBV replication declines as a sign of partial viral control. However, sustained, complete infection control with seroconversion of HBsAg to anti-HBs (hepatitis B surface antibody), the so-called "functional cure", is a rare event in chronic HBV infection [7]. As immunologic selective pressure increases and HBV is progressively cleared from the body, HBeAg-negative viral variants emerge [8]. Various mutations in the BCP (basal core promoter) and preC (precore) regions have been described as the molecular basis for HBeAg negativity. Although HBV shows relatively high sequence heterogeneity due to replication by reverse transcription, with at least nine different genotypes described to date [9], a certain level of intrahost genetic stability over time in chronic infections has been reported [10].
Several immunologic mechanisms facilitating chronic HBV infection have been described [11]; in particular, virus-specific CD8+ (cluster of differentiation) T lymphocytes play a pivotal role in achieving sustained viral control by cytolytic and noncytolytic (cytokine-mediated) effects. In chronically infected patients, T-cell dysfunction has been well described, possibly as a consequence of exhaustion or anergy in the context of permanent overstimulation of HBV-specific T-cell signalling and upregulation of inhibitory and inactivating pathways [12]. Moreover, the selection of mutations in viral epitopes targeted by CD8+ T cells may further contribute to immune escape [13]. However, the frequency of these cells declines when HBsAg clearance is achieved [14]. Depending on the stage of HBV infection, low frequencies of HBV-specific CD8+ T cells are common in chronically infected patients.
HBV infection presents a significant risk to all patient groups with dysfunctional immune defences, whether due to disease or iatrogenic causes. Oncology patients were among the first in whom HBV was described [15]. There is extensive knowledge of HBV reactivation during pharmaceutical immunosuppression [16] such as cytostatic or immunomodulatory therapy; however, little systematic research and experience exist regarding the course of primary HBV infection acquired under cytostatic treatment. The cohort examined in this study is exceptional, as it consists of former paediatric oncology patients who have acquired a primary HBV infection during their cytostatic treatment, all of them with an identical virus strain. Preexisting older data from this cohort suggest a course similar to that of perinatally acquired infections [17,18]. The current work aims to determine whether the long-term clinical course of this unique group of chronic HBV infections acquired under cytostatic treatment still exhibits the characteristics of a perinatally acquired infection over 30 years later. Furthermore, the emergence of viral mutations under these circumstances and virus-specific Tcell immunity was examined.
## 1 | Materials and Methods
## 1.1 | Patient Population
We included 16 individuals who had been infected with HBV during an outbreak between 1983 and 1986, affecting 74 patients aged 3-18 years who were undergoing cytostatic chemotherapy at the Department of Paediatric Haematology and Oncology in Giessen, Germany [19]. All patients were infected with an identical virus strain (Genotype D1 [20], HBsAg subtype ayw2, EMBL accession number: Y07587 [21]), suggesting a single source of infection and patient-to-patient transmission rather than a transfusion-transmitted infection [22]. Until 1989, 20 children died of their malignancy, with no negative prognostic influence of HBV infection in cases with acute lymphoblastic leukaemia [23]. Five years after termination of cytostatic treatment, > 90% of the surviving patients (49 out of 54) still showed a high replicative, low inflammatory HBeAg-positive chronic HBV infection (formerly termed "immune tolerant carrier state" [24]), while < 10% (5 out of 54) showed HBsAg to anti-HBs seroconversion, indicating functional cure. Remarkably, anti-HBc (hepatitis B core antibody) was not detected in the serum of 20% of the patients during the first follow-up period and was negative on repetitive testing for several years [18]. None of the patients had elevated liver enzyme levels that exceeded the intermittent, slight elevations typically observed during chemotherapy. Liver biopsies taken from 36 patients in 1989 only showed signs of "minimal hepatitis" in some cases [18].
Figure 1 provides an overview of the cohort and recruitment process. We were able to contact 29 former survivors via the German Childhood Cancer Registry. Seventeen (59%) agreed to participate, of which 16 were included in the study. The 17th person was one of the five survivors who initially had anti-HBs seroconverted; i.e., they had experienced an acute HBV infection and never developed a chronic infection.
Written consent was obtained from all participants, and all research was conducted in accordance with both the Declarations of Helsinki and Istanbul. The study protocol was approved by the ethical review committee of Friedrich-Alexander University, Erlangen-Nuremberg (Reference No: 126_19 B).
## 1.2 | Clinical Follow-Up
All patients were interviewed, and pre-existing medical findings were collected from the patients and attending physicians. Serum tests were performed in Erlangen and Giessen. Liver enzymes, bilirubin, albumin, plasma coagulation markers (international normalised ratio and activated partial thromboplastin time), electrolytes (Na + , K + , Ca 2+ , Cl -, PO 4 -, and Mg 2+ ), and HBV markers were tested. In the presence of HBsAg, HBV DNA levels were also determined (see 1.3 Virologic assessment). Furthermore, we determined the IgG antibody titres against HCV (hepatitis C virus) and HEV (hepatitis E virus) and total antibody titres (IgG and IgM) against HDV (hepatitis D virus). In the case of positive HEV IgG, IgM antibody and RNA levels were determined to rule out acute infection. Subsequently, 15 patients were medically examined by hepatologists at the departments of Internal Medicine of the University Hospital Erlangen, Germany or University Hospital Giessen, Germany, including abdominal ultrasonography with non-invasive measurement of liver stiffness using Acoustic Radiation Force Impulse imaging (Acuson S2000 Virtual Touch tissue quantification, Siemens Healthineers). The cut-off values for the application of the METAVIR (Meta-analysis of Histological Data in Viral Hepatitis) fibrosis score [25] to elastography results were adopted from Friedrich-Rust et al. [26] For the findings of abdominal ultrasonography, we defined two categories: "steatosis" and "parenchymal damage" (i.e., signs of fibrosis or cirrhosis).
The phase of chronic HBV infection was determined using serological HBV parameters and ALT (alanine transaminase) levels, and, if available, a non-invasive assessment of liver disease, following the nomenclature introduced in 2017 by the EASL (European Association for the Study of the Liver) [24]. The indications for antiviral treatment were defined according to the current clinical practice guidelines of the EASL [24].
## 1.3 | Virologic Assessment
The qualitative determinations of HBsAg (HBsAg Qualitative II, LoD (Limit of detection) 0.02 IU/mL), HBeAg, anti-HBc (Anti-HBc II, LoD 0.5 PEI U/mL), anti-HBe (hepatitis B e antibody) as well as the quantitative determinations of anti-HBs (LoD 0.98 IU/L) and HBsAg (LoD 0.05 IU/mL; carried out only if qualitatively tested positive) were performed by commercial chemiluminescent microparticle immunoassay (ARCHITECT System, Abbott). Anti-HDV total antibodies were tested using a chemiluminescent immunoassay (LIAISON murex Anti-HDV, DiaSorin), anti-HCV virus total immunoglobulin with the ARCHITECT System (Abbott), and anti-HEV IgG and IgM antibodies were determined using ELISA (HEV ELISA 4.0 and HEV IgM ELISA 3.0, MP Biomedicals). HEV RNA was quantitatively assayed using reverse transcription PCR (RealStar HEV RT-PCR Kit 2.0, Altona). HBV DNA was quantified using realtime PCR (LoD < 10 IU/mL, 1 IU/mL = 5.4 genomes/mL, linear range 20 to 1.7x10 8 IU/mL, Cobas TaqMan HBV Test, Roche).
If the DNA concentration was ≥ 35 IU/mL, a 621 bp fragment of the HBV genome spanning the S and polymerase gene regions (s74 to s226 and rt83 to rt288, respectively) was amplified by PCR as previously described [27] for Sanger sequencing (Eurofins Genomics) and subsequent genotyping and mutational analysis. The 621 bp fragment allows screening for immune escape and drug resistance mutations, as it encodes the major hydrophilic region of the S protein, including the a determinant, which is believed to be a main target of the antibody-mediated immune response, and the main part of the reverse transcriptase domain of the DNA polymerase protein, which is the target of nucleos(t)ide analogues used in HBV therapy. Genotyping was performed based on the aforementioned DNA sequence using the Geno2Pheno online tool (https:// www. geno2 pheno. org/ ). The polymerase and S gene overlapping open reading frames were analysed for known drug resistance mutations [28] and immune escape mutations [27], respectively. To determine the molecular basis of HBeAg negativity, we sequenced the BCP and preC regions to characterise mutations known to reduce or abolish HBeAg expression. HBV relaxed circular DNA from virions in serum samples was purified using the innuPREP Virus DNA/ RNA kit (Analytik Jena) and then amplified by hemi-nested PCR spanning the BCP, preC, and entire Core open reading frame using a primer set modified from Sato et al. [29] (outer forward primer 5′-CAT AAG AGA CTC TTG GAC T-3′, inner forward primer 5′-TGT CAA CGA CCG ACC-3′, and reverse primer 5′-GGA AAR GAD GGD GTT TDC C-3′) and Platinum Superfi II Polymerase (Thermo Fisher Scientific). The PCR products were subjected to agarose gel electrophoresis and gel extraction (Nucleospin Gel Extraction Kit, Takara Bio), followed by Sanger sequencing (LGC genomics). If no amplificates were obtained from the PCR outlined above, in a second attempt, the reverse primer was exchanged (5′-AGA CTC TAA GGC TTC CCG-3′), resulting in shorter amplificates spanning only BCP, preC, and one-third of the Core open reading frame. In a third attempt, DNA was subjected to nested PCR as described by Sato et al. [29]
## 1.4 | HBV-Specific CD8+ T-Cell Response
The HBV core-specific reactivity of cytotoxic T cells was tested in 13 patients; no samples could be obtained from the remaining three patients. IFN-γ (interferon γ) secretion by CD8+ T cells was measured using flow cytometry after in vitro antigenspecific expansion and intracellular cytokine staining as previously described [30]. Briefly, peripheral blood mononuclear cells were isolated from EDTA-anticoagulated blood samples by density gradient centrifugation (ROTISep 1077 human; Carl Roth GmbH, Germany), conserved in 40% fetal bovine serum (FBS), 40% RPMI, and 20% DMSO, and stored at -80°C until further analysis. A total of 21 18mer peptides (EMC, Tübingen, Germany) covering the precore/core region of HBV genotype D were divided into four pools and used for antigen-specific stimulation of CD8+ T cells. A CEF (human cytomegalovirus, Epstein-Barr virus, and influenza A virus) peptide pool (peptides & elephants, Germany) was used as a positive control. After thawing, the cells were suspended in complete medium R10 (RPMI 1640 containing 10% fetal calf serum, 100 U/mL penicillin, 100 μg/mL streptomycin, and 10 mmol/L HEPES buffer) for stimulation with HBV core peptides (1 μg/mL) in the presence of IL-2 (25 U/mL, Roche, Germany) and anti-CD28/CD49d (0.5 μg/ mL, BD Biosciences, Germany). Fresh medium and IL-2 were added after 7 days. After 14 days of antigen-specific expansion, CD8+ T cells were re-stimulated with the same peptide pool and tested for IFN-γ secretion using flow cytometry on a FACSCanto (BD Biosciences, Germany) after intracellular cytokine staining with anti-human IFN-γ (BD Biosciences, Germany). The percentage of IFN-γ+ CD8+ T cells detected in the unstimulated culture (negative control) was subtracted from the results of the stimulated culture to obtain the actual percentage of IFN-γ+ CD8+ T cells reactive to the CEF peptide pool or the different HBV peptide pools. A T-cell response was considered positive when the frequency of IFN-γ+ T cells was > 0.1% and at least 3-fold above the background.
## 2 | Results
An overview of clinical and virological findings is given in Table 1.
## 2.1 | Clinical Follow-Up
Current clinical practice guidelines from 2017 [24] recommend regular checks for ALT and HBV DNA levels, and liver ultrasonography. The assessment of clinical monitoring in our cohort showed that only six patients (38%) had been continuously monitored in recent years. Of the remaining 10 patients, four had at least received regular ALT and/or HBV DNA determinations, while six underwent no follow-up at all. From a treatment status perspective, the four patients who were under antiviral treatment at the start of the study had an almost seamless follow-up, with ultrasonography missing in only one case. In contrast, only three of the 12 untreated patients had a complete recent follow-up.
Most patients (n = 12; 75%) presented with HBeAg-negative chronic HBV infection, including four patients who reached this inactive phase of HBV infection by receiving antiviral treatment; three (19%) showed an HBeAg-negative CHB (chronic hepatitis B), and one (6%; patient #16) showed functional cure. An overview of participants' HBV infection status is shown in Figure 2.
Four patients (25%) of the cohort met the criteria for the initiation of antiviral therapy; among them were all three patients with HBeAg-negative CHB (#01-03) plus one (#14) with previously known severe fibrosis and oesophageal varices due to severe ethanol abuse for 18 years. Assessment of liver disease revealed relevant disease activity in all three patients with HBeAg-negative CHB: One patient (#03) showed laboratory signs of an acute hepatitis flare with ALT elevation to 44-fold the upper limit of normal, but no signs of structural damage to the liver on ultrasonography or elastography. The other two patients only had mild ALT elevation (below twice the upper limit of normal), but both showed structural changes in the liver, i.e., steatotic parenchymal damage on ultrasonography or severe fibrosis on elastography (median ≥ 1.55 m/s, METAVIR F ≥ 3), respectively. All three patients had very high levels of HBV DNA (> 10 6 IU/mL), in contrast to low (n = 8; < 2,000 IU/mL) or below the detection limit (n = 4; < 10 IU/mL) levels in all other patients. In the remaining 13 patients (of whom ultrasonography and elastography could only be obtained in 12 patients), signs of marked liver disease activity or damage to the liver were present in only two patients (#10 and #14), and three (#06, #07, and #16) exhibited slight sonographic changes (steatosis). These findings were clinically assessed as either due to lifestyle factors rather than HBV infection or were previously known and stable: One patient (#14) qualified for the initiation of antiviral therapy due to ethanol toxic damage to the liver with signs of severe fibrosis on elastography (median ≥ 1.55 m/s, METAVIR F ≥ 3). Although she had no laboratory signs of chronic hepatitis (HBsAg weakly positive 40 IU/mL, HBV DNA undetectable, normal ALT) at the time of assessment, thereby formally not meeting the criteria for antiviral therapy according to the EASL guidelines, antiviral therapy was recommended. The reason for this clinical decision was that the patient had detectable HBV DNA (28 IU/mL) only 15 months earlier, thereby meeting the treatment criteria. The watch-andwait approach with regular monitoring provided no realistic alternative considering the patient's poor compliance. The second patient with relevant liver disease (#10) showed steatosis and signs of significant fibrosis (median ≥ 1.34 m/s, METAVIR F ≥ 2) and slight ALT elevation below twice the upper limit of normal.
In this case, fibrosis was diagnosed histologically more than 12 years ago in the context of an acute hepatitis flare and has shown no progression under antiviral therapy since then. The other three patients with mild steatosis on ultrasonography were overweight (#16, BMI ≥ 25.5 kg/m 2 ), obese (#07, BMI ≥ 30 kg/ m 2 ), or consumed relevant amounts of alcohol (#06, 40 g/d).
## 2.2 | Virologic Assessment
All patients tested positive for anti-HBc. 13 patients were HBsAg-positive and anti-HBs negative. Patients #05 and #06, with low HBsAg levels (18 and 7 IU/mL, respectively), showed parallel detection of anti-HBs at 4.7 and 2.4 IU/L. This level of < 10 IU/L is considered non-protective and negative by most authors, but reveals a low level of active B-cell response insufficient to saturate the circulating HBsAg. Patient #16 tested positive for anti-HBs at 84 IU/L and negative for HBsAg and HBeAg, indicating a functional cure. HBV DNA levels ranged from negative to very high: Four patients had levels below the detection limit of < 10 IU/mL, three of whom were under antiviral treatment. One patient under antiviral treatment had a positive HBV PCR result but below the lower limit of quantification of < 20 IU/mL (#07), seven patients had low HBV DNA levels between 38 and 849 IU/ mL (all treatment-naive), and three patients presented with high HBV DNA levels between 1.16x10 6 and 4.10x10 7 IU/mL. HBV DNA was not tested in the patient who showed functional cure serologically. All 15 patients without a functional cure were HBeAg-negative, 11 of whom also tested positive for anti-HBe antibodies. Of the four patients without concomitant anti-HBe positivity, three were under antiviral treatment, and the fourth was highly viraemic, fulfilling the criteria for the start of antiviral therapy. Regarding other hepatotropic viruses, all patients tested negative for HCV, HDV, and HEV, except for two patients who had IgG antibodies against HEV with negative IgM and RNA testing, suggesting a previous HEV infection.
Ten patients had sufficient HBV DNA levels (≥ 35 IU/mL) to determine the genotype of the infecting strain and were classified as sub-genotype D1. These patients were all treatment-naive. Primary drug resistance mutations in the reverse transcriptase domain of the P gene or immune escape mutations in the major hydrophilic region of the HBV surface protein (s99 to s169) [31] were not detected in any of the 10 patients. However, several non-synonymous amino acid substitutions were identified, with none occurring in more than two patients. There was only one amino acid substitution (Y134F; patient #11) in the major hydrophilic region of the S protein. Sequencing and analysis of the BCP, preC, and Core regions were possible in two patients (#01 and #03) with high viral loads (> 10 6 IU/mL), both of whom were HBeAg-negative, anti-HBe-positive, and treatment-naive. Both sequences showed a T1753C mutation in BCP. In #01, we found the A1762T/G1764A BCP mutation, and
## Note:
In the clinical data section, pathologic findings are in bold writing. Abbreviations: AATD, α-
in #03, the C1817T mutation, which generates a stop codon in the preC sequence (Figure 3). The material was qualitatively insufficient for the third highly viraemic patient in the cohort (#02, HBeAg-negative, without concomitant anti-HBe). In five patients with low HBV DNA levels (123-849 IU/mL), none of the three alternative PCR methods produced amplification products for sequencing.
$$T A B L E 1 (Continued)$$
## 2.3 | HBV-Specific CD8+ T-Cell Response
CD8+ T cells from all patients showed IFN-γ secretion after stimulation with the CEF pool (positive control), indicating an overall intact functionality of the cells after transport and thawing. The median percentage of IFN-γ+ CD8+ cells after incubation and stimulation with CEF peptides was 23.3%, with a rather wide scatter (minimum 0.9%; maximum 59.2%; IQR 31.4).
The T-cell responses to different HBV core peptide pools were as follows: After antigen-specific expansion and stimulation, a T-cell response (IFN-γ+ CD8+ cells) to at least two of the four examined peptide pools was detectable in all but one (#16). This patient had a serologic profile of an anti-HBs-positive, resolved/controlled infection and showed no measurable HBVspecific T-cell reaction, while having 55.5% IFN-γ+ CD8+ cells detectable after CEF-specific expansion and stimulation. The frequency of HBV-specific T cells against the four different peptide pools was low, with a percentage of IFN-γ+ CD8+ cells below 0.8% in all but two patients (#08 and #13). Patients #08 and #13 both had untreated and stable HBeAg-negative, low viraemic (849 and 38 IU/mL, respectively) chronic infections with no signs of clinical disease (liver enzyme elevation, abnormal sonographic findings, impaired hepatic function). Depending on the HBV peptide pool, they showed a percentage of HBV-specific T cells up to > 5%. An overview of the T-cell response to each peptide pool is shown in Figure 4. The detailed results grouped by clinical/serological phases of HBV infection are listed in Table 2. A formal statistical analysis of the differences between the patient groups was not possible owing to the small sample size. A complete genome sequence of the infecting strain, deposited in GenBank as Y07587 in 1996, served as the parental strain. HBc, hepatitis B core gene; preC, precore region.
## 3 | Discussion
This study revealed that the majority of the patients experienced a relatively benign course of their chronic HBV infection, which had persisted for around 35 years. In 4/16 patients, this favourable course was supported by antiviral therapy with entecavir or tenofovir. However, regarding regular clinical monitoring of patients with chronic HBV infection, this study revealed large gaps, especially in patients not currently receiving antiviral treatment. This is important to note, as over 30 years after primary infection, close follow-up has proven to remain necessary and justified for this cohort because of the known long-term risks of chronic HBV infection, particularly liver fibrosis and hepatocellular carcinoma.
In addition, the clinical follow-up revealed highly relevant new findings in five patients, which have direct consequences on their lives: Three new indications for antiviral therapy of a highly replicative CHB and, on the positive side, HBsAg/anti-HBs seroconversion in one patient, indicating a functional cure of the HBV infection. In one patient, monitoring revealed advanced alcoholic liver disease. As this patient was HBsAg-positive and had detectable HBV DNA in the past, she was assigned to antiviral therapy according to the EASL guidelines, although she had no detectable HBV DNA in the serum at the time of our study.
The number of patients available for follow-up has decreased since the last study of this unusual cohort approximately 25 years ago. Nevertheless, the newly gathered data confirm the previous conclusion that the course of HBV infection is most likely comparable to that of infections acquired during the perinatal period or early in life. This is remarkable because the children were up to 18 years old at the time of their primary infection. Loss of HBeAg has occurred in all patients throughout the 25 years since the initial follow-up, when they were still HBeAg-positive.
In contrast, only one HBsAg/anti-HBs seroconversion was observed. This is congruent with what is known about the natural history of chronic HBV infection acquired in early childhood [7,32,33]. The concept of "immune tolerance" of the adaptive immune system (i.e., T cells) towards viral antigens in early life and adolescence is increasingly being challenged. Instead, there is an alternative interpretation of a lacking capacity to mount a generally robust pro-inflammatory environment in which the adaptive immune system can achieve viral control and clearance [34]. Considering this, it seems consistent to expect a similar clinical course of chronic HBV infection acquired under cytostatic treatment, as the general inflammatory responses are temporarily impaired under chemotherapy.
Sequencing of the BCP and preC regions in two cases with high viraemia (> 10 6 IU/mL) revealed known mutations that can promote HBeAg negativity in highly viraemic patients by preventing preC translation by generating a stop codon (C1817T preC mutation) [35] or suppressing the transcription of preC (↓), while favouring the transcription of pregenomic (↑) mRNA. This results in reduced HBeAg expression, but enhanced virion production (A1762T/G1764A BCP double mutation) [36][37][38][39]. These well-known BCP mutations are often associated with another BCP mutation pattern, T1753V (V = C, A, or G; T1753C in our study, in both cases). All three BCP mutations have been reported to be associated with an increased risk of advanced liver disease during chronic HBV infection of genotype C [40] but also of genotype D [41]. Interestingly, these BCP mutations also affect the overlapping coding sequence of the multifunctional HBx (hepatitis B virus regulatory protein X) protein, causing amino acid exchanges in HBx that lead to changes in transactivation activity of HBx, which might contribute to carcinogenesis and development of hepatocellular carcinoma in genotypes C [42] and D [43].
Although HBeAg-negative patients are believed to rarely have high viraemia, our study identified 3/16 such patients. The potential infectivity of these patients is of particular concern, because acute infection with such variants often causes severe or fulminant hepatitis [44]. Finding no known immune escape or drug resistance mutations in 10 patient sera hints at a certain level of long-term persistent genomic stability over 30 years of chronic infection and supports the idea of "self-normalising" mutational activity of HBV formulated by Tedder et al. [10] Regarding the analysis of drug resistance mutations, we could only examine treatment-naive patients with no therapy-induced selection pressure and our results are consistent with other studies that report low rates of drug resistance mutations in treatment-naive patients [45].
The serologic finding of concurrent HBsAg and low-titre anti-HBs in two patients (#5 and #6) is puzzling. Parallel detection of HBsAg and anti-HBs has been reported in a wide range of cases of chronic HBV infections, including (i) HBeAg-negative patients with low viral load and appearance of (escape)mutations in the antigenic loop of HBsAg [46] and (ii) HBeAgpositive patients with high viral load but without significant mutations within the antigenic loop of HBsAg [47]. Patients #5 and #6 do not fall under either constellation. Despite being HBeAg-negative with low viral loads and low HBsAg levels, they do not exhibit any relevant mutations in the antigenic loop of HBsAg. However, it is possible that a minority of mutated HBsAg is being produced by integrated HBV DNA and is therefore not detectable in the wild-type genomes of circulating HBV virions.
Investigation of the cytotoxic T-cell response revealed a quantitatively weak HBV-specific immunity, with a low frequency of CD8+ T cells reactive to HBV core peptides after in vitro expansion. This pattern is typically observed in chronic HBV infection, especially in HBeAg-positive, highly viraemic chronic HBV infections [48][49][50]. Interestingly, all patients in our cohort in whom a weak T-cell response could be examined were HBeAg-negative, which is in line with previous reports [49,51]. At the individual level, we had one patient with an acute hepatitis flare (relevant ALT elevation) at the time of sample collection, showing no higher CD8+ frequency than the other patients, which is consistent with earlier findings that the number of HBV-specific T cells does not correlate with hepatic inflammatory activity [52].
## 4 | Conclusion
More than 30 years after the primary infection, patients in this unique cohort exhibit a clinical course of chronic HBV infection that is most likely comparable to that of perinatally infected individuals. All of them have seroconverted from HBeAg-positive chronic infection (formerly termed "immune tolerant" infection) to HBeAg-negative chronic infection, with some of them entering the "immune active" phase of chronic infection with higher levels of HBV DNA and ALT elevation as a sign of increased hepatic inflammatory activity. Due to the small number of patients, the general validity of our virological findings and the assessment of T-cell immunity is limited. However, re-examining this special cohort after 30 years of chronic infection contributes to the limited knowledge on primary HBV infections under conditions of iatrogenic immunosuppression. Furthermore, our study underlines the necessity for regular clinical monitoring of chronically HBV-infected patients, even decades after the primary infection, as there were a relevant number of patients requiring antiviral treatment due to study findings.
## References
1. (2019) "Global, Regional, and National Burden of Hepatitis B, 1990-2019: A Systematic Analysis for the Global Burden of Disease Study" *Lancet Gastroenterology & Hepatology*
2. Mcmahon, Alward, Hall (1985) "Acute Hepatitis B Virus Infection: Relation of Age to the Clinical Expression of Disease and Subsequent Development of the Carrier State" *Journal of Infectious Diseases*
3. Edmunds, Medley, Nokes et al. (1993) "The Influence of Age on the Development of the Hepatitis B Carrier State" *Proceedings of the Biological Sciences*
4. Beasley (2009) "Rocks Along the Road to the Control of HBV and HCC" *Annals of Epidemiology*
5. Weinbaum, Williams, Mast (2008) "Recommendations for Identification and Public Health Management of Persons With Chronic Hepatitis B Virus Infection" *MMWR -Recommendations and Reports*
6. Livingston, Simonetti, Bulkow (2007) "Clearance of Hepatitis B e Antigen in Patients With Chronic Hepatitis B and Genotypes" *Gastroenterology*
7. Simonetti, Bulkow, Mcmahon (2010) "Clearance of Hepatitis B Surface Antigen and Risk of Hepatocellular Carcinoma in a Cohort Chronically Infected With Hepatitis B Virus" *Hepatology*
8. Kramvis, Kostaki, Hatzakis et al. (2018) "Immunomodulatory Function of HBeAg Related to Short-Sighted Evolution, Transmissibility, and Clinical Manifestation of Hepatitis B Virus" *Frontiers in Microbiology*
9. Kramvis (2014) "Genotypes and Genetic Variability of Hepatitis B Virus" *Intervirology*
10. Tedder, Bissett, Myers et al. (2013) "The 'Red Queen' Dilemma-Running to Stay in the Same Place: Reflections on the Evolutionary Vector of HBV in Humans" *Antiviral Therapy*
11. Yuen, Chen, Dusheiko (2018) "Hepatitis B Virus Infection" *Nature Reviews Disease Primers*
12. Shin, Sung, Park (2016) "Immune Responses and Immunopathology in Acute and Chronic Viral Hepatitis" *Nature Reviews*
13. Kefalakes, Budeus, Walker (2015) "Adaptation of the Hepatitis B Virus Core Protein to CD8(+) T-Cell Selection Pressure" *Hepatology*
14. Kefalakes, Jochum, Hilgard (2015) "Decades After Recovery From Hepatitis B and HBsAg Clearance the CD8+ T Cell Response Against HBV Core Is Nearly Undetectable" *Journal of Hepatology*
15. Sutnick, London, Blumberg et al. *Australia Antigen*
16. (1970) "Leukemia"
17. Loomba, Liang (2017) "Hepatitis B Reactivation Associated With Immune Suppressive and Biological Modifier Therapies: Current Concepts, Management Strategies, and Future Directions" *Gastroenterology*
18. Repp (1994) "Hepatitis B Virusinfektionen bei Kindern mit malignen Erkrankungen [habilitation dissertation] Medicine, Justus-Liebig-Universität Gießen"
19. Repp, Von Horsten, Csecke (1993) "Clinical and Immunological Aspects of Hepatitis B Virus Infection in Children Receiving Multidrug Cancer Chemotherapy" *Archives of Virology. Supplementum*
20. Bertram, Repp, Lampert et al. (1990) "Hepatitis B Endemic in Children Treated With Cytostatic Drugs" *Deutsche Medizinische Wochenschrift*
21. Norder, Courouce, Coursaget (2004) "Genetic Diversity of Hepatitis B Virus Strains Derived Worldwide: Genotypes, Subgenotypes, and HBsAg Subtypes" *Intervirology*
22. Stoll-Becker, Repp, Glebe (1997) "Transcription of Hepatitis B Virus in Peripheral Blood Mononuclear Cells From Persistently Infected Patients" *Journal of Virology*
23. Repp, Rhiel, Heermann (1993) "Genotyping by Multiplex Polymerase Chain Reaction for Detection of Endemic Hepatitis B Virus Transmission" *Journal of Clinical Microbiology*
24. Lampert, Willems, Bertram et al. (1987) "No Adverse Prognostic Influence of Hepatitis B Virus Infection in Acute Childhood Lymphoblastic Leukemia" *Blut*
25. (2017) "EASL 2017 Clinical Practice Guidelines on the Management of Hepatitis B Virus Infection" *Journal of Hepatology*
26. Bedossa, Poynard (1996) "An Algorithm for the Grading of Activity in Chronic Hepatitis C. The METAVIR Cooperative Study Group" *Hepatology*
27. Friedrich-Rust, Nierhoff, Lupsor (2012) "Performance of Acoustic Radiation Force Impulse Imaging for the Staging of Liver Fibrosis: A Pooled Meta-Analysis" *Journal of Viral Hepatitis*
28. Abdelnabi, Saleh, Baraghithi et al. (2014) "Subgenotypes and Mutations in the s and Polymerase Genes of Hepatitis B Virus Carriers in the West Bank, Palestine" *PLoS One*
29. Geipel, Seiz, Niekamp (2015) "Entecavir Allows an Unexpectedly High Residual Replication of HBV Mutants Resistant to Lamivudine" *Antiviral Therapy*
30. Sato, Suzuki, Akahane (1995) "Hepatitis B Virus Strains With Mutations in the Core Promoter in Patients With Fulminant Hepatitis" *Annals of Internal Medicine*
31. Ren, Esser, Jochum (2012) "Interleukin 21 Augments the Hepatitis B Virus-Specific CD8+ T-Cell Response in Vitro in Patients Coinfected With HIV-1" *AIDS*
32. Lazarevic, Banko, Miljanovic et al. (1109) "Immune-Escape Hepatitis B Virus Mutations Associated With Viral Reactivation Upon Immunosuppression" *Viruses*
33. Mcmahon, Holck, Bulkow et al. (2001) "Serologic and Clinical Outcomes of 1536 Alaska Natives Chronically Infected With Hepatitis B Virus" *Annals of Internal Medicine*
34. Di Bisceglie, King, Lisker-Melman (2019) "Age, Race and Viral Genotype Are Associated With the Prevalence of Hepatitis B e Antigen in Children and Adults With Chronic Hepatitis B" *Journal of Viral Hepatitis*
35. Bertoletti, Hong (2014) "Age-Dependent Immune Events During HBV Infection From Birth to Adulthood: An Alternative Interpretation" *Frontiers in Immunology*
36. Kramvis, Kew, Bukofzer (1998) "Hepatitis B Virus Precore Mutants in Serum and Liver of Southern African Blacks With Hepatocellular Carcinoma" *Journal of Hepatology*
37. Jammeh, Tavner, Watson et al. (2008) "Effect of Basal Core Promoter and Pre-Core Mutations on Hepatitis B Virus Replication" *Journal of General Virology*
38. Sendi, Mehrab-Mohseni, Zali et al. (2005) "T1764G1766 Core Promoter Double Mutants Are Restricted to Hepatitis B Virus Strains With an A1757 and Are Common in Genotype D" *Journal of General Virology*
39. Erhardt, Reineke, Blondin (2000) "Mutations of the Core Promoter and Response to Interferon Treatment in Chronic Replicative Hepatitis B" *Hepatology*
40. Buckwold, Xu, Chen et al. (1996) "Effects of a Naturally Occurring Mutation in the Hepatitis B Virus Basal Core Promoter on Precore Gene Expression and Viral Replication" *Journal of Virology*
41. Chen, Lee, Lu (2005) "Clinical Significance of Hepatitis B Virus (HBV) Genotypes and Precore and Core Promoter Mutations Affecting HBV e Antigen Expression in Taiwan" *Journal of Clinical Microbiology*
42. Sharma, Sharma, Singla (2010) "Clinical Significance of Genotypes and Precore/Basal Core Promoter Mutations in HBV Related Chronic Liver Disease Patients in North India" *Digestive Diseases and Sciences*
43. Yuan, Zhou, Tanaka (2007) "Hepatitis B Virus (HBV) Genotypes/Subgenotypes in China: Mutations in Core Promoter and Precore/Core and Their Clinical Implications" *Journal of Clinical Virology*
44. Elkady, Tanaka, Kurbanov et al. (2008) "Virological and Clinical Implication of Core Promoter C1752/V1753 and T1764/G1766 Mutations in Hepatitis B Virus Genotype D Infection in Mongolia" *Journal of Gastroenterology and Hepatology*
45. Seiz, Slanina, Ziebuhr et al. (2015) "Studies of Nosocomial Outbreaks of Hepatitis B in Nursing Homes in Germany Suggest a Major Role of Hepatitis B e Antigen Expression in Disease Severity and Progression" *International Journal of Medical Microbiology*
46. Gomes-Gouvea, Ferreira, Teixeira (2015) "HBV Carrying Drug-Resistance Mutations in Chronically Infected Treatment-Naive Patients" *Antiviral Therapy*
47. Lada, Benhamou, Poynard et al. (2006) "Coexistence of Hepatitis B Surface Antigen (HBs ag) and Anti-HBs Antibodies in Chronic Hepatitis B Virus Carriers: Influence of "a" Determinant Variants" *Journal of Virology*
48. Zhang, Xu, Wang (2007) "Coexistence of Hepatitis B Surface Antigen (HBsAg) and Heterologous Subtype-Specific Antibodies to HBsAg Among Patients With Chronic Hepatitis B Virus Infection" *Clinical Infectious Diseases*
49. Boni, Fisicaro, Valdatta (2007) "Characterization of Hepatitis B Virus (HBV)-Specific T-Cell Dysfunction in Chronic HBV Infection" *Journal of Virology*
50. Park, Wong, Wahed (2016) "Hepatitis B Virus-Specific and Global T-Cell Dysfunction in Chronic Hepatitis B" *Gastroenterology*
51. Webster, Reignat, Brown (2004) "Longitudinal Analysis of CD8+ T Cells Specific for Structural and Nonstructural Hepatitis B Virus Proteins in Patients With Chronic Hepatitis B: Implications for Immunotherapy" *Journal of Virology*
52. Walker, Schwarz, Brinkmann-Paulukat (2022) "Immune Escape Pathways From the HBV Core(18-27) CD8 T Cell Response Are Driven by Individual HLA Class I Alleles" *Frontiers in Immunology*
53. Maini, Boni, Lee (2000) "The Role of Virus-Specific CD8(+) Cells in Liver Damage and Viral Control During Persistent Hepatitis B Virus Infection" *Journal of Experimental Medicine* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12777568&blobtype=pdf | # The effect of antacid and mineral supplements on bictegravir pharmacokinetics: results from a Phase 1, open-label, drug-drug interaction study
Priyanka Arora, Jason Hindman, Steve West, John Ling, Ramesh Palaparthy, Dhananjay Marathe
## Abstract
The mechanism of action of integrase strand transfer inhibitors involves binding to magnesium ions in the active site of the HIV integrase enzyme, making them susceptible to chelation-type drug-drug interactions with metal cation-containing medications. This study evaluated the potential of metal cation-containing antacids and mineral supplements to impact bictegravir (BIC) exposure and assessed alternative approaches for combined use. This was an open-label, single-dose, Phase 1 study in adult participants without HIV. The pharmacokinetics and safety of BIC (administered as part of a single-tablet combination with emtricitabine [F] and tenofovir alafenamide [TAF; B/F/TAF]) were assessed when co-administered with maximum-strength alumi num/magnesium-containing antacid (referred to as "aluminum/magnesium-containing antacid"), calcium carbonate, or ferrous fumarate under fasted and fed conditions, and administered 2 hours before or after the antacid. Pharmacokinetic parameters were compared using analysis of variance to calculate geometric least-squares mean ratios and 90% confidence intervals. Forty-two participants were enrolled. BIC exposure (area under the plasma concentration-time curve extrapolated to infinity) was reduced by 79%, 33%, and 63%, respectively, when co-administered with aluminum/magnesiumcontaining antacid, calcium carbonate, and ferrous fumarate under fasted conditions. Co-administration of B/F/TAF with calcium carbonate or ferrous fumarate with a meal and administration of B/F/TAF 2 hours before the antacid reduced the impact of the interactions. B/F/TAF was well tolerated alone or in combination with metal cation-con taining medications. Co-administration of BIC and calcium/iron-containing supplements with a meal and administration of BIC 2 hours or more before aluminum/magnesiumcontaining antacids are some of the effective strategies to mitigate chelation effects on BIC exposure.CLINICAL TRAILS This study was registered at NCT05502341/NCT06333808. KEYWORDS bictegravir, pharmacokinetics, drug-drug interactions, metal cations, human immunodeficiency virus, integrase strand transfer inhibitor I ntegrase strand transfer inhibitors (INSTIs) are recommended as part of the first-line treatment for people with HIV (PWH) (1, 2). Bictegravir (BIC) is a potent INSTI approved for the treatment of HIV-1 infection as part of an oral, once-daily, fixed-dose, singletablet regimen, B/F/TAF, which also incorporates two nucleoside reverse transcriptase inhibitors, emtricitabine (F) and tenofovir alafenamide (TAF) (1-3). Clinical trials (4-7) and real-world studies (8, 9) have shown B/F/TAF to be effective and well tolerated in PWH. In addition, BIC is currently in Phase 2/3 clinical development in combination with lenacapavir, an HIV-1 capsid inhibitor (NCT05502341/NCT06333808) (10).
A key element of INSTI action involves chelation of magnesium ions in the active site of the HIV integrase enzyme, displacing the HIV viral DNA and preventing insertion into the host cell genome (11). Therefore, INSTIs, including BIC, are susceptible to chelationtype drug-drug interactions (DDIs) with metal cation-containing therapeutic products (12). Previous studies have shown reductions in the exposure of other INSTIs when co-administered with metal cation-containing therapeutic products (13)(14)(15)(16)(17).
PWH often take metal cation-containing medications in addition to antiretroviral therapy (ART); for example, gastroesophageal reflux and other gastrointestinal disorders are common among PWH, leading to frequent use of gastric acid modifiers, including antacids (18). Furthermore, HIV-1 infection and ART are associated with reductions in bone mineral density (19), and calcium supplements may be taken by PWH to improve bone health. PWH may also use iron supplements as they are at high risk of iron-deficiency anemia due to infection, chronic inflammation, and the effects of ART (20). Finally, as for the general population, PWH may also take general multivitamin/multimineral supplements containing metal cations (21).
This study evaluated the potential of metal cation-containing antacids and mineral supplements to impact BIC exposure in adult participants without HIV and assessed alternative approaches for combined use if an interaction was observed.
## MATERIALS AND METHODS
## Study design and participants
Study GS-US-380-3909 was an open-label, single-dose, fixed-sequence, multiplecohort, multiple-period, adaptive Phase 1 study conducted in adult participants without HIV (typically referred to as healthy volunteers) at a single center within the United States between April 4, 2016, and May 20, 2016. Eligible participants were adult (18-45 years) males and non-pregnant, non-lactating females in good general health at a screening evaluation performed no more than 28 days before the first study dose. Key inclusion criteria included body mass index ≥19.0 and ≤30.0 kg/m 2 ; normal 12-lead electrocardiogram (ECG) or clinically insignificant ECG abnormalities (as judged by the study investigator); normal renal function; estimated glomerular filtration rate using the Cockcroft-Gault method (22) ≥90 mL/min, based on serum creatinine and actual body weight; and no significant medical history. Full eligibil ity criteria are available in the supplement. All prescription and over-the-counter medications except vitamins, acetaminophen, ibuprofen, and hormonal contracep tives were not permitted during study participation.
The primary objectives were to evaluate the effects of simultaneous co-adminis tration of aluminum/magnesium-containing antacid, or calcium or iron supplements, with B/F/TAF fixed-dose combination (FDC) under fasted and fed conditions on the pharmacokinetics (PK) of BIC. Additionally, the study aimed to compare the PK of BIC between staggered administration of the antacid and B/F/TAF compared with B/F/TAF FDC alone, under fasted conditions. The secondary objective was to evaluate the safety and tolerability of single doses of B/F/TAF FDC administered alone or in combination with aluminum/magnesium-containing antacid, or calcium or iron supplements.
Three dosing cohorts were evaluated (Table 1). All cohorts were initiated in parallel and received a single dose of B/F/TAF (50/200/25 mg) FDC as one tablet administered orally on Day 1 under fasted conditions, followed by a 7-day washout period. This was followed by co-administration of single doses of B/F/TAF with single doses of aluminum/magnesium-containing antacid, or calcium or iron supplements on Days 9, 17, and 25, respectively (Cohorts 1 and 3 only), with all treatments separated by a 7-day washout period. Vitamins and/or other supplements were not permitted on the dosing day.
Cohorts 1 and 3 were designed to assess the effect of single-dose antacid, calcium, or iron supplements simultaneously co-administered with single-dose B/F/TAF under fasted (Cohort 1) or fed (Cohort 3) conditions. Participants received maximum-strength antacid oral suspension (four teaspoons; 20 mL; 1,600 mg aluminum hydroxide/1,600 mg magne sium hydroxide/160 mg simethicone; referred to as "aluminum/magnesium-containing antacid") on Day 9, calcium carbonate (1,200 mg; 2 × 600 mg tablets) on Day 17, and ferrous fumarate (1 × 324 mg tablet) on Day 25. Cohort 2 assessed the effect of staggered administration of single-dose, aluminum/magnesium-containing antacid oral suspension with single-dose B/F/TAF under fasted conditions. Participants received B/F/TAF 2 hours before the antacid on Day 9 and 2 hours after the antacid on Day 17. If the 2-hour separation window was found to be insufficient to minimize the interaction with BIC, a fourth, adaptive cohort was planned to evaluate the effect of 4-hour staggered administration of B/F/TAF and aluminum/magnesium-containing antacid.
A fixed dose of 50/200/25 mg B/F/TAF was chosen as this was the FDC formulation selected for Phase 3 clinical development (23). The highest recommended dose of aluminum/magnesium-containing antacid (24) was chosen because of its high divalent metal cation content, to evaluate the worst-case scenario for interaction. Commonly recommended daily doses for a calcium supplement (calcium carbonate) and an iron supplement (ferrous fumarate) were chosen, with elemental calcium and iron content of approximately 480 mg and 107 mg, respectively (25,26). Given the half-life for BIC of ~17 hours (3), a crossover study with a 7-day washout between periods was considered appropriate.
Participants were confined to the clinic for the duration of the study, from Day -1 until completing study procedures on Day 29 for Cohorts 1 and 3, or Day 21 for Cohort 2. Follow-up phone calls were conducted 7 (± 2) days after clinic discharge to collect information on any adverse events (AEs) that may have occurred and any concomitant medications taken since discharge.
On Day 1, all participants received B/F/TAF administered following an observed overnight fast (for at least 10 hours). Subsequently, participants in Cohort 1 received study drugs on Days 9, 17, and 25, following an observed overnight fast; participants in Cohort 2 received study drugs on Days 9 and 17, following an observed overnight fast; and participants in Cohort 3 received study drugs on Days 9, 17, and 25, following an observed overnight fast and within 5 minutes of the participant finishing a moderate-fat breakfast (~600 kcal; ~27% fat).
## BIC PK analyses
Serial blood samples for PK assessment were taken on Days 1, 9, and 17 for Cohorts 1-3 and additionally on Day 25 for Cohorts 1 and 3 pre-B/F/TAF administration, and at 0. plasma samples were determined by QPS, LLC (Newark, DE) using fully validated liquid chromatography-tandem mass spectroscopy bioanalytical methods. All samples were analyzed within the timeframe determined using frozen stability storage data. The lower and upper limits of quantification for BIC were 0.020 µg/mL and 20 µg/mL, respectively. The interassay precision (coefficient of variation) and accuracy (relative error) ranges for BIC were 3.9%-5.7% and 2.4%-5.8%, respectively. PK parameters were estimated using Phoenix WinNonlin software (Version 6.4, Pharsight Corporation, Princeton, NJ, USA) using standard noncompartmental analy sis methods. Assessed key PK parameters included area under the plasma concentra tion-time curve extrapolated to infinity (AUC inf ), plasma concentration after 24 h (C 24 ), and maximum observed plasma concentration (C max ).
The BIC PK analysis set included all enrolled participants who received at least one dose of the study drug and had at least one post-dose PK concentration value.
## Safety analyses
Assessments included periodic clinical laboratory tests, ECGs, and physical examinations (including vital signs) at Day -1 and/or Day 1, then every 8 days and at the end of the study (Day 29 for Cohorts 1 and 3, and Day 21 for Cohort 2), and daily monitoring of AEs and concomitant medications.
The safety analysis set included all participants who received at least one dose of the study drug.
## Statistical analyses
Participant demographics and baseline characteristics were summarized using descriptive statistics.
Plasma concentrations and PK parameters for BIC were summarized using descriptive statistics, by treatment within each cohort.
AUC inf , C 24 , and C max for test treatments (B/F/TAF plus study drug) were compared with reference treatment (B/F/TAF alone under fasted or fed conditions) using mixedeffects analysis of variance to calculate geometric least-squares mean (GLSM) ratios and associated 90% confidence intervals (CIs). The estimated 90% CIs were compared to a prespecified lack of DDI boundary of 70%-143%. Previously reported clinical data support the use of a 70%-143% boundary over the traditional 80%-125% bioequiva lence boundary for evaluation of quantitative significance for DDI effects; BIC has a long dissociation half-life (~163 hours), a flat exposure-response curve for efficacy, and no additional safety concerns over a broad dose/exposure range (27)(28)(29)(30)(31). Generally, test drugs with 90% CIs entirely contained within this range were deemed to have no significant effect on the PK of BIC. A minimum of 12 evaluable participants in each cohort was calculated to be needed to reject the null hypothesis (i.e., the 90% CI for the GLSM ratio for each PK parameter was outside the prespecified boundary) with ≥90% power.
AEs, laboratory parameters, graded laboratory abnormalities, and vital signs were summarized using descriptive statistics.
## RESULTS
## Study participants
The study took place between April 4 and May 20, 2016, with all participants enrolled between April 13 and 16, 2016. A total of 42 individuals participated in the study (14 participants in each of Cohorts 1-3), received at least one dose of study drug, and were included in the safety and BIC PK analysis sets. Overall, 41 participants (98%) completed the study. One participant in Cohort 2 discontinued study drug because of an AE of Grade 2 urticaria (Fig. 1).
Baseline demographics and characteristics are shown in Table S1. Most participants were male (69%), White (71%), and of Hispanic/Latino ethnicity (69%); median age (quartile [Q]1, Q3) was 34 (29,40)
## BIC PK outcomes
## Cohort 1: Simultaneous co-administration (fasted)
Simultaneous co-administration of B/F/TAF with aluminum/magnesium-containing antacid, calcium carbonate, or ferrous fumarate resulted in reduced BIC exposure compared with B/F/TAF alone under fasted conditions (Tables 2 and3; Fig. 2a). BIC AUC inf , C 24 , and C max were reduced by 79%, 78%, and 80%, respectively, with alumi num/magnesium-containing antacid; by 33%, 35%, and 42%, respectively, with calcium carbonate; and by 63%, 63%, and 71%, respectively, with ferrous fumarate, compared with B/F/TAF alone. The 90% CIs of the GLSM ratios were outside the prespecified lack of DDI boundary of 70%-143%, indicating significant DDIs (Table 3).
The median time of C max (T max ) was delayed by 1 hour and 1.5 hours with alu minum/magnesium-containing antacid and ferrous fumarate, respectively, compared with B/F/TAF alone, but the median terminal elimination half-life (T 1/2 ) was similar between treatments. The mean BIC apparent oral clearance (CL/F) and apparent volume of distribution (V z /F) for each treatment increased following simultaneous co-admin istration of aluminum/magnesium-containing antacid (404% and 421%, respectively), calcium carbonate (56% and 84%, respectively), and ferrous fumarate (174% and 216%, respectively) with B/F/TAF compared with B/F/TAF alone.
## Cohort 2: Staggered administration of aluminum/magnesium-containing antacid (fasted)
Administration of B/F/TAF under fasted conditions 2 hours before aluminum/magne sium-containing antacid counteracted the interaction and resulted in BIC exposure similar to that seen with B/F/TAF alone (Tables 2 and3; Fig. 2b). The 90% CIs of the GLSM ratios for BIC AUC inf , C 24 , and C max were within the prespecified lack of DDI boundary, indicating a lack of significant impact on BIC exposure.
Administration of B/F/TAF 2 hours after aluminum/magnesium-containing antacid resulted in a less marked reduction in BIC exposure than when administered simultaneously under fasted conditions (52%, 53%, and 59% decreases in AUC inf , C 24 , and C max , respectively, compared with B/F/TAF alone; Tables 2 and3; Fig. 2b). However, this was not sufficient to completely counteract the interaction, and the 90% CIs of the GLSM ratios for BIC AUC inf , AUC from time zero to the last quantifiable concentration (AUC last ), and C max still spanned the prespecified lack of DDI boundary.
The median T max was delayed by 1.5 hours and 1 hour when B/F/TAF was admin istered 2 hours before and 2 hours after aluminum/magnesium-containing antacid, respectively, compared with B/F/TAF alone, but the median BIC T 1/2 was similar between treatments. The mean BIC CL/F and V Z /F were generally similar following B/F/TAF administration 2 hours before aluminum/magnesium-containing antacid but were increased following administration 2 hours after aluminum/magnesium-containing antacid (124% and 180%, respectively) compared with B/F/TAF alone.
Since administration of B/F/TAF 2 hours before aluminum/magnesium-containing antacid administration was deemed sufficient to counteract the DDI effect, the fourth, adaptive cohort, assessing the effect of 4-hour staggered administration, was not initiated.
## Cohort 3: Simultaneous co-administration (fed)
Simultaneous co-administration of B/F/TAF with calcium carbonate or ferrous fumarate under fed conditions generally resulted in BIC exposure similar to that seen with BIC alone under fasted conditions (Tables 2 and3; Fig. 2c). The 90% CIs of the GLSM ratios for AUC inf , C 24 , and C max were within the prespecified lack of DDI boundary, with the Co-administration of B/F/TAF with aluminum/magnesium-containing antacid under fed conditions resulted in a less marked reduction in BIC exposure than when adminis tered under fasted conditions (47%, 44%, and 49% decreases in AUC inf , C 24 , and C max , respectively, compared with B/F/TAF alone under fasted conditions; Tables 2 and3; Fig. 2c). However, this was not sufficient to completely counteract the interaction, and the 90% CIs of the GLSM ratios for BIC AUC inf , C 24 , and C max were still outside of the prespecified lack of DDI boundary.
The median T max was delayed by 2 hours with simultaneous co-administration of B/F/TAF with aluminum/magnesium-containing antacid or calcium carbonate and by 2.5 hours with ferrous fumarate under fed conditions, compared with B/F/TAF alone under fasted conditions, but the median T 1/2 was similar between treatments. The mean BIC CL/F and V z /F increased following co-administration of aluminum/magnesiumcontaining antacid with B/F/TAF under fed conditions (89% in both cases) but were similar for calcium carbonate and ferrous fumarate under fed conditions, compared with B/F/TAF alone under fasted conditions.
## Safety
AEs and laboratory abnormalities are shown in Table 4. Across all cohorts and treatments, there were no Grade 3 or 4 AEs, serious AEs, or deaths. The only AE considered related to B/F/TAF by the investigator was a single event of Grade 2 urticaria experienced by one participant in Cohort 2, which began 1 day after receiving B/F/TAF alone and resulted in study drug discontinuation. The participant was treated with diphenhydramine, and the event resolved on Day 7.
## DISCUSSION
PWH often take antacids to treat heartburn and gastroesophageal reflux disease (18) and calcium/iron supplements to improve bone health and treat anemia associated with HIV infection and ART (19,20); therefore, it is important to understand the effects of these treatments on ART regimens to provide dosing recommendations and DDI mitigation strategies to minimize the risk of virologic failure and the development of INSTI resistance.
In this study, BIC exposure was reduced following simultaneous co-administration of B/F/TAF with aluminum/magnesium-containing antacid, calcium carbonate, or ferrous fumarate under fasted conditions. These effects were mitigated by staggering adminis tration and/or administration under fed conditions (32).
The largest reduction in BIC exposure was observed upon simultaneous co-adminis tration of BIC (as B/F/TAF) with aluminum/magnesium-containing antacid, which had the highest metal cation concentration tested in this study; an intermediate decrease was seen with ferrous fumarate, and the smallest decrease was seen with calcium carbonate. The 90% CIs of the GLSM ratios for all of these medications were outside of the prespeci fied lack of DDI boundary; therefore, all could be deemed to have significant DDIs. However, all available cumulative data for BIC efficacy, including the overall risk-benefit balance, were taken into consideration to inform clinical acceptability and contextualize the observed data. As BIC has been shown to have a long dissociation half-life (~163 hours) and a flat exposure-response curve for efficacy over a wide dose/exposure range (27)(28)(29)(30)(31), the scenarios for staggered administration and/or co-administration with a meal, as conveyed in the dosing recommendations, were deemed to have no clinically meaningful effect based on the magnitude of changes observed in the study. The observed reductions in BIC exposure are likely to be due to chelation of BIC by the metal cations, limiting BIC absorption (33). In support of this, both the mean BIC CL/F and V z /F were increased by a similar extent following co-administration of B/F/TAF with the aluminum/magnesium-containing antacid, or calcium or iron supplements, indicating that these changes are a result of reductions in bioavailability (F).
The effect of metal cations on BIC exposure was reduced when administration was staggered or under fed conditions. Administration of B/F/TAF 2 hours before aluminum/ magnesium-containing antacid under fasted conditions mitigated the chelation effects seen with co-administration under fasted conditions. As the aluminum/magnesiumcontaining antacid had the highest metal cation content tested in this study, resulting in the greatest reduction in BIC exposure, it is likely that staggered administration would also mitigate the chelation effects of the calcium/iron-containing supplements.
In addition, co-administration of B/F/TAF with calcium carbonate and ferrous fumarate after a moderate-fat meal counteracted the chelation effects seen with co-administration under fasted conditions. The decrease in BIC exposure was less marked when B/F/TAF was simultaneously co-administered with the aluminum/magnesium-con taining antacid after a moderate-fat meal compared with co-administration under fasted conditions. However, the 90% CIs of the GLSM ratios were still outside of the prespecified lack of DDI boundary, indicating that there was a significant effect on BIC exposure. As BIC has been shown to have a long dissociation half-life (~163 hours) and a flat exposure-response curve for efficacy over a wide dose/exposure range (27)(28)(29)(30)(31), this effect was not considered clinically meaningful based on the magnitude of the observed changes.
As seen in previous studies, B/F/TAF was well tolerated when administered alone (23,34) and when given in combination with aluminum/magnesium-containing antacid, calcium carbonate, or ferrous fumarate.
Our findings suggest that staggering the administration of B/F/TAF and metal cation-containing antacids or calcium/iron-containing supplements and co-administra tion of B/F/TAF and calcium/iron-containing supplements with a meal results in BIC plasma exposure similar to when B/F/TAF is administered alone and could be a strategy to mitigate the effects of antacids and mineral supplements on BIC exposure.
Reductions in BIC exposure may be of particular relevance in pregnant women with HIV who, in addition to reduced BIC exposure associated with pregnancy (31), may be more likely to take antacids and metal cation-containing supplements than non-preg nant PWH. We have previously reported lower BIC exposure in pregnant women with HIV-1 receiving B/F/TAF than in non-pregnant adults with HIV-1 (3,31). Therefore, as gastroesophageal reflux disease affects approximately two-thirds of pregnant women and calcium-containing antacids are the preferred medication (35), it is important to consider mitigation strategies for the concomitant use of B/F/TAF with antacids and mineral supplements in this population.
Like BIC, the mechanism of action of other INSTIs, such as raltegravir, dolutegra vir, and elvitegravir, involves the binding of magnesium at the integrase active site (11), and thus, they are also susceptible to the chelating effects of metal cation-con taining therapeutic products. Indeed, similar to our findings, previous studies have reported reductions in the exposure of all three of these INSTIs when administered with metal cation-containing therapeutic products (13)(14)(15)(16)(17). These studies have also shown that the DDI effects of aluminum/magnesium-containing antacids were counteracted when dolutegravir (15) and elvitegravir (16) were administered 2 hours after antacid treatment. In addition, the DDI effects of ferrous fumarate and calcium carbonate on dolutegravir were prevented when dolutegravir was administered 2 hours before ferrous fumarate and calcium carbonate or simultaneously under fed conditions (17). For this reason, staggered administration of INSTIs and metal cation-containing therapeutic products and/or administration with food is recommended (36)(37)(38)(39)(40). Effects of antacids on raltegravir PK may be compounded by the fact that, unlike other INSTIs, raltegravir absorption is pH dependent (14), and co-administration of raltegravir with aluminum-, magnesium-, and/or calcium carbonate-containing antacids is not recommended (40).
In light of this study's findings and in conjunction with the established expo sure-response relationship for BIC over a wide exposure range, as well as cumulative data on BIC efficacy, including its long dissociation half-life (27)(28)(29)(30)(31), current recommen dations are that B/F/TAF is to be administered at least 2 hours before or 2 hours after taking aluminum/magnesium-containing antacids or calcium-or iron-containing supplements, or that they are to be co-administered with a meal. Following individual regulatory agency reviews and labeling discussions, these dosing recommendations, and regional variations thereof, have been incorporated into the region-specific labels for B/F/TAF (3,41). In addition, findings from this study have been contextualized to inform pregnancy-related considerations, with the recommendations for pregnant individuals, where BIC exposures from B/F/TAF are lowered compared with non-pregnant individuals (31), that B/F/TAF is to be administered at least 2 hours before or 6 hours after taking aluminum/magnesium-containing antacids, or calcium-or iron-containing supplements, or that calcium-or iron-containing supplements are to be co-administered with a meal (3,41).
The recommendation to separate administration of B/F/TAF and aluminum/magne sium-containing antacids/supplements or calcium-or iron-based medications/supple ments by at least 6 hours was to mitigate further lowering of BIC exposure and was based on previous experience with other drugs, including dolutegravir (36) and cabotegravir (39) (other INSTIs) as well as ciprofloxacin (42) and delafloxacin (43). These drugs are known to bind to polyvalent cation-containing medications via similar mechanisms to those of BIC.
Limitations of this study include the potential risks of bias associated with the open-label study design. In addition, findings in participants without HIV may not be directly applicable to the target population; however, since disease status is not expected to significantly impact BIC PK, these data are likely to translate to PWH. Furthermore, this study looked at single doses of B/F/TAF and metal cation-contain ing therapeutic products but did not assess the effects of metal cations on BIC exposure when these products are taken more than once daily or over a prolonged period of time. However, as single doses are generally more sensitive in assessing the absorption phase, which is expected to be most impacted by chelation-type DDIs, therapeutic products taken more than once daily are unlikely to have any further impact on the exposure of once-daily BIC. In addition, given the median T 1/2 of approximately 17 hours for BIC (3), the data are expected to be extrapolatable to steady-state dosing. Finally, the small participant number and short study timeframe may not be adequate to identify rare AEs.
## Conclusions
In conclusion, single doses of B/F/TAF, alone or combined with aluminum/magnesiumcontaining antacid, calcium carbonate, or ferrous fumarate, were well tolerated in this study. The effect of aluminum/magnesium-containing antacid on BIC exposure was prevented by administering 2 hours after B/F/TAF. Co-administration of B/F/TAF with a meal counteracted the chelating effects of calcium carbonate and ferrous fumarate; in addition, the chelating effect of aluminum/magnesium-containing antacid was also reduced under fed conditions. These results support the recommendations that B/F/TAF is to be administered at least 2 hours before or 2 hours after taking aluminum/mag nesium-containing antacids or calcium-or iron-containing supplements, or that they are to be co-administered with a meal. Furthermore, findings from this study were contextualized to inform pregnancy-related considerations in labeling, where additional staggering of B/F/TAF at least 6 hours after taking metal cation-containing medications or supplements is recommended to minimize reductions in BIC exposure.
## References
1. Dhhs (2024) "Guidelines for the use of antiretroviral agents in adults and adolescents with HIV"
2. (2024) "Guidelines version 12.1. Available from"
3. (2025) *Full-Length Text Antimicrobial Agents and Chemotherapy*
4. Molina, Ward, Brar et al. (2018) "Switching to fixed-dose bictegravir, emtricita bine, and tenofovir alafenamide from dolutegravir plus abacavir and lamivudine in virologically suppressed adults with HIV-1: 48 week results of a randomised, double-blind, multicentre, active-controlled, phase 3, non-inferiority trial" *Lancet HIV*
5. Orkin, Dejesus, Sax et al. "GS-US-380-1490 study investigators. 2020. Fixed-dose combination bictegravir, emtricitabine, and tenofovir alafenamide versus dolutegravir-containing regimens for initial treatment of HIV-1 infection: week 144 results from two randomised, double-blind, multicentre, phase 3, non-inferiority trials" *Lancet HIV*
6. Sax, Rockstroh, Luetkemeyer et al. "Switching to bictegravir, emtricitabine, and tenofovir alafenamide in virologically suppressed adults with human immunodeficiency virus" *Clin Infect Dis*
7. Stellbrink, Arribas, Stephens et al. (2019) "Co-formulated bictegravir, emtricitabine, and tenofovir alafena mide versus dolutegravir with emtricitabine and tenofovir alafenamide for initial treatment of HIV-1 infection: week 96 results from a rando mised, double-blind, multicentre, phase 3, non-inferiority trial" *Lancet HIV*
8. Ambrosioni, Liévano, Berrocal et al. (2022) "Real-life experience with bictegravir/emtricitabine/tenofovir alafenamide in a large reference clinical centre" *J Antimicrob Chemother*
9. Mounzer, Brunet, Fusco et al. (2022) "Advanced HIV infection in treatment-naive individuals: effectiveness and persistence of recommended 3-drug regimens" *Open Forum Infect Dis*
10. Mounzer, Slim, Ramgopal et al. (2024) "Phase II study of switch to daily BIC + LEN in individuals on a multitablet HIV treatment regimen (Poster 642)"
11. Jóźwik, Passos, Lyumkis (2020) "Structural biology of HIV integrase strand transfer inhibitors" *Trends Pharmacol Sci*
12. Scarsi, Havens, Podany et al. (2020) "HIV-1 integrase inhibitors: a comparative review of efficacy and safety" *Drugs (Abingdon Engl)*
13. Kiser, Bumpass, Meditz et al. (2010) "Effect of antacids on the pharmacokinetics of raltegravir in human immunodeficiency virus-seronegative volunteers" *Antimicrob Agents Chemother*
14. Krishna, East, Larson et al. (2016) "Effect of metal-cation antacids on the pharmacokinetics of 1200 mg raltegravir" *J Pharm Pharmacol*
15. Patel, Song, Borland et al. (2011) "Pharmacokinetics of the HIV integrase inhibitor S/GSK1349572 co-administered with acid-reducing agents and multivitamins in healthy volunteers" *J Antimicrob Chemother*
16. Ramanathan, Mathias, Wei et al. (2013) "Pharmacokinetics of once-daily boosted elvitegravir when administered in combination with acid-reducing agents" *J Acquir Immune Defic Syndr*
17. Van Lunzen, Liess, Arastéh et al. (2007) "Concomitant use of gastric acid-reducing agents is frequent among HIV-1-infected patients receiving protease inhibitor-based highly active antiretroviral therapy" *HIV Med*
18. Lopes, Farinatti, Lopes G De et al. (2021) "Muscle mass, strength, bone mineral density and vascular function in middle-aged people living with HIV vs. age-matched and older controls" *Braz J Infect Dis*
19. Abioye, Andersen, Sudfeld et al. (2020) "Anemia, iron status, and HIV: a systematic review of the evidence" *Adv Nutr*
20. Yetley (2007) "Multivitamin and multimineral dietary supplements: definitions, characterization, bioavailability, and drug interactions" *Am J Clin Nutr*
21. Cockcroft, Gault (1976) "Prediction of creatinine clearance from serum creatinine" *Nephron*
22. Gallant, Lazzarin, Mills et al. (2017) "Bictegravir, emtricitabine, and tenofovir alafenamide versus dolutegravir, abacavir, and lamivudine for initial treatment of HIV-1 infection (GS-US-380-1489): a double-blind, multicentre, phase 3, randomised controlled non-inferiority trial" *Lancet*
23. (2025) "Aluminum hydroxide / magnesium hydroxide / simethicone dosage" *Drugs.com*
24. (2025) "Calcium carbonate dosage" *Drugs.com*
25. (2025) "Ferrous fumarate dosage"
26. Arora, Zhang, Hindman et al. (2024) "Pharmacokinetics, safety, and efficacy study in pregnancy and existing cumulative data/evidence to support clinical use and labeling of B/F/TAF in pregnant women with HIV (Poster 26)"
27. Gallant, Thompson, Dejesus et al. (2017) "Antiviral activity, safety, and pharmacokinetics of bictegravir as 10-day monotherapy in HIV-1infected adults" *J Acquir Immune Defic Syndr*
29. (2018) "BIKTARVY NDA 210251 uni-review"
30. White, Osman, Cuadra-Foy et al. (2023) "Long dissociation of bictegravir from HIV-1 integrase-DNA complexes" *Antimicrob Agents Chemother*
31. Zhang, Hindman, Lin et al. (2024) "A study of the pharmacokinetics, safety, and efficacy of bictegravir/emtricitabine/ tenofovir alafenamide in virologically suppressed pregnant women with HIV" *AIDS*
32. Mathias, Lutz, West et al. "Pharmacokinetics of bictegravir in combination with polyvalent cation-containing antacids and supplements (Poster P260). Presented at: HIV Drug Therapy Glasgow"
33. Lu, Bednarczyk, Catanzaro et al. (2021) "Pharmacokinetic drug interactions of integrase strand transfer inhibitors" *Curr Res Pharmacol Drug Discov*
34. Sax, Dejesus, Crofoot et al. (2017) "Bictegravir versus dolutegravir, each with emtricitabine and tenofovir alafenamide, for initial treatment of HIV-1 infection: a randomised" *Full-Length Text Antimicrobial Agents and Chemotherapy*
35. *Lancet HIV*
36. Altuwaijri (2022) "Evidence-based treatment recommendations for gastroesophageal reflux disease during pregnancy: a review" *Medicine (Abingdon)*
37. Healthcare (2022)
38. Healthcare (2024)
39. Healthcare (2025)
40. Sharp, Dohme (2022) "Isentress"
41. Therapeutics, Llc (2025) |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12764842&blobtype=pdf | # Salvianolic acid A from Salvia miltiorrhiza identified as a capdependent endonuclease inhibitor for pathogenic arenaviruses
Xiao Gao, Yan Wu, Xiao-Xue He, Guo-Long Liu, Hai-Xia Yang, Jia Lu, Xue-Rui Zhu, Xin-Lan Chen, Chen-Shu Zhao, Hao-Yu Li, Zhong-Fa Zhang, Chan Yang, Shu Shen, Fei Deng, Wei Xu, Shu-Wen Liu, Geng-Fu Xiao, Xiao-Yan Pan
## Abstract
Negative-stranded segmented RNA viruses (NSVs) employ a cap-snatching mechanism for transcription, which makes capdependent endonuclease (CEN) an attractive target for drug development. Pathogenic arenaviruses pose a serious threat to humans, yet no approved treatments exist, underscoring the importance of discovering novel compounds targeting arenaviral CENs. Therefore, this study aimed to identify novel CEN inhibitors for arenaviruses and investigate their antiviral mechanisms. A high-throughput screening system based on enzymatic activity of CEN was established for discovering inhibitors of lymphocytic choriomeningitis virus (LCMV). Several hit compounds were screened from a vast natural product library, and then evaluated for both toxicity and inhibition through cellular and animal experiments. One candidate compound was finally identified, and its mechanism of action on CEN was elucidated through simulation analysis and biochemical studies. Moreover, its broad-spectrum effects were investigated among pathogenic arenaviruses as well as representative NSVs. Consequently, salvianolic acid A (SAA) from Salvia miltiorrhiza was identified as a promising compound that effectively inhibited LCMV infection and significantly reduced the viral load via intravenous administration. It was shown to bind to the active pocket of arenaviral CENs while chelating their metal ions through its acid carboxyl group, acting in a substrate-competitive manner. Additionally, SAA exhibited broad-spectrum inhibition of pathogenic arenaviruses as well as representative viruses from the order Bunyavirales. This study identified SAA as a novel CEN inhibitor, particularly for pathogenic arenaviruses, showcasing its promise for antiviral drug development.
## INTRODUCTION
Murine-borne mammarenaviruses, comprising the Old World (OW) and New World (NW) arenaviruses, sporadically cause outbreaks in Africa and South America, respectively, thereby posing significant threats to human health. These viruses disproportionately affect individuals in resource-limited regions with restricted access to medical care [1,2]. Mammarenaviruses are classified in the family Arenaviridae of the order Bunyavirales [3], among which lymphocytic choriomeningitis virus (LCMV) is globally distributed and can cause severe symptoms, particularly in immuno-compromised individuals and during pregnancies [4]. Other mammarenaviruses include NW arenaviruses Machupo (MACV), Junin virus (JUNV), Guanarito virus (GTOV), Chapare virus (CHAPV), and Sabia virus (SABV), as well as the OW arenaviruses, Lassa virus (LASV), Lujo virus (LUJV), and Dandenong virus (DANV). These are virulent pathogens that cause severe hemorrhagic fever with high mortality and necessitate handling in bio-safety level 4 facilities [5,6]. However, no approved vaccines or drugs are available for most of the aforementioned viruses, thereby highlighting a definite need for drug development.
Arenaviruses, members of the family Arenaviridae, are classical negative-stranded segmented RNA viruses (NSVs). For example, the LCMV genome consists of two segments: the small segment (S) and the large segment (L) [7]. The S segment encodes the glycoprotein (GP) and nucleoprotein (NP), which facilitates viral entry and is essential for genome assembly, respectively. The L segment encodes the matrix protein (ZP) for virion assembly and a large RNA-dependent RNA polymerase (LP) crucial for genome replication and mRNA transcription [8,9]. Specifically, the LP includes three known functional domains: the cap-dependent endonuclease domain (CEN), RNA-dependent RNA polymerase (RdRP) domain, and cap-binding domain (CBD). These domains collectively mediate 5' cap snatching from host RNAs, consequently initiating capdependent transcription and viral mRNA synthesis [10]. This mechanism is characteristic of NSVs, including orthomyxoviruses (e.g., influenza virus) and viruses from families such as Phenuividae, Nairoviridae, Hantaviridae, and Peribunyaviridae (for example, severe fever with thrombocytopenia syndrome virus (SFTSV), Heartland virus (HRTV), Guertu virus (GTV), Crimean-Congo hemorrhagic fever virus (CCHFV), Hantaan virus (HTNV), and La Crosse virus (LACV)) [11,12]. Therefore, CEN is a reliable target for drug development as well as broad-spectrum drug discovery. Currently, baloxavir (BXA) is the only marketed drug targeting the influenza virus CEN. This highlights the urgency for the exploration of more extensivespectrum cap-dependent endonuclease inhibitors (CENis) to combat against a broader range of NSVs.
In recent years, few drug candidates for arenaviruses have advanced to clinical trials [13], except for ARN-75039 which targets the fusion domain of GP and has completed its phase I trial. Additionally, favipiravir (T-705), an anti-influenza virus drug, has been repurposed and evaluated in a phase II trial. Therapeutic antibodies [14 -16], CENis, derived from BXA [17] and other strategies [18,19] were reported to be effective against arenaviruses either in vitro or in vivo. Despite the historical use of ribavirin (RBV) in treating arenavirus infections, conclusive clinical data supporting its efficacy is lacking, and its use is associated with notable potential adverse effects such as thrombocytopenia and anemia [20]. Given these challenges and limitations, there is an urgent need to develop safe and effective antiviral treatments specifically targeting arenaviruses to improve patient outcomes and ensure public health preparedness in the face of potential outbreaks.
In this study, a natural product library comprising 3519 compounds was employed to screen CENis for arenaviruses. This library showcases a diverse range of chemical structures and contains numerous general acids with potential chelation abilities towards the divalent metal ions of CENs. By establishing a highthroughput screening system based on in vitro enzymatic reactions, we unveiled several compounds that effectively inhibit both enzymatic activity and viral infection. Noteworthy among these is the compound salvianolic acid A (SAA), a key pharmacological component of the traditional Chinese medicinal plant Salvia miltiorrhiza, identified as a broad-spectrum antiviral agent against arenaviruses as well as pathogens in the order of Bunyavirales. Subsequent investigations confirmed its capacity to bind to the active pocket of CEN leading to the inhibition of enzymatic function and LCMV infection both in vitro and in vivo. Overall, this study introduces a potential CENi for combating arenaviruses, thus, highlighting the potential of discovering CENis from plant natural products.
## MATERIALS AND METHODS
## Natural product library and compounds
The natural product library was acquired from MedChemExpress (MCE, Shanghai, China). It comprised 5413 natural products, mainly sugars, phenylpropanoids, quinones, flavonoids, terpenoids, glycosides, steroids, alkaloids, phenols, acids, and aldehydes. A smaller library of 3519 compounds, selected by removing compounds with poor solubility or instability, was used in this study. SAA, chebulinic acid, tannic acid, and punicalagin, with purities exceeding 98.8%, were obtained from MedChemExpress (MCE, Shanghai, China). T-705, RBV, and BXA were obtained from Target Mol Chemicals (Shanghai, China).
Cell lines and viruses BHK-21, A549, VERO, and VERO E6 cells used in this study were sourced from the American Type Culture Collection (Manassas, VA, USA). Cells were maintained in Dulbecco's modified Eagle's medium (DMEM) (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS) (Gibco, Grand Island, NY, USA) at 37 °C with 5% CO 2 .
The LCMV, SFTSV, HRTV, GTV, HTNV, and CCHFV strains were obtained from the National Viral Resource Center (Wuhan, China). LCMV displays cytotropism for BHK-21 and A549 cells and is propagated in BHK-21. In contrast, SFTSV, HRTV, GTV, and CCHFV exhibit cytotropism for VERO cells, are propagated in them, and are cultured in DMEM supplemented with 2% FBS. HTNV shows cytotropism for VERO E6 cells and is propagated with them.
## Protein expression purification
The MACV L fragment N-terminal (1-201 aa), LASV L fragment N-terminal (1-169 aa), lymphocytic choriomeningitis virus (LCMV), Junin virus (JUNV), Guanarito virus (GTOV), Chapare virus (CHAPV), Sabia virus (SABV), and LUJV L fragment N-terminal (1-196 aa), DANV L fragment N-terminal (1-171 aa), HTNV L fragment N-terminal (1-180 aa), and LACV L fragment N-terminal (1-185 aa) were inserted into the pET-28a-sumo expression vector. The gene sequence encoding amino acids 1-210 at the N-terminal of the SFTSV L fragment was cloned into the pET-42a vector, while the gene sequence covering amino acids 587-782 at the N-terminal of the CCHFV L fragment was inserted into the pGEX-6P-1 vector. The plasmids were separately introduced into BL21 (DE3) strains (Beyotime, Shanghai, China). LCMV CEN expression was induced with 1 mM isopropyl-β-d-thiogalactoside (IPTG) at 37 °C for 4 h. The expression of other CENs was induced with 0.5 mM IPTG at 16 °C for 20 h. Following centrifugation to harvest the bacterial cells, sonication was employed for cell lysis. Supernatants were obtained after high-speed centrifugation to separate the precipitate, and the soluble proteins were subsequently purified via nickel column or GST column affinity chromatography. The mutant LCMV CEN proteins underwent purification using conditions identical to those of the wild-type.
High-throughput screening by fluorescence resonance energy transfer (FRET) In a 384-well plate, 1 µM LCMV CEN proteins diluted in reaction buffer (50 mM HEPES, 150 mM KCl, 1 mM MnCl 2 , pH = 7.8) was coincubated with 50 µM compounds at 37 °C for 30 min in a 20 µL reaction system. Subsequently, the single-stranded RNA substrate (5'FAM-AGGAAGAUUAAUAAUUUUUUUUUCCU-BHQ13') at a final concentration of 0.3 µM was introduced, and the fluorescence signals were immediately measured at λ ex /λ em = 485 nm/ 535 nm (Synergy H1, BioTek, Winooski, VT, USA). LCMV CEN inhibition was assessed by measuring changes in fluorescence intensity and calculated using the following formula:
where V is the enzyme reaction rate, t 1 is the reaction starting time, t 2 is the reaction ending time, F 1 is the fluorescence value at t 1 , and F 2 is the fluorescence value at t 2 .
Inhibitionð%Þ ¼ ðV experimental group =V control group Þ 100 Cellular antiviral assay The cells were seeded at a density of 1 × 10 4 cells per well in 96well plates overnight. Preceding viral inoculation, cells were incubated with gradient-diluted compounds for 1 h. BHK-21 and A549 cells were infected with LCMV at a multiplicity of infection (MOI) of 0.1, while VERO cells were infected with SFTSV, HRTV, GTV, and CCHFV at an MOI = 0.1, A549 cells were infected with HTNV at MOI = 1. Following 1 h incubation, the supernatants were removed, and cells were replenished with 100 µL of culture medium containing the respective compound concentrations. After 48 h, the supernatants were harvested for viral copy quantification using real time-quantitative PCR (RT-qPCR).
In preparation for RT-qPCR analysis, viral RNA from the culture supernatants was extracted using an automatic nucleic acid extractor with a compatible kit (Vazyme, Nanjing, China) in accordance with the manufacturer's guidelines. Viral RNA copies were then quantified using a one-step quantification reagent (Vazyme, Nanjing, China), along with specific primers and template plasmids containing the relevant genes, on an ABI platform (QuantStudio 6 Pro, Thermo Fisher, Waltham, MA, USA).
$$V ¼ ðF 2 À F 1 Þ=ðt 2 À t 1 Þ$$
## Cytotoxicity assay
Cells were seeded at a density of 1 × 10 4 cells per well in 96-well plates overnight. The following day, cells were treated with gradientdiluted compounds. Forty-eight hours later, the cultural supernatants were aspirated, and 100 µL diluted Cell Counting Kit-8 reagent (CCK8) (GLPBIO, Montclair, CA, USA) was added to each well. After incubation at 37 °C for 1 h, the absorbance at 450 nm was measured to assess cell viability (Synergy H1, BioTek, Winooski, VT, USA).
## Acrylamide-urea gel electrophoresis
The CEN of LCMV, MACV, or LASV was incubated with SAA at concentrations of 50, 17, 5.6, and 0 µM for 30 min. Next, a singlestranded RNA substrate (sequence: 5'-AGGAAGAUUAAUAAUU UUCCU-3') was introduced, and cleavage of the RNA substrate was assessed via acrylamide-urea gel electrophoresis after 5, 10, and 20 min. The acrylamide-urea gel was formulated with 4.5 mL of 40% polyacrylamide solution, 1 mL of 10× Tris-Borate-Ethylenediaminetetraacetic acid (EDTA) (TBE) buffer, 4.2 g urea, 0.05 mL of 10% ammonium persulphate, and 0.005 mL of coagulant promoter. Diethyl pyrocarbonate (DEPC) water was added to reach a final volume of 10 mL. Samples from different time points were loaded into the gel wells. RNA electrophoresis was then conducted in TBE buffer at 180 V for 0.5 h on ice. Images were captured using a ChemiDoc MP Imaging System (ChemiDoc Touch, Bio-Rad, Hercules, CA, USA).
## Surface plasmon resonance (SPR)
The affinity of compounds to CEN proteins was measured using the Biacore 1 K system (Cytiva, Marlborough, MA, USA) at 25 °C. CENs of LCMV, MACV, LASV, SFTSV or mutant LCMV were immobilized on the CM5 chip using an amino coupling kit (Cytiva, Marlborough, MA, USA) with a loading capacity of approximately 15,000 RU. SAA was diluted at concentrations of 400, 200, 100, 50, 25, 12.5, and 6.25 µM, and flowed over the chip at a rate of 30 µL/ min for 120 s of association and 90 s of dissociation. Data were analyzed using the Biacore evaluation software (version 1 K) fitted to a 1:1 binding model, and the fitted curves were drawn using GraphPad Prism 9.0.
## Enzyme kinetic analysis
The effects of SAA on the kinetics of CENs derived from LCMV, MACV, and LASV were investigated using a multifunctional automatic microplate reader (Synergy H1, BioTek, Winooski, VT, USA). Fluorescence values were measured by fixing the substrate concentration at 1 µM while altering the CEN concentration from 0.2 to 1.4 µM. Alternatively, fluorescence values were measured by fixing the enzyme concentration while altering the substrate concentration. The reaction rates at different enzyme concentrations and at different substrate concentrations were fitted separately. Kinetic equations for the enzymatic reactions were used to determine the interference mode of the compounds on CENs.
## Molecular docking
The 3D structure of SAA was retrieved in SDF format from the PubChem database, and the SDF file was subsequently converted to PDB format. Protein PDB structures of LCMV CEN (5t2t), MACV CEN (7elc), and LASV CEN (4miw) were obtained from the RCSB PDB. The compound and protein PDB files were converted to PDBQT files using the AutoDock Vina software. Boxes of the CEN active pocket were then established for docking, and conformations with relatively low binding energies were chosen for analysis using the online PLIP tool and mapped using PyMOL.
Molecular dynamics (MD) simulations MD simulations were conducted using Gromacs software (version 2022.3). The GAFF force field was incorporated into the small molecules via AmberTools22, followed by hydrogenation and calculation of RESP potential using Gaussian16W. Simulations were carried out at a fixed temperature of 300 K and atmospheric pressure (1 bar), applying the Amber99sb-ildn force field, with water (Tip3p water model) as the solvent. The system's total charge was balanced through the addition of Na + ions. Initially, the molecular dynamics simulation system underwent energy minimization using the steepest descent method, followed by 100,000 steps of isothermal-isovolumic system equilibrium and isothermal-isobaric system equilibrium, each maintaining a coupling constant of 0.1 ps over a duration of 100 ps. Subsequently, a free molecular dynamics simulation was carried out for 5,000,000 steps with a step size of 2 fs, totaling a duration of 100 ns. Following the simulation, trajectory analysis was performed using the software tool, the root mean square deviation of each amino acid trajectory was calculated.
In vivo efficacy experiments All animal experimental procedures were performed according to ethical guidelines and were approved by the Animal Care Committee of the Wuhan Institute of Virology (WIVA25202305). Female BALB/c mice aged 6-8 weeks were obtained from GemPharmatech (Nanjing, China) and housed in a pathogen-free facility. The animal challenge experiments were performed in a biosafety Level 2 laboratory. Mice were randomly divided into groups (n = 5), and each group received an intraperitoneal injection of 1 × 10 5 plaque-forming units (PFUs) of LCMV. Upon challenge, mice in the vehicle group were administered the solvent, and mice in the positive control group were orally administered 300 mg/kg T-705 or 30 mg/kg RBV. In the treatment groups, mice were orally administered SAA, chebulinic acid, tannic acid, and punicalagin at a dose of 20 mg/kg, or severally administered SAA through intragavage (i.g.), intraperitoneal injection (i.p.), and tail vein injection (i.v.) at both 20 mg/kg and 40 mg/kg. The treatments were given once daily for 3 d. At the experimental endpoint, the liver or spleen was dissected for viral RNA extraction using a QIAGEN RNA extraction kit (QIAGEN, Venlo, Netherlands), and viral RNA copies were quantified as described above.
## Pharmacokinetic experiment
The pharmacokinetics of SAA were investigated in rats. Six female rats were divided evenly into two groups. SAA was administered via both i.g. and i.v. at a dose of 40 mg/kg, and blood samples were collected at 2, 10, 20, and 30 min and at 1, 2, 4, 6, and 24 h post-administration. Plasma samples (10 µL) were mixed with an internal standard (25 µL), followed by the addition of acetonitrile (300 µL), vortexed for 10 min, and centrifuged at 1700× g for 15 min in a refrigerated high-speed centrifuge at 4 °C. The supernatant (50 µL) was mixed with acetonitrile and water (2:8 ratio) containing 100 µg/mL VC, vortexed for 5 min, and prepared for LC-MS/MS analysis using the TRIPLE QUAD 6500 + AB SCIEX system (Sciex, Framingham, MA, USA). Pharmacokinetic parameters were calculated using the WinNonlin software (Version 8.3) by applying the statistical method of moments.
## Statistical analysis
Data were obtained from at least three independent biological replicates, unless otherwise specified. Statistical analyses were conducted using the t-test or one-way ANOVA, with a Pvalue < 0.05 considered statistically significant. Plot and histograms were generated and visualized using GraphPad Prism 9.0. Statistical significance was indicated as follows: ns for P > 0.05, * for P < 0.05, ** for P < 0.01, and *** for P < 0.001.
## RESULTS
SAA was hit from a natural product library for LCMV A high-throughput screening system for LCMV CEN was established in a 384-well plate, optimized for enzyme concentration, substrate selection, and reaction duration. A standardized set of conditions (enzyme concentration at 1 µM, substrate concentration at 0.3 µM, and a reaction time of 1 h) in a 20 μL reaction system was validated and utilized for swift screening, following the entire screening process (Fig. 1a). Initially, 150 compounds were selected from the 3519 natural compounds based on a threshold of ≥40% anti-CEN inhibition at 50 μM (Fig. 1b). Subsequently, a dual criterion based on both antiviral inhibition on live LCMV and anti-CEN inhibition was applied to identify hits, resulting in 23 selected compounds (the experiments were independently performed at least three times, Fig. 1c). The 23 hits were tested in BHK-21 cells to evaluate antiviral efficacy and cytotoxicity, calculating 50% effective concentration (EC 50 ) and 50% cytotoxic concentration (CC 50 ) values. Selectivity indices (SI, CC 50 /EC 50 ) are detailed in Table 1. Notably, four compounds-SAA, chebulinic acid, tannic acid, and punicalagin exhibited promisingly low EC 50 values and high CC 50 values, with SIs exceeding 30.
The chemical structures of the four compounds, depicted in Fig. 1d, were used for further study. SAA exhibited the most favorable EC 50 values in antiviral assays conducted on BHK-21 (the experiments were independently performed at least three times, Fig. 1e) and A549 cells (the experiments were independently performed at least three times, Fig. 1f), while tannic acid showed relatively high cytotoxicity. Additionally, immunofluorescence analysis performed on BHK-21 and A549 cells (n = 3, Supplementary Figs. S1a andS1b) illustrated concentration-dependent viral inhibition by the four compounds, with SAA achieving complete inhibition on LCMV infection at a relatively low concentration. Therefore, SAA was selected for further research.
In vivo antiviral effect of SAA on LCMV The in vivo efficacy of SAA and three other compounds was evaluated in BALB/c mouse model. Viral copies in the targeted organs were detected using RT-qPCR 3 days post-infection (dpi) Fig. 1 Four hit compounds were identified from a natural product library, and their antiviral activity against LCMV infection was assessed. a Flowchart of the screening process for cap-dependent endonuclease (CEN) inhibitors from a natural product library, based on both in vitro anti-CEN assays and cellular antiviral assays. b Inhibition of the enzymatic activity of the lymphocytic choriomeningitis virus (LCMV) CEN was detected using our established high-throughput screening system; 150 compounds at 50 µM inhibited LCMV CEN, with over 40% selected for cellular analysis. c The inhibition of viral infection was detected using real-time quantitative PCR; 23 compounds at 50 µM inhibited LCMV infection, with over 40% selected from the initial 150. d Chemical structures of the four hits: salvianolic acid A (SAA), chebulinic acid, tannic acid, and punicalagin. e, f Antiviral effects and cytotoxicity of SAA, chebulinic acid, tannic acid, and punicalagin on LCMV were determined using BHK-21 cells (e) and confirmed using A549 cells (f); EC 50 s and CC 50 s, noted along with the curves, were calculated from at least three independent biological replicates.
after diverse drug treatments (Fig. 2a). Mice challenged with LCMV were treated with 20 mg/kg of SAA, chebulinic acid, tannic acid, punicalagin, or 300 mg/kg of T-705 [21], via i.g. The body weight of the mice in the vehicle group decreased for 3 days, while this decrease was halted by T-705 and SAA (Fig. 2b). In addition, T-705 and SAA significantly reduced viral copies in the liver, unlike the other three treatments, which had no effect (Fig. 2c).
Further experiments were conducted to investigate the influence of the administration route on the efficacy of SAA. Specifically, mice were subjected to i.g., i.p., and i.v. of either 20 or 40 mg/kg of SAA once a day, and viral copies from the liver and spleen were taken for analysis at 3 dpi (Fig. 2a). It was found that the positive control, T-705, significantly decreased viral loads in both the liver and spleen by approximately two logs, while RBV [17] only mildly reduced viral loads in the liver, which is consistent with previous reports [21,22]. For SAA, i.v. administration reduced viral loads in the liver or spleen by 1-1.5 logs, while i.g. administration showed mild reductions, and i.p. administration was ineffective (Fig. 2d). Pharmacokinetic analysis in rats revealed that the half-life of SAA following i.v. administration was 2.34 h, which exceeded that observed with i.g. administration. This discrepancy suggests that the in vivo effectiveness of SAA is impacted by its bioavailability across diverse administration routes, thereby implying the suboptimal oral bioavailability of SAA (Fig. 2e). Collectively, these data demonstrate the definite in vivo efficacy of SAA on LCMV and highlight the necessity for an optimized administration route for SAA.
## SAA inhibited LCMV CEN in a substrate-competing manner
To elucidate the underlying mechanism, a series of biochemical assays were performed on LCMV CEN. It was observed that, in the presence of equal concentrations of single-stranded RNA, SAA markedly impeded the cleavage of LCMV CEN on the RNA substrate in a dose-dependent and time-dependent manner (n = 3, Fig. 3a and Supplementary Fig. S2a). FRET based enzymatic assay revealed that SAA inhibited LCMV CEN with an IC 50 at the micromolar level (n = 3, Fig. 3b). Moreover, the mechanism of action of SAA on CEN was investigated through a series of traditional enzyme kinetic experiments with varying enzyme concentrations at a fixed substrate level, and substrate concentrations at a fixed enzyme level. Enzyme kinetic experiments collectively demonstrated that the interaction between SAA and LCMV CEN was reversible and involved substrate-competition (n = 3, Fig. 3c).
Subsequent surface plasmon resonance (SPR) experiments demonstrated that SAA bound to LCMV CEN with micromolarlevel affinity, thereby indicating rapid association and dissociation kinetics (Fig. 3d). Moreover, molecular docking analysis suggested that SAA interacted with key residues, such as Lys43, Ser46, Ile86, Asp88, Glu101, Cys102, Lys114, and Lys121, situated within the active site of LCMV CEN. Additionally, the docking analysis suggested that SAA chelates Mn 2+ ions in the active center, with a predicted binding energy of -7.9 kcal/mol (Fig. 3e). For further validation of the molecular docking outcomes of SAA and LCMV CEN, MD simulations indicated that SAA maintained stability upon binding to LCMV CEN for a duration of 10 ns (Fig. 3f). These findings confirm that SAA exerts antiviral effects by targeting LCMV CEN, competitively binding to the enzymatic active site and chelating divalent metal ions.
SAA exerted a broad-spectrum effect on pathogens in the order Bunyavirales including arenaviruses Thereafter, we explored the broad-spectrum effects of SAA on representative NSVs in Arenaviridae, Phenuiviridae, Nairoviridae, Hantaviridae, and Peribunyaviridae in the order Bunyavirales (Fig. 4a). Initially, we analyzed the primary and secondary structures of CENs from LCMV, NW arenavirus MACV, and OW arenavirus LASV, revealing classical features of the PD-D/ExK nuclease superfamily and over 70% sequence similarity (Fig. 4b). Further research revealed that SAA notably inhibited the cleavage of single-stranded RNA substrates by MACV CEN and LASV CEN within 20 min (n = 3, Fig. 4c). The IC 50 values of SAA on MACV CEN and LASV CEN were 6.9 µM and 6.5 µM, respectively (n = 3, Fig. 4d), which was comparable to LCMV CEN. Moreover, SAA competitively inhibited MACV CEN and LASV CEN in a substratecompeting manner (n = 3, Fig. 4e) and bound to MACV CEN and LASV CEN with K D of 85.4 μM and 46.0 μM, respectively (Fig. 4f). Subsequent results revealed comparable inhibitory effects of SAA on the CEN of JUNV, GTOV, CHAPV, SABV, LUJV, and DANV within the family Arenaviridae, with its potency exceeding that of BXA (n = 3, Fig. 4g and Supplementary Fig. S2b). These findings indicated that SAA possesses broad-spectrum anti-CEN effects against the pathogenic arenaviruses.
Additionally, we investigated the antiviral activity of SAA against NSVs beyond Arenaviridae. The effects of SAA on Phenuiviridae were determined using cellular antiviral assays, revealing significant antiviral activity of SAA against SFTSV (n = 3, Fig. 5a andb). Subsequent enzymatic assays revealed that SAA effectively inhibited SFTSV CEN, with an IC 50 of 7.8 µM (n = 3, Fig. 5c). Moreover, the docking analysis predicted that it bound to the active pocket of SFTSV CEN with an affinity of -7.1 kcal/mol (Fig. 5d), and SPR experiment indicated a robust affinity between SAA and SFTSV CEN, with a K D = 33.5 μM (Fig. 5e). MD simulations further validated the outcomes of SAA docking with SFTSV CEN, demonstrating the complex reaching equilibrium after 60 ns (Fig. 5f). These results collectively indicate that SAA exhibits antiviral activity by targeting SFTSV CEN. Subsequently, GTV and HRTV, both belonging to the same family as SFTSV, were selected to confirm the antiviral activity. Our results also showed that SAA displayed antiviral activity against GTV and HRTV (n = 3, Fig. 5g). Besides, we chose CEN of CCHFV from Nairoviridae, HTNV from Hantaviridae, and LACV from Peribunyaviridae, to further validate the broad-spectrum effects of SAA. The inhibitory and antiviral activities of SAA against the CEN of NSVs across various families were also confirmed (n = 3, Fig. 5h andi). From the IC 50 s (n = 3, Supplementary Fig. S2c), we found that SAA exhibited comparative inhibitory activity on CENs from Arenaviridae, Phenuiviridae, Hantaviridae, and Peribunyaviridae, while demonstrating suboptimal inhibitory activity on CEN from Nairoviridae, indicating a limited selectivity of SAA in vitro. In conclusion, these findings illustrated the broad-spectrum inhibition and potential antiviral activity of SAA against NSVs in the order of Bunyavirales.
## Binding mechanism of SAA on LCMV cap-dependent endonuclease
We conducted a detailed analysis to validate the binding sites and interaction patterns of SAA, enhancing the understanding of how it influences LCMV CEN. From the target view, we utilized docking analysis to predict the binding sites on CEN across the six most favorable conformations, followed by quantifying the frequency of In detail, the mice either daily received 20 mg/kg SAA, chebulinic acid, tannic acid, and punicalagin via intragavage (i.g.), or daily received 20 or 40 mg/kg SAA via i.g., intraperitoneal injection (i.p.) or tail vein injection (i.v.) upon challenge of 1 × 10 5 plaque-forming units of LCMV via i.p. Mice received 300 mg/kg Favipiravir (T-705, a nucleic acid analog drug with broad-spectrum antiviral activity to target viral RNA-dependent RNA polymerase) or 30 mg/kg of Ribavirin (RBV) daily via i.g., as controls. Viral copies in spleens or livers were taken for analysis at 3 days post-infection (dpi). b Body weight changes in mice after treatment with 20 mg/kg of SAA, chebulinic acid, tannic acid, or punicalagin via i.g. upon challenge. 300 mg/kg T-705 administered via i.g. was used as a positive control. c, d Viral copies in livers and spleens were determined at 3 dpi using RT-qPCR through the standard curve method, after the treatments of 20 or 40 mg/kg SAA via i.g., i.p., or i.v., or the treatment of 300 mg/kg T-705 or 30 mg/kg RBV via i.g. Data were represented as mean ± standard deviation from five animals. e Metabolic kinetic data of SAA in rats (n = 3) were determined through both i.g. and i.v. administration routes. The table presents the half-life (t 1/2 ), maximum concentration time (T max ), maximum blood concentration (C max ), area under curve (AUC), and bioavailability (F) of SAA across various administration methods. Data were represented as mean ± standard deviation from three animals. amino acid interactions with SAA. As a result, we identified 17 amino acids that likely to interact with SAA. Among these, residues Lys43, Arg47, Ile86, Asp88, Glu101, Lys114, and Lys121 were observed three or more times and were subsequently subjected to affinity analysis (Figs. 4b and6a). Significantly, these amino acids represent key sites within the active center of CEN. Specifically, Lys43, Asp88, Glu101, Lys114, and Lys121 were conserved across the CENs of LCMV, MACV, and LASV. Furthermore, Asp88, Glu101, Lys114, and Lys121 were conserved across all arenaviruses [23]. Through a series of SPR assays, we observed that mutations at Glu101 and Lys114 significantly reduced the affinity of SAA for CEN, whereas mutations at the remaining five sites had a less pronounced impact (Fig. 6b). These findings confirm that SAA definitely interacts with CEN, thus, suggesting a potential mechanism for its broad-spectrum effects.
From the compound perspective, we examined the functional group responsible for chelating the divalent metal ions within CENs leading to deactivation of the metalloenzyme. Utilizing a derivative of SAA known as methyl salvionolate A (MSA), where the carboxyl group had been esterified by methanol thereby eliminating its acidity and chelating ability, we substantiated the role of the lone exposed carboxyl group of SAA, which was predicted to interact with the Mn 2+ ion in LCMV CEN (Fig. 6c). Consequently, MSA displayed a weaker affinity (-6.9 kcal/mol) for LCMV CEN than that for SAA (-7.9 kcal/mol) (Fig. 6d). Moreover, MSA entirely abrogated its inhibitory impact on LCMV CEN and its antiviral efficacy against LCMV (n = 3, Fig. 6e). These findings collectively elucidated the action mechanism of SAA on CENs, thus, indicating the prospect of SAA as a promising compound for drug development.
Fig. 3 SAA inhibits LCMV CEN activity through competing substrate. a The inhibitory effect of SAA on the cleavage of LCMV CEN on ssRNA substrate was examined at 5, 10, and 20 min via acrylamide-urea gel electrophoresis. Gray analysis of the corresponding strips was calculated for the inhibition rate, shown in the right panel. The divalent metal ion chelator, EDTA, was used as a positive control. The representative image was derived from three replicate experiments. b Dose-dependent inhibitory effect of SAA on LCMV CEN activity was detected using fluorescence resonance energy transfer (FRET), and an IC 50 was calculated based on the fluorescent signal value. Baloxavir (BXA) was used as a control. Data were from three independent experiments. c Enzymatic reaction rate was determined by adding different concentrations of LCMV CEN (left) to the reaction system while maintaining a constant substrate concentration, or adding different concentrations of substrate (right) to the reaction system in the presence of 1 µM of LCMV CEN. SAA was used as 5, 10, and 20 µM. Data were from three independent experiments. d Affinity of SAA to LCMV CEN was measured using surface plasmon resonance (SPR); K D was calculated and is shown in the fitted curves. Data were from once representative experiment. e Interaction between SAA and LCMV CEN was analyzed by docking. SAA is shown in green, the blue sticks indicate amino acids interacting with SAA, and the Mn 2+ is shown in purple. f Molecular dynamics simulations of SAA with LCMV CEN at 100 ns.
## DISCUSSION
Cap-snatching mediated by viral endonuclease is a distinctive mechanism utilized by NSVs. This mechanism a critical role in viral transcription, thus, rendering CEN a critical target for drug discovery and development [11]. Until now, only BXA has been approved for the clinical treatment of influenza virus infection due to its excellent efficacy and pharmacokinetics [24]. However, BXA has been shown to induce mutations in CEN leading to drug resistance [25]. In addition, BXA has been demonstrated to inhibit SFTSV and HRTV with EC 50 s of 0.26 µM and 0.25 µM, respectively, and inhibit LCMV and JUNV with an EC 50 > 1 µM [26]. Thus, there is a pressing need of CENis that are effective against NSVs beyond influenza virus.
Recent studies have highlighted three main routes for discovering CENis: high-throughput screening of metal-chelating agent compound libraries [17], de novo synthesis based on rational design [27], and screening of natural products that contain diverse general acids [28]. Historically, natural products and their structural analogs have been pivotal in therapies for cancer and infectious diseases [29]. Salvianolic acid C derived from Salvia miltiorrhizafrom has been previously proved to inhibit severe acute respiratory syndrome coronavirus 2 by blocking membrane fusion [30]. Our previous study has also screened a mini natural product library containing 71 compounds, and found that flavonoids such as tanshione I from Salvia miltiorrhiza exhibits broad-spectrum antiviral effects on NSVs including SFTSV, influenza A virus, and LCMV, which was associated with CEN inhibition [28]. In addition, our recent work has also revealed Licoflavone C as an alternative inhibitor of SFTSV, thereby offering insights into targeting CEN with flavonoids in drug discovery [31].
Here, we reported a polyphenol component of Salvia miltiorrhiza, called SAA, inhibiting CENs from viruses in Arenaviridae, Phenuiviridae, Hantaviridae, Peribunyaviridae, and Nairoviridae, through screening a large natural compound library containing 3519 compounds. In comparison to the previous study, SAA discovered in this study exhibits a broader antiviral spectrum and higher efficacy against NSV CENs in comparison to tanshinone I and Licoflavone C. Moreover, the mechanism of SAA was shown to bind with conserved sites in the CEN active pocket while chelating metal ion via the carboxyl group, thus, demonstrating explicit in vitro and in vivo effects. In view of the above, natural products from Salvia miltiorrhiza are important resources for antiviral discovery, and medicinal ingredients such as flavonoids and polyphenols are worth in-depth study.
To the best of our knowledge, this study represents the first report of SAA as a novel CENi to inhibit NSVs, although previous studies have reported its multifaceted roles. Specifically, SAA alleviates oxidative stress-induced osteoporosis by acting as an antioxidant and modulating bone metabolic pathways [32]. Moreover, in the context of acute cerebral ischemia-reperfusion injury, SAA migrates neuroinflammation by inhibiting the microglial TLR2/4 pathway [33]. Additionally, SAA has been shown to alleviate heart failure with preserved ejection fraction (HFpEF) through the modulation of TLR/Myd88/TRAF/NF-κB and p38MAPK/CREB signaling pathways [34]. Thus, SAA has been documented to exert anti-inflammatory and antioxidant effects. Studies have indicated that viral infection triggers a cascade of pro-inflammatory factors by activating the Toll-like receptor (TLR) and RIG-I-like receptor (RLR) pathways, and that mitochondrial dysfunction leads to burst of reactive oxygen species (ROS) [35,36]; therefore, it is plausible to hypothesize that SAA might play a role in viral infections through its anti-inflammatory and antioxidant properties. Thus, the contribution of off-target effects due to the anti-inflammatory actions of SAA cannot be entirely excluded in this study. In addition to the aforementioned antiinflammatory effects, SAA may exert a direct inactivation effect on enveloped RNA/DNA viruses by interfering with the phospholipid bilayer, thereby representing an inhibition at the viral entry stage by disrupting the virion integrity. This mechanism is widely found in natural products [37][38][39], and may also constitute a part of its common anti-viral mechanism. Thus, an off-target effect resulting from envelope glycoprotein interference may also contribute to the antiviral performance of SAA against arenavirus in this study.
Moreover, previous studies have also demonstrated that SAA exerts antiviral activity by targeting envelope glycoproteins, for example, binding specifically to glycoprotein B of HSV and the receptor-binding domain (RBD) of SARS-CoV-2 [40,41] to block viral entry. Thus, SAA may interact with different viral proteins to exhibit unique mechanisms while combating various viruses. In this study, SAA exhibited a high affinity for the arenavirus CEN (K D = 5.1 µM) demonstrating an antiviral effect by inhibiting viral replication. Consequently, we reported a unique mechanism of its antiviral activity, which may enhance the understanding of SAA pharmacology.
Although the off-target effects of natural products are often considered drawbacks compared to targeted synthetic compounds, these effects may offer unexpected benefits in combating complex viral infections because they allow for targeting multiple viral replication stages simultaneously. Despite this, data on the toxicity of SAA in dogs has indicated that SAA lacks mutagenic properties but exhibits hepatotoxicity and nephrotoxicity; however, these effects have been shown to be transient and reversible [42]. Hence, further research on developing SAA as an antiviral targeted at CEN must fully consider its potential primary and side effects, as well as the mechanism of action derived from its multitarget properties.
Notably, our pharmacokinetic investigation of SAA in rats unveiled a markedly low oral bioavailability (F) utilization of SAA (F = 1%), aligning with previous research on the in vivo metabolic pathways of SAA. Thus, an effective concentration might not be achieved in the liver or spleen through administration via i.g. [43]. This low bioavailability can be attributed to poor membrane permeability in the gastrointestinal tract and predominant fecal metabolism [44]. Furthermore, existing clinical studies on SAA have demonstrated its favorable tolerability in humans within a single dose range of 10-300 mg [43]. Hence, SAA holds promise for development as an antiviral drug targeting NSVs CEN. Further research and consideration are needed to address the challenges related to enhancing its oral bioavailability and antiviral activity. A potential optimization strategy involves adding large side-chain groups to SAA to enhance binding affinity and activity at the CEN active site. The other is to reduce the hydroxyl groups of SAA to retard metabolism, thus, improving its in vivo bioavailability and safety. Although modifications based on natural products can be challenging, increasing plant-or microbe-based fermentation techniques may help [45].
In this study, we performed in-depth research on the mechanism of action of SAA on LCMV CEN by both biochemical and computational analysis, and identified key amino acid sites in Fig. 4 Broad-spectrum inhibitory effect of SAA on arenaviral CENs. a Phylogenetic tree of the pathogens from family Phenuiviridae, Nairoviridae, Arenaviridae, Hantaviridae, and Peribunyaviridae in the order of Bunyavirales were made based on L protein sequences, pathogens mentioned in this study were marked. b Alignment of the primary amino acid sequences and secondary spatial structures of LCMV, MACV, and LASV CENs, and analysis on enzyme active pocket. Blue triangles indicate the active sites of LCMV CEN. Red triangles indicate mutational sites of LCMV CEN. Amino acid sites of PD-D/E(X)K are marked by red pentagrams. c The inhibitory effect of SAA on the cleavage of MACV and LASV CENs on ssRNA substrate was examined at a number of different time points via acrylamide-urea gel electrophoresis. Gray analysis of the corresponding strips was calculated into the inhibition rate, shown in the right panel. The divalent metal ion chelator, EDTA, was used as a positive control. The representative image was derived from three replicate experiments. d IC 50 of SAA on MACV and LASV CENs were determined using FRET. BXA was used as a control. e SAA inhibited MACV and LASV CENs in a substrate-competing manner. f Affinity of SAA to MACV and LASV CENs was detected using SPR. Data were from once representative experiment. g In vitro inhibitory activity of SAA against CHAPV CEN, DANV CEN, GTOV CEN, JUNV CEN, LUJV CEN and SABV CEN in Arenaviridae. BXA was used as a control. Data from enzymatic assays were obtained from at least three independent experiments. the active pocket of LCMV CEN engaging with SAA binding, where a naked SAA carboxyl group might chelate with the divalent ions in it. Interestingly, we found that mutations at positions Glu101 and Lys114 noticeably weakened the affinity of SAA to LCMV CEN. Importantly, these two sites are not only conserved among CENs of LCMV, MACV, and LASV, but also conserved among CENs of all arenaviruses [23]. Additionally, the broad-spectrum effects were validated by enzymatic or antiviral experiments on representative NSVs from different families in the order Bunyavirales. Besides, SAA demonstrated superior inhibitory activity against NSV CENs compared to BXA. Notably, we found that SAA shares a similar ternary ring structure with BXA and its analogs. These broadly target arenaviral CENs when the hydroxyl group at position 24 of SAA forms a ring with the carbon atom at position 32 on the benzene ring [17]. This potentiates the applicability of SAA and implies a possibility of discovering broad-spectrum antivirals against CEN based on natural products, although this should be subjected to further modification. Moreover, the variations in IC 50 s and EC 50 s between different viruses' prompt concerns about the potential off-target effects of SAA, which should be considered and possibly addressed by structure-activity optimization.
This study examined a wide range of natural products and identified SAA as a promising compound. SAA demonstrated its in vitro and in vivo efficacy, broad-spectrum antiviral activity, and detailed mechanism of action on CEN. These findings underscore the potential of SAA against pathogenic arenaviruses, and highlight the feasibility of targeting CENs for drug discovery against bunyaviruses.
## References
1. Charrel, Lamballerie (2003) "Arenaviruses other than Lassa virus" *Antivir Res*
2. Luo, Lu, Qin et al. (2023) "Emergence of an ancient and pathogenic mammarenavirus" *Emerg Microbes Infect*
3. (2024) "International Committee on Taxonomy of Viruses (ICTV"
4. Kinori, Schwartzstein, Zeid et al. (2018) "Congenital lymphocytic choriomeningitis virus-an underdiagnosed fetal teratogen" *J Am Assoc Pediatr Ophthalmol Strabismus*
5. Mclay, Liang, Ly (2014) "Comparative analysis of disease pathogenesis and molecular mechanisms of New World and Old World arenavirus infections" *J Gen Virol*
6. Shao, Liang, Ly (2015) "Human hemorrhagic Fever causing arenaviruses: molecular mechanisms contributing to virus virulence and disease pathogenesis" *Pathog Basel Switz*
7. Riviere, Ahmed, Southern et al. (1985) "The S RNA segment of lymphocytic choriomeningitis virus codes for the nucleoprotein and glycoproteins 1 and 2" *J Virol*
8. Kranzusch, Schenk, Rahmeh et al. (2010) "Assembly of a functional Machupo virus polymerase complex" *Proc Natl Acad Sci*
9. Morin, Coutard, Lelke et al. (2010) "The N-terminal domain of the arenavirus L protein is an RNA endonuclease essential in mRNA transcription" *PLoS Pathog*
10. Velthuis, Grimes, Fodor (2021) "Structural insights into RNA polymerases of negative-sense RNA viruses" *Nat Rev Microbiol*
11. Olschewski, Cusack, Rosenthal (2020) "The cap-snatching mechanism of Bunyaviruses" *Trends Microbiol*
12. Holm, Kopicki, Busch et al. (2018) "Biochemical and structural studies reveal differences and commonalities among cap-snatching endonucleases from segmented negative-strand RNA viruses" *J Biol Chem*
13. Cross, Fenton, Woolsey et al. (2024) "Monoclonal antibody therapy protects nonhuman primates against mucosal exposure to Lassa virus" *Cell Rep. Med*
14. Pan, Wu, Wang et al. (2020) "Development of horse neutralizing immunoglobulin and immunoglobulin fragments against Junín virus" *Antivir Res*
15. Pan, Wu, Wang et al. (2018) "Novel neutralizing monoclonal antibodies against Junin virus" *Antivir Res*
16. Toba, Sato, Kawai et al. (2022) "Identification of capdependent endonuclease inhibitors with broad-spectrum activity against bunyaviruses" *Proc Natl Acad Sci*
17. Zhang, Cao, Cai et al. (2019) "Structure-activity relationship optimization for lassa virus fusion inhibitors targeting the transmembrane domain of GP2" *Protein Cell*
18. Lan, Zhang, Jia et al. (2022) "Screening and identification of Lassa virus endonuclease-targeting inhibitors from a fragmentbased drug discovery library" *Antivir Res*
19. Gowen, Naik, Westover et al. (2021) "Potent inhibition of arenavirus infection by a novel fusion inhibitor" *Antivir Res*
20. Hickerson, Westover, Jung et al. (2018) "Effective treatment of experimental lymphocytic choriomeningitis virus infection: consideration of favipiravir for use with infected organ transplant recipients" *J Infect Dis*
21. Zadeh, Afowowe, Abe et al. (2022) "Potential and action mechanism of favipiravir as an antiviral against Junin virus" *PLoS Pathog*
22. Saez-Ayala, Yekwa, Carcelli et al. (2018) "Crystal structures of Lymphocytic choriomeningitis virus endonuclease domain complexed with diketo-acid ligands" *IUCrJ*
23. Abraham, Morton, Saravolatz (2020) "Baloxavir: a novel antiviral agent in the treatment of influenza" *Clin Infect Dis Publ Infect Dis Soc Am*
24. Hickerson, Petrovskaya, Dickensheets et al. (2023) "Impact of baloxavir resistance-associated substitutions on influenza virus growth and drug susceptibility" *J Virol*
25. Wang, Shin, Zhang et al. (2020) "The capsnatching SFTSV endonuclease domain is an antiviral target" *Cell Rep*
26. Credille, Dick, Morrison et al. (2018) "Structure-activity relationships in metal-binding pharmacophores for influenza endonuclease" *J Med Chem*
27. He, Yang, Wu et al. (2023) "Identification of tanshinone I as cap-dependent endonuclease inhibitor with broad-spectrum antiviral effect" *J Virol*
28. Atanasov, Zotchev, Dirsch (2021) "International natural product sciences taskforce, supuran CT. Natural products in drug discovery: advances and opportunities" *Nat Rev Drug Discov*
29. Yang, Pan, Xu et al. (2020) "Salvianolic acid C potently inhibits SARS-CoV-2 infection by blocking the formation of six-helix bundle core of spike protein" *Signal Transduct Target Ther*
30. Gao, He, Zhu et al. (2025) "Identification of Licoflavone C as a cap-dependent endonuclease inhibitor against severe fever with thrombocytopenia syndrome virus" *Acta Pharmacol Sin*
31. Khan, Sabella, Mandlem et al. (2025) "Salvianolic acid-A alleviates oxidative stress-induced osteoporosis" *Life Sci*
32. Ling, Ma, Huang et al. (2021) "Salvianolic acid A alleviated inflammatory response mediated by microglia through inhibiting the activation of TLR2/4 in acute cerebral ischemia-reperfusion" *Phytomedicine*
33. Dawuti, Sun, Wang et al. (2023) "Salvianolic acid A alleviates heart failure with preserved ejection fraction via regulating TLR/Myd88/ TRAF/NF-κB and p38MAPK/CREB signaling pathways" *Biomed Pharmacother*
34. Huang, Cheng, Wang et al. (2022) "Viruses utilize ubiquitination systems to escape TLR/RLR-mediated innate immunity" *Front Immunol*
35. To, Erlich, Liong et al. (2020) "Mitochondrial reactive oxygen species contribute to pathological inflammation during influenza A virus infection in mice" *Antioxid Redox Signal*
36. Lu, Pan, Chen et al. (2021) "Broad-spectrum antivirals of protoporphyrins inhibit the entry of highly pathogenic emerging viruses" *Bioorg Chem*
37. Chen, Lu, Yang et al. (2020) "The seafood Musculus senhousei shows anti-influenza A virus activity by targeting virion envelope lipids" *Biochem Pharmacol*
38. Chen, Li, Wang et al. (2024) "Salvianolic acid A inhibits pseudorabies virus infection by directly inactivating the virus particle" *Phytomedicine*
39. Hu, Wang, Zhang et al. (2021) "Three salvianolic acids inhibit 2019-nCoV spike pseudovirus viropexis by binding to both its RBD and receptor ACE2" *J Med Virol*
40. Yan, Li, Chen et al. (2025) "Salvianolic acid A acts as a herpes simplex virus dual inhibitor by blocking glycoprotein B-mediated adsorption and membrane fusion" *Phytomedicine*
41. Yang, Song, Gan et al. (2022) "Non-clinical safety evaluation of salvianolic acid A: acute, 4-week intravenous toxicities and genotoxicity evaluations" *BMC Pharm Toxicol*
42. Chen, Ruan, Lou et al. (2022) "First-in-human study to investigate the safety and pharmacokinetics of salvianolic acid A and pharmacokinetic simulation using a physiologically based pharmacokinetic model" *Front Pharmacol*
43. Sun, Song, Zhang et al. (2018) "Some pharmacokinetic parameters of salvianolic acid A following single-dose oral administration to rats" *Pharm Biol*
44. Li, Wang, Fan et al. (2020) "The application of fermentation technology in traditional chinese medicine: a review" *Am J Chin Med*
45. "Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12809628&blobtype=pdf | # Archiv der Pharmazie
Martin Juhás, Gerhard Ecker
## Abstract
The SLC5 family of solute carriers is of significant interest for drug development due to its role in many disease processes. Building on the recent elucidation of SGLT2's structure, we developed a proteochemometric model for SLC5 inhibitors in order to gain information on selectivity-driving amino acids in the binding site. Ensemble-based algorithms, namely random forest (RF) and gradient-boosted trees, proved the best suited for the task reaching high accuracy in both activity and selectivity predictions with Morgan circular fingerprints and Z-scales for ligand and protein features, respectively. Inclusion of protein sequence as input parameters for the PCM modeling allowed identification of Leu286 in hSLGT2 as a new potential key binding site residue crucial for selectivity. Furthermore, the PCM model also performed well in predicting the effect of single-point mutations at hSGLT2 on the binding affinity of empagliflozin. The obtained models are available in the form of a Jupyter notebook.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
## 1 | Introduction
Solute carrier transporters (SLCs) are one of the largest superfamilies of proteins facilitating substrate movement across cellular membranes, predominantly involved in cellular uptake [1,2]. To this date, the human SLC superfamily is organized into 66 families based on sequence similarity and transport mechanisms [3], comprising passive and secondary active transporters. Function-wise, the SLC families are diverse and transport a wide range of organic and inorganic molecules, including hormones, neurotransmitter, amino acids, glucose, and xenobiotics (for reviews, see references [1,2,[4][5][6][7]). SLC transporters are involved in multiple pharmacological and toxicological processes, and defects in some members have been implicated in severe disorders like amyotrophic lateral sclerosis (ALS) or diabetes mellitus (DM) [2,6]. Although roughly 30% of the SLC-transporter are still orphan [8], several members comprise versatile targets in drug development. These include, among others, the GABA transporter GAT1 for epilepsy, the serotonin transporter SERT for depression, and the sodiumglucose co-transporter SGLT2 (SLC5A2) for DM. SGLT2 is expressed in the kidney and inhibited by a drug class called gliflozins, leading to improved glucose metabolism in DM patients. However, the favorable effects of the SGLT2-inhibition seem to go beyond the treatment of DM as observed, for example, in patients with heart failure [9,10]. Recently, dual inhibition of SGLT1 and SGLT2 has been reinvestigated for the treatment of other glucose-linked conditions, thus potentially uncovering other positive effects of SLGT2 inhibition in the future [11,12]. Further drug development efforts will be noticeably improved due to a very recent high-quality co-crystal structure of SGLT2 with empagliflozin (PDB ID: 7VSI [13]), which was also used in this study.
Quantitative structure-activity relationship (QSAR) is a wellestablished method in computational drug design and the method of choice when structural information for the target is missing. However, traditional QSAR works best on a relatively small chemical space, commonly on a congeneric series of compounds, and is limited to one target per model. These limitations are overcome by proteochemometric modeling (PCM), which simultaneously considers ligands as well as their targets in one model. Thus, PCM uses a more complex and more detailed feature space and can produce more reliable models compared with traditional QSAR [14]. Another advantage of PCM is that by introducing features of the targets, it is not limited to one target only. On the condition that targets' features of interest are different (sequences or e.g., residues in the binding sites), PCM modeling can be used to study selectivity of ligands across a set of targets, and potentially extrapolate toward new ligands and ideally also toward homologous targets [14]. Furthermore, PCM also allows the use of other organisms' data for direct extrapolation, for example, activities on rat can be used for extrapolation to humans [15]. Thus, PCM has already been applied on various targets, including, for example, also lead optimizations as a part of late-stage preclinical development [14,[16][17][18][19].
PCM modeling on SLC5 family members had already been successfully attempted by Burggraaff et al. [20], who aimed to discover new inhibitors of hSLGT1. In their work, a mixture of public domain (ChEMBL v.23) and in-house data were used to produce a classification model, which was validated experimentally, identifying 30 notable inhibitors of hSGLT1 out of 77 classified hits.
However, at that time, the SLC5 members were lacking any experimental (co)crystal structures, so the PCM modeling lacked any structure-based information, like the exact binding site defined based on the recent hSGLT2 cocrystal [13]. Thus, in this study, we developed a regression-based PCM model using data from ChEMBL v.30, focusing on the interpretation of the model with respect to residues influencing transporter selectivity. Descriptor importance ranking of the final model allowed identification of Leu286 in hSLGT2 as a new potential key binding site residue important for ligand-transporter selectivity. Furthermore, the PCM model also allowed to predict the effect of single-point mutations at hSGLT2 on the binding affinity of empagliflozin.
## 2 | Results and Discussion
An overall workflow of the data preparation is visualized in Figure 1. Full details of the appropriate steps are described in Section 4.
## 2.1 | Data Retrieval
In total, 30 reviewed SLC5 genes were identified in Uniprot for human, mouse, and rat. However, only 12 were also found in ChEMBL v.30 [21], and these are presented in Table 1. The sequences of all 12 targets were aligned using the COBALT web service (https://www.ncbi.nlm.nih.gov/tools/cobalt/re_cobalt. cgi) and processed as described in Section 4.
Based on the obtained alignment and subsequent superposition of the SGLT-2 cocrystal (PDB ID: 7VSI) and AlphFold2 structures, we identified 16 positions in the binding sites that were variable (different amino acids in at least one studied protein), and thus were used for the PCM modeling. The superposition of the selected structures is presented in Figure 2, residues are listed in Table 2.
## 2.2 | ChEMBL Data Processing
In total, ChEMBL v30 contained 5148 activities datapoints (including possible duplicates) for the 12 SLC5s. The dataset was manually curated and filtered to obtain bioactivity value (pIC50) for each compound-target pair. For a detailed description of the data processing, please see Section 4.
Based on the multiple sequence alignment, datapoints from RnSLC5A5 and HsSLC5A7 were removed as they contained gaps at positions of the binding site. Upon activity data processing, three additional targets got eliminated as they contained permanently charged molecules (MmSLC5A7), missing standard_values (RnSLC5A6), or did not contain IC50 or EC50 values (RnSLC5A7). Finally, datapoints from MmSLC5A2 were removed as MmSLC5A2 and HsSLC5A2 had the same residues in the active site, but different activities for the same compounds were noted, which would confuse PCM modeling. The final dataset contained 2310 datapoints representing 1734 unique molecules and activities against six targets. Out of these, 548 molecules had pIC50 values for more than one target and thus represented the most valuable part of the dataset from the PCM perspective. A detailed datapoint distribution based on targets is presented in Table 3.
As can be seen in Table 3 and also from the distribution of the pIC50 values in Figure 3, the obtained dataset was not uniform from the target perspective and biased toward more active compounds. This is an unfortunate problem related to working with public datasets, such as the ChEMBL database [22].
## 2.3 | Machine Learning Modeling
A wide range of ligand features have been investigated in PCM modeling with various degree of success [14]. In this study, we used the three most common: Physico-chemical descriptors, Morgan fingerprints, and MACCSkeys. For details, please refer to Section 4. For encoding the different proteins in the data matrix, we focused on the Z3-scales, as they not only allow us to trace back which amino acid(s) in the binding site drive ligand selectivity, but also provide information on the respective physicochemical property. Of course, if the main intention is to build a reasonable machine learning model, also, for example, one-hot encoding usually works quite well.
The training set and test set were obtained by a target-based stratified split (70/30). Final distribution based on the target is presented in Table 4.
Many machine learning algorithms showed promising results in the PCM; among the most successful were partial least squares (PLS), support vector machine (SVM), random forest (RF), and neural networks (NN) [14]. Here, we investigated only the "classical" machine learning algorithms: SVM, RF, and gradient boosted trees (hereby referred to as XGB based on the used implementation XGBoost [23]).
SVM has previously been successfully applied in PCM, and it also worked well in this study [14,24]. Best predictions were obtained with physico-chemical descriptors (R 2 = 0.76 for the final model using 49 phys-chem features, see Section 4), closely followed by MACCSkeys (R 2 = 0.75). Despite overall good performance, there were more strong outliers (pIC pICpred exp > 2, see below) in the SVM models com- pared with the later investigated ensemble methods. Morgan fingerprints in SVM achieved the lowest performance with R 2 = 0.71. Results obtained for all models are presented in Table 5 and in Figure 4.
Ensembles methods such as RF or gradient boosted trees belong to the decision tree-based algorithms and represent one of the most versatile machine learning algorithms applicable in a wide range of applications. Both RF and XGB performed well, significantly outperforming SVM. The best results were obtained using Morgan circular fingerprints, with R 2 = 0.80 and R 2 = 0.82 for RF and XGB, respectively. Physico-chemical and MACCSKey descriptors scored slightly worse but still outperformed SVM.
Despite XGB achieving (marginally) better overall accuracy, RF models were faster and therefore used for all detailed investigations that follow. If not indicated otherwise, the presented results always apply to the Morgan fingerprints-based models (random_state=1) as the highest performing ones.
Observing the high accuracy of the RF Morgan fingerprints model, the robustness of the model was further investigated by the DummyRegressor implemented in scikit-learn. We applied "mean" and "median" prediction strategy; in both cases, the accuracy of the model reached R 2 ≈ 0.0.
As shown in Figure 4, models for SLC transporter contributing a considerable amount of data (HsSLC5A1 and HsSLC5A2) generally show good performance for the test set, while R2 values for those with only a few data points (HsSLC5A4 and RnSLC5A2) vary considerably. However, this might be due to the different pIC50 ranges covered by the respective data sets rather than their limited size.
## 2.3.1 | Outliers in the Random Forest Models
To investigate the capabilities of the RF PCM models, we trained three randomized models (different initial random_state parameter) for each of the used ligand features using the same train/test split. The goal was to assess the robustness of the ML parameters. The predicted outliers with errors greater than two pIC units (pIC pICpred exp > 2, hereby referred to as strong outliers) were gathered and compared within runs and among all models with different ligand featurizations.
Overall, the performance was consistent within the randomized runs, and very few strong outliers were observed. As can be seen in Table 6, the dataset is biased toward active compounds, which is often observed when working with public domain databases like ChEMBL [22]. All the outliers were thus overpredicted, as seen in Tables 789. Good overlap of the outliers was observed within different runs.
The highest number of unique outliers was observed in MACCS-based models (four, Table 9), but interestingly also in the Morgan fingerprints model (four, Table 8). The lowest number was in the models using Phys-Chem properties (two, Table 7). This was surprising since fingerprint-based models overall scored better than Phys-Chem properties-based. One outlier compound appeared in all models, CHEMBL2397443, while two outliers (including CHEMBL2397443) were shared in Morgan and Phys-Chem (CHEMBL1779206, CHEMBL2397443), and two were shared in MACCS and Phys-Chem (CHEMBL2397443, CHEMBL3288757). However, upon closer inspection, in most cases the difference of pIC exp and pIC pred was only slightly above the defined cutoff as exemplified by compounds CHEMBL1779206 and CHEMBL1784419 in Morgan models and/or MACCS models, where the pIC50 prediction was approx. 2.1 pIC50 units overpredicted (see Tables 8,9). Not considering these slight overpredictions, we observed only a handful of outliers.
The most notable outlier is CHEMBL3288757, a highly selective SGLT-1 inhibitor (pIC50 = 6.31). While its SGLT-1 activity was present in the test set, the SGLT-2 (inactive) datapoint (pIC50 = 4.05) was in the training set. At first, we thought the overprediction of CHEMBL3288757, which occured in Phys-Chem and MACCS models, was due to the inability of distinguishing these two targets. However, upon closer inspection, we discovered that this compound comes from a larger congeneric series of compounds containing several activity cliffs, with some of these compounds being in the training set. As an example, we present CHEMBL3660004, a very potent but non-selective inhibitor of SGLT-1 and SGLT-2 (pIC50 = 7.08 and 6.83, respectively) that is a positional isomer of CHEMBL3288757 (Figure 5). Both activities were in the training set, thus this overprediction was due to the presence of an activity cliff that was insufficiently described by some descriptors. Similar activity cliffs were also observed for CHEMBL2397443, which was also observed as an outlier in all models.
Activity cliffs are generally challenging in drug design and present an exceptional challenge for QSAR modeling. However, as investigated, for example, by Tilborg and colleagues [25], use of circular fingerprints seems appropriate for cases where activity-cliffs are expected, supporting why Morgan fingerprints outperformed other descriptors in the current study.
In the following tables, the chemical structure represents structures after final standardization for the ML modeling, that is, without chirality assigned.
## 2.4 | Importance of the Target Residues
Studying selectivity, we were particularly interested in the importance of the chosen pocket residues, that is, the target features. The main rationale was to find which positions (residues) are crucial for predicting the correct pIC50, and could thus potentially be responsible for the selectivity of the inhibitors toward (investigated) isoforms of SLC5s. This approach has already been successfully applied in PCM modeling by others, for example, on serine proteases [26], GPCRs [27], or HDAC [28] receptors, as well as on GABA transporters by our group [29]. Numbering of residues/positions in the following section refers to the position of the residue in the SLGT2, since this can be easily visualized using the crystal structure PDB ID: 7VSI.
For reference, we are also including the position in the COBALT MSA. Full mapping of the binding site residues of the investigated targets, including their position in the COBALT MSA is presented in Table 2. A big advantage of RF, and other decision tree-based methods, is their natural feature importance awareness and thus ability to rank the features' impact on the prediction. As Morgan fingerprints had the most accurate predictions, this model was used for the investigations. Little attention was paid to the ligand features since comparison and interpretation of bits in Morgan fingerprints would be a nontrivial matter.
Two different methods for importance investigations were used. In the first one (1), the ranking of the features was obtained directly from scikit-learn. The disadvantage of this method was that it could not distinguish if other feature(s) could substitute the impact of the feature (i.e., specific amino acid residue represented by the Z-scale). We tried to mitigate this by the second method (2), doing an exhaustive
permutation of all positions, retraining the model with the best obtained parameters (no grid searching) and monitoring the performance.
The feature importances obtained from the trained model are expressed as the mean decrease in impurity of the trees in the RF model, with the higher values meaning higher importance (see feature_importances in scikit-learn). For our investigations, considering there were 1443 features in the models, we found working with the ranks more illustrative of the features' importance. Note that by ranking the importances in the ascending order their original meaning is retained; the higher the importance, the higher its rank.
The first method found position 286 (position 312 in COBALT MSA) to be the most important and was thus ranked the highest. More precisely, it was its Z2-scale value (steric properties). This position was closely followed by positions 95 (with Z1-scale=lipophilicity and Z3-scale=electronic properties) and 460 (Z1, Z2, and Z3-scale). Table 10 shows the mean, median, and standard deviation of the ranks from the top 10 ranked positions within five randomized runs (different random_state parameter). The mean in this case represents the average ranking of the feature as returned by scikit-learn. Considering that the majority of 1443 descriptors in the model are represented by the bits of the Morgan fingerprints, the rank 1438 of position 286 is significant, that is, the feature is very important for model performance.
In the second method, we focused only on the selected amino acid position; Z-scale values were not distinguished. The same residues as in the first method were identified as important.
Due to ignoring the Z-scales, a slightly different ranking was observed. The following six positions were essential for the model, in the order of decreasing importance: 460 (488 in COBALT MSA), 286 (312 in COBALT MSA), 95 (106 in COBALT MSA), 287 (313 in COBALT MSA), 157 (168 in COBALT MSA), 283 (309 in COBALT MSA). Removal of any of these residues from the feature space strongly deteriorated the performance, effectively reaching the performance of a model where no positions were included (R 2 ≈ 0.36, Q 2 ≈ 0.37). When only one position was added back, the best performance was again observed with residue 460 R 2 = 0.79 (Q 2 = 0.79), closely followed by 286, 95, 287, 157, and 283, which achieved the worst score, R 2 = 0.77 (Q 2 = 0.75). A model containing only the six residues performed equivalently to the model with all selected binding site residues R 2 = 0.80 (Q 2 = 0.79), thus showing a relatively small (statistical) impact of other residues toward activity prediction.
The identified important positions and the respective amino acids present in the investigated SLC5 proteins are indicated in Table 11.
Yet, a mathematical significance does not necessarily translate to biological relevance. Unfortunately, to our best knowledge, there has been no experimental mutation analysis involving residue 286 that we saw consistently being the most impactful (see Figure 6). However, other impactful positions, notably the close by residues at positions 95, 157, and 283 were experimentally validated by point-mutation analyses in hSGLT2. The mutation of these residues to the corresponding amino acids from hSGLT1 (V95I, V157A, and L283M) showed reduced inhibitory activity of empagliflozin (highly selective SGLT2 inhibitor) while retaining the glucose absorption of the mutant [13]. We thus conclude that the observed residue 286 could also bear a strong biological significance for the SLC5 family, and perhaps also participate or govern the selectivity of inhibition among its members.
$$RnSLC5A1 P53790 CHEMBL5374 72$$
$$HsSLC5A11 Q8WWX8 CHEMBL1744524 21$$
$$HsSLC5A2 P31639 CHEMBL3884 893383$$
$$RnSLC5A2 P53792 CHEMBL4316 94$$
$$HsSLC5A4 Q9NY91 CHEMBL1770047 18 8$$
## 2.5 | Ability of the Model to Correctly Predict Selectivity
Finally, we have closely analyzed the ability of the model to correctly predict the selectivity of compounds. Out of all 548 unique molecules with activity values on more than one SLC transporter (1124 datapoints in the whole dataset), the test set contained 278 molecules (330 datapoints). The activity prediction of these compounds was satisfactory, overall reaching R 2 = 0.84, MSE = 0.29, with 92% being within ±1 pIC50% and 70% within ±0.5 pIC50 (test set metrics). Thus, the model was able to successfully predict activity for specific targets on unseen data, despite the compound had multiple, often largely different activities against specific SLC5 isoforms.
To simulate the ability to correctly predict target selectivity, we analyzed a set of 52 compounds that had more than one pIC50 value, but were present only in the test set (i.e., the model could not have learned it from the training set). Selectivity was judged as , where pIC = 0 diff means the model predicted a compound to be selective against target 1 compared with target 2 with the same magnitude as observed experimentally. The quality of prediction of the pIC50 itself was investigated above. Thus, 92% of the predictions were within ±1% and 58% within ±0.5 pIC diff (R 2 = 0.78, MSE = 0.39). As evident from Figure 7, the model tends to overpredict (pIC pIC > pred exp ), which is the unfortunate result of the insufficient number of low activity datapoints in ChEMBL.
The presented model was therefore able to predict both the activity and degree of selectivity of the compound-target pairs with reasonable accuracy and correctness, and none of the observed compounds was identified as a strong outlier (difference in pIC50 > 2).
As a proof of concept, we used the in silico point-mutated pocket as investigated by Niu et al. [13] for the protein features and predicted the pIC50 of empagliflozin, for which the pointmutation analyses were available. It is important to note that the mutated proteins' activities were not part of the dataset. As shown in Table 12, the model was able to correctly predict the decreased activity against the mutated SGLT2 compared with the wild type. More rigorous testing would require additional experimental
## 2.6 | Investigation of the Chemical Space and Applicability Domain
It has been previously noted that ChEMBL covers a large but also realistic chemical space [30]. That is the reason why ChEMBL alone or in combination with other databases serve as a starting point for the ligand-based drug design.
The chemical space of SLC5 inhibitors, as used in this study, is unfortunately biased by the public unavailability of (in)active data. The most abundant chemotypes present in the dataset, denoted as A and B in Figure 8, correspond to the arylsubstituted saccharides (structure A) and diaryl-substituted pyrazoles (structure B). In both cases, the substituents R 1-4 represent mostly short alkyl, hydroxy, and/or halogen substitution, while R 5,6 represent mostly a further decorated fragment and/or another (annelated) ring.
Still, as seen in Figure 9, the majority of molecules share only very low whole-structure similarity with a Tanimoto index < 0.4. We thus believe that there is sufficient chemical variability for the purpose of studying the selectivity of SLC5 inhibitors.
## 3 | Conclusions
We have shown that PCM is a valuable method for deeper insights into the selectivity of inhibitors in the SLC5 family. This study produced three machine learning models using traditional algorithms, SVM, RF, and XGB, capable of predicting the activity and, more importantly selectivity of SLC5 inhibitors based on the input structure and receptor pocket residues. The models were validated using cross-validation methodology and proved accurate, with the best Q 2 = 0.80. We have shown a potential applicability of the model on predicting selectivity on point-mutated receptors with qualitatively correct predictions, although additional point-mutation training data would be required to obtain a quantitative model. By studying the selectivity, we uncovered a new potential binding site residue of interest, possibly responsible for the selectivity of SLC5 inhibitors that could be prioritized in further research of the new SLC5 inhibitors as potential drugs.
## 4 | Experimental
All data manipulation was done in Python 3.7 using JupyterLab.
$$HsSLC5A1 I A M L T T RnSLC5A1 M A L L A T HsSLC5A11 V V M P S S HsSLC5A2 V V L V S S RnSLC5A2 V V L V S S HsSLC5A4 T A M T A S$$
## 4.1 | Data Retrieval
Uniprot accession codes for all reviewed human (Homo sapiens, ID: 9606), mouse (Mus musculus, ID: 10090) and rat (Rattus norvegicus, ID: 10116) slc5 genes were gathered manually and translated to ChEMBL target IDs using the REST API (ChEMBL v.30 was used, https://www.ebi.ac.uk/chembl/). Upon identifying available targets in ChEMBL, the activities data were downloaded using the REST API and saved. All REST API calls were done using an inhouse Python script.
## 4.2 | Targets Multiple Sequence Alignment
Primary sequences of all targets available in ChEMBL were aligned using the COBALT webservice (https://www.ncbi.nlm. nih.gov/tools/cobalt/re_cobalt.cgi), using default settings. Subsequently, AlphaFold2 [31] structures of all identified proteins without available crystal structure were retrieved and superimposed on the crystal structure of human SGLT2 complexed with empagliflozin (PDB ID: 7VSI). Binding site residues were defined in MOE 2020.09 using the Pocket definition (residues with atoms within 4.5 Å of the co-crystallized ligand), and the corresponding residues in the other transporter were identified based on the COBALT alignment. Only positions with amino acid variability were considered for PCM modeling. For the proteins with gaps within the chosen positions, AlphaFold2 structures were aligned to PDB ID: 7VSI one-by-one based only on the structure in MOE to improve the alignment. If there were still gaps observed, the receptors were excluded. COBALT alignment of the final chosen targets in FASTA format is available in the Supporting Information S1.
## 4.3 | Dataset Processing and Molecule Standardization
All Finally, compounds were checked manually, and compounds containing fluorescent probes (CHEMBL449778, CHEMBL443616, CHEMBL445247) and triphenyl protection (CHEMBL3288753) were removed. The data set was then aggregated based on the InChIKeys. For InChIKeys having less than five activity datapoints for a particular target, the average was taken if values were within SEM < = 0.3, otherwise all activity data for that molecule-target pair were removed. For InChIKeys with five or more activities, the average was taken.
## 4.4 | Investigation of the Chemical Space and Applicability Domain
The chemical space was investigated using da imensionality reduction plot by Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) [34]. The python package umap-learn 0.5.3.
First, the embeddings for the training set were obtained by extracting unique molecules from the training set. Note that without this step, there would be instances of the same molecule due to having activity on different targets.
Then, the obtained model was applied on the test set to compare the covered chemical space as well as the applicability domain of the PCM models. Compounds were represented by the same Morgan fingerprints as during the featurization, that is, radius=2 and nBits=2048. For the UMAP hyperparameters, "jaccard" metric was used with n_neighbors=50 and min_distance=0.001. Value for the n_neighbors was chosen as best from the tested values of 2, 5, 10, 25, 50, 80, and 100. The obtained model is available with the prediction Jupyter Notebook code in the Supporting Information S2.
## 4.5 | Machine Learning Modeling
For the receptors, that is, the binding site residues, the first three Z-scales described by Sandber and colleagues [35], related to lipophilicity, steric, and electronic properties were used as features.
For ligands, physico-chemical descriptors, Morgan fingerprints, and MACCS keys were used. The phys-chem descriptors were calculated from available descriptors in RDKit 2021.03.4. Out of 208 available descriptors, all fragment descriptors (starting with "fr_") were discarded, and from the rest, descriptors with obvious redundant information were filtered out manually, for example, from MolWt, HeavyAtomMolWt, ExactMolWt, only MolWt was retained. Thus, we obtained 108 features that were used in the initial models; the descriptors are listed in the Supporting Information S2. Later, the features were manually revised. Highly correlated descriptors (absolute of the Pearson coefficient > 0.7) were removed, and from the rest 49 descriptors representing common molecular properties (electronic, lipophilicity, shape, etc.) were handpicked. The list of final descriptors is also provided in the Supporting Information S2.
The simpler models were found to perform similarly to those with all physico-chemical descriptors, and only these models are therefore commented in this article. A vector of 2048 bits with a radius of 2 was used for Morgan circular fingerprints. MACCS keys were calculated as bit vectors. In the case of all descriptors, only non-constant features were used in model building. No cross terms were used.
The ML modeling was done using algorithms as implemented in scikit-learn 1.0.2 and xgboost v.1.6.1. The 10-fold crossvalidation monitored overfitting. Final cross-validated R 2 , hereby reported as Q 2 , represents the average of the 10-fold R 2 values. The performance was judged based on R 2 , Q 2 , and MSE. Optimal hyperparameters for the ML algorithms were obtained by performing a grid search using GridSearchCV as implemented in scikit-learn. Details for hyperparameter optimization as well as final hyperparameters used for all the models can be found in the Supporting Information S2. SVR, Ran-domForestRegressor, and XGBRegressor classes were used for the ML modeling. For the SVM models, the data were scaled using StandardScaler function. The r2_score, cross_val_score, and mean_squared_error functions were used to evaluate the model performance, 10-fold cross-validation was used. The final plots were done using matplotlib 3.5.2 and seaborn 0.11.2.
## References
1. Lai (2013) "1 -Membrane Transporters and the Diseases Corresponding to Functional Defects"
2. Lin, Yee, Kim et al. (2015) "SLC Transporters as Therapeutic Targets: Emerging Opportunities" *Nature Reviews Drug Discovery*
3. Denecke, Driva, Luong (2020) "The Identification and Evolutionary Trends of the Solute Carrier Superfamily in Arthropods" *Genome Biology and Evolution*
4. He, Vasiliou, Nebert (2009) "Analysis and Update of the Human Solute Carrier (SlC) Gene Superfamily" *Human Genomics*
5. (2025) "SLC superfamily of solute carriers. IUPHAR/BPS Guide to PHARMA-COLOGY"
6. Liu (2019) "SLC Family Transporters"
7. Digles, Ingles-Prieto, Dvorak (2024) "Advancing Drug Discovery Through Assay Development: A Survey of Tool Compounds Within the Human Solute Carrier Superfamily" *Frontiers in Pharmacology*
8. Superti-Furga, Lackner, Wiedmer (2020) "The RESOLUTE Consortium: Unlocking SLC Transporters for Drug Discovery" *Nature Reviews Drug Discovery*
9. Gyimesi, Pujol-Giménez, Kanai et al. (2020) "Sodium-Coupled Glucose Transport, the SLC5 Family, and Therapeutically Relevant Inhibitors: From Molecular Discovery to Clinical Application" *Pflügers Archiv -European Journal of Physiology*
10. Hsia, Grove, Cefalu (2017) "An Update on Sodium-Glucose Co-transporter-2 Inhibitors for the Treatment of Diabetes Mellitus"
11. Rendell (2018) "Sotagliflozin: A Combined SGLT1/SGLT2 Inhibitor to Treat Diabetes" *Expert Review of Endocrinology & Metabolism*
12. Tsimihodimos, Filippas-Ntekouan, Elisaf (2018) "SGLT1 Inhibition: Pros and Cons" *European Journal of Pharmacology*
13. Niu, Liu, Guan (2022) "Structural Basis of Inhibition of the Human SGLT2-MAP17 Glucose Transporter" *Nature*
14. Van Westen, Wegner, Ijzerman et al. (2011) "Proteochemometric Modeling as a Tool to Design Selective Compounds and for Extrapolating to Novel Targets" *MedChemComm*
15. Van Westen, Van Den Hoven, Van Der Pijl (2012) "Identifying Novel Adenosine Receptor Ligands by Simultaneous Proteochemometric Modeling of Rat and Human Bioactivity Data" *Journal of Medicinal Chemistry*
16. Van Westen, Wegner, Geluykens (2011) "Which Compound to Select in Lead Optimization? Prospectively Validated Proteochemometric Models Guide Preclinical Development" *PLoS One*
17. Dakshanamurthy, Issa, Assefnia (2012) "Predicting New Indications for Approved Drugs Using a Proteochemometric Method" *Journal of Medicinal Chemistry*
18. Naeem, Dakshanamurthy, Walthieu (2020) "Predicting New Drug Indications for Prostate Cancer: The Integration of an In Silico Proteochemometric Network Pharmacology Platform With Patient-Derived Primary Prostate Cells" *Prostate*
19. Qiu, Tian, Liu (2021) "Revealing the Mutation Patterns of Drug-Resistant Reverse Transcriptase Variants of Human Immunodeficiency Virus Through Proteochemometric Modeling" *Biomolecules*
20. Burggraaff, Oranje, Gouka (2019) "Identification of Novel Small Molecule Inhibitors for Solute Carrier SGLT1 Using Proteochemometric Modeling" *Journal of Cheminformatics*
21. Zdrazil, Felix, Hunter (2023) "The CheMBL Database in 2023: A Drug Discovery Platform Spanning Multiple Bioactivity Data Types and Time Periods" *Nucleic Acids Research*
22. Smajić, Rami, Sosnin et al. (2023) "Identifying Differences in the Performance of Machine Learning Models for Off-Targets Trained on Publicly Available and Proprietary Data Sets" *Chemical Research in Toxicology*
23. Chen, Guestrin (2016) "XGBoost: A Scalable Tree Boosting System"
24. Heikamp, Bajorath (2014) "Support Vector Machines for Drug Discovery" *Expert Opinion on Drug Discovery*
25. Van Tilborg, Alenicheva, Grisoni (2022) "Exposing the Limitations of Molecular Machine Learning With Activity Cliffs" *Journal of Chemical Information and Modeling*
26. Ain, Méndez-Lucio, Ciriano et al. (2014) "Modelling Ligand Selectivity of Serine Proteases Using Integrative Proteochemometric Approaches Improves Model Performance and Allows the Multi-Target Dependent Interpretation of Features" *Integrative Biology*
27. Gao, Huang, Wu (2013) "Study on Human GPCR-Inhibitor Interactions by Proteochemometric Modeling" *Gene*
28. Wu, Huang, Zhang (2012) "Screening of Selective Histone Deacetylase Inhibitors by Proteochemometric Modeling" *BMC Bioinformatics*
29. Kickinger, Seiler, Digles et al. (2008) "Proteochemometric Modeling Strengthens the Role of Q299 for Gaba Transporter Subtype Selectivity" *bioRxiv*
30. Cauchy, Leguy, Mota (2023) "Definition and Exploration of Realistic Chemical Spaces Using the Connectivity and Cyclic Features of CheMBL and ZINC" *Digital Discovery*
31. Jumper, Evans, Pritzel (2021) "Highly Accurate Protein Structure Prediction With Alphafold" *Nature*
32. Xu, Feng, Cheng "C-Aryl Glucosides Substituted at the 4′-position as Potent and Selective Renal Sodium-Dependent Glucose Co-Transporter 2 (SGLT2) Inhibitors for the Treatment of Type 2"
33. (2011) *Bioorganic & Medicinal Chemistry Letters*
34. Lee, Kim, Choi et al. (2010) "Pyrimidinylmethylphenyl Glucoside as Novel C-Aryl Glucoside SGLT2 Inhibitors" *Bioorganic & Medicinal Chemistry Letters*
35. Healy, Mcinnes (2024) "Uniform Manifold Approximation and Projection" *Nature Reviews Methods Primers*
36. Sandberg, Eriksson, Jonsson et al. (1998) "New Chemical Descriptors Relevant for the Design of Biologically Active Peptides. A Multivariate Characterization of 87 Amino Acids" *Journal of Medicinal Chemistry* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12307093&blobtype=pdf | # A Framework for Evaluating the Use of Surveillance Systems for Short-Term Influenza Forecasting
Negin Maroufi, Lucy Telfar Barnard, | Qiu, Sue Huang, | Dobbie, Nayyereh Aminisani, Steffen Albrecht, Nhung Nghiem, Michael Baker
## Abstract
Background: Public health surveillance systems need to monitor influenza activity and guide measures to mitigate its high impact on morbidity, mortality and healthcare systems. There is an increasing expectation that surveillance data will support the modeling of future short-term disease scenarios using artificial intelligence (AI) and machine learning (ML). This study examines how influenza surveillance can support AI/ML-based short-term forecasting for influenza at the community and hospital levels in a high-income country setting (Aotearoa/New Zealand). Methods: This study used a two-phase approach. The first phase involved a comprehensive review of government reports, official websites, and literature to characterize existing influenza surveillance systems. The second phase evaluated systems against eight key attributes-timeliness, sensitivity, specificity, representativeness, coverage, robustness, completeness, and historical data-using a five-level ranking system. Attribute selection was informed by experts' knowledge, ML requirements, and established frameworks. Weighted scores for training and short-term forecasting capabilities were calculated to determine alignment with AI/ML requirements. Results: The Southern Hemisphere Influenza and Vaccine Effectiveness Research and Surveillance (SHIVERS) community cohort and Severe Acute Respiratory Infection (SARI) hospital surveillance emerged as the most useful systems, achieving the highest scores in both training and short-term forecasting in community and hospital settings, respectively. The National Minimum Dataset of hospitalizations and mortality datasets demonstrated strong training potential but are limited in short-term forecasting due to timeliness constraints. Additionally, laboratory-based surveillance performs a useful role in bridging community and hospital datasets. Conclusions: A set of key attributes is useful for assessing which influenza surveillance systems are best aligned with AI/ML training and short-term forecasting requirements. These attributes distinguished systems that are likely to be the most suitable for modeling future short-term disease scenarios for influenza at the community and hospital levels in New Zealand. Integrating these data sources could enhance influenza forecasts to improve public health responses and intervention planning.This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
## 1 | Introduction
Influenza remains a significant global public health challenge, causing seasonal epidemics and occasional pandemics, which result in considerable morbidity, mortality, and healthcare burden [1]. Surveillance systems can help mitigate its impact by guiding timely and informed interventions [2][3][4]. According to the World Health Organization (WHO), influenza surveillance involves the systematic collection, compilation, and analysis of data to monitor influenza activity within defined populations, ranging from national to regional or group-specific levels [2]. Surveillance systems range from simple, single-source data collection to complex multisource electronic systems and community-based approaches [2,4]. These systems provide insights into when and where influenza occurs, monitoring virus changes, and assessing the disease's impact on illness, hospitalizations and mortality [2][3][4].
Aotearoa/New Zealand has had established respiratory disease surveillance since 1989, as part of the WHO Global Influenza Surveillance Network [5][6][7]. The country employs multiple surveillance approaches to gathering information across various levels of severity [6,8]. Effective interpretation of this data supports public health decision making, particularly in identifying trends and forecasting disease spread [2]. Surveillance system evaluations can help ensure their quality, efficiency, and usefulness [7]. Such evaluations focus on reviewing the system's outputs and overall performance in achieving its goals by applying standards and performance measures [9]. Various frameworks exist for evaluating surveillance systems in the field of public health [3,[9][10][11][12][13][14][15][16]. These frameworks focus on assessing how well systems meet their objectives and are well established and valuable for system performance assessment. However, they were not designed to assess the suitability of surveillance systems for AI and ML applications.
The emergence of artificial intelligence (AI) and machine learning (ML) technologies offers transformative potential for enhancing the efficiency and predictive capabilities of surveillance systems' data [17]. By integrating real-world data from existing systems, these technologies can improve influenza forecasting by providing insights into the near future and support timely and informed public health decision making [13,18,19]. Certain attributes such as timeliness, data accuracy (incorporating sensitivity and specificity), data representativeness, coverage, robustness, completeness, and longitudinal data availability are key to ensuring the utility of these datasets in predictive models [11,13,17,18].
Building on existing surveillance evaluation frameworks, this paper proposes a complementary approach focusing on which attributes matter for AI/ML applications. The study provides a method to identify which surveillance systems are best suited for ML applications and discusses how they might be improved or integrated. It uses New Zealand influenza surveillance systems as a case study, evaluating and comparing their suitability for short-term forecasting and predictive model development. Study findings have the potential to improve the usefulness of surveillance data for generating likely scenarios for short-term influenza forecasts. Such scenarios are important for understanding the influenza season's trajectory in the community and its shortterm impact on healthcare systems, particularly hospitals and primary care services. This study emphasizes achieving optimal predictive accuracy, that is, forecasts that closely reflect actual outcomes, while addressing practical considerations such as timeliness and the efficient use of available resources.
## 2 | Method
This study evaluated influenza surveillance systems in New Zealand in two phases.
The first phase aimed to produce a comprehensive description of influenza surveillance systems in New Zealand. Sources included governmental reports from the Ministry of Health, Health New Zealand (Te Whatu Ora), and the New Zealand Institute for Public Health and Forensic Science (formerly the Institute of Environmental Science and Research [ESR]), along with official websites for each surveillance system and relevant published academic literature. A detailed web search was conducted to gather information on these systems (Appendix S1). This search collected definitions and characteristics of the existing influenza surveillance systems in New Zealand (Appendix S1).
The second phase aimed to evaluate the suitability of these surveillance systems to support predictive analysis for influenza short-term forecasts at various levels from community to hospitalization and death. An in-depth review of established frameworks, including those proposed by the Centers for Disease Control and Prevention (CDC), the WHO, and other studies, was conducted to identify structured criteria for assessment [3,[9][10][11][12][13][14][15][16] (Appendix S2).
Eight key attributes-timeliness, sensitivity, specificity, representativeness, coverage, robustness, completeness, and historical data-were selected from 16 data quality-related attributes, drawn from a larger pool of 31 attributes (Appendix S2). Attributes were selected based on their alignment with machine learning requirements, their relevance to AI/ML applications, their consistent use in existing surveillance evaluation frameworks, and background knowledge of the expert panel [3,[9][10][11][12][13][14][15][16]. The expert panel included epidemiologists, public health specialists, virologists, and data scientists and reached consensus through iterative discussions, ensuring balance between AI/ ML suitability and epidemiological and public health considerations. This multidisciplinary approach ensured that the attributes were evidence-based, practical, and relevant to the study's predictive modeling objectives.
Each surveillance system was assigned scores for each attribute based on Phase 1 data and Phase 2 criteria. The scores were weighted using predefined multipliers to reflect the importance of each attribute for training or short-term forecasting. Weighted scores were averaged to quantify system alignment with predictive model requirements. Two primary metrics were defined to assess system utility:
• Usefulness for training: This metric evaluates how well surveillance systems support ML model training by prioritizing the availability of historical data, along with sensitivity, specificity, and completeness. These attributes give models high predictive accuracy generalizable to the wider population under surveillance.
• Usefulness for short-term forecasting: Short-term forecasting, which spans approximately 1 to 4 weeks ahead, relies on real-time or near real-time data and prioritizes timeliness along with robustness, sensitivity, and specificity to provide outputs with high predictive accuracy in realtime or near real-time.
Detailed process, definitions, explanations, and the five-level ranking system for each attribute are provided in Appendix S3.
## 3 | Findings
Our research identified 10 surveillance systems that provide data on influenza in New Zealand (Table 1) across four levels (from community to mortality). Strengths and limitations of these systems were assessed across eight critical attributes. They were then evaluated for training and short-term forecasting applications, respectively (Table 2).
Community-based systems, such as HealthStat and FluTracking, demonstrated strong timeliness, necessary for short-term forecasting (scores: 3.15 and 3.05). However, their reliance on voluntary self-reporting of symptoms results in lower specificity that limits their effectiveness in training ML models. Healthline, though real-time, faced similar constraints because it relies on self-reported symptoms and uses a broader ILI definition than the WHO standard (see Appendix S1). This broader definition does not require fever and may capture many noninfluenza illnesses, which reduces its influenza virus-specific accuracy. Among community systems, Southern Hemisphere Influenza and Vaccine Effectiveness Research and Surveillance (SHIVERS) stood out with the highest training (4.09) and short-term forecasting (4.20) scores. Its influenza virus-specific laboratory-confirmed data provide sensitivity, specificity, and historical depth, making it highly effective for both applications.
Hospital-based systems scored highly in both categories. SARI achieved the highest short-term forecasting score (4.05) and strong training performance (4.10), due to its available longitudinal data, sensitivity, and specificity. The National Minimum Data Set (NMDS), although useful for training (4.33) due to its extensive historical data, was less suitable for short-term forecasting because of its limited timeliness. Laboratory-based surveillance, represented by the National Influenza Center, bridges community and hospital data, offering moderate utility for both training and short-term forecasting. Although not the highest performer, its role in integrating data from various sources improves its utility.
The mortality dataset has historical depth but is constrained by poor timeliness, making it unsuitable for short-term forecasting. Nonetheless, it remained valuable for long-term trend analysis.
SHIVERS and SARI emerged as the most useful systems for community and hospital settings, respectively. These results highlight the complementary roles of different systems in supporting AI/ML applications.
## 4 | Discussion
This study evaluates the suitability of New Zealand's influenza surveillance systems for predictive models, focusing on training and short-term forecasting capabilities. While systematic reviews have looked at general data quality, to our knowledge, this is the first study looking at which surveillance system attributes matter for AI/ML applications, bringing together both public health and data science perspectives [42]. Although we acknowledge that each surveillance system is designed to meet its own objectives, the growth of AI/ML in health decision making highlights the need to consider how these systems might also support predictive analytics and early-warning applications.
For predictive models, the quality of input data is as important as the model design [18,43]. To ensure outputs achieve optimal predictive accuracy, models must be trained with sufficiently well-characterized data that meet specific criteria of usefulness for training [18,43]. Selecting an optimal combination of data sources therefore contributes to effectively supporting influenza forecasting. To the best of our knowledge, previous influenza forecasting studies have used single data sources, occasionally supplemented by laboratory data, and typically integrating search engines or social media data [44][45][46]. In New Zealand, the SARI dataset has already been integrated for ML-based short-term forecasting, and the impact of integrating laboratory data was investigated [47]. While individual systems exhibit distinct strengths, leveraging a combination of community, laboratory, and hospital-based surveillance systems could optimize forecasting accuracy [19,44,48].
Evidence from prior studies shows integrating syndromic and laboratory-confirmed data enhances model accuracy, supporting a multisystem approach [44].
It is possible that the predictive value could be improved by using multiple surveillance systems instead of one. In New Zealand, influenza surveillance covers different levels of severity. Using SHIVERS for community-level data and SARI for hospital-level data could enhance the predictive value of models for forecasting influenza burden and support effective resource allocation. Also, combining syndromic and etiological surveillance systems at different stages-using the high accuracy of etiological data to train machine learning models and the real-time capabilities of syndromic data for short-term forecasting-may enhance predictive value.
This study also suggests reviewing systems with lower scores to determine whether they need to be improved, integrated into other data sources, or discontinued based on their respective objectives and cost-effectiveness. Such a review could increase their value for training and short-term forecasting, improve resource use, and help preserve valuable data sources.
Additionally, the NMDS dataset showed strong training potential due to its historical depth. Future research could explore its forecasting capability, and if predictions were accurate, Weight assigned to each attribute for evaluating the system's utility for short-term forecasting.
e Laboratory-based surveillance is defined as a separate category. While it primarily captures severe hospital cases, it also includes data from community sources, such as SHIVERS and general practice sentinel surveillance.
improving its timeliness to make it a near real-time system could be justified.
Another key implication of this study is the need to address the underrepresentation of certain populations in existing surveillance systems. This gap may limit the accuracy and generalizability of predictive models and subsequent public health decisions. Integrating lower-cost syndromic surveillance systems like FluTracking with other data sources, particularly systems providing etiological laboratory data, could strengthen forecasting capabilities. Such enhancements would support more comprehensive early detection and prediction of surges in influenza infection at both community and hospital levels.
The evaluation framework developed in this study allows comparisons between real-world surveillance systems, defining a hypothetical optimal data source for predictive model training and short-term forecasting. By identifying characteristics key to predictive models, the framework could inform strategic data collection efforts and surveillance investments, enhancing the usefulness of forecasting with better data quality.
In future research, we aim to develop a training dataset aligned with different levels of influenza severity, incorporating wellsuited data sources to enhance ML-based forecasting models. A promising approach is the application of transfer learning within a multivariate framework to integrate time series data across severity levels. This approach could make models leverage cross-learning dependencies, improving forecasts in multivariate-to-multivariate models (e.g., DeepAR) or multivariate-to-univariate models (e.g., temporal fusion transformer). Additionally, instead of traditional ensemble methods, forecasting could be strengthened by training specific models on diverse datasets, capturing distinct features of influenza activity to improve predictive accuracy.
This study has limitations. Although the attribute rankings of surveillance systems in this study were validated by co-authors with expertise in epidemiology, virology, public health, data science, and machine learning, the subjective nature of these evaluations may limit the validity and generalizability of the conclusions. We have provided transparent reasoning and justifications to allow critical review and further refinement of the evaluation framework. Also, sensitivity and specificity were not measured directly but were instead inferred from proxy indicators like laboratory confirmation and the use of case definitions. Moreover, the findings of this study raise several hypotheses, particularly regarding multisource datasets integrating syndromic and etiological data from different levels of severity, warranting further practical investigations to assess their feasibility and explore their implications for resource planning and decision making.
## 5 | Conclusions
This study proposes an evaluation framework for assessing the suitability of surveillance systems for modeling future short-term disease scenarios for influenza. It emphasizes their alignment with predictive model requirements, notably for training and short-term forecasting. Key findings demonstrate the strengths of specific systems, which in New Zealand include SHIVERS and SARI, which perform well in community and hospital settings, respectively. The framework lays the groundwork for more accurate and timely forecasting by providing a foundation for creating datasets aligned with AI/ML techniques, integrating multisource datasets, and addressing gaps in representativeness. This multidisciplinary study not only advances academic understanding but also ensures that public health surveillance systems can be improved to better support measures to reduce the burden of influenza.
## References
1. (2024) *Influenza (Seasonal) (World Health Organization)*
2. (2024) "enzasurve illan ce. html#: ~: text= Influ enza% 20sur veill ance% 20is% 20the% 20col lecti on"
3. German, Lee, Horan et al. (2001) "Updated Guidelines for Evaluating Public Health Surveillance Systems: Recommendations From the Guidelines Working Group" *MMWR. Recommendations and Reports: Morbidity and Mortality Weekly Report*
4. Ghendon (1991) "Influenza Surveillance" *Bull World Health Organ*
5. (2024) "tious -disea se-intel ligen ce-surve illan ce/ respi rator y-surve illan ce-senti nel-pract ice-infor mation"
6. Jennings (1999) "Influenza Surveillance in New Zealand" *Vaccine*
7. Ziegler, Mamahit, Cox (2018) "65 Years of Influenza Surveillance by a World Health Organization-Coordinated Global Network" *Influenza and Other Respiratory Viruses*
8. Jennings, Barr (2019) "Future Pandemic Influenza Virus Detection Relies on the Existing Influenza Surveillance Systems: A Perspective From Australia and New Zealand"
9. Peyre, Salman, Steneroden (2022) "Frameworks and Tools for Evaluating Health Surveillance Systems. Principles for Evaluation of One Health Surveillance: The EVA Book"
10. Baker, Easther, Wilson (2010) "A Surveillance Sector Review Applied to Infectious Diseases at a Country Level" *BMC Public Health*
11. Cox, Mendoza (2024) "AI, Machine Learning & Big Data Laws and Regulations 2024" *Global Legal Insights*
12. Muellner, Stärk, Watts (2016) "Surveillance Evaluation Framework (SurF): Main Document"
13. Muley, Muzumdar, Kurian et al. (2023) "Risk of AI in Healthcare: A Comprehensive Literature Review and Study Framework" *Asian Journal of Medicine and Health*
14. Pearson (2022) "How to Use AI and Personal Data Appropriately and Lawfully (Information Commissioner's" *Office*
15. Peyre, Hoinville, Njoroge (2019) "The RISKSUR EVA Tool (Survtool): A Tool for the Integrated Evaluation of Animal Health Surveillance Systems" *Preventive Veterinary Medicine*
16. (2017) "Instructions for the National Infection Prevention and Control Assessment Tool 2 (IPCAT2)"
17. Murtaza, Ahmed, Khan et al. (2023) "Synthetic Data Generation: State of the Art in Health Care Domain" *Computer Science Review*
18. Mckee, Rosenbacke, Stuckler (2024) "The Power of Artificial Intelligence for Managing Pandemics: A Primer for Public Health Professionals" *International Journal of Health Planning and Management*
19. Cheng, Wu, Lin (2020) "Applying Machine Learning Models With an Ensemble Approach for Accurate Real-Time Influenza Forecasting in Taiwan: Development and Validation Study" *Journal of Medical Internet Research*
20. Healthstat, Healthstat (2024)
21. Adnan, Peterkin, Lopez et al. (2017) "Electronic Sentinel Surveillance of Influenza-Like Illness" *Applied Clinical Informatics*
22. Huang, Baker, Mcarthur (2014) "Implementing Hospital-Based Surveillance for Severe Acute Respiratory Infections Caused by Influenza and Other Respiratory Pathogens in New Zealand" *Western Pacific Surveillance and Response Journal*
23. Lopez, Wood, Prasad et al. (2016) "Influenza Surveillance in New Zealand"
24. New, Snomed Ct (2024)
25. Healthline, Health, Zealand
26. Wilson, Pienaar, Large et al. (2024) "Enhancing Aotearoa, New Zealand's Free Healthline Service Through Image Upload Technology" *Int J Telemed Appl*
27. "Sharing Excellence in Health and Disability Information Management"
28. Dalton, Carlson, Butler et al. (2013) "Building Influenza Surveillance Pyramids in Near Real Time, Australia" *Emerging Infectious Diseases*
29. Harvey, Trent, Mackenzie et al. (2022) "Calculating Incidence of Influenza-Like and COVID-Like Symptoms From Flutracking Participatory Survey Data" *MethodsX*
30. (2024) "Public Health Group from the Ministry of Health of New Zealand" *About FluTracking*
31. Huang, Turner, Baker (2015) "Southern Hemisphere Influenza and Vaccine Effectiveness Research and Surveillance"
32. Huang, Bandaranayake, Wood (2019) "Risk Factors and Attack Rates of Seasonal Influenza Infection: Results of the Southern Hemisphere Influenza and Vaccine Effectiveness Research and Surveillance (SHIVERS) Seroepidemiologic Cohort Study" *Journal of Infectious Diseases*
33. Wilson, Mason, Tobias et al. (2009) "Google Flu Trends" Data for Pandemic H1N1 Influenza: The New Zealand Experience" *Eurosurveillance*
34. (2015) "Flu Search Activity"
35. "& ctype = l& strai l= false & nselm = h& met_ y= flu_ index & scale_ y= lin& ind_ y= false & rdim= count ry& idim= count ry: NL: CL: DE: NZ& tstar t= 11935 29600 000& tunit = D& tlen= 1106& hl= en& dl= en& iconS"
36. Dictionary "cation/ natio nalminim um-datas et-hospi tal-event s-data-dicti onary"
37. Huang, Lopez, Mccallum et al. (1997) *Influenza Surveillance and Immunisation in New Zealand*
38. Schwabe, Becker, Seyferth et al. (2024) "The METRIC-Framework for Assessing Data Quality for Trustworthy AI in Medicine: A Systematic Review" *Npj Digital Medicine*
39. Sarker (2021) "Machine Learning: Algorithms, Real-World Applications and Research Directions" *SN Computer Science*
40. Moss, Zarebski, Dawson et al. (2017) "Retrospective Forecasting of the 2010-2014 Melbourne Influenza Seasons Using Multiple Surveillance Systems" *Epidemiology and Infection*
41. Yang, Cowling, Lau et al. (2015) "Forecasting Influenza Epidemics in Hong Kong" *PLoS Computational Biology*
42. Santillana, Nguyen, Dredze et al. (2015) "Combining Search, Social Media, and Traditional Data Sources to Improve Influenza Surveillance" *PLoS Computational Biology*
43. Albrecht, Broderick, Dost (2024) "Forecasting Severe Respiratory Disease Hospitalizations Using Machine Learning Algorithms" *BMC Medical Informatics and Decision Making*
44. Ertem, Raymond, Meyers "Optimal Multi-Source Forecasting of Seasonal Influenza" *PLoS Computational Biology*
45. (2018) |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12255040&blobtype=pdf | # Sexual networks, sexual practices, and sexual health among youths in WHO-South East Asia Region: a scoping review protocol
Amrita Rao, Rashmi Shinde, Mohammed Shahabuddin
## Abstract
Background South-East Asia Region has one of the largest youth populations in the world. All countries are striving to achieve the sustainable development goal by 2030; hence, it is important to prioritize healthcare services for youths. Youths in the age bracket of 18 to 24 years often engage in high-risk behaviors such as unsafe injecting practices/substance abuse. These high-risk practices lead to increased transmission of sexually transmitted infections including HIV among them. It is imperative to understand the dynamics around sexual transmission of diseases among youth. This review will map the available evidence and identify the gaps in sexual health interventions related to the sexual networks, sexual practices, and sexual health among youths across the World Health Organization (WHO)-South East Asia Region (SEAR).
MethodsThe scoping review is guided by the Arksey and Malley framework. Peer-reviewed articles focusing on youths in the age groups of 18 to 24 years, from the 11 countries of SEAR, will be accessed from three databases, namely PubMed, Scopus, and Journals@Ovid. Additionally, grey literature from 2015 to date will also be accessed. Two reviewers will independently screen the articles based on pre-defined eligibility criteria in Rayyan software. Data extraction will be carried out based on pre-specified variables aligned with the objectives. We will synthesize the evidence from the relevant qualitative, quantitative, and mixed method studies. The reporting will follow Preferred Reporting Items for Systematic reviews and Meta-Analysis extension for Scoping Reviews (PRISMA-ScR).Discussion This review will help to generate evidence focusing on the current sexual networks, sexual practices, and sexual health among youths in WHO SEAR with a focus on India, highlighting the gaps in sexual health interventions that need to be bridged. The insights from this review will assist in designing larger evidence-based intervention studies for improving the sexual networks, sexual practices, and sexual health among youths in this region. The findings from this review will be disseminated through peer-reviewed journals and conferences. Systematic review registration Open Science Framework. The link is https:// doi. org/ 10. 17605/ OSF. IO/ 2JSMC.
## Background
Globally, young people who are in the age bracket of 10 to 24 years comprise 1.8 million; 16% of the total population [1]. Young people comprise of youths in the age bracket of 15 to 24 years and adolescents between the ages of 10 and 19 [2]. Young people are considered a priority group, and engaging them for community health strengthening will be a step closer to achieving the targets of sustainable development goals by 2030 [3]. Further, adolescence is associated not only with rapid changes in growth and development but also with an increased desire for independence, stronger peer influence, and a heightened tendency for risk-taking and experimentation. This could escalate their risk-taking ability for substance use, unprotected sexual activity, and mental health challenges during this critical developmental period.
In Bhutan and Bangladesh, some of the issues highlighted were that the program staff were reluctant to distribute condoms and talk about sexually transmitted infections and HIV testing to adolescent clients [4]. These inhibitions need to be resolved among the caregivers so that better care and services are provided to adolescents and extended to youths. The recent World Health Organization (WHO) guidance on accelerated action for adolescent health is based on the current health status, mortality, and morbidity. This will help planning and implementing strategies focusing on this age group [5]. While adolescent health care remains a focus, youths may be neglected. Further, in a conservative society, sexrelated issues are a taboo for discussion; hence, youths do not actively seek counselling regarding sexual health. Social ostracism and disease-associated stigma have created an attitude of negativity and shame in the minds of this population. Vulnerability to high-risk behaviors, including substance abuse and high-risk sexual behavior such as lack of condom use and multiple sexual partners, often leads to increased chances of sexually transmitted infections (STI), including HIV [6,7]. Youths may seek care at the adult health care facilities, and a few may not even approach these facilities for treating STIs. Obtaining sexual history, especially from youths, requires counselling skills in addition to clinical skills. Hence, gathering this history is a challenge at times. Partner seeking and history of sexual exposure among youths need to be sought in the privacy and comfort of the youth.
Recent estimates indicate that around 0.21 million new HIV infections were reported among adolescent girls and young women aged 15 to 24, while among adolescent boys and young men (15 to 24 years), the sub-Saharan region contributed to nearly 63% of these infections [8]. Developed countries like the USA are working to reduce disparities and improve HIV outcomes among youths aged 13 to 24, who account for nearly 21% of the 37,968 new HIV diagnoses [9]. Even in India, the recent NFHS-5 data revealed that among the unmarried women and men of age 15-29 years, 4.2% of women and 23.3% of men had more than one sexual partner, and 17.24% of men had paid sex [10].
A few studies across the world have revealed that youths have poor knowledge of sexual and reproductive health and many of them do not even access the services. It is a common practice to discuss these issues among their own peers [11,12]. In addition, studies from various regions worldwide have highlighted a significant prevalence of sexual activity among unmarried adolescents and youths of both sexes. These studies indicate a decreasing age of sexual initiation, evolving sexual practices and preferences, and engagement in high-risk behaviors such as unprotected intercourse with multiple partners, raising serious public health concerns. Sexual behavior is shaped by a complex interplay of physiological factors, alongside rapidly shifting cultural and social influences that vary across generations [13,14]. Recent technological advancements and modernization have expanded partner seeking beyond the physical locations into the virtual world [15], making partner tracing complex.
Hence, it is important to study the risk potential of infectious disease transmission among specific populations. Understanding sexual behaviors, tracing sexual partners, are an important focus to prevent the spread of infections. These interactions and understanding of this dynamics will help to develop interventions for specific population groups that will prevent the spread of infections from high-risk groups to low-risk populations. Studying these sexual and social networks has to be specific for each locality or environment, and it cannot be generalized across all populations. This is an essential component to help improve health, especially related to HIV and STI [16].
Based on this background, this review aims to conduct a scoping review on literature published on sexual networks, sexual practices, and sexual health among youths in the WHO South East Asia Region (SEAR).
The objectives of this review are as follows:
1. To comprehensively map the available evidence and understand the sexual networks, sexual practices, and sexual health among youths across the WHO-South East Asia Region (SEAR). 2. To identify research gaps to facilitate larger multicentric projects related to the sexual networks, sexual practices, and sexual health among youths across the WHO South East Asia Region (SEAR).
## Methods
## Protocol design
The scoping review adopts an Arksey and O'Malley framework [17]. This framework involves the following steps: (1) formulating the research questions.
(2) Developing the search strategy and identifying relevant articles. (3) Selecting eligible studies. (4) Extracting data.
(5) Collating, organizing, and synthesizing the evidence.
We have registered the protocol in Open Science Framework (OSF). The link is https:// doi. org/ 10. 17605/ OSF. IO/ 2JSMC. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) checklist has been used while developing and reporting the protocol (Supplementary file 1) [18].
## Step 1: identifying the research question
The research question focuses on sexual networks, sexual practices, and sexual health among youths in WHO-South East Asia Region, as well as identifying and characterizing the gaps in the sexual health interventions.
The following questions will guide the review.
1. What strategies/approaches are used to identify sexual networks, sexual practices and sexual health among youths in WHO-SEAR? 2. What are the gaps in existing sexual health interventions?
Step 2: developing the search strategy and identifying the relevant articles
## Eligibility criteria
The inclusion and exclusion criteria are defined in Table 1. Since we are looking into the current practices of sexual networks, sexual practices, and sexual health, we have restricted the search strategy to include studies conducted over the last decade (from 1 January 2015 till 31 December 2024). We will search articles from three databases: PubMed, Scopus, and Ovid (Journals@Ovid). Both published and grey literature will be accessed. We will access the grey literature reports through the stakeholders and the experts during our consultations, and also, if accessible, from the World Health Organization during the same time period. The review will include studies with quantitative, qualitative, and mixed-method study designs. Publications focusing on youths (18 to 24 years) and on the eleven countries of WHO-SEAR will be included. We shall include studies that involve a wide age range and multi-centric studies that have disaggregated data pertaining to youths (18 to 24 years) and from the WHO SEAR countries. We shall include studies that are published only in English or that are translated into English since the team is fluent in only the English language.
## Conceptual framework and definition
In this review we will use the following three definitions: Sexual networks "are the network structures that emerge when individuals have sexual contact with each other" [20].
Sexual practices "are those actions, which people define as sexual, and their relationship to the architecture of society. These practices vary widely between and within societies, and change significantly through time" [21].
Sexual health "a state of physical, emotional, mental and social well-being in relation to sexuality; it is not merely the absence of disease, dysfunction or infirmity. Sexual health needs a positive and respectful approach to sexuality and sexual relationships, as well as the possibility of having pleasurable and safe sexual experiences, free of coercion, discrimination and violence. For sexual health to be attained and maintained, the sexual rights of all persons must be respected, protected and fulfilled" [22].
The search terms will focus on the broader domains of "sexual networks, " "sexual practices", "sexual health", "gaps in sexual health intervention", "youths", and the "11 countries belonging to WHO SEAR." This strategy will be developed by the core team, which includes two subject experts and will be peer-reviewed by the librarian (information specialist). Each domain will be conceptualized using PCC (population, concept and context) framework, Types of context-we will search for studies that focus on the 11 member countries of the WHO South-East Asia Region, namely Bangladesh, Bhutan, Myanmar, India, Indonesia, Maldives, Nepal, Sri Lanka, Thailand, Timor-Leste, and the Democratic People's Republic of Korea.
## Search strategy
A librarian, with experience in data management and literature retrieval, will help develop the search strategy and also access relevant databases such as PubMed, Scopus, and Journals @Ovid (through Ovid platform) for data retrieval. While PubMed is a platform offering access to biomedical literature, especially citations from Medline indexed journals and papers in PubMed Central, Scopus provides access to indexing, abstracting, and citation data of scientific literature across various disciplines. Besides, Ovid platform allows access to Journal@Ovid database, which indexes journals from 50 publishers and societies. For our study, we have considered young adults in the age bracket of 18 to 24 years. However, in PubMed, we have included articles that define "young adults" as individuals aged 19 to 24 years. For the other two databases, we have used the term youths and young adults to retrieve articles. The search is structured using the Joanna Briggs Institute(JBI) recommended PCC (population/concept/ context) framework. Every element of the framework is clearly defined using keywords and index terms (such as MESH terms in PubMed) so that they can be used along with Boolean operators to search different databases effectively. The search strategy used to search PubMed is given (Supplementary file 2). The Preferred Reporting Items for Systematic Reviews and Meta-Analysis extension for Scoping Reviews (PRISMA-ScR). PRISMA-ScR will be followed while reporting [19].
## Step 3: selecting eligible articles
The search results from all the databases will be imported into Rayyan ((https:// rayyan. ai/). This will be followed by removing duplicates, followed by title and abstract screening in Rayyan. This will be done by two reviewers independently to select studies that are relevant and fit the definition, inclusion criteria fitting into the PCC framework (population/concept/context). The articles that have been selected during the title and abstract screening will undergo full text screening. All relevant articles, qualitative, quantitative, and mixed-method studies will be included. Mixed-methods studies will be considered if the data from the quantitative or qualitative components can be extracted separately. The reasons for excluding the articles will be noted. Data extraction will be done by two independent reviewers. Any conflicts of opinion between the reviewers will be resolved mutually or, if needed, by the third person (co-author). The final search results will be reported in the PRISMA flow diagram (Supplementary file 3). The study is limited to the articles, which are published in the English language.
## Stage 4: extracting data
Two independent reviewers will be involved in the data extraction. A data extraction form adopted from Joanna Briggs Institute's template will be used for data extraction [23]. The key information collected from relevant studies will include date of publications, country/state, study design, objectives, methodology, and results. The data extraction will cover the details on the participants, including the socio-demographic profile and inclusion and exclusion criteria. The central concept of the study, including the operational definition if any, will be extracted. The socio-cultural, institutional, and economic context information will also be collected. We will also note the relevant findings and outcomes in terms of the PCC framework. The quantitative outcomes will represent the knowledge, awareness, and other factors related to sexual practices and sexual health, such as the risk of sexually transmitted infections and partner tracing. The data extracted from the qualitative component will include verbatim extract, interpretation of the results, and illustrations.
The data extraction form is prepared and will be pilot tested (Supplementary file 4). Data extraction agreement will be performed at regular intervals of 10%, 50%, and after completion of the selected studies. This would be done in consultation with all the team members. Necessary adjustments and changes will be made in consultation with the team members. Since this is a scoping review, quality assessment of articles is not mandatory.
## Stage 5: collating, organizing, and synthesizing the evidence
In this stage, the scoping review will synthesize evidence on sexual networks, sexual practices, and sexual health among youth in the WHO South East Asia Region. This will help identify the current sexual health interventions and the existing gaps to help countries draw upon the interventions that have worked and that have not worked for youths. This will play an important role in identifying factors that prevent the transmission of STIs, including HIV. It will also support the development of mechanisms to identify and trace sexual networks, increase the uptake of sexual health services, and create awareness among the youths.
Data extraction will be done using MS Excel. The outcomes from these studies will be tabulated and summarized narratively. For quantitative studies, the summary of findings and descriptive analysis will be reported based on the type of data extracted. Descriptive statistics will be used to present the findings. We shall try to pool the data and conduct meta-analysis on studies with common outcomes. Where statistical pooling is impossible, the findings will be presented in narrative form, including tables and figures where applicable, to aid in data presentation. For the qualitative studies, we will use charts to identify different concepts and maps of descriptive themes to populate analytical domains. This framework will be modified based on the emerging themes during analysis and in consultation with the researchers in the team. Coding of data will be conducted based on the themes identified. Each study will be indexed based on the themes identified. If newer themes emerge irrespective of their reconciliation with findings from other studies, they will be included in the interpretative framework.
The synthesized quantitative and qualitative findings will be integrated using a configurative analysis. These findings will be juxtaposed against each other to organize or link the evidence generated.
## Discussion
This review is an effort to summarize the evidence around sexual networks, sexual practices, and sexual health among youths in the WHO-SEAR. Generating evidence around sexual networks, sexual practices, and sexual health among youths will help gain insights into the current sexual networks, sexual health, and sexual practices among youths. This comprehensive understanding will help to develop appropriate interventions.
This scoping review may miss articles that focus on adolescents, as we have looked into youths in the age bracket of 18 to 24 years in the WHO-SEAR member countries. We shall not undertake the quality assessment as it is not mandatory for a scoping review. This may limit the strength of the evidence generated. Additionally, relevant articles in other languages will be missed as we have included articles published in English language only. Furthermore, we will not take into consideration published protocols, case reports, and systematic reviews even though they focus on sexual networks, sexual health, and sexual practices in WHO SEAR countries.
## References
1. (2020) "Young social entrepreneurship and the agenda for 2030"
2. (2024) "Adolescent health in the South-East Asia Region"
3. Melles, Rciker (2018) "Youth participation in HIV and sexual and reproductive health decision-making, policies, programmes: perspectives from the field" *Int J Adolesc Youth*
4. (2008) "Accelerating implementation of Adolescent Friendly Health Services (AFHS) in the South-East Asia Region: report of the meeting of the National Adolescent Health Programme Managers in member countries of the South-East Asia Region"
5. "Global Accelerated Action for the Health of Adolescents (AA-HA!): guidance to support country implementation"
6. Rao, Mamulwar, Shahabuddin et al. (2022) "HIV epidemic in Mizoram, India: A rapid review to inform future responses" *Indian J Med Res*
7. Oinam (2008) "Exploring the links between drug use and sexual vulnerability among young female injecting drug users in Manipur. Health and Population Innovation Fellowship Programme Working Paper"
8. (2023) "United Nations. The path that ends AIDS: UNAIDS Global AIDS Update"
9. "HIV national strategic plan for the United States: a roadmap to end the epidemic"
10. (2021) "3fs-public/ HIV-Natio nal-Strat egic-Plan-2021-2025"
11. (2021) *National Family Health Survey (NFHS-5)*
12. Saraçoğlu, Erdem, Doğan et al. (2014) "Youth sexual health: sexual knowledge, attitudes, and behavior among students at a university in Turkey" *Noro Psikiyatr Ars*
13. Agarwal, Brar, Kumar et al. (2021) "Sexual behaviour and practices among adolescents and young people: study and results from atertiary care centre of north India" *Int J Community Med Public Health*
14. Okpani, Okpani (2000) "Sexual activity and contraceptive use among female adolescents -a report from Port Harcourt" *Nigeria Afr J Reproduct Health*
15. Tumwakire, Ashaba, Mubangizi et al. (2022) "Sexual and reproductive health knowledge and practices among youth with and without mental illness in Uganda: a comparative study" *Trop Med Health*
16. (2022) "Mapping, size estimation and risk behaviour survey among key population groups in virtual space -A basic guide"
17. Friedman, Aral (2001) "Social networks, risk-potential networks, health, and disease" *J Urban Health*
18. Arksey, Malley (2005) "Scoping studies: towards a methodological framework" *Int J Soc Res Methodol*
19. Moher, Shamseer, Clarke et al. (2015) "Preferred reporting items for systematic review and metaanalysis protocols (PRISMA-P) 2015 statement" *Syst Rev*
20. Tricco, Zarin, Brien et al. (2018) "PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation" *Ann Intern Med*
21. Liljeros (2009) "Encyclopedia of complexity and systems science"
22. Adriaenssens, Hendrickx (2012) "Sex, price, and preferences: accounting for unsafe sexual practices in prostitution markets" *Sociol Health Illn*
23. (2025) "Pan American Health Organization. Sexual and reproductive health"
24. (2024) "JBI manual for evidence synthesis" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12692795&blobtype=pdf | # Design and Synthesis of Structurally Modified Analogs of 24Z-Isomasticadienonic Acid with Enhanced Anti-Proliferative Activity
Panagiota Stamou, Leentje Persoons, Dominique Schols, Steven De Jonghe, Leandros Skaltsounis, Ioannis Kostakis
## Abstract
Triterpenic acids represent a prominent class of bioactive compounds, with a wide range of biological properties, including anti-inflammatory, antiviral, and anticancer effects. Among them, 24Z-isomasticadienonic acid (IMNA), a major constituent of Chios Mastic Gum, has attracted little attention compared with other well-studied triterpenes such as oleanolic or betulinic acid, largely because its isolation in sufficient purity and quantity was only recently achieved. In this study, a series of IMNA analogs was synthesized through targeted modifications at the A-ring. These included the introduction of heteroatoms at position 2, the incorporation of heterocyclic rings such as an oxazole and a thiazole, and rearrangements of the ring structure. The new compounds were evaluated for their antiproliferative activity against a diverse panel of cancer cell lines (Capan-1, HCT-116, LN-229, NCI-H460, DND-41, HL-60, K-562, Z-138). Among the synthesized analogs, compounds 3, 7 and 9 demonstrated selective anticancer activity toward the Capan-1 cell line, whereas compounds 6 and 10 exhibited broad-spectrum cytotoxic effects across multiple cancer cell lines. Overall, these findings highlight IMNA as a promising scaffold for anticancer drug design and demonstrate the value of A-ring modifications in improving activity and selectivity.
## 1. Introduction
In recent years, natural products have played a significant role in medicine, serving not only as a source of potential chemotherapeutic agents but also as lead compounds for the semi-synthesis or total synthesis of new drugs [1][2][3]. Characteristic examples are the pentacyclic and tetracyclic triterpenic acids, a class of chemical compounds found in plants, marine organisms, bacteria and fungi, known for their diverse biological activities, including anti-inflammatory, antiviral and anticancer effects [4][5][6][7]. Among the most studied triterpenic acids are oleanolic, ursolic, moronic and betulinic acids, which have attracted significant scientific interest due to their broad spectrum of biological activities and their widespread occurrence in numerous plant species [8][9][10][11][12] (Figure 1). Despite their notable bioactivity and ability to interact with biological systems through multiple mechanisms, triterpenes exhibit certain limitations [13][14][15]. One of the major challenges is their low aqueous solubility, which significantly reduces their absorption and bioavailability in vivo. Consequently, although triterpenes often demonstrate potent activity in vitro, they often show limited therapeutic efficacy when evaluated in vivo. Therefore, the development of more potent analogs based on the bioactive triterpene scaffold is required [16].
In recent years, numerous triterpene analogs have been synthesized, many of which exhibit potent biological activities, amongst others, anticancer effects. Reflecting their therapeutic potential, an increasing number of patents has been filed to protect these newly developed triterpenoid compounds ( [6,17,18] and references herein). A prominent group among these includes substituted derivatives with a heteroatom at position 2 of the A-ring, as well as those incorporating heterocyclic or polycyclic structures fused to the triterpenoid backbone [16,19,20] (Figure 2). Specifically, due to their chemical structure, heterocyclic compounds are often more stable and resistant to metabolic degradation. As a result, they remain active in the body for longer periods, enhancing therapeutic efficacy. Additionally, the presence of aromatic systems, heteroatoms and heterocyclic rings enables absorption in the ultraviolet (UV) region, facilitating compound detection and monitoring. Consequently, these structural modifications offer clear advantages in terms of both anti-proliferative potential and clinical applicability [21]. 24Z-isomasticadienonic acid (IMNA, 1) is a tetracyclic triterpene and one of the main constituents of Chios Mastic Gum (CMG), along with 24Z-masticadienonic acid (MNA) [22]. Although IMNA is naturally abundant in the resin, its isolation in high purity and sufficient quantities remained particularly challenging until recently. As a result, its biological properties remain underexplored and only a limited number of analogs have been synthesized to date [23]. In contrast, other triterpenic acids-such as oleanolic, betulinic, and ursolic acid-have been extensively studied, and numerous analogs of these compounds have been developed [9,[24][25][26]. These observations underscore the relevance of investigating IMNA's anti-proliferative potential and support the rationale for designing novel analogs based on its scaffold.
Therefore, the aim of the study is to develop IMNA analogs for the purpose of investigating their potential activity against various cancer cell lines. Drawing from the literature and the chemical structures of similar bioactive triterpenoid analogs, a series of IMNA derivatives was designed, featuring changes on the A-ring. Specifically, these modifications include the introduction of heterocyclic rings such as an oxazole and thiazole, along with substitution by a heteroatom at position 2. Furthermore, considering that the structure and size of the A-ring significantly influence the molecule's conformation and, consequently, its biological activity, additional analogs possessing a rearrangement of the A-ring were also synthesized. These modifications aim to enhance pharmacological properties and provide insight into structure-activity relationships.
## 2. Results and Discussion
2.1. Chemistry 2.1.1. Semi-Synthesis of 24Z-Isomasticadienonic Acid (1) Compound 1 (IMNA) was synthesized following a procedure previously reported by our group. Briefly, 24Z-masticadienonic acid (MNA) was isolated from the resin of CMG by preparative high-performance liquid chromatography (preparative HPLC). In the presence of BBr 3 , isomerization of the internal double bond of MNA took place, affording compound 1 (Scheme 1) [23,27].
## 2.1.2. Synthesis of 24Z-Isomasticadienonic Acid Analogs
The synthesis of compounds 2-7 is outlined in Scheme 2. Treatment of IMNA with ethyl formate in the presence of sodium ethoxide afforded 2-hydroxymethylen-3oxotirucalla-8,24Z-dien-26-oic acid (2) via a Claisen condensation [28]. Subsequent reaction with hydroxylamine hydrochloride in absolute ethanol, afforded the corresponding oxime, which upon intramolecular cyclization yielded the isoxazole derivative 3 [29,30]. Consequently, treatment of 1 with pyridinium perbromide (PyHBr 3 ) afforded 4 as a mixture of αand β-epimers in a 6.5:3.5 ratio, complicating NMR analysis and complete structural assignment [31]. Nonetheless, the 1 H NMR spectrum displayed a peak at 5.10 ppm, integrating for one proton, characteristic of the proton at position 2 (Figure S13, see Supplementary Material). Bromine incorporation was further confirmed by mass spectrometry, where under negative ionization mode, the characteristic isotopic pattern of bromine was observed together with a pseudo-molecular ion at m/z 531.2479, consistent with the desired product. Subsequent treatment of 4 with thiourea in absolute ethanol afforded the aminothiazole derivative 5 [32,33]. Its structural elucidation was supported by the HMBC spectrum (Figure S21, see Supplementary Material). The protons at C-1, resonating at 2.61 and 2.30 ppm, correlated with the quaternary carbons C-2 and C-3 of the aminothiazole ring, observed at 115.92 and 149.91 ppm, respectively. Notably, C-3 showed HMBC correlations with the protons of the methyl group at positions 28 and 29. In addition, treatment of 4 with potassium thiocyanate in DMSO yielded 2-thiocyanate-3-oxotirucalla-8,24Z-dien-26-oic acid (6). Compound 6, similarly to 4, was obtained as an inseparable mixture of epimers (α and β) in a ratio of 6.7:3.3. Its structure was confirmed by mass spectrometry, which in negative ionization mode showed a pseudo-molecular ion at m/z 510.3027, consistent with the expected product. Finally, the morpholino derivative 7 was obtained by treatment of compound 6 with morpholinium acetate [34].
The synthesis of compounds 8-11 is depicted in Scheme 3. Treatment of compound 1 in t-BuOH with potassium tert-butoxide under an oxygen atmosphere, followed by the addition of aqueous potassium hydroxide, afforded the oxidized compound 2-hydroxy-3-oxotirucalla-8,24Z-dien-26-oic acid (8), as well as 1α-hydroxy-3-oxanor-tirucalla-8,24Z-dien-26-oic acid (9) and 3β-hydroxy-1(2→3)-abeotirucalla-8,24Z-dien-26-oic acid (10), through benzilic acid rearrangement [35]. Compounds 8 and 10 have been previously reported, whereas 9 represents a novel analog of the naturally occurring 24Z-isomasticadienonic acid [23]. Structural elucidation of compound 9 was performed using mass spectrometry and one-and two-dimensional NMR spectroscopy. In the mass spectrum under negative ionization mode, a pseudo-molecular ion was detected at m/z 471.3152, corresponding to the molecular formula C 29 H 43 O 5 . This indicates the presence of two additional oxygen atoms and one carbon atom less, when compared with 1. The 13 C NMR spectrum confirmed these features, showing both the expected number of carbon signals and chemical shifts consistent with additional oxygenated carbons (Figure S39, see Supplementary Material). The proton at C-1 (δ H 5.49 ppm), as well as the methyl protons at C-29 and C-30, exhibited HMBC correlations to the carbonyl carbon at C-3 (δ C 181.4 ppm).
The hydroxyl group at C-1 accounted for both the de-shielding of the proton at this position (5.49 ppm) and the chemical shift of C-1 at 102.1 ppm. Furthermore, NOESY data established the stereochemistry of the hydroxyl substituent at C-1, with the proton showing NOE correlations to the methyl group at C-19 (Figure S40, see Supplementary Material). Finally, treatment of the α-carboxylic acid 10 with palladium tetraacetate triggered decarboxylation, followed by oxidation of the hydroxyl group, furnishing the final compound 11 [36].
## 2.2. Antitumoral Evaluation
IMNA and the newly synthesized derivatives were evaluated for potential antitumoral activity using a panel of solid cancers, including pancreatic adenocarcinoma (Capan-1), colorectal carcinoma (HCT-116), glioblastoma (LN-229) and lung carcinoma (NCI-H460). In parallel, the cytotoxicity versus various hematological cancers, including acute lymphoblastic leukemia (DND-41), acute myeloid leukemia (HL-60), chronic myeloid leukemia (K-562) and non-Hodgkin's lymphoma (Z-138) was also determined. Etoposide, used as positive control, displayed potent antiproliferative activity against all cancer cell lines (Table 1). Notably, the Capan-1 pancreatic cell line emerged as the most sensitive, with all compounds showing cytotoxicity in the 6-39 µM range. Compounds 3, 7, and 9 demonstrated selective cytotoxicity toward Capan-1, with minimal effects on the other tested cell lines. Compounds 6 and 10 showed broader antiproliferative activity, affecting most solid and hematological cancer lines. Compounds 4, 5, and IMNA exhibited intermediate activity across multiple cell lines, whereas compounds 8 and 11 displayed a more limited and restricted cytotoxic profile. To ensure a selective antitumoral effect, the effect of the compounds on the viability of normal (i.e., non-cancerous) peripheral blood mononuclear cells (PBMCs) was studied (Figure 3). Etoposide (as reference compound) at a concentration of 10 µM did not display toxicity to PMBC. At a concentration of 50 µM, none of the analogues (compounds 3-11 and IMNA) decreased the viability of PBMCs, suggesting a selective effect on cancer cells.
## 3. Materials and Methods
Ethyl acetate (EtOAc), n-Hexane, methanol (MeOH), dichloromethane (DCM) and cyclohexane (c-hexane) used for extraction and purification were of analytical grade (Fisher Scientific, Loughborough, Belgium), while water (H 2 O) was distilled. For the pH adjustment sodium hydroxide pellets (NaOH-penta CHEMICALS UNITED) and hydrochloric acid (HCl-analytical grade; Fisher Scientific, Loughborough, Belgium) were used. Acetonitrile (ACN, Avantor Performance Materials, Gliwice, Poland), and H 2 O (Fisher Scientific, Loughborough, Belgium) used for preparative High-Performance Liquid Chromatography (prep-HPLC) analysis were of HPLC grade. Reaction progresses were monitored by thin-layer chromatography on pre-coated silica gel 60 F254 plates from Merck (0.25 mm thickness) and were visualized on 254 and 366 nm (UV lamp). Flash chromatography was performed using Merck silica gel 60A (0.040-0.060 mm) (Merck, Kenilworth, NJ, USA) with the specified solvent system, typically applying gradients of increasing polarity. All commercially available reagents were purchased from Alfa Aesar and used without any further purification. The isolation of MNA was performed using preparative RP-HPLC (Buchi, Flawil, Switzerland, Pure C-850 FlashPrep) hyphenated to a Photo Diode Array (PDA) detector. All solvent evaporations were performed using a rotary evaporator (Buchi) with a water bath at 40 • C. Optical rotations were measured with a Jasco P-2000 Polarimeter (Jasco, Tokyo, Japan). Measurements were performed at a wavelength of 589 nm (Na D-line) and a path length of 50 mm. 1 H NMR, 13 C NMR and 2D spectra were recorded on a Bruker Avance III 600 MHz spectrometer (Bruker Biospin AG, Faellanden, Switzerland) and on a Bruker Avance NEO 400 MHz spectrometer (Bruker, Faellanden, Switzerland) in deuterated solvents (see Supplementary Materials). Chemical shifts are expressed as δ values in parts per million (ppm), and the coupling constants (J) are given in Hertz (Hz). The signals of 1 H and 13 C NMR spectra were unambiguously assigned by using 2D NMR techniques: 1 MNA was obtained from the Total Mastic Extract Without Polymer (TMEWP) after removal of cis-1,4-poly-β-myrcene. From 10.0 g of TMEWP, 5.1 g of the acidic fraction of triterpenes (AF) were recovered through liquid-liquid extraction with gradual pH adjustment. In brief, TMEWP was partitioned in a separatory funnel between a polar phase (20% NaOH in H 2 O/MeOH, 1:1, pH = 11) and a non-polar phase (n-hexane/EtOAc, 8:2). The aqueous phase was then acidified with 1 N HCl to pH 3 and extracted with EtOAc, affording the AF. Finally, a RP-HPLC-PDA method was employed using a gradient elution system consisting of ACN and H 2 O, through which 25.0 mg of pure MNA were isolated from 250.0 mg of AF per injection. All separations were carried out on a Buchi Pure C-850 FlashPrep system coupled to a PDA detector, using a reversed-phase column (Agilent 5 Prep-C18, Agilent Technologies, Santa Clara, CA, USA, 50 × 50 mm). NMR data were consistent with the literature [22,23].
## 3.1.2. Synthesis of 24Z-Isomasticadienonic Acid (1)
MNA (60.0 mg, 0.13 mmol) was dissolved in anhydrous DCM (23.0 mL), and BBr 3 (1.0 eq) was added at 0 • C under an argon atmosphere. The reaction mixture was stirred for 15 min at 0 • C, after which water was added and the product extracted with EtOAc. The organic layer was dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure to afford 50.0 mg of IMNA as a white solid. (yield = 83%). NMR data were consistent with the literature [23]. To a solution of 1 (124.4 mg, 0.274 mmol) in HCOOEt (5.0 mL) under argon, was added NaOEt (25.0 eq) and the mixture was stirred at room temperature for 17 h. Upon completion of the reaction, H 2 O was added, followed by acidification with 6 N HCl and extraction with EtOAc. After drying the organic phase over anhydrous Na 2 SO 4 and concentrating it under reduced pressure, 120.0 mg of 2 were obtained as a white solid (yield = 90%).
## 3.2. Synthesis of 24Z-Isomasticadienonic
[a] 25 D +15 (c 0.243, CHCl 3 ); 1 To a solution of 2 (248.7 mg, 0.515 mmol) in abs EtOH (12.4 mL) and H 2 O (1.8 mL) NH 2 OH•HCl was added (10.0 eq) and the mixture was stirred at 80 • C for 2 h. Subsequently, the EtOH was evaporated under reduced pressure, and the residue was extracted with EtOAc and H 2 O. The organic phase was then dried over anhydrous Na 2 SO 4 and concentrated. A solid residue of 210.0 mg was obtained, which was further purified by column chromatography (silica gel) using a c-hexane/EtOAc 95.5:4.5 v/v, as the eluent. Finally, 190.0 mg of 3 were obtained as a pale yellow solid (yield = 76%). [a] 25 D +13 (c 0.560, CHCl 3 ); 1 To a solution of 1 (210.0 mg, 0.462 mmol) in ACN (15.0 mL) under argon, PyHBr 3 (1.5 eq) is added, and the mixture is stirred at room temperature for 12 h. Upon completion of the reaction, the solvent is removed under reduced pressure, and the resulting oily residue is dissolved in EtOAc, washed with H 2 O, dried over anhydrous Na 2 SO 4 , filtered, and concentrated under reduced pressure. A solid residue of 250.0 mg is obtained, which is purified by column chromatography (silica gel) using c-hexane/DCM: 5:95→100:0% v/v, as the eluent. Finally, a mixture of the 2-bromo-substituted IMNA isomers is obtained (164.0 mg) as a yellow solid (yield = 66%).
[a] To a solution of 4 (52.0 mg, 0.097 mmol) in absolute EtOH, 10.0 eq. of thiourea were added. The resulting mixture was stirred at room temperature for 5 days under inert atmosphere. After completion of the reaction and the evaporation of the EtOH, the residue was dissolved in EtOAc, washed with H 2 O, dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude product was purified by flash chromatography (silica gel), using DCM/MeOH, 100:0→98:2 v/v as the eluent, to afford 40.0 mg of 5 (yield: 80%). To a solution of 4 (122.8 mg, 0.230 mmol) in 2.0 mL DMSO, KSCN (156 mg, 1.61 mmol, 7.0 eq) were added. The resulting mixture was stirred at 90 • C for 4 h under inert atmosphere. After completion of the reaction the residue was dissolved in EtOAc, washed with H 2 O, dried over anhydrous Na 2 SO 4 , and concentrated under reduced pressure. The crude product was purified by flash chromatography (silica gel), using c-hexane/DCM: 5:95→100:0% v/v, as the eluent, to afford 105.0 mg of 6 (yield: 89%). [a] 25 D -80 (c 0.320, CHCl 3 ); LC-ESI-HRMS (-) calcd. for C 31 H 44 NO 3 S -: m/z 510.3047, found 510.3027.
3.2.6. Synthesis of (S,Z)-2-Methyl-6-((1S,3aS,5aR,10aS,12aS)-3a,6,6,10a,12a-pentamethyl-8morpholino-2,3,3a,4,5,5a,6,10,10a,11,12,12a-dodecahydro-1Hcyclopenta [7,8]phenanthro [2,3-d]
To a solution of 6 (83.0 mg, 0.162 mmol) in CHCl 3 , 5.0 eq. of freshly prepared Morpholinium acetate were added. The resulting mixture was stirred at room temperature for 7 days. After completion of the reaction the residue was dissolved in DCM, washed with H 2 O, dried over anhydrous Na 2 SO 4 and concentrated under reduced pressure. The crude product was purified by flash chromatography (silica gel), using c-hexane/DCM: 20:80→100:0% v/v as the eluent, to afford 40.0 mg of 3 (yield: 40%). [a] 25 D +13 (c 0.701, CHCl 3 ); 1 H NMR (600 MHz, CDCl 3 ) δ (ppm) 6.08 (1H, dt, J = 1.08/7.55 Hz, H-24), 3.79 (4H, m, H-33, H-34), 3.40 (4H, m, H-32, H-35), 2.64 (1H, d, J = 15.25 Hz, H-2a), 2.57 (1H, m, H-23a), 2.47 (1H, m, H-23b), 2.37 (1H, d, J = 15.06 Hz, H-2b), 2.14 (1H, m, H-7a), 2.07 (2H, m, H-16), 1.98 (1H, m, H-7b), 1.92 (3H, s, CH 3 -27), 1.76 (2H, m, H-12), 1.74 (1H, m, H-6a), 1.58 (2H, m, H-11), 1.58 (1H, m, H-5), 1.54 (1H, m, H-22a), 1.51 (1H, m, H-17), 1.44 (1H, m, H-6b), 1.44 (1H, m, H-20), 1.35 (2H, m, H-15), 1.22 (3H, s, CH 3 -29), 1.15 (4H, m, H-22b), 1.14 (3H, s, CH 3 -28), 0.95 (3H, s, CH 3 -19), 0.94 (3H, d, J = 6.28 Hz, CH 3 -21), 0.90 (3H, s, CH 3 -30), 0.77 (3H, s, CH 3 -18); 13 3.2.7. Synthesis of 2-Hydroxy-3-oxotirucalla-8,24Z-dien-26-oic acid (8), 1α-hydroxy-3-oxa-nor-tirucalla-8,24Z-dien-27-oic acid (9), 3β-hydroxy-1(2→3)-abeotirucalla-8,24Z-dien-26-oic Acid (10) To a solution of 1 (200.0 mg, 0.439 mmol) in t-BuOH, t-BuOK (359 mg, 3.2 mmol, 7.3 eq) was added under continuous oxygen flow and the resulting mixture was stirred at 40 • C for 2 h. Then, KOH (123.0 mg, 2.2 mmol, 5.0 eq) in 1.4 mL of H 2 O were added, and the mixture was stirred at 75 • C for 24 h. After completion of the reaction, H 2 O and 6N HCl were added until the pH reached 4. The mixture was extracted with EtOAc, the organic phase was washed with H 2 O (×5), dried over anhydrous Na 2 SO 4 , and concentrated to dryness. This afforded 190.0 mg of an oily residue, which was purified by column chromatography (silica gel) using a DCM/MeOH. Analog 8 eluted with 1% MeOH (20.0 mg, yield = 10%), 9 with 3% MeOH (40 mg, yield = 21%), 10 with 5-10% MeOH (60 mg, yield = 30%). %). NMR data of compounds 8 and 10 were consistent with the literature [23].
3.2.8. Spectroscopic Data of 1α-Hydroxy-3-oxa-nor-oxotirucalla-8,24Z-dien-27-oic acid (9) [a] 3.2.9. Synthesis of 3-One-1(2→3)-abeotirucalla-8,24Z-dien-26-oic Acid (11) To a solution of 10 (168.8 mg, 0.347 mmol) in AcOH (6.0 mL) Pb(C 2 H 3 O 2 ) 4 (309.6 mg, 0.694 mmol, 2.0 eq) was added, and the resulting mixture was stirred at room temperature for 5 h. After completion of the reaction, AcOH was evaporated under reduced pressure, the residue was redissolved in EtOAc and washed with H 2 O (x5). After drying the organic phase over anhydrous Na 2 SO 4 and concentrating under reduced pressure, 191.6 mg of solid residue was obtained. This was purified by column chromatography (silica gel) using c-hexane/EtOAc: 96:4 → 92:8 v/v, as the eluent. Finally, 95.0 mg of 11 was obtained as a pale yellow solid (yield = 62%). [a]
$$thiazol-1-yl)hept-2-enoic Acid (7)$$
## 3.3. Antitumoral Assays
Human cancer cell lines used in this study included Capan-1, HCT-116, NCI-H460, LN-229, HL-60, K-562, and Z-138, all obtained from the American Type Culture Collection (ATCC). The DND-41 cell line was sourced from the Deutsche Sammlung von Mikroorganismen und Zellkulturen (DSMZ). All cell lines were cultured in media obtained from Gibco Life Technologies, supplemented with 10% fetal bovine serum (HyClone, Logan, UT, USA). Adherent cell lines were seeded in 384-well tissue culture plates at the following densities: Capan-1 at 500 cells/well, and HCT-116, NCI-H460, and LN-229 at 1500 cells/well. After overnight incubation to allow cell attachment, cells were treated with a seven-point serial dilution of test compounds, ranging from 100 µM to 0.006 µM. Suspension cell lines were seeded as follows: HL-60, K-562, and Z-138 at 2500 cells/well, and DND-41 at 5500 cells/well, using the same compound concentration range. All treatments were performed in 384-well plates under identical conditions. Following 72 h of compound exposure, cell viability was assessed using the CellTiter 96 ® AQueous One Solution Cell Proliferation Assay (MTS, Promega, Leiden, The Netherlands), following the manufacturer's protocol. The final assay mixture contained 333 µg/mL MTS and 25 µM phenazine methosulfate (PMS). Absorbance was measured at 490 nm using a SpectraMax Plus 384 microplate reader (Molecular Devices, San Jose, CA, USA). Optical density values were used to calculate the half-maximal inhibitory concentration (IC 50 ) for each compound. All compounds were tested in a minimum of two independent biological replicates to ensure reproducibility.
## 3.4. Cell Viability of PBMCs
A buffy coat preparation sourced from a healthy donor was acquired from the Blood Transfusion Center in Leuven, Belgium. Peripheral blood mononuclear cells (PBMCs) were separated using density gradient centrifugation over Lymphoprep (density: 1.077 g/mL) from Nycomed. These cells were then cultured in cell culture medium (RPMI, Gibco, Waltham, MA, USA) supplemented with 10% FBS. The PBMCs were seeded at a density of 28,000 cells per well in 384-well tissue culture plates containing the test compounds at concentrations ranging from 50 µM to 3.2 nM (for compounds 3-11 and IMNA) and from 10 µM to 0.6 nM (for etoposide). Following a 72 h incubation period, cell viability was assessed using the MTS cell viability assay. Each compound was evaluated in duplicate.
## 4. Conclusions
In this study, a focused library of 11 IMNA analogs was synthesized through targeted A-ring modifications, including the introduction of heteroatoms, heterocyclic substitutions, and structural rearrangements. These modifications were designed to explore the impact of A-ring structural changes on anticancer activity, while preserving key pharmacophoric elements of the IMNA scaffold. The newly synthesized compounds were evaluated for their antiproliferative effects across a panel of solid and hematological cancer cell lines. Among the solid tumors, the pancreatic adenocarcinoma cell line (Capan-1) emerged as the most sensitive, with all compounds showing cytotoxicity in the 6-39 µM range. Compounds 3, 7, and 9 displayed selective activity toward Capan-1, whereas compounds 6 and 10 exhibited broad-spectrum antiproliferative effects across all tested cell lines. IMNA itself, along with several other derivatives, showed intermediate activity, while compounds 8 and 11 had a more restricted cytotoxic profile. These findings underscore IMNA as a promising scaffold and illustrate that rational A-ring modifications can meaningfully enhance both activity and selectivity in triterpenoid-based anticancer agents. These results are based on initial in vitro studies, and they provide a solid foundation for further research, which will focus on optimizing pharmacological properties, expanding structural modifications beyond the A-ring, and integrating computational and mechanistic approaches to elucidate the mode of action and to guide the design of more potent and selective analogs.
## References
1. Newman, Cragg (2016) "Natural Products as Sources of New Drugs from 1981 to 2014" *J. Nat. Prod*
2. Reddy, Odhav, Bhoola (2003) "Natural Products for Cancer Prevention: A Global Perspective" *Pharmacol. Ther*
3. Atanasov, Zotchev, Dirsch et al. (2021) "Natural Products in Drug Discovery: Advances and Opportunities" *Nat. Rev. Drug Discov*
4. Cassels, Asencio (2011) "Anti-HIV Activity of Natural Triterpenoids and Hemisynthetic Derivatives 2004-2009" *Phytochem. Rev*
5. Agra, Ferro, Barbosa et al. (2015) "Triterpenes with Healing Activity: A Systematic Review" *J. Dermatol. Treat*
6. Shanmugam, Nguyen, Kumar et al. (2012) "Targeted Inhibition of Tumor Proliferation, Survival, and Metastasis by Pentacyclic Triterpenoids: Potential Role in Prevention and Therapy of Cancer" *Cancer Lett*
7. Safayhi, Sailer (1997) "Anti-Inflammatory Actions of Pentacyclic Triterpenes" *Planta Med*
8. Hostettmann-Kaldas, Nakanishi (1979) "Moronic Acid, a Simple Triterpenoid Keto Acid with Antimicrobial Activity Isolated from Ozoroa Mucronata" *Planta Med*
9. Liu (1995) "Pharmacology of Oleanolic Acid and Ursolic Acid" *J. Ethnopharmacol*
10. Woźniak, Sk Ąpska, Marszałek (2015) "Ursolic Acid-A Pentacyclic Triterpenoid with a Wide Spectrum of Pharmacological Activities" *Molecules*
11. Eichenmüller, Von Schweinitz, Kappler (2009) "Betulinic Acid Treatment Promotes Apoptosis in Hepatoblastoma Cells" *Int. J. Oncol*
12. Genet, Strehle, Schmidt et al. (2010) "Structure-Activity Relationship Study of Betulinic Acid, A Novel and Selective TGR5 Agonist, and Its Synthetic Derivatives: Potential Impact in Diabetes" *J. Med. Chem*
13. Xu, Pang, Zhou et al. (2014) "Proteomic Investigation into Betulinic Acid-Induced Apoptosis of Human Cervical Cancer HeLa Cells" *PLoS ONE*
14. Li, Du, Kong et al. (2013) "Lamin B1 Is a Novel Therapeutic Target of Betulinic Acid in Pancreatic Cancer" *Clin. Cancer Res*
15. Bernardo, Cunha-Oliveira, Serafim et al. (2013) "Dimethylaminopyridine Derivatives of Lupane Triterpenoids Cause Mitochondrial Disruption and Induce the Permeability Transition" *Bioorg. Med. Chem*
16. Borkova, Gurska, Dzubak et al. (2016) "Lupane and 18α-Oleanane Derivatives Substituted in the Position 2, Their Cytotoxicity and Influence on Cancer Cells" *Eur. J. Med. Chem*
17. De Vasconcelos Cerqueira, Braz, Carvalho Nascimento Juúnior et al. (2020) "Terpenes with Antitumor Activity: A Patent Review" *Recent Pat. Anticancer Drug Discov*
18. Yan, Gong, Zheng et al. (2014) "Triterpenoids as Reversal Agents for Anticancer Drug Resistance Treatment" *Drug Discov. Today*
19. Xu, Li, Luo et al. (2012) "Synthesis and Biological Evaluation of Heterocyclic Ring-Fused Betulinic Acid Derivatives as Novel Inhibitors of Osteoclast Differentiation and Bone Resorption" *J. Med. Chem*
20. Herrera-España, Us-Martín, Hernández-Ortega et al. *Structure Analysis and Activity against Breast and Cervix Cancer Cells of a Triterpenoid Thiazole Derived from Ochraceolide A. J. Mol*
21. Kvasnica, Urban, Dickinson et al. (2015) "Pentacyclic Triterpenoids with Nitrogen-and Sulfur-Containing Heterocycles: Synthesis and Medicinal Significance" *Nat. Prod. Rep*
22. Paraschos, Magiatis, Mitakou et al. (2007) "In Vitro and in Vivo Activities of Chios Mastic Gum Extracts and Constituents against Helicobacter Pylori" *Antimicrob. Agents Chemother*
23. Stamou, Gianniou, Trougakos et al. (1618) "Anti-Inflammatory Activity of the Major Triterpenic Acids of Chios Mastic Gum and Their Semi-Synthetic Analogues" *Biomolecules*
24. Ayeleso, Matumba, Mukwevho (1915) "Oleanolic Acid and Its Derivatives: Biological Activities and Therapeutic Potential in Chronic Diseases" *Molecules*
25. Aiken, Chen (2005) "Betulinic Acid Derivatives as HIV-1 Antivirals"
26. Mlala, Oyedeji, Gondwe et al. (2019) "Ursolic Acid and Its Derivatives as Bioactive Agents"
27. Hubert, Reimlinger (1970) "The Isomerization of Olefins Part II. Thermal and Catalytic Lsomerization of Olefins Using Acids, Metals, Metal Complexes, or Boron Compounds as Catalysts" *Synthesis*
28. Leal, Wang, Salvador et al. (2012) "Synthesis of Novel Ursolic Acid Heterocyclic Derivatives with Improved Abilities of Antiproliferation and Induction of P53, P21waf1 and NOXA in Pancreatic Cancer Cells" *Bioorg. Med. Chem*
29. Fu, Zhou, Bao et al. (2014) "Tryptophan Hydroxylase 1 (Tph-1)-Targeted Bone Anabolic Agents for Osteoporosis" *J. Med. Chem*
30. Kang, Hu, Gao et al. (2012) "Synthesis, Anti-Proliferative and Proapoptotic Activity of Novel Oleanolic Acid Azaheterocyclic Derivatives" *MedChemComm*
31. Chen, Li, Zheng et al. (2017) "Discovery of FZU-03,010 as a Self-Assembling Anticancer Amphiphile for Acute Myeloid Leukemia" *Bioorg. Med. Chem. Lett*
32. Wang, Yang, Fan et al. (2019) "Design and Synthesis of the Novel Oleanolic Acid-Cinnamic Acid Ester Derivatives and Glycyrrhetinic Acid-Cinnamic Acid Ester Derivatives with Cytotoxic Properties" *Bioorg. Chem*
33. Wang, Zhang, Ma et al. (2022) "Design, Synthesis, and Biological Evaluation of Ocotillol Derivatives Fused with 2-Aminothiazole via A-Ring as Modulators of P-Glycoprotein-Mediated Multidrug Resistance" *Eur. J. Med. Chem*
34. Borkova, Adamek, Kalina et al. (2017) "Synthesis and Cytotoxic Activity of Triterpenoid Thiazoles Derived from Allobetulin, Methyl Betulonate, Methyl Oleanonate, and Oleanonic Acid" *ChemMedChem*
35. Kacharov, Yemets, Nemykin et al. (2013) "Stereoselectivity of A-Ring Contraction for 3-Oxotriterpenoids"
36. Banerjee (2021) "Lead Tetraacetate in Organic Synthesis" *Org. Med. Chem. Int. J*
37. "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=PMC12375836&blobtype=pdf | # Magnesium Oxide Nanoparticles: A New Frontier in Antiviral Therapy Against Herpes Simplex Virus Type 1
Abdulhussain Jwaziri, Zahra Salavatiha, Seyed Kiani, Pegah Khales, Masoud Vazirzadeh, Ahmad Tavakoli
## Abstract
Background and Aims: Herpes simplex virus Type 1 (HSV-1) causes a wide spectrum of diseases in humans, including skin and mucosal ulcers, encephalitis, and keratitis. Acyclovir is regarded as the gold standard for treating infections with this virus. However, there are certain drawbacks to using this drug, such as its inefectiveness against treatment-resistant virus strains. Terefore, the development of novel and efective drugs to combat this virus is urgently needed. Te present work aims to explore the efcacy of magnesium oxide nanoparticles (MgONPs) against HSV-1 in vitro as a potential novel antiviral agent. Methods: MgONPs were characterized by X-ray difraction, energy-dispersive X-ray spectroscopy, feld-emission scanning electron microscope, ultraviolet-visible spectrophotometry, Fourier-transform infrared spectroscopy, dynamic light scattering, and zeta potential. To assess the cytotoxic efects of MgONPs on Vero cells, the neutral red uptake assay was used. Te efects of MgONPs at nontoxic concentrations on HSV-1 were then examined using a quantitative real-time PCR assay. Results: No toxic efect was observed in all used concentrations of MgONPs (up to a concentration of 1000 μg/mL). Tree-hour incubation of HSV-1 with MgONPs at concentrations of 900 and 1000 μg/mL resulted in a remarkable decrease in viral load with an inhibition rate of 93.6% and 96.8%, respectively. Te results from the posttreatment assay also showed that MgONPs at concentrations of 300 and 1000 μg/mL led to a signifcant decrease in viral load with an inhibition rate of 99.5% and 99.7%, respectively. Conclusion: MgONPs can exert their inhibitory efects on HSV-1 in a dose-dependent manner, both directly and through interfering with the replication cycle of the virus.
## 1. Introduction
Herpes simplex virus Type 1 (HSV-1) belongs to the subfamily Alphaherpesvirinae of the Herpesviridae family with a relatively high prevalence of infection and is particularly prone to causing numerous disorders in humans (1). HSV-1 infection is primarily acquired orally in children and frequently appears as cold sores, but it can also cause more signifcant neurological, ocular, and mucocutaneous problems. Adults who were not infected orally as children can get genital HSV-1 infection. Sexual transmission of HSV-1 infection has become more common, especially in highincome nations, and it is currently the most common cause of frst-episode genital herpes in a number of these nations. Almost 4 billion people, or two-thirds of the world's population between the ages of 0 and 49, were infected with HSV-1 in 2020, primarily orally, with over 120 million new cases reported in this year [1,2].
Acyclovir, valacyclovir, and famciclovir are nucleoside analogs that target the viral DNA polymerase and are among the frst-line treatments for HSV infections [3]. Te gold standard for treating HSV-1 infections is still acyclovir, a guanosine analog with minimal toxicity and good selectivity [4]. Systemic treatment for HSV infections, such as labial and genital herpes, involves the use of acyclovir. For the treatment of HSV, other nucleoside analogs such as famciclovir and trifuridine are utilized in addition to acyclovir and its prodrug valacyclovir [5]. However, these medications have signifcant drawbacks, including resistance, insufcient suppression, low bioavailability, and short half-life [6,7]. Currently, no drug is available to remove a latent infection, and the extended therapeutic use of antivirals in immunocompromised individuals can contribute to the occurrence of treatment failure due to the emergence of antiviral-resistant virus strains [8].
Nanotherapeutics could revolutionize the development of antiviral medications and solve problems with strainspecifc targeting, resistance, new viruses, and incurable viral diseases [9]. It makes use of nanoparticles (NPs) between 1 and 100 nm in size as a way for drug delivery, infectious disease diagnostics, and therapy [9][10][11]. Tere are two types of therapeutic NPs: inorganic (such as metal NPs) and organic (such as polymeric, liposomes, micelles, and ferritin). For a range of medical disorders, both kinds of NPs have shown efcacy in preclinical research and clinical settings [12,13]. Since metals can "attack" multiple targets on viruses with little efect on the later development of resistance, the use of metal NPs as antiviral medicines has expanded quickly in recent years [9,14].
NPs have emerged as promising alternatives due to their unique physicochemical properties and ability to target bacteria through multiple mechanisms, reducing the likelihood of resistance development [15]. Nanomaterials, particularly metallic NPs functionalized with sulfonates or polyphenols such as tannic acid, ofer promising alternatives due to their ability to inhibit viral entry and replication through diverse mechanisms, such as physical blocking or disruption of the viral capsid [16]. NPs, such as silver, gold, zinc oxide, and carbon-based nanomaterials, exhibit unique physicochemical properties that enable them to inhibit HSV-1 at various stages of its life cycle, including viral attachment, entry, replication, and cell-to-cell spread. Furthermore, NPs serve as efective nanocarriers for ACV, enhancing its bioavailability, reducing side efects, and enabling targeted delivery [6].
Magnesium oxide NPs, or MgONPs, stand out among other metal oxide NPs because they are less harmful to the host and have a number of advantageous characteristics. MgONPs have several qualities, including low toxicity, biodegradability, biocompatibility, strong antimicrobial activity, biomedical applications, high chemical stability, electrical permeability, and photocatalytic activity. Tey are also inexpensive and easily accessible. Furthermore, MgO is currently regarded by the United States' Food and Drug Administration (FDA) as a substance that is safe for human ingestion [17].
Some research has been performed on MgONPs' antifungal and antibacterial qualities, and data are accessible in this area. Nevertheless, only one study so far examined the antiviral efects of MgONPs and evaluated their antiviral efcacy against foot and mouth disease virus (FMDV), which is another viral infection that afects cattle. As a result, there is insufcient evidence known about MgONPs' antiviral efects, and more research is needed in this feld. Te purpose of this work is to examine the in vitro activity of MgONPs against HSV-1 as a possible novel antiviral drug, given the signifcance and high prevalence of HSV-1 and the related health and economic issues they present worldwide.
## 2. Materials and Methods
## 2.1. Characterization of MgONPs.
US Research Nanomaterials Inc. provided highly purifed MgO nanopowders (> 99%). Te X-ray difraction (XRD) (PHILIPS PW1730) with Cu-Kα radiation (λkα �1.54 Å) was used to examine the crystal structure, chemical composition, and purity of the MgONPs. Using Fourier-transform infrared spectroscopy (FTIR), the surface functional group in MgONPs was assessed in the 450-4000 cm -1 range. Potassium bromide (KBr) and MgONP powders were mixed in a ratio of 1:19. Te samples underwent analysis using FTIR (360 Nicolet AVATAR spectrometer, Termo Scientifc, USA). To assess the size and shape of the NPs, a feld-emission scanning electron microscope (FESEM) (MIRA3TESCAN-XMU) was used. Moreover, energy-dispersive X-ray (EDX) spectroscopy was used to assess the sample's chemical purity. Moreover, the surface area of the MgONPs was examined using Brunauer-Emmett-Teller (BET) (BELsorp-MINI II, BEL, Japan) after degassing at 175 °C for 90 min in streaming nitrogen, followed by measuring the adsorption isotherm of nitrogen gas at a temperature of 77 K. Te optical absorption range of MgONPs was measured using an ultraviolet-visible (UV-vis) spectrophotometer (AVaSpec 2048 TEC) in the wavelength range of 250-800 nm. Te dynamic light scattering (DLS) method was used to evaluate the MgONPs' hydrodynamic measure and zeta potential at room temperature.
## 2.2. Cells and Viruses.
African green monkey kidney (Vero) cells were grown in high-glucose Dulbecco's Modifed Eagle's Medium (DMEM) (Gibco, Invitrogen, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (Gibco, Invitrogen, USA), 2 mM l-glutamine (Merck, Germany), 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma-Aldrich, USA) at 37 °C with 5% CO 2 . Te HSV-1 KOS strain was used to measure antiviral activity. After the virus was propagated using Vero cells, the Reed and Muench method was used to determine the propagated viral stock infectious titer as TCID 50 mL -1 . Following titration, the viral stock was aliquoted in sterile microtubes and kept at -70 °C for further use.
## 2
Advances in Virology 2.3. Determination of Cell Cytotoxicity. Te cytotoxicity of MgONPs was evaluated through the neutral red uptake assay, a method used to measure cell viability by quantifying the cytotoxic efects of MgONPs in vitro. Tis assay depends on the capacity of viable cells to incorporate and bind neutral red, a weakly cationic dye, within their lysosomes [18]. Consequently, cytotoxicity is indicated by a concentrationdependent decrease in neutral red uptake following exposure to MgONPs. A 96-well microtiter plate was seeded with 1.5 × 10 4 Vero cells per well. Te growth media were removed from the wells, and diferent concentrations of MgONPs (100-1000 μg/mL) were added to them after 24 h of incubation at 37 °C in a humid environment with 5% CO 2 .
Te plate was incubated for 48 h at 37 °C in a humidifed incubator with 5% CO 2 . Following the incubation period, the growth medium containing various MgONP concentrations was removed. After adding 100 μg/mL of neutral red to each well, the cells were incubated for 3 h. Te integrated dye was released from the cells following the addition of desorption solution (1% glacial acetic acid solution, 50% EtOH, and 49% H 2 O) to each well. A microplate reader (Hiperion MPR 4+, Roedermark, Germany) was used to measure the absorbance at a test wavelength of 550 nm. Data are presented as a percentage of cell viability relative to untreated cells (negative control, set at 100% viability).
## 2.4. Determination of Antiviral Activity
2.4.1. Virucidal Assay. In brief, 100 μL of HSV-1 suspension (100 TCID 50 /mL) was incubated with 100 μL of MgONPs in two nontoxic concentrations for 3 h at 37 °C in a humidifed 5% CO 2 atmosphere. In parallel, the same amount of the viral solution was incubated with cell culture media without MgONPs and was used as a virus control in this assay. A 96-well microtiter plate was seeded with 1.5 × 10 4 Vero cells per well. After 24 h of incubation at 37 °C in 5% CO 2 , Vero cell monolayers were then treated with the above mixtures and incubated for an additional hour at 37 °C. Following incubation, the supernatant was removed, and the cells were gently washed three times with PBS to eliminate any nonabsorbed viruses. Fresh DMEM media containing 2% FBS was then added, and the cells were incubated at 37 °C for 48 h, and the viral load of HSV-1 was calculated using the quantitative real-time PCR (qPCR) assay.
## 2.4.2. Cell Posttreatment Assay. Monolayers of Vero cells
were prepared in a 96-well plate and infected with 100 TCID 50 /mL of HSV-1 solution for one hour at 37 °C. Following the infection, the cells were rinsed with PBS to remove any noninternalized viruses. Diferent nontoxic concentrations of MgONPs were then added to the infected Vero cells, and the plate was further incubated for 48 h at 37 °C with 5% CO 2 . Cell and virus controls were included in the experiment using the same conditions, and the viral load of HSV-1 were calculated using the qPCR assay.
2.5. qPCR Analysis. qPCR was used to confrm the impact of MgONPs on HSV-1 infection of Vero cells. HSV viral DNA was extracted from the supernatants of virus-infected Vero cells taken at virucidal and cell posttreatment experiments using the BehPerp Viral Nucleic Acid Extraction Kit (BehGene Biotechnology, Iran) according to the manufacturer's guidelines. Te forward, reverse, and probe sequences targeting the UL30 gene of HSV-1 were 5′-ATCGGCGAG TACTGCATACA-3′,5′-GAGCTCCAGATGGGGCAA-3′, and 5′-HEX-ATTCCCTGCTGGTGGGCCA-BHQ1-3′, respectively, with an amplicon size of 75 bp. Te real-time PCR was performed in a fnal volume of 25 μL reaction including 12.5 μL of the RealQ Plus 2x Master Mix for Probe, without ROX ™ (Ampliqon, Denmark), 2 μL of forward primer (10 μM), 1 μL of reverse primer (10 μM), 1 μL of probe (10 μM), 5 μL of template DNA, and 3.5 μL of ddH 2 O. Te experiment was carried out using the Rotor-Gene Q instrument (QIAGEN, Germany) under the following conditions: 10 min for initial denaturation at 95 °C, followed by 40 cycles of 10 s at 95 °C and 30 s at 60 °C [19].
A reference standard was prepared by using the pUC57 vector containing the corresponding specifc viral sequence. Tenfold dilutions equivalent to 10 -1 -10 -10 concentrations of the synthesized plasmid were prepared to generate a calibration curve and to be run in parallel with the test samples. Te copy number of the plasmid was calculated using an online DNA copy number calculator (Technology Networks). Te limit of detection (LOD) of this real-time PCR was 3 copies/μL. Each run had a positive and a negative control, and all reactions were carried out in triplicate.
## 2.6. Statistical Analysis.
All statistical analyses were performed using the statistical software SPSS, Version 22.0 (IBM SPSS Statistics, Chicago, IL, USA). Te statistical signifcance of the data was assessed by conducting a oneway analysis of variance (ANOVA), followed by a Dunnett's post hoc test to compare multiple groups against a control group. All p values ≤ 0.05 were considered statistically signifcant. All graphs and charts were created using GraphPad Prism 8.
## 3. Results
## 3.1. Characterization of MgONPs.
Figure 1(a) presents the XRD pattern of the synthesized MgONPs. Comparison of the XRD pattern of MgONPs with the standard pattern reveals characteristic peaks of MgO at 36.9 °, 42.9 °, 62.2 °, 74.6 °, and 78.6 °, confrming successful synthesis. According to JCPDS no. 43-1022, the highest intensity peaks occur at 2θ values of 42.9 °and 62.2 °, with Miller indexes of 200 and 220, respectively, indicating the cubic structure of the singlephase MgO. Utilizing the Debye-Scherrer formula, the average crystallite size of the produced MgONPs was estimated, yielding a calculated particle size ranging from 40 to 65 nm. Tese fndings align with previous studies on MgONP synthesis [20].
## Advances in Virology
Figure 1(b) depicts the FTIR spectra of MgONPs. Tis analysis aimed to identify the diferent functional groups of MgONPs by detecting characteristic peaks within the wavelength range of 480-4000 cm -1 . Te bands observed at 3446 and 1637 cm -1 are associated with moisture content (O-H stretching bond vibration) present in either the analyzed sample or the precursor solution. As reported in various studies, it is widely recognized that NPs such as MgO possess a high specifc surface area due to their elevated surface-to-volume ratio. Consequently, when exposed to the atmosphere, these NPs can readily absorb H 2 O molecules [21][22][23]. In addition, the bond vibrations observed at approximately 1108 cm -1 correspond to O-H bending vibrations. Te absorption peak at 3701 cm -1 further indicates the presence of hydroxyl groups on the surface of MgO particles, representing a single OH species [24]. Moreover, the broad peak observed at lower frequencies near 520 cm -1 is characteristic of MgO vibrations, which is consistent with the fndings reported in other studies [25][26][27].
Te optical characteristics of MgONPs were assessed utilizing a UV-vis spectrophotometer. Figure 2(a) presents the UV-vis absorption spectroscopy of MgONPs, depicting absorbance as a function of wavelength within the range of 200-500 nm. Te presence of MgONPs was verifed by the distinct absorption peak observed at 280 nm. Te obtained UV-vis absorption spectra align with the expected absorption band range of MgONPs, falling within the nonvisible light range, which is consistent with previous literature fndings [26,28,29]. Te zeta potential range for the MgONP sample was -40 to +20 mV, with the peak observed at -12.1 mV, indicating good stability of the dispersed MgONPs in water (Figure 2(b)). However, the negative zeta potentials indicate that the surfaces of MgONPs carry a negative charge [29]. Te surface topography, morphology, and size distribution of MgONPs were examined using the SEM technique at various magnifcations (10 and 100 kx). Te FESEM images of MgO powder (Figures 3(a) and 3(b)) distinctly display the presence of crystalline NPs characterized by a polyhedral structure, with some evidence of agglomeration while maintaining uniformity and density. Te observed aggregation of these NPs may be attributed to the electrostatic attraction among the MgONPs. It is worth noting that the polyhedral particles exhibit well-defned facets, with a majority of them being hexagonal crystals.
Figure 3(c) reveals that the diameters of the polyhedra, calculated from the images, were approximately 59 nm, which is consistent with the fndings from the XRD analysis. In addition, Figure 3(d) demonstrates the qualitative and quantitative analyses of the elemental structure and purity of the MgONPs using EDX spectroscopy (EDX or EDS). Te corresponding EDS spectra display strong peaks, indicating the presence of magnesium (Mg) and oxygen (O) elements in the sample without any impurities. Te aforementioned analysis verifes the purity of the MgONPs. To gain further insights into the presence and distribution of elements in the sample, elemental mapping images of MgONPs were obtained. Figure 3(e) distinctly illustrates the presence of metallic magnesium atoms (blue color) and nonmetallic oxygen atoms (yellow color). Based on these images, it can be concluded that the presence of magnesium and oxygen atoms corresponds to the presence of MgONPs.
## 3.2. Cytotoxicity Assay.
Based on the results of the cytotoxicity test, no toxic efects were observed on Vero cells at all concentrations of MgONPs (up to a concentration of 1000 μg/mL). Incubation of HSV-1 with MgONPs at both concentrations, 900 and 1000 μg/mL, for three hours was associated with remarkable reductions in the formation of CPEs (Figure 4). Te results of qPCR assay also showed that three-hour incubation of HSV-1 with MgONPs at concentrations of 900 and 1000 μg/mL resulted in a decrease in the number of copies of HSV-1 genomic DNA with an inhibition rate of 93.6% (p value < 0.001) and 96.8% (p value < 0.001), respectively, compared to the virus control. Accordingly, MgONPs showed their antiviral activity in a dose-dependent manner (Figure 5).
## 3.3.2. Posttreatment Assay.
Following virus entry into the cells, the viral load was reduced in comparison to the virus control after 48 h of incubation with concentrations of 100, 300, and 1000 μg/mL of MgONPs. Based on the calculated viral loads and comparing them with the virus control, the percentage of viral inhibition by MgONPs at concentrations of 100, 300, and 1000 μg/mL was 23%, 99.5%, and 99.7%, respectively (p value < 0.001) (Table 1). Similar to the virucidal assay, MgONPs showed their antiviral activity in a dose-dependent manner in the posttreatment assay (Figure 6).
## 4. Discussion
To assess the antiviral efects of MgONPs against HSV-1 in this work, concentrations that maintained over 90% cell viability were utilized. As a result, all tested concentrations in the neutral red uptake assay were included in the antiviral tests since they were able to completely preserve cell viability. Terefore, the cytotoxicity test fndings revealed that MgONPs exhibited a very high level of biocompatibility with cells and had no detrimental impacts on living cells. Tus, they can be considered a highly safe nanomaterial.
Te direct efect of MgONPs on viral particles is evaluated in virucidal activity assessments, whilst the efect of MgONPs on various stages of viral replication is examined in posttreatment activity assessments. Our study found that the addition of MgONPs after cell infection considerably decreased the viral load of HSV-1, especially at the highest concentrations (300 and 1000 μg/mL), showing the largest antiviral efects. However, a lower concentration (100 μg/ mL) of MgONPs was associated with a lower reduction in HSV-1 viral load, indicating that the antiviral action of MgONPs is dose-dependent.
Numerous studies have been conducted to investigate the antibacterial characteristics and efcacy of MgONPs [30][31][32][33]. However, only one study has so far precisely studied the antiviral activities of MgONPs. Tat study looked at the antiviral activity of MgONPs against FMDV [34]. Te results of Rafei et al. showed that MgONPs could decrease viral activity by more than 90% in the early phases of infection, including attachment and penetration. However, after the virus entered the cells, no antiviral efects were seen. At 3 and 6 hours after attachment, they added the NPs to infected cells to evaluate their infuence on the viral replication cycle; however, no discernible change was seen at these phases of the FMDV infection cycle. Tese results imply that MgONPs prevent viruses from entering cells. Our research showed that treatment of HSV-1-infected cells with MgONPs, particularly at the higher doses, resulted in considerable reductions in the CPE and HSV-1 viral load compared to the virus control, which is in contradiction to the fndings of the study by Rafei et investigated the efects of these NPs on an enveloped virus. Te absence of a lipid envelope makes nonenveloped viruses, such as FMDV, generally more resistant to environmental stressors [35,36]. As opposed this, the membrane of enveloped viruses (such as HSV-1) is essential to their infectivity [37]. Enveloped viruses may be especially vulnerable to MgONPs because their lipid membrane is a target for oxidative stress. Te fndings of our investigation indicate that the viral load has decreased when the virus was incubated with MgONPs for 3 h outside of the cellular environment. Te results of this study demonstrate that MgONPs can directly afect virus particles. MgONPs and viral capsid proteins may directly interact to cause structural alterations that hinder viral attachment or entrance into host cells. Another portion of our study results has indicated that the addition of MgONPs after the entry of viruses into cells leads to inhibitory efects on the viral load. Tis result implies that MgONPs may employ other modes of action against HSV-1 in addition to direct impacts on the structure and viral particles.
MgONPs have been demonstrated to cause the production of intracellular reactive oxygen species (ROS) [38]. Te uncontrolled ROS formation is thought to be caused by the release of free Mg2+ ions from the NP [39]. Previous research has demonstrated that the antibacterial activities of MgONPs are related to the generation of ROS. Tese ROS lead to lipid peroxidation, protein denaturation, bacterial cell death, intracellular leakage, and oxidative DNA damage [17,32,38]. Given the infuence of MgONPs on bacterial cell membranes, it is reasonable to speculate about their efect on the structure of HSV-1. ROS and oxidative stress may cause damage to viral genetic material and proteins, making the virus noninfectious. MgONPs have the capacity to degrade the protein capsid structure or viral envelope of HSV-1, resulting in the disruption of its structure and a decrease in viral infectious titer. MgONPs may impair vital HSV-1 enzymes, hence impeding intracellular virus replication.
## 5. Conclusion
For the frst time, the antiviral activity of MgONPs against HSV-1 was examined in this study. Te data showed that MgONPs, especially at the highest concentrations, have no harmful impact on cell viability. In addition, there were notable inhibitory efects on HSV-1 from both direct exposure of the NPs to the virus and the addition of MgONPs following cell infection, especially at higher concentrations. Tese data indicate that MgONPs may be regarded as a promising and safe antiviral agent against several diseases caused by HSV-1. However, additional research on experimental models and laboratory animals is required to attain this goal.
## References
1. Su, Han, Shi (2024) "An Updated Review of HSV-1Infection-Associated Diseases and Treatment, Vaccine Development, and Vector Terapy Application" *Virulence*
2. Harfouche, Almukdad, Alareeki (2025) "Estimated Global and Regional Incidence and Prevalence of Herpes Simplex Virus Infections and Genital Ulcer Disease in 2020: Mathematical Modelling Analyses" *Sexually Transmitted Infections*
3. Birkmann, Zimmermann (2016) "HSV Antivirals-Current and Future Treatment Options" *Current Opinion in Virology*
4. Vere Hodge, Field (2013) "Antiviral Agents for Herpes simplex Virus" *Advances in Pharmacology*
5. Birkmann, Hewlett, Rübsamen-Schaef et al. (2011) "Helicase-Primase Inhibitors as the Potential Next Generation of Highly Active Drugs Against Herpes simplex Viruses" *Future Virology*
6. Ahmadi, Bolourchi, Rezaei et al. (2021) "Nanoparticles: A Promising Treatment for Herpes Simplex Virus Type-1 Infections" *Mediterranean Journal of Infection, Microbes and Antimicrobials*
7. Morone, Chianese, Dell'annunziata (2024) "Ligand-Free Silver Nanoparticles: An Innovative Strategy Against Viruses and Bacteria" *Microorganisms*
8. Whitley (2006) "New Approaches to the Terapy of HSV Infections" *Herpes: Te Journal of the IHMF*
9. Aderibigbe (2017) "Metal-Based Nanoparticles for the Treatment of Infectious Diseases" *Molecules*
10. Joob, Wiwanitkit (2017) "Nanotechnology for Health: A New Useful Technology in Medicine" *Medical Journal of Dr DY Patil Vidyapeeth*
11. Prasad, Lambe, Brar (2018) "Nanotherapeutics: An Insight Into Healthcare and Multi-Dimensional Applications in Medical Sector of the Modern World" *Biomedicine & Pharmacotherapy*
12. Nikalje (2015) "Nanotechnology and Its Applications in Medicine" *Medicinal Chemistry*
13. Anselmo, Mitragotri (2016) "Nanoparticles in the Clinic" *Bioengineering & translational medicine*
14. Galdiero, Falanga, Vitiello et al. (2011) "Silver Nanoparticles as Potential Antiviral Agents"
15. Morone, Dell'annunziata, Giugliano (2022) "Pulsed Laser Ablation of Magnetic Nanoparticles as a Novel Antibacterial Strategy Against Gram Positive Bacteria" *Applied Surface Science Advances*
16. Tomaszewska, Ranoszek-Soliwoda, Bednarczyk (2022) "Anti-HSV Activity of Metallic Nanoparticles Functionalized With Sulfonates vs. Polyphenols" *International Journal of Molecular Sciences*
17. Cai, Chen, Liu et al. (2018) "Magnesium Oxide Nanoparticles: Efective Agricultural Antibacterial Agent Against Ralstonia Solanacearum" *Frontiers in Microbiology*
18. Ates, Vanhaecke, Rogiers et al. (2017) "Assaying Cellular Viability Using the Neutral Red Uptake Assay" *Methods in Molecular Biology*
19. Hamidzade, Monavari, Kiani et al. (2024) "Enhanced Synergistic Antiviral Efects of Termally Expanded Graphite and Copper Oxide Nanosheets in the Form of a Novel Nanocomposite Against Herpes simplex Virus Type 1" *Microbial Pathogenesis*
20. Wang, Hu, Shao (2017) "Te Antimicrobial Activity of Nanoparticles: Present Situation and Prospects for the Future" *International Journal of Nanomedicine*
21. Almatar, Makky, Var et al. (2018) "Te Role of Nanoparticles in the Inhibition of Multidrug-Resistant Bacteria and Bioflms" *Current Drug Delivery*
22. Bassegoda, Ivanova, Ramon et al. (2018) "Strategies to Prevent the Occurrence of Resistance Against Antibiotics by Using Advanced Materials" *Applied Microbiology and Biotechnology*
23. Hemeg (2017) "Nanomaterials for Alternative Antibacterial Terapy" *International Journal of Nanomedicine*
24. Natan, Banin (2017) "From Nano to Micro: Using Nanotechnology to Combat Microorganisms and Teir Multidrug Resistance" *FEMS Microbiology Reviews*
25. Jeevanandam, San Chan, Danquah (2019) "Evaluating the Antibacterial Activity of Mgo Nanoparticles Synthesized From Aqueous Leaf Extract" *Med One*
26. Abdel-Aziz, Emam, Elsherbiny (2019) "Bioactivity of Magnesium Oxide Nanoparticles Synthesized From Cell Filtrate of Endobacterium Burkholderia rinojensis Against Fusarium oxysporum" *Materials Science and Engineering: C*
27. Khan, Shabir, Ahmad et al. (2020) "Biosynthesis and Antibacterial Activity of MgO-NPs Produced From Camellia-Sinensis Leaves Extract" *Materials Research Express*
28. Mccusker, Carvalho (2018) "Nano-Strategies to Fight Multidrug Resistant Bacteria: A Battle of the Titans" *Frontiers in Microbiology*
29. Ammulu, Vinay, Viswanath et al. (2021) "Phytoassisted Synthesis of Magnesium Oxide Nanoparticles from Pterocarpus Marsupium Rox. B Heartwood Extract and Its Biomedical Applications" *Journal of Genetic Engineering and Biotechnology*
30. Rodríguez-Hernández, Vega-Jiménez, Vázquez-Olmos et al. (2023) "Antibacterial Properties in Vitro of Magnesium Oxide Nanoparticles for Dental Applications" *Nanomaterials*
31. Nguyen, Grelling, Wetteland et al. (2018) "Antimicrobial Activities and Mechanisms of Magnesium Oxide Nanoparticles (Nmgo) Against Pathogenic Bacteria, Yeasts, and Bioflms" *Scientifc Reports*
32. Hirphaye, Bonka, Tura et al. (2023) "Biosynthesis of Magnesium Oxide Nanoparticles Using Hagenia Abyssinica Female Flower Aqueous Extract for Characterization and Antibacterial Activity" *Applied Water Science*
33. Rotti, Sunitha, Manjunath (2023) "Green Synthesis of Mgo Nanoparticles and Its Antibacterial Properties" *Frontiers of Chemistry*
34. Rafei, Rezatofghi, Ardakani et al. (2015) "In Vitro Anti-Foot-And-Mouth Disease Virus Activity of Magnesium Oxide Nanoparticles" *IET Nanobiotechnology*
35. Narula, Lokshman, Pathak et al. (2024) "Chemical Inactivation of Two Non-Enveloped Viruses Results in Distinct Termal Unfolding Patterns and Morphological Alterations" *BMC Microbiology*
36. Zupanc, Zevnik, Filipić (2023) "Inactivation of the Enveloped Virus phi6 With Hydrodynamic Cavitation" *Ultrasonics Sonochemistry*
37. Zeng, Li, Lv (2023) "Environmental Stability and Transmissibility of Enveloped Viruses at Varied Animate and Inanimate Interfaces" *Environmental Health*
38. Tabrez, Khan, Hoque et al. (2022) "Investigating the Anticancer Efcacy of Biogenic Synthesized Mgonps: An in Vitro Analysis" *Frontiers in Chemistry*
39. Bhattacharya, Dey, Neogi (2021) "An Insight into the Mechanism of Antibacterial Activity by Magnesium Oxide Nanoparticles" *Journal of Materials Chemistry B* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12510926&blobtype=pdf | # Bibliometric analysis of traditional Chinese medicine for viral infections through immune modulation (2015-2025)
Lei Zhang, Shuang Jin, Chuang Qin, Dabao Ma, Jinsheng Ye, Qingquan Liu, Jinzhi Duan, Jianjun Sun
## Abstract
Objectives: The immunomodulatory properties of traditional Chinese medicine (TCM) have attracted significant attention as a strategy for addressing viral infections. However, a comprehensive bibliometric analysis is still lacking. This study aims to systematically identify research trends, knowledge hotspots, and emerging themes in TCM applications for viral infections through immune modulation from 2015 to 2025. Methods: We collected publications from the Web of Science database from 2015 to 2025 and performed a comprehensive analysis using R, VOSviewer, and CiteSpace. In addition, clinical trial records published during this period were obtained from the PubMed database to assess clinical advancements in this field. Results: A total of 3,370 publications were analyzed in this study. Between 2015 and 2021, the number of publications in this field showed two distinct stepwise increases, separated by a period of relative stability, followed by a modest decline from 2021 to 2025. China contributed the highest volume of publications and demonstrated the broadest international collaborations, establishing itself as the leading country in this area. Frontiers in Immunology published the largest number of articles, while the Journal of Virology was the most frequently cited journal. Core topics included "Infection," "COVID-19," "Expression," "Antiviral," and "Protein." The primary research focus centered on TCM's antiviral effects and its modulation of immune responses, investigating its regulatory impact on inflammation and cytokine storms during viral infections, and examining TCM's role in modulating immune responses to viral vaccines. Clinical trials in this field focus on improving the management of viral infections, and immune reconstitution strategies for chronic infections.
## 1 Introduction
Viral infections pose a significant challenge to global public health, substantially contributing to worldwide morbidity and mortality. Notable viruses such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) (1), influenza viruses (2), hepatitis viruses (3)(4)(5), the human immunodeficiency virus (HIV) (6,7), and poliovirus (8), continue to threaten public health. Currently, about 270 viruses are known to infect humans (9). Worryingly, Viral infections are closely associated with increased mortality; for instance, the global SARS-CoV-2 pandemic caused an estimated 8.8 million deaths in 2021, ranking it as the second leading cause of death after ischemic heart disease. This outbreak also reduced global life expectancy at birth to 71.4 years and healthy life expectancy to 61.9 years, levels last seen in 2012 (1,10). The devastating impact of viral pandemics extends beyond SARS-CoV-2, with historical outbreaks consistently leading to severe consequences. In response, the international community has prioritized various viral diseases in health-related sustainable development goals, including acquired immune deficiency syndrome (AIDS), hepatitis, and poliomyelitis as core indicators for monitoring progress (10). The prevalence of these outbreaks not only escalates medical and financial challenges but also signifies a transition in the global disease burden towards infectious diseases (1,2,10).
Current strategies for combating viral infections fall into three main categories: targeted antiviral drugs, prophylactic vaccines, and supportive care. Despite significant advancements, major challenges persist: (1) the specificity, resistance, and timing issues of antiviral agents (11,12). ( 2) vaccines' limited efficacy against variant strains (13,14). (3) The absence of efficient therapeutic options for particular viruses, such as respiratory syncytial virus (15) and emerging pathogens. These limitations severely constrain clinical management. Encouragingly, high-quality clinical research increasingly supports the efficacy of traditional Chinese medicine (TCM) in treating viral infections (16)(17)(18).
Over a century ago, Canadian internist Sir William Osler observed that mortality from infections is primarily due to the host's response rather than the pathogen itself (19). Immune dysregulation subsequent to viral infection remains a key factor in severe complications (20). While the concept of "immunity" was officially defined in Chinese medical literature during the Ming Dynasty in the Mian Yi Lei Fang, the connection between infectious diseases and the immune response was already suggested in the pre-Qin period. The Yellow Emperor's Canon of Internal Medicine examines the transmission during epidemic outbreaks, whereas the Treatise on Cold Damage Diseases presents a systematic account of the progression of externally contracted diseases, both providing early insights into the relationship between immunity and infectious diseases.
The immune protection conferred by TCM against viral infections is primarily achieved through the bidirectional regulation of innate and adaptive immune responses (21). By modulating Tolllike receptors (TLRs), natural killer (NK) cells, and the functions of neutrophils and macrophages, the active components of TCM offer a robust strategy for addressing viral infections within innate immunity (22)(23)(24). Additionally, these Chinese herbal compounds are pivotal in adaptive immunity, promoting immune homeostasis and recovery by regulating T cells, B cells, and cytokines (22,24,25). Despite extensive research on the immunomodulatory effects of TCM in viral infections, analyses of research hotspots and emerging trends are lacking. This gap prevents new researchers from quickly grasping the field and impedes the advancement of TCM immunotherapy for viral infections.
Bibliometric analysis serves as a systematic tool to explore the evolution of academic fields (26). Utilizing extensive databases like Web of Science, it identifies key aspects of a field, including contributing countries, significant publications, research focuses, and collaborative networks (27). Such analyses establish a robust basis for comprehending the progression of knowledge development and anticipating forthcoming research domains (28). Despite growing interest in TCM immune modulation for viral infections, systematic bibliometric studies are scarce, hindering a comprehensive understanding of the field's evolution. This study analyzes literature from 2015 to 2025 to identify trends, key contributors, main research themes, and emerging frontiers. By addressing this gap, we aim to provide an integrated overview of the field, foster international collaboration, guide future research, and support evidence-based application of TCM immune modulation in antiviral therapy.
## 2 Materials and methods
## 2.1 Data collection
Due to its broad multidisciplinary coverage being included (with over 12,000 high-impact journals being covered) and its comprehensive metadata being made available for citation analysis and collaboration network construction, the Web of Science Core Collection (WoSCC) was chosen as the primary data source for this bibliometric analysis (29). Data were retrieved via Capital Medical University's institutional subscription on May 4, 2025. After removing irrelevant records, a total of 3,370 eligible publications were identified. In addition, relevant clinical trial data were obtained from the PubMed database, as such information is not available in WoSCC. The search strategies are detailed in Annex 1. Duplicate entries were removed, and the remaining articles were saved in plain text format with cited references exported as complete records. Results of clinical trials were exported in PubMed format.
## 2.2 Data analysis
This study utilized sophisticated data visualization and scientific knowledge mapping tools for the bibliometric analysis, specifically employing Origin 2018, R software (version 4.5.0, http:// www.bibliometrix.org) (30), VOSviewer (version 1.6.20) (31), and CiteSpace (version 6.4.R1) (32). National and institutional coauthorship networks, as well as source co-citation and keyword co-occurrence analyses, were visualized using VOSviewer. The specific parameters were as follows: (1) The national co-authorship network included countries with at least 5 publications; (2) The institutional co-authorship network included institutions with at least 13 publications; (3) Source cocitation analysis considered sources with a minimum of 189 citations; (4) Keyword co-occurrence analysis included keywords that appeared at least 23 times, with synonymous terms merged. The impact factors used in this study were obtained from Journal Citation Reports (JCR) for the year 2023.
## 3 Results
## 3.1 General landscapes of global publications
The dataset consisted of 3,370 publications sourced from WoSCC. As shown in Figure 1A, the annual publication count on TCM for viral infections through immune regulation displayed two major growth phases: an initial rise from 2015 to 2016, with an increase of 130 publications (representing 105.6% growth), followed by a more substantial surge between 2019 and 2021, during which output grew by 238 publications (94.4% increase over two years), peaking at 490 articles in 2021. These periods of expansion were separated by a plateau from 2017 to 2019, characterized by minimal annual variation in publication numbers (approximately ±1.6%). After 2021, a gradual decline in annual publications was observed. This trend is closely linked to major global events, including the 2015 Nobel Prize awarded to Professor Tu Youyou (33) and the outbreaks of Zika, Ebola (34,35), and COVID-19 (36)(37)(38), all of which heightened interest in TCM-mediated immune modulation. As the COVID-19 pandemic has stabilized, research activity and publication output in this area have shown a slight decline since 2021.
An analysis of corresponding authors' countries showed that China (n = 1008) was the primary contributor, followed by the USA (n = 499), India (n = 180), Iran (n = 136), and Korea (n = 101). Furthermore, as presented in Figure 1B and detailed in Table 1, 15.7% of publications from China and 28.1% from the USA involved multi-country collaborations (MCPs). Notably, China not only leads in publication volume but also maintains an extensive international collaboration network, as illustrated in 2). These findings indicate that researchers in China prioritize the exploration of the immunomodulatory effects of TCM concerning viral infections. This trend appears to be influenced by China's unique context, which encompasses a rich heritage in herbal medicine, strong governmental backing for TCM research, and proactive policies that encourage international scientific collaboration.
## 3.2 Journals and co-cited journals
In order to ascertain the journals exhibiting the greatest publication and citation impact within the domain of TCM pertaining to viral infections via immune modulation, we utilized the Bibliometrix package in R software (version 4.5.0). The visualizations were generated utilizing the ggplot2 package. Furthermore, a journal co-citation analysis was conducted utilizing VOSviewer (version 1.6.20).
The present study revealed a comprehensive collection of 3,370 documents distributed among 1126 academic journals (refer to Supplementary Material S2 for further details). As demonstrated in Table 3 and represented in Figure 3A, Frontiers in Immunology (n = 116, IF = 5.7) has emerged as the predominant publisher, succeeded by Fish & Shellfish Immunology (n = 66, IF = 4.1), PLoS One (n = 59, IF = 2.9), Viruses-Basel (n = 58, IF = 3.8), and Journal of Virology (n = 57, IF = 4.0). Table 4 and Figure 3B present an analysis of the most frequently cited journals, which include the Journal of Virology (n = 6997, IF = 4.0), PLoS One (n = 4002, IF = 2.9), PNAS (n = 3375, IF = 9.4), Nature (n = 2884, IF = 50.5), and the Journal of Immunology (n = 2855, IF = 3.6). It is noteworthy that the co-cited journals map presented in Figure 4 illustrates that the Journal of Virology, PLoS One, and Frontiers in Immunology serve as pivotal collaboration hubs. The collective findings highlight the significant contribution of the Journal of Virology to the domain of TCM in the treatment of viral infections via immune system modulation.
## 3.3 Citation burst
In order to conduct a comprehensive examination of the frontier areas and focal points within TCM concerning viral infections through immune modulation, we utilized CiteSpace to identify 127 references exhibiting significant citation bursts according to established criteria (top 25; status count: 2; minimum duration: 2). A selection of 25 references is illustrated in Figure 5. The complete compilation of these citations along with their respective DOIs can be found in Supplementary Material S3. It is noteworthy that the three references exhibiting the most significant citation bursts were: (1) "G3BP1 Is a Tunable Switch that Triggers Phase Separation to Assemble Stress Granules" (strength: 12.38); (2) "Antiviral innate immunity and stress granule responses" (strength: 11.51); (3) "Critical Role of an Antiviral Stress Granule Containing RIG-I and PKR in Viral Detection and Innate Immunity" (strength: 11.28). Additionally, the three latest emerging citation bursts have been identified as follows: (1) "Mechanisms of SARS-CoV-2 entry into cells"; (2) "SARS-CoV-2 N Protein Antagonizes Stress Granule Assembly and IFN Production by Interacting with G3BPs to Facilitate Viral Replication"; (3) "RNase L promotes the formation of unique ribonucleoprotein granules distinct from stress granules".
In summary, illustrating from the citation burst analysis, we have discerned three principal research focal points within the domain of TCM concerning viral infections through immune system modulation: (1) The relationship between stress granules and antiviral immune responses, emphasizing the interplay between cellular stress mechanisms and innate immunity; (2) The phenomenon of cytokine storm and immune modulation in viral infections, investigating the role of TCM formulations in regulating inflammatory responses during severe viral illnesses; (3) The mechanisms of antiviral compounds, analyzing the dual action of TCM preparations in direct viral inhibition and the regulation of the immune system.
## 3.4 Keyword clusters and evolution
Keyword cluster analysis serves as a valuable method for pinpointing research hotspots and developmental trends within scholarly disciplines. This research employed VOSviewer software to extract a total of 15,952 keywords from the existing literature. Table 5 presents a comprehensive analysis of keyword frequency, revealing that 15 terms have been recorded with over 150 occurrences. Notably, "Infection" leads the list with 424 instances, succeeded by "COVID-19" (n=389), "Expression" (n=306), "Antiviral" (n=210), "Protein" (n=204), "In-Vitro" (n=203), "Cells" (n=198), and "Activation" (n=195). Added to that, we pointed out 197 keywords that met a minimum frequency threshold of 23 occurrences, which were subsequently utilized to create a keyword cluster map (Figure 6). The map delineates five distinct clusters, each indicated by a unique color. Utilizing cluster analysis, we discerned five unique clusters: (1) Cluster 1 (red dots): This cluster centers on the epidemiological characteristics of viral infections and explores advances in immunization strategies, with particular reference to TCM adjuvant therapies and their influence on vaccine efficacy, antibody generation, and overall immune response. Key terms in this group include infection, vaccine, antibody, adjuvant, and efficacy. Beyond that, we developed a dynamic thematic progression chart applying the bibliometrix toolkit to discern the evolving trends in research (Figure 7). The evolution of the field exhibits a
## 3.5 Clinical progress analysis
A total of 8 clinical trials were retrieved from the PubMed database (Annex 5). These studies can be broadly categorized into two main research themes: (1) The application of TCM for treating viral infections through immune modulation; (2) TCM as an immune reconstitution strategy for patients with chronic viral infections.
## 4 Discussion
## 4.1 General information
In this study, we analyzed 3,370 publications from 2015 to 2025 using bibliometric and visual methods. The results reveal a growth pattern from 2015 to 2021, with a notable surge in 2016 when publications doubled to 253. Output then stabilized for three years, followed by a rapid increase, peaking at 490 articles in 2021 (238 more than in 2019). Post-2021, the number of publications gradually declined, although it remained above pre-2019 levels. Data for 2025 is incomplete due to the May 4 cutoff.
First of all, the award of the 2015 Nobel Prize in Physiology or Medicine to Professor Tu Youyou for her discovery of artemisinin was a pivotal event that elevated scholarly interest and policy support for TCM research (33). Professor Tu's work, influenced by historical Chinese medical texts such as the "Vade Mecum with Prescriptions for All Emergencies" by Old Immortal Ge, was recognized on the international stage, spurring increased funding and policy initiatives for TCM investigations. Additionally, the emergence of public health crises, such as the Zika and Ebola virus epidemics (34), heightened awareness of the potential role of TCM in viral prevention (35), leading to a significant increase in academic publications on the topic.
In the wake of the global COVID-19 pandemic that emerged in 2020 (36,37), TCM gained prominence in China's strategies for managing and mitigating the crisis (38,39). This highlighted the profile of TCM's role in modulating immune responses to viral infections, sparking growing global interest in its applications for immune regulation and viral prevention (40). Consequently, research institutions and journals increased support for TCMrelated studies, fueling rapid growth in the literature. However, as the COVID-19 pandemic came under control, research enthusiasm waned, leading to fewer large-scale studies and publications. With most initial findings already published, innovation slowed, and research shifted to synthesis, further reducing output. Additionally, diversified funding and priorities have redirected resources to other fields. It is worth noting that the apparent decline in publications observed in 2025 may be due to incomplete data, as the data collection for this study ended on May 4 of that year, providing a technical explanation for the observed downward trend.
China leads globally in scholarly publications on TCM related to viral infections and immune modulation, highlighting significant engagement by Chinese researchers. This focus is driven by TCM's local prominence and supportive national policies. Furthermore, by The 25 most cited references on traditional Chinese medicine for viral infection through immune modulation.
## 4.2 Hotspots and development trends
As described above, a thorough bibliometric analysis has revealed emerging research hotspots related to the application of TCM in viral infection through immune system modulation. The findings indicate that research frontiers and focal areas in this domain mainly concentrate on three themes. First, contemporary research has increasingly focused on delineating the antiviral actions of TCM, not only through direct inhibition of viral replication and entry but also through immunomodulation that encompasses both innate and adaptive immune responses, thereby enhancing host antiviral defenses. Second, substantial effort has been devoted to characterizing the capacity of TCM to regulate inflammation and mitigate cytokine storm during viral infections, highlighting its significance in curbing excessive immune responses and reducing immunopathological consequences. Finally, accumulating evidence indicates that TCM can modulate vaccineinduced immune responses and may function as an adjuvant to enhance vaccine immunogenicity and effectiveness. Collectively, existing studies provide supportive evidence for these applications across therapeutic and preventive contexts.
Expanding upon the fundamental concept that TCM plays a role in antiviral defense through both direct and indirect mechanisms, we initially outline its direct antiviral effects. TCM have demonstrated notable direct antiviral effects through multiple mechanisms targeting different stages of the viral life cycle. Initially, numerous components of TCM impede viral entry by obstructing attachment to host-cell receptors or by inhibiting membrane fusion. For instance, glycyrrhizin and glycyrrhizic acid, which are extracted from licorice, demonstrate a significant ability to obstruct viral adsorption and penetration (41,42). In a similar vein, the active constituents of Artemisia vulgaris and perilla leaf extracts possess the ability to directly inactivate viral particles or disrupt their interactions with host receptors (43,44). Secondly, for viruses that have effectively infiltrated cells, constituents of TCM can function by inhibiting viral replication, focusing on essential viral enzymes (such as polymerases) and the production of viral nucleic acids and proteins. The utilization of glycyrrhizic acid-based carbon dots, Lianhua Qingwen Capsule, Huashi Baidu Decoction, baicalin, and associated natural flavonoid derivatives has been shown to diminish viral load and efficiently inhibit replication through various mechanisms (41,(45)(46)(47). Additionally, compounds like emodin and dandelion extracts have been shown to downregulate the expression of essential viral genes during postentry stages, consequently hindering protein synthesis and viral proliferation (48,49). Furthermore, a significant aspect of direct action involves direct virucidal activity, in which the physical disruption of viral structures results in their inactivation. For example, various herbal extracts obtained from forsythia and honeysuckle have demonstrated the ability to compromise the viral envelope, significantly diminishing infectivity (47). In conclusion, the findings suggest that TCM alleviates viral infection via several direct mechanisms, such as obstructing viral entry, inhibiting replication, compromising structural integrity, and interfering with viral release.
Beyond direct mechanisms, TCM also significantly contributes to the enhancement of the host immune response, which serves as a vital indirect approach in the fight against viral infections. Through the modulation of both innate and adaptive immune mechanisms, TCM enhances the overall immune defense, thereby facilitating more effective viral clearance and bolstering host resistance (21). The innate immune system, the first defense against pathogens, includes macrophages and NK cells. TCM activates pattern recognition receptors to bolster innate immunity (50). Xuanfei Baidu Decoction modulates the PD-1/IL17A pathway, reducing pro-inflammatory cytokine secretion and neutrophil and macrophage infiltration (51,52). Lianhua Qingwen promotes M2 macrophage infiltration, decreases M1 macrophage markers, and alleviates inflammation in Raw264.7 macrophages, significantly enhancing their phagocytic capacity (53,54). In H1N1-infected mice, San Wu Huangqin decoction boosted NK cell activity, accelerated the phagocytic function of macrophages (55). Overactive innate immune responses, such as excessive macrophage infiltration post-RSV infection, worsen outcomes like pneumonia and asthma; Xuanfei Fang formula can counteract these effects (56). It is essential to emphasize that research has demonstrated that TRIM29 and PARP9 serve as significant regulators of antiviral immunity. The absence of TRIM29 confers protection against fatal infections caused by influenza virus and other DNA/RNA viruses by augmenting the antiviral innate immune response in alveolar macrophages (57), dendritic cells (58,59), and intestinal epithelial cells (60). Furthermore, PARP9 functions as a noncanonical sensor for RNA viruses, playing a significant role in the defense against RNA virus infections (61). The incorporation of these mechanisms will deepen the comprehension of TCM-mediated immunomodulation within the framework of viral infection (62).
Furthermore, the adaptive immune system, primarily composed of T and B lymphocytes, is crucial for antigen-specific responses and immunological memory. Astragaloside, the main active component of Astragalus membranaceus, exerts immunoregulatory effects by promoting T-cell activation, balancing effector and regulatory T cells, enhancing CD45 phosphatase activity, and inhibiting pro- inflammatory cytokine production and the NF-kB pathway (63). Beyond astragalus saponin, numerous other Chinese medicines also exhibit immunoregulatory properties. Zhiyi Xie et al. reviewed how various Chinese herbal polysaccharides, such as those from Atractylodes macrocephala Koidz. and Cordyceps sinensis, bolster immunity by influencing both adaptive and innate responses (64). A systematic review of randomized controlled trials indicates that TCM modulates adaptive immune responses in post-viral fatigue, potentially increasing CD4 T lymphocyte proportions and reducing serum IL-6 levels (65). In addition to regulating humoral immunity, TCM can promote antibody production and enhance immune responses when used as a vaccine adjuvant (66,67). Moreover, It also helps regulate excessive immune responses and lowers proinflammatory factors like IL-6, IL-8, and IL-17A, mitigating systemic inflammation and cytokine storms (56).
Another critical consideration is that immune dysregulation following viral infection is a major driver of severe complications and mortality (20). As the regulation of immune responses is crucial for eliminating pathogens while minimizing immunopathological harm, the targeted modulation of excessive inflammation represents a key therapeutic focus in which TCM may confer substantial benefit. Uncontrolled inflammatory responses, characterized by excessive release of pro-inflammatory cytokines like IL-6 and TNF-a, are primarily driven by persistent activation of signaling pathways such as NF-kB (50). In addition, there is a growing concern regarding TRIM29, which has recently been reported to enhance PERK-mediated endoplasmic reticulum stress immune responses, thereby promoting the production of proinflammatory cytokines (68,69). These responses are critical in the progression of viral diseases, including influenza and coronaviruses, leading to tissue damage, acute lung injury, multi-organ failure, and increased mortality (20, 21 50). TCM offers notable benefits in managing hyperinflammatory conditions during various stages of viral infections. Isoliquiritigenin, a flavonoid compound, inhibits viral replication and reduces virus-induced inflammation by activating the NRF2 signaling pathway, thereby decreasing oxidative stress and the inflammatory cascade (70). The findings suggest that specific TCM monomers exert targeted regulatory effects on inflammation associated with viral infections. Classical TCM compounds similarly demonstrate notable antiviral and antiinflammatory properties. Sangju Cold Granule significantly inhibits influenza A virus replication and its related inflammatory response in both in vitro and in vivo settings. This inhibition is primarily mediated through the suppression of the RIG-I/NF-kB/ IFN (I/III) signaling pathway, which is essential for preventing the amplification of cytokine storms (71). Additionally, Qingjin Huatan Decoction decreases IL-6 and TNF-a expression in lung tissues of experimental models, thereby mitigating local inflammatory damage and protecting pulmonary function (72). XiaoEr LianHuaQingGan and Xijiao Dihuang Decoction, widely used for viral infections, significantly reduce fever duration and improve clinical symptoms by suppressing inflammatory mediator production, regulating immune responses, and enhancing mitophagy (73,74). Chaihu Guizhi Decoction has demonstrated efficacy in diminishing systemic inflammation and modulating immune function in viral diseases (75). Additionally, Liang-Ge-San mitigates lung inflammation by modulating key cytokine levels and reducing neutrophil infiltration in inflamed tissues (76). These findings demonstrate that TCM modulates inflammation and cytokine storms in viral infections through various mechanisms. The regulatory effects on multiple targets and pathways highlight TCM's unique advantages in preventing and managing virusinduced hyperinflammation and cytokine storms.
Building upon its established capacity to modulate both innate and adaptive immune responses, TCM presents a promising strategy as an immunoadjuvant to augment vaccine-induced immunity and enhance protective efficacy. Vaccines represent a cornerstone of preventive medicine, with their efficacy in mitigating viral diseases, such as polio and hepatitis B, well-documented (77). Adjuvants play a crucial role in shaping the nature and magnitude of the immune response elicited by vaccines (78,79). In recent years, the application of TCM as vaccine adjuvants has gained increasing attention (67,80,81). Polysaccharides derived from TCM plants, including ginseng (82), astragalus (83,84), Ganoderma lucidum (85,86), Codonopsis pilosula (87), Rehmannia glutinosa (88), lentinan (89), longan (90), Radix Cyathulae Officinalis (90), and Angelica sinensis (90), have demonstrated significant potential as vaccine adjuvants. The immunomodulatory properties, low toxicity, and favorable safety profiles of these polysaccharides underlie their potential therapeutic applications. They can enhance both humoral and cellular immune responses by activating key immune cells, such as macrophages, T cells, B cells, and NK cells, as well as by regulating cytokine and antibody production (67,80,81,90). Furthermore, these polysaccharides can modulate the intestinal microbiome, promote dendritic cell maturation, and enhance antigen presentation, thereby strengthening both Th1 and Th2 immune pathways (91). Their mechanisms of action also involve interactions with pathogen recognition receptors, including TLRs and NOD-like receptors, which initiate intracellular signaling cascades and immune activation (92,93). Comprehensive studies have demonstrated the efficacy of these polysaccharides in enhancing vaccine immunogenicity against viral diseases, with minimal adverse effects (83,84).
In addition to the extensive research on polysaccharides as adjuvants, various active compounds from TCM show significant promise for the development of new vaccine adjuvants. Active components such as saponins (e.g., ginsenosides) (66), flavonoids (e.g., Epimedium flavonoids) (94), tannins and organic acids (e.g., tannic acid) (95), and so on, have been demonstrated to modulate host immune responses, enhance vaccine immunogenicity, and exhibit promising application prospects (67,96). The multitargeted and immune-balancing properties of TCM adjuvants are anticipated to play a crucial role in enhancing vaccine efficacy, broadening immune coverage, and bolstering public health protection in the future.
Taken together, available evidence suggests that TCM may function as a multifaceted antiviral modality across therapeutic and preventive domains. By concurrently engaging antiviral, antiinflammatory, and memory-enhancing mechanisms, it offers an integrated approach consonant with the complexity of viral infectious diseases.
## 4.3 Clinical progress
A review of the PubMed database revealed 8 clinical trials examining the use of TCM in modulating immune responses to viral infections. Analysis of these studies highlights key trends and focal areas: (1) The immunomodulatory properties of TCM and natural products in managing viral infections. Clinical evidence suggests that TCM and natural products are increasingly integral to viral infection management, as they modulate inflammatory cytokines, activate immune cell subsets, and facilitate immune recovery. For instance, in a murine model of HCoV-229E infection, Shufeng Jiedu capsules demonstrated significant antiviral and immunoregulatory effects, including reduction of pro-inflammatory cytokines (IL-6, TNF-a, IFN-g) and elevation of CD4 + and CD8 + T cell counts (25). Notably, a follow-up clinical real-world study presented in the same publication indicated that Shufeng Jiedu capsules, when administered in conjunction with standard antiviral therapies, facilitated a more rapid resolution of symptoms (such as fatigue and cough) in patients experiencing moderate COVID-19 (25). Additionally, agents such as Qiliqiangxin have demonstrated efficacy in modulating the balance of pro-and anti-inflammatory cytokines (e.g., reducing IFN-g, IL-17, TNF-a, IL-4 and increasing IL-10) and enhancing cardiac function in viral cardiomyopathy (97). Extracts from Perilla and Portulaca oleracea have been shown to augment NK cell activity and Th1 cytokine production (IL-12, IFN-g) in healthy individuals (98). (2) Recent advancements in immune reconstitution for chronic viral infections have highlighted innovative strategies to restore immune function. Chinese herbal formulations, including Mianyi granules and the Wenshen Jianpi recipe, have been reported to increase CD4 + T cell counts and improve NK cell subpopulations in patients unresponsive to antiretroviral therapy (99,100). TCM has also been effective in reducing HBV DNA levels and enhancing HBeAg clearance and seroconversion rates in HBeAg-positive chronic hepatitis B patients with normal alanine aminotransferase levels, likely through modulation of the host immune response (101).
## 4.4 Future directions
Despite the demonstrated antiviral, anti-inflammatory, and immunomodulatory effects of various TCM compounds and formulations, their molecular mechanisms remain largely undefined. Current investigations provide only a cursory examination of the interactions between TCM components and viral or host targets, as well as the detailed pathways involved in immune regulation and adjuvant activity. Future research must systematically delineate specific molecular targets and signaling pathways, and explore the relationship between antiviral effects and immune modulation. Advanced structural biology, immunophenotyping, and multi-omics technologies, including transcriptomics, metabolomics, proteomics, and metagenomics, provide robust methodologies to systematically tackle these deficiencies. These methodologies can accurately delineate interaction networks between TCM and host/virus, characterize dynamic immune responses. Moreover, it is essential to prioritize rational integration strategies, such as the combination of TCM with direct-acting antivirals to shorten treatment duration or with vaccines to bolster mucosal immunity, in order to enhance clinical applicability. By employing mechanistic elucidation and evidencebased combination design, TCM can be strategically integrated into contemporary antiviral frameworks, enhancing its contribution to personalized infection management and public health initiatives.
## 4.5 Limitations
This study utilized data from the WoSCC database to offer a comprehensive overview of the research landscape, highlighting major themes, focal points, and emerging trends. This method enhances our comprehension of the field and aids in identifying forthcoming research priorities. Nonetheless, several limitations should be acknowledged. Firstly, relying solely on the WoSCC database may have omitted relevant literature, despite its widespread recognition for quality and suitability for bibliometric analysis. Secondly, the analysis was confined to English-language articles, potentially introducing language bias; however, given English's predominant position in academia, this limitation is deemed justifiable. Notwithstanding these constraints, the study's findings are robust and provide valuable insights for guiding future research in this area.
## 5 Conclusion
This study offers the initial comprehensive analysis of worldwide research patterns in utilizing TCM for viral infections through immune modulation in the last decade.
## References
1. Koelle, Martin, Antia et al. (2022) "The changing epidemiology of SARS-CoV-2" *Science*
2. Ryu, Cowling (2021) "Human influenza epidemiology. Cold Spring Harb Perspect Med"
3. Hsu, Huang, Nguyen (2023) "Global burden of hepatitis B virus: current status, missed opportunities and a call for action" *Nat Rev Gastroenterol Hepatol*
4. Stockdale, Kreuels, Henrion et al. (2020) "The global prevalence of hepatitis D virus infection: Systematic review and metaanalysis" *J Hepatol*
5. Alberts, Clifford, Georges et al. (2022) "Worldwide prevalence of hepatitis B virus and hepatitis C virus among patients with cirrhosis at country, region, and global levels: a systematic review" *Lancet Gastroenterol Hepatol*
6. Hemelaar, Elangovan, Yun et al. (2020) "Global and regional epidemiology of HIV-1 recombinants in 1990-2015: a systematic review and global survey" *Lancet HIV*
7. Haeuser, Serfes, Cork et al. "Mapping age-and sex-specific HIV prevalence in adults in sub-Saharan Africa"
8. (2022) *BMC Med*
9. Thompson, Kalkowska (2020) "Review of poliovirus modeling performed from 2000 to 2019 to support global polio eradication" *Expert Rev Vaccines*
10. Forni, Cagliani, Clerici et al. (2022) "Disease-causing human viruses: novelty and legacy" *Trends Microbiol*
11. (2024) *World Health Organization. World health statistics*
12. Jeng, Liaw (2021) "Finite antiviral therapy in chronic hepatitis B patients with cirrhosis"
13. Batool, Chokkakula, Song (2023) "Influenza treatment: limitations of antiviral therapy and advantages of drug combination therapy" *Microorganisms*
14. Doi, Kanto, Hosseinipour et al. (2021) "Phase 1 human immunodeficiency virus (HIV) vaccine trial to evaluate the safety and immunogenicity of HIV subtype C DNA and MF59-adjuvanted subtype C envelope protein" *Clin Infect Dis*
15. Zhang, Zhang, Hua et al. (2024) "Expert consensus on the diagnosis, treatment, and prevention of respiratory syncytial virus infections in children" *World J Pediatr*
16. Zou, Chang, Fan et al. (2024) "Effectiveness and safety of Sanhan Huashi granules versus nirmatrelvir-ritonavir in adult patients with COVID-19: A randomized, open-label, multicenter trial"
17. Wang, Cao, Liu et al. (2011) "Oseltamivir compared with the Chinese traditional therapy maxingshigan-yinqiaosan in the treatment of H1N1 influenza: a randomized trial" *Ann Intern Med*
18. Liu (2021) "Treating older patients with AIDS using Traditional Chinese Medicine Combined with Conventional Western Medicine in China" *Aging Dis*
19. Osler (1909) "The principles and practice of medicine"
20. Aschenbrenner, Bauer, Bock et al. (2024) "The pathophysiology of sepsis and precision-medicine-based immunotherapy" *Nat Immunol*
21. Bai, Liu, Zhang et al. (2025) "Traditional chinese medicine for viral pneumonia therapy: pharmacological basis and mechanistic insights" *Int J Biol Sci*
22. Shi, Xu, Xiao et al. (2020) "Jinhua Qinggan granules attenuates acute lung injury by promotion of neutrophil apoptosis and inhibition of TLR4/MyD88/NF-kappaB pathway" *BioMed Pharmacother*
23. Shi, Ge, Yan et al. (2021) "Shufeng Jiedu, a promising herbal therapy for moderate COVID-19:Antiviral and anti-inflammatory properties, pathways of bioactive compounds, and a clinical real-world pragmatic study" *Exploratory Res Hypothesis Med*
24. Wang, Lu, Song et al. (2000) "From past to future: Bibliometric analysis of global research productivity on nomogram"
25. Kokol, Blazun, Zavrsnik (2021) "Application of bibliometrics in medicine: a historical bibliometrics analysis" *Health Info Libr J*
26. Ullah, Asghar, Griffiths (2022) "An integrated methodology for bibliometric analysis: A case study of internet of things in healthcare applications" *Sensors (Basel)*
27. Alryalat, Malkawi, Momani (2019) "Comparing bibliometric analysis using pubMed, scopus, and web of science databases" *J Vis Exp*
28. Aria, Cuccurullo, Bibliometrix ; Van Eck et al. (2010) "Software survey: VOSviewer, a computer program for bibliometric mapping" *Scientometrics*
29. Chen, Citespace (2006) "Detecting and visualizing emerging trends and transient patterns in scientific literature" *J Am Soc Inf Sci Technol*
30. Wang, Yu (2022) "Light people: elizabeth rogan" *Light Sci Appl*
31. Baker, Mahmud, Miller et al. (2022) "Infectious disease in an era of global change" *Nat Rev Microbiol*
32. Lai, Hung, Hong et al. (2024) "Vitro Antiviral Activity of Rhodiola crenulata Extract against Zika Virus and Japanese Encephalitis Virus: Viral Binding and Stability. Pharm (Basel)"
33. Xia, Yao, Zhao et al. (2022) "Publication trends of research on COVID-19 and host immune response: A bibliometric analysis"
34. Baloch, Baloch, Zheng et al. (2020) "The coronavirus disease 2019 (COVID-19) pandemic" *Tohoku J Exp Med*
35. (2020) *Pharmacol Res*
36. Lyu, Xiao, Wang et al. (2019) "A bibliometric analysis of research related Chinese Medicine in the prevention and treatment of corona virus disease" *Acta Pharm Sin B*
37. Tong, Hu, Zhou et al. (2020) "Glycyrrhizic-acid-based carbon dots with high antiviral activity by multisite inhibition mechanisms" *Small*
38. Wang, Yang, Yuan et al. (2015) "The antiviral and antimicrobial activities of licorice, a widely-used Chinese herb" *Acta Pharm Sin B*
39. Xiao, Liu, Hu et al. (2023) "Antiviral activities of Artemisia vulgaris L. extract against herpes simplex virus" *Chin Med-UK*
40. Tang, Tsai, Chang et al. (2021) "Perilla (Perilla frutescens) leaf extract inhibits SARS-CoV-2 via direct virus inactivation" *BioMed J*
41. Fan, Wang, Song et al. (2023) "Baicalin inhibits the replication of the hepatitis B virus by targeting TRIM25" *J Tradit Complementary Med*
42. Xu, Li, Liu et al. (2020) "Bioactive compounds from Huashi Baidu decoction possess both antiviral and anti-inflammatory effects against COVID-19" *Proc Natl Acad Sci*
43. Cai, Shao, Wang et al. (2023) "Antiviral activity of dandelion aqueous extract against pseudorabies virus both in vitro and in vivo" *Front Veterinary Sci*
44. Cai, Wang, Li et al. (2023) "Emodin as an inhibitor of PRV infection in vitro and in vivo" *Molecules*
45. Zhang, Ye, Liu et al. (2023) "Research progress on the effect of traditional Chinese medicine on the activation of PRRs-mediated NF-kappaB signaling pathway to inhibit influenza pneumonia" *Front Pharmacol*
46. Zhao, Liu, Wang et al. (2023) "Multimodal identification by transcriptomics and multiscale bioassays of active components in xuanfeibaidu formula to suppress macrophage-mediated immune response"
47. Wang, Wang, Li et al. (2022) "Xuanfei Baidu Decoction reduces acute lung injury by regulating infiltration of neutrophils and macrophages via PD-1/IL17A pathway" *Pharmacol Res*
48. Li, Feng, Zhang et al. (2024) "Lianhua Qingwen protects LPS-induced acute lung injury by promoting M2 macrophage infiltration" *J Ethnopharmacol*
49. Shen, Yin (2021) "The mechanisms and clinical application of Traditional Chinese Medicine Lianhua-Qingwen capsule" *BioMed Pharmacother*
50. Ma, Ren, Luo (2021) "San Wu Huangqin decoction regulates inflammation and immune dysfunction induced by influenza virus by regulating the NF-kappaB signaling pathway in H1N1-infected mice" *J Ethnopharmacol*
51. Qin, Luo, Zhao et al. (2024) "Xuanfei Formula inhibited RSV infection by normalizing the SREBP2-mediated cholesterol synthesis process" *Front Microbiol*
52. Xing, Weng, Yuan et al. (2016) "Identification of a role for TRIM29 in the control of innate immunity in the respiratory tract" *Nat Immunol*
53. Xing, Zhang, Minze et al. (2018) "TRIM29 negatively regulates the type I IFN production in response to RNA virus" *J Immunol*
54. Xing, Zhang, Zhang et al. (2017) "TRIM29 promotes DNA virus infections by inhibiting innate immune response" *Nat Commun*
55. Wang, Wang, Lu et al. (2025) "TRIM29 controls enteric RNA virus-induced intestinal inflammation by targeting NLRP6 and NLRP9b signaling pathways" *Mucosal Immunol*
56. Xing, Zhang, Du et al. (2021) "Identification of poly (ADP-ribose) polymerase 9 (PARP9) as a noncanonical sensor for RNA virus in dendritic cells" *Nat Commun*
57. Li, Yi, Zhu et al. (2025) "Immunomodulatory actions of tonifying polysaccharides: pharmacological effects, mechanisms and therapeutic applications" *Front Immunol*
58. Qi, Gao, Hou et al. (2017) "Anti-inflammatory and immunostimulatory activities of astragalosides" *Am J Chin Med*
59. Xie, Jiang, Lin et al. (2023) "The mechanisms of polysaccharides from tonic chinese herbal medicine on the enhancement immune function: A review" *Molecules*
60. Hu, Cai, Li et al. (2024) "Chinese herbal medicine for post-viral fatigue: A systematic review of randomized controlled trials" *PloS One*
61. Silva, Kato, Melo et al. (2021) "A particulate saponin/TLR agonist vaccine adjuvant alters lymph flow and modulates adaptive immunity" *Sci Immunol*
62. Zhao, Chen, Wang et al. (2023) "Improvement influenza vaccine immune responses with traditional Chinese medicine and its active ingredients" *Front Microbiol*
63. Wang, Lu, Zhang et al. (2024) "Loss of TRIM29 mitigates viral myocarditis by attenuating PERK-driven ER stress response in male mice" *Nat Commun*
64. Zhuo, He, Wang et al. (2023) "Pedunculoside targets P2X7R to protect against myocarditis by regulating the NLRP3/PIP2/MAPK signaling pathway" *Front Pharmacol*
65. Gao, Liu, Huang et al. (2024) "Sangju Cold Granule exerts anti-viral and anti-inflammatory activities against influenza A virus in vitro and in vivo" *J Ethnopharmacol*
66. Liu, Zhao, Xu et al. (2023) "Qingjin Huatan decoction protects mice against influenza a virus pneumonia via the chemokine signaling pathways" *J Ethnopharmacol*
67. Li, Li, Zhao et al. (2022) "XiaoEr LianHuaQinqGan alleviates viral pneumonia in mice infected by influenza A and respiratory syncytial viruses" *Pharm Biol*
68. Deng, Zhao, Liu et al. (2024) "Xijiao Dihuang decoction combined with Yinqiao powder promotes autophagy-dependent ROS decrease to inhibit ROS/NLRP3/pyroptosis regulation axis in influenza virus infection" *Phytomedicine*
69. Feng, Xu, Fu et al. (2025) "An integrative pharmacology-based study on the efficacy and mechanism of essential oil of Chaihu Guizhi Decoction on influenza A virus induced pneumonia in mice" *J Ethnopharmacol*
70. Yang, Zhou, Liu et al. (2025) "Liang-Ge-San attenuates virusinduced acute lung injury by targeting FXR-mediated ACE2 downregulation to modulate the formation of the cytokine storm" *Phytomedicine*
71. Pollard, Bijker (2021) "A guide to vaccinology: from basic principles to new developments" *Nat Rev Immunol*
72. Pulendran, Hagan (2013) "Emerging concepts in the science of vaccine adjuvants" *Nat Rev Drug Discov*
73. Hioki, Hayashi, Natsume-Kitatani et al. (2011) "Structure-function relationship of the saponins from the roots of Platycodon grandiflorum for hemolytic and adjuvant activity" *Int Immunopharmacol*
74. Feng, Han, Han et al. (2024) "Structural analysis of a soluble polysaccharide GSPA-0.3 from the root of Panax ginseng C. A. Meyer and its adjuvant activity with mechanism investigation" *Carbohydr Polym*
75. Du, Wan, Huang et al. (2022) "A critical review of Astragalus polysaccharides: From therapeutic mechanisms to pharmaceutics" *BioMed Pharmacother*
76. Shan, Sun, Dalloul et al. (2019) "Effect of the oral administration of astragalus polysaccharides on jejunum mucosal immunity in chickens vaccinated against Newcastle disease" *Microb Pathog*
77. Shen, Xu, Liu et al. (2024) "Synthesis, characterization and MAFLD prevention potential of Ganoderma lucidum spore polysaccharide-stabilized selenium nanoparticles" *Int J Biol Macromol*
78. Jiang, Chang, Liu et al. (2017) "Overview of Ganoderma sinense polysaccharide-an adjunctive drug used during concurrent Chemo/Radiation therapy for cancer treatment in China" *BioMed Pharmacother*
79. Wan, Zhang, Luo et al. (2020) "Codonopsis pilosula polysaccharide attenuates Abeta toxicity and cognitive defects in APP/PS1 mice" *Aging (Albany NY*
80. Wu, Peng, Liu et al. (2024) "The structures of two polysaccharides from Lepidium meyenii and their immunomodulatory effects via activating NF-kappaB signaling pathway" *Int J Biol Macromol*
81. Li, Guo, Dong et al. (2024) "Polysaccharides from Chinese herbal medicine: a review on the hepatoprotective and molecular mechanism" *Chin J Nat Med*
82. Wang, Kwak, Lee et al. (2018) "Rehmannia glutinosa polysaccharide promoted activation of human dendritic cells" *Int J Biol Macromol*
83. Feng, Ding, Liu et al. (2021) "Astragalus polysaccharide enhances the immune function of RAW264.7 macrophages via the NF-kappaB p65/MAPK signaling pathway" *Exp Ther Med*
84. Nie, Cai, Shao et al. (2018) "Toll-like receptors, associated biological roles, and signaling networks in non-mammals" *Front Immunol*
85. Wu, Ou, Yao et al. (2025) "The immunostimulatory activity of Epimedium flavonoids involves toll-like receptor 7/8" *Front Pharmacol*
86. Cabral-Hipolito, Molina-Ramirez, Castillo-Maldonado et al. (2022) "Tannic acid exhibits adjuvant activity by enhancing humoral and cell-mediated immunity against BSA as a protein antigen" *Protein Pept Lett*
87. Nazeam, Singab (2022) "Immunostimulant plant proteins: Potential candidates as vaccine adjuvants" *Phytother Res*
88. Liao, Yuan, Jin et al. (2024) "Qiliqiangxin alleviates imbalance of inflammatory cytokines in patients with dilated cardiomyopathy: A randomized controlled trial" *Curr Med Sci*
89. Jung, Baek, Park et al. (2025) "Immunomodulatory effects of supercritical CO(2) extracted oils from Portulaca oleracea and Perilla frutescens (PPCE) in healthy individuals: a randomized double-blind clinical trial" *Food Funct*
90. Ying, Wen, Qingfei et al. (2022) "Efficacy and safety of Mianyi granules (+mianyi+) for reversal of immune nonresponse following antiretroviral therapy of human immunodeficiency virus-1: a randomized, double-blind, multi-center, placebocontrolled trial" *J Tradit Chin Med*
91. Zhuang, Jian, Xin et al. (2022) "Wenshen Jianpi recipe induced immune reconstruction and redistribution of natural killer cell subsets in immunological non-responders of human immunodeficiency virus/acquired immune deficiency syndrome: a randomized controlled trial" *J Tradit Chin Med*
92. Xing, Wei, Zhou et al. (2020) "Efficacy of a Chinese herbal formula on hepatitis B e antigen-positive chronic hepatitis B patients" *World J Gastroenterol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584745&blobtype=pdf | # Performance analysis of the GeneXpert respiratory panel prototype assay for the diagnosis of viral and bacterial upper respiratory tract infections
Linda Chan, Michael Tang, Eddie Leung, Nicole Lee, Viola Chow
## Abstract
GeneXpert respiratory panel (GX-RP) is a new in vitro qualitative multi plexed nucleic acid amplification test designed to simultaneously detect and identify 26 common respiratory pathogens from nasopharyngeal swabs (NPSs) in patients exhibiting signs and symptoms of upper respiratory tract infection. This is a retrospective study that compares the prototype assay with two FDA-approved respiratory panelsthe BioFire FilmArray (FA) Respiratory 2.1 plus and Hologic Panther Fusion (PF) respira tory assay. A total of 292 NPS specimens from patients with upper respiratory tract infection symptoms were collected from three hospitals in Hong Kong, SAR. There was concordance in 269/292 specimens (92.1%), reaching full agreement with Influenza A, Bordetella pertussis, and Mycoplasma pneumoniae. Among the discordant specimens, 7 specimens were only positive for GX-RP, and 16 specimens were only positive for their comparators. Codetections were present in 60.9% of the discordant results. GX-RP has an overall positive percent agreement of 93.1%, negative percent agreement of 99.9%, and a prevalence-adjusted bias-adjusted kappa of 99.0%. It showed good agreement when compared with FA respiratory panel or PF respiratory assay. IMPORTANCE Multiplex respiratory pathogen panels are a diagnostic mainstay in patients with upper respiratory tract symptoms. New platforms and improved versions of previous platforms emerge over time. Early evaluation of their diagnostic performance using real-world data is essential for the necessary revisions to be made, which would facilitate public access to improved panels. Our retrospective study provides prelimi nary evidence that the GeneXpert respiratory panel prototype assay has comparable performance to the BioFire FilmArray and Hologic Panther Fusion respiratory assays and may be an additional candidate in our future toolkit against upper respiratory tract pathogens.
enables robust surveillance of pathogen epidemiology nationally and internationally (1,6). The recent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic has demonstrated the importance of rapid testing in combating URI, and particularly paramount to control epidemics and pandemics (7).
Since respiratory symptoms are non-specific, there is a paradigmatic shift towards syndromic diagnostics where specimens can be tested for multiple respiratory patho gens concurrently. Multiplex real-time polymerase chain reaction (RT-PCR) testing platforms incorporate selected primers of common respiratory pathogens within one reaction tube. It has a fast turnaround time, as fast as 1 h, and the diagnostic yield has greatly increased for viral and atypical bacterial pathogens (8,9). The recent multicenter RP2+ study showed that a wide range of respiratory viruses were actively circulating during the SARS-CoV-2 pandemic, marking the necessity of syndromic respiratory panels for precise diagnosis (10).
Most URIs are primarily caused by viruses, but can also be caused by bacteria. Common pathogens include influenza A and B (Flu A and B), parainfluenza (PIV), adenovirus (AdV), respiratory syncytial virus (RSV), human metapneumovirus (hMPV), and rhinovirus/enterovirus (RV/EV), although there is an apparent epidemiological shift in URI pathogens before and after the SARS-CoV-2 pandemic (11,12). The first FDAapproved respiratory panel in 2008 could detect 14 viruses and took 5 h to process (13). Platform performances have improved substantially since then and encompass improvements such as the inclusion of bacterial targets, higher throughput, the ability to distinguish RV/EV, shortened hands-on operator time, and rapid processing time as fast as 1 h (14).
The objective of this study is to examine the diagnostic performance of a prototype version of a new respiratory panel, GeneXpert respiratory panel (GX-RP), by comparing it with two FDA-approved respiratory panels-the BioFire FilmArray (FA) and Hologic Panther Fusion (PF) respiratory panel assays. A comparison of detectable pathogens by them is shown in Table 1.
$$/EV ✓ M/Ct ✓ M RV Ct SARS-COV-2 ✓ Ct ✓ M ✓ Ct Bacteria B. pertussis ✓ M ✓ M - - B. parapertussis ✓ M ✓ M - - C. pneumoniae ✓ M ✓ M - - M. pneumoniae ✓ M ✓ M - - a AdV =$$
## RESULTS
## Demographics
A total of 292 nasopharyngeal swab (NPS) specimens were collected and compared using the GX-RP assay against its two comparators-PF/FA. The median age was seven, ranging from 1 month to 97 years old. Specimens from children (age < 18) constituted 65% (191/292) of the overall sample. The sex distribution between females and males was 123/292 (42.1%) and 169/292 (57.9%), respectively.
## GeneXpert respiratory panel testing
The relative prevalence of identified organisms in this study in descending order was RSV (20.5%), PIV (17% 2. There were 236 detected organism results with GX-RP and 246 for its comparators, with an overall positive percent agreement (PPA) of 93.1% (229/246). All analytes have a PPA higher than 80% except for B. parapertussis, which has a PPA of 66.7% but is limited by a small number of positive cases (n = 2). The negative percent agreement (NPA) was above 99% for all except for RSV, which had a NPA of 96.6%. The prevalence-adjusted bias-adjusted kappa (PABAK) was all above 0.97. GX-RP and its comparators showed similar PPA, NPA, PABAK, and McNemar values for each specific analyte. When the analyte results were pooled together, all agreement statis tics remained similar, but the McNemar test became statistically significant (P = 0.041). Bacterial targets were few in this study (n ≤ 2), with the exception of M. pneumoniae (n = 6). All were single analyte detections with no other bacterial targets. Three specimens were implicated in discordant results, and two of the detections were codetected with another virus.
## Retrospective analysis of discordant results
Of all the specimens, 269/292 (92.1%) had concordant results between GX-RP and comparator tests, of which 197/292 (67.5%) were consensus positive and 72/292 (24.7%) were consensus negative. 23/292 (7.9%) had discordant results. Among concordant results, children contributed 136/269 (50.6%) of the total specimens and were found positive in all. In contrast, adults contributed to all the negative specimens and 61/197 (31.0%) of the positive specimens. There were fewer females than males (38% vs 62%). Among discordant results, 18/23 (78.3%) were from children, and the sex distributions were similar. Concordant and discordant results stratified by age and sex are shown in Table 3. Discordant specimens had significantly more codetections (14/23, 60.9%) than concordant specimens (16/197, 8.1%) with an odds ratio of 17.1 (95% CI: 5.9-52.8). The most common respiratory pathogens involved in discordant codetection results are RV/EV (n = 5) and hMPV (n = 3). There were seven specimens with seven pathogens detected only by GX-RP, including hMPV (n = 2), RV/EV (n = 2), RSV (n = 1), AdV (n = 1), and B. parapertussis (n = 1). Many of the specimens showed late amplification, suggestive of low analyte levels, ranging from cycle threshold (Ct) values of 35.6 to 44.4. The only exceptions were B. parapertussis and a case of RV/EV, which were detected with the melting curve method. According to the manufacturer's guide and further enquiry with Cepheid, the melting curve method would be utilized when RV/EV is not detected by the amplification method, but specific internal algorithms are not elaborated. On the other hand, there were 16 specimens with 17 pathogens detected only by comparators (one specimen with three positive targets), including AdV (n = 3), RV/EV (n = 3), CoV (n = 2), Flu B (n = 2), PIV (n = 2), RSV (n = 2), hMPV (n = 1), SARS-CoV2 (n = 1), and B. parapertussis (n = 1). Among these, two AdV demonstrated late amplification with Ct > 36. Other AdVs were detected by melting curve analysis. Details of discordant results can be found in Table 4.
All specimens tested with GX-RP and PF were further analyzed to compare their Ct values (Table 5). Concordant results showed similar Ct values between the two assays, whereas discordant cases had generally higher Ct values than concordant cases, although the sample size is small.
## DISCUSSION
## Comparison of respiratory panels
This study compares GX-RP with either FA or PF. There are no specific respiratory panels that are designated as the gold standard for URI testing. Both FA and PF are FDA-approved for standard laboratory testing. A study that compares PF with an older version of FA (2.0) showed good agreement with a PPA of 96.5% and NPA of 98.4% (15), but a head-to-head comparison of their latest versions is lacking. The two platforms have historically delivered reliable performance, even if tested after multiple freeze-thaw cycles (9,(16)(17)(18)(19). Both have been demonstrated to improve clinical outcomes and yield positive economic impacts (20)(21)(22)(23). Despite the similarities in diagnostic performance, there are key differences that could influence the choice of respiratory panels. First, GX-RP and FA have a processing time of only 45-60 min as opposed to approximately 2.5 h for PF (Table 6). A faster turnaround time would yield greater benefit in high-throughput platforms. Second, each assay covers a different set of pathogens, which makes it more appropriate in certain populations depending on the clinical suspicion. GX-RP and FA detect atypical bacteria, which may be more useful in the paediatric population. If MERS is suspected, only FA can reliably detect the pathogen. In a Flu A outbreak, GX-RP and FA allows for subtype differentiation and may improve epidemiological surveillance. PF allows a modular syndromic approach to virus detection but does not detect MERS or bacterial targets (Table 1). Third, PF utilizes real-time amplification for all virus targets, and GX-RP for certain viruses, allowing quantifiable comparisons for Ct values where applicable. It is not known in this study why PA or FA was chosen in each of the cases, but the above reasons are some potential considerations.
## Diagnostic performance
The performance characteristics of GX-RP were similar to FA and PF. Agreement statistics showed NPA, PPA, and PABAK all exceeding 90%. Traditional agreement statistics, such as Kappa, are affected when specimens are dominated by agreement cells. The McNemar test was employed as a more sensitive test to detect differences, which showed a small but statistically significant difference when tests from each analyte were summed. This finding reflects an imbalance between the discordant cells, such that GX-RP is less likely to detect a pathogen (7 vs 17) when it disagrees with its comparators. Nevertheless, it represents only a small number and studies with a larger number of discordant results are needed to better ascertain genuine detection differences across the platforms. Targets such as Flu A, PIV, RSV, and SARS-CoV-2 showed 95%-100% concordance. For Flu A, false positives are possible if vaccines were used, such as FluMist or other live-attenuated influenza vaccines, but these factors were not explored in our study. As Flu A demonstrated 100% agreement when reported up to group level, further strain-specific accuracy could be explored, especially for reporting imported strains with potential public health threats, such as suspected avian flu strains (H5 and H7).
## Discrepant analysis
Earlier studies have suggested that discordant results could be due to low viral load, specimen degradation, cross-reactivity, reduced amplification efficiency due to competition in specimens with multiple positive targets, or multiple freeze-thaw steps of archived specimen processing (9,17,18,(24)(25)(26). Clinical history, such as previous infection, could also be important, highlighted by the propensity of AdV and RV/EV to have prolonged residual presence in the upper respiratory tract, causing false positives (27,28). These are plausible explanations, as most of the discrepant results in our study showed late amplification, raising suspicion that analyte levels are beyond the limit of detection threshold (LoD). We believe codetection may also be a key factor in affecting accuracy, which was present in 14/23 (60.9%) of our discordant specimens (Table 4).
AdV has historically demonstrated low sensitivity for other respiratory panel assays, including BioFire FA version 1.6, with improved sensitivity in subsequent commercial versions by inclusion of more serotypes (17,18,29). Our study showed that AdV has one of the lowest PPAs. Possible explanations may include low level of analytes, which may be below the LoD or differences in subtype coverage by the assays, in which the GX-RP detects group A to E, whereas the comparators cover up to group A to F (24,27). Since subgroups F and G cause gastrointestinal symptoms, and one AdV (Table 4, case 8) not detected by GX-RP had watery diarrhea, a potential explanation would be due to AdV subgroup F not detected by GX-RP but detected by comparators, although this could be over-analysis of one case and beyond the context of respiratory pathogen testing.
## Limitations
There are several limitations due to the retrospective nature of this study. Specimens in this study were freeze-thawed before testing with GX-RP, as compared to fresh specimens subjected to FA or PF. This could potentially affect GX-RP's performance, but our study has found good overall performance. The Ct values comparing GX-RP with PF were similar (Table 5). This is consistent with previous studies that one additional freeze-thaw cycle appears to have a limited impact (24). Testing a fresh specimen is still desirable, and a prospective study design would obviate this issue.
Discrepant results are conventionally resolved by repeat testing or ascertained by other independent methods. This could not be done in this study because of insufficient specimen volume, which could again be alleviated if it was planned in a prospective study. The study is further hampered by the low prevalence of some bacterial targets, such as B. pertussis and B. parapertussis, which makes meaningful interpretation difficult. There were no detected C. pneumoniae or MERS to allow further analysis.
Additionally, the initial study design does not attempt to account for clinical variables such as age, gender, or co-existing infections that may have a significant impact on performance. Positivity rates are known to be higher in children when compared to adults, but it was striking that specimens from children had a 100% positivity rate in this study. We found most of the discrepant results to occur in children (Table 3). Future studies could explore whether there is differential performance across different subgroups of patients.
## Conclusions
Our study is a preliminary analysis of the diagnostic performance of GX-RP, which shows overall good agreement when compared with FA or PF for rapid diagnoses of URI pathogens. The study is limited by its retrospective study design and small sample size. Future studies could meaningfully inform practice by including larger sample sizes and a prospective study design.
## MATERIALS AND METHODS
## Study design and clinical specimens
This retrospective study was conducted at the Prince of Wales Hospital with specimens collected from three hospitals (Prince of Wales Hospital, North District Hospital, and Alice Ho Miu Ling Nethersole Hospital) in Hong Kong, SAR, between January 2020 and September 2023. NPSs were collected from patients presenting with URI symptoms such as cough, rhinorrhea, and sore throat. They were placed in 3 mL viral transport medium and sent to the laboratory for immediate standard-of-care laboratory testing by FDA-approved commercial assays BioFire FA 2.1 plus respiratory panel (bioMérieux, Durham, NC) or PF respiratory panel (Hologic, San Diego, CA) according to the manufac turer's instructions. Aliquots of these specimens were stored at -70°C after standard testing. Basic demographic information, including age and sex, was collected at the time of specimen collection and retrieved from the digital database retrospectively at the time of study. Specimens were first randomly selected for evaluation with the GX-RP prototype assay (Cepheid, Sunnyvale, CA). For pathogens with low prevalence, such as B. pertussis, B. parapertussis, and M. pneumonia, specimens that tested positive by PF or FA were added to allow better characterization of the PPA. A total of 292 specimens were included. Some pathogens were not available in our locality, such as Flu A H1-pdm 2009 strain, C. pneumoniae, and MERS, and could not be included. There was no attempt to further add positive specimens for other purposes such as better distribution of age, sex, or testing by comparator. The selected archived frozen specimens were retrieved and thawed at ambient temperature for testing with GX-RP. Of those selected for this study, 81 had been tested by FA and 211 by PF. The specimens were then subjected to agreement testing. A specimen was defined as consensus positive or negative if all pathogens in the assay had concordant results between GX-RP and the comparator. Discordant specimens were further analyzed for possible explanations, including Ct value (wherever available) and presence of codetections. Traditional discrepant analysis was not performed as there was inadequate volume for many of the samples. The workflow is shown in Fig. 1.
## GX-RP multiplex RT-PCR prototype assay
GX-RP test is an automated in vitro diagnostic test for qualitative detection and differentiation for 26 common respiratory pathogens; AdV (A-E), Coronavirus (CoV HKU1, NL63, 229E, and OC43), Flu A with differentiation of H1-pdm 2009 and Flu B, hMPV, PIV 1-4, RSV A and B, RV/EV, SARS-CoV-2, B. pertussis, B. parapertussis, M. pneumoniae, and C. pneumoniae. Targets are similar for comparator assays (Table 1). The test is performed on Cepheid GeneXpert Instrument Systems equipped with GeneXpert 10 color modules. The test was performed on retrieved remnant specimens. Approximately 300 µL of specimen was dispensed in the cartridge for automated specimen processing, nucleic acid amplification, and detection of the target sequences in specimens using RT-PCR and melt peak detection. The following pathogens are detected using melt curve analysis: Metapneumovirus, Flu A H1-pdm 2009, M. pneumoniae, C. pneumoniae, B. parapertussis, and B. pertussis. GX-RP requires approximately 1 h/run.
## Comparator assays
Two FDA-approved commercial assays were used as comparators, which include either FA or PF respiratory panels. The FA respiratory panel is a multiplex nested PCR, per formed in a closed and autonomous system, allowing the simultaneous detection of 19 viruses and 4 bacteria (Table 1). A 300 µL of specimen is mixed with specimen buffer and injected into a test pouch containing all necessary reagents for nucleic extraction, amplification, and target detection. The pouch is then placed in the FA instrument. The system software automatically interprets the endpoint melting curve data to provide a qualitative result for each target. A microorganism is reported as detected if at least one of its corresponding assays is positive. The test completion time is approximately 1 h. The PF respiratory panel detects a similar range of viruses as FA without bacterial detection (Table 1). Three different multiplexed RT-PCR tests were used to detect and differentiate respiratory viruses: (i) Flu A/B/RSV; (ii) AdV/hMPV/RV; and (iii) PIV 1-4. The separate panels allow modular syndromic testing. A 500 µL of specimen was transferred to a specimen lysis tube and loaded directly onto the Hologic PF System. This plat form performs automated nucleic acid extraction and amplification of the gene target sequences by RT-PCR. The results will be ready in approximately 2.5 h. See Table 6 for comparison of GX-RP and comparator methods.
## Results and statistical analysis
Statistical analyses were performed using R version 4.2.1 with the R packages "epiR" and "fmsb. " PF and FA testing were grouped collectively under the term "comparator. " The type of agreement statistics included in this study include PPA, NPA, PABAK, and McNemar's chi-square test. Confidence intervals of 95% for PPA and NPA were construc ted using exact intervals. Additional comparative statistics included the Fisher's exact test with a level of statistical significance set at P value of < 0.05.
## References
1. Sirota, Doxey, Dominguez et al. (2025) "Global, regional, and national burden of upper respiratory infections and otitis media, 1990-2021: a systematic analysis from the Global Burden of Disease Study 2021" *Lancet Infect Dis*
2. Jin, Ren, Li et al. (1990) "Global burden of upper respiratory infections in 204 countries and territories" *EClinicalMedicine*
3. 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" *Lancet*
4. Bender, Sirota, Swetschinski et al. (2024) "Global, regional, and national incidence and mortality burden of non-COVID-19 lower respiratory infections and aetiologies, 1990-2021: a systematic analysis from the global burden of disease study 2021" *Lancet Infect Dis*
5. Clark, Lindsley, Wigmosta et al. (2023) "Rapid multiplex PCR for respiratory viruses reduces time to result and improves clinical care: results of a systematic review and meta-analysis" *J Infect*
6. Wils, Saegeman, Schuermans (2022) "Impact of multiplexed respiratory viral panels on infection control measures and antimicrobial stewardship: a review of the literature" *Eur J Clin Microbiol Infect Dis*
7. Peeling, Heymann, Teo et al. (2022) "Diagnostics for COVID-19: moving from pandemic response to control" *Lancet*
8. Pierce, Elkan, Leet et al. (2012) "Comparison of the idaho technology filmarray system to real-time PCR for detection of respiratory pathogens in children" *J Clin Microbiol*
9. Popowitch, Neill, Miller (2013) "Comparison of the biofire filmarray RP, genmark eSensor RVP, luminex xTAG RVPv1, and luminex xTAG RVP fast multiplex assays for detection of respiratory viruses"
10. Duclos, Hommel, Allantaz et al. (2022) "Multiplex PCR detection of respiratory tract infections in SARS-CoV-2-negative patients admitted to the emergency department: an international multicenter study during the COVID-19 pandemic" *Microbiol Spectr*
11. Li, Zhang, Ren et al. (2021) "Etiological and epidemiological features of acute respiratory infections in China" *Nat Commun*
12. Maison, Omony, Rinderknecht et al. (2024) "Old foes following news ways?pandemic-related changes in the epidemiology of viral respiratory tract infections" *Infection*
13. Mahony, Chong, Merante et al. (2007) "Development of a respiratory virus panel test for detection of twenty human respiratory viruses by use of multiplex PCR and a fluid microbead-based assay" *J Clin Microbiol*
15. Diaz-Decaro, Green, Godwin (2018) "Critical evaluation of FDAapproved respiratory multiplex assays for public health surveillance" *Expert Rev Mol Diagn*
16. Boerger, Binnicker (2020) "Comparison of the panther fusion respiratory panels to routine methods for detection of viruses in upper and lower respiratory tract specimens" *Diagn Microbiol Infect Dis*
17. Stellrecht, Cimino, Wilson et al. (2019) "Panther fusion respiratory virus assays for the detection of influenza and other respiratory viruses" *J Clin Virol*
18. Doern, Lacey, Huang et al. (2013) "Evaluation and implementa tion of filmarray version 1.7 for improved detection of adenovirus respiratory tract infection" *J Clin Microbiol*
19. Leber, Everhart, Daly et al. (2018) "Multicenter evaluation of biofire filmarray respiratory panel 2 for detection of viruses and bacteria in nasopharyngeal swab samples" *J Clin Microbiol*
20. Chen, Lam, Yip et al. (2016) "Clinical evaluation of the new highthroughput luminex NxTAG respiratory pathogen panel assay for multiplex respiratory pathogen detection" *J Clin Microbiol*
21. Mahony, Blackhouse, Babwah et al. (2009) "Cost analysis of multiplex PCR testing for diagnosing respiratory virus infections" *J Clin Microbiol*
22. Nelson, Stockmann, Hersh et al. (2015) "Economic analysis of rapid and sensitive polymerase chain reaction testing in the emergency department for influenza infections in children" *Pediatr Infect Dis J*
23. Rogers, Shankar, Jerris et al. (2015) "Impact of a rapid respiratory panel test on patient outcomes" *Arch Pathol Lab Med*
24. Brendish, Malachira, Armstrong et al. (2017) "Routine molecular point-of-care testing for respiratory viruses in adults presenting to hospital with acute respiratory illness (ResPOC): a pragmatic, open-label, randomised controlled trial" *Lancet Respir Med*
25. Popowitch, Kaplan, Wu et al. (2022) "Comparative performance of the luminex NxTAG respiratory pathogen panel, genMark eSensor respiratory viral panel, and bioFire filmarray respiratory panel" *Microbiol Spectr*
26. Penela-Sánchez, González-De-Audicana, Armero et al. (2021) "Lower respiratory tract infection and genus Enterovirus in children requiring intensive care: clinical manifesta tions and Impact of viral co-infections" *Viruses*
27. Upadhyay, Reddy, Proctor et al. (2014) "Expanded PCR panel testing for identification of respiratory pathogens and coinfections in influenza-like illness" *Diagnostics (Basel)*
28. Asner, Rose, Petrich et al. (2015) "Is virus coinfection a predictor of severity in children with viral respiratory infections?" *Clin Microbiol Infect*
29. Mansbach, Piedra, Teach et al. (2012) "Prospective multicenter study of viral etiology and hospital length of stay in children with severe bronchiolitis" *Arch Pediatr Adolesc Med*
30. Huang, Tsai, Chang et al. (2018) "Multiplex PCR system for the rapid diagnosis of respiratory virus infection: systematic review and meta-analysis" *Clin Microbiol Infect* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724141&blobtype=pdf | # Direct airway delivery of a humanized anti-H7N9 neutralizing antibody broadly protects against divergent H7 influenza viruses in the mouse model
Wang Yu, Xiaozheng He, Jiangyan Zhao, Yunlong Dou, Tingyu Hu, Xia Chen, Xuran Ma, Xiaoquan Wang, Shunlin Hu, Jiao Hu, Xiufan Liu, Zenglei Hu
## Abstract
Passive administration of broadly neutralizing anti-influenza monoclonal antibodies (mAbs) before or after virus infection can prevent or alleviate disease. Unlike seasonal influenza, infection with zoonotic avian influenza viruses can lead to acute respiratory distress syndrome and high mortality in humans. Respiratory tract-targeting antibody delivery appears to be more clinically relevant and effective for zoonotic influenza treatment. In this study, the efficacy of an anti-H7N9 murine mAb 4B7 and its humanized form (chi4B7) against H7 subtype influenza viruses administered through the intranasal route was investigated in mice. 4B7 recognizes critical residues in the vestigial esterase domain and receptor-binding sites in the hemagglutinin of H7N9 virus. The antibody had cross-H7 binding, hemagglutination inhibition, and neutralizing activities. In particular, the dose of 4B7 required for prophylactic protection against H7N9 infection was significantly reduced in mice treated locally (intranasal) compared with those treated systemically (intraperitoneal). Intranasal delivery of the antibody also enhanced therapeutic efficacy against H7N9 infection compared to intraperitoneal administration. Chi4B7 generated by grafting the variable regions onto the human IgG1 backbone sustained cross-reactivity with different H7 viruses of the parental murine antibody. Airway-delivered chi4B7 provided broad prophylactic and therapeutic protection against divergent H7 viruses in mice. Moreover, intranasal administration of chi4B7 had a long effective prophylaxis window against H7N9 infection. Our results suggest that airway delivery of the humanized anti-H7 antibody is a favorable approach for broad-spectrum prophylaxis and therapy against the H7 subtype influenza. IMPORTANCE Infection of zoonotic H7 avian influenza viruses can cause severe respiratory symptoms and high mortality in humans. Monoclonal antibody adminis tration is an effective approach for treatment of zoonotic influenza infection, while systematic routes of antibody administration (typically intravenous infusion) have several shortcomings. However, there are no approved anti-H7 antibody therapies, and the efficacy of antibodies administered through the airway route against H7 viruses has not been fully investigated. Herein, we report a murine broadly neutralizing monoclo nal antibody against divergent H7 viruses and reveal that intranasal administration enhanced prophylactic and therapeutic efficacy of this antibody against H7N9 virus compared to systemic administration. Airway delivery of the humanized antibody conferred broad protection against diverse strains of H7 virus in mice. Our study presents new candidates of broad antiviral agents against H7 avian influenza viruses and highlights airway delivery as a more potent manner of administering antibodies for clinical treatment of influenza.
KEYWORDS H7 subtype influenza virus, neutralizing monoclonal antibody, airway delivery, humanized antibody, broad protection A vian influenza viruses (AIVs) pose a significant threat to both animal and human health due to their ability to cross species barriers and cause zoonotic infections (1). Among AIVs, the highly pathogenic H5N1 subtype (clade 2.3.4.4b) has recently caused an unprecedented outbreak in dairy cows and other mammalian animals in the United States (2,3), with spillover events leading to dozens of confirmed human infections (4). While H5N1 has garnered significant attention, the H7 subtype AIVs also circulate and cause disease outbreaks in poultry worldwide (5)(6)(7)(8)(9). In addition, the H7 subtype AIVs have also emerged as a public health concern. Between 2013 and 2017, five epidemic waves of H7N9 virus in China resulted in over 1,500 laboratory-confirmed human infections, resulting in severe respiratory illness and a mortality rate approaching 40% (1). In addition to the H7N9 subtype, viruses of the H7N2, H7N3, H7N4, and H7N7 subtypes also cause human infection cases with mild-to-severe respiratory symptoms (1,7). In 2018, a human case infected with a novel H7N4 subtype AIV was reported in China (10). Currently, H7N4 viruses are circulating in shorebirds, and a strain is pathogenic in mice without prior adaptation, indicating a potential risk to other mammals and humans (11). Humans remain immunologically naïve to the H7 subtype AIVs, raising concerns that zoonotic H7 viruses could acquire pandemic potential through adaptation to human hosts. Therefore, it is vital to develop effective antiviral agents against the H7 subtype viruses.
Despite ongoing efforts to develop vaccines against the H7 subtype AIVs (12)(13)(14), no licensed vaccines are currently available for human use. Moreover, existing vaccination strategies of H7 vaccines often induce suboptimal levels of hemagglutination inhibition (HI) and virus-neutralizing (VN) antibodies (15,16), underscoring the need for improved vaccine design and optimization. Moreover, the efficacy of vaccines is impaired in immunocompromised and elderly populations. For severe influenza infections, including those caused by zoonotic H5 and H7 viruses, effective antiviral therapeutics are critical for reducing morbidity and mortality (17,18). While the neuraminidase inhibitors (such as oseltamivir) and the cap-dependent endonuclease inhibitor (baloxavir marboxil) are currently used to treat influenza, their efficacy is limited by time-sensitive administration and the emergence of drug-resistant virus strains (17,19). Administration of neutraliz ing monoclonal antibodies (mAbs) is an attractive alternative for influenza treatment, especially for immunocompromised and elderly populations and individuals who are at high risk of virus exposure (20,21). Moreover, mAbs targeting the conserved neutralizing epitopes are desired to protect against diverse influenza viruses.
MAbs have emerged as a promising approach for prophylaxis and therapy of viral infections, with several mAb therapies already approved for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and Ebola virus infections (22). Notably, numerous antibodies against the H7 subtype AIVs have demonstrated potent antiviral activity in preclinical studies (23)(24)(25)(26). Perhaps due to the effectiveness of systematic passive immunization using infused antisera, the general approach for delivering mAbs has been through the systemic route, mainly intravenous infusion. However, influenza virus infection is initiated in and mainly restricted to the respiratory tract, and thus local administration of anti-influenza mAbs to the patient's airway appears to be a clinically relevant and effective approach. Moreover, the concentration of systematically administered antibodies in the respiratory tract is low, and thus a high concentration of antibodies are needed for protection from lethal infection. Both the manufacturing process and the antibody amount required for protection make mAb therapy expensive and inaccessible for large-scale implementation. Previous studies have shown that direct administration in the respiratory tract can improve the efficacy of broadly neutralizing mAbs against influenza A (H1N1, H1N1pdm, H3N2, and H5N1) and B viruses compared to the systematic route of administration (27,28). Infection of zoonotic H5 or H7 viruses can lead to acute respiratory distress syndrome (ARDS) and high mortality, while the efficacy of anti-H7 mAbs through airway delivery remains to be fully assessed.
Antibody humanization and human antibody production are important approaches in passive immune therapy because murine antibodies can trigger a human anti-mouse antibody response (29). Therefore, minimizing the murine immunogenic components in antibodies while preserving the functional antigen-binding domains of the original murine antibodies is a common strategy for the design of therapeutic antibodies. Several human antibodies against the H7N9 subtype AIV were generated from vacci nated-healthy donors or virus-infected patients (23)(24)(25)30). However, the source of human specimens is limited, and collection and handling of such samples are strictly regulated. Therefore, humanization of murine antibodies is still an indispensable method for developing clinically relevant antibodies.
Although there are many reports on the generation of antibodies against H7 viruses, studies on airway delivery and humanization of anti-H7 mAbs are limited. In this study, a neutralizing murine mAb 4B7 against the hemagglutinin (HA) of the H7N9 subtype AIV was generated. This antibody targets the vestigial esterase domain (VED) and receptor-binding sites (RBS) in H7 HA and can cross-react to divergent H7 viruses. Systemic and intranasal administration of the antibody conferred protection against H7N9 virus infection in mice, while airway delivery significantly enhanced the efficacy of the antibody. A humanized antibody generated by grafting the variable regions onto a human IgG1 backbone sustained antigen binding, HI, and neutralizing activities of the murine antibody. More importantly, airway delivery of the humanized antibody provided broad protection against H7N9, H7N4, and H7N3 viruses. Our results suggest that the airway delivery of a humanized anti-H7 neutralizing antibody could be an effective approach to protect against divergent H7 subtype influenza viruses.
## RESULTS
## Generation and characterization of H7N9-reactive mAbs
To screen neutralizing mAbs against H7N9 virus, the frozen hybridomas prepared in our previous study (31) were recovered and screened using enzyme-linked immunosorbent assay (ELISA) and microneutralization test. After a three-round screening, a mAb 4B7 with potent neutralizing activity against H7N9 virus was obtained. Minimal binding concentrations of the mAb with the H7N9 HA protein and H7N9 viruses were 0.01 and 0.001 µg/mL, respectively, indicating high binding affinity of the antibody to the antigens (Fig. 1A through C). VN and HI titers of the antibody against three H7N9 viruses were determined, and amino acid identity of the HA protein among these viruses is above 97% (Fig. S1A). Half-maximal inhibitory concentrations (IC 50 ) of 4B7 against a mouse-adapted A/chicken/China/SDL124/2015 (maSDL124), PR8/SF003 (a PR8 6:2 reassortant with the HA and NA genes deriving from A/Guangdong/17SF003/2016), and A/chicken/Guangdong/GD15/2016 (GD15) were 0.03, 0.036, and 0.34 µg/mL, respectively (Fig. 1D andE). In addition, the minimal HI concentrations (MHC) of 4B7 against maSDL124, PR8/SF003, and GD15 were 0.94, 0.47, and 2.5 µg/mL, respectively (Fig. 1F). These results indicate that 4B7 is an H7N9 HA-reactive mAb with potent neutralizing and HI activities against H7N9 viruses.
## Interaction between 4B7 and the H7N9 HA analyzed by molecular docking
Sequence analysis and isotype determination revealed that the mAb 4B7 is a murine IgG1 antibody. The antibody uses IGHV1-71-6*01 for the heavy chain with 9% somatic hypermutation (SHM) and IGKV8-27*01 for the light chain with 6% SHM (Fig. 2A). Three complementary determining regions (CDRs) in the heavy and light chains of 4B7 were identified, and CDR3 in the heavy (HCDR3) and light chains (LCDR3) are 14 and 8 amino acids, respectively (Fig. 2A). Molecular docking demonstrated that a total of 48 amino acids were present in the interface between the 4B7 single-chain antibody (scFv) and H7N9 HA, and their interactions involved the VED, RBS (130-loop and 150-loop), and the antigenic site A (Fig. 2B through D; Table S1). Most interactions are driven by the heavy chains and consist of hydrogen and salt bridge bonds. Amino acids in the HCDR1 (D36 and W38) and HCDR2 (D57, S59, D62, and S63) form hydrogen or salt bridge bonds with key residues in the 130-loop (R139, T140, N141, and V143) (Table S2). Three residues in the HCDR2 (T65, Y67, and R72) mainly bind critical residues in the 140-loop in the antigenic site A (R149, S150, and S152) through hydrogen bonds (Fig. 2E; Table S2). Y112 in the HCDR3 and S59 in the HCDR2 bind to L159 and T165 in the 150-loop via hydrogen bonds, respectively (Table S2). The light chain is less involved in interaction, and residues in the LCDR1 and LCDR3 form hydrogen bonds with the 128-KEPMG-132 motif close to the 130-loop (Table S2). These results suggest that the mAb 4B7 interacts with the VED and critical domains in the RBS of the H7N9 HA protein.
## G151E mutation in the antigenic site A mediates H7N9 virus escape from 4B7 neutralization
To validate epitope prediction by molecular docking, escape mutants were generated in chicken embryos by incubating a mixture of the H7N9 GD15 strain and 4B7 at a neutraliz ing concentration. After a three-round screening, there was a 16-fold reduction in the HI titer of the antibody against the escape mutants compared to that against the parental virus (Table S3). The escape mutants were plaque-purified, and two mutations, G151E and I335V, in the HA protein were identified (Table S3).
To determine the role of these mutations in virus escape, two reassortant H7N9 viruses carrying the G151E or I335V mutation in HA were generated. The mAb 4B7 had strong neutralizing activities against GD15 (IC 50 : 0.03 µg/mL) and GD15-HAI335V (IC 50 : 0.02 µg/mL) but did not neutralize GD15-HAG151E at up to 10 µg/mL (Fig. 2F). Similarly, MHCs of 4B7 against GD15 (1.31 µg/mL) and GD15-HAI335V (0.98 µg/mL) were compara ble, while the MHC against GD15-HAG151E was significantly increased by around 40-fold (52.1 µg/mL) (Fig. 2G). In addition, the reactivity of the mAb 4B7 with GD15-HAG151E was markedly decreased compared to that with GD15, as determined by immunofluorescence assay (IFA) (Fig. 2H). G151 is a critical residue in the antigenic site A (148-RRSGSS-153, based on H7 numbering) in H7 HA, and mutation at this position resulted in prominent conformational changes of residues in this domain, including R148, S150, and G151 (Fig. 2I andJ). Moreover, molecular docking revealed no interface between the HA protein carrying the G151E mutation and the 4B7 scFv (Fig. 2K andL), which correlated to the IFA data and validated the role of this mutation in virus escape. Therefore, the G151E Full-Length Text mutation in the antigenic site A, a major epitope targeted by 4B7, was associated with H7N9 virus escape from 4B7 neutralization.
## The mAb 4B7 cross-reacts with diverse H7 viruses
The G151E mutation in the antigenic site A mediated virus escape, indicating an important role of this domain in antibody-virus binding. To investigate conservation of this domain, the amino acid sequences of the H7 HA were aligned, and a phylo genetic tree was generated. A stark divergence could be observed between the HA genes of North American (NA) and Eurasian (EA) lineages, whereas the antigenic site A is a common and conserved motif in all H7 HA sequences (Fig. 3A). The amino acid sequence of the antigenic site A (148-RRSGSS-153) was found in the vast majority of EA lineage H7 HAs, while a small proportion of the isolates have 148-KRSGSS-153, such as H7N2-DK2007 used in this study (Fig. 3A). The 148-RRSGSS-153 sequence was found in half of NA lineage isolates, and the second-most commonly observed sequence is 148-TRSGSS-153 (Fig. 3A), which is found in the HA gene of A/feline/New York/16-040082-1/2016 (H7N2-fel2016). Therefore, although there is sequence variation, the antigenic site A, especially G151, targeted by the mAb 4B7 is highly conserved in the H7 subtype.
Subsequently, cross-reactivity of 4B7 with different influenza subtypes was examined. Reactivity of the antibody with the HA proteins of representative subtypes in group 1 (H1N1, H2N2, H5N1, and H9N2) and group 2 (H3N2, H4N6, H7N7, H7N9, and H15N8) was tested in ELISA, and 4B7 only reacted with H7N7 and H7N9 HAs (Fig. 3B). To investigate the cross-reactivity of 4B7 with different H7 viruses, 6:2 reassortant viruses based on the PR8 internal genes were generated (Table S4). The HA and NA genes of PR8/H7N4, PR8/ H7N3, and PR8/felH7N2 viruses were derived from A/Chicken/Jiangsu/2018 (H7N4-JS2018), A/Mexico/InDRE7218/2012 (H7N3-Mex2012), and A/feline/New York/ 16-040082-1/2016 (H7N2-fel2016), respectively. H7N9-GD15, H7N4-JS2018, and H7N2-DK2007 represent EA lineage, and H7N2-fel2016 and H7N3-Mex2012 represent the NA lineage (Fig. 3A). There are high similarities of HA amino acid sequences within H7 viruses of EA (93.4-95.2%) and NA lineages (90.9%) and low identities ranging from 80.7% to 85.7% between the viruses of EA and NA lineages (Fig. S1B). Of note, the amino acid at position 151 in HA of these viruses is G, whereas residue 148 in H7N2-fel2016 and H7N2-DK2007 HAs is T and K, respectively (Fig. 3A). 4B7 had high VN activities against H7N9-GD15, PR8/H7N4, and PR8/H7N3 viruses, with minimal neutralizing concentrations (MNC) of 0.1, 0.01, and 0.06 µg/mL, respectively (Fig. 3C). Nevertheless, the antibody showed significantly lower VN titers against H7N2-DK2007 (MNC: 10 µg/mL) and undetectable neutralizing activity against PR8/felH7N2 at up to 100 µg/mL (Fig. 3C). Similar patterns were observed for HI activity of the antibody against H7 viruses (Fig. 3D). Additionally, the mAb 4B7 had reactivity with H7N9 and PR8/H7N3 in IFA, as expected, and unexpect edly, PR8/H7N2 was also recognized by the antibody (Fig. 3E). Interestingly, molecular docking analysis demonstrated that the 4B7 scFv did not interact with the antigenic site A of H7N2-fel2016 (Fig. 3F) but turned to bind a domain in the stalk distant from the antigenic site A (Fig. 3G andH). These findings suggest that the mAb 4B7 broadly reacts to diverse H7 viruses, and antibody reactivity with the viruses is affected by the variation of residue 148 in the antigenic site A.
## The mAb 4B7 protects mice from H7N9 virus infection in prophylactic and therapeutic settings via intraperitoneal administration
One potential use of anti-influenza mAb therapy is the prophylactic administration of antibodies to people who are likely exposed to infected humans or animals. To assess in vivo protection efficacy conferred by 4B7, an antibody transfer study was performed in mice in prophylactic and therapeutic settings. BALB/c mice were administered with the antibody at 30, 20, 10, and 5 mg/kg via the intraperitoneal (i.p.) route and then infected with 10 50% mouse lethal dose (MLD 50 ) of the H7N9 maSDL124 virus through the intranasal (i.n.) route after 2 h (Fig. S2A). Body weight of the mice administrated with the antibody at all doses steadily increased (Fig. S2B), and administration of 4B7 completely protected against mortality (100% survival) (Fig. S2C). The PBS-inoculated mice started to lose body weight from day 3 post-challenge (p.c.) and all succumbed to infection at day 9 p.c. (Fig. S2B andC). At days 3 and 5 p.c., lung virus loads around 10 3.5 50% tissue culture infectious dose (TCID 50 )/0.1 mL were detected in the PBS-treated mice and were significantly decreased in the antibody-treated mice (Fig. S2D). At day 3 p.c., lung virus titers in the 4B7-treated mice were below the detection limit, and at day 5 p.c., a low amount of the virus was isolated from the lung of one mouse that received 30 mg/kg of the antibody (Fig. S2D).
In terms of pathology, no obvious lesions were observed in the lungs of the noninfected mock mice, while severe tissue lesions such as extensive infiltration of lympho cytes and neutrophils marked incrassation of the alveolar wall, and cellular necrosis was seen in the lungs of the PBS-inoculated mice (Fig. S2E). The antibody at all doses significantly reduced the severity of lung tissue damage compared to the PBS-inoculated mice (Fig. S2F andG), whereas mild incrassation of the alveolar wall, cellular necrosis, and infiltration of inflammatory cells were still observed in the mice treated with low doses (5 and 10 mg/kg) of 4B7 (Fig. S2E).
Subsequently, a therapeutic regimen was tested to simulate a clinical scenario of treatment of influenza virus infection in humans. Mice were infected with 10 MLD 50 of maSDL124 12 h or 24 h before mAb administration (Fig. S3A andD). The PBS-treated mice lost their weight fast, and all died within 5 days p.c. (Fig. S3B, C, E andF). The mice received the antibody (5 and 10 mg/kg) at 12 h p.c. rapidly lost weight, and none survived beyond day 8 p.c. (Fig. S3B andC). Treatment with 4B7 at 20 and 30 mg/kg reduced weight loss, and the mice regained weight from day 9 p.c. (Fig. S3B), conferring 60% and 100% protection, respectively (Fig. S3C). When the antibody was administered 24 h after infection, all the mice that received 5 mg/kg of 4B7 quickly died, and the antibody at 10 and 20 mg/kg provided 20% protection. Antibody treatment at 30 mg/kg protected 40% of the mice (Fig. S3E andF).
Taken together, these findings indicated that the mAb 4B7 is prophylactically and therapeutically protective against H7N9 infection via systemic delivery, and an early administration of a high dose was required for a potent therapeutic protection.
## Intranasal administration of the mAb 4B7 is superior to intraperitoneal administration in a prophylactic setting
Influenza virus infection is mainly restricted in the respiratory tract, and the efficacy of topically (i.n.) and systematically (i.p.) applied 4B7 against H7N9 virus was compared. Groups of mice were treated with the antibody at 3, 1, 0.3, and 0.1 mg/kg through the i.p. or i.n. route 2 h prior to infection (Fig. 4A). As shown in Fig. 4B, the PBS-inoculated mice exhibited dramatic weight loss after H7N9 virus infection, and all died at day 9 p.c. Body weight loss of the mice receiving the antibody at 3 and 1 mg/kg via the i.p. route was significantly reduced compared to that of the PBS group, whereas only i.p. administration of 4B7 at 3 mg/kg can significantly prevent weight loss compared to the mock group. Significant weight loss was detected in the mice treated with the antibody at 0.3 and 0.1 mg/kg via the i.p. route. In addition, i.n. administration of 4B7 at 3, 1, 0.3, and 0.1 mg/kg significantly reduced body weight loss compared to the PBS group, and there were no significant differences between the antibody-treated and mock control mice. When comparing the same doses, there was a significant difference in body weight between the mice treated with 0.1 mg/kg of 4B7 through the i.p. and i.n. routes. In terms of mortality, survival of the mice receiving the antibody at 3, 1, and 0.3 mg/kg via the i.p. route was significantly higher than that of the PBS group, and administration of 4B7 at 0.1 mg/kg protected 12.5% of the mice (Fig. 4C). By contrast, i.n.-administered antibody at all doses conferred complete protection against mortality caused by H7N9 virus infection (Fig. 4C). Moreover, at days 3 and 5 p.c., virus loads in the lungs of the mice receiving the antibody at 3 mg/kg through the i.p. route were significantly reduced compared to that of the PBS-treated mice (Fig. 4D andE). At day 5 p.c., i.p.-administered antibody at 1 and 0.1 mg/kg significantly decreased virus titers in the lungs (Fig. 4E). It is noted that the H7N9 virus could not be isolated from the lungs of the mice receiving the antibody at all doses via the i.n. route (Fig. 4D andE). Delivery of the antibody at all doses through the i.p. or i.n. route (except 0.1 mg/kg via the i.p. route) significantly alleviated lung pathology at day 3 and 5 p.c. (Fig. 4F through H).
Taken together, these results suggest that airway delivery of the mAb 4B7 significantly enhanced efficacy against H7N9 infection compared to systemic administration in a prophylactic setting.
## Intranasal delivery of 4B7 enhances therapeutic efficacy against H7N9 virus
To compare the therapeutic efficacy of the mAb 4B7 delivered through the systemic or topical route, mice were infected with H7N9 virus and then treated with 4B7 via either the i.p. or i.n. route 48 and 72 h later (Fig. 5A andD). We found that body weight of all the infected mice gradually decreased from day 1 to 5 p.c. and was regained from day 6 p.c. (Fig. 5B andE). Compared to the PBS-inoculated mice, treatment with the antibody at all doses at 48 and 72 h post H7N9 infection failed to significantly reduce body weight loss (Fig. 5B andE). In addition, i.n. administration of 4B7 at 10 mg/kg at 48 h p.c. conferred complete protection, whereas 40% of the mice receiving the same dose via the i.p. route survived (Fig. 5C). When given at 72 h p.c., 10 mg/kg of the antibody conferred no protection (i.p.) or 20% protection (i.n.) against H7N9 virus infection (Fig. 5F). When the antibody dose was decreased to 3 mg/kg, i.n. administration at 48 h p.c. provided 60% protection, whereas a 40% survival was observed for the i.p. route (Fig. 5C). The antibody at 3 mg/kg administered at 72 h provided no (i.p.) or 40% protection (i.n.) against H7N9 virus infection (Fig. 5F). Moreover, i.n.-administered 4B7 at 1 mg/kg still protected 40% of the mice at 48 h, while i.p. delivery conferred 20% protection (Fig. 5C). The antibody at 1 mg/kg administered via the i.p. route provided no protection at 72 h, while i.n. delivery at the same dosage still protected 20% of the mice (Fig. 5F). These data demonstrated an enhanced therapeutic efficacy of locally administered 4B7 at 48 h post H7N9 virus infection.
## Humanized 4B7 sustains similar reactivity to H7 viruses as the murine 4B7
The goal of this study was to develop a neutralizing antibody for H7 influenza prevention and treatment in humans. For clinical use, the immunogenicity of murine antibodies should be diminished to reduce anti-antibody immune responses. To this end, humani zation of the murine antibody 4B7 was performed. A chimeric antibody (chi4B7) was generated by grafting the VH and VL regions onto a human IgG1 and kappa backbone (Fig. 6A). The antibody chi4B7 was expressed in Chinese hamster ovary (CHO) cells and purified. The molecular sizes of the heavy chain and light chain were around 55 and 25 kDa, respectively, under reducing conditions and 150 kDa under non-reducing conditions (Fig. 6B). The H7 HA-binding, HI, and VN activities of chi4B7 against H7N9, PR8/H7N3, and PR8/H7N4 viruses were comparable to those of the parental mouse antibody (Fig. 6C through E). These findings suggest that humanization of the murine mAb 4B7 did not alter its activities with H7 viruses.
## The humanized antibody confers cross-protection against divergent H7 viruses
The humanized antibodies can broadly react to H7 viruses, and thus cross-protection provided by this antibody against different H7 viruses was assessed in mice in prophy lactic and therapeutic settings. In a prophylaxis model, mice were i.n. administered with chi4B7 and then infected with H7N9, PR8/H7N4, and PR8/H7N3 viruses 2 h later (Fig. 7A). The PBS-inoculated mice rapidly lost their body weight after challenge, and all succumbed to virus infection (Fig. 7B through G). Body weight loss of all the mice receiving 3, 1, and 0.3 mg/kg of chi4B7 was significantly reduced compared to the PBS-inoculated mice after infection with H7N9, PR/H7N4, and PR8/H7N3 viruses (Fig. 7B through D). Compared to the mock control mice, there were no significant differences in the body weight of the antibody-treated mice post-infection with H7N9 and PR8/H7N4 viruses, except those receiving chi4B7 at 1 mg/kg after H7N9 virus infection, while significant weight loss was detected for the mice administered with the antibody at all doses after PR8/H7N3 challenge (Fig. 7B through D). It is noted that the humanized antibody completely conferred protection against mortality after infection with H7N9, PR8/H7N4, and PR8/H7N3 viruses (Fig. 7E through G). For virus load determination, because PR8/H7N4 had a low virus titer in MDCK cells (Table S4), the virus in the lungs could not be isolated in MDCK cells, and thus quantitative real-time PCR was used to measure lung virus loads of this virus. Lung virus loads of all antibody-treated mice were below the detection limit and significantly lower than that of the PBS-inoculated mice (Fig. 7H through J). Lung pathology caused by H7N9 and PR8/H7N4 viruses was significantly alleviated in the mice receiving 3 and 1 mg/kg of the antibody (Fig. S4 andS5). For the PR8/H7N3 virus, although the severity of lung lesions in the antibody-treated mice was significantly decreased, marked pathological changes including hemorrhage, infiltration of inflammatory cells, and incrassation of alveolar wall were still seen in the mice receiving the antibody (Fig. S6).
To assess the therapeutic efficacy of the humanized antibody against different H7 viruses, mice were infected with H7 viruses and then treated with chi4B7 through the i.n. route 48 h later (Fig. 8A). The H7N9 subtype AIVs cause the largest number of human infection cases and are still prevalent in poultry (1,6,9). The H7N4 subtype AIV has caused a human infection in 2018 (10), and a recent study highlighted circulation of this subtype in shorebirds, their pathogenicity in mice without prior adaptation, and potential threat to public health (11). H7N9 and PR8/H7N4 viruses were selected for the therapeutic study for these reasons. Body weight of the virus-infected mice decreased, and those who received the antibody (10 and 3 mg/kg) at day 2 p.c. recovered shortly and gained weight on days 4 to 6 after treatment (Fig. 8B andE). One mouse succumbed to H7N9 infection at days 4 and 5 after receiving the antibody at 10 and 3 mg/kg, respectively. The humanized antibody at these two dosages conferred 80% protection against H7N9 virus infection (Fig. 8C). Antibody therapy also significantly reduced H7N9 virus titers in the lungs (Fig. 8D). In addition, chi4B7 treatment conferred complete protection against PR8/H7N4 virus infection (Fig. 8F), and lung virus loads were also significantly decreased after treatment with 10 mg/kg of the antibody (Fig. 8G). However, moderate-to-severe lung lesions were still observed in the mice therapeutically treated with the antibody at 10 and 3 mg/kg after infection with H7N9 and PR8/H7N4 viruses, despite that treatment with 10 mg/kg of the antibody significantly alleviated lung lesions in the PR8/H7N4-infected mice (Fig. 8H through Q). These data showed that the humanized antibody conferred prophylactic and therapeutic cross-protection against divergent H7 viruses and reduced virus burden in the lungs, whereas antibody therapy did not significantly alleviate lung pathology.
## Airway-delivered humanized antibody has a long effective prophylaxis window
It is of interest to determine how long the effective prophylaxis window is for airwaydelivered chi4B7. A survival study was performed in the H7N9 infection model with antibody administration initiated at 48 and 72 h before virus infection (Fig. 9A andD). The antibody at 3 and 0.3 mg/kg delivered at 48 and 72 h before infection significantly reduced body weight loss when compared to the PBS-inoculated mice, whereas only the 3 mg/kg dose given at 48 h prior to infection can confer protection against weight loss compared to the mock control mice (Fig. 9B andE). In addition, i.n. administration of the antibody at 3 and 0.3 mg/kg at 48 and 72 h before infection conferred full protection against mortality (Fig. 9C andF). These findings indicate that the humanized antibody has a long effective window when prophylactically delivered to the airway.
## DISCUSSION
In this study, a murine mAb against the HA of H7N9 virus was generated, and a humanized version of this antibody was produced. The antibody binds critical residues in the VED and RBS in H7 HA and particularly, the G151E mutation in the antigenic site A mediated virus escape from the antibody. The murine and humanized antibodies exhibited cross-binding, HI, and neutralizing activities with diverse H7 viruses. Airway delivery of the murine antibody reduced the dose required for potent prophylactic and therapeutic protection against H7N9 virus in mice compared to systemic administration. Intranasal administration of the humanized antibody conferred broad protection against diverse H7 viruses and had a long effective prophylaxis window. Our findings suggest that airway delivery of the humanized antibody is a feasible and cost-effective approach for prevention and therapy against H7 influenza.
Many mAbs against the H7N9 subtype AIV were generated to develop novel antiviral agents, and systemic administration of antibodies can protect from overt clinical signs and mortality caused by H7N9 viruses (23,25,32). Respiratory virus infection, including influenza infection, in humans is typically initiated and limited in the respiratory tract (17,33). Thus, concentrations of antibodies in the site of infection are essential for protection. However, the bioavailability of systemically administered antibodies in the respiratory tract is low, and high antibody doses must be administered for protection against fatal infection (28). This provides a rationale for delivering anti-influenza antibodies directly to the airway side of the respiratory tract.
Consistent with this idea, we found that intranasal administration of the mAb 4B7 via the i.n. route improved prophylactic efficacy against H7N9 virus infection in mice compared to the i.p. route. Particularly, 3 mg/kg of the antibody administered via the i.p. route was required for significant protection against body weight loss and mortal ity, whereas i.n. delivery of 4B7 at 0.1 mg/kg was sufficient to confer protection. In addition, i.n. delivery also enhanced the therapeutic efficacy of the antibody against H7N9 virus. Previous studies showed that compared to systemic administration, airway delivery enhanced the efficacy of antibodies against respiratory pathogens, including influenza virus, respiratory syncytial virus, and SARS-CoV-2 (27,28,34,35). In this regard, our findings correlate with these reports, highlighting the advantage of airway delivery over systemic administration of antibodies in treating respiratory infections. Besides enhancing efficacy, airway delivery that allows dose sparing is a highly desirable approach to reduce the cost and increase the accessibility of antibody therapy for large-scale administration. Although airway delivery of the mAb 4B7 is superior to systemic administration in mice, it is still necessary to compare the efficacy of the antibody delivered via these two routes in ferrets, the optimal disease model for human influenza. Friesen et al. demonstrated prophylactic and therapeutic efficacy of a human mAb CR6261 at 30 mg/kg against H5N1 subtype AIV in ferrets through intravenous injection (36). Another study showed that i.p. administration of MEDI8852 at 25 mg/kg, a broadly reactive mAb against the HA stem, was effective for lethal H5N1 and H7N9 infection and can protect naive ferrets from airborne transmission of H1N1pdm09 (37). Therefore, head-to-head comparison studies are required to determine whether airway delivery of anti-influenza antibodies outperforms systemic administration in ferrets. Our study also presented new findings supporting antibody airway delivery for zoonotic influenza treatment. Previous reports on mAb delivery are mainly focused on seasonal influenza, including H1N1, H1N1pdm09, and H3N2 viruses. Of particular note, human infections with zoonotic H7 viruses have distinct clinical manifestations compared to seasonal influenza. H7N9 infection has usually presented with severe viral pneumonia, and some cases were complicated by ARDS and multiorgan failure. Unlike patients with seasonal or H5N1 influenza, patients with H7N9 disease were more likely to be older and have underlying comorbidities. Similar clinical features were also observed for the H7N4 human infection case. Therefore, it is essential to assess the efficacy of anti-H7 antibodies administered through airway delivery. A previous report analyzed the efficacy of a neutralizing antibody (Mab 62) against the H7N7 subtype AIV via the i.n. and i.p. routes (26), while our study presented more systemic findings compared to that study. First, the mAb 4B7 and humanized antibody were assessed in a more clinically relevant manner. The current standard-of-care anti-influenza therapy, oseltamivir, is usually administered 48 h post-exposure or post-infection. The Mab 62 was administered through the i.p. or i.n. route at 24 h before H7N7 virus infection, whereas the efficacy of i.n. delivery of the humanized 4B7 at 48 and 72 h prior to H7N9 infection was determined. In a therapeutic setting, one dose of the Mab 62 was given at 24 h after H7N7 virus infection, and the efficacy of the murine and humanized 4B7 was assessed at 48 and 72 h p.c. The humanized antibody administered via the i.n. route at 72 h (0.3 mg/kg) before infection or 48 h (10 mg/kg) post infection provided good protection, indicating its potential as antiviral prophylaxis and therapy against H7 influenza infection. Moreover, the Mab 62 recognizes a highly conserved epitope (K715) in H7 HAs and has cross-HI and VN activities with divergent H7 viruses, whereas cross-protection against H7 viruses was not determined. In this study, the humanized antibody chi4B7 conferred cross-protection against infection with different H7 viruses in mice, highlighting its potential to be used as a broad antiviral agent against H7 viruses.
For passive antibody therapy, the immunogenicity of murine antibodies in humans should be minimized to reduce potential side effects. Antibody humanization is a common approach for this purpose, and it is critical to preserve antigen-binding activities in this process. Herein, grafting the variable regions onto a human IgG1 backbone was performed for 4B7 humanization, causing no alterations in antigen binding, HI, and VN activities of the antibody. More importantly, the humanized antibody provided prophylactic and therapeutic protection against various H7 viruses via i.n. administration. Nevertheless, the antibody generated through variable region grafting is not a full humanized antibody. To further reduce anti-drug antibodies against murine antibody components, new humanization methods, including CDR grafting, specificity-determining residues grafting, or framework (FR) shuffling, can be employed (38). However, antigen affinity of the humanized antibodies generated with these strategies should be carefully evaluated due to incompatibility between human FRs and mouse CDRs.
The mAb 4B7 footprint covers crucial domains in the RBS (130-loop and 150-loop) and a major antigenic region (site A). The antigenic site A is highly conserved in H7 HAs and is a dominant target of neutralizing antibodies. Interestingly, the G151E mutation in the antigenic site A mediated H7N9 virus escape from 4B7 neutralization. Mutation at G151 is also associated with immune escape of H7N9 viruses from other neutralizing antibodies (30,(39)(40)(41). A recent report revealed that an H7N9 field isolate (SD001) with a natural G151E variation in the HA can escape from two anti-H7N9 neutralizing antibod ies, strengthening the critical role of G151 in affecting antibody neutralization (32). Moreover, variations in other residues in the antigenic site A also impact the reactivity of antibodies with H7 HA. The mAb 4B7 showed significantly lower neutralizing and HI activities against an EA lineage isolate H7N2-DK2007 with K148 and undetectable neutralizing and HI activities against an NA lineage isolate PR8/felH7N2 with T148. Intriguingly, the reactivity of the antibody with PR8/felH7N2 was detected in IFA, which supported 4B7-H7N2 HA interaction in molecular docking. It is noted that 4B7 turned to recognize a new epitope in the HA stalk due to the presence of T148 in the antigenic site A. Epitope switching may lead to reduced neutralizing and HI activities but no impacts on HA binding. However, further experiments are required to verify the docking results, and it is interesting to investigate why mutations in residues 151 and 148 exert distinctive effects on 4B7 activities.
Many previously reported murine and human anti-H7 mAbs select escape mutant viruses with mutations at five key residues in the antigenic site A (25, 30,[41][42][43][44]. In terms of neutralizing mechanisms of 4B7, the antibody processes HI activity against H7 viruses, indicating that it can block virus attachment to cell receptors. This may be associated with recognition of the 130-and 150-loop in the RBS. Of note, although the antigenic site A is outside the RBS, antibodies targeting this region can also inhibit virus binding to the receptors. A pan-H7 human mAb (rH7-235) recognizes residues in the antigenic site A outside of the RBS but near the edge of the RBS (45). rH7-235 IgG potently neutralizes H7N9 viruses and protects against H7N9 infection due to avidity effect and Fc steric hindrance. However, the role of critical domains targeted by 4B7 in virus neutralizing activity and protection needs to be determined by further studies. Therefore, variation of the five residues in the antigenic site A may affect activities of anti-H7 antibodies, and antibodies targeting this region may neutralize H7 viruses through distinct mechanisms, informing the rational design of therapeutics and vaccines against the H7 subtype influenza viruses.
Currently, the wide spreading of the H5N1 subtype AIV (clade 2.3.4.4b) worldwide and its spillover to mammals and humans poses a great challenge to public health. Besides the H5 subtype, avian influenza outbreaks caused by different H7 subtypes have occurred in five countries in the last century, and a total of 1,687 human cases with H7 influenza virus infection have been documented from 1959 to 2019 (1). Recently, several poultry outbreaks associated with highly pathogenic H7 subtype AIVs were reported in different countries, highlighting the pandemic potential of the H7 subtype AIVs (7,8). The murine and humanized antibodies reported herein target conserved epitopes in the H7 subtype, which afford cross-reactivity of the antibodies with multiple H7 viruses. More importantly, the humanized 4B7 conferred broad protection against diverse H7 viruses when administered via the i.n. route. Therefore, the humanized antibody can serve as a broad antiviral therapy against the H7 subtype influenza.
In conclusion, we generated and characterized murine and humanized antibodies with cross-H7 binding, HI, and neutralizing activities. Airway delivery of the human ized antibody provided broad prophylactic and therapeutic protection against diverse H7 viruses. These features render the humanized antibody a favorable candidate for prophylaxis and therapy against the H7 subtype influenza. Moreover, epitope profiling of the antibodies also presented novel insights into understanding antibody response to H7 viruses and for vaccine design.
## MATERIALS AND METHODS
## Viruses, plasmids, proteins, and cells
The mouse-adapted H7N9 subtype AIV strain maSDL124 derived from A/chicken/ Shandong/SDL124/2015 (Genbank accession numbers: MW397099 to MW397114) and the reassortant H7N9 virus GD15 derived from a highly pathogenic strain (A/chicken/ Guangdong/GD15/2016) (EPI_ISL_305597) were generated previously (46,47). The H7N2 subtype AIV of duck origin (H7N2-DK2007) was reported previously (48) and obtained from Prof. Liping Yan (Key Animal Virology Laboratories of the Ministry of Agriculture and Rural Affairs of China). The viruses were propagated in 10-day-old specific pathogen-free (SPF) embryonated chicken eggs (ECEs). The plasmids encoding the six internal proteins of PR8 were used for generation of the reassortant influenza viruses. The HA protein of the GD15 strain was expressed in the baculovirus expression system (31). Madin-Darby canine kidney (MDCK) cells and human embryonic kidney 293T cells were cultured in Dulbecco's modified Eagle medium (DMEM) (ThermoFisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (FBS) (LONSERA, Suzhou, China) at 37°C with 5% CO2. CHO cells were cultured in DMEM/F12 medium (ThermoFisher Scientific) supplemented with 10% FBS at 37°C with 5% CO2.
## MAb screening and generation
MAbs against the H7N9 HA were generated using the hybridoma method previously, but only one antibody with a low VN titer was obtained (31). In that study, a total of 8 mice were immunized with the HA protein, and spleen cells were collected for fusion. The hybridomas derived from 5 mice were screened, and the remaining hybridomas were stored in liquid nitrogen. Herein, the frozen hybridomas were recovered and screened using ELISA and microneutralization assay. The positive clones were then subcloned using the limiting dilution method. Positive hybridomas were intraperitoneally injected into BALB/c mice. Ascites was collected, and antibody purification was performed using protein G affinity chromatography (GE Healthcare, Piscataway, NJ, USA).
## ELISA
ELISA was performed as previously reported (49). The HA protein (0.25 µg/mL) or the H7N9 viruses at 32 hemagglutination units (HAU) was immobilized onto 96-well plates in a carbonate buffer and incubated at 4°C overnight. The plates were washed three times with PBST (PBS with 0.05% Tween-20) and then blocked with 200 µL of 5% (wt/vol) skim milk in PBST at 37°C for 1 h. The plates were washed three times with PBST and then incubated with the antibodies at serial concentrations at 37°C for 1 h (100 µL/well). After washing, 100 µL of horseradish peroxidase-conjugated goat anti-mouse or -human IgG antibodies (1:5,000) was added and incubated for 1 h at 37°C. The plates were washed with PBST three times and developed with 100 µL of 3,3' ,5,5'-tetramethylbenzi dine substrate at room temperature (RT) for 15 min. Then, 50 µL of 3 M HCl was added to stop the reaction, and the absorbance at 450 nm was measured.
## HI test
HI tests were conducted according to the protocol of World Organization of Animal Health with some modifications. In brief, the antibodies were initially diluted by 100-fold, followed by 2-fold serial dilutions. The H7 viruses (4 HAU) were used as the antigens. HI titers were defined as the minimal concentrations of the antibodies required to completely inhibit the hemagglutination activity of the viruses.
## Microneutralization assay
MDCK cells at a density of 30,000 were seeded onto 96-well plates and cultured overnight to reach a confluency of 80%. The antibodies were 10-fold serially diluted and incubated with 100 TCID 50 of H7 viruses at 37°C for 1 h. DMEM containing 1 µg/ml of tosyl-sulfonyl phenylalanyl chloromethyl ketone (TPCK)-treated trypsin was used for dilution of the viruses and antibody. The antibody-virus mixtures were transferred to MDCK cells and cultured for 48 h. The supernatants were then collected, and virus infection was determined using the hemagglutination test. The IC 50 was calculated using nonlinear regression analysis with the GraphPad Prism (GraphPad Software, San Diego, CA).
## Antibody sequencing and molecular docking
RNA was extracted from the 4B7 hybridomas and then transcribed into cDNA. Antibody VH and VL genes were amplified using PCR and sequenced. Genetics composition of the antibody 4B7 was analyzed using the IMGT/V-QUEST program (https://www.imgt.org/ IMGT_vquest/input). Homology models of the 4B7 scFv and the H7 HA protein struc tures were generated using the ABodyBuilder-ML server (https://opig.stats.ox.ac.uk/ webapps/sabdab-sabpred/sabpred/abodybuilder/) and SWISS-MODEL server (https:// swissmodel.expasy.org), respectively. Molecular docking of the 4B7 scFv and the HA proteins was performed using the ZDOCK server (https://zdock.wenglab.org/), and the complex was visualized and handled using the PyMOL software. The interaction surface of the scFv-HA complex was analyzed on the web PISA server (https://www.ebi.ac.uk/ pdbe/pisa/).
## Escape mutant studies
To identify the epitope recognized by 4B7, escape mutants derived from the H7N9 GD15 strain were generated as reported previously (50). Briefly, 10 6 50% embryo infectious dose (EID 50 ) of the virus in 0.5 mL was mixed with the antibody at 20 µg/mL and was incubated at RT for 1 h. Then, 0.1 mL of the mixture was inoculated into 10-day-old SPF ECEs. After incubation for 4 days, the allantoic fluids were collected for hemagglutination assay. These procedures were repeated twice by incubating serial dilutions (10 5 to 10 8 ) of the allantoic fluids with the antibody. Subsequently, HI titers of the antibody against the allantoic fluids were measured, and the viruses were defined as the escape mutants when HI titers of the mAb reduced by at least 8-fold compared to that against the parental virus. The HA gene of the escape mutants was sequenced.
## Generation of the reassortant H7 viruses
To validate the results of escape mutant studies, the reassortant H7N9 viruses carry ing the G151E and I335V mutations in HA were generated. In addition, the HA and NA genes of the H7N9 (A/Guangdong/17SF003/2016, SF003) (EPI_ISL_280902), H7N2 (A/feline/New York/16-040082-1/2016, fel2016) (EPI_ISL_260817), H7N3 (A/Mex ico/InDRE7218/2012, Mex2012) (EPI_ISL_128317), and H7N4 (A/chicken/Jiangsu/1/2018, JS2018) (EPI_ISL_332358) viruses were synthesized and cloned into the pHW2000 vector. Notably, none of the HA sequences harbored a polybasic cleavage site. The 6:2 reassor tant viruses based on the PR8 internal genes were generated using reverse genetics.
For virus rescue, co-cultures of MDCK and 293T cells were co-transfected with the plasmids. These plasmids contained the HA and NA genes, along with six internal genes of PR8 (400 ng of each plasmid). The TransIT-X2 Dynamic Delivery System (Mirus, Madison, WI, USA) was used for transfection purposes. At 24 h post-transfection, TPCK-treated trypsin (1 µg/mL) was added to the cell culture. At day 3 post-transfection, both the cells and supernatants were collected. Subsequently, 0.3 mL of this mixture was subjected to freeze-thaw cycles and then inoculated into 10-day-old SPF ECEs. The presence of the virus was detected using the hemagglutination assay.
## Pathogenicity of the reassortant H7 viruses in a PR8 backbone in mice
Morbidity and mortality of the rescued reassortant H7 viruses in the PR8 backbone in mice were determined. Groups of 6-to 8-week-old female BALB/c mice (n = 5) were intranasally inoculated with the PR8/H7N3 and PR8/H7N4 viruses at 10 6.0 , 10 5.0 , 10 4.0 , 10 3.0 , and 10 2.0 EID 50 . Another group of mice (n = 5) was inoculated with PBS as the non-infected mock control. Clinical signs and body weight were monitored daily for 14 days, and mice that lost 25% or more of their initial body weight were euthanized and scored dead. The MLD 50 was calculated.
## Prophylactic and therapeutic efficacy of the mAb 4B7 against H7N9 virus in mice via intraperitoneal administration
Passive transfer studies were performed to investigate the prophylactic and therapeutic efficacy of the mAb 4B7 in mice. To evaluate the prophylactic efficacy, 6-to 8-week-old female BALB/c mice (n = 11) were inoculated with the antibody at 30, 20, 10, and 5 mg/kg via the i.p. route. Another group of mice (n = 11) was inoculated with PBS as the negative control. Naïve mice (n = 5) were used as the non-infected mock control. After 2 h, the mice were anesthetized by i.p. injection with 0.1 mL of pentobarbital sodium and then inoculated with 10 MLD 50 of maSDL124 virus via the i.n. route. Body weight was monitored daily for 14 days, and mice that lost 25% or more of their initial body weight were euthanized and scored dead. At days 3 and 5 p.c., three mice from the virus-infected groups were euthanized, and the lungs were collected for virus titration and histopathologic examination.
Moreover, in a therapeutic setting, 6-to 8-week-old female BALB/c mice (n = 5) were inoculated with 10 MLD 50 of maSDL124 through the i.n. route and then treated with the antibody at 30, 20, 10, and 5 mg/kg at 12 and 24 h p.c. via the i.p. administration. The mice were monitored daily for 14 days, and weight loss and survival were recorded.
## Prophylactic and therapeutic efficacy of the mAb 4B7 against H7N9 virus in mice via intranasal administration
For prophylactic studies, the antibody at 3, 1, 0.3, and 0.1 mg/kg was administered to 6to 8-week-old female BALB/c mice (n = 18) via the i.p. route for systematic administration or via the i.n. route for topical administration. Another group of mice (n = 18) was inoculated with PBS as the negative control, and eight naïve mice were used as the non-infected mock control. After 2 h, the mice were anesthetized and then infected with 10 MLD 50 of maSDL124. Weight loss and survival of the mice were monitored for 14 days. At days 3 and 5 p.c., 5 mice from the virus-infected groups were euthanized, and the lungs were harvested for virus titration and histopathologic analysis. In a therapeutic setting, groups of mice (n = 5) were infected with 10 MLD 50 of maSDL124 and then treated with 4B7 at 10, 3, and 1 mg/kg via the i.p. or i.n. route at 48 and 72 h p.c. Body weight and survival were monitored for 14 days.
## Expression and characterization of the humanized antibody chi4B7
To generate the humanized antibody, the VH and VL segments of the mAb 4B7 were PCR-amplified and then individually cloned into the pCDNA3.4 expression vector encoding human IgG1 constant regions. The heavy-and light-chain expression plasmids were co-transfected into CHO cells. At day 7 post-transfection, the supernatants were harvested and centrifuged at 5,000 rpm for 20 min, followed by filtration through 0.45 µm filter apparatus to remove remaining cell debris. Antibodies were purified from the supernatants using Protein G columns (GE Healthcare). The concentration of the purified antibody was measured, and the antibody was assessed by sodium dodecyl sulfate polyacrylamide gel electrophoresis under the reducing and non-reduc ing conditions. In addition, binding, HI, and VN activities with H7 viruses of the human ized antibody were measured.
## Prophylactic and therapeutic efficacy of the humanized antibody against divergent H7 viruses in mice
Protective efficacy of the humanized antibody (chi4B7) against different H7 viruses was assessed in mice in prophylactic and therapeutic settings. In prophylactic studies, groups of 6-to 8-week-old female BALB/c mice (n = 11) were administered with chi4B7 at 3, 1, and 0.3 mg/kg via the i.n. route. Another group of mice (n = 11) was inoculated with PBS as the negative control, and five naïve mice were used as the non-infected mock control. The mice were inoculated with 10 MLD 50 of the maSDL124, PR8/H7N4, and PR8/H7N3 viruses 2 h after antibody administration. Weight loss and survival were monitored for 14 days. At days 3 and 5 p.c., 3 mice from the virus-infected groups were euthanized, and the lungs were harvested for virus load measurement and histopathologic analysis. For therapeutic experiments, groups of mice (n = 8) were first i.n. inoculated with 10 MLD 50 of the maSDL124 and PR8/H7N4 viruses and administered with chi4B7 at 10 and 3 mg/kg through the i.n. route 48 h later. Body weight and survival of the mice were monitored for 14 days. At day 5 p.c., 3 mice from the virus-infected groups were euthanized, and the lungs were collected for virus titration and histopathologic examination. The data of the mock control, virus load, and lung pathology at day 5 p.c. of the PBS-inoculated mice were shared by the prophylactic and therapeutic experiments. In addition, to assess the effective prophylactic window of airway-delivered chi4B7, groups of mice (n = 7) were i.n. administered with the antibody at 3 and 0.3 mg/kg. Another group of mice (n = 7) was inoculated with PBS as the negative control, and five naïve mice were used as the non-infected mock control. The mice were infected with the H7N9 maSDL124 virus at 48 or 72 h post antibody administration, and body weight and survival were monitored for 14 days.
## Quantitative real-time PCR
Since PR8/H7N4 had a low virus titer in MDCK cells (Table S4) and virus in the lung samples was not isolated in MDCK cells, quantitative real-time PCR was used to measure NP gene copies in the lungs. A plasmid containing the PR8 NP gene was used as the standard. The standard plasmid was serially diluted, and the samples containing the NP gene at 10 1 -10 7 copies/µL were used in PCRs to generate the standard curve. Total RNA was extracted from the tissue samples using TRIzol reagent (Vazyme) and then reverse-transcribed into cDNA using the HiFiScript SuperFast gDNA Removal cDNA Synthesis Kit (CWBIO, Taizhou, China). The system of PCR was composed of 1 µL of the tissue cDNA or the standards, 0.4 µL of the primers, 0.2 µL of the probe, and 10 µL of 2 × AceQ qPCR Probe Master Mix (Vazyme). The sequences of the primers and probe were as follows: forward primer, 5′-AAGGTGGTCCCAAGAGGGA-3′; reverse primer, 5′-GCTGCCA TAACGGTTGTTCTG-3′; probe, FAM-TACAGAGAAATCT-MGB. PCRs were conducted using the CFX Connect Real-Time System with the following cycles: one cycle for denaturing at 95 °C for 5 min, 40 cycles for PCR at 95 °C for 10 s and 60 °C for 30 s. The standard curve was generated using the Bio-Rad CFX Maestro software, and the NP gene copies were calculated based on the standard curve.
## Statistical analysis
For ELISA, HI, and VN assays, mean values ± standard error of the mean (SEM) of at least two independent experiments were shown. For body weight and survival, mean values ± SEM of at least five mice per group were shown. Virus loads and lung lesion scores of three or five individual animals were shown as mean values ± SEM. The data were analyzed using one-way ANOVA in GraphPad Prism software. Multiple comparisons were conducted by comparing the mean of each group with every other group using Tukey's test for multiple comparison correction.
## FUNDING
## References
1. Shi, Zeng, Cui et al. (2023) "Alarming situation of emerging H5 and H7 avian influenza and effective control strategies" *Emerging Microbes Infections*
2. Ison, Marrazzo (2025) "The emerging threat of H5N1 to human health" *N Engl J Med*
3. 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"
4. Caserta, Frye, Butt et al. (2024) "Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle" *Nature*
5. Chang, Sadeyen, Bhat et al. (2023) "Risk assessment of the newly emerged H7N9 avian influenza viruses"
6. Yin, Deng, Zeng et al. (2021) "Genetic and biological properties of H7N9 avian influenza viruses detected after application of the H7N9 poultry vaccine in China" *PLoS Pathog*
7. Luczo, Spackman (2025) "Molecular evolution of the H5 and H7 highly pathogenic avian influenza virus haemagglutinin cleavage site motif" *Rev Med Virol*
8. Ahrens, Pohlmann, Grund et al. (2024) "Out of the blue: detection of a unique highly pathogenic avian influenza virus of subtype H7N5 in Germany" *Emerging Microbes Infections*
9. Hou, Deng, Cui et al. (2024) "Evolution of H7N9 highly pathogenic avian influenza virus in the context of vaccination" *Emerging Microbes Infections*
10. (2025) *Full-Length Text Journal of Virology*
11. Tong, Weng, Xue et al. (2018) "First human infection by a novel avian influenza A(H7N4) virus" *J Infec*
12. Chai, Li, Li et al. (2022) "Emergence, evolution, and pathogenicity of influenza A(H7N4) virus in shorebirds in China" *J Virol*
13. Bart, Hohenboken, Cioppa et al. (2014) "A cell culture-derived MF59-adjuvanted pandemic A/H7N9 vaccine is immunogenic in adults" *Sci Transl Med*
14. Hörner, Fiedler, Bodmer et al. (2023) "A protective measles virus-derived vaccine inducing long-lasting immune responses against influenza A virus H7N9" *NPJ Vaccines*
15. Wang, Li, Peng et al. (2024) "Dual N-linked glycosylation at residues 133 and 158 in the hemaggluti nin are essential for the efficacy of H7N9 avian influenza virus like particle vaccine in chickens and mice" *Vet Microbiol*
16. Kamal, Blanchfield, Belser et al. (2017) "Inactivated H7 influenza virus vaccines protect mice despite inducing only low levels of neutralizing antibodies" *J Virol*
17. Jackson, Campbell, Frey et al. (2015) "Effect of varying doses of a monovalent H7N9 influenza vaccine with and without AS03 and MF59 adjuvants on immune response: a randomized clinical trial" *JAMA*
18. To, Chan, Yuen (2014) "Viral lung infections: epidemiology, virology, clinical features, and management of avian influenza A(H7N9)"
19. Zheng, Tang, Gao et al. (2018) "Benefit of early initiation of neuraminidase inhibitor treatment to hospitalized patients with avian influenza A(H7N9) virus"
20. Sato, Takashita, Katayose et al. (2021) "Detection of variants with reduced baloxavir marboxil and oseltamivir susceptibility in children with influenza A during the 2019-2020 influenza season" *J Infect Dis*
21. Mcbride, Lim, Burgess et al. (2017) "Phase 2 randomized trial of the safety and efficacy of MHAA4549A, a broadly neutralizing monoclonal antibody, in a human influenza A virus challenge model" *Antimicrob Agents Chemother*
22. Plotnik, Sager, Aryal et al. (2024) "A phase 1 study in healthy volunteers to investigate the safety, tolerability, and pharmacokinetics of VIR-2482: a monoclonal antibody for the prevention of severe influenza A illness" *Antimicrob Agents Chemother*
23. Jin, Lei, Hu et al. (2017) "Human monoclonal antibodies as candidate therapeutics against emerging viruses" *Front Med*
24. Jia, Zhao, Morano et al. (2024) "Human neutralizing antibodies target a conserved lateral patch on H7N9 hemagglutinin head" *Nat Commun*
25. Yu, Song, Wu et al. (2016) "A potent germline-like human monoclonal antibody targets a pH-sensitive epitope on H7N9 influenza hemagglutinin"
26. He, Kumar, Khader et al. (2013) "Effective intranasal therapeutics and prophylactics with monoclonal antibody against lethal infection of H7N7 influenza virus" *Antiviral Res*
27. Vigil, Frias-Staheli, Carabeo et al. (2020) "Airway delivery of anti-influenza monoclonal antibodies results in enhanced antiviral activities and enables broad-coverage combination therapies" *J Virol*
28. Leyva-Grado, Tan, Leon et al. (2015) "Direct administration in the respiratory tract improves efficacy of broadly neutralizing anti-influenza virus monoclonal antibodies" *Antimicrob Agents Chemother*
29. Marasco, Sui (2007) "The growth and potential of human antiviral monoclonal antibody therapeutics" *Nat Biotechnol*
30. Li, Yang, Wang et al. (2019) "Rapid isolation of a potent human antibody against H7N9 influenza virus from an infected patient" *Antiviral Res*
31. Hu, Zhao, Shi et al. (2021) "Identification of the dominant non-neutralizing epitope in the haemagglutinin of H7N9 avian influenza virus"
32. Zhao, Sun, Wang et al. (2025) "Structural basis of different neutralization capabilities of monoclonal antibodies against H7N9 virus" *J Virol*
33. Kuiken, Taubenberger (2008) "Pathology of human influenza revisited" *Vaccine (Auckl)*
34. Prince, Hemming, Horswood et al. (1987) "Effectiveness of topically administered neutralizing antibodies in experimental immunotherapy of respiratory syncytial virus infection in cotton rats" *J Virol*
35. Halwe, Kupke, Vanshylla et al. (2021) "Intranasal administra tion of a monoclonal neutralizing antibody protects mice against SARS-CoV-2 infection" *Viruses*
36. Friesen, Koudstaal, Koldijk et al. (2010) "New class of monoclonal antibodies against severe influenza: prophylactic and therapeutic efficacy in ferrets" *PLoS One*
37. Paules, Lakdawala, Mcauliffe et al. (2017) "The hemagglutinin A stem antibody MEDI8852 prevents and controls disease and limits transmission of pandemic influenza viruses" *J Infect Dis*
38. Wang, Chen, Xu et al. (2024) "Comparison of "framework Shuffling" and "CDR grafting" in humanization of a PD-1 murine antibody" *Front Immunol*
39. Chang, Lukosaityte, Sealy et al. (2023) "Antigenic characterization of human monoclonal antibodies for therapeutic use against H7N9 avian influenza virus" *J Virol*
41. Ito, Yamayoshi, Murakami et al. (2019) "Characterization of mouse monoclonal antibodies against the HA of A" *Viruses*
42. Li, Chen, Wang et al. (2019) "A cross-reactive human monoclonal antibody targets the conserved H7 antigenic site A from fifth wave H7N9-infected humans"
43. (2025) *Full-Length Text Journal of Virology*
44. Thornburg, Zhang, Bangaru et al. (2016) "H7N9 influenza virus neutralizing antibodies that possess few somatic mutations" *J Clin Invest*
45. Tan, Leon, Albrecht et al. (2016) "Broadly-reactive neutralizing and non-neutralizing antibodies directed against the H7 influenza virus hemagglutinin reveal divergent mechanisms of protection" *PLoS Pathog*
46. Schmeisser, Vasudevan, Verma et al. (2015) "Antibodies to antigenic site A of influenza H7 hemagglutinin provide protection against H7N9 challenge"
47. Gilchuk, Dong, Irving et al. (2025) "Pan-H7 influenza human antibody virus neutralization depends on avidity and steric hindrance" *JCI Insight*
48. Wang, Liu, Hu et al. (2021) "Mutations during the adaptation of H7N9 avian influenza virus to mice lungs enhance human-like sialic acid binding activity and virulence in mice" *Vet Microbiol*
49. Liu, Zhang, He et al. (2018) "Characteristics of the emerging chicken-origin highly pathogenic H7N9 viruses: a new threat to public health and poultry industry" *J Infect*
50. Yao, Chen, Wang et al. (2020) "Identification of antigenic epitopes in the haemagglutinin protein of H7 avian influenza virus" *Avian Pathol*
51. Hu, Zhao, Zhao et al. (2020) "Hemagglutinin-specific non-neutralizing antibody is essential for protection provided by inactivated and viralvectored H7N9 avian influenza vaccines in chickens" *Front Vet Sci*
52. Sun, Shi, Meng et al. (2018) "Development of a colloidal gold-based immunochromatographic strip for rapid detection of H7N9 influenza viruses" *Front Microbiol*
53. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12270885&blobtype=pdf | # Study of the 16S microbiome of swans died during the H5N1 outbreak in the Caspian seashore
Kobey Karamendin, Sardor Nuralibekov, Temirlan Sabyrzhan, Yermukhammet Kasymbekov, Symbat Suleymenova, Aidyn Kydyrmanov
## Abstract
In 2023 and 2024, mass mortalities of swans occurred on the Caspian coast of Kazakhstan, which affected more than seven hundred birds of a local population of 10-15 thousand. It is widely known that viral infections significantly affect the microbiome content of various organisms, but the influence of H5N1 infection in the gut microbiota of wild birds remains little studied. Almost no information is available on postmortem microbial changes after the devastating impact of H5N1 influenza.Methods: In addition to standard routine virological studies, we were interested in investigating the microbiological changes resulting from infection with the highly pathogenic H5N1 using 16S rRNA gene sequencing.Results: Virological studies of samples taken from the dead swans identified the highly pathogenic influenza virus H5N1 subtype as the primary cause of mortality. 16S analysis of samples from freshly dead swans revealed patterns of microbial dysbiosis caused by the overwhelming dominance of Campylobacter and Fusobacterium genera in the microbiome.Discussion: Unlike previous fecal microbiome studies in live H5N1-infected birds, this is the first post-mortem analysis revealing systemic dysbiosis across respiratory and digestive tracts in swans, dominated by Campylobacter (mean 74.7% ± 19.3) and Fusobacterium (mean 15.9% ± 12.2).
## Introduction
Influenza virus H5N1 is a highly pathogenic subtype that has a devastating effect on wild bird populations. This virus, which primarily infects the respiratory and gastrointestinal tracts of birds, has been identified in numerous wild bird species globally. Swans are particularly susceptible to H5N1, and infected birds often exhibit severe listlessness and neurological dysfunction consisting of seizures, tremors, and marked incoordination (1). Swans' migratory behavior further exacerbates the virus's dissemination, as they can carry and shed the virus over long distances, impacting other bird populations and ecosystems.
For the first time in Kazakhstan, a highly pathogenic avian influenza H5N1 outbreak significantly impacted wild swan populations in 2006 (2). This outbreak coincided with the seasonal migration of wild birds in the Caspian Sea, likely facilitating the virus's spread across regions (3). Furthermore, mass mortality among swans caused by H5N1 was observed at Lake Karakol (Caspian seashore) in Kazakhstan during the winter of 2023 and 2024 (4). The Caspian Sea is not only an important crossing point for the migratory routes of wild birds, but also serves as a nesting and wintering area. Mass mortalities of gulls and terns regularly occur in this region due to H5N1 (5) and adenoviruses (6).
It is well established that infection with pathogenic viruses induces significant changes in the microbiome of the affected organism (7). Avian influenza virus (AIV) infection has been shown to reduce diversity and alter dominant taxa in the gut microbial composition in wild birds and poultry (8). In H5N1-infected swans, severe necrosis in multiple organs, along with lymphoid depletion in Peyer's patches, which play a crucial role in the immune system, was observed (9).
In this study, we aimed to investigate the state of the swan microbiome following the lethal impact of highly pathogenic H5N1 in 2023 and 2024. The effect of H5N1 infection on the gut microbiota of migratory birds has been minimally studied (10); such research is crucial for understanding the microbiological processes that occur during epidemics.
## Materials and methods
## Sample collection
Swab samples were collected from swans (Cygnus olor and Cygnus cygnus) that died from confirmed H5N1 infection in December 2023 and December 2024 (Table 1). The samples were obtained from freshly deceased swans that died the same day or 1 to 2 days prior. The air temperature during sampling ranged from 0 to 2°C, and the carcass condition was assessed as very fresh. Two types of swab samples were obtained: cloacal and tracheal from juveniles and adults. The samples were collected using sterile swabs, placed in vials with transportation media, and stored for 1-2 days in liquid nitrogen at -195°C until delivery to the laboratory. In the laboratory, the samples were placed in a freezer and stored at -80°C until library preparation.
## 16S rRNA gene amplification and sequencing
Microbial DNA was extracted using the PureLink Microbiome DNA Purification kit (Invitrogen, USA). The concentration of extracted DNA was assessed with a Qubit 4 fluorometer (Invitrogen, USA). The V3-V4 region of the 16S rRNA gene was amplified using the primer pair containing Illumina adaptors, following the PCR thermal cycling conditions recommended by the manufacturer (Illumina, USA). Amplicons were verified via agarose gel electrophoresis and purified with AxyPrep Mag PCR Clean-Up beads (Axygen, USA). Purified amplicons were indexed using the Illumina Nextera XT Index Kit, followed by an additional round of purification. Libraries were quantified using the Qubit 1x dsDNA HS Assay Kit (Invitrogen, USA) and pooled in equimolar concentrations. Sequencing was performed using the MiSeq v.3 (2 × 300 bp) chemistry kit. Adapter removal was carried out by the sequencer's software. Generated raw 16S rRNA gene sequences were uploaded to the Sequence Read Archive (BioProject ID: PRJNA1240282).
## Bioinformatic analysis
The DADA2 pipeline in RStudio was used for quality filtering, error correction, and amplicon sequence variant (ASV) inference. Taxonomic classification was performed using the SILVA 138.1 ribosomal RNA database (silva_nr99_v138.1_wSpecies_train_set.fa). Relative abundances of bacterial genera were visualized using ggplot2 in RStudio, and statistical differences between swab types were analyzed using the Wilcoxon rank-sum test. Alpha diversity metrics (Shannon and Simpson indices) and beta diversity (Bray-Curtis dissimilarity) were computed using the phyloseq package in R. Principal Coordinates Analysis (PCoA) was performed to assess community composition differences. Statistical tests were performed using PERMANOVA in the adonis2 function of the R Vegan package
## Results
Taxonomic composition of the affected swan microbiome
Analysis of taxonomic composition at the genus level revealed an obvious dominance of specific bacterial taxa across the dataset (Figure 1). Campylobacter emerged as the overwhelmingly predominant genus in both cloacal and tracheal samples, followed by Fusobacterium as the second most abundant taxon (Supplementary Table S1). Other consistently prevalent bacteria included Streptobacillus and Bacteroides, which are commonly associated with gut and mucosal microbiota. This pattern of dominance remained remarkably consistent in all sixteen examined samples, suggesting a systemic influence of these bacteria throughout the respiratory and digestive systems of the infected swans.
The visualization of relative abundances through stacked bar plots confirmed this taxonomic profile, with Campylobacter (represented in mustard yellow) constituting the majority of the bacterial community in all sample types, while other genera were present in substantially lower proportions. This consistent dominance across sample types represents a significant finding, as it suggests that H5N1 infection may facilitate similar patterns of microbial dysbiosis regardless of the physiological environment within the host.
## Microbial diversity patterns in H5N1-infected swans
The analysis of alpha diversity metrics revealed distinct patterns in the microbial communities of swans that were affected by H5N1 infection. Both Shannon and Simpson diversity indices consistently demonstrated that cloacal samples harbored more diverse microbial communities compared to tracheal samples across the dataset. The Shannon index, which accounts for both richness and evenness of taxa, showed elevated values in cloacal samples regardless of the swan's age category. Similarly, Simpson diversity values approached 1 for most samples, indicating communities with relatively balanced compositions despite being dominated by a few key taxa.
Age-related differences were apparent in the diversity metrics, with juvenile swans exhibiting a notably wider range of diversity values compared to adult birds. This variability suggests that younger swans potentially host more heterogeneous microbial communities, which could be attributed to differences in immune system development or varied environmental exposures prior to infection.
## FIGURE 1
Relative abundance at the genus level. The most abundant genus appears to be one represented in brownish-green. Other noticeable genera include those in green, pink, and blue, but they are present in lower proportions. n = 16 samples (9 cloacal, 7 tracheal; 11 adults, 5 juveniles).
Interestingly, when comparing samples collected in 2023 versus 2024, no pronounced year-based clustering was evident, indicating relatively stable microbial diversity patterns across the 2 years of the outbreak despite potential environmental or viral strain variations (Figure 2).
## Community composition and beta diversity analysis
Principal Coordinates Analysis (PCoA) based on the Bray-Curtis dissimilarity matrix revealed significant patterns in community structure across various sample categories. The visualization showed moderate separation between cloacal and tracheal samples, indicating that the sampling site is a major driver of microbial community differences in H5N1-infected swans. This anatomical differentiation was more pronounced than any clustering based on age categories, as juvenile and adult samples exhibited considerable overlap in the ordination space (Figure 3).
PERMANOVA testing provided statistical confirmation of these observed patterns. The analysis revealed that swab type was indeed a significant factor (F = 2.337, R 2 = 0.143, p = 0.01), explaining approximately 14.3% of the variation in microbial composition across
## Heatmap visualization of microbial abundances
The hierarchical clustering analysis visualized through a heatmap further illustrated the patterns of microbial distribution across samples (Figure 4). The color gradient, ranging from red (highest abundance) through yellow (moderate abundance) to blue (low abundance), effectively demonstrated that again, Campylobacter and Fusobacterium genera maintained high abundance across most samples while others showed more variable or sample-specific patterns. Log10 transformation of abundance values normalized the data and highlighted differences in relative abundance across the dataset. The heatmap clustering suggested that while anatomical sampling sites (cloacal versus tracheal) influenced overall community structure, certain bacterial signatures were consistently associated with H5N1 infection regardless of other variables. This observation supports the hypothesis that avian influenza infection may drive predictable shifts in the avian microbiome, potentially contributing to disease pathogenesis. This heatmap successfully highlights microbiome composition differences across samples.
## Discussion
## Predominant bacterial genera in H5N1-infected swans
The most striking finding of this study was the overwhelming dominance of Campylobacter and, to a lesser extent, Fusobacterium Campylobacter and Fusobacterium emerged as the predominant genera in all examined samples, suggesting a systemic spread throughout the birds' bodies. Their significant presence in all samples raises questions about their potential involvement in tissue damage following H5N1 infection. Supposedly, high abundances of Campylobacter and Fusobacterium in the swan samples are a consequence of dysbiosis rather than an increase in their pathogenicity. Campylobacter species are known to inhabit avian intestinal tracts, often as commensals, but typically represent a much smaller proportion of the normal gut microbiome in healthy birds (11). These bacteria are recognized as important zoonotic pathogens, with wild birds serving as natural reservoirs for species that can cause gastroenteritis in humans (12,13). In poultry and waterfowl, Campylobacter commonly colonizes their guts without causing clinical signs. Exceptions such as C. hepaticus and C. bilis can cause "spotty liver disease" in layer hens, but these are specialized cases (14). Fusobacterium, the second most abundant genus identified, comprises anaerobic bacteria that are normally minor components of avian gut microbiota (15) but in some microbiomes of wild bird species, it can be one of the common species (16). Thus, further research is needed to substantiate the significant increase of these bacteria in swans as a result of infection with the highly pathogenic H5N1 influenza virus.
Streptobacillus and Bacteroides were also consistently detected, though at lower abundances. Streptobacillus is less commonly reported in avian microbiome studies (16), while Bacteroides species are typical members of healthy gut communities in many vertebrates, including birds, where they play important roles in carbohydrate metabolism (17).
## Statistical confirmation of community patterns
The PERMANOVA analysis provided robust statistical confirmation of observed patterns in microbial community structure. Swab type emerged as a significant factor (F = 2.337, R 2 = 0.143, p = 0.01), accounting for approximately 14.3% of the variation in microbial composition across samples. This finding indicates that despite the consistent dominance of certain genera across sample types, there remain significant anatomical differences in microbial communities between the respiratory and digestive systems of H5N1infected swans. Interestingly, age did not significantly influence microbial composition (F = 0.6495, R 2 = 0.044, p = 0.865), suggesting The heatmap visualization of the genus-level microbiome composition across multiple swan samples. Hierarchical clustering is applied to both rows (genera) and columns (samples), grouping similar patterns together. The log10 transformation of values helps normalize data and highlight differences in relative abundance. Color Interpretation-the color gradient represents the transformed abundance values: Red-highest abundance. Yellowmoderately high abundance, Blue-Low abundance, White-Very low or near-zero abundance. n = 16 samples (9 cloacal, 7 tracheal; 11 adults, 5 juveniles).
that H5N1 infection may override age-related differences in the microbiome that would typically be observed in healthy birds. The year of collection showed only a marginal effect (F = 1.5349, R 2 = 0.099, p = 0.098), indicating relatively stable microbial patterns across the 2 years of the outbreak. Collectively, the tested factors explained approximately 29.67% of the total variation in microbial community structure, suggesting that additional factors not captured in this study, such as viral load, individual variation, or environmental conditions, also contribute to the observed patterns.
## Comparison with previous studies on H5N1-infected swan and other avian species microbiomes
Our findings differ significantly from those reported by Zhao et al. (10), who investigated the influence of H5N1 virus infection on the fecal microbiota of migrating whooper swans. While they observed shifts in microbial communities following H5N1 infection, they did not report the overwhelming dominance of Campylobacter that we identified. Instead, Zhao et al. found that H5N1 infection significantly increased the relative abundance of Aeromonas while decreasing Lactobacillus in the gut microbiota. The discrepancy between our results and those of Zhao et al. could reflect differences in sample collection timing (live birds versus freshly dead birds), geographical variables influencing baseline microbiome composition, or variations in viral strains and their effects on host-microbe interactions. Additionally, Zhao et al. focused exclusively on fecal samples, whereas our study incorporated both cloacal and tracheal samples, providing a more comprehensive view of microbial changes throughout the avian body following lethal H5N1 infection in swans.
Avian influenza virus compromises gut integrity, disturbs microbial balance, and triggers inflammation in the intestinal mucosa (18). In poultry, even low-pathogenic avian influenza virus (H9N2) leads to gut dysbiosis, characterized by reduced microbial alpha-diversity during the acute phase of infection and a delayed return to microbial equilibrium (8,19). The severe illness observed in influenza-bacterial co-infections is primarily driven by impaired antibacterial immune responses and synergistic interactions between the pathogens. (20,21).
## Comparison with normal avian microbiome
The microbiome composition observed in our H5N1-infected swans deviates substantially from the typical avian gut microbiome. Healthy bird gastrointestinal tracts are generally dominated by members of Firmicutes, Bacteroidetes, and Proteobacteria (22), with considerable variation across species, diet, and habitat. In contrast, our findings show an extreme dominance of specific genera (particularly Campylobacter and Fusobacterium) that are usually present in much lower abundances in healthy birds. This pattern suggests severe dysbiosis associated with H5N1 infection. The consistent presence of these potentially pathogenic bacteria across both respiratory and digestive tracts indicates a profound disruption of the normal microbiome throughout multiple body systems. The altered microbiome likely contributes to disease pathogenesis either through direct pathogenic effects or by compromising normal physiological functions dependent on healthy microbial communities, such as nutrient absorption, barrier protection, and immune regulation.
## Significance of findings
To our knowledge, this study represents the first comprehensive characterization of the microbiome in swans that died specifically from H5N1 infection. Our findings reveal a consistent pattern of microbial dysbiosis characterized by the dominance of Campylobacter and Fusobacterium genera across different anatomical sites. These results provide important insights into how highly pathogenic avian influenza may interact with the host microbiome, potentially contributing to disease severity and mortality. The identification of specific bacterial signatures associated with fatal H5N1 infection could potentially serve as markers for disease progression or indicators of poor prognosis in infected birds. Furthermore, our findings suggest that considering bacterial co-infections or secondary bacterial overgrowth may be important in understanding the full pathology of H5N1 infection in wild bird populations. The consistency of these microbial patterns across samples collected in different years (December 2023 and December 2024) indicates a stable hostpathogen-microbiome interaction that may be characteristic of H5N1 outbreaks in swan populations in the Caspian region.
## Limitations
A notable limitation of this study is that samples were collected from dead swans, which could potentially influence the microbiome composition due to post-mortem changes. However, several factors mitigate this concern. Samples were collected in cold December conditions (0-2°C), which would significantly slow bacterial overgrowth and post-mortem decomposition processes. Additionally, we specifically selected freshly dead swans, with some samples collected from birds that had died the same day or were in the process of dying. The carcass condition was assessed as "very fresh, " further minimizing potential post-mortem alterations. The immediate storage of samples in liquid nitrogen at -195°C would have effectively halted any further microbial changes after collection. While acknowledging this limitation, we believe the consistent patterns observed across multiple samples and the extreme dominance of specific taxa suggest genuine infection-associated dysbiosis rather than post-mortem artifacts. A notable limitation is the absence of a healthy control group, which prevents definitive conclusions that the observed microbiome changes are solely due to H5N1 infection and not influenced by other factors. Future studies including samples from healthy swans and birds at different stages of infection would further validate these findings.
## Conclusion
This study provides novel insights into the microbiome alterations associated with fatal H5N1 infection in wild swans in the Caspian Sea region. The consistent dominance of Campylobacter and Fusobacterium across different sampling sites suggests a profound and systemic impact of viral infection on the avian microbiome. These findings extend our understanding of how highly pathogenic avian influenza affects host-microbe interactions in wild birds, potentially identifying new factors that influence disease severity and mortality in these vulnerable populations. Further research into the specific mechanisms by which H5N1 infection drives these microbiome Frontiers in Veterinary Science 08 frontiersin.org changes and how these alterations contribute to disease pathogenesis will be valuable for expanding our knowledge of avian influenza ecology and may inform future approaches to wildlife disease management.
## References
1. Brown, Stallknecht, Swayne (2008) "Experimental infection of swans and geese with highly pathogenic avian influenza virus (H5N1) of Asian lineage" *Emerg Infect Dis*
2. Burashev, Strochkov, Sultankulova et al. (2020) "Near-complete genome sequence of an H5N1 avian influenza virus strain isolated from a swan in Southwest Kazakhstan in 2006" *Microbiol Resour Announc*
3. Kydyrmanov, Sayatov, Karamendin et al. (2017) "Monitoring of influenza a viruses in wild bird populations in Kazakhstan in 2002-2009" *Arch Virol*
4. Sultankulova, Argimbayeva, Aubakir et al. (2024) "Reassortants of the highly pathogenic influenza virus a/H5N1 causing mass swan mortality in Kazakhstan from 2023 to 2024" *Animals*
5. Kydyrmanov, Karamendin, Kassymbekov et al. (2024) "Mass mortality in terns and gulls associated with highly pathogenic avian influenza viruses in Caspian Sea"
6. Karamendin, Kydyrmanov, Fereidouni (2021) "High mortality in terns and gulls associated with infection with the novel Gull adenovirus" *J Wildl Dis*
7. Lv, Xiong, Shi et al. (2021) "The interaction between viruses and intestinal microbiota: a review" *Curr Microbiol*
8. Chrzastek, Leng, Zakaria et al. (2021) "Low pathogenic avian influenza virus infection retards colon microbiota diversification in two different chicken lines" *Anim Microbiome*
9. Teifke, Klopfleisch, Globig et al. (2007) "Pathology of natural infections by H5N1 highly pathogenic avian influenza virus in mute (Cygnus olor) and whooper (Cygnus cygnus) swans" *Vet Pathol*
10. Zhao, Wang, Li et al. (2018) "Influence of novel highly pathogenic avian influenza a (H5N1) virus infection on migrating whooper swans fecal microbiota" *Front Cell Infect Microbiol*
11. Keller, Shriver (2014) "Prevalence of three campylobacter species, C. jejuni, C. Coli, and C. lari, using multilocus sequence typing in wild birds of the mid-Atlantic region, USA" *J Wildl Dis*
12. Ahmed, Gulhan, Kahar-Bador et al. (2009) "Fatal influenza a (H3N2) and Campylobacter jejuni coinfection"
13. "Frontiers in Veterinary Science 09 frontiersin"
14. Zhang (2023) "Campylobacteriosis in birds"
15. Waite, Taylor, Herder et al. (2015) "Elevation correlates with significant changes in relative abundance in hummingbird fecal microbiota, but composition changes little" *Front Microbiol*
16. El Kaoutari, Armougom, Gordon et al. (2013) "The abundance and variety of carbohydrate-active enzymes in the human gut microbiota" *Nat Rev Microbiol*
17. El-Hack, El-Saadony, Alqhtani et al. (2022) "The relationship among avian influenza, gut microbiota and chicken immunity: an updated overview" *Poult Sci*
18. Yitbarek, Weese, Alkie et al. (2018) "Influenza a virus subtype H9N2 infection disrupts the composition of intestinal microbiota of chickens" *FEMS Microbiol Ecol*
19. Jia, Xie, Zhao et al. (2017) "Mechanisms of severe mortality-associated bacterial co-infections following influenza virus infection" *Front Cell Infect Microbiol*
20. Zhang, Zhou, Liu et al. (2024) "Avian influenza and gut microbiome in poultry and humans: a "one health" perspective" *Fundam Res*
21. Sun, Chen, Liu et al. (2022) "The avian gut microbiota: diversity, influencing factors, and future directions" *Front Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724197&blobtype=pdf | # B cell-intrinsic interleukin 17 receptor A signaling supports the establishment of chronic murine gammaherpesvirus 68 infection
Tahir Majeed, Nicholas Huss, Samantha Bradford, Christopher Jondle
## Abstract
The human-specific gammaherpesviruses Epstein-Barr virus and Kaposi's sarcoma-associated herpesvirus infect >95% of all adults, establish lifelong infections, and are associated with multiple different cancers, including B-cell lymphomas. These viruses naturally infect B cells and drive a unique and robust germinal center response to establish their latent viral reservoir in memory B cells. The unique and robust nature of the virus-driven germinal center response increases the risk for B-cell transformation, which is why many gammaherpesvirus-associated cancers are derived from germinal center or post-germinal center B cells. We have previously reported that both global and T cell-specific host IL-17RA signaling is proviral in the context of gammaherpesvirus infection by promoting the virus-driven germinal center response and establishment of latent viral infection. In this study, we examine the role of B cell-specific IL-17RA signaling in the context of gammaherpesvirus infection, given these viruses' reliance on B cells to establish and maintain lifelong infection. Similarly, to what we observed in the context of global and T cell-specific IL-17RA signaling, we found that B cell-intrinsic IL-17RA signaling supports the establishment of viral latency and reactivation in the spleen and peritoneal cavity, as well as promotes the virus-driven germinal center response. Our study reveals the proviral role of B cell-intrinsic IL-17RA signaling in supporting the establishment of chronic gammaherpesvirus infection. IMPORTANCE Gammaherpesviruses are lifelong pathogens that are prevalent in over 95% of all adults. These viruses are tumorigenic and associated with multiple cancers, including B-cell lymphomas. They are B-cell tropic viruses that manipulate the germinal center response to establish their latent viral reservoir in memory B cells. This manipu lation of the germinal center response is thought to be the target of viral transforma tion, leading to lymphomagenesis as most gammaherpesvirus-associated cancers are germinal center or post-germinal center derived. In this study, we developed a new mouse model to understand the B cell-intrinsic role of IL-17RA signaling in gammaher pesvirus infection. Previous studies have shown that IL-17RA signaling is proviral in the context of gammaherpesvirus infection, and this study found that while B cell-intrinsic IL-17RA signaling is not the sole factor behind the systemic proviral role of IL-17RA signaling, it plays an important role in the virus-driven germinal center response, the expansion of B-1 B cells, and the establishment of chronic gammaherpesvirus infection. KEYWORDS germinal center B cells, B cells, IL-17RA, murine gammaherpesvirus 68, MHV68, gammaherpesvirus G ammaherpesviruses are ubiquitous, DNA viruses that establish lifelong infections in their host. Epstein-Barr virus (EBV) and Kaposi's Sarcoma-associated herpesvirus (KSHV) are the two human-specific gammaherpesviruses that infect >95% of all adults and are associated with multiple cancers, including B-cell lymphomas (1, 2). These viruses have a biphasic life cycle consisting of lytic replication during the acute phase
of infection, followed by lifelong latent infection often characterized as chronic infection (3). The study of chronic EBV and KSHV infection in vivo is challenging due to the strict species specificity of gammaherpesviruses. This study utilizes murine gammaher pesvirus 68 (MHV68) as a tractable animal model of gammaherpesvirus infection. MHV68 is a natural rodent pathogen that is genetically, biologically, and pathologically related to EBV and KSHV (3)(4)(5)(6). It provides a model system to study chronic gammaherpesvirus infection and pathogenesis in an intact host, while also allowing for the manipulation of host genetics to understand potential host mechanisms that impact gammaherpesvirus infection.
EBV, KSHV, and MHV68 are B-cell-tropic viruses, with EBV and MHV68 establishing a latent viral reservoir in memory B cells by usurping the germinal center response (7)(8)(9)(10)(11). These gammaherpesviruses, unlike most other viral infections, usurp B-cell differentiation to establish a latent viral reservoir in memory B cells (7)(8)(9). They achieve this by infecting naïve B cells and inducing a germinal center response, which includes both virus-infected and uninfected B cells (7,12,13) that subsequently differentiate into class-switched plasma cells, where viral reactivation occurs, or memory B cells that host life-long latent infection (14). The germinal center stage of B-cell differentiation is susceptible to cellular transformation, as germinal center B cells rapidly divide while downregulating tumor suppressors (15) and increasing expression of mutagenic enzymes (16,17). Consequently, many gammaherpesvirus-driven B-cell lymphomas are of germinal center or post-germinal center origin (18). Furthermore, increased viral reactivation often precedes tumorigenesis (19)(20)(21)(22). Importantly, the mechanisms by which gammaherpesviruses induce the germinal center response and factors that promote viral reactivation are poorly understood.
IL-17A is a member of the IL-17 family of cytokines, consisting of IL-17A through IL-17F (23). It signals through the heterodimeric receptor complex of IL-17 receptor A (IL-17RA) and IL-17RC (24)(25)(26). The signaling events downstream of the IL-17RA-IL-17RC receptor utilize the adapter Act1 to recruit and ubiquitinate TRAF6, leading to the activation of the NF-kB and MAPK pathways (27)(28)(29). Many IL-17RA target genes have NF-kB promoters, and activation of the canonical NF-kB pathway is the primary mediator of inflammatory gene activation downstream of IL-17RA (23,30). IL-17A plays a diverse role across a number of different conditions. It is critical for the clearance of multiple bacterial and fungal pathogens (31,32), while also being associated with various autoimmune diseases, such as psoriasis, rheumatoid arthritis, and Crohn's disease (33). There have been four FDA-approved IL-17 targeting treatments against various inflammatory diseases (34,35). The role of IL-17A signaling in viral infections is less understood. It inhibits rhinovirus replication, promotes an IFN-γ response against herpes simplex virus 2, and is an important mediator of B-1a B-cell natural antibody response against pulmonary influenza infection (36)(37)(38), highlighting an antiviral function of IL-17A signaling. Intriguingly, SARS-CoV-2's viral protein Orf8 acts as a viral mimic of host IL-17 to activate IL-17RA and IL-17RC (39). Likewise, herpesvirus saimiri, a simian gammaherpesvirus, encodes a viral IL-17, which functions similarly to host IL-17A and IL-17F in cultured cells (40)(41)(42). Other gammaherpesviruses, including EBV, KSHV, and MHV68, do not encode a viral IL-17A homolog, yet in the context of infectious mono nucleosis, a clinical disease associated with recent EBV infection, there is a significant increase in IL-17A-producing CD4+ T cells that persists at least 1 month following the resolution of clinical symptoms (43). Furthermore, MHV68 infection leads to increased IL-17A production in the lungs during the acute stage of infection and the spleen and peritoneal cavity during the chronic stage of infection by multiple cell types (44,45). Finally, increases in IL-17A in the lungs following nontypeable Haemophilus influenzae secondary infection promote the establishment of MHV68 latency (46). These data suggest that IL-17A signaling may be proviral in the context of some viral infections and that gammaherpesvirus infection induces an IL-17 response.
Global loss of IL-17A signaling during MHV68 infection led to a significant attenua tion of viral latency and reactivation in the spleen and peritoneal cavity, as well as a reduction in the viral-driven germinal center response and decreased latent infection in activated/germinal center B cells in the spleen (45). Furthermore, loss of IL-17A signaling during MHV68 infection also impaired the production of irrelevant and self-directed antibody responses, which is a hallmark of EBV and MHV68 infection (47,48), with the detection of antibodies against horse red blood cells being diagnostic for recent EBV infection (49). Finally, loss of IL-17A signaling specifically in T cells during MHV68 infection resulted in a similar attenuation of viral latency and reactivation as well as the germinal center response in the spleen (50). Together, these data highlight a proviral role for global and more specifically T cell-intrinsic IL-17A signaling during MHV68 infection (51).
IL-17A signaling in B cells in the context of autoimmunity and bacterial infection is known to promote B-cell activation, proliferation, as well as germinal center formation and migration in addition to antibody class switching (52)(53)(54)(55)(56)(57)(58). In regard to viral infection, IL-17A signaling in B-1a cells promotes natural antibody production during pulmonary influenza infection (38). Given the significance of B cells to gammaherpesvirus infection (7)(8)(9)(10)(11)(12)(13), the role of IL-17A signaling in B cells (38,(52)(53)(54)(55)(56)(57)(58), and the noted proviral role for global IL-17RA signaling during gammaherpesvirus infection (45), in this study, we examine the B cell-intrinsic role of IL-17RA signaling during MHV68 infection. B cellspecific IL-17RA deficiency had no impact on B220+ B-cell numbers in naïve animals. Upon infection with MHV68, B cell-specific IL-17RA-deficient mice had significantly attenuated viral latency and reactivation in both the spleen and peritoneal cavity compared to IL-17RA B cell-sufficient control mice. Furthermore, the germinal center response in the infected B cell-specific IL-17RA-deficient mice was significantly decreased compared to the control. Interestingly, the decreased germinal center response in the B cell-specific IL-17RA-deficient mice did not correlate with a reduction in the titers of irrelevant and self-directed antibodies, as was previously observed in the global and T cell-specific IL-17RA-deficient mice (45,50). This is potentially accounted for by the increase in extrafollicular antibody-secreting B cells observed in the infected B cell-spe cific IL-17RA-deficient mice, which could compensate for the decrease in other antibodysecreting B cells. Furthermore, IL-17RA signaling in B cells during MHV68 infection supported the expansion of both B-1a and B-1b B cells in the spleen and peritoneal cavity. B-1b B cells are particularly important in the peritoneal cavity for supporting latent infection (59,60). In summary, our findings show that B cell-intrinsic IL-17RA signaling is proviral in the context of MHV68 infection. It helps promote the virus-driven germinal center response and the infection of B cells, both of which are necessary to establish viral latency. Furthermore, B cell-intrinsic IL-17RA signaling supports viral reactivation. This study, along with our previous ones, indicates that IL-17RA signaling in multiple cell types is critical to fully support the establishment of chronic gammaherpes virus infection.
## RESULTS
## Mouse model of B cell-specific IL-17RA deficiency
Having discovered a proviral role of global IL-17RA signaling during gammaherpesvirus infection (45), we next investigated the effect of B cell-intrinsic IL-17RA signaling. B cell-specific IL-17RA deficiency was generated by crossing IL-17RA fl/fl mice (61) to a mouse strain where expression of the Cre recombinase is driven by CD19 (B6.129P2-Cd19 tm1(cre)Cgn /J) (62). Use of the CD19 Cre allows for the recombination of the conditional IL-17RA allele at the earliest stages of B-cell development (63). Successful generation of the B cell-specific IL-17RA-deficient mice was confirmed by examining protein expres sion of IL-17RA on magnetically sorted CD19-positive B cells from naïve mice (Fig. 1A). Furthermore, expression of IL-17RA in CD19-positive B cells (Fig. 1B) and CD19-negative cells (Fig. 1C) from the spleen of naïve CD19 Cre-negative and CD19 Cre-positive mice was examined via flow cytometry (61). Loss of IL-17RA in B cells had no impact on splenic size (Fig. 1D) nor the frequency (Fig. 1E) or absolute number (Fig. 1F) of B220+ B cells in the spleen of naïve CD19 Cre-positive as compared to CD19 Cre-negative littermates.
Similarly, there was no difference in the frequency or absolute number of CD3+ T cells (Fig. S1A andB) and CD11b+ monocytes (Fig. S1C andD) in the spleen of naïve CD19 Cre-positive as compared to CD19 Cre-negative littermates.
## B cell-intrinsic IL-17RA deficiency leads to attenuated establishment of chronic MHV68 infection
We previously found that both global and T cell-intrinsic IL-17RA signaling support the establishment of chronic MHV68 infection (45,50). To determine whether B cellintrinsic IL-17RA signaling was required for optimal establishment of chronic MHV68 infection, viral latency and reactivation were measured at 16 days post-infection, the peak of viral latency (3), in the spleen and peritoneal cavity of CD19 Cre-positive and CD19 Cre-negative littermates. Consistent with what was observed in the global and T cell-intrinsic IL-17RA-deficient models (45,50), the frequency of MHV68 DNA-positive splenocytes (Fig. 2A) and the frequency of MHV68 reactivation from splenocytes (Fig. 2B) were significantly attenuated in CD19 Cre-positive mice compared to CD19 Cre-nega tive littermates. In the peritoneal cavity, the frequency of MHV68 DNA-positive cells (Fig. 2C) and MHV68 reactivation (Fig. 2D) was also attenuated in CD19 Cre-positive mice compared to CD19 Cre-negative littermates. Thus, B cell-intrinsic IL-17RA signaling supports the establishment of chronic MHV68 infection. Following peak viral latency at 16 days post-infection, the MHV68 latent viral reservoir constricts and stabilizes. This reduction in the latent viral reservoir also contracts the MHV68-driven germinal center response and reduces viral reactivation down to unde tectable levels by 42 days post-infection. Similar to what was observed in the global and T cell-intrinsic IL-17RA-deficient models (45,50), viral reactivation was undetectable in the spleen or peritoneal cavity of CD19 Cre-positive mice and CD19 Cre-negative littermates at 42 days post-infection (data not shown). Furthermore, there was no longer a difference in the frequency of MHV68-infected splenocytes (Fig. 2E) and peritoneal cells (Fig. 2F) between CD19 Cre-positive mice and CD19 Cre-negative littermates at the longterm infection time point of 42 days post-infection. Thus, B cell-intrinsic IL-17RA signaling helps promote peak MHV68 latency at 16 days post-infection but does not impact the long-term maintenance of the latent viral reservoir.
## B cell-intrinsic IL-17RA signaling promotes the MHV68-driven germinal center response during the establishment of latency
The germinal center response is critical for the establishment of chronic MHV68 infection (7)(8)(9). With the attenuation of viral latency observed at 16 days post-infection in CD19 Cre-positive mice compared to CD19 Cre-negative mice (Fig. 2), the germinal center response was investigated next in the spleen. Loss of B cell-intrinsic IL-17RA signal ing during MHV68 infection resulted in a decrease in splenomegaly (Fig. 3A), with a significant reduction in the frequency (Fig. 3B) and total number (Fig. 3C) of overall B cells in the spleen of infected CD19 Cre-positive mice compared to CD19 Cre-negative mice. Looking at the germinal center response, there was a reduced frequency and number of germinal center B cells (Fig. 3D through F) and T follicular helper cells (Fig. 3G through I) in CD19 Cre-positive mice compared to CD19 Cre-negative mice at 16 days post-infection. The germinal center B cells can be organized into the dark and light zones of the germinal center, with rapidly proliferating centroblasts making up the dark zone and non-proliferating centrocytes in the light zone (64). Loss of B cell-intrinsic IL-17RA signaling in naïve CD19 Cre-positive mice caused an increase in the frequency of proliferating centroblasts and a subsequent decrease in centrocytes compared to naïve CD19 Cre-positive mice (Fig. S2A andC). That difference in the ratio of centroblasts to centrocytes observed in naïve mice disappeared upon MHV68 infection, with no difference in the frequency of centroblasts (Fig. S2A) and centrocytes (Fig. S2C) between CD19 Cre-positive mice and Cre-negative mice at 16 days post-infection. Figure S3 shows the gating schemes for the determination of all reported cellular frequencies in the spleen.
At the long-term infection time point of 42 days post-infection, the frequency and number of germinal center B cells (Fig. 3J andK) and T follicular helper cells (Fig. 3L andM) were still reduced in CD19 Cre-positive mice compared to CD19 Cre-negative mice. Taken together, these data indicate that B cell-intrinsic IL-17RA signaling during MHV68 infection supports the establishment and maintenance of the virus-driven germinal center response.
## B cell-intrinsic IL-17RA signaling has no effect on the antibody response, including the irrelevant and self-directed class-switched antibodies stimula ted by MHV68 infection
Given the decrease in the germinal center response in MHV68-infected CD19 Crepositive mice (Fig. 3), we next examined the antibody response as an outcome of B-cell differentiation. Unlike what was previously observed by several groups with the global IL-17RA-deficient mice (45,65) and our group in T cell-specific IL-17RA-deficient mice (50), CD19 Cre-positive mice did not have a baseline increase in IgG levels. Furthermore, there was no difference in the total IgG (Fig. 4A) and IgM (Fig. 4B), nor viral-specific IgG (Fig. 4C) and IgM (Fig. 4D) antibody titers between MHV68-infected CD19 Cre-positive and CD19 Cre-negative mice at 16 days post-infection.
Both EBV and MHV68 are known to drive a robust non-viral specific B-cell differentiation that leads to a rapid, though transient increase in antibody titers against self-and irrelevant (foreign) antigen (47,48). The presence of irrelevant antibodies against horse red blood cells is diagnostic for recent EBV infection (49). In the context of MHV68 infection, global and T cell-specific IL-17RA signaling both supported the induction of irrelevant and self-directed antibodies (45,50). To determine whether B cell-intrinsic IL-17RA signaling helps support the induction of self-directed and irrelevant antibodies, we repurposed a clinical assay used in the diagnosis of autoimmune diseases (antinu clear antibody or ANA) (48). Sera from mock and 16 day post-infected CD19 Cre-positive and CD19 Cre-negative mice were used on ANA slides coated with Hep-2 cells, which express a wide array of antigens. In contrast to what was observed in the global and T cell-specific IL-17RA-deficient models at 16 days post-infection, loss of B cell-intrinsic signaling had no impact on the MHV68-driven induction of self-directed and irrelevant antibodies as observed through ANA staining (Fig. 4E andF) and anti-double-stranded DNA (dsDNA)-directed antibodies (Fig. 4G). At 42 days post-infection, representing long-term latency, the germinal center response continues to be reduced in CD19 Cre-positive mice (Fig. 3J through M). Despite this persistent attenuation, we do not find any measurable differences in antibody responses between infected CD19 Cre-positive and CD19 Cre-negative mice (Fig. 4H through L). Taken together, B cell-intrinsic IL-17RA signaling has no impact on the general, viral-specific, or the self-directed and irrelevant antibody response driven by MHV68 infection.
## B cell-intrinsic IL-17RA signaling during MHV68 infection modulates some splenic B-cell populations
Given the overall reduction in splenomegaly (Fig. 3A), total B cells (Fig. 3B andC), and the germinal center response (Fig. 3D through I) in the spleen, along with no appreciable difference in the antibody response in MHV68-infected CD19 Cre-positive mice compared to CD19 Cre-negative mice (Fig. 4), the impact of B cell-intrinsic IL-17RA signaling on B-cell populations during MHV68 infection was determined at 16 days post-infection. Figure S3 shows all gating strategies used to obtain frequencies shown in Fig. 5. Follicular and marginal zone B cells were first examined with loss of B cell-intrinsic IL-17RA, causing an increase in the frequency (Fig. 5A), but not number (Fig. 5B) of follicular B cells in naïve and MHV68-infected mice. No significant difference was observed in the frequency of marginal zone B cells (Fig. 5C) between naïve and MHV68infected CD19 Cre-positive and CD19 Cre-negative mice. There was, however, a statistical reduction in the total number of marginal zone B cells (Fig. 5D) between naïve and MHV68-infected CD19 Cre-positive and CD19 Cre-negative mice.
Next, antibody-secreting cells were examined. There was a significant reduction in the frequency (Fig. 5E) and number (Fig. 5F) of plasma cells in MHV68-infected CD19 Crepositive compared to CD19 Cre-negative mice. Interestingly, loss of B cell-intrinsic IL-17RA expression during MHV68 infection led to a significant increase in the frequency (Fig. 5G) and number (Fig. 5H) of extrafollicular antibody-secreting cells (66). Ageassociated B cells also produce antibodies (67), and no difference in the frequency (Fig. 5I) or number (Fig. 5J) of age-associated B cells was observed in the spleen between naïve and MHV68-infected CD19 Cre-positive and CD19 Cre-negative mice. Finally, B-1 cells, which secrete natural antibodies (68,69) and can be subdivided into B-1a and B-1b cells on the basis of CD5 expression (70), were examined. Loss of IL-17RA signaling on B cells resulted in a significant reduction in the frequency (Fig. 5K andM) and number (Fig. 5L andN) of B-1a and B-1b B cells in naïve and MHV68-infected CD19 Cre-positive mice compared to CD19 Cre-negative mice. Taken together, these data suggest that B cellintrinsic IL-17RA signaling not only promotes the germinal center response (Fig. 3) and subsequent plasma cell differentiation (Fig. 5E andF), but it also restricts the expansion of extrafollicular antibody-secreting B cells (Fig. 5G andH). Furthermore, B cell-intrinsic IL-17RA contributes to the expansion of B-1a (Fig. 5K andL) and B-1b (Fig. 5M andN) B-1 cell subsets in naïve mice and following MHV68 infection.
## B cell-intrinsic IL-17RA signaling during MHV68 infection supports infection of germinal center B cells
Having observed reduced latency in the spleen as well as a decreased MHV68-driven germinal center response in CD19 Cre-positive mice, the efficiency of germinal center Bcell infection was examined next, given that germinal center B cells contain a majority of the splenic latent viral reservoir at 16 days post-infection. The MHV68.ORF73bla reporter virus, which expresses a fused mLANA-b-lactamase protein in latently infected cells (71,72), was used to label latently infected cells by flow cytometry using a cell-permeable blactamase substrate (CCF2). There were significantly fewer CD19 Cre-positive splenic B cells harboring MHV68 latent virus compared to CD19 Cre-negative mice (8.2-fold decrease) (Fig. 5A andB), which is in line with the observed attenuation of splenic latency in the CD19 Cre-positive mice (Fig. 2A).
There was also a significant decrease (8.2-fold) in the frequency of MHV68 latently infected germinal center B cells from CD19 Cre-positive mice compared to MHV68 latently infected germinal center B cells from CD19 Cre-negative mice (Fig. 5C andD). Taken together, these data indicate that B cell-intrinsic IL-17RA signaling not only promotes the MHV68-driven germinal center response (Fig. 3), but it also supports MHV68 latent infection of germinal center B cells as the attenuation of the germinal center response alone (~2-fold decrease) in MHV68-infected CD19 Cre-positive does not fully account for the over eightfold decrease in MHV68 latently infected cells in the spleen.
## B cell-intrinsic IL-17RA signaling in the peritoneal cavity during MHV68 infection promotes B-cell latency and expansion of B-1 B cells
The impact of B cell-intrinsic IL-17RA signaling on the peritoneal cavity during MHV68 infection was examined next. Loss of B cell-intrinsic IL-17RA signaling during MHV68 infection resulted in approximately 1 million fewer peritoneal cells on average. That reduction in total peritoneal cells was not statistically significant (Fig. 7A). To determine the impact loss of B cell-intrinsic IL-17RA signaling had on the distribution of latently infected cells in the peritoneal cavity at 16 days post-infection, given the decrease in overall viral latency observed there (Fig. 2C) in the absence of B cell-intrinsic IL-17RA signaling, we magnetically sorted CD19-positive B cells from MHV68-infected CD19 Cre-positive and CD19 Cre-negative mice and performed limiting dilution-PCR to determine the frequency of latently infected peritoneal B cells (Fig. 7B) and non-B cells (Fig. 7C). B cells lacking IL-17RA harbored significantly less latent MHV68 (Fig. 7B), while there was no difference in the frequency of MHV68 latency in non-B cells between CD19 Cre-positive and CD19 Cre-negative mice (Fig. 7C).
In the peritoneal cavity, macrophages and B-1 B cells make up the majority of the latent viral reservoir (59,60,73,74). Loss of IL-17RA B cell-intrinsic signaling had no impact on the frequency of CD11b+ macrophages (Fig. 7D), yet there were significantly fewer overall CD11b+ macrophages in MHV68-infected CD19 Cre-positive mice compared to CD19 Cre-negative mice (Fig. 7E). Looking at B cells, there was no difference in the frequency (Fig. 7F) or total number (Fig. 7G) of B220+ B cells in the peritoneal cavity of naïve or MHV68-infected CD19 Cre-positive and CD19 Cre-negative mice. B-1 cells are the predominant B cells to foster latent infection in the peritoneal cavity and are subdivided into B-1a (CD5 neg) and B-1b (CD5 pos) cells based on CD5 expression, with B-1b cells harboring most, if not all, of the latent MHV68 compared to B-1a cells (59,60). Unlike what was observed in the spleen at 16 days post-infection (Fig. 5K through N), there was no significant difference in the frequency (Fig. 7H andJ) or total number (Fig. 7I andK) of B-1a and B-1b cells in naïve CD19 Cre-positive and CD19 Cre-negative mice. Upon MHV68 infection, there is a significant reduction in the frequency (Fig. 7H andJ) and total number (Fig. 7I andK) of B-1a and B-1b cells in CD19 Cre-positive mice compared to CD19 Cre-negative mice. Taken together, these data indicate that B cell-intrinsic IL-17RA signaling supports the establishment of viral latency in B cells as well as the expansion of both B-1 B cell subsets during MHV68 infection in the peritoneal cavity. Figure S4 shows the gating schemes for the determination of all reported cellular frequencies in the peritoneal cavity.
## DISCUSSION
In this study, we uncovered the proviral role of B cell-intrinsic IL-17RA signaling during MHV68 infection through the generation of a B cell-specific IL-17RA-deficient mouse model. Loss of IL-17RA signaling in B cells during MHV68 infection led to an attenuation of viral latency and reactivation and suppression of B-1 cell populations in both the spleen and peritoneal cavity, as well as a significant reduction in the MHV68-driven germinal center response in the spleen. These data indicate that B cell-intrinsic IL-17RA signaling helps facilitate the establishment of chronic MHV68 infection, supports the MHV68-driven germinal center response, and expansion of B-1 cell populations.
## B cell-intrinsic function of IL-17RA signaling during MHV68 infection
Our previous studies were the first to demonstrate that global IL-17RA signaling and T cell-intrinsic IL-17RA signaling are proviral in the establishment of chronic MHV68 infection (45,50). Those previous studies found that both global and T cell-intrinsic IL-17RA signaling supported the establishment of viral latency, viral reactivation, and the MHV68-driven germinal center response. As EBV and MHV68 are B-cell-tropic viruses (7)(8)(9)(10)(11), we set out to better understand the impact B cell-intrinsic IL-17RA signaling had on MHV68 infection. Loss of IL-17RA signaling in B cells was sufficient to attenuate MHV68 latency and reactivation in the spleen (Fig. 2A andB), and unlike what was observed in the T cell-specific deficiency (50), in the peritoneal cavity as well (Fig. 2C andD). The decrease in viral reactivation and latency in the peritoneal cavity is similar to what was observed in the global IL-17RA-deficient mouse model (45).
Loss of B cell-intrinsic IL-17RA signaling during MHV68 infection led to a reduction in overall spleen size following infection (Fig. 3A). Furthermore, there were significantly fewer B220+ B cells in the spleen of mice lacking B cell-intrinsic IL-17RA signaling during MHV68 infection (Fig. 3B andC). Similar to what was observed in the global and T cell-deficient IL-17RA signaling models (45,50), loss of B cell-specific IL-17RA signaling also significantly attenuated the MHV68-driven germinal center response at 16 days post-infection (Fig. 3D through I). The germinal center response was still attenuated with the loss of IL-17RA B cell-intrinsic signaling at 42 days post-infection (Fig. 3J through M). These findings indicate that B cell-intrinsic IL-17RA signaling is also important in sustaining the germinal center response during long-term latent MHV68 infection. Given that both B cell-intrinsic IL-17RA signaling (Fig. 3) and T cell-intrinsic IL-17RA signaling (50) alone supported the expansion of MHV68-driven germinal center, while impacting signaling in only one cell type, either germinal center B cells or the T follicular helper cells, highlights the interconnectedness of germinal center B cells and T follicular helper cells in the germinal center response. If either the expansion of germinal center B cells or T follicular helper cells is restricted due to lack of IL-17RA signaling in one of those cell types, then we get a similar reduction in the other cell type as they work together to support the expansion of the germinal center response.
To determine whether attenuation (twofold reduction) of the germinal center response in IL-17RA-deficient B cells (Fig. 3) was solely responsible for the reduced MHV68 latency in the spleen (Fig. 2A), the MHV68.ORF73bla reporter virus was utilized. Lack of IL-17RA signaling in B cells resulted in an eightfold reduction in the frequency of MHV68-infected germinal center B cells (Fig. 6C andD). The frequency of germinal center B cells was only reduced twofold in infected B cell-specific IL-17RA-deficient mice compared to IL-17RA-sufficient mice, which indicates that IL-17RA signaling in B cells not only supports the MHV68-driven germinal center response, but it also helps facilitate the establishment of latency in germinal center B cells.
Interestingly, loss of B cell-specific IL-17RA signaling had no impact on the humoral response to MHV68 infection at 16 and 42 days post-infection (Fig. 4), despite the attenuated germinal center response (Fig. 3). This was surprising, particularly for irrelevant and self-directed antibody production, given that global and T cell-specific deficient IL-17RA signaling loss resulted in a significant defect in the generation of irrelevant and self-directed antibodies (45,50), a process that is unique to gammaherpes virus infection (47,48). These data indicate that the irrelevant and self-directed antibody response during gammaherpesvirus infection driven by IL-17RA signaling is T celldependent, as only when IL-17RA signaling is lost in T cells do we see the attenuation of these responses (45,50). Future studies will examine the T cell-dependent role of IL-17RA signaling in promoting the irrelevant and self-directed antibody response during gammaherpesvirus infection.
One potential explanation as to why there was no difference in the humoral response in the absence of B cell-intrinsic IL-17RA signaling during gammaherpesvirus infection is the role B cell-intrinsic IL-17RA signaling has in modulating B-cell populations. Loss of B cell-intrinsic IL-17RA signaling resulted in a significant decrease in the frequency and number of plasma cells in MHV68-infected mice (Fig. 5E andF). Interestingly, loss of B cell-intrinsic IL-17RA signaling increased the frequency and number of extrafollicular antibody-secreting B cells (Fig. 5G andH). Age-associated B cells also produce antibodies (67), and loss of IL-17RA signaling in B cells had no impact on the age-associated B cells during MHV68 infection (Fig. 5I andJ). B-1 cells are a group of B cells that produce natural antibodies (68,69) and loss of IL-17RA signaling in B cells resulted in significant reduc tions in both B-1a (Fig. 5K andL) and B-1b (Fig. 5M andN) subsets of B-1 B cells in naïve and MHV68-infected mice in the spleen. In the peritoneal cavity, a significant reduction in both B-1a (Fig. 7H andI) and B-1b (Fig. 7J andK) was only observed following MHV68 infection in the CD19 Cre-positive mice compared to CD19 Cre-negative mice. Loss of IL-17RA signaling in B cells during MHV68 infection resulted in a loss of around 50%-60% of the plasma cells (Fig. 5F) and B-1a (Fig. 5L) and B-1b (Fig. 5N) cells in the spleen. The loss of antibody-secreting plasma cells (Fig. 5F) and B-1a (Fig. 5L) and B-1b (Fig. 5N) cells was accounted for by a nearly 100% increase in extrafollicular antibody-secreting B cells (Fig. 5H) in MHV68-infected mice lacking B cell-intrinsic IL-17RA signaling. These data indicate that despite a decrease in B-1a and B-1b cells and a reduced germinal center response leading to fewer plasma cells in the absence of B cell-intrinsic IL-17RA signaling during MHV68 infection, the loss of those antibody-secreting cells is made up for in an increase in extrafollicular antibody-secreting cells. Furthermore, our data demonstrate that B cell-intrinsic IL-17RA signaling not only supports the germinal center response (Fig. 3) and differentiation of plasma cells (Fig. 5E andF); it restricts the expansion of extrafollicular antibody-secreting B cells (Fig. 5G andH) during MHV68 infection, while promoting the expansion of B-1a (Fig. 5K, L, 7H andI) and B-1b (Fig. 5M, N, 7J and K) cells. To our knowledge, this is the first time IL-17RA signaling has been implicated in impact ing extrafollicular antibody-secreting B cells. Our data also support and expand on findings by Wang et al. (38) that found IL-17RA signaling promotes B-1a cell differentiation and natural antibody production in the lungs during influenza virus infection (38). We show that IL-17RA signaling in B cells also promotes B-1b B cell (Fig. 5M, N, 7J and K) activation in addition to B-1a B cell (Fig. 5K, L, 7H and I) differentiation in naïve mice in the spleen and during MHV68 infection in the spleen and peritoneal cavity. Loss of IL-17RA signaling in B cells during MHV68 infection also led to significant attenuation in viral latency (Fig. 2C) and reactivation (Fig. 2D) in the peritoneal cavity. To determine how B cell-intrinsic IL-17RA signaling supports the establishment of latency in the peritoneal cavity, we examined viral latency in B cells (Fig. 7B) and non-B cells (Fig. 7C). Loss of IL-17RA signaling in B cells during MHV68 infection only impacted viral latency in B cells (Fig. 7B) with viral latency in non-B cells being identical in the presence or absence of IL-17RA signaling in B cells (Fig. 7C). Viral latency in non-B cells is primarily in macrophages (73,74), while B-1b cells make up the majority of peritoneal B cells harboring latent virus (59,60). Loss of B cell-intrinsic IL-17RA signaling during MHV68 infection led to a reduction in the number of CD11b+ macrophages (Fig. 7E) as well as the frequency and number of B-1a (Fig. 7H andI) and B-1b (Fig. 7J andK). Taken together, one can see how despite no difference in the frequency of viral latency in macrophages (Fig. 7C) in the absence of B cell-intrinsic IL-17RA signaling during MHV68 infection, there Full-Length Text is an overall decrease in viral latency in the absence of IL-17RA signaling in B cells in the peritoneal cavity given the fewer macrophages (Fig. 7E) and significantly fewer B-1b B cells (Fig. 7K). These data indicate that B cell-intrinsic IL-17RA signaling during MHV68 infection supports the establishment of latency in peritoneal B cells (Fig. 7B) as well as the expansion of B-1a (Fig. 7H andI) and B-1b (Fig. 7J andK) B cells.
This study revealed that some of the proviral effects observed in global IL-17RA signaling deficiency during MHV68 infection (45) can be attributed to B cell-intrinsic IL-17RA signaling. B cell-intrinsic IL-17RA signaling supported the establishment of MHV68 latency and viral reactivation in the spleen and peritoneal cavity (Fig. 2). It helps promote the MHV68-driven germinal center response (Fig. 3) and supports the establishment of MHV68 latency in germinal center B cells (Fig. 6) and peritoneal cavity B cells (Fig. 7B), all of which were phenotypes observed in the global IL-17RA-deficient model (45). When comparing MHV68-dependent phenotypes between the global-deficient, T cell-deficient, and now B cell-deficient IL-17RA signaling models, it is clear that multiple cell types are responsible for the proviral phenotype seen in the global-deficient model. Both B and T cell-intrinsic IL-17RA signaling support the establishment of MHV68 latency and viral reactivation in the spleen (Fig. 2A andB) (50), while B cell-intrinsic signaling supports the establishment of MHV68 latency and viral reactivation in the peritoneal cavity (Fig. 2C andD). Both help support the MHV68-driven germinal center response (Fig. 3), while T cell-intrinsic (50), but not B cell-intrinsic (Fig. 4), IL-17RA signaling helps support the generation of irrelevant and self-directed antibodies. In the future, studies will continue to determine the combination of cell subsets that require IL-17RA signaling to support MHV68 infection as well as investigate potential mechanisms of IL-17RA signaling that support MHV68 infection.
## B cell-intrinsic IL-17RA signaling function in the B-cell responses and establishment of MHV68 latency
There are multiple ways in which B cell-intrinsic IL-17RA signaling may help promote the establishment of MHV68 latency. Previous reports have shown that IL-17RA signaling helps promote B-cell activation (52) and, importantly, germinal center formation and migration (53)(54)(55)(56)(57) in the context of autoimmunity and bacterial infection. Furthermore, IL-17RA signaling via Blimp-1 expression and NF-kB activity in B-1a cells promotes differentiation and natural antibody production during pulmonary influenza virus infection (38). Both B-cell activation and germinal center formation, which IL-17RA signaling can affect (52)(53)(54)(55)(56)(57)(58), are critical for the establishment of latent gammaherpes virus infection (7)(8)(9)12). We found that loss of B cell-intrinsic IL-17RA signaling in the context of MHV68 infection led to attenuated MHV68 latency (Fig. 2 and6) and MHV68-driven germinal center response (Fig. 3) in the spleen. The reduction in the MHV68-driven germinal center response was not solely responsible for the full attenua tion of MHV68 latency in the spleen (Fig. 6). One potential way in which B cell-intrinsic IL-17RA signaling could promote the MHV68-driven germinal center response is through activation of NF-kB downstream of the receptor (27)(28)(29). NF-kB activity has been shown to support the establishment of MHV68 latency and the germinal center response during MHV68 infection (75)(76)(77). Attenuation of NF-kB activity in the absence of IL-17RA signaling in B cells could lead to both a reduction in the MHV68-driven germinal center response and MHV68 latent viral levels. Furthermore, we observed that B cell-intrinsic IL-17RA signaling was important for the expansion of both B-1a (Fig. 5K, L, 7H andI) and B-1b (Fig. 5M, N, 7J and K) B cells in the spleen and peritoneal cavity during MHV68 infection. B-1b cells in particular are an important latent viral reservoir in the peritoneal cavity (59,60). A previous report showed that, via Blimp-1 expression and NF-kB activity in B-1a cells, downstream of IL-17RA signaling is important for B-1a cell differentiation (38). The lack of B-1a (Fig. 5K, L, 7H and I) and B-1b (Fig. 5M, N, 7J and K) B cells expansion in the spleen and peritoneal cavity during MHV68 infection could be due to the lack of NF-kB activity downstream of IL-17RA signaling in those cells. Future studies will be done to determine if that is the mechanism driving B-1b B-cell expansion during MHV68 infection. The lack of B-1b cell expansion in the peritoneal cavity (Fig. 7H andI) is potentially due to loss of NF-kB activity downstream of IL-17RA signaling in B cells, along with the decrease in efficiency to infect B cells in the peritoneal cavity (Fig. 7B) seen during MHV68 infection in mice lacking IL-17RA signaling in B cells, which would explain the decreased viral latency observed in the peritoneal cavity (Fig. 2C). These results highlight another aspect of NF-kB activity downstream of IL-17RA signaling in B cells other than germinal center B cells, which may directly impact the establishment of MHV68 latency this time in the peritoneal cavity.
NF-kB activity is not the only potential route by which B cell-intrinsic IL-17RA signaling could promote MHV68 latency establishment. IL-17RA signaling can stimulate glucose uptake through the induction of Iκbζ (78), which can stimulate immune cell activation. Glucose metabolism can promote gammaherpesvirus infection (79)(80)(81)(82) as well as play an important role in supporting the exponential expansion of B cells in the germinal center response (80)(81)(82) and other important B-cell functions (83,84). IL-17RA signaling also promotes mRNA stabilization through the universally expressed RNA-binding protein HuR (also known as ELAVL1) (85)(86)(87). Loss of efficient mRNA stabilization may explain the decreased MHV68 latency efficiency observed in IL-17RA-deficient germinal center B cells (Fig. 6). Future studies will work to identify downstream signaling mechanisms of IL-17RA in B cells and other cells, which could explain its proviral role in MHV68 infection.
This study is the first, to our knowledge, to directly examine the B cell-intrinsic role of IL-17RA signaling in the context of viral infection and show a role for it in the establishment of MHV68 latent infection, the MHV68-driven germinal center response, and expansion of both B-1a and B-b B cells. Previous studies looking at the impact of IL-17RA signaling on B cells have relied on global-deficient mouse models and IL-17A or IL-17RA neutralizing antibodies. The mice generated in this study will provide an excellent tool to understand the significance of IL-17RA signaling in B cells in a wide range of autoimmune and infection models.
## MATERIALS AND METHODS
## Animals used
To generate mice with B cell-specific IL-17RA deficiency, IL-17RA fl/fl (B6.Cg-IL-17ra tm2.1Koll /J stock 031000) mice (61) were bred to CD19 Cre-positive (B6.129P2-Cd19 tm1(cre)Cgn /J) mice (62) from The Jackson Laboratories (Bar Harbor, ME). All mice were housed and bred in a specific pathogen-free facility. All experimental manipulations of mice were approved by the Institutional Animal Care and Use Committee of Western Michigan University Homer Stryker M.D. School of Medicine (2022-0026). Both male and female mice were used, with no sex-specific phenotypes noted.
## Magnetic sorting
CD19+ splenic B cells and B cells in the peritoneal cavity were negatively selected according to the manufacturer's instructions (MojoSort Mouse CD19 B cell Isolation Kit, BioLegend, San Diego, CA).
## Western blot
Magnetically sorted CD19+ B cells from naïve CD19 Cre-negative and CD19 Cre-positive mice were used to assess IL-17RA protein expression. Sorted B cells were lysed in Laemmli buffer and subjected to Western analyses as previously described (88). The following antibodies were used: anti-IL-17RA (1:5,000), anti-β actin (1:40,000), a goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (1:15,000), and a goat anti-mouse horseradish peroxidase-conjugated secondary antibody (1:15,000) (Thermo Fisher Scientific, Waltham, MA).
## Infections
Between 6 and 10 weeks of age, mice were intranasally inoculated with 1,000 PFU of MHV68 (WUMS) diluted in sterile serum-free Dulbecco's modified Eagle's medium (15 µL/ mouse), under light anesthesia. MHV68 viral stock was prepared and titered on NIH 3T12 cells. The spleen and peritoneal cells were harvested from euthanized mock-treated and MHV68-infected animals at the indicated times post-infection. Mice were euthanized by CO 2 inhalation from a compressed gas source in a non-overcrowded chamber. Mice were bled prior to euthanasia via a submandibular route, and serum was isolated using BD Microtainer blood collection tubes (Becton, Dickinson and Company, Franklin Lakes, NJ).
## Limiting dilution assays
Frequency of cells harboring viral DNA was determined by limiting dilution PCR analysis, while the frequency of ex vivo reactivation was determined by limiting dilution assay as previously described (89). Briefly, to determine the frequency of cells reactivating virus ex vivo, serial two-fold dilutions of splenocytes or peritoneal cell suspensions harves ted from infected mice were plated onto monolayers of mouse embryonic fibroblasts (MEFs) immediately following harvest, at 24 replicates per dilution. To control for preformed infectious virus, twofold serial dilutions of mechanically disrupted spleno cytes or peritoneal cells were plated as above. MHV68 was allowed to reactivate from explanted cells, and the virus was further amplified within the same well via subsequent replication in MEF. At 21 days post-plating, all replicates were scored in a binary fashion for the presence of live fibroblasts (no viral reactivation/replication) or absence of such (cytopathic effect driven by lytic replication). Because primary MEFs were used to amplify the virus, the sensitivity of the limiting dilution reactivation assay was below a single PFU of MHV68, as defined using a 3T12-based plaque assay. Because the endpoint of viral amplification in MEF was measured, the limiting dilution reactivation assay was not susceptible to variability of titers released from primary cells upon viral reactivation ex vivo (90).
## Flow cytometry
Single-cell suspensions of splenocytes and peritoneal cells from individual mice were prepared in FACS buffer (phosphate-buffered saline [PBS] + 2% FCS +0.05% sodium azide) at 1 × 10 7 nucleated cells/mL. A total of 2 × 10 6 cells were treated with Fc block (24G2) prior to extracellular staining for 30 minutes on ice. Data acquisition was performed on an Attune NxT flow cytometer (Thermo Fisher Scientific, Waltham, MA) and a Fortessa flow cytometer (BD Biosciences, San Jose, CA) and analyzed using FlowJo software (Becton, Dickinson & Company, Ashland, OR). The following antibodies were purchased from BioLegend (San Diego, CA) for use in this study: CD3 (17A2), CD4 (RM4-5), CD5 (53-7.3), CD11b (M1/70), CD11c (N418), CD19 (6D5), CD21 (7E9), CD22 (OX-97), CD23 (B3B4), CD38 (88), CD43 (S11), CD44 (IM7), CD86 (GL-1), PD-1 (29 f.1A12), B220 (RA3-6B2), GL7 (GL-7), CXCR4 (L276F12), CXCR5 (L138D7), IRF-4 (IRF4.3E4), T-bet (4B10), IgD (11-26c.2a), and IgM (RMM-1). CD95 (JO2) was purchased from BD Hori zon (San Jose, CA). IL-17RA (PAJ-17R) was purchased from Thermo Fisher Scientific (Waltham, MA. Cat #12-7182-82). Compensation controls were done using OneComp eBeads (Thermo Fisher Scientific, Waltham, MA). Briefly, a negative control (beads alone) was used to establish a baseline PMT (photomultiplier tube) voltage and fluorescent background. Positive controls for each fluorochrome (beads with a single fluorochrome) were used to establish spill-over of the individual fluorochrome into the other channels being used. PMT values are adjusted for each fluorochrome to minimize spillover.
## MHV68.ORF73βla studies
Mice were infected intranasally with 10,000 PFU MHV68.ORF73βla. Mice were euthanized at 16 days post-infection, with splenocytes pooled from each group (3 mice/group). Cells (2 × 10 7 ) from each group were Fc blocked and then stained with B220 and GL7 for 30 minutes on ice as described above. The cells were washed twice and loaded with CCF2-AM (GeneBLAzer kit; ThermoFisher Scientific, Waltham, MA) at room temperature for 1 hour. The cells were then washed twice and suspended in FACS buffer prior to analysis via flow cytometry.
## ELISA
Total, MHV68-specific, and dsDNA immunoglobulin levels were determined as previously described (91). Briefly, Nunc Maxisorp plates (Fisher Scientific, Pittsburgh, PA) were coated with anti-IgG (Heavy+Light) or anti-IgM antibodies (Jackson ImmunoResearch, West Grove, PA), UV-irradiated MHV68 virus stock in PBS (740,000 microjoules/cm 2 × 2) (Stratalinker UV Crosslinker 1800; Agilent Technologies, Santa Clara, CA), or dsDNA from Escherichia coli (12•5 µg/mL; Sigma-Aldrich, St. Louis, MO) overnight at 4°C. Plates were washed with PBS-Tween (0.05%), blocked for 1 hour with PBS-Tween (0.05%)-BSA (3%), incubated with fivefold serial dilutions of serum in PBS-Tween (0.05%)-BSA (1.5%) for 2 hours, and washed with PBS-Tween (0.05%). Bound antibody was detected with horseradish peroxidase-conjugated goat anti-mouse total IgG (heavy + light chain) or IgM (Jackson ImmunoResearch, West Grove, PA) using 3,3′, 5,5′-tetramethylbenzidine substrate (Life Technologies, Gaithersburg, MD). HRP enzymatic activity was stopped by the addition of 1 N HCl (Sigma-Aldrich, St. Louis, MO), and the absorbance was read at 450 nm on BioTek EPOCH2 Plate Reader (Agilent Technologies, Santa Clara, CA).
## ANA panels
Antinuclear antibodies were assessed with an Antinuclear Antibody (ANA) Test Kit (Antibodies Inc., Davis, CA). Following the manufacturer's protocol, serum was diluted (1:40 in PBS) and incubated over slides coated with fixed HEp-2 cells. Following serum incubation, the slides were rinsed and stained with anti-mouse IgG Alexa Fluor-488 (H+L) (ThermoScientific, Waltham, MA). Fluorescent images were captured using NIS Elements software. Corrected fluorescence was quantified using ImageJ software from a randomly chosen field of ~10 cells in each sample.
## Statistical analyses
Statistical analyses were performed using Student's t test when comparing two groups and one-way ANOVA with Tukey's post hoc test when comparing more than two groups (Prism, GraphPad Software, Inc.).
## References
1. Cesarman (2014) "Gammaherpesviruses and lymphoproliferative disorders" *Annu Rev Pathol*
2. Jha, Banerjee, Robertson (2016) "The role of gammaherpesviruses in cancer pathogenesis" *Pathogens*
3. Barton, Mandal, Speck (2011) "Pathogenesis and host control of gammaherpesviruses: lessons from the mouse" *Annu Rev Immunol*
4. Efstathiou, Ho, Hall et al. (1990) "Murine herpesvirus 68 is genetically related to the gammaherpesviruses Epstein-Barr virus and herpesvirus saimiri" *J Gen Virol*
5. Tarakanova, Suarez, Tibbetts et al. (2005) "Murine gammaherpesvirus 68 infection is associated with lymphoproliferative disease and lymphoma in BALB β2 microglobulin-deficient mice" *J Virol*
6. Virgin Hw 4th, Latreille, Wamsley et al. (1997) "Complete sequence and genomic analysis of murine gammaherpesvirus 68" *J Virol*
7. Flaño, Kim, Woodland et al. (2002) "Gamma-herpesvirus latency is preferentially maintained in splenic germinal center and memory B cells" *J Exp Med*
8. Willer, Speck (2005) "Establishment and maintenance of long-term murine gammaherpesvirus 68 latency in B cells in the absence of CD40" *J Virol*
9. Babcock, Decker, Freeman et al. (1999) "Epsteinbarr virus-infected resting memory B cells, not proliferating lympho blasts, accumulate in the peripheral blood of immunosuppressed patients" *J Exp Med*
10. Mesri, Cesarman, Arvanitakis et al. (1996) "Human herpesvirus-8/Kaposi's sarcomaassociated herpesvirus is a new transmissible virus that infects B cells" *J Exp Med*
11. Monini, Colombini, Stürzl et al. (1999) "Reactivation and persistence of human herpesvirus-8 infection in B cells and monocytes by Th-1 cytokines increased in Kaposi's sarcoma" *Blood*
12. Roughan, Thorley-Lawson, Da (2009) "The intersection of Epstein-Barr virus with the germinal center" *J Virol*
13. Collins, Boss, Speck (2009) "Identification of infected B-cell populations by using a recombinant murine gammaherpesvirus 68"
14. (2025) *Full-Length Text Journal of Virology*
15. *J Virol*
16. Stebegg, Kumar, Silva-Cayetano et al. (2018) *Regulation of the germinal center response. Front Immunol*
17. Phan, Dalla-Favera (2004) "The BCL6 proto-oncogene suppresses p53 expression in germinal-centre B cells" *Nature*
18. Muramatsu, Sankaranand, Anant et al. (1999) "Specific expression of activation-induced cytidine deaminase (AID), a novel member of the RNA-editing deaminase family in germinal center B cells" *J Biol Chem*
19. Martin, Bardwell, Woo et al. (2002) "Activation-induced cytidine deaminase turns on somatic hypermutation in hybridomas" *Nature*
20. Küppers (2003) "B cells under influence: transformation of B cells by Epstein-Barr virus" *Nat Rev Immunol*
21. Campbell, Borok, Gwanzura et al. (2000) "Relationship of human herpesvirus 8 peripheral blood virus load and Kaposi's sarcoma clinical stage" *AIDS*
22. Meerbach, Wutzler, Häfer et al. (2008) "Monitoring of Epstein-Barr virus load after hematopoietic stem cell transplantation for early intervention in post-transplant lymphoproliferative disease" *J Med Virol*
23. Feng, Cohen, Fischer et al. (2004) "Reactivation of latent Epstein-Barr virus by methotrexate: a potential contributor to methotrexateassociated lymphomas" *J Natl Cancer Inst*
24. Orlandi, Paulli, Viglio et al. (2001) "Epstein-Barr virus-positive aggressive lymphoma as a consequence of immunosuppression after multiple salvage treatments for follicular lymphoma" *Br J Haematol*
25. Kolls, Lindén (2004) "Interleukin-17 family members and inflammation" *Immunity*
26. Wright, Guo, Quazi et al. (2007) "Identification of an interleukin 17F/17A heterodimer in activated human CD4+ T cells" *J Biol Chem*
27. Chang, Dong (2007) "A novel heterodimeric cytokine consisting of IL-17 and IL-17F regulates inflammatory responses" *Cell Res*
28. Toy, Kugler, Wolfson et al. (2006) "Cutting edge: interleukin 17 signals through a heteromeric receptor complex" *J Immunol*
29. Amatya, Garg, Gaffen (2017) "IL-17 signaling: the Yin and the Yang" *Trends Immunol*
30. Cohen (2014) "The TLR and IL-1 signalling network at a glance" *J Cell Sci*
31. Chang, Park, Dong (2006) "Act1 adaptor protein is an immediate and essential signaling component of interleukin-17 receptor" *J Biol Chem*
32. Shen, Hu, Goswami et al. (2006) "Identification of common transcriptional regulatory elements in interleukin-17 target genes" *J Biol Chem*
33. Veldhoen (2017) "Interleukin 17 is a chief orchestrator of immunity" *Nat Immunol*
34. Huppler, Conti, Hernández-Santos et al. (2014) "Role of neutrophils in IL-17-dependent immunity to mucosal candidiasis" *J Immunol*
35. Kuwabara, Ishikawa, Kondo et al. (2017) "The role of IL-17 and related cytokines in inflammatory autoimmune diseases" *Mediators Inflamm*
36. Adams, Maroof, Baker et al. (2020) "Bimekizumab, a novel humanized IgG1 antibody that neutralizes both IL-17A and IL-17F" *Front Immunol*
37. Țiburcă, Bembea, Zaha et al. (2022) "The treatment with interleukin 17 inhibitors and immunemediated inflammatory diseases" *Curr Issues Mol Biol*
38. Graser, Ekici, Sopel et al. (2016) "Rhinovirus inhibits IL-17A and the downstream immune responses in allergic asthma" *Mucosal Immunol*
39. Bagri, Anipindi, Nguyen et al. (2017) "Novel role for interleukin-17 in enhancing type 1 helper T cell immunity in the female genital tract following mucosal herpes simplex virus 2 vaccination" *J Virol*
40. Wang, Ma, Chen et al. (2016) "IL-17A promotes pulmonary B-1a cell differentiation via induction of blimp-1 expression during influenza virus infection" *PLoS Pathog*
41. Wu, Xia, Shin et al. (2022) "Viral mimicry of interleukin-17A by SARS-CoV-2 ORF8" *mBio*
42. Knappe, Hiller, Niphuis et al. (1998) "The interleukin-17 gene of herpesvirus saimiri" *J Virol*
43. Yao, Fanslow, Seldin et al. (1995) "Herpesvirus saimiri encodes a new cytokine, IL-17, which binds to a novel cytokine receptor" *Immunity*
44. Yao, Spriggs, Derry et al. (1997) "Molecular characterization of the human interleukin (IL)-17 receptor" *Cytokine*
45. Gao, Shao (2017) "Investigation for the roles of TLR2, TLR9 and Th17/ Treg in the pathogenesis of infectious mononucleosis" *Int J Clin Exp Med*
46. Mukhopadhyay, Swan, Pridjian et al. (2024) "Gammaherpesvirus infection stimulates lung tumor-promoting inflammation" *Pathogens*
47. Jondle, Johnson, Aurubin et al. (2021) "Gammaherpesvirus usurps host il-17 signaling to support the establishment of chronic infection" *mBio*
48. Huss, Majeed, Wills et al. (2024) "Nontypeable Haemophilus influenzae challenge during gammaherpesvirus infection enhances viral reactivation and latency" *Virology (Auckl)*
49. Sangster, Topham, Costa et al. (2000) "Analysis of the virus-specific and nonspecific B cell response to a persistent B-lymphotropic gammaherpesvirus" *J Immunol*
50. Gauld, Santis, Kulinski et al. (2013) "Modulation of B-cell tolerance by murine gammaherpesvirus 68 infection: requirement for Orf73 viral gene expression and follicular helper T cells" *Immunology*
51. Fleisher, Collins, Fager (1983) "Limitations of available tests for diagnosis of infectious mononucleosis" *J Clin Microbiol*
52. Jondle, Tarakanova (2022) "T cell-intrinsic interleukin 17 receptor a signaling supports the establishment of chronic murine gammaherpes virus 68 infection" *J Virol*
53. Majeed, Jondle (2024) "Bringing balance: immune interactions regulating murine gammaherpesvirus 68 latency" *Curr Clin Micro Rpt*
54. Subbarayal, Chauhan, Zazzo et al. (2016) "IL-17 Augments B cell activation in ocular surface autoimmunity" *J Immunol*
55. (2025) *Full-Length Text Journal of Virology*
56. Ferretti, Carlo, Ognio et al. (2015) "Interleukin-17A promotes the growth of human germinal center derived non-Hodgkin B cell lymphoma" *Oncoimmunol ogy*
57. Algood, Allen, Washington et al. (2009) "Regulation of gastric B cell recruitment is dependent on IL-17 receptor A signaling in a model of chronic bacterial infection" *J Immunol*
58. Hsu, Yang, Wang et al. (2008) "Interleukin 17-producing T helper cells and interleukin 17 orchestrate autoreactive germinal center develop ment in autoimmune BXD2 mice" *Nat Immunol*
59. Ferretti, Ponzoni, Doglioni et al. (2016) "IL-17 superfamily cytokines modulate normal germinal center B cell migration" *J Leukoc Biol*
60. Xie, Li, Wang et al. (2010) "IL-17 activates the canonical NF-κB signaling pathway in autoimmune B cells of BXD2 mice To upregulate the expression of regulators of G-protein signaling 16" *J Immunol*
61. Mitsdoerffer, Lee, Jäger et al. (2010) "Proinflammatory T helper type 17 cells are effective B-cell helpers" *Proc Natl Acad Sci*
62. Rekow, Darrah, Mboko et al. (2016) "Gammaherpesvirus targets peritoneal B-1 B cells for long-term latency" *Virology (Auckl)*
63. Darrah, Kulinski, Mboko et al. (2017) "B cell-specific ATM expression promotes chronic gammaherpesvirus infection" *J Virol*
64. Kumar, Monin, Castillo et al. (2016) "Intestinal Interleukin-17 receptor signaling mediates reciprocal control of the gut microbiota and autoimmune inflammation" *Immunity*
65. Rickert, Roes, Rajewsky (1997) "B lymphocyte-specific, cremediated mutagenesis in mice" *Nucleic Acids Res*
66. Otero, Anzelon, Rickert (2003) "CD19 function in early and late B cell development: I. Maintenance of follicular and marginal zone B cells requires CD19-dependent survival signals" *J Immunol*
67. Allen, Okada, Cyster (2007) "Germinal-center organization and cellular dynamics"
68. Yu, Ruddy, Wong et al. (2007) "An essential role for IL-17 in preventing pathogeninitiated bone destruction: recruitment of neutrophils to inflamed bone requires IL-17 receptor-dependent signals" *Blood*
69. Lam, Baumgarth (2023) "Toll-like receptor mediated inflammation directs B cells towards protective antiviral extrafollicular responses" *Nat Commun*
70. Mouat, Goldberg, Horwitz (2022) "Age-associated B cells in autoimmune diseases" *Cell Mol Life Sci*
71. Mattos, Vandendriessche, Waisman et al. (2024) "The immunology of B-1 cells: from development to aging" *Immun Ageing*
72. Rothstein, Griffin, Holodick et al. (2013) "Human B-1 cells take the stage" *Ann N Y Acad Sci*
74. Smith, Baumgarth (2019) "B-1 cell responses to infections" *Curr Opin Immunol*
75. Nealy, Coleman, Li et al. (2010) "Use of a virus-encoded enzymatic marker reveals that a stable fraction of memory B cells expresses latency-associated nuclear antigen throughout chronic gammaherpesvirus infection" *J Virol*
76. Lange, Jondle, Darrah et al. (2019) "LXR alpha restricts gammaherpesvirus reactivation from latently infected peritoneal cells" *J Virol*
77. Weck, Kim, Virgin et al. (1999) "Macrophages are the major reservoir of latent murine gammaherpesvirus 68 in peritoneal cells" *J Virol*
78. Li, Ikuta, Sixbey et al. (2008) "A replication-defective gammaherpesvirus efficiently establishes long-term latency in macrophages but not in B cells in vivo" *J Virol*
79. Krug, Collins, Gargano et al. (2009) "NF-κB p50 plays distinct roles in the establishment and control of murine gammaherpes virus 68 latency" *J Virol*
80. Krug, Moser, Dickerson et al. (2007) "Inhibition of NF-κB activation in vivo impairs establishment of gammaherpesvirus latency" *PLoS Pathog*
81. Cieniewicz, Santana, Minkah et al. (2016) "Interplay of murine gammaherpesvirus 68 with NF-κB signaling of the host" *Front Microbiol*
82. Majumder, Amatya, Revu et al. (2019) "IL-17 metabolically reprograms activated fibroblastic reticular cells for proliferation and survival" *Nat Immunol*
83. Thaker, Ch'ng, Hr (2019) "Viral hijacking of cellular metabolism" *BMC Biol*
84. Jayachandran, Mejia, Sheikholeslami et al. (2018) "TAPP adaptors control B cell metabolism by modulating the phosphatidylinositol 3-kinase signaling pathway: A novel regulatory circuit preventing autoimmunity" *J Immunol*
85. Blair, Dufort, Chiles (2012) "Protein kinase Cβ is critical for the metabolic switch to glycolysis following B-cell antigen receptor engagement" *Biochem J*
86. Lam, Bhattacharya (2018) "Metabolic links between plasma cell survival, secretion, and stress" *Trends Immunol*
87. Vivas-García, Efeyan (2022) "The metabolic plasticity of B cells" *Front Mol Biosci*
88. Johnstone, Yazicioglu, Clarke (2024) "Fuelling B cells: dynamic regulation of B cell metabolism" *Curr Opin Immunol*
89. Ma, Cheng, Campbell et al. (1996) "Cloning and characterization of HuR, a ubiquitously expressed Elav-like protein" *J Biol Chem*
90. Brennan, Steitz (2001) "HuR and mRNA stability" *Cell Mol Life Sci*
91. Sun, Mayeda, Hampson et al. (1993) "General splicing factor SF2/ASF promotes alternative splicing by binding to an exonic splicing enhancer" *Genes Dev*
92. Mounce, Tsan, Droit et al. (2011) "Gammaherpesvirus gene expression and DNA synthesis are facilitated by viral protein kinase and histone variant H2AX" *Virology (Auckl)*
93. Tarakanova, Stanitsa, Leonardo et al. (2010) "Conserved gammaherpesvirus kinase and histone variant H2AX facilitate gammaherpesvirus latency in vivo" *Virology (Auckl)*
94. Mboko, Olteanu, Ray et al. (2015) "Tumor suppressor interferon-regulatory factor 1 counteracts the germinal center reaction driven by a cancer-associated gammaherpesvirus" *J Virol*
95. Darrah, Jondle, Johnson et al. (2019) "Conserved gammaherpesvirus protein kinase selectively promotes irrelevant B cell responses" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645995&blobtype=pdf | # A glycoprotein D-targeted lipid nanoparticle-encapsulated mRNA vaccine elicits strong protective immunity against pseudorabies virus
Yue Sun, Shi-Jia Xu, Yongfei Zhou, Yanhe Zhang, Hongliang Zhang, Ting Le, Yuan-Zhe Bai, Cui-Hong Rao, Shanshan Huo, Tianceng Zhou, Tong-Qing An, Xin Yin, Fei Yu, Xue-Hui Cai, Yan-Dong Tang
## Abstract
The pseudorabies virus (PRV), a highly contagious pathogen with zoo notic potential, continues to threaten swine production and public health due to the emergence of virulent variants and insufficient protection conferred by conventional live attenuated vaccines. Although commercial vaccines are safe for pigs, their resid ual pathogenicity in other susceptible species underscores the demand for universally safe alternatives. Here, we engineered a lipid nanoparticle-encapsulated mRNA vaccine (mRNA-LNPs) expressing PRV glycoprotein D (gD) and evaluated its efficacy in murine and porcine models. In mice, vaccination with gD mRNA-LNPs elicited potent neutral izing antibodies and provided complete protection against lethal PRV challenge. In piglets, immunization induced rapid humoral immune responses, significantly reduced viral loads in tissues and viral shedding post-challenge, and alleviated histopathologi cal damage. Mechanistically, except for its ability to elicit neutralizing antibodies, the vaccine also stimulated antigen-specific CD3 + CD4 + T-cell and CD3 + CD8 + T-cell prolifera tion and enhanced IFN-γ production, demonstrating robust activation of both humoral and cellular immunity. These findings establish gD mRNA-LNPs as a safe, effective, and broadly applicable vaccine candidate for PRV control across susceptible species, with advantages in scalability and biosafety over traditional platforms.
IMPORTANCEThe emergence of virulent pseudorabies virus (PRV) variants and the insufficient cross-species protection conferred by conventional live attenuated vaccines pose significant challenges to global swine production and zoonotic biosecurity. Here, we developed a lipid nanoparticle-encapsulated mRNA vaccine (gD mRNA-LNPs) targeting PRV glycoprotein D (gD), a critical mediator of viral entry. This vaccine elicits robust neutralizing antibodies and potent T-cell responses, providing com plete protection against lethal PRV challenge in both murine and porcine models. Unlike traditional vaccines, gD mRNA-LNPs eliminates residual pathogenicity risks and demonstrates broad efficacy against diverse PRV strains, including emerging variants. Its scalable production platform and ability to differentiate vaccinated from infected animals via serological diagnostics align with One Health strategies for PRV eradication. This study establishes mRNA-LNPs technology as a versatile, safe, and effective solution for combating PRV, with implications for improving livestock health and reducing zoonotic spillover threats. KEYWORDS mRNA vaccine, pseudorabies virus, glycoprotein D, vaccine P seudorabies virus (PRV), a member of the Alphaherpesvirinae subfamily (Herpesvir idae), is a highly virulent pathogen that imposes substantial burdens on global swine industries and zoonotic biosecurity (1). As the primary natural host, pigs
exhibit a spectrum of clinical manifestations upon PRV infection, ranging from respira tory distress and neurological dysfunction to reproductive failure and mortality, with severity modulated by age, immunological competence, and viral genotype (2,3). While pigs serve as the natural reservoir, PRV demonstrates alarming cross-species tropism, infecting ruminants, carnivores, and rodents with near-uniform lethality, thereby amplifying its ecological threat (4)(5)(6). The emergence of genotype II variants in China since 2011 has exacerbated economic losses in pig farming and heightened spillover risks, including sporadic human infections, underscoring the urgency for next-genera tion prophylactic strategies (7,8). The most widely utilized commercial PRV vaccines are based on the Bartha K61 strain, which was developed through the attenuation of virulent strains via serial passaging on chicken embryo fibroblasts, similar to other traditional live attenuated vaccines (9)(10)(11). The prevention and control of PRV in China and worldwide mainly rely on the administration of the live attenuated vaccine Bartha K61 (2,3,10). However, since 2011, PRV variants have emerged in China, rendering the efficacy of the Bartha K61 vaccine inadequate in providing sufficient immune protec tion (6). Moreover, different animal species exhibit varying degrees of susceptibility to PRV, with some animals still experiencing high pathogenicity toward live attenuated vaccine (12). Therefore, it is imperative to develop a universal vaccine that ensures safety and effectiveness across all susceptible animal hosts for combating this inter-species transmission.
mRNA vaccine technology has emerged as a transformative platform for combat ing viral pathogens, offering unparalleled advantages in safety, scalability, and rapid adaptability (13). mRNA vaccines can be classified into two types: self-amplifying RNA (saRNA) and non-replicating mRNA. The conventional non-replicating mRNA consists of a cap, 5′-untranslated regions (UTR), open reading frame (ORF) encoding protective antigens, 3′-UTRs, and poly(A) tail. Apart from the ORF, these structural elements play a crucial role in maintaining mRNA stability and transcriptional efficiency. Moreover, they can be modified to extend the half-life of mRNA in vivo and minimize undesired immune responses (13,14). The success of mRNA vaccines against SARS-CoV-2 (15,16), influenza (17), Zika virus (18), and porcine coronaviruses (e.g., PEDV, PDCoV) underscores their versatility and positions them as a paradigm-shifting solution for addressing various viral challenges (19,20).
Here, we present a lipid nanoparticle-encapsulated mRNA vaccine encoding the glycoprotein D (gD) of PRV (gD mRNA-LNPs), engineered to address the limitations of existing vaccines. By leveraging gD as the immunodominant antigen, this candidate induces potent neutralizing antibodies and antigen-specific T-cell responses, conferring cross-species protection against high-dose PRV challenge in mice and piglets. This research not only redefines the role of gD in PRV immunoprotection but also establishes a blueprint for mRNA-based interventions against alphaherpesviruses, with implications for mitigating zoonotic transmission and advancing One Health initiatives.
## RESULTS
## Design and in vitro validation of gD mRNA-LNP
To identify the optimal antigenic target for PRV vaccine development, we first evaluated the immunogenicity of glycoproteins B (gB) and D (gD), which have been reported to contribute significantly to the induction of neutralizing antibodies (21,22). We first expressed the gB and gD proteins in CHO cells. The expression of these proteins was confirmed via SDS-PAGE analysis and by using specific antibodies against gB and gD, respectively (Fig. 1A andB). We believe that an effective vaccine should stimulate the production of antibodies capable of inhibiting the virus's interaction with its receptors, thereby neutralizing it. Our experiment aims to investigate the interactions between viral proteins gB and gD with cell receptors. Subsequently, we test which protein binds to and competes for these receptors. To this end, various concentrations of gB and gD were co-incubated with the PRV-EGFP reporter virus (23). It was observed that a high concentration of the gD protein effectively competes for receptor-binding sites and Five mice per group were immunized, with vaccinations at weeks 0 and 2. Serum samples were collected 2 weeks after the booster. PRV-EGFP reporter virus (200 TCID 50 ) was incubated with 50 µL of twofold serially diluted sera for 2 hours at 37°C. The mixtures were then transferred to a 96-well plate containing a monolayer of Vero E6 cells and incubated for an additional 2 hours at 37°C. After washing, the cells were cultured in DMEM supplemented with 2% FBS. Forty-eight hours post-infection, infected cells were examined under a microscope. Neutralization was observed when serum dilution reached a concentration of 1:16. Scale bars represent 400 µm. This experiment was conducted three times, and a representative result is presented.
(F) Neutralizing antibodies were calculated using the Reed-Muench method at 48 h post-infection. The data represent the means ± standard deviation (SD) from five animals. inhibits PRV replication in Vero E6 cells, whereas the impact of gB is minimal (Fig. 1C). The relative infection rate was quantified by measuring the EGFP-positive area, which demonstrated that the purified gD protein significantly suppressed PRV replication (Fig.
## Full-Length Text
Journal of Virology 1D). Following this, individual mice were immunized with either gB or gD proteins. After the booster immunization, antibodies from the gD-immunized group demonstra ted a significantly higher titer of neutralizing antibodies and completely inhibited PRV infection in Vero E6 cells. In contrast, the antibodies derived from the gB-immunized group exhibited a relatively lower titer of neutralizing antibodies (Fig. 1E andF). Based on these findings, we selected the viral gD gene as our mRNA target for subsequent experiments.
The gD mRNA vaccine was designed according to the schematic shown in Fig. 2A. The gD ectodomain of PRV HeN1 variant strain was cloned into an mRNA vector containing a 5′-cap, as well as 5′ and 3′ UTRs. Furthermore, the tissue plasminogen activator signal sequence was incorporated to enhance the efficient secretion of gD proteins from eukaryotic cells. Subsequently, the gD mRNA transcripts were encapsulated within lipid nanoparticles (LNPs) comprising cationic lipid (SM102), phospholipid with two stearoyl chains and a phosphorylcholine headgroup (DSPC), cholesterol, and DMG-2000. The morphology of the mRNA vaccines is depicted in Fig. 2B. The resulting gD mRNA-LNPs formulations were evaluated for particle size, a critical parameter influencing delivery efficiency and biodistribution. Smaller particles can more readily penetrate tissues and cells, while larger particles may exhibit distinct pharmacokinetic properties. The optimal particle size range for effective delivery is generally between 50 and 200 nanometers. Furthermore, the polydispersity index (PDI), which indicates the uniformity of the particle size distribution, was assessed. A lower PDI value signifies a narrower size distribution, which is essential for ensuring consistent performance and safety. The results demonstra ted that the gD mRNA-LNPs exhibited an average particle size of 100.62 nm with a narrow distribution characterized by a PDI of 0.146 (Fig. 2C andD). To further validate gD expression in vitro, we conducted an indirect immunofluorescence assay (IFA) (Fig. 2E) and Western blot analysis (Fig. 2F) after incubating HEK293T cells with gD mRNA-LNPs or Empty LNPs. The results demonstrated efficient expression of the gD protein, with GAPDH serving as a loading control.
## gD mRNA-LNPs induces sustained humoral and cellular immunity in mice
We next evaluated the immunogenicity of gD mRNA vaccines in mice. A total of 12 mice were randomly divided into two groups (n = 6), with one group receiving immunization using a 10 µg dose of gD mRNA vaccine and subsequent boosting with an equivalent quantity of mRNA after 2 weeks from the initial immunization. Another group was designated as the control group and received empty LNPs for vaccination (Negative control, NC). All mice were vaccinated via intramuscular injection. Serum samples were collected at different time points for PRV antibody detection (Fig. 3A). The presence of PRV gD-specific antibodies was detected using indirect ELISA, and the results demon strated that the gD mRNA immunized group exhibited significantly elevated levels of PRV gD-specific antibodies, which persisted for at least 10 weeks post-vaccination (Fig. 3B). In contrast, no detectable PRV gD-specific antibodies were observed in the control group (Fig. 3B). Additionally, no detectable PRV gB-specific antibodies were observed in the gD mRNA immunized group or control group using indirect ELISA (Fig. 3C). Collectively, these data suggest that the gD mRNA vaccine elicited a robust-specific immune response against gD. Furthermore, a high level of PRV neutralizing antibody was elicited by this vaccine as evidenced by the detected neutralizing antibody titer in the gD mRNA immunized group (Fig. 3D).
To evaluate the T-cell response stimulated by gD mRNA, mouse splenic lymphocytes were isolated and subjected to flow cytometry analysis of antigen-specific lymphocyte proliferation. Therefore, CD3 + CD4 + T-cells and CD3 + CD8 + T-cells were gated for analysis of antigen-specific lymphocyte proliferative responses (Fig. 3E). Subsequently, these lymphocytes were stimulated in vitro with gD mRNA-LNPs (10 µg) or empty LNPs. Additionally, we employed gE mRNA-LNPs (10 µg), another viral gene of PRV, as a more rigorous control. The results demonstrated that the groups immunized with gD mRNA exhibited a significantly higher rate of CD3 + CD4 + T-cell and CD3 + CD8 + T-cell proliferation compared to the control group (Fig. 3F andG). We further confirmed cellular immune responses by measuring IFN-γ secretion from lymphocytes stimulated with gD mRNA-LNPs. It was observed that lymphocytes from mice vaccinated with gD mRNA exhibited significantly higher levels of IFN-γ production compared to controls, which were stimulated with empty LNPs (Fig. 3H). These findings suggest that gD mRNA elicits robust PRV-specific humoral and T-cell responses in mice.
## Complete protection against high-dose PRV challenge in mice
We next investigated the protective efficacy of gD mRNA vaccine against PRV challenge in mice (Fig. 4A). The gD mRNA group received two doses of the gD mRNA vaccine, each at a dosage of 10 µg via intramuscular injection. Our results demonstrated that the gD mRNA immunized group exhibited significantly elevated levels of PRV gD-specific antibodies (Fig. 4B andC) and neutralizing antibodies for the PRV HeN1 strain (Fig. 4D). Next, mice were challenged with the wild-type PRV HeN1 strain via intramuscular injection. In this study, we challenged mice with a high dose of wild-type PRV (2 × 10 4 PFU). We observed that all mice immunized with the gD mRNA vaccine survived after challenge; however, the empty LNPs immunized group experienced 100% mortality following challenge (Fig. 4E). The clinical signs were assessed using clinical scores (Fig. S1). Additionally, viral loads in indicated tissues were detected, revealing significantly lower viral copies in the gD mRNA vaccine group compared to the empty LNPs-immu nized group (Fig. S2). Furthermore, histopathological changes in the brain and lungs were evaluated. Our findings indicated that no histopathological alterations were observed in either the gD mRNA vaccine group or the negative control group. However, the group immunized with empty LNPs exhibited significant microscopic lesions in both the brain and lungs. Specifically, the alterations observed in the brains of mice treated with empty LNPs are characterized by glial cell proliferation and marked degeneration of neuronal cells (Fig. 4F). In the lungs of mice administered empty LNPs, congestion and infiltration by inflammatory cells lead to widening of alveolar septa, accompanied by partial degeneration and necrosis of bronchial epithelial cells. Additionally, fibrinous exudate is present within the alveolar spaces (Fig. 4F). Overall, the gD mRNA vaccine provided full protection against PRV challenge.
## gD mRNA vaccine provides protection against PRV infection in mice via respiratory route
PRV typically spreads through the respiratory route. Therefore, we subsequently evaluated whether this vaccine induces mucosal protection. Following two rounds of immunization, we challenged the vaccinated mice with wild-type PRV (1 × 10 3 PFU) via the intranasal route (Fig. 5A). The results demonstrated that all mice immunized with the gD mRNA vaccine survived following the intranasal challenge. In contrast, the group that received empty LNPs exhibited a 100% mortality rate after undergoing the same intranasal challenge (Fig. 5B). The clinical signs were further evaluated using clinical scores (Fig. 5C). Additionally, we assessed viral loads in the specified tissues. The results indicated a significantly lower number of viral copies in the gD mRNA vaccine group compared to the empty LNPs immunized group (Fig. 5D). Due to the potential for PRV infection to induce latent infections within the nervous system, we also evaluated viral loads in the trigeminal nerve, dorsal root ganglion, and sciatic nerve of mice that were immunized with gD mRNA-LNPs or empty LNPs following intranasal challenge. The results indicated that the gD mRNA vaccine provided complete protection against PRV intranasal challenge (Fig. 5E through G).
## gD mRNA vaccine elicits high levels of PRV-specific humoral and T-cell response in piglets post vaccination
We next evaluated the immunogenicity of gD mRNA vaccines in piglets. The experimen tal timeline and design are schematically illustrated in Fig. 6A. There was no significant difference observed in the average daily weight gain between the group immunized with gD mRNA vaccines and the control group which was vaccinated with empty LNPs (Fig. 6B). Additionally, gD-specific antibodies were detectable 14 days post-vac cination (Fig. 6C). As a control, we also assessed the presence of gB antibodies; how ever, no gB antibodies were detected (Fig. 6D). Moreover, a robust PRV neutralizing antibody response was elicited at day 21 after gD mRNA immunization (Fig. 6E). To characterize cellular immunity, we performed lymphocyte proliferation assays (Fig. S3). Flow cytometric analysis showed significantly enhanced CD3 + CD4 + T-cell and CD3 + CD8 + T-cell proliferation indices in vaccinated piglets compared to controls (Fig. 6F andG). Furthermore, we confirmed cellular immune responses by measuring IFN-γ secretion from lymphocytes stimulated with gD mRNA-LNPs. Notably, piglets vaccinated with gD mRNA exhibited significantly elevated levels of IFN-γ production compared to control piglets (Fig. 6H). These findings indicate that in piglets, gD mRNA vaccine elicits potent PRV-specific humoral and T-cell responses.
## gD mRNA vaccine prevents clinical disease and viral shedding in piglets
We next investigated the protective efficacy of gD mRNA vaccine against PRV challenge in piglets. gD mRNA-LNPs-immunized group and the empty LNPs-vaccinated group were challenged with wild-type PRV HeN1 strain. The mock-infected group, serving as a negative control, was not exposed to the virus during the experimental period.
It was observed that all piglets immunized with the gD mRNA vaccine survived after challenge; however, the empty LNPs-vaccinated group experienced 100% mortality following challenge (Fig. 7A). The rectal temperature and clinical signs were recorded (Fig. 7B andC). Furthermore, gB-specific antibodies and gE-specific antibodies were detectable 10 days post-challenge in the group vaccinated with gD mRNA vaccine (Fig. 7D). Interest ingly, there was a significant increase in neutralizing antibodies in the group vaccina ted with gD mRNA-LNPs post-challenge (Fig. 7E). Furthermore, viral loads in serum were assessed and demonstrated lower viral copies in the gD mRNA vaccine group compared to the empty LNPs-vaccinated group (Fig. 7F). Additionally, virus shedding was evaluated by detecting viral copies in nasal and anal swabs, revealing a lower level of virus shedding in the gD mRNA vaccine group (Fig. 7G). Moreover, viral loads in indicated tissues were also evaluated, demonstrating lower viral loads in the gD mRNA vaccine group compared to the empty LNPs-vaccinated group (Fig. S4). Furthermore, histopathological changes in the brain and lungs were examined, with no significant histopathological alterations observed in the gD mRNA vaccine group, which were similar to those in the mock-infected group (Fig. S5). However, severe microscopic lesions were observed in the brain and lungs of piglets challenged with wild-type PRV in the empty LNPs-vaccinated group (Fig. S5). In pigs treated with empty LNPs, notable changes in brain tissue include proliferation of glial cells, extensive degeneration and necrosis of neuronal cells, presence of vascular cuffs, as well as occasional necrosis among glial cells. The alterations identified in the lungs of pigs receiving empty LNPs encompass congestion, significant perivascular edema, infiltration by inflammatory cells, along with fibrinous exudate within the alveolar lumen that is associated with necrotic shedding from alveolar epithelial cells (Fig. S5). The results of immunohistochemistry indicate that, in the gD mRNA vaccine group, viral antigens were undetectable (Fig. 7H). Overall, the gD mRNA vaccine provided complete protection against PRV challenge in piglets.
## gD mRNA vaccine elicits cross-neutralizing antibodies against diverse PRV strains
To evaluate the breadth of humoral immunity induced by the gD mRNA vaccine, sera collected at 28 days post-vaccination (dpv) from immunized mice and piglets were analyzed for cross-reactivity and neutralizing activity against heterologous PRV strains. Initially, we evaluated the conservation of gD amino acid sequences across various PRV strains. We retrieved 37 gD amino acid sequences from the NCBI GenBank database, which included 4 sequences from genotype I, 5 sequences from genotype II classic strains, and 28 sequences from genotype II variant strains. Subsequently, we employed MegaAlign software to analyze the conservation of the gD protein's amino acid sequence among different PRV strains. The results indicated that the sequence of the PRV gD protein is conserved, with only a few amino acids exhibiting mutations across all aligned sequences (Fig. 8A andB). To evaluate the cross-reactivity and broad-spectrum neutralizing activity elicited by the gD mRNA vaccine, we selected HeN1, SC, TJ, and Bartha K61 from various branches within the phylogenetic tree. IFA revealed that sera from both gD mRNA-vaccinated mice (Fig. S6A) and piglets (Fig. S6B) exhibited robust cross-reactivity with four genetically distinct PRV strains: HeN1, SC, TJ, and Bartha K61.
Neutralization assays further demonstrated broad-spectrum protective potential. Serum from vaccinated mice (Fig. 8C) and piglets (Fig. 8D) effectively neutralized SC, TJ, and Bartha K61 strains, with neutralization titers comparable to those observed against the homologous HeN1 strain. This cross-neutralizing activity correlated with the conserved epitope recognition profile observed in IFA, confirming that gD mRNA immunization elicits antibodies targeting conserved antigenic regions across divergent PRV strains.
## DISCUSSION
In this study, we have demonstrated the safety and efficacy of an mRNA vaccine against PRV in both mice and piglets. Additionally, our study has shown that these vaccines not only induce potent neutralizing antibodies but also stimulate robust specific T-cell responses. Indeed, several DNA vaccines have been developed utilizing plasmid DNA encoding the gB, gC, gD, or gE proteins of PRV; however, these vaccines have demonstra ted a moderate level of protection in pigs against PRV infection (24,25). In our study, we demonstrated that the gD subunit protein elicited a robust humoral response (Fig. 1F). We think there are two primary advantages of the gD mRNA-LNPs vaccine compared to the gD subunit protein vaccine. First, the gD mRNA-LNPs can effectively enter host In fact, for PRV infection, both neutralizing antibodies and T-cell responses are essential components of vaccine-mediated protection. Neutralizing antibodies primarily target extracellular viral particles, effectively preventing the spread of the virus; however, their impact on intracellular viruses is limited (26). Conversely, T-cell immune responsesparticularly those mediated by CD8 T cells-are capable of eliminating infected target cells, thereby reducing potential sites for viral replication. It has been reported that both gB and gD contribute significantly to the induction of neutralizing antibodies (21,22). The primary role of gD is to directly interact with cellular receptors and other cellular molecules, facilitating the attachment and entry of PRV (27,28). Meanwhile, gB primarily facilitates fusion between the viral envelope and the host cell membrane, thereby nucleocapsids and tegument proteins are released into the host cell cytoplasm for subsequent replication (29). Our current study demonstrates that gD elicits a robust and high level of neutralizing antibodies, in contrast to gB. While purified soluble gD and gB, overexpressed in CHO cells, exhibit differential abilities to induce neutralizing antibodies, this observation does not preclude the potential of gB mRNA-based vaccines to provide protection. Furthermore, the gB antigen may be effective in inducing T-cell-mediated immunity or potentially facilitating the production of non-neutralizing antibodies. These antibodies have demonstrated significant efficacy in mediating effector functions such as antibody-dependent cellular cytotoxicity and antibody-dependent cellular phagocy tosis. Additionally, our recent findings suggest that except for gB, other viral antigens may be critical for PRV protection (30). We utilized anticodon-engineered transfer RNAs to generate live but replication-defective viruses lacking gB expression, which still provided full protection against PRV challenge (30). In fact, the levels of neutralizing antibodies induced by this experiment were also higher than those stimulated by replication-defective vaccines (30). However, immunization with replication-defective viruses in mice still provides adequate immune protection, indicating that T-cell immune responses are also crucial for immune protection against PRV in mice (31). In our study, we also demonstrated that the gD mRNA vaccine effectively stimulated a robust T-cell response.
PRV has been successfully eradicated in several countries, primarily through the implementation of vaccine-based serological diagnostics (3). Consequently, the gD mRNA vaccine becomes more feasible using serological diagnostics to differentiate from PRV infection. This vaccine will accelerate the global eradication of PRV. Therefore, it is helpful to develop a serological diagnostic method targeting gD in the future. Furthermore, by examining antibodies against other viral proteins, it becomes possible to differentiate between immune responses resulting from gD mRNA vaccination and by gD mRNA-LNPs and flow cytometry analysis of antigen-specific lymphocyte proliferation. Peripheral blood lymphocytes from piglets were isolated and stained with CellTrace Violet Cell Proliferation Kit. Subsequently, 1,000,000 stained cells were stimulated with gD mRNA-LNPs (10 µg) at 37°C with 5% CO2 for 72 hours. Empty LNPs stimulated cells were used as negative controls. The cells were then stained with anti-porcine CD3, CD4, and CD8 antibodies. After three washes with PBS, flow cytometric analysis was performed. The data represent the means ± standard deviation (SD) from three independent experiments. (H) IFN-γ production in piglets vaccinated with gD mRNA vaccine. The cellular immune responses in the vaccinated piglets were evaluated using IFN-γ enzyme-linked immunospot (ELISPOT). Briefly, 300,000 PBMCs were stimulated with gD mRNA-LNPs (5 µg), or empty LNPs, followed by incubation at 37°C for 48 hours. Empty LNPs stimulated cells were used as negative controls. The plates were washed and then incubated with a detection antibody. Streptavidin-HRP was added to the plates and incubated. After final washes, the TMB substrate solution was applied and subsequently stopped with water. The air-dried plates were analyzed using an ELISPOT reader to determine the number of spot-forming cells per million cells. The data represent the means ± standard deviation (SD) from three independent experiments. those arising from wild virus infection. In conclusion, we developed a gD mRNA-LNPs vaccine, and this vaccine provides protective immunity against PRV challenge in both mice and piglets and elicits high levels of neutralizing antibodies and specific T-cell responses, indicating its potential as a new and promising vaccine candidate for the fight against PRV.
## MATERIALS AND METHODS
## Cells, viruses, proteins, and antibodies
Human embryonic kidney cells (HEK293T, ATCC CRL-11268) and African green monkey kidney cells (Vero-E6, ATCC CRL-1586) were cultured in Dulbecco's modified Eagle's medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Excell, Australia) at a temperature of 37°C with a CO 2 concentration of 5%. The PRV HeN1 strain (GenBank Accession No. KP098534), SC (GenBank Accession No. KT809429), TJ strain (GenBank Accession No. KJ789182), and Bartha K61 strain (GenBank Accession No. JF797217) were stored at -80°C in our laboratory. The PRV HeN1 EGFP strain was maintained as described in our previous report (29,32). The Anti-GAPDH antibody was procured from Proteintech (Proteintech, China). PE/Cyanine5 anti-mouse CD3 antibody and APC/Cyanine7 anti-mouse CD8a antibody were procured from Biolegend (Biole gend, USA). The FITC rat anti-mouse CD4 antibody was sourced from BD-Pharmingen (BD-Pharmingen, USA). Mouse anti-porcine CD3E-SPRD antibody, mouse anti-porcine CD4-FITC antibody, and mouse anti-porcine CD8A-PE antibody were obtained from SouthernBiotech (SouthernBiotech, USA). Anti-mouse IgG-FITC was acquired from Sigma (Sigma, USA). Additionally, mouse monoclonal antibodies specific for the PRV gB were generously provided by Prof. Zhi-jun Tian and Jinmei Peng at our institute.
## Construction of the expression plasmids and purification of the protein
The gB and gD genes of the PRV HeN1 strain (GenBank Accession: KP098534) were codon-optimized and synthesized by Sangon Biotech (Sangon Biotech, China). Subsequently, the ectodomain of either gB or gD was cloned into the pb513B vector, which incorporates a Flag tag along with a 6× His tag. A stable CHO cell line expressing these constructs was then established as previously described (33). The CHO cells stably expressing the recombinant protein were established through puromycin selection. The recombinant protein present in the supernatant of cultured cells was collected and purified using Ni-NTA resin affinity chromatography (GenScript, USA) according to the manufacturer's instructions. SDS-PAGE and Western blot analyses were conducted to confirm the purity of the isolated protein. The concentration of the purified proteins was quantified utilizing a BCA Protein Assay Kit (Solarbio, China), after which these proteins were stored at -80°C.
## PRV replication was inhibited by protein
To investigate the effect of gB or gD on viral replication, different concentrations of gB or gD protein were co-incubated with the PRV-EGFP reporter virus (multiplicity of infection = 0.01) on Vero E6 cells for 2 hours. After washing three times with cold PBS, the samples were cultured in DMEM containing 2% FBS. PBS was used as the control. After 36 hours of infection, the infected cells were examined under a microscope. The extent of infection was determined by quantifying the area occupied by EGFP-positive cells using Image J software.
## Design of the mouse vaccination experiments by the gB and gD protein
For the vaccination of BALB/c mice, a total of 15 female BALB/c mice were randomly assigned to three groups. The mice received intramuscular injections of 100 µL (100 µg) of purified gB or gD protein in the hind leg, or phosphate-buffered saline (PBS), at weeks 0 and 2. During the initial immunization, the proteins were emulsified with Freund's complete adjuvant (Sigma, USA). In subsequent booster administrations, they were combined with Freund's incomplete adjuvant (Sigma, USA). Blood samples were collected from the retro-orbital sinus using glass capillaries. Serum samples were obtained 2 weeks after booster immunization and stored at -20 °C for neutralizing antibody detection.
## Immunofluorescence assay
HEK293T cells treated with gD mRNA-LNPs or empty LNPs were washed with PBS and fixed with 4% paraformaldehyde (PFA) for 15 minutes at room temperature and washed twice with PBS as described in our previous studies (34)(35)(36)(37)(38)(39). Then, the cells were permeabilized with 0.25% Triton X-100 for 10 minutes at 4°C and blocked with 2% bovine serum albumin (BSA) (Coolaber, China) for 30 minutes at 37°C. After washing three times with PBS, the cells were subsequently incubated with PRV gD-specific primary antibodies at 37°C for 1 hour and washed three times with PBS. The cells were then incubated with appropriate secondary antibodies at 37°C for 1 hour and washed three times with PBS. The cell nuclei were stained with 4′,6-diamino-2-phenylindole (DAPI) for 15 minutes at 37°C and washed three times with PBS. Finally, the fluorescence signals were detected via microscope.
## Western blot
The western blot was performed as described in our previous reports (33,(40)(41)(42). Briefly, HEK293T cell lysates or PRV gB and gD proteins were mixed with loading buffer for SDS-PAGE. The separated proteins were transferred to PVDF membranes (Millipore, USA). The membranes were blocked with 5% nonfat milk in PBS for 1 hour and incubated with an antibody overnight at 4°C, followed by washing and incubation with the appropriate secondary antibody for 1 h at room temperature. Finally, the membranes were visualized using the Odyssey CLx imaging system.
## mRNA production and LNP encapsulation
The glycoprotein D of the HeN1 strain of PRV was employed as the reference amino acid sequence (GenBank Accession: KP098534). The ectodomain (1-355 aa) was cloned into an mRNA vector, followed by in vitro transcription using T7 RNA polymerase and a linearized plasmid DNA template containing optimized codons for the glycoprotein D. The lipids, including cationic lipid (SM102), phospholipid with two stearoyl chains and a phosphorylcholine headgroup (DSPC), cholesterol, and DMG-2000, were dissolved in anhydrous ethanol at a mass ratio of 50:10:38.5:1.5. Subsequently, the mRNA was dissolved in an aqueous phase prepared using citrate buffer solution at pH 4.0 (50 mM). The microfluidic device (LNP-S1-L, Fluidiclab, China) was utilized to package the mRNA with an alcohol phase to water phase ratio of 1:3 (43). The resulting mixture was then diluted with RNase-free PBS buffer and concentrated through ultrafiltration employing a 30 kDa cutoff membrane. To adjust the mRNA concentration to 60 µg/mL and the sucrose concentration to 10%, an equal volume of PBS solution containing 20% sucrose was added. Finally, the gD mRNA-LNPs vaccine preparation underwent filtration through a 0.22 µm membrane before being stored at -20°C after packaging. The particle size and PDI were measured by NanoBrook 90Plus device (Brookhaven Instruments Corporation, USA) according to the instructions.
## Expression of mRNA encoding the PRV gD
HEK293T cells were seeded in 12-well plates. When the cells reached 70%-80% confluence, cells were treated with 2 µg gD mRNA-LNPs. The empty LNPs were served as the control group. After 24 hours of incubation, the expression of mRNA was analyzed by IFA and western blot (30,44).
## Design of the mouse vaccination experiments and viral challenge study
## Safety evaluation
In all, 12 BALB/c female mice (6 weeks old) were randomly assigned to two groups of 6 mice per group. The immunization schedule is shown in Fig. 3A, one group vaccinated via intramuscular (i.m.) injection with a volume of 150 µL (10 µg) gD mRNA-LNPs at the hind leg. Mice in the control group were given 150 µL empty LNPs. A booster immuni zation was performed 2 weeks after primary immunization. The blood samples were obtained from the retro-orbital sinus using glass capillaries. Serums were collected at 0, 2, 3, 4, 5, 6, 7, 8, 9, and 10 weeks post-vaccination for antibody analysis. The mice were euthanized with inhaled CO 2 , and then the antigen-specific lymphocyte proliferation assays and IFN-γ detection were conducted using splenocytes that were extracted at 4 weeks post-first vaccination.
## Intramuscular challenge
To investigate the protective efficacy of gD mRNA-LNPs against PRV challenge via intramuscular infection, 15 BALB/c female mice were randomly divided into three groups. The experimental schedule is shown in Fig. 4A, groups 1 and 2 (n = 5) were immunized with a volume of 150 µL gD mRNA-LNPs (10 µg) or 150 µL empty LNPs via i.m. injection at the hind leg, respectively. Group 3 was given a volume of 150 µL empty LNPs and designated as the negative control. All groups were immunized twice with the same dose, 14 days apart. The blood samples were obtained from the retro-orbital sinus using glass capillaries. Serums were collected for antibody analysis. Groups 1 and 2 were challenged with a volume of 100 µL PRV HeN1 strain (2 × 10 4 PFU) via i.m. injection at the hind leg at 28 days post the first immunization. Group 3 served as the mock challenge group and was not subjected to viral challenge during experiments. After the challenge, the mice were observed daily, and the clinical symptoms were scored as in previous research (22). At 14 days post-challenge (dpc), the surviving mice were euthanized with inhaled CO 2 , and samples of six tissues, including brain, lung, heart, liver, spleen, and kidney, were collected for viral detection via quantitative real-time polymerase chain reaction (qRT-PCR). The brain and lung were subjected to histopathological examination following hematoxylin and eosin (H&E) staining.
## Mucosal challenge
To investigate the mucosal protective effects of gD mRNA-LNPs against PRV challenge via the respiratory route, 10 male and 5 female BALB/c mice were randomly assigned to three groups. The experimental schedule is illustrated in Fig. 5A. Groups 1 and 2 (n = 5) received immunizations with gD mRNA-LNPs or empty LNPs, respectively, while Group 3 was administered empty LNPs and served as the negative control. The immunization dose, route of administration, and intervals for the mice were conducted as previously described. At 28 days post-initial immunization, Groups 1 and 2 underwent intranasal challenge with a volume of 20 µL containing the PRV HeN1 strain (1 × 10 3 PFU). Following the challenge, daily observations were made on the mice to assess clinical symptoms, which were subsequently scored. 14 days post-challenge (dpc), surviving mice were euthanized using inhaled CO 2 ; tissues were then collected for viral detection through qRT-PCR analysis.The brain and lung were subjected to histopathological examination following hematoxylin and eosin (H&E) staining.
## Design of the pig vaccination experiments and viral challenge study
## Safety evaluation
In all, 15 piglets (28 days old, 9 males and 6 females) were free of PRV, PRRSV, ASFV, PCV2, PCV3, and CSFV. The piglets were randomly divided into three groups (n = 5). The experimental schedule is shown in Fig. 6A, group 1 was immunized via intramuscular (i.m.) injection with a volume of 1 mL (50 µg) gD mRNA-LNPs at neck muscles. Groups 2 and 3 were given a volume of 1 mL empty LNPs at the neck muscles. The vaccinated piglets also received a booster dose at 14 dpv. The piglets' body weights were measured weekly. Furthermore, serum samples were collected from each piglet at 0, 7, 14, 21, and 28 dpv for antibody detection. The piglet peripheral blood lymphocytes were isolated to measure the antigen-specific T-cell proliferation and IFN-γ detection.
## Intramuscular challenge
At 28 dpv, groups 1 and 2 were challenged with a volume of 1 mL PRV HeN1 strain via intramuscular injection (1.0 × 10 6.0 PFU per piglet) in the neck muscles. Group 3 served as the mock challenge group and was not subjected to viral challenge during experi ments. The animals were recorded for clinical signs and rectal temperatures. The clinical symptoms were scored as in previous research (45). The rectal temperature ≥40.5°C was defined as fever. Furthermore, the blood samples, nasal swab, and anal swab were collected from piglets at 0, 3, 5, 7, 10, and 14 days post-challenge (dpc). All surviving piglets were euthanized using Zoletil 50 (Virbac, France) (7 mg/kg) via i.m. injection at the neck muscles and autopsied at 14 dpc. The serum samples were collected from individual piglets to test for the presence of viremia. The nasal swab, anal swab, and samples of nine tissues, including brain, heart, liver, spleen, lung, kidney, tonsil, submandibular lymph node, and inguinal lymph node, were collected for viral detection via qRT-PCR. The brain and lung were collected and subjected to histopathology and immunohistochemistry analysis. Our institute is equipped with a team of professional pathologists. When we provide tissue samples, we assign random numbers and do not disclose detailed information to ensure the integrity of the analysis. Consequently, histopathological slides are examined in a blinded manner.
## Enzyme-linked immunosorbent assay
PRV gD-, gB-, and gE-specific antibodies were detected via indirect enzyme-linked immunosorbent assay (ELISA) as shown in previous research (46)(47)(48). The proteins were expressed in CHO cells, and purified proteins (100 ng/well) were coated onto 96-well ELISA plates (Corning, USA) overnight at 4°C. The plates were blocked with 3% BSA for 2 h at 37°C. The diluted serum sample (1:200) was added to the plates and incubated at 37°C for 1 h. After washing three times with PBST, 5 min for each time, and incubating at 37°C for 45 min with HRP-conjugated secondary antibodies (ZSGB-BIO, China). After another washing step, TMB solution (Abcam, USA) was added to each well and incubated under dark conditions for 10 min at room temperature. The reaction was stopped with 2 M H 2 SO 4 , and the absorbance was read on a microtiter plate reader (PE, USA) at 450 nm within 10 min.
## Virus-neutralizing antibody detection
Serum-neutralizing antibodies against PRV were determined by performing a neutraliza tion assay. The assay was performed as previously described with minor modifications (30,49). Serum samples from mice and piglets were inactivated by heating at 56°C for 30 min and then twofold serial dilutions with DMEM. A similar amount of PRV HeN1 EGFP strain, HeN1 strain, SC strain, TJ strain, or Bartha K61 strain (200 TCID 50 ) was mixed thoroughly with 50 µL of the diluted inactivated serum for 2 h at 37°C. Subsequently, the mixtures of 100 µL were added to a 96-well plate with monolayer Vero E6 cells for 2 h at 37°C. Two hours later, the cells were washed three times with cold PBS and cultured in DMEM containing 2% FBS. Blank cells were set up as a negative control. The plates were incubated for 48 hours at 37°C with 5% CO 2 . Neutralizing antibodies were calculated using the Reed-Muench method.
## Enzyme-linked immunospot assay
The cellular immune responses in the vaccinated piglets were evaluated using IFN-γ precoated enzyme-linked immunospot (ELISPOT) kits (Mabtech, USA) following the manufacturer's instructions. Briefly, the plates were blocked with RPMI 1640 medium (Gibco, USA) containing 10% FBS and incubated for 30 minutes. Peripheral blood lymphocytes from piglets were seeded at a density of 300,000 cells per well and stimulated with gD mRNA-LNPs (5 µg) or empty LNPs followed by incubation at 37°C with 5% CO 2 for 48 hours.
Empty LNPs stimulated cells were used as negative controls. The plates were washed five times with PBS and then incubated at room temperature for 2 hours with a detection antibody. Streptavidin-HRP was added to the plates and incubated for an additional hour. After final washes, the TMB substrate solution was applied and subsequently stopped with water. The air-dried plates were analyzed using an ELISpot reader (AID, Germany) to determine the number of spot-forming cells per million cells.
## IFN-γ detection
The spleen lymph single-cell suspensions from mice were seeded into 12-well plates at a density of 1,000,000 cells per well and incubated with gD mRNA-LNPs (10 µg) and Empty LNPs for 72 hours at 37°C. The cells stimulated by empty LNPs were used as the negative control. Subsequently, the cells were centrifuged and the supernatants were collected to measure IFN-γ levels using commercial mouse IFN-γ ELISA kits (Cusabio, China), following the manufacturer's instructions. The concentration of IFN-γ in the samples was determined based on standard curves.
## Quantitative real-time polymerase chain reaction
The viral DNA from blood, tissues, and swab samples was extracted using the TIANamp Genomic DNA Kit (Tiangen, China) following the manufacturer's instructions. Quantification of the copy numbers of PRV's gB gene was performed using qRT-PCR, as previously described in our reports (23,29,40).
## Assessment of the broad-spectrum efficacy of gD mRNA vaccine against various PRV strains
A total of 37 gD amino acid sequences were obtained from the NCBI GenBank database, comprising 4 sequences of genotype I, 5 sequences of genotype II classic strains, and 28 sequences of genotype II variant strains. Subsequently, the MegaAlign software was employed to analyze the conservation of the gD protein amino acid sequence across various PRV strains, while phylogenetic analysis was conducted using MEGA software. Sera collected at 28 days post-vaccination (dpv) from immunized mice and piglets were evaluated through IFA and neutralization assays against several PRV strains, including HeN1, SC, TJ, and Bartha K61, which represent different branches in the phylogenetic tree.
## Statistical analysis
Statistical analysis was conducted via GraphPad Prism 8.0 software. Statistical significance was analyzed via t tests. A P value < 0.05 was considered statistically significant. Statistical significance is indicated by * P < 0.05, ** P < 0.01, *** P < 0.001, and **** P < 0.0001.
## References
1. Davison, Eberle, Ehlers et al. (2009) "The order Herpesvirales"
2. Freuling, Müller, Mettenleiter (2017) "Vaccines against pseudorabies virus (PrV)" *Vet Microbiol*
3. Mettenleiter (2020) "Aujeszky's disease and the development of the marker/DIVA vaccination concept" *Pathogens*
4. Ye, Zhang, Tian et al. (2015) "Genomic characterization of emergent pseudorabies virus in China reveals marked sequence divergence: evidence for the existence of two major genotypes" *Virology (Auckl)*
5. Tan, Yao, Yang et al. (2021) "Current status and challenge of pseudorabies virus infection in China" *Virol Sin*
6. An, Peng, Tian et al. (2012) "Pseudorabies virus variant in Bartha-K61vaccinated pigs" *Emerg Infect Dis*
7. Wong, Lu, Zhang et al. (2019) "Pseudorabies virus: a neglected zoonotic pathogen in humans?" *Emerg Microbes Infect*
8. Guo, Chen, Zhang (2021) "Human PRV infection in China: an alarm to accelerate eradication of PRV in domestic pigs" *Virol Sin*
9. Tang, Li, Cai et al. (2025) "Viral live-attenuated vaccines (LAVs): past and future directions" *Adv Sci (Weinh)*
10. Delva, Nauwynck, Mettenleiter et al. (2020) "The attenuated pseudorabies virus vaccine strain bartha K61: a brief review on the knowledge gathered during 60 years of research" *Pathogens*
11. Bai, Xu, Liu et al. (2025) "Nsp2 replicasemediated viral uncoating in porcine alveolar macrophages contributes to the attenuation of PRRSV-2 live attenuated vaccine" *J Virol*
12. Yin, Li, Zhang et al. (2020) "Construction of a US7/US8/UL23/US3-deleted recombinant pseudorabies virus and evaluation of its pathogenicity in dogs" *Vet Microbiol*
13. Wang, Zhang, Luo et al. (2021) "mRNA vaccine: a potential therapeutic strategy" *Mol Cancer*
14. Pardi, Hogan, Weissman (2020) "Recent advances in mRNA vaccine technology" *Curr Opin Immunol*
15. Baden, Sahly, Essink et al. (2021) "Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine" *N Engl J Med*
16. Jackson, Anderson, Rouphael et al. (2020) "An mRNA vaccine against SARS-CoV-2 -preliminary report" *N Engl J Med*
17. Kelvin, Falzarano (2022) "The influenza universe in an mRNA vaccine" *Science*
18. Richner, Himansu, Dowd et al. (2017) "Modified mRNA vaccines protect against zika virus infection" *Cell*
19. Zhao, Song, Gao et al. (2024) "PEDV-spike-protein-expressing mRNA vaccine protects piglets against PEDV challenge" *mBio*
20. Li, Xiao, Chen et al. (2024) "A spike-based mRNA vaccine that induces durable and broad protection against porcine deltacoronavirus in piglets" *J Virol*
21. Li, Yang, Hu et al. (2017) "Two classes of protective antibodies against Pseudorabies virus variant glycoprotein B: Implications for vaccine design" *PLoS Pathog*
22. Zhang, Liu, Chen et al. (2020) "A single dose glycoprotein D-based subunit vaccine against pseudorabies virus infection" *Vaccine (Auckl)*
23. Tang, Guo, Wang et al. (2018) "CRISPR/Cas9-mediated 2-sgRNA cleavage facilitates pseudorabies virus editing" *FASEB J*
24. Gerdts, Jöns, Mettenleiter (1999) "Potency of an experimental DNA vaccine against Aujeszky's disease in pigs" *Vet Microbiol*
25. Gerdts, Jöns, Makoschey et al. (1997) "Protection of pigs against Aujeszky's disease by DNA vaccination" *J Gen Virol*
26. Na, Zheng, Yang et al. (2025) "The neonatal Fc receptor (FcRn): guardian or trojan horse in viral infection?" *PLoS Pathog*
27. Campadelli-Fiume, Cocchi, Menotti et al. (2000) "The novel receptors that mediate the entry of herpes simplex viruses and animal alphaherpesviruses into cells" *Rev Med Virol*
28. Cairns, Connolly (2021) "Entry of alphaherpesviruses" *Curr Issues Mol Biol*
29. Chen, Wang, Shen et al. (2022) "Neuropilin-1 facilitates pseudorabies virus replication and viral glycoprotein B promotes its degradation in a furin-dependent manner" *J Virol*
30. Wang, Meng, Sang et al. (2023) "A novel viral vaccine platform based on engineered transfer RNA" *Emerg Microbes Infect*
31. Tang, Yu, Cai (2024) "Novel technologies are turning a dream into reality: conditionally replicating viruses as vaccines" *Trends Microbiol*
32. Tang, Liu, Fang et al. (2016) "Recombinant pseudorabies virus (PRV) expressing firefly luciferase effectively screened for CRISPR/Cas9 single guide RNAs and antiviral compounds" *Viruses*
33. Yang, Liu, Wang et al. (2021) "Aminopeptidase N is an entry co-factor triggering porcine deltacoronavirus entry via an endocytotic pathway" *J Virol*
34. Bai, Sun, Liu et al. (2024) "Minor envelope proteins from GP2a to GP4 contribute to the spread pattern and yield of type 2 PRRSV in MARC-145 cells" *Front Cell Infect Microbiol*
35. Rao, Huang, Bai et al. (2024) "MARCH8 inhibits pseudorabies virus replication by trapping the viral cell-to-cell fusion complex in the trans-Golgi network" *Int J Biol Macromol*
36. Huang, Rao, Bai et al. (2024) "MARCH1 and MARCH2 inhibit pseudorabies virus replication by trapping the viral cell-to-cell fusion complex in trans-Golgi network" *Vet Microbiol*
37. Zhang, Liang, Wang et al. (2022) "SARS-CoV-2 hijacks macropinocytosis to facilitate its entry and promote viral spikemediated cell-to-cell fusion" *J Biol Chem*
38. Zhang, Li, Sun et al. (2021) "Evaluating angiotensin-converting enzyme 2-mediated SARS-CoV-2 entry across species" *J Biol Chem*
39. Bai, Wang, Sun et al. (2025) "The full-length nsp2 replicase contributes to viral assembly in highly pathogenic PRRSV-2" *J Virol*
40. Wang, Yang, Feng et al. (2020) "Pseudorabies virus UL24 abrogates tumor necrosis factor alphainduced NF-κB activation by degrading P65" *Viruses*
41. Sun, Sun, Yang et al. (2024) "Multivalent nanobody-based sandwich enzyme-linked immunosorbent assay for sensitive detection of porcine reproductive and respiratory syndrome virus" *Int J Biol Macromol*
42. Yang, Meng, Qin et al. (2020) "Trypsin promotes porcine deltacoronavirus mediating cell-to-cell fusion in a cell type-dependent manner" *Emerg Microbes Infect*
43. Li, Jia, Lai et al. (2025) "Injectable microenvironment-responsive hydrogels encapsulating engineered NF-κB-targeting circular RNA for osteoarthritis therapy" *J Control Release*
44. Wang, Fang, Cong et al. (2019) "The Nsp12-coding region of type 2 PRRSV is required for viral subgenomic mRNA synthesis" *Emerg Microbes Infect*
45. Ren, Wang, Zhou et al. (2020) "Glycoproteins C and D of PRV strain HB1201 contribute individually to the escape from Bartha-K61 vaccine-induced immunity" *Front Microbiol*
46. Wang, Wang, Zhang et al. (2019) "Intranasal administration with recombinant Bacillus subtilis induces strong mucosal immune responses against pseudorabies" *Microb Cell Fact*
47. Cao, Zhang, Zhang et al. (2022) "Efficacy of a gB + gD-based subunit vaccine and the adjuvant granulocyte-macrophage colony stimulating factor for pseudorabies virus in rabbits" *Front Microbiol*
48. Ren, Cao, Tian et al. (2023) "A self-assembled nanoparticle vaccine based on pseudorabies virus glycoprotein D induces potent protective immunity against pseudora bies virus infection" *Vet Microbiol*
49. Zhao, Chen, Luo et al. (2023) "Pseudorabies gD protein protects mice and piglets against lethal doses of pseudorabies virus" *Front Microbiol*
50. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12808154&blobtype=pdf | # Handling editor Ana Cristina Bratanich
Emilia Pulkkinen, Reilly Jackson, Ruut Joensuu, Essi Korhonen, Moses Muia Masika, Omu Anzala, Joseph Ogola, Paul Webala, Tamika Lunn, Kristian Forbes, Olli Vapalahti, Tuure Kinnunen, Tarja Sironen, Anne Jääskeläinen
## Abstract
To enhance preparedness against existing and new zoonotic viruses such as Hendra virus, Nipah virus, and other paramyxoviruses, screening tools and efficient diagnostic methods are needed. Here, we established a conventional nested pan-PCR assay using previously described primers for screening of human and bat samples. Additionally, we developed specific real-time RT-PCR (RT-qPCR) assays to detect Nipah virus and Hendra virus genotypes 1 and 2. Both PCR methods demonstrated good performance and could be used for screening of paramyxoviruses. Human serum and cerebrospinal fluid samples from 558 Finnish patients and 60 serum samples from Kenyan patients were screened using the nested-pan-PCR assay, and all were negative. In addition, we screened 340 synanthropic bat samples collected during 2021 and 2023 from Kenya, resulting in the discovery of two parajeilongviruses in Angolan free-tailed bat (Mops condylurus) samples.
## Introduction
Comprehensive preparedness is essential for a rapid response to emerging zoonotic pathogens and future pandemics. Research shows that 60.3% of emerging pathogens have a zoonotic origin, 71.8% of which originate from wildlife [1]. One of the families of high-risk viruses is Paramyxoviridae, in particular, the subfamilies Orthoparamyxovirinae and Feraresvirinae, which both deserve special attention when preparedness is established. These two subfamilies include diverse paramyxoviruses (PMVs) that infect mammals of various species. A notable PMV that infects humans is measles virus (Morbillivirus hominis). This virus spreads extremely efficiently via droplets and still continues to cause childhood mortality worldwide. Measles virus can also cause a wide spectrum of non-fatal complications, including central nervous system (CNS) infections and pneumonia [2]. Less-severe and more-common human PMV infections include respiratory tract infections caused by parainfluenza viruses.
Many PMVs possess zoonotic potential because of their high mutation rate, a typical characteristic of RNA viruses, and their ability to switch host species [3]. Two examples of such zoonotic viruses are the bat-borne viruses Hendra virus Tarja Sironen and Anne J. Jääskeläinen these senior authors contributed equally to this article.
(HeV) and Nipah virus (NiV), which have caused severe, often fatal infections in humans in Southeast Asia and Australia [4,5]. Fruit bats serve as the primary reservoirs for these viruses [6]. However, domesticated animals, such as horses and pigs, have been implicated as intermediate hosts facilitating virus transmission to humans [5]. HeV infections typically manifest as equine-borne sporadic spillover events in Australia, whereas NiV causes annual outbreaks linked to the consumption of contaminated date palm sap or through pigs, for example, in Bangladesh and India [7,8]. A new emerging paramyxovirus with the pathogenicity of a henipavirus and the transmissibility of measles virus would pose a severe public-health risk, potentially triggering an epidemic -or even a pandemic.
Fruit bats are known to host various paramyxoviruses [9,10]. For example, in Africa, henipavirus RNAs have been detected in straw-colored fruit bats (Eidolon helvum) in Ghana [11] and the Republic of Congo [12]. Moreover, serological evidence suggests the presence of henipa-like viruses circulating in domesticated pigs from several different pig farms in Uganda [13]. While no henipavirus infections in humans have been reported in Africa, Sosuga virus -a closely related pathogen belonging to the subfamily Rubulavirinae that is in Rousettus aegyptiacus bats in Uganda -has been transmitted to a human, causing acute febrile illness [14]. In Europe, PMVs have been detected in different mammalian species, including insectivorous bats, shrews, and hedgehogs [15][16][17][18]. However, spillover events from reservoir animals to humans have not been reported to date. The bat populations in Finland consist only of insectivorous bats [19], and it is currently unknown whether these bat species would carry potentially zoonotic PMVs. To date, no zoonotic PMV infections have been reported in Finland.
In order to study human samples collected from patients in Finland and Kenya with suspected CNS infections or other acute or febrile infections, we established a nested pan-PCR assay using previously designed primers [20] for the detection of PMVs. This assay was also used to screen bat samples from Kenya for the presence of PMVs. As nested assays are prone to contamination, i.e., due to crosscontamination from the positive PCR control, we designed a PCR control with a built-in modification that can be identified by sequencing to avoid false positives in diagnostic settings. For emergency settings, the nested pan-PCR assay is too slow and laborious to be practical. Accordingly, we established fast and specific RT-qPCR methods for the detection and discrimination of NiV and HeV. The HeV RT-qPCR was designed to detect both HeV genotype 1 (G1) and the recently discovered genotype 2 (G2) [21], in order to cover all of the viruses known to cause human disease.
## Materials and methods
## Human samples
Altogether, 679 samples from Finnish and Kenyan patients were collected and screened. From Finland, 606 serum samples and 13 cerebrospinal fluid samples were collected retrospectively from 558 individuals (median age, 46 years; range, 1-88 years; sex, data not available) between August 2018 and November 2018 from 16 different hospital districts and the Åland Islands. The samples were collected from anonymized sample archives with limited patient data. All samples were from patients with suspected severe acute viral infections, particularly central nervous system infections. Symptoms and differential diagnostics from Finnish patients are listed in Table 1. All data were anonymized, and the research was carried out with ethical approval (permits HUS/32/2018, HUS/211/2020, HUS/244/2021, and 159/ HUS/151/2022) from the HUS Diagnostic Center, Helsinki, Finland. A subset of 60 serum samples from 60 Kenyan patients (median age, 19 years; range, 7 months-83 years; sex, 30 female, 29 male, 1 unknown) was obtained from a larger sample collection [22] from Taita-Taveta County, southeastern Kenya, in October 2016. All patients had an acute febrile illness, and the symptoms and differential diagnostics are listed in Table 2. Most of the Kenyan patients had been in contact with either domesticated farm animals or wild animals, including bats and rodents. The characteristics of the participants were reported previously by Masika et al. [22].
RNA was extracted from human samples using a semiautomated MagNA Pure 96 System (Roche) and a Total Nucleic Acid Kit according to the manufacturer's instructions. Extracted RNA samples were stored at -70°C.
## Bat samples
All bat samples were collected in Taita-Taveta County in southeastern Kenya between August 2021 and May 2023. Permission to conduct bat fieldwork was granted through permits from the National Commission for Science, Technology and Innovation (#NACOSTI/P/21/9267), the Kenya Wildlife Service (#KWS/BRM/500 and WRTI/RP/118.6), and the University of Arkansas Institutional Animal Care and Use Committee (#22012).
Bats were captured in flyways, building roosts, and inside roosting buildings, using mist-nets or by hand. Bats were placed into individual cotton bags upon capture and stored in a humid cool location overnight. The following morning, the species, sex, and age of the bats were recorded (Table 3). Fecal samples were collected from inside the bags and directly from the excreting bats during the sampling 1 3
process. This capturing protocol followed previously described standards [23]. At dusk on the day after capture, the bats were returned to the collection site and released. The fecal pellets were stored in 2-ml tubes containing 0.5 ml of RNAlater, first at -20°C for a short initial period, and then at -80°C for long-term storage. Later, a subset of 340 samples was selected for RNA extraction using Invitrogen TRIzol Reagent (Thermo Fisher Scientific, Waltham, Massachusetts, USA) according to the manufacturer's guidelines.
## Nested pan-PCR assay and screening
Previously designed PCR primers [20] were used to establish a nested pan-PCR protocol for detection of different members of the family Paramyxoviridae. In addition, for this assay, we designed a PCR control with an artificial built-in modification as a contamination control to avoid control. This plasmid contained part of the genome of CDV, with a recognizable insertion of artificial and non-coding nucleotides (CDV plasmid; Integrated DNA Technologies) to further discriminate possible contamination from the original viruses (Supplementary Fig. S1). To assess the sensitivity of the nested pan-PCR assay, a tenfold dilution series of CDV plasmid was tested, as well as RNAs from HeV G1 and NiV (Fig. 2).
All human and bat samples were screened using the nested pan-PCR assay, followed by agarose gel electrophoresis (AGE), sequencing and annotation using the core nucleotide collection from the National Centre for Biotechnology Information (NCBI) with Basic Local Alignment Tool (BLAST).
## Henipavirus RT-qPCR assays
Specific RT-qPCR assays for NiV and HeV G1 and G2 were established using the CFX96 platform (Bio-Rad) (see Appendix). Primers were designed to recognize the region of the genome encoding the nucleocapsid protein. All of the RT-qPCR assays were validated using viral RNA of NiV and HeV G1, commercial quantified plasmid controls of HeV G1 (IDT) and G2 (IDT), and a negative control panel (Table 4). The sensitivity was determined by calculating the limit of detection (LOD) and intra-assay repeatability for 10 parallel reactions.
## Results
## Nested pan-PCR testing of human samples
In total, 606 serum and 13 CSF samples from Finnish patients, as well as 60 serum samples from Kenyan patients (total N = 679) were screened using the nested pan-PCR assay. PCR products (~ 650 base pair long) were obtained from 16 sera (16/666, 2.4%) and two CSF samples (2/13, 15.4%), all of which were successfully sequenced. No viral nucleic acids were detected among the sequenced amplicons, and since all of these sequences corresponded to human DNA, they were interpreted as nonspecific background amplifications. No PMVs were detected in the Finnish and Kenyan patient samples.
## Nested pan-PCR testing of bat samples
In total, 340 bat fecal samples from 340 individuals comprising 16 different bat species (Table 3) from Taita-Taveta County (Kenya) were screened by nested pan-PCR, and amplification products of PMVs were detected in two fecal samples of Angolan free-tailed bats (Mops condylurus). The false positives (see Appendix). According to the taxonomy nomenclature at the time of publication, the viruses amplified with these PCR primers included members of the genera Henipavirus, Morbillivirus, Respirovirus, and Rubulavirus. Under the current taxonomy, the viruses detected using these primers would be classified in several genera of the two subfamilies Orthoparamyxovirinae and Feraresvirinae. Using these primers and various PMV controls, PCR products ranging in length from 600 to 700 base pairs were produced and then sequenced by the Sanger method. As the protocol was established for diagnostic purposes, it was further tested with human serum and CSF samples. To enhance the specificity of the assay for clinical samples, the level of nonspecific amplification of human DNA was reduced by using a touch-down PCR as the second PCR cycle, by first allowing flexible primer binding followed by more-specific amplification conditions towards the end of the procedure (see Appendix).
To test the performance of the nested pan-PCR assay, viral RNA from measles virus, canine morbillivirus (previously canine distemper virus, CDV), HeV G1, and NiV were used. Because HeV and NiV are classified as biosafety-level-4 pathogens, they were received as inactivated virus stocks from the Commonwealth Scientific and Industrial Research Organisation (CSIRO, Australia). In addition, we designed a cDNA plasmid as an assay contamination Tanzania. In addition, these sequences shared a high degree of similarity with partial viral sequences from free-tailed bats (Tadarida sp.). Unfortunately, no remaining sample material was available for further NGS sequencing attempts with these two positive bats. Extracted RNA from CDV, measles virus, HeV, NiV, and a commercial plasmid control for CDV with a built-in contamination control were all successfully amplified using the nested pan-PCR and were subsequently sequenced and correctly annotated (see Appendix). The negative control and negative panel were all negative; no amplification was detected by AGE using these CSF or blood samples. Tenfold dilution series of CDV plasmid and HeV and NiV RNAs positive fecal samples were collected from two adult male bats in May 2023 from the main building of an orphanage house. The 457-base-pair-long sequences were first annotated using the BLAST annotation tool, followed by construction of a Bayesian maximum-likelihood phylogenetic tree using IQ-TREE [24,25], which was visualized using FigTree [26] (Fig. 1 and Supplementary Fig. S2). This analysis revealed that these sequences shared a high degree of similarity with viral sequences from the genera Jeilongvirus and Parajeilongvirus of the subfamily Orthoparamyxovirinae. The highest sequence identity values -99.11% and 98.66% -were observed with partial parajeilongvirus sequences obtained from M. condylurus samples from In this study, as expected, we did not find measles virus or other PMVs in the Finnish samples (N = 619). The Kenyan patient samples were collected from April to August 2016, before the large-scale measles-rubella vaccination campaign started in October 2016 in Kenya. Nevertheless, we did not find any measles virus or other PMVs in the Kenyan samples. However, it is worth noting that we did not have any vaccination data from these 60 Kenyan patients. Although measles vaccines have been used for decades in Kenya, measles virus continues to cause infections every year. For example, in 2016, the vaccination coverage for the first and second dose of the measles vaccine was 93% and 35%, respectively. In the same year, 128 measles cases were reported in Kenya [28]. Of note, there is currently a measles epidemic ongoing in the USA [29]. Complete genome sequences have been obtained for paramyxoviruses from bat samples collected in Kenya, including samples from Triaenops, Hipposideros, Coleura, Miniopterus, Otomops, Nycteris, Rousettus, Cardioderma and Chaerephon sp., but none have been obtained from Mops species [30,31]. In our study, a parajeilongvirus was detected in two individual Mops condylurus fecal samples, using nested pan-PCR and sequencing. The sequences from these bat samples exhibited high sequence similarity (90-99% identity) to other partial genome sequences of putative paramyxoviruses found in different bat species, i.e., Mops condylurus, Mops pumilus, and Tadarida sp., in northern Africa (Fig. 1). Partial genome sequences of parajeilongviruses found in Tanzania, obtained from the NCBI Gen-Bank database, were 98-99% identical to our sequences, suggesting the circulation of this parajeilongvirus type in the African Great Lakes region. The similarity of the viral sequences in Mops sp. and closely related Tadarida sp. suggests potential for spillover between these species. However, due to the lack of whole genome sequences, and because only a highly conserved part of the sequence was used to study the relationships among different virus strains, no definitive conclusions are possible. Although Tadarida aegyptiaca is generally found in more-southern parts of Africa than M. condylurus, they are both present in the African Great Lakes region, especially in Kenya (Fig. 3) [32,33].
Taxonomically, the genus Parajeilongvirus was separated from the genus Jeilongvirus in 2023 based on differences in their host associations and genomic structure (International Committee on Taxonomy of Viruses). In the current taxonomy, all of the parajeilongviruses originate mostly from bats, which is consistent with our findings, since the sequences obtained from M. condylurus samples were more similar to parajeilongviruses than jeilongviruses (Fig. 1). Both genera consist of newly discovered viruses, and therefore, in-depth research data on these viruses is showed increased sensitivity with the nested pan-PCR assay compared to the initial PCR step of the assay (Fig. 2). The CDV control demonstrated clear amplification already after the first PCR step, whereas HeV and NiV could be detected only after the nested step. The nested PCR step increased performance by 1000-fold with the CDV plasmid control (Fig. 2).
## Henipavirus RT-qPCR assays
The assays were validated using extracted viral RNA from inactivated HeV and NiV (CSIRO) and commercial quantified plasmid controls of NiV, HeV G1, and HeV G2 (IDT). The henipavirus RT-qPCR assays were able to discriminate NiV from HeV. A negative panel comprised of 14 sera, five cerebrospinal fluid samples, 20 EDTA-blood samples, and nine respiratory samples, in addition to a panel of 20 non-henipaviral RNAs (Table 4), were all negative (68/68, 100%). By using 95% confidence intervals (SPSS, Probit, IBM), the LOD for the henipavirus RT-qPCRs was 14.1, 7.2, and 9.8 copies per PCR reaction for HeV G1 (IDT), HeV G2 (IDT), and NiV (IDT), respectively. At a copy level of 430-440 with 10 parallel reactions for HeV G1, the average Ct value was 33.4, the standard deviation (STDEV) was 1.1, and the coefficient of variation (CV) was 3%. At a copy level of 8400 for HeV G2, the Ct value was 35.5, the STDEV was 0.78, and the CV was 2%, and for NiV at a copy level of 430-440, the Ct value was 32.5, the STDEV was 0.70, and the CV was 2%.
## Discussion
In Finland, no endemic zoonotic infections caused by PMVs have been reported to date. However, another PMV -measles virus -is still causing sporadic infections almost annually. The highest rate of measles virus infections was in 2011, with 27 laboratory-confirmed cases [27]. The coverage of measles, mumps, and rubella (MMR) vaccination varies geographically in Finland but is high overall [27].
Fig. 1 Maximum-likelihood phylogenetic tree based on partial sequences of L gene of the two detected parajeilongviruses and related viruses with midpoint rooting. The tree was constructed using sequences from orthoparamyxoviruses (blue), jeilongviruses (green), and parajeilongviruses (red), as well as the 100 top BLAST hits (black) for the two detected parajeilongviruses (purple). Strains found in the official ICTV taxonomy are labeled with their taxonomic nomenclature, whereas the sequences identified using BLAST are named using their geographical location and host species. The best-fitting model was chosen for these 475-base-pair-long sequences using IQTREE ModelFinder [35]. The analysis was performed in IQTREE [36], using the GTR + F + R6 model with 1000 bootstrap replicates, and the tree was visualized using FigTree v1.4.4 [26]. A full non-collapsed tree is shown in Supplementary Fig. S2) samples typically include a greater diversity of background nucleic acids than human blood or CSF samples.
For emergency settings, molecular methods based on nested PCR and sequencing are not recommended due to their time-consuming nature and susceptibility to crosscontamination. For situations in which results are needed within hours rather than days, RT-qPCR assays with a short turnaround time and easily interpreted results should be designed. In this study, we designed, performed, and validated specific RT-qPCR assays for NiV, HeV G1, and HeV G2. These assays demonstrated good performance, including efficient detection of the new HeV G2. These methods are now available for use in clinical diagnostics. Although RT-qPCR assays are efficient, conventional PCR assays combined with Sanger sequencing remain valuable tools for distinguishing new virus variants and genotypes from existing ones [21].
Preparedness for zoonotic infections requires basic research and surveillance to enhance our understanding of viral circulation in humans and wildlife. The family Paramyxoviridae is likely to include numerous undiscovered limited. Although new viruses are detected continually, the most common screening methods in use result in short nucleotide sequences from conserved areas of the genome, leaving many open questions, including the detailed characteristics of the detected viruses.
Overall, the procedure for the nested pan-PCR assay fits well into a routine diagnostic setting, and in this study, no cross-contamination was detected. The nested PCR step includes a contamination control that appears to be suitable for routine diagnostic purposes. The assay produced some background amplification (human DNA), more frequently with CSF samples (15.4%) than with serum samples (2.4%) , but this was technically straightforward to exclude by Sanger sequencing. The process of implementing a plasmid control improved the practicality of the laborious contamination-prone nested PCR in routine diagnostics, highlighting its potential application, for example, in resource-limited settings where RT-qPCR or next-generation sequencing may not be available. More nonspecific amplification was detected by AGE when using samples from bats than when using human samples. This was expected, as bat fecal
## References
1. Albariño (2012) "Novel Paramyxovirus Associated with Severe Acute Febrile Disease, South Sudan and Uganda" *Emerg Infect Dis*
2. Kurth (2012) "Novel paramyxoviruses in free-ranging European bats" *PLoS ONE*
3. Haring (2024) "Detection of novel orthoparamyxoviruses, orthonairoviruses and an orthohepevirus in European whitetoothed shrews DATA SUMMARY" *Microb Genom*
4. Horemans, Van Bets, Maes et al. (2023) "Discovery and genome characterization of six new orthoparamyxoviruses in small Belgian mammals" *Virus Evol*
5. Vanmechelen, Vergote, Merino et al. (2020) "Common occurrence of Belerina virus, a novel paramyxovirus found in Belgian hedgehogs" *Sci Rep*
6. Liukko, Stjernberg (2019) "Atlas of Finnish Bats" *Ann Zool Fennici*
7. Tong, Chern, Li et al. (2008) "Sensitive and broadly reactive reverse transcription-PCR assays to detect novel paramyxoviruses" *J Clin Microbiol*
8. Annand (2022) "Novel Hendra Virus Variant Detected by Sentinel Surveillance of Horses in Australia" *Emerg Infect Dis*
9. Masika (2020) "Detection of dengue virus type 2 of Indian origin in acute febrile patients in rural Kenya" *PLoS Negl Trop Dis*
10. Sikes (2016) "Guidelines of the American Society of Mammalogists for the use of wild mammals in research and education" *J Mammal*
11. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: Fast Model Selection for Accurate Phylogenetic Estimates" *Nat Methods*
12. Minh (2015) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Mol Biol Evol*
13. Rambaut (2018)
14. Conrardy (2014) "Molecular Detection of Adenoviruses, Rhabdoviruses, and Paramyxoviruses in Bats from Kenya" *Am J Trop Med Hyg*
15. Waruhiu (2017) "Molecular detection of viruses in Kenyan bats and discovery of novel astroviruses, caliciviruses and rotaviruses" *Virol Sin*
16. Taxonomy (2023) "Mops condylurus (A.Smith, 1833) in GBIF Secretariat"
17. Jones (2008) "Global trends in emerging infectious diseases" *Nat*
18. Kitchen, Shackelton, Holmes (2011) "Family level phylogenies reveal modes of macroevolution in RNA viruses" *Proc Natl Acad Sci U S A*
19. Mahalingam (2012) "Hendra virus: an emerging paramyxovirus in Australia" *Lancet Infect Dis*
20. Eaton, Broder, Middleton et al. (2006) "Hendra and Nipah viruses: different and dangerous" *Nat Rev Microbiol*
21. Halpin (2011) "Pteropid bats are confirmed as the reservoir hosts of henipaviruses: a comprehensive experimental study of virus transmission" *Am J Trop Med Hyg*
22. Islam (2011) "Nipah Virus Transmission from Bats to Humans Associated with Drinking Traditional Liquor Made from Date Palm Sap" *Emerg Infect Dis*
23. Thiagarajan (2023) "Nipah virus: India's Kerala state moves quickly to control fresh outbreak" *BMJ*
24. Drexler (2012) "Bats host major mammalian paramyxoviruses" *Nat Commun*
25. Maganga (2014) "Identification of an unclassified paramyxovirus in Coleura afra: a potential case of host specificity" *PLoS ONE*
26. Drexler "RNA in African Bats"
27. Weiss (2012) "Henipavirus-related Sequences in Fruit Bat Bushmeat, Republic of Congo -18"
28. Atherstone (2019) "Evidence of exposure to henipaviruses in domestic pigs in Uganda Funding information CGIAR Research Program on Agriculture for Nutrition and Health" *Transbound Emerg Dis*
29. Minh (2015) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Mol Biol Evol*
30. "Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations"
31. Taxonomy, Geoffroy (1818) "Tadarida aegyptiaca"
32. Zhang (2022) "A Zoonotic Henipavirus in Febrile Patients in China" *N Engl J Med*
33. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: Fast Model Selection for Accurate Phylogenetic Estimates" *Nat Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12502735&blobtype=pdf | # Teropavimab and zinlirvimab sensitivity in people living with multidrug-resistant HIV-1: data from the PRESTIGIO Registry
Vincenzo Spagnuolo, Laura Galli, Jiani Li, Keith Dunn, Filippo Lagi, Roberta Gagliardini, Loredana Sarmati, Anna Cattelan, Andrea Giacomelli, Maria Santoro, Maurizio Zazzi, Christian Callebaut, Antonella Castagna, Laurie Vanderveen
## Abstract
We characterized sensitivity to teropavimab (TAB) and zinlirvimab (ZAB) in people living with four-class drug-resistant HIV (4DR-PWH). This was a multicenter, observational study using plasma or peripheral blood mononuclear cells collected from 50 4DR-PWH (25 with HIV-1 RNA > 1,000 copies/mL matched by age, sex, nadir CD4+, and years on ART to 25 virologically suppressed [HIV-1 RNA < 50 copies/mL]) enrolled in the PRESTIGIO Registry (NCT04098315) with a documented 4DR (NRTIs, NNRTIs, PIs, and INSTIs). Phenotypic sensitivity to bNAbs was determined using the PhenoSense monoclonal antibody assay (Monogram), with susceptibility defined as IC 90 ≤ 2 µg/mL. The HIV-1 envelope was genotyped by next-generation sequencing, and sequences were analyzed for the presence of multi-position HIV-1 envelope amino acid signatures associated with in vitro phenotypic susceptibility to TAB and ZAB. Of 46/50 (92%) participants with PhenoSense mAb assay results, 35 (76%) were phenotypically sensitive to TAB, 23 (50%) to ZAB, and 19 (41%) to both bNAbs; seven (15%) were phenotypically resistant to both bNAbs. The proportion of individuals with sensitivity to both bNAbs was similar in participants with viremia (41%) and those with virologic suppression (42%; P = 0.99). We observed marginal correlations between TAB 90% inhibitory concentration (IC 90 ) values and years since HIV diagnosis at the time of sample collection (Spearman r = 0.29, P = 0.05) as well as between ZAB IC 90 values and CD8+ cell count (Spearman r = -0.32, P = 0.05). A significant number of the 4DR-PWH analyzed were found to have virus susceptible to TAB and ZAB. These data provide proof-of-concept that selected multidrug-resistant PWH may be candidates for future trials investigating bNAbs-con taining regimens to achieve or maintain virologic suppression. IMPORTANCE Multidrug-resistant HIV presents significant challenges for treatment, often leaving individuals with a limited range of therapeutic alternatives. This study provides crucial insights into the efficacy of two promising broadly neutralizing antibodies, teropavimab (TAB) and zinlirvimab (ZAB), in individuals living with HIV who have developed resistance to multiple drug classes. The findings indicate that a significant proportion of the population remains susceptible to these novel treatments, irrespective of their viral suppression status. These results offer a promising basis for developing new therapeutic strategies to improve outcomes for individuals with multidrug-resistant HIV who have a history of extensive treatment, paving the way for future clinical trials aimed at achieving long-term viral suppression with novel drug regimens.
H arboring multidrug-resistant (MDR) virus is a clear risk factor for clinical progression and death in people living with HIV (PWH) (1,2). In these individuals, it is often very difficult to initiate a suppressive regimen (in the case of virologic failure) or simplify an ongoing regimen (in the case of suppressive therapy) due to the unavailability of fully active drugs, previous antiretroviral therapy (ART)-related toxicities, or lack of adherence (3).
Therefore, the availability of drugs with novel mechanisms of action and long-lasting efficacy would represent a substantial advancement in therapy for this population.
Lenacapavir (LEN), a novel HIV-1 capsid inhibitor, is currently approved for twiceyearly dosing in heavily treatment-experienced (HTE) PWH in combination with other antiretrovirals and being studied with various compounds for long-acting oral and subcutaneous (SC) injection every 3-6 months (4).
Teropavimab (formerly GS-5423 or 3BNC117-LS; TAB) and zinlirvimab (formerly GS-2872 or 10-1074-LS; ZAB) are broadly neutralizing antibodies (bNAbs) that target non-overlapping HIV-1 envelope spike sites (CD4 binding site and V3 loop, respectively) and designed to have long half-lives, potentially allowing twice-yearly dosing (5,6). TAB and ZAB and their parental antibodies 3BNC117 and 10-1074 can neutralize a significant proportion of diverse global HIV-1 strains, although the breadth of neutralization varies by subtype (7)(8)(9)(10).
In addition, selective immune pressure during untreated infection may drive the HIV-1 envelope evolution, resulting in viral populations that can evade recognition by bNAbs and limit their effectiveness. Thus, bNAbs susceptibility testing before initiating treatment with TAB and ZAB may improve clinical success. While there is currently no standard for determining whether viral strains are susceptible to bNAbs, the genotypic and phenotypic evaluations of viral susceptibility have been explored in clinical trials of 3BNC117, 10-1074, and their long-acting derivatives (11)(12)(13)(14)(15).
In a phase 1b proof-of-concept study (NCT04811040) in chronically treated, virologically suppressed PWH who discontinued oral ART and initiated a regimen of SC LEN plus weight-based intravenous infusions of TAB and ZAB, 18/20 (90%) partici pants maintained virologic suppression at week 26 (15). In an ongoing phase 2 study (NCT05729568), efficacy of the every-6-months regimen of LEN with fixed doses of TAB and ZAB was similar to that of daily oral ART through week 26 (16). In both studies, which evaluated PWH susceptible to both TAB and ZAB, as determined by in vitro phenotype, the proportion of PWH susceptible to both bNAbs was approximately 50% (15,16).
Interestingly, the 26 week efficacy of the combination of LEN plus TAB and ZAB appears to be maintained, even in the presence of susceptibility to either TAB or ZAB alone (17).
Given that HIV-1 susceptibility TAB and ZAB may vary across populations, future studies should aim to evaluate diverse cohorts. Although TAB and ZAB may be useful for the treatment of PWH harboring MDR strains, data in this specific group of PWH are limited. In this study, we characterized the susceptibility to TAB and ZAB and HIV-1 envelope diversity in people living with four-class drug-resistant HIV (4DR-PWH).
## RESULTS
Fifty 4DR-PWH were evaluated (25 viremic and 25 non-viremic). Phenotypic assay failure was observed in four individuals (three in the viremic group and one in the non-viremic group), allowing the analysis of TAB and ZAB susceptibility in a total of 46 PWH. The characteristics of the individuals with analyzable samples (median age 55 [interquartile range: 48-58] years; 80% male) were indicative of a long history of HIV-1 infection (26 [23][24][25][26][27][28][29][30][31][32] years), extensive treatment history (23 [21-27] Additional demographic, virologic, and therapeutic characteristics of 4DR-PWH at the time of sample collection and according to HIV RNA viral load are detailed in Table 1.
Of the 46 participants with PhenoSense monoclonal antibody assay results, 35 (76%) were phenotypically sensitive to TAB, 23 (50%) to ZAB, and 19 (41%) to both bNAbs, while seven (15%) were phenotypically resistant to both bNAbs. The 90% inhibitory concentration (IC 90 ) values for TAB and ZAB in the 4DR PWH enrolled in the study are shown in detail in Fig. 1.
Of 22 viremic participants, 19 (86%) were phenotypically sensitive to TAB, 10 (45%) to ZAB, nine (41%) to both bNAbs, and two (9%) to neither. Of the 24 participants with virologic suppression, 16 (67%) were phenotypically sensitive to TAB, 13 (54%) to ZAB, 10 (42%) to both bNAbs, and five (21%) to neither. The proportion of participants with sensitivity to both bNAbs was similar (P = 0.99) in viremic participants (9/22 [41%]) compared to those with virologic suppression (10/24 [42%]).
Average pairwise distance was calculated among the individual variants of HIV-1 env detected in plasma HIV-1 RNA and proviral DNA per participant. The HIV env genetic sequence diversity was high in both plasma virus (median 0.62, range 0.26-1.5) and PBMC provirus (median 1.7, range 0.56-3.6) from 4DR-PWH, consistent with diverse HIV-1 populations arising during extensive treatment history and in agreement with prior observations (18, 19) (Fig. S1).
Viral susceptibility to bNAbs was assessed by the application of HIV-1 envelope amino acid signatures known to predict sensitivity to TAB or ZAB. The analysis of phenotypic data by the presence of HIV-1 envelope signatures showed a good correla tion between genotypic sensitivity predictions and observed phenotypic sensitivity, with more complex signatures predicting phenotypic susceptibility to TAB (Fig. 2; panel A) and ZAB (Fig. 2; panel B) with increasing accuracy.
The demographic, virologic, and therapeutic characteristics of 4DR-PWH at the time of sample collection and according to susceptibility to TAB and ZAB are detailed in Table 2.
## DISCUSSION
In our study evaluating susceptibility to TAB and ZAB in a cohort of 4DR-PWH, we observed that phenotypic susceptibility to both bNAbs was present in approximately 40% of individuals. This rate of susceptibility is similar to that observed by Selzer et al. who reported 50% susceptibility to both TAB and ZAB in individuals screened for the phase 1b proof-of-concept study that evaluated the efficacy of a combination of LEN plus TAB and ZAB in PWH with chronic infection and suppressed viremia on ART (10,15). However, in that study, the total duration of HIV infection and the ART duration were 8.2 and 2.6 years, respectively, which is markedly lower compared to the 4DR PWH in our study (25.7 and 22.7 years, respectively).
These findings indicate that a long history of infection and ART exposure may not significantly affect viral susceptibility to bNAbs, even when associated with high viral diversity. Recent data from individuals with PHI, when viral diversity is expected to be low, revealed that the proportion of individuals susceptible to both bNAbs was 31% based on the presence of specific amino acid signatures linked to TAB and ZAB susceptibility (20); however, some individuals with virus susceptible to bNAbs may be missed due to the low sensitivity performance of these signature predictions. In other studies, higher bNAb susceptibility was associated with lower HIV-1 env diversity in individuals who initiated ART during acute and early HIV-1 infection vs. chronic infection, which did not differ pre-and post-treatment (19).
Furthermore, our study did not identify a correlation between co-receptor usage (CCR5 or CXCR4) at the time of sample collection and susceptibility to TAB and ZAB in accordance with the findings of other research studies (20). However, in our study, we examined a single time point, and we were, therefore, not able to evaluate the evolution of bNAbs susceptibility or HIV-1 genetic diversity over time. In addition, only nearly 25% of the PWH evaluated were receiving treatment with maraviroc or fostemsavir. While we did not observe any association between the use of these agents and bNAbs susceptibility, we cannot exclude antiviral-driven env evolution with an impact on TAB and ZAB susceptibility.
Our research suggests a marginal correlation between a longer history of HIV infection and higher IC 90 values for TAB. While the proportion of TAB susceptibility was similar to that observed in other studies of PWH without multidrug resistance, a longer duration of HIV infection-and consequently, exposure to ART and detectable viremia-may slightly impact TAB susceptibility. However, the PRESTIGO Registry only collects complete immunovirological data from the time the four-drug class resistance is detected. This prevents us from evaluating the possible impact of cumulative exposure to periods of detectable viremia on bNAb susceptibility. The median average pairwise distances of plasma viral and PBMC proviral sequences from 4DR PWH were 0.62 and 1.7, respectively, which are similar to or higher than those reported in previous studies that evaluated HIV-1 env diversity during chronic infection using a similar methodology (18,19). It is notable that the differences in genetic diversity between plasma and PBMC virus did not correlate with the differences in phenotypic susceptibility to bNAbs. This may reflect limitations in short-read sequencing to differentiate between intact env necessary for successful phenotyping and defective env sequences that may be archived in the viral reservoir.
Susceptibility testing may be important to identify individuals more likely to respond to bNAbs. However, there is no standardized method or interpretive framework for determining susceptibility to TAB and ZAB. Here, we compared phenotypic suscepti bilities in the PhenoSense mAb assay to genotypic susceptibility predictions derived from multi-position HIV-1 envelope amino acid signatures. Our study showed that more complex HIV-1 envelope amino acid signatures predicted phenotypic susceptibility with increasing specificity, consistent with prior reports (10,21). However, a non-negli gible proportion of 4DR PWH without HIV-1 envelope amino acid signatures remains susceptible to bNAbs in the phenotypic assay, indicating suboptimal sensitivity of the genotypic signature method.
Potential limitations of our study include the modest sample size and the cross-sec tional design, which did not allow assessment of the evolution of TAB and ZAB suscepti bility over time; however, for the first time, we have described susceptibility to bNAbs in a cohort of PWH with multidrug-resistant virus and extensive history of ART exposure and found that a relatively high percentage of these individuals retains susceptibility to these two novel long-acting agents. The confirmation of these findings in a larger sample may clarify whether heavily treatment-experienced PWH with multidrug-resistant HIV could be considered candidates for future trials evaluating bNAbs-containing regimens to achieve or maintain virologic suppression.
## MATERIALS AND METHODS
This was a multicenter, observational, cross-sectional study that used plasma or peripheral blood mononuclear cells (PBMCs) collected from 50 4DR-PWH (25 PWH with HIV-1 RNA > 1,000 copies/mL matched by age, sex assigned at birth, nadir CD4+, and years on ART to 25 virologically suppressed PWH [defined as HIV-1 RNA < 50 cop ies/mL]) enrolled in the PRESTIGIO Registry (NCT04098315; https://registroprestigio.org). The PRESTIGIO registry is an ongoing, observational, prospective, Italian, multicenter, annual collection of biological samples, and data on clinical, laboratory, treatment, and virological characteristics of 4DR-PWH (defined as genotypically resistant to nucleoside reverse transcriptase inhibitors [NRTIs], non-NRTIs [NNRTIs], protease inhibitors [PIs], and integrase strand transfer inhibitors [INSTIs]). Plasma and PBMC samples are collected on an annual basis from the date of enrollment and cryopreserved in a biobank (BioRep, https://www.biorep.it/). Clinical, laboratory, treatment, and virological data are collected annually from the date of evidence of 4DR (defined as baseline) (22). The PRESTIGIO Registry has been approved by the ethics committees of all participating centers, and all participants (n = 270 as of March 2025) have given written informed consent for their data and samples to be used for research purposes.
Phenotypic sensitivity to TAB and ZAB was determined using the PhenoSense monoclonal antibody assay from Monogram Biosciences (South San Francisco, CA, USA) with susceptibility defined as IC 90 ≤ 2 µg/mL for both bNAbs, consistent with previ ous clinical investigations (10,(15)(16)(17). Briefly, expression vectors containing plasma-or peripheral blood mononuclear cell-derived HIV-1 env are co-transfected with an HIV-1 genomic luciferase reporter in HEK293 cells to produce pseudovirions. Neutralizing antibody susceptibility is assessed as inhibition of pseudovirus infection of target cells following pre-incubation with bNAbs measured by luciferase activity (23).
The HIV-1 env gene from PhenoSense Env expression vectors was genotyped at Monogram Biosciences using Mi-Seq (Illumina, San Diego, CA, USA) next-generation sequencing. HIV-1 env sequences were analyzed using a previously described analysis pipeline (19). A custom-developed APOBEC hypermutation algorithm was applied to the deep sequencing data. Briefly, each read was evaluated by comparing the G->A mutations and other mutations. Reads were classified as hypermutated if they contained ≥4 G->A mutations and ≤2 non-G->A mutations. Hypermutated reads were excluded from downstream mutation analysis.
Genetic diversity within these sequences was assessed by average pairwise distance analysis using a sliding windows approach across the HIV env gene (19). Briefly, HIV-1 env was divided into 50 base pair genomic intervals with a 25-base pair overlap between two adjacent windows to reduce the impact of signal noise. The average nucleotide difference between different reads in a sliding window was calculated using Nei and Li's method (24).
Sequences were analyzed for the presence of multi-position HIV-1 Env amino acid signatures associated with in vitro phenotypic susceptibility to TAB and ZAB (21). Briefly, in vitro neutralization data combined with virus sequence information for >200 subtype B viruses were used to identify HIV Env amino acid positions important for susceptibility (IC 50 < 1 µg/mL). Here, genotypic signatures were applied using the more stringent threshold of IC 90 ≤ 2 µg/mL. Only base-pair positions with variability <2% in viral quasi-species were considered to be part of the signature. Sensitivity of the signatures was defined as the probability that the amino acid signature was present when the virus is susceptible to the bNAb. Specificity of the signature was defined as the probability that the amino acid was not present when the virus is not susceptible to the bNAb.
## Statistical analyses
The participants' characteristics at the time of sample collection were described using median (interquartile range [IQR]) or frequency (percentage), either overall or in each group (viremic and non-viremic 4DR-PWH). Comparisons among groups were calculated with the Kruskal-Wallis test or Wilcoxon rank-sum test for continuous variables, χ test, or Fisher's exact test for categorical ones, as appropriate. The Spearman's rank test was used to test linear associations between phenotypic susceptibility and continuous clinical variables. Two-sided P values < 0.05 were considered statistically significant. All analyses were performed using SAS release 9.4 (SAS Institute, Cary, NC, USA).
## References
1. Comi, Valenti, Bologna: Pierluigi et al.
2. Busto Arsizio: Barbara, Menzaghi, Farinazzo et al. "Sassi Serena; MODENA: Cristina Mussini, Adriana Cervo, Giulia Nardini; NAPOLI: Elio Manzillo, Antonella Gallicchio; PADOVA: Anna Maria"
3. Galli, Parisi, Poli et al. (2020) "Burden of disease in PWH harboring a multidrug-resistant virus: data from the PRESTIGIO Registry" *Open Forum Infect Dis*
4. Hsu, Fusco, Henegar et al. (2023) "Heavily treatment-experienced people living with HIV in the OPERA cohort: population characteristics and clinical outcomes" *BMC Infect Dis*
5. Spivack, Pagkalinawan, Samuel et al. (2022) "HIV: how to manage heavily treatment-experienced patients" *Drugs Context*
6. Tailor, Chahine, Koren et al. (2024) "Lenacapavir: a novel long-acting capsid inhibitor for HIV" *Ann Pharmacother*
7. Waters, De, Buckley et al. (2023) "Broadly neutralizing antibodies for human immunodeficiency virus treatment: broad in theory, narrow in reality" *Clin Infect Dis*
8. Gautam, Nishimura, Gaughan et al. (2018) "A single injection of crystallizable fragment domain-modified antibodies elicits durable protection from SHIV infection" *Nat Med*
9. Caskey, Klein, Lorenzi et al. (2015) "Viraemia suppressed in HIV-1-infected humans by broadly neutralizing antibody 3BNC117" *Nature*
10. Caskey, Schoofs, Gruell et al. (2017) "Antibody 10-1074 suppresses viremia in HIV-1-infected individuals" *Nat Med*
11. Scheid, Mouquet, Ueberheide et al. (2025) *Research Article Microbiology Spectrum*
12. Bjorkman, Chait, Nussenzweig (2011) "Sequence and structural convergence of broad and potent HIV antibodies that mimic CD4 binding" *Science*
13. Selzer, Vanderveen, Parvangada et al. (2025) "Susceptibility screening of HIV-1 viruses to broadly neutralizing antibodies, teropavimab and zinlirvimab, in people with HIV-1 suppressed by antiretroviral therapy" *J Acquir Immune Defic Syndr*
14. Gunst, Højen, Pahus et al. (2023) "Impact of a TLR9 agonist and broadly neutralizing antibodies on HIV-1 persistence: the randomized phase 2a TITAN trial" *Nat Med*
15. Gunst, Pahus, Rosás-Umbert et al. (2022) "Early intervention with 3BNC117 and romidepsin at antiretroviral treatment initiation in people with HIV-1: a phase 1b/2a, randomized trial" *Nat Med*
16. Gaebler, Nogueira, Stoffel et al. (2022) "Prolonged viral suppression with anti-HIV-1 antibody therapy" *Nature*
17. Scheid, Horwitz, Bar-On et al. (2016) "HIV-1 antibody 3BNC117 suppresses viral rebound in humans during treatment interruption" *Nature*
18. Eron, Little, Crofoot et al. (2024) "Safety of teropavimab and zinlirvimab with lenacapavir once every 6 months for HIV treatment: a phase 1b, randomised, proof-of-concept study" *Lancet HIV*
19. Ogbuagu, Gaur, Mcmahon et al. (2025) "Efficacy and safety of lenacapavir, teropavimab, and zinlirvimab: phase 2 week 26 primary outcome"
20. Eron, Cook, Mehrotra et al. (2025) "Lenacapavir plus two broadly neutralizing antibodies, teropavi mab and zinlirvimab, for people with HIV-1 highly susceptible to either teropavimab or zinlirvimab" *J Infect Dis*
21. Vanderveen, Selzer, Moldt et al. (2024) "HIV-1 envelope diversity and sensitivity to broadly neutralizing antibodies across stages of acute HIV-1 infection" *AIDS*
22. Moldt, Günthard, Workowski et al. (2022) "Evaluation of HIV-1 reservoir size and broadly neutralizing antibody susceptibility in acute antiretrovi ral therapy-treated individuals" *AIDS*
23. Chaix, Terracol, Nere et al. (2024) "Susceptibility to lenacapavir, fostemsavir and broadly neutralizing antibodies in French primary HIV-1 infected patients in 2020-2023" *J Med Virol*
24. Moldt, Parvangada, Martin et al. (2021) "Evaluation of broadly neutralizing antibody sensitivity by genotyping and phenotyping for qualifying participants to HIV clinical trials" *J Acquir Immune Defic Syndr*
25. Clemente, Galli, Lolatto et al. (2024) "Cohort profile: PRESTIGIO, an Italian prospective registry-based cohort of people with HIV-1 resistant to reverse transcriptase, protease and integrase inhibitors" *BMJ Open*
26. Pahus, Zheng, Olefsky et al. (2025) "Evaluation and real-world experience of a neutralization susceptibility screening assay for broadly neutralizing anti-HIV-1 antibodies" *J Infect Dis*
27. Nei, Li (1979) "Mathematical model for studying genetic variation in terms of restriction endonucleases" *Proc Natl Acad Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12846271&blobtype=pdf | # Analysis of the Specific Expression Profile of Immune Cells in Infants and Young Children Infected with RSV and Construction of a Disease Prediction Model
John Frean, José Leija-Martínez, Adriana Monsiváis-Urenda, Fausto Sánchez-Muñoz, Kai Ren, Honggang Sun, Tian Ren, Kailun Ma, Jizheng Chen
## Abstract
It has been demonstrated that infants and young children exhibit immune tolerance as a consequence of immature immune systems, which are characterized by a natural Th2 bias. RSV infection has been reported to result in acute lower respiratory infection (ALRI), while formalin-inactivated vaccination has been observed to exacerbate Th2 responses, consequently leading to enhanced respiratory disease (ERD). Transcriptomic data from three independent cohorts of RSV-infected infants were analyzed (GSE246622 served as the discovery and train set; GSE105450 and GSE188427 were used as validation sets). Immune infiltration analysis revealed immunological characteristics, which were then used to perform unsupervised clustering using feature-related genes. WGCNA was used to identify co-expressed gene modules, while Mfuzz and TCseq were employed to analyze temporal expression patterns. Machine learning models were developed using a refined panel of candidate genes. Severe symptoms of RSV infection exhibited a strong correlation with age, with younger infants demonstrating more intense inflammatory responses from neutrophils, macrophages, mast cells and dendritic cells. A predictive model was constructed using ten co-expressed genes: The following genes were identified: MCEMP1, FCGR1B, ANXA3, FAM20A, CYSTM1, GYG1, ARG1, SLPI, BMX and SMPDL3A. It was observed that infants of a younger demographic demonstrated a heightened degree of immunosuppression and pronounced innate immune activation in patients of severe symptoms with RSV infection. However, eosinophils exhibited minimal involvement in these processes. These gene models pertaining to the neutrophil, macrophage or mast cell was found to be a relatively effective predictor in patients of severe symptoms.
## 1. Introduction
Respiratory syncytial virus (RSV) infection poses a significant threat to the health of infants and young children, which imposes a substantial global burden, with an estimated 33 million new cases of acute lower respiratory infection (ALRI) occurring each year among children under five [1,2]. RSV The clinical presentation typically involves the acute onset of coughing or dyspnea, often accompanied by tachypnea [1,3]. The incidence of severe disease peaks between 2 and 4 months of age, with the majority of critical cases occurring in infants under 6 months [4]. The consequences of this phenomenon are significant, with the number of hospital admissions and mortalities reaching approximately 3.2 million and 120,000 per annum, respectively [1].
Incomplete maturation of the immune system in early life predisposes infants to pulmonary infections by causing a dysfunctional adaptive immune response that fails to generate effective and durable memory [4]. Neonatal immunity is marked by a Th2polarized predisposition, evident in the cytokine response of TLR-stimulated dendritic cells and monocytes, compounded by a lymphocyte profile rich in recent thymic emigrants that are inherently biased toward Th2 effector differentiation [5,6].
It has been demonstrated that infants' innate Th2-polarized immunity can result in severe ERD following immunization with early formalin-inactivated RSV vaccines [7]. This is characterized by parenchymal tissue damage, bronchopneumonia with atelectasis/pneumothorax, and pulmonary neutrophilia with macrophage/lymphocyte infiltration and eosinophilia [7]. In addition, the immunological characteristic of infants and young children infected with respiratory syncytial virus (RSV) lies in the incomplete maturation of both innate and adaptive immune systems, which not only results in their relatively weak viral clearance ability but also is often accompanied by excessive inflammatory responses [8]. Upon infection with respiratory syncytial virus (RSV), neonates exhibit reduced Toll-like receptor (TLR) signaling, altered antigen-presenting cell (APC) function, decreased expression of innate antiviral cytokines (interferons), and increased production of inflammation-related factors. This may skew the adaptive immune response toward Th2 and Th17 subsets, impairing the protective antiviral functions of Th1 cells and cytotoxic T lymphocytes (CTLs) [9]. However, the majority of research on RSV infection focuses exclusively on the immunological distinctions between infants and young children, adults, and the elderly, with limited attention paid to the variations among infants of different months within one year of age. Exploration of the immunological characteristics of RSV infection in infants is therefore of paramount importance for the prevention and treatment of antivirals, as well as for the research and development of vaccines in this age stage.
In this study, we analyzed transcriptome sequencing data from whole blood samples of infants under one year old hospitalized due to RSV infection, which were obtained from an open database GSE246622. Seven machine learning algorithm models were constructed by training on 87 samples from the GSE246622 dataset, with three independent validation sets utilized, including 170 samples from GSE246622, 65 samples from GSE105450, and 122 samples from GSE188427. To clarify the immunological features of infants aged less than one year in the context of RSV infection and develop a predictive model for predicting severe symptoms of RSV infection, the following analytical methods were adopted: immune infiltration analysis, consensus matrix analysis, WGCNA, Mfuzz, TCseq, and machine learning algorithms. We revealed distinct immune cell profiles in infants featuring T/B cell dysfunction with compensatory inflammatory hyperactivation. A total of ten genes, including MCEMP1, FCGR1B, ANXA3, FAM20A, CYSTM1, GYG1, ARG1, SLPI, BMX and SMPDL3A, were found to demonstrate significant correlations with age and severe symptoms of RSV infection, thus providing novel insights with regard to clinical diagnostics, therapeutics and vaccine design. The specific analysis process is shown in Figure 1.
## 2. Materials and Methods
## 2.1. Data Collection and Immune Cell-Related Genes (IRGs)
The RNA expression matrix and the corresponding clinical information were obtained from the public database, GEO, which is maintained by the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/ (accessed on 15 August 2025)). Our study included three transcriptome datasets of RSV-infected infants (GSE246622 [10], GSE188427, GSE105450 [11]). The 47 immune-related genes (IRGs) were derived from a previously published study, where they were utilized as markers for flow cytometry to identify immune cell populations in RSV-infected infants' blood samples [10].
## 2.2. Analysis of RSV-Infected Cohorts
A total of 533 infants under 12 months from GSE246622 were included in the study, consisting of 56 healthy individuals, 208 convalescent patients, and 257 RSV-infected individuals. The severity of infection was then subjected to further grouping according to ReSVinet score [12], resulting in the identification of 87 mild cases, 116 moderate cases, and 54 severe cases. Principal Component Analysis (PCA) was performed to visualize the sample distribution in each group. The identification of differentially expressed genes (DEGs) between healthy individuals and RSV-infected patients was undertaken in accordance with the following criteria: The first criterion is that |logFC| must be greater than 1; the second is that p-value must be less than 0.05. The application of these criteria resulted in the identification of 86 DEGs. The R package 1.28.4 "enrichplot" was utilized to conduct Gene Ontology (GO) enrichment analysis of DEGs. The generation of heatmaps of the DEGs was achieved by utilizing the complexheatmap R package [13].
## 2.3. Immune Cell Landscape
CIBERSORT [14], EPIC [15], QUANTISEQ [16], XCELL [17], and Single Sample Gene Set Enrichment Analysis (ssGSEA) was employed for immune cell infiltration analysis. Following the establishment of correlations between IRGs and various immune cells, a subset exhibiting a correlation coefficient greater than 0.4 was selected. This subset was then utilized for matrix multiplication with the IRG expression matrix to obtain the final results for the immune cell landscape.
## 2.4. Clustering
The R package 'ConsensusClusterPlus ' [18] was utilized to analyze the GSE246622 cohort, which was divided into three clusters based on the 47 IRGs. The clustering analysis was configured with a maximum of 9 clusters. The process involved 50 iterations of resampling, wherein 80% of the samples were randomly selected in each iteration. Feature sampling was disabled (feature ratio = 1), and the Partitioning Around Medoids (PAM) algorithm was employed using Euclidean distance as the metric. The identification of differentially expressed genes between each cluster was conducted in accordance with the criteria delineated in Section 2.2.
## 2.5. Weighted Correlation Network Analysis (WGCNA) [19]
The R package WGCNA was utilized to investigate the co-expression gene network. The 'pickSoftThreshold' function was applied in order to calculate the soft threshold, with the optimal threshold identified as 12. The minimum number of genes that could be accommodated within each module was set at ten. A total of 1161 genes across the five modules were selected for further analysis. Subsequently, 61 genes were selected for further analysis, as they were present in both 86 differentially expressed genes (DEGs) and 1161 genes from the five modules.
## 2.6. Signatures Identification
The R package "Time course sequencing data analysis" (Tcseq) was employed for the purpose of conducting a clustering analysis of the 61 genes that have been demonstrated to be correlated with clinical symptoms. In Tcseq, fuzzy c-means clustering was performed, and the algorithm was configured to generate four gene clusters. In addition, Mfuzz [20] was utilized for the identification of genes whose expression varies with age. The optimal fuzzifier value was determined using the mestimate function, and the number of clusters was set to generate four clusters as output. Four gene clusters exhibited consistent up-regulation or down-regulation, with either worsening symptoms or increasing age identified upon analysis. A total of 33 genes were identified using Tcseq, whereas 52 genes were identified using Mfuzz. Twenty-five candidate signatures exhibiting a continuous trend in both age and clinical symptoms were selected for further analysis.
## 2.7. Receiver Operating Characteristic (ROC) and Predictive Model Construction
The patients were divided into two individuals, namely inpatients and outpatients. The Area Under the Curve (AUC) of the ROC curves was utilized to evaluate the accuracy of the 25 signatures. Signatures with an AUC > 0.65 were selected for the construction of a predictive model. The model was trained on 87 samples from the GSE246622 dataset using 5-fold cross-validation with ten repeats. A total of three independent validation sets were used: 170 samples from GSE246622, 65 samples from GSE105450, and 122 samples from GSE188427. For model construction, we employed and tuned seven distinct machine learning algorithms using the R Mime1 package [21]. The hyperparameter search space for each algorithm was configured as follows. The Naïve Bayes (nb) model was tuned with a Laplace correction (fL) tested at 0, 0.5, 1, 1.5, and 2, while kernel density estimation was enabled with a bandwidth adjustment factor (adjust) ranging from 0.5 to 1.5. For the weighted Support Vector Machine with a radial basis function kernel (svmRadialWeights), we evaluated a sigma parameter from 5 × 10 -4 to 0.05, a cost parameter from 1 to 20, and class Weights from 0.1 to 10. The Random Forest (rf) algorithm was optimized by varying the number of features considered at each split (mtry), which was sampled across ten values from 2 to 369. The Kernel k-Nearest Neighbors (kknn) model was trained with a fixed distance of 2 and an "optimal" kernel, while the maximum number of neighbors (kmax) was tested with odd values from 5 to 13. Two boosting algorithms were also implemented: AdaBoost, for which the number of iterations (nIter) was tested from 50 to 250 and the method was set to either "Adaboost.M1" or "Real adaboost"; and LogitBoost, which was tuned over a range of nIter values from 11 to 101. Finally, the cancerclass method was applied using the welch.test for feature selection prior to classification.
## 3. Results
## 3.1. Gene Expression Was Different Among Individuals with Different Infection Status and Severity
The population was divided into three statuses based on infection status: healthy, convalescent, and RSV-infected. The severity of infection in RSV-infected individuals was also divided into three subgroups based on ReSVinet score: mild, moderate, and severe. PCA is performed on the expression profile of different subsets. As demonstrated in Figure 2A, patients with different infection severities were shown to exhibit grouping characteristics, as indicated by PCA. Furthermore, the PCA indicates that RSV-infected individuals are clustered separately from healthy and convalescent individuals (Figure 2B). A significant age difference was found among groups of differing symptom severity, with greater severity associated with younger age (Figure 2C). In order to characterize the transcriptome features of individuals infected with RSV, a screening of 86 differentially expressed genes (|logFC| > 1 & p < 0.05) was conducted (Figure 2D,F) Among which, 4 genes (FCRL3, CHI3L1, CXCL8, IL5RA) were downregulated upon RSV infection, while the other 82 differentially expressed genes were all upregulated. GO analysis indicated that the differentially expressed genes between healthy and RSV-infected statuses primarily involved virus-related response genes, which are associated with respiratory mucosal immunity (Figure 2E). This indicates that mucosal immunity plays a core antiviral role in RSV-infected process, and these 86 differentially expressed genes (DEGs) can well reflect the immunological characteristics of infants and young children with RSV infection.
## 3.2. Immune Cell-Related Genes Exhibit the Close Correlation Between the Severity of Symptoms in RSV-Infected Individuals and Immune Cells
Correlations between 47 immune cell-related genes and multiple immune cells were analyzed using five methods: CIBERSORT, EPIC, QUANTISEQ, SSGSEA, and XCELL (Supplementary-Figure S1). Among these, the correlations of 41 genes were greater than 0.4 (Figure 3A). As illustrated in Figure 3B, a correlation was identified between distinct severity groups and immune cells. The functional activity of T and B cells showed a higher correlation with healthy individuals, while the function of neutrophils, macrophages, mast cells, and dendritic cells was more closely associated with severe infections. Because the majority of T and B cell-related functions appeared to be impaired, the functions of neutrophils, macrophages, mast cells, and dendritic cells were enhanced (Figure 3B). In addition, A central finding of this study is the systematic explanation of how the spectrum of immune cells is related to the severity of symptoms in RSV infection. (Figure 3B). The study also reveals that disease severity in immunologically immature infants with RSV infection is associated with two key elements: the spectrum of T and B cells and various types of immunoregulatory cells (Tregs, pDCs, MDSCs, and M2 cells). In order to ascertain the differences in mRNA expression levels among healthy (blue dots), mild (orange dots), moderate (red dots), and severe (purple dots) groups, a principal component analysis (PCA) was conducted (A). Furthermore, a PCA was conducted in order to ascertain the differences in RSVinfected (red dots), healthy (orange dots), and convalescent (blue dots) statuses (B). The correlation between severe symptoms of RSV infection and age is demonstrated in (C). As illustrated in (D,F), the DEGs are displayed on the volcano plot and the heatmap, respectively. The DEGs are analyzed using GO enrichment (E). (ns > 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001; t-test).
## 3.3. Clustering and Annotation of Immune Cell-Related Gene Expression
The expression matrix of 47 immune cell-related genes was extracted from healthy and infected individuals. Three clusters (1 = B, 2 = C, 3 = A) were identified via consensus matrix and principal component analyses. (Figure 4A,B). Figure 4D illustrates distinct severity of infection (healthy, mild, moderate, severe) in clusters A, B, and C. The subjects in Cluster C, who were of a younger age, exhibited a higher prevalence of severe cases during RSV infection. Furthermore, these subjects showed significant disparities in the expression of 47 IRGs (Figure 4C,E). Cluster B consists mainly of healthy individuals, while Cluster C is predominantly composed of moderate-to-severe infected patients. As shown in Heatmap, the 47 immune-related genes exhibit significant differences in expression levels. In comparison with Cluster B, 30 immune-related genes displayed reduced expression, while another 17 had elevated expression. Moreover, a comparison of each pair of clusters A, B, and C revealed significant disparities in gene expression (|logFC| > 1 and p < 0.05) (Figure 4F). A comprehensive compendium of data pertaining to the enrichment of differentially expressed genes in GO analysis is appended (Supplementary-Figure S2). The WGCNA method was employed to construct a gene co-expression network. For all the previously referenced grouped samples, the Sankey diagram presents the correlations between the four categories: group, cluster, status, and age group (Figure 5A). The severity of infection (group) exhibited a close correlation with age and cluster (Figure 4B). Cluster C had predominantly moderate/severe infections, with most cases in those less than 3 months of age (Figure 5B). Subsequently, we employed dynamic hybridization cutting to construct a hierarchical clustering tree and form gene modules. These branches exhibited numerous genes that demonstrated analogous expression profiles. Each individual gene may be regarded as analogous to a leaf in a tree (Figure 5D). The construction of twenty-seven modules was facilitated with transcriptome data (Figure 5E). "Age" showed a significant negative correlation with the "Group" and "Cluster" in co-expressed genes, indicating their expression levels reflect age and infection severity. The sample grouping and clustering is presented in (A,B). The mean connectivity and scale-free fit index of soft threshold power are illustrated in (C). The hierarchical clustering tree of genes based on topological overlap is confirmed in (D). The correlation between these gene modules and age, group, cluster, and status (E) is examined.
## 3.5. TCseq and Mfuzz Were Used to Analyze the Co-Expressed Gene Module
Among the 1161 genes analyzed using WGCNA, 61 exhibited significant differential expression (Figure 6A). The 61 genes were analyzed by TCseq and Mfuzz, respectively (Figure 6B,C). Both TCseq and Mfuzz are analytical tools based on time-series transcriptome data. By analyzing severe symptoms of RSV infection and age as temporal correlations, it was found that only Cluster 2 identified by both methods exhibited a highly positive correlation. The 25 genes constituting Cluster 2 were found to be contingent on age and the severe symptoms of RSV infection (Figure 6B-E).
## 3.6. Ten Genes That Could Distinguish the Severity of RSV Infection Were Screened
The 25 genes that were screened based on Mfuzz and TCseq were utilized for the calculation of the ROC curve. The 10 genes, including MCEMP1, FCGR1B, ANXA3, FAM20A, CYSTM1, GYG1, ARG1, SLPI, BMX and SMPDL3A, which showed an AUC area greater than 0.65, are displayed in Figure 7A. Due to the single-center nature of the analyzed dataset and the limited sample size, the AUC value was relatively low (AUC < 0.7). The expression levels of the ten selected genes in three datasets are illustrated in Figure 7B. Data demonstrate that the differences in the expression levels of the ten genes can distinguish the severity of infection to a certain extent, thereby facilitating the determination of the necessity of hospitalization.
## 3.7. Constructing Machine Learning Algorithm Models
To further improve the credibility of the prediction, the ten screened genes mentioned above were used to construct a classification model for predicting severe symptoms of RSV infection using machine learning algorithms (Figure 8). Consequently, seven machine learning algorithm models were obtained. A portion of the data from the GSE246622 dataset was used as the training set, while the other portion, along with GSE105450 and GSE188427, was employed as the validation set. Compared with the prediction based on the expression level of a single gene, GSE246622, as the validation set, showed that the predictive model constructed using the ten genes together-except for the "cancerclass" model-significantly improved the prediction credibility. The seven constructed predictive models which aimed to determine whether the expression levels of the ten genes can predict severe symptoms of RSV infection in infants under one year old following RSV infection, thereby assisting clinicians in judging the necessity of hospitalization. Integrating data from the seven predictive models can offset the shortcomings of insufficiently significant predictive performance in some models, thereby facilitating more accurate judgmentmaking. Due to variations in multiple factors-including sample genetic backgrounds, sampling time points, and diagnostic criteria for severe symptoms of RSV infection-this led to suboptimal prediction results with GSE105450 and GSE188427, which also constitutes a limitation to the promotion of these prediction models. 'svmRadialWeights': Support Vector Machine (SVM) is capable of processing nonlinear data via kernel functions (Radial Basis Function (RBF) kernel and linear kernel), mapping low-dimensional non-separable data into a high-dimensional space. 'rf': Random Forest is a tree-based algorithm based on "ensemble learning", constructed by multiple decision trees via "bootstrap sampling" and "random feature selection". 'kknn': K-nearest Neighbors is a type of lazy learning algorithm without an explicit training process, with its core being "nearest neighbor voting". 'adaboost': AdaBoost Classification Trees iteratively trains multiple weak classifiers (usually decision stumps) and assigns higher weights to the misclassified samples from the previous iteration. 'LogitBoost': Boosted Logistic Regressions is a linear classification algorithm based on the sigmoid function, which outputs classification results in the form of probabilities (usually using 0.5 as the threshold for binary classification). 'cancerclass': Cancerclass is constructed based on an ensemble algorithm integrating random forest and support vector machine, which is capable of analyzing gene expression profiles. The dashed line represents a 50% probability for both true and false predictions.
## 4. Discussion
RSV infection in infants and young children has been shown to result in viral lower respiratory tract infections (LRI) characterized by a Th2 tendency and increased pulmonary eosinophils in the immune system [4,22]. This has been demonstrated to result in a significant increase in the rates of hospitalization and mortality among infants and young children [4,22]. Furthermore, the severe ERD effect that has been observed in infants and young children following vaccination with formalin-inactivated vaccine (FI-RSV) has become a significant impediment to the development of RSV vaccines for this demographic, with the result that there are currently no RSV vaccines available on the market [23]. Consequently, it is imperative to undertake comprehensive and meticulous investigation of the immunological characteristics exhibited by infants and young children.
In this study, the objective was to utilize the GSE246622 transcriptome data for the purpose of conducting an expression difference analysis and a GO analysis. This analysis revealed significant disparities in differentially expressed genes among the healthy, convalescent, and RSV-infected status. Furthermore, genes relating to respiratory mucosal immunity were shown to be activated in antiviral immunity-related responses. Cell infiltration analysis indicated that younger infants were more likely to develop severe illness when infected with RSV. As demonstrated in the relevant literature, the immune system of an infant exhibits an immune tolerance state, which in turn results in the suppression of adaptive immunity [4,24,25]. In addition, immune tolerance-related cells, including Tregs, Bregs, and M2 macrophages, have been shown to play an important role in limiting severe allergic reactions and asthma caused by RSV infection [24,26]. The application of correlation analysis revealed a negative correlation between the severity of RSV infection and both T and B cells. Furthermore, immune tolerance-related Tregs, pDCs, MDSCs, and M2 cells exhibited a strong positive correlation with the severity of lower respiratory tract infections. During the acute phase of RSV infection, a shift towards a Th2-type response, accompanied by the suppression of IFN-γ antiviral immunity, underlies airway hyperresponsiveness in a subset of susceptible infants and young children [27]. Activated Th2 cells have been observed to secrete large quantities of IL-4, IL-5, and IL-13, which in turn chemotactically attract and activate neutrophils, mast cells, basophils, and eosinophils. These cells have been shown to induce B cell antibody class switching, resulting in the secretion of substantial amounts of IgE antibodies, thereby triggering type I hypersensitivity reactions [4,28]. Our analysis results also showed a positive correlation between the severity of RSV infection and the counts of neutrophils, macrophages, mast cells, and dendritic cells. Thus, we hypothesize that the younger the infant, especially preterm infants, the more severe this immune tolerance becomes, the more prone the innate immune system is to activation, and the more inclined the cytokine secretion of the entire immune system is to Th2 polarization. This Th2 polarization tendency may facilitate the development of tolerance to self-antigens and other foreign antigens in the body, but it may also increase susceptibility to viral infections [29]. When infected with RSV, these immune system characteristics are enhanced, resulting in more severe early-onset respiratory disease (ERD) in younger infants. The results of the correlation analysis demonstrated a positive correlation between age and the severity of RSV infection, as well as between age and the number of neutrophils and mast cells. However, low correlation was observed between eosinophil levels and the severity of infection. The reason why RSV infection is more severe in younger infants is attributed to environmental exposure-particularly RSV infection itself-which may induce airway remodeling in infancy and impair the function of the developing immune system [25]. In addition to direct virus-host interactions, certain bacterial members of the respiratory microbiota may modulate the host's response to RSV, thereby regulating inflammation and potentially influencing disease severity [30]. We have reason to believe that the severe ERD in infants and young children caused by RSV infection is the result of multiple factors. However, the immature development of the immune system is undoubtedly an extremely important contributing factor.
A total of 47 genes associated with immune cells were identified through immunological infiltration analysis. These genes were used to determine the correlation between samples and immune cells through unsupervised clustering analysis, which resulted in the population being divided into three categories. Clusters A and B were compared with cluster C, which was found to have a younger age demographic and a higher proportion of subjects in the severe group. Furthermore, a heat map displaying the expression levels of 47 immune-related genes revealed significant differences among clusters A, B and C. Given that the proportion of individuals manifesting severe and moderate symptoms in cluster B is the least substantial, the volcano plot (Figure 4F) demonstrates that, in comparison with the expression difference of cluster A/cluster C, the number and intensity of differentially expressed genes in cluster B/cluster C are considerably more pronounced. GO analysis of differentially expressed genes shows that, in comparison with cluster A/cluster B, cluster B/cluster C exhibits increased intensity of granulocyte degranulation (Supplementary-Figure S2). Consistent with previous reports, acute infection with respiratory syncytial virus (RSV) can induce degranulation of mast cells, basophils, and eosinophils. This process results in the release of a significant quantity of intracellular active mediators into surrounding tissues, leading to immune damage [4]. The expression matrices of healthy and infected individuals were used for WGCNA co-expression analysis. The purpose of this step was to screen out co-expression modules related to multiple clinical information through co-expression analysis and take the intersection with the 87 differentially expressed genes screened to screen out 61 genes.
In the context of severe symptoms of RSV infection as a quasi-time continuous index, the Mfuzz and TCseq algorithms are employed to undertake time series analysis of the expression matrix, with the objective of evaluating the two indicators of severe symptoms of RSV infection and age. The intersection of the genes obtained from the two time-series analyses was performed to obtain 25 candidate genes. The application of ROC calculation to the screening of ten genes (MCEMP1, FCGR1B, ANXA3, FAM20A, CYSTM1, GYG1, ARG1, SLPI, BMX and SMPDL3A) enabled the distinction of the severe symptoms of RSV infection. The construction of seven machine-learning algorithm models was achieved using ten genes that have been demonstrated to be capable of predicting severe symptoms of RSV infection. The prediction models were verified using three datasets: a portion of GSE246622, GSE105450 and GSE188427. The finding demonstrate that the developed model exhibits a certain predictive capacity for severe symptoms of RSV infection; however, multiple AUC values failed to exceed 0.7 due to disparities in genetic backgrounds, sampling time points, and diagnostic criteria for severe symptoms of RSV infection among the validation datasets GSE105450 and GSE188427, indicating that the predictive model still has inherent limitations that warrant further improvement.
## 4.1. MCEMP1 and CYSTM1
Mast Cell Expressed Membrane Protein 1 (MCEMP1) is a single-channel transmembrane protein involved in regulating the differentiation activities and immune responses of mast cells. Mast cells aggravate sepsis by interfering with the phagocytic activity of resident macrophages and increasing the release of inflammatory cytokines [31]. Cystinosin 1 (CYSTM1) is a novel cysteine-rich transmembrane module that plays a role in stress tolerance across eukaryotes and is significantly associated with a wide variety of immune cell types [32,33].
## 4.2. FCGR1B
Homo sapiens Fc fragment of IgG receptor 1 B (FCGR1B) is highly expressed in neutrophils. Tuberculosis promotes phagocytosis and induces severe inflammatory responses and pathological damage [34]. Neutrophils are the most abundant cell type in the airways of children [25], and it can be speculated that FCGR1B causes severe pathological damage by activating neutrophils to release a large number of inflammatory factors.
## 4.3. ANXA3 and GYG1
The protein encoded by Annexin A3 (ANXA3) is called lipocalin 3. It is a member of the calcium-binding protein family and contributes to inflammation-induced lung injury by activating nuclear factor-κB (NF-κB) [31]. In a transcriptomic study of neutrophils in peripheral blood, it was found that the expression of ANXA3 significantly increased throughout the course of sepsis [31]. Glycogen synthase 1 (GYG1) belongs to the glycogenin family and is primarily responsible for initiating glycogen synthesis. In addition, ANXA3, GYG1 and Arginase 1 (ARG1) were predicted to participate in neutrophil degranulation [35].
## 4.4. FAM20A and ARG1
There is a correlation between Family with Sequence Similarity 20 Member A (FAM20A) and ARG1, and an increase in neutrophil abundance [36,37]. In humans, ARG1 is mainly released by the liver and neutrophils. ARG catalyzes the degradation of arginine into ornithine and urea. By depleting arginine in the extracellular environment, it downregulates the expression of the CD3ζ chain in T lymphocytes, thereby inhibiting the activation and proliferation of T cells through the CD3/T-cell receptor (TCR) complex, thus inhibiting T cell activation and producing strong immunosuppression [38]. The expression of ARG1 is mainly induced by type 2 cytokines (IL-4, IL-13) and immunosuppressive cytokines (TGF-β, IL-10) [39].
## 4.5. SLPI
Secretory Leukocyte Protease Inhibitor (SLPI) is mainly expressed in the lungs, cervical mucosa, body fluids, and the skin. LPS, IL-1, TNF-α, Neutrophil elastase (NE) and neutrophil α-defensin can increase protein expression levels. Its main function is to act as a serine protease inhibitor, which can protect tissues from degradation by a variety of proteases such as cathepsin G, elastase, trypsin, chymotrypsin, chymase, and tryptase. Among them, NE, which is mainly produced by neutrophils, is regarded as the main protease target of SLPI [40]. In monocytes, SLPI can prevent the activation of NF-κB by inhibiting the degradation of NF-κB (IκB-α and IκB-β), thereby restricting the release of inflammatory factors [40].
## 4.6. BMX
Bone marrow tyrosine kinase on chromosome X (BMX) is a member of the TEC family of non-receptor tyrosine kinases [41]. In progenitor cell populations in the bone marrow and mature hematopoietic cell populations of the granulocyte/monocyte lineage, Bmx expression increases with maturation and differentiation. High levels of BMX are also found in mature peripheral neutrophils and monocytes/macrophages [42]. It regulates various cellular processes and participates in the inflammatory response cascade by regulating Toll-like receptor-induced interleukin (IL)-6 production [41,43]. Neutrophils are the most abundant cell type in the airways of children with bronchiolitis caused by respiratory syncytial virus infection; however, their exact role in the pathological response remains unclear [25].
## 4.7. SMPDL3A
Sphingomyelin phosphodiesterase acid-like 3A (SMPDL3A), a member of the acid sphingomyelinase (aSMase) family, is strongly regulated by cholesterol loading [44]. Cholesterol-activated LXR upregulates SMPDL3A expression and then selectively hydrolyzes 2 ′ ,3 ′ -cGMP to inhibit type I interferon and NF-κB signaling pathways, thus achieving the effect of inhibiting inflammation [44,45].
Since neutrophils, macrophages and mast cells are positively correlated with the age and severe symptoms of RSV infection in infants and young children, it is reasonable to use the seven neutrophil-related genes FCGR1B, ANXA3, GYG1, FAM20A, ARG1, SLPI, and BMX as markers to predict the severe symptoms of RSV infection in infants and young children. In addition, MCEMP1, CYSTM1 and SMPDL3A also showed a high correlation with inflammation-related macrophages, mast cells and NF-κB signaling pathways, which are logically selected for prediction.
## 5. Conclusions
In conclusion, analysis of the whole blood cell transcriptome dataset GSE246622 of infantile RSV infection showed that there was a high correlation between the severe symptoms of RSV infection and the age of the infected. Immune infiltration has been demonstrated to be a marker of severity, with younger patients exhibiting more severe symptoms. This phenomenon is believed to be associated with the immune tolerance state of adaptive immunity and the over-activation of innate immunity. It is noteworthy that, although eosinophils have been documented as being present in abundance in the lungs during RSV-induced high reactivity in the lower respiratory tract, there is low correlation with severe symptoms of RSV infection. Following a thorough analysis, seven machine learning algorithm models were constructed utilizing the WGCNA, Mfuzz, TCseq, and machine learning algorithms. In the future, the scope of the study population (individuals with similar genetic backgrounds) will be further refined and restricted, the classification criteria for infection severity standardized, and the sampling process optimized to enhance the predictive performance of the models. Finally, the seven predictive models will be integrated to further improve prediction accuracy, thereby facilitating their translation into clinical applications.
## References
1. Mejias, Rodriguez-Fernandez, Oliva et al. (2020) "The journey to a respiratory syncytial virus vaccine" *Ann. Allergy Asthma Immunol*
2. Hall, Weinberg, Iwane et al. (2009) "The burden of respiratory syncytial virus infection in young children" *N. Engl. J. Med*
3. Nair, Nokes, Gessner et al. (2010) "Global burden of acute lower respiratory infections due to respiratory syncytial virus in young children: A systematic review and meta-analysis" *Lancet*
4. Eichinger, Kosanovich, Lipp et al. (2021) "Strategies for active and passive pediatric RSV immunization" *Ther. Adv. Vaccines Immunother*
5. Kollmann, Crabtree, Rein-Weston et al.
6. Hajjar (2009) "Neonatal innate TLR-mediated responses are distinct from those of adults" *J. Immunol*
7. Opiela, Koru-Sengul, Adkins (2009) "Murine neonatal recent thymic emigrants are phenotypically and functionally distinct from adult recent thymic emigrants" *Blood*
8. Halstead "Vaccine-Associated Enhanced Viral Disease: Implications for Viral Vaccine Development" *BioDrugs*
9. Ruckwardt, Morabito, Graham (2016) "Determinants of early life immune responses to RSV infection" *Curr. Opin. Virol*
10. Lambert, Sagfors, Openshaw et al. (2014) "Immunity to RSV in Early-Life" *Front. Immunol*
11. Zivanovic, Oner, Abraham et al. "Single-cell immune profiling reveals markers of emergency myelopoiesis that distinguish severe from mild respiratory syncytial virus disease in infants" *Clin. Transl. Med*
12. Heinonen, Velazquez, Ye et al. "Immune profiles provide insights into respiratory syncytial virus disease severity in young children" *Sci. Transl*
13. Justicia-Grande, Pardo-Seco, Cebey-Lopez et al. (2016) "Development and Validation of a New Clinical Scale for Infants with Acute Respiratory Infection: The ReSVinet Scale" *PLoS ONE*
14. Gu, Eils, Schlesner (2016) "Complex heatmaps reveal patterns and correlations in multidimensional genomic data" *Bioinformatics*
15. Newman, Liu, Green et al. (2015) "Robust enumeration of cell subsets from tissue expression profiles" *Nat. Methods*
16. Racle, De Jonge, Baumgaertner et al. (2017) "Simultaneous enumeration of cancer and immune cell types from bulk tumor gene expression data"
17. Finotello, Mayer, Plattner et al. (2019) "Molecular and pharmacological modulators of the tumor immune contexture revealed by deconvolution of RNA-seq data" *Genome Med*
18. Aran, Hu, Butte et al. (2017) "Digitally portraying the tissue cellular heterogeneity landscape" *Genome Biol*
19. Wilkerson, Hayes (2010) "ConsensusClusterPlus: A class discovery tool with confidence assessments and item tracking" *Bioinformatics*
20. Zhang, Horvath (2005) "A general framework for weighted gene co-expression network analysis" *Stat. Appl. Genet. Mol. Biol*
21. Kumar, Futschik, Mfuzz (2007) "A software package for soft clustering of microarray data" *Bioinformation*
22. Liu, Zhang, Zhang et al. (2024) "Mime: A flexible machine-learning framework to construct and visualize models for clinical characteristics prediction and feature selection" *Comput. Struct. Biotechnol. J*
23. Shay, Holman, Newman et al. (1999) "Bronchiolitis-associated hospitalizations among US children, 1980-1996" *JAMA*
24. Polack, Teng, Collins et al. (2002) "A role for immune complexes in enhanced respiratory syncytial virus disease" *J. Exp. Med*
25. Pieren, Boer, De Wit (2022) "The adaptive immune system in early life: The shift makes it count" *Front. Immunol*
26. Heinonen, Rodriguez-Fernandez, Diaz et al. (2019) "Infant Immune Response to Respiratory Viral Infections" *Immunol. Allergy Clin. N. Am*
27. Lotz, Peebles, Jr (2012) "Mechanisms of respiratory syncytial virus modulation of airway immune responses" *Curr. Allergy Asthma Rep*
28. Binns, Tuckerman, Licciardi et al. (2022) "Respiratory syncytial virus, recurrent wheeze and asthma: A narrative review of pathophysiology, prevention and future directions" *J. Paediatr. Child. Health*
29. Prescott (2006) "The development of respiratory inflammation in children" *Paediatr. Respir. Rev*
30. Restori, Srinivasa, Ward et al. (1249) "Neonatal Immunity, Respiratory Virus Infections, and the Development of Asthma" *Front. Immunol*
31. De Steenhuijsen Piters, Heinonen, Hasrat et al. (2016) "Nasopharyngeal Microbiota, Host Transcriptome, and Disease Severity in Children with Respiratory Syncytial Virus Infection" *Am. J. Respir. Crit. Care Med*
32. Huang, Cheng, Deng et al. (2024) "Integrated Bioinformatics Exploration and Preliminary Clinical Verification for the Identification of Crucial Biomarkers in Severe Cases of COVID-19" *J. Inflamm. Res*
33. Venancio, Aravind (2010) "CYSTM, a novel cysteine-rich transmembrane module with a role in stress tolerance across eukaryotes" *Bioinformatics*
34. Zhang, Chen, An et al. (2023) "Analysis and validation of diagnostic biomarkers and immune cell infiltration characteristics in pediatric sepsis by integrating bioinformatics and machine learning" *World J. Pediatr*
35. Wu, Li, Chen et al. "Probe Signal Values in mRNA Arrays Imply an Excessive Involvement of Neutrophil FCGR1 in Tuberculosis"
36. Cao, Zhang, Liu et al. (2024) "Identification of hub genes and pathogenesis in Kawasaki disease based on bioinformatics analysis. Indian" *J. Pathol. Microbiol*
37. Zhang, Sun, Liang et al. (2024) "FAM20A: A potential diagnostic biomarker for lung squamous cell carcinoma" *Front. Immunol*
38. Cane, Barouni, Fabbi et al. (2023) "Neutralization of NET-associated human ARG1 enhances cancer immunotherapy" *Sci. Transl. Med*
39. Rotondo, Barisione, Mastracci et al. (2009) "IL-8 induces exocytosis of arginase 1 by neutrophil polymorphonuclears in nonsmall cell lung cancer" *Int. J. Cancer*
40. Grzywa, Sosnowska, Matryba et al. (2020) *J. Myeloid Cell-Derived Arginase in Cancer Immune Response. Front. Immunol*
41. Mckiernan, Mcelvaney, Greene (2011) "SLPI and inflammatory lung disease in females" *Biochem. Soc. Trans*
42. Huang, Zhang, Liao et al. "BMX deletion mitigates neuroinflammation induced by retinal ischemia/reperfusion through modulation of the AKT/ERK/STAT3 signaling cascade"
43. Weil, Power, Smith et al. (1997) "Predominant expression of murine Bmx tyrosine kinase in the granulo-monocytic lineage" *Blood*
44. Wang, Xia, Fang et al. (2017) "A novel BMX variant promotes tumor cell growth and migration in lung adenocarcinoma" *Oncotarget*
45. Traini, Quinn, Sandoval et al. (2014) "Sphingomyelin phosphodiesterase acid-like 3A (SMPDL3A) is a novel nucleotide phosphodiesterase regulated by cholesterol in human macrophages" *J. Biol. Chem*
46. Shin, Chung (2023) "SMPDL3A links cholesterol metabolism to the cGAS-STING pathway" *Immunity*
47. "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=PMC12751290&blobtype=pdf | # Structural and dynamic insights into F13L variations in monkeypox virus and their impact on tecovirimat resistance
Yingzhi Wang, Yao Wang, Cuiling Wu, Yingzhi Wang, Tao Li
## Abstract
The recent global outbreak of Monkeypox virus since 2022, particularly outside Africa, has underscored its significant threat to public health. Genomic surveillance from multiple countries has revealed diverse genetic variations among circulating Monkeypox virus strains. Notably, a common F13L (E353K) mutation was identified in strains responsible for concentrated outbreaks in the United Kingdom and North America. The F13L is the current target of the only approved drug for poxvirus treatment, tecovirimat (previously known as ST-246), Previous studies have reported that the F13L (G277C) mutation significantly increases resistance. However, the impact of the shared F13L (E353K) mutation on viral transmission and its potential role in conferring tecovirimat resistance remains unclear. To address this, we employed homology modeling, molecular docking, and molecular dynamics simulations to investigate the phospholipase activity of the F13L protein. Our findings suggest that resistance to tecovirimat in F13L and its mutants may be attributed to altered flexibility of the drug-binding pocket and changes in the distance between the H334 residue and the fluorine atom in tecovirimat. This study provides a framework for rapidly assessing the resistance of emerging Monkeypox virus variants to tecovirimat and for guiding the rational modification of existing therapeutics to counter new viral threats.
## Introduction
Monkeypox virus (MPXV) belongs to the genus Orthopoxvirus (OPXV), which includes variola virus (smallpox) and cowpox virus, poses a significant public health threat by causing diseases characterized by fever, rash, and lymphadenopathy [1,2]. Two primary genetic clades of MPXV have been identified: the more virulent Congo Basin clade and the less severe West Africa clade, which historically demonstrated lower human-to-human transmissibility [3]. The global threat posed by MPXV was starkly illustrated in 2022, when over 3,300 confirmed cases were reported worldwide by June, marking the largest monkeypox epidemic recorded outside of Africa [4]. A major outbreak in the United Kingdom between May and September 2022 alone accounted for more than 3,500 cases [5]. prompting the World Health Organization to declare the situation a Public Health Emergency of International Concern in July [6]. The rapid emergence of MPXV in numerous non-endemic countries, including the UK, Portugal, and Australia, highlighted a heightened and widespread threat compared to historical patterns in endemic African regions [7]. This urgent situation underscores the critical need for developing effective antimonkeypox vaccines and therapeutics.
MPXV is a large double-stranded DNA virus with a genome of approximately 197 kb encoding nearly 190 genes. Recent genomic sequencing efforts, such as those by the UK Health Security Agency, have been pivotal in tracking the outbreak [5]. Sequencing data revealed that the dominant 2022 strain circulating in the UK possesses 48 single-nucleotide polymorphisms compared to a 2018 reference strain, 21 of which result in amino acid changes [8]. Among these, four mutations in viral proteins C23L (S105L), C22L (S54F), C19L (D266N), and F13L (E353K) were classified as medium priority [8]. Notably, the F13L (E353K) mutation was also identified in US outbreak strains and was conserved across all viral sequences within the predominant 2022 global cluster [9], suggesting a potential role in the virus's widespread transmission.
The F13L protein is an envelope phospholipase that is critical for viral dissemination [10,11]. It facilitates the wrapping of virus particles by trans-Golgi components, leading to the formation of triple-membrane-enveloped viruses that are essential for efficient cell-to-cell spread [12]. Impairment of F13L function through deletion or mutation can reduce virus yield by up to 100-fold by blocking the formation of these enveloped virions. Tecovirimat, an FDA-approved antiviral, specifically targets the F13L protein and has demonstrated efficacy against MPXV in vitro and in animal models [13,14]. However, the potential for resistance is a concern, as evidenced by a study showing that a single F13L (G277C) mutation in cowpox virus conferred over 800-fold resistance to tecovirimat [15]. It remains unclear whether the widespread F13L (E353K) mutation affects the binding affinity of tecovirimat or contributes to viral fitness. Elucidating the binding mechanism of tecovirimat to both wild-type F13L and its mutants is therefore crucial, as understanding the functional consequences of the E353K mutation is vital for assessing its impact on the 2022 outbreak and guiding treatment strategies.
In this study, we employed computational approaches to investigate the molecular interactions between tecovirimat and the F13L protein. We characterized the binding patterns and evaluated the effects of point mutations, including E353K, on these interactions. Our findings aim to provide insights for epidemic response and the design of new therapeutics against evolving MPXV strains.
## Results
## Monkeypox virus F13L possesses the conserved HxK(x)4D(x)6GSxxN motif of the phospholipase D enzyme family
We first obtained the full-length amino acid sequence of F13L from NCBI and used AlphaFold2, AlphaFold3, and RoseTTAFold for structure prediction. The predicted full-length model comprises 372 amino acids and exhibits an overall structure consisting of 17 β-sheets and 14 α-helices (Fig. 1A). The three models demonstrated strong structural alignment, except for the N-terminal signal peptide region, supporting the reliability of the predictions (Figure S1 & Table S1). Subsequent sequence analysis revealed that F13L shares high similarity with the phospholipase D (PLD) superfamily. To further investigate this, we performed a homologous protein structure search using the SWISS-MODEL server [16] and selected three representative PLD structures for comparison: human (PDB: 6U8Z), Streptomyces antibioticus (PDB: 7JRU), and Serratia plymuthica (PDB: 7E0M) [17][18][19]. Structural alignment showed that the predicted F13L model and the reference PLD proteins share a conserved folding pattern, characterized by two mirror-symmetric domains. Each domain is composed of seven β-strands and six α-helices interconnected by loops, forming a central pocket at the symmetry interface. Structural alignment revealed that this central pocket exhibits high conservation with the catalytic pockets of canonical PLD enzymes, which are known to bind phosphate groups and phospholipid substrates such as phosphatidic acid [20]. Specifically, the geometric and physicochemical properties of this pocket including its size, shape, and arrangement of key polar residues are consistent with the binding modes observed in homologous PLD structures complexed with phosphate ions or phospholipid analogs (PDB: 6U8Z, 7JRU, 7E0M) [17][18][19]. Furthermore, the presence of the conserved HxK(x)₄D(x)₆GSxxN motif, which coordinates phosphate moieties during catalysis in PLD enzymes, supports the functional relevance of this pocket. We therefore propose that this pocket serves as the active site in F13L and is capable of binding phosphate, sulfate, or phosphatidic acid substrates (Fig. 1B).
We next performed a multiple sequence alignment of PLD proteins from Monkeypox virus, Vaccinia virus, Buffalopox virus, Sheeppox virus, and diverse species including Serratia, Arabidopsis, Streptomyces, Klebsiella, Mouse, and Human (Fig. 2). This analysis confirmed that F13L and other PLD family proteins share the conserved HxK(x)4D(x)6GSxxN motif (where x refers to any amino acid residue) [21]. Notably, the canonical histidine (H) in this motif is substituted by an asparagine (N) in the viral orthologs. Given the similar polarity and nucleophilic capacity of these residues, this substitution is unlikely to impair the catalytic degradation of phospholipids. These findings strongly support the classification of F13L within the PLD protein family and suggest it retains broad phospholipase activity, potentially acting through diverse metabolic pathways.
## Comparative analysis of the substrate binding pocket in Monkeypox virus F13L
While numerous eukaryotic phospholipase structures have been characterized, the first crystal structure of a prokaryotic PLD superfamily member, endonuclease (Nuc), was solved by Stuckey and Dixon in 1999 [22]. Their work identified a conserved HxK(x)4D(x)6GSxN motif, where residues His (H), Lys (K), Asp (D), Gly (G), Ser (S), and Asn (N) form an extensive hydrogen bond network within the active site. This network facilitates substrate binding, charge neutralization, and the formation of a phosphoenzyme intermediate. Later, Leiros and Hough determined a more canonical PLD structure from Streptomycessp. (PDB: 7JRU) [23]. We therefore used this well-defined 7JRU structure as a reference to analyze the amino acid composition and polarity of the putative active site pocket in our AlphaFold2-predicted F13L model.
As shown in Fig. 3A and S2, the 7JRU structure contains a defined phosphatidic acid binding pocket. Similarly, we identified a deep, putative substrate-binding pocket at the interface of the two mirror-symmetric domains on the F13L surface, formed by multiple loops. Notably, we observed that residues R89 and K281 form a lock-like lid at the entrance of this pocket (Fig. 4A), which may enhance stability for substrates or small molecule inhibitors. We further analyzed the amino acid composition and polarity of the 7JRU active site, which is lined with polar residues including Y466, Y491, D467, D200, R385, K443, K170, N459, H448, and H168 (Fig. 4B). In contrast, the putative active pocket of F13L exhibits a unique amphipathic character, comprising polar residues (H334, H338, N329, K314, S327, N312, N133, S135, C120) interspersed with hydrophobic residues (T137, I144, F52, L239, L118). This divergence in composition and polarity suggests potential differences in substrate specificity and functional mechanisms between viral and bacterial PLDs. We speculate that the hydrophobic residues in F13L may mediate interactions with substrate lipid chains, while the polar residues are likely responsible for catalytic activity.
In addition, we analyzed the amino acid composition and polarity of the 7JRU structure and found that the structure consists of a large number of polar amino acids such as Y466, Y491, D467, D200, R385, K443, K170, N459, H448, H168 (Fig. 4B). While for F13L, there is a unique and hydrophobic active pocket composed of polar amino acids such as H334, H338, N329, K314, S327, N312, N133, S135, C120 and Hydrophobic residues such as T137, I144, F52, L239, L118. The differences in amino acid composition and polarity of phospholipase active centers between Monkeypox virus and Streptomyces suggest that there may be some differences in phospholipase binding substrate and function among different species. We speculate that the hydrophobic residues in monkeypox virus F13L are related to the interaction of substrate hydrophobic chains, and the polar residues should play a major catalytic role.
## Molecular Docking of Tecovirimat with F13L and its mutants
As F13L is the primary target of tecovirimat, and mutations such as G277C and E353K have been associated with potential drug resistance, it is critical to first characterize the binding interactions between tecovirimat and both wild-type F13L and its mutant variants. Molecular docking was therefore employed as an initial computational approach to predict the optimal binding conformations of tecovirimat in the active site of each F13L protein, compare the binding affinities across wildtype and mutants, and identify key residue-drug interactions that might be disrupted by mutations.
The lowest-energy binding poses of tecovirimat within the wild-type F13L and its G277C and E353K mutant structures are shown in Fig. 5. The G277C mutation constricted the binding pocket, whereas the wild-type and E353K mutant maintained a wider pocket conformation. Despite this, tecovirimat adopted a similar binding mode across all three variants. Specifically, the hydrophobic core of tecovirimat occupied the upper region of the pocket, while the polar fluorine atom was oriented toward the lower part of the binding site.
Analysis of specific interactions revealed that in the wild-type protein, tecovirimat forms a hydrogen bond with Asn312. Furthermore, the imidazole ring of His334 engages in an anion-π interaction with the fluorine atom. Asn312 is part of the conserved PLD family motif, and His334 is speculated to be a key catalytic residue within a charge relay system. These interactions suggest that tecovirimat inhibits the catalytic function by disrupting the
## Molecular dynamics simulations reveal the structural basis of Tecovirimat resistance
To investigate the binding stability and dynamics of tecovirimat, we performed 100 ns molecular dynamics (MD) simulations. Structural superimposition of 10 snapshots extracted at 10 ns intervals revealed distinct behaviors (Fig. 6). The wild-type and E353K complexes exhibited clustered, stable conformations throughout the simulation, indicating a conserved and tight binding mode. In contrast, the G277C mutant showed significant fluctuation and displacement of tecovirimat within the binding pocket. This instability can be attributed to the proximity of residue 277 to the ligand binding site, whereby the G277C mutation exerts a more direct steric and energetic impact on small molecule binding compared to the more distal E353K mutation.
We quantified the stability of the key anion-π interaction by measuring the distance between the fluorine atom of tecovirimat and the His334 residue (Fig. 7). The F-His334 distance remained stable for the E353K mutant (0.4-0.8 nm) and wild-type protein (0.3-0.8 nm) throughout the simulation. However, this distance fluctuated markedly in the G277C mutant, indicating a disrupted and unstable interaction that would likely impair binding affinity.
Further analysis of the simulation trajectories confirmed the destabilizing effect of the G277C mutation (Fig. 8). The root-mean-square deviation (RMSD, a measure of structural deviation from the initial conformation), radius of gyration (Rg, an indicator of protein compactness), and solvent-accessible surface area (SASA, a measure of the protein surface exposed to solvent) values were consistently higher for the G277C complex than for the wild-type and E353K complexes. Additionally, the root-mean-square fluctuation (RMSF, a measure of structural stability) profile showed increased flexibility specifically around residues 260-270, a region proximal to the active center. This collective instability in the G277C mutant complex underpins the mechanism of drug resistance.
Finally, we employed the Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) method-a widely used end-state approach for calculating binding . Black, red, and blue lines represent wild-type, G277C, and E353K, respectively. (E) Binding free energy and its components (kcal/mol) calculated using the MM/GBSA method free energies in molecular simulations-to compute the binding free energies from 200 snapshots of the last 20 ns of simulation. The energy components are detailed in Fig. 8E. The results demonstrate a substantial decrease in binding free energy for the G277C mutant relative to the wild-type, confirming the development of strong drug resistance. In contrast, the E353K mutation had a minimal effect on binding, with a slight increase in affinity compared to the wild-type, indicating this variant remains susceptible to tecovirimat.
## Discussion
Since the COVID-19 pandemic began in 2019, the global spread of MPXV has highlighted the persistent threat of emerging viral pathogens [24]. The increasing frequency of MPXV outbreaks worldwide has been accompanied by the continuous evolution of the virus, with human populations providing a favorable environment for its adaptation and mutation [25]. The current therapeutic arsenal remains limited, with tecovirimat being one of the few approved drugs targeting a single viral protein, consistent with recent findings showing that tecovirimat specifically disrupts the F13L activity by promoting its dimerization [10]. This underscores the urgent need to accelerate drug development efforts to address the emerging challenge of drug-resistant viral variants.
Computational approaches, including molecular docking and molecular dynamics simulations, offer powerful tools for rapidly assessing and understanding potential resistance mechanisms. In this study, we demonstrated that the MPXV F13L gene encodes a functional phospholipase, characterized by a conserved HxK(x)4D(x)6GSxN motif common to the PLD family (Figs. 1, 2 and3). Furthermore, we identified a potential small-molecule binding pocket in F13L, formed by key residues including Phe52, Thr137, Asn312, Lys314, Asn329, His334, and His338, which provides the structural basis for tecovirimat binding (Fig. 5).
Building on this structural foundation, we further explored the complexity of tecovirimat-F13L interactions-an area recently advanced by a 2025 study that complements yet expands our understanding [10]. This study demonstrated that tecovirimat inhibits F13L function by promoting F13L dimerization: specifically, the drug binds at F13L's dimer interface, and clinical tecovirimat-resistant mutations (e.g., sF13(A295E), sF13(N267D, A288P, A290V, D294V), sF13(ΔN267)) all localize to this interface, with such mutations abrogating tecovirimat-induced dimerization. Notably, this study also uncovered a critical unresolved question relevant to our work: the G277C mutation (a well-documented tecovirimat-resistant variant) is spatially distant from F13L's dimer interface, yet it still confers resistance-suggesting the existence of dimerization-independent resistance mechanisms. Given this gap, our MD simulations of monomeric F13L provide targeted insights: we observed tecovirimat-induced local conformational changes in F13L's functional domains, which may contribute to resistance when disrupted by mutations like G277C.
We systematically evaluated the impact of the F13L E353K mutation on tecovirimat susceptibility, using the known drug-resistant mutant G277C as a positive control and the wild-type protein as reference (Figs. 6, 7 and 8). Our docking and MD simulations revealed that the G277C mutation induces increased flexibility in the binding pocket and disrupts a critical anion-π interaction by increasing the distance between the fluorine atom of tecovirimat and the positively charged His334 residue. This reduction in binding stability provides a mechanistic explanation for the observed resistance of the G277C mutant to tecovirimat. In contrast, the E353K mutation, located distal to the binding pocket, had minimal impact on pocket flexibility and preserved the F-His334 interaction distance. Consequently, the E353K variant maintained sensitivity to tecovirimat. These computational findings are strongly supported by recently published experimental data showing comparable binding affinities of tecovirimat for wild-type F13L (0.007731 ± 0.002 mM) and the E353K mutant (0.002860 ± 0.001 mM), The concordance between our simulations and experimental results validates our computational approach and suggests that analysis of binding pocket microenvironments and specific drug-residue interactions (particularly the F-His334 distance) could serve as predictive indicators of tecovirimat resistance in F13L variants. This framework also provides a strategic approach for evaluating potential resistance conferred by other point mutations in F13L and other viral targets.
Beyond its role as a drug target, F13L is known to facilitate the formation of viral membrane precursors through its phospholipase activity in vaccinia virus [26]. While our data indicate that the E353K mutation does not impair the structural stability of the phospholipase active center, we hypothesize that this mutation might enhance viral transmission by potentially modulating the formation of viral envelope precursors. This could provide a mechanistic explanation for the widespread transmission of E353K-containing strains during the 2022 outbreak. However, this hypothesis requires further experimental validation through virological and biochemical studies.
## Conclusion
In this study, we employed a comprehensive computational approach-integrating homology modeling, molecular docking, and molecular dynamics simulations-to investigate the molecular mechanisms underlying tecovirimat resistance associated with mutations in the monkeypox virus F13L protein. Our findings provide critical insights for addressing the challenge of rapidly evolving viral variants and offer a strategic framework for the rational modification of existing therapeutics to counteract emerging resistance.
Our analyses demonstrate that the drug resistance conferred by the G277C mutation arises from increased flexibility of the binding pocket and disruption of a key interaction between the drug's fluorine atom and the His334 residue. In contrast, the prevalent E353K mutation, located distal to the binding site, does not significantly alter the binding affinity or dynamics of tecovirimat. This mechanistic understanding enables rapid in silico assessment of newly identified mutations for their potential to cause drug resistance.
Future work should focus on refining these computational models through advanced sampling techniques and machine learning algorithms to enhance predictive accuracy. The development of such rapid computational assessment technologies will be crucial for providing timely guidance in response to new clinical outbreaks and informing the design of next-generation antiviral drugs.
## Computational methods
## 1. Sequence alignment
To investigate the evolutionary conservation of F13L, we performed a comprehensive sequence analysis. The amino acid sequence of Monkeypox virus F13L (Uniprot entry: Q50LD4) was used as a query for BLASTP searches against the non-redundant protein database with an E-value cutoff of 1e-10. Representative sequences from different viral strains and species were selected for comparison, including: Vaccinia virus (P04021), Buffalopox virus (A0A2P1JPL8), Sheeppox virus (Q9QCQ2), as well as phospholipase D homologs from Serratia plymuthica (Q53728), Arabidopsis thaliana (Q38882), Streptomyces antibioticus (Q53728), Mouse (Q9Z280), and Human (Q13393, O14939). Multiple sequence alignment was performed using the MUSCLE algorithm with default parameters, and the resulting alignment was visualized using ESPript 3.0 (https://espript.ibcp.fr) to g e n e r a t e Fig. 2.
## 2. System preparation and molecular docking
The three-dimensional structure of the F13L protein was constructed using AlphaFold2, RoseTTA fold and Alpha-Fold3 [27,28]. The amino acid sequence of F13L was retrieved from NCBI (Accession: UTG40738.1). Given the high structural similarity among the three predicted models (Figure S1). the top-ranked model (ranked_0) generated by AlphaFold2 was selected as the initial structure for subsequent analyses. The F13L(E353K) and F13L(G277C) mutant structures were generated from the wild-type model using the mutagenesis tool in PyMOL v2.5 [29]. Each mutant structure underwent a 50 ns molecular dynamics (MD) simulation to achieve conformational relaxation, and the resulting stabilized structure was used as the starting conformation for molecular docking.
The three-dimensional structure of tecovirimat was obtained from PubChem (CID: 16124688) and energetically minimized using Gaussian 09 software at the HF/6-31G(d) level of theory. Molecular docking was performed with AutoDock 4.2.6 [30]. using a grid box of 60 Å × 60 Å × 60 Å centered on the active site, with a grid spacing of 0.375 Å. The Lamarckian Genetic Algorithm (LGA) was employed with the following parameters: 150 individuals in the initial population, random torsion initialization for each individual, and a maximum of 2,500,000 energy evaluations per docking run [31]. A total of 1000 independent docking runs were conducted for each protein system (wild-type and mutants). During the docking of tecovirimat with wild-type F13L and its E353K/G277C mutants, the protein backbone and all residues (including those in the binding pocket) were treated as rigid. All other parameters were set to their default values.
## 2. Molecular dynamics Simulation
The docked pose with the lowest binding energy for each complex (wild-type, F13L(E353K), and F13L(G277C)) was selected for 100 ns MD simulations. Atomic partial charges for tecovirimat were derived using the Restrained Electrostatic Potential (RESP) method. First, electrostatic potential (ESP) charges were computed with Gaussian 09 at the HF/6-31G(d) level, which were then converted to RESP charges using the antechamber module in Amber-Tools21 [32,33].
The F13L(E353K) and F13L(G277C) mutant structures were generated from the wild-type AlphaFold2 model (ranked_0) using the "Mutagenesis Wizard" plugin in PyMOL v2.5 [29]. To ensure the physiological relevance of side-chain conformations, the Dunbrack 2010 rotamer library was employed to predict the most stable orientation of the mutated amino acid residues. After mutation construction, a local energy minimization step was performed on all residues within a 5 Å radius of the mutation site to eliminate steric clashes. This local optimization used the steepest descent algorithm for 10,000 steps, with a force constant of 50 kcal/mol•Å² applied to non-mutated residues to preserve the global structure of the protein. A 50 ns pre-equilibrium MD simulation was conducted for each mutant to achieve conformational relaxation before molecular docking.
All MD simulations were performed using GRO-MACS 2021.5 under periodic boundary conditions in the NPT ensemble. The protein was described with the AMBER99SB-ILDN force field, while parameters for tecovirimat were assigned using the General AMBER Force Field (GAFF) via the ACPYPE program [34]. The system was solvated in a cubic box with TIP3P water molecules, maintaining a minimum distance of 1.0 nm between the solute and box boundaries. Sodium and chloride ions were added to neutralize the system and achieve a physiological concentration of 0.15 M.
Energy minimization was conducted using the steepest descent algorithm for 50,000 steps, with positional restraints applied to protein and ligand atoms (force constant: 100 kcal/mol•Å²). The simulation time step was set to 2 fs, and bonds involving hydrogen atoms were constrained using the LINCS algorithm [35]. Long-range electrostatic interactions were treated with the Particle Mesh Ewald (PME) method [36], with a cutoff of 1.2 nm for both Coulomb and van der Waals interactions. Temperature was maintained at 310 K using the v-rescale thermostat, and pressure was controlled at 1 bar using the Parrinello-Rahman barostat [37]. Trajectories were saved every 10 ps for subsequent analysis. Simulations were considered equilibrated once the RMSD of the protein-ligand complex reached a stable plateau.
## 3. Binding free energy calculation
The binding free energy (ΔGbind) for each complex was calculated using the MM/GBSA method [38] as implemented in gmx_MMPBSA [39] a tool designed for endstate free energy calculations with GROMACS trajectory files. The ΔGbind was computed as the difference in free energy between the complex and the sum of the isolated protein and ligand, as described by:
This can be decomposed into contributions from molecular mechanical energy in the gas phase (ΔGMM), solvation free energy (ΔGSolv), and an entropic term (TΔS):
The molecular mechanical energy comprises electrostatic (ΔGele) and van der Waals (ΔGvdw) components:
The solvation free energy includes polar (electrostatic) and nonpolar contributions:
$$∆ Gbind = Gcomplex -Gprotein -Gligand$$
$$∆ Gbind = ∆ G MM + ∆ G Solv -T∆ S$$
$$∆ G MM = ∆ G ele gas + ∆ G vdw gas$$
$$∆ G Solv = ∆ G ele Solv + ∆ G nonpolar$$
## Solv
The polar solvation term (ΔGSolvele) was calculated using the Generalized Born (GB) model, while the nonpolar term (ΔGSolvnonpolar) was estimated from the SASA:
where γ = 0.00542 kcal/mol•Å² and b = 0.92 kcal/mol are empirical constants.
$$∆ G nonpolar Solv = γ SASA + b$$
## References
1. Xiang, White (1080) "Monkeypox virus emerges from the shadow of its more infamous cousin: family biology matters" *Emerg Microbes Infect*
2. Huang, Mu (1038) "Monkeypox: epidemiology, pathogenesis, treatment and prevention"
3. Brien, Lebreton, Doty et al. (2016) "Clinical manifestations of an outbreak of Monkeypox virus in captive chimpanzees in Cameroon" *J Infect Dis*
4. Luo, Han (2022) "Preparedness for a monkeypox outbreak" *Infect Med*
5. Ward, Christie, Paton et al. (1136) "Transmission dynamics of monkeypox in the United Kingdom: contact tracing study" *BMJ*
6. Wenham, Eccleston-Turner (2022) "Monkeypox as a PHEIC: implications for global health governance" *Lancet*
7. Adler, Gould, Hine et al. (1016) "Clinical features and management of human monkeypox: a retrospective observational study in the UK" *Lancet Infect Dis*
8. Elsheikh, Makram, Vasanthakumaran et al. (2023) "Monkeypox: a comprehensive review of a multifaceted virus" *Infect Med*
9. Gigante, Korber, Seabolt et al. "Multiple lineages of Monkeypox virus detected in the United States"
10. (2022) *Science*
11. Vernuccio, León, Poojari et al. (0193) "Structural insights into Tecovirimat antiviral activity and poxvirus resistance" *Nat Microbiol*
12. Rout, Mishra, Dey et al. (2023) "Exploiting the potential of natural polyphenols as antivirals against monkeypox envelope protein F13 using machine learning and all-atoms MD simulations" *Comput Biol Med*
13. Alakunle, Moens, Nchinda et al. (2020) "Monkeypox virus in Nigeria: infection biology, epidemiology, and evolution"
14. Delaune, Iseni (2020) "Drug development against smallpox: present and future" *Antimicrob Agents Chemother*
15. Duraffour, Lorenzo, Zöller et al. (2015) "ST-246 is a key antiviral to inhibit the viral F13L phospholipase, one of the essential proteins for orthopoxvirus wrapping" *J Antimicrob Chemother*
16. Delaurentis, Kiser, Zucker (2022) "New perspectives on antimicrobial agents: Tecovirimat for treatment of human Monkeypox virus" *Antimicrob Agents Chemother*
17. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-MODEL: homology modelling of protein structures and complexes" *Nucleic Acids Res*
18. Bowling, Salazar, Bell et al. (0499) "Crystal structure of human PLD1 provides insight into activation by PI(4,5)P2 and RhoA" *Nat Chem Biol*
19. Samantha, Damnjanović, Iwasaki et al. "Structures of an engineered phospholipase D with specificity for secondary alcohol transphosphatidylation: insights into plasticity of substrate binding and activation"
20. (2021) *Biochem J*
21. Wang, Liu, Mao et al. (2021) "Crystal structure of a phospholipase D from the plant-associated bacteria Serratia plymuthica strain AS9 reveals a unique arrangement of catalytic pocket" *IJMS*
22. Selvy, Lavieri, Lindsley et al. (2011) "Phospholipase D: enzymology, functionality, and chemical modulation" *Chem Rev*
23. Bryk, Brewer, Ward et al. (2018) "Vaccinia virus phospholipase protein F13 promotes rapid entry of extracellular virions into cells" *J Virol*
24. Jones, Morgan, Cockcroft (1999) "Phospholipase D and membrane traffic" *Biochim Et Biophys Acta (BBA) -Mol Cell Biology Lipids*
26. Yao, Li, Lin et al. (2021) "Structural insights into phospholipase D function" *Prog Lipid Res*
27. Hossain, Mda, Rahman et al. (2025) "Mpox (monkeypox): a comprehensive updated of current epidemic evidence" *Science in One Health*
28. Otieno, Ruis, Onoja et al. (2025) "Global genomic surveillance of Monkeypox virus" *Nat Med*
29. Husain, Moss (2002) "Similarities in the induction of post-Golgi vesicles by the vaccinia virus F13L protein and phospholipase D" *J Virol*
30. Guo, Perminov, Bekele et al. (1038) "AlphaFold2 models indicate that protein sequence determines both structure and dynamics" *Sci Rep*
31. Jumper, Evans, Pritzel et al. (2021) "Highly accurate protein structure prediction with alphafold" *Nature*
34. Lill, Danielson (2011) "Computer-aided drug design platform using PyMOL" *J Comput Aided Mol Des*
36. Forli, Huey, Pique et al. (2016) "Computational protein-ligand docking and virtual drug screening with the AutoDock suite" *Nat Protoc*
37. Fuhrmann, Rurainski, Lenhof et al. (2010) "A new Lamarckian genetic algorithm for flexible ligand-receptor docking" *J Comput Chem*
38. Frisch (2013) "Gaussian 09"
39. Pu, He, Chen et al. (2022) "Apple polyphenols attenuate the binding ability of angiotensin converting enzyme 2 to viral proteins: computer simulation and in vitro experiments" *Food Biosci*
40. Da Silva, Vranken (0367) "ACPYPE -AnteChamber PYthon Parser inter-facE" *BMC Res Notes*
41. Loubet, Kopec, Khandelia (2014) "Accelerating all-atom MD simulations of lipids using a modified virtual-sites technique" *J Chem Theory Comput*
42. Wang, Gao, Fang (2016) "Multiple staggered mesh ewald: boosting the accuracy of the smooth particle mesh Ewald method" *J Chem Theory Comput*
43. Ke, Gong, Liao et al. (2022) "Effects of thermostats/barostats on physical properties of liquids by molecular dynamics simulations" *J Mol Liq*
44. Gao, Wang, Chen et al. (1080) "Integrating molecular dynamics simulation and molecular mechanics/generalized born surface area calculation into pharmacophore modeling: a case study on the proviral integration site for Moloney murine leukemia virus (Pim)-1 kinase inhibitors" *J Biomol Struct Dyn*
45. Ischak, Aman, Hasan et al. (2023) "In silico screening of andrographis paniculata secondary metabolites as anti-diabetes mellitus through PDE9 inhibition" *Res Pharm Sci* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724363&blobtype=pdf | # Contrasted impacts of commercial diets and rearing water on Aedes aegypti fitness and microbiota
Elodie Calvez, Isaure Quétel, Ludmina Saint-Alban, Gladys Gutiérrez-Bugallo, Christelle Dollin, Cédric Ramdini, Anubis Vega-Rúa
## Abstract
Mosquito rearing optimization in laboratory conditions is crucial for both vector research and control. Although the addition of nutrients is important for Aedes aegypti development from immature stages to adult mosquitoes, little is known about the nutrient composition of commercial diets used for mosquito rearing and their influence on Ae. aegypti life traits. Here, we evaluated the influence of four commercial diets commonly used to rear Ae. aegypti in the laboratory on its fitness, lifespan, and microbiota. We also compared the effect of these diets on this mosquito when com bined with two different rearing waters (laboratory versus field-collected waters). Our investigations demonstrated that higher levels of protein and lipid in commercial diets promote better Ae. aegypti development, lifespan, and size in both water. Metagenomic analysis revealed specific modulations of adult microbiota composition according to both diet and rearing water. Chryseobacterium dominated the microbiota of female mosquitoes reared in laboratory water, except for yeast condition, where a more diverse microbiota was observed. When reared in larval site water, the microbiota diversity was overall higher despite diet addition, except for fish food, which promoted Sphingobacte rium dominance. Given the pivotal influence of diet addition during the larval stage on Ae. aegypti microbiota and life traits, rearing conditions should be carefully chosen according to the goals of the research (i.e., vectorial capacity estimations) or vector control intervention. IMPORTANCE Aedes aegypti is the main vector of arbovirus, such as dengue, yellow fever, and chikungunya viruses. Vector research and control are primarily carried out in laboratories, with larval stage rearing conducted using commercial diet. If many nutrients are essential for Ae. aegypti development, gaining insight into the influence of these diets and their nutrient levels is important to promote optimized rearing worldwide. In this study, our results indicated a significant impact of commercial diet on Ae. aegypti development, lifespan, size, and microbiota related to contrasted protein, lipid, and carbohydrate levels in these diets. This study will help people working with Ae. aegypti raise awareness in staff working with Ae. aegypti to select optimized diets for their specific purpose.
recommended to choose appropriate rearing conditions for this species in the labora tory.
Optimization of mosquito immature larvae rearing is important for (i) entomological surveillance (i.e., identification of adults that were field-collected as larvae), (ii) research purposes, such as dissecting pathogen transmission abilities in mosquitoes or evaluat ing their levels of insecticide resistance, and (iii) the implementation of vector control approaches requiring mosquito mass-rearing (i.e., sterile insect technique, Wolbachiabased population introgression strategy (7). Mosquito rearing is conducted, for the immature stages, in water containing a given microbial composition and supplemented with diet. Essential nutrients for mosquito development and growth are classified in two groups: macronutrients (carbohydrates, amino acids, and polyunsaturated fatty acids) and micronutrients (vitamins, salt, sterols, and metals) (6,(8)(9)(10)(11). The addition of diet and, therefore, of specific nutrients during the larval stage has been shown to influence several mosquito traits, such as larval development, adult longevity, ability to fly, or body size (11,12). Additionally, bacteria present in the larval site water could also influence Ae. aegypti microbiota (13)(14)(15). Moreover, the presence of microorganisms is also an important factor for nutrient assimilation and influences mosquito development (13,(16)(17)(18)(19).
Many options are available for larval feeding in the laboratory. One of the most common is the use of commercial diets, such as yeast or animal feed (i.e. ,TetraMin flakes, rabbit pellets) (9,20,21). Although the influence of nutrient content on vector development has been previously investigated (15,22), little is known about the nutrient composition of the commercial diets and their influence on Ae. aegypti life traits and microbiota (12,21). Here, we assess how commercial diets with different macronutrient content and microbiota impacted Ae. aegypti and provide a holistic overview of Ae. aegypti phenotypic differences associated with diets commonly used in laboratory. To achieve this goal, we characterized the influence of four commercial diets added into waters with different microbial compositions (laboratory and field collected waters) on Ae. aegypti fitness, lifespan, and microbiota, all these being important factors influencing the mosquito vectorial capacity. Beyond the contribution to fundamental knowledge on microbiota-macronutrients-mosquito interaction, we believe the data obtained in this study will raise awareness in staff working with Ae. aegypti to select optimized diets for their specific purpose.
## RESULTS
## Rabbit food, lacking in proteins and lipids, delayed Ae. aegypti development compared to the other commercial diets
The four commercial diets used in the experiments were fish food flakes (FF), yeast, rabbit food pellets (RF), and a mix 1:1 of fish and rabbit food (Mix FF/RF). These diets contain different macronutrient levels (Fig. 1). In our experimental condition (0.2 g of diet in 1 L of water), macronutrient dosage indicated the presence of carbohydrate in all the diets selected (>2.4 mg/L). The presence of protein and lipid was also detected in FF and mix FF/RF (between 0.61 and 3.2 mg/L), whereas lipid appeared absent in yeast (<0.0001 mg/L). In RF, protein and lipid were both undetectable (<0.0001 mg/L). To assess the difference in Ae. aegypti development according to these diets used for rearing the immature stages, we compared the time needed for 50% of larvae to pupate and for 50% of pupae to emerge as adults. For rearing conditions using laboratory water, between 7.36 and 7.94 days were needed to obtain 50% pupation with FF, yeast, and the mix FF/RF diets, while 11.97 days on average were needed for RF, highlighting a significant development delay for this latter diet (P < 0.0001, analysis of variance [ANOVA], Tukey test) (Fig. 2A). Regarding the time required for 50% of emergence rate, homogeneous results were found for FF and yeast (mean of 10.66 days) (Fig. 2B), both being significantly shorter than for the mix FF/RF (mean of 12.03 days; respectively P = 0.0008 and P = 0.0038, ANOVA, Tukey test) and for RF (mean of 14.51 days) that exhibit the longest emergence time of all diets (P < 0.0001 compared to FF and yeast, P = 0.0002 compared to mix FF/RF, ANOVA, Tukey test).
In field-collected water, the time needed for 50% pupation also differed between diets (Fig. 2D). The longest time was recorded for RF, as previously observed with laboratory water (mean = 12.05 days; P < 0.0001, ANOVA, Tukey test). Significant differences were also found between the mix FF/RF (mean = 9.49 days) and the two other diets (mean = 8.37 days for yeast [P = 0.0018] and mean = 7.89 days for FF [P = 0.0004], ANOVA, Tukey test). Time for 50% pupation also significantly differed between yeast and FF (P = 0.0395, ANOVA, Tukey test). Regarding emergence rates (Fig. 2E), the time needed for 50% emergence was similar for yeast and FF (mean of 10.66) but significantly shorter compared to RF (mean of 14.51 days; P < 0.0001, ANOVA, Tukey test) and the mix FF/RF (mean = 12.03 days; P = 0.0008 compared to FF, P = 0.0038 compared to yeast and P = 0.0002 compared to RF, ANOVA, Tukey test).
Interestingly, time for 50% pupation was significantly shorter in laboratory water compared to field-collected water for yeast and mix FF/RF diets. For yeast, 50% pupation was delayed in field-collected water (8.37 days) compared to laboratory water (7.39 days) (P = 0.0019, ANOVA, Tukey test). For mix FF/RF, an increase of the mean time needed for 50% pupation was recorded from 7.85 to 9.49 days for laboratory and field-collected waters, respectively (P < 0.0001, ANOVA, Tukey test). No significant differences were recorded for the other diets or at the emergence stage.
## Higher Ae. aegypti female survival with diets containing a higher amount of protein and lipid independently of water origin
To evaluate the influence of larval diet and the water used for rearing on female Ae. aegypti lifespan, we recorded daily the number of dead adult mosquitoes and estimated survival for each of the eight rearing conditions. In laboratory water, no significant differences were found between the FF diet and the mix FF/RF (Fig. 2C). However, lifespan was significantly shorter with RF and yeast compared to the condition with FF (RF: P = 0.0003 and yeast: P = 0.025, Gehan-Breslow-Wilcoxon test). In field-collected water, lifespan was more homogeneous with a shorter survival only recorded for RF compared to the mix FF/RF (P = 0.0053, Gehan-Breslow-Wilcoxon test) (Fig. 2F). Water origin did not influence Ae. aegypti survival when reared with FF, yeast, and RF. However, for the mix FF/RF, a longer lifespan was observed after rearing in field-collected water compared to laboratory water (P = 0.0175, Gehan-Breslow-Wilcoxon test).
## Fish food and yeast diets led to Ae. aegypti female with longer wing length in both laboratory and field-collected waters
To investigate the influence of both commercial diets and water origin on Ae. aegypti body size, we measured the wing length (from the tip to the distal end of the allula) of females and males. In both Ae. aegypti females and males reared in laboratory water (Fig. 3), wings of mosquitoes fed with FF and yeast (mean length for females 0.28 cm and for males 0.22 cm for both) were significantly longer than those of mosquitoes fed with RF (mean length for females 0.26 cm and for males 0.20 cm) and the mix FF/RF (mean length for females 0.27 cm and for males 0.21 cm) (P < 0.0001, ANOVA, Tukey test). Regarding mosquitoes reared in field-collected water, wings of females fed with RF (mean length 0.26 cm) were significantly smaller compared to those fed on FF (length 0.27 cm; P = 0.0006, ANOVA, Tukey test), yeast (length 0.27 cm; P = 0.0027, ANOVA, Tukey test), and the mix FF/RF (length 0.28 cm; P < 0.0001, ANOVA, Tukey test) (Fig. 3). Interestingly, no significant differences were found between the four diets for males (mean length 0.21 cm; ranging from 0.21 cm for yeast to 0.20 cm for RR).
Significant differences were also found for some conditions between the rearing in laboratory and field-collected waters. Indeed, both females and males fed with FF had significantly longer wing length after rearing in laboratory water (P < 0.0001 and P = 0.0057, respectively, ANOVA, Tukey test). Similar results were also recorded for females fed with yeast, with longer wing length observed for females reared in laboratory water compared to those reared in field-collected water (P < 0.0001, ANOVA, Tukey test).
## Emerging mosquitoes harbor a more diverse microbiota composition when their larvae are reared with yeast in laboratory water
To determine the influence of diet addition on Ae. aegypti microbiota in laboratory water, 16S rRNA gene metabarcoding was separately performed on water, larvae, and newly emerged females reared with the different diets in dechlorinated tap water. In total, 1,860 amplicon sequencing variants (ASVs) were identified from the samples tested (water, larvae, and adult females) belonging to 16 phyla and 210 genera. Whatever the diet used, bacterial communities' richness (alpha diversity) decreased after diet addition (from mean 141.5 before to lower than 76 after diet addition) but remained stable between the water and the larvae (mean richness 74.6, min = 34, max = 133; Fig. 4A). Interestingly, an increase in the richness was observed at the adult stage ranging from 153 (mix FF/RF) to 316.5 (yeast). However, alpha diversity based on Shannon and Simpson indices decreased for all the conditions, except for yeast, between larvae and adult mosquitoes (Fig. 4A).
The variation of microbiota diversity (alpha diversity) in each sample demonstrated a significant difference at the water level between the diets tested (P = 0.0024, Shannon and P = 0.0099, Simpson, ANOVA) (Fig. 4A). Alpha diversity was significantly higher for condition "Before diet addition" compared to FF (P = 0.008, Shannon, ANOVA, Tukey test), yeast (P = 0.0175, Shannon, ANOVA, Tukey test), and mix FF/RF (P = 0.0018, Shannon and P = 0.0072, Simpson, ANOVA, Tukey test). Furthermore, alpha diversity was also significantly higher for RF compared to mix FF/RF (P = 0.0153, Shannon and P = 0.0318, Simpson, ANOVA, Tukey test). Difference between the diet among the microbial composition (beta diversity) was observed (R²= 0.1981, P = 0.01, permutational multivariate analysis of variance [PERMANOVA] statistical test) as in water (R²= 0.7168, P = 0.004, PERMANOVA statistical test) (Fig. 4B).
Interestingly, a high amount of specific ASVs (N = 228) was found before the addition of any diet in the water (Fig. 4C). Microbiota analysis of laboratory water demonstrated a predominance of minor genera with less than 1% of relative abundance (24.70%) (Fig. 5A). After the addition of the diet, the numbers of ASVs were lower (<70.5), and the most abundant ASVs appeared to be specific to each condition (between 17 for mx FF/RF and 38 for RF) (Fig. 4C). Indeed, the addition of FF or yeast enriched microbiota with Sphingobacterium (~40%) was mainly associated with Pseudomonas (21.40%) for FF and Acinetobacter (23.91%) for yeast (Fig. 5A). For the RF condition, Sphingobium (25.36%) and Sphingobacterium (22.62%) were the most abundant genera, whereas in mix FF/RF, there were Acinetobacter (41.97%) and Nubsella (36.59%) (Fig. 5A). The presence of specific ASVs was also recorded in mosquitoes at larval and adult stages (Fig. 4C). At the larval stage, the microbiota composition differed between the four diets used (Fig. 5B). Indeed, larvae rearing with RF led to a high abundance of Microbacte rium (46.83%). Microbacterium (25.27%), Acinetobacter (23.94%), and Salmonella (21.59%) were predominantly found in larvae reared with mix FF/RF (Fig. 5B). Larvae reared with yeast presented a predominance of Salmonella (36.95%), while for the FF condition, Acinetobacter (27.33%), Sphingobacterium (19.73%), and Chryseobacterium (18.20%) were mainly detected. At the adult stage, the microbiota of Ae. aegypti females reared in water supplemented with FF, RF, and mix FF/RF was mainly composed of Chryseobacterium at contrasted proportions (47.87, 88.57, and 95.07%, respectively) (Fig. 5C). The microbiota composition diversity of female previously reared in the presence of yeast was higher compared to the other conditions with a predominance of unidentified bacteria (27.30%) and minor genera (23.74%), followed by 10 other bacteria with similar relative abundan ces (ranging from 1.80 to 11.93%) (Fig. 5C).
## Diets containing lipids promote the presence of specific predominant genera in Ae. aegypti microbiota at the adult stage after rearing in field-collected water
We also investigated how diet addition influenced the microbiota of field-collected water, as well as that of larvae and newly emerged females reared under these condi tions. In field-collected water, a total of 3,679 ASVs was detected among the samples tested (water, larvae, and newly emerged females) belonging to more than 40 phyla and 462 genera. Analysis of the mean richness revealed a dramatic decrease after diet addition in the water from mean 1,017 to <331 (Fig. 6A). The richness then remained stable between water, larvae, and female adults (between 89 and 426) (Fig. 6A).
Significant differences in microbiota diversity were only found in water samples (P < 0.0001, Shannon; P = 0.0087, observed; P = 0.0087, Chao1, and P = 0.0011, Simpson, ANOVA), especially before and after diet addition (P < 0.0001, Shannon; P < 0.0242, observed; P < 0.0243, Chao1 and P < 0.006; Simpson ANOVA, Tukey test) (Fig. 6A). Microbiota composition was significantly different (R²= 0.1994, P = 0.013, PERMANOVA statistical test) as in the water according to the diet status (R²= 0.7846, P = 0.002, PERMANOVA statistical test) (Fig. 6B).
As for laboratory water, most abundant ASVs were mostly specific to each condition in water at the larval and female adult stages (Fig. 6C). Analysis of microbiota composition demonstrated a high proportion of minor genera (73%) in the field-collected water before the addition of a diet (Fig. 7A). After the addition of yeast or mix FF/RF, Sphingo bacterium was detected as the major genera (>35.32%). In FF condition, Sphingobium and Sediminibacterium were the most abundant genera (27.08 and 19.78%, respectively), whereas in RF condition, it was Novosphingobium and Flavobacterium (22.75 and 24.09%, respectively). At the larval stage, the most abundant genera were Salmonella for FF condition (36.34%) and Microbacterium for yeast (36.06%), while genus diversity was high for RF and mix FF/RF conditions (Fig. 7B). After larval rearing in field-collected water, newly emerged female adults presented a high genus diversity in the microbiota composition for the RF and yeast conditions. For FF and mix FF/RF conditions, even if the genera diversity was also high, Sphingobacterium and Chryseobacterium represented 47.07 and 27.20% of relative abundance, respectively (Fig. 7C).
## Differential impact of diet addition after larval rearing in laboratory or fieldcollected water on Ae. aegypti microbiota
Overall, the mean ASV richness was significantly higher in field-collected water com pared to laboratory water both before and after diet addition (P = 0.00589, ANOVA) (Fig. 4A and6A). The same observation was done at the larvae stage, except for FF with a homogenous richness for both water conditions (laboratory mean = 84.5; field-collected mean = 121). At the adult stage, the richness appeared to be dependent on both water and diet combinations. In both waters, ASVs appeared to be specific to the conditions studied in the water and at the larval and adult stages (Fig. 4C and6C). After diet addition, microbiota composition in water was specific to FF, RF, and mix FF/RF condi tions but not for yeast that showed a common predominance of Sphingobacterium whatever the water used (>38.58%) (Fig. 5A and7A). At the larval stage, with diverse proportions, Microbacterium seemed to be an abundant genus in both waters for yeast, RF, and mix FF/RF (abundance between 7.97% for mix FF/RF in field-collected water and 46.83% for RF in laboratory water) (Fig. 5B and7B). At the adult stage (Fig. 5C and7C), the microbiota composition of female Ae. aegypti seemed diverse in both waters for yeast, whereas Chryseobacterium appeared to be predominant in adults after a rearing with mix FF/RF in both laboratory and field-collected water (95.07 and 27.20%, respectively). For RF, adult microbiota composition was mainly composed of Chryseobacterium (88.57%) after rearing in laboratory water and highly diverse after rearing in field-collected water. For FF, Chryseobacterium was the predominant genus in adult microbiota after a rearing in laboratory water (47.87%), whereas it was Sphingobacterium after a rearing in fieldcollected water (47.07%). Overall, these results demonstrated the importance of both water origin and diet in the specific microbiota establishment in Ae. aegypti.
## DISCUSSION
For decades, mosquito rearing has been performed in laboratories using commercial diets for research or vector control strategies (7). If special attention has been given to homemade diets and nutrients for mosquito rearing optimization, little is known about the influence of commercial diets used at the larval stage and their macronutrient content on Ae. aegypti life traits and microbiota. This lack of knowledge could be explained by the diversity of rearing methods used and the complexity of interactions between the environmental and rearing parameters selected (e.g., temperature, diet, larval density) (23). In this study, we investigated under controlled and standardized conditions the impact of four commercial diets using two rearing waters (laboratory tap water or field-collected water) on Ae. aegypti development, size, lifespan, and microbiota.
First, we evaluated the influence of these diets with contrasted concentrations of macronutrient content (lipids, proteins, and carbohydrates) on Ae. aegypti development and survival after rearing them in laboratory water. Furthermore, we evaluated these aspects using field-collected water, as Ae. aegypti larval sites could also influence its development and microbiota (13). Heterogeneous levels of nutrients, physicochemical properties, and a high level of diversity in microbiota composition have also been observed in larval sites in the field (13,14,24,25). In this study, we evaluated whether microbiota differences in the water used for rearing (laboratory vs field-collected waters) could modify the outcome of diet addition on Ae. aegypti life traits. The contrasted nutrient content of our diets allowed us to conclude that when larvae were fed with an RF diet that contains a low amount of protein and lipid concentrations, the development of Ae. aegypti was delayed compared to other diets in both water conditions. However, a lack of lipids can be counterbalanced by carbohydrates as observed with yeast. Previous studies demonstrated that a high carbohydrate amount is required to allow an optimized (quick and homogenous) rearing, especially for the larval growth, when the threshold concentration of lipid and protein is too low (6,9,(26)(27)(28). However, an excess of protein could be harmful for Ae. aegypti development probably due to a toxic production of ammonia during protein digestion (6). Additionally, larvae fed with FF and mix FF/RF demonstrated a longer lifespan compared to larvae fed with other diets, especially in laboratory water, demonstrating the importance of larval diet on Ae. aegypti survival (29,30). The presence of carbohydrates appeared to be an important factor for Ae. aegypti growth, especially when protein concentration is low. Lipids can be produced from carbohydrate resources, which could explain the complete development of the mosqui toes in the RF condition in both waters (6,26). Furthermore, wing analysis revealed that both male and female Ae. aegypti reared with FF and yeast had longer wing length compared to the specimens reared with other diets in laboratory water. In field water, smaller females were obtained with RF compared to other diets, while no impact was observed in male mosquitoes' wing length. Diet seemed to impact more drastically Ae. aegypti female size as previously demonstrated by Van Schoor et al. (6).
The second aim of this study was to evaluate how the larval diet influenced Ae. aegypti microbiota during a standardized rearing procedure. The influence of the larval site on Ae. aegypti microbiota was previously highlighted in field and laboratory water (13,14,(31)(32)(33). Our results demonstrated both in laboratory and field-collected waters a decrease of ASV richness after diet addition. Water microbiota diversity and composi tion were also significantly modified by the diet and influenced Ae. aegypti microbiota composition during its development until adulthood. In our study, we found a predom inance of different bacteria, such as Sphingobacterium, after addition of FF, RF, and yeast in laboratory water, which suggests a diet-associated selection of bacteria. Hence, the genera best adapted to the nutrients added by the diet will more likely become predominant. Protein, carbohydrate, and lipid are essential for bacteria growth and used, for example, as a source of carbon (34). In our study, nutrient quantification was performed on the total lipid, protein, or carbohydrate content. It could be interesting in further experiments to determine if the presence of specific nutrients (e.g., glucose, fructose, or maltose for carbohydrate) enhances the presence of predominant genera by the activation of specific bacterial metabolic pathways (e.g., glycolysis) (35). Interest ingly, this diet-associated impact on microbiota is less observable in adult mosquitoes. Our experiments conducted with laboratory water showed that females harbored a similar microbiota characterized by a Chryseobacterium predominance whatever the diet used, except for yeast. When fed with yeast, females showed a highly diverse microbiota composed mainly of an unidentified bacteria and minor genera. Conversely, when reared in field-collected water, female microbiota composition was highly diverse, except when nourished with FF, where Chryseobacterium predominated, as previously observed in experiments with laboratory water. This genus is commonly found in mosquito microbiota, especially in newly emerged Ae. aegypti, as in our study (17,36). Overall, despite the diet status, microbiota composition was more diverse after larval rearing in field-collected water. Given that microbiota composition and diversity could influence mosquito immunity, fitness, blood digestion, or virus transmission (37)(38)(39)(40)(41)(42), further investigations are needed to evaluate the impact of the nutrients used for immature stages rearing on Ae. aegypti ability to transmit arboviruses. This is particularly important when alternative vector control strategies based on mosquito mass-rearing are increasingly being implemented (43). In view of our findings, it could be interest ing to investigate further Chryseobacterium-arboviruses interactions. Previously, it was found that the abundance of this genus was higher in Ae. aegypti mosquitoes fed with ZIKV-spiked blood compared to the noninfectious blood and sucrose/water groups (44).
Taken together, our data demonstrated how diets commonly used in laboratories impacted Ae. aegypti development, lifespan, and microbiota. Our findings demonstra ted that protein content in commercial diets (i.e., FF, yeast, and mix FF/RF) is pivotal to ensure a quick and homogenous Ae. aegypti rearing (pupation and emergence) and a good mosquito survival. Conversely, when the commercial diet did not contain lipids and proteins (i.e., RF), the development was delayed. Secondly, FF and yeast provided mosquitoes with longer wing length for both females and males. Finally, the different diets influenced the establishment of Ae. aegypti microbiota through important modifications of rearing water microbiota. However, yeast yielded adults harboring a more diverse microbiota composition closer to those that can be found in the field (14). Based on our results (Table 1), for example, to incriminate a local mosquito as a vector for a given pathogen and assess the local epidemic risk, yeast diet can be a good choice because it allows a very good mosquito development and yields adult mosquitoes with a more diverse microbiota like the ones observed in the field. However, if the goal is to conduct a mosquito mass rearing and to have a quick larval development and adult mosquitoes of bigger size (i.e., to increase flight ability and dispersal), diets with at least protein and carbohydrate like yeast, fish food, and mix FF/RF could be more adapted. Finally, adding diet to field-collected water could be interesting in certain contexts given that the adult mosquitoes could develop well, and their microbiota could be diverse as found in the field (17).
All these diet-induced modifications should be carefully considered when design ing research experiments to limit any associated bias or before implementing vector control activities that heavily rely on mosquito mass-rearing (i.e., sterile insect techni que, Wolbachia-based population introgression strategy). For instance, diet influence on mosquito microbiota and size might bias the evaluation of virus transmission risk through vector competence changes (42,45,46). On the contrary, it could also influence Ae. aegypti flight ability or sexual competitiveness (29), phenotypic traits that are essential for alternative vector control strategies that require successful reproduction between released and wild-type mosquitoes (47). We, therefore, highlight the impor tance of optimizing rearing methods according to the expected outcomes to limit the bias introduced by mosquito laboratory rearing.
## MATERIALS AND METHODS
## Mosquitoes
The Ae. aegypti population was collected at the immature stage (larvae and pupae) in artificial larval sites in October 2021 in Petit-Bourg (16°13′04.9″N; 61°36′07.7″W), Guadeloupe. The immature stages (F 0 ) were reared in dechlorinated tap water supple mented with brewer's yeast capsules, and adults were fed ad libitum with 10% sucrose solution. Mosquitoes were reared under controlled conditions at 27 ± 1°C, 12/12 h light/ dark photoperiod, and 70% relative humidity. To obtain the F 1 generation used in this study, females were blood-fed twice a week using the Hemotek system (Hemotek, Ltd., UK).
## Diet description and nutrient quantification
Three diets commonly used in laboratory were selected: fish food (FF; TetraMin, Tetra, Germany), yeast (Gayelord Hauser, France), and rabbit food (RF; GMA, Guadeloupe). Manufacturers indicated that FF contains 46% protein, 11% fat, 3% cellulose, and 6% water, as well as several additives. For 4 g of yeast, the composition displayed by the manufacturer is the following: fat <0.5 g, protein 1.4 g, carbohydrates 0.8 g, 47 kcal of energy value, and several additives. RF contains 16.43% protein, 5.89% fat, 14.94% cellulose, and also several additives. As the heterogeneity of the information provided by the manufacturers may be associated with different nutrient dosage techniques, a standardized nutrient quantification was performed in our study. As larvae filter the nutrients in the water (48), in our study, we estimated the whole macronutrient amount (proteins, lipids, and carbohydrates) using diet-derived solutions. For that, each diet was ground, and a solution at 40g/L was prepared in 1 L of distilled water. The solutions were sent to the AgroQual laboratory that subcontracted the analysis to the Capinov laboratory. The Capinov laboratory is accredited by French authorities for food analysis, including nutrient composition quantification (ISO9001 and Cofrac n°1-6211). Nutrient quantification obtained by these laboratories is reported in Table 2.
## Experimental rearing conditions
Three commercial diets and one mix of two of them (mix 1:1 of FF and RF; mix FF/RF) were selected for this study (Fig. 1). Each diet was ground, and aliquots of 0.2 g were prepared and autoclaved before use in the experiments.
In total, eight rearing conditions were used in this study, including all combinations between two waters and four diets. Laboratory water collected in 2021 consisted of dechlorinated tap water (chlorinated tap water left standing for at least 72 h), while field-collected water was taken from two drums in November 2021 in Lauricisque (16°15′09.3″N; 61°32′53.6″W), Pointe-à-Pitre, Guadeloupe. Once in the laboratory, both field water samples were mixed (1:1) to obtain the field water used in this study. Water was stored at 27 ± 1°C before use.
The four diets were separately used to feed the larvae. Containers were disinfected with 70% ethanol before transferring 1 L of water (either laboratory dechlorinated tap water or field-collected water). F 1 Ae. aegypti eggs were transferred in batches in the containers, and 0.2 g of diet was added to each container. Three days after the hatching, 250 first-instar larvae were counted and placed in a new container containing 1 L of water and 0.2 g of the same diet. Water and diet were renewed every 3 days.
For each condition (four diets and two rearing waters per diet), two biological replicates were performed. To reduce experiment bias and optimize the standardization of environmental conditions, rearing with the four diets was conducted in parallel for each water type. Larvae were reared under controlled conditions at 27 ± 1°C, 12/12 h light/dark photoperiod. Daily, the number of pupae and adults was counted until the end of the pupation or the end of the emergence. The pupation rate was estimated as the proportion of pupae counted per day among the total number of larvae initially placed in the container, while the emergence rate referred to the proportion of adults counted daily among the number of pupae previously counted.
Pupation and emergence curves were generated using nonlinear regression. An ordinary one-way ANOVA, followed by a Tukey's test, was used to compare 50% pupation and 50% emergence values for each diet using both replicate data.
## Survival assays
Female survival was estimated for each of the eight diet/water combinations using between 21 and 30 females randomly collected from the rearing replicates. The emerging females were maintained under controlled conditions at 27 ± 1°C, 12/12 h light/dark photoperiod, and 70% relative humidity and fed ad libitum with 10% sucrose solution previously autoclaved. Deaths were recorded daily for 90 days. Ae. aegypti survival curves were generated by the Kaplan-Meier method and analyzed using a log-rank test and a Gehan-Breslow-Wilcoxon test.
## Wing measurement
To assess the influence of water/diet combinations on mosquito body size, wing measurements were conducted on mosquitoes. After adult emergence, 40 females and 20 males were randomly collected at equal proportions from both replicates for each diet/water combination. Wing dissection was performed using a scalpel and a binocular magnifying glass. Wing length considered as a proxy to evaluate mosquito size (49) was measured from the tip to the distal end of the allula as previously described by Dickson et al. (13) using ImageJ. To evaluate the influence of diet on Ae. aegypti wing length, an ordinary one-way ANOVA, followed by a Tukey's test, was performed after verification of data normal distribution by Shapiro-Wilk test.
## Bacterial DNA extraction from water and mosquito samples
To identify the bacterial microbiota associated with the eight rearing conditions (four diets/two waters), 2 mL of water was collected before diet addition (N = 2) and from each container (N = 16) at the fourth instar larvae stage. Water was then centrifuged at 8,000 rpm for 10 min at 4°C. Supernatant was removed, and pellets were kept at -20°C before analysis. For each replicate, 10 fourth instar larvae and 10 females not previously fed with sugar or blood were randomly collected, separately pooled per sample type, and stored at -20°C, as for the rearing water. Before the DNA extraction, mosquitoes (larvae and adults) were surface-sterilized as previously described by Hery et al. (14). Briefly, whole mosquitoes were rinsed three times using 2 mL of sterile water, one time using 2 mL of 70% ethanol for 10 min, five times using sterile water, and finally one time using 0.8% NaCl. Then, a mechanical lysis was performed in 400 µL of sterile phosphate-buffered saline (PBS) using a bead beater (MM 400, Retsch, France) at 30 Hz for 30 s (three times) and a centrifugation at 12,000 pm for 3 min. Total genomic DNA was extracted from the water and the mosquitoes using the RNeasy PowerMicrobiome Kit following the manufacturer's instructions (Qiagen, Germany). Extraction checking was performed using 16S rDNA universal bacteria primers 27F (5′ AGA GTT TGA TCC TGG 3′), 1492R (5′ GGT TAC CTT GTT ACG 3′) (50) and DreamTaq DNA polymerase (Thermo Scientific, USA) according to the manufacturer's instructions. Amplicons were visualized under ultraviolet light by 1.5% agarose gel electrophoresis stained with gel red (Biotium, USA).
## Bacterial microbiota analysis
The 16S rRNA paired-end sequencing (300 bp read length) was performed using the universal prokaryote-specific primers targeting the V3-V4 region: 341F (5′CCT ACG GGN GGC WGC AG3′) and 785R (5′GAC TAC HVG GGT ATC TAA TCC3′) (51) with Illumina MiSeq at the Biomics Platform, C2RT (Institut Pasteur, Paris, France). After sequencing, the R package DADA2 was used for cleaning raw sequences obtained (fastq) (52)(53)(54). This step includes filtering, merging, clustering, chimera and singleton deletion, as well as taxonomic ASV assignation. The taxonomic affiliations were deleted if the bootstrap value was lower than 50%. The Silva database was used for the taxonomy assignment (55). After the standardization, a total of 1,860 ASVs were obtained, and 122,343.5 reads per sample were retained for laboratory water (N = 26). For field water, 3,679 ASVs were obtained, and 88,158 reads per sample were maintained (N = 26). Metrics as the alpha (Chao1, Shannon, and Simpson) and the beta (Bray-Curtis dissimilarity matrix; PCOA representation) diversity metrics were generated using R packages ggplot2 (56), phyloseq (57), and vegan (58). For alpha and beta metrics, statistical analysis was performed using ANOVA, a post-hoc Tukey test, and PERMANOVA.
## Statistical analysis
The level of statistical significance for all analyses on Ae. aegypti life traits was set at P ≤ 0.05. Graphs and statistical analyses were carried out in GraphPad Prism v9 (GraphPad Software Inc.) and R version 4.3.0 (R Core Team, Vienna, Austria).
## References
1. Kraemer, Sinka, Duda et al. (2015) "The global distribution of the arbovirus vectors Aedes aegypti and Ae. albopictus. eLife 4:e08347"
2. Souza-Neto, Powell, Bonizzoni (2019) "Aedes aegypti vector competence studies: a review" *Infect Genet Evol*
3. Viglietta, Bellone, Blisnick et al. (2021) "Vector specificity of Arbovirus transmission" *Front Microbiol*
4. Couret, Dotson, Benedict (2014) "Temperature, larval diet, and density effects on development rate and survival of Aedes aegypti (Diptera: Culicidae)" *PLoS One*
5. Souza, Virginio, Riback et al. (2019) "Microorganism-based larval diets affect mosquito development, size and nutritional reserves in the yellow fever mosquito Aedes aegypti (Diptera: Culicidae)" *Front Physiol*
6. Van Schoor, Kelly, Tam et al. (2020) "Impacts of dietary nutritional composition on larval development and adult body composition in the yellow fever mosquito (Aedes aegypti)" *Insects*
7. Benedict, Knols, Bossin et al. (2009) "Colonisation and mass rearing: learning from others" *Malar J*
8. Singh, Brown (1957) "Nutritional requirements of Aedes aegypti L" *J Insect Physiol*
9. Linenberg, Christophides, Gendrin (2016) "Larval diet affects mosquito development and permissiveness to Plasmodium infection" *Sci Rep*
10. Rivera-Pérez, Clifton, Noriega (2017) "How micronutrients influence the physiology of mosquitoes" *Curr Opin Insect Sci*
11. Ratnayake, Chotiwan, Saavedra-Rodriguez et al. (2023) "The buzz in the field: the interaction between viruses, mosquitoes, and metabolism" *Front Cell Infect Microbiol*
12. Carvajal-Lago, Ruiz-López, Figuerola et al. (2021) "Implications of diet on mosquito life history traits and pathogen transmission" *Environ Res*
13. Dickson, Jiolle, Minard et al. "Lambrechts L. 2017. Carryover effects of larval exposure to different environmental bacteria drive adult trait variation in a mosquito vector" *Sci Adv*
15. Hery, Guidez, Durand et al. (2021) "Natural Variation in physicochemical profiles and bacterial communities associated with Aedes aegypti breeding sites and larvae on Guadeloupe and French Guiana" *Microb Ecol*
16. Macleod, Dimopoulos, Short (2021) "Larval diet abundance influences size and composition of the midgut microbiota of Aedes aegypti mosquitoes" *Front Microbiol*
17. Gaio A De, Gusmão, Santos et al. (2011) "Contribution of midgut bacteria to blood digestion and egg production in Aedes aegypti (Diptera: Culicidae) (L.)" *Parasit Vectors*
18. Coon, Vogel, Brown et al. (2014) "Mosquitoes rely on their gut microbiota for development" *Mol Ecol*
19. Coon, Valzania, Mckinney et al. (2017) "Bacteria-mediated hypoxia functions as a signal for mosquito development" *Proc Natl Acad Sci*
20. Steven, Hyde, Lareau et al. (2021) "The axenic and gnotobiotic mosquito: emerging models for microbiome host interactions" *Front Microbiol*
21. Souza, Diaz-Albiter, Dillon et al. (2016) "Digestion of yeasts and beta-1,3-glucanases in mosquito larvae: physiological and biochemical considerations" *PLoS One*
22. Benedict, Hunt, Vella et al. (2020) "Pragmatic selection of larval mosquito diets for insectary rearing of Anopheles gambiae and Aedes aegypti" *PLoS One*
23. Martinson, Strand (2021) "Diet-microbiota interactions alter mosquito development" *Front Microbiol*
24. Couret, Benedict (2014) "A meta-analysis of the factors influencing development rate variation in Aedes aegypti (Diptera: Culicidae)" *BMC Ecol*
25. Ngugi, Mutuku, Ndenga et al. (2017) "Characteri zation and productivity profiles of Aedes aegypti (L.) breeding habitats across rural and urban landscapes in western and coastal Kenya" *Parasit Vectors*
26. Medeiros-Sousa, De Oliveira-Christe, Camargo et al. (2020) "Influence of water's physical and chemical parameters on mosquito (Diptera: Culicidae) assemblages in larval habitats in urban parks of São Paulo" *Acta Trop*
27. Briegel (1990) "Metabolic relationship between female body size, reserves, and fecundity of Aedes aegypti" *J Insect Physiol*
28. Bond, Ramírez-Osorio, Marina et al. (2017) "Efficiency of two larval diets for mass-rearing of the mosquito Aedes aegypti" *PLoS One*
29. Pooraiiouby, Sharma, Beard et al. (2018) "Nutritional quality during development alters insulin-like peptides' expression and physiology of the adult yellow fever mosquito" *Aedes aegypti. Insects*
30. Lang, Idugboe, Mcmanus et al. (2018) "The effect of larval diet on adult survival, swarming activity and copulation success in male Aedes aegypti (Diptera: Culicidae)" *J Med Entomol*
31. Yan, Kibech, Stone (2021) "Differential effects of larval and adult nutrition on female survival, fecundity, and size of the yellow fever mosquito" *Aedes aegypti. Front Zool*
32. Dickson, Ghozlane, Volant et al. (2018) "Diverse laboratory colonies of Aedes aegypti harbor the same adult midgut bacterial microbiome" *Parasites Vectors*
33. Zouache, Martin, Rahola et al. (2022) "Larval habitat determines the bacterial and fungal microbiota of the mosquito vector Aedes aegypti" *FEMS Microbiol Ecol*
34. Rodpai, Boonroumkaew, Sadaow et al. (2023) "Microbiome composition and microbial community structure in mosquito vectors Aedes aegypti and Aedes albopictus in Northeastern Thailand, a dengueendemic area" *Insects*
35. Bonnet, Lagier, Raoult et al. (2020) "Bacterial culture through selective and non-selective conditions: the evolution of culture media in clinical microbiology" *New Microbes New Infect*
36. Watkins, Maiden, Gupta (2016) "Metabolic competition as a driver of bacterial population structure" *Future Microbiol*
37. Muturi, Dunlap, Ramirez et al. (2019) "Host bloodmeal source has a strong impact on gut microbiota of Aedes aegypti" *FEMS Microbiol Ecol*
38. Hegde, Rasgon, Hughes (2015) "The microbiome modulates arbovirus transmission in mosquitoes" *Curr Opin Virol*
39. Scolari, Casiraghi, Bonizzoni (2019) "Aedes spp. and their microbiota: a review" *Front Microbiol*
40. Cansado-Utrilla, Zhao, Mccall et al. (2021) "The microbiome and mosquito vectorial capacity: rich potential for discovery and translation" *Microbiome*
41. Gómez, Martinez, Muñoz et al. (2022) "Aedes aegypti and Ae. albopictus microbiome/virome: new strategies for controlling arboviral transmission?" *Parasit Vectors*
42. Gabrieli, Caccia, Varotto-Boccazzi et al. (2021) "Mosquito trilogy: microbiota, immunity and pathogens, and their implications for the control of disease transmis sion" *Front Microbiol*
43. Ferreira, Lemos, Moura et al. (2023) "Role of the microbiome in Aedes spp. vector competence: what do we know?" *Viruses*
44. Achee, Grieco, Vatandoost et al. (2019) "Alternative strategies for mosquito-borne arbovirus control"
45. Shi, Beller, Wang et al. (2022) "Bidirectional Interactions between Arboviruses and the bacterial and viral microbiota in Aedes aegypti and Culex quinquefasciatus" *mBio*
46. Alto, Reiskind, Lounibos (2008) "Size alters susceptibility of vectors to dengue virus infection and dissemination" *Am J Trop Med Hyg*
47. Schneider, Chadee, Mori et al. (2011) "Heritability and adaptive phenotypic plasticity of adult body size in the mosquito Aedes aegypti with implications for dengue vector competence" *Infect Genet Evol*
48. Oliva, Benedict, Collins et al. (2021) "Sterile insect technique (SIT) against Aedes species mosquitoes: a roadmap and good practice framework for designing, implementing and evaluating pilot field trials" *Insects*
49. Rashed, Mulla (1990) "Comparative functional morphology of the mouth brushes of mosquito larvae (Diptera: Culicidae)" *J Med Entomol*
50. Nasci (1990) "Relationship of wing length to adult dry weight in several mosquito species (Diptera: Culicidae)" *J Med Entomol*
51. Feldman, Black, Cary et al. (1997) "Molecular phylogenetics of bacterial endosymbionts and their vestimentiferan hosts" *Mol Mar Biol Biotechnol*
52. Beckers, De Beeck, Thijs et al. (2016) "Performance of 16s rDNA primer pairs in the study of rhizosphere and endosphere bacterial microbiomes in metabarcod ing studies" *Front Microbiol*
53. Ihaka, Gentleman (1996) "R: a language for data analysis and graphics" *J Comput Graph Stat*
54. Callahan, Mcmurdie, Rosen et al. (2016) "DADA2: high-resolution sample inference from Illumina amplicon data" *Nat Methods*
55. (2016) "R core team. R: a language and environment for statistical computing"
56. Quast, Pruesse, Yilmaz et al. (2013) "The SILVA ribosomal RNA gene database project: improved data processing and web-based tools" *Nucleic Acids Res*
57. Wickham (2016) "Ggplot2: Elegant Graphics for Data analysisSpringer-Verlag New York"
58. Mcmurdie, Holmes (2013) "Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data" *PLoS One*
59. Oksanen, Simpson, Blanchet et al. (2022) |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724254&blobtype=pdf | # The functional antibody landscape in HIV post-treatment controllers is heterogeneous
Nicholas Webb, Matthew Gorman, Lily Parker, Wonyeong Jung, Dansu Yuan, Jonathan Li, Galit Alter, Boris Julg
## Abstract
HIV-1 establishes a latent reservoir early in infection that persists during antiretroviral therapy (ART), typically leading to rapid viral rebound upon treatment interruption in most people living with HIV-1. However, a rare group known as posttreatment controllers (PTCs) can maintain low or undetectable viral loads for extended periods after ART cessation, offering critical insights into durable HIV remission. While immune responses are thought to play a key role in this control, the contribution of humoral immunity-particularly Fc-mediated antibody functions-remains insufficiently defined. In this study, we explored the influence of humoral responses on viral rebound in PTCs using a comprehensive systems serology approach. We profiled Env-spe cific antibody characteristics and Fc-effector functions-including antibody-dependent cellular phagocytosis, antibody-dependent neutrophil phagocytosis, antibody-depend ent complement deposition, and antibody-dependent NK cell activation-in PTCs and matched non-controllers (NCs). Baseline antibody features prior to ART interruption did not clearly distinguish PTCs from NCs, and antibody responses evolved dynamically and in a highly individualized manner during viral rebound. Notably, distinct and individualspecific shifts in Fc-functional antibody signatures occurred during rebound, indicating that the humoral immune system actively adapts during this critical period. These findings highlight the complexity and adaptability of antibody-mediated responses, shaped by dynamic processes during rebound rather than static baseline features. This work underscores the value of longitudinal immune profiling to delineate how humoral immunity evolves in relation to, but not necessarily driving, post-treatment control. IMPORTANCE Most people living with HIV experience a quick return of the virus after stopping antiretroviral therapy (ART). However, a small group-post-treatment controllers (PTCs)-can keep the virus suppressed without ongoing treatment. This study examined whether antibody responses, especially those involving Fc-mediated functions, play a role in this control. This study compared antibody features and immune functions in PTCs and matched individuals who did not control the virus after stopping ART. They found that antibody profiles before treatment interruption did not predict who would control the virus. Instead, antibody responses changed over time in unique ways during viral rebound. These shifts suggest that the immune system adapts dynamically rather than relying on fixed traits. The findings highlight the importance of tracking immune changes over time to better understand how long-term HIV control may occur, even if antibodies alone may not be the driving force behind post-treatment remission.KEYWORDS system serology, antibody responses, HIV-1, post-treatment controllers D uring the early stages of HIV infection, a latent viral reservoir is established within CD4+ T cells, which persists despite years of highly effective antiretroviral therapy (ART) (1, 2). Upon interruption of ART, this reservoir can be reactivated to produce virus, leading to viral rebound. However, rare instances of natural control over HIV infection
occur in the absence of treatment. Notably, elite controllers (ECs) maintain undetectable viremia without ever receiving ART (3), while post-treatment controllers (PTCs) are able to control viral rebound to low levels after discontinuing ART (4,5). Defining the mechanisms that underlie these natural forms of viral control can provide valuable insights into the development of novel strategies aimed at achieving a functional cure.
The mechanisms underlying viral control in ECs appear to involve the targeted elimination of infected cells. For instance, certain ECs exhibit robust T-cell responses and rare human leukocyte antigen (HLA) alleles that facilitate cytotoxic T lymphocyte (CTL) responses against conserved HIV peptides presented via major histocompatibility complex (MHC) class I molecules (3). Additionally, polyfunctional antibodies capable of mediating antibody-dependent cellular cytotoxicity against infected cells have been identified (6). Although the precise mechanisms enabling post-treatment control remain poorly understood, emerging evidence suggests that viral control in PTCs may also involve targeted cell elimination through strong HIV-specific CD4+ T-helper responses and natural killer (NK) cell activity (7), as well as a limited size of the viral reservoir both before and during rebound (5,8). However, since PTCs exhibit considerably weaker HIV-specific CTL responses compared to ECs (5), it is plausible that antibody-mediated mechanisms may play a significant role in viral control.
Antibodies mediate a variety of immunological functions through their Fc region, including recruitment and activation of innate effector cells by engagement of cellsurface Fc-gamma receptors (FcyRs) directly or by binding complement and then engaging complement receptors. These functions play a critical role in antiviral immunity to influenza (9), respiratory syncytial virus (10), SARS-CoV-2 (11), Ebola (12,13), and others. Recent humoral profiling studies have found that both FcyR3a-mediated NK activity and specific Fc-glycan profiles are associated with delayed rebound after treatment interruption (14). It's possible that in the absence of a strong CTL response, PTC may be driven by similar antibody-dependent cell-targeting mechanisms.
Additional factors contributing to post-treatment control (PTC) include the size and diversity of the latent viral reservoir, as PTC is more commonly observed in individuals who received treatment during acute primary infection, a period when the reservoir is small and less diverse. However, PTC has also been documented in individuals treated later, during the chronic phase of infection, when a diverse latent reservoir has been established and the humoral response begins to develop broader specificity. Individuals treated during chronic infection also represent the majority of people living with HIV in the world. Because chronically treated individuals possess diverse humoral and viral repertoires, the mechanisms underlying post-treatment control for these individuals potentially differ substantially from individuals treated early during primary infection, as a greater reliance on the breadth and function of the humoral immune response may be more important for chronically treated individuals.
We sought to understand the humoral repertoires of chronically treated PTCs and study-matched non-controllers (NCs) identified in the ontrol of HIV after antiretroviral medication pause (CHAMP) study, which includes longitudinal PTC and NC serum samples from analytical treatment interruption (ATI) studies (15). Using a panel of diverse HIV-1 Envs, we profiled antibody subclass, isotype, and FcyR binding. We also quantified Env-specific antibody effector functions such as antibody-dependent complement deposition (ADCD), antibody-dependent cellular phagocytosis (ADCP), antibody-dependent neutrophil phagocytosis (ADNP), and antibody-dependent NK cell activation (ADNKA). Our results reveal highly variable humoral profiles in both PTCs and NCs prior to ATI. During rebound, these profiles follow individual trajectories that may reflect differences in viral load kinetics.
## RESULTS
## Virological and immunological characteristics
To understand the impact of humoral immunity in chronically treated individuals both before ATI and as viral load evolves after ATI, we focused on individuals with longitudinal samples up to 50 weeks post interruption (10 PTCs, 7 NCs, Table 1). All individuals had undetectable viral loads at baseline (≤2 weeks prior to ATI) except for two PTCs whose viral loads were detectable below 1,000 cp/mL at ATI (321568 and 621422, viral load = 127 cp/mL and 793 cp/mL, respectively, Table 1).
Both PTCs and NCs had similar CD4+ T cell counts at the time of antiretroviral treatment interruption (ATI) (Fig. S1a) suggesting comparable immune states. About half of the NCs (3 of 7) showed the highest levels of cell-associated DNA (Fig. 1a). However, overall levels of cell-associated DNA or RNA did not differ significantly, as several NCs had low or undetectable amounts.
After ATI, all NCs experienced a rapid viral rebound during the first 10-12 weeks, followed by a stable viral set point (Fig. 1b). Rebound viral load profiles among PTCs varied: some individuals reached high viral loads (>10,000 cp/mL) within the first 10 weeks after interruption, while others had much lower levels. Rebound kinetics also reflected this variability, as nearly half of PTCs (4 of 10) maintained <200 cp/mL for longer periods post-ATI, while the rest did not (Fig. S1b). Viral setpoints in PTCs, however, were 100-fold lower on average than the ones in NCs (P = 8×10 -8 , t-test, Fig. 1c), consistent with viremic post-treatment control.
## Heterogeneous pre-ATI HIV-specific antibody profiles
To map humoral profiles, we measured IgM, IgA, and IgG subclass binding titers to 13 monomeric HIV gp120 Envs, 3 trimeric gp140 Envs, JRFL SOSIP, gp41, and p24 (Table S1). Positivity in antigen-specific binding titers was defined using non-specific binding to albumin as a threshold plus two standard deviations (see Methods). Prior to ATI (baseline), all individuals were positive for total IgG specific to p24 and gp41 (Fig. S2a), while HIV-specific IgA and IgG4 binding titers were not detectable above threshold for any individual (data not shown). Baseline IgG1 and IgG3 binding titers were not significantly different between PTC and NC (Fig. 2a; Fig. S2b andc). However, we noted a broad heterogeneity in binding titers for PTCs and NCs, where both IgG subtypes varied by up to 50-fold.
To understand whether baseline Fc-mediated effector functions might have an impact on rebound control, we measured HIV-specific antibody binding to FcyRs 2a, 2b, 3a and 3b (Fig. 2b,left and Fig. S3 andb). We found no significant differences in FcyR binding between PTCs and NCs. FcyR binding generally reflected IgG1 breadth as opposed to IgG3, as most individuals had detectable IgG1 binding titers to Envs p1012, ZM53M, and 01 CM but not IgG3. FcyR binding titers were also strongly correlated across FcyRs (Pearson R ≥ 0.93 across all antigens).
To determine whether there were direct functional differences between PTC and NC baseline repertoires, we measured (ADCD, Fig. 2b, right and Fig. S3c), neutrophil and monocyte phagocytosis (ADNP and ADCP, respectively, Fig. 2c), and (ADNKA, Fig. 2d). As with IgG1 and IgG3 titers, HIV-specific ADCD was also heterogeneous for both PTCs and NCs with no significant differences. Similarly, ADCP or ADNP specific to trimeric JRFL gp140 Env (Fig. 2c) did not differ, as both PTCs and NCs varied broadly in phagoscores. HIV-specific ADNKA was measured by degranulation (CD107a), as well as chemokine and cytokine expression (MIP-1b and IFNy) against four HIV Envs (Fig. 2d). We found a high degree of variability among both PTCs and NCs with no significant differences in ADNKA overall.
## Humoral responses during rebound
We next examined the interplay between viral rebound and humoral responses after ATI. With the samples available, we defined three time periods for which we could compare both virological and humoral data: baseline (B) is prior to ATI, mid-rebound (M) spans weeks 7-18 of ATI, and late rebound (L) includes weeks 20-28 (Fig. 3a B,M, and L labels). To understand how humoral responses are associated with these rebound dynamics, we measured immunoglobulin titers, Fc-receptor binding, and effector function from longitudinal serum samples for each individual through mid and late rebound. Similar to baseline, we found no significant difference in either antibody or FcyR binding titers (Fig. S4 and S5, respectively) or functions (Fig. S6) at mid or late rebound.
We noticed that PTCs experienced a variety of different rebound viral load profiles, suggesting highly individualized viral dynamics. To better describe and interpret humoral responses in light of these diverse rebound viral load kinetics, we classified all individuals based on their post-interruption viral load profiles (groups A-E, Fig. 3a). Group A was composed exclusively of NCs, who all experienced rapid increases in viral load followed by a stable set point. Groups B-E show the different viral load profiles that we observed among PTCs for the first 50 weeks of ATI. Like NCs, group B (n = 3) also Humoral time courses specific to HIV Envs JRFL (gp140) and 6535 (gp120) are shown in Fig. 3b andc, while changes in humoral time courses from baseline to early rebound and early to late rebound for all antigens are shown in Fig. S7 through S9.
Prior to ATI, highly variable IgG1 and IgG3 titers were observed for NCs. This trend continued through early and late rebound (Fig. 3b andc, group A), where some individuals experienced 10-to 100-fold increases or decreases in IgG1 and IgG3 titers specific to JRFL and 6535 Envs. Changes in FcyR binding titers were highly correlated through rebound (data not shown) and tended to reflect the wider breadth of IgG1 as opposed to IgG3 (compare Fig. S4b andc to Fig. S5), similar to baseline Fc-receptor binding (Fig. 2b). Functional antibody trajectories were variable among NCs, particularly ADCD, where we observed both transient changes specific to early rebound and steady increases through the early and late rebound time points against Env 6535, depending on the individual (Fig. 3c). ADNKA (measured by expression of CD107a) and ADNP, however, did appear to follow a broader trend of decreased NK and increased NP over time. For example, NCs with ≥10% CD107a+ NK cells at baseline experienced a decrease in ADNKA by late rebound (Fig. 3b andc). Conversely, individuals with low ADNP phagoscores (≤50) tended to develop higher ADNP phagoscores by late rebound (Fig. 2B). These trends, however, were not statistically significant.
PTC rebound group B was defined by a rapid increase in viral load during the early post-ATI time point (similar to NCs) followed by a rapid decrease and viremic control by the late post-ATI time points. Similar to NCs, there were no clear rebound-associated trends in IgG1 or IgG3 titers or FcyR binding. For example, one individual in group B experienced a ~100-fold decrease in 6535-specific IgG1 levels at early rebound, while another in group B experienced a nearly 100-fold increase at late rebound (Fig. 3c).
Group C also consisted of largely heterogeneous humoral trajectories. We did observe a possible trend of decreased ADCD for individuals in both groups B and C and, for some individuals, increased ADNKA specific to 6535 (Fig. 3c). Likewise, individuals from both groups tended to experience a loss of JRFL-specific ADNP from baseline to late rebound.
PTC groups D and E shared an initial period of unstable viral load through early post-ATI and then differed in whether this trend continued (group D) or viral load was controlled during late post-ATI time points (group E). Unlike groups A, B, and C, individuals in groups D and E had very stable IgG1/G3 titers, FcyR binding, and ADCD over time. Interestingly, although the individuals had detectable viral load at baseline and greater fluctuations in viral load during rebound, the humoral profiles of group D appeared to be more stable over time than group E (Fig. S7 through S9).
Finally, we examined correlations between humoral functions and binding titers across baseline, mid, and late rebound. We found significant correlations between p24-specific ADCD and total p24-IgG binding titers across all three time points in PTCs but not in NCs (Fig. 3d, Spearman r ≥ 0.915, P < 0.005). In PTCs, this was accompanied by correlations between p24-specific FcyR binding titers and ADCD (Spearman 0.3 < r< 0.5).
## DISCUSSION
Here, we aimed to determine whether humoral responses exhibited enhanced Fc-medi ated effector functions in PTCs compared to NCs prior to or after ATI. Previous stud ies have suggested that antibody-dependent functions, but not necessarily titers, are associated with elite control of HIV (6), with PTC in individuals treated during acute infection (16), and with delayed rebound among NCs during ATI (14,17). To investi gate this, we employed a panel of diverse Env variants to capture humoral responses to the diversifying viral reservoir during rebound viremia. Our results reveal substan tial heterogeneity in antibody titers, Fc-receptor binding, and Fc-mediated functions, with variability across individuals being more pronounced than differences based on controller status. While all individuals exhibited detectable p24 and gp41-specific IgG, Env-specific humoral signatures varied dramatically both at baseline and during rebound, with variation observed in IgG subtypes. This heterogeneity likely reflects the unique dynamics of rebound viremia occurring in each individual and the diversity of latent viral isolates that are reactivated in the absence of treatment.
We observed that baseline HIV-specific antibody binding titers, FcyR binding, and Fc-mediated effector functions were not associated with post-treatment control (PTC) status, indicating that the humoral functional profiles at the time of treatment inter ruption do not predict control following ART cessation, at least within this cohort. Previous studies, including our own, have shown that Fc-mediated effector functions, particularly NK cell activity, prior to ART interruption can be linked to delayed viral rebound during ATI (17). However, in that study, antibodies with distinct functional properties also mirrored the transcriptional activity of the viral reservoir, which itself can influence viral rebound dynamics. In contrast, in the present study, PTCs and NCs, although heterogeneous in their baseline reservoir measures, exhibited similar cellular HIV RNA levels. Furthermore, although non-neutralizing antibody functions pre-ART interruption were not enriched or predictive of PTC, it is likely that an individual's baseline humoral repertoire influences viral rebound by selecting for the replication of viral variants resistant to immune control. In fact, autologous neutralizing antibodies (aNAbs) have been reported to contribute to the control of susceptible reservoir viruses (18), and a stronger aNAb response was reported as a pre-ATI feature distinguishing PTC from NC (19).
Consistent with studies of acutely treated individuals (16,19), we find that the humoral profiles of chronically treated PTCs and NCs evolve in response to viral replication during rebound. Because viral exposure can drive humoral evolution ( 16), we stratified the PTCs in our cohort into groups that reflect similar post-ATI viral load profiles in terms of kinetics and stability. These groups were not thought to demonstrate clear biological differences but rather allow to interpret the antibody data based on viral kinetic patterns. Indeed, we found that individuals in PTC groups D and E were characterized by persistent but low-level viremia that was either unstable (group D) or transient (group E), and these individuals had some of the most stable post-ATI humoral profiles in our study, despite the fact that the individuals in group D had detectable viral load at baseline. Conversely, NCs and PTC group B, whose initial viral loads were similarly high had dynamically changing humoral profiles during rebound. PTC group C, however, experienced low-level viremia (similar to group D) but experienced changing humoral profiles like groups A and B. The viral load trajectories of individuals in group C steadily increased, while those of group D were unstable and fluctuated during mid and late rebound. It's possible that low-level viremia can drive humoral adaptation when exposure is steady and consistent. Future studies are needed to identify humoral signatures associated with viral control by investigating time-dependent correlations between viral load dynamics and changes in humoral features.
The diverse viral load dynamics observed in this small cohort of 10 PTCs suggest that viral control can manifest at varying levels and times during rebound. Although we detected changes in individual humoral profiles during both early and late rebound, these changes did not consistently correlate with viral load dynamics, highlighting the complexity of the relationship between humoral immune responses and viral control.
Interestingly, p24-specific IgG associated with ADCD was highly correlated in PTCs but not NCs. As p24 is an internal capsid protein, these antibodies are unlikely to mediate direct antiviral activity. Their presence may instead reflect immune recognition of viral debris following the lysis of virus-producing cells via other mechanisms (i.e., by cytotoxic T cells or NK cells). Such responses could signify effective immune clearance or low-level antigen exposure. Indeed, previous studies have associated p24-specific IgG1 and ADCP with viral control (20). Future mechanistic studies are needed to define whether these antibodies actively participate in viral control or simply mirror a more contained infection state.
Our study is limited by the small sample size and the variability in sampling time points across individuals from different studies. The sample size was further reduced when individuals were stratified based on their rebound viral load profiles. Additionally, we were unable to profile humoral responses to contemporaneous rebound isolates due to the lack of available viral sequences, which limited our characterization of Env-specific humoral characteristics and breadth.
Overall, our findings demonstrate that both the biophysical and functional humoral profiles of chronically treated PTCs and NCs exhibit considerable diversity prior to treatment interruption, and that the changes occurring during rebound are similarly variable across NCs and certain subgroups of PTCs. Since Env is the most accessible HIV antigen, a more comprehensive understanding of the humoral mechanisms of control will likely necessitate concurrent, detailed characterization of the rebound isolates, including sequencing of integrated Env variants across various tissues and their blood-borne descendants. A more refined approach to categorizing different types of PTCs, based on viral load profiles, could help elucidate the relationships between viral load kinetics, control mechanisms, and the adaptive interplay between the virus and host immune responses.
## MATERIALS AND METHODS
## Samples
Pre-ATI and post-ATI samples were obtained from previously completed ACTG (Advanc ing Clinical Therapeutics Globally for HIV/AIDS and Other Infections) ATI trials, including ACTG A5068 and A5170 (21,22). Posttreatment controllers were defined as individu als who remained off ART for ≥24 weeks posttreatment interruption and maintained viral loads ≤ 400 copies/mL for at least two-thirds of the time points as described in the CHAMP study (15), regardless of the viral load at ATI. Viral loads > 400 HIV-1 RNA copies/mL were acceptable if the participant was subsequently able to suppress to ≤400 HIV-1 RNA copies/mL and maintained virologic control through week 24 posttreatment interruption.
## Antibody and Fc-receptor binding titers
Antigen-specific antibody isotypes and FcyR were quantified through a multiplexed Luminex assay, as previously described (23). A panel of HIV antigens (insert antigens and source here) was covalently linked to carboxyl-modified MagPlex microspheres (Luminex) via N-hydroxysuccinimide (NHS)-ester linkages via EDC (Thermo Scientific) and Sulfo-NHS (Thermo Scientific). Serum samples were diluted (dilution here) for isotypes and FcyR binding and incubated with coupled beads in 384-well plates at 850 rpm for 2 h at room temperature. After the formation of immune complexes, microspheres were washed three times in 0.1% bovine serum albumin (BSA) and 0.05% Tween20 (Luminex assay buffer) with an automated plate washer (Tecan). Antibody isotypes were stained with mouse anti-human IgG-PE antibodies (Southern BioTech) for 1 h at room tempera ture. FcyR proteins were biotinylated (BirA500, Avidity) per manufacturer's instructions and tagged with streptavidin-PE (Agilent) and incubated for 1 h at room temperature. Plates were washed again, and beads were resuspended and run on flow cytometry instruments (iQue, Intellicyt) to determine the geometric mean fluorescent intensity. All experiments were run in duplicate fashion. Positive reactivity was defined as a signal that was 2 × s.d. over non-specific binding to human albumin.
## Antibody-dependent phagocytosis
ADNP and ADCP assays were performed as previously described (24,25). HIV antigens listed in Table S1 were purchased from ImmuneTech except for BG505 SOSIP, which was manufactured by the protein core at Duke University. HIV antigens were biotinylated using the EZ-link Sulfo-NHS-LC-LC-Biotin kit (Thermo Fisher) and coupled to fluorescent neutravidin beads (Thermo Fisher). Serum was diluted (1:100) and incubated for 2 h at 37°C, following incubation, unbound antibody was washed. For ADCP, immune complexes were incubated overnight (16-20 h) at 37° with cultured THP-1 (ATCC). For ADNP, primary neutrophils were isolated from whole blood with ammonium-chlor ide-potassium lysis buffer. Immune complexes and primary neutrophils were incubated for 1 h at 37°C. After incubation, cells were washed, THP-1 (ADCP) was subsequently fixed with 4% paraformaldehyde (PFA), primary neutrophils (ADNP) were stained for CD66b+ marker (BioLegend) and fixed with 4% PFA. Subsequent flow cytometry was performed with an iQue (IntelliCyt). A phagocytosis score for both assays was determined as % cells positive × geometric mean fluorescence intensity of positive cells. All assays were performed in duplicate fashion; for ADNP, two healthy donors were utilized.
## Antibody-dependent NK activation
ADNKA assays were performed as previously described (24). ELISA plates were coated with HIV antigens (insert antigens and source here) at 3 μg/mL and incubated at 37°C for 2 h. Plates were washed and blocked with 5% BSA and incubated at 4°C overnight. Primary NK cells were isolated from buffy coats (MGH) from two healthy donors using the RosetteSep isolation kit (StemCell Technologies). NK cells were incubated overnight with supplemented IL-15 (Miltenyi Biotec) at 37°C. Serum was diluted (1:50) and incubated on ELISA plates for 2 h at 37°C. During incubation, a staining cocktail was prepared with anti-CD107a-PE-Cy5 (BD), brefeldin-A (Sigma), and GolgiStop (BD) and added to NK cells. After the incubation plates were washed, NK cells were added at a concentration of 5 × 10 6 per well. Serum and NK cells were incubated for 5 h at 37°C. NK cells were subsequently stained with Perm and B (Thermo Fisher) and surface markers were stained with anti-CD16 APC-Cy7 (BD) and anti-CD56-PE-Cy7 (BD). Intracellular markers were stained with anti-IFNy APC (BD) and anti-MIP-1B PE (BD). Subsequent flow cytometry was performed with an iQue (IntelliCyt). NK cells were verified as CD56+, CD16+, and CD3-and activation was determined as the percentage of NK cells CD107a+, IFNy+, or MIP-1B+. Healthy donors were performed in replicate fashion for all assays.
## Antibody-dependent complement deposition
ADCD assays were performed as previously described (26). HIV antigens (insert antigens and source here) were biotinylated using the EZ-link Sulfo-NHS-LC-LC-Biotin kit (Thermo Fisher) and coupled to fluorescent neutravidin beads (Thermo Fisher). Diluted serum (1:10) and coupled beads were incubated at 37°C for 2 h to form immune complexes. Lyophilized guinea pig complement (Cedarlane) was reconstituted with water and added to gelatin veronal buffer containing Mg2+ and Ca2+ (GVB++, Sigma Aldrich). After incubation, plates were washed, and the complement solution was added and incubated for 50 min at 37°C. The reaction was then quenched with two washes with 15 mM EDTA in PBS. Immune complexes were stained with fluorescein-conjugated goat IgG fraction to guinea pig complement C3 (Mp Bio). All experiments were performed in duplicate.
## Data processing and statistics
Luminex and ADCD thresholds of detection were defined by nonspecific binding of samples to human albumin plus two standard deviations. Data below this threshold were excluded from statistical comparisons. Univariate distributions of each feature were confirmed to be non-parametric using Shapiro and Levene tests. Univariate comparisons between PTCs and NCs were performed using Mann-Whitney, and P values were corrected for each timepoint and readout group (e.g., P values for baseline IgG1 were corrected for all antigens measured) using the Bonferroni method. Corre lations were calculated using the Spearman method due to the presence of outlier data points. Spearman correlations were calculated for each humoral function across antigen-matched binding titers (total IgG, IgG1, IgG3, and FcyR2a/2b/3a/3b) and P values were corrected using the Bonferroni method.
## References
1. Cohn, Chomont, Deeks (2020) "The biology of the HIV-1 latent reservoir and implications for cure strategies" *Cell Host Microbe*
2. Siliciano, Siliciano (2022) "In vivo dynamics of the latent reservoir for HIV-1: new insights and implications for Cure" *Annu Rev Pathol*
3. Lambotte, Boufassa, Madec et al. (2005) "HIV controllers: a homogeneous group of HIV-1-infected patients with spontaneous control of viral replication" *Clin Infect Dis*
4. Hocqueloux, Prazuck, Avettand-Fenoel et al. (2010) "Long-term immunovirologic control following antiretroviral therapy interruption in patients treated at the time of primary HIV-1 infection" *AIDS*
5. Sáez-Cirión, Bacchus, Hocqueloux et al. (2013) "Post-treatment HIV-1 controllers with a long-term virological remission after the interruption of early initiated antiretroviral therapy ANRS VISCONTI Study" *PLoS Pathog*
6. Ackerman, Mikhailova, Brown et al. (2016) "Polyfunctional HIV-specific antibody responses are associated with spontaneous HIV control" *PLoS Pathog*
7. Etemad, Sun, Li et al. (2023) "HIV post-treatment controllers have distinct immunological and virological features" *Proc Natl Acad Sci*
8. Sharaf, Lee, Sun et al. (2018) "HIV-1 proviral landscapes distinguish posttreatment controllers from noncontrollers" *J Clin Invest*
9. Boudreau, Alter (2019) "Extra-neutralizing FcR-mediated antibody functions for a universal influenza vaccine" *Front Immunol*
10. Bartsch, Cizmeci, Kang et al. (2022) "Antibody effector functions are associated with protection from respiratory syncytial virus" *Cell*
11. Mackin, Desai, Whitener et al. (2023) "Fc-γR-dependent antibody effector functions are required for vaccine-mediated protection against antigen-shifted variants of SARS-CoV-2" *Nat Microbiol*
12. Gunn, Yu, Karim et al. (2018) "A role for Fc function in therapeutic monoclonal antibody-mediated protection against ebola virus" *Cell Host Microbe*
13. Ilinykh, Huang, Santos et al. (2020)
14. (2025) *Full-Length Text Journal of Virology*
15. "Non-neutralizing antibodies from a marburg infection survivor mediate protection by Fc-effector functions and by enhancing efficacy of other antibodies" *Cell Host Microbe*
16. Offersen, Yu, Scully et al. (2020) "HIV antibody Fc N-linked glycosylation is associated with viral rebound" *Cell Rep*
17. Namazi, Fajnzylber, Aga et al. (2018) "The control of HIV after antiretroviral medication pause (CHAMP) study: posttreatment controllers identified from 14 clinical studies" *J Infect Dis*
18. Molinos-Albert, Lorin, Monceaux et al. (1944) "Transient viral exposure drives functionallycoordinated humoral immune responses in HIV-1 post-treatment controllers" *Nat Commun*
19. Bartsch, Loos, Rossignol et al. (2021) "Viral rebound kinetics correlate with distinct HIV antibody features" *mBio*
20. Bertagnolli, Varriale, Sweet et al. (2020) "Autologous IgG antibodies block outgrowth of a substantial but variable fraction of viruses in the latent reservoir for HIV-1" *Proc Natl Acad Sci*
21. Esmaeilzadeh, Etemad, Lavine et al. (2023) "Autologous neutralizing antibodies increase with early antiretroviral therapy and shape HIV rebound after treatment interruption" *Sci Transl Med*
22. Chung, Mabuka, Ndlovu et al. (2018) "Viral control in chronic HIV-1 subtype C infection is associated with enrichment of p24 IgG1 with Fc effector activity" *AIDS*
23. Jacobson, Bucy, Spritzler et al. (2006) "Evidence that intermittent structured treatment interrup tion, but not immunization with ALVAC-HIV vCP1452, promotes host control of HIV replication: the results of AIDS Clinical Trials Group 5068" *J Infect Dis*
24. Skiest, Su, Havlir et al. (2007) "Interruption of antiretroviral treatment in HIV-infected patients with preserved immune function is associated with a low rate of clinical progression: a prospective study by AIDS Clinical Trials Group 5170" *J Infect Dis*
25. Brown, Dowell, Boesch et al. (2017) "Multiplexed Fc array for evaluation of antigen-specific antibody effector profiles" *J Immunol Methods*
26. Ackerman, Moldt, Wyatt et al. (2011) "A robust, high-throughput assay to determine the phagocytic activity of clinical antibody samples" *J Immunol Methods*
27. Karsten, Mehta, Shin et al. (2019) "A versatile highthroughput assay to characterize antibody-mediated neutrophil phagocytosis" *J Immunol Methods*
28. Fischinger, Fallon, Michell et al. (2019) "A high-throughput, bead-based, antigen-specific assay to assess the ability of antibodies to induce complement activation" *J Immunol Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548458&blobtype=pdf | # The BTK-DDX41 axis of the STING pathway is activated during cytomegalovirus lytic infection
Vargab Baruah, Christine O'connor
## Abstract
The innate immune response is the first line of defense against invad ing pathogens, including the betaherpesvirus human cytomegalovirus (CMV). The host's innate response acts as the first line of defense, and CMV, like other viruses, has consequently evolved multiple mechanisms to manipulate host interferon (IFN) responses. DEAD-box helicase 41 (DDX41) is an intracellular dsDNA sensor that, upon activation by Bruton's tyrosine kinase (BTK), triggers type I IFN production through the stimulator of interferon genes (STING) signaling pathway. Here, we show the activation of this signaling pathway during lytic CMV infection, wherein BTK, DDX41, and STING are activated through tyrosine phosphorylation, and both DDX41 and BTK interact with STING. Furthermore, CMV infection re-localizes a fraction of DDX41 from the nucleus to the cytoplasm. Here, DDX41 phosphorylation is attenuated, suggesting that cytoplasmic redistribution leads to a less active or inactive form. Additionally, a pool of DDX41 co-localizes in the virus assembly compartment (vAC) with the CMV tegument proteins, pp65 and pp71, each of which interact with DDX41 in immunoprecipitation assays. We further demonstrate the protective role of this signaling pathway, as treatment with the BTK inhibitor orelabrutinib attenuates DDX41 phosphorylation/activation and supports increased expression of viral proteins and virus replication. In sum, our work highlights the important role of BTK-DDX41-STING signaling in the innate immune response against CMV, which the virus subverts by attenuating its cytoplasmic activity, thereby diverting it from its typically protective function.
IMPORTANCE Human cytomegalovirus (CMV) is an ubiquitous pathogen that poses a significant threat to immunocompromised individuals, highlighting the critical role of innate immunity in controlling this viral infection. Despite extensive research, the complex mechanisms underlying innate immunity against CMV and the virus's strat egies for evading immune detection remain only partially understood. This study identifies the activation of the cellular BTK-DDX41-STING innate signaling axis during lytic CMV infection, which ultimately results in protective interferon responses. Our findings show that CMV infection triggers the cytoplasmic redistribution of the cellular protein, DDX41, leading to reduced phosphorylation and activity, thereby undermining its protective function. Additionally, pharmacological inhibition of BTK enhances viral protein expression and replication, highlighting the importance of this pathway in immune defense. Our work identifies BTK-and DDX41-dependent STING signaling as important for innate immune responses against CMV and further advances understand ing of CMV's manipulation of these responses.
myeloid cells, characterized by viral genome maintenance and suppressed infectious virus production. Latent CMV can intermittently reactivate under certain condi tions of immune system dysfunction and cellular differentiation, resulting in the lytic replication of infectious virus (2). While primary infection and reactivation events are typically asymptomatic, they can be severe in immunocompromised, immunosup pressed, or immuno-naïve individuals, potentially causing serious organ-specific diseases or even death. A substantial proportion of the population, even in developed countries, is immunologically vulnerable. In the United States, about 3% of adults are immunocom promised, and 6.6% are immunosuppressed due to health conditions or their treatments (3,4). Among immuno-naïve neonates, approximately 1 in 200 babies has congenital CMV infection (5), and up to 15% of these infants develop long-term sequelae (6). These data underscore the need to more completely understand the viral and cellular factors that support CMV replication.
The severity and outcome of CMV infection largely depend on the early immune events, when the host's innate immunity plays a crucial role (7). It can prevent the establishment of infection by overwhelming CMV upon initial contact or con fine the virus to the latent phase, preventing progression to the lytic phase. Early immune responses often involve recognition of pathogen-associated molecular patterns (PAMPs)-conserved molecular signatures found on pathogens-by the host's pattern recognition receptors (PRRs), which heavily influence the subsequent immune strategy. CMV's double-stranded DNA (dsDNA) genome is detected by dsDNA-sensing PRRs, triggering an immune signaling cascade that results in the production of anti-viral interferons (IFNs) and interferon-stimulated genes (ISGs) (8). Understanding these key immune responses at the molecular level is crucial, particularly in the absence of a vaccine, to determine how CMV might subvert these mechanisms as it infects the host.
Activation of key components of the innate immune response is intricately linked to the CMV life cycle, with different viral components targeted based on their abundance and accessibility. CMV dsDNA can bind cytosolic DNA receptors, such as cyclic GMP-AMP synthase (cGAS) and interferon-γ inducible protein 16 (IFI16). This DNA sensing is followed by activation of stimulator of interferon genes (STING; also known as MITA and ERIS), a key mediator in subsequent downstream immune signaling. STING activation leads to the recruitment and activation of TANK-binding kinase 1 (TBK1), which, in turn, phosphorylates interferon regulatory factor 3 (IRF3). This activated IRF3 then translocates to the nucleus, where it drives the expression of type I interferons (IFN) and other inflammatory cytokines, thereby mounting a defense against the invading virus (9).
Host-driven immune responses against CMV are complex since CMV has evolved impressively sophisticated methods of manipulating host factors. For instance, CMV's tegument protein pp65 and late protein pUL31 bind cGAS, thereby inhibiting its activity (10,11). Additionally, the membrane protein pUL42 also binds cGAS, blocking both its DNA binding and its ability to oligomerize/activate (12). Similar observations are reported for IFI16, where pp65 binds IFI16 and inhibits its DNA sensing by preventing its aggregation and signal amplification (13,14). CMV also modulates restriction factors downstream of DNA sensing. The viral glycoprotein US9, the tegument protein pp71, and the latency-associated protein pUL138 each bind STING and impair immune signaling (15)(16)(17). In addition to directly inhibiting immune signaling through binding, CMV also subverts the immune response by misdirecting the localization of restriction factors. For instance, while nuclear IFI16 binds CMV DNA early in infection, during later stages, it is phosphorylated by pUL97 and mislocalized to the cytoplasmic vAC (18). While these studies highlight the intricate and diverse mechanisms the host employs against CMV, as well as those the virus uses to counter them, our understanding of the array of innate immune sensors triggered by CMV remains incomplete.
DEAD-box helicase 41 (DDX41) is a helicase with a conserved Asp-Glu-Ala-Asp (DEAD) motif that functions as a DNA sensor (19). Similar to other DNA sensors, such as cGAS and IFI16, DDX41 activity is regulated by tyrosine phosphorylation. Lee et al. demonstrated that Bruton's tyrosine kinase (BTK) phosphorylates DDX41, enhancing its binding to adenovirus dsDNA and STING (20). Additionally, herpes simplex virus type 1 (HSV-1) infection results in upregulation of upregulates DDX41, which then interacts with cGAS, activates the STING signaling pathway, and promotes inflammatory cytokine production (19). However, the role of the BTK-DDX41-STING signaling axis in CMV infection remains unknown. Given that DDX41 shares many downstream effectors with cGAS and IFI16, and that CMV has a dsDNA genome, we posited that this signaling pathway could function as part of the innate immune response against CMV.
Here, we show that the BTK-DDX41-STING signaling axis is activated during lytic CMV infection, involving tyrosine phosphorylation of the restriction factors, and interaction of STING with DDX41 and BTK. Following infection, a pool of DDX41 re-localizes from the nucleus to the cytoplasm, where it co-localizes with the tegument protein, pp65, in the perinuclear vAC. Additionally, cytoplasmic-localized DDX41 displays reduced phosphorylation, indicating that its cytoplasmic redistribution leads to a less active form of the protein. Finally, we demonstrate that inhibiting BTK with orelabrutinib reduces phosphorylation and activation of BTK and DDX41, which increases viral protein expression and replication. Our work reveals an important role of the BTK-DDX41-STING pathway in the innate immune response to CMV and demonstrates a mechanism by which the virus disrupts this response.
## MATERIALS AND METHODS
## Cells and viruses
Primary newborn human foreskin fibroblasts (NuFF-1, passages 13-28; GlobalStem, Rockville, MD, USA) were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine, 0.1 mM nonessen tial amino acids, 10 mM HEPES, and 100 U/mL each of penicillin and streptomycin. The cells were maintained at 37°C with 5% CO 2 .
The bacterial artificial chromosome (BAC)-derived viruses, TB40/EmCherry (WT) (21) and TB40/EmCherry-UL99eGFP (UL99gfp) (22,23), were previously characterized. Virus stocks were prepared as previously described (24). In brief, p0 stocks were generated by transfecting BAC DNA into naïve NuFF-1 cells. When 100% cytopathic effect (CPE) was observed, p1 stocks were expanded on naïve NuFF-1 cells and harvested at 100% CPE. Virus was purified through a 20% sorbitol (VWR) cushion, resuspended 1:1 in XVIVO-15 (Lonza) and 3% bovine serum albumin (BSA), aliquoted, flash-frozen in liquid nitrogen, and stored at -80°C. Virus stocks were titered on naïve NuFF-1 cells by 50% tissue culture infectious dose (TCID 50 ) assay.
Cells were infected at the multiplicity of infection (MOI) and for the duration indicated in the text. Cells were infected in DMEM containing viral inoculum for 1 h at 37°C, with gentle agitation every 15 minutes (min). The cells were washed twice with 1× phosphate-buffered saline (PBS) and replenished with fresh DMEM at 37°C.
## Protein, RNA, and DNA analyses
For protein analyses, cells were treated and harvested at the times indicated in the text and/or figure legends. Where indicated, cells were transfected with 10 µg/mL of either poly(dA:dT) interferon-stimulatory DNA (ISD) or control ISD using Lipofectamine 2000. In all cases, cells were lysed on ice for 1 hour (h) in radioimmunoprecipitation assay (RIPA) buffer (1.0% NP-40; 1.0% sodium deoxycholate; 0.1% SDS; 0.15 M NaCl; 0.01 M NaPO4 [pH 7.2]; 2 mM EDTA), with vortexing every 15 min. For protein isolation in nuclear/cytoplasmic fractions, cell pellets were resuspended in lysis buffer (10 mM HEPES, pH 7.5; 5 mM MgCl2; 15 mM KCl; 1 mM PMSF) and subjected to three freeze-thaw cycles in liquid nitrogen. The resulting suspension was passed through a 27 G needle 15 times. The lysed cells were layered over 50% sucrose (wt/vol) and centrifuged. The supernatant was collected as the cytoplasmic fraction. The pellet containing nuclei was resuspended in lysis buffer and sonicated three times. After centrifugation, the supernatant was collected as the nuclear fraction. In all cases, protein concentrations were measured by Bradford assay with Protein Assay Dye Reagent Concentrate (Bio-Rad). Samples were denatured at 95°C for 10 min. Equal protein concentrations of the samples were separated by SDS-PAGE and transferred to 0.45 µm nitrocellulose membranes (Amersham) using a semi-dry transfer method. The following antibodies were used at the specified concentrations: anti-BTK (Cell Signaling Technology [CST], 1:5,000), anti-STING (CST, 1:5,000; MilliporeSigma, 1:5,000), anti-p-STING (CST, 1:1,000), anti-DDX41 (CST, 1:5,000), anti-TBK (CST, 1:5,000), anti-p-TBK (CST, 1:1,000), anti-CMV IE1 (clone 1B12, 1:100, (25)), anti-CMV UL44 (Virusys, 1:2,500), anti-CMV pp65 (clone 8A8, 1:50 ( 26)), anti-CMV pp71, anti-tubulin (MilliporeSigma, 1:5,000), anti-H3 (Abcam, 1:5,000), anti-beta-actinperoxidase (MilliporeSigma, 1:10,000), goat-anti-mouse horseradish peroxidase (HRP) secondary, and goat-anti-rabbit horseradish peroxidase (HRP) secondary (both from Jackson ImmunoResearch Labs, 1:10,000). Relative protein levels were quantified by densitometry using ImageJ (27).
To quantify transcripts, total RNA was isolated and purified using the High Pure RNA Isolation Kit (Roche) in accordance with the manufacturer's protocol. Equal RNA concentrations (1.0 μg) were then used to generate cDNA with random hexamer primers and SYBR™ Green I Reagents (Thermo Fisher Scientific). Equal volumes of cDNA were used for quantitative polymerase chain reaction (qPCR). The following primers were used for transcript quantification: IFNB, forward 5′-CTTGGATTCCTACAAAGAAGCAGC-3′ and reverse 5′-TCCTCCTTCTGGAACTGCTGCA-3′; UL123, forward 5′-GCCTTCCCTAAGAC CACCAAT-3′ and reverse 5′-ATTTTCTGGGCATAAGCCATAATC-3′; GAPDH, forward 5′-CTGTTGCTGTAGCCAAATTCGT-3′ and reverse 5′-ACCCACTCCTCCACCTTTGAC-3′. Samples were analyzed in triplicate using a 96-well-format CFX Connect Real Time PCR machine (Bio-Rad), and expression quantified relative to host GAPDH expression.
To quantify CMV genomic DNA, NuFF-1 fibroblasts were infected (MOI = 0.5 TCID 50 / cell) with WT virus. At each indicated time point, cells were harvested, and sam ples were treated with RNaseA (100 µg/mL) for 1 h at 37°C to remove RNA. DNA was isolated by phenol-chloroform-isoamyl alcohol extraction, followed by ethanol precipitation and resuspension in 10 mM Tris-HCl, pH 8.0 with 0.1 mM EDTA. Equal sample volume was quantified by qPCR using the following primers: viral UL69, forward 5′-GGCTGATGATCTTGCGGGAA-3′ and reverse 5′-CGAGAGTCTACGTCTGGCAC-3′; cellular MDM2, forward 5′-CCCCTTCCATCACATTGCA-3′ and reverse 5′-AGTTTGGCTTTCTCAGA GATTTCC-3′. UL69 and MDM2 abundance was quantified against a viral BAC standard (TB40/E-BACstandard), which contains an MDM2 gene fragment (28). All samples were analyzed in triplicate and then normalized to cellular MDM2.
## Immunofluorescence assays
Cells were cultured on coverslips in 6-well tissue culture plates (3 × 10 5 cells per well) and infected (MOI = 0.5 TCID 50 /cell) as described above. After 96 hpi, cells were washed three times with 1× PBS, fixed with 2% paraformaldehyde for 15 min at 37°C, and then permeabilized with 0.1% Triton X-100 for 15 min at room temperature. The cells were washed three times with 0.25% (wt/vol) saponin and subsequently blocked with 1× PBS containing 10% (vol/vol) human serum (Sigma Aldrich) and 0.25% (wt/vol) saponin. Cells were incubated with primary antibodies against STING, BTK, phospho-BTK-Y551 (all from Cell Signaling Technology, 1:1,000), DDX41 (Cell Signaling Technology, 1:1,000), and STING (MilliporeSigma, 1:1,000) in 10% (vol/vol) human serum and 0.25% (wt/vol) saponin for 1 h at room temperature. Following this, the cells were washed three times with 0.25% (wt/vol) saponin and incubated with Alexa-488-conjugated anti-mouse and Alexa-647-conjugated anti-rabbit secondary antibodies (both from Thermo Fisher, 1:1,000) for 1 h at room temperature, shielded from light. Finally, coverslips were mounted with FluorSave reagent (MilliporeSigma) on slides and visualized using a Leica SP8 inverted confocal microscope with a 63× oil-immersion objective. Colocalization in immunofluorescent images was quantified using Volocity software. To assess colocaliza tion of DDX41 or TBK1 with STING, we calculated the percentage of mCherry-positive, CMV-infected cells that exhibited overlap between the Volocity-defined compartments of each protein and STING. Quantification was performed on three fields of view per condition, and the resulting percentages were plotted as bar graphs. Where indicated, Pearson's correlation coefficient was calculated for DDX41 co-occurrence with pp65 using ImageJ/Fiji.
## Co-immunoprecipitation
Cells were lysed in 1.0 mL of immunoprecipitation (IP) buffer (25 mM Tris [pH 7.4], 1 mM MgCl₂, 2 mM EDTA; 1.0% Triton-X 100; 150 mM NaCl; Complete Protease Inhibitor Cocktail [Roche]) on ice for 2 h. Protein G beads (100 μL; Protein G-sepharose Fast Flow, MilliporeSigma) were washed twice with IP buffer (500 × g, 5 min), blocked for 1 h in 5.0% BSA in IP buffer, and then resuspended in 1.0 mL IP buffer. The samples were clarified by centrifugation (12,000 × g, 10 min, 4°C), and 50 μL of the sample (10%) was retained as the "input" control. The remaining sample was divided into two aliquots, each incubated overnight at 4°C with protein G beads and either 2 μg of the specific precipitating antibody or an isotype-matched IgG control antibody. The next day, the samples were centrifuged at 500 × g for 5 min, and 50 μL was reserved as the "unbound" sample. The samples were then washed four times with IP wash buffer (25 mM Tris [pH 7.4], 10 mM MgCl₂, 2 mM EDTA, 0.1% Triton-X 100), with 50 μL retained after washes 1 and 4. After the final wash, proteins in all samples were denatured using 2× SDS and heating at 95°C for 10 min. The samples were then separated by SDS-PAGE and transferred to nitrocellulose as described above.
## Toxicity assay
NuFF-1 cells (1.75 × 10 4 cells/well) were plated onto 96-well opaque white CulturPlate (Perkin Elmer) and allowed to adhere overnight. Cells were then treated with orelab rutinib (GLPBio) in nine two-fold dilutions (0.078 to 20 µM) for 24 h, followed by three washes with 1× PBS. Cell viability was assessed after 2 and 5 days (i.e., 3 and 6 days post-treatment) using CellTiter-Glo (Promega) according to the manufacturer's guidelines, and the plates were read with a Cytation 5 (Agilent) plate reader. DMSO (equivalent vol/vol) served as the diluent and vehicle control, while puromycin (ranging from 20 µM to 0.08 µM) was used as a positive control.
## Quantification of viral lytic replication
NuFF-1 cells were treated with 0.5 µM orelabrutinib (GLPBio) prior to infection with TB40/EmCherry (MOI = 0.01 TCID 50 /cell) for 1 h, with gentle agitation every 15 min. After adsorption, cells were washed three times with 1× PBS, and fresh media (V f = 2.0 mL) containing 0.5 µM orelabrutinib was added. Cell-free virus (350 µL) was collected over a 16-day (d) time course, and 250 µL of fresh media with 100 µL of 10 µM orelabrutinib was added to restore the total volume to 2 mL. DMSO (equivalent vol/vol) served as the diluent and vehicle control. Cell-associated virus was harvested at the final time point (16 days post-infection; dpi) by lysing the cells through three freeze-thaw cycles. The clarified supernatant was then used to determine the cell-associated viral titer. Cell-free and cell-associated viruses were quantified by TCID 50 assay on naïve NuFF-1 fibroblasts.
## RESULTS
## BTK-DDX41-STING signaling is activated during lytic CMV infection
Emerging data reveal an antiviral role for DDX41 (19,(29)(30)(31), though whether DDX41 functions as a restriction factor during CMV infection remains unclear. To begin to address this, we first assessed the expression and activation of restriction factors associated with the BTK-DDX41-STING signaling pathway during lytic CMV replication. We mock-or lytically infected NuFF-1 fibroblasts with TB40/EmCherry (wild type; WT) over a 96-hour (h) time course, after which we measured protein BTK, STING, DDX41, and TBK1 abundance and phosphorylation, which is essential for activation of these restriction factors (20,(32)(33)(34). Additionally, we probed for the viral protein, IE1. As a control for BTK and DDX41 activation, we transfected parallel cultures with poly(dA:dT) (29). Since there is no phospho-specific antibody for DDX41, and the phospho-BTK antibody is ineffective in immunoblot assays, we performed immunoprecipitation with a pan-phospho-tyrosine antibody, followed by immunoblot using antibodies directed against the total forms of either BTK or DDX41. Our data indicate that BTK is activated by 6 h post-infection (hpi), with activation increasing through 96 hpi (Fig. 1A). Similarly, DDX41 is also activated at 6 hpi, with sustained increased phosphorylation through the 96 h time point. Additionally, total DDX41 is also increased over the time course relative to mock-infected cultures (Fig. 1B). Levels of both total and phosphorylated forms of BTK and DDX41 were comparable to those induced by poly(dA:dT) transfection over a similar time course, although poly(dA:dT) transfection triggered an earlier and more rapidly peaked response (Fig. 1A andB). We also observed an increase in phospho-forms of STING and TBK1 to levels consistent with their activation following interferon stimula tory DNA (ISD) transfection (Fig. 1C), which serves as a positive control for STING-TBK1 activation (35). Finally, we quantified mRNA expression of IFNB, which encodes IFNβ, a protein upregulated in response to innate immune signals. Infection induces a time-dependent increase in IFNB transcription, which peaks at 3 hpi and is sustained throughout the time course (Fig. 1D). We consistently observed increased activation of BTK and DDX41 at 96 hpi, with sustained activation of STING and TBK1 through this time point. Overall, these data indicate that CMV induces both the abundance and phosphorylated activation of BTK, DDX41, STING, and TBK1, which results in increased IFNB expression in infected fibroblasts.
To investigate the signaling mechanism and confirm pathway activation, we used immunofluorescence assays (IFA) to examine colocalization and co-immunoprecipitation (co-IP) to assess interactions among restriction factors, with a particular focus on STING. Previous studies have shown that STING physically interacts with many of its signaling partners, such as TBK1 and IRF3, bringing them into close proximity for phosphorylation and subsequent activation (36). We mock-or WT-infected NuFF-1 fibroblasts for 96 h and visualized colocalization of each restriction factor (BTK, pBTK-Y551, DDX41, and TBK) with STING by immunofluorescence. We found BTK (Fig. 2A), its activated form pBTK-Y551 (Fig. 2B), DDX41 (Fig. 2C), and TBK1 (Fig. 2D) each colocalize with STING in the perinuclear region of the infected cell. We quantified the colocalization of STING with upstream DDX41 and downstream TBK1 in infected cells expressing mCherry between 24 and 96 hpi and found consistent colocalization during this period in approximately 40-50% of cells (Fig. 2C andD). Quantification in mCherry-positive cells was not possible between 3 and 12 hpi due to the absence of mCherry expression (Fig. S1). Furthermore, we confirmed the interaction between STING and BTK (Fig. 3A; Fig. S2A), as well as STING and DDX41 (Fig. 3B; Fig. S2B) by co-IP, consistent with previous findings (20,37). We also confirmed CMV infection by IE1 protein abundance (Fig. S2C andD). Collectively, our data reveal that BTK-DDX41-STING signaling is activated following WT infection.
## CMV redistributes a fraction of DDX41 to the cytoplasm.
Our data reveal that CMV infection results in distinct changes in the subcellular localization of the restriction factors, including DDX41 (Fig. 2). In the absence of infection (mock control), DDX41 primarily displays nuclear localization; however, CMV infection results in the progressive migration of a fraction of this host restriction factor to the perinuclear region as early as 24 hpi. DDX41 cytoplasmic localization increases as infection progresses and is observed in about 50-60% of all infected cells at 72 and 96 hpi (Fig. 2C). This is consistent with previous work revealing that DNA stimulation triggers DDX41 relocalization (19). Furthermore, the CMV-induced perinuclear relocalization of DDX41 suggested that this host protein may become co-localized within the CMV vAC. To test this, we infected fibroblasts with a recombinant virus, TB40/EmCherry-UL99gfp, which contains eGFP fused to the carboxy terminus of the UL99 ORF (22,23). UL99 encodes pp28, a viral protein that localizes to the vAC during lytic infection (38), which we also confirmed (Fig. 4A). Similar to our data in Fig. 2C, we noted that a portion of DDX41 localizes to the vAC (Fig. 4A). Additionally, the other DDX41-associated restriction factors (BTK, STING, and TBK1) also relocalize to the vAC following CMV lytic infection (Fig. S3). To gain a better understanding of the abundance of DDX41 that was relocalized to the cytoplasm in response to CMV infection, we analyzed protein lysates derived from nuclear and cytoplasmic fractions of mock-or WT-infected cells over a time course of infection. In mock-infected cells, DDX41 retains its nuclear localization. However, in CMVinfected fibroblasts, a portion of DDX41 is translocated to the cytoplasm as early as 24 hpi, which increases and persists through the time course we evaluated, while a pool of DDX41 is retained in the nucleus (Fig. 4B). Furthermore, a portion of the cytoplasmic DDX41 pool is detected in the perinuclear vAC between 24 and 96 hpi, although not at earlier time points (Fig. 4C; Fig. S4). Localization of a portion of DDX41 to the vAC is supported by moderate Pearson's coefficient values for colocalization with pp65, a viral protein that also localizes to the vAC during lytic infection (Fig. 4C andD). This indicates a fraction of DDX41 is relocalized to the cytoplasm, where a portion is localized to a perinuclear region, consistent with vAC staining (Fig. 4C andD). Next, we hypothesized that since DDX41 is a dsDNA sensor, perhaps a portion of DDX41 is retained within the nucleus by the host cell to target the viral genome. To test this, we evaluated whether DDX41 colocalizes with the CMV DNA polymerase subunit, UL44, which we used to identify sites of viral genome replication by IFA. Although a fraction of DDX41 remains in the nucleus during infection, nuclear DDX41 does not specifically colocalize with UL44 (Fig. 4E). Taken together, our data indicate that a fraction of DDX41 is relocalized from the nucleus to the cytoplasm, where a portion is localized to perinuclear cytoplasmic regions consistent with the vAC.
In addition to relocalizing this antiviral protein, we hypothesized that CMV infection may also result in altered activity of the cytoplasmic portion of DDX41. Thus, we next asked whether cytoplasmic-localized DDX41 exhibited altered signaling potency compared to its nuclear counterpart. To this end, we analyzed the tyrosine phosphoryla tion of DDX41 in nuclear and cytoplasmic fractions of cells that were either mock-or WT-infected, first confirming the fractionation using tubulin and H3 as cytoplasmic and nuclear loading controls, respectively (Fig. 5A). We then performed immunoprecipitation with a phospho-tyrosine antibody and probed for DDX41 in these fractions. Under mock conditions, DDX41 is localized to the nucleus, consistent with our findings above, where it is phosphorylated. During infection, although the total DDX41 is more abundant in the cytoplasm than in the nucleus, the phosphorylated/activated fraction in the cytoplasm is greatly reduced, as indicated by a reduction in the p-Tyr DDX41 pulldown from the cytoplasmic fraction (Fig. 5B). Taken together, these data indicate that CMV infection induces a relocalization of a fraction of DDX41 in its inactivated form to the cytoplasm.
## DDX41 interacts with CMV-encoded tegument proteins
Previous reports have noted a similar CMV-induced perinuclear distribution of the dsDNA sensor, IFI16. At later stages of infection, IFI16 is localized to the vAC, where it interacts with the tegument protein, pp65, to collectively subvert IFI16-mediated DNA sensing (10,13,18). Indeed, pp65 also interacts with cGAS, thereby interfering with cGAS-STING interaction and innate sensing activity (10). In parallel, UL82-encoded tegument protein, pp71, interacts with host restriction factors, ultimately preventing the proper assembly of the STING-TBK1-IRF complex, whose activity is required for innate sensing via this pathway (39). Based on these observations, as well as our IFAs showing colocalization of a pool of DDX41 and pp65 between 24 and 96 hpi (Fig. 4C), we hypothesized that a similar interaction between CMV-encoded tegument proteins and DDX41 regulates this host restriction factor's localization and activity following CMV infection. To this end, we infected fibroblasts and assessed the potential co-localization and interaction between DDX41 and either pp65 or pp71. Our findings reveal that a portion of DDX41 co-localizes with pp65 (Fig. 6A) and pp71 (Fig. 6B) in the perinuclear vAC. Furthermore, DDX41 interacts with pp65 (Fig. 6C) and pp71 (Fig. 6D), although detection of the latter interaction is far weaker than the former. These results suggest that pp65 and, to a lesser extent, pp71 interact with DDX41. While additional work is required, this is perhaps a mechanism by which the virus retains a pool of this host restriction factor in the cytoplasm in its inactive form.
## Activated BTK-DDX41-dependent STING signaling limits CMV protein expression and replication
Since our data indicate that BTK, DDX41, and STING are activated upon lytic CMV infection, we next assessed the impact of this branch of STING signaling on infection. To this end, we utilized pharmacological inhibition using orelabrutinib, a highly specific and irreversible BTK inhibitor (40). Since BTK activates DDX41 (20), we posited this would serve as a reasonable approach to inhibit DDX41, as our attempts of both knockout and knockdown of this protein were unsuccessful (data not shown). This is perhaps not surprising, given that DDX41 is embryonic lethal (30,41). Using the pharmacological approach instead, we first evaluated the toxicity and efficacy of orelabrutinib in our model system and observed minimal toxicity over a concentration range of 0.078 to 20 µM. To validate the assay, we included puromycin treatment as a positive control (Fig. S6A). We also evaluated the impact of orelabrutinib treatment on the signaling pathway using a multipronged approach. We assessed DDX41 protein abundance following orelabrutinib treatment at various concentrations (0.1-10 µM) and durations (24-120 h at 0.5 µM). DDX41 abundance was attenuated following 0.5 µM orelabrutinib treatment and was further decreased with increasing concentration of orelabrutinib (Fig. S6B) and longer treatment times (Fig. S6C). Additionally, orelabrutinib treatment reduced IFNB transcription in WT-infected cells, although this reduction was not statistically significant (Fig. S6D). This is perhaps unsurprising, as IFNB increases prior to BTK-DDX41 activation (Fig. 1), indicating that IFNB is likely activated early post-infection by other arms of the STING pathway. Despite this, we did observe diminished phosphorylation of BTK (Fig. S7A) and DDX41 (Fig. S7B) in orelabrutinib-treated, WT-infected cells compared to untreated counterpart cultures in IP-western assays (Fig. S7). Our data also reveal orelabrutinib-treated, WT-infected cells display a reduction in total DDX41 abundance (Fig. S7B), which likely also contributes to attenuated phosphorylation levels of this protein. Collectively, these data suggest that orelabrutinib inhibits BTK and DDX41 activity without causing toxicity in fibroblasts. Having an effective means by which to target BTK-DDX41, we leveraged orelabruti nib to assess the impact of inhibiting BTK-DDX41-dependent STING signaling on lytic CMV infection. To this end, to determine the impact of orelabrutinib-mediated DDX41 inhibition on viral protein translation, we evaluated the abundance of representative viral proteins from the early and late kinetic classes. Orelabrutinib-treated, WT-infected cells displayed a reduction in the expression of BTK, DDX41, and STING compared to infected cells treated with DMSO (-orelabrutinib; Fig. 7A). Concurrent with the reduced abundance of these restriction factors, we observed increased abundance of the viral early protein, UL44, and the viral late protein, pp65 (Fig. 7A). In parallel, we evaluated CMV genome copy number in WT-infected fibroblasts treated with 0.5 µM orelabrutinib or DMSO (-orelabrutinib) over a 96 h time course. We find a statistically significant increase of viral genomes at 72 hpi in orelabrutinib-treated cells (Fig. 7B). We also evaluated multistep viral growth over a 16-day time course, which revealed that orelabrutinib treatment resulted in an increase in cell-free virus production from 8 to 16 dpi, with a statistically significant increase observed at 12 dpi (Fig. 7C). Additionally, cell-associated virus also increased in the presence of orelabrutinib, although not to statistically significant levels (Fig. 7D). In sum, these data show that attenuation of BTK-DDX41-dependent signaling promotes CMV lytic replication, indicating a role for this branch of STING signaling during CMV infection.
## DISCUSSION
Here, we show that BTK-DDX41-dependent STING signaling functions in the innate immune response against lytic CMV infection. Specifically, our data reveal that this branch of STING signaling is activated through the interaction of BTK and DDX41 with STING (Fig. 1 to 3). Furthermore, CMV infection induces a cytoplasmic redistrib ution of a portion of DDX41 and its associated restriction factors (Fig. 4), where it exhibits reduced activation, as evidenced by decreased tyrosine phosphorylation (Fig. 5) and interacts with the CMV tegument proteins pp65 and pp71 (Fig. 6). Additionally, we demonstrate the protective nature of this signaling pathway by pharmacological inhibition of BTK using orelabrutinib, an inhibitor that attenuates this signaling axis and results in increased viral protein abundance and CMV replication (Fig. 7). Collectively, our findings underscore the significance of this signaling pathway and reveal a mechanism by which CMV evades the immune response (Fig. 8).
dsDNA sensors are of considerable interest in the context of CMV, as several, including cGAS, IFI16, AIM2, and ZBP1, trigger IFN production upon detection of the CMV genome (42)(43)(44)(45). Our findings herein now add DDX41 to this growing repertoire of sensors that are involved in the innate response against CMV. Our study shows that lytic CMV infection induces the abundance of BTK, DDX41, STING, and TBK1, as well as accumula tion of IFNB transcripts. We confirmed the activation of the pathway, as evidenced by the phosphorylation of BTK, DDX41, and STING, with both BTK and DDX41 colocalizing and interacting with STING. It is important to note that while IFNB transcripts are increased upon CMV infection, this precedes BTK-DDX41 activation. This is consistent with the ability of other arms of the STING pathway, which function earlier post-infection to trigger IFN production. It is also likely that functional redundancy exists in IFN regula tion (e.g., cGAS-STING and IFI16-STING). While a direct role for BTK-DDX41 in driving IFN remains unclear, our data support a role for this axis in repressing infection via STING. Indeed, previous research showed that DDX41 interacts with STING at its second to fourth transmembrane domains in poly(dA)-stimulated myeloid dendritic cells (29).
Additionally, in a HEK293T-based overexpression system, the BTK SH3/SH2-interaction domain binds the transmembrane region of STING (20). This mediatory role of STING is well documented; for example, STING interacts with TBK1 and IRF3 and then facilitates the phosphorylation and activation of IRF3 by TBK1 (36). We hypothesize a similar mechanism based on the results from our study. After BTK and DDX41 bind STING, BTK could activate DDX41 through phosphorylation. Although it remains unclear whether STING directly induces this phosphorylation or merely facilitates the proximity of BTK and DDX41, our current data suggest that STING likely plays a more active role in the process. Multiple studies have highlighted CMV's diverse mechanisms to counteract DNA sensor-dependent immunity at various levels. At the DNA sensing level, CMV UL31 interacts with cGAS, causing the dissociation of DNA from cGAS and negatively regulating the STING signaling pathway (11). Similarly, UL42 binds cGAS, in turn inhibiting its DNA binding, oligomerization, and enzymatic activity (12). The tegument protein, pp65, also binds cGAS, preventing its interaction with STING and thus inactivat ing STING signaling (10). pp65 also interacts with IFI16's PYRIN domain, blocking its oligomerization upon DNA sensing and subsequent immune signaling (14). Here, we observed the formation of the nascent vAC using pp65 beginning at 24 hpi-the same time point at which a fraction of DDX41 was detected in the cytoplasmic fraction. This observation, together with the documented propensity of viral proteins, particularly pp65, to bind dsDNA sensors, led us to focus on tegument proteins and their potential interaction with DDX41. Indeed, we found that DDX41 associates with both pp65 and, to a lesser extent, pp71 during lytic CMV infection. The decreased detection of pp71 does not necessarily imply a weaker or less probable interaction, but rather could result from diminished availability of pp71 for interaction with DDX41 compared to pp65. Importantly, while a portion of DDX41 is inactive in the cytoplasm, it is important to point out that our data show a pool of DDX41 that retains its active state. This is consistent with our data revealing that DDX41 inhibition with orelabrutinib results in increased viral protein abundance and viral replication. Thus, while CMV, perhaps via pp65 and pp71, restricts a portion of DDX41, this host factor retains an active form still capable of targeting the virus. What remains outstanding is whether the interaction between DDX41 and pp65 and/or pp71 is required for relocalizing DDX41 to the cytoplasm or whether these interactions serve to keep DDX41 in the cytoplasm once DDX41 is already there. While possible, it is unlikely that the tegument proteins aid in DDX41 redistribution. Others have reported DDX41 relocalization to perinuclear regions in response to DNA stimuli (19,46), which Singh and colleagues demonstrated to be dependent upon acetylation of DDX41's most amino-terminal nuclear localization signal (NLS; (19)). It is thus likely that DDX41 cytoplasmic localization during CMV infection is stimulated in a similar fashion, dependent upon DNA sensing, as opposed to specific, CMV-encoded factors. Why would CMV-encoded tegument proteins interact with cytoplasmic-localized DDX41? It is possible that the tegument proteins interact with DDX41 to retain this host protein in a less active state late in infection. We showed that cytoplasmic DDX41 activity is attenuated, despite its colocalization with STING and BTK in the perinuclear region. It is attractive to hypothesize that the interaction with the tegument proteins precludes DDX41 activation, similar to what is reported for cGAS and IFI16 (10,14). Alterna tively, the interaction between CMV tegument protein(s) and DDX41 results in DDX41's incorporation into the mature viral particle, similar to that reported for IFI16 (18). This could ensure that DDX41 is delivered to newly infected cells, allowing for its immedi ate action for very early innate responses following infection. The latter hypothesis could explain the late-phase phenotypes we observed. Furthermore, virion-delivered DDX41 may indicate additional functions beyond innate immune responses, similar to IFI16, which interacts with pp65 to activate viral gene transcription at the major immediate early promoter (14). DDX41 may have additional functions during CMV infection, as this host protein additionally coordinates RNA splicing and transcriptional elongation, thereby preventing DNA replication stress (47). Thus, one could posit that DDX41's functions during CMV infection are more extensive, raising the possibility that its potential inclusion in the virion serves a biologically important purpose. These are compelling questions warranting further exploration, and experiments to address these possibilities are ongoing.
Our findings herein highlight the protective role of BTK-DDX41-dependent STING signaling in the immune response to CMV lytic infection and further reveal a mechanism of immune evasion through disrupted activation and localization of a fraction of DDX41. Our work advances our understanding of innate responses against CMV, providing insights that could inform strategies to counteract CMV's immune evasive strategies.
## References
1. Griffiths, Reeves (2021) "Pathogenesis of human cytomegalovirus in the immunocompromised host" *Nat Rev Microbiol*
2. Smith, Chan, Connor (2021) "Modulation of host cell signaling during cytomegalovirus latency and reactivation" *Virol J*
3. Antinori, Bausch-Jurken (2023) "The Burden of COVID-19 in the immunocompromised patient: implications for vaccination and needs for the future" *J Infect Dis*
4. Martinson, Lapham (2024) "Prevalence of immunosuppression among US adults" *JAMA*
5. Jones, Bailey, Bamford et al. (2023) "Managing challenges in congenital CMV: current thinking" *Arch Dis Child*
6. Boppana, Ross, Fowler (2013) "Congenital cytomegalovirus infection: clinical outcome" *Clin Infect Dis*
7. Dell'oste, Biolatti, Galitska et al. (2020) "Tuning the orchestra: HCMV vs" *innate immunity. Front Microbiol*
8. Marques, Ferreira, Ribeiro (2018) "The interplay between human cytomegalovirus and pathogen recognition receptor signaling" *Viruses*
9. Almine, Hare, Dunphy et al. (2017) "IFI16 and cGAS cooperate in the activation of STING during DNA sensing in human keratinocytes" *Nat Commun*
10. Biolatti, Dell'oste, Pautasso et al. (2018) "Human cytomegalovirus tegument protein pp65 (pUL83) dampens type I interferon production by inactivating the DNA sensor cGAS without affecting STING" *J Virol*
11. Huang, Zou, Liao et al. (2018) "Human cytomegalovirus protein UL31 inhibits DNA sensing of cGAS to mediate immune evasion" *Cell Host Microbe*
12. Fu, Guo, Zou et al. (2019) "Human cytomegalovirus protein UL42 antagonizes cGAS/MITAmediated innate antiviral response" *PLoS Pathog*
13. Cristea, Moorman, Terhune et al. (2010) "Human cytomegalovirus pUL83 stimulates activity of the viral immediate-early promoter through its interaction with the cellular IFI16 protein" *J Virol*
14. Li, Chen, Cristea (2013) "Human cytomegalovirus tegument protein pUL83 inhibits IFI16-mediated DNA sensing for immune evasion" *Cell Host & Microbe*
15. Choi H Jin, Park, Kang et al. (2018) "Human cytomegalovirus-encoded US9 targets MAVS and STING signaling to evade type I interferon immune responses" *Nat Commun*
16. Nukui, Roche, Jia et al. (2020) "Protein S-nitrosylation of human cytomegalovirus pp71 inhibits its ability to limit STING antiviral responses" *J Virol*
17. Albright, Mickelson, Kalejta (2021) "Human cytomegalovirus UL138 protein inhibits the STING pathway and reduces interferon beta mRNA accumulation during lytic and latent infections"
18. (2025) *Full-Length Text Journal of Virology*
19. Dell'oste, Gatti, Gugliesi et al. (2014) "Innate nuclear sensor IFI16 translocates into the cytoplasm during the early stage of In vitro human cytomegalovirus infection and is entrapped in the egressing virions during the late stage" *J Virol*
20. Singh, Vidhyasagar, Yang et al. (2022) "DDX41 is required for cGAS-STING activation against DNA virus infection" *Cell Rep*
21. Lee, Kim, Kui et al. (2015) "Bruton's tyrosine kinase phosphorylates DDX41 and activates its binding of dsDNA and STING to initiate type 1 interferon response" *Cell Rep*
22. O'connor, Shenk (2011) "Human cytomegalovirus pUS27 G proteincoupled receptor homologue is required for efficient spread by the extracellular route but not for direct cell-to-cell spread" *J Virol*
23. Nukui, Connor, Murphy (2018) "The natural flavonoid compound deguelin inhibits HCMV lytic replication within fibroblasts" *Viruses*
24. O'connor, Nukui, Gurova et al. (2016) "Inhibition of the FACT complex reduces transcription from the human cytomegalovirus major immediate early promoter in models of lytic and latent replica tion" *J Virol*
25. O'connor, Shenk (2012) "Human cytomegalovirus pUL78 G proteincoupled receptor homologue is required for timely cell entry in epithelial cells but not fibroblasts" *J Virol*
26. Zhu, Shen, Shenk (1995) "Human cytomegalovirus IE1 and IE2 proteins block apoptosis" *J Virol*
27. Bechtel, Shenk (2002) "Human cytomegalovirus UL47 tegument protein functions after entry and before immediate-early gene expression" *J Virol*
28. Schneider, Rasband, Eliceiri (2012) "NIH Image to ImageJ: 25 years of image analysis" *Nat Methods*
29. Krishna, Wass, Connor (2020) "Activator protein-1 transacti vation of the major immediate early locus is a determinant of cytomega lovirus reactivation from latency" *Proc Natl Acad Sci*
30. Zhang, Yuan, Bao et al. (2011) "The helicase DDX41 senses intracellular DNA mediated by the adaptor STING in dendritic cells" *Nat Immunol*
31. Stavrou, Aguilera, Blouch et al. (2018) "DDX41 recognizes RNA/DNA retroviral reverse transcripts and is critical for in vivo control of murine leukemia virus infection" *mBio*
32. Duan, Zeng, Fan et al. (2019) "Herpes simplex virus type 1-encoded miR-H2-3p manipulates cytosolic DNA-stimulated antiviral innate immune response by targeting DDX41" *Viruses*
33. Rawlings, Scharenberg, Park et al. (1996) "Activation of BTK by a phosphorylation mechanism initiated by SRC family kinases" *Science*
34. Wang, Wang, Veleeparambil et al. (2020) "EGFR-mediated tyrosine phosphorylation of STING determines its trafficking route and cellular innate immunity functions" *EMBO J*
35. Wang, Sharma, Veleeparambil et al. (2021) "STING-mediated interferon induction by herpes simplex virus 1 requires the protein tyrosine kinase Syk"
36. Zhang, Chan, Samarina et al. (2016) "Cytoplasmic isoforms of Kaposi sarcoma herpesvirus LANA recruit and antagonize the innate immune DNA sensor cGAS" *Proc Natl Acad Sci*
37. Tanaka, Chen (2012) "STING specifies IRF3 phosphorylation by TBK1 in the cytosolic DNA signaling pathway" *Sci Signal*
38. Parvatiyar, Zhang, Teles et al. (2012) "The helicase DDX41 recognizes the bacterial secondary messengers cyclic di-GMP and cyclic di-AMP to activate a type I interferon immune response" *Nat Immunol*
39. Sanchez, Sztul, Britt (2000) "Human cytomegalovirus pp28 (UL99) localizes to a cytoplasmic compartment which overlaps the endoplas mic reticulum-golgi-intermediate compartment" *J Virol*
40. Fu, Su, Gao et al. (2017) "Human cytomegalovirus tegument protein UL82 inhibits STING-mediated signaling to evade antiviral immunity" *Cell Host Microbe*
41. Dhillon (2021) "Orelabrutinib: first approval" *Drugs (Abingdon Engl)*
42. Stepanchick, Wilson, Sulentic et al. (2024) "DDX41 haploinsufficiency causes inefficient hematopoiesis under stress and cooperates with p53 mutations to cause hematologic malignancy" *Leukemia*
43. Paijo, Döring, Spanier et al. (2016) "cGAS senses human cytomegalovirus and induces type I interferon responses in human monocyte-derived cells" *PLoS Pathog*
44. Gariano, Dell'oste, Bronzini et al. (2012) "The intracellular DNA sensor IFI16 gene acts as restriction factor for human cytomegalovirus replication" *PLoS Pathog*
45. Huang, Ma, Huang et al. (2017) "Interaction between HCMV pUL83 and human AIM2 disrupts the activation of the AIM2 inflammasome" *Virol J*
46. Defilippis, Alvarado, Sali et al. (2010) "Human cytomegalovirus induces the interferon response via the DNA sensor ZBP1" *J Virol*
47. Ma, Li, Fan et al. (2018) "Identification of DEAD-box RNA helicase DDX41 as a trafficking protein that involves in multiple innate immune signaling pathways in a zebrafish model" *Front Immunol*
48. Shinriki, Hirayama, Nagamachi et al. (2022) "DDX41 coordinates RNA splicing and transcriptional elongation to prevent DNA replication stress in hematopoietic cells" *Leukemia*
49. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607910&blobtype=pdf | # Could wastewater-based surveillance be key to combating antimicrobial resistance?
José Balcázar
## Abstract
Antimicrobial resistance (AMR) is one of the most urgent health threats of the 21st century. Surveillance is needed to enable timely interventions, close knowledge gaps, and anticipate long-term trends. Current frameworks rely heavily on clinical data, which often fail to capture population-level dynamics. Wastewater-based surveillance (WBS) offers a complementary approach by detecting antibiotic resistance genes (ARGs) in sewage. In AMR surveillance, early warning includes the detection of novel or clinically relevant ARGs, including those carried by mobile genetic elements (MGEs) before they affect clinical outcomes. WBS can also reveal resistome composition, dissemination routes, and ecological drivers of AMR. This is especially relevant in settings with poor sanitation, high exposure, and limited clinical reporting. Unlocking its potential will require harmonized protocols, sustained investment, and strong ethical measures. Within a One Health framework, WBS can strengthen equitable and evidence-based strategies against AMR.
## WBS CAN SUPPORT EFFORTS TO COMBAT THE GLOBAL AMR CRISIS
WBS, first applied in poliovirus detection in 1939, was widely adopted during the coronavirus disease 2019 (COVID-19) pandemic, particularly for monitoring severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission at the population level (8). Advances in real-time PCR, high-throughput sequencing, and bioinformatics now enable rapid and cost-effective sewage analysis covering a broad range of biological markers. As AMR emerges as one of the most pressing global health threats, WBS provides an accessible approach that complements conventional monitoring systems. Clinical surveillance is essential but limited, particularly in low-and middle-income countries (LMICs) where reliable AMR data are scarce.
A global metagenomic survey of sewage from 79 sites in 60 countries revealed strong associations between the abundance and diversity of ARGs and socioeconomic, health, and environmental conditions, including sanitation infrastructure and access to clean water (3). Another study analyzing 757 sewage samples from 243 cities across 101 countries revealed marked regional variations in resistome profiles (9). Notably, 49 common ARGs were detected across diverse genetic environments, including plasmids, suggesting both localized circulation and potential for global dissemination. These data support targeted public health interventions and optimize resource allocation for antimicrobial stewardship.
Given this potential, WBS is especially relevant in regions where AMR is intensified by inadequate sanitation, high levels of antibiotic pollution, and limited access to healthcare infrastructure (5,6,10). However, implementation remains challenging in many LMICs, where non-sewered sanitation systems, including latrines, septic tanks, and open drains, complicate systematic sampling and may introduce biases (10,11).
In addition to surveillance, WBS can actively guide local interventions. Spatiotem poral mapping of ARG abundance can identify environmental AMR hotspots, support ing targeted wastewater treatment improvements, such as upgrading disinfection processes, optimizing biological treatment to reduce ARG loads, or implementing advanced oxidation technologies, alongside hygiene promotion initiatives (12). At the hospital level, outbreaks may involve MGEs that transfer several ARGs across different pathogens in sink drains or wastewater pipes, creating hotspots of resistance that can guide prescribing and hygiene measures (13). Detecting such events could help physicians adjust empirical prescribing and alert hygiene teams to mitigate localized hotspots of resistance. Moreover, when integrated with data on pharmaceutical residues and pathogen loads, WBS enables comprehensive assessments of how antibiotic use influences the emergence and spread of AMR across diverse ecological and socioeco nomic settings (14,15).
The adaptability of WBS across spatial scales, from individual buildings to entire cities, makes it a key tool for One Health surveillance, which integrates human, animal, and environmental health (16). Monitoring wastewater from hospitals, dermatology clinics, livestock operations, aquaculture systems, and pharmaceutical manufacturing facilities alongside municipal sewage therefore offers a comprehensive view of environmental AMR pressures (5). This integrated perspective is especially relevant as ARGs circulate among human, animal, and environmental reservoirs, particularly in regions experienc ing rapid urbanization and agricultural intensification (17). To fully realize this potential, advances in sequencing technologies are essential. Metagenomics provides cultureindependent insights into the resistome, but challenges include high detection limits and the difficulty of assigning ARGs to hosts (3,6,18). Harmonization of pipelines is actively being developed (19). Short-read metagenomics continues to be valuable for large-scale mapping of resistome diversity and comparative analysis across regions, although assemblies can be incomplete and may limit the ability to resolve the genomic environment of ARGs. Long-read sequencing offers greater resolution by linking ARGs to hosts or plasmids, enhancing the study of co-selection processes and improving recovery of MGEs (19). Complementary approaches such as targeted whole-genome sequencing of priority ESKAPEE pathogens (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, Enterobacter spp., and Escherichia coli), are also essential for identifying emerging multidrug-resistant lineages. Wastewater composition is highly variable across catchments, with inputs from hospitals, farms, industrial facilities, or municipal sources, and systematic metadata collection is critical for robust interpretation (15). Building on these technological and methodologi cal advances, wastewater-based epidemiology (WBE) has emerged as a key application of WBS data, providing insights into population-level trends (20,21). For AMR, however, WBS also informs on ecological and genetic processes that extend beyond epidemiologi cal estimates.
Strengthening the role of WBS in AMR surveillance will require long-term investment and international coordination to ensure standardization, data sharing, and robust ethical frameworks (21,22). Although sequencing data are typically filtered, personal information can still appear in raw data sets, especially in small catchments. Transpar ent governance is essential to minimize risks of stigmatization. Along with expand ing infrastructure in low-resource settings, global efforts should prioritize integrating WBS with clinical and environmental data streams to deliver actionable insights into resistance dynamics. With such support, WBS can become an integral component of One Health surveillance (23), bridging critical data gaps and reinforcing coordinated responses at both local and global levels.
## OPPORTUNITIES AND RISKS: ETHICS, EQUITY, AND INFRASTRUCTURE
Integrating WBS into AMR surveillance presents major opportunities, but also technical, ethical, and infrastructural challenges that must be addressed to ensure responsible and equitable implementation globally.
## TRANSFORMATIVE PRIORITIES: FUNDING, INTEGRATION, AND POLICY FRAMEWORKS
To unlock the full potential of WBS for AMR surveillance, coordinated changes are needed in funding, institutional integration, ethical governance, and technological infrastructure.
First, funding mechanisms and institutional structures should position WBS as an integral part of AMR monitoring. This includes support from international financing programs such as the Global AMR Innovation Fund and environmental health initiatives, while also prioritizing pilot projects in low-resource settings. Demonstrating feasibil ity and cost-effectiveness in such settings is fundamental to support broader policy adoption and ensure long-term sustainability (27). Efforts should also include awareness and education programs to strengthen public understanding of AMR and build local capacity for sustained surveillance.
Second, national surveillance frameworks should formally incorporate WBS as a complementary source of AMR data. Integration into systems, such as the Global Antimicrobial Resistance and Use Surveillance System (GLASS), is essential (28). Establishing sentinel sites at strategically selected urban centers and high-risk facili ties would enable more representative monitoring. Regular reporting of WBS-derived AMR indicators, aligned with clinical and agricultural data sets, would strengthen data triangulation and improve comparability across sectors. In developing countries, international support should prioritize the creation of locally adapted guidelines and capacity-building programs to ensure feasible and equitable implementation.
Third, robust ethical and governance mechanisms are essential. International bodies, such as the World Health Organization (WHO) and the United Nations Environment Program (UNEP), should lead the development of global guidelines that address privacy protections, consent procedures for catchment-level monitoring, data stewardship, and standards for public reporting. At the local level, culturally sensitive, community-driven consent models can help build public trust and mitigate the risk of stigmatization associated with environmental surveillance.
Fourth, international standardization is crucial to ensure the reliability and intero perability of WBS data. Clear International Organization for Standardization (ISO)-like protocols should guide sampling, processing, sequencing, and interpretation. Openaccess, FAIR-compliant pipelines that integrate with resistance gene databases, such as the Comprehensive Antibiotic Resistance Database (CARD) (29) or ResFinder (30), would facilitate cross-jurisdictional data use.
Fifth, targeted investment in technological innovation is necessary to ensure the feasibility of WBS across diverse global settings. This includes supporting the develop ment of portable, low-cost molecular diagnostic platforms, such as digital PCR, for the rapid detection of ARGs in resource-limited environments. Innovations in treatment processes are equally important, such as solar-driven advanced oxidation or tailored disinfection methods, which can reduce the environmental persistence and spread of ARGs (31).
Finally, the One Health approach, integrating human, animal, and environmental health, should guide all coordination efforts. This approach is promoted by WHO and recognized in G7 and G20 agendas, as it links AMR indicators across sectors and strengthens policy coherence. Cross-sectoral collaboration involving public health, veterinary medicine, agriculture, and environmental protection is essential for the effective interpretation and application of WBS data. Indicators derived from WBS may serve as real-time metrics to assess the impact of AMR policies, including limits on antibiotic discharge, improvements in sanitation infrastructure, and public awareness campaigns. Clear guidelines on the safe disposal of unused or excess antibiotics from hospitals, clinics, and households are also critical to limit environmental contamination and downstream selection pressures.
The emergence of initiatives such as the U.S. National Wastewater Surveillance System (NWSS) highlights the feasibility of large-scale AMR monitoring through WBS (32). However, global adoption will ultimately depend on sustained political commitment, regulatory alignment, and explicit integration of WBS into multilateral health financing agendas. Together, these priorities highlight that WBS is not only a technical innovation but also a societal and policy challenge, requiring global collaboration to translate surveillance data into effective action against AMR.
## CONCLUSIONS
WBS has evolved rapidly from a pandemic-era innovation into a robust tool for AMR surveillance. Aligning advances in sewage monitoring with ethical frameworks, policy integration, and equitable infrastructure development offers a path to detect and mitigate emerging resistance before the next crisis. Here, emerging resistance includes novel ARGs not yet reported clinically, the transfer of known ARGs into new pathogenic hosts, and their introduction into previously unaffected regions. With the scientific evidence and technical capacity already available, the next step is to integrate WBS within international AMR strategies and long-term response planning.
## References
1. Naghavi, Vollset, Ikuta et al. (2024) "Global burden of bacterial antimicrobial resistance 1990-2021: a systematic analysis with forecasts to 2050" *Lancet*
2. Lambrou, South, Ballou et al. (2023) "Early detection and surveillance of the SARS-CoV-2 variant BA.2.86 -worldwide" *Morbid Mortal Wkly Rep*
3. Hendriksen, Munk, Njage et al. (2019) "Global monitoring of antimicrobial resistance based on metagenomics analyses of urban sewage" *Nat Commun*
4. Hart, Warren, Wilkinson et al. (2023) "Environmental surveillance of antimicrobial resistance (AMR), perspectives from a Perspective mBio November"
5. "national environmental regulator in 2023" *Euro Surveill*
6. Larsson, Flach (2022) "Antibiotic resistance in the environment" *Nat Rev Microbiol*
8. Pruden, Vikesland, Davis et al. (2021) "Seizing the moment: now is the time for integrated global surveillance of antimicrobial resistance in wastewater environments" *Curr Opin Microbiol*
9. Partridge, Kwong, Firth et al. (2018) "Mobile genetic elements associated with antimicrobial resistance" *Clin Microbiol Rev*
10. Singer, Thompson, Filho et al. (2023) "A world of wastewater-based epidemiology" *Nat Water*
11. Munk, Brinch, Møller et al. "Global Sewage Surveillance Consortium. 2022. Genomic analysis of sewage from 101 countries reveals global landscape of antimicrobial resistance" *Nat Commun*
12. Iskandar, Molinier, Hallit et al. (2021) "Surveillance of antimicrobial resistance in low-and middle-income countries: a scattered picture" *Antimicrob Resist Infect Control*
13. Balcázar (2025) "Wastewater-based epidemiology as a complementary tool for antimicrobial resistance surveillance: overcoming barriers to integration" *Bioessays*
14. Mosaka, Unuofin, Daramola et al. (2022) "Inactivation of antibiotic-resistant bacteria and antibiotic-resistance genes in wastewater streams: current challenges and future perspec tives" *Front Microbiol*
15. Constantinides, Chau, Quan et al. (2020) "Genomic surveillance of Escherichia coli and Klebsiella spp. in hospital sink drains and patients" *Microb Genom*
16. Prieto Riquelme, Garner, Gupta et al. (2022) "Demonstrating a comprehensive wastewater-based surveillance approach that differentiates globally sourced resistomes" *Environ Sci Technol*
17. Sims, Kasprzyk-Hordern (2020) "Future perspectives of wastewaterbased epidemiology: monitoring infectious disease spread and resistance to the community level" *Environ Int*
18. Musicha, Morse, Cocker et al. (2024) "Time to define One Health approaches to tackling antimicrobial resistance" *Nat Commun*
19. Martak, Henriot, Hocquet (2024) "Environment, animals, and food as reservoirs of antibiotic-resistant bacteria for humans: One Health or more?" *Infect Dis Now*
20. Kim, Cha (2021) "Antibiotic resistome from the One-Health perspective: understanding and controlling antimicrobial resistance transmission" *Exp Mol Med*
21. Sherry, Lee, Giulieri et al. (2025) "Genomics for antimicrobial resistance-progress and future directions" *Antimicrob Agents Chemother*
22. Clarke, Brien, Murray et al. (2024) "A review of wastewater-based epidemiology for antimicrobial resistance surveillance" *J Environ Expo Assess*
23. Conforti, Pruden, Acosta et al. (2025) "Strength ening policy relevance of wastewater-based surveillance for antimicrobial resistance" *Environ Sci Technol*
24. Miłobedzka, Ferreira, Vaz-Moreira et al. (2022) "Monitoring antibiotic resistance genes in wastewater environments: the challenges of filling a gap in the One-Health cycle" *J Hazard Mater*
25. Punch, Azani, Ellison et al. (2014) "The surveillance of antimicrobial resistance in wastewater from a one health perspective: a global scoping and temporal review" *One Health*
26. Ahmed, Brien, Keshaviah et al. (2023) "Wastewater-based monitoring could help guide responses to the USA opioid epidemic" *Nat Water*
27. Gholipour, Shamsizadeh, Halabowski et al. (2024) "Combating antibiotic resistance using wastewater surveillance: significance, applications, challenges, and future directions" *Sci Total Environ*
28. Khan, Wurzbacher, Uchaikina et al. (2025) "A perspective on wastewater and environmental surveillance as a public health tool for low-and middle-income countries" *Microorgan isms*
29. Chau, Barker, Budgell et al. (2022) "Systematic review of wastewater surveillance of antimicrobial resistance in human populations" *Environ Int*
30. Ajulo, Awosile (2024) "Global antimicrobial resistance and use surveillance system (GLASS 2022): investigating the relationship between antimicrobial resistance and antimicrobial consumption data across the participating countries" *PLoS One*
31. Alcock, Huynh, Chalil et al. (2023) "CARD 2023: expanded curation, support for machine learning, and resistome prediction at the Comprehensive Antibiotic Resistance Database" *Nucleic Acids Res*
32. Florensa, Kaas, Clausen et al. (2022) "ResFinder -an open online resource for identification of antimicrobial resistance genes in next-generation sequencing data and prediction of phenotypes from genotypes" *Microb Genom*
33. Starling, Neto Rp De, Pires et al. (2021) "Combat of antimicrobial resistance in municipal wastewater treatment plant effluent via solar advanced oxidation processes: achievements and perspectives" *Sci Total Environ*
34. Adams, Bias, Welsh et al. (2024) "The National Wastewater Surveillance System (NWSS): from inception to widespread coverage" *Sci Total Environ* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282987&blobtype=pdf | # Expanding the offer of Vaccinology Education with the Master of Science in Immunology and Vaccinology
Paulo Bettencourt
## Abstract
The COVID-19 pandemic revealed the urgency for well-trained professionals in vaccinology. With the purpose of building capacity in the field, the Catholic University of Portugal created the Masters in Immunology and Vaccinology. Here, I describe this innovative program and compare it to the current Vaccinology programs worldwide to reveal its uniqueness, and how it will contribute to reduce the burden of disease, improve disease prevention and promote healthier lives globally.
immune response, in health and diseases, including the usage of bioinformatics tools, artificial intelligence, and machine learning. Students will learn to apply this knowledge to the design, development, and preclinical evaluation of vaccines. Scale-up for industrial production, clinical trials, regulatory affairs and licensing, distribution of vaccines, and the Public Health issues associated with vaccination, particularly accessibility to vaccines, and vaccine hesitancy will also be addressed. In this way, students will gain a comprehensive understanding of fundamental, translational, and applied science, in parallel with the practical requirements for successful vaccine implementation in the market, resulting in individuals with a critical mind-set to initiate scientific research and development activities in immunology and vaccinology. 6 The Master of Science degree adopts a model that combines fundamental theoretical knowledge with the practical and laboratory skills needed to carry out immunology and vaccinology studies. A unique characteristic of this program, compared to others, shown in Table 1, is the strong focus in immunology as the basic knowledge to understand the principles required to improve current vaccines and design new and more effective ones. The first year of the Master's degree is divided into two semesters. In the first, students will take four general Curricular Units (CU) that form the structuring component and one CU of Introduction to Vaccinology.
Data and Biostatistics develops students' ability to independently apply statistics and data analysis tools, including methods for data collection, organization, visualization, and both exploratory and inferential analyses. In the second semester, five specialized CUs are offered, focused on acquiring in-depth knowledge in relevant areas of research. Students will participate in practical laboratory sessions to familiarize with basic methods of immunology and vaccinology in the CU Laboratory of Immunology and Vaccine Immunogenicity Analysis, including 1) Laboratory Techniques in Virology; 2) PCR analysis to quantify viral vector yields; 3) Host-Pathogen Interactions: Infection of macrophages with Mycobacterium bovis BCG and quantification of infection by Colony Forming Units assay; 4) Innate Immunology: Analysis of macrophage gene expression after activation by PAMPs; 5) Adaptive Immunology: Analysis of immune cell populations by flow cytometry; 6) Vaccine-induced immunogenicity (Antibodies): Quantification of antibodies by Enzyme-Linked Immunosorbent Assay (ELISA); 7) Vaccine-induced immunogenicity -(T Cells): Quantification of cellular immunity by Interferon-gamma Enzyme-linked immunosorbent spot (ELISpot).
A particular relevant course, Translational Vaccinology, trains students in in silico vaccine design, including the selection of a suitable vaccine platform for a particular target. Students will perform database sequence mining, using available bioinformatics tools to identify regions of interest, and determine the most appropriate vaccine types for generating protective responses. Real-world challenges and lessons from the COVID-19 pandemic and similar relevant examples will be used to illustrate applied concepts.
The objectives of the course Contemporary Epidemiology and Global Health are to introduce students to a holistic perspective on current global challenges and computation-based approaches to pathogen transmission, surveillance and control. The goal is to strengthen understanding of the methodological and research landscapes of contemporary epidemiology and global health.
The course Virology Research for the Future offers both theoretical and practical methods in virology, including mechanisms of surveillance and the dynamics of viral transmission, the impact of antivirals and vaccines, and practical computing skills in virology research.
The Dissertation Project on the second semester consists of the preparation of the thesis work to be carried out during the second year.
The second year is fully dedicated to thesis research that can be conducted in an academic setting at the UCP or in an industrial setting. A distinctive feature of this degree is the opportunity for selected students to conduct their thesis work at the pharmaceutical company Zendal. This collaboration allows students to develop their research in an industrial setting, under the joint supervision of a Zendal mentor and a UCP faculty member.
Upon graduation, students will be individuals with scientific independence from the perspective of planning, executing, and analyzing research in Vaccinology, as well as producing and formally presenting scientific data. 6 Graduates in Immunology and vaccinology will be encouraged to pursue further studies through the PhD Program in Medical Sciences at the UCP, continuing their academic development in the field.
Vaccinology is a timely topic, and a highly beneficial Public Health instrument to combat infectious and noninfectious diseases, and the only available weapon to prevent emerging pandemics in the future. The purpose of the MSc in Immunology and Vaccinology is to contribute to reducing the knowledge gap in Vaccinology, and at the same time, train the much-needed specialized technicians and scientists. The future graduates will work either in academia or industry, contributing actively to improve disease prevention. This will eventually have an impact on the reduction of the burden of disease and consequently promote healthier lives globally.
## Disclosure statement
No potential conflict of interest was reported by the author(s).
## References
1. Shattock, Johnson, Sim et al. (2024) "Contribution of vaccination to improved survival and health: modelling 50 years of the expanded programme on immunization" *Lancet*
2. Watson, Barnsley, Toor et al. (2022) "Global impact of the first year of COVID-19 vaccination: a mathematical modelling study" *Lancet Infect Dis*
3. Bussink-Voorend, Hautvast, Vandeberg et al. (2022) "A systematic literature review to clarify the concept of vaccine hesitancy" *Nat Hum Behaviour*
4. Dochez, Duclos, Macdonald et al. (2022) "Advanced vaccinology training globally: update and impact of the COVID-19 crisis" *Vaccine*
5. Poland, Levine, Clemens (2010) "Developing the next generation of vaccinologists" *Vaccine*
6. (2024) "Master of Science in Immunology and Vaccinology" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724129&blobtype=pdf | # Identification and characterization of a novel plaque-invisible lytic single-stranded RNA phage
Yuer Wang, Fengjuan Tian, Jinbei Zhang, Shan Xu, Mengzhe Li, Yigang Tong
## Abstract
The RNA phages offer promising applications in biotechnology, including vaccine development and drug delivery. However, their potential remains underex plored due to the limited number of known RNA phages, partly because conventional methods fail to identify plaque-invisible lytic phages that do not form plaques. Here, we introduced a novel method that combines RNA-inclusive metagenomic studies and quantitative reverse transcription-PCR (RMS-RT-qPCR) to identify and characterize active RNA phages from environmental samples. This study led to the discovery of a new active Qbeta-like phage, named Cute. Genomic analysis revealed that Cute is a new member of the Qubevirus genus. Although Cute does not form plaques, it can be observed to continuously release into the supernatant when co-cultured with the host by RT-qPCR detection. This discovery underscores the potential diversity of RNA phages in nature and the limitations of traditional culture-dependent techniques. Our findings suggest that RMS-RT-qPCR could aid in the discovery of active RNA phages with significant biotechnological applications.
IMPORTANCEThe discovery and characterization of RNA phages might be historically constrained by traditional culture-based methods. Our study provides a powerful tool for identifying active RNA phages by combining RNA-inclusive metagenomic analysis with RT-qPCR. This method expands our understanding of the diversity and ecological roles of RNA phages, which are often overlooked in microbiome studies. This research highlights the importance of RNA phages in natural ecosystems and their potential applications in biotechnology and medicine, such as antimicrobial therapies and vaccine development. By expanding our understanding of RNA phage diversity, this study opens new avenues for their utilization in various fields, emphasizing the need for continued exploration of these versatile biological entities.
smaller, which may reflect not only their biological features but also methodological and other limiting factors in their discovery.
The double-layer plate method is the conventional method for identifying phages (12). However, it is ineffective for detecting plaque-invisible lytic RNA phages, as some do not form plaques on agar plates. The traditional double-layer plate method may hinder the wide discovery of RNA phages. It is necessary to supplement or replace the traditional methods with other approaches. Although some studies have isolated the RNA phage phiNY that does not form plaque through CF11 cellulose (13), CF11 cellulose is only suitable for extracting dsRNA (only dsRNA can bind to cellulose) (14). Therefore, a more universal method for identifying RNA phages should be developed.
Metagenome sequencing based on next-generation sequencing (NGS) technology has been extensively employed for detecting pathogens (15) and drug-resistant genes (16). This technology is particularly advantageous for identifying novel viruses, sig nificantly expanding the known viral repertoire (17)(18)(19). In 2016, Krishnamurthy et al. identified more than 120 novel RNA phages by analyzing existing metagenomic databases, which previously contained only 11 ssRNA and five dsRNA phage genome sequences in the NCBI (20). In 2020, Wolf et al. identified more than 4,500 different RNA viruses based on metagenomic analysis of water samples (21). In addition, Callanan et al. identified complete genome sequences of more than 1,000 novel ssRNA phages in publicly available metatranscriptomic databases (11). These studies demonstrated that only a limited portion of RNA phages present in the environment has been characterized. These studies have significantly enriched the known diversity of RNA phages, uncovering several structural domains that had not been previously reported in these phages, thereby facilitating evolutionary analysis and the exploration of gene structure-function relationships. However, the analysis is limited to the genetic level due to the lack of active phages (22). Obtaining active phages is necessary for a more in-depth investigation of their evolutionary relationships with hosts and for exploring their practical applications.
In our study, we combined RNA-inclusive metagenomic studies with quantitative reverse transcription-PCR (RMS-RT-qPCR) to identify and culture a novel RNA phage, determine its host, and monitor its lifecycle through RT-qPCR.
## RESULTS AND DISCUSSION
## Metagenomic analysis and identification of ssRNA phage sequence
Using the RMS-RT-qPCR method (Fig. 1), we identified a novel lytic ssRNA phage, named Cute, from the metagenomic data of the soil sample. Since it is challenging to identify RNA phages by analyzing metagenomic data directly from environmental samples due to their low abundance (11), we addressed this issue by increasing phage abundance within samples through adding pure cultures of different bacteria and incubating them overnight before constructing the metagenomic library.
We initially processed raw sequencing data from the metagenomic library construc ted on the Illumina platform, yielding a total of 70,391,221 reads with an average length of 150 bp and an average GC content of 62%. After quality control and assembly using SPAdes, we identified 473 phage contigs using VirSorter2 and checkV for viral sequence identification. Of these, 389 contigs were classified into seven different viral classes. We calculated the transcripts per million for each contig, summed them by category, and then calculated the relative abundance ratios for each classification: Caudoviricetes (82.91%), Leviviricetes (17.01%), Faserviricetes (0.06%), Papovaviricetes (0.004%), Alsuviri cetes (0.002%), and Tolucaviricetes (0.001%) (Fig. 2). Notably, the significant presence of Leviviricetes suggests a potential abundance of ssRNA phages in this sample. To further verify our hypothesis and focus on RNA viruses, we reanalyzed the sequences using Virsorter2 for RNA virus detection and aligned the reads against the NCBI non-redundant nt database using BLASTn. This reanalysis identified a sequence with 74.91% homol ogy (Query Cover 79%) with Enterobacteria phage Qbeta (NC_001890.1) and 95.49% homology (Query Cover 99%) with Escherichia phage Qbeta GIII_C RNA (LC710219.1).
Based on this high degree of sequence similarity, we hypothesize the presence of a novel ssRNA phage in our metagenomic data set.
## Genomic analysis of phage Cute
In our analysis of the novel phage, named Cute, we identified a genome spanning 4,387 nt with a G+C content of 48%. The genome architecture of Cute is linear, containing three open reading frames (ORFs) and four conserved structural domains. Specifically, the ORFs encode the A2 maturation protein, the A1 minor capsid protein, and the RNA replicase (Fig. 3A). The conserved domains are characterized as the Phage_mat-A superfamily, the Levivirus coat protein, the Read-through superfamily, and the RNA_rep licase_B superfamily. Phylogenetic analysis based on RNA replicase amino acid (aa) sequences positions the phage Cute within the genus Qubevirus in the family Fiersviridae (Fig. 3B). The phage Cute clustered in a well-supported clade, including the Escheri chia phage Qbeta (GenBank accession no. NC_001890.1 ) and the Escherichia phage FrHibiscus (GenBank accession no. PP430143.1). The translational exception of the phage Cute at nucleotides 1750-1752, where the stop codon TGA is occasionally bypassed, allows for continued translation. As previously reported, this mechanism results in the production of the A1 minor capsid protein, albeit in low quantities (about 3-10 copies per virus particle) (23). This discovery of Cute expands our understanding of RNA phage genome organization and highlights the diversity within the Fiersviridae family, offering new insights into the genetic and functional complexities of ssRNA phages.
## Morphological characterization and biological characteristics of phage Cute
We used the specific primer of phage Cute A1 gene to detect the Ct value of phage Cute at the initial time (blue column, 0 min) and in the supernatant after co-culturing with the original host Klebsiella pneumoniae (K. pneumoniae) 100, 37°C for 200 min (red column, 200 min) by RT-qPCR method. The results showed that the Ct value decreased significantly after co-culture with the original host K. pneumoniae 100 (Fig. 4B), indicating that although the phage Cute did not form plaques (Fig. 4A), it could be released into the supernatant after co-culture with the host. Transmission electron microscopy (TEM) shows that the phage Cute is a tailless phage with a diameter of approximately 28 nm (Fig. 5A), which resembles the morphology of the previously described (1). After negative staining, phage Cute appears as uniformly sized bright round particles. Due to its icosahedron structure, it may present a hexagonal appearance at certain angles. The growth experiment showed a latent period of 60 min and a burst period of 100 min, followed by the plateau phase after 160 min (Fig. 5B).
## Host range analysis
Since phage Cute does not form any plaques on these strains, we, respectively, used the machine learning software vpf-class (24,25) and the CRISPR-based analysis method (26,27) to predict the hosts of phage Cute, but both failed. Predicting RNA phages using bioinformatics remains a challenge (28,29). To overcome these limitations, we employed RT-qPCR to ascertain the host of phage Cute (30). We co-cultured the phage Cute with 13 strains of K. pneumoniae and 8 strains of Escherichia coli (E. coli) and subsequently performed RT-qPCR targeting the A1 gene of phage Cute. The results showed that several K. pneumoniae strains (e.g., 100, 370, 409, 510, 655, 760, 814, and 828) supported phage replication, as evidenced by Ct values after co-culture markedly reduced (Fig. 4C). In contrast, no comparable changes were observed with E. coli strains (Fig. 4D). Taken together, these results corroborate that K. pneumoniae serves as the host of phage Cute, with 8 out of 13 strains being hosts of phage Cute. Given that it has been shown that most Qubevirus phages with E. coli as the host, this would be the first evidence that the host range of Qubevirus phages has been extended to include K. pneumoniae.
## The TraA protein serves as the receptor of the phage Cute
Previous studies have shown that F-pilus is the attachment site of RNA phages R17 and Qbeta (31), and the A2 protein of the phage Qbeta is responsible for adsorption on F-pilus (2) (32,33). The F-pilus is composed of F-pilin. The F-pilin subunit is processed from a 121 aa propilin encoded by the traA gene. Previous studies have shown that four random mutations in the traA gene can affect the attachment sites of RNA phages on F-pilus (34). To investigate whether the infection of K. pneumoniae by phage Cute might be associated with the F-like conjugation plasmid pKpQIL, we sequenced the genomes The red color denotes the novel phage, Cute. of 13 K. pneumoniae strains. The results showed that only the original host K. pneumoniae 100 and seven other susceptible bacteria (i.e., strains that can be adsorbed and infected by phage Cute) have the F-like conjugation plasmid pKpQIL, which carries the traA gene. This observation suggests a potential relationship between the presence of TraA and phage Cute susceptibility. The aa sequences of the TraA protein of 13 strains of K. pneumoniae were aligned, and it was found that these sequences were highly similar (Fig. 6A). We discovered three conserved aa motifs from it, which may determine the attachment of the A2 protein of phage Cute (Fig. 6B).
Although the development of cryo-electron microscopy has contributed to the study of the interaction mechanism of ssRNA phages binding to their receptors (35), there are fewer studies on the specific molecular mechanism of the interaction between the A2 protein and TraA. Since the A2 maturation/lysis protein of phage Qbeta is responsible for adsorption and lysis, changes in the host range of the phage Cute may be related to alterations in the aa sequence of the A2 protein. Initially, we aligned the nucleotide and aa similarities between phage Cute A2 and Qbeta A2, which are 72.79% and 63.90%, respectively. To further explore the reasons, we aligned the aa sequence of the Cute A2 protein with other Enterobacteria phages in Qubevirus with E. coli as host (Fig. S1). The aa highlighted in red represents non-conserved aa residues in the phage Cute A2 sequence. We speculate that the changes in these aa positions are the key factors in the variation of the phage host range. However, the specific mechanism still requires further investigation.
## Exploring the potential application value of the phage Cute
The virus-like particles (VLPs) based on the phage Cute have a broad perspective in applications, such as bioimaging, drug delivery, and vaccines (36), due to their ability to self-assemble in vitro and display antigenic epitopes on the coat protein surface (37,38). The formation of functional VLPs depends on the specific binding of coat protein to the RNA operator. Therefore, predicting the RNA operators in the genome of phage Cute is of crucial importance for exploring the application potential of ssRNA phages.
The phage Qbeta has a similar structure in the RNA operators to the phage MS2 (Fig. 7A) (39). The RNA operators of both Qbeta and MS2 consist of an adenine-containing loop and a stem, which differ in the size of the loop and the length of the stem. The coat protein of MS2 recognizes the Qbeta RNA operator by replacing the glutamic acid in residue 89 with threonine or lysine (40). The key structure of the Qbeta RNA operator is a three-nucleotide loop and an eight-base-pair stem, with the last adenine in the loop necessary to occupy the adenine-binding pockets of the CP. Notably, we found that phage Cute shares the same RNA operator sequence (AUGCAUGUCUAAGACAGCAU) with phage Qbeta (Fig. 7B), which is located between 2,364 and 2,383 nt in the genome of phage Cute.
To confirm the accuracy of the phage Cute RNA operator sequence, we simulated the molecular docking between the three-dimensional structure of the phage Cute coat protein dimer and the RNA operator with HDOCK software (Fig. 7C). The results demonstrate that the binding site between the coat protein dimer and the RNA operator is located on the surface of the β-sheet, which is consistent with previous research findings (39). As reported previously, A+1 and A+7 bases are stacked together with the Tyr of the two monomers in the model of coat protein dimer and RNA operator, respectively (Fig. 8A andB) (41). The adenine base of A+8 nucleotides is suitable for the adenine binding bag formed by Ser, Gln, and Lys of the A chain in the capsid protein dimer (Fig. 8B). In summary, our predicted RNA operator demonstrates a strong ability to bind within the protein pocket, confirming the reliability of the predicted RNA operator sequence. To further confirm the accuracy of the RNA operator sequence of the phage Cute, we prepared Cute VLP using prokaryotic expression vectors (Fig. S2A). In addition, enhanced green fluorescent protein (EGFP) carrying RNA operator was used as the target sequence, and RT-qPCR was used to verify whether the Cute VLP was packaged with cargo RNA after extracting nucleic acid. Naked RNA samples extracted from Cute VLPs were treated without reverse transcriptase (the reverse transcriptase was replaced with RNase-free water) as a control (42). Compared with the unreverse transcription samples (non-RT), the Ct value of the reverse transcription (RT) was significantly decreased, which demonstrated that the target RNA EGFP was indeed encapsulated in Cute VLP (Fig. S2B).
In conclusion, we successfully developed an innovative approach for identifying and characterizing plaque-invisible lytic phage. Our method, RMS-RT-qPCR, combines RNA-inclusive metagenomic studies with RT-qPCR analysis, which not only facilitates the discovery of novel RNA phage genomes but also enables the identification of active phage. Utilizing this approach, we identified a new Qbeta-like phage from soil samples, and it belonged to the Qubevirus genus in the Fiersviridae family. Host range assays revealed that it could infect K. pneumoniae. Our findings can expand the known diversity of RNA phages and emphasize their potential ubiquity in the environment. This study paves the way for further exploration of RNA phage ecology and evolution, and discovered more RNA phages as candidate vehicles for vaccine delivery, gene therapy, and antimicrobial treatments.
## MATERIALS AND METHODS
## Sample collection and preprocessing
Soil samples were collected from an area near the 307 Hospital of the People's Liberation Army in Beijing, China. The 100 g of soil samples were then resuspended in 500 mL of phosphate-buffered saline (PBS) and centrifuged at 10,000 × g for 10 min. Next, 115 bacterial strains representing 18 different species (Table S1) in the Tong Lab bacterial library were selected as candidate hosts for RNA phages. The bacteria were, respectively, cultured in 10 mL shaking tubes to the logarithmic phase. Specifically, we added 50 µL of soil suspension per tube to 115 shaking tubes, each containing 5 mL of bacterial culture in the logarithmic phase. After overnight incubation, 50 µL from each culture was transferred to a clean centrifuge tube, resulting in a total of 5,750 µL of enriched mixture. Then, 200 µL of the enriched mixture was used to extract nucleic acids, and one library was constructed for subsequent NGS analysis. to 20 mL of filtrate to prepare a 30% wt/vol sucrose solution. Add the prepared sucrose buffer to a 38.5 mL Ultra-Clear centrifuge tube (Beckman Coulter). Centrifuge the tubes in a SW 32 Ti rotor (Beckman Coulter) at 35,000 × g for 2 h. Take the bottom solution and resuspend it with 200 µL PBS. The purified phage Cute solution was incubated on a carbon-coated copper grid for 10 min and then dried using filter paper. The sample was then negatively stained with uranyl acetate for 90 s and air-dried at room temperature. Phage morphology was subsequently observed by TEM (JEM-1400plus, Japan) at 120 kV (52).
## Growth curve experiment
Phage Cute was mixed with its host K. pneumoniae for 5 min, and the mixture was then centrifuged at 10,000 × g for 5 min to remove unadsorbed phages. The precipitate was resuspended with Luria-Bertani (LB) and cultured at 37°C with shaking at 200 rpm. The co-cultures were subsequently centrifuged at 10,000 × g for 10 min, and the supernatant was filtered through the 0.22 µm filters. RNA extraction, cDNA synthesis, and qPCR were performed as described in "Host range analysis, " below. The relative quantity curve of phage Cute was plotted using the 2 ΔCt method (53) by GraphPad Prism 9.
## Double-layer plate method
The double-layer plate method was experimentally performed as described previously (54). Briefly, the base layer of LB solid medium (1.5% agar) was first prepared. Then, 100 µL of the overnight cultured K. pneumoniae strain 100 and 100 µL of phage Cute (filtered through a 0.22 µm filter) were mixed with 5 mL of semi-solid LB medium (0.75% agar). This mixture was then overlaid onto the base layer and incubated overnight at 37°C.
## Host range analysis
The 50 µL of each bacterial strain and 50 µL of the phage Cute (through a 0.22 µM filter) were mixed and incubated for 5 min, followed by centrifugation at 10,000 × g for 5 min to remove unadsorbed phage. Resuspend the precipitate with 5 mL LB. At the initial moment, samples were taken and passed through the 0.22 µm filter. After co-culture at 37°C for 6 h, samples were taken again and passed through the 0.22 µm filter. Subsequently, the filtrate was lysed at 100°C for 5 min to release the RNA. HiScript II Q RT SuperMix for qPCR (+gDNA wiper) (Vazyme, China) was used to prepare cDNA. The qPCR primers were designed to detect the phage Cute A1 gene (Table S3). The qPCR was performed using Taq Pro Universal SYBR qPCR Master Mix (Vazyme, China) on the 7500 real-time PCR system.
## Cute VLP plasmid construction and expression system
The DNA sequences of the phage Cute's coat protein dimer with his tag, A2 gene, and the sequences of EGFP were synthesized by General Biol Co., Ltd., and the target sequences were constructed into the PACYC plasmid by homologous recombination. The obtained plasmids were verified by Sanger sequencing (General Biol Co., Ltd., China) and named coatdimer-EGFP-pac. The coatdimer-EGFP-pac plasmid was transformed into the BL21(DE3) strain as a prokaryotic expression system. BL21(DE3)-coatdimer-EGFP-pac was cultured at 37°C in LB (400 mL) supplemented with 50 µg/mL chloramphenicol. The expression was induced by adding 160 µL 1 M isopropyl-beta-D-thiogalactopyranoside (IPTG) at OD 600 = 0.8 at 16°C and 200 rpm for 24 h. The bacteria were precipitated by centrifugation (7,000 rpm, 10 min, 4°C), resuspended in 10 mL PBS, and centrifuged at 6,000 rpm for 10 min (4°C). The following were added to the bacterial sediment: 20 mL of Tris-HCl NaCl buffer containing 500 µL of 30% Triton X-100, 6 µL of 100 mg/mL RNase (Vazyme Biotech Co., Ltd.), 600 µL of 1 mg/mL DNase I (Solarbio), 3 mL of glycerol, and 300 µL of protease inhibitor (Selleckchem). The bacterial precipitate suspension containing VLPs was disrupted by ultrasonic homogenizer and then centrifuged in the ultracentrifuge (SW32 Ti; 38.6 mL tube capacity) at 4°C, 18,000 rpm for 20 min. The supernatant containing his-tagged VLPs was purified by Ni-NTA Agarose (QIAGEN). To remove the plasmids from the solution, we pre-incubated VLP solution with 150 U of DNase I at 37°C for 1 h. Subsequently, DNase I was thermally inactivated at 68°C for 10 min. Then, the impurity proteins were eluted successively with Solution A (30 mM imidazole) and Solution B (60 mM imidazole). Finally, VLP was eluted with the eluents (500 mM imidazole), and the purified VLP was stored at 4°C.
## References
1. Golmohammadi, Fridborg, Bundule et al. (1996) "The crystal structure of bacteriophage Q beta at 3.5 A resolution"
2. Newton, Zhao, Finn et al. (2023) "Diversity in Qβ Vvirus-like particle cage assembly via coat protein monomers and AYGG-linked dimers" *Microsc Microanal*
3. Fang, Ramos, Holguin et al. (2017) "Functional RNAs: combined assembly and packaging in VLPs" *Nucleic Acids Res*
4. Fiedler, Fishman, Brown et al. (2018) "Multifunctional enzyme packaging and catalysis in the Qβ protein nanoparticle" *Biomacromolecules*
5. Das, Zhao, Crooke et al. (2020) "Stabilization of near-infrared fluorescent proteins by packaging in virus-like particles" *Biomacromolecules*
6. Rhee, Hovlid, Fiedler et al. (2011) "Colorful virus-like particles: fluorescent protein packaging by the Qβ capsid" *Biomacromolecules*
7. Li, Sun, Jia et al. (2014) "Messenger RNA vaccine based on recombinant MS2 virus-like particles against prostate cancer" *Int J Cancer*
8. Benjamin, Chen, Brohlin et al. (2020) "Using FRET to measure the time it takes for a cell to destroy a virus" *Nanoscale*
9. Steinmetz, Hong, Spoerke et al. (2009) "Buckyballs meet viral nanoparticles: candidates for biomedicine" *J Am Chem Soc*
10. Trashi, Durbacz, Trashi et al. (2023) "Self-assembly of a fluorescent virus-like particle for imaging in tissues with high autofluorescence" *J Mater Chem B*
11. Callanan, Stockdale, Shkoporov et al. (2020) "Expansion of known ssRNA phage genomes: from tens to over a thousand" *Sci Adv*
12. Kamer, Abdelaziz, Nosair et al. (2022) "Characterization of newly isolated bacteriophage to control multi-drug resistant Pseudomonas aeruginosa colonizing incision wounds in a rat model: in vitro and in vivo approach" *Life Sci*
13. (2025) *Full-Length Text Journal of Virology*
14. Cai, Tian, Teng et al. (2021) "Cultivation of a lytic double-stranded RNA bacteriophage infecting Microvirgula aerodenitrificans reveals a mutualistic parasitic lifestyle" *J Virol*
15. Peyambari, Roossinck (2018) "Characterizing mycoviruses" *Methods Mol Biol*
16. Jiang, Lang, Duan et al. (2020) "Intestinal virome in patients with alcoholic hepatitis" *Hepatology*
17. Raza, Shin, Hur et al. (2022) "Higher abundance of core antimicrobial resistant genes in effluent from wastewater treatment plants" *Water Res*
18. Benler, Koonin (2021) "Fishing for phages in metagenomes: what do we catch, what do we miss?" *Curr Opin Virol*
19. Lu, Yang, Zhang et al. (2022) "Metagenomic analysis of viral community in the Yangtze River expands known eukaryotic and prokaryotic virus diversity in freshwater" *Virol Sin*
20. Prado, Shubo, Freitas et al. (2022) "Virome in roof-harvested rainwater of a densely urbanized lowincome region" *Sci Total Environ*
21. Krishnamurthy, Janowski, Zhao et al. (2016) "Hyperexpansion of RNA bacteriophage diversity" *PLoS Biol*
22. Wolf, Silas, Wang et al. (2020) "Doubling of the known set of RNA viruses by metagenomic analysis of an aquatic virome" *Nat Microbiol*
23. Coclet, Roux (2021) "Global overview and major challenges of host prediction methods for uncultivated phages" *Curr Opin Virol*
24. Rumnieks, Tars (2011) "Crystal structure of the read-through domain from bacteriophage Qβ A1 protein" *Protein Sci*
25. Pons, Paez-Espino, Riera et al. (2021) "VPF-class: taxonomic assignment and host prediction of uncultivated viruses based on viral protein families" *Bioinformatics*
26. Young, Rogers, Robertson (2020) "Predicting host taxonomic information from viral genomes: a comparison of feature representa tions" *PLoS Comput Biol*
27. Edwards, Mcnair, Faust et al. (2016) "Computational approaches to predict bacteriophage-host relationships" *FEMS Microbiol Rev*
28. Biswas, Staals, Morales et al. (2016) "CRISPRDetect: a flexible algorithm to define CRISPR arrays" *BMC Genomics*
29. Libbrecht, Noble (2015) "Machine learning applications in genetics and genomics" *Nat Rev Genet*
30. Ali, Mahas, Mahfouz (2018) "CRISPR/Cas13 as a tool for RNA interference" *Trends Plant Sci*
31. Schmidtke, Hickey, Wirbel et al. (2025) "The prototypic crAssphage is a linear phage-plasmid" *Cell Host Microbe*
32. Manchak, Anthony, Frost (2002) "Mutational analysis of F-pilin reveals domains for pilus assembly, phage infection and DNA transfer" *Mol Microbiol*
33. Woody, Cliver (1995) "Effects of temperature and host cell growth phase on replication of F-specific RNA coliphage Q beta" *Appl Environ Microbiol*
34. Woody, Cliver (1997) "Replication of coliphage Q beta as affected by host cell number, nutrition, competition from insusceptible cells and non-FRNA coliphages" *J Appl Microbiol*
35. Frost, Paranchych (1988) "DNA sequence analysis of point mutations in traA, the F pilin gene, reveal two domains involved in F-specific bacteriophage attachment" *Mol Gen Genet*
36. Gorzelnik, Zhang (2021) "Cryo-EM reveals infection steps of singlestranded RNA bacteriophages" *Prog Biophys Mol Biol*
37. Jelínková, Jhun, Eaton et al. (2021) "An epitope-based malaria vaccine targeting the junctional region of circumsporozoite protein" *NPJ Vaccines*
38. Pang, Huang, Chen et al. (2023) "Bioengineered bacteriophage-like nanoparticles as RNAi therapeutics to enhance radiotherapy against glioblastomas" *ACS Nano*
39. Ortega-Rivera, Shin, Chen et al. (2021) "Trivalent subunit vaccine candidates for COVID-19 and their delivery devices" *J Am Chem Soc*
40. Lim, Spingola, Peabody (1996) "The RNA-binding site of bacteriophage Qbeta coat protein" *J Biol Chem*
41. Spingola, Peabody (1997) "MS2 coat protein mutants which bind Qbeta RNA" *Nucleic Acids Res*
42. Rumnieks, Tars (2014) "Crystal structure of the bacteriophage Qβ coat protein in complex with the RNA operator of the replicase gene" *J Mol Biol*
43. Qi, Zhang, Wang et al. (2023) "One-plasmid double-expression system for preparation of MS2 virus-like particles packaging SARS-CoV-2 RNA" *Front Cell Infect Microbiol*
44. Mu, Fei, Xu et al. (2025) "RASP v2.0: an updated atlas for RNA structure probing data" *Nucleic Acids Res*
45. Ru, Zha, Gao (2025) "Incorporating neighboring protein features for enhanced drug-target interaction prediction: a comparative analysis of similarity-based alignment methods" *J Chem Inf Model*
46. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: a fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol Biol Evol*
47. Zhaoyang, Guowei, Jing et al. (2022) "Clinical characteristics of the host DNAremoved metagenomic next-generation sequencing technology for detecting SARS-CoV-2, revealing host local immune signaling and assisting genomic epidemiology" *Front Immunol*
48. Church, Extavour (2022) "Phylotranscriptomics reveals discordance in the phylogeny of Hawaiian Drosophila and Scaptomyza (Diptera: Drosophilidae)" *Mol Biol Evol*
49. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-MODEL: homology modelling of protein structures and complexes" *Nucleic Acids Res*
50. Yan, Zhang, Zhou et al. (2017) "HDOCK: a web server for protein-protein and protein-DNA/RNA docking based on a hybrid strategy" *Nucleic Acids Res*
51. Martelli, Macera, Spezia et al. (2018) "Torquetenovirus detection in exosomes enriched vesicles circulating in human plasma samples" *Virol J*
52. Regnard, Rybicki, Hitzeroth (2017) "Recombinant expression of beak and feather disease virus capsid protein and assembly of virus-like particles in Nicotiana benthamiana" *Virol J*
53. Liu, Zheng, Quan et al. (2022) "Biological characteristics and genomic analysis of a novel Vibrio parahaemolyticus phage phiTY18 isolated from the coastal water of Xiamen China" *Front Full-Length Text Journal of Virology*
54. *Cell Infect Microbiol*
55. Lin, Li, Sun et al. (2020) "A novel freshwater cyanophage vB_MelS-Me-ZS1 infecting bloom-forming cyanobacterium Microcystis elabens" *Mol Biol Rep*
56. Mao, Yang, Yan et al. (2024) "Proteomic analysis of carbapenem-resistant Klebsiella pneumoniae outer membrane vesicles under the action of phages combined with tigecycline" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12656789&blobtype=pdf | # Effect of Additional Aluminum Filtration on the Image Quality in Cone Beam Computed Tomographic Studies of Equine Distal Limbs Using Visual Grading Characteristics Analysis: A Pilot Study
Micaela Sgorbini, Massimo Vignoli, Luca Papini, Mathieu De Preux, Frederik Pauwels, Joris Missotten, Elke Van Der Vekens
## Abstract
This pilot study aimed to improve the image quality in cone beam computed tomography (CBCT) for the diagnosis of fractures in equine distal limbs. Specifically, it evaluated the effects of varying tube current and aluminum filter thickness on image quality, including reductions in artifacts. Equine cadaver limbs were scanned using a mobile CBCT unit (O-arm ® ), with tube currents ranging from 10 to 100 mA and aluminum filters of 13 and 25 mm thickness. A board-certified veterinary radiologist and a board-certified equine surgeon, both blinded to the applied exposure and filters, evaluated the reconstructed images, focusing on the distinctness of anatomical structures and the presence of imaging artifacts. The results showed that higher tube currents generally improved image quality by reducing noise and enhancing the visibility of cortical and trabecular bone structures. However, overexposure was present at the periphery of the cortices at the highest tube current. For the metacarpal and metatarsal bones, the highest image quality was achieved without filters, particularly at tube currents of 50 or 64 mA. In these regions, the application of filters actually reduced image quality, likely due to diminished X-ray penetration through thick cortical bone. In contrast, for the proximal phalanx (P1), the use of thicker aluminum filters (19-25 mm) at 50 mA enhanced image quality and reduced artifacts. These findings provide preliminary recommendations for optimizing CBCT protocols in equine distal limb imaging using the O-arm ® . Moderate tube currents (50-64 mA) without filters were ideal for scanning the metacarpal and metatarsal bones, while thick aluminum filters used at 50 mA yielded superior results for P1. These suggested settings balance diagnostic quality with radiation safety, representing a first step toward more accurate detection of subtle bone lesions and fractures in long bones in the distal limb, and aim to contribute to earlier diagnosis and improved clinical outcomes in equine orthopedic practice.
Abstract(1) Background: Cone beam computed tomography (CBCT) is increasingly used in equine practice to diagnose musculoskeletal injuries, including fractures in the distal limb. However, limited detail in the thick cortical bone of the metacarpus/metatarsus hinders accurate diagnosis. In human medicine, the addition of aluminum filters (AF) enhanced
## 1. Introduction
Computed tomography (CT) and cone beam CT (CBCT) are cross-sectional diagnostic imaging techniques that have been increasingly used in clinical equine practice in recent years [1,2]. Their application has been described for the identification of pathological changes in the equine skull, specifically lesions affecting osseous structures, teeth, nerves, and peripheral soft tissues, and for lameness work-ups revealing musculoskeletal injuries [3][4][5][6][7]. The higher contrast resolution of CT images compared to radiographic images, and the elimination of superimposition, provide CT with significant diagnostic value in the diagnosis of fractures and fissures occurring at the level of the equine distal limbs and the equine head [8]. However, at the Vetsuisse Faculty of the University of Bern, it was observed that osseous fissures can go undiagnosed in the acute stage on CBCT images acquired with the O-arm ® imaging unit (Medtronic, Louisville, CO, USA). This is due to blooming and/or insufficient spatial resolution in the thick cortex of long bones, specifically in the metacarpus/metatarsus, which can have a detrimental effect on the diagnosis and treatment of clinical cases. Pre-filtering of the X-ray beam as a tool to improve image quality and decrease patient dose has been mainly studied in humans, specifically in pediatric medicine [9][10][11]. This process, known as "beam hardening", uses filters to absorb low-energy X-rays, which reduces beam hardening artifacts and patient dose. The thickness and material of the filter influence the degree of filtration and the effect on image quality; primarily, aluminum and copper filters have been evaluated [9, 12,13]. A possible disadvantage of excessive filtration is an increase in statistical noise due to the reduced number of photons available for image formation and the associated decrease in image quality [9]. Previous studies have evaluated the effect of exposure values on the image quality of equine head CT [14]. However, to the best of the authors' knowledge, the potential benefits of added filtration and different tube currents have not yet been evaluated in CBCT of equine distal limbs.
The purpose of this study was to determine the optimal exposure parameters and filter thickness for CBCT images acquired with the O-arm ® imaging unit of the Vetsuisse Faculty of the University of Bern (O-arm ® ; Medtronic) in order to improve cortical bone detail in equine distal limbs.
## 2. Materials and Methods
## 2.1. Image Acquisition and Reconstruction
Adult equine cadaver limbs (n = 3), donated by the owners with a signed consent for research purposes and euthanized at the Swiss Institute of Equine Medicine (ISME), University of Bern, for reasons unrelated to this study, were used. Four anatomical regions were scanned for this study: 2 metacarpi (MC and MC-S), 1 metatarsus (MT), and 1 proximal phalanx (P1). The same distal forelimb was used for the MC and P1 regions, and also the MT originated from the same warmblood horse, while MC-S originated from a Shetland pony. Each limb was positioned on a carbon fiber table (Opera Swing; General Medical Merate SPA, Seriate, Italy). Image acquisition was performed using a mobile CBCT unit (O-arm ® ; Medtronic). Orthogonal fluoroscopic views were obtained to center the chosen region in the center of the field of view before starting the sequence of scans. Each anatomical region was subsequently scanned using the Standard 3D-knee mode, with a fixed power setting of 125 kV and manually varied tube currents (11 settings: 10, 12, 16, 20, 25, 32, 40, 50, 64, 80, and 100 mA). After the native scanning sequence, a mushroom-shaped aluminum filter consisting of a large square (7.7 × 7.7 cm) and a smaller square (4.7 × 4.7 cm), produced by RECO Mecanique SA, Salgesch, Switzerland, was installed on the tube housing. For fixation, a velcro-type reclosable fastener (3M SJ-3560 Dual Lock, Industrial Adhesives and Tapes Division, 3M Center, Maplewood, MN, USA) was applied on both the larger square of the filter and the tube housing, as shown in Figure 1A,B. For the latter, the O-arm ® gantry was opened, and the tube housing was made accessible using the available lever arm, as performed during technical maintenance. The smaller square filter fits inside the tube housing, allowing thicker filters to be used without risking contact between the filter and the inside of the CT gantry during tube rotation. A total of 5 filters with different thicknesses (13 mm (F13), 16 mm (F16), 19 mm (F19), 22 mm (F22), and 25 mm (F25)) were applied consecutively (Figure 2A,B); the same sequence of scans was repeated for each filter using the parameters listed above. Therefore, a total of 264 image series were created (4 bones, 1 native, 5 filters, and 11 current settings). Each series consisted of 192 reconstructed isometric transverse CT images numbered iNr 1 to 192, based on 391 exposures made during a 13 s acquisition period and a 360 • tube rotation. These reconstructed images had a slice thickness of 0.833 mm and an in-plane resolution of 0.415 mm.
All computed tomographic image series were exported and stored in a picture archiving and communication system (PACS, DeepUnity Diagnost, Dedalus Healthcare Group) and evaluated on dedicated imaging workstations in DICOM format.
## 2.2. Data Recording and Analysis
To allow blinded image evaluation by external observers, a postgraduate equine clinician (LP) selected, in consensus with a non-blinded board-certified veterinary radiologist (EVdV), the three image series with different tube currents that provided the best subjective image quality. This selection was made for each of the 4 anatomical regions and for both the native scans and all filter thicknesses. The selected series were anonymized and subsequently assigned a random identification number (n = 72; 18 series per bone, including 3 series per filter and 3 series for the 'native' group). Fixed window width (WW) and window level (WL) settings (WL: 2500; WW: 6000), subjectively considered appropriate during preliminary evaluation and selection of the scans by the non-blinded board-certified veterinary radiologist, were used during all further evaluations.
## 2.2.1. Visual Image Quality Assessment
Blinded, independent quality assessments of the 72 reconstructed series were performed by one board-certified veterinary radiologist (FP) and one board-certified equine surgeon (MdP). For each series, either 3 areas for the metacarpi/-tarsi (MC/MC-S/MT) and 2 areas for P1 of 10 consecutive slices each, were provided for evaluation: the proximal meta-/diaphysis (iNr. 32-42), mid diaphysis (iNr. 91-101) and distal meta-/diaphysis (iNr. 150-160) of MC/MC-S/MT, respectively, proximal meta-/diaphysis and mid to distal diaphysis of P1 (iNr. 88-98 and 115-125). Image quality was assessed according to a four-point visual grade scale shown in Table 1. Prominent artifact, severely impaired recognition of the IS This grading system was adapted from the system described by Demehri et al. ( 2015) [15]. Specifically, the visibility of trabecular pattern within the cortex of long bones (distinctness of anatomical structures) was graded on a four-point scale as a sign of sufficient penetration of the X-ray beam: (3) complete internal structure (IS), 0% loss of IS; (2) nearly intact IS, 1-25% loss of IS; (1) 25-75% loss of IS; and (0) >75% loss of IS (Figure 3).
Each observer recorded their scores in a predesigned spreadsheet (Microsoft Excel; Supplementary Material). In addition, the observers had to indicate whether any artifacts were visible that hampered cortical evaluation; these included significant quantum noise, cone beam artifacts, streak artifacts, or other artifacts (Figure 4). Also, the overexposure, as seen in Figure 5, needed to be noted, as it could affect the visibility of the outer cortex and thereby hinder the accurate diagnosis of a possible fissure and potentially associated early periosteal reaction. After a first individual assessment, the two board-certified experts re-evaluated the areas where their scores differed by more than 1 grade. For these areas, a second assessment was performed by both experts together to reach a consensus.
## 2.2.2. Clinical Case
The right metatarsus (MtF) of a 12-year-old Dutch warmblood horse, euthanized following the radiographic diagnosis of a metatarsal fissure and rejection by the owners to opt for surgical treatment, was used. Image acquisition was performed using the same scan protocol as for the first part of the study, including the 11 different current settings to acquire a native series, and repeated scans applying the same 5 filters. The scans were then evaluated by the postgraduate equine clinician (LP) together with the non-blinded board-certified veterinary radiologist (EVdV) to identify the series with the best image quality and clearest visibility of the fissure using the same software and settings.
## 2.2.3. Statistics
The interobserver variation in the visual image quality assessment of both observers was compared using a paired t-test. The best performing filter and tube current were determined using a Bayesian ordinal regression model (JASP software, Version 0.19.3.0 for Windows) based on the visual grade analysis scores of both observers, with the score as the ordinal factor. This model was applied for the MC, MC-S, and MT data with either the filter or the tube current as a fixed factor. Both calculated filter and current effects were then combined, assuming additive contributions from both predictors, into a performance matrix.
## 3. Results
## 3.1. Visual Image Quality Assessment
Both blinded reviewers assessed all 72 reconstructed CBCT scans. In cases where their initial scores differed by more than one grade, a joint re-evaluation was conducted to reach consensus. However, even after this targeted reassessment, statistically significant differences remained between the observers' scores in both evaluation categories-image quality and artifact presence (t-test, p < 0.05). The percentage of agreement, calculated as the proportion of identical ratings out of the total, was 61% for "Distinctness of anatomical structures" and 70% for "CT artifacts". As a result, average scores were not computed; individual scores were retained for subsequent evaluations and statistical analysis.
The results of the comparisons between the combinations of the different filters, including no filter and the different tube currents for the MC, MT, and MC-S, are shown in the performance matrix in Table 2. The native scan series clearly outperformed all filters, with results similar for tube currents of 64 and 50 mA. There was a slight preference for 50 mA, as despite a lower visibility score for IS, little to no peripheral overexposure was observed. When only IS grades were taken into account, the native scans with a tube current of 64 mA scored better than those with 50 mA, with average scores of 2.66 versus 2 on a maximum grade of 3, respectively. Also, both F16 and F13, when combined with 50 mA of tube current, performed better than average. Filter thicknesses from 13 mm to 22 mm in combination with 64 mA, as well as filters F19 and F22 in combination with 50 mA, performed around average. The tube current setting of 100 mA, combined with a 25 mm aluminum filter, performed worst, followed by F19 and F22.
For P1, tube currents 40, 50, and 64 were evaluated for the native and all filters except F13, where 50, 64, and 80 were selected. None of the scans with filters showed artifacts that influenced the evaluation of the internal structure, and all were graded the maximum score of 3 by both observers. However, the native scans at both 50 and 64 mA showed peripheral overexposure, and 1 observer noted streak artifacts. Therefore, both 50 mA and 64 mA scored worse with average artifact scores of 1.5 and 0.5 out of a maximum of 3, compared to 40 mA, which attained an average artifact score of 2.5. All filters from 19 mm thickness upwards had a perfect score for both artifacts and internal structure at 50 mA, while either the lower (for F22) or higher current (for F19) was scored slightly worse for IS by at least one of the observers, creating a mean overall score of 2.75 and 2.5, respectively. For filter thicknesses of 16 mm or less, including the native scans, none of the evaluated tube currents produced images that received a perfect score for internal structure from either observer.
Applying a filter reduced artifacts across all four anatomical regions, primarily preventing peripheral overexposure.
## 3.2. Clinical Case-Visual Image Quality Assessment
When comparing the different scans obtained in MtF using the different filters and current settings, the results from the first part of the study could be confirmed. The fissure was indeed best visible in the native scans (Figure 6) and on the scans with the highest mA (Figure 7). Five native cone beam computed tomographic scans of a right metatarsal bone with a naturally occurring acute fissure using a fixed tube potential of 125 kV, but different current settings:16 mA, 32 mA, 50 mA, 64 mA, and 80 mA, respectively. A tube current of 64 mA was considered the lowest exposure that provided a good visibility of the fissure as well as the internal structure of the adjacent dorsal cortex.
## 4. Discussion
This study focused on identifying CBCT acquisition parameters that enhance the detection of subtle fissures by using variable tube currents and additional filtration, specifically in challenging anatomical regions characterized by dense, thick cortical bone, such as the third metacarpal/-tarsal bones and the proximal phalanx of equine limbs. Modifying only the tube current directly influences the image noise and contrast without altering other acquisition parameters that could confound the interpretation of results. By isolating this variable, we were able to specifically assess its effect on fissure detection, ensuring that any observed differences in image quality could be attributed solely to changes in tube current. In our standard protocol, the tube power (125 kV) was fixed and could not be modified. Alternative scan modes, such as high-resolution, are possible on the O-arm ® and would allow different power settings but also require substantially longer scan times, making them unsuitable for scanning standing sedated horses.
Higher tube currents were generally associated with improved image quality across all evaluated regions. This supports the need for high tube currents to avoid missing fine fissures. However, higher currents entail increased radiation exposure for animal handlers in the room, who must position the limb correctly during scanning of standing, sedated horses [6].
While native (unfiltered) scans provided the best image quality in the MC, MC-S, and MT regions, thicker aluminum filters (19 mm, 22 mm, or 25 mm) enhanced image quality in the P1 region without requiring increased exposures. The filters also had a positive effect by reducing artifacts, especially overexposure at the edges of the cortex in native scans, but the decreased penetration and associated reduced diagnostic quality of the thick cortical bone in the third metacarpal/-tarsal bones outweighed the benefits of reducing artifacts. As a compensatory means, achieving a high-quality image with filters required increasing the tube current beyond that used for native images. Therefore, the use of filters is not in line with the ALARA principle of radioprotection for personnel and helpers during scanning of equine metacarpi and -tarsi.
Based on the results of our study, we propose practical guidelines for CBCT imaging of equine limbs acquired using the O-arm ® imaging unit, as outlined in Table 3. A key component of this study was the use of a custom visual grading scale for image quality analysis. Visual Grading Analysis (VGA) is a robust, observer-based tool frequently used in radiological research to evaluate the clinical adequacy of images, particularly in relation to dose optimization strategies [15,16]. In our study, the grading scale was tailored to focus on anatomical detail relevant to equine limb pathology, with particular emphasis on trabecular pattern visibility and artifact interference.
The interobserver agreement of this study was fair to moderate, preventing the combination of their results. This result is lower than a previous study comparing CBCT and multidetector CT for the visualization of anatomical structures in the fetlock region [17]. However, that study focused on the visibility of a structure-its margins, shape, size, and overall attenuation-without specifically evaluating the internal structure of the cortices. Our results mirror findings in human CBCT imaging, where interobserver agreement levels below the commonly accepted threshold of 80% have also been reported for visual assessments of bone and soft tissue structures [15]. These results underline the subjective nature of visual image quality assessment and the importance of using combined quantitative and qualitative evaluation approaches or of additional validation, especially when introducing technical modifications like filtration. While the gray scale values of the O-arm ® CBCT system do not represent Hounsfield units (HUs), quantitative measurements may still be applied when comparing regions of diseased versus non-diseased bones [18]. However, that approach cannot be used in this study when evaluating the effects of imaging parameters and filters on the same bone. We used post-mortem CBCT results from a clinical case with an in vivo-diagnosed fissure (MtF) to provide additional validation for part of our findings from the visual image quality assessment study. The metatarsal bone fissure was optimally visualized with a tube current of 64 mA without additional filtration.
This study has several limitations that should be acknowledged. Although both the number of limbs and evaluators are the same as a previous study comparing the diagnostic quality of CBCT and fan-beam CT in the equine metacarpophalangeal joint [19], and the number of reviewers is similar to other studies evaluating image quality in equine CT examinations [20,21], the sample size was small. We evaluated only three cadaver limbs and one clinical case, which limits the extrapolation of the findings to the broader equine population. We specifically included a pony limb as one of the normal cadaver limbs, but no limbs of draft horses were included. Although previous studies have shown differences in density between horse breeds in the navicular and carpal bones, a recent study did not detect such differences in the metacarpus, which is in line with our observations [22][23][24]. However, evaluating a larger number of limbs from different breeds would allow further confirmation of these results. Similarly, only two blinded evaluators performed the analysis of visual grading characteristics. Despite these low numbers, clear trends toward higher required tube currents and no added filtration were observed in the metacarpus and -tarsus regions in this preliminary study. However, future studies should include a larger number of equine limbs and a greater number of blinded evaluators to strengthen the reliability of the results. Another limitation is the lack of quantitative image quality metrics, such as signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR). Previous studies have used the mean gray scale voxel value and its standard deviation to calculate CNR [18,25]. In addition, scanning phantoms can be used to calculate the SNR. Performing these evaluations could provide additional quantitative information on the effect of filtration on CBCT image quality. Finally, these results are specific to the O-arm ® system and cannot be extrapolated to other CBCT systems. Specifically, post-processing varies across CBCT systems, as the selected protocol and associated reconstruction method significantly influence the appearance of the generated transverse CBCT images. In the authors' experience with the O-arm 2® system, these post-processing algorithms can play a major role in the conspicuity of internal structures within the cortex, and they cannot be altered retrospectively. Therefore, future work should explore the use of different post-processing algorithms, including artificial intelligence-based image processing techniques, which have been shown to improve image quality and diagnostic value in other fields [26]. Radiation dose optimization studies would also be valuable to ensure the safe and effective clinical use of CBCT in equine practice.
erinary Medicine, Vetsuisse Faculty, University of Bern, for providing the resources to perform this study.
## References
1. De Preux, Klopfenstein Bregger, Brunisholz et al. (2020) "Clinical use of computer-assisted orthopedic surgery in horses" *Vet. Surg*
2. Dakin, Lam, Rees et al. (2014) "Technical set-up and radiation exposure for standing computed tomography of the equine head"
3. Keane, Paul, Sturrock et al. (2017) "Computed Tomography in Veterinary Medicine: Currently Published and Tomorrow's Vision"
4. Puchalski (2012) "Advances in equine computed tomography and use of contrast media" *Vet. Clin. N. Am. Equine Pract*
5. Bregger, Koch, Zimmermann et al. (2019) "Cone-beam computed tomography of the head in standing equids" *BMC Vet. Res*
6. Pauwels, Van Der Vekens, Christan et al. (2021) "Feasibility, indications, and radiographically confirmed diagnoses of standing extremity cone beam computed tomography in the horse" *Vet. Surg*
7. Stewart, Siewerdsen, Nelson et al. (2021) "Use of cone-beam computed tomography for advanced imaging of the equine patient" *Equine Vet. J*
8. Crijns, Martens, Bergman et al. (2014) "Intramodality and intermodality agreement in radiography and computed tomography of equine distal limb fractures" *Equine Vet. J*
9. Perks, Dendere, Irving et al. (2015) "Filtration to reduce paediatric dose for a linear slot-scanning digital X-ray machine" *Radiat. Prot. Dosim*
10. Perks, Trauernicht, Hartley et al. (2013) "Effect of aluminium filtration on dose and image quality in paediatric slot-scanning radiography"
11. Hansson, Finnbogason, Schuwert et al. (1997) "Added copper filtration in digital paediatric double-contrast colon examinations: Effects on radiation dose and image quality" *Eur. Radiol*
12. Ay, Mehranian, Maleki et al. (2013) "Experimental assessment of the influence of beam hardening filters on image quality and patient dose in volumetric 64-slice X-ray CT scanners" *Phys. Med*
13. Prionas, Huang, Boone (2011) "Experimentally determined spectral optimization for dedicated breast computed tomography" *Med. Phys*
14. Davies, Skelly, Puggioni et al. (2020) "Standing CT of the equine head: Reducing radiation dose maintains image quality" *Vet. Radiol. Ultrasound*
15. Demehri, Muhit, Zbijewski et al. (2015) "Assessment of image quality in soft tissue and bone visualization tasks for a dedicated extremity cone-beam CT system" *Eur. Radiol*
16. Lang, Neubauer, Fritz et al. (2016) "A retrospective, semi-quantitative image quality analysis of cone beam computed tomography (CBCT) and MSCT in the diagnosis of distal radius fractures" *Eur. Radiol*
17. Bierau, Cruz, Koch et al. (2024) "Visualization of anatomical structures in the fetlock region of the horse using cone beam computed tomography in comparison with conventional multidetector computed tomography" *Front. Vet. Sci*
18. Siewerdsen, Uneri, Hernandez et al. (2020) "Cone-beam CT dose and imaging performance evaluation with a modular, multipurpose phantom" *Med. Phys*
19. Stewart, Siewerdsen, Selberg et al. (2023) "Cone-beam computed tomography produces images of numerically comparable diagnostic quality for bone and inferior quality for soft tissues compared with fan-beam computed tomography in cadaveric equine metacarpophalangeal joints" *Vet. Radiol. Ultrasound*
20. Mcquillan, Kearney, Hoey et al. (2022) "A threshold volume of 10 mL is suggested for detecting articular cartilage defects in equine carpal joints using CT arthrography: Ex vivo pilot study" *Vet. Radiol. Ultrasound*
21. Ogden, Winderickx, Bennell et al. (2018) "Computed tomography of the equine caudal spine and pelvis: Technique, image quality and anatomical variation in 56 clinical cases" *Equine Vet. J*
22. Gabriel, Jolly, Detilleux et al. (1998) "Morphometric study of the equine navicular bone: Variations with breeds and types of horse and influence of exercise" *J. Anat*
23. Abdunnabi, Ahmed, Philip et al. (2012) "Morphometrical Variations of the Carpal Bones in Thoroughbreds and Ponies" *Anat. Histol. Embryol*
24. Goldstein, Engiles, Rezabek et al. (2021) "Locomotion on the edge: Structural properties of the third metacarpal in Thoroughbred and Quarter Horse racehorses and feral Assateague Island ponies" *Anat. Rec*
25. Carrino, Muhit, Zbijewski et al. (2014) "Dedicated Cone-Beam CT System for Extremity Imaging" *Radiology*
26. Jankowski, Chan (2024) "Advances in imaging (Intraop Cone-Beam Computed Tomography, Synthetic Computed Tomography, Bone Scan, Low-Dose Protocols)" *Neurosurg. Clin*
27. "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=PMC12619308&blobtype=pdf | # Facilitator and barrier perspectives on learning and implementing high-resolution anoscopy in Abuja, Nigeria: a qualitative study
Bmc Cancer, Megan Mansfield, Connor Volpi, Sylvia Adebajo, Rebecca Nowak, Chama John, Ruxton Adebiyi, Andrew Mitchell, Jumoke Aigoro, Yerima Bawa, Kazeem Kolawole, Uchenna Ononaku, Paul Jibrin, Oluwole Olaomi, Francis Agbo, Abayomi Aka, Søren Bentzen, Stephen Goldstone, Patrick Dakum, Joel Palefsky, Cheryl Knott
## Abstract
Background Early detection and treatment of anal precancer via high resolution anoscopy (HRA) is paramount to prevent anal cancer, particularly for populations at heightened risk like sexual minority men (SMM) living with HIV. Successful training and sustainability of cancer screening requires attention to local contexts, best captured by qualitative research. Using the Consolidated Framework for Implementation Research (CFIR), this study investigated factors that challenged or fostered learning and implementing HRA across a variety of stakeholder groups in Abuja, Nigeria.Methods Using in-depth qualitative methodology, nineteen semi-structured interviews were conducted in September 2023 with stakeholders -patients who underwent HRA, HRA providers, and health system representatives in Nigeria. Thematic analysis, guided by CFIR, was employed to identify key themes related to the barriers and facilitators to practicing HRA as guided by the International Anal Neoplasia Society.
ResultsEight themes were identified across three domains. Barriers included low knowledge and understanding of HRA, with participants explicitly noting the need for more research in low resource settings to garner local acceptance. Participants were concerned about financial costs for the clinic and the patients. Facilitators included organizational buy-in, SMM social networks, and a safe clinic environment to support HRA engagement. Facilitators important for sustainability included acceptance of the research evidence for HRA and recognition of the health benefits. Overall, participants from all stakeholder groups welcomed HRA as a new evidence-based intervention as part of HIV care services. Conclusions Our study highlighted the need for localized research, cultural sensitivity, and resource allocation to improve the adoption of HRA in a Nigerian HIV care setting.
## Introduction
Anal cancer is the most prevalent human papillomavirus (HPV)-associated cancer among sexual minority men (SMM) living with HIV [1,2], and its burden is expected to rise in low-and middle-income countries (LMICs) due to the high prevalence of HIV, oncogenic HPV, poor vaccine coverage, and improved survival [3][4][5][6]. With 30,146 new cases of anal cancer reported in 2020 [7,8], many could be prevented with an expansion of anal cancer screening [9]. As demonstrated by the ANCHOR study [10], anal cancer can be prevented through the timely identification and treatment of high-grade squamous intraepithelial lesions (HSIL), the precursors to anal cancer, using high resolution anoscopy (HRA).
Anal cancer remains a significant but often overlooked public health concern in Nigeria, with an estimated crude incidence rate of 0.65 per 100,000 men and increasing age-specific rates observed particularly among those aged 45 years and older [11]. In 2020, over 100 new cases were reported annually among men aged 40-59, with corresponding mortality rates rising steeply with age [11]. Despite this growing burden, there are currently no official national recommendations or organized screening programs for anal cancer in Nigeria (e.g., usage of HRA) underscoring the urgent need for targeted prevention and early detection strategies.
Initiating anal cancer screening with HRA presents its own unique challenges because of the steep learning curve in correctly identifying anal HSIL [12,13]. To help standardize anal HSIL detection, the International Anal Neoplasia Society (IANS) developed guidelines on clinic setup for HRA services, training and mentorship in HRA, practical competencies during HRA procedures, target practice volumes and performance metrics, and monitoring patient experiences [14]. However, the IANS guidelines were developed by experts from highincome settings who had clinical infrastructure, integrated pathology services, and accessible mentorship. Bringing anal cancer screening to Nigeria where there are unique LMIC context-specific challenges [9,15] prompted an evaluation of whether the IANS guidelines and the screening could be readily integrated into HIV care services.
Qualitative methodology offers a nuanced and in-depth investigative approach for identifying context-specific barriers and facilitators for implementing a novel procedure [16], such as HRA-guided detection of HSIL, previously unavailable in Nigeria. Emerging qualitative research highlights the importance of obtaining three perspectives: patients, health providers, and health system for overall HRA sustainability [17][18][19][20][21]. Barriers highlighted by patients included lack of knowledge, limited access to screening, HIV or HPV-related stigma, and discomfort. Health provider barriers included gaps in knowledge and expertise, as well as challenges in communicating and building relationships with patients. System-level barriers encompassed societal stigma, limited healthcare capacity, and competency with new clinical procedures [18,[22][23][24]. Furthermore, barriers are resolved when a variety of bridges are built across stakeholder groups [18]. When providers adopt communication to convey knowledge to address patient-reported uncertainty, SMM express a willingness to engage in anal cancer screening [25]. However, most anal cancer screening research focuses on the patient [18][19][20][21] or provider perspective [26] but often lacks integration from all stakeholders. Sustaining anal cancer screening in Nigeria where the population who would benefit from the service is highly marginalized will require support from members of the targeted population, healthcare providers, and other health sector stakeholders [27,28].
This study is part of a larger implementation research study, the Integrated Model for Prevention of Anal Cancer using screen and Treat for HSIL (IMPACT). One of the aims of the study is to identify barriers and facilitators to training and implementing HRA per the IANS guidelines in the local Nigerian setting [14]. The current study contributes to this aim by capturing the perspectives of patients, providers and health sector stakeholders in a setting naïve to HRA. Specifically, we were guided by the conceptual framework, Consolidated Framework for Implementation Research (CFIR) [29], because its consideration of the internal and external environment as well as the process and the complexity of the efficacious procedure (e.g., HRA) aligned with our goal to explore future sustainability in an LMIC setting.
## Methods
## Study setting
This research was conducted at the TRUST Clinic, a community-based research and care center established in 2012 in Abuja, Nigeria through a tripartite partnership between the International Center for Advocacy and Rights to Health (ICARH), the Institute of Human Virology Nigeria, and the University of Maryland Baltimore [30]. Through this collaboration, ICARH recruited safe healthcare environments facilitated trust and patient engagement and would promote long term sustainability. Overall, the study provided perspectives from various stakeholders that strengthen clinical proficiency and sustainability of anal cancer screening in Nigeria. Keywords Anal cancer prevention, Implementation science, Africa a cohort of nearly 2,800 SMM using respondent-driven sampling, in which initial participants received coupons to recruit peers in their social networks. This peer-driven model fostered a sense of trust and confidentiality critical for engaging a highly stigmatized population [31]. The success of the TRUST cohort is in part attributed to the safety and freedom felt by participants who come to the clinic. Our prior qualitative research highlighted the TRUST clinic as a place where SMM could freely express themselves, a place they love to come to, and where it allowed them to readily engage with study and healthcare staff [32]. The clinic offers HIV and STI testing and treatment services and has become a trusted hub for healthcare and research among SMM in Abuja [33][34][35]. The broader study environment includes key health system actors and providers operating within this context. These stakeholders, providers, policymakers, and government officials are embedded in a health system where services are largely paid out of pocket, infrastructure is often limited, and training for specialized procedures such as HRA is scarce [36,37].
A key component of the IMPACT study is the Implementation Science (IS) Team, which includes local stakeholders who have been involved with the study since its inception. IS Team members comprise patients, healthcare providers, researchers, and representatives from the government and non-governmental organizations. This multisectoral, community-engaged team provides a unique opportunity to explore both structural and interpersonal dynamics that shape the feasibility and acceptability of implementing HRA screening among Nigerian SMM. The primary objective of the IS Team is to identify and adapt elements of the IANS guidelines to the Nigerian context. To do this they meet regularly to review guideline components, discuss contextual barriers, and co-develop locally appropriate adaptations, such as modifying training benchmarks, adjusting language in patient-facing materials, and identifying culturally sensitive approaches to screening, to ensure the IANS guidelines are acceptable and feasible within the Nigerian healthcare and sociocultural context. As a result, of their knowledge of and engagement with the IANS guidelines, many members of the IS Team participated in the interviews conducted for the current study.
## Conceptual framework
CFIR was the theoretical framework used to guide this inquiry, allowing for a comprehensive examination of the contextual factors affecting implementation processes in an LMIC setting [38,39]. CFIR categorizes conceptual elements from various theories and disciplines into 39 constructs, further organized into five key domains. These domains encompass intervention characteristics, outer setting, inner setting, characteristics of individuals, and implementation process [40], all of which interact to influence the implementation effectiveness of an evidence-based intervention [29]. Our qualitative interview guide prompted participants to provide nuanced perspectives on CFIR constructs in terms of their function as both barriers and facilitators for the implementation of HRA as per the IANS guidelines. We also asked participants what recommendations they had for overcoming barriers and leveraging facilitators. Specifically, they were prompted to consider their own experiences and opinions based on their area of expertise.
## Study design and sample
The semi-structured interviews (SSI) were conducted in September 2023 with participants from our three groups of interest -patients, health providers, and health system representatives (Table 1). Patients, aged 24-39, were purposefully sampled to include those who had good and moderate antiretroviral adherence, in order to capture variability in how participants follow health recommendations, including HIV medication adherence [41].
Patients were further stratified based on self-reported levels of pain experienced during the HRA procedure to ensure a range of procedural experiences. All participating patients had completed HRA at the TRUST Clinic within the prior 2-4 months. Most of these patients identified as bisexual or pansexual (75%) and the remaining identified as gay or homosexual (25%). Eligibility criteria for the SSI included being willing to attend an in-person interview at the TRUST Clinic and providing written informed consent. Participants received refreshments and approximately 8500 Naira (approximately $5.00 USD) to cover transportation costs. The health providers were clinicians directly involved in learning, conducting, and refining HRA procedures for detecting HSIL at the TRUST Clinic. Health system representatives included individuals holding roles at local hospitals, non-governmental organizations, and government agencies, all of whom had direct involvement with or oversight of the implementation of the HRA program, such as participating in strategic planning, training, or monitoring activities. Eligibility criteria for providers and health system representative groups included having direct knowledge of HRA procedures, being responsive to the interview invitation (via email or phone), and providing written informed consent to participate Interviews were conducted at a mutually agreed location (e.g., participant's office), with the majority conducted at the TRUST Clinic.
Data collection continued until thematic saturation [42,43] was reached within each stakeholder group, defined as the point at which no new themes or concepts emerged during interviews. The study team conducted ongoing, iterative reviews of the interviews throughout data collection to assess the emergence and repetition of themes specific to each group. Although formal member checking was not conducted, preliminary themes were reviewed and discussed by the broader team to support the credibility and contextual validity of the findings. Of note, potential stakeholders who were not responsive to multiple contact attempts were no longer pursued.
## Data collection
The publicly available CFIR Interview Tool [44] was adapted for the SSI guide (Appendix A). Two trained qualitative researchers conducted the interviews. The interviewers were both White, cisgender, and academically trained researchers from the U.S. who identify as sexual minorities and have prior experience conducting research with sexual and gender minority communities. The study team engaged in reflexivity throughout the study, recognizing how personal and professional standpoints shaped the design, conduct, and interpretation of the interviews [45,46]. Prior to data collection and during analysis, team members discussed positionality and critically examined how their lived experiences may have influenced participant interactions and analytic decisions. The interview guide included directions on questions that were relevant to the different groups of participants and their areas of expertise to ensure their responses were relevant and appropriate (e.g., patients were not asked about provider training or clinical implementation procedures). This adaptive approach, grounded in reflexive practice [45,46], supported the credibility and trustworthiness of the data by ensuring that participants were only asked questions aligned with their lived experience and contributions to HRA implementation. The interviews lasted 30-60 min and were audio recorded and transcribed verbatim by local members of our study team. As needed, a translator was present to support converting Hausa or Nigerian Pidgin into English. The translator was a bilingual member of the study team who had prior experience working with the TRUST Clinic population. To promote comfort and safety, particularly given the stigmatized identities of many participants, interviews were conducted in private settings, and participants were encouraged to use pseudonyms or initials instead of their real names during the interviews.
## Data analyses
We conducted a thematic analysis of the transcripts [47]. Thematic analysis is a flexible analytic procedure that focuses on the experiences of stakeholders in their own words [47,48]. This approach allowed us to identify key themes related to both facilitators and barriers to implementing HRA per the IANS guidelines in Nigeria.
The analyses began with the researchers familiarizing themselves with the data through review of interviewer notes and transcripts. They then collaborated on an initial codebook with codes deductively specified based on CFIR. The initial codebook was pilot-tested on a subset of the transcripts (n = 5) using ATLAS.ti [49]. Piloting the codebook allowed the researchers to ensure they were using the codes consistently (i.e., calibration). Iterative discussions during the piloting process resulted in minor adjustments of the codebook to further facilitate consistent use of the codebook. These adjustments included identification of emergent codes and coming to agreement on coding practices (e.g., coding larger segments of text at the level of sentences or paragraphs to retain as much context as possible). Once calibrated, the researchers each independently coded the remaining 14 transcripts.
After coding was completed, the coded texts were grouped into themes and sub-themes and linked with CFIR constructs. The identified themes are presented with support from participant quotations. Throughout the coding and thematic analyses process, the coders engaged with local members of the research team to promote culturally sensitive interpretations. Reporting for this study was guided by the Consolidated Criteria for Reporting Qualitative Research (Table 2) [50].
## Results
We completed 19 SSI, eight with patients enrolled in the IMPACT study (42%), four with providers at the TRUST clinic (21%), and seven with Nigeria health system representatives (37%). A primarily deductive thematic analysis of the codes revealed a total of eight themes across three domains (see Table 3). The themes reflect CFIR constructs in terms of both barriers and facilitators of implementing HRA based on the IANS guidelines.
## Domain 1: barriers to HRA implementation per the IANS guidelines
Overall, the two themes constituting barriers to implementing HRA screening using the IANS guidelines centered on feasibility, especially considering the potential expansion of HRA beyond the TRUST clinic.
## Evidence strength and quality: more research on anal cancer screening is needed in Nigeria and other LMICs
Provider and health system stakeholders acknowledged that, even though they trusted the existing research that supports the IANS guidelines, this research is limited. Specifically, the guidelines were written primarily by Westerners and little to none of the research was conducted in Nigeria or other LMICs. One health system representative noted, "all the evidence is there [in high income countries [but] …here [in Nigeria] you have to start from scratch to get that evidence. " As a result, the feasibility and appropriateness of some of the aspects may require changes for improved training and ultimately implementation in Nigeria. For example, one provider suggested adjusting some of the milestones, such as detecting at least 50 HSILs during the first year of training: "With time and then the longer we do it, 50 [HSIL detection] might begin to seem more realistic. But for now, given that it's our first year we might not be hitting those figures. " Additionally, another provider suggested adjusting the language used in some of the materials to be more sensitive to different cultural contexts, stating, "there is this great diversity of tradition that we have…Nigeria is very, very large. So, some of the things you may need to tweak due to their beliefs of that area or that region…the language. For example, [for] some people the word "sex," "anus," or "cervix" may [or may not] be offensive to them. "
## Cost: costs implications are a concern relating to necessary infrastructure to implement HRA and relating to the out-ofpocket expenses for patients
Stakeholders from all three groups discussed barriers associated with costs. One provider stated that "funding will be key…in terms of the actual training of these persons and letting them have a work tool. " The tools refer to the costs of getting a facility equipped to complete HRA screenings as per the IANS guidelines as a notable barrier. Many commented that such expenses would require facilities to have funders, such as one health system representative who stated, "I was asking myself particularly around things that need to be in place for the anoscopy. I'm asking myself who bears the costs of those things?" But another health system representative disagreed stating that "the necessary equipment [is] not too expensive…[for] smaller clinics, that should not be expensive. "
However, both providers and health system representatives agreed that the mentorship component of HRA implementation is a notable challenge for training of Nigerian providers, noting "it's very expensive training one person" (health system representative). To their knowledge, there were no mentors on the African continent, let alone in Nigeria. So, mentors must be flown in from abroad, which drastically increases the cost as compared to contexts wherein mentors are more localized.
Additionally, participants brought up the expenses associated with conducting HRA. Who pays for the screening, pathology, and treatment for any illnesses that are detected? Patients described difficulty getting to HRA appointments due to the cost of transportation. Although, as part of the larger study, their transportation was covered, many suspected that without that financial support most people would not be able to have access to HRA screening. One patient stated, "what I think that can stop people can be cost of transportation because many people are living far away…the transportation cost that is what I see as a major setback. " There was concern that, even if the screening was free, people would be hesitant to get screened given the potential implications for the cost of treatment based on the results of the screening. If anal cancer were detected, many would likely be unable to afford such treatments. One patient stated, "the treatment -I know it's a lot, it's a very expensive thing. It is not an easy experience [receiving a cancer diagnosis] and there is no solution coming from anywhere. " One health system representative further explained that "most of the health services [in Nigeria] are paid out of pocket. So, if [patients] are going to bear the cost, you will see that it is going to be difficult. " Two white American members of the IMPACT research team.
## 2. Credentials
What were their credentials?
Both interviewers had master's level degrees related to social science research.
## 3. Occupation
What was their occupation at the time of the study?
One interviewer was a Senior Qualitative Analyst and a PhD student. The other interviewer was a PhD student.
## 4. Gender
Were they a man or woman?
One interviewer was a woman, and she conducted all internal team interviews. The other was a man, and he conducted all interviews with IMPACT participants. Both interviewers conducted interviews with external stakeholders.
## 5. Experience and training
What experience or training did they have?
One interviewer had 11 years of experience in conducting and analyzing qualitative research. She had a BA in LGBT studies and psychology, a master's degree with a co-concentration in applied research methods and evaluation and applied social psychology. At the time of data collection, she was pursuing their PhD in applied social psychology and had been working with members of the research team for over a year. The other interviewer had over 7 years of experience in conducting and analyzing qualitative research. He has a BS in psychology, a master's degree in public health with a focus in epidemiology. At the time of data collection, he was pursuing his PhD in Social & Behavioral Sciences and had been working with members of the research team for over a year. Relationship with Participants 6. Relationship established Was a relationship established prior to study commencement.
For external stakeholders and IMPACT participants, the semi-structured interview conducted for this study was the first time the interviewer had met the interviewee. For internal team interviews, the interviewer had an established relationship with two of the four interviewees as they had worked together (virtually) for about one year.
## 7. Participant knowledge of the interviewer
What did the participants know about them?
Most participants knew the interviewers as members of the IMPACT study team and that their role was to conduct these semi-structured interviews. Some members of the internal team were professionally familiar with the interviewers. Where was the data collected?
Most interviews were conducted in private spaces at the TRUST Clinic. Some external stakeholders were interviewed at their offices in other locations in Abuja 15. Presence of non-participants Was anyone else present beside the participants and the researchers?
All IMPACT participant interviews included a translator who was also a member of the research team. The same translator was present for some of the external stakeholder interviews. A study coordinator was present for all internal team interviews.
## 16. Description of sample
What are the important characteristics of the sample? Stakeholder group (i.e., patients, health provider, health system representative)
## Domain 2: facilitators of implementing HRA per the IANS guidelines
Overall, the three themes in this domain demonstrate the importance of an awareness of the context in which HRA is being implemented. In the context of this study, with an emphasis on the SMM community in Nigeria, the reputation of the clinic and dynamics between staff and patients emerged as important facilitators.
## Self-efficacy and individual stage of change: providers are supportive of and feel competent with the IANS guidelines
The providers were very supportive of and motivated to learn HRA and use the IANS guidelines. One provider described the clinic staff as being all "on board. I believe everybody gets the message and everybody's working to ensure that it's been implemented the way it's supposed [to be]. " Similarly, health system representatives expressed strong support for the implementation of anal cancer HRA screening with one representative stating "I believe that it [anal cancer screening] enhances -help[s] the client to access care and get treated. " Providers and health system representatives also felt confident about their knowledge of the IANS guidelines, noting that their knowledge, understanding and confidence were growing over time. They especially emphasized the important role mentorship plays in supporting their increased knowledge and confidence. One provider described how another provider "told me he was also not confident when he started, like we were just doing everything blindly…The [PROVIDER] is now very confident. "
## Compatibility and culture: the TRUST clinic is a safe space for patients
Participants from all groups emphasized the importance of the TRUST facility in terms of patient comfort. In Nigeria, SMM experience stigma and discrimination in community and societal contexts. As a result, SMM patients, the target population for this study, need to feel safe and accepted when attending appointments for a potentially stigmatizing procedure. One patient explained that "If this procedure is taken to another hospital they may stigmatize or shame someone. " So, many participants emphasized that a facility without this kind of acceptance and trust will not be able to implement anal cancer screening HRA well. For example, one health system representative explained that the "TRUST Clinic is very much welcoming because you have the population over there. They understand what they are doing there. But if you take it to a normal general hospital, how would that affect their accessibility to the care? Their willingness to go? Will they be blackmailed? Will they be threatened?"
## Networks and communications: social networks play an important role
Patients discussed the important role of community social networks for increasing awareness of and accessing anal cancer screening. For example, one patient noted that "if I have a friend that has it but does not want to undergo the screening, I will ask why and advise him to do it because it is an issue that can claim his life. " Patients also described learning and sharing with others about HRA via WhatsApp groups and interpersonal relationships/conversations, emphasizing the importance of Social networks play a crucial role not only in sharing information but also in assisting patients with transportation to clinics for screening appointments. As members of a stigmatized group, one patient explained that traveling to the clinic in groups felt more comfortable: "I'm even thinking bringing him here because sometimes they do fear in our community but they're hiding. They don't want to be out. So if you tell them, okay, go by yourself to the clinic, they will say, this is our first time, come lets go together" with another stating that leveraging social connections can also increase attendance and help mitigate costs: "I don't have money to go with you if you can go That is the same challenge I'm talking about. Bringing them here is the same challenge to me and you. "
## Domain 3: facilitators of sustainability and broader implementation of HRA
The three themes in this domain demonstrate the utility and acceptance of anal cancer screening through HRA implementation among all stakeholder groups. Stakeholders respected the training and recognized the associated health benefits, which motivate patients to complete the procedures.
## Evidence strength & quality: existing research is well accepted
Providers and health system representatives were confident and accepted the robustness and legitimacy of the IANS guidelines. One provider stated that they "trust the IANS guidelines. " Overall, they trusted the scientific rigor of the existing research. For example, one health system representative described reviewing presentations and journal articles about HRA and "the findings have helped to know that these things can be practiced in other countries or in other areas. I feel it's something that should be done and executed. " The participants described confidence in the applicability of the IANS guidelines with limited adaptations.
## Relative advantage & patient needs and resources: HRA serves the needs of patients
Participants from all three groups acknowledged that HRA is beneficial for the health of patients and Nigerians more broadly. One provider described that patients find benefits in completing the screening and are happy they did it, stating "a lot of clients are happy that they just even performed the process. Even those who have some benign results, more of them are happy. So, I think it is satisfying that more services are being added." This was further supported by patients who noted that they "will come back for [the screening] 100%. " And a health system representative noted that the benefits of the HRA guidelines are that they are "patient-centered" and "everything is to be tailored to a patient. So, it meets the needs of patients rather than the needs of the healthcare worker. " Patients agreed, with one noting that "80% of our community members are not protecting themselves during sex…and some of us, we don't take our health serious. So, I think there is need for us to continue doing tests…I feel there is a need for us to do the test, we need it. "
## Cosmopolitanism: stakeholder engagement is crucial
Participants from all three groups acknowledged the importance of engaging relevant people for their support in promoting acceptance and broader implementation of HRA as per the IANS guidelines. They emphasized the need to partner not only with established communitybased organizations that relate to and are trusted by SMM at risk for anal cancer, but also with key institutional stakeholders. As one participant explained, "If you get the Ministry of Health, led by the Minister for Health, on board, then the rest is done. " A health system representative simply stated, "You have to be in the community as well. The linkage is the reason we have [access]. " Participants noted that this success has been observed in at the TRUST clinic due to the collaboration and engagement of the community-based organization ICARH. As one provider explained, "we have gatekeepers, influencers, and key opinion leaders working with us, collaborating with over the years. So, these people were the persons who gave us access and gave us a channel to pass this information to the members within the community about something like this [screening.] Imagine we didn't have those channels; it would be quite difficult to actually get the message out there and get people coming in for the procedure. "
## Discussion
Using HRA to detect and treat HSIL is paramount to the prevention of anal cancer [1,2] and with the availability of the IANS guidelines, its adoption is within reach in LMICs. Specifically, settings like Nigeria have the most to gain from this secondary prevention strategy, given the population of SMM living with HIV and the lack of clinical expertise [8]. Our analyses integrated perspectives from patients, HRA providers, and health system representatives to capture a comprehensive understanding of the facilitators and barriers to HRA implementation as per the IANS guidelines. Although presenting the data collectively allowed us to identify overarching themes, we acknowledge that each stakeholder group offered unique insights. Capturing local perspectives from various stakeholders on the facilitators and barriers to learning and implementing anal cancer screening is necessary for long term success and sustainability [51]. Nuanced differences provided from the different stakeholder groups offer valuable context for tailoring implementation strategies to address the distinct needs and priorities of each group. Specifically, we identified existing research limitations and concerns about the cost of implementing HRA as a barrier. General facilitators included provider and patient acceptability when the screening was based at a clinic where patients felt safe. Concerning overall sustainability, our analyses found evidence for acceptability across stakeholder groups for implementing HRA per IANS guidelines suggesting a good foundation for building further buy-in from entities beyond the inner context of the TRUST Clinic.
The findings from this study shed light on the complex landscape surrounding the roll out of anal cancer screening in Nigeria, particularly within the context of the SMM community. Through an exploration of barriers, facilitators, and sustainability factors, several crucial themes emerged, offering relevant insights for public health practitioners, policymakers, and researchers alike. These insights can ultimately be used to inform sustainable implementation of HRA in Nigeria and serve as the foundation for building capacity in HRA in other LMICs.
The identified barriers underscore the need for contextual adaptation and resource allocation to ensure the successful integration of HRA into existing healthcare settings. The lack of localized research and cultural sensitivity within the existing guidelines present significant hurdles. These barriers were recently highlighted by Blair et al., who emphasized the need for data and strategies specific to LMICs [4]. Further, although there has been some research on HRA and anal cancer screening in Latin America, we are not aware of any study implementing HRA in Nigeria or any other country in Africa [52][53][54][55]. Efforts to bridge this gap should prioritize collaborative research endeavors that actively involve a variety of stakeholders (e.g., community members, researchers, clinicians, policy makers) and reflect the diverse sociocultural landscape of the country. Furthermore, mitigating the financial burdens associated with facility setup, mentorship costs, and patient out-of-pocket expenses is critical, as these barriers pose a substantial threat to the broader implementation and long-term sustainability of HRA programs in Nigeria and similar LMIC settings. Without strategic financial support, effort to scale up screening services may stall, particularly in resource-constrained contexts where most health services are paid for out-of-pocket. Strategies such as securing local or donor funding, integrating cancer screening into existing public health infrastructures, and implementing subsidized or free screening and treatment programs have been shown to enhance accessibility and feasibility in comparable LMIC settings [56][57][58]. These approaches can not only reduce financial strain on patients and providers but also build more sustainable and equitable screening programs. For the IANS guidelines specifically, the ability to obtain proper mentorship presented a significant hurdle. There are a limited numbers of HRA providers globally and they are primarily located in non-LMICs, making access for LMICs disproportionately costly [59].
The IANS guidelines [60] emphasize that effective implementation of HRA requires structured provider training, including supervised clinical practice and ongoing mentorship to ensure diagnostic accuracy and safety. Leveraging tele-mentorship platforms and identifying regional HRA providers to become future mentors can help address workforce limitations, reduce reliance on international experts, and enhance sustainability in resource-limited settings such as Nigeria [61,62]. These strategies align with the preferences expressed by Nigerian providers in our study, who emphasized the importance of mentorship and context-specific adaptation of training benchmarks to match the local burden of disease and available infrastructure.
Conversely, facilitators identified in this study offer important leverage points for optimizing the implementation process. The supportive stance of HRA providers and other stakeholders, coupled with their increasing confidence in adhering to the IANS guidelines, is a testament to the importance of organizational buy-in and capacity development. This support is particularly needed for SMM who are hesitant to access care in more traditional settings [31]. The pursuit of supportive and safe spaces for Nigerian SMM in this cohort has been documented and reflects their intra-community solidarity despite stigma [32]. SMM relied on peers for various forms of support, while also valuing the services and sanctuary provided by the research site [32]. Creating a safe and inclusive environment within HIV care facilities is essential, particularly for marginalized populations, and is reliant on the pivotal role of healthcare providers in fostering trust and patient engagement. Moreover, harnessing community social networks not only enhances awareness but also fosters a sense of collective empowerment and support, bolstering HRA uptake. Community engagement has been reported to increase other types of cancer prevention in Nigeria with recent studies reporting that such engagements led to notable rises in cervical cancer screening, clinical breast examination, and HPV vaccination uptake [63]. In addition to community engagement, engagement of policy makers and national organizations (e.g., Ministry of Health, National Institute for Cancer Research and Treatment) will be crucial for ensuring the sustainability and equitable accessibility of HRA in Nigeria. Although participants acknowledged this reality, especially in terms of mitigating patient costs, more discussion and work needs to be completed on how to effectively engage these entities of the outer context. Many efforts were taken to ensure methodological rigor, including verbatim transcription of interviews, independent coding with subsequent consensus-building, and the engagement of local team members. However, the limited number of coders may have introduced bias or overlooked certain nuances in the data analyses process, especially given that the coders were both non-Nigerians. Additionally, although saturation (i.e., consistent responding such that additional interviews were unlikely to uncover new information or themes) was reached, the findings of this study are specific to the TRUST clinic and may not generalize to other contexts in Nigeria or other LMICs. Furthermore, the absence of consistent demographic data such as age, gender identity, and educational attainment limits our ability to contextualize participants' perspectives and examine how experiences with HRA implementation may vary across subgroups. This omission reflects the original design of the interviews, which prioritized confidentiality and participant comfort, particularly given the sensitivity of the population and topic. These limitations highlight the need for additional research to explore a broader range of perspectives and employ more diverse methodological approaches to enhance contextualized understandings of facilitators and barriers to implementation of anal cancer screening in other LMICs.
This study examined barriers and facilitators to the implementation of HRA as per the IANS guidelines in Nigeria by integrating multiple perspectives from patients, providers, and healthcare system representatives. The study stakeholders' unwavering confidence in the IANS guidelines and the perceived benefits of HRA affirm the true potential to address critical health needs related to anal cancer within the Nigerian population. Sustaining this momentum requires ongoing collaboration with key stakeholders, including established organizations with existing ties to target populations. By fostering strategic partnerships and leveraging the infrastructure established through initiatives like the IMPACT study, the scalability and sustainability of HRA programs can be significantly enhanced.
## Conclusion
Successful implementation of HRA in Nigeria requires a multifaceted approach that addresses both structural barriers and leverages existing community facilitators. By prioritizing contextually relevant adaptations, fostering organizational support, and cultivating community engagement, stakeholders can work towards realizing the full potential of HRA in advancing public health outcomes and reducing the burden of preventable diseases. These insights can inform sustainable HRA implementation in Nigeria and provide a foundation for scaling anal cancer screening in other LMICs.
## References
1. Machalek, Poynten, Fairley et al. (1016) "Anal human papillomavirus infection and associated neoplastic lesions in men who have sex with men: a systematic review and meta-analysis" *Lancet Oncol*
2. Souza, Wiley, Li et al. (2008) "Incidence and epidemiology of anal cancer in the multicenter AIDS cohort study (MACS)" *J Acquir Immune Defic Syndr*
3. Deshmukh, Suk, Shiels et al. (2020) "Recent trends in squamous cell carcinoma of the anus incidence and mortality in the united States, 2001-2015" *J Natl Cancer Inst*
4. Blair, Valderrama-Beltran, Bautista-Arredondo et al. (1016) "Anal cancer screening in low-income and middle-income countries" *Lancet Gastroenterol Hepatol*
5. Borda, Bloem, Akaba et al. (2024) "Status of HPV disease and vaccination programmes in lmics: introduction to special issue" *Vaccine*
6. Tchelebi, Eng, Messick et al. (2022) "Current treatment and future directions in the management of anal cancer" *CA Cancer J Clin*
7. Sung, Ferlay, Siegel et al. (2021) "Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries" *CA Cancer J Clin*
8. Deshmukh, Damgacioglu, Georges et al. (2023) "Global burden of HPV-attributable squamous cell carcinoma of the anus in 2020, according to sex and HIV status: a worldwide analysis" *Int J Cancer*
9. Bamodu, Chung (2024) "Cancer care disparities: overcoming barriers to cancer control in Low-and Middle-Income countries" *JCO Glob Oncol*
10. Palefsky, Lee, Jay et al. (2022) "Treatment of anal high-grade squamous intraepithelial lesions to prevent anal cancer" *N Engl J Med*
11. Bruni, Albero, Serrano et al. (2021) "ICO/IARC Information Centre on HPV and Cancer (HPV Information Centre). Human papillomavirus and related diseases in the world" *Summary Report*
12. Palefsky (2012) "Practising high-resolution anoscopy" *Sex Health*
13. Richel, Prins, De Vries (2014) "Screening for anal cancer precursors: what is the learning curve for high-resolution anoscopy?" *Aids*
15. Hillman, Cuming, Darragh et al. (2016) "2016 IANS international guidelines for practice standards in the detection of anal cancer precursors" *J Low Genit Tract Dis*
17. Nowak, Ndembi, Dauda et al. (2019) "Implementation of and early outcomes from anal cancer screening at a community-engaged health care facility providing care to Nigerian men who have sex with men" *J Global Oncol*
18. Buchbinder, Timmermans (1177) "Newborn screening for metabolic disorders: parental perceptions of the initial communication of results" *Clin Pediatr (Phila)*
19. Kutner, Hou, Giguere et al. (1007) "To improve motivational barriers to retention in high resolution anoscopy, patients and providers recommend social and environmental changes: a sequential explanatory mixed-methods pilot study in a federally qualified health center" *AIDS Behav*
20. Apaydin, Nguyen, Panther et al. (2018) "Facilitators of and barriers to high-resolution anoscopy adherence among men who have sex with men: a qualitative study" *Sex Health*
21. Cruz, Ramos-Cartagena, Torres-Russe et al. (1186) "Barriers and facilitators to anal cancer screening among people living with HIV in Puerto Rico" *BMC Public Health*
22. Diez-Martínez, Perpiñá-Galvañ, Fernández-Alcántara et al. (2024) "Knowledge and perceived emotions regarding an anal cancer preventive program in people diagnosed with HIV: a qualitative study" *Cancer Nurs*
23. Hou, Giguere, Dewitt et al. (1007) "To improve motivational barriers to retention in high resolution anoscopy, patients and providers recommend social and environmental changes: a sequential mixed methods pilot study in a federally qualified health center"
24. Armstrong (2019) "Navigating the uncertainties of screening: the contribution of social theory" *Social Theory Health*
25. Gaspar, Grennan, Salit et al. (2018) "Confronting comorbidity risks within HIV biographies: gay men's integration of HPV-associated anal cancer risk into their narratives of living with HIV" *Health Risk Soc*
26. Silvera, Martinson, Gaisa et al. (2021) "The other side of screening: predictors of treatment and follow-up for anal precancers in a large health system" *AIDS*
27. Acree, Mcnulty, Blocker et al. (1080) "Shared decisionmaking around anal cancer screening among black bisexual and gay men in the USA" *Cult Health Sex*
28. Ong, Temple-Smith, Chen et al. (2015) "Why are we not screening for anal cancer routinely-HIV physicians' perspectives on anal cancer and its screening in HIV-positive men who have sex with men: a qualitative study" *BMC Public Health*
29. Petersen, Jaca, Ginindza et al. (1186) "Barriers to uptake of cervical cancer screening services in low-and-middleincome countries: a systematic review" *BMC Womens Health*
30. Schliemann, Ramanathan, Matovu et al. (0880) "The implementation of colorectal cancer screening interventions in low-and middle-income countries: a scoping review" *BMC Cancer*
31. Kirk, Kelley, Yankey et al. (2015) "A systematic review of the use of the consolidated framework for implementation research" *Implement Sci*
32. Cohen (2018) "Building trust in an LGBTQ-hostile country" *Science*
33. Schwartz, Nowak, Orazulike et al. (2015) "The immediate effect of the Same-Sex marriage prohibition act on stigma, discrimination, and engagement on HIV prevention and treatment services in men who have sex with men in Nigeria: analysis of prospective data from the TRUST cohort" *Lancet HIV*
34. Rodriguez-Hart, Bradley, German et al. (2023) "Building that strong energy: an exploration of stigma coping strategies among sexual and gender minorities in Nigeria" *Stigma Health*
35. Robbins, Dauda, Kokogho et al. (2020) "Oral sex practices among men who have sex with men and transgender women at risk for and living with HIV in Nigeria" *PLoS ONE*
36. Baral, Ketende, Schwartz et al. (2015) "Evaluating respondent-driven sampling as an implementation tool for universal coverage of antiretroviral studies among men who have sex with men living with HIV" *JAIDS J Acquir Immune Defic Syndr*
37. Charurat, Emmanuel, Akolo et al. (2015) "Uptake of treatment as prevention for HIV and continuum of care among HIV-positive men who have sex with men in Nigeria" *JAIDS J Acquir Immune Defic Syndr*
38. Osain (2011) "The Nigerian health care system: need for integrating adequate medical intelligence and surveillance systems" *J Pharm Bioallied Sci*
39. Scott-Emuakpor (2010) "The evolution of health care systems in nigeria: which way forward in the twenty-first century" *Nigerian Med*
40. Damschroder, Aron, Keith et al. (2009) "Fostering implementation of health services research findings into practice: a consolidated framework for advancing implementation science" *Implement Sci*
41. Varsi, Ekstedt, Gammon et al. (2015) "Using the consolidated framework for implementation research to identify barriers and facilitators for the implementation of an internet-based patient-provider communication service in five settings: a qualitative study" *J Med Internet Res*
42. Vandevanter, Kumar, Nguyen et al. (2017) "Application of the consolidated framework for implementation research to assess factors that May influence implementation of tobacco use treatment guidelines in the Viet Nam public health care delivery system" *Implement Sci*
43. Tuot, Sim, Nagashima-Hayashi et al. (2023) "What are the determinants of antiretroviral therapy adherence among stable people living with HIV? A cross-sectional study in Cambodia" *AIDS Res Ther*
44. Francis, Johnston, Robertson et al. (2010) "What is an adequate sample size? Operationalising data saturation for theory-based interview studies" *Psychol Health*
45. Ph, Ness (2015) "Are we there yet? Data saturation in qualitative research"
46. Lam, Quinn, Cipriano-Steffens et al. (2021) "Identifying actionable strategies: using consolidated framework for implementation research (CFIR)-informed interviews to evaluate the implementation of a multilevel intervention to improve colorectal cancer screening" *Implement Sci Commun*
47. Rankl, Johnson, Vindrola-Padros (2021) "Examining what we know in relation to how we know it: a team-based reflexivity model for rapid qualitative health research" *Qual Health Res*
48. Olmos-Vega, Stalmeijer, Varpio et al. (2023) "A practical guide to reflexivity in qualitative research: AMEE guide 149" *Med Teach*
50. Braun, Clarke (2006) "Using thematic analysis in psychology" *Qualitative Res Psychol*
51. Charmaz (2006) "Constructing grounded theory: A practical guide through qualitative analysis"
52. Smit (2002) "Atlas. ti for qualitative data analysis" *Perspect Educ*
53. Tong, Sainsbury, Craig (2007) "Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups" *Int J Qual Health Care*
54. Volpi, Chama, Mansfield et al. (2025) "Strengthening anal cancer prevention in Abuja, Nigeria: identifying barriers and potential strategies to improve training" *PLoS Glob Public Health*
55. Charúa-Guindic, Esquivel-Ocampo, Villanueva-Herrero et al. (2009) "La neoplasia intraepitelial anal y La infección Por virus Del papiloma humano En Pacientes anorreceptivos" *Rev Gastroenterol Mex*
56. Boldrini, Volpini, De Freitas et al. (2018) "Anal HPV infection and correlates in HIV-infected patients attending a sexually transmitted infection clinic in Brazil" *PLoS ONE*
57. Cardinal, Carballo, Lorenzo et al. (2014) "A six-year experience with anal cytology in women with HPV in the lower genital tract: utility, limitations, and clinical correlation" *Diagn Cytopathol*
58. Blair, Martinez-Vernaza, Ordonez-Blanco et al. (2021) "Screening with anal cytology in Colombia: initial experience and need for High-Resolution anoscopy" *J Surg Res*
60. Welch, Salicrup (2015) "Funding for Cancer Research and Clinical Studies in Low-and Middle-Income Countries" *Cancer Research and Clinical Trials in Developing Countries: A Practical Guide. edn*
61. Sivaram, Majumdar, Perin et al. (2018) "Population-based cancer screening programmes in low-income and middleincome countries: regional consultation of the international cancer screening network in India" *Lancet Oncol*
62. Gelband, Sankaranarayanan, Gauvreau et al. (2016) "Costs, affordability, and feasibility of an essential package of cancer control interventions in low-income and middle-income countries: key messages from disease control priorities" *Lancet*
64. Albuquerque, Nathan, Cappello et al. (1016) "Anal cancer and precancerous lesions: a call for improvement" *Lancet Gastroenterol Hepatol*
65. Stier, Clarke, Deshmukh et al. (2024) "International anal neoplasia society's consensus guidelines for anal cancer screening" *Int J Cancer*
66. Awuah, Tan, Bharadwaj et al. (1002) "Surgical mentorship in low-resource environments: opportunities and challenges, a perspective" *Health Sci Rep*
67. Hawkins, Fang (2022) "Anal squamous intraepithelial lesions and anal cancer management in low resource settings" *Clin Colon Rectal Surg*
68. Chigbu, Onyebuchi, Onyeka et al. (2017) "The impact of community health educators on uptake of cervical and breast cancer prevention services in Nigeria" *Int J Gynaecol Obstet* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12589879&blobtype=pdf | # Melioidosis clinically unresponsive to meropenem in a traveller returning from Thailand
Ilka Grewe, Anette Hennigs, Flaminia Olearo, Marylyn Addo, Michael Ramharter, Sabine Jordan
## Abstract
Burkholderia pseudomallei can cause severe systemic infection with involvement of multiple organs and abscess formation. Carbapenems are considered superior to ceftazidime in severe cases.We here report a case of Melioidosis in a previously healthy 57-year-old traveller. Despite in vitro testing confirming susceptibility of the isolate for meropenem, the patient's clinical condition did not improve during seven days of meropenem treatment. After changing the treatment regimen to high-dose ceftazidime, the patient's clinical condition and laboratory parameters rapidly improved. Thus, we here report a rare case of Melioidosis clinically unresponsive to meropenem.
## Introduction
Melioidosis is a common cause of pneumonia and septicaemia in South-East-Asia and Australia, but is rarely diagnosed in travellers returning from endemic areas [1]. Clinically, the infection can cause abscess formation and multiple organ involvement. Risk factors for a severe disease course include diabetes mellitus, chronic lung diseases and chronic kidney injury [2]. Burkholderia pseudomallei, the gram-negative bacteria causing melioidosis, is naturally resistant to a range of antibiotics. Resistance occasionally evolves during antibiotic treatment, but since B. pseudomallei is not spread from human-to-human resistant strains are not rapidly spread. Treatment includes an intensive intravenous antibiotic regimen during the acute phase, followed by an oral eradication therapy to prevent relapse [3,4]. High-dose ceftazidime has shown efficacy for initial therapy during the acute phase in a randomized clinical trial and may be used for non-critically-ill patients [5]. Since observational studies point towards superiority of carbapenems compared to ceftazidime, to date meropenem is considered to be the gold-standard for severe infections [6][7][8][9]. However, prospective clinical studies comparing carbapenem and cephalosporine therapy for melioidosis are scarce [10]. Carbapenem resistances have been described to evolve during therapy [11]. Here, we describe a case of severe infection with B. pseudomallei, that did not improve under anti-infective treatment with meropenem, but responded well to ceftazidime.
## Case
A 57-year-old female with no known pre-existing condition presented to our outpatient clinic with fever up to 39,7 • C, headache and dyspnea. One week prior to consultation she had returned from a sevenweek travel to Thailand and Cambodia, where she spent majority of her time in Bangkok, Phuket, Koh Samui and Siem Reap. Laboratory results revealed CrP-elevation of 218 mg/L, white blood count of 11.0 × 10 9 /L and an increased blood sedimentation rate of 81 mm (Fig. 1A). A chest X-ray showed a consolidation in the left upper lung. An empirically initiated treatment with amoxicillin and clavulanic acid (875/125 mg orally every eight hours) for suspected community-acquired pneumonia did not lead to improvement of the patient's clinical condition. The radiographic finding in the chest X-ray presented as a solid consolidation with a caverna in a consecutive CT-scan (Fig. 1B). Since the patient's condition continuously worsened, including progressive dyspnea and somnolence, she was admitted to our clinic and antibiotic therapy was escalated to meropenem (1000 mg intravenously every eight hours). Notably, fever, dyspnea and headache continuously worsened and CrP increased under meropenem treatment (Fig. 1A). In a bronchoalveolar lavage Burkholderia pseudomallei was identified via culture and subsequent Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF). Although antimicrobial susceptibility testing showed the isolate to be susceptible to meropenem with a Minimum Inhibitory Concentration (MIC) of 1 μg/mL, the patient's condition did not improve during seven days of meropenem treatment. Therefore, antibiotic treatment was changed to high-dose ceftazidime (2000 mg intravenously every six hours), for which the isolate was also tested susceptible with a MIC of 1 μg/mL. Subsequently, the patient's clinical condition rapidly improved and laboratory parameters normalized. A control CT-scan showed regression of the cavernous consolidation in the left upper lung (Fig. 1C). Following 14 days of intravenous ceftazidime, oral cotrimoxazole/trimethoprim was initiated during which the patient experienced a relapse of fever. Eradication therapy was therefore switched to doxycycline, which was administered for six months, resulting in full recovery.
## Discussion
B. pseudomallei is highly endemic in Southeast Asia and northern Australia, where B. pseudomallei is commonly found in soil and water [7]. While most severe cases of Melioidosis occur in patients with chronic conditions such as chronic lung disease, diabetes mellitus or chronic kidney injury, we here present a severe case of Melioidosis in a previously healthy traveller returning from Thailand to Germany [12,13].
Natural resistances of B. pseudomallei towards carbapenems have not been reported previously, but carbapenem resistances occasionally evolve during treatment [11,14]. In vitro susceptibility testing showed that the vast majority of isolates in Thailand were sensitive to meropenem (98 %), along with high susceptibility rates to ceftazidime [15]. Epidemiological data on clinical responsiveness to antibiotic treatment in the area are scarce. Also globally, meropenem resistance is extremely rare in patients without previous meropenem exposure [16,17]. Since our patient did not have any history of previous carbapenem treatment, this case represents a rare case of Melioidosis that did not respond to initial treatment with meropenem.
Previous studies showed superiority of carbapenems compared to ceftazidime in treatment of severe melioidosis [10,18]. This case emphasises that in carbapenem unresponsive cases, a treatment regimen with high-dose ceftazidime can be considered. In vitro susceptibility testing did not confirm meropenem-resistance. Nevertheless, disease severity continuously worsened during seven days of meropenem treatment and rapidly mitigated after change to a high-dose ceftazidime regimen. The reason for the discrepancy between in vitro antimicrobial resistance testing and clinical treatment failure remains unclear. A previous study reported growth-defective B. pseudomallei associated with ceftazidime treatment failure, leading to a lack of in vitro detection of resistant isolates [19]. This phenomenon has not been described for meropenem-resistant isolates yet.
A key contributor to the carbapenem resistance in B. pseudomallei is the active efflux of antibiotics mediated by resistance-nodulation-celldivision (RND) efflux pumps [11,20]. Upregulation of expression of efflux pumps usually occurs during antibiotic treatment and is only rarely observed in initial isolates. While this may represent a possible underlying mechanism for the observed clinical treatment failure, gene expression of efflux pumps was unfortunately not assessed in the present case. Furthermore, mutations upstream of the penA gene, leading to an overexpression of class A β-lactamase can result in ceftazidime resistance and also in decreased meropenem sensitivity [21]. Given the patient's clinical response to ceftazidime, a penA-mediated mechanism of resistance appears unlikely in this case. Due to the single-case-design of the study, no further direct conclusions on antibiotic therapy can be drawn. However, this case emphasises the need for awareness and further investigation on carbapenem unresponsive Melioidosis.
## Conclusion
We here report treatment failure of meropenem as initial therapy for a severe case of B. pseudomallei infection. Although resistance was not confirmed microbiologically, clinical symptoms and laboratory results reflect a better response to ceftazidime. Thus, this is a rare case in which ceftazidime treatment was superior to meropenem.
## CRediT authorship contribution statement
## Ethical approval
The study was conducted in accordance with the local legislation and institutional requirements. The patient granted permission to publish her medical history.
## Patient consent
The patient granted permission to publish her medical history.
## References
1. Limmathurotsakul, Golding, Dance et al. (2016) "Predicted global distribution of burkholderia pseudomallei and burden of melioidosis" *Nat Microbiol*
2. Koshy, Jagannati, Ralph et al. (2019) "Clinical manifestations, antimicrobial drug susceptibility patterns, and outcomes in melioidosis cases" *India. Emerg Infect Dis*
3. Chaowagul, Suputtamongkol, Dance et al. (1993) "Relapse in melioidosis: incidence and risk factors" *J Infect Dis*
4. Limmathurotsakul, Chaowagul, Chierakul et al. (2006) "Risk factors for recurrent melioidosis in northeast Thailand" *Clin Infect Dis*
5. White, Dance, Chaowagul et al. (1989) "Halving of mortality of severe melioidosis by ceftazidime" *Lancet*
6. Dance (2014) "Treatment and prophylaxis of melioidosis" *Int J Antimicrob Agents*
7. Meumann, Limmathurotsakul, Dunachie et al. (2024) "Burkholderia pseudomallei and melioidosis" *Nat Rev Microbiol*
8. Simpson, Opal, Angus et al. (2000) "Differential antibiotic-induced endotoxin release in severe melioidosis" *J Infect Dis*
9. Cheng, Fisher, Anstey et al. (2004) "Outcomes of patients with melioidosis treated with meropenem" *Antimicrob Agents Chemother*
10. Simpson, Suputtamongkol, Smith et al. (1999) "Comparison of imipenem and ceftazidime as therapy for severe melioidosis" *Clin Infect Dis*
11. Sarovich, Webb, Pitman et al. (2018) "Raising the stakes: loss of efflux pump regulation decreases meropenem susceptibility in burkholderia pseudomallei" *Clin Infect Dis*
12. Currie, Mayo, Ward et al. (2021) "The Darwin prospective melioidosis study: a 30-year prospective, observational investigation" *Lancet Infect Dis*
13. Pacheco, Chea, Saunders et al. (2024) "Case report: risk factors associated with mortality in adults with burkholderia pseudomallei bacteremia: a retrospective case series of melioidosis in Cambodia" *Am J Trop Med Hyg*
14. Zamani, Zueter, Besari et al. (2020) "Reduced susceptibility of burkholderia pseudomallei following exposure to carbapenem" *Trop Biomed*
15. Paveenkittiporn, Apisarnthanarak, Dejsirilert et al. (2009) "Five-year surveillance for burkholderia pseudomallei in Thailand from 2000 to 2004: prevalence and antimicrobial susceptibility" *J Med Assoc Thai*
16. Crowe, Mcmahon, Currie et al. (2014) "Current antimicrobial susceptibility of first-episode melioidosis burkholderia pseudomallei isolates from the Northern Territory" *Australia. Int J Antimicrob Agents*
17. Rao, Hu, Chen et al. (2019) "Molecular epidemiology and antibiotic resistance of burkholderia pseudomallei isolates from hainan, China: a STROBE compliant observational study" *Med (Baltim)*
18. Smith, Wuthiekanun, Walsh et al. (1996) "In-vitro activity of carbapenem antibiotics against beta-lactam susceptible and resistant strains of burkholderia pseudomallei" *J Antimicrob Chemother*
19. Chantratita, Rholl, Sim et al. (2011) "Antimicrobial resistance to ceftazidime involving loss of penicillin-binding protein 3 in burkholderia pseudomallei" *Proc Natl Acad Sci*
20. Kumar, Mayo, Trunck et al. (2008) "Expression of resistance-nodulation-cell-division efflux pumps in commonly used burkholderia pseudomallei strains and clinical isolates from Northern Australia" *Trans R Soc Trop Med Hyg*
21. Suchartlikitwong, Saninjuk, Tirapattanun et al. (2025) "Emergence of ceftazidime resistance in burkholderia pseudomallei during therapy: clinical, phenotypic and genotypic insights from paired isolates" *J Glob Antimicrob Resist* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12596415&blobtype=pdf | # Utilizing HIV Proviral DNA to Assess for the Presence of HIV Drug Resistance
Annemarie Wensing, Charlotte Charpentier, Vincent Calvez, Francesca Ceccherini-Silberstein, Huldrych Günthard, Donna Jacobsen, Roger Paredes, Robert Shafer, Douglas Richman
## Abstract
The improved efficacy and tolerability of newer antiretroviral drugs, as well as the introduction of long-acting regimens, have prompted more frequent therapy switches in individuals on suppressive antiretroviral therapy (ART). For these individuals, the assessment of HIV drug resistance using DNA from peripheral blood lymphocytes has become increasingly popular. However, compared with HIV RNA-based analyses, implementation of HIV DNA testing as an alternative approach in clinical care requires new documented quality-assessment procedures and clinical validation. Furthermore, the use of HIV DNA to assess drug resistance has some distinct technical and biologic challenges that are relevant to the clinical management of people with HIV. This Viewpoint article addresses the issues relevant to clinical virologists and treating physicians for the interpretation of drug-resistance testing or subtype assessment based on DNA analysis, when HIV RNA genotypic assessment is not possible.
Human immunodeficiency virus (HIV) RNA genotypic resistance testing is a crucial clinically validated tool to optimize antiretroviral treatment-related decisions. The International Antiviral Society-USA (IAS-USA) HIV Drug Resistance Group maintains a list of relevant mutations in HIV that impact drug susceptibility [1] and has previously recommended HIV-RNA resistance testing for individuals who are drug naive at the time of presentation and in those experiencing failure of therapy who have an HIV-RNA level above 200 copies/mL [2]. However, there are specific situations in which a resistance test may be warranted when HIV-RNA testing is not possible. Such situations could include repeated failure of RNA amplification at lower levels of viremia or a switch to a new antiretroviral regimen in the setting of viral suppression. These may include a switch to a new regimen in the absence of a clearly documented treatment history, or insufficient access to historical resistance test results performed at baseline, or at earlier times of failure, or a switch of therapy for which information is needed that is not documented in a previous resistance report. Some of these situations may be overcome by performing HIV-RNA resistance testing on stored plasma samples; however, consistent storage of past samples is expensive and often unavailable.
Recently, there has been an increased use of proviral DNA as an alternative source for HIV resistance testing in clinical trials and in routine practice [3,4]. Although the rationale for the potential use of resistance testing using viral DNA isolated from infected cells in the situations described above seems reasonable, several technical and biologic challenges associated with this approach need to be appreciated.
## TECHNICAL ASPECTS OF HIV DNA RESISTANCE TESTING
To perform sequencing of proviral HIV DNA variants, DNA can be extracted directly from a whole-blood sample or from peripheral blood mononuclear cells (PBMCs) and then amplified using nested polymerase chain reaction (PCR) [5]. The amplified DNA is sequenced using dideoxy-terminator Sanger sequencing or next-generation sequencing (NGS). Sanger sequencing reads individual DNA fragments, but NGS allows for massively parallel sequencing of DNA fragments, enabling higher throughput and greater sensitivity. The number of viral genomes analyzed in the procedure depends on the PBMC proviral load, the number of PBMCs from which DNA is extracted, and the efficiency of DNA extraction and PCR [5].
Amplification has been reported to be successful even in individuals with low CD4+ cell counts, likely because these individuals tend to have higher proviral DNA loads [6].
There is more random variation in the detection of drug-resistance mutations (DRMs) within a single sample when analyzing PBMCs than when analyzing plasma. This is because DRMs in PBMCs are more likely to be mixed with ancestral virus populations [5,7]. Indeed, the mean reproducibility at detecting DRMs on repeat assays was approximately 80% in 2 different studies [8,9]. Sampling bias is also likely to affect reproducibility when variants are present in low proportions within the proviral DNA, because fewer HIV sequences are available for interrogation. One approach to mitigate this limitation has been to perform triplicate-nested PCR [10].
In several countries, academic laboratories offer DNA resistance testing as a technically validated procedure [6]. Sanger sequencing can detect the presence of mutations at levels of approximately 20% in the viral population [11,12]. In contrast, the proportion at which variants can be detected by NGS depends on the threshold selected by the laboratory, which usually ranges between 1% and 10%. One commercially available, technically validated DNA resistance assay using NGS has a mutation detection threshold of approximately 3% to 10%, with the lower threshold applied to mutations in sequences that are not consistent with APOBEC (apolipoprotein B mRNA editing enzyme) alterations [13], as described below.
Regardless of the assay used, the interpretation of DRMs in proviral DNA requires an approach that considers the near-universal presence of APOBEC-mediated G-to-A hypermutation at some level in PBMC HIV DNA [14,15]. APOBEC 3F and 3G are host enzymes that cripple viral genomes by indiscriminately mutating the dinucleotides GG to AG and GA to AA, respectively. In most cases, sequences that contain APOBEC3-mediated G-to-A hypermutation also have stop codons resulting from G-to-A changes at tryptophan amino acids or changes from aspartate to asparagine at 1 or more active site positions. The DRMs resulting from G-to-A hypermutation (rather than from drug selection pressure) are likely to be present in replication-incompetent viruses that also contain stop codons and other crippling mutations.
Eighteen DRMs arise within an APOBEC3-mediated dinucleotide context (Table 1) [16]. For NGS, 2 approaches can be applied to account for APOBEC3-mediated DRMs: (1) excluding sequence reads that are hypermutated or (2) determining the proportion of sequences that are hypermutated and then reporting only those APOBEC3-context DRMs that occur above this proportion. Of note, studies using long-read NGS showed that stop codons and G-to-A hypermutation are mostly present on the same read [17]. It is essential that laboratory drug-resistance reports confirm that APOBEC3-mediated mutations have been purged from DNA resistance results when using NGS sequencing.
For Sanger sequencing it can be more challenging to discriminate whether resistance mutations are present in replicationcompetent variants or in APOBEC-mediated defective virus [18]. DNA resistance reports based on Sanger sequencing should acknowledge this limitation and indicate which mutations may be mediated by APOBEC. In addition, it is crucial The number of viral genomes able to be sampled depends on the PBMC proviral copy number, the number of PBMCs from which DNA is extracted, and the efficiencies of DNA extraction and PCR amplification. c There is greater intra-sample stochastic variation in the detection of DRMs in PBMCs than in plasma, because DRMs in PBMCs are more likely to coexist with ancestral wild-type virus populations. In addition, plasma is a homogeneous liquid source, but the number of HIV-infected CD4+ cells can vary in the PBMCs according to the immunologic status of the individual.
that publications and presentations about studies using HIV DNA drug-resistance testing indicate that the effects of APOBEC editing have been accounted for in the results.
## POTENTIAL INDICATIONS, INTERPRETATION, AND LIMITATIONS OF HIV DNA RESISTANCE TESTS
## Potential Indications
There are several situations in which HIV DNA resistance analysis may have added value (see Text box), as follows:
1. Repeated failure of RNA amplification at baseline or time of virologic failure, especially at lower levels of viremia when amplification of HIV-RNA may fail. 2. A switch to a new regimen in individuals who are virally suppressed in the absence of a clearly documented treatment history, or insufficient access to historic resistance tests performed at baseline or during previous failure: (a) When a switch is considered to a regimen with a lower genetic barrier to resistance than the current suppressive regimen. (b) When recycling a drug class that was previously used at times of insufficient or unknown levels of viral suppression. (c) When using a drug in the nonnucleoside analogue reverse transcriptase inhibitor (NNRTI) class if an NNRTI-based regimen was previously interrupted without immediate switch to a new regimen since NNRTI resistance may have been selected after interruption due to prolonged presence of insufficient drug levels [19,20]. (d) When the viral subtype is not recorded in previous reports. This information may be useful when selecting antiretroviral drugs in certain geographic regions. Also, an earlier reported subtype may not be adequately classified based on current knowledge. This is particularly relevant for subtype A6, which has been often misclassified as subtype A1. Subtype A6 has been recently identified as 1 of the risk factors for virologic failure of the injectable long-acting combination of the NNRTI rilpivirine and the integrase strand transfer inhibitor (InSTI) cabotegravir [21]. (e) When insight regarding resistance-related polymorphisms is needed. Circulating wild-type viral polymorphisms may diminish antiviral activity or lower the genetic barrier to clinically relevant resistance to certain drugs. Particularly the reverse transcriptase (RT) E138A polymorphism related to rilpivirine and the integrase E157Q polymorphism related to low-level resistance to the first-generation InSTIs raltegravir and elvitegravir. Each of these mutations occurs in 1% to 6% of isolates from antiretroviral therapy (ART)-naive persons depending on subtype [22][23][24]. (f) When co-receptor tropism is required for a switch to the CCR5-receptor inhibitor maraviroc. In general, a high concordance of co-receptor tropism assignment based on RNA and DNA has been reported in sample sets with predominantly CCR5-tropic viruses in the plasma. However, CXCR4-tropic viruses may be more often detected in DNA than in RNA [25,26]. 3. The initiation of therapy in individuals not on ART, since in the absence of selective pressure, resistant variants may be archived in the DNA but replaced in the plasma by more replication-competent variants with less or no resistance.
(a) When, in individuals who are previously ART exposed, ART is reinitiated after discontinuation for several weeks without assessment of resistance at time of earlier (presumed) therapy failure and without access to a stored plasma sample at the time of failure [27,28]. (b) When individuals who are chronically infected and therapy-naive initiate ART while there is a high likelihood of initial infection with drug-resistant HIV. In cases of infection with resistant virus, no wild-type virus is archived in the DNA. The chance of detection of resistance mutations in the plasma differs depending on the fitness costs they induce. Mutations may be still detectable in DNA once they have cleared from plasma [13,[29][30][31].
## Interpretation and Limitations
Various randomized clinical studies have shown the added value of HIV RNA resistance testing to guide treatment decisions, but no such studies have been performed for HIV DNA resistance testing [32][33][34]. As HIV DNA resistance tests are now commercially available and used in clinical trials to assess resistance, it is important that clinicians are aware of the limitations and uncertainties regarding the use of proviral DNA to detect drug resistance for guiding treatment decisions (Table 1). HIV DNA resistance test results thus need to be interpreted with caution. HIV DNA resistance testing should only be considered when HIV DNA testing may be useful when RNA resistance testing results cannot be generated due to inability to perform RNA amplification.
HIV DNA testing could provide supplementary information on previously selected resistance in individuals who are suppressed with antiretroviral therapy when there is an incomplete therapeutic history, absence of sequence results, or missing information on earlier reports about subtype or coreceptor tropism.
HIV DNA testing could provide supplementary information on baseline resistance either due to resistance-related polymorphisms or transmitted resistance.
plasma HIV RNA testing is not technically feasible. When making treatment decisions, one must take into account the treatment history, the history of previous treatment failures, and the alternative treatment options available for that particular individual. It is uncertain whether HIV DNA resistance testing information will lead to better treatment choices than clinical judgment based on these considerations. On the one hand, clinicians must be aware that drug-resistant variants may be missed by proviral DNA testing. At the time of therapy failure, RNA resistance testing is preferred since resistance mutations are detected in the plasma earlier than in DNA [35,36]. Also, in ART-suppressed individuals, the negative-predictive value of HIV DNA resistance tests is lower than the evaluation of cumulative historical RNA-based resistance tests performed at the time of failure [37,38]. This is likely based on the fact that decline in archived resistant viruses in the viral reservoir increases with the duration of virologic suppression, once an effective ART regimen is initiated after virologic failure [39,40]. The kinetics of decay depend on the DRMs and several other factors, such as the duration and level of viral replication at the time of virologic failure [41]. In addition, only a subset of as much as 2 to 3 million PBMCs is assessed with HIV DNA resistance tests, representing a very small fraction of the viral reservoir. In addition, these cells may not be informative of the viral reservoir that is present in lymph nodes or other anatomic compartments. If Sanger technology is used for sequencing, only viral variants representing more than 20% of all viral quasi-species may be detected and minority resistant variants may be overlooked. This limitation is likely more relevant considering the overall decreased sensitivity of HIV DNA testing compared with RNA testing. Selection of therapy based on the absence of DRMs in an HIV DNA test may therefore be risky, especially when switching to a regimen with a lower genetic barrier to resistance.
On the other hand, DNA resistance testing may overestimate resistance. The viral reservoir is constituted primarily of defective proviruses, mainly due to large internal deletions and not just due to APOBEC-induced mutations [14]. Some DRMs detected in proviral DNA may occur in nonviable viruses and thus unlikely to cause virologic failure. Avoiding a certain treatment based on the detection of DRMs with an HIV DNA test may be too restrictive and disqualify beneficial therapeutic options.
## CONCLUSIONS
The use of HIV DNA resistance tests can lead to an underestimation of DRMs due to technical limitations of the methodology and to a low negative-predictive value, as well as possible overestimation of resistance mutations due to the presence of APOBEC3-induced defective proviruses. HIV drug-resistance testing using DNA may have added value for the indications described in this article. Given the limitations and technical challenges, vigorous quality control is warranted. It is important that resistance reports clearly indicate when resistance results are based on DNA. Implementation of antiretroviral drug-resistance testing using viral DNA in clinical practice with increased confidence will require additional rigorous clinical investigation.
## Notes
## References
1. Wensing, Calvez, Ceccherini-Silberstein (2025) "2025 Update of the drug resistance mutations in HIV-1" *Top Antivir Med*
2. Günthard, Calvez, Paredes (2019) "Human immunodeficiency virus drug resistance: 2018 recommendations of the International Antiviral Society-USA panel" *Clin Infect Dis*
3. Kityo, Mambule, Musaazi (2024) "Switch to long-acting cabotegravir and rilpivirine in virologically suppressed adults with HIV in Africa (CARES): week 48 results from a randomised, multicentre, open-label, non-inferiority trial" *Lancet Infect Dis*
4. Botha, Byott, Spyer (2023) "Sensitive HIV-1 DNA pol next-generation sequencing for the characterisation of archived antiretroviral drug resistance" *Viruses*
5. Chu, Armenia, Walworth et al. (2022) "Genotypic resistance testing of HIV-1 DNA in peripheral blood mononuclear cells" *Clin Microbiol Rev*
6. Jaha, Schenkel, Jorimann (2023) "Prevalence of HIV-1 drug resistance mutations in proviral DNA in the Swiss HIV cohort study, a retrospective study from 1995 to 2018" *J Antimicrob Chemother*
7. Milliere, Bocket, Tinez (2021) "Assessment of intra-sample variability in HIV-1 DNA drug resistance genotyping" *J Antimicrob Chemother*
8. Antoni, Andreatta, Acosta et al. (2021) "HIV-1 DNA genotyping is often variable in repeat testing from single blood draws [CROI abstract 438"
9. Curanovic, Cai, Toma et al. (2020) "Results of repeat HIV-1 DNA resistance tests are highly concordant" *Open Forum Infect Dis*
10. Lubke, Cristanziano, Sierra (2015) "Proviral DNA as a target for HIV-1 resistance analysis" *Intervirology*
11. Woods, Brumme, Liu (2012) "Automating HIV drug resistance genotyping with RECall, a freely accessible sequence analysis tool" *J Clin Microbiol*
12. Schuurman, Demeter, Reichelderfer et al. (1999) "Worldwide evaluation of DNA sequencing approaches for identification of drug resistance mutations in the human immunodeficiency virus type 1 reverse transcriptase" *J Clin Microbiol*
13. Curanovic, Martens, Rodriguez et al. (2023) "HIV-1 DNA testing in viremic patients identifies more drug resistance than HIV-1 RNA testing" *Open Forum Infect Dis*
14. Bruner, Murray, Pollack (2016) "Defective proviruses rapidly accumulate during acute HIV-1 infection" *Nat Med*
15. Li, Etemad, Dele-Oni (2021) "Drug resistance mutations in HIV provirus are associated with defective proviral genomes with hypermutation" *AIDS*
16. Tzou, Pond, Avila-Rios et al. (2020) "Analysis of unusual and signature APOBEC-mutations in HIV-1 pol nextgeneration sequences" *PLoS One*
17. Jorimann, Tschumi, Zeeb (2023) "Absence of proviral human immunodeficiency virus (HIV) type 1 evolution in early-treated individuals with HIV switching to dolutegravir monotherapy during 48 weeks" *J Infect Dis*
18. Armenia, Gagliardini, Alteri (2023) "Temporal trend of drug-resistance and APOBEC editing in PBMC genotypic resistance tests from HIV-1 infected virologically suppressed individuals" *J Clin Virol*
19. Jourdain, Ngo-Giang-Huong, Coeur (2004) "Intrapartum exposure to nevirapine and subsequent maternal responses to nevirapine-based antiretroviral therapy" *N Engl J Med*
20. Cressey, Green, Khoo (2008) "Plasma drug concentrations and virologic evaluations after stopping treatment with nonnucleoside reverse-transcriptase inhibitors in HIV type 1-infected children" *Clin Infect Dis*
21. Cutrell, Schapiro, Perno (2021) "Exploring predictors of HIV-1 virologic failure to long-acting cabotegravir and rilpivirine: a multivariable analysis" *AIDS*
22. Charpentier, Malet, Garnier (2018) "Phenotypic analysis of HIV-1 E157Q integrase polymorphism and impact on virological outcome in patients initiating an integrase inhibitor-based regimen" *J Antimicrob Chemother*
23. Lambert-Niclot, Charpentier, Storto (2013) "Prevalence of pre-existing resistance-associated mutations to rilpivirine, emtricitabine and tenofovir in antiretroviral-naive patients infected with B and non-B subtype HIV-1 viruses" *J Antimicrob Chemother*
24. Rhee, Gonzales, Kantor et al. (2003) "Human immunodeficiency virus reverse transcriptase and protease sequence database" *Nucleic Acids Res*
25. Fabeni, Berno, Svicher (2015) "Genotypic tropism testing in HIV-1 proviral DNA can provide useful information at low-level viremia" *J Clin Microbiol*
26. Meini, Rossetti, Bianco (2014) "Longitudinal analysis of HIV-1 coreceptor tropism by single and triplicate HIV-1 RNA and DNA sequencing in patients undergoing successful first-line antiretroviral therapy" *J Antimicrob Chemother*
27. Paquet, Baxter, Weidler (2011) "Differences in reversion of resistance mutations to wild-type under structured treatment interruption and related increase in replication capacity" *PLoS One*
28. Trignetti, Sing, Svicher (2009) "Dynamics of NRTI resistance mutations during therapy interruption" *AIDS Res Hum Retroviruses*
29. Pingen, Nijhuis, De Bruijn et al. (2011) "Evolutionary pathways of transmitted drug-resistant HIV-1" *J Antimicrob Chemother*
30. Yang, Kouyos, Boni (2015) "Persistence of transmitted HIV-1 drug resistance mutations associated with fitness costs and viral genetic backgrounds" *PLoS Pathog*
31. Kuhnert, Kouyos, Shirreff (2018) "Quantifying the fitness cost of HIV-1 drug resistance mutations through phylodynamics" *PLoS Pathog*
32. Baxter, Mayers, Wentworth (2000) "A randomized study of antiretroviral management based on plasma genotypic antiretroviral resistance testing in patients failing therapy. CPCRA 046 study team for the Terry Beirn community programs for clinical research on AIDS" *AIDS*
33. Tural, Ruiz, Holtzer (2002) "Clinical utility of HIV-1 genotyping and expert advice: the Havana trial" *AIDS*
34. Durant, Clevenbergh, Halfon (1999) "Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled trial" *Lancet*
35. Kroodsma, Kozal, Hamed et al. (1994) "Detection of drug resistance mutations in the human immunodeficiency virus type 1 (HIV-1) pol gene: differences in semen and blood HIV-1 RNA and proviral DNA" *J Infect Dis*
36. Havlir, Gamst, Eastman et al. (1996) "Nevirapine-resistant human immunodeficiency virus: kinetics of replication and estimated prevalence in untreated patients" *J Virol*
37. Wirden, Soulie, Valantin (2011) "Historical HIV-RNA resistance test results are more informative than proviral DNA genotyping in cases of suppressed or residual viraemia" *J Antimicrob Chemother*
38. Delaugerre, Braun, Charreau (2012) "Comparison of resistance mutation patterns in historical plasma HIV RNA genotypes with those in current proviral HIV DNA genotypes among extensively treated patients with suppressed replication" *HIV Med*
39. 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*
40. Nouchi, Nguyen, Valantin (2018) "Dynamics of drug resistance-associated mutations in HIV-1 DNA reverse transcriptase sequence during effective ART" *J Antimicrob Chemother*
41. Abdi, Palich, Seang (2024) "Clearance of archived integrase strand transfer inhibitors resistance mutations in people with virologically suppressed HIV infection" *JAC Antimicrob Resist* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12802262&blobtype=pdf | # Early mucosal IFN-α, IP-10, and IL-1RA and synchronized mucosal and systemic immune responses mediate COVID-19 disease progression
Mona Agrawal, Armando Flores-Torres, John Franks, Sarah Lang, Thomas Fabrizio, Kristin Mcnair, Laura Boywid, Ashley Blair, Chloe Hundman, Nicholas Hysmith, Michael Whitt, Rachael Keating, Paul Thomas, Richard Webby, Amanda Green, Heather Smallwood, Paul 75n93021c00018, Thomas
## Abstract
Mucosal immunity plays a crucial role in protection against respiratory viruses. However, the mechanisms underlying mucosal responses and their impact on COVID-19 outcomes are not well understood, as mucosal immunity is compartmental ized and not always reflected in the bloodstream. This study examined primary immune responses in 584 mucosal and blood specimens collected over a month from previ ously naïve adults and children with COVID-19. Various laboratory techniques were utilized to quantify and characterize viral RNA, antigens, antibodies, and cytokines in the samples, including PCR, sequencing, ELISA, and Luminex. Comprehensive system analysis uncovered distinctive characteristics associated with mild COVID-19 disease progression, including markedly early and elevated induction of mucosal IFN-α and IP-10, followed by increased levels of IL-1RA and IgG. Individuals experiencing mild COVID-19 demonstrated synchronized mucosal and systemic immune responses, with a gradual increase in antibody production that resulted in enhanced neutralization potency, potentially conferring greater protection against future infection. In contrast, individuals with moderate and severe COVID-19 exhibited diminished IFN-α and IP-10 responses and dysregulated mucosal and systemic immune responses marked by rapid and robust yet less effective humoral immunity, potentially driven by high antigen and cytokine levels in both compartments. Collectively, these findings underscore that early mucosal immune responses may play a pivotal role in attenuating COVID-19 disease severity. Additionally, they suggest that primary mucosal immune responses to novel viruses influence clinical outcomes, providing critical insights necessary for developing prognostic indicators, treatments, and mucosal vaccines that confer protection against SARS-CoV-2 and emerging respiratory pathogens. IMPORTANCE This research is crucial for understanding the intricate interplay between mucosal immunity and SARS-CoV-2 infection. By examining the distinct systemic and mucosal immune responses during COVID-19, this study addresses the critical gap in our knowledge of how the body defends itself at the primary site of infection: the respiratory mucosa. The findings shed light on the specific characteristics of the mucosal immune response, including the roles of different antibody isotypes, immune cells, and local factors in controlling viral entry and replication. Furthermore, because this study focuses on de novo immune responses, the results may have broad implications for understanding immune responses to future novel pathogens. Ultimately, this research will contribute to the development of more effective diagnostic tools, therapeutic strategies, and mucosal vaccines to prevent and control COVID-19. By focusing on the often-overlooked mucosal compartment, this work offers a new perspective on
## RESULTS
## Population and specimens
To assess primary immune responses to a novel respiratory viral infection, 77 partici pants, enrolled between July 15, 2020, to March 9, 2021, were classified as having mild (n = 27), moderate (n = 30), or severe (n = 20) COVID-19 (Table 1). Demographics and clinical data are summarized in Table 2. Subjects were equally distributed among severity groups. We used 584 longitudinally collected mucosal and blood samples in our analysis (Fig. 1).
## Efficient viral and antigen clearance associated with mild disease progression
All subjects with mild symptoms cleared the virus within a week of diagnosis, whereas 10% and 15% of those who progressed to moderate and severe disease remained PCR positive (Fig. S1A). Individuals with mild or moderate COVID-19 significantly reduced mucosal viral load over time, but this was not the case for those with severe illness (Fig. S1A). Only ancestral B.1 lineages were detected in patients' swabs, including B.1.223, B.1.369, B.1.2, and related sub-lineages (Fig. S1B). Importantly, no Alpha (B. 1.1.7) or other variants of concern were detected during the study period. N and S antigens were detected in the mucosa within one week of diagnosis (acute infection phase); 14% and 48% of subjects were S and N positive, respectively. Systemic S was significantly higher in severe cases (Fig. S1C). N was not detected in plasma. Notably, within a week of diagnosis, class switching from IgM+/IgG-to IgM+/IgG+ was indicative of disease severity. Nearly 50% of participants who developed severe disease, compared to under 10% of those with mild disease, were IgM+/IgG+ (Fig. S1D). By four weeks, approximately 90% of severe and 50% of mild cases were IgM+/IgG+ (Fig. S1D). Severe cases failed to rapidly reduce viral load, exhibited higher antigen levels, and showed faster isotype switching. These data suggest that individuals who rapidly initiate mucosal viral clearance develop mild illness.
## Controlled mucosal and peripheral humoral responses linked to mild disease progression
To determine the relationship between mucosal and peripheral humoral responses and disease progression, we quantified S-and N-specific antibodies over time. Nasoconver sion rates and mucosal IgA levels were high, irrespective of severity (Fig. 2A; Fig. S2A andC). However, individuals with mild to moderate illness significantly increased mucosal IgG production over time (Fig. S2C). Seroconversion rates and systemic IgA, IgG, and IgM levels during week 1 significantly increased as clinical outcomes worsened (Fig. 2B; Fig. S2B). Systemic IgA, IgG, and IgM production steadily increased over time with mild disease (Fig. S2D). Mild cases also displayed a transient increase in IgM that peaked at week 2, reflecting class switching in week 3, yet IgM levels remained elevated in moderate and severe cases, indicating continued antigen recognition and antibody synthesis (Fig. S2D). In recovery, despite similar ACE2 receptor binding inhibition and titers, subjects with mild illness produced antibodies with significantly higher neutraliza tion potency (Fig. 2C). They also exhibited a significant positive correlation between IgG levels and neutralization (Fig. 2D). Only the mild cases had strong and significant positive correlations between local and peripheral humoral responses (Fig. 2E).
## Distinct mucosal and systemic immune response dynamics associated with clinical outcomes
To study how disease progression in previously naïve individuals correlates with mucosal and systemic immunity, we quantified immune factors in NRF and plasma (Fig. S3A and B, respectively). To understand the involvement of immune factors at mucosa and periphery during different stages of viral infection and immune responses, we analyzed immune factors among severity groups during early (0-4 days), adaptive (5-10 days), resolution (11-21 days), and convalescence (>21 days) phases. Individuals who devel oped mild illness produced significantly more mucosal IFN-α2 and IP-10 during the early response phase (Fig. 3A andB). Patients with moderate to severe COVID-19 significantly increased the production of mucosal pro-inflammatory factors, including IL-8, TNF-β, IL-2, IL-9, IL-17A, Flt-3L, GM-CSF, IL-12p40, MDC, VEGF, IL-15, sCD40L, and TGF-α, during the early response phase (Fig. 3C through O). Severe cases also showed increased levels of mucosal MDC, IL-1β, and MCP-3 during adaptive response phase (Fig. 3K, P and Q), mucosal IL-2, IL-9, IL-17A, Flt-3L, GM-CSF, IL-12p40, IL-15, and EGF during symptoms resolution phase (Fig. 3E through J, M and R) and mucosal eotaxin, IL-4, and GRO during convalescence phase (Fig. 3S through U), compared to mild and moderate cases. We input these cytokine data into energy flow diagrams that indicate the magnitude of each factor through their convergence on cellular targets (Fig. S4). During the early response phase, those with mild COVID-19 displayed enhanced targeting of T cells, dendritic cells (DC), and natural killer (NK) cells in the upper airways largely via chemo kines (Fig. S4A green). The total pg/mg of factors targeting T cells, DCs, and NK cells declined from 28,037 to 15,844 pg/mg in mild versus severe cases, respectively. The most striking evidence of enhanced mucosal antiviral responses was IP-10. Mild cases had twice and almost five times higher levels compared to moderate and severe cases, respectively. Factors targeting endothelial, epithelial, and fibroblast cells totaled 31,117 pg/mg in mild and 56,897 pg/mg in severe cases. This increase suggests airway damage in these patients (Fig. S4A).
Distinct from the mucosa, no systemic immune factors were elevated in the mild group compared to the moderate and severe groups. During the early response phase, as severity outcomes worsened, there was a significant increase in anti-inflammatory IL-1RA and IL-10 (Fig. 4A andB), chemoattractants MIP-1β (Fig. 4C), and pro-inflammatory factors TNF-α and IL-15 (Fig. 4D andE). During the adaptive response phase, systemic IL-1RA, IL-10, MIP-1β, TNF-α, IL-15, IP-10, fractalkine, IL-1α, IL-6, IL-8, TNF-β, IL-4, MCP-3, IL-9, and IL-13 were significantly increased as severity outcomes worsened (Fig. 4A through O). Systemic IL-1RA, IL-10, MIP-1β, TNF-α, IL-15, fractalkine, MIP-1α, and VEGF were elevated during the resolution or convalescence, or in both phases, as severity outcomes worsened (Fig. 4A through E, G, P andR). Systemic IL-1α, IL-8, IL-13, and IL-5 significantly increased in moderate cases compared to mild cases after adaptive response phase (Fig. 4H, J, O andQ). In contrast to the mucosa, peripheral cytokines targeting T cells, DC, and NK cells progressively increased from mild to moderate to severe cases during the early response period (Fig. S4B). Cytokine levels targeting these cells were twofold higher in severe cases compared to mild cases, with a 3.2-fold increase in systemic IP-10. This was the opposite of NRF. In the early response phase, the mean total cytokine magnitudes were 2,709, 3,502, and 4,132 pg/mL for mild, moderate, and severe disease progression groups, respectively (Fig. S4B).
## Unraveling the coordination of mucosal and systemic responses to novel viruses
We assessed viral and immune dynamics in both the compartments using Pearson's correlation analysis across all participants. Significant positive and negative correlations during the acute phase of infection were traced between the mucosa and systemic compartments (Fig. 5 depicted in yellow and blue, respectively). Interestingly, only mild cases showed mucosal and systemic viral correlations (Fig. 5A, top black). Many mucosal and systemic immune factors were significantly positively correlated in individuals with mild illness, including mucosal adaptive, pro-inflammatory, anti-inflammatory, chemoat tractant, viral, and growth factors, which correlated positively to systemic cytokines, except mucosal IgG, EGF, and FGF-2, which correlated negatively with systemic eotaxin, fractalkine, and IP-10, respectively (Fig. 5A). Patients with moderate illness showed increased positive and negative correlations (mVEGF, mTNF-α, mMIP-1α, mMCP-3, mEotaxin with sEotaxin, mMCP-1 with sIL-1α, and mIP-10 with sIL-6 negatively correla ted) between both the compartments compared to mild illness (Fig. 5B). Although significant positive correlations were detected, we observed a dramatic increase in significant negative correlation between mucosal and systemic cytokines in severe COVID-19 cases (Fig. 5C). In the recovery phase, mild and moderate cases lacked significant negative correlations between mucosal and systemic cytokines (r < 0.5). In contrast, the significant negative correlations between mucosal and systemic compart ments increased from the acute to recovery phases in severe cases (Fig. S5A andB).
## Prognostic indicators: IFN-α2, IP-10, IL-1RA, IL-6, eotaxin, IgA, and IgM
To identify potential prognostic biomarkers among naïve individuals encountering a novel virus for the first time, we performed systems analysis combining the mucosal and plasma data collected during week 1, week 2, week 3, and week 4+ from diagnosis corresponding to the early acute, late acute, recovery, and convalescent phases of infection, respectively. Within a week of diagnosis, individuals who developed asympto matic or mild COVID-19 could be distinguished by significantly increased mucosal IP-10 and decreased systemic IL-1RA, MIP-1β, IL-6, fractalkine, IgA, and IgM within one week of positive diagnostic (Fig. 6A andC). In this early acute phase, patients who progressed to severe COVID-19 were distinguishable from those who developed mild and moderate illness by their significantly higher mucosal CD40L, FGF-2, Flt-3L, IL-2, MDC, and VEGF, as well as systemic TNF-α, IL-10, IL-15, MIP-1β, IL-1RA, and IgA, and decreased systemic eotaxin (Fig. 6A andB). In week 2, mild cases displayed significantly less mucosal PDGF-AB/BB and systemic IP-10, IL-8, IL-1RA, and MIP-1β compared to those with severe and moderate illness (Fig. 6D andF). At this time, severe cases produced significantly more mucosal IFN-α2, sCD40L, PDGF-AA/BB, MCP-1, MCP-3, IL-5, and FGF-2 and peripheral IFN-α2, IL-1RA, IP-10, IL-10, fractalkine, MIP-1α, MIP-1β, VEGF, and IgA (Fig. 6D andE). Notably, in week 2, individuals with mild to moderate symptoms produced significantly more mucosal IL-1RA than severe cases (Fig. 6D andE). Remarkably, patients with severe COVID-19 maintained significantly higher systemic IP-10, IL-1RA, IL-10, IL-15, and MIP-1β levels and mucosal IL-1RA, IL-12p40, and fractalkine through recovery and convalescence (Fig. S6A andB).
## Distinct immune responses in vulnerable understudied populations
To better understand the immune response to novel respiratory viruses in vulnerable populations, subjects were first grouped by outpatient or hospitalized based on COVID-19 severity scores. After that, subjects were grouped by age, sex, or race. The subjects were categorized by age into children (aged 8 months to 19 years) and adults (20-65 years), respectively. When outpatient children were compared to outpatient adults, children produced significantly more mucosal IL-2p70 and low levels of mucosal EGF-2 (Fig. S7A). Surprisingly, hospitalized children produced elevated levels of systemic cytokines and reduced levels of mucosal cytokines IFN-Ƴ, sCD40L, fractalkine, and Flt.3L compared to hospitalized adults (Fig. S7B).
Subjects were grouped by sex into female or male based on CRF data. Similarly, we compared outpatient females and males with COVID-19 illness. Outpatient females participants with asymptomatic or mild symptoms (mild) were determined compared to those with moderate illness (bottom panel). Differential expression analysis was performed using XLSTAT analysis set to parametric test type and Tukey (HSD). Volcano plots were graphed using OriginPro. The X-axis represents log2 fold change of protein with dashed lines intersecting at the twofold cutoff point. The Y-axis represents -log10 P value with dashed lines intersecting at the cutoff for significance (P < 0.05). Proteins that significantly increased ≥2-fold are colored red, significantly decreased ≥2-fold are colored green, and significantly different proteins with fold changes between -2 to +2 are colored orange. Mucosal (M) and systemic (S) compartments are indicated before protein symbols. produced increased levels of systemic IgM with mucosal eotaxin and IgA and reduced levels of mucosal IL-1α, IL-17A, and IFN-α2 compared to outpatient males (Fig. S7A). Hospitalized females produced elevated levels of mucosal cytokines and reduced levels of systemic cytokines GM-CSF, IL-2, and IL-12p40 compared to hospitalized males (Fig. S7B). Interestingly, females showed elevated levels of mucosal eotaxin compared to males for both outpatient and hospitalization (Fig. S7A andB), Based on CRF data, subjects were grouped by race as African Americans (AA) or European Americans (EA). AA had significantly higher severity outcome scores (mean difference = 3.453, P < 0.0001). Outpatient AA had significantly higher systemic cytokines, with increased mucosal IL-1α and reduced levels of mucosal RANTES, MDC, TGF-α, VEGF, and GM-CSF and systemic MCP-1 compared to outpatient EA (Fig. S7A). When hospitalized AA were compared to hospitalized EA, AA showed high levels of mucosal sCD40L, MCP-1, MCP-3, IL-1α, GM-CSF, IL-10, MDC, and EGF with systemic IgA and reduced levels of systemic EGF, GM-CSF, IL-13, sCD40L, and MCP-3 with mucosal IgA and IgG (Fig. S7B). Interestingly, AA produced elevated levels of mucosal IL-1α compared to their counterparts for both outpatient and hospitalization (Fig. S7A andB).
## DISCUSSION
COVID-19 remains a public health threat especially in the winter when hospitalizations are highest for older adults and children. Peripheral vaccination has improved outcomes but fails to elicit lasting protection against SARS-CoV-2 and durable mucosal immunity. Mucosal responses are the first line of defense against respiratory viruses and can limit their spread. In this study, we investigated the dynamics and function of mucosal and systemic immune responses to SARS-CoV-2. We used a cohort representative of the general unvaccinated, naïve population and examined the role of synchronicity between these compartments in disparate clinical outcomes. We verified that the subjects were infected by the ancestral B.1 lineages and were at the same stage of infection based on IgM seropositivity (25).
Systemic and mucosal humoral immune responses are crucial for combating respiratory viral infections and may be especially important against newly emerging viral pathogens. We found that individuals with mild COVID-19 gradually increased mucosal IgG production. Disease severity was independent of early viral load and mucosal IgA levels, indicating they do not account for desperate outcomes. We also found periph eral S levels, rate of antibody class switching, and early antibody levels significantly increased with illness severity. Antibody affinity has been reportedly lower in severe COVID-19 cases (26). We found that individuals with mild symptoms slowly increased antibody production over time, yielding significantly higher neutralization potency. Their systemic IgG levels significantly and positively correlated with neutralization, suggesting enhancement by IgG. We and others detected very high levels of systemic IgA in severe COVID-19 cases (26). IgA interference, first identified in humans following HIV vaccination (27), could negatively impact COVID-19 outcomes. Our data showed that early systemic IgA in mild cases was below the limit of detection and significantly higher in moder ate and severe patients, suggesting that IgA interference may contribute to COVID-19 disease progression. We observed equivalent levels of mucosal IgA and IgG, as well as systemic IgG, concomitant with undetectable IgA and IgG in mild cases during acute infection. These initial humoral responses, which are classically considered to be more specific, did not improve clinical COVID-19 outcomes. Additionally, the higher neutrali zation potency we observed in mild cases is likely to improve their recall responses and enhance protection against subsequent infections. Overall, among previously naïve individuals, more gradual and synchronous mucosal and systemic humoral responses were associated with mild illness, whereas premature excessive, uncoordinated, and less potent humoral responses were observed as COVID-19 severity escalated. Our results shed new light on the discrepancy between individuals who have restrained humoral responses to novel viruses and those who initiate early vigorous peripheral IgA production and develop more severe symptoms. They also support the notion that mucosal vaccines could serve as a valuable option for future pandemics and may offer more sustained and potent protection than current COVID-19 booster strategies that increase systemic IgA.
Neutralizing antibodies alone are insufficient for viral clearance. We found that naïve individuals with mild COVID-19 cleared the virus in week 1 and produced significantly more mucosal IFN-α and IP-10 within four days of diagnosis, indicating robust very early antiviral responses. This is consistent with other COVID-19 studies associating severe disease with impaired epithelial interferon responses, fewer local IFN-α-producing plasmacytoid dendritic cells, and mucosal IP-10 levels that correlated with interferon responses (28)(29)(30). Mild cases were marked by the immediate production of IFN-α and IP-10 at the site of infection, enhancing the recruitment of cells expressing the cognate CXCR3 receptor, such as T and NK cells. This can expedite viral clearance, the production of germinal center-derived high-affinity antibodies, and the control of systemic immune responses. In week 2, we observed that mild and moderate cases had significantly higher mucosal anti-inflammatory IL-1RA compared to those with severe illness, suggesting that IL-1RA mediated the return of local immune responses to baseline. Systemic IFN-I responses were reportedly diminished or absent in hospitalized patients with COVID-19 (31). However, we found delayed mucosal and systemic IFN responses in patients with severe COVID-19, which could allow for deeper virus penetration in the respiratory tract early in the infection. Our analysis revealed that naïve individuals with synchronized mucosal and systemic cytokine responses in the acute phase develop mild illness, whereas those with uncoupled local and peripheral early responses progress to more severe illness and fail to adequately resolve immune responses to novel viruses in the recovery phase. Overall, these data strongly support a role for early mucosal IFN-α and IP-10 in antiviral protection and, in conjunction with subsequent IL-1RA production, in limiting disease severity in the naïve population.
While the number of participants enrolled in first wave infections limited the statistical power of our subgroup analyses, our data suggest novel differences in COVID-19 cytokine response and outcomes in race, sex, and age subgroups. Children, females, and AA may have unique immune responses to SARS-CoV-2 that could benefit from specific clinical interventions. The high levels of IL-6 and IL-8 observed here indicate that tocilizumab and siltuximab may be particularly effective for treating hospitalized pediatric and female patients with COVID-19. Male sex and low levels of eotaxin have been reported in severe COVID-19 illness (32). Our results show that high levels of eotaxin in females could be protective against poor COVID-19 outcomes. IL-1 receptor antagonism via anakinra has improved clinical outcomes for severe COVID-19 (33). We found that AA and hospitalized children had elevated IL-1, suggesting that anakinra may be especially effective in these populations. Our analysis of the entire cohort suggests that IFN-α2, IP-10, IL-1RA, IL-6, eotaxin, IgA, and IgM are prognostic markers that can be monitored in mucosal fluids and blood at the time of diagnosis or shortly after to forecast disease progression and aid in treatment plans following infection with a novel respiratory virus. Importantly, these findings came from a cohort with equal representation across outcomes, sex, age, and race subgroups, which increases the generalizability of this study. To our knowledge, this is the most diverse cohort to have a COVID-19 mucosal immunity assessment and the first reported comparison of mucosal and peripheral immune kinetics extending into convalescent infection.
This study has several limitations. First, although our cohort was racially diverse and balanced by sex and age, the overrepresentation of severe cases among Afri can American participants and mild cases among European Americans limits the generalizability of race-stratified subgroup analyses. Second, we did not assess prior exposure to seasonal human coronaviruses, which could influence immune priming. While pre-existing humoral and cellular immunity to SARS-CoV-2 has been reported in unexposed individuals, the literature remains mixed on whether such responses modulate disease severity or confer protection. Importantly, the current consensus suggests that pre-existing cross-reactive T-cell responses may offer modest benefits (34,35), while cross-reactive antibodies are less likely to be protective (34,36,37). Third, participants were enrolled between July 2020 and March 2021, when cumulative case counts in the area remained low (between 0.93% and 10% of the population, respectively), and epidemiologic evidence indicates that prior infection conferred strong protection against reinfection for at least 6 to 12 months. Given this low prevalence and the durability of post-infection immunity during the pre-Omicron period, it is highly likely-though not certain-that participants were experiencing a primary SARS-CoV-2 infection. Consistent with this, nearly all were IgM-positive by week 2, supporting recent infection and appropriate cohort classification. Fourth, we standardized sampling time points using the date of first positive SARS-CoV-2 PCR test rather than symptom onset due to variability and underreporting in self-reported symptoms-especially among Black participants. This decision was supported by serological validation and enabled consistent temporal alignment across the cohort. Finally, this observational study establishes associations-but not causality-between immune parameters (e.g., mucosal IFN-α, IP-10, IL-1RA) and clinical outcomes in SARS-CoV-2-naïve individuals. Nonetheless, our findings are consistent with emerging reports in vaccinated and breakthrough cases that identified these cytokines as systemic markers of vaccine responsiveness or protection (38)(39)(40). Thus, although our findings reflect immune responses in SARS-CoV-2-naïve individuals, particularly elevated mucosal IFN-α and IP-10 followed by IL-1RA, they may have broader translational relevance.
In conclusion, our data provide strong evidence of the vital role of early mucosal responses in controlling disease progression following infection with a novel respi ratory virus, such as SARS-CoV-2, and highlight the potential power of targeting mucosal immunity in vaccination efforts and monitoring mucosal responses in the naïve population. Individuals with high mucosal IFN-α, IP-10, and IL-1RA production quickly synchronized local and systemic immune responses, whereas those lacking these protective mucosal responses had poor or dysregulated immune responses with delayed resolution that were associated with more severe illness. We began collecting nasophar yngeal rinses and blood in 2020 from participants with a broad range of COVID-19 severity, a cohort that will be impossible to replicate in the future due to acquired immunity. This unique study revealed previously unknown aspects of mucosal responses to SARS-CoV-2 associated with disease progression in the naïve population. These data and insights may prove useful for developing next-generation COVID-19 vaccines and prognostic indicators when the next novel respiratory virus emerges.
## MATERIALS AND METHODS
Additional details are provided in the online data supplement.
## Study participants
Participants with COVID-like illnesses were recruited from Le Bonheur Children's Hospital, Methodist University Hospital, and outpatient testing sites in Memphis, Tennessee. Pregnant women and those who could not provide informed consent were excluded. Participants with clinical laboratory polymerase chain reaction or antigen positive results within 72 hours of enrollment were included. Two severe subjects died during the study. Two seniors were partially vaccinated against SARS-CoV-2. Electronic medical record data were collected using a standardized data record form.
## Study procedures
Mid-turbinate swabs (MT-swabs) were collected and placed in viral transport media on ice. Then, 0.1% saline was flushed through the nares to collect nasopharyngeal rinse fluid (NRF), followed by the addition of ice-cold BEGM. The sample was centrifuged, cells were removed, and protease inhibitor cocktail was added. Blood was drawn into vacutainer cell preparation tubes and immediately processed following the manufactur er's guidelines. Aliquots were stored at -80°C.
## Primary outcomes
COVID-19 outcome was assigned based on retrospective chart review at least 28 days from diagnosis. A modified World Health Organization (WHO) COVID-19 case definition was used to score severity outcome (41,42), based on hospitalization and symptoms (Table 1). Participants with asymptomatic infections were classified as mild. All outpa tients were classified as mild, and hospitalized participants were designated as moderate or severe.
To standardize temporal analyses, the date of the first positive SARS-CoV-2 PCR test was designated as day 0 for all participants. Time from symptom onset to diagnosis varied across individuals and was not used as an anchor due to reliance on self-report. To confirm that PCR-based day 0 reflected early-stage infection, SARS-CoV-2-specific IgM levels were measured across time points. The majority of participants were IgM-positive by week 2, consistent with expected early seroconversion (Fig. S2).
## SARS-CoV-2 RNA RT-qPCR and sequencing
RNA was isolated from swabs and subjected to reverse transcription real-time quantita tive polymerase chain reaction (RT-qPCR). Primers targeted either ORF1b-nsp14 or Spike (43). Average cycle threshold (Ct) value was determined with a negative threshold of 40. PCR-positive swabs were sequenced following the manufacturer's protocols. RNA was transcribed to single-stranded cDNA using random hexamers. SARS-CoV-2 sequence libraries were prepared and sequenced using paired-end 2 × 150 bp reads with the MiSeq Reagent Kit v2 for 300 cycles (Illumina, #MS-102-2002). The sequenced libraries were assembled, and SARS-CoV-2 lineages were determined as previously described (44).
## Antigen and antibody quantification
S, N, and antibodies were quantified by ELISA. Undiluted NRF and plasma were quantified and analyzed per manufacturer recommendations. Protein concentrations were extrapolated from standard curves.
## Neutralization assays
Plasma neutralization of spike-containing pseudovirus was performed as previously described (45). Antibody neutralization potency was calculated as a ratio of neutraliza tion titer 50 (NT 50 ) to the sum of plasma antibodies (46). Neutralizing antibody activ ity was quantified using the BioPlex Pro Human SARS-CoV-2 Neutralization Antibody 2-Plex Panel (BioRad #12016848), which operates via a bead-based competitive assay measuring inhibition of ACE2binding to RBD and S1 antigens. Antibody concentration and percent inhibition were calculated based on MFI values relative to negative controls and standard curves.
## Immune factor quantification
Forty-one human immune factors were quantified on a Luminex 200 instrument according to the manufacturer's instructions. The absolute quantity of immune factors was reported as pg/mL of plasma and pg/mg protein of concentrated NRF.
## Statistical analysis
GraphPad Prism software was used for basic statistical analysis: Log-rank Mantel-Cox test, one-way ANOVA with Fisher's least significant difference (LSD) procedure and posttest for linear trends, and two-way ANOVA with Tukey's honestly significant difference (HSD). The XLSTAT package was used for Pearson correlations and expression analysis. Post hoc power analyses were conducted using standard statistical methods to assess the sensitivity of our sample sizes for detecting group differences or correlations across various experiments summarized in the figures. The power analyses summarized below were performed using the statistical software G*Power 3.1 (47).
## Humoral immune analyses
Group comparisons of N-and S-specific Ab levels in mucosal (NRF) and systemic (plasma) samples were analyzed using one-way ANOVA with Fisher's LSD post hoc test. Sample sizes per severity group were as follows: mild (n = 102 NRF, 86 plasma), moderate (n = 77 NRF, 62 plasma), and severe (n = 23 NRF and plasma). At α = 0.05, these group sizes provide ≥80% power to detect moderate-to-large differences in mean antibody levels (Cohen's f ≥ 0.4). The severe group had ≥70% power for large effects (d ≥ 0.8). Correlation analyses using 130 paired NRF-plasma samples had >90% power to detect moderate associations (r ≥ 0.3) at α = 0.05.
## Cytokine analyses
Cytokine levels were compared across severity groups at defined time windows using two-way ANOVA with Tukey's HSD correction. Mucosal cytokines were quantified in 190 nasopharyngeal rinse fluid (NRF) samples from mild (n = 59), moderate (n = 44), and severe (n = 43) cases. Systemic cytokines were quantified in 169 plasma samples from mild (n = 67), moderate (n = 31), and severe (n = 35) cases. Based on these sample sizes and α = 0.05, the analyses had ≥80% power to detect moderate-to-large between-group differences in cytokine levels (Cohen's f ≥ 0.4) at early (0-4 days) and adaptive (5-10 days) time points. To assess early associations between mucosal and systemic cytokines within severity groups, we analyzed NRF and plasma samples collected during the acute phase (≤7 days from diagnosis): mild (n = 24 NRF, 24 plasma), moderate (n = 25 NRF, 21 plasma), and severe (n = 10 NRF, 12 plasma). Power analysis (α = 0.05, two-tailed) indicates ≥80% power to detect correlations (r ≥ 0.4) in mild and moderate groups and (r ≥ 0.62) in the severe group.
## Between-group immune comparisons at defined time windows
Between-group comparisons of mucosal and systemic immune responses-including cytokines, chemokines, growth factors, and N/S-specific IgA, IgM, and IgG antibodieswere conducted using parametric statistical tests with Tukey's HSD post hoc correction at two defined time windows (week 1: ≤7 days post-diagnosis; week 2: 8-14 days). Sample sizes were mild (n = 161), moderate (n = 121), and severe (n = 66). Based on these group sizes and α = 0.05, the estimated statistical power for detecting differences in immune markers was as follows: mild vs. moderate: ≥95% power for moderate effects (Cohen's d ≥ 0.5), ≥99% for large effects (d ≥ 0.8); mild vs. severe: ≥80% power for moderate effects, ≥98% for large effects; and moderate vs. severe: ≥72% power for moderate effects, ≥95% for large effects.
## References
1. Li, Hannah, Durbin et al. (2020) "Multivariate analysis of Black race and environmental temperature on COVID-19 in the US" *Am J Med Sci*
2. Ludvigsson (2020) "Systematic review of COVID-19 in children shows milder cases and a better prognosis than adults" *Acta Paediatr*
3. Zimmermann, Curtis (2020) "Coronavirus infections in children including COVID-19: an overview of the epidemiology, clinical features, diagnosis, treatment and prevention options in children" *Pediatr Infect Dis J*
4. Argyropoulos, Serrano, Hu et al. (2020) "Association of initial viral load in severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) patients with outcome and symptoms" *Am J Pathol*
5. Heald-Sargent, Muller, Zheng et al. (2020) "Age-related differences in nasopharyngeal severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) levels in patients with mild to moderate coronavirus disease 2019 (COVID-19)" *JAMA Pediatr*
6. Yang, Rao (2021) "Structural biology of SARS-CoV-2 and implications for therapeutic development" *Nat Rev Microbiol*
7. Li, Xiong, Deng et al. (2022) "The utility of SARS-CoV-2 nucleocapsid protein in laboratory diagnosis" *J Clin Lab Anal*
8. Isho, Abe, Zuo et al. (2020) "Persistence of serum and saliva antibody responses to SARS-CoV-2 spike antigens in COVID-19 patients" *Sci Immunol*
9. Cervia, Nilsson, Zurbuchen et al. (2021) "Systemic and mucosal antibody responses specific to SARS-CoV-2 during mild versus severe COVID-19" *J Allergy Clin Immunol*
10. Sasson, Campo, Carpenter et al. (2021) "Diverse humoral immune responses in younger and older adult COVID-19 patients" *mBio*
11. Ragab, Salah Eldin, Taeimah et al. (2020) "The COVID-19 cytokine storm; what we know so far" *Front Immunol*
12. Chen, Wu, Guo et al. (2019) "Clinical and immunological features of severe and moderate coronavirus disease" *J Clin Invest*
13. Layne, Walters, Kash et al. (2022) "More autopsy studies are needed to understand the pathogenesis of severe COVID-19" *Nat Med*
14. Du, Yuan (2020) "Mathematical modeling of interaction between innate and adaptive immune responses in COVID-19 and implications for viral pathogenesis" *J Med Virol*
15. Gallo, Locatello, Mazzoni et al. (2021) "The central role of the nasal microenvironment in the transmission, modulation, and clinical progression of SARS-CoV-2 infection" *Mucosal Immunol*
16. Chen, Magri, Grasset et al. (2020) "Rethinking mucosal antibody responses: IgM, IgG and IgD join IgA" *Nat Rev Immunol*
17. Alfi, Yakirevitch, Wald et al. (2021) "Human nasal and lung tissues infected ex vivo with SARS-CoV-2 provide insights into differential tissue-specific and virus-specific innate immune responses in the upper and lower respiratory tract" *J Virol*
18. Smith, Goncalves, Charbit et al. (2021) "Distinct systemic and mucosal immune responses during acute SARS-CoV-2 infection" *Nat Immunol*
19. Cass, Nicolau, Jr et al. (2024) "Coordinated nasal mucosa-mediated immunity accelerates recovery from COVID-19" *ERJ Open Res*
20. Selva, Ramanathan, Haycroft et al. (2023) "Preexisting immunity restricts mucosal antibody recognition of SARS-CoV-2 and Fc profiles during breakthrough infections" *JCI Insight*
21. Verheul, Kaczorowska, Hofstee et al. (2024) "Protective mucosal SARS-CoV-2 antibodies in the majority of the general population in the Netherlands" *Mucosal Immunol*
22. Lindeboom, Worlock, Dratva et al. (2024) "Human SARS-CoV-2 challenge uncovers local and systemic response dynamics" *Nature*
23. Russell, Moldoveanu, Ogra et al. (2020) "Mucosal immunity in COVID-19: a neglected but critical aspect of SARS-CoV-2 infection" *Front Immunol*
24. Russell, Mestecky (2022) "Mucosal immunity: the missing link in comprehending SARS-CoV-2 infection and transmission" *Front Immunol*
25. Deeks, Dinnes, Takwoingi et al. (2020) "Van den Bruel A, Cochrane COVID-19 Diagnostic Test Accuracy Group"
26. *Cochrane Database Syst Rev*
28. Hendriks, Schasfoort, Koerselman et al. (2022) "High titers of low affinity antibodies in COVID-19 patients are associated with disease severity" *Front Immunol*
29. Tomaras, Ferrari, Shen et al. (2013) "Vaccineinduced plasma IgA specific for the C1 region of the HIV-1 envelope blocks binding and effector function of IgG" *Proc Natl Acad Sci*
30. Ziegler, Miao, Owings et al. (2021) "Impaired local intrinsic immunity to SARS-CoV-2 infection in severe COVID-19" *Cell*
31. Laurent, Yang, Rendeiro et al. (2022) "Sensing of SARS-CoV-2 by pDCs and their subsequent production of IFN-I contribute to macrophage-induced cytokine storm during COVID-19" *Sci Immunol*
32. Cheemarla, Watkins, Mihaylova et al. (2021) "Dynamic innate immune response determines susceptibility to SARS-CoV-2 infection and early replication kinetics" *J Exp Med*
33. Viox, Bosinger, Douek et al. (2024) "Harnessing the power of IFN for therapeutic approaches to COVID-19" *J Virol*
34. Ling, Chen, Lui et al. (2021) "Longitudinal cytokine profile in patients with mild to critical COVID-19" *Front Immunol*
35. Montazersaheb, Khatibi, Hejazi et al. (2022) "COVID-19 infection: an overview on cytokine storm and related interventions" *Virol J*
36. Galipeau, Siragam, Laroche et al. (2021) "Relative ratios of human seasonal coronavirus antibodies predict the efficiency of cross-neutralization of SARS-CoV-2 spike binding to ACE2" *EBioMedicine*
37. Murray, Ansari, Frater et al. (2023) "The impact of pre-existing cross-reactive immunity on SARS-CoV-2 infection and vaccine responses" *Nat Rev Immunol*
38. Nasrallah (2022) "Do preexisting antibodies against seasonal coronaviruses have a protective role against SARS-CoV-2 infections and impact on COVID-19 severity?" *EBioMedicine*
39. Woudenberg, Pelleau, Attia et al. (2021) "Humoral immunity to SARS-CoV-2 and seasonal coronaviruses in children and adults in north-eastern France" *EBioMedicine*
40. Chan, Pinedo, Stabile et al. "Prior vaccination prevents overactivation of innate immune responses during COVID-19 break through infection" *Sci Transl Med*
41. Almeida, Nagahara, Bueno et al. (2024) "Interleukin-1β, interleukin-1Ra, and interleukin-8 in patients with SARS-CoV-2 infection: a correlation between vaccination and clinical outcome" *GPD*
42. Alkhattabi, Alharbi, Basabrain et al. (2024) "Studying the correlation of inflammatory cytokines to COVID-19 disease" *Pathol Res Pract*
43. Zhang, Dong, Cao et al. (2020) "Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan" *Allergy*
44. Fragaszy, Shrotri, Geismar et al. "Virus Watch Collaborative. 2022. Symptom profiles and accuracy of clinical case definitions for COVID-19 in a community cohort: results from the Virus Watch study" *Wellcome Open Res*
45. Chu, Pan, Cheng et al. (2020) "Molecular diagnosis of a novel coronavirus (2019-nCoV) causing an outbreak of pneumonia" *Clin Chem*
46. Zhu, Chang, Webby et al. (2020) "idCOV: a pipeline for quick clade identification of SARS-CoV-2 isolates" *bioRxiv*
47. Wohlgemuth, Whitt, Cherry et al. (2021) "An assessment of serological assays for SARS-CoV-2 as surrogates for authentic virus neutralization" *Microbiol Spectr*
48. Garcia-Beltran, Lam, Astudillo et al. (2021) "COVID-19-neutralizing antibodies predict disease severity and survival" *Cell*
49. Faul, Erdfelder, Buchner et al. (2009) "Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses" *Behav Res Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12826974&blobtype=pdf | # Ubiquitin-specific protease 5 promotes EV-A71 replication by de-ubiquitinating MAVS and IRF3
Shumin Zhang, Yuan Fang, Shuai Ren, Xuhua Zhang, Chenggong Zheng, Zhipeng Qin, Wenqiang Wei, Huabin Zheng, Chuntian Li, Zekun Wang, Yujie Ren, Virologica Sinica
## Abstract
Human enterovirus A71 (EV-A71) is a major causative agent of hand, foot and mouth disease (HFMD), which poses a significant public health threat, particularly among young children. Mitochondrial antiviral signaling protein (MAVS) and interferon regulatory factor 3 (IRF3) are vital proteins for the induction of type I interferons (IFN-I) and downstream interferon-stimulated genes (ISGs) during EV-A71 infection. While posttranslational modifications are known to critically influence viral infection processes, the mechanisms by which EV-A71 exploits host deubiquitinases (DUBs) for immune evasion remain poorly understood. In this study, we demonstrated that EV-A71 infection upregulated ubiquitinspecific protease 5 (USP5) expression. Knockdown of USP5 not only inhibited EV-A71 replication but also observably increased the production of IFN-I and ISGs. Furthermore, USP5 also regulated the replication of EV-D68 and CVA16 and the production of IFN-I and ISGs. Mechanistically, USP5 physically interacted with MAVS and IRF3 and reduced the K63-linked polyubiquitination of MAVS and IRF3. Conversely, USP5 knockdown increased the K63-linked polyubiquitination of MAVS and IRF3, thereby accelerating the phosphorylation of IRF3 and increasing IFN-I production during EV-A71 infection. Furthermore, pharmacological inhibition of USP5 with the small-molecule inhibitor PR-619 significantly potentiated the antiviral effects of IFN against EV-A71. Collectively, our findings reveal a previously unrecognized role of USP5 in facilitating EV-A71 immune evasion by dampening MAVS-and IRF3-mediated antiviral signaling. These insights provide a novel therapeutic avenue for combating EV-A71 infection through targeted modulation of the USP5-IRF3 axis.
## INTRODUCTION
Innate immune responses constitute the primary defence mechanism against pathogen invasion, particularly during the initial phase of viral infection. Key pathogen recognition receptors (PRRs), including RIG-Ilike receptors (RLRs), melanoma differentiation-associated protein 5 (MDA5), Toll-like receptors (TLRs), NOD-like receptors (NLRs), and cyclic GMP-AMP synthase (cGAS), detect pathogen-associated molecular patterns (PAMPs) to trigger signaling cascades that activate transcription factors such as IRF3 and NF-κB, triggering IFN-I production and the subsequent expression of interferon-stimulated genes (ISGs), establishing potent antiviral states (Schoggins et al., 2011;Iwasaki, 2012;Gao, D. et al., 2021;He et al., 2025).
Enterovirus 71 (EV-A71), a member of the Picornaviridae family, is the primary causative agent of severe hand, foot, and mouth disease (HFMD) in children under seven years of age (Solomon et al., 2010;Saguil et al., 2019). In severe cases, EV-A71 infection can lead to neurological complications, posing a significant public health concern. EV-A71 outbreaks have been reported in numerous countries, with China documenting annual case reports (Zhang, J. et al., 2011). The EV-A71 genome consists of a positive-strand RNA, which includes two noncoding regions (a 5 ′ UTR and a 3 ′ UTR) and a large open reading frame (ORF). This ORF encodes four structural proteins (VP1-VP4) and seven nonstructural proteins (2A, 2B, 2C, 3A, 3B, 3C and 3D). These viral components play distinct yet coordinated roles in viral replication, immune evasion, and pathogenesis (Han, Y. et al., 2016;Yao et al., 2019;Cao et al., 2020;Wen et al., 2020). Notably, the 2A protease disrupts antiviral innate immunity by targeting MAVS and TRAF3, thereby suppressing IFN-I responses (Wang, B. et al., 2013;Zheng et al., 2023). Similarly, the 3C protease cleaves OAS3, a key effector in the OAS/RNase L antiviral pathway, further facilitating immune evasion (Zhou, X. et al., 2022). The 2C protein, which is essential for viral RNA replication, exhibits ATPase and helicase activities. Additionally, the N-terminal domain of 2C not only interacts with viral-encoded proteins (Yin et al., 2007;Liu, Y. et al., 2010), but also with host proteins, such as Reticulon 3 (RTN3) (Tang et al., 2007). Furthermore, 2C has been shown to inhibit the TNF-α-induced NF-κB signaling pathway, modulating host immune responses (Li, Q. et al., 2016). Given the intricate interplay between EV-A71 proteins, viral RNA, and the host cellular machinery, further research is crucial to uncover novel mechanisms that enhance antiviral innate immunity and develop effective therapeutic strategies.
Ubiquitination is a multistep enzymatic process that involves the covalent attachment of monoubiquitin or polyubiquitin chains to target proteins. As a classical posttranslational modification, protein ubiquitination plays crucial regulatory roles in diverse cellular processes (Calistri et al., 2014;Clague et al., 2015). The biological consequences of ubiquitination depend on the specific ubiquitin linkage type: K48-based polyubiquitin chains typically target substrates for proteasomal degradation (Kaiho-Soma et al., 2021;Zhang, S. et al., 2022), whereas K63-based ubiquitin chains predominantly modulate protein function and signaling pathways (Yau et al., 2017;Wu et al., 2022). This dynamic modification is counterbalanced by de-ubiquitinating enzymes (DUBs), which cleave ubiquitin moieties and reverse ubiquitination events (Sun, T. et al., 2020). The ubiquitination-deubiquitination axis represents a sophisticated regulatory mechanism governing fundamental cellular processes, including cell proliferation, DNA damage repair, lipid metabolism, and immune responses (Han, S. et al., 2022;Jiang et al., 2023;Zhou, B. et al., 2024). Notably, emerging evidence suggests that specific DUBs can modulate type I IFN pathways through their enzymatic activity. For example, ubiquitin-specific protease (USP) 24 has been shown to suppress EV-A71 replication by negatively regulating the production of IFN-I (Zang et al., 2023a). USP13 enhances antiviral immunity by stabilizing STAT1 and potentiating IFN activity against DEN-2 infection (Yeh et al., 2013). In contrast to these findings, our study revealed that USP5 unexpectedly promotes EV-A71 replication, suggesting a distinct regulatory role in host-virus interactions.
USP5, a multifunctional DUB, regulates epithelial-mesenchymal transition and tumor progression via its catalytic activity (Meng et al., 2019;Wan et al., 2024). Previous studies have reported that USP5 participates in the regulation of non-small cell lung cancer, myocardial ischemia-reperfusion injury (Sun, W. et al., 2024), and DNA mismatch repair (Mao et al., 2024), primarily by targeting immune checkpoints and immune molecules (Xiao et al., 2023). Moreover, USP5 also inhibits IFN-induced P-STAT1 activation and downstream antiviral gene expression of STAT1 (Qian et al., 2020). Similarly, USP5 can inhibit innate immunity induced by vesicular stomatitis virus (VSV), sendai virus (SeV), and influenza virus A by deconjugating K48-linked unanchored and K63-linked anchored ubiquitin from IRF3 (Qiao et al., 2025). USP5 can interact with Unc51-like kinase 1 (ULK1) and negatively regulate autophagy (Pai et al., 2023). Although USP5 has been found to act as a scaffold to regulate innate immunity (Liu, Q. et al., 2018), further investigation is needed to determine whether it has other viral targets. In this study, we found that USP5 interacted with MAVS and IRF3, inhibiting the production of type I IFNs and disassembling the K63-linked polyubiquitin chains on MAVS and IRF3 during EV-A71 infection. Furthermore, we revealed that PR-619, a USP5 inhibitor, can increase the production of IFN-I to inhibit the replication of EV-A71.
## RESULTS
## EV-A71 infection upregulates USP5 expression to promote viral replication
The proteomic analysis of EV-A71-infected cells revealed that the expression of USP5 was upregulated (Supplementary Fig. S1A). To further confirm this phenomenon, we first infected HEK293T cells with EV-A71 (MOI = 1) for 24 h and assessed USP5 protein levels via Western blotting and USP5 mRNA levels via RT-qPCR. Compared to mock infection, the results of Western blotting (Fig. 1A andB) and RT-qPCR (Fig. 1D) revealed that the expression levels of USP5 were increased 24 h after infection with EV-A71. Again, the upregulated USP5 protein (Fig. 1C) and mRNA (Fig. 1E) levels can also be observed in EV-A71 infected mouse peritoneal macrophages compared to mock infection. These results revealed that USP5 is upregulated during EV-A71 infection.
To investigate the regulatory role of USP5 after EV-A71 infection, the expression of USP5 and VP1 of EV-A71 were markedly upregulated upon viral infection in HEK293T cells (Fig. 1F), HeLa cells and mouse peritoneal macrophages (Supplementary Fig. S1B-C). Again, HEK293T cells were transfected with increasing amounts of Myc-USP5 plasmids and subsequently infected with EV-A71 (MOI = 1). Western blot analysis revealed that USP5 upregulated the expression of the EV-A71-encoded VP1 protein in a dose-dependent manner (Fig. 1G). RT-qPCR analysis revealed that USP5 increased EV-A71 viral RNA levels in a dosedependent manner (Fig. 1H). Moreover, overexpression of USP5 upregulated the expression of EV-A71-encoded VP1 proteins in a timedependent manner (Fig. 1I). Similarly, EV-A71 viral titers (Fig. 1J) were also increased by overexpression of USP5 in HEK293T cells. And the expression of USP5 was also upregulated by EV-A71 in an MOIdependent manner (Fig. 1K). These results demonstrated that USP5 was a positive regulator of EV-A71 proliferation.
## USP5 knockdown inhibits EV-A71 replication
We next investigated the mechanism by which USP5 promotes EV-A71 replication. First, two shRNAs were designed to knock down USP5 expression and were used to construct USP5-knockdown cell lines. The results of the CCK-8 cell counting assay indicated that USP5 knockdown did not result in inhibition of cell proliferation for 24 h (Fig. 2A). RT-qPCR analysis (Fig. 2B) and Western blot analysis (Fig. 2C) revealed that both shRNAs were highly efficient at knocking down USP5 in HEK293T cells. Next, we infected shscramble or shUSP5 treated HEK293T cells with EV-A71 and found that the levels of EV-A71 VP1 protein (Fig. 2E) and viral RNA (Fig. 2D) were significantly lower in shUSP5-treated cells than those in shscramble-treated cells. Moreover, USP5 knockdown in HeLa cells also downregulated the expression of EV-A71-encoded VP1 proteins (Fig. 2F) in a time-dependent manner during EV-A71 infection. Together, these findings suggested that the knockdown of USP5 downregulated the replication of EV-A71.
## USP5 negatively regulates the IFN-I production induced by EV-A71
We next investigated the mechanism by which USP5 regulates EV-A71 replication. IFN-I plays a very important role in innate immunity, so we speculated that USP5 may regulate the replication of EV-A71 via IFN-I responses. To confirm this, we analyzed the mRNA expression of IFNA and IFNB in HeLa cells overexpressing USP5 during EV-A71 infection and found that the mRNA levels of IFNB and IFNA were downregulated (Fig. 3A andB). Since the production of IFN-I stimulates the expression of IFN-stimulated genes (ISGs), we next determined whether the overexpression of USP5 could regulate the expression of ISGs during EV-A71 infection. RT-qPCR analysis (Fig. 3C andD) revealed that the mRNA levels of typical ISGs, including ISG15 and RANTES, were inhibited by USP5 during EV-A71 infection. Consistent with our previous findings, the mRNA levels of IFN-I and ISGs were significantly increased in USP5-knockdown HeLa cells (Fig. 3E-H). These findings indicate that USP5 downregulates the innate immune pathway in HeLa cells infected with EV-A71. Given that IRF3 is an important member of the IFN regulatory family that controls multiple interferon induction pathways, we investigated its role in this process. Pathogens can induce IRF3 phosphorylation (P-IRF3), and P-IRF3 can then translocate into the nucleus to induce the production of IFN-I. We noted that the knockdown of USP5 increased the phosphorylation of IRF3 during EV-A71 infection in HeLa cells but did not affect the total protein level of IRF3 (Fig. 3I), whereas the overexpression of USP5 inhibited the phosphorylation of IRF3 (Fig. 3J). And similar results were found in HEK293T cells with USP5 knockdown (Supplementary Fig. S2A). Meanwhile, restoring wild-type USP5 expression in USP5knockdown cells rescues EV-A71 replication and suppress IFN/ISG responses (Fig. 3K). We also observed that the replication and innate immunity induced by Enterovirus D68 (EV-D68) (Supplementary Fig. S3A-B) and Coxsackievirus A16 (CVA16) (Supplementary Fig. S3C-D) are similarly regulated by USP5. Taken together, these results demonstrated that USP5 negatively regulates IRF3 activation and the production of IFN-I, which may reveal the influence of USP5 on the replication of several enteroviruses.
## USP5 interacts with MAVS and IRF3
Given that RNA viruses can be recognized by RIG-I and activate the IFN-I innate immune response, we hypothesized that USP5 might modulate the RIG-I/MAVS signaling pathway during EV-A71 infection. Thus, we transfected pivotal components of the RIG-I/MAVS pathway, including RIG-I, MAVS, TBK1, and IRF3 into HEK293T cells, and the potential USP5-interacting proteins were subsequently examined via immunoprecipitation experiments. The data revealed that exogenously Fig. 1. EV-A71 infection increases the mRNA and protein levels of USP5. A HEK293T cells were infected with EV-A71 (MOI = 1) for 24 h, after which the cells were subjected to immunoblotting with anti-USP5, anti-VP1 and anti-β-actin antibodies. B USP5 protein expressions were represented by Western blot bands in (A) and were quantified using ImageJ software. β-actin was used as control. n = 3 independent experiments were performed. C Mouse peritoneal macrophages were infected with EV-A71 (MOI = 5) for 24 h, after which the cells were subjected to immunoblotting with anti-USP5, anti-VP1 and anti-β-actin antibodies. D RT-qPCR analysis of USP5 mRNA levels in HEK293T cells infected with EV-A71 (MOI = 1) for 24 h. E RT-qPCR analysis of USP5 mRNA levels in mouse peritoneal macrophages infected with EV-A71 (MOI = 5) for 24 h. F Western blot analysis of USP5 and VP1 protein in HEK293T cells infected with EV-A71 (MOI = 1) at the indicated times. G HEK293T cells were transfected with different amount of Myc-USP5. After 24 h of transfection, the cells were infected with EV-A71 (MOI = 1) for 24 h. The resulting cell lysates were immunoblotted with the indicated antibodies. H RT-qPCR analysis of the EV-A71 mRNA levels in (G). I Western blot analysis of EV-A71 VP1 protein in HEK293T cells transfected with Myc-USP5 for 24 h and then infected with EV-A71 (MOI = 1) for different durations. J EV-A71 titers were titrated in HEK293T cells transfected with Myc-USP5 for 24 h and then infected with EV-A71 (MOI = 1) for 24 h. Mean ± SEM, statistical analysis was performed via unpaired two-tailed Student's t-test. *, P < 0.05; **, P < 0.01.K Western blot analysis of USP5 in HEK293T cells infected with different MOIs (1, 5 or 10) of EV-A71.
overexpressed MAVS and IRF3 could interact with USP5 (Fig. 4A andB) but not RIG-I or TBK1 (Fig. 4C andD). Meanwhile, the endogenous potential proteins that interact with endogenous USP5 were subsequently examined via immunoprecipitation experiments, and the results showed that USP5 could interact with MAVS and IRF3 (Fig. 4E andF), but not RIG-I or TBK1 (Fig. 4G andH). Moreover, our data demonstrated that USP5 interacts with MAVS and IRF3 during EV-A71 infection (Fig. 4I). Overall, we concluded that USP5 could interact with MAVS and IRF3 to regulate the IFN-I antiviral immune response.
## USP5 reduces the K63-linked polyubiquitination of MAVS and IRF3
Given that USP5 is a de-ubiquitinating enzyme that prefers to cleave Lys48-linked polyubiquitin and Lys63-linked polyubiquitin (Dayal et al., 2009), we further verified how USP5 regulates the ubiquitination and expression levels of MAVS and IRF3. Moreover, USP5 knockdown did not affect the expression of MAVS or IRF3 but increased the total polyubiquitination levels of IRF3 (Fig. 5A) and MAVS (Fig. 5B). Furthermore, to determine which types of ubiquitination of MAVS and IRF3 are regulated by USP5, we transfected HEK293T cells with HA-WT-Ub, HA-K48-Ub, or HA-K63-Ub constructs for immunoprecipitation analysis. The results revealed that the K63-linked but not the K48-linked polyubiquitination of IRF3 (Fig. 5C) and MAVS (Fig. 5E) were significantly increased in USP5-knockdown cells. To investigate whether the deubiquitinase activity of USP5 is crucial for regulating antiviral immunity, USP5 with mutated enzymatic activity (USP5-C335A) and USP5 lacking two ubiquitin-associated domains (USP5-ΔUBA) plasmids were constructed. The USP5-C335A and USP5-ΔUBA mutants did not affect the interaction between USP5 and IRF3 or MAVS but abolished USP5's ability to regulate K63-linked polyubiquitination of IRF3 (Fig. 5D) and MAVS (Fig. 5F). Moreover, the ability of USP5 to upregulate EV-A71 replication was dependent on its enzymatic activity, as the USP5-C335A and USP5-ΔUBA mutants failed to do so compared to wild-type USP5 (Fig. 5G andH). Together, our data demonstrated that USP5 can target MAVS and IRF3, affect the K63-linked polyubiquitination of MAVS and IRF3, and then modulate IRF3 activation and ISG production after EV-A71 infection in a manner dependent on its deubiquitinase activity.
## PR-619, a USP5 inhibitor, inhibits EV-A71 replication by activating the IFN-I signaling pathway
To explore the application of USP5 as a target to fight against EV-A71, we chose a selective small-molecule inhibitor of USP5, named PR-619, for further investigation. First, we treated HeLa cells with different concentrations of PR-619 (0-10 μM) for 24 h. Cell viability tests, assessed by the CCK-8 cell counting assay, revealed that PR-619 was nontoxic to cells (Fig. 6A). Then, HeLa cells were treated with different concentrations of PR-619 for 24 h after infection with EV-A71. The cytopathic effect (CPE) induced by EV-A71 revealed that PR-619 inhibited EV-A71 proliferation in HeLa cells in a dose-dependent manner (Fig. 6B). Since the above results showed that USP5 may regulate IFN-I responses activated by EV-A71 infection, as we expected, treatment with PR-619 increased the mRNA levels of IFNA and IFNB during EV-A71 infection (Fig. 6C andD). Given that high production of IFN-I can stimulate the production of IFN-stimulated genes (ISGs), RT-qPCR analysis revealed that PR-619 could increase the mRNA levels of typical ISGs, including ISG15 and RANTES in HeLa cells infected with EV-A71 (Fig. 6E andF). In contrast, PR-619 did not inhibit EV-A71 replication via IFN-I responses in USP5 knockout HeLa cells (Fig. 6G-L). Furthermore, we rescued USP5 in USP5 knockout cells treated with PR-619, and found that PR-619 exerted its inhibitory effects on EV-A71 replication and EV-A71-induced innate immunity via USP5 (Supplementary Fig. S4A-E). Consistent with our previous findings, noncytotoxic concentrations of PR-619 inhibited the replication of EV-A71 in HEK293T cells (Supplementary Fig. S2B) and mouse peritoneal macrophages (Fig. 7A andB), and upregulated IFN-I responses, including the mRNA levels of IFNA, IFNB, ISG15 and RANTES in mouse peritoneal macrophages (Fig. 7C-F). Together, these findings demonstrated that the inhibitor of USP5 positively regulated IFN-I expression and subsequent ISGs production, which resulted in reduced replication of EV-A71.
## DISCUSSION
Previous studies have demonstrated that the production of IFN-I activated by RNA viruses is an important mechanism for innate immunity (Iwasaki and Pillai, 2014;Wang, S. et al., 2021). To circumvent host antiviral defence, numerous viruses have evolved molecular strategies to antagonize critical nodes within the IFN-I signaling Fig. 3. USP5 knockdown negatively regulates IRF3 activation. A-D HeLa cells were transfected with vector plasmids or Myc-USP5 plasmids for 24 h, after which the cells were infected with EV-A71 (MOI = 1) for 24 h. The mRNA levels of IFNB (A), IFNA (B), ISG15 (C) and RANTES (D) were tested via RT-qPCR. E-H Shscramble or shUSP5-treated HeLa cells were infected with EV-A71 (MOI = 1) for 24 h. The mRNA levels of IFNB (E), IFNA (F), RANTES (G) and ISG15 (H) were tested via RT-qPCR. Independent experiments were performed in triplicate. Mean ± SEM, statistical analysis was performed via unpaired two-tailed Student's t-test. *, P < 0.05; **, P < 0.01. I Shscramble or shUSP5-treated HeLa cells were infected with EV-A71 (MOI = 1) for 0, 12 or 24 h. The protein levels of P-IRF3 (pSer396-IRF3), total IRF3, P-TBK1 (pSer172-TBK1) total TBK1, MxA and ISG15 were measured via Western blotting at the indicated times. J HeLa cells were transfected with vector plasmids or Myc-USP5 plasmids for 24 h, after which the cells were infected with EV-A71 (MOI = 1) for 0, 12 or 24 h. The protein levels of P-IRF3 (pSer396-IRF3), total IRF3, P-TBK1 (pSer172-TBK1) and total TBK1 were measured via Western blotting at the indicated times. K Shscramble, shUSP5-treated HeLa cells or shUSP5-treated HeLa cells overexpressed with wild type USP5 were respectively infected with EV-A71 for 24 h. The protein levels of P-IRF3 (pSer396-IRF3), total IRF3, P-TBK1 (pSer172-TBK1), total TBK1, MxA and ISG15 were measured via Western blotting. cascade. EV-A71 2C protein impairs TNF-α-induced NF-κB activation via an interaction with IκB kinase β (IKKβ), which leads to the suppression of IKKβ phosphorylation (Li, Q. et al., 2016). The EV-A71 3C proteinase can cleave IRF7, preventing RIG-I-induced innate immunity (Lei et al., 2013). Meanwhile, EV-A71 2A protease cleaves TRAF3 to inhibit STING activation (Zheng et al., 2023). EV-A71 actually has many ways of evading the host's natural immune response (Li, H. et al., 2023;Cui et al., 2024;Wei et al., 2024).
Posttranslational modifications play crucial roles in virus infection and cancer development (Kumar et al., 2020). Deubiquitination, which reverses the process of ubiquitination, is an important process (Sun, T. et al., 2020). USP11 can interact with the NP of influenza A virus (IAV) and de-ubiquitinate the NP to increase IAV replication (Liao et al., 2010). USP14 can also increase alphaherpesvirus proliferation by targeting VP16 via ER stress-induced selective autophagy (Ming et al., 2022). With respect to the posttranslational modification of EV-A71, the ubiquitin E3 ligase SPOP negatively regulates the EV-A71-encoded 2A protease (Zang et al., 2023b). Moreover, posttranslational modifications can also regulate EV-A71 by modulating innate immune pathways. USP24 can target TBK1 and restrict the K63-linked polyubiquitination of TBK1 to promote EV-A71 infection (Zang et al., 2023a). These studies demonstrated that posttranslational modifications of EV-A71 and immune molecules are important for the antagonistic mechanism between the host and the virus.
In this study, we found that the expression of USP5, a multifunctional DUB, was upregulated after EV-A71 infection and promoted EV-A71 replication. Mechanistically, our data revealed that USP5 knockdown could activate the RIG-I/MAVS signaling pathway, thereby inhibiting the replication of EV-A71. USP5 downregulates K63-linked polyubiquitination of MAVS and IRF3, but not K48-linked polyubiquitination, thereby stabilizing IRF3 expression, inhibiting IRF3 activation, and attenuating antiviral innate immune responses. Moreover, the influence of USP5 on the upregulation of EV-A71 expression mainly depends on its de-ubiquitinating enzyme activity. Future research should explore whether USP5 can interact with the structural or nonstructural proteins of EV-A71 to promote its replication. Meanwhile, we found that EV-D68 and CVA16 can also upregulate the expression of USP5, which in turn inhibits the innate immune response and promotes viral replication. These findings suggest that the regulation of USP5 is likely mediated primarily by host factors induced by these enteroviruses.
On the other hand, two important questions merit further investigation. First, it is unclear whether USP5 can regulate the proliferation of other enterovirus clades. Second, the specific E3 ligases that attach ubiquitin chains to MAVS and IRF3 need to be identified. Furthermore, we found that an inhibitor of USP5 inhibited EV-A71 replication in a dose-dependent manner. Owing to the high degree of homology among mammals, the regulation of USP5 in the production of IFN-I may be demonstrated in mouse models or humans infected with EV-A71. Therefore, an inhibitor of USP5 may be a potential treatment for hand, foot, and mouth disease patients.
The process of viral infection is intricate, and our investigation focused mainly on the relationships among EV-A71, USP5, and IFN-I. Whether USP5 can regulate other RNA virus-related immune pathways needs further study. Moreover, an inhibitor of USP5, PR-619, can suppress the replication of EV-A71, which is highly important for exploring the interaction between EV-A71 and the host and provides new antiviral targets. Although our results showed that USP5 significantly inhibits the EV-A71-induced antiviral immune response in both human and mouse cells, whether similar results can be obtained in vivo remains to be verified. What's more, our results and the previous report (Qiao et al., 2025) focus only on RNA viruses; whether USP5 has any effects on the replication of DNA viruses and other innate immune pathways requires further research.
## CONCLUSIONS
In conclusion, our findings demonstrated that the de-ubiquitinating enzyme USP5 downregulates IFN-I antiviral immunity and promotes viral replication during EV-A71 infection. Mechanistic studies revealed that USP5 acts as a negative regulator of innate immunity by specifically de-ubiquitinating K63-linked polyubiquitin chains on both MAVS and IRF3, thereby suppressing IRF3 activation. Moreover, PR-619, a USP5 inhibitor, inhibited the proliferation of EV-A71. These findings provide novel strategies to control EV-A71 infection.
## MATERIALS AND METHODS
## Cell culture and reagents
HEK293T and HeLa cells were commercially obtained from ATCC and cultured with Dulbecco's modified Eagle's medium (DMEM, Gibco), and mouse peritoneal macrophages were obtained from Wuhan SUNN-CELL Biotechnology Co., Ltd., and cultured with RPMI 1640 (Corning) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco) and 100 U/mL penicillin/streptomycin at 37 • C in an incubator with 5% CO 2 . DMSO and PR-619 were obtained from MCE. HeLa cells were treated with different concentrations of PR-619. After 24 h of treatment, the cell viability was tested via a CCK8 assay (G). The mRNA levels of EV-A71 (H), IFNA (I), IFNB (J), ISG15 (K) and RANTES (L) in USP5 -/-HeLa cells infected with EV-A71 and then treated with PR-619 (0-10 μM) were tested via RT-qPCR. The data are shown relative to β-actin expression. And the above results are compared to cells treated with PR-619 with 0 μM. Mean ± SEM, statistical analysis was performed via unpaired two-tailed Student's t-test. N.S, no significance; *P < 0.05, **P < 0.01 and ***P < 0.001.
## Virus infection and titration assay
EV-A71 strain H (VR-1432) was prepared as previously described (Qiu et al., 2017). HEK293T cells were infected with viruses after being transfected with Myc-tagged USP5 plasmids for 24 h. Cells were incubated with viruses for 1 h, washed twice with phosphate-buffered saline (PBS) and then added with complete medium. And the supernatants were collected after infection with 24 h. In addition, the collected culture supernatants was detected via a TCID 50 assay. TCID 50 assay was performed on vero cells in 96-well plates infected with a 10-fold serial dilution of viruses for 72 h. Viral titers were calculated via the Reed and Muench method (Gao, Q. et al., 2015).
Plaque assay was performed on vero cells in 24-well plates infected with a 10-fold serial dilution of viruses. The plates were incubated at 37 • C for 1 h to allow adsorption. Then the supernatant was removed, and cells were overlaid with 1% Carboxymethyl cellulose (Sigma-Aldrich) in DMEM containing 2% FBS. After further incubation at 37 • C for 72 h, the cells were fixed with 4% formaldehyde and stained with 0.2% crystal violet to visualize the plaques.
## Cell viability test
A CCK-8 assay was used to evaluate the cytotoxicity of PR-619 in mouse peritoneal macrophages, HeLa cells and sgUSP5-treated HeLa cells. Briefly, the cells seeded into 96-well plates were incubated with increasing concentrations of PR-619. After incubation at 37 • C for 12 h, the cell mixture was replaced with fresh DMEM. After incubation for another 12 h, CCK-8 solution was added to detect the absorbance at 450 nm. HEK293T sh-scramble cells and USP5-knockdown HEK293T cells seeded into 96-well plates were tested with CCK-8 solution as described above.
## Coimmunoprecipitation, western blotting and antibodies
Coimmunoprecipitation and Western blotting were conducted as previously described with minor modifications (Xu et al., 2021). In brief, the cell monolayers were washed with PBS and incubated on ice with lysis buffer containing 20 mM Tris-HCl, 150 mM NaCl, 1 mM EDTA, 1% NP-40 and a 1% protease and phosphatase inhibitor cocktail (Roche).
For each sample, 600 μL of protein lysate was incubated with 1 μg of antibody and 30 μL of protein A/G magnetic beads (MCE) overnight at 4 • C. The magnetic beads were washed 3 times with 1 mL of lysis buffer, after which the precipitates were detected by Western blotting.
For Western blotting, the cells were harvested and lysed as described above. Equal amounts of protein samples were denatured for 15 min in 5 × SDS-PAGE loading buffer (Beyotime). Proteins were separated on SDS-PAGE gels and then electrotransferred onto polyvinylidene fluoride membranes (Millipore), which were then blocked for 1 h at room temperature in Tris-buffered saline (Sigma-Aldrich) containing 5% nonfat milk powder and 0.1% Tween 20 (Sigma-Aldrich). Next, the membranes were incubated with primary antibodies at 4 • C overnight and then with the corresponding secondary antibodies conjugated to horseradish peroxidase at room temperature for 1 h. The protein bands were detected via an enhanced chemiluminescence (ECL) kit.
The following antibodies were used: anti-FLAG (Sigma), antiphospho-TBK1 (CST), anti-phospho-IRF3 (CST), anti-HA, anti-β-actin, anti-MAVS, anti-IRF3, anti-TBK1, anti-myc, anti-USP5, anti-RIG-I (from Protech, Wuhan, China), and anti-VP1 (made in our laboratory).
## Quantitative real-time PCR
Total RNA was extracted using TRIzol reagent (TaKaRa, Dalian, China) according to the manufacturer's instructions. cDNA was acquired by using SuperScript II Reverse Transcriptase (TaKaRa, Dalian, China). cDNA was diluted in nuclease-free water, and gene expression was analyzed via qPCR SYBR Green Master Mix (Yeasen) as described previously (Qiu et al., 2017(Qiu et al., , 2020;;Fang et al., 2021). The RT-qPCR primers used are listed in Supplementary Table S1. Differences between the experimental and control groups were tested via Student's t-test. P values less than 0.05 were considered statistically significant.
## Lentiviral shRNA or sgRNA packaging, infection and selection
For packaging lentiviruses, 1 μg of pLKO.1 shRNA plasmid or sgRNA plasmids (Addgene), 750 ng of psPA × 2 packaging plasmids (Addgene), and 250 ng of pMD2.G envelope plasmids (Addgene) were cotransfected into 7 × 10 4 HEK293T cells with 6 μL of Lipofectamine ( 2000) Reagent (Life Technologies). The supernatants containing lentiviruses were collected, filtered and stored at -80 • C. For infection, HEK293T and HeLa cells were incubated with viral stocks supplemented with 8 mg/mL polybrene (Solarbio) for 10 h and then supplied with fresh medium. The cells were selected with puromycin (Invitrogen) at 24 h.p.i .. The following primers were used: human USP5 1# shRNA: GACCACACGATTTGCCTCATT; human USP5 2# shRNA: GATAGACATGAACCAGCGGAT; human USP5 1# sgRNA: GAGTCTACTTGCACCTCCGG; human USP5 2# sgRNA: GGAGGTGCAAGTAGACTCGC.
## References
1. Calistri, Munegato, Carli et al. (2014) "The ubiquitin-conjugating system: multiple roles in viral replication and infection"
2. Cao, Qu, Liu et al. (2020) "Myristoylation of EV71 VP4 is essential for infectivity and interaction with membrane structure" *Virol. Sin*
3. Clague, Heride, Urb� E (2015) "The demographics of the ubiquitin system" *Trends Cell Biol*
4. Cui, Yang, Yan et al. (2024) "UBE3C restricts EV-A71 replication by ubiquitination-dependent degradation of 2C" *J. Virol*
5. Dayal, Sparks, Jacob et al. (2009) "Suppression of the deubiquitinating enzyme USP5 causes the accumulation of unanchored polyubiquitin and the activation of p53" *J. Biol. Chem*
6. Fang, Liu, Qiu et al. (2021) "Inhibition of viral suppressor of RNAi proteins by designer peptides protects from enteroviral infection in vivo" *Immunity*
7. Gao, Ciancanelli, Zhang et al. (2021) "TLR3 controls constitutive IFN-β antiviral immunity in human fibroblasts and cortical neurons" *J. Clin. Investig*
8. Gao, Yuan, Zhang et al. (2015) "Discovery of itraconazole with broad-spectrum in vitro antienterovirus activity that targets nonstructural protein 3A" *Antimicrob. Agents Chemother*
9. Han, Wang, Zhang et al. (2022) "The role of ubiquitination and deubiquitination in tumor invasion and metastasis" *Int. J. Biol. Sci*
10. Han, Wang, Cui et al. (2016) "SIRT1 inhibits EV71 genome replication and RNA translation by interfering with the viral polymerase and 5'UTR RNA" *J. Cell Sci*
11. He, Kuang, Xu et al. (2025) "TRIM38 inhibits zika virus by upregulating RIG-I/MDA5 pathway and promoting ubiquitin-mediated degradation of viral NS3 protein" *Viruses*
12. Iwasaki (2012) "A virological view of innate immune recognition" *Annu. Rev. Microbiol*
13. Iwasaki, Pillai (2014) "Innate immunity to influenza virus infection" *Nat. Rev. Immunol*
14. Jiang, Li, Peng et al. (2023) "Ubiquitin ligase enzymes and de-ubiquitinating enzymes regulate innate immunity in the TLR, NLR, RLR, and cGAS-STING pathways" *Immunol. Res*
15. Kaiho-Soma, Akizuki, Igarashi et al. (1417) "TRIP12 promotes small-moleculeinduced degradation through K29/K48-branched ubiquitin chains" *Mol. Cell*
16. Kumar, Mehta, Mishra et al. (2020) "Role of host-mediated posttranslational modifications (PTMs) in RNA virus pathogenesis" *Int. J. Mol. Sci*
17. Lei, Xiao, Xue et al. (2013) "Cleavage of interferon regulatory factor 7 by enterovirus 71 3C suppresses cellular responses" *J. Virol*
18. Li, Wang, Wang et al. (2023) "Secreted LRPAP1 binds and triggers IFNAR1 degradation to facilitate virus evasion from cellular innate immunity" *Signal Transduct. Targeted Ther*
19. Li, Zheng, Liu et al. (2016) "proteins of enteroviruses suppress IKKβ phosphorylation by recruiting protein phosphatase 1" *J. Virol*
20. Liao, Wu, Su et al. (2010) "Ubiquitination and deubiquitination of NP protein regulates influenza A virus RNA replication" *EMBO J*
21. Liu, Wu, Qin et al. (2018) "Broad and diverse mechanisms used by deubiquitinase family members in regulating the type I interferon signaling pathway during antiviral responses" *Sci. Adv*
22. Liu, Wang, Mueller et al. (2010) "Direct interaction between two viral proteins, the nonstructural protein 2C and the capsid protein VP3, is required for enterovirus morphogenesis" *PLoS Pathog*
23. Mao, Li, Che et al. (2024) "The ubiquitin ligase UBR4 and the deubiquitylase USP5 modulate the stability of DNA mismatch repair protein MLH1" *J. Biol. Chem*
24. Meng, Ai, Lei et al. (2019) "USP5 promotes epithelial-mesenchymal transition by stabilizing SLUG in hepatocellular carcinoma" *Theranostics*
25. Ming, Zhang, Wang et al. (2022) "Inhibition of USP14 influences alphaherpesvirus proliferation by degrading viral VP16 protein via ER stress-triggered selective autophagy" *Autophagy*
26. Pai, Lin, Peng et al. (2023) "The deubiquitinase Leon/USP5 interacts with Atg1/ULK1 and antagonizes autophagy" *Cell Death Dis*
27. Qian, Zhu, Huang et al. (2020) "Ubiquitin specific protease 5 negatively regulates the IFNs-mediated antiviral activity via targeting SMURF1" *Int. Immunopharmacol*
28. Qiao, Li, Zhang et al. (2025) "USP5 inhibits anti-RNA viral innate immunity by deconjugating K48-linked unanchored and K63-linked anchored ubiquitin on IRF3" *PLoS Pathog*
29. Qiu, Xu, Zhang et al. (2017) "Human virus-derived small RNAs can confer antiviral immunity in mammals" *Immunity*
30. Qiu, Xu, Wang et al. (2020) "Flavivirus induces and antagonizes antiviral RNA interference in both mammals and mosquitoes" *Sci. Adv*
31. Saguil, Kane, Lauters et al. (2019) "Hand-Foot-and-Mouth disease: rapid evidence review" *Am. Fam. Phys*
32. Schoggins, Wilson, Panis et al. (2011) "A diverse range of gene products are effectors of the type I interferon antiviral response" *Nature*
33. Solomon, Lewthwaite, Perera et al. (2010) "Virology, epidemiology, pathogenesis, and control of enterovirus 71" *Lancet Infect. Dis*
34. Sun, Liu, Yang (2020) "The role of ubiquitination and deubiquitination in cancer metabolism" *Mol. Cancer*
35. Sun, Lu, Ma et al. (2024) "Deubiquitinase USP5 regulates RIPK1 driven pyroptosis in response to myocardial ischemic reperfusion injury" *Cell Commun. Signal*
36. Tang, Yang, Wu et al. (2007) "Reticulon 3 binds the 2C protein of enterovirus 71 and is required for viral replication" *J. Biol. Chem*
37. Wan, Li, Liu et al. (2024) "USP5 promotes tumor progression by stabilizing SLUG in bladder cancer" *Oncol. Lett*
38. Wang, Xi, Lei et al. (2013) "Enterovirus 71 protease 2Apro targets MAVS to inhibit anti-viral type I interferon responses" *PLoS Pathog*
39. Wang, Dai, Qin et al. (2021) "Targeting liquid-liquid phase separation of SARS-CoV-2 nucleocapsid protein promotes innate antiviral immunity by elevating MAVS activity" *Nat. Cell Biol*
40. Wei, Lv, Wang et al. (2024) "Recent progress in innate immune responses to enterovirus A71 and viral evasion strategies" *Int. J. Mol. Sci*
41. Wen, Qi, Wang (2020) "The function and mechanism of enterovirus 71 (EV71) 3C protease" *Curr. Microbiol*
42. Wu, Berlemann, Bader et al. (2022) "LUBAC assembles a ubiquitin signaling platform at mitochondria for signal amplification and transport of NF-κB to the nucleus" *EMBO J*
43. Xiao, Shi, He et al. (2023) "ERK and USP5 govern PD-1 homeostasis via deubiquitination to modulate tumor immunotherapy" *Nat. Commun*
44. Xu, Kong, Lyu et al. (2021) "The capsid protein of Rubella virus antagonizes RNA interference in mammalian cells" *Viruses*
45. Yao, Hu, Xi et al. (2019) "Transcriptomic analysis of cells in response to EV71 infection and 2A(pro) as a trigger for apoptosis via TXNIP gene" *Genes Genomics*
46. Yau, Doerner, Castellanos et al. (2017) "Assembly and function of heterotypic ubiquitin chains in cell-cycle and protein quality control" *Cell*
47. Yeh, Yu, Yang et al. (2013) "Ubiquitin-specific protease 13 regulates IFN signaling by stabilizing STAT1" *J. Immunol*
48. Yin, Liu, Wimmer et al. (2007) "Complete protein linkage map between the P2 and P3 non-structural proteins of poliovirus" *J. Gen. Virol*
49. Zang, Gu, Yang et al. "2023a. Ubiquitin-specific protease 24 promotes EV71 infection by restricting K63-linked polyubiquitination of TBK1" *Virol. Sin*
50. Zang, Yang, Chen et al. (2023) "Ubiquitin E3 ligase SPOP is a host negative regulator of enterovirus 71-encoded 2A protease" *J. Virol*
51. Zhang, Sun, Chang et al. (2008) "Characterization of hand, foot, and mouth disease in China between" *Biomed. Environ. Sci*
52. Zhang, Zhang, Bie et al. (2022) "STUB1 regulates antiviral RNAi through inducing ubiquitination and degradation of dicer and AGO2 in mammals" *Virol. Sin*
53. Zheng, Zhou, Rui et al. (2023) "TRAF3 activates STING-mediated suppression of EV-A71 and target of viral evasion" *Signal Transduct. Targeted Ther*
54. Zhou, Luo, Bi et al. (2024) "Amelioration of nonalcoholic fatty liver disease by inhibiting the deubiquitylating enzyme RPN11" *Cell Metab*
55. Zhou, Tian, Wang et al. (2022) "EV71 3C protease cleaves host anti-viral factor OAS3 and enhances virus replication" *Virol. Sin* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12509679&blobtype=pdf | # Analysis of the genomes and phylogenetic relationships of lytic Pseudomonas aeruginosa PA14 phages BL5, BL8, and BL9 from aquatic sources
Fahareen Mosharraf, Enoch Ghosh, Carson Bellew, Austen Rowell, Lisa Bono
## Abstract
Three lytic bacteriophages targeting Pseudomonas aeruginosa PA14 were isolated from Lubbock, TX, playa lakes. Genomic sequencing and bioinformatic analysis classified these phages as unclassified species in the Pbunavirus genus of the Caudoviri cetes class. KEYWORDS bacteriophages, environmental isolates, genomic characterization, phylogenetic analysis P seudomonas aeruginosa PA14 is a standard laboratory strain and a highly drug-resist ant, Gram-negative opportunistic pathogen commonly found in environments, such as soil and water (1-4). We report the isolation, sequencing, and characterization of three bacteriophages infecting PA14, sourced from playa lakes (ephemeral, rain-filled ponds and wetlands) in Lubbock, Texas.On 11 August 2022, water samples were collected from Lubbock, TX, at depths of 7-15 cm, at coordinates 33.566°, -101.803° (BL5 and BL8) and 33.511°, -102.002° (BL9). Samples were filtered successively through 0.45-micron and 0.22-micron syringe filters to eliminate debris and non-intended microbial cells, respectively. Each filtered sample (200 µL) was mixed with 400 µL of stationary PA14 culture and incubated at 37°C overnight. Post-incubation, samples were centrifuged at 1,786×g for 30 min and filtered again through a 0.22-micron filter. Phages were isolated via the agar-overlay method on PA14 lawns (5), followed by triple plaque purification. Phage DNA was extracted using the Quantabio Extraction DNA Kit, with quality and concentration assessed via the BioTek Take3 microvolume plate. DNA libraries were prepared with the Illumina DNA Prep Tagmentation Kit and sequenced on the Illumina NextSeq 2000 using a 300-cycle flow cell kit, producing 150-bp paired-end reads.Raw sequence data were assembled, and the genome was annotated, following our earlier announcements (6, 7). The software fastp v0.23.4 (8) was used for trimming and quality control. Host contaminants (NCBI GeneBank accession no. NZ_CP127126.1) were removed using bowtie2 (9). Genomes were assembled with SPAdes v3.15.5 (10), and coverage and depth were evaluated using SAMtools v1.6 (11) and BEDTools v2.31.0 (12). Assembly metrics were generated with Quast v2.2.4 (13), and genome completeness was verified using CheckV v1.0.1 (14). Taxonomic classification was performed with NCBI BLASTn, aligning sequences to the NCBI nucleotide database. Genomes were annotated using PHROGS via Pharokka v1. 3.2 (15, 16). We assessed the sequences for alternative codon usage using both Pharokka Prodigal algorithm (17) and NCBI ORF finder with standard genetic code and alternative start/stop codon usage option. FastANI v1.34 (18) was used for pairwise genome comparisons to identify closest relatives. All software was used with default settings. Phage morphology was examined using a Hitachi H-7650 TEM at TTU's CASM facility, with 1% uranyl acetate staining.
consequently, the genome sequence for BL5 was categorized as partial; however, BL8 and BL9 were categorized as complete. Sequence analysis (see Table 1) placed the phages in an unclassified species within the genus Pbunavirus in the class Caudoviricetes. Despite being classified similarly, these phages do not share the same host as those identified from earlier work (6,7). Electron microscopy showed phage particles with lengths of 160 nm and head diameters of 50 nm (Fig. 1A through C). Whole genome sequences from the NCBI database (retrieved 12 June 2025) (19) were aligned with MAFFT v7.526 (20), and phylogenetic trees were inferred using IQ-TREE v2. 3.6 (21), visualized with FigTree v1.4.4 (22) (Fig. 1D).
## References
1. Takeya, Amako (1966) "A rod-shaped pseudomonas phage" *Virology (Auckl)*
2. Bradley (1972) "A study of pili on Pseudomonas aeruginosa" *Genet Res*
3. Arora, Bangera, Lory et al. (2001) "A genomic island in Pseudomonas aeruginosa carries the determinants of flagellin glycosyla tion" *Proc Natl Acad Sci*
4. Vallet, Olson, Lory et al. (2001) "The chaperone/ usher pathways of Pseudomonas aeruginosa: identification of fimbrial gene clusters (cup) and their involvement in biofilm formation"
5. Mosharraf, Marpa, Rojas et al. (2024) "The genome sequences of lytic Pseudomonas aeruginosa bacteriophages BL1, BL2, and BL3 isolated from the environment" *Microbiol Resour Announc*
6. Mosharraf, Rowell, Bono (2024) "The genomic and phylogenetic characterization of lytic Pseudomonas aeruginosa PAK bacteriophages BL4, BL6, and BL7 collected from aquatic samples" *Microbiol Resour Announc*
7. Chen, Zhou, Chen et al. (2018) "Fastp: an ultra-fast all-in-one FASTQ preprocessor" *Bioinformatics*
8. Langmead, Salzberg (2012) "Fast gapped-read alignment with Bowtie 2" *Nat Methods*
9. Bankevich, Nurk, Antipov et al. (2012) "SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing" *J Comput Biol*
10. Li, Handsaker, Wysoker et al. (2009) "1000 genome project data processing subgroup. The sequence alignment/map format and SAMtools" *Bioinformatics*
11. Quinlan, Hall (2010) "BEDTools: a flexible suite of utilities for comparing genomic features" *Bioinformatics*
12. Gurevich, Saveliev, Vyahhi et al. (2013) "QUAST: quality assessment tool for genome assemblies" *Bioinformatics*
13. Nayfach, Camargo, Schulz et al. (2021) "CheckV assesses the quality and completeness of metagenomeassembled viral genomes" *Nat Biotechnol*
14. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J Mol Biol*
15. George, Nepal, Houtak et al. (2023) "Pharokka: a fast scalable bacteriophage annotation tool" *Bioinformatics*
16. Cook, Telatin, Bouras et al. (2023) "Predicting stop codon reassignment improves functional annotation of bacteriophages"
17. Jain, Lm, Phillippy et al. (2018) "High throughput ANI analysis of 90K prokaryotic genomes reveals clear species boundaries" *Nat Commun*
18. Nucleotide (2004) "National Library of Medicine (US), National Center for Biotechnology Information"
19. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
20. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: new models and efficient methods for phylogenetic inference in the genomic era" *Mol Biol Evol*
21. Rambaut (2009) "FigTree. Tree figure drawing tool" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12098820&blobtype=pdf | # Correction: Therapeutic treatment of hepatitis E virus infection in pigs with a neutralizing monoclonal antibody
Isabella Hrabal, Elmira Aliabadi, Sven Reiche, Saskia Weber, Cora Holicki, Laura Schmid, Christine Fast, Charlotte Schröder, Benjamin Gutjahr, Patrick Behrendt, Martin Groschup, Martin Eiden, Scientifc Reports
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282416&blobtype=pdf | # Coadministration of ribavirin and arenaviral entry inhibitor LHF-535 enhances antiviral benefit against authentic lassa virus
Cheng Peng, Jialing Hu, Yuan Bai, Wei Wu, Wenting Mao, Yang Liu, Yi Wan, Lei Zhang, Wei Li, Tingting Tian, Tiezhu Liu, Yanhai Wang, Mifang Liang, Jun Han, Zhiming Yuan, Jiandong Li, Chao Shan, Fei Deng, Wei Wang, Virologica Sinica
Dear Editor, Lassa virus (LASV) is the causative agent of the acute viral hemorrhagic Lassa fever (LF), which is classified into Mammarenavirus within the Arenaviridae family, with a single-stranded, negative-sense, bisegmented RNA genome. Due to its high pathogenicity and lethality, LASV is considered as a priority threat to public health, with an estimated cases of 300,000 infections and 5000 deaths annually. LASV was first isolated and described as a clinical entity in 1969 in Lassa, Nigeria (Garry, 2023). LASV isolates of different geographic and host origins are highly diverse in genomic sequences and phylogenetically classified into up to seven lineages, with each lineage predominately localized in specific countries. Although the research on LF has been carried out for decades since the pathogen first characterized, there is no approved antiviral drugs or vaccines for clinical use against LASV to date (Grant et al., 2023). One possible reason that hindered the development of countermeasures is that the preclinical studies on authentic LASV are restricted in high bio-containment biosafety level 4 (BSL-4) facilities. In this letter, we describe isolation, and characterization of the LASV from the clinical samples. And we applied a coadministration assay of antiviral drugs for LASV by using a clinically isolated Mammarenavirus lassaense strain in the BSL-4 facility, aiming to investigate new therapeutic strategies for LASV infection.
The LASV positive specimen (Ran et al., 2024;Li and Luo, 2025) was centrifuged at 15,000Âg for 10 minutes in BSL-4 laboratory; the supernatant was then diluted 10 times in minimal essential medium and inoculated onto monolayers of African green monkey kidney (Vero E6) cells. The cells were cultured at 37 C in 5 % carbon dioxide and observed daily to monitor for cytopathic effect.
As suggested by the results of IFAs, LASV was isolated and amplified after third passages on Vero E6 cells (Fig. 1A). Viral titers were quantified via plaque assay. The second-generation LASV isolate yielded 1.0 Â 10 5 PFU/mL, while the third-generation isolate reached 1.0 Â 10 6 PFU/mL. The TEM analysis showed that LASV is a particle with a diameter of approximately 100 nm although the precise structure of spikes was not obviously shown on the virus surface (Fig. 1B). The immature LASV particles without envelope were mainly located in cytoplasmic vesicles of Vero E6 cells (Fig. 1C). On the third passage a total number of nonredundant 9,444,636 reads were obtained by RNAseq from the third passage supernatant. The sequence comparison analysis showed that 153,402 reads were closely related to the LASV, which had never been reported in China. The full-length genomic sequence of a new LASV strain, which contains two single-stranded negative-sense RNA segments (L and S). The L segment is 7.2 kb in length, and the S segment is 3.4 kb in length. Phylogenetic analysis revealed that the L and S segments of this LASV isolate cluster with strains from lineage IV, indicating that this strain belongs to lineage IV (Supplementary Fig. S1). The complete genome sequence of the LASV strain (Lassa_HX strain) has been deposited in Science Data Bank under the CSTR identifier: 31253.11.sciencedb.22517 (https://www.scidb.cn/s/veQvMr).
Although approved antiviral agents against LASV infections are still lacking, the nucleoside analog ribavirin is currently the primary treatment available for LF (Eberhardt et al., 2019;Groger et al., 2023). As an inhibitor of viral RNA-dependent RNA polymerase (RdRp), ribavirin can inhibit viral genome replication and upregulate host IFN type I responses (Feld and Hoofnagle, 2005). Although commonly used, ribavirin is considered only effective when administered in the early course of disease and cause significant side effects in clinical trials (Cheng et al., 2022;Grant et al., 2023). Interestingly, the entry step of LASV appears to be an effective target for the development of antiviral inhibitors. LHF-535, an analog of the viral entry inhibitor ST-193, which displays an inhibitory activity against the viral envelope glycoproteins from all LASV lineages (Madu et al., 2018), was perceived as a leading antiviral candidate for LF treatment (Amberg et al., 2022;Garry, 2023).
For the development of pharmacotherapy, the combination of drugs targeting different steps in viral life cycle may provide a promising treatment for LF. As Fig. 2 shows, the combined effect of antiviral reagents ribavirin and LHF-535 against LASV were evaluated in vitro by using the immunofluorescence assay. After infecting Vero E6 cells with LASV, the numbers of IFA-positive cells in the presence of antiviral reagents were compared with those in the vehicle control group at 48 hours post-infection. The results suggested that both ribavirin and LHF-535 inhibit LASV infection in Vero E6 cells (Fig. 2A). Moreover, the combination of ribavirin and LHF-535 showed a higher inhibitory effect compared to the use of either drug alone (Fig. 2A). According to the analysis by the MacSynergy II, the combinations of LHF-535 with ribavirin had significant synergistic antiviral effects with the volume of 377.57 nM 2 %, which mechanistically reflects their complementary inhibition of viral entry and RNA-dependent RNA polymerase activity (Fig. 2B).
The combination of the RdRp inhibitor ribavirin and the viral entry inhibitor LHF-535 appears synergistic, with drug combination assay showing enhanced protection against LASV infection. With synergistic interactions, the combination of antiviral drugs may achieve stronger therapeutic effects. Our findings establish a relationship between viral RdRp and entry inhibitors, highlighting the clinical significance of ribavirin-based combinations in LASV infection treatment.
In a clinical case of pharmacotherapy, a combination of ribavirin and another RdRp inhibitor favipiravir was proved a survival improvement on patients with LF (Raabe et al., 2017), indicating effective antiviral inhibition on LASV replication. The coadministration of drugs targeting different processes in LASV lifecycle are seldom reported. The fusion inhibitor LHF-535 and the nucleoside analog favipiravir have shown better promise in animal models of LF than administered the drugs individually (Westover et al., 2024), evidencing the benefits of the drug combination. The data presented in this study expand the possible pharmaceutical compositions in preventing LASV infection, providing promising progress to address the need for new therapeutics of LF.
Even tremendous efforts have been made in antiviral drug development, the emergence of drug resistance in pathogenic virus hampers the effectiveness of antiviral agents and brings challenges to the treatment of infectious viral diseases. In LASV, generation of the resistance to entry inhibitors in mutants were also reported (Wang et al., 2018;Zhang et al., 2019;Liu et al., 2021). Upon LASV cell entry, the binding of LASV envelope glycoprotein complex (GPC) to the cell surface receptor α-dystroglycan (α-DG) is the major process mediates the entry of the virus. As reports showed, the alternation of single amino acids within GPC complex modulates sensitivity of LASV on entry inhibitors (Madu et al., 2018;Zhang et al., 2024), especially the amino acids in and near the predicted transmembrane domain of the GP2 subunit (Larson et al., 2008). Building on the data from our previous screen of adaptive mutants, the F446L mutation was proved to be a barrier to the antiviral activity of bergamottin and compound 57 (Zhang et al., 2019;Liu et al., 2021). Besides, the V434I mutation on GP2 attenuates the effect of LHF-535 on LASV inhibition, according to the divergent drug sensitivity of the LP strain carrying this mutant (Madu et al., 2018). In addition, the replacement of amino acid T40 located in the stable-signal peptide (SSP) subunit endowed LASV resistance to the entry inhibitor lacidipine (Wang et al., 2018). Moreover, ribavirin has recently been characterized as an RNA mutagen, inducing mutagenesis in intracellular viral genomes to inhibit LASV infection (Hu et al., 2024). In our results, the great inhibition of LASV on the application of drug combinations provides new therapeutic strategies to combat the pathogen, and provides several alternatives to prevent the emergence of antimicrobial resistance. In future studies, we would employ long-term serial passaging of LASV under escalating ribavirin/LHF-535 concentrations to systematically characterize resistance pathways and identify potential compensatory mutations.
In summary, we isolated a strain of LASV. Moreover, we described the viral characteristics of the isolation and evaluated the co-administration efficacy of nucleoside analog and entry inhibitor against the LASV infection, demonstrating enhanced antiviral activity. Fig. 2. Synergistic effect of ribavirin and LHF-535. A Effects of ribavirin and LHF-535 on inhibiting the LASV infection. Vero E6 cells were preincubated with compounds or vehicle at 37 C for 1 h, followed by incubation with LASV (MOI, 0.175) at 4 C for 1 h. Cells were stained with an anti-LASV NP antibody (green), and the nuclei were stained with DAPI (blue). Scale bars, 1 mm. B Differential surface plots at 95 % confidence level (CI) were calculated and generated using MacSynergy II for the drug-drug interactions for evaluating combinations of ribavirin with LHF-535 targeting LASV. MacSynergy II defines nM 2 % values as follows: strong synergy: >100 nM 2 %; slight synergy: 50-100 nM 2 %; additive effect: À50 to 50 nM 2 %; antagonism: <À50 nM 2 %.
## References
1. Amberg, Snyder, Vliet-Gregg et al. (2022) "Safety and pharmacokinetics of LHF-535, a potential treatment for lassa fever, in healthy adults" *Antimicrob. Agents Chemother*
2. Cheng, French, Salam et al. (2022) "Lack of evidence for ribavirin treatment of lassa fever in systematic review of published and unpublished studies(1)" *Emerg. Infect. Dis*
3. Eberhardt, Mischlinger, Jordan et al. (2019) "Ribavirin for the treatment of Lassa fever: a systematic review and meta-analysis" *Int. J. Infect. Dis*
4. Feld, Hoofnagle (2005) "Mechanism of action of interferon and ribavirin in treatment of hepatitis C" *Nature*
5. Garry (2023) "50 Years of lassa fever research" *Curr. Top. Microbiol. Immunol*
6. Grant, Samuels, Garry et al. (2023) "Lassa fever natural history and clinical management" *Curr. Top. Microbiol. Immunol*
7. Groger, Akhideno, Kleist et al. (2023) "Pharmacokinetics of ribavirin in the treatment of lassa fever: an observational clinical study at the irrua specialist teaching hospital, edo state" *Clin. Infect. Dis*
8. Hu, Bai, Tian et al. (2024) "A novel bsl-2 lassa virus reverse genetics system modelling the complete viral life cycle" *Emerg. Microb. Infect*
9. Larson, Dai, Hosack et al. (2008) "Identification of a broad-spectrum arenavirus entry inhibitor" *J. Virol*
10. Li, Luo (2025) "The first imported case of lassa fever in China: a case description" *Quant. Imag. Med. Surg*
11. Liu, Guo, Cao et al. (2021) "Screening of botanical drugs against lassa virus entry" *J. Virol*
12. Madu, Files, Gharaibeh et al. (2018) "A potent Lassa virus antiviral targets an arenavirus virulence determinant" *PLoS Pathog*
13. Raabe, Kann, Ribner et al. (2017) "Favipiravir and ribavirin treatment of epidemiologically linked cases of lassa fever" *Clin. Infect. Dis*
14. Ran, Chen, Huang et al. (2024) "The first case of Lassa fever in China" *Electronic J. Emerging Infectious Dis*
15. Wang, Liu, Zhang et al. (2018) "Screening and identification of lassa virus entry inhibitors from an FDA-approved drug library" *J. Virol*
16. Westover, Bailey, Wasson et al. (2024) "Coadministration of LHF-535 and favipiravir protects against experimental Junín virus infection and disease" *Antivir. Res*
17. Zhang, Cao, Cai et al. (2019) "Structure-activity relationship optimization for lassa virus fusion inhibitors targeting the transmembrane domain of GP2" *Protein Cell*
18. Zhang, Takenaga, Fehling et al. (2024) "Hexestrol, an estrogen receptor agonist, inhibits Lassa virus entry" *J. Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12846582&blobtype=pdf | # Detection and Molecular Characterisation of Protoparvovirus carnivoran1 in Golden Jackals (Canis aureus) in Croatia
Ivona Coric, Gorana Miletic, Dean Konjevic, Ivica Boskovic, Miljenko Bujanic, Alenka Skrinjaric, Snjezana Kovac, Ljubo Barbic, Andreja Jungic, Vladimir Stevanovic
## Abstract
Protoparvoviruses are highly contagious pathogens that cause severe, often fatal diseases in both domestic and wild carnivores. Golden jackal (Canis aureus) populations have experienced expansion in recent years, increasingly occupying urban and peri-urban areas. Despite this, they remain largely overlooked in scientific research. This study aimed to detect and characterise Protoparvovirus carnivoran1 circulating in a golden jackal population in Croatia and to assess their role in the epidemiology of parvovirus infections in companion animals. Small intestines from 55 jackals hunted in 2024 and 2025 were tested for Protoparvovirus carnivoran1 using real-time PCR. Positive samples were found across all sampling sites, with an overall positivity rate of 40%. Based on characteristic amino acid residues within the VP2 protein, the viruses detected in jackals were classified as feline panleukopenia virus (FPV). Phylogenetic analysis of the VP2 protein demonstrated considerable genetic diversity among strains circulating in Croatia. Additionally, a distinct group was identified, shared exclusively by Croatian domestic cats and golden jackals. Amino acid analysis revealed the novel A91T mutation, found only in jackals, and the E411Q mutation, unique to Croatian FPV strains. Structural modelling of the VP2 protein indicates that the observed mutations are located on the protein surface, within the antibody-binding site. These findings highlight the potential role of wild carnivores in parvovirus epidemiology and underscore the importance of including them in future surveillance and research efforts.
## 1. Introduction
Protoparvovirus carnivoran1 is a small, non-enveloped, single-stranded DNA virus classified within the family Parvoviridae [1,2]. Its genome is approximately 5.2 kb long and contains two main open reading frames (ORFs). ORF1 encodes the non-structural proteins NS1 and NS2, while ORF2 encodes the structural proteins VP1 and VP2. VP2 plays a key role in determining viral antigenicity and host range. Mutations in VP2 can alter receptor binding and facilitate immune escape. The parvoviral capsid is composed of 60 copies of a single structural protein organised into icosahedral symmetry. The main capsid component, VP2, has an eight-stranded antiparallel β-barrel core, from which approximately two-thirds of the polypeptide chain extends as loop insertions connecting β-strands. These loop regions form spike-like protrusions located at or near the icosahedral threefold axes. The β-barrel core creates a cylindrical depression around the icosahedral fivefold axes [3].
The two most important pathogens in the species Protoparvovirus carnivoran1 are feline panleukopenia virus (FPV) and canine parvovirus type 2 (CPV-2). FPV was first described in 1928 as a potential causative agent of acute gastroenteritis and leukopenia in cats [4]. FPV infects animals of all ages but particularly young ones, with mortality rates reported between 50% and 80%. Its high pathogenicity has established FPV as one of the most significant viral pathogens in domestic cats [5].
Canine parvovirus is closely related to FPV and is believed to have originated from it [6,7]. The specific ancestral viral strain that gave rise to CPV-2 has not yet been identified. A role for wildlife reservoirs in the emergence of CPV-2 has been proposed, but conclusive evidence remains lacking. Since its emergence, CPV-2 has undergone multiple evolutionary events, resulting in new antigenic variants that have rapidly spread worldwide [7]. In contrast, FPV has remained largely antigenically stable since its discovery, with studies indicating a substantially lower rate of evolution compared to CPV-2 [8].
The Protoparvovirus carnivoran1 host range is mainly determined by the interaction between VP2 and the transferrin receptor, which facilitates viral entry into host cells [9]. While CPV-2 has a wide host range, FPV has traditionally been considered to be confined to domestic and wild felids and incapable of infecting dogs [10]. However, in 2019, a strain of FPV with the A300P mutation was isolated and demonstrated the ability to infect canines [11].
Changes in host range can significantly influence the epidemiology of infectious diseases, including Protoparvovirus carnivoran1 infections. For instance, recent studies have documented an expansion of FPV's host range to include other carnivorous species [12][13][14][15][16]. The significance of changes in host range in the epidemiology of Protoparvovirus carnivoran1 infections is underscored today, as urbanisation brings wildlife into close contact with humans. Wild animals, like golden jackals, increasingly rely on human-derived food sources [17,18], enabling them to thrive in urban and peri-urban environments [19].
In Croatia, two distinct populations of golden jackal have been identified. The longestablished Dalmatian population inhabits coastal areas along the Adriatic Sea and has remained relatively stable. On the other hand, the continental population, once considered extinct, is currently undergoing rapid expansion, forcing jackals to share habitat with humans and domestic animals [20,21]. This may create opportunities for pathogen spillover from the domestic to wild carnivore population, as well as in the opposite direction. Investigating the role of golden jackals in the epidemiology of Protoparvovirus carnivoran1 is thus essential for understanding parvovirus dynamics in domestic and wild species, as they may serve as bridging hosts.
This study aimed to investigate the circulation of Protoparvovirus carnivoran1 in golden jackals in continental Croatia and to compare the obtained sequences with strains currently circulating in domestic animals from the same area.
## 2. Materials and Methods
## 2.1. Clinical Samples
As part of a game management plan during 2024 and 2025, 55 small intestine samples were obtained from golden jackals by local hunting associations. These samples were https://doi.org/10.3390/v18010123
Viruses 2026, 18, 123 collected from continental Croatia: Osijek-Baranja County, Sisak-Moslavina County and Zagreb County (Figure 1). The samples were submitted to the Faculty of Veterinary Medicine, University of Zagreb, on the same day and stored at -70 • C until subsequent viral DNA extraction. No information on pathological findings, sex or age of animals was available. In addition to the jackal samples, 11 archived samples from routine diagnostics at the Virology Laboratory, Faculty of Veterinary Medicine, University of Zagreb, were included in this study. These rectal swabs originated from domestic cats and had previously tested positive for FPV during routine diagnostics. Samples were collected in 2024 and 2025 from continental Croatia (Figure 1).
## 2.2. DNA Extraction and Polymerase Chain Reaction (PCR)
Viral DNA from small intestine samples was extracted using the DNeasy Blood and Tissue Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Viral DNA from rectal swabs was extracted using the IndiSpin Pathogen Kit (INDICAL BIOSCIENCE, Leipzig, Germany) according to the manufacturer's instructions.
Viral DNA was detected using a minor groove binder real-time PCR assay, as previously described [22]. Briefly, following the activation of GoTaq ® Hot Start Polymerase included in GoTaq ® qPCR Master Mix (Promega Corporation, Madison, WI, USA) at 95 • C for 2 min, 45 cycles of a two-step PCR were performed, consisting of denaturation at 95 • C for 30 s and primer annealing/extension at 60 • C for 1 min. Fluorescence was measured during the annealing/extension step, and samples with a cycle threshold value below 35 were considered positive. Reactions were performed on a Rotor-Gene Q real-time cycler (QIAGEN, Hilden, Germany).
In positive samples, the full-length VP2 gene was amplified using previously published, overlapping PCR primer pairs [23], with PCR conditions optimised for the present study. Following activation of TaKaRa Taq™ DNA Polymerase Hot Start Version (Takara Bio, Shiga, Japan) at 95 • C for 5 min, touchdown PCR was performed with annealing temperatures decreasing from 55 • C to 45 • C over 45 cycles. Primers and the expected amplicon sizes are listed in Supplementary Table S2. PCR products were visualised by electrophoresis on a 1% agarose gel stained with ethidium bromide and viewed under UV illumination using a Uvidoc HD6 (UVITEC, Cambridge, UK). Amplified VP2 segments were subjected to Sanger sequencing using the identical primer pairs (LGC Genomics GmbH, Berlin, Germany).
## 2.3. Sequence Alignment and Mutation Site Analysis
Nucleotide sequences obtained in this study were manually visualised and assembled using MEGA version 11.0 [24]. The resulting consensus sequences were compared against the National Centre for Biotechnology Information (NCBI) nucleotide database using the Basic Local Alignment Search Tool 2. 17. 0 (BLAST) [25]. Amino acids at positions 80, 93, 103, 323, 564, and 568 were then analysed to characterise Protoparvovirus carnivoran1 subtype [26].
Amino acid sequences of Croatian FPV isolates were compared to the reference strain CU-4 (GenBank accession number M38246) to identify positions of amino acid substitutions. Additionally, all available VP2 sequences from golden jackals in the NCBI database were retrieved, and mutation sites were compared with those of the Croatian isolates using MEGA software.
## 2.4. Phylogeny Analysis
The phylogenetic analysis included full-length VP2 sequences of all 18 Croatian isolates, along with 103 FPV sequences from various countries and time periods retrieved from the NCBI database. A list of all sequences obtained from NCBI is provided in Supplementary Table S3.
A maximum-likelihood phylogenetic tree of the complete FPV VP2 gene was constructed in IQ-TREE2 with 1000 bootstrap replicates [27]. The Tamura-3 parameter model [28] was selected based on the Akaike Information Criterion and the Bayesian Information Criterion using ModelFinder. The resulting phylogenetic tree was visualised using iTOL software version 7.2.2 [29].
## 2.5. Three-Dimensional Modelling Prediction of FPV VP2 Protein Structure
To assess the position of the amino acid substitution in the three-dimensional structure of the VP2 protein, we modelled these substitutions using the mutagenesis function in PyMOL version 3.1.6 [30]. The crystal structure of FPV (PDB ID: 1FPV) was used for molecular modelling [31]. The modelled structure allowed assessment of structural changes resulting from observed amino acid substitutions.
## 2.6. GenBank Accession Numbers
All Croatian isolates obtained in this study were submitted to the NCBI database, and accession numbers were generated. A list of all accession numbers is provided in Table 1.
## 3. Results
## 3.1. Detection and Characterisation of Protoparvovirus carnivoran1
Out of 55 golden jackals, 22 animals (40.0%) tested positive for the presence of Protoparvovirus carnivoran1 using real-time PCR. This represents the first detection of Protoparvovirus carnivoran1 in golden jackals in Croatia. Positive cases were detected in all sampled locations. In Zagreb County, 3 out of 9 samples were positive (33.3%), while in Osijek-Baranja County, 11 out of 35 samples tested positive (31.4%). The highest positivity rate was observed in Sisak-Moslavina County, where seven out of 11 samples were positive (63.6%).
Complete VP2 gene sequences (1755 base pairs, encoding 584 amino acids) were obtained from seven of the positive golden jackals and all domestic cats. Sequences were obtained from golden jackal samples with Ct values ranging from 22.31 to 31.70, whereas samples with higher Ct values did not yield sequences (Supplementary Table S1). Amino acid analysis at positions 80, 93, 103, 323, 564, and 568 [26] indicated that all Croatian sequences examined in this study matched the characteristic amino acid sites of the FPV. Based on NCBI BLAST analysis, it was found that all sequences from the golden jackals and domestic cats exhibited 99-100% similarity to the available FPV sequences.
## 3.2. Phylogenetic Analysis
Phylogenetic analysis of the VP2 gene identified seven distinct FPV groups with bootstrap values ranging from 70.00 to 91.75% (Figure 2). CPV-2 strains formed a separate group, with no isolates from this study present in it. Most Croatian isolates were closely related and belonged to the FPV-5, FPV-6, and FPV-7 groups. The FPV-5 group, supported by a bootstrap value over 90, consisted exclusively of Croatian isolates obtained from both domestic cats (V_21_25, V_149_24, V_49_25, V_61_24, FV_12_24) and golden jackals (C_1_25). The FPV-6 group consisted of Croatian domestic cat isolates (FV_2_24, FV_9_24, FV_15_24, FV_24_24) and Croatian golden jackal isolates (C_12_25, C_19_25, C_39_25), which were clustered together with Italian strains obtained from domestic cats. The FPV-7 group contained one Croatian domestic cat isolate (V_150_24), which was most closely related to the domestic cat FPV strains from Italy and Belgium.
In addition to these main clusters, one Croatian domestic cat isolate (FV_17_24) was clustered within the FPV-1 group, along with three Croatian golden jackal isolates (C_4_24, C_18_25, C_36_25). https://doi.org/10.3390/v18010123
## 3.3. Amino Acid Mutations in the VP2 Gene
Croatian golden jackal FPV isolates exhibit several amino acid mutations compared to the FPV CU-4 reference strain. These mutations include A5T, A91T, I101T, and E411Q, as shown in Table 2. Of particular interest, two golden jackal isolates (C_12_25 and C_39_25) carry the A91T mutation, previously reported in FPV strains in golden jackals but not further studied [15,16]. None of the domestic cat isolates carry this mutation; however, one isolate harbours the A91S mutation, which has previously been identified in FPV strains from China [26,32]. The E411Q mutation, also not previously reported, was identified in seven isolates (39%), including both golden jackals (C_12_25, C_19_25, C_39_35) and domestic cats (FV_2_24, FV_9_24, FV_15_24, FV_29_24). The I101T mutation, common in many FPV strains [33], was present in all Croatian isolates, including those from golden jackals.
Comparison of Croatian isolates with partial golden jackal VP2 sequences from NCBI showed that one Italian and one Serbian jackal share the A91T mutation, but none carry the E411Q mutation (Table 2).
## 3.4. Modelling of the VP2 Structure
Structural VP2 visualisation using PyMOL version 3.1.6.1 revealed the locations of amino acid substitutions at positions 91, 101, and 411 in the three-dimensional structure of VP2 (Figure 3). Residues 91 and 411 are surface-exposed and situated near the threefold spike, in a region referred to as the shoulder [3], whereas residue 101 is buried beneath the surface. Residues 91 and 101 are located within loop one, while residue 411 is positioned within loop four.
## 4. Discussion
In the 20th century, the golden jackal population significantly declined in the Balkan Peninsula due to widespread poisoning campaigns targeting wolves, foxes, and other pests, as well as excessive hunting and habitat loss [34]. By the late 1990s, the population in eastern Croatia began to recover, likely moving in from Bulgaria along the Danube River valley and eventually extending into the Sava and Drava valleys. Jackals primarily inhabit lowland, marshy areas overgrown with shrubs, and the war left many of these areas undisturbed, allowing them to raise offspring successfully. As the population increases, jackals during sexual maturation and during juvenile roaming, spreading into less densely populated areas, often inhabiting the edges of human settlements and living alongside people [20]. Despite this, they are frequently overlooked in the epidemiology of infectious diseases, as evidenced by the lack of information on Protoparvovirus carnivoran1 infection in jackals.
To this date, the only published reports originated from northeastern Italy in 2022 [16] and from Serbia in 2023 [15]. In both countries and in this study, jackals were infected with FPV, unlike other canid hosts, which are usually infected with CPV-2. FPV cannot efficiently bind to the canine transferrin receptor (TfR), which the virus uses to enter host cells; consequently, FPV cannot infect most canids. A critical factor that facilitated the host barrier jump of the FPV-like virus, leading to the emergence of CPV-2, was its ability to bind canine TfR [9]. Interestingly, jackals seem to retain ancestral features of the transferrin receptor. Their receptors probably represent an intermediate stage between feline-like and canine-like receptors, potentially explaining their susceptibility to FPV [35,36].
Our study assessed the presence of Protoparvovirus carnivoran1 infection in jackal populations along the Danube and Drava (Osijek-Baranja County) and the Sava (Sisak-Moslavina County and Zagreb County) in Croatia. The high overall positivity rate of FPV infection (40%) observed in the Croatian golden jackal population is partially consistent with findings from Italy, which reported a positivity rate of 25%. In contrast, research from Serbia indicated a considerably lower positivity rate of only 4.4%. This discrepancy is somewhat surprising, especially given that genetic testing suggests the Croatian continental jackal population and the Serbian population are, in fact, the same [20]. The authors argued that the low positivity rate in their study may be due to the jackals' advanced age. Since our research did not include age data for the tested animals, this reasoning cannot be further investigated. A more likely explanation for the observed difference is that they used spleen as the sample material, where viral DNA is detected less frequently than in the intestine [37].
The phylogenetic analysis identified seven FPV groups. Croatian golden jackals belonged to the FPV-1 and were closely related to the FPV-5 and FPV-6 groups. All groups included strains from Croatian domestic cats and golden jackals. This suggests that at least some of the FPV-positive jackals may have acquired the virus through direct contact with domestic cats or a shared source of infection, such as a contaminated environment or other susceptible hosts. This is further supported by amino acid analysis of isolates in this study, which revealed that Croatian domestic cats and golden jackals share the previously unreported E411Q and I101T mutations. The E411Q mutation appears to be unique to Croatian strains, potentially indicating an independent evolutionary event among the local population. This assumption is further supported by the presence of the FPV-5 group, which includes only Croatian isolates. Three Croatian jackals belonged to the phylogenetically distant FPV-1 group, along with one Croatian domestic cat. The cat exhibited an A91S mutation previously described as the predominant strain in China [26,32]. Since this mutation has not been reported outside China, it appears unlikely that this case was imported from there. There is a possibility that FPV strains share a common mutation pattern, leading to sequences from various continents converging. Regardless, our research indicates that the viral population currently circulating in Croatia is not uniform and that multiple introductions of the virus may have occurred. A high diversity of FPV strains in jackals has previously been reported in both Italy and Serbia. Research from Italy attributed this diversity to cross-species transmission or to different origins of jackals repopulating the area. Although Croatia's continental region was repopulated in the 1990s after jackals were brought to the brink of extinction, the population shows low genetic diversity [38]. As a result, the latter explanation is less likely for FPV diversity in jackals.
The data from this study support the idea of cross-species transmission facilitated by host adaptation. In line with this, the previously undescribed A91T substitution, found exclusively in jackals, may reflect an evolutionary process unique to this host, as the structural regions surrounding residues 93, 300, and 323 are known to contribute to host range determination [39][40][41][42][43]. These residues are part of surface-exposed loop regions that influence differences in antigenicity, host range, and tissue tropism. The A91T mutation in Loop 1, which, with Loops 2 and 4, forms the spike of the threefold capsid axis and contributes to the formation of the antigen A site, one of two major antigenic sites of FPV [41,44].
The E411Q mutation, which had not previously been reported, was identified in both golden jackals and domestic cats. E411Q is located in Loop 4. As antigen A serves both as the site for host receptor binding and as a target for virus neutralisation, mutations at this position are likely to promote host adaptation and help the virus evade the immune response. However, experimental studies are required to determine the biological significance of these changes.
Although the limited number of analysed sequences precludes more definitive conclusions, our results demonstrate considerable genetic diversity among Croatian FPV strains. The high FPV positivity rate, together with the unique features of their transferrin receptor, indicates that golden jackals are highly susceptible to FPV and may serve both as a reservoir and a bridging host, facilitating virus maintenance and transmission between wildlife and domestic animals.
## 5. Conclusions
Our study, for the first time, confirmed the presence of FPV in the Croatian golden jackal population, with a high positivity rate among the tested animals. The data reveal a certain limited degree of heterogeneity among the FPV strains circulating in Croatia, including a distinctive Croatian strain. The detection of the A91T amino acid mutation unique to jackals, along with mutations shared by both domestic cats and jackals, highlights the importance of wildlife research for the understanding of the epidemiology of parvoviruses.
## 6. Research Limitations
The main limitation of the study was the small number of golden jackal sequences, which restricted the ability to draw more definitive conclusions about the significance of the observed data. Data on sex, age and pathological findings would allow further understanding of Protoparvovirus carnivoran1 infection epidemiology in golden jackals and its clinical relevance. Additionally, obtaining complete viral genomes would allow a more detailed characterisation of the detected viruses. Finally, further experimental studies are needed to fully understand the importance of the mutations observed in this research.
## References
1. Adams, Lefkowitz, King et al. (2014) "Ratification vote on taxonomic proposals to the International Committee on Taxonomy of Viruses" *Arch. Virol*
2. Cotmore, Agbandje-Mckenna, Chiorini et al. (2014) *The family Parvoviridae. Arch. Virol*
3. Muzyczka, Berns (2001) "Parvoviridae: The Viruses and Their Replication"
4. Verge, Christoforoni (1928) "La gastro-entérite infectieuse des chats; est-elle due à un virus filtrable?" *CR Seances Soc. Biol. Fil*
5. Barrs (2019) "Feline panleukopenia: A re-emergent disease" *Vet. Clin. Small Anim. Pract*
6. Truyen, Parrish (1995) "The evolution and control of parvovirus host ranges" *Semin. Virol*
7. Truyen, Evermann, Vieler et al. (1996) "Evolution of canine parvovirus involved loss and gain of Feline Host Range" *Virology*
8. Hoelzer, Parrish (2010) "The emergence of parvoviruses of carnivores" *Vet. Res*
9. Hueffer, Parker, Weichert et al. (2003) "The natural host range shift and subsequent evolution of canine parvovirus resulted from virus-specific binding to the canine transferrin receptor" *J. Virol*
10. (2026) *Viruses*
11. Parrish (1990) "Emergence, natural history and variation of canine, mink, and feline parvoviruses" *Adv. Virus Res*
12. Li, Peng, Zeng et al. "Isolation of a feline-derived feline panleukopenia virus with an A300P substitution in the VP2 protein and confirmation of its pathogenicity in dogs"
13. Yi, Liu, Meng et al. "Feline Panleukopenia Virus With G299E Substitution in the VP2 Protein First Identified From a Captive Giant Panda in China" *Front. Cell. Infect. Microbiol*
14. Diakoudi, Lanave, Berjaoui et al. (2024) "Feline Panleukopenia Virus in a Marsican Brown Bear and Crested Porcupine" *Emerg. Infect. Dis*
15. Canuti, Britton, Graham et al. (2017) "Epidemiology and molecular characterization of protoparvoviruses infecting wild raccoons (Procyon lotor) in British Columbia" *Canada. Virus Res*
16. Milićević, Glišić, Veljović et al. (2024) "Protoparvovirus carnivoran 1 infection of golden jackals Canis aureus in Serbia" *Vet. Res. Commun*
17. Leopardi, Milani, Cocchi et al. (2612) "Carnivore protoparvovirus 1 (CPV-2 and FPV) circulating in wild carnivores and in puppies illegally imported into north-eastern Italy" *Viruses*
18. Bošković, Šperanda, Florijančić et al. (2013) "Dietary habits of the golden jackal (Canis aureus L.) in the eastern Croatia" *Agric. Conspect. Sci*
19. Penezić, Ćirović (2015) "Seasonal variation in diet of the golden jackal (Canis aureus) in Serbia" *Mamm. Res*
20. Rotem, Berger, King et al. (2011) "The effect of anthropogenic resources on the space-use patterns of golden jackals" *J. Wildl. Manag*
21. Bošković, Florijančić, Ozimec et al. (2025) "Stručna Podloga za Utvr divanje Osnovnih Odrednica Obitavanja, Statusa i Smjernica Gospodarenja Čagljem (Canis aureus L.) u Republici Hrvatskoj"
22. Kryštufek, Murariou, Kurtonur (1997) "Present distribution of the Golden Jackal Canis aureus in Balkans and adjacent regions" *Mammal Rev*
23. Decaro, Desario, Lucente et al. (2008) "Specific identification of feline panleukopenia virus and its rapid differentiation from canine parvoviruses using minor groove binder probes" *J. Virol. Methods*
24. Horiuchi, Yamaguchi, Gojobori et al. (1998) "Differences in the evolutionary pattern of Feline Panleukopenia virus and canine parvovirus" *Virology*
25. Tamura, Stecher, Kumar (2021) "MEGA11: Molecular Evolutionary Genetics Analysis version 11" *Mol. Biol. Evol*
26. Altschul, Gish, Miller et al. (1990) "Basic local alignment search tool" *J. Mol. Biol*
27. Wang, Yan, Liu et al. (2021) "Prevalence and molecular evolution of parvovirus in cats in eastern Shandong"
28. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Mol. Biol. Evol*
29. Tamura (1992) "Estimation of the number of nucleotide substitutions when there are strong transition-transversion and G+C-content biases" *Mol. Biol. Evol*
30. Letunic, Bork (2021) "Interactive Tree Of Life (iTOL) v7: An online tool for the display and annotation of phylogenetic and other trees" *Nucleic Acids Res*
31. (2025) "The PyMOL Molecular Graphics System, Version 3.1.6. Schrödinger: New York"
32. Agbandje, Mckenna, Rossmann et al. (1993) "Structure determination of feline panleukopenia virus empty particles" *Proteins*
33. Chen, Wang, Zhou et al. (2022) "Circulation of heterogeneous Carnivore protoparvovirus 1 in diarrheal cats and prevalence of an A91S feline panleukopenia virus variant in China" *Transbound. Emerg. Dis*
34. (2026) *Viruses*
35. Zhou, Zeng, Zhang et al. (2017) "The genetic evolution of canine parvovirus-A new perspective" *PLoS ONE*
36. Spassov (1989) "The position of Jackals in the Canis genus and life-history of the Golden Jackal (Canis aureus L.) in Bulgaria and on the Balkans" *Hist. Nat. Bulg*
37. Palermo, Hafenstein, Parrish (2006) "Purified feline and canine transferrin receptors reveal complex interactions with the capsids of canine and feline parvoviruses that correspond to their host ranges" *J. Virol*
38. Kaelber, Demogines, Harbison et al. (2012) "Evolutionary reconstructions of the transferrin receptor of Caniforms supports canine parvovirus being a re-emerged and not a novel pathogen in dogs" *PLoS Pathog*
39. Ndiana, Lanave, Desario et al. (2021) "Circulation of diverse protoparvoviruses in wild carnivores" *Italy. Transbound. Emerg. Dis*
40. Fabbri, Caniglia, Galov et al. (2014) "Genetic structure and expansion of golden jackals (Canis aureus) in the north-western distribution range (Croatia and eastern Italian Alps)" *Conserv. Genet*
41. Chang, Sgro, Parrish (1992) "Multiple amino acids in the capsid structure of canine parvovirus coordinately determine the canine host range and specific antigenic and hemagglutination properties" *J. Virol*
42. Horiuchi, Goto, Ishiguro et al. (1994) "Mapping of determinants of the host range for canine cells in the genome of canine parvovirus using canine parvovirus/mink enteritis virus chimeric viruses" *J. Gen. Virol*
43. Strassheim, Gruenberg, Veijalainen et al. (1994) "Two dominant neutralizing antigenic determinants of canine parvovirus are found on the threefold spike of the virus capsid" *Virology*
44. Llamas-Saiz, Agbandje-Mckenna, Parker et al. (1996) "Structural analysis of a mutation in canine parvovirus which controls antigenicity and host range" *Virology*
45. Parker, Parrish (1997) "Canine parvovirus host range is determined by the specific conformation of an additional region of the capsid" *J. Virol*
46. Parrish, Carmichael (1983) "Antigenic structure and variation of canine parvovirus type-2, feline panleukopenia virus, and mink enteritis virus" *Virology*
47. "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=PMC12837457&blobtype=pdf | # Comprehensive Identification and Male-Biased Expression Analysis of Odorant-Binding Protein Genes in the Hawaiian Flower Thrips, Thrips hawaiiensis (Thysanoptera: Thripidae)
Natraj Krishnan, Ying Lin, Qingqing Fan, Yanjun Li, Xiaodi Hu
## Abstract
The Hawaiian flower thrips, Thrips hawaiiensis, is a widespread pest that infests the flowers of numerous horticultural crops. We characterized the previously unknown olfactory system of Thrips hawaiiensis by analyzing its odorant-binding proteins (OBPs). Our genome survey revealed 12 OBP genes, a count comparable to other thrips but low relative to most insects. Subsequent transcriptomic and RT-qPCR analyses identified consistent male-biased expression, implicating these OBPs in male-specific olfactory behaviors, such as mate location and foraging. We further identified 11 CSPs, the majority of which showed a male-biased expression pattern similar to that of the OBPs.
## 1. Introduction
Thrips hawaiiensis (Morgan) (Thysanoptera: Thripidae), a common flower-dwelling thrips species native to the Oriental and Pacific regions, has expanded its geographical occidentalis [30] were retrieved from GenBank (National Center for Biotechnology Information, NCBI) and used as query sequences for a BLASTP (v2.5.0) search against the thrips amino acid database, applying an identity cutoff of 30%. Then, gene annotation was performed using hmmscan (v3.1b1) with the Pfam-A.hmm database, applying an E-value threshold of 1 × 10 -5 to identify putative OBP genes (PF01395.18). To improve accuracy, the hmmscan results were combined with BLAST-based annotations by retaining only overlapping gene predictions, yielding a high-confidence gene set for downstream analyses. The putative N-terminal signal peptides of T. hawaiiensis OBPs were identified using the SignalP 6.0 online server (https://services.healthtech.dtu.dk/services/SignalP-6.0/, accessed on 28 October 2025) [42]. The OBP sequences were aligned using MUSCLE (v5.1) [43] software. The resultant alignment file was then imported into Jalview software (v2.11.5.0) [44] for visualization.
## 2.2. Characteristics Analysis of OBPs
The OBP sequences of thrips were aligned using MUSCLE (v5.1) [43]. Following alignment, a maximum likelihood (ML) phylogenetic tree was constructed on the aligned dataset with IQ-TREE (v2.2.5) [45], employing 1000 ultrafast bootstrap replicates (-B 1000). The chromosomal locations of OBP genes were retrieved from the T. hawaiiensis GFF annotation file using a custom Python (v3.10.13) script. They were subsequently visualized on the chromosomes via the online tool MapGene2Chrom (http://mg2c.iask.in/mg2c_v2.1 /, accessed on 28 October 2025) [46,47]. The exon-intron structures of the ThawOBP genes were determined through an in-house Python script.
## 2.3. Insect Sample Collection
A laboratory population of Thrips hawaiiensis was established from individuals collected on kidney beans (Phaseolus vulgaris) at the Institute of Plant Protection, Fujian Academy of Agricultural Sciences, China (119 • 34 ′ E, 26 • 13 ′ N), and has been continuously reared on this host since 2016. The kidney bean diet was prepared by soaking in water, coating with a 10% honey solution, and air-drying. Thrips were maintained in MGC-350HP artificial climate incubators (Yiheng Scientific Instruments, Shanghai, China) under the following conditions: 27 ± 1 • C, 60 ± 5% RH, and a 16:8 (L:D) photoperiod.
## 2.4. RNA Extraction and Real-Time Quantitative PCR Analysis of ThawOBPs
A total of 300 adult male and female T. hawaiiensis samples were selected for analysis of relative mRNA expression levels. Total RNA was extracted from the samples using the Trizol method. Subsequently, the RNA was reverse transcribed into cDNA using a reverse transcription kit (Accurate Biology, Changsha, Hunan, China).
A primer design tool (https://www.ncbi.nlm.nih.gov/tools/primer-blast/, accessed on 28 October 2025) was used to design primers for the ThawOBP gene of Thrips hawaiiensis, with β-actin as the internal reference gene [48]. RT-qPCR was performed on an ABI QuantStudio 5 system (Thermo Fisher Scientific, Waltham, MA, USA), with the entire process conducted on ice. Each experiment included at least two technical replicates and three biological replicates. The expression levels in male and female Thrips hawaiiensis were calculated using the 2 -∆CT method. Data analysis and visualization were conducted using GraphPad Prism (v9.5.0) and R (v4.2.0) for correlation calculations. All the primers used in this study are listed in Table 1.
## 2.5. Transcriptome Sequencing
To obtain a comprehensive overview of gene expression, total RNA was isolated from adult male and female subjects. Criteria for cDNA library qualification are as follows: the AD260/280 absorbance ratio should be between 1.8 and 2.0, the A260/230 ratio should be between 1.9 and 2.4, and the concentration measured by Qubit should be between 0.95 and 3.0. Each biological replicate comprised a pool of approximately 800 individuals. RNA extraction was performed using TRIzol Reagent (Thermo Fisher Scientific, USA). Subsequently, Illumina paired-end libraries were constructed with the TruSeq RNA Library Preparation Kit (Illumina, San Diego, CA, USA) according to the manufacturer's instructions and sequenced on an Illumina NovaSeq 6000 platform. This generated approximately 14.3 GB of high-quality 150 bp paired-end sequence data.
## 2.6. Phylogenetic Analysis of OBP Genes
A maximum likelihood phylogeny was reconstructed using IQ-TREE (v2.3.3) [45]. To determine the most appropriate substitution model, we used the built-in ModelFinder tool, which selected the optimal model from a candidate set based on the Bayesian Information Criterion. The analysis was subsequently conducted under the best-fit model, WAG + R4. Branch support was assessed using ultrafast bootstrap approximation, with 1000 replicates.
## 3. Results
## 3.1. Identification and Sequence Analysis of OBPs in T. hawaiiensis
The genome of T. hawaiiensis was assembled using a hybrid approach that integrated the technologies of Oxford Nanopore long-read sequencing, Illumina short-read sequencing, and Hi-C chromatin conformation capture. This strategy produced a final assembly of 287.59 Mb, with a scaffold N50 of 13.84 Mb. According to BUSCO analysis, the assembled genome exhibits a high completeness of 98.7% [31]. As shown in Table 2, a total of 12 ThawOBP genes were identified through our BLASTP analysis. The genes, named ThawOBP1-ThawOBP12 based on their chromosomal locations (Figure 1), each possessed a complete open reading frame (ORF). The putative proteins range from 133 (ThawOBP12) to 233 (ThawOBP3) amino acids, while most are approximately 150 amino acids long. The analysis classified these genes into two subfamilies: the Classic subfamily (10 genes) and the Minus-C subfamily (two genes, namely ThawOBP3 and ThawOBP8) (Figure 2). All encoded proteins possess a putative N-terminal signal peptide, with the cleavage site predicted between amino acids 17 and 28. These putative ThawOBPs share from 31.4% (ThawOBP2) to 86.8% (ThawOBP5) sequence identity with their closest matches in the database, supported by highly significant E-values ranging from 8.28 × 10 -128 to 3.67 × 10 -8 (Table 2). Chromosomal mapping of the 12 identified ThawOBP genes across major genomic scaffolds was conducted. The genes are unevenly distributed, with sup_sca_14 harboring a significant number of ThawOBP loci. Several additional scaffolds also contain ThawOBP genes (Figure 1), suggesting a dispersed genomic organization of the odorant-binding protein family in this species. Investigation of the phylogenetic relationships and exon-intron structures of OBP genes in T. hawaiiensis unveiled considerable diversity within this gene family. A phylogenetic tree was constructed from full-length OBP sequences. The tree reveals the evolutionary relationships among the 12 ThawOBP genes, which are categorized into three distinct clades (Figure 3A). ThawOBP11 and ThawOBP12 form a closely related pair with high bootstrap support, suggesting a strong functional constraint. Similarly, ThawOBP4 and ThawOBP10 cluster together, indicating a close evolutionary relationship. ThawOBP1 and ThawOBP7 are positioned more distantly from the other groups, with ThawOBP1 and ThawOBP7 appearing as two of the more divergent sequences in the dataset (Figure 3A). The number of exons in these OBP genes varied between five and nine, with the majority containing six or seven (Figure 3B). Our analysis of the ThawOBP family showed that the average exon length was 69.97 bp. Furthermore, we found that the exon lengths across all 12 genes were relatively constrained, ranging from 57.9 to 82.8 bp (Figure 4A). Among the 12 ThawOBP genes, ThawOBP10 possessed the shortest average exon length (57.9 bp) and contained 7 exons. In contrast, ThawOBP3 had the highest exon count (9) and an average exon length of 78.0 bp. Furthermore, ThawOBP4, ThawOBP11, and ThawOBP12 shared an identical exon number of 6 and exhibited similar average exon lengths (Figure 4B). These structural features are entirely consistent with the clustering pattern observed in Figure 2.
## 3.3. Phylogenetic Relationship Analysis of All OBPs
Using the same identification methodology, we conducted analyses in species including Dendrothrips minowai [34], Frankliniella occidentalis [36], Frankliniella intonsa [40], Frankliniella fusca [35], Odontothrips loti [37], Thrips palmi [38], Stenchaetothrips biformis [41], Thrips tabaci [39], Megalurothrips usitatus (Hainan) [32], Megalurothrips usitatus (Zhejiang) [33], and Acyrthosiphon pisum [49]. The results revealed that the number of OBPs identified ranges from 10 to 17 across these species (Figure 5A). To assess the phylogenetic relevance between Thrips hawaiiensis OBPs and other OBPs, all OBPs were aligned to generate unrooted trees. As shown in Figure 5B, the OBPs from the same species (Zhejiang and Hainan populations) were tightly clustered, whereas those from other thrips species were interspersed, a pattern likely attributable to their close phylogenetic relationship within the same family, Thripidae. https://doi.org/10.3390/biology15020170
## 3.4. Expression Patterns of T. hawaiiensis OBP Genes
The transcriptomic analysis was conducted using a comprehensive reference-based pipeline. Raw sequencing reads were quality-trimmed and adapter-removed using Trimmomatic (v0.39) [50] to ensure data quality for downstream analyses. Subsequently, the processed reads were aligned to the reference genome using HISAT2 (v2.2.1) [51], a splice-aware aligner optimized for RNA-seq data. Finally, gene-level read counts were quantified from the aligned reads using HTSeq-count (v2.0.2) [52] with default union-counting mode, generating a count matrix for subsequent differential expression analysis. To validate the gene expression profiles obtained from transcriptome sequencing, we selected 12 OBP genes in Thrips hawaiiensis for confirmation by RT-qPCR. The results demonstrated that 10 of these genes exhibited malebiased expression (Figure 6A), consistent with transcriptomic predictions. A strong correlation (Pearson correlation coefficient r = 0.83) was observed between the RNA-Seq and RT-qPCR results, validating the reliability of the transcriptomic data (Figure 6B). The male-biased expression patterns were consistent across both methods for most genes, except for ThawOBP11 and ThawOBP12. For these two genes, a discrepancy was noted: they showed minimal expression bias by RT-qPCR (Log2FC ≈ 0 or negative) but were indicated as low-level male-biased by RNA-Seq, a divergence potentially due to their low expression levels or technical limitations. To investigate whether other olfaction-related gene families follow similar patterns, we also identified genes encoding chemosensory proteins (CSPs) (PF03392.9) using the same pipeline. A total of 11 CSP genes were identified in the T. hawaiiensis genome. The expression patterns of the 11 identified CSP genes were analyzed based on transcriptome sequencing and validated by RT-PCR. Transcriptomic data revealed that eight genes were significantly upregulated in males compared to females of T. hawaiiensis (Figure S1A). The RT-qPCR validation confirmed the accuracy of the transcriptomic expression profiles (Figure S1B). All primers used in this study are listed in Table S1. The results demonstrate that over half of the genes show significantly higher expression in males, consistent with the expression pattern observed for OBP genes.
## 4. Discussion
The identification of olfactory genes is fundamental to elucidating the molecular mechanisms of olfaction. In many insect species, odorant-binding proteins (OBPs) have been identified through transcriptomic and genomic analyses. Notably, studies have reported 51 OBP genes in Drosophila melanogaster [53], 65 in Anopheles gambiae [54], 64 in Aedes aegypti [54], 53 in Culex quinquefasciatus [54], 44 in Bombyx mori [55], and 50 in Tribolium castaneum [56]. In the present study, by analyzing our previously published genome data, we expanded the repertoire of OBPs in the insect species Thrips hawaiiensis to 12. While the 12 OBPs in T. hawaiiensis are substantially fewer than those in model insects such as D. melanogaster, this number is comparable to other thrips species, such as M. usitatus (14,17), F. occidentalis (10), T. palmi (15), F. intonsa (15), and O. loti (13) (Figure 5A). The number of OBP genes identified in other insects is significantly larger than that in thrips. This comparative reduction in thrips could be due to a simplified olfactory system or a unique evolutionary trajectory. Furthermore, phylogenetic analysis revealed that ThawOBP11 and ThawOBP12 cluster into a clade and exhibit the highest degree of sequence similarity (Figure 3A), a finding supported by sequence alignment (Figure 2).
For adult thrips, the most critical behaviors are host plant seeking, mating, and reproduction. These behaviors involve the extensive detection of both plant volatiles and thrips pheromones [57,58]. Male adults of both thrips species synthesize an aggregation pheromone detectable by both sexes that mediates attraction [59,60]. Consequently, these genes likely play a role in mediating the detection of both plant volatiles and the malederived aggregation pheromone in Thrips hawaiiensis. Extensive research has demonstrated that odorant exposure can alter the expression levels of OBPs [61][62][63]. Through in vitro and in vivo functional assays in Bactrocera dorsalis, OBP83g-2 was identified as a key OBP involved in γ-octalactone perception, which was further confirmed to play a significant role in γ-octalactone-mediated oviposition behavior [64]. In Hyphantria cunea, OBP2 plays a crucial role in guiding larvae toward food sources that contain adult sex pheromones [65]. While the knockdown of SfruOBP18 did not impair larval survival or development, our combined RNAi and bioassay approach uncovered its critical function in conferring tolerance to multiple insecticides. This suggests a novel and non-canonical role for SfruOBP18 in insecticide susceptibility in Spodoptera frugiperda [66]. In our study, gene expression pattern analysis shows strong male-specific expression of OBP and CSP genes in T. hawaiiensis (Figure 6A and Figure S1). This pattern, also reported for OBPs and CSPs in M. usitatus, Frankliniella occidentalis, and Frankliniella intonsa [26,30], further supports their functional role in male-driven behaviors such as mate location and foraging. In experiments with Rhynchophorus ferrugineus, females injected with OBP-dsRNA showed a significant decrease in the expression of both RferOBP3 and RferOBP1768, which led to impaired perception of the odorants trans-2,4-nonadienal and trans-2-nonenal [67]. Docking results suggested a role for FoccOBP4/FintOBP4, FoccOBP6/FintOBP6, and FoccCSP2/FintCSP2 in trans-porting the major pheromone neryl (S)-2-methylbutanoate, with FoccOBP6/FintOBP6, Foc-cCSP2/FintCSP2, and FoccCSP3/FintCSP3 also implicated in binding the minor component (R)-lavandulyl acetate [30]. Based on these findings, we hypothesize that the majority of OBP and CSP genes in T. hawaiiensis are expressed at higher levels in males than in females, suggesting that males may possess a heightened sensitivity to environmental odors.
## 5. Conclusions
In summary, we report the first genome-wide identification of odorant-binding protein (OBP) genes and chemosensory protein (CSP) genes in Thrips hawaiiensis and characterize their expression profiles in both female and male T. hawaiiensis. This work provides crucial data for elucidating the functions of olfactory proteins in T. hawaiiensis, thereby paving the way for novel, targeted management strategies against thrips pests.
## References
1. Murai (2001) "Development and reproductive capacity of Thrips hawaiiensis (Thysanoptera: Thripidae) and its potential as a major pest" *Bull. Entomol. Res*
2. Reynaud, Balmès, Pizzol (2008) "Thrips hawaiiensis (Morgan, 1913) (Thysanoptera: Thripidae), an Asian pest thrips now established in Europe"
3. Aliakbarpour, Che Salmah (2011) "Seasonal abundance and spatial distribution of larval and adult thrips (Thysanoptera) on weed host plants in mango orchards in Penang" *Malaysia. Appl. Entomol. Zool*
4. Goldarazena (2011) "First record of Thrips hawaiiensis (Morgan, 1913) (Thysanoptera: Thripidae), an Asian pest thrips in Spain"
5. Mound, Thysanoptera (2005) "Diversity and interactions" *Annu. Rev. Entomol*
6. Fu, Qiu, Li et al. (2020) "Flower injection of imidacloprid and spirotetramat: A novel tool for the management of banana thrips Thrips hawaiiensis" *J. Pest Sci*
7. Zhang, Fu, Liu et al. (2014) "The effect of temperature on the development and survival of Thrips hawaiiensis (Morgan)" *Acta Ecol. Sin*
8. (2026) *Biology*
9. Kurosawa, Takaoka, Naito (1964) "Thrips infesting the tobacco plants in Japan (Thysanoptera)" *Jpn. J. Entomol*
10. Woo, Paik (1971) "Studies on the thrips (Thysanoptera) unrecorded in Korea (I)" *Korean J. Plant Prot*
11. Chen, Lo (1987) "Differential preference of the flower-dwelling thrip, Thrips hawaiiensis (Morgan) (Thysanoptera: Thripidae), to some gladiolus cultivars" *J. Agric. Res. China*
12. Chandra, Lal (1973) "Record of thrips on some vegetables and ornamental plants from Kulu Valley, Himachal Pradesh" *Himachal Pradesh Indian J. Entomol*
13. Ketavan (1978) "Thrips of Economic Importance and Their Control in Thailand"
14. Chen, Gao, Han et al. (2018) "Species composition and distribution of thrips (Thysanoptera) in mango orchards of China" *Anim. Biol*
15. Chiu, Shen, Wu (1991) "Occurrence and damage of thrips in Citrus orchards in southern Taiwan" *Chin. J. Entomol*
16. Tsai, Hwang, Chen (1992) "Occurrence and damage of Thrips hawaiiensis in banana orchards" *Chin. J. Entomol*
17. Jhala, Borad, Bharpoda (2004) "Incidence of Thrips hawaiiensis (Morgan) on banana in Gujarat"
18. Marullo, De Grazia (2017) "Thrips hawaiiensis a pest thrips from Asia newly introduced into Italy" *Bull. Insectology*
19. Andersson, Löfstedt, Newcomb (2015) "Insect olfaction and the evolution of receptor tuning" *Front. Ecol. Evol*
20. Zhou (2010) "Odorant-binding proteins in insects" *Vitam. Horm*
21. Rützler, Zwiebel (2005) "Molecular biology of insect olfaction: Recent progress and conceptual models" *J. Comp. Physiol. A*
22. Zhu, Guo, Ban et al. (2018) "Niemann-Pick C2 proteins: A new function for an old family" *Front. Physiol*
23. Sun, Wang, Wang et al. (2017) "Identification and characterization of odorant binding proteins in the forelegs of Adelphocoris lineolatus (Goeze)" *Front. Physiol*
24. Brito, Moreira, Melo (2016) "A look inside odorant-binding proteins in insect chemoreception" *J. Insect Physiol*
25. Tang, Tai, Dai et al. (2019) "Expression and functional analysis of two odorant-binding proteins from Bradysia odoriphaga (Diptera: Sciaridae)" *J. Agric. Food Chem*
26. Pelosi, Maida (1995) "Odorant-binding proteins in insects" *Comp. Biochem. Physiol. Part B Biochem. Mol. Biol*
27. Xia, Yang, Li et al. (2025) "Genome-Wide Identification and Expression Profiling of Odorant-Binding Protein Genes in the Bean Flower Thrips Megalurothrips usitatus (Bagnall) (Thysanoptera: Thripidae)" *Insects*
28. Xu, Zwiebel, Smith (2003) "Identification of a distinct family of genes encoding atypical odorant-binding proteins in the malaria vector mosquito, Anopheles gambiae" *Insect Mol. Biol*
29. Zhou, He, Pickett et al. (2008) "Identification of odorant-binding proteins of the yellow fever mosquito Aedes aegypti: Genome annotation and comparative analyses" *Insect Mol. Biol*
30. Vogt, Riddiford (1981) "Pheromone binding and inactivation by moth antennae" *Nature*
31. Li, Cheng, Chen et al. (2022) "Comparison and functional analysis of odorant-binding proteins and chemosensory proteins in two closely related thrips species, Frankliniella occidentalis and Frankliniella intonsa (Thysanoptera: Thripidae) based on antennal transcriptome analysis" *Int. J. Mol. Sci*
32. Hu, Tian, Chen et al. "Genome assembly of Hawaiian flower thrips Thrips hawaiiensis (Thysanoptera: Thripidae). Sci. Data 2025"
33. Ma, Liu, Wei et al. (2023) "Chromosome-level genome assembly of bean flower thrips Megalurothrips usitatus (Thysanoptera: Thripidae)" *Sci. Data*
34. Zhang, Bao, Chen et al. (2023) "The Chromosome-Level Genome Assembly of Bean Blossom Thrips (Megalurothrips usitatus) Reveals an Expansion of Protein Digestion-Related Genes in Adaption to High-Protein Host Plants" *Int. J. Mol. Sci*
35. Xiu, Zhao, Zhang et al. (2025) "Chromosome-level genome assembly of Dendrothrips minowai and genomic analysis highlights distinct adaptations to high polyphenols in tea plants" *Pest Manag. Sci*
36. Catto, Labadie, Jacobson et al. (2023) "Pest status, molecular evolution, and epigenetic factors derived from the genome assembly of Frankliniella fusca, a thysanopteran phytovirus vector" *BMC Genom*
37. (2026) *Biology*
38. Song, Cao, Chen et al. (2024) "Chromosome-level genome assembly of the western flower thrips Frankliniella occidentalis" *Sci. Data*
39. Luo, Wei, Dai et al. (2024) "Chromosome-level genome assembly of Odontothrips loti Haliday (Thysanoptera: Thripidae)" *Sci. Data*
40. Guo, Cao, Song et al. (2020) "Chromosome-level assembly of the melon thrips genome yields insights into evolution of a sap-sucking lifestyle and pesticide resistance" *Mol. Ecol. Resour*
41. Gao, Ji, Xu et al. (1003) "Chromosome-level genome assembly of cotton thrips Thrips tabaci (Thysanoptera: Thripidae)" *Sci. Data*
42. Song, Wang, Cao et al. (2024) "A chromosome-level genome for the flower thrips Frankliniella intonsa" *Sci. Data*
43. Hu, Ye, Zhuo et al. (2023) "A chromosome-level genome assembly of Stenchaetothrips biformis and comparative genomic analysis highlights distinct host adaptations among thrips" *Commun. Biol*
44. Almagro Armenteros, Tsirigos, Sønderby et al. (2019) "SignalP 5.0 improves signal peptide predictions using deep neural networks" *Nat. Biotechnol*
45. Edgar, Muscle (2004) "Multiple sequence alignment with high accuracy and high throughput" *Nucleic Acids Res*
46. Waterhouse, Procter, Martin et al. (2009) "Jalview Version 2-A multiple sequence alignment editor and analysis workbench" *Bioinformatics*
47. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: A fast and effective stochastic algorithm for estimating maximum-likelihood phylogenies" *Mol. Biol. Evol*
48. Chao, Kong, Wang et al. (2015) "MapGene2Chrom, a tool to draw gene physical map based on Perl and SVG languages" *Hereditas*
49. Chao, Li, Sun et al. "MG2C: A user-friendly online tool for drawing genetic maps" *Mol. Hortic*
50. Chen, Tian, Lin et al. (2023) "Sublethal effects of emamectin benzoate on development, reproduction, and vitellogenin and vitellogenin receptor gene expression in Thrips hawaiiensis (Thysanoptera: Thripidae)" *J. Insect Sci*
51. Consortium (2010) "Genome sequence of the pea aphid Acyrthosiphon pisum" *PLoS Biol*
52. Bolger, Lohse, Usadel (2014) "Trimmomatic: A flexible trimmer for Illumina sequence data" *Bioinformatics*
53. Kim, Paggi, Park et al. (2019) "Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype" *Nat. Biotechnol*
54. Anders, Pyl, Huber (2015) "HTSeq-A Python framework to work with high-throughput sequencing data" *Bioinformatics*
55. Hekmat-Scafe, Scafe, Mckinney et al. (2002) "Genome-wide analysis of the odorant-binding protein gene family in Drosophila melanogaster" *Genome Res*
56. Manoharan, Ng Fuk Chong, Vaïtinadapoulé et al. (2013) "Comparative genomics of odorant binding proteins in Anopheles gambiae, Aedes aegypti, and Culex quinquefasciatus" *Genome Biol. Evol*
57. Gong, Zhang, Zhao et al. (2009) "The odorant binding protein gene family from the genome of silkworm, Bombyx mori"
58. Dippel, Oberhofer, Kahnt et al. (1141) "Tissue-specific transcriptomics, chromosomal localization, and phylogeny of chemosensory and odorant binding proteins from the red flour beetle Tribolium castaneum reveal subgroup specificities for olfaction or more general functions"
59. Terry, Schneider (1993) "Copulatory behaviour and mating frequency of the western flower thrips, Frankliniella occidentalis (Insecta: Thysanoptera)" *J. Pure Appl. Zool*
60. Steenbergen, Abd-El-Haliem, Bleeker et al. (2018) "Thrips advisor: Exploiting thrips-induced defences to combat pests on crops" *J. Exp. Bot*
61. Zhang, Zhu, Lu (2011) "Behavioural and chemical evidence of a male-produced aggregation pheromone in the flower thrips Frankliniella intonsa" *Physiol. Entomol*
62. (2026) *Biology*
63. James, Hamilton, Hall et al. (2005) "Identification of a male-produced aggregation pheromone in the western flower thrips Frankliniella occidentalis" *J. Chem. Ecol*
64. Yin, Wang, Fang et al. (2019) "Functional characterization of odorant-binding proteins from the scarab beetle Holotrichia oblita based on semiochemical-induced expression alteration and gene silencing" *Insect Biochem. Mol. Biol*
65. Wu, Liu, Shi et al. (2020) "Transcriptome analysis of antennal cytochrome P450s and their transcriptional responses to plant and locust volatiles in Locusta migratoria" *Int. J. Biol. Macromol*
66. Mappin, Bellantuono, Ebrahimi et al. "Odor-evoked transcriptomics of Aedes aegypti mosquitoes" *PLoS ONE*
67. Chen, Lei, Liang et al. (1355) "A case study on the γ-octalactone induced expression of Obp83g-2 in Bactrocera dorsalis (Hendel) revealed the transcriptional regulation of insect odorant binding protein" *Commun. Biol*
68. Yuan, Wang, Zhang et al. "General Odorant Binding Protein 2 and Odorant Binding Protein 36 facilitate the recognition of adult sex pheromone components by Hyphantria cunea larvae"
69. Zhao, Wang, Wang et al. (2024) "Omics Analysis of Odorant-Binding Proteins and Cuticle-Enriched SfruOBP18 Confers Multi-Insecticide Tolerance in Spodoptera frugiperda" *J. Agric. Food Chem*
70. Yuan, Rao, Zhong et al. (2024) "Exploring the functional profiles of odorant binding proteins crucial for sensing key odorants in the new leaves of coconut palms in Rhynchophorus ferrugineus" *Int. J. Biol. Macromol*
71. "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=PMC12479675&blobtype=pdf | # Handling editor Zhongjie Shi
Mert Erdin, E Altan, Nina Suomalainen, Hussein Alburkat, Tuomas Aivelo, Tarja Sironen, Teemu Smura
## Abstract
We report complete and partial genome sequences of rat hepatitis E virus (RHEV, Rocahepevirus ratti), from archived brown rats captured in Helsinki, Finland. Phylogenetic analysis confirmed the presence of the pathogenic RHEV genotype C1 in the Helsinki region. Finnish strains clustered together with strains from South Korea and Spain. However, the polytomous topology of phylogenetic trees and the large genetic distances between spatially distinct strains suggest that RHEV has remained inadequately sampled on a global scale. Further surveillance of rocahepeviruses is needed to assess their threat to public health and to understand their diversity and evolutionary patterns.
known as ferret hepatitis E virus, FrHEV), and foxes (order Carnivora), as well as humans [2][3][4][5][6][7][8][9][10][11]. Currently, there are two officially classified RHEV genotypes according to the International Committee on Taxonomy of Viruses (ICTV): RHEV-C1 in rats and RHEV-C2 in mustelids. In addition, a putative genotype infecting field mice has been proposed [12].
While human hepatitis E virus (HEV; subfamily Orthohepevirinae, species Paslahepevirus balayani) is one of the most common causes of acute hepatitis [13,14], rocahepeviruses have garnered attention due to their zoonotic potential. In addition to acute hepatitis, HEV is associated with severe hepatitis in women during pregnancy [15]. Some adverse outcomes during pregnancy, such as fulminant liver failure, premature delivery, postpartum hemorrhage, and low birth weight, may be observed and pose a threat to both mothers and infants [15]. Thus, understanding the diversity of HEV and RHEV will help to improve diagnostic methods for early detection and consequently enhance our capability to prevent these adverse outcomes. Genotype RHEV-C1 has been reported to cause mild self-limiting liver dysfunction, but it also has the potential to cause persistent infections, especially in immunocompromised patients [9,11,16]. Its ability to cause infections in humans may add to the global human hepatitis E burden and makes this virus a public health concern. Human case reports of RHEV-C1 from Hong Kong, Spain, and France and one reported infection in a patient traveling from Africa to Canada have provided insight into the global distribution and high divergence of Hepatitis E viruses are single-stranded positive-sense RNA viruses belonging to the family Hepeviridae, which is further divided into two subfamilies: Orthohepevirinae and Parahepevirinae [1]. Rocahepevirus is one of the four genera in the subfamily Orthohepevirinae, and this genus is known to include genotypes that are pathogenic to humans [1]. Initially, rocahepeviruses were discovered in rodents, particularly in rats [2], but they have recently been found to circulate also in mice, voles, and even carnivores such as ferrets [3]. Currently, the genus Rocahepevirus is divided into two species: Rocahepevirus ratti (rat hepatitis E virus, RHEV) and Rocahepevirus eothenomi (vole hepatitis E virus, VHEV). VHEV has been reported in voles (family Cricetidae, order Rodentia) [4,5], whereas RHEV has been detected in rodents, ferrets (ferret-borne RHEV is also these viruses [8][9][10][11]. The incidence of RHEV-C1 infections seems to be higher in Europe and Asia than on the other continents [17], but more studies are needed to confirm this.
Rats are distributed globally, and therefore, human pathogens carried by these animals are also widespread [18]. Rats co-inhabit areas with humans, increasing the likelihood of direct or indirect contact between the two. This highlights the importance of pathogen surveillance in rats. The genus Rocahepevirus seems to have high genetic diversity with varying zoonotic potential, as it includes some genotypes that are pathogenic to humans and others that are not. Hence, genetic surveillance of these viruses is of immense importance. Here, we report complete genome sequences and phylogenetic analysis of RHEV from brown rats captured in Helsinki, Finland.
RHEV was detected recently in brown rats (Rattus norvegicus) in Helsinki, Finland, and a detailed description of rodent capture, identification, and RHEV screening was provided in a previous report [18]. In that study, the rats were collected by pest management professionals and identified morphologically. A total of 285 rat livers were used for total RNA extraction followed by hepevirus-specific PCR screening, resulting in four RHEV-RNA-positive samples. Here, we subjected these archived RHEV RNA-positive samples to metatranscriptomic sequencing.
We used an NEBNext rRNA Depletion Kit v2 and an Ultra II RNA Library Prep Kit for construction of a nextgeneration sequencing (NGS) library, which was sequenced using an Illumina NovaSeq 6000 system. The raw data were then quality-filtered and assembled de-novo, and the resulting contigs were annotated using Lazypipe software [19]. Two partial and two complete RHEV genome sequences were obtained from a total of four samples. The sequence read and mapping statistics are shown in Supplementary Table S1.
Multiple sequence datasets were used for phylogenetic analysis. The first two datasets consisted of the complete open reading frame 1 (ORF1) and ORF2 sequences of members of the subfamily Orthohepevirinae available in the GenBank database. These ORFs encode the non-structural proteins and the capsid protein, respectively. The sequences in each dataset were aligned using MAFFT, and maximumlikelihood (ML) trees were constructed using IQ-TREE2 [20]. ModelFinder [21], implemented in IQ-TREE2, was used to find the best-fitting model for tree construction. This analysis showed the phylogenetic placement of the Finnish strains within the genus Rocahepevirus. After this initial characterization, we analyzed a dataset consisting of all of the complete genome sequences of RHEV from brown rats (n = 12). Because of the small number of complete sequences available for RHEV from brown rats, we also included partial sequences of both ORF1 and ORF2 within the complete genome dataset for a total of 221 sequences. The partial sequences from ORF1 and ORF2 ranged in size from 151 to > 6600 nucleotides. These sequences were aligned, and ML trees were constructed as described above. Based on the clustering pattern, we subsampled the sequences to focus on the characterization of the Finnish sequences. Because of the inconsistent length of the sequences in the dataset, we also subsampled the complete sequences and calculated the p-distances, using the ape package in R. The phylogenetic trees were visualized using iTOL or the ggtree package in R.
In a previous study [18], we identified four rat liver samples, one from 2020, two from 2022, and one from 2023, that were found to be positive for HEV using a molecular screening test [18]. These samples were from three different locations: the Vantaa waste incineration plant, from which one positive sample was collected in 2020, the Konala neighborhood, from which two positive samples were collected in 2022, and the Ruskeasuo community garden, from which one positive sample was collected in 2023. The positive samples were sequenced, resulting as two complete RHEV genome sequences and two partial sequences (accession numbers PP839290-93). While one of the partial sequences was a nearly complete genome sequence with 96.5% coverage in comparison to the complete RHEV genome, the other one had only 8% coverage and corresponded to the nonstructural protein coding region of the RHEV genome. The nucleotide sequence identity values between the sequences were high (99.3%). In addition to RHEV, two of the samples contained partial genome sequences that were related to Fesa-like virus V5B (MG571885), with 86% nucleotide and 94.4% amino acid identity. Fesa-like virus V5B is an unclassified member of the order Picornavirales that was detected in fecal samples collected from Amerindian children living in an isolated Amazonian village in Venezuela [22]. Further studies are needed to assess the prevalence of this virus in rats and to confirm the complete genome sequences. In addition, some sequence reads mapping to members of the family Partitiviridae were also detected.
Maximum-likelihood phylogenetic trees based on ORF1 or ORF2 sequences from members of the subfamily Orthohepevirinae showed that the RHEV genomes from Finland formed a subcluster together with sequences from Germany, Hungary, the Netherlands, the USA, China, and Indonesia, suggesting a global distribution of this subcluster (Fig. 1). There were only minor differences in the topologies between the ORF1 and ORF2 trees.
Beacuse the total number of complete genome sequences for RHEV from brown rats (clade RHEV-C1) was small (Supplementary Fig. S1), we combined the complete and partial sequence datasets to better capture the diversity of RHEV-C1. The initial dataset, which included all of the RHEV-C1 sequences, contained 2198 distinct patterns with 1 3 1856 parsimony-informative, 607 singleton, and 4562 constant sites. The resulting maximum-likelihood tree exhibited two major clades (Fig. 2). The clade including the Finnish strains was then subsampled and found to contain 2096 distinct patterns with 1801 parsimony-informative, 568 singleton, and 4656 constant sites. In the resulting ML tree, the Finnish strains formed a distinct cluster that shared an ancestral node with complete genome sequences from South Korea and partial sequences from Spain. Notably, the phylogenetic tree exhibited long branch lengths between the tips and internal nodes, suggesting that the full extent of RHEV genetic diversity may not be represented in the tree. Concordantly, relatively high pairwise p-distances indicated a high degree of divergence among the complete sequences (Supplementary Fig. S2)
In addition, low bootstrap values in the basal nodes of the tree indicated poor resolution of deep evolutionary relationships. This suggests that a large number of divergent strains have not yet been identified, especially in Eurasian countries.
In this study, we sequenced the complete genomes of RHEV-C1 genotype isolates from previously identified archived brown rat samples from Helsinki, Finland. The zoonotic potential of brown-rat-associated RHEV raises concerns for human health. Most of the human cases reported to date have been caused by members of genotype C1, to which the Finnish strains also belong. The capability of this genotype to cross host species barriers and the distribution of its main hosts, brown rats, in the areas where humans might come into close contact, makes this genotype a potential public-health threat and highlights the importance of continuous surveillance. It should also be noted that RHEV infections may be under-diagnosed due to the mild and self-limiting disease it typically causes. Further research and more genomic sequencing are needed to understand the spatiotemporal patterns and evolution of RHEV.
Fig. 1 Maximum-likelihood phylogenetic trees based on the ORF1 and ORF2 genes of members of the subfamily Orthohepevirinae for genus-level placement of the Finnish strains. The ORF1 tree was constructed using the model GTR + F + I + R5, which was identified as the best-fitting model by ModelFinder, and the ORF2 tree was constructed using TIM2 + F + I + R5 as the best-fitting model. For both trees, 1000 ultrafast bootstrap replicates were performed. The Finnish strains from this study are indicated by black circles, and the subcluster including the Finnish sequences is indicated by a red rectangle
## References
1. Tse, Zee, Tsang et al. (2021) "Transmission of Rat Hepatitis E Virus Infection to Humans in Hong Kong: A Clinical and Epidemiological Analysis" *Hepatology*
2. Wu, Zhou, Wang et al. (2024) "Molecular epidemiology and phylogeny of the emerging zoonotic virus Rocahepevirus: A global genetic analysis" *Infect Genet Evol*
3. Aslan, Balaban (2020) "Hepatitis E virus: Epidemiology, diagnosis, clinical manifestations, and treatment" *World J Gastroenterol*
4. Kamar, Izopet, Pavio et al. (2017) "Hepatitis E virus infection" *Nat Rev Dis Primers*
5. Wu, Wang, Xiang et al. "Chinese Consortium for the Study of Hepatitis E (2024) Role of viral hepatitis in pregnancy and its triggering mechanism" *J Transl Int Med*
6. Casares-Jimenez, Rivero-Juarez, Lopez-Lopez et al. (2024) "Rat hepatitis E virus (Rocahepevirus ratti) in people living with HIV" *Emerg Microbes Infect*
7. Benavent, Carlos (2023) "Rocahepevirus ratti as an emerging cause of acute hepatitis worldwide" *Microorganisms*
8. Aivelo, Alburkat, Suomalainen et al. (2018) "Potentially zoonotic pathogens and parasites in opportunistically sourced urban brown rats (Rattus norvegicus) in and around Helsinki" *Euro Surveill*
9. Plyusnin, Vapalahti, Sironen et al. (2023) "Enhanced viral metagenomics with lazypipe 2"
10. Minh, Schmidt, Chernomor et al. (2020) "IQ-TREE 2: New Models and Efficient Methods for Phylogenetic Inference in the Genomic Era" *Mol Biol Evol*
11. Kalyaanamoorthy, Minh, Wong et al. (2017) "ModelFinder: fast model selection for accurate phylogenetic estimates" *Nat Methods*
12. Siqueira, Dominguez-Bello, Contreras et al. (2018) "Complex virome in feces from Amerindian children in isolated Amazonian villages" *Nat Commun*
13. Purdy, Drexler, Meng et al. (2022) "ICTV virus taxonomy profile: hepeviridae 2022" *J Gen Virol*
14. Widen, Ayral, Artois et al. (2014) "PCR detection and analysis of potentially zoonotic Hepatitis E virus in French rats" *Virol J*
15. Eiden, Dahnert, Spoerel et al. (1993) "Spatial-Temporal Dynamics of Hepatitis E Virus Infection in Foxes (Vulpes vulpes) in Federal State of"
16. Wang, Li, Zhou et al. (2018) "Chevrier's Field Mouse (Apodemus chevrieri) and Pere David's Vole (Eothenomys melanogaster) in China Carry Orthohepeviruses that form Two Putative Novel Genotypes Within the Species Orthohepevirus C" *Virol Sin*
17. Ryll, Heckel, Corman et al. (2019) "Genomic and spatial variability of a European common vole hepevirus" *Arch Virol*
18. Sridhar, Yip, Lo et al. (2022) "Hepatitis E Virus Species C Infection in Humans"
19. Raj, Smits, Pas et al. (2012)
20. Rivero-Juarez, Frias, Perez et al. (2022) "Orthohepevirus C infection as an emerging cause of acute hepatitis in Spain: First report in Europe" *J Hepatol*
21. Sridhar, Yip, Wu et al. (2018) "Rat Hepatitis E Virus as Cause of Persistent Hepatitis after Liver Transplant" *Emerg Infect Dis*
22. Andonov, Robbins, Borlang et al. (2019) "Rat Hepatitis E Virus Linked to Severe Acute Hepatitis in an Immunocompetent Patient" *J Infect Dis*
23. Sridhar, Yip, Wu et al. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645945&blobtype=pdf | # Galectin 3-binding protein suppresses PRRSV replication via Cullin3-mediated ubiquitination degradation of non-structural protein 12
Xinrong Wang, Wenli Zhang, Juan Zhang, Rui Li, Longxiang Zhang, Nan Yan, Junhai Zhu, Lizhi Fu, Yue Wang
## Abstract
Porcine reproductive and respiratory syndrome virus (PRRSV) poses a major threat to the global swine industry, yet effective antiviral strategies remain limited. This study identifies galectin 3-binding protein (LGALS3BP) as a critical host factor inhibit ing PRRSV infection through targeting the viral conserved non-structural protein 12 (nsp12), a key component of the viral replication-transcription complex. Overexpression of LGALS3BP significantly suppressed PRRSV replication, while its knockdown enhanced viral replication. Mechanistically, LGALS3BP recruits the Cullin3 E3 ubiquitin ligase via its BACK domain to mediate the ubiquitination of nsp12 at lysine residue 91, lead ing to proteasomal degradation. This process disrupts nsp12-dependent synthesis of viral subgenomic RNA, thereby disrupting replication. Additionally, LGALS3BP enhances antiviral innate immunity by upregulating interferon (IFN)-β and IFN-stimulated genes (ISGs). The antiviral effect of LGALS3BP is conserved across diverse PRRSV strains, highlighting its broad-spectrum potential. These findings reveal a dual mechanism whereby LGALS3BP restricts PRRSV through direct degradation of a critical viral enzyme and modulation of host immune responses, highlighting LGALS3BP as a promising therapeutic avenue for PRRSV control.IMPORTANCE Porcine reproductive and respiratory syndrome virus (PRRSV) remains a major challenge to global swine production due to its genetic diversity, rapid mutation rate, and ability to evade host immunity. The nsp12 is highly conserved across PRRSV strains and plays a crucial role in viral RNA synthesis. This study identifies LGALS3BP as a critical host factor that inhibits PRRSV infection by targeting nsp12 via the ubiquitin-pro teasome pathway. By uncovering this novel antiviral mechanism, the research highlights LGALS3BP as a promising therapeutic target for PRRSV control. Moreover, it contributes to our understanding of how host factors modulate viral replication and immunity, opening new avenues for developing host-targeted antiviral strategies. These findings have the potential to mitigate PRRSV-driven economic losses and improve swine health worldwide.
genome mutations, insertions/deletions, and recombination, making epidemic control a challenge (5). In China, classic PRRSV strains, such as CH-1a and BJ-4, were first identified after 1995. The highly pathogenic PRRSV variant (HP-PRRSV), including strains JXA1 and HuN4, emerged in 2006 (6,7). Recent surveillance indicates that HP-PRRSV-like and NADC30-like strains are now the dominant variants, with increasing detection rates (8).
PRRSV has a single-stranded, positive-sense RNA genome of approximately 15 kb, encoding at least 10 open reading frames (ORFs). ORF1a and ORF1b are translated into large polyproteins that undergo autocleavage to generate 16 non-structural proteins (nsp1α, nsp1β, nsp2N, nsp2TF, nsp2-nsp6, nsp7α, nsp7β, and nsp8-nsp12) (9,10). The replication and transcription complexes (RTCs), composed of viral and host proteins, are crucial for PRRSV replication (11,12). ORF1b, the most conserved region of the arterivirus genome, encodes nsp9-12, which is a key component of the RTC (13). Among them, nsp12 was initially uncharacterized but is known to regulate subgenomic mRNA synthesis (+sgmRNA and -sgmRNA) without affecting minus-strand genomic RNA (-gRNA) synthesis (14).
Galectin 3-binding protein (LGALS3BP), also known as 90K or Mac-2BP, is a highly glycosylated, disulfide-linked oligomeric protein in the scavenger receptor cysteine-rich (SRCR) superfamily (15). Its multi-domain structure, comprising SRCR, BTB/POZ, and BACK domains, enables interactions with extracellular matrix components and contrib utes to immune regulation, cell adhesion, and tumor progression (16,17). Recent studies report that LGALS3BP is upregulated during viral infections, such as HIV and herpesvirus, where it acts as a modulator of antiviral immunity (18,19). In our previous study (20,21), PRRSV infection could upregulate multiple host cellular genes, including LGALS3BP. However, its antiviral function in PRRSV remains unclear. This study investigates the role of LGALS3BP in PRRSV infection.
## RESULTS
## PRRSV infection induces LGALS3BP expression
Based on our previous RNA-seq results (20,21), LGALS3BP was significantly upregula ted in response to both HuN4 and SD53 infections at 3 days post-infection (dpi), as shown in a heatmap (Fig. 1A). We then verified the LGALS3BP mRNA expression in lung homogenates. Reverse transcription quantitative PCR (RT-qPCR) analysis revealed that the LGALS3BP mRNA levels were significantly higher in pigs infected with either the HP-PRRSV strain HuN4 or NADC30-like strain SD53 compared with controls (Fig. 1B andC). To further validate this upregulation at protein levels, enzyme-linked immunosorbent assay (ELISA) was performed on lung homogenates, revealing a significant increase in LGALS3BP protein following infection with both strains (Fig. 1D andE). These data suggest that the upregulation of LGALS3BP may be part of the host's antiviral response to PRRSV infection.
## LGALS3BP suppresses PRRSV replication
To assess the role of LGALS3BP in PRRSV infection, we constructed a plasmid encoding LGALS3BP. Marc-145 cells were transfected with either empty vector (EV) or LGALS3BP plasmids for 24 h, followed by infection with 0.1 multiplicity of infection (MOI) PRRSV HuN4 for an additional 24 h. RT-qPCR analysis revealed significant, dose-dependent reduction in PRRSV mRNA levels with LGALS3BP overexpression (Fig. 2A). Western blot analysis confirmed this, as LGALS3BP suppressed PRRSV-N protein levels in cells treated with one or 2 µg LGALS3BP plasmids, seen via weaker band than EV controls (Fig. 2B). Additionally, LGALS3BP-overexpressed cells had significantly lower viral titers (Fig. 2C). To further evaluate the effect of LGALS3BP on PRRSV replication, we examined viral RNA, protein, and titers at different time points and MOIs. The results showed significant reductions at 12, 24, and 36 h post-infection (hpi) (Fig. 2D through F). Moreover, LGALS3BP overexpression resulted in a significant decrease in PRRSV replication across different MOIs (Fig. 2G through I).
LGALS3BP also inhibited replication of multiple PRRSV strains, including HP-PRRSV JX, classic strains BJ-4 and CH-1R, and NADC30-like strain HNhx, as shown by RT-qPCR (Fig. 2J). The antiviral activity of LGALS3BP was validated in porcine alveolar macrophages (PAMs) using a recombinant lentivirus expressing LGALS3BP. Following PRRSV infection, RT-qPCR and Western blot analyses showed successful overexpression of LGALS3BP (Fig. 2K andM), which resulted in significant reductions in viral RNA (Fig. 2L), protein (Fig. 2M), and titer (Fig. 2N).
Furthermore, knockdown of LGALS3BP using specific siRNAs in Marc-145 cells resulted in a significant increase in PRRSV RNA, protein, and titer levels (Fig. 3A through D). And the result of knocking down LGASL3BP in immortalized PAMs (iPAMs) is similar to that of Marc-145 cells (Fig. 3E through G). Taken together, these findings confirm LGALS3BP as a host factor with antiviral activity against PRRSV.
## LGALS3BP regulates PRRSV infection at the stage of replication
To clarify LGALS3BP's role in the PRRSV life cycle, Marc-145 cells were transfected with EV or LGALS3BP for 24 h, followed by PRRSV infection for 1 h at 4°C to allow viral attachment. After removing unbound virions, cells were shifted to 37°C and harvested at designated times for viral quantification. As shown in Fig. 4A andB, LGALS3BP overexpression did not affect PRRSV RNA levels at 1 hpi (4°C) or between 1 and 3 hpi (37°C), suggesting no impact on virus attachment or penetration. However, a significant reduction in viral RNA was observed at 9 hpi (Fig. 4C), indicating that LGALS3BP regulates PRRSV infection during the viral biosynthesis stage after entry into the host cell.
Inhibition of viral biosynthesis is commonly associated with the production of type I interferons (IFN-I, including IFN-α/β) or the activation of IFN-stimulated genes (ISGs). Therefore, we investigated the impact of LGALS3BP on IFN-I responses by transfect ing Marc-145, HEK-293T-CD163 cells, and iPAMs with EV, LGALS3BP, and poly(I:C) via measuring IFN-β and ISG expression by RT-qPCR. The results revealed that overexpres sion of LGALS3BP significantly upregulated IFN-β and several ISGs, including IFN-stimula ted gene 15 (ISG15) and IFN-induced protein with tetratricopeptide repeats 1 (IFIT1) (Fig. 4D through L). These results suggest that LGALS3BP enhances innate antiviral responses.
## LGALS3BP degrades nsp12 through the ubiquitin-proteasome pathway
Previous studies have demonstrated that LGALS3BP possesses ubiquitination activity and can promote the degradation of target proteins (22,23). Based on this, we hypothesized that the inhibitory effect of LGALS3BP in PRRSV may also be mediated by proteolytic activity. To test this, we co-transfected expression plasmids for major PRRSV nonstructural or structural proteins with LGALS3BP into HEK-293T cells and analyzed viral protein expression at 24 h. The results showed that LGALS3BP specifically reduced the levels of nsp12. In contrast, other nonstructural proteins, including nsp1α, nsp1β, nsp4, nsp5, nsp7, nsp9, nsp10, and nsp11, and structural proteins, such as GP2(ORF2), GP3(ORF3), GP4(ORF4), GP5(ORF5), M(ORF6), and N(ORF7), showed no significant changes (Fig. 5A). A dose-dependent experiment revealed that increasing LGALS3BP levels led to a gradual decline in nsp12 expression (Fig. 5B). Concurrently, a time-course experiment exhibited a significant reduction in nsp12 expression as well (Fig. 5C).
To determine the degradation pathway, we co-transfected HEK-293T cells with LGALS3BP and nsp12 plasmids with MG132, a proteasome inhibitor, and 3-MA, an autophagy inhibitor. As shown in Fig. 5D, LGALS3BP-mediated nsp12 degradation was effectively blocked by MG132 but not by 3-MA, indicating that nsp12 degradation occurs through the ubiquitin-proteasome pathway. Furthermore, we observed a significant increase in nsp12 ubiquitination in the presence of LGALS3BP, further enhanced by MG132 (Fig. 5E). Co-immunoprecipitation (Co-IP) experiments exhibited an interac tion between LGALS3BP and nsp12 (Fig. 5F), and laser confocal microscopy revealed their co-localization (Fig. 5G), suggesting that LGALS3BP facilitates nsp12 degradation through direct interaction.
## LGALS3BP mediates nsp12 degradation through the BACK domain
We mapped the domains of LGALS3BP (SRCR, BTB, and BACK) and created truncation mutants (Fig. 6A). Co-transfection of these truncation mutants with nsp12 into HEK-293T cells demonstrated that the BACK domain significantly reduced nsp12 expression as determined by Western blot (Fig. 6B). Co-IP analysis showed that nsp12 ubiquitination was primarily dependent on the BACK domain, while the SRCR and BTB domains had no significant effects (Fig. 6C). To identify the critical ubiquitination site, we generated lysine (K)-to-arginine (R) mutants within the 154 amino acids of nsp12 (K59R, K75R, K91R, K125R, K127R, K130R) (Fig. 6D). Western blot analysis showed that the K91R mutant was resistant to LGALS3BP-mediated degradation, implicating that lysine 91 (K91) is the key ubiquitination site (Fig. 6D andE). These findings demonstrate that LGALS3BP mediates nsp12 degradation through BACK domain-dependent ubiquitination at the K91 residue.
## LGALS3BP mediates nsp12 degradation by recruiting the Cullin3 E3 ubiquitin ligase
The BACK domain of BTB-BACK-kelch proteins, such as KEAP1 (Kelch-like ECH-associated protein 1), stabilizes interactions with Cullin3 (24-28), facilitating the formation of the Cullin3-RING E3 ligase complex, which targets specific substrates for ubiquitination and proteasome degradation. Based on this, we hypothesized that LGALS3BP recruits Cullin3 to degrade nsp12 via its BACK domain. To test this, we knocked down Cullin3 using specific siRNA, which significantly attenuated LGALS3BP-mediated nsp12 degradation (Fig. 7A andB). Co-IP analysis confirmed interactions between LGALS3BP and Cullin3 (Fig. 7C), as well as between Cullin3 and nsp12 (Fig. 7D). Further Co-IP of co-transfected LGALS3BP, Cullin3, and nsp12 revealed that LGALS3BP pulled down both nsp12 and Cullin3, indicating the formation of ternary complexes (Fig. 7E). Confocal microscopy analysis also validated the co-localization of Cullin3 with both LGALS3BP and nsp12 (Fig. 7F through G). Collectively, these findings identify Cullin3 as the key E3 ubiquitin ligase responsible for LGALS3BP-driven nsp12 ubiquitination and degradation, providing novel insights into the antiviral mechanisms of LGALS3BP in PRRSV infection.
LGALS3BP inhibits the synthesis of PRRSV subgenomic RNA PRRSV nsp12 has been reported to play a key role in the synthesis of viral subge nomic mRNA (14). To determine whether LGALS3BP-mediated nsp12 degradation affects subgenomic mRNA synthesis, Marc-145 cells were transfected with LGALS3BP plasmids for 24 h and then infected with PRRSV (0.1 MOI) for 4 h. Cells were harvested to analyze viral RNA levels. Reverse transcription was performed using universal primers (for total viral RNA) or strand-specific primers (for subgenomic mRNA), followed by qPCR to quantify viral genomic RNA, plus-and minus-strand subgenomic mRNA levels. The results revealed that LGALS3BP significantly reduced subgenomic mRNA production before impacting viral genomic RNA levels (Fig. 8A through C), indicating that LGALS3BP inhibits viral replication by blocking subgenomic mRNA synthesis.
## DISCUSSION
The genetic, pathogenic, and antigenic variability of PRRSV strains remains a major challenge to global swine health and sustainable pork production (29,30), underscor ing the need for effective antiviral strategies. In this study, we identify LGALS3BP as a potent antiviral host factor that inhibits PRRSV infection. We demonstrate that LGALS3BP targets the viral non-structural protein nsp12, a critical component of the replicationtranscription complex, leading to its proteasomal degradation. Our results show that overexpression of LGALS3BP significantly suppresses PRRSV replication, whereas its knockdown enhances viral replication. Furthermore, LGALS3BP enhances antiviral innate immune responses by upregulating IFN-β and ISGs. Together, these findings suggest that LGALS3BP restricts PRRSV through direct degradation of nsp12 and modulation of host immune responses, making it a promising therapeutic target for PRRSV control. Our findings align with previous studies showing that LGALS3BP is upregulated during viral infections such as HIV, herpesvirus, influenza A virus, vesicular stomatitis virus, and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), where it modulates antiviral immunity (15,18,19,31,32). In these cases, LGALS3BP has been linked to type I interferon responses. However, our study uniquely highlights LGALS3BP's direct role in PRRSV inhibition through the degradation of nsp12. To our knowledge, this is the first report of LGALS3BP targeting a key viral replicase protein in a nidovi rus like PRRSV. This contrasts with findings in SARS-CoV-2, where LGALS3BP's antiviral activity appears to be mainly immune-modulatory, without direct degradation of viral enzyme responsible for subgenomic RNA synthesis. In contrast, LGALS3BP's antiviral effect in SARS-CoV-2 is primarily immune-modulatory. We hypothesize that these two mechanisms work in parallel to restrict PRRSV replication: (i) degradation of nsp12 via the ubiquitin-proteasome pathway and (ii) upregulation of IFN-β and ISGs; however, future studies are needed to dissect their interplay. The dual activity of LGALS3BP suggests it acts both as a direct viral suppressor and as an enhancer of antiviral immunity, providing a robust defense against PRRSV.
Initially identified for its role in cancer development,
LGALS3BP has emerged as a potential therapeutic target (33,34). Our findings show that LGALS3BP significantly upregulates IFN-β and ISGs, including ISG15 and IFIT1, in addition to its direct antiviral effect. This suggests that LGALS3BP not only inhibits viral replication but also enhances the host's innate immune response. While this immune-modulatory role appears to be part of a general antiviral mechanism (35), the specific degradation of PRRSV nsp12 suggests that LGALS3BP has a targeted, virus-specific antiviral effect. This combination of immune regulation and viral protein degradation makes LGALS3BP an effective host factor against PRRSV and potentially other viral infections.
LGALS3BP, a multifunctional glycoprotein featuring SRCR, BTB/POZ, and BACK domains, participates in diverse biological processes (36)(37)(38)(39). While homologous domains in other proteins have been characterized, the precise roles of LGALS3BP's individual domains remain incompletely understood. The SRCR domain, a constituent of the SRCR protein receptor superfamily, mediates protein-ligand interactions (40). The BTB/POZ domain, prevalent in adaptor proteins, governs transcription, cytoske letal dynamics, and ubiquitination (15). The BACK domain, conserved in BTB-kelch proteins, facilitates E3 ubiquitin ligase assembly (28,(41)(42)(43). Although LGALS3BP induces E-cadherin downregulation (23), the domain(s) responsible for this activity have yet to be identified. Here, we show that LGALS3BP mediates nsp12 degradation by recruiting Cullin3, an E3 ubiquitin ligase, via its BACK domain. Specifically, LGALS3BP forms a ternary complex with Cullin3 and nsp12, promoting K91-specific ubiquitination of nsp12 and subsequent proteasomal degradation. Co-IP and confocal microscopy confirmed the direct interactions among these proteins, while Cullin3 knockdown abolished LGALS3BPdependent nsp12 degradation, establishing Cullin3 as essential for this process.
Nsp12, a highly conserved component of the RTC, is essential for PRRSV subge nomic RNA synthesis (44)(45)(46)(47). Additionally, nsp12 is often targeted by the host restric tion factors, such as PSMB1, RNF114, and GAL3 (48)(49)(50), underscoring its vulnerability as a viral dependency factor. We identified K91 in nsp12 as a conserved ubiquitina tion site across PRRSV strains, linking LGALS3BP-mediated ubiquitination to antiviral activity. These findings provide a novel mechanism by which LGALS3BP restricts PRRSV replication and highlight the ubiquitin-proteasome system as a key regulator of viral pathogenesis.
While our study provides compelling in vitro evidence for the antiviral activity of LGALS3BP, the findings should be validated in in vivo models. As PRRSV infects PAMs in vivo, future studies using PRRSV-infected pigs are needed to determine the therapeu tic potential of LGALS3BP. These studies should assess viral loads, immune responses, and clinical outcomes to evaluate the efficacy of LGALS3BP in a more physiologically relevant system. Furthermore, future work should investigate whether LGALS3BP can be administered as a therapeutic agent to reduce PRRSV replication and improve disease outcomes in swine.
In conclusion, our study identifies LGALS3BP as a critical antiviral host factor that suppresses PRRSV infection by directly targeting nsp12 and modulating immune responses (Fig. 9). This dual mechanism, degradation of a viral protein and activation of antiviral immunity, suggests that LGALS3BP could serve as a novel therapeutic target for PRRSV control. Given its broad-spectrum antiviral activity, LGALS3BP holds promise as a candidate for host-directed therapies, not only for PRRSV but potentially for other viral infections as well.
at Harbin veterinary research institute (51). Both PAMs and iPAMs were cultured in RPMI-1640 medium (Gibco) containing 10% FBS, 1% antibiotics. All cell cultures were incubated at 37°C in a humidified atmosphere with 5% CO₂.
The PRRSV strains employed in this study, including HuN4 (GenBank: EF635006), JX (GenBank: JX317649.1), BJ-4 (GenBank: AF331831.1), HNhx (GenBank: KX766379.1), and CH-1R (GenBank: EU807840), were preserved in our laboratory. The HuN4 strain served as the primary viral model throughout this investigation. Virus propagation and titration were performed in Marc-145 cells following standard protocols (52).
## Lentiviral preparation
Lentiviral production was performed as previously described with minor modifications (53). Briefly, HEK-293T cells were co-transfected with the lentiviral expression plasmid pLVX-IRES-ZsGreen1/LGALS3BP and the packaging plasmids psPAX2 and pMD2.G at a 3:2:1 mass ratio. Viral supernatants were collected at 48 and 72 h post-transfection, centrifuged at 4,000 × g for 5 min (4°C) to remove cell debris, and subsequently filtered through 0.45-µm PVDF membranes (Millipore) to obtain the clarified lentiviral stock. The harvested lentivirus was titrated in HEK293T cells with serial 10-fold dilution in the presence of 6 µg/mL polybrene. After 3-4 days of incubation, fluorescent cells from the last two dilution gradients were counted, and viral titer was calculated using the following formula: TU/mL = (X + Y × 10) × dilution factor × 10, where X and Y denote the number of fluorescent wells at each respective dilution. For transduction experiments, the PAMs were exposed to lentivirus at a MOI of 5, achieving greater than 50% fluorescent-positive cells.
## Antibodies and reagents
The anti-PRRSV-N antibody was prepared in our laboratory. Commercial antibodies included anti-Flag (MA1-91878), anti-HA (26183), goat anti-mouse IgG (H+L) cross-adsor bed secondary antibody (G-21040), and goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody (G-21234) from Invitrogen; anti-β-actin (A2228) from Sigma-Aldrich; BeyoMag Protein G beads (P2105), Western/IP lysis buffer (P0013), 4′,6′-diamidino-2phenylindole (DAPI, C1002), Alexa Fluor 555-labeled donkey anti-rabbit IgG (H+L) (A0452), and Alexa Fluor 488-labeled goat anti-mouse IgG (H+L) (A0428) from Beyotime Biotechnology. Pharmacological agents MG132 (HY-13259) and 3-methyladenine (3-MA) (HY-19312) were purchased from MedChemExpress (MCE). The porcine LGALS3BP ELISA kit was sourced from Shanghai Coibo Biotechnology.
## Construction of plasmids
PrimeSTAR HS DNA polymerase was used to amplify the LGALS3BP gene (GenBank: XM_021066516.1) from PAMs' cDNA before cloning it into the pCAGGS expression vector with a C-terminal-HA. Similarly, the PRRSV nsps' genes were amplified from cDNA of PRRSV HuN4-infected Marc-145 cells using PrimeSTAR HS DNA polymerase and cloned into pCAGGS with a C-terminal-Flag. Plasmids for Cullin3 were purchased from Miaoling Biotechnology (Wuhan, China).
## RNA-mediated interference (RNAi)
Small-interfering RNAs (siRNAs) against LGALS3BP and Cullin3 were synthesized by GenePharma (Shanghai, China). Marc-145 or HEK-293T cells were transfected with 50 nM of either siRNA-LGALS3BP or siRNA-Cullin3 using the siRNA-MATE plus reagent (Gene pharma) following the recommended procedure. The siRNA sequences are listed in Table 1.
## Reverse transcription-quantitative PCR
Total RNA was extracted from cells, using RNA extraction kits (BioFlux, China), and reverse transcribed into cDNA using PrimeScript RT reagent Kit (Takara, Japan) according to the manufacturer's instructions. The ChamQ Universal SYBR qPCR Master Mix (Vazyme, China) was used to quantify mRNA levels. The primer sequences are listed in Table 2. Fold changes were determined using the cycle threshold (ΔΔCT) method.
## Transfection and Western blotting
The cells were transfected with indicated plasmids using Lipofectamine 2000 (Invi trogen). After 24 h, the cells were collected and lysed in Western and IP lysis buf fer (Beyotime) with 1% PMSF (Beyotime). The cell supernatants were collected after centrifugation for 10 min at 12,000 × g and mixed with SDS-PAGE loading buffer, followed by boiling at 100°C for 10 min. The samples were separated with SDS-PAGE and then transferred to polyvinylidene difluoride (PVDF) membranes (Merck Millipore, USA), which were blocked in 5% skim milk and incubated with the indicated primary and secondary antibodies.
## Co-immunoprecipitation assay
After 24 h of transfection, cells were collected and lysed in Western and IP lysis buffer supplemented with 1% PMSF. The cell supernatants were collected after centrifugation for 10 min at 12,000 × g. A 100-µL aliquot of supernatant was taken as the input sample, mixed with SDS-PAGE loading buffer, and boiled at 100°C for 10 min. The remaining supernatant was incubated with monoclonal anti-Flag, anti-HA, or anti-GFP antibodies for 2 h. Following incubation, 20 µL Beyomag protein G beads was added, and the mixture was incubated for an additional 2 h. The beads were then washed five times with PBST, and the samples were analyzed by Western blotting.
## Confocal imaging
HEK-293T cells were co-transfected with indicated plasmids for 24 hpi. The cells were fixed in 4% paraformaldehyde for 30 min, permeabilized with 0.1% Triton X-100 for 15 min, and blocked with 3% bovine serum albumin for 1.5 h. The transfected cells were incubated with mouse anti-HA MAb and rabbit anti-Flag MAb for 1 h at RT and washed three times with PBST. The cells were then incubated at 37°C for 1 h with goat anti-mouse IgG (H+L) antibody conjugated with Alexa Fluor 488 and donkey anti-rabbit IgG (H+L) antibody labeled with Alexa Fluor 555. Finally, the cells were stained with 1 µg/mL of DAPI for 5 min and examined using a Zeiss confocal system.
## RT-qPCR of sgmRNA
Marc-145 cells were transfected with LGALS3BP plasmids for 24 h, followed by infection with PRRSV at 0.1 MOI. Cells were harvested at 4 hpi for RNA extraction. For +sgmRNA detection, cDNA was amplified with oligo-d(T). For -sgmRNA detection, cDNA was amplified using the primer: 5′-GTGTTGGCTCATGCCACGGC-3′. Strand-specific quantification of +sgmRNA and -sgmRNA was then conducted by qPCR using the primers provided in Table 1.
## Statistical analysis
Statistical analyses were carried out using Prism 8.0. Most experiments were conducted with at least three independent replicates. Data are presented as means ± standard errors (SD) from three or more independent experiments. One-way analysis of variance or Student's t-test (two-tailed) was used for statistical comparisons. Significance levels were defined as: ns, P > 0.05; *, P < 0.05; **, P < 0.01; ***, P < 0.001; and ****, P < 0.0001.
## References
1. Li, Li, Gong et al. (2025) "A lineage 1 branch porcine reproductive and respiratory syndrome virus live vaccine candidate provides broad crossprotection against HP-like PRRSV in piglets" *Virulence*
2. Cohen (2024) "Meat from gene-edited pigs could hit the market" *Science*
3. Meulenberg, Hulst, De Meijer et al. (1993) "Lelystad virus, the causative agent of porcine epidemic abortion and respiratory syndrome (PEARS), is related to LDV and EAV" *Virology (Auckl)*
4. Qiu, Qiu, Li et al. (2025) "Emergence, prevalence and evolution of porcine reproductive and respiratory syndrome virus 1 in China from 1994 to 2024" *Virology (Auckl)*
5. Cui, Xia, Luo et al. (2024) "Recombination of porcine reproductive and respiratory syndrome virus: features, possible mechanisms, and future directions"
6. Han, Wu, Deng et al. (2009) "Molecular mutations associated with the in vitro passage of virulent porcine reproductive and respiratory syndrome virus" *Virus Genes*
7. Tian, An, Zhou et al. (2009) "An attenuated live vaccine based on highly pathogenic porcine reproductive and respiratory syndrome virus (HP-PRRSV) protects piglets against HP-PRRS" *Vet Microbiol*
8. Guo, Chen, Li et al. (2018) "The prevalent status and genetic diversity of porcine reproductive and respiratory syndrome virus in China: a molecular epidemiological perspective" *Virol J*
9. Zhang, Li, Li et al. (2015) "Importation and recombination are responsible for the latest emergence of highly pathogenic porcine reproductive and respiratory syndrome virus in China" *Front Vet Sci*
10. Song, Liu, Gao et al. (2018) "Mapping the nonstructural protein interaction network of porcine reproductive and respiratory syndrome virus" *J Virol*
11. Kappes, Faaberg (2015) "PRRSV structure, replication and recombination: origin of phenotype and genotype diversity" *Virology (Auckl)*
12. Nan, Lan, Tian et al. (2018) "The network of interactions among porcine reproductive and respiratory syndrome virus non-structural proteins" *Front Microbiol*
13. Wang, Fang, Cong et al. (2019) "The Nsp12-coding region of type 2 PRRSV is required for viral subgenomic mRNA synthesis" *Emerg Microbes Infect*
14. Loimaranta, Hepojoki, Laaksoaho et al. (2018) "Galectin-3binding protein: a multitask glycoprotein with innate immunity functions in viral and bacterial infections" *J Leukoc Biol*
15. Gallo, Arienzo, Iacobelli et al. (2022) "Gal-3BP in viral infections: an emerging role in severe acute respiratory syndrome coronavirus 2" *Int J Mol Sci*
16. Capone, Iacobelli, Sala (2021) "Role of galectin 3 binding protein in cancer progression: a potential novel therapeutic target" *J Transl Med*
17. Gröschel, Braner, Funk et al. (2000) "Elevated plasma levels of 90K (Mac-2 BP) immunostimulatory glycoprotein in HIV-1-infected children" *J Clin Immunol*
18. Rasmussen, Draborg, Houen et al. (2022) "Human herpesvirus infections and circulating microvesicles expressing galectin-3 binding protein in patients with systemic lupus erythemato sus" *Clin Exp Rheumatol*
19. (2025) *Full-Length Text Journal of Virology*
20. Zhang, Wang, Zhang et al. (2024) "Comparison of pathogenicity and host responses of emerging porcine reproductive and respiratory syndrome virus variants in piglets" *J Virol*
21. Zhang, Ma, Pan et al. (2023) "Characterization of Rongchang piglets after infection with type 2 porcine reproductive and respiratory syndrome virus strains differing in pathogenicity" *Front Microbiol*
22. Zhao, Zhou, Li et al. (2024) "UBE2G2 inhibits vasculogenic mimicry and metastasis of uveal melanoma by promoting ubiquitination of LGALS3BP" *Acta Pharm Sin B*
23. Park, Yoon, Sun et al. (2017) "Glycoprotein 90K promotes E-cadherin degrada tion in a cell density-dependent manner via dissociation of E-cadherinp120-catenin complex" *Int J Mol Sci*
24. Canning, Bullock (2014) "New strategies to inhibit KEAP1 and the cul3-based E3 ubiquitin ligases" *Biochem Soc Trans*
25. Gupta, Beggs (2014) "Kelch proteins: emerging roles in skeletal muscle development and diseases" *Skelet Muscle*
26. Szabó, Papin, Cornu et al. (2018) "Ubiquitylation dynamics of the clock cell proteome and TIMELESS during a circadian cycle" *Cell Rep*
27. Gao, Pallett, Croll et al. (2019) "Molecular basis of cullin-3 (Cul3) ubiquitin ligase subversion by vaccinia virus protein A55" *J Biol Chem*
28. Nan, Rao, Tang et al. (2024) "Burkholderia pseudomallei BipD modulates host mitophagy to evade killing" *Nat Commun*
29. Nan, Wu, Gu et al. (2017) "Improved vaccine against PRRSV: current progress and future perspective" *Front Microbiol*
30. Murtaugh, Stadejek, Abrahante et al. (2010) "The ever-expanding diversity of porcine reproductive and respiratory syndrome virus" *Virus Res*
31. Xu, Xia, Deng et al. (2019) "Inducible LGALS3BP/90K activates antiviral innate immune responses by targeting TRAF6 and TRAF3 complex" *PLoS Pathog*
32. De Jarcy, Akbil, Mhlekude et al. (2023) "90K/ LGALS3BP expression is upregulated in COVID-19 but may not restrict SARS-CoV-2 infection" *Clin Exp Med*
33. Kim, Park, Kim et al. (2025) "Secreted LGALS3BP facilitates distant metastasis of breast cancer" *Breast Cancer Res*
34. Li, Zhao, Li et al. (2022) "Osteosarcoma exocytosis of soluble LGALS3BP mediates macrophages toward a tumoricidal phenotype" *Cancer Lett*
35. Lodermeyer, Ssebyatika, Passos et al. (2018) "The antiviral activity of the cellular glycoprotein LGALS3BP/90K is species specific" *J Virol*
36. Sasaki, Brakebusch, Engel et al. (1998) "Mac-2 binding protein is a cell-adhesive protein of the extracellular matrix which self-assembles into ring-like structures and binds beta1 integrins, collagens and fibronectin" *EMBO J*
37. Kong, Lin, Li et al. (2010) "Cyclophilin Cassociated protein/Mac-2 binding protein colocalizes with calnexin and regulates the expression of tissue transglutaminase" *J Cell Physiol*
38. Hellstern, Sasaki, Fauser et al. (2002) "Functional studies on recombinant domains of Mac-2-binding protein" *J Biol Chem*
39. Grassadonia, Tinari, Fiorentino et al. (2004) "The 90K protein increases major histocompatibility complex class I expression and is regulated by hormones, gamma-interferon, and double-strand polynucleotides" *Endocrinology*
40. Martínez, Moestrup, Holmskov et al. (2011) "The conserved scavenger receptor cysteine-rich superfamily in therapy and diagnosis" *Pharmacol Rev*
41. Stogios, Privé (2004) "The BACK domain in BTB-kelch proteins" *Trends Biochem Sci*
42. Canning, Cooper, Krojer et al. (2013) "Structural basis for Cul3 protein assembly with the BTB-Kelch family of E3 ubiquitin ligases" *J Biol Chem*
43. Ji, Privé (2013) "Crystal structure of KLHL3 in complex with cullin3" *PLoS One*
44. Huang, Sun, Zhu et al. (2024) "Identification of new antigenic epitopes of porcine reproductive and respiratory syndrome virus nsp12 protein using monoclonal antibodies" *Int J Biol Macromol*
45. Zhu, Xu, Chen et al. (2023) "Bergamottin inhibits PRRSV replication by blocking viral non-structural proteins expression and viral RNA synthesis" *Viruses*
46. Wang, Yi, Guo et al. (2023) "PCNA promotes PRRSV replication by increasing the synthesis of viral genome" *Vet Microbiol*
47. Cook, Brown, Shang et al. (2022) "Ribosome profiling of porcine reproductive and respiratory syndrome virus reveals novel features of viral gene expression" *Elife*
48. Li, Zhou, Jiang et al. (2018) "Galectin-3 inhibits replication of porcine reproductive and respiratory syndrome virus by interacting with viral Nsp12 in vitro" *Virus Res*
49. Li, Bai, Zhou et al. (2023) "PSMB1 inhibits the replication of porcine reproductive and respiratory syndrome virus by recruiting nbr1 to degrade nonstructural protein 12 by autophagy" *J Virol*
50. Bai, Li, Shan et al. (2020) "Proteasomal degradation of nonstructural protein 12 by RNF114 suppresses porcine reproductive and respiratory syndrome virus replication" *Vet Microbiol*
51. Wang, Liu, Li et al. (2018) "Porcine alveolar macrophage CD163 abundance is a pivotal switch for porcine reproductive and respiratory syndrome virus infection" *Oncotarget*
52. Zhang, Zhang, Pan et al. (2021) "Downregulation of miR-218 by porcine reproductive and respiratory syndrome virus facilitates viral replication via inhibition of type I interferon responses" *Journal of Biological Chemistry*
53. Pan, Zhang, Ibrahim et al. (2024) "miR-191-5p suppresses PRRSV replication by targeting porcine EGFR to enhance interferon signaling" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12811730&blobtype=pdf | # Candida auris in Heart Transplant Donor: First Isolation in a Southern Italy Heart Transplant Center
Lorenzo Giovannico, Giuseppe Fischetti, Luca Savino, Giuseppina Caggiano, | Chironna, Silvio Tafuri, Tomaso Bottio
## Abstract
Candida auris is an emerging multidrug-resistant fungal pathogen increasingly implicated in healthcare-associated outbreaks worldwide. Its presence in organ donors poses a significant threat to transplant recipients due to the risk of invasive infection and limited antifungal treatment options. We report the first isolation of Candida auris in a heart transplant donor at a transplant center in Southern Italy. The donor, a 45-year-old woman from Greece, was colonized with C. auris in the bronchoaspirate sample collected at the time of organ retrieval. Despite this colonization, the donor heart was successfully transplanted into a 62-year-old male recipient with end-stage heart failure secondary to myocardial infarction and cardiogenic shock. The recipient received targeted perioperative prophylaxis and was placed under strict isolation protocols. Repeated microbiological surveillance, including blood, urine, and mucosal cultures, revealed no evidence of C. auris transmission. Environmental surveillance of the operating room and ICU also tested negative. The patient recovered uneventfully, showing good cardiac function and no signs of graft rejection or infection. This case emphasizes the critical importance of early detection, thorough microbiological assessment, and stringent infection control in transplantation involving donors colonized with multidrug-resistant organisms. It also raises the question of whether C. auris should be routinely screened in potential donors and if specific transplant guidelines should be developed to address such emerging threats.
## 1 | Introduction
Heart failure (HF) is a global health issue, affecting over 64.3 million people worldwide in 2017. Its prevalence is expected to rise due to improved survival following HF diagnosis and increasing life expectancy [1,2]. Heart transplantation remains the gold standard for end-stage HF patients, despite the limited availability of donors. The use of marginal donors, when carefully matched with recipients, is a viable option [3,4]. This report presents the first case of Candida auris isolation in a heart transplant donor in southern Italy, emphasizing the importance of microbiological surveillance and infection control measures in transplant settings.
## 2 | Case History/Examination
A 62-year-old man with a history of arterial hypertension and smoking habits presented to the emergency room with epigastric pain, asthenia, and hypotension. He was diagnosed with inferior myocardial infarction and underwent emergency angioplasty with stenting.
Despite the intervention, his cardiac function did not improve, and he experienced frequent ventricular tachycardia episodes. Three days postinfarction, he developed ventricular fibrillation, which was resuscitated after cardiac defibrillation. Due to cardiogenic shock, he was transferred to the cardiac surgery unit for veno-arterial extracorporeal membrane oxygenation (V-A ECMO) and emergency heart transplant listing.
## 3 | Differential Diagnosis, Investigations, and Treatment
A donor was identified 4 days later: a 45-year-old woman from Greece with matching anthropometric parameters. Her blood tests showed normal myocardial necrosis indices, an echocardiogram revealed preserved function (EF 60%, TAPSE 22 mm), and coronary angiography was normal.
Although the donor remained afebrile, leukocytosis and increased inflammation indices were noted. Blood cultures were negative, but bronchocultures were positive for Acinetobacter baumannii and Klebsiella pneumoniae. Based on these findings, the national transplant center approved the heart for transplantation but required additional microbiological investigations at the time of organ retrieval. Lungs and other organs were not allocated due to colonization risk and microbiological findings. All centers receiving organs from this donor were promptly informed about the detection of Candida auris to ensure appropriate posttransplant monitoring and infection control protocols.
The recipient underwent orthotopic heart transplant using the standard bicaval technique (ischemic time: 4 h). Postoperatively, he received prophylactic antibiotics (Piperacillin/Tazobactam and Vancomycin) and was initiated on an immunosuppression protocol.
## 4 | Conclusion and Results (Outcome and Follow-Up)
The microbiology laboratory later reported that the donor's blood and urine cultures were negative, but the bronchoaspirate was positive for Candida auris. Antifungal susceptibility testing showed high resistance to fluconazole (> 256 g/mL) and variable susceptibility to other antifungal agents. Identification of Candida auris was performed using matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS; Bruker Daltonics). Colonies were grown on CHROMagar Candida medium and incubated at 37°C for 48 h. Identification was confirmed by the species-specific spectral profile in the Bruker Biotyper database (DB-5989). Although MALDI-TOF MS was used in our laboratory, we acknowledge that sequencing-based methods such as ITS region amplification remain the gold standard for species-level identification of rare yeasts.
Given the isolation of Candida auris in the donor's bronchoaspirate, the recipient was started on antifungal prophylaxis with micafungin (100 mg daily intravenously), initiated immediately after transplantation and continued for 14 days. (Table 1) This choice was based on the antifungal susceptibility profile and current recommendations for echinocandin use in C. auris colonization. No adverse events or breakthrough fungal infections occurred during or after prophylaxis.
Despite donor colonization, strict infection control measures prevented transmission to the recipient. He was placed in isolation and monitored for 15 days with repeated cultures (urine, axillary swabs, wound, and mucocutaneous swabs), all of which remained negative.
Environmental surveillance was conducted in the operating room and ICU, with swabs taken from medical devices and high-touch surfaces, all testing negative for Candida auris.
The patient was later transferred to a regular ward, achieving good functional recovery. Echocardiography confirmed preserved biventricular function and the absence of valve disease. After four endomyocardial biopsies ruling out rejection (ISHLT'04 0R), he was discharged in excellent condition.
## 5 | Discussion
Candida auris, a yeast species first isolated in 2009 in Japan, is known for its multidrug resistance and potential to cause invasive infections, including bloodstream infections with high mortality rates (30%-60%). Most strains exhibit resistance to at least one major antifungal class, with some strains resistant to all three major classes (azoles, echinocandins, and polyenes) [5][6][7].
## Summary
• Candida auris, a multidrug-resistant fungal pathogen, was identified in a heart donor in Southern Italy.
• Rigorous infection control and microbiological surveillance successfully prevented transmission.
• This case highlights the need for systematic donor screening and infection prevention strategies in transplant programs facing emerging fungal threats.
## 6 | Conclusion
This case demonstrates that successful heart transplantation from a Candida auris-colonized donor is possible when rigorous infection prevention and microbiological monitoring protocols are implemented. The absence of transmission to the recipient underscores the effectiveness of strict isolation measures, targeted prophylaxis, and postoperative surveillance.
Given the global emergence of C. auris and its multidrug resistance, this case supports the need to include C. auris in routine microbiological assessments of organ donors, particularly those with risk factors or originating from high-prevalence regions.
Moreover, early detection, thorough environmental decontamination, and clinical vigilance are essential to minimizing transmission risk and ensuring patient safety. Transplant centers should consider developing specific guidelines for managing colonized donors to standardize care in the context of emerging infectious threats.
## References
1. Mcdonagh, Metra, Adamo (2021) "2021 ESC Guidelines for the Diagnosis and Treatment of Acute and Chronic Heart Failure" *European Heart Journal*
2. Savarese, Becher, Lund et al. (2023) "Global Burden of Heart Failure: A Comprehensive Review" *Cardiovascular Research*
3. Bifulco, Bottio, Caraffa (2022) "Marginal Versus Standard Donors in Heart Transplantation" *Journal of Clinical Medicine*
4. Giovannico, Parigino, Ramirez (2024) "World's Oldest Heart Transplant Donor: Age Is Just a Number" *Journal of Cardiovascular Medicine*
5. Satoh, Makimura, Hasumi et al. (2009) "Candida auris sp. Nov" *Microbiology and Immunology*
6. Sticchi, Raso, Ferrara (2023) "Increasing Cases of Candida auris in North Italy" *Journal of Clinical Medicine*
7. Giacobbe, Magnasco, Sepulcri (2021) "Advances in Candida auris Treatment" *Expert Review of Clinical Pharmacology*
8. Magiorakos, Burns, Baño (2017) "Infection Prevention and Control Measures and Tools for the Prevention of Entry of Carbapenem-Resistant Enterobacteriaceae into Healthcare Settings: Guidance from the European Centre for Disease Prevention and Control" *Antimicrobial Resistance and Infection Control* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548462&blobtype=pdf | # Variation in virion phosphatidylserine content drives differential GAS6 binding among closely related flaviviruses
Lizhou Zhang, Byoung-Shik Shim, Claire Kitzmiller, Young-Chan Kwon, Audrey Richard, Michael Farzan, Hyeryun Choe
## Abstract
Many enveloped viruses engage phosphatidylserine (PS) receptors to enter cells, a phenomenon known as "apoptotic mimicry. " We previously reported that Zika virus (ZIKV), but not closely related West Nile virus (WNV) or dengue virus (DENV), utilized AXL to infect cells because only ZIKV could bind the AXL ligand GAS6, a PSbinding protein. In this study, we investigated the mechanisms underlying the differ ential ability of these viruses to bind GAS6. Although immature virions expose larger patches of the viral membrane than do mature ones, our data show that virion maturity levels did not contribute to GAS6 binding. Surprisingly, while ZIKV contains PS compara ble to cellular membranes, PS on WNV and DENV is markedly reduced. These findings explain why only ZIKV can bind GAS6 and provide insights into a novel mechanism by which closely related flaviviruses differentially utilize cellular entry factors. IMPORTANCE Among flaviviruses, Zika virus uniquely causes microcephaly and congenital defects. While no flavivirusspecific entry receptors have been identified, they commonly take advantage of phosphatidylserine (PS) receptors to enter cells. Our previous studies revealed that Zika virus uniquely utilizes AXL, found in immuneprivileged sites, such as the brain and placenta, via binding to its ligand, GAS6. Our current study shows that despite being produced from the same cells, the Zika virus has substantially higher PS content than closely related dengue and West Nile viruses, which likely explains the Zika virus's unique ability to bind GAS6. These findings provide insight into how closely related flaviviruses can vary substantially in their use of cellular entry factors, potentially contributing to the distinct diseases they cause.
animal studies (5)(6)(7)(8)(9)(10)(11)(12) supports a causal relationship between ZIKV and microcephaly, distinguishing it from other flaviviruses.
Flaviviruses replicate in the cytoplasm. Their ~11 kb genome is translated as a single polypeptide and cleaved by cellular and viral proteases into the structural and non-structural proteins. The structural proteins include capsid (C), pre-membrane (prM), and envelope (E), and the non-structural proteins consist of NS1, NS2A, NS2B, NS3, NS4A, NS4B, and NS5. Flavivirus virions assemble on the cytoplasmic side of the expanded endoplasmic reticulum (ER), bud into the ER lumen, and undergo maturation as they transit through the ER and Golgi apparatus. During virion maturation, the prM protein, which is essential for E protein folding and assembly (13), is cleaved into the pr peptide and M protein by the cellular protease furin in the Golgi apparatus (14). The pr peptide remains associated with the virion within the cell but is released when the virion reaches the neutral pH of the extracellular environment (15). Immature virions have a spiky surface because E proteins that are associated with the uncleaved prM proteins stand upright as trimers on the virion surface. In contrast, mature virions are smooth because after the cleaved pr peptides are released, E-proteins form dimers that lie flat on the virion surface. There are 90 such E-protein dimers per virion, and each E protein is paired with an M protein underneath (16). This arrangement completely covers the virion, leaving minimal membrane exposure. However, the dynamic "breathing" motion of viral proteins and membranes (17,18) transiently exposes patches of membrane, particularly at 37°C (19). Compared to mature virions, immature virions expose more of their membrane due to the upright configuration of prM-E trimer spikes (20). In addition to fully mature and immature virions, chimeric or mosaic forms, which are partly mature and immature in patches, have been observed by electron microscopy (21,22). While the mature state is generally required for infectivity, immature virions can still become infectious if pr peptide cleavage occurs by target cell furin during infection (23). Moreover, mosaic virions are infectious even without target-cell furin (24).
Virus entry into the target cell is typically initiated by the viral entry glycoprotein binding to a receptor expressed on the target-cell surface. In the case of flaviviruses, no virusspecific entry receptor has been identified, but many cell-surface molecules are known to facilitate their infection. Best known entry factors include C-type lec tins (namely, DC-SIGN/CD209 and L-SIGN/CD299) and various phosphatidylserine (PS) receptors (25)(26)(27)(28)(29)(30)(31)(32)(33). Lectins recognize carbohydrate moieties on the virion surface, while PS receptors bind PS present in the virion membrane (27,28). The primary physiological function of PS receptors is to bind the PS exposed on the apoptotic cells and mediate their phagocytic clearance, a process known as efferocytosis (34)(35)(36)(37)(38). This process avoids immune activation to prevent responses against self-proteins. Thus, many enveloped viruses exploit these PS receptors to gain entry into cells, and this strategy is known as "apoptotic mimicry" (39,40). Among PS receptors, the T-cell immunoglobulin mucin (TIM) family (TIM1, TIM3, and TIM4) and the TAM family (TYRO3, AXL, and MERTK) receptors play key roles in flavivirus infection (26-28, 31-33, 40). Of the three human TIM-family members (TIM1, 3, and 4), TIM4 is expressed on professional phagocytes, such as dendritic cells and macrophage sub-populations (36,41). In contrast, TIM1 is induced to be expressed on non-professional phagocytes, namely, epithelial cells in tissues, such as the lung, kidney, mammary gland, and placenta, where it clears neighboring cells when they undergo apoptosis (42)(43)(44)(45)(46)(47)(48)(49)(50). The TAM-family members are broadly expressed across various tissues, particularly in regions subject to continuous challenge and renewal. They are expressed in the endothelium and immune-privileged sites, such as the brain, eye, placenta, and testis, where they facilitate apoptotic cell clearance while maintaining the integrity of vasculature and barrier cells of those sensitive areas (9,47,(51)(52)(53)(54)(55).
While TIM-family receptors directly bind PS, TAM family receptors engage PS indirectly through soluble mediators, specifically, growth arrest specific gene 6 (GAS6) and Protein S (ProS), which are present in serum and other bodily fluids (56,57). GAS6 serves as a ligand for all three TAM receptors, whereas ProS binds only to TYRO3 and MERTK (57). Consequently, GAS6 is the sole ligand for AXL. Although PS is the primary target for these PS-binding receptors, a subset of them, such as human TIM1, MFGE8, and Annexin A5, also bind phosphatidylethanolamine (PE) with equal or greater efficiency compared to PS (28,(58)(59)(60). Like PS, PE is typically restricted to the cytoplasmic leaflet of the membrane lipid bilayer and flips to the extracellular side during apoptosis (61), which also facilitates efferocytosis and virus entry (28).
GAS6/AXL-mediated TAM receptor usage by enveloped viruses was reported for the first time by Morizono et al. (32) using lentiviral vectors pseudotyped with alphavirus glycoproteins derived from Sindbis or Ross River virus. We also made similar observa tions: retroviral vectors pseudotyped with glycoproteins from filoviruses, arenaviruses, or alphaviruses showed markedly enhanced entry mediated by AXL (26). Surprisingly, in the same study, we found that WNV virus-like particle (VLP) did not use AXL, although it efficiently used TIM1 (26). In fact, among all TIM1-using viruses, WNV VLP was the only one that failed to utilize AXL. This was the first time we noticed WNV's inability to engage AXL. Later, we observed that DENV, like WNV, also failed to use AXL and demonstrated that such differential use of AXL was due to ZIKV's unique ability to bind GAS6, while WNV and DENV do not (31). The mechanism behind this intriguing observation remains unclear. In the current study, we investigated the differential ability of these viruses to bind GAS6 and found that WNV and DENV contain significantly lower PS levels in their virions. This low PS content in WNV and DENV virions does not support their interaction with GAS6, whereas the higher PS content in ZIKV facilitates it.
## RESULTS
## ZIKV, but not WNV or DENV, binds GAS6
We have previously reported that ZIKV, but not WNV or DENV, efficiently uses AXL and that such differential AXL use is due to the unique ability of ZIKV to bind GAS6: e.g., WNV and DENV do not bind GAS6 (31). However, the underlying mechanism remains unknown. To investigate the mechanism behind this unexpected phenomenon, we first confirm the reproducibility of our previous observations using viruses produced in human cell lines, because in our previous study, the viruses were grown in the monkey cell line, Vero (31). In this study, viruses were propagated in A549, a human lung epithelial cell line, or in Huh7, a human hepatic epithelial cell line. To measure GAS6 binding, these viruses were incubated with the Ig-fusion form of GAS6 (GAS6-Ig), and captured virus particles were precipitated using Protein A-Sepharose beads. These captured viruses were then visualized by western blots using anti-E protein antibodies or quantified by reverse transcription-quantitative PCR (RT-qPCR) of viral RNA. Ig-fusion forms of the C-terminal half of GAS6 (C-GAS6-Ig), which binds AXL but does not bind PS, and TIM1-Ig, which binds all three viruses, were included as a negative control and positive control, respectively (28,31). The reason TIM1-Ig could bind all three viruses, although it is a PS-binding protein, like GAS6, is that it is able to bind PE in addition to PS, and because the PE levels are typically much higher than those of PS (28). Figure 1 shows that ZIKV, but not WNV or DENV, bound GAS6-Ig as shown by both WB (Fig. 1A) and RT-qPCR (Fig. 1B). The results obtained from the viruses produced in A549 and Huh7 are nearly identical. As expected, C-GAS6-Ig did not bind any virus, and TIM1-Ig bound all viruses. These data confirm, using live viruses produced from human cell lines, our previous report that ZIKV, but not WNV or DENV, can bind GAS6 and utilize AXL as an entry factor (31).
## Virion maturity does not alter GAS6 binding
To explore the mechanism with which closely-related flaviviruses differentially bind GAS6, we first focused on virion maturity, because immature virions expose more membrane than do mature viruses, owing to different E protein arrangement on the virion surface (20). We therefore reduced WNV virion maturity to promote binding of GAS6 by immature WNV. To do so, we interfered with WNV maturation by treating infected cells with a furin inhibitor, Decanoyl-RVKR-CMK. This treatment reduces virion maturation by inhibiting the cleavage between the pr peptide and M protein. Vero cells treated with Decanoyl-RVKR-CMK produced an immature virion population, as indicated by the uncleaved prM band, while untreated Vero cells produced mature WNV virions (Fig. 2A, left panel). WNV so treated still did not bind GAS6-Ig (Fig. 2A, right panel), while it bound TIM1-Ig. We then tried an alternative method to produce immature virions. Based on our previous experience, flaviviruses grown at a lower temperature exhibit lower maturity. Accordingly, cells infected with WNV were cultured at 37°C or 28°C. As expected, WNV virions produced at 28°C were less mature than those produced at 37°C, demonstrated by the uncleaved prM band detected only in the WNV grown at 28°C (Fig. 2B, left panel). However, this 28°C-produced WNV again did not bind GAS6-Ig, while it did bind TIM1-Ig (Fig. 2B, right panel). Finally, assuming the WNV produced under these two conditions may not be sufficiently immature, we produced completely immature WNV from LoVo cells. This human colon carcinoma cell line lacks functional furin due to a frameshift mutation (62). WNV produced from LoVo cells was therefore fully immature, as indicated by the absence of the cleaved M protein (Fig. 2C, left panel). Nonetheless, this completely immature WNV still did not bind GAS6-Ig (Fig. 2C, right panel). These data demonstrate that virion maturity is unlikely to contribute to GAS6 binding.
## N-glycans on the E protein do not affect GAS6 binding
We next investigated whether N-linked glycans on the virion contribute to GAS6 binding. N-glycans on viral proteins, including flavivirus E proteins, play more than just a shielding role against immunity. They have been linked to enhanced virus production in mosquito or avian hosts, tissue tropism, and an increased ability for neuroinvasion (63)(64)(65)(66). Lack of N-glycosylation has also been linked to attenuated phenotype or reduced virulence (67). Although GAS6 has not so far been shown to directly bind carbohydrates, we nonetheless sought to investigate N-glycans' contribution to differential GAS6 binding because N-glycans on the virion surface could sterically hinder GAS6 binding to the virion membrane. Therefore, to study their effect, N-linked glycans on the virions were removed by incubating WNV with PNGase F. As Fig. 2D shows, PNGase F treatment reduced the apparent size of WNV E protein as expected from the removal of N-glycans. However, PNGase F treatment did not make WNV bind GAS6-Ig (Fig. 2E), indicating that Full-Length Text the different number of N-glycans on the three viruses does not account for differential GAS6 binding by these viruses.
## Virion phospholipids, not viral proteins, mediate ZIKV binding to GAS6
To confirm that capturing ZIKV by GAS6-Ig in our assays is mediated by PS, not by proteins, on the virion, we treated the ZIKV-GAS6-Ig complex with phospholipase C (PLC) after the virions were captured by GAS6-Ig. PLC cleaves the bond between phosphate and glycerol in all phospholipids, removing their headgroup from the diacylglycerol backbone (Fig. 3A). Therefore, if ZIKV binds GAS6 through PS, PLC treatment should dissociate the virions from GAS6-Ig (Fig. 3B), but if ZIKV-GAS6 binding occurs through viral proteins, PLC treatment will not make any difference. As shown in Fig. 3C (left panel vs right panel), PLC digestion removed a majority of ZIKV bound to GAS6 or TIM1. Similarly, pre-treating ZIKV with PLC before its binding to GAS6-Ig or TIM1-Ig yielded similar results (Fig. 3C, middle panel). Virus binding to TIM1 is reduced or eliminated by PLC digestion, because PLC cleaves the head group of all phospholipids. Quantification by RT-qPCR of captured ZIKV digested by PLC over a range of concentrations produced similar results (Fig. 3D). These findings confirm that ZIKV binding to GAS6 is mediated by phospholipids rather than viral proteins.
## PS content of ZIKV is substantially higher than that of WNV and DENV
Having confirmed that ZIKV binds GAS6 via phospholipids in the virion membrane, we next investigated whether the phospholipid content of DENV, WNV, and ZIKV is different. Initially, we did not pursue this hypothesis because it seemed the least likely, as the membranes of enveloped viruses are generally believed to derive from cellular membranes. In addition, all three viruses were grown in the same cell line, and all of them bud from the same organelle, the ER. Nevertheless, in the absence of a clear mechanistic explanation, we compared the phospholipid contents of purified ZIKV, WNV, and DENV grown in Vero cells, using two-dimensional thin-layer chromatography (2D TLC). Before analyzing viral phospholipids, we verified the positions of the four major phospholipids-phosphatidylcholine (PC), PE, PS, and phosphatidylinositol (PI)on the 2D TLC plate. Although sphingomyelin (SPH) is not a phospholipid, it contains a phosphate group and thus is stainable by iodine, so it was included in the assay. These lipids were purchased from Avanti Polar Lipids, mixed in a ratio (PC:PE:PS:SPH:PI = 10:5:2:2:1) roughly comparable to that of mammalian cell membrane (68)(69)(70), separated by 2D TLC on a silica plate, and stained with iodine vapor. The location of each of these phospholipids is indicated in the upper panel of Fig. 4A. We then analyzed the lipids extracted from the ER membrane of Vero cells in the same way and compared the results to those of the commercial phospholipids. The five lipids derived from Vero ER membrane are located at the same positions as the commercial ones (Fig. 4A, lower panel). Commercial lipids formed more compact spots compared to those from the Vero ER membrane, as they contain homogenous fatty acyl chains (we purchased those with dioleic acid), whereas the fatty acyl chains of cellular lipids are heterogeneous in length.
Having analyzed phospholipids of Vero ER membrane, we then analyzed the lipids from purified ZIKV, WNV, and DENV. These viruses were propagated in Vero cells and purified using potassium tartrate gradient centrifugation. Their purity was assessed by Coomassie blue staining of viral proteins after they were separated by electropho resis. Figure S1 shows that these virus preparations were highly pure. Total lipids were extracted from these viruses and analyzed by 2D TLC (Fig. 4B). One immediately noticeable difference is that the PS spots of WNV and DENV are barely visible, while those of ZIKV and Vero ER membrane are clearly visible with comparable intensity. When quantified (Fig. 4C; Fig. S2), the PS content of ZIKV (5.8%) was comparable to that of Vero ER (6.7%), while the PS contents of WNV and DENV were much lower (1.3% and 0.9%, respectively). ZIKV has significantly higher PS content (4.5-to 6.4-fold) compared to WNV and DENV (Fig. 4D). PC and PE contents of all viruses were roughly comparable to those of Vero ER, although PE of WNV (22%) is modestly lower than that of ZIKV and DENV (25%-28%) and Vero ER (27%). Interestingly, as Fig. 4C shows, the PI level of all three viruses was considerably higher (13.5%-17.3%) than that of the Vero ER membrane (9.2%). However, since it increased in all three viruses, PI cannot be the factor responsible for the differential GAS6 binding by the viruses. Because PI is also negatively charged like PS and because a subset of PS-binding proteins, such as TIM1 and MFGE8, bind additional phospholipid PE (28,58), we nonetheless analyzed the binding profile of GAS6 to each phospholipid to verify that PS, but not other phospholipids, is responsible for GAS6 binding. We conducted lipid enzyme-linked immunosorbent assay (ELISA) with PC, PE, PS, and PI. As Fig. 4E shows, GAS6 binds only PS, while TIM1 binds both PE and PS as we reported previously (31). Although PI weakly binds GAS6 and TIM1 at high concentrations, it cannot explain why WNV and DENV do not bind GAS6, as PI is modestly more abundant in WNV and DENV than in ZIKV (Fig. 4C). These data show that the PS content of ZIKV is higher than that of WNV or DENV and correlates with ZIKV's ability to bind GAS6.
## PL composition of the ER membrane is not differentially altered by different viruses
To further investigate the mechanism by which ZIKV, WNV, and DENV differentially incorporate PS into their virion, we first examined the changes in the ER lipidome induced by virus infection. The ER lipids were extracted when the majority of cells were infected, which was 2 days after infection for WNV and ZIKV and 6 days for DENV. The viruses used in this study were also harvested 2 days (WNV and ZIKV) or 6 days (DENV) after infection. The ER lipids from the infected Vero cells were compared to those extracted from uninfected cells. As Fig. 5A and Fig. S3 show, no substantial alteration in the phospholipid composition of the ER membrane was induced by virus infection or by different viruses. Moderately higher PS level was observed in the cells infected with WNV or DENV, compared to uninfected cells or those infected with ZIKV (Fig. 5B andC), which cannot explain the low PS content observed with WNV and DENV. These results demonstrate that virus infection does not significantly alter the ER lipidome, at least for the phospholipid ratio, and thus cannot explain the apparent difference in PS content observed among the three viruses. These data also suggest that viruses themselves are likely responsible for differential PS incorporation into their virion.
## DISCUSSION
Many enveloped viruses, including flaviviruses, exploit PS receptors to gain entry to target cells. However, not all PS receptors can be utilized equally well. For instance, as we previously reported, ZIKV, but not WNV or DENV, can use AXL, as ZIKV is the only one capable of binding GAS6 (31). Because all PS receptors recognize PS in the viral membrane rather than viral proteins, and because the membrane of enveloped viruses is derived from the host cell membrane, the differential use of PS receptors by closely related viruses is puzzling, particularly when those viruses are grown in the same cells. In the current study, we found that the PS content of the virion membrane varies among these viruses, even when produced in the same cells. The higher PS content of ZIKV compared to WNV and DENV explains why only ZIKV can bind GAS6. This finding reveals a previously unrecognized mechanism by which lipid composition of the virion membrane drives receptor usage. Although there are multiple lipidomic studies on virus-infected cells, only a few have focused on virion lipid composition (71)(72)(73)(74)(75). Two studies report PS contents of nonflaviviral virions: while HIV-1 contains higher PS than cellular membranes (71), HCV shows no detectable PS (74). One study on flavivirus found that WNV grown in HeLa cells had higher PS compared to the host cell membrane (75), which contradicts our findings. This discrepancy may, in part, be due to differences in analytical methods: this lipidomic study used mass spectrometry, whereas we used 2D TLC. While mass spectrometry is a powerful tool for many applications, it is less reliable for quantifying complex lipids by class due to variable ionization efficiencies of different molecular species (e.g., those sharing the same head group but differing in fatty acyl chains) and the limited availability of comprehensive internal standards. In contrast, 2D TLC directly separates and visualizes lipids by class without being affected by these factors.
Searching for a potential explanation for why WNV and DENV cannot bind GAS6, we initially investigated virion maturity. In mature flaviviruses, the virion surface is com pletely covered with E protein homodimers arranged in a herringbone pattern, leaving little membrane exposed (16,76,77). However, even in the mature virion, patches of virion membrane are transiently exposed, especially at 37°C (19), because virion shells are in a continuous dynamic motion (21,78), which is described as "breathing" (22). Immature virions, on the other hand, expose larger portions of their membrane because of the standing configuration of prM-E heterotrimers (21). While these immature virions can bind PS receptors, their infectivity remains unclear. Mosaic virions, which are partially immature and partially mature (21,22,79,80), have both immature regions capable of binding PS receptors and mature regions that can mediate infection. Therefore, if the ZIKV population contains a higher proportion of mosaic or immature virions compared to WNV and DENV, this could explain why only ZIKV binds GAS6. However, our data show that WNV engineered to be immature still does not bind GAS6, indicating that virion maturity does not contribute to GAS6 binding. Rather, we found that WNV and DENV have barely detectable levels of PS in their virion, while ZIKV has PS levels comparable to those of the ER membrane of the Vero cells in which the viruses were propagated. Because WNV and DENV have low PS content, even when WNV was made more immature, it could not bind GAS6.
We hypothesized that at least three different scenarios could explain lower PS content in WNV and DENV compared to ZIKV and Vero ER. First, infection of cells by different flaviviruses differentially alters the cellular lipidome, especially in the expanded ER, where flaviviruses replicate and assemble. We investigated this possibility and found that this was not the case. PL composition of the ER membrane did not notably change following infection by any of the three viruses. We deduce from these results that the viruses themselves or virus replication-assembly processes must be responsible for their distinct virion phospholipid compositions. One possible mechanism for the distinct virion PS content could be the location of virus budding. Flavivirus infection of cells markedly increases the synthesis of lipids, including phospholipids, to expand the ER membrane, where virus replication takes place. Although all flaviviruses replicate inside the pockets formed on the expanded ER membrane and bud into the ER lumen, it is unknown whether all three viruses form replication pockets at similar locations on the expanded ER membrane.
Although the PS content of the ER membrane is known to be approximately 4%-5% (70,81,82), it does not mean PS is evenly distributed in the ER. Rather, differ ent compartments of the ER membrane are believed to have a distinct phospholipid composition. For example, two major enzymes synthesizing PS-Phosphatidyl Synthase 1 (PSS1) that synthesizes PS from PC and Phosphatidyl Synthase 2 (PSS2) that produces PS from PE-are located in the mitochondria-associated membrane (MAM) in mammals (83) and synthesize the majority of cellular PS. Much of the PS produced in this location is funneled to the mitochondria and converted to PE by phosphatidylserine decarboxylase that resides inside the mitochondria (83), and PE produced therein is transported back to the ER through MAM and converted to PS and PC. Thus, MAM, a subdomain of ER that connects ER to the nearby mitochondria (84), is enriched with PS and PE. Therefore, although all three viruses bud from the ER membrane, it is possible that ZIKV forms replication sites closer to MAM than do DENV or WNV, receiving a higher PS level than do DENV and WNV.
In summary, our data demonstrate that the PS content of the closely related ZIKV, DENV, and WNV viruses is quite different and that higher PS content is linked to ZIKV structural proteins.
## MATERIALS AND METHODS
## Cell lines
HEK293T (human embryonic kidney), Huh-7 (human hepatoma), and Vero (monkey kidney) cells were grown in high-glucose DMEM, and A549 (human lung) and LoVo (human colon) cells were grown in F-12K medium. All cells were cultured in medium supplemented with 10% FBS, at 37°C with 5% CO 2 .
## Virus production
ZIKV Brazilian strain PB-81 was obtained from the World Reference Center for Emerging Viruses and Arboviruses (WRCEVA) at the University of Texas Medical Branch (UTMB), DENV type 2 New Guinea C strain was obtained from ATCC (VR-1584), and WNV lineage I New York 1999 (NY99) strain was kindly provided by R. Tesh, UTMB, Galveston, TX. All these flaviviruses were propagated in Vero cells at 37°C in DMEM supplemented with 10% FBS, and the culture supernatants containing viruses were clarified using 0.45 um filters. Flaviviruses were also produced from Huh7, A549, and LoVo cell lines in the respective cell culture medium and filtered. To assess virus titers, viral RNA was extracted and quantified by RT-qPCR as described below, using the primers and probe listed in Table 1. Aliquoted viruses were kept at -80°C. Although not required for ZIKV and DENV, all live virus experiments were conducted side by side with WNV in the BSL3 facility at UF Scripps Institute for Biomedical Research, with approval from the Institutional Biosafety Committee.
## Immature virus production
Immature WNV particles were produced by culturing WNV-infected Vero cells in DMEM containing 10% FBS and 50 uM Decanoyl-RVKR-CMK (a furin inhibitor), or by culturing the cells at 28°C without the furin inhibitor, or growing WNV in the LoVo cell line that has a mutation in the furin gene (85). Virus was harvested from Vero cells 2 days later, regardless of culturing temperature, and 3 days later from LoVo cells. Viruses were quantified by RT-qPCR as described below, and the immature status of WNV virions 2 mL 24% sucrose cushion, and centrifuged in an SW41 rotor at 175,000 × g for 2 h at 10°C. Pelleted viruses were resuspended in 1 mL NTE buffer and subjected to a linear 10%-35% (wt/vol) potassium tartrate gradient centrifugation at 175,000 × g for 2 h at 10°C. The virus band, located approximately at 20% in the potassium tartrate gradient, was harvested and pelleted by centrifugation at 175,000 × g for 2 h at 10°C. Purified viruses were resuspended in NT buffer (120 mM NaCl, 20 mM Tris-HCl, pH 8.0) and stored at -80°C.
## Endoplasmic reticulum isolation
Total ER (both rough and smooth ER) was isolated from the mock-and virus-infected Vero cells using the ER Enrichment Extraction kit (Novus Biologicals, Cat# NBP2-29482) by following the manufacturer's instructions. Briefly, three T175 flasks of over 90% confluent uninfected Vero cells or those infected with viruses were harvested at day 2 post-WNV and ZIKV infection or day 6 post-DENV infection. Cells were washed once with PBS, pelleted at 500 × g for 5 min, and then resuspended in 1× isosmotic homoge nization buffer supplemented with 1× protease inhibitor cocktail (PIC) provided by the kit. These cells were homogenized on ice in a clean PTFE tissue grinder (Cole-Parmer, Cat# UX-44468-02). Homogenates were centrifuged at 1,000 × g for 10 min at 4°C to eliminate pelleted nuclei and cell debris, followed by another centrifugation at 12,000 × g for 15 min at 4°C to further remove cell debris and mitochondria. Then, supernatants were transferred to ultracentrifuge tubes, centrifuged at 130,000 × g for 1 h at 4°C, and resuspended in 200 µL suspension buffer in the kit containing 1× PIC from the kit.
## Lipid extraction from the ER membrane or purified viruses
Total lipids were extracted from purified flavivirus particles or isolated ER using the Bligh and Dyer method (87). For comparing virion lipids, lipids were extracted from 1 × 10¹³ genome copies of ZIKV, DENV, or WNV particles. For ER lipids, an isolated ER equivalent of 1 mg of total protein was used for lipid extraction. Specifically, 1.5 mL of methanol and chloroform mixed at a 2:1 ratio was added to a viral or ER sample diluted in 0.4 mL PBS in a glass vial. The mixture was vortexed briefly and allowed to sit for 5 min at room temperature. 0.5 mL of chloroform was added to it, and the solution was vortexed for 30 s. A volume of 0.5 mL of double-distilled H 2 O (ddH 2 O) was then added, and the solution was vortexed again before being centrifuged at 250 × g in a GH-3.8 rotor for 5 min at room temperature. The bottom chloroform phase containing the lipids was carefully collected with a Pasteur pipette and transferred to a new glass vial. An equal volume of a 2:2:1 mixture of methanol, chloroform, and ddH 2 O was added, and the extraction procedure was repeated. After centrifugation, the bottom phase was collected into a 1.5 mL Eppendorf tube and dried overnight in a fume hood.
## Two-dimensional thin-layer chromatography
For phospholipid analysis by 2D-TLC, dried viral or ER lipids were resuspended in 100 µL chloroform and loaded onto a TLC plate coated with silica gel 60 F 254 (0.5 mm, 20 × 20 cm, Millipore, Cat# 1.05744.001) at a spot 2 cm away from either edge. For a sharper separation, a small amount of sample was loaded and dried completely before another aliquot was loaded. To verify the position of each lipid, a mixture containing synthetic DOPC, DOPE, DOPS, DOPI, and SPH was separated by 2D-TLC in the same way. For firstdimensional separation, the plate with loaded sample was placed in a sealed TLC chamber for 90 min with a solvent composed of chloroform, methanol, ammonium hydroxide, and water at a ratio of 13:5:1:19. The plate was air-dried for 30 min and run for the second dimension for 90 min in a solvent containing chloroform, acetone, methanol, acetic acid, and water at a ratio of 10:4:2:2:1. After drying, the plate was stained with iodine vapor produced from iodine crystals heated at 40°C in a sealed chamber placed inside a fume hood. The plate was removed from the iodine chamber, placed in a fume hood, and allowed iodine to evaporate until the background yellow color dissipated and phospholipid spots became clear. During this process, images were captured every 2 min in a ChemiDoc imager (Bio-Rad), and the intensity of phospholipid spots was analyzed using the Image Lab software (Bio-Rad). Several images of varying intensity were analyzed and averaged per plate.
## Statistical analysis
All data were analyzed with GraphPad Prism version 9.0 (GraphPad Software Inc.) and expressed as mean ± standard deviation(SD). Statistical significance among the groups was analyzed by one-way analysis of variance (ANOVA) with Tukey's multiple compari sons test or two-way ANOVA (ns, not significant; *P < 0.05, **P < 0.01, ***P < 0.005, and ****P < 0.005).
## References
1. Swanson, Ii, Db (2015) "Viral diseases of the central nervous system" *Curr Opin Virol*
2. Winkelmann, Luo, Wang (2016) "West Nile virus infection in the central nervous system"
3. Julander, Siddharthan, Blatt et al. (2007) "Effect of exogenous interferon and an interferon inducer on western equine encephalitis virus disease in a hamster model" *Virology (Auckl)*
4. O'leary, Kuhn, Kniss et al. (2006) "Birth outcomes following West Nile Virus infection of pregnant women in the United States: 2003-2004" *Pediatrics*
5. Cugola, Fernandes, Russo et al. (2016) "The Brazilian Zika virus strain causes birth defects in experimental models" *Nature*
6. Mlakar, Korva, Tul et al. (2016) "Zika virus associated with microce phaly" *N Engl J Med*
7. Hirsch, Roberts, Grigsby et al. (2018) "Zika virus infection in pregnant rhesus macaques causes placental dysfunction and immunopathology" *Nat Commun*
8. Nguyen, Antony, Dudley et al. (2017) "Highly efficient maternal-fetal Zika virus transmission in pregnant rhesus macaques" *PLoS Pathog*
9. Miner, Cao, Govero et al. (2016) "Zika virus infection during pregnancy in mice causes placental damage and fetal demise" *Cell*
10. Calvet, Aguiar, Melo et al. (2016) "Detection and sequencing of Zika virus from amniotic fluid of fetuses with microcephaly in Brazil: a case study" *Lancet Infect Dis*
11. Dudley, Aliota, Mohr et al. (2016) "A rhesus macaque model of Asian-lineage Zika virus infection" *Nat Commun*
12. Martinot, Abbink, Afacan et al. (2018) "Fetal neuropathology in Zika virus-infected pregnant female rhesus monkeys"
13. Lorenz, Allison, Heinz et al. (2002) "Folding and dimeriza tion of tick-borne encephalitis virus envelope proteins prM and E in the endoplasmic reticulum" *J Virol*
14. Stadler, Allison, Schalich et al. (1997) "Proteolytic activation of tick-borne encephalitis virus by furin" *J Virol*
15. Yu, Zhang, Holdaway et al. (2008) "Structure of the immature dengue virus at low pH primes proteolytic maturation" *Science*
16. Kuhn, Zhang, Rossmann et al. (2002) "Structure of dengue virus: implications for flavivirus organization, maturation, and fusion" *Cell*
17. Dowd, Jost, Durbin et al. (2011) "A dynamic landscape for antibody binding modulates antibody-mediated neutralization of West Nile virus" *PLoS Pathog*
18. Sabo, Luca, Ray et al. (2012) "Hepatitis C virus epitope exposure and neutralization by antibodies is affected by time and temperature" *Virology (Auckl)*
19. Fibriansah, Ibarra, Ng et al. (2015) "Cryo-EM structure of an antibody that neutralizes dengue virus type 2 by locking E protein dimers" *Science*
20. Pierson, Diamond (2012) "Degrees of maturity: the complex structure and biology of flaviviruses" *Curr Opin Virol*
21. Cherrier, Kaufmann, Nybakken et al. (2009) "Structural basis for the Full-Length Text Journal of Virology October"
22. "preferential recognition of immature flaviviruses by a fusion-loop antibody" *EMBO J*
23. Plevka, Battisti, Junjhon et al. (2011) "Maturation of flaviviruses starts from one or more icosahedrally independent nucleation centres" *EMBO Reports*
24. Rodenhuis-Zybert, Van Der Schaar, Da et al. (2010) "Immature dengue virus: a veiled pathogen?" *PLoS Pathog*
25. Mukherjee, Lin, Dowd et al. (2011) "The infectivity of prM-containing partially mature West Nile virus does not require the activity of cellular furin-like proteases" *J Virol*
26. Hamel, Dejarnac, Wichit et al. (2015) "Biology of Zika virus infection in human skin cells" *J Virol*
27. Jemielity, Wang, Chan et al. (2013) "TIM-family proteins promote infection of multiple enveloped viruses through virion-associated phosphatidylserine" *PLoS Pathog*
28. Meertens, Carnec, Lecoin et al. (2012) "The TIM and TAM families of phosphatidylserine receptors mediate dengue virus entry" *Cell Host Microbe*
29. Richard, Zhang, Park et al. (2015) "Virionassociated phosphatidylethanolamine promotes TIM1-mediated infection by Ebola, dengue, and West Nile viruses" *Proc Natl Acad Sci*
30. Davis, Nguyen, Hanna et al. (2006) "West Nile virus discriminates between DC-SIGN and DC-SIGNR for cellular attachment and infection" *J Virol*
32. Tassaneetrithep, Burgess, Granelli-Piperno et al. (2003) "DC-SIGN (CD209) mediates dengue virus infection of human dendritic cells" *J Exp Med*
33. Richard, Shim, Kwon et al. (2017) "AXL-dependent infection of human fetal endothelial cells distinguishes Zika virus from other pathogenic flaviviruses" *Proc Natl Acad Sci*
34. Morizono, Xie, Olafsen et al. (2011) "The soluble serum protein Gas6 bridges virion envelope phosphatidyl serine to the TAM receptor tyrosine kinase Axl to mediate viral entry" *Cell Host Microbe*
35. Morizono, Chen (2014) "Role of phosphatidylserine receptors in enveloped virus infection" *J Virol*
36. Bosurgi, Bernink, Cuevas et al. (2013) "Paradoxical role of the protooncogene Axl and Mer receptor tyrosine kinases in colon cancer" *Proc Natl Acad Sci*
37. Dekruyff, Bu, Ballesteros et al. (2010) "T cell/transmembrane, Ig, and mucin-3 allelic variants differentially recognize phosphatidylserine and mediate phagocytosis of apoptotic cells" *J Immunol*
38. Kobayashi, Karisola, Peña-Cruz et al. (2007) "TIM-1 and TIM-4 glycoproteins bind phosphatidylserine and mediate uptake of apoptotic cells" *Immunity*
39. Kuchroo, Umetsu, Dekruyff et al. (2003) "The TIM gene family: emerging roles in immunity and disease" *Nat Rev Immunol*
40. Miyanishi, Tada, Koike et al. (2007) "Identification of Tim4 as a phosphatidylserine receptor" *Nature*
41. Mercer, Helenius (2008) "Vaccinia virus uses macropinocytosis and apoptotic mimicry to enter host cells" *Science*
42. Amara, Mercer (2015) "Viral apoptotic mimicry" *Nat Rev Microbiol*
43. Meyers, Chakravarti, Schlesinger et al. (2005) "TIM-4 is the ligand for TIM-1, and the TIM-1-TIM-4 interaction regulates T cell proliferation" *Nat Immunol*
44. Brooks, Yeung, Brooks et al. (2015) "KIM-1-/TIM-1-mediated phagocytosis links ATG5-/ ULK1-dependent clearance of apoptotic cells to antigen presentation" *EMBO J*
45. Ichimura, Asseldonk, Humphreys et al. (2008) "Kidney injury molecule-1 is a phosphatidylserine receptor that confers a phagocytic phenotype on epithelial cells" *J Clin Invest*
46. Ichimura, Bonventre, Bailly et al. (1998) "Kidney injury molecule-1 (KIM-1), a putative epithelial cell adhesion molecule containing a novel immunoglobulin domain, is upregulated in renal cells after injury" *J Biol Chem*
47. Sandahl, Hunter, Strunk et al. (2010) "Epithelial celldirected efferocytosis in the post-partum mammary gland is necessary for tissue homeostasis and future lactation" *BMC Dev Biol*
48. Sizing, Bailly, Mccoon et al. (2007) "Epitope-dependent effect of anti-murine TIM-1 monoclonal antibodies on T cell activity and lung immune responses" *J Immunol*
49. Tabata, Petitt, Puerta-Guardo et al. (2016) "Zika virus targets different primary human placental cells, suggesting two routes for vertical transmission" *Cell Host Microbe*
50. Umetsu, Dekruyff (2004) "Regulation of tolerance in the respiratory tract: TIM-1, hygiene, and the environment" *Ann N Y Acad Sci*
51. Yang, Brooks, Xiao et al. (2015) "KIM-1-mediated phagocytosis reduces acute injury to the kidney" *J Clin Invest*
53. Arai, Kitada, Yamazaki et al. (2016) "Apoptosis inhibitor of macrophage protein enhances intraluminal debris clearance and ameliorates acute kidney injury in mice" *Nat Med*
54. Deng, Chen, Han (2016) "The roles of TAM receptor tyrosine kinases in the mammalian testis and immunoprivileged sites" *Front Biosci (Landmark Ed)*
55. Dransfield, Zagórska, Lew et al. (2015) "Mer receptor tyrosine kinase mediates both tethering and phagocytosis of apoptotic cells" *Cell Death Dis*
56. Prasad, Rothlin, Burrola et al. (2006) "TAM receptor function in the retinal pigment epithelium" *Mol Cell Neurosci*
57. Xiong, Chen, Wang et al. (2008) "Gas6 and the Tyro 3 receptor tyrosine kinase subfamily regulate the phagocytic function of Sertoli cells" *Reproduction*
58. Melaragno, Wuthrich, Poppa et al. (1998) "Increased expression of Axl tyrosine kinase after vascular injury and regulation by G protein-coupled receptor agonists in rats" *Circ Res*
59. (2025) *Full-Length Text Journal of Virology*
60. Nagata, Ohashi, Nakano et al. (1996) "Identification of the product of growth arrestspecific gene 6 as a common ligand for Axl, Sky, and Mer receptor tyrosine kinases" *J Biol Chem*
61. Stitt, Conn, Goret et al. (1995) "The anticoagulation factor protein S and its relative, Gas6, are ligands for the Tyro 3/Axl family of receptor tyrosine kinases" *Cell*
62. Hanayama, Tanaka, Miwa et al. (2002) "Identification of a factor that links apoptotic cells to phagocytes" *Nature*
63. Stuart, Reutelingsperger, Frederik (1998) "Binding of annexin V to bilayers with various phospholipid compositions using glass beads in a flow cytometer" *Cytometry*
64. Swairjo, No, Kaetzel et al. (1995) "Ca 2+bridging mechanism and phospholipid head group recognition in the membrane-binding protein annexin V" *Nat Struct Biol*
65. Emoto, Toyama-Sorimachi, Karasuyama et al. (1997) "Exposure of phosphatidylethanolamine on the surface of apoptotic cells" *Exp Cell Res*
66. Takahashi, Nakagawa, Kasai et al. (1995) "A second mutant allele of furin in the processing-incompetent cell line, LoVo. Evidence for involvement of the homo B domain in autocatalytic activation" *J Biol Chem*
67. Yoshii, Yanagihara, Ishizuka et al. (2013) "N-linked glycan in tick-borne encephalitis virus envelope protein affects viral secretion in mammalian cells, but not in tick cells" *J Gen Virol*
68. Jiang, Lowe, Higgs et al. (1993) "Single amino acid codon changes detected in louping ill virus antibody-resistant mutants with reduced neurovirulence" *J Gen Virol*
69. Yan, Shi, Wang et al. (2018) "A single mutation at position 156 in the envelope protein of Tembusu virus is responsible for virus tissue tropism and transmissibility in ducks" *J Virol*
70. Wen, Li, Dong et al. (2018) "Nglycosylation of viral E protein is the determinant for vector midgut invasion by flaviviruses" *mBio*
71. Prow, May, Westlake et al. (2011) "Determinants of attenuation in the envelope protein of the flavivirus Alfuy" *J Gen Virol*
72. Vance (2015) "Phospholipid synthesis and transport in mammalian cells" *Traffic*
73. Renkonen, Gahmberg, Simons et al. (1972) "The lipids of the plasma membranes and endoplasmic reticulum from cultured baby hamster kidney cells (BHK21)" *Biochim Biophys Acta*
74. Renkonen, Kääriäinen, Gahmberg et al. (1972) "Lipids of Semliki Forest virus and of host cell membranes" *Biochem Soc Symp*
75. Brügger, Glass, Haberkant et al. (2006) "The HIV lipidome: a raft with an unusual composition" *Proc Natl Acad Sci*
76. Kalvodova, Sampaio, Cordo et al. (2009) "The lipidomes of vesicular stomatitis virus, semliki forest virus, and the host plasma membrane analyzed by quantitative shotgun mass spectrometry" *J Virol*
77. Gerl, Sampaio, Urban et al. (2012) "Quantitative analysis of the lipidomes of the influenza virus envelope and MDCK cell apical membrane" *J Cell Biol*
78. Merz, Long, Hiet et al. (2011) "Biochemical and morphological properties of hepatitis C virus particles and determination of their lipidome" *J Biol Chem*
79. Martín-Acebes, Merino-Ramos, Blázquez et al. (2014) "The composition of West Nile virus lipid envelope unveils a role of sphingolipid metabolism in flavivirus biogenesis" *J Virol*
80. Zhang, Chipman, Corver et al. (2003) "Visualization of membrane protein domains by cryo-electron microscopy of dengue virus" *Nat Struct Mol Biol*
81. Sirohi, Chen, Sun et al. (2016) "The 3.8 Å resolution cryo-EM structure of Zika virus" *Science*
82. Lok, Kostyuchenko, Nybakken et al. (2008) "Binding of a neutralizing antibody to dengue virus alters the arrangement of surface glycopro teins" *Nat Struct Mol Biol*
83. Zhang, Corver, Chipman et al. (2003) "Structures of immature flavivirus particles" *EMBO J*
84. Nelson, Jost, Xu et al. (2008) "Maturation of West Nile virus modulates sensitivity to antibody-mediated neutraliza tion" *PLoS Pathog*
85. Leventis, Grinstein (2010) "The distribution and function of phosphatidylserine in cellular membranes" *Annu Rev Biophys*
86. Van Meer, Voelker, Feigenson (2008) "Membrane lipids: where they are and how they behave" *Nat Rev Mol Cell Biol*
87. Stone, Vance (2000) "Phosphatidylserine synthase-1 and -2 are localized to mitochondria-associated membranes" *J Biol Chem*
88. Vance (1990) "Phospholipid synthesis in a membrane fraction associated with mitochondria" *J Biol Chem*
89. Takahashi, Kasai, Hatsuzawa et al. (1993) "A mutation of furin causes the lack of precursor-processing activity in human colon carcinoma LoVo cells" *Biochem Biophys Res Commun*
90. Zhang, Richard, Jackson et al. (2020) "Phosphatidyle thanolamine and phosphatidylserine synergize to enhance GAS6/AXLmediated virus infection and efferocytosis" *J Virol*
91. Bligh, Dyer (1959) "A rapid method of total lipid extraction and purification" *Can J Biochem Physiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548417&blobtype=pdf | # A novel ZIKV-targeted scRNA-seq method for precise quantification of ZIKV RNA
Yang Zhou, Libo Liu, Wei Yang, Yanhua Wu, Chongyao Zhong, Yuxuan Liu, Kunqi Lin, Dongying Fan, Yisong Wang, Peigang Wang, Jing An
## Abstract
Zika virus (ZIKV), transmitted by mosquitoes, poses a serious public health threat. Currently, precise quantitative diagnostic methods are lacking. Existing single-cell RNA-sequencing (scRNA-seq) techniques are challenged to detect ZIKV RNA due to its absence of a poly(A) tail, hindering the identification of infected cells. In this study, we developed a novel ZIKV-targeted scRNA-seq method that enables precise quantification of ZIKV RNA in individual cells. Immunocompetent suckling mice were intracerebrally infected with ZIKV to establish persistent infection in the brain. Samples were collec ted at 10-day post infection for ZIKV-targeted and 10× Genomics scRNA-seq analysis. Comparative analysis identified 17 distinct cell types in both ZIKV-infected and control suckling mouse brains, with significant changes in cell type distribution and proportion post-infection. The ZIKV-targeted scRNA-seq method suggested higher efficiency in capturing exogenous cells compared to the 10× Genomics scRNA-seq method. Both methods identified multiple endogenous and exogenous cell types susceptible to ZIKV, with peripheral blood-derived monocytes/macrophages (PBDMMs), neurons, and T cells as primary cell types expressing ZIKV RNA. IFA validated scRNA-seq findings, revealing that neurons and microglia could be infected by ZIKV, with significant reductions in their numbers post-infection. This study presents a novel ZIKV-targeted scRNA-seq that enables accurate quantification of ZIKV RNA within individual cells, identifies key susceptible cell types, and offers advantages in detecting exogenous cells, making it a scalable solution for providing valuable insights into therapeutic and vaccine develop ment.
IMPORTANCEThis study marks the first use of a scRNA-seq method tailored for ZIKV, allowing accurate measurement of ZIKV RNA in individual cells and identification of critical susceptible cell types. A comparative analysis with the 10× Genomics scRNA-seq method highlighted the advantages of ZIKV-targeted scRNA-seq in terms of accuracy and practicality, particularly its superior ability to capture exogenous cells. Beyond ZIKV, this method also helps establish precise quantification of viral RNA at the single-cell level for other viruses by designing target-specific beads based on conserved regions of the viral genome. This advancement is set to greatly enhance studying pathogenesis of ZIKV infection and then significantly contribute to improve prevention and research in therapeutics and vaccines.
vomiting (9), arthralgia (10)(11)(12), rash (13,14), and lymphadenopathy (15). However, ZIKV exhibits pronounced neurotropism that leads to congenital Zika syndrome in newborns (16)(17)(18)(19) and Guillain-Barré syndrome in adults (20)(21)(22). Previous studies showed that the viral envelope (E) protein facilitates ZIKV entry into host cells (23,24), while the nonstructural protein 1 (NS1), NS3, and NS5 play a significant role in viral replication and immune evasion (25). Nonetheless, traditional virological approaches and bulk RNA sequencing techniques face the challenges in dissecting the dynamics of ZIKV infection, replication, and the variation in cellular susceptibility to viruses. In particular, lacking the resolution to accurately distinguish cell states across various stages of infection may hinder the comprehensive understanding of ZIKV-host interactions.
The development of single-cell RNA sequencing (scRNA-seq) in 2009 represented a major breakthrough, overcoming the limitations of traditional bulk RNA sequencing, which measures gene expression at the population level (26). By enabling in-depth and precise exploration of gene expression at the single-cell level, scRNA-seq has provided invaluable insights into biological and medical research. Over the past decade, this technology has continuously advanced our understanding of development, disease mechanisms, and cellular functions (26,27).
However, current scRNA-seq methods, such as those developed by 10× Genomics, primarily rely on Oligo(dT)-modified magnetic beads to capture RNA from individual cells (28). Since ZIKV RNA lacks a poly(A) tail (29), it is theoretically unable to be captured by Oligo(dT)-modified beads for reverse transcription, which limits the identification and quantification of target cells during ZIKV infection in subsequent analyses, thereby restricting research on pathogenesis of ZIKV infection (30,31).
In this study, we designed a ZIKV-targeted scRNA-seq method that enabled the precise quantification of ZIKV-infected cells and identified the majority of cell types affected during ZIKV infection. We compared the ZIKV-targeted scRNA-seq data set with data set generated by the 10× Genomics scRNA-seq method and found no significant differences in the distribution and proportion of ZIKV RNA expression between the two approaches. Additionally, we validated part of the ZIKV-targeted scRNA-seq results through immunofluorescence assay (IFA), confirming the accuracy and feasibility of this method. Our results suggested ZIKV-targeted scRNA-seq has the advantages of being cost-effective, having low environmental and equipment requirements and, thus, offers broader application prospects.
## RESULTS
## The brain of suckling mice can support ZIKV replication
In this study, suckling mice at postnatal day 2 (P2) were intracranially infected with ZIKV to establish the immunocompetent animal model (Fig. 1A). The feasibility, similari ties, and differences between ZIKV-targeted scRNA-seq and 10× Genomics scRNA-seq methods for ZIKV detection were subsequently evaluated. By 10-day post infection (dpi) following ZIKV infection, a continuous upward trend in body weight was observed in both infected-and control groups. However, a significant reduction in weight gain was noted starting from 3 dpi, with marked differences compared to the control group (P < 0.05) (Fig. 1B). Quantitative analysis revealed that the viral load in the brain of ZIKV-infected mice was approximately 1 × 10 6 copies/µg by 10 dpi, with no significant differences detected among different brain regions (P > 0.05) (Fig. 1C). Furthermore, IHC demonstrated the presence of ZIKV E antigen (Fig. 1D) and NS1 antigen (Fig. 1E) in the brain of infected mice, thereby confirming that ZIKV can successfully infect and replicate within the brain of suckling mice.
## Design of magnetic beads and construction of a double complementary DNA library for ZIKV-targeted scRNA-seq
ZIKV-infected and control BALB/c suckling mouse brain cells were captured at 10 dpi using the FocuSCOPE microfluidic chip. control to enable targeted capture of ZIKV, we designed ZIKV Barcoding Beads incorporating Illumina Read 1 sequencing primer sequences, unique cell barcodes, unique molecular identifiers (UMIs), poly(T) nucleotide sequences, and ZIKV probes (Fig. 2A). The ZIKV probe specifically identifies ZIKV RNA, exhibiting no significant cross-reactivity with other species or virus, as confirmed by the Nucleotide Basic Local Alignment Search Tool (BLASTN). Notably, by combining the poly(T) sequence at the end of the oligonucleotide with the ZIKV-specific targeting probes, we successfully achieved precise capture and labeling of host RNA and ZIKV RNA from the brain cells of suckling mice.
To obtain a sufficient quantity of full-length complementary DNA (cDNA) products, the RNA enriched with magnetic beads underwent treatment with a template-switching oligonucleotide (TSO) during the reverse transcription process. Subsequently, the PCR amplification of the synthesized cDNA products was performed by adding PCR handle sequences (compatible with Illumina next-generation sequencing primers) to the 5′ end of the ZIKV Barcoding Beads. This approach facilitated the enrichment of full-length cDNA products (Fig. 2B).
On the one hand, to meet the length requirements for the sequencing library in next-generation sequencing (NGS), the fragmentation enzymes were employed to disrupt the full-length cDNA products obtained from the reverse transcription PCR amplification. Each cDNA fragment was approximately 500 bp in length and was utilized for the construction of both the host transcriptome library and the ZIKV-targeted enrichment library. For the host transcriptome library, we first performed end repair of the fragmented cDNA and added an A-tail to the 3′ end. Next, the cDNA fragments with P5 and P7 adapters at both ends were ligated, and the index (i7) at the 5′ end of the P7 adapter was introduced through PCR amplification. Finally, by filtering out low-qual ity fragments and selecting high-quality fragments, a high-quality host transcriptome library was obtained (Fig. 3A).
On the other hand, to obtain precise expression for ZIKV RNA in host cells, we successfully constructed a ZIKV-targeted enrichment library using ZIKV RNA captured by ZIKV Barcoding Beads. The reverse transcription and PCR amplification on the ZIKV RNA transcripts specifically captured was performed by the probes on the ZIKV Barcoding Beads, resulting in significant enrichment of cDNA products. Subsequently, using the cDNA as a template, the enrichment of ZIKV RNA was targeted to construct the enrichment library with FocuSCOPE-specific enrichment primers successfully (Fig. 3B). By constructing high-quality host transcriptome and ZIKV libraries, the scRNA-seq was performed based on the Illumina sequencing platform.
## Two single-cell RNA sequencing methods reveal similarities and differences in cellular heterogeneity
To further investigate the similarities and differences in cell types and distribution characteristics between ZIKV-targeted and 10× Genomics scRNA-seq methods in the brain of ZIKV-infected and control suckling mice at 10 dpi, we conducted a compara tive analysis of the host transcriptome data sets. First, low-quality cells and those with extreme expression levels were filtered out to obtain clean and high-quality data set. Briefly, low-quality cells were defined as those with fewer than 500 UMIs per cell, fewer than 200 detected genes, or a mitochondrial gene proportion exceeding 20%. These criteria were designed to exclude cell debris, dead cells, or doublets, thereby ensuring the reliability of downstream analyses.
The ZIKV-targeted scRNA-seq method captured 9,939 cells from the control group and 13,683 cells from the ZIKV-infected group, whereas 10× Genomics scRNA-seq captured 9,661 cells from the control group and 9,029 cells from the ZIKV-infected group, indicating a discrepancy in cell capture efficiency between the two methods.
Cell type annotation was based on the Shared Nearest Neighbor (SNN) modularity optimization algorithm (24), CellMarker 2.0 (25), and scRNA-seq/transcriptome sequenc ing references (26)(27)(28)(29)(30)(31). Cells captured by both methods were categorized into 17 distinct clusters, showing variations in the number of cells within each cluster. In the ZIKV-targeted scRNA-seq data set, the identified cell types included Endothelial cells (Endo), Ependymal cells (ECs), Fibroblasts, pericytes, neurons, Neural Progenitor cells (NPCs), Oligodendrocytes (OLs), Peripheral Blood Derived Monocytes/Macrophages (PBDMM), Perivascular Macrophages (pvMs), microglia, neutrophils, Antigen-Presenting cells (APCs), Natural Killer (NK) cells, B cells, T cells, Dendritic Cells (DCs), and Mast cells (MCs), all of which were present both with and without ZIKV infection (Fig. 4A andB). In contrast, the 10× Genomics data set identified cell types including ECs pericytes, neurons, NPCs, neuroblasts, OLs, microglia, Activated/Quiescent Neural Stem cells (a/qNSCs), Choroid Plexus epithelial cells (CPCs), Vascular Endothelial cells (VECs), meningeal cells, PBDMMs, B cells, and T cells. Notably, NK cells were detected in the brain only after ZIKV infection, whereas pvMs disappeared following the infection (Fig. 5A andB).
Further comparative analysis (Fig. 4A andB) revealed that both methods identified a common set of cell types, including neurons, NPCs, ECs, pericytes, OLs, PBDMMs, microglia, NK cells, B cells, and T cells, suggesting overall similarity in cell-type profiles.
However, the 10× Genomics data set captured a broader diversity of endogenous cells, while the ZIKV-targeted method had an advantage in detecting exogenous cells such as MCs, neutrophils, DCs, and APCs, indicating some differences in detected cell types between the two datasets.
To compare the two methods in terms of changes in cell proportions and dominant cell types, we analyzed cellular composition in both control and ZIKV-infected groups.
In the ZIKV-targeted data set, Endo and microglia were the dominant cell types before infection, whereas the 10× Genomics data set indicated that neurons and microglia predominated (Fig. 4C and5C; Table 1). Following ZIKV infection, both data sets showed a disruption of the original cell distribution, with PBDMMs becoming the predominant cell type in the brain.
Regarding changes in cellular composition, both scRNA-seq methods indicated a reduction in the proportions of endogenous cells, such as neurons, microglia, and NPCs post-infection, and significantly increase in exogenous cells, including T cells, B cells, NK cells, and PBDMMs. However, the 10× Genomics data set captured a higher proportion of endogenous cells (e.g., neurons, NPCs, OLs) both with or without infection, while the ZIKV-targeted method showed a greater proportion of exogenous cells, such as PBDMMs, T cells, and B cells (Fig. 4C and5C; Table 1). This comparative analysis high lights important similarities and differences in cell capture efficiency, endogenous/exog enous cell identification between the two scRNA-seq methods, offering their respective strengths in mapping ZIKV-infected brain tissues.
## PBDMM, neuron, and T cell are the target cells in ZIKV infection
To investigate the levels of ZIKV RNA in the brain of ZIKV-infected suckling mice, a gene integration analysis was performed using ZIKV-targeted and 10× Genomics scRNA-seq data set. Data set was extracted from the ZIKV-infected group for each scRNA-seq method. In the ZIKV-targeted analysis, host data set from infected cells was merged with viral-enriched library data set, facilitating the visualization of ZIKV RNA expression across various cell types based on enriched ZIKV cDNA counts. For the 10× Genomics data set, the ZIKV cDNA sequence was directly integrated into the host data set for visualization. Our analysis for ZIKV targeted scRNA-seq data set revealed predominant ZIKV RNA expression in endogenous cell types, including neurons, Endo, and microglia, as well as exogenous cells, such as PBDMM, T cells, and DCs. Notably, OLs, pvMs, MCs, and pericytes showed no detectable ZIKV RNA expression (Fig. 6A). In the 10× Genomics data set, ZIKV RNA was similarly expressed in a range of endogenous cells, including neurons and NPCs, as well as exogenous cell PBDMM. However, in the 10× Genomics data set, ECs and CPCs exhibited no expression (Fig. 6B), which is different from ZIKV targeted scRNA-seq data set. Despite some differences in cell type identification between the two dataset, consistent ZIKV RNA expression was observed in neurons, PBDMM, T cells, NK cells, B cells, microglia, and NPCs, with no expression in pericytes in either data set.
Further analysis identified PBDMM, neurons, and T cells as primary ZIKV infection targets. The ZIKV-targeted data set indicated that neurons accounted for the highest proportion of ZIKV RNA expression (Fig. 6C), whereas the 10× Genomics data set showed that PBDMM was identified the highest expression levels (Fig. 6D). Furthermore, the proportions of cells expressing ZIKV RNA were investigated to gain deeper insights into ZIKV RNA expression in specific cell types (Table 2). In the ZIKV-targeted data set, 13,681 cells were analyzed, with approximately 2.21% (n = 302) expressing ZIKV RNA (Fig. 6E). In the 10× Genomics data set, out of 9,029 analyzed cells, approximately 3.78% (n = 341) tested positive for ZIKV RNA (Fig. 6F). In terms of the absolute number of cells expressing ZIKV RNA, PBDMMs were indicated by both scRNA-seq methods, followed by neurons, T cells, and other cell types (Fig. 6G andH).
Subsequently, the PBDMM, neurons, and T cells were analyzed independently to elucidate their expression characteristics regarding ZIKV RNA. The proportion of PBDMM expressing ZIKV RNA was not significantly different across data set. In the ZIKV-targeted data set, 3.18% (176 cells) of 5,530 PBDMMs expressed ZIKV RNA, predominantly located in the central and upper regions of the cell cluster (Fig. 7A; Table 2). In the 10× Genomics data set, 6.95% (183 cells) of 2,648 PBDMMs expressed ZIKV RNA, concentrated in the upper right quadrant of the cell cluster (Fig. 7B; Table 2). For neurons, the ZIKV-targeted scRNA-seq data set indicated that 25% (71 cells) of 284 neurons expressed ZIKV RNA, primarily on the right side of the cell cluster (Fig. 7A; Table 2), whereas the 10× Genomics data set showed that 2.55% (32 cells) of 1,257 neurons expressed ZIKV RNA, displaying a scattered distribution (Fig. 7B; Table 2). Regarding T cells, 0.66% (26 cells) of 3,967T cells expressed ZIKV RNA in the ZIKV-targeted data set (Fig. 7A; Table 2), compared to 1.59% (26 cells) of 1,632T cells in the 10× Genomics scRNA-seq data set (Fig. 7B; Table 2). However, it is worth noting that the absolute number and proportion of cells carrying ZIKV RNA remained unchanged.
## Neuron and microglia could be infected by ZIKV
To validate the scRNA-seq findings on the susceptibility of neurons and microglia to ZIKV infection, IFA tests were conducted for ZIKV NS1 antigen, along with the neuronal marker NeuN and microglial marker TMEM119, respectively (Fig. 8A andB). Following ZIKV infection, we observed a significant reduction in staining intensity for TMEM119 (Fig. 8C) and NeuN (Fig. 8D) compared to the control groups, with minimal co-localization between ZIKV NS1 and these cellular markers (Fig. 8E andF). These results corroborate our findings from scRNA-seq data set, indicating that ZIKV infection damages neurons and microglia and reduces the cellular number in the suckling mouse brain, further confirming that both cell types are susceptible to ZIKV infection.
## ZIKV-targeted scRNA-seq method has multiple advantages
The value and scenarios of the ZIKV-targeted scRNA-seq method were assessed from multiple dimensions (Table 3). First, the ZIKV-targeted scRNA-seq supports manual operation for single-cell sorting, and the FocuSCOPE microfluidic chip is portable due to the lightweight and small-sized. In contrast, while the 10× Genomics scRNA-seq was allowed to complete single-cell sorting automatically via the 10× Genomics Chromium Single-Cell 3′ kit, the equipment is heavy and bulky and requires high stability, which makes frequent transportation inconvenient and restricts it to fixed locations. Second, both methods require a horizontal centrifuge suitable for 15-50 mL centrifuge tubes and an automated cell counter to obtain high-quality single-cell suspensions before library construction. In addition, the 10× Genomics scRNA-seq completes these steps in an integrated workflow via the Chromium Single-Cell 3′ kit including mRNA reverse transcription and PCR amplification, whereas the ZIKV-targeted scRNA-seq relies on a metal bath and thermal cycler. However, in terms of equipment costs, the ZIKV-targeted scRNA-seq is significantly less expensive than the 10× Genomics method. Moreover, both methods can ensure high-quality sequencing data based on the Illumina sequencing platform, while the thermal cyclers and metal baths are becom ing increasingly miniaturized and portable currently, which indicates the ZIKV-targeted scRNA-seq combining low cost and device portability, is suitable for laboratories, field investigations, and single-cell transcriptomics in remote areas, demonstrating considera ble practical application value. In contrast, the 10× Genomics scRNA-seq method still depends on a fixed laboratory environment, limiting its application scenarios. Although tissue cryopreservation solutions have been developed and applied to transport fresh tissues, they may significantly affect intracellular gene expression profiles and cell viability due to mechanical forces.
## DISCUSSION
ZIKV is primarily transmitted through mosquito bites and continues to pose a major public health threat globally. Since 2015, large-scale outbreaks in Brazil, Colombia (32), and Puerto Rico have been associated with a surge in congenital Zika syndrome (CZS) among neonates and Guillain-Barré syndrome (GBS) in adults (33,34), raising serious concerns about its long-term neurological impact. Mechanistic studies suggest that ZIKV hijacks peripheral blood monocytes to traverse the blood-brain barrier, subse quently infecting both central and peripheral nervous systems (35). This neurotropism results in the infection and apoptosis of neural cells, ultimately leading to severe and often irreversible neurological damage (36)(37)(38). Moreover, recent studies have revealed potential associations between ZIKV infection and various host cellular events, including cell cycle arrest (39), DNA damage (40), mitochondrial fragmentation (41), and endoplas mic reticulum (ER) stress (42). However, the pathogenesis of ZIKV infection is highly complex. The intricate regulatory networks and cellular heterogeneity within host tissues become the formidable challenges to traditional virological methods and bulk RNA sequencing.
In recent years, the emergence of scRNA-seq technology has transformed virologi cal research, offering unparalleled insights into host-virus interactions and identifying specific cell types infected by viruses (43)(44)(45). However, ZIKV, as the member of RNA viruses, was used to be considered poorly captured by oligo(dT)-based methods due to their lack of poly(A) tails (46). Interestingly, our previous study demonstrated that oligo(dT)-modified beads were capable of detecting and quantifying ZIKV RNA in the mouse testis by integrating the ZIKV genome sequence into the host transcriptome data set during downstream analysis (47). However, the specificity in oligo(dT)-based capture of ZIKV RNA and the underlying mechanisms remain to be clarified, highlighting the urgent need for the development of a ZIKV-specific scRNA-seq method to overcome the current limitations in ZIKV research.
In this study, we developed a novel ZIKV-targeted scRNA-seq method that accu rately quantifies ZIKV RNA within individual cells, revealing key cell types suscepti ble to infection and offering a refined understanding of ZIKV infection dynamics. The comparative analysis between the ZIKV-targeted scRNA-seq and the 10× Genom ics scRNA-seq methods highlights significant differences in cell capture efficiency. Specifically, the ZIKV-targeted method exhibited higher efficacy in capturing exogenous cells, such as PBDMM, T cells, and MCs, whereas the 10× Genomics scRNA-seq method demonstrated broader coverage of larger endogenous cells, including neurons and glial cells (Table 1). The characteristics of cell type indicate that there is a difference in cell capture efficiency between the two methods. Depending on the research objective, different scRNA-seq methods can be chosen.
Possible reasons underlying the differences in cell types captured by the two methods include several technical factors. First, a 70 µm filter was used to process the dissociated single-cell suspension from mouse brains in the ZIKV-targeted scRNA-seq method, manually, which can effectively block larger cells and debris (Table 3). In ZIKV-infected suckling mouse brains, the majority of small-sized cells are infiltrating exogenous cells. Therefore, this method yields a higher capture rate of these exogenous cells compared to the automated dissociation and cell capture process performed by the 10× Genomics scRNA-seq method. This represents one of the primary factors contribu ting to the observed differences in cell type composition between the two methods.
Additionally, in terms of sample preparation, the ZIKV-targeted scRNA-seq method involves manually injecting the single-cell suspension into a microfluidic chip, where individual cells are randomly distributed into wells based on a Poisson distribution (Table 3). ZIKV barcoding beads and lysis buffer are then manually introduced to capture both cells and their mRNA. In contrast, the 10× Genomics scRNA-seq method relies on the automated 10× Chromium system for cell encapsulation and capture (Table 3), indicating that the differences in sample preparation workflows may contribute to variations in both cell number and cell type representation. Notably, the manual approach allows for greater control over sample handling and offers a gentler, customizable alternative to automated systems. Regarding capture efficiency, the ZIKV barcoding beads used in the targeted method may possess higher RNA capture efficiency compared to those used in the 10× Genomics scRNA-seq method. This may further contribute to the discrepancies in cell capture rates and cellular diversity observed between the two methods.
Given that the ZIKV-targeted scRNA-seq consistently captured a greater number of cells in both infected and control groups, we conclude that the differences in cell number and composition are primarily driven by technical variation between the methods rather than underlying biological differences.
Our findings showed that ZIKV RNA was detected in various cell types, primarily expressed in neurons, PBDMM, and T cells using both methods, while the capture mechanisms of ZIKV RNA in 10× Genomics scRNA-seq were not clear. It is speculated that this may be associated with the potential post-infection modifications of viral RNA, such as polyadenylation, which warrant further investigation. These modifications may not be mere artifacts of viral gene expression but could play crucial roles in host-virus interactions, which need further investigation. Considering the specificity of the ZIKV probe, ZIKV-targeted scRNA-seq method demonstrates scientific validity and representa tiveness in quantifying intracellular ZIKV RNA expression and identifying target cell types. Despite these advantages, a limitation of the ZIKV-targeted scRNA-seq method is the exclusion of larger cells, such as neurons and glial cells, during the filtration process. This could lead to a biased representation of certain cell populations. Future studies may benefit from combining this approach with other technologies, such as flow cytometry or traditional Oligo(dT)-based capture methods, to achieve a more compre hensive representation of both small and large cell types. Such integration would further enhance the versatility and applicability of the ZIKV-targeted method.
Although cell preservation solutions have been developed to maintain cell viability during long-distance transport; however, external factors, such as mechanical vibrations and temperature fluctuations, may compromise genomic stability and cellular pheno types (48,49). It can be hypothesized that long-distance transport may increase the risk of phenotypic and genomic changes in cells. Such changes could undermine the accuracy of experimental outcomes and negatively affect the correct identification of cell types and their corresponding ZIKV RNA expression levels in downstream analyses. By employing a manual approach for cell sorting, cDNA reverse transcription, PCR amplification, and library construction, the ZIKV-targeted scRNA-seq method circum vents these challenges and allows to perform on-site, which significantly reduces the risks of physical and genomic changes during transport and offers a more affordable and accessible alternative (Table 3). This characteristic is particularly crucial in regions where the cost and availability of advanced laboratory infrastructure pose significant barriers to effective viral surveillance and provide a feasible solution in diverse research scenarios.
In addition to identifying ZIKV-infected cell types, the ZIKV-targeted scRNA-seq method is able to quantify viral RNA expression across diverse cell populations provides critical insights into how ZIKV interacts with different cellular environments. These findings can shed light on potential mechanisms of immune evasion or persistence within the host, which are key factors in developing targeted therapeutic interventions. For instance, understanding the viral load in neurons and glial cells could offer valua ble data for the development of neuroprotective therapies in ZIKV-infected patients. Additionally, the data generated by the ZIKV-targeted scRNA-seq method could play a pivotal role in future vaccine development, particularly in designing vaccines that target specific cell types or tissues most affected by ZIKV infection. Further research could explore how this information could be leveraged to develop cell-specific immune responses, offering broader applicability not only for ZIKV but also for related viruses.
Overall, the ZIKV-targeted scRNA-seq method offers a broadly applicable and cost-effective strategy for studying host-ZIKV interactions, with implications for both therapeutic and vaccine development. Future work can build upon these findings to further explore viral mechanisms and improve global responses to ZIKV and related viruses. However, given the absence of reported ZIKV infection cases in China since 2019, a limitation of this method is the challenge of validating it with clinical sam ples from ZIKV-infected individuals. Nonetheless, with the anticipated advancement of international collaborations in vector-borne infectious disease prevention and control, this method holds promise for eventual clinical validation in ZIKV-endemic regions.
## MATERIALS AND METHODS
## ZIKV infection animal model
BALB/c suckling mice at postnatal day 2 were infected with the SMGC-1 ZIKV strain via intracranial injection with an infection dose of 20 µL containing 100 PFU. The mice were humanely euthanized at the 10 dpi, and brain tissues were collected for further analysis.
## Preparation of single-cell suspensions from suckling mouse brain
Suckling mouse brain tissues were minced into 1 mm 3 pieces using ophthalmic scissors and digested in Collagenase IV (SIGMA, USA) under the conditions of 37°C, 250 rpm shaking for 15 min. The digested brain tissue was filtered through a 70 µm mesh to collect the cell suspension. The cell suspension was then resuspended in calcium-and magnesium-free 1× PBS and washed by centrifugation at 350 g, 4°C, for 5 min with no brake. According to the manufacturer's instructions, the SCelLive Debris Removal Kit (Singleron Biotechnologies, China, 13200066) was used to remove cell debris and dead cells from the cell suspension. Cell viability was assessed using AOPI staining (Counter Star, China,RE010212), and samples with cell viability ≥85% and cell aggregation rate ≤10% were deemed suitable for further single-cell library preparation.
## 10× Genomics scRNA-seq
According to the manufacturer's instructions for the 10× Genomics Chromium Single-Cell 3′ Kit (V3), the single-cell suspensions were loaded onto the 10× Chromium system to capture individual cells. Subsequent cDNA amplification and library construction steps were performed according to the standard protocol. The libraries were sequenced on the Illumina NextSeq 500 platform (paired-end, 150 bp) at LC-Bio Technology Co., Ltd. (Hangzhou, China), using a multiplexed sequencing run.
## ZIKV-targeted scRNA-seq
We collaboratively developed a Zika virus (ZIKV)-targeted single-cell RNA sequencing kit with Singlera Genomics Co., Ltd. and performed sorting and cDNA library construc tion on single cells derived from the brains of suckling mice. To overcome the gap in current ZIKV-host interactions analysis methods in sc-RNA seq, probes binding to ZIKV sequence and oligo-dT were added to the magnetic capture beads, which allows to capture and reverse transcribing mRNA from host and virus sequence. The ZIKV probe was designed to map current ZIKV lineages based on the National Center for Biotech nology Information (NCBI). The Nucleotide Basic Local Alignment Search Tool (BLASTN) was applied to evaluate the specificity of ZIKV Probe. Following the manufacturer's recommendations, the single-cell suspension was briefly added to a microfluidic chip, introducing ZIKV-targeted modified capture magnetic beads (ZIKV Probe: TTTGCATGT CCACCGCCATCTGAGCTGGAA) along with cell lysis buffer into the microfluidic chip to capture single cells and RNA. The mixture was then incubated in a heated metal bath for cDNA reverse transcription. Subsequently, cDNA products were amplified using the 7500 Real-Time PCR System (Applied Biosystems, USA). After purification with Ampure XP magnetic beads (Beckman Coulter Life Sciences, A63880, USA), the purified cDNA amplification products were used for constructing GEXSCOPE single-cell transcriptomic libraries for both host and virus. Quality control of the cDNA libraries was performed using a Qubit 4.0, and sequencing was conducted based on the Illumina PE150 strategy on the Novaseq platform.
## The quality control of raw data set for 10× Genomics scRNA-seq
Sequencing results were demultiplexed and converted to FASTQ format using the Illumina bcl2fastq software. Sample demultiplexing, barcode processing, and single-cell 3′ gene counting were performed using Cell Ranger v5.0.1, and the scRNA-seq data set was aligned to the Ensembl genome GRCm38 reference genome (release-95). The Cell Ranger output was subsequently loaded into Seurat v4.1.1 for dimensionality reduction, clustering, and analysis of the scRNA-seq data set. In total, 22,988 cells passed the quality control thresholds: genes expressed in fewer than one cell were removed, a low cutoff of more than 500 genes expressed per cell was applied, and the percentage of expression derived from mitochondrial DNA was kept below 25%.
## Immunohistochemistry
To investigate the distribution of ZIKV E + and ZIKV NS1 + cells in the cortical regions of suckling mouse brains, brain sections were incubated overnight at 4°C with mouse anti-ZIKV NS1 monoclonal antibody (Invitrogen, California, USA, 1:500) and mouse anti-ZIKV 4G2 (1:500) as primary antibodies. After washing with PBS, the sections were stained with a secondary antibody, HRP-conjugated goat anti-mouse antibody (Zhongshan Golden Bridge Biotechnology Co., Ltd., China) for 1 h. The reaction was visualized by adding 3,3′-diaminobenzidine (DAB) as the chromogenic substrate, and the reaction was terminated by removing DAB.
## Cell quantification
The positive staining cells in the IFA were analyzed using ImageJ v1.8.0.112 software, with each group comprising at least three suckling mice. For each mouse, sections were analyzed across a minimum of five fields of view (×400 magnification), with each field containing 100 or more cells. The cell count was expressed as the number of cells per mm².
## Data set statistics and reproducibility
Images from the IFA were obtained from at least three independent experiments. All statistical analyses were conducted using SPSS version 17.0 software (IBM, Armonk, NY, USA) and were subsequently verified using Microsoft Excel 2016. A two-tailed Student's t-test was employed to analyze quantitative data set between two groups that followed a normal distribution. A P-value of less than 0.05 was considered statistically significant between the two groups.
## References
1. Fauci, Morens (2016) "Zika virus in the Americas-yet another arbovirus threat" *N Engl J Med*
2. Duffy, Chen, Hancock et al. (2009) "Zika virus outbreak on Yap Island, Federated States of Micronesia"
3. Pielnaa, Al-Saadawe, Saro et al. (2020) "Zika virus-spread, epidemiology, genome, transmission cycle, clinical manifestation, associated challenges, vaccine and antiviral drug development" *Virology (Auckl)*
4. Wang, Wang (2016) "Zika virus and Zika fever" *Virol Sin*
5. Song, Yun, Woolley et al. (2017) "Zika virus: history, epidemiol ogy, transmission, and clinical presentation" *J Neuroimmunol*
6. Mwaliko, Nyaruaba, Zhao et al. (2021) "Zika virus pathogenesis and current therapeutic advances" *Pathog Glob Health*
7. Camacho-Concha, Santana-Román, Sánchez et al. (2023) "Insights into Zika virus pathogenesis and potential therapeutic strategies" *Biomedicines*
8. Araújo Psr De, Júnior, De et al. (2019) "Co-infection ZIKV and HSV-1 associated with meningoencephalitis: case report and literature review" *J Infect Public Health*
9. Pougnet, Thill, Pougnet et al. (2016) "Infection à virus Zika: à propos d'un cas importé en métropole" *Ann Biol Clin*
10. Wimalasiri-Yapa, Yapa, Huang et al. (2020) "Zika virus and arthritis/arthralgia: a systematic review and meta-analysis" *Viruses*
11. Tiwari, Bergman (2023) "Viral arthritis"
12. Gupta, Kodan, Baruah et al. (2023) "Zika virus in India: past, present and future" *QJM*
13. Muzumdar, Rothe, Jm (2018) "The rash with maculopa pules and fever in adults" *Clin Dermatol*
14. Côrtes, Lira, Prates-Syed et al. (2023) "Integrated control strategies for dengue, Zika, and Chikungunya virus infections" *Front Immunol*
15. Haddow, Perez-Sautu, Wiley et al. (2020) "Modeling mosquito-borne and sexual transmission of Zika virus in an enzootic host, the African green monkey" *PLoS Negl Trop Dis*
16. Freitas, Souza-Santos, Carvalho et al. (2020) "Congenital Zika syndrome: a systematic review" *PLoS One*
17. Robinson, Galvan, Trujillo et al. (1985) "Congenital Zika syndrome: pitfalls in the placental barrier" *Rev Med Virol*
18. Chen, Wilson (2019) "Zika circulation, congenital syndrome, and current guidelines: making sense of it all for the traveller" *Curr Opin Infect Dis*
19. Paixao, Cardim, Costa et al. (2022) "Mortality from congenital Zika syndrome: nationwide cohort study in Brazil" *N Engl J Med*
20. Leonhard, Mandarakas, Gondim et al. (2019) "Diagnosis and management of Guillain-Barré syndrome in ten steps" *Nat Rev Neurol*
21. Rodríguez, Rojas, Pacheco et al. (2018) "Guillain-Barré syndrome, transverse myelitis and infectious diseases" *Cell Mol Immunol*
22. Deseda (2017) "Epidemiology of Zika" *Curr Opin Pediatr*
23. Agrelli, De Moura, Crovella et al. (2019) "ZIKA virus entry mechanisms in human cells" *Infect Genet Evol*
24. Giraldo, Xia, Aguilera-Aguirre et al. (2020) "Envelope protein ubiquitination drives entry and pathogenesis of Zika virus" *Nature*
25. Kam, Leite, Amrun et al. (2019) "ZIKV-specific NS1 epitopes as serological markers of acute Zika virus infection" *J Infect Dis*
26. Samad, Wu (2021) "Single cell RNA sequencing approaches to cardiac development and congenital heart disease" *Semin Cell Dev Biol*
27. Paolillo, Londin, Fortina (2019) "Single-cell genomics" *Clin Chem*
28. Kim, Booth, Saunders et al. (2022) "Nuclear oligo hashing improves differential analysis of singlecell RNA-seq" *Nat Commun*
29. Kümmerer (2018) "Establishment and application of flavivirus replicons" *Adv Exp Med Biol*
30. Swaminath, Russell (2024) "The use of single-cell RNA-seq to study heterogeneity at varying levels of virus-host interactions" *PLoS Pathog*
31. Chang, Liu, Wang (2024) "Single-cell RNA sequencing to understand host-virus interactions" *Virol Sin*
32. Plourde, Bloch (2016) "A literature review of Zika virus" *Emerg Infect Dis*
33. Krauer, Riesen, Reveiz et al. (2017) "Zika virus infection as a cause of congenital brain abnormalities and Guillain-Barré syndrome: systematic review" *PLoS Med*
34. Giraldo, Gonzalez-Orozco, Rajsbaum (2023) "Pathogenesis of Zika Virus infection" *Annu Rev Pathol*
35. Yan, Zheng, Jiang et al. (2023) "Transcriptomic reveals the ferroptosis features of host response in a mouse model of Zika virus infection" *J Med Virol*
36. Van De Beek, Brouwer (2017) "CNS Infections in 2016: 2016, the year of Zika virus" *Nat Rev Neurol*
37. Liang, Guida, Do Nascimento et al. (2019) "Host and viral mechanisms of congenital Zika syndrome" *Virulence*
38. Cugola, Fernandes, Russo et al. (2016) "The Brazilian Zika virus strain causes birth defects in experimental models" *Nature*
39. Li, Zhu, Ren et al. (1962) "DEFA1B inhibits ZIKV replication and retards cell cycle progression through interaction with ORC1" *Life Sci*
40. Wang, Song, Li et al. (2024) "Erp57 facilitates ZIKV-induced DNA damage via NS2B/NS3 complex formation" *Emerg Microbes Infect*
41. Yang, Gorshkov, Lee et al. (2020) "Zika virus-induced neuronal apoptosis via increased mitochondrial fragmentation" *Front Microbiol*
42. Muthuraj, Sahoo, Kraus et al. (2021) "Zika virus infection induces endoplasmic reticulum stress and apoptosis in placental trophoblasts" *Cell Death Discov*
43. Zanini, Robinson, Croote et al. (2018) "Virus-inclusive single-cell RNA sequencing reveals the molecular signature of progression to severe dengue" *Proc Natl Acad Sci*
44. Wu, Huang, Sun et al. (2023) "Zika virus targets human trophoblast stem cells and prevents syncytialization in placental trophoblast organoids" *Nat Commun*
45. Zanini, Pu, Bekerman et al. (2018) "Single-cell transcriptional dynamics of flavivirus infection"
46. Yang, Duff, Graveley et al. (2011) "Genome wide characterization of non-polyadenylated RNAs" *Genome Biol*
47. Yang, Liu, Liu et al. (2023) "Single-cell RNA sequencing reveals the fragility of male spermatogenic cells to Zika virus-induced complement activation" *Nat Commun*
48. Cleri, Landuzzi, Blossey (2018) "Mechanical evolution of DNA double-strand breaks in the nucleosome" *PLoS Comput Biol*
49. Santos, Cook, Gough et al. (2021) "DNA damage alters nuclear mechanics through chromatin reorganiza tion" *Nucleic Acids Res* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12481070&blobtype=pdf | # International Journal of Surgery Case Reports
M Ali-Hassanzadeh
## Abstract
Letter to the editor Commentary on "digital amputations in a child with multisystem inflammatory syndrome (MIS-C): A case report" by Davoodi et al., 2025
The report by Davoodi and colleagues describing digital amputations in a paediatric patient with multisystem inflammatory syndrome in children (MIS-C) underscores one of the most devastating vascular sequelae associated with post-COVID-19 hyperinflammatory states [1]. Although MIS-C has been well recognized since the early phases of the pandemic, reports of critical ischemic complications requiring amputation remain exceptionally rare. This case therefore serves as a stark reminder of the thrombo-inflammatory burden in paediatric populations and highlights the urgent need for heightened clinical vigilance, early recognition, and aggressive intervention. Beyond the clinical singularity of this report, however, several important issues warrant further reflection and discussion.
First, a deeper exploration of the pathophysiological underpinnings of digital necrosis in MIS-C would enrich the clinical significance of this report. Hyperinflammatory cytokine cascades, endothelial dysfunction, and thrombotic microangiopathy are increasingly recognized as central drivers of vascular injury in post-COVID inflammatory syndromes [2]. These mechanisms not only explain the catastrophic progression to digital ischemia and amputation but also highlight potential avenues for targeted interventions ranging from timely anti-inflammatory therapy to aggressive anticoagulation strategies. Framing this complication within such mechanistic context is essential to inform both acute management and future preventive strategies.
Second, a more comprehensive account of the diagnostic work-up including imaging modalities, vascular assessments, and laboratory markers would enhance both the reproducibility and the educational value of this case [3]. Such detail is crucial to guide clinicians faced with similarly ambiguous presentations of ischemia in MIS-C.
Third, the therapeutic rationale and timing of interventions deserve fuller exploration. Clarifying the decision-making process surrounding anticoagulation, immunomodulatory therapy, and ultimately surgical intervention would not only inform clinical practice but also illuminate the thresholds at which medical management gives way to irreversible tissue compromise [4].
Fourth, long-term outcomes warrant closer attention. Functional recovery, rehabilitation strategies, and psychosocial support are indispensable components of care for paediatric patients who undergo limb loss. Addressing these dimensions would provide valuable guidance for multidisciplinary teams navigating the profound physical and psychological sequelae of such complications.
Finally, the broader implications of this case extend well beyond its rarity. Digital amputation in MIS-C underscores both the severity of vascular complications in post-COVID inflammatory states and the existing gaps in management guidelines. Greater emphasis on prevention strategies, early recognition of ischemic warning signs, and systematic follow-up could substantially mitigate the burden of such catastrophic outcomes in children [5].
In summary, Davoodi and colleagues present a striking and clinically significant complication of MIS-C. By situating this case within a broader discussion of pathophysiology, diagnostic pathways, therapeutic reasoning, and long-term care, the report could provide invaluable lessons for paediatric and critical care communities worldwide, and prompt renewed consideration of strategies to anticipate and prevent severe thrombo-ischemic events in vulnerable children.
## Ethical approval
Not applicable.
## References
1. Davoodi, Abootalebi (2025) "Digital amputations in a child with multisystem inflammatory syndrome (MIS-C): a case report" *Int. J. Surg. Case Rep*
2. Kuchler, Günthner, Ribeiro et al. (2023) "Persistent endothelial dysfunction in post-COVID-19 syndrome and its associations with symptom severity and chronic inflammation" *Angiogenesis*
3. Cattalini, Taddio, Bracaglia et al. (2021) "Childhood multisystem inflammatory syndrome associated with COVID-19 (MIS-C): a diagnostic and treatment guidance from the rheumatology study Group of the Italian Society of pediatrics" *Ital. J. Pediatr*
4. Peng, Zhou (2025) "Progress on diagnosis and treatment of multisystem inflammatory syndrome in children" *Front. Immunol*
5. Mahmoud, El-Kalliny, Kotby et al. (2022) "Treatment of MIS-C in children and adolescents" *Curr. Pediatr. Rep* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12528060&blobtype=pdf | # New onset refractory status epilepticus diagnosed in the second trimester: A case report
Vesna Sokol Karadjole, Dareen Alshaer, John Snelgrove, Laurence Carmant, Ginette Moores
## Abstract
New-onset refractory status epilepticus (NORSE) is a rare condition in which a previously healthy individual develops refractory seizures without an identifiable cause. In pregnancy, management is particularly challenging due to the need to control seizures while minimizing teratogenic risk for the fetus. We report a 22-year-old woman who developed NORSE at 19 weeks' gestation following recurrent tonic-clonic seizures. Treatment included multiple antiseizure medications: levetiracetam, oxcarbazepine, lacosamide, clobazam, and lamotrigine. Due to super-refractory status, she required intubation and sedation with propofol and midazolam and was extubated once seizure-free. Following a breakthrough seizure and suicide attempt, levetiracetam was replaced with brivaracetam. Fetal growth and biophysical profile remained appropriate on serial surveillance. She underwent term induction of labor, delivering a healthy neonate without signs of withdrawal. This is the first reported case of second-trimester NORSE with favorable perinatal outcomes, underscoring the need for a multidisciplinary approach to balance seizure control and fetal safety.
## Introduction
New-onset refractory status epilepticus (NORSE) is rare and characterized by status epilepticus unresponsive to standard treatment without an apparent structural, toxic, or metabolic cause. 1 If an autoimmune or infectious trigger is later identified, it is still classified as NORSE; if no cause is found, it is termed cryptogenic. 2 NORSE accounts for 20% of refractory status epilepticus (RSE) cases and carries a high risk of neurological morbidity and mortality. 3 Though uncommon, it can affect young individuals, including pregnant women, where altered antiseizure medication (ASM) metabolism and teratogenicity complicate management.
Only one case of NORSE in early pregnancy has been reported, requiring vagal nerve stimulation for seizure control. 4
## Case presentation
A healthy 22-year-old woman in her first pregnancy presented to the emergency department at 18 weeks and 6 days of gestation, with a couple of episodes of tonic-clonic seizure lasting less than 2 min. She had no history of epilepsy or seizure risk factors. She was started on levetiracetam 500 mg twice daily (BID). Initial labs, brain imaging, and CSF testing were unremarkable (Tables 1 and2).
Despite treatment, she continued to have focal seizures with automatisms. Levetiracetam was increased and oxcarbazepine and lacosamide were added sequentially. Electroencephalogram (EEG) showed left temporal seizures without clinical correlation. Given the subclinical seizures refractory to several ASM, she was transferred to a tertiary care center for continuous EEG monitoring.
By Day 9, she experienced worsening tonic-clonic seizures requiring intubation and sedation with propofol and midazolam. Her ASM regimen included levetiracetam 2 g twice daily, clobazam 20 mg twice daily, lacosamide 2000 mg twice daily, oxcarbazepine 600 mg twice daily, and lamotrigine 50 mg twice daily. Repeat infectious work-up and malignancy screening were negative (Table 1).
On Day 12, she was started on intravenous immunoglobulin and a five-day course of intravenous methylprednisolone, followed by oral prednisone (1 mg/kg). EEG showed multifocal discharges without seizures. Sedation was weaned, but stopping midazolam (Day 23) led to recurrent electrographic seizures. Midazolam was resumed and oxcarbazepine increased to 900 mg twice daily. By Day 26, she was extubated and seizure-free.
On Day 29, the woman was transferred to the tertiary obstetric center, fetal ultrasound showed normal growth and biophysical parameters (Table 2). One week later, she was discharged home with ASM therapy and slow prednisone taper.
## Outpatient course and delivery
She remained seizure-free initially but developed ASM-related side effects (ataxia, drowsiness). Lamotrigine was gradually discontinued given its subtherapeutic dose and concern for multiple sodium channel blockade.
One month later, she had a breakthrough seizure following fasting for her oral glucose tolerance test. Shortly after, she had mood dysregulation with a suicide attempt by overdosing on her ASM. Psychiatry attributed mood instability to levetiracetam and corticosteroids. Levetiracetam was switched to brivaracetam (150 mg twice daily) in addition to ongoing prednisone taper with clinical improvement.
A multidisciplinary meeting was held for delivery planning. At 38 weeks, she underwent an uncomplicated induced vaginal delivery of a healthy male neonate. Due to prolonged steroid use in pregnancy, stress dose steroids were
## Discussion
This is the first reported case of second-trimester NORSE with a good perinatal outcome. RSE is a medical emergency whereby recurrent seizures fail to respond to initial treatment with ASM. While rare during pregnancy, it is of concern as it is associated with increased risks of fetal and maternal mortality. 5,6 Two studies have linked tonic-clonic seizures during pregnancy to increased risk of preterm birth and small for gestational age. 7,8 In one case, seizures during labor were demonstrated to prolong uterine contraction, compromising placental blood flow and causing fetal hypoxia. 9 Other risks of maternal seizures include fetal injury, placental abruption, or intrauterine fetal death due to maternal trauma sustained during a seizure. 5 Unfortunately, limited research has been conducted to assess the direct impact of maternal seizures on fetal well-being, and there is a lack of information regarding the frequency or duration of seizures that may endanger the fetus.
In nonpregnant women, early immunotherapy is recommended for NORSE, ideally within 72 h of SE onset. 10 No pregnancy-specific guidelines exist; consensus suggests applying standard SE management principles during pregnancy. The International League Against Epilepsy advises prompt and aggressive treatment of tonic-clonic SE in pregnancy. 11 In our case, the initial delay in immunotherapy likely resulted from diagnostic uncertainty and initial concerns regarding fetal safety prior to the involvement of maternal-fetal medicine. EEG improved rapidly thereafter, enabling sedation wean and extubation.
Five ASMs were ultimately required, with close multidisciplinary monitoring supporting fetal well-being.
Seizure management in pregnancy must balance fetal drug risks against the dangers of uncontrolled seizures for both mother and fetus. Pregnancy-induced pharmacokinetic changes-especially in lamotrigine, levetiracetam, and oxcarbazepine-may necessitate frequent dose adjustments. 12 Lamotrigine, despite favorable safety, requires slow titration over several weeks, limiting its role in SE. Additionally, safe intravenous ASM options in pregnancy are limited.
Benzodiazepines are the first-line treatment for SE with levetiracetam, valproic acid and phenytoin representing the second-line agents. Levetiracetam is preferred over valproic acid and phenytoin due to a lower teratogenic profile. 13 Polytherapy increases teratogenicity risk, particularly if valproic acid is included. 14,15 For RSE requiring sedation, propofol and midazolam are considered safe in pregnancy. 16 Our patient's suicide attempt illustrates psychiatric vulnerability among pregnant individuals with epilepsy. One prospective study found increased rates of depression, anxiety, and suicidal ideation in this population. 17 Maternal anxiety is also linked to adverse neurodevelopmental outcomes. 18 Levetiracetam, in particular, has been associated with behavioral side effects, necessitating psychiatric support and mood screening in prenatal epilepsy care. Limited safety data exist for brivaracetam in pregnancy, but current evidence shows no specific teratogenic or fetal risks. 19 After discussing risks and benefits, our patient was successfully switched from levetiracetam to brivaracetam. This underscores the need for continued research on newer ASMs during pregnancy. 20 NORSE in pregnancy requires prompt recognition, early immunotherapy, and multidisciplinary coordination. Management should follow established SE protocols with individualized ASM risk-benefit assessment. With close monitoring, favorable outcomes are achievable.
the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
## Informed consent
The patient provided written informed consent for the publication of this case report.
## References
1. Hirsch, Gaspard, Van Baalen (2018) "Proposed consensus definitions for new-onset refractory status epilepticus (NORSE), febrile infection-related epilepsy syndrome (FIRES), and related conditions" *Epilepsia*
2. Ritter, Nashef (2021) "New-onset refractory status epilepticus (NORSE)" *Pract Neurol*
3. Gaspard, Foreman, Alvarez (2015) "New-onset refractory status epilepticus: etiology, clinical features, and outcome" *Neurology*
4. Jindal, Delaj, Winston (2023) "Safe and effective implantation and use of vagal nerve stimulation in new-onset refractory status epilepticus in early pregnancy: a case report" *Front Neurol*
5. Mazzone, Hogg, Weir (2023) "Comparison of perinatal outcomes for women with and without epilepsy: a systematic review and meta-analysis" *JAMA Neurol*
6. Macdonald, Bateman, Mcelrath (2015) "Mortality and morbidity during delivery hospitalization among pregnant women with epilepsy in the United States" *JAMA Neurol*
7. Chen, Chiou, Lin (2009) "Affect of seizures during gestation on pregnancy outcomes in women with epilepsy" *Arch Neurol*
8. Rauchenzauner, Ehrensberger, Prieschl (2013) "Generalized tonic-clonic seizures and antiepileptic drugs during pregnancy-a matter of importance for the baby?" *J Neurol*
9. Sveberg, Svalheim, Taubøll (2015) "The impact of seizures on pregnancy and delivery" *Seizure*
10. Wickstrom, Taraschenko, Dilena (2022) "International consensus recommendations for management of new onset refractory status epilepticus (NORSE) including febrile infection-related epilepsy syndrome (FIRES): summary and clinical tools" *Epilepsia*
11. Trinka, Cock, Hesdorffer (2015) "A definition and classification of status epilepticus-report of the ILAE task force on classification of status epilepticus" *Epilepsia*
12. Voinescu, Park, Chen (2018) "Antiepileptic drug clearances during pregnancy and clinical implications for women with epilepsy" *Neurology*
13. Veroniki, Cogo, Rios (2017) "Comparative safety of anti-epileptic drugs during pregnancy: a systematic review and network meta-analysis of congenital malformations and prenatal outcomes" *BMC Med*
14. Shi, Wang, Zhang (2022) "Effects of antiepileptic drugs polytherapy on pregnancy outcomes in women with epilepsy: an observation study in northwest China" *Epilepsy Behav*
15. Holmes, Mittendorf, Shen (2011) "Fetal effects of anticonvulsant polytherapies: different risks from different drug combinations" *Arch Neurol*
16. Roberti, Rocca, Iannone (2022) "Status epilepticus in pregnancy: a literature review and a protocol proposal" *Expert Rev Neurother*
17. Meador, Pennell, May (2018) "Changes in antiepileptic drug-prescribing patterns in pregnant women with epilepsy" *Epilepsy Behav*
18. Mølgaard-Nielsen, Hviid (1996) "Newer-generation antiepileptic drugs and the risk of major birth defects" *Jama*
19. Devin, Shaughnessy, Sardana (2025) "The use of newer anti-seizure medicines in women with epilepsy in pregnancy: a case series" *Epilepsy Behav Rep*
20. Honybun, Cockle, Malpas (2024) "Neurodevelopmental and functional outcomes following in utero exposure to antiseizure medication: a systematic review" *Neurology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12151991&blobtype=pdf | # Editorial: Global excellence in virology: Latin America
Laura Mendoza, Henry Puerta-Guardo, Latin America, Curtis Brandt
## Abstract
The Americas are endemic for a wide spectrum of infectious diseases that makes it vulnerable to health emergencies, largely driven by pathogens such as viruses, parasites and bacteria (Pujol and Paniz-Mondolfi, 2024). Population growth and increased migration along with climate change contribute to the increasing emergence and re-emergence of new and old pathogens with high frequency and severity leading to significant impacts on public health systems and well-being of its populations worldwide. According to the World Health Organization (WHO) (WHO, 2025a), Latin American countries are a hotbed of microbial diversity, leading to continuous emergence of new pathogens responsible for large-scale epidemics mainly driven by zoonotic viral diseases (de Thoisy et al., 2024). Of this, RNA viruses are well-known to affect northern, central and suthamerican countries including arthropod-borne viruses of public health significance such as dengue (DENV), Zika (ZIKV), and chikungunya (CHIKV) viruses as well as respiratory viruses such as influenza, and coronaviruses among others. These viruses are diverse in nature, and their host and vectors can adapt easily to new environments, constantly creating newly emerging zoonotic threats due to their expanding geographical circulation (Robert et al., 2020). Some 60% of emerging infectious diseases that are reported globally are zoonoses. Over 30 new human pathogens have been detected in the last three decades, 75% of which have originated in animals, it is of great importance to enhance the capacity for detection and diagnosis in the areas where they are most likely to emerge (WHO).This Research Topic on "Global Excellence in Virology in Latin America: A short journey through air-borne-, vector-borne-, and sexually-transmitted viruses" describes relevant examples of these achievements and discusses ongoing limitations yet existing in the region regarding epidemiological, virological and clinical characterization of infectious diseases such as COVID 19, arthropod-borne virus (arboviral) diseases and other neglected diseases such as the multifocal epithelial hyperplasia having human papillomavirus as the main cause.Over the past three decades, the incidence of arboviral diseases has increased markedly, driven by factors such as climate change, population growth, and global travel, which collectively create favorable conditions for viral transmission. Geographical expansion of viruses transmitted by the Aedes mosquitoes including DENV, ZIKV, and CHIKV in the Americas, represents a burden for healthcare systems in tropical and subtropical regionsFrontiers in Cellular and Infection Microbiology frontiersin.org 01
. One critical issue regarding arboviral infections is the lack of and proper diagnosis particularly in endemic areas. The high rate of asymptomatic or mild cases, and the overlapping symptoms with other diseases, are challenges in accessing timely and accurate testing, particularly in resourcelimited areas such as the primary health cares in many Latinamerican countries which results in reduced accuracy of diagnostics (Coronel-Ruiz et al., 2023).
In this Research Topic, the results of Soto-Garita et al., show the co-circulation of these three viruses (DENV, ZIKV, CHIKV) in Costa Rica during a febrile disease outbreak that occurred between July 2017 and May 2018 (Soto-Garita et al.). More importantly, this study highlights the high number of cases misdiagnosed or not confirmed and exemplifies how difficult the diagnosis of febrile diseases has become in arboviral hyperendemic areas. Therefore, it sheds light on how critical healthcare strategies become in managing arboviral outbreaks, and emphasizes the importance of using comprehensive molecular and serological diagnostic approaches, as well as molecular characterization. Another study by Chen-German et al., shows the importance of the genomic surveillance of DENV to facilitate the detection of new introductions or reintroduction of DENV serotypes and genotypes, as well as genotype replacement, or association of circulating genotypes with new clinical manifestations or higher transmission in the endemic area of Panamá(Chen-German et al.). On note, this study focused on the first cases of autochthonous DENV-4 detected in Panama through the National Surveillance System of Arbovirus after 23 years of no circulation between September 2023 and April 2024. It also highlights how this reemerged DENV-4 genotype had a high similarity to DENV-4 sequences circulating in Nicaragua and El Salvador during the same year 2023.
In 2019, an abrupt outbreak caused by a new coronavirus (SARS-CoV-2) led to a respiratory viral disease pandemic (COVID-19) that rapidly spread worldwide resulting in more than 700 million of cases and 7 million deaths (WHO, 2025b). Despite this, COVID-19 provided an opportunity to reinforce public health capacities, in infrastructure, capacity building, and training for molecular biologists to improve reporting transparency, and enhance regional coordination which leads to strengthening of genomic surveillance strategies based on molecular methods that have been developed in Latin America. Since then, more consistent but still limited research has been carried out in the region to address the viral threats that account for a significant portion of health concerns.
COVID-19 pandemic underlined the importance of molecular diagnosis through RT-PCR and of genomic surveillance for characterizing and monitoring SARS-CoV-2 variant circulation. The regional genomic surveillance of SARS-CoV-2 in Latin America, a collaboration between countries under Pan-American Health Organization (PAHO) guidance, has been focused on characterizing the virus from symptomatic patients and asymptomatic individuals (Gräf et al., 2024). The study by Gaitań et al., presents the initial findings of SARS-CoV-2 detection in sewage water in Panama's capital city and its surrounding areas, comparing these results with data obtained from the clinical genomic surveillance program (Gaitań et al.). The identification of a new variant in the country through wastewater surveillance highlights the critical value of maintaining both SARS-CoV-2 genomic surveillance in clinical samples and wastewater monitoring to support a more comprehensive and integrated surveillance system.
On the other hand, genetic association studies of the COVID-19 phenotype are very limited in Latin America (Ferreira de Araujo et al., 2022). The study by Chavez-Veĺez et al, helps to clarify the role of genetic factors in the COVID-19 phenotype. The experience with COVID-19 provided an opportunity to identify an ethnicitybased approach to recognize genetically high-risk individuals in different populations for emerging diseases (Chavez-Veĺez et al).
Identifying high-risk individuals who need urgent medical attention is especially important during epidemics. It can be very helpful in formulating policies and allocating resources.
Furthermore, the study by De la Cruz Montoya et al., through phylogenetic analyses observed genomic transitions of SARS-CoV-2 viruses from several lineages dominant in the first wave versus a second wave of the COVID-19 pandemic in Mexico city. This study contributes to a better understanding of the evolutionary dynamics and selective pressures that act at the genomic level, the prediction of more accurate variants of clinical significance, and a better comprehension of the molecular mechanisms driving the evolution of SARS-CoV-2 to improve vaccine and drug development (De la Cruz Montoya et al).
Finally but not least, Conde-Ferraéz and Gonzaĺez-Losa gave us an updated view of the Multifocal Epithelial Hyperplasia (MEH), also known as Heck's disease, a rare pathology of the oral mucosa associated with the infection of human papillomavirus types 13 and 32, which affects significantly indigenous groups around the world, particularly in the Yucatan peninsula (Conde-Ferraéz and Gonzaĺez-Losa). This minireview highlights how the MEH is considered as neglected by research indicating that only clinical cases of MEH have been reported so far, with yet undetermined biological factors involved. Furthermore, it warrants that additional studies must be performed in these communities particularly those inhabiting rural/remote areas in order to better understand the epidemiological and host determinants associated with its high incidence in indigenous populations.
The continuous evolution of viruses through the concerted action of mutational forces that challenge human immunity and vaccine development poses an enormous burden on the public health systems worldwide. In many Latin-ameŕica countries, we still have numerous limitations in our public health capacity to implement genomic surveillance program, or to increase the access to molecular test for improve diagnosis and characterization of important health problems such as arboviral diseases (e.g. dengue, Zika), respiratory infections (e.g. COVID-19) or better understanding the host or viral determinants conditioning viral rare diseases such as the multifocal epithelial hyperplasia (MEH). Overall, this selection of scientific manuscripts and reviews highlights the importance of building research networks Latin American countries bringing together academic and non-academic institutions to enhance capacity building and to generate the data needed to design or strengthen infectious disease control strategies. We believe this set of manuscripts will facilitate further interactions among all involved institutions and researchers that contributed to this Research Topic. As editors of the "Global Excellence in Virology in Latin America" Research Topic, we would like to acknowledge all contributing authors for providing insight into the exciting research field of the viral infectious diseases for better understanding of the virus-hostpathogen interactions at molecular and ecological levels, and translation into new interventions for the control of many viral infections that significantly affect the well-being of Latin American populations.
## References
1. Abbasi (2025) "Global expansion of Aedes mosquitoes and their role in the transboundary spread of emerging arboviral diseases: A comprehensive review" *IJID. One Health*
2. Coronel-Ruiz, Velandia-Romero, Calvo et al. (2023) "Improving dengue diagnosis and case confirmation in children by combining rapid diagnostic tests, clinical, and laboratory variables"
3. De Thoisy, Gräf, Mansur et al. (2024) "The risk of virus emergence in south america: A subtle balance between increasingly favorable conditions and a protective environment" *Annu. Rev. Virol. 1)*
4. Ferreira De Araujo, Menezes, Saraiva-Duarte et al. (2022) "Systematic review of host genetic association with Covid-19 prognosis and susceptibility: What have we learned in 2020?" *Rev. Med. Virol*
5. Gräf, Martinez, Bello et al. (2024) "Dispersion patterns of SARS-CoV-2 variants Gamma, Lambda and Mu in Latin America and the Caribbean"
6. Robert, Stewart-Ibarra, Estallo (2020) "Climate change and viral emergence: evidence from Aedes-borne arboviruses" |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12282424&blobtype=pdf | # Corrigendum to "Evaluating the performance of the PREDAC method in flu vaccine recommendations over the past decade (2013-2023)" [Virol. Sin. 40 (2025) 288-291]
Yousong Peng, Lei Yang, Weijuan Huang, Mi Liu, Xiao Ding, Xiangjun Du, Yuelong Shu, Taijiao Jiang, Dayan Wang, Virologica Sinica
## Abstract
During the final proofing stage of the paper, the wrong version of Fig. 2 was accidently used when replacing it with a high-resolution version. The star and circle marks were missing in the published version.The correct Fig. 2 is given below. We apologize for our oversight when preparing the figure and state that this does not change the scientific conclusions of the article in any way.
## Fig. 2. Validation of vaccine recommendations of influenza A(H3N2
) viruses by PREDAC method and WHO. Antigens were colored as indicated by the figure legend in the bottom. The gray refers to minor antigenic clusters which didn't predominate in any influenza season. The ticks refer to that the recommended vaccine strains had the same antigenicity as the dominant strains; the stars refer to that the recommended vaccine strains were antigenic similar to the dominant strains; the gray circles refer to no recommendations by PREDAC in 2021 due to the COVID-19. The WHO reported antigens were compiled from the vaccine recommendation reports by WHO. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584662&blobtype=pdf | # STING-mediated antiviral response: insights into MVA replication control in avian cells
Teresa Brusco, Valentina Menci, Carmen Caiazza, Anna Petrone, Renata Palladino, Matteo Faticanti, Veronica Bignone, Concetta Ambrosino, Elisa Scarselli, Massimo Mallardo, Loredana Siani, Valentino Ruzza
## Abstract
The safety-tested Modified Vaccinia virus Ankara (MVA) is a well-character ized mutant virus widely used in fundamental research to elucidate the functions of Poxvirus host-interaction factors. Beyond its safety profile, MVA is an attractive viral vector for vaccine development due to its genetic stability and ability to efficiently infect antigen-presenting cells, such as dendritic cells and tumor cells. In this report, we investigated the interplay between MVA and the cyclic GMP-AMP synthase-stimula tor of interferon genes (cGAS-STING) antiviral pathway in chicken fibroblast cell lines (wild-type DF-1 and knock-out STING) to verify whether manipulation of the STING axis could impact MVA replication and cell responses. Our findings demonstrate that STING-mediated signaling plays a role in contrasting the replication of MVA. Upon MVA infection, the loss of STING hampered the expression of type I interferons (IFNs) and, in turn, interferon-stimulated gene 15 (ISG15) and interferon-induced transmembrane protein 3 (IFITM3). In line with these results, the expression of early and late MVA genes was enhanced, and DNA replication occurred earlier and was more abundant. Interferon regulatory factor 1 (IRF1) and myeloid differentiation primary response 88 (MyD88) were significantly induced by MVA infection in STING-KO cells, indicating that their responses to MVA infection are independent of the cGAS/STING axis. Collectively, these results refine our knowledge of MVA-host interaction in chicken fibroblasts and offer insights to guide strategies for enhancing Poxvirus vaccine vector production. IMPORTANCE Given the context-dependent nature of STING antiviral activity, it is critical to broaden the investigation in order to clarify the virus-host response mechanisms across different species, particularly in chicken fibroblasts, to provide insights into MVA-based vaccine production improvements. KEYWORDS viral-host interaction, STING, MVA, avian cells, innate immunity T he properties that made Modified Vaccinia Ankara (MVA) virus an attractive antigen-delivery vehicle are essentially attributable to the high-level transgene expression and the cargo capacity (1), the avirulence and inability to replicate in human cells (2), and the general safety profile, already observed in the over 120,000 healthy volunteers vaccinated during the last smallpox eradication campaign (1, 3).MVA was derived in the seventies from the parental strain Chorioallantois Vaccinia Ankara (CVA) virus after more than 570 sequential passages in chicken embryo fibroblast cells (CEFs) (4, 5). During this adaptation, MVA lost approximately 30 Kbp of its genome, consisting of six major deletions and further gene truncations/mutations. These changes were shown to be decisive for its attenuated virulence, restricted host range replication, and safety while preserving its high immunogenicity as a viral vector (6, 7).Although Vaccinia virus has the ability to contrast host innate immune response by releasing soluble inhibitors that interfere with cytokine, chemokine, and interferon
activities (8)(9)(10), MVA has the unique immunological property to stimulate adequately host innate immune response without releasing any interferon α/β (IFNα/β), interferon γ (IFNγ), and tumor necrosis factor (TNF) inhibitor protein, thus representing an interesting profile for an immunogenic but safe vaccine (8).
Indeed, MVA stimulates IFNα/β production in murine conventional dendritic cells mostly via the cytosolic DNA-sensing axis cyclic GMP-AMP synthase/ stimulator of interferon genes (cGAS)/STING/Interferon regulatory factor 3 (IRF3), with interferon regulatory factor 7 (IRF7) and interferon alpha and beta receptor subunit 1 (IFNAR1) amplifying the signal, but with the toll-like receptor 7-9/myeloid differentiation primary response 88 (TLR7-9/MyD88) pathway playing a minor role in activating innate immune response (9).
The STING is a transmembrane protein located in the endoplasmic reticulum and encoded by the transmembrane protein 173 (TMEM173) gene (10). Initially identified as a crucial player in host defense against viral infections (11), it has since been implicated in a broader range of innate immune processes, as well as in adaptive immunity regulation and in immunotherapy (12,13). Cytosolic DNA, originating either from infections, endogenously through phagocytosis, or released from mitochondria, initiates the production of 2′3′-cGAMP (cG[2′-5′]pA[3′-5′]p) via the enzyme cGAS (14). This 2′3′-cGAMP is recognized by the adaptor protein STING, facilitating its translocation to the Golgi apparatus and promoting the interaction with TANK-binding kinase 1 (TBK1), which next activates the transcription factor IRF3 (15,16). IRF3 dimerizes and enters the nucleus to initiate a type I Interferon (IFN-I) response, leading to the expression of a set of interferon-stimulated genes (ISGs) and the establishment of antimicrobial immunity and inflammatory state (17). Moreover, the activation of cGAS-STING signaling also leads to nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) activation, whose regulation relies upon TNF receptor-associated factor 6 (TRAF6), NF-κB essential modulator (NEMO), IkappaB kinase beta (IKKβ), and TBK1 (18). Thus, the immune pathway generated by the dsDNA-sensing cGAS-STING can both counteract viral replication and spreading. As confirmation of these issues, we have previously demonstrated that STING knockout (KO) cell lines are more permissive to the infection of an oncolytic herpes simplex virus (HSV) compared with their wild-type counterparts (19). Moreover, in the STING knockout cell line, the HSV spread was faster and evidenced the formation of larger lysis plaques compared with the wild-type cell line (19).
In this article, we demonstrate that STING-mediated signaling plays a role in contrasting the replication of MVA in the spontaneously immortalized chicken embryo fibroblast DF-1 cells. Remarkably, following MVA infection in DF-1 STING Knock-out (DSK) cells, we observed reduced expression of IFNα and IFNβ genes along with Interferonstimulated gene 15 (ISG15), a major protein induced by type I interferons (20), interferoninduced transmembrane protein 3 (IFITM3), which interferes with vaccinia virus entry into the cytoplasm (21), and other interferon-stimulated genes (ISGs).
Consequently, upon MVA infection, DSK cells exhibit more cytoplasmic replication centers, higher expression of early and late viral genes, improved transgene expression, and an overall increase in viral titer compared with the DF-1 WT cells.
Notably, TLR signaling genes interferon regulatory factor 1 (IRF1) and MyD88 are significantly induced by MVA infection in DSK cells, probably to counterbalance the reduced IFN-I response of the cytosolic DNA-sensing cGAS/STING/IRF3 axis.
## RESULTS
## STING knockout cell line generation and characterization
To investigate the cGAS-STING signaling role as mediator of the cellular response upon MVA infection, a DF-1 chSTING KO cell line was generated. Two guide RNAs (gRNAs) targeting the third exon of the STING chicken gene were designed (Fig. 1A), and each gRNA was cloned into the PX458 plasmid. The resulting plasmids contain the gRNA and the Cas9 protein fused at the N-terminus with an enhanced green fluorescent protein (EGFP) tag. Following co-transfection with the two plasmids, single-cell clones were isolated by limiting dilution and screened by PCR (Fig. S1) using primers designed to amplify a region spanning from the 5' untranslated region (UTR) to the intron 4 of the genomic DNA sequence (Fig. 1A, zoomed-in section). One clone, designated DSK, exhibited heterozygous mutations in the STING locus. As confirmed by Sanger sequenc ing, the mutations in this clone consist of a single point deletion on one allele and a 107 bp deletion on the other, thus resulting in predicted nonfunctional peptides, lacking all the functional domains (i.e., the serine 366) and even the first transmembrane and loop domains essential for the ER translocation (Fig. 1B). Consequently, quantitative PCR (qPCR) analysis revealed reduced STING transcriptional levels (Fig. 1C), using both oligo pairs within the 107 bp deletion sequence and targeting exon 4 (Fig. 1A). Taken together, these results indicate that in DSK cells, the STING function is impaired.
## Disruption of the STING gene enhances MVA viral growth
In order to evaluate whether the STING knock-out had an impact on MVA viral cycle, a viral growth assessment by infecting DSK and DF-1 cells at MOI (Multiplicity Of Infec tion) 0.03 with MVA expressing HcRed fluorescent protein (MVA-red) was performed. Seventy-two hours post-infection, the cells were collected by scraping and lysed with one freeze-thaw cycle and sonicated for 10 minutes before the virus was titrated by immunoassay (Fig. 2A). The analysis revealed a 1.6-fold increase in the infectious unit per cell (ifu/cell) production in DSK compared with DF-1 WT cells. To better characterize the viral growth increase, an infection time-course experiment was performed, harvesting cells at 24, 48, and 72 h post-infection. Immunoassay titration confirmed an approxi mately 1.6-fold increase in MVA ifu/cell titer in DSK not only at 72 h post-infection but also at 24 and 48 h post-infection (Fig. 2B). However, since the infectious titer may not correlate with viral DNA levels, a qPCR assay amplifying the MVA DNA polymerase gene (E9L) was performed to measure only encapsidated viral genomes. The results showed that the viral particle per cell (vp/cell) concentration was similar between KO and WT cells at 48 and 72 h, whereas at 24 h, it was approximately 1.9 times higher in DSK compared with WT cells (Fig. 2C). Considering the flattening of the viral particle concentration at 48 and 72 h, the virus exhibits higher infectivity in DSK cells at later time points as indicated by the vp/ifu ratio (Fig. 2D).
To support these observations, the cells were infected with MVA-red at different MOIs and harvested at 6, 24, and 48 h post-infection for flow cytometry analysis, looking at the expression of the viral red fluorescent protein. As shown in Fig. 2E, mean fluorescent intensity (MFI) measurements revealed no differences at 6 h but were significantly higher in DSK compared with DF-1 WT cells at later time points. On the other side, Fig. 2F revealed that DSK cells exhibited a higher number of infected cells compared with DF-1 WT at all three times of harvest.
Together, these findings demonstrate that the absence of STING promotes MVA viral growth and infectivity. Furthermore, the higher percentage of infected cells at 6 h may reflect the enhanced expression of early genes in STING-lacking cells.
## Early and late MVA gene expression is enhanced in DSK-infected cells
In order to assess whether STING loss of function affects early and late stages of the MVA replicative cycle, the expression of early and late viral genes was analyzed in infected DF-1 WT and DSK cells. We selected E3L, K1L, and F1L as MVA early genes, and B16R and B8 as late genes. To avoid interference in gene expression analysis due to the presence of multiple viral particles in a single cell, DF-1 and DSK cells were infected at 0.2 MOI. Cells were harvested at different time points, and the total RNA was analyzed by quantitative reverse-transcription PCR (qRT-PCR). As shown in Fig. 3 (panels A,B), all the tested early genes showed earlier and more abundant expression in DSK cells compared with the WT. This effect was evident as early as 30 min post-infection, reaching peak fold induction at 60 min, whereas at 2 h post-infection, the differences in the early gene expression between DF-1 and DSK were observed to a lesser extent. Similarly, the late gene expression showed an increase in the expression of the DSK cells from 2 to 4 h post-infection, peaking at 3 h (B16R) and 4 h (D8) (Fig. 3A andB).
## The lack of STING influences MVA DNA replication
We have previously demonstrated that there is a strict correlation between Vaccinia Virus early genes expression and the replication of viral DNA (22). Thus, we wanted to assess whether the lack of STING could also affect the DNA replication of MVA. To achieve this, MVA infection with 0.1 MOI was performed, harvesting the cells at the indicated time points post-infection. Viral entry was synchronized as described in the methods, and following incubation for 15 min at 37°C, the unbound viral particles were washed. As shown in Fig. 4A, in DSK cells, some DNA replication centers were already visible at 60 min post-infection (mpi) compared with the 90-120 mpi observed in DF-1 cells. Moreover, the number of DNA replication centers per nucleus was higher and larger in size in the DSK cells at 90 mpi, reaching peak induction at 120 mpi (Fig. 4B).
Altogether, these data confirm that STING is also involved in the modulation of MVA replication center number and formation.
## Analysis of STING downstream gene expression upon MVA infection
Several genes are regulated by STING activation upon viral infection. Among them, we wanted to analyze the expression of interferon α and β, IFITM3, ISG15, MyD88, and IRF1 in DF-1 and DSK cells upon MVA infection and verify whether their expression is affected when STING functionality is impaired. Type I interferon subtypes IFNα and IFNβ are transcriptionally activated by the infection-mediated phosphorylation of STING (23,24). By qRT-PCR, we observed that both genes are activated at 4 h post-MVA infection in DF-1 cells. Conversely, in DSK cells, the INFα and INFβ expression was downregulated at 4 h post-MVA infection (Fig. 5A andB). The same result was observed when analyzing the expression of IFITM3 (Fig. 5C), a gene that is induced by STING via the activation of type I IFN (25). We further analyzed the expression of ISG15, a gene encoding for a ubiquitinlike molecule that is highly induced by type I IFN during infection by viral and bacterial pathogens (26). As expected, ISG15 was upregulated by MVA infection in DF-1 cells, whereas its expression rate was comparable in mock-infected vs. infected DSK cells (Fig. 5D). The myeloid differentiation marker MyD88 is able to complex with STING, prevent ing its autophagic degradation (27). Moreover, MyD88-STING complex is required for lipopolysaccharide (LPS)-induced aconitate decarboxylase 1 (ACOD1) expression during septic shock. We wanted to investigate the expression of MyD88 upon MVA infection. As shown in Fig. 5E, MyD88 is slightly upregulated at 4 h post-infection in the DF-1 cells, whereas it is strongly upregulated in the DSK cells. Finally, we analyzed the expression rate of IRF1. Our analysis reveals that at 4 h post-MVA infection, IRF1 expression is specifically upregulated in DSK cells, whereas no such upregulation is observed in WT DF-1 cells (Fig. 5F). These findings collectively demonstrate that in chicken fibroblasts, impairment of STING function significantly affects the primary downstream effectors of the cGAS/STING signaling pathway.
## DISCUSSION
MVA has the unique immunological property to fine-stimulate host innate immune response, representing an interesting profile for an immunogenic but safe vaccine (9).
STING is one of the key host genes, mediating the innate immune response upon MVA infection (9). However, the mechanisms behind its role in counteracting and restricting viral replication remain not completely understood.
Specifically, the role of the cGAS-STING signaling pathway in activating antiviral responses within chicken cells following MVA infection has yet to be thoroughly investigated (28). Indeed, we generated a DF-1 STING Knock-out (DSK) cell line by means of CRISPR CAS9-mediated genome editing technique and evaluated the impact on MVA replication efficiency and cell interferon response in comparison to wild-type counterpart.
The loss of STING activity in DSK cells leads to a significant increase in the infective titer of MVA across all tested time points (Fig. 2B). Conversely, the number of genome particles per cell in DSK cells is higher only during the initial phase of infection; over time, it becomes comparable with that of wild-type cells, suggesting that the DNA replication curve has likely reached a plateau (Fig. 2C). As a result, during the initial phase of infection, the ratio of viral particles to infectious particles (vp/ifu) is similar between DSK and DF-1 cells; however, at later stages, this ratio is significantly lower in DSK cells, indicating higher infectivity (Fig. 2D). These data suggest that STING may have a dual impact on the MVA replication cycle, affecting both the initial phases of infection and the processes of viral assembly and maturation at the ER and/or trans-Golgi network levels.
The ability of MVA to replicate in the absence of STING was also evaluated through flow cytometry, by infecting DSK and DF-1 cells with an MVA-red virus at both high (10) and low (0.03) MOI, examining both the early and late phases of infection. At high MOI and during the early phase of infection, the percentage of fluorescent cells is meaningfully higher in DSK cells, indicating that MVA replication is enhanced in a STING knockout (KO) cell background (Fig. 2F). However, the MFI at this time point is comparable between the two cell lines (Fig. 2E), probably because the selected time point is still too premature to detect a measurable increase in mean fluorescence intensity between WT and DSK cells. At low MOI and late infection phase, an increase in both fluorescent cell percentage and MFI in DSK compared with DF-1 cells can be appreciated, highlighting again the inhibiting impact of STING on MVA replication and maturation.
In summary, with high MOI at the early phase of infection, it is evidenced that MVA gene expression is faster in the absence of STING activity. On the other hand, in the advanced phase of infection, both the percentage of the infected cells and the MFI are consistently higher in DSK cells, highlighting that STING can intervene at different steps of the MVA replication cycle.
A distinctive feature of Poxviruses lies in their ability to replicate in the cytoplasm of the host cell. In particular, vaccinia virus localizes in discrete cytoplasmic nuclei called virosomes or factories (29), where, after DNA replication and late gene expression (~6 h), the intracellular mature virus (IMV) assembly starts with the support of mem branes borrowed from the ER. Later on, some extracellular enveloped virus particles are eventually formed in the trans-Golgi network (29).
We observed a notable increase in the number of MVA factories in DSK compared with DF-1 cells, detectable as early as 60 min post-infection (mpi). By 180 mpi, DSK cells show an average of ~3-4 factories per cell, whereas DF-1 cells exhibit less than one (Fig. 4A andB). These findings further support the hypothesis that MVA replication is enhanced in a cellular environment lacking STING activity.
MVA early gene transcription begins within 20 min of infection, and approximately 2 h later, DNA replication and the transcription of late genes start. By 6 h, new virus particles are ready to be assembled (29). To further investigate the dynamics of STING's impact on MVA replication, we assessed the kinetics of early and late viral gene expressions in DSK compared with DF-1 cells (Fig. 3A andB). The results showed that in DSK cells, both early and late gene expressions were higher, indicating that the absence of STING enhances the early stages of the viral replication cycle. However, the expression of early, intermediate, and late genes is not synchronous because some genes are expressed at multiple stages due to their hybrid promoters containing more than one promoter motif (early and late), which have not been completely characterized yet (30).
Microarray analyses revealed the induction of host resistance and immune modu lation genes following infection of HeLa cells with MVA (31). To further explore the mechanisms of MVA infection, we assessed the interferon response and interferon-stimu lated genes, including interferon α and β, IFITM3, ISG15, MyD88, and IRF1 in DSK cells and examined how their expression is affected when STING functionality is impaired.
Type I interferons α and β are the main effector genes of the antiviral c-GAS-STING pathway (24). Using qRT-PCR, we analyzed their expression levels after 4 h of MVA infection in the DF-1 and DSK cell lines. As expected, both genes were upregulated in the WT DF-1, whereas the ablation of STING impaired their expression (Fig. 5A andB). Interferon-inducible transmembrane 3 (IFITM3) is an intrinsic antiviral effector able to restrict viral cytosolic entry by blocking endosomal and membrane fusion pathways. Indeed, IFITM3 overexpression significantly reduces vaccinia virus (VACV) levels in 293T and Vero cells (21,32). Here, we demonstrate that upon MVA infection, mRNA levels of IFITM3 are reduced in DSK cells compared with DF-1 (Fig. 5C), highlighting the role of STING in amplifying and sustaining antiviral response following Poxvirus infection. Moreover, the MVA-mediated upregulation of the interferon-stimulated gene 15 (ISG15) observed in the DF-1 was also hampered in the DSK cells. The lack of ISG15 activation can be explained by the fact that it is highly induced by type I IFN. It has been shown that in ISG15-deficient THP-1 cells, the infection of HIV-1 was enhanced in both undifferentiated and phorbol-12-myristate-13-acetate (PMA)-differentiated ISG15-deficient THP-1 cells compared with the control (33). Thus, the lack of STING, affecting the expression of type I IFN and ISG15, may explain the observed enhancement of MVA replication.
MyD88 and STING may form a complex that acts as adaptor proteins in innate immunity (27). However, it has been shown that the depletion of STING had no effect on the level of MyD88 protein. As we observed that MyD88 is only slightly upregulated at 4 h post-infection in the DF-1 cell but is strongly upregulated in the DSK cells, we may speculate that the mechanism by which MyD88 is involved in the innate immunity is not mediated by the transcriptional regulation of the encoding gene. Moreover, we analyzed the expression rate of the interferon regulatory factor 1 (IRF-1). Surprisingly, we found that IRF1 is not transcriptionally activated in DF-1, whereas its transcription is enhanced in DSK cells. However, it is reported that MVA infection induces the production of IFN-I in dendritic cells, mainly via STING activation of the transcription factors IRF3 and IRF7 (27). In addition, it is reported that TLR9-triggered MyD88 activation caused not only nuclear translocation of IRF1 but also activated its transcription factor function in murine macrophages (34). In summary, in chicken cells, we observed that IFN-I-mediated activation of MyD88/IRF1 occurs only in the absence of STING, highlighting that this is secondary to the IRF3/IRF7 pathway in responding to MVA infection.
## Conclusions
Overall, several pieces of evidence suggest that STING plays multiple roles in counteract ing MVA during the viral replication cycle.
First, viral replication is facilitated in DSK cells, as evidenced by a higher number of infective and viral particles during the early phase of infection, an early and substantial increase in both early and late viral gene expression, and a concurrent increase in the number of cytosolic replication centers.
Second, the initial augmented number of viral particles in DSK cells tends to equalize the number of viral particles in DF-1 cells, probably due to a saturation phenomenon. Nevertheless, the DSK cells present an advantage during the viral maturation phase as the number of infective particles is significantly and constantly higher than that of DF-1, indicating that STING intervenes not only at the viral entry phase but also during the final morphogenesis of new infectious virions.
Interestingly, it has been demonstrated that STING does not impact the replication of MVA in the permissive cell line BHK21. In particular, the growth kinetics of MVA in BHK21 STING KO cells are indistinguishable from those shown by their wild-type counterpart (35). In BHK21 cells, although the expression of some ISGs is upregulated upon MVA infection, the cGAS/STING axis does not serve as the primary mediator of the antiviral response, suggesting alternative sensing mechanisms.
Conversely, in our report, we show that the transcriptional activation of type I IFN, IFITM3, and ISG15 is hampered in MVA-infected DSK cells. Moreover, it has been shown that in duck cGAS-KO fibroblasts, duck adenovirus replication is enhanced, accompa nied by the inactivation of type I IFN and ISGs, suggesting a species-specific response downstream the cGAS/STING axis (36).
Given the context-dependent nature of their antiviral activity, it is critical to broaden the investigation in order to clarify the virus-host response mechanisms in the different species, particularly in chicken fibroblasts, to give insights into MVA-based vaccine production improvements.
In summary, the results suggest that STING could counteract MVA at different levels of its replication cycle. Interestingly, by using an optimized version of the proximitydependent biotin identification (BioID) technique applied to detect protein-protein interactions in living cells, Motani and Kosako identified several STING interactors, including IFITM3 (37). Likely, in collaboration with IFITM3, which is dysregulated upon infection in DSK cells, STING may counteract MVA, particularly during the entry phase, by targeting viral particles for elimination through the endosomal pathway.
CCATTTAGTACCGATTCT-3′, and probe 5′-AGGACGTAGAATGATCTTGTA-3′ (FAM-TAMRA) (from Baker and Ward, 2014). The qPCR assay was performed using TaqMan Universal PCR Master Mix (Applied Biosystems). A standard curve was generated with a plasmid containing 1,000 bp of the E9L viral gene.
MVA infectious titer was obtained with an immunostaining assay executed on monolayers of Vero cells. At 48 h post-infection, the infected cells were detected using a rabbit polyclonal anti-vaccinia primary antibody (Abcam), followed by an anti-rabbit horseradish peroxidase (HRP)-conjugated secondary antibody (Sigma).
## FACS analysis
For fluorescent-activated cell sorter (FACS) analysis, the cells (infected or not infected) were scraped and centrifuged. After washing cell pellets with PBS, the samples were stained with LIVE/DEAD Fixable Near-IR Dead Cell Stain Kit (Thermo Fisher Scientific) 1:100 in PBS for 15-20' . The cells were centrifuged again and resuspended in PBS. Flow cytometry was performed using BD FACS Canto II, and data were analyzed using FlowJo v10 software.
## Immunofluorescence assay
To assess the micronuclei formation, DF-1 cells were infected with MVA-red at an MOI of 0.1 under the following conditions. The cells were seeded onto glass coverslips and incubated overnight under standard growth conditions. The next day, MVA-red at a MOI of 0.1 was added to the culture medium, and the cells were incubated on ice for 30 min, followed by incubation at 37°C for the designated times. The cells were then washed three times with PBS and fixed with 3.7% formaldehyde for 30 min at room temperature, followed by a 10 min incubation with 0.1 M glycine. Nuclear staining was performed using Hoechst (Invitrogen H3570) for 5 min at room temperature, protected from light. Coverslips were subsequently washed three times with PBS and mounted onto glass slides with a 1:1 PBS/glycerol solution. Images were acquired using a laser scanning microscope (LEICA DMi8) and analyzed using LEICA LAS X software.
## Gene expression analysis
RNA was extracted using Direct-zol RNA Miniprep (Zymo Research), and cDNA synthe sis was performed with SuperScript IV VILO Master Mix (Invitrogen) according to the manufacturer's instructions.
RT-PCR was performed using the PowerUp SYBR Green Master Mix (Applied Biosystems) using the oligonucleotides listed in Table S1.
Relative gene expression was analyzed with the 2 -ΔΔCt method using GAPDH as an endogenous housekeeping gene.
## Statistical analyses
Statistical analyses were performed using unpaired t-test method. P values less than 0.05 were considered significant. The results are represented as mean ± standard error of the mean (SEM).
## References
1. Sutter, Moss (1992) "Nonreplicating vaccinia vector efficiently expresses recombinant genes" *Proc Natl Acad Sci*
2. Sancho, Schleich, Griffiths et al. (2002) "The block in assembly of modified vaccinia virus Ankara in HeLa cells reveals new insights into vaccinia virus morphogenesis" *J Virol*
3. Stickl, Hochstein-Mintzel, Mayr et al. (1974) "MVA vaccination against smallpox: clinical tests with an attenuated live vaccinia virus strain (MVA)"
4. *Dtsch Med Wochenschr*
5. Volz, Sutter (2017) "Modified vaccinia virus ankara: history, value in basic research, and current perspectives for vaccine development" *Adv Virus Res*
6. Meisinger-Henschel, Schmidt, Lukassen et al. (2007) "Genomic sequence of chorioallantois vaccinia virus Ankara, the ancestor of modified vaccinia virus Ankara" *J Gen Virol*
7. Meisinger-Henschel, Späth, Lukassen et al. (2010) "Introduction of the six major genomic deletions of modified vaccinia virus Ankara (MVA) into the parental vaccinia virus is not sufficient to reproduce an MVA-like phenotype in cell culture and in mice" *J Virol*
8. Meyer, Sutter, Mayr (1991) "Mapping of deletions in the genome of the highly attenuated vaccinia virus MVA and their influence on virulence" *J Gen Virol*
9. Orlova, Glazkova, Bogoslovskaya et al. (2022) "Development of modified vaccinia virus ankara-based vaccines: advantages and applications" *Vaccines (Basel)*
10. Dai, Wang, Cao et al. (2014) "Modified vaccinia virus Ankara triggers type I IFN production in murine conventional dendritic cells via a cGAS/STING-mediated cytosolic DNA-sensing pathway" *PLoS Pathog*
11. Keskitalo, Haapaniemi, Einarsdottir et al. (2019) "Novel TMEM173 mutation and the role of disease modifying alleles" *Front Immunol*
12. Liu, Pang, Zhang (2021) "The cGAS-STING pathway in bacterial infection and bacterial immunity" *Front Immunol*
13. Caiazza, Brusco, 'alessio et al. (2022) "The lack of STING impairs the MHC-I dependent antigen presentation and JAK/STAT signaling in murine macrophages" *Int J Mol Sci*
14. Xu, Xiong (2024) "Targeting STING signaling for the optimal cancer immunotherapy" *Front Immunol*
15. Chen, Sun, Chen (2016) "Regulation and function of the cGAS-STING pathway of cytosolic DNA sensing" *Nat Immunol*
16. Ikeda, Hecker, Rozenknop et al. (2007) "Involvement of the ubiquitin-like domain of TBK1/IKK-i kinases in regulation of IFN-inducible genes" *EMBO J*
17. Motwani, Pesiridis, Fitzgerald (2019) "DNA sensing by the cGAS-STING pathway in health and disease" *Nat Rev Genet*
18. Tamura, Yanai, Savitsky et al. (2008) "The IRF family transcription factors in immunity and oncogenesis" *Annu Rev Immunol*
19. Liu, Zhang, Joo et al. (2017) "NF-κB signaling in inflammation" *Signal Transduct Target Ther*
20. Froechlich, Caiazza, Gentile et al. (2020) "Integrity of the antiviral STING-mediated DNA sensing in tumor cells is required to sustain the immunotherapeutic efficacy of herpes simplex oncolytic virus" *Cancers (Basel)*
21. Perng, Lenschow (2018) "ISG15 in antiviral immunity and beyond" *Nat Rev Microbiol*
22. Li, Du, Tian et al. (2018) "The host restriction factor interferon-inducible transmembrane protein 3 inhibits vaccinia virus infection" *Front Immunol*
23. Mallardo, Leithe, Schleich et al. (2002) "Relationship between vaccinia virus intracellular cores, early mRNAs, and DNA replication sites" *J Virol*
24. Guerra, Cáceres, Knobeloch et al. (2008) "Vaccinia virus E3 protein prevents the antiviral action of ISG15" *PLoS Pathog*
25. Ishikawa, Ma, Barber (2009) "STING regulates intracellular DNAmediated, type I interferon-dependent innate immunity" *Nature*
26. Diamond, Farzan (2013) "The broad-spectrum antiviral functions of IFIT and IFITM proteins" *Nat Rev Immunol*
27. Lin, Kuffour, Fuchs et al. (2023) "Regulation of STING activity in DNA sensing by ISG15 modification" *Cell Rep*
28. Chen, Wu, Liu et al. (2022) "The STING1-MYD88 complex drives ACOD1/IRG1 expression and function in lethal innate immunity"
29. Cheng, Sun, Wang et al. (2015) "Chicken STING mediates activation of the IFN gene independently of the RIG-I gene" *J Immunol*
30. Tolonen, Doglio, Schleich et al. (2001) "Vaccinia virus DNA replication occurs in endoplasmic reticulum-enclosed cytoplasmic mini-nuclei" *Mol Biol Cell*
31. Deng, Navarro-Forero, Yang (2025) "Temporal expression classes and functions of vaccinia virus and mpox (monkeypox) virus genes"
32. Guerra, López-Fernández, Conde et al. (2004) "Microarray analysis reveals characteristic changes of host cell gene expression in response to attenuated modified vaccinia virus Ankara infection of human HeLa cells" *J Virol*
33. Ren, Wang, Xie et al. (2024) "Analysis of chicken IFITM3 gene expression and its effect on avian reovirus replication" *Viruses*
34. Kuffour, König, Häussinger et al. (2019) "ISG15 deficiency enhances HIV-1 infection by accumulating misfolded p53" *mBio*
35. Schmitz, Heit, Guggemoos et al. (2007) "Interferon-regulatory-factor 1 controls Toll-like receptor 9-mediated IFN-beta production in myeloid dendritic cells" *Eur J Immunol*
36. Hood, Sumner (2022) "Disruption of the cGAS/STING axis does not impair sensing of MVA in BHK21 cells" *J Gen Virol*
37. Lin, Zheng, Wang et al. (2023) "Duck cGAS inhibits DNA and RNA virus replication by activating IFNs and antiviral ISGs" *Front Immunol*
38. Motani, Kosako (2020) "BioID screening of biotinylation sites using the avidin-like protein Tamavidin 2-REV identifies global interactors of stimulator of interferon genes (STING)" *J Biol Chem*
39. Lullo, Soprana, Panigada et al. (2009) "Marker gene swapping facilitates recombinant modified vaccinia virus Ankara production by host-range selection" *J Virol Methods* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12724325&blobtype=pdf | # Ceramide synthase 4 interferes with replication of influenza virus but is downregulated by infection
Il Kwang, Jung, Chuan Xia, Savannah Mckenna, Ying He, Vijayamahantesh Vijayamahantesh, Jennifer Wolf, Bumsuk Hahm
## Abstract
Influenza continues to pose a serious threat to humans. Influenza-host interaction is incompletely understood, requiring identification of host factors that regulate viral pathogenicity. Ceramide synthases (CerSs) are responsible for produc ing and controlling ceramide levels within cells. Ceramides are structural and signal ing sphingolipid components that mediate various biological functions and affect the infectivity of multiple viruses. However, the role of CerSs during virus infections remains unclear. In this study, we investigated the possible function of CerSs in host defense against influenza virus infection. Cells stably expressing CerS4 poorly supported influenza virus replication, whereas CerS1 did not affect replication. Transient overex pression of CerS4 also impaired the efficient production of viral proteins as well as infectious progeny viruses. In support of these results, knockdown of endogenous CerS4 in cells enhanced virus replication. Intriguingly, CerS4 impeded virus-induced activation of cellular c-Jun N-terminal kinase (JNK), which interfered with influenza viral replication. On the other hand, influenza virus infection was shown to induce CerS4 ubiquitination and downregulation, which could limit the antiviral activity of CerS4. Collectively, these findings reveal a new function of CerS4 that restricts influenza virus infection and provides valuable insights into influenza-host defense interactions. IMPORTANCE Seasonal influenza causes serious public health problems in the world with substantial annual morbidity and mortality. Further, there have been persistent concerns about potential development of an influenza pandemic. Current antiviral drugs are limited in their efficacy, especially due to the rapid emergence of drug-resistant variants. Host protein-directed therapy is an alternative or complementary approach to broadly controlling influenza virus infections but requires a deeper understanding of influenza-host interplay. Host ceramide synthase 4 regulates the level of ceramides that possess both structural and signaling mediator functions. Our study reveals that ceramide synthase 4 displays an antiviral activity against influenza virus infection by regulating JNK activation. However, influenza virus triggers degradation of ceramide synthase 4, which could favor virus replication. The findings advance our knowledge about the ceramide network interaction with influenza and provide a framework for developing a host-targeted therapy to cure influenza. KEYWORDS influenza virus, ceramide synthase 4, JNK regulation, influenza-host interaction I nfluenza remains a major public health concern, causing significant morbidity and mortality worldwide through both seasonal and pandemic influenza (1-4). The 2009 influenza pandemic, along with recurrent avian influenza outbreaks, has heightened the awareness about the risk of future influenza pandemics (5-9). Due to frequent genetic mutations of the influenza virus leading to antigenic variation, influenza vaccines must
be updated yearly, and the vaccine's efficacy varies year by year. Drugs that inhibit activity of viral proteins, such as NA, PA, and M2, have been used to mitigate the effects of influenza, but these treatments are constrained by the rise of drug-resistant strains (10)(11)(12)(13)(14). These drug-resistant viral variants, which include various seasonal and avian influenza viruses, undermine the effectiveness of current antiviral therapies (11,13,15). Therefore, there is an urgent need to design new therapeutics and complementary approaches, such as host-directed therapy, by identifying cellular targets to effectively control influenza virus infections. This requires a better understanding of influenza-host interaction and influenza viral pathogenesis.
Ceramide synthase (CerS) enzymes, which are integral membrane proteins found in the endoplasmic reticulum, consist of six distinct members designated as CerS1 through CerS6 (16)(17)(18). Each CerS enzyme has specificity to fatty acyl CoA chains of varying lengths and has differential tissue expression patterns (17,(19)(20)(21). The function of CerS has been studied in multiple disease conditions such as cancer, diabetes, and obesity, likely due to ceramide's relevance to the regulation of cell death mechanisms or fatty acid metabolism (22)(23)(24). The inhibitor fumonisin B1, which blocks the activity of most of the CerS enzymes with limited specificity, was tested during influenza virus replication (25). However, the positive and negative results in separate studies yielded controversy (25)(26)(27), raising a question about the role of these enzymes upon influenza virus infection. CerS4, also known as LASS4 or TRH1, is the least studied CerS and has substrate specificity toward C18-C20 CoAs (19,28,29). It is primarily expressed in the lung and liver but also detected in multiple other tissues including the skin, nervous system, and pancreas (30). Unlike CerS1, CerS4 does not affect cellular sensitivity to chemotherapeutic drugs (31), but it is elevated in the brain of an Alzheimer's disease mouse model (32). The function of CerS4 in viral infection has not been explored.
In this study, we investigated the possible role of CerSs, such as CerS4, during influenza virus replication. We found that CerS4 displays antiviral activity as it interferes with the replication and propagation of influenza viruses. The underlying mechanism involves the inhibition of JNK activation by CerS4. However, influenza virus induces a decrease of CerS4 at the protein level upon infection, suggesting the presence of a complex interplay between host CerS4 and influenza.
## RESULTS
## Stable overexpression of CerS4, but not CerS1, inhibits the replication of influenza virus
Previously, CerS1 and CerS4 were studied using cells stably overexpressing each enzyme with the treatment of anticancer agents to determine their functions during the stress response (31). In this study, these cells were used to determine the possible roles of CerS1 and CerS4 during influenza virus infection. Since CerS1, but not CerS4, regulated the cell death induced by the anticancer agent cisplatin, we initially tested CerS1 by comparing control and CerS1-overexpressing HEK cells in terms of cellular susceptibility to IAV infection. However, stable overexpression of CerS1 did not affect the expressions of viral proteins such as NS1 and M2 of 2009 pandemic influenza A/CA/04/09 (H1N1) (pH1N1) in Western blot analysis (Fig. 1A through D). Interestingly, stable overexpression of CerS4 inhibited the expressions of viral proteins upon IAV infection (Fig. 1E andF). The inhibition of viral protein expression by CerS4 was observed upon influenza A/WSN/33 (H1N1) (IAV WSN) (Fig. 1E) or pH1N1 at multiple time points after infection (Fig. 1F). As IAV WSN replicates more rapidly than pH1N1 in the culture condition, a lower multiplicity of infection (MOI) (MOI = 0.1) was used for infection with WSN than pH1N1 (MOI = 1.0). The results indicate that stable expression of CerS4 impairs IAV replication.
## Transient overexpression of CerS4 impairs replication and propagation of influenza viruses
We next determined if transient overexpression of CerS4 regulates influenza virus replication similar to the stable overexpression of CerS4. To this end, cells were transfec ted with CerS4-encoding plasmid or vector control DNA and then infected with pH1N1 at an MOI of 1.0 for 24 hours (h). Transiently overexpressed CerS4 interfered with the replication of IAV, as evidenced by the reduced expression levels of many viral proteins detected by immunoblotting (Fig. 2A). The inhibition of viral protein expression by CerS4 was also observed when cells were infected with pH1N1 at an MOI of 3.0 for 8 h, representing a single-cycle infection condition (Fig. 2B). Importantly, CerS4 overexpres sion substantially decreased the production of infectious influenza virus particles from IAV-infected human lung epithelial A549 cells compared to the control in multiple experimental conditions, as measured by plaque assay (Fig. 2C through E): CerS4, but not CerS1, interfered with the production of infectious viruses at 1, 2, and 3 dpi when cells were infected with pH1N1 at an MOI of 0.01 (Fig. 2C). Further, CerS4 repressed the generation of infectious virus progeny when cells were infected with pH1N1 at an MOI of 3.0 for 8 h (Fig. 2D) or infected with IAV WSN strain at an MOI of 0.001 (Fig. 2E). To determine whether the effects of CerS4 on influenza are limited to IAV, influenza HA-tagged CerS4 (CerS4) were either left uninfected (Mock) or infected with IAV (pH1N1) at an MOI of 1. At 0, 6, and 9 hpi, the cells were harvested, and Western blotting was performed to detect the levels of viral NS1, CerS4, and GAPDH. In all experiments, GAPDH levels were measured to serve as an internal loading control. n.s., not significant. In (D), A549 cells were infected with IAV pH1N1 at an MOI of 3 for 8 hours, followed by supernatant collection for titration by plaque assay (n = 3/group). In (E), A549 cells were infected with IAV WSN at an MOI of 0.001. Cell supernatants were collected at 2 and 3 dpi for assessing viral titers using plaque assay B/Lee/40 virus (IBV) and influenza A/Hong Kong/8/68 (H3N2) virus (IAV H3N2) were also examined (Fig. 2F). The overexpression of CerS4 led to the inhibition of viral protein expression when cells were infected by IBV or IAV H3N2 (Fig. 2F), suggesting that CerS4 regulates the replication of multiple influenza virus types and subtypes. Taken together, these data indicate that overexpressed CerS4 displays antiviral function and impairs the productive infection of influenza viruses.
## Knockdown of endogenous CerS4 enhances the replication of influenza virus
To investigate the role of endogenous CerS4 during influenza virus infection, we implemented a knockdown method utilizing small interfering RNA (siRNA) designed to specifically target CerS4 (si-CerS4). Following the downregulation of CerS4 with si-CerS4, an increase in viral protein expression levels was observed at both 12 and 24 hours post-infection (hpi) with IAV pH1N1 compared to scrambled siRNA control (SCR) (Fig. 3A andB). Similarly, viral protein expression was elevated in cells infected with the IAV WSN strain when CerS4 expression was knocked down (Fig. 3C). Furthermore, downregulation of CerS4 using siRNA resulted in increased production of infectious IAV particles, which was assessed by plaque assay (Fig. 3D).
To confirm the antiviral role of endogenous CerS4 during influenza virus infection, recombinant lentiviruses were constructed to deliver shRNA targeting CerS4 (sh-CerS4) and inhibit the expression of endogenous CerS4. A549 cells were transduced with lentivirus encoding sh-CerS4, a mixture of multiple sh-CerS4 (sh-CerS4 Mix) or control lentivirus, and then infected with IAV to assess the impact of CerS4 downregulation on the level of viral replication. sh-CerS4-mediated CerS4 downregulation (Fig. 3E andF) led to an increase in the expressions of viral proteins HA and PB1 (Fig. 3G). The results further demonstrate that endogenous CerS4 acts as an antiviral host factor during influenza virus infection.
## CerS4 inhibits JNK activation, which is associated with CerS4's antiviral activity during infection
The main stages of the influenza virus life cycle include viral entry, viral replication, and the release of virus particles from the host cell (33,34). To further understand the mechanism of CerS4 inhibition of IAV replication, we conducted reverse transcription (RT) with viral (-) strand RNA-complementary primers followed by real-time qPCR using IAV-infected CerS4-overexpressing cells and control cells. The synthesis of viral (-) strand RNAs specific for NP (Fig. 4A) and NS1 (Fig. 4B) was inhibited at 4 hpi in CerS4-overex pressing cells compared to control cells. However, no changes in the levels of viral (-) strand RNA were observed between control and CerS4-overexpressing cells at 1 hpi or 2 hpi (Fig. 4A andB). A similar pattern of viral RNA regulation was detected when CerS4 was transiently overexpressed in A549 cells followed by pH1N1 infection (Fig. 4C andD). The results imply that the inhibition of viral replication by CerS4 did not occur at very early steps of viral infection such as viral entry.
While it is unknown whether CerS4 regulates specific cellular signaling pathways, ceramide has been reported to activate multiple cellular signaling proteins depending on the disease condition (35)(36)(37). To investigate the underlying mechanism by which CerS4 regulates IAV replication, the activations of several signaling components impor tant for IAV replication or host defense were monitored in control cells and CerS4overexpressing cells at 0, 3, 6, and 9 hours after IAV infection (Fig. 5A). When viral proteins were detected at 6 and 9 hpi, correlative decreases in phosphorylated forms of JNK (p-JNK) were observed in CerS4-overexpressing cells compared to those in control cells. The experiment was repeated with similar results.
However, the activation status of MEK, NF-κB, STAT1, and p38 MAPK (pMEK, pNF-κB, pSTAT1, and p-p38) remained largely unchanged by CerS4 overexpression during IAV infection. Since JNK inhibition was reported to repress IAV replication (38,39), the finding led us to hypothesize that CerS4 regulates JNK activation and impairs viral replication. Next, we determined if transient overexpression of CerS4 affects the phosphorylation of JNK during IAV replication. Infection by IAV WSN increased the levels of p-JNK. Transient overexpression of CerS4 did not affect the levels of total JNK proteins but impaired the increase in p-JNK during WSN infection over time, which was associated with the inhibition of viral protein expression (Fig. 5B). The inhibition of p-JNK levels by transiently overexpressed CerS4 during infection was also observed when A549 cells were infected by pH1N1 at an MOI of 0.1 and 1.0 (Fig. 5C through E), where virus-induced increase of p-JNK was more evidently observed at an MOI of 1.0, presumably due to the higher MOI used for infection. Since immunoblotting can detect a one-time measure ment, we used two different experimental conditions (MOI of 0.1 and 1.0) to find that CerS4 mediates the inhibition of p-JNK in multiple experimental conditions. Further, when the endogenous CerS4 was downregulated using sh-CerS4 encoding lentivirus, the levels of IAV-induced p-JNK, as well as viral protein expression, increased (Fig. 5F through H). To further support the result, JNK activation was blocked by the treatment with SP600125, a specific JNK inhibitor. In the presence of the JNK inhibitor, the effects of CerS4 downregulation on both p-JNK and viral protein expressions were abrogated (Fig. 5F through H), suggesting that CerS4's regulation of JNK is important for CerS4-mediated suppression of IAV replication.
Although our experiments were performed with kinetic monitoring of signaling components and incorporation of JNK inhibitor, it may be difficult to fully exclude the possibility that CerS4 inhibits IAV replication via an as-yet undefined mechanism, which leads to suppression of IAV-induced JNK activation. Indeed, CerS4 or ceramide-mediated inhibition of p-JNK has not been reported. Therefore, we determined whether CerS4 regulates p-JNK under another condition in the absence of viruses. For this purpose, A549 cells were stimulated with thapsigargin (TG), which is known to activate JNK. CerS4 overexpression was demonstrated to inhibit TG-induced JNK activation (Fig. 6A through C). The results indicate that CerS4 can inhibit JNK activation in multiple conditions.
## Influenza virus infection induces degradation of CerS4
While studying the role of CerS4 during IAV replication, we observed that IAV infection induced downregulation of endogenous CerS4 in multiple cell types including A549 (Fig. 7A) and HEK293 cells (Fig. 7B). However, the levels of mRNA encoding CerS4 did not significantly decrease upon IAV infection (Fig. 7C), suggesting that CerS4 regulation takes place at the post-transcriptional level during infection. These observations led us to hypothesize that IAV infection induces degradation of CerS4 protein, which could be beneficial for IAV replication. To determine if IAV infection triggers ubiquitination of CerS4, endogenous CerS4 was immunoprecipitated using anti-CerS4 antibody followed by the detection of ubiquitinated forms of CerS4 by Western blotting. Indeed, IAV infection increased the level of ubiquitinated CerS4 protein in A549 cells (Fig. 7D). To further evaluate CerS4 degradation during infection, we used inhibitors that block proteasomal (MG132) or lysosomal (NH 4 Cl and Bafilomycin A1) protein degradation pathways. Cells were infected with influenza virus and only treated with inhibitor for 6 hours prior to harvest as earlier treatment could affect IAV replication, which would subsequently impair virus-induced CerS4 degradation. A549 cells treated with any of the inhibitors (MG132, NH 4 Cl, or bafilomycin A1) inhibited IAV-induced CerS4 downregula tion (Fig. 7E). The results corroborate the conclusion that IAV induces degradation of CerS4.
## DISCUSSION
Using multiple genetic modification methods of stable or transient overexpression of CerS4 as well as CerS4 knockdown by siRNA or shRNA, this study revealed that CerS4 acts as an antiviral protein during influenza virus infection. However, influenza virus may strive to counteract the antiviral activity of CerS4 as infection triggers degradation of CerS4, which in turn could promote a cellular environment advantageous for the replication of influenza viruses (Fig. 8).
In prior cancer research, CerS1, but not CerS4, increased cellular sensitivity to cisplatin, a chemotherapeutic drug, with enhanced cell death. Activation of p38 MAPK, which is a well-known stress kinase, is increased by CerS1, but not by CerS4 in response to cisplatin treatment, explaining the enhanced cell death by CerS1 (31). However, upon influenza virus infection, p38 MAPK was not activated by CerS4. The difference in activation of signaling components by various stress stimuli or pathogen may explain the involvement of specific CerS under distinct disease conditions. Ceramide has been reported to increase cell stress responses and activate multiple signaling proteins, including ASK1, p38 MAPK, and JNK (40)(41)(42)(43). Therefore, the inhibition rather than activation of JNK by CerS4 was surprising. The inhibition of JNK activation by CerS4 was also detected when cells were stimulated by TG. Thus, the results suggest that JNK inhibition may represent a CerS4-specific regulatory mechanism. It is currently unknown if CerS4 or the metabolic product, such as C20 ceramide, directly regulates JNK activa tion. As JNK can be activated by multiple signaling pathways that include upstream kinases (44), it is also possible that CerS4 targets a specific signaling molecule upstream of JNK. The regulatory mechanism could be common to the pathway activated by TG and influenza virus infection (45). The detailed mechanism remains to be explored.
The role of ceramide during virus infections has been previously investigated, mainly using cell-permeable short-chain ceramide analogs, such as C-2, C-,6 or C-8 ceramide, and inhibitors that block the synthesis of ceramide (i.e., fumonisin B1) or ceramide precursor 3-ketosphinganine (i.e., myriocin). However, during influenza virus infection, the inhibition of de novo ceramide synthesis by inhibitor treatment was shown to affect virus replication both positively and negatively during different in vitro experiments (25)(26)(27). These results suggest that the ceramide pathway is important for virus replication or host defense in diverse ways. Opposing results could potentially be explained by the complexity of the sphingolipid metabolic pathway and the use of an in vitro experimen tal condition as the complete inhibition of ceramide synthesis may lead to the inhibition of other metabolites, including sphingomyelin, over time. Further, the treatment of cells with C-6 ceramide analog, which could increase many ceramide species in cells and change the lipid components of membranes, suppressed influenza virus replication and suggests an antiviral role of ceramides. On the other hand, local administration of C-8 ceramide in mice did not affect influenza virus propagation in vivo but increased dendritic cell stimulation and virus-specific T cell responses during infection (46). While the de novo sphingolipid synthesis pathway could be critical for regulating influenza virus infection, completely blocking the early stage of sphingolipid synthesis may result in diverse outcomes based on prior inhibitor studies. Currently, there are no studies investigating the role of CerS enzymes during influenza infection using genetic modification approaches. As the in vitro overexpression of CerS4, but not CerS1, inhibited influenza virus replication (Fig. 1), it is possible that specific ceramide subspecies synthesized in cells may be critical for the host defense against virus infection. Therefore, interrogating the role of individual CerSs during influenza virus infection could increase the comprehensive understanding of the ceramide network in explicitly regulating influenza virus replication. CerS4 was shown to display antiviral activity against the replication of multiple strains (WSN and pH1N1) and subtypes (H1N1 and H3N2) of influenza A virus as well as influenza B virus. CerS4 activity is unaffected by the well-known CerS inhibitor fumonisin B1 (28). The function of CerS4 during infection with other viruses has not yet been reported to the best of our knowledge. Thus, it would be interesting to determine if CerS4 has similar or different roles during other virus infections for a deeper understand ing of ceramide/CerS4-virus interactions.
Influenza virus has been reported to induce degradation of many host proteins (47-52). The phenomena have been linked to influenza viral evasion or nullification of the host defense to favor virus propagation. However, the systemic and detailed pathways for influenza viral mechanisms to induce degradation of host antiviral proteins remain poorly understood (53). In this study, the antiviral host factor CerS4 was shown to be degraded during influenza virus infection, where influenza virus most likely utilizes additional host factors to cause CerS4 ubiquitination and subsequent degradation. Thus, probing the mode of action by which influenza virus induces degradation of CerS4 may help identify host pro-influenza proteins and facilitate the design of host-directed antiviral therapeutics against influenza.
## MATERIALS AND METHODS
## Viruses and cells
The influenza viruses A/WSN/33 (H1N1) and pandemic A/CA/04/09 (H1N1) were utilized as described in prior studies (54,55). Both the A/Hong Kong/8/68 (H3N2) virus (ATCC VR-1679) and the B/Lee/40 (IBV) virus (ATCC VR-1535) were purchased through ATCC. The viruses were propagated in Madin-Darby canine kidney (MDCK) cells following methods previously reported (51,54,55). Virus titration was conducted using a plaque assay. CerS1-or CerS4-overexpressing HEK293 cells were a kind provision from Dr. Stephen Alexander (University of Missouri-Columbia). A549, HEK293, HEK293T, and 293 FT cells were maintained in Dulbecco's modified Eagle's medium (DMEM; Gibco), while MDCK cells were grown in DMEM with MEM NEAA (non-essential amino acids; Invitrogen), as described earlier (51,54,56,57). For the influenza A/WSN/33 (H1N1) virus, cells were incubated with 10% fetal bovine serum (FBS)-containing medium; for the influenza A/CA/04/09 (H1N1) virus, cells were incubated with FBS-free medium containing 0.3% BSA and TPCK-treated trypsin (1 µg/mL) (55). All cells were incubated at 37°C in a CO 2 incubator, with media supplemented with 10% FBS (Sigma) and penicillin-streptomycin (100 U/mL and 100 µg/mL, respectively; Invitrogen).
## Reagents and antibodies
Bafilomycin A1 was sourced from Cayman Chemical; MG132 and NH4Cl were purchased from Fisher Scientific. Thapsigargin was purchased from Tocris Bioscience (Bristol, UK). Anti-CerS4 antibodies were purchased from Origene (Rockville, MD). Antibodies against IAV HA were purchased from Santa Cruz Biotechnology (Dallas, TX). Antibodies specific to vial NS1 and PB1 were purchased from GeneTex (Irvine, CA). Antibodies targeting IAV M1, IAV M2, and IBV NP were purchased from Abcam (United Kingdom, Cambridge). Antibodies for p-JNK, total JNK, p-MEK, p-NFκB, p-p38, p-STAT1, HA tag, Myc tag, and GAPDH were purchased from Cell Signaling Technology (Danvers, MA).
## Plasmid constructs and transfection
The plasmid DNAs encoding human CerS1 and human CerS4, tagged with Myc and Flag, were purchased from Origene, and the pCMV-Flag plasmid was provided by Dr. David Pintel (University of Missouri-Columbia). Cultured cells were plated in 24-well plates at densities of 1 × 10⁵ cells/well at 18 hours prior to transfection. Cells were transiently transfected with plasmids encoding Myc and Flag-tagged CerS4 at a concentration of 500 ng/mL using TurboFect transfection reagent (Thermo Scientific), following the manufacturer's recommended protocols. In all transfection experiments, pCMV-Flag (Empty vector) was used as controls to ensure that the total amount of DNA was consistent across all samples.
## Western blot analysis
Western blotting was conducted as described in previous studies (51,54,56,57). In summary, denatured polypeptides from cell lysates or immunoprecipitation (IP) samples were resolved by SDS-PAGE and transferred onto nitrocellulose membranes (Bio-Rad). Membrane-bound primary antibodies were detected using HRP-conjugated secondary antibodies (Cell Signaling Technology). The signals were visualized using an Odyssey Fc imaging system (Li-Cor) and processed with Image Studio software (Li-Cor). At least two independent experiments were performed, which produced comparable results.
## Denatured immunoprecipitation assay
Detection of CerS4 ubiquitination during IAV infection was performed using A549 cells (1 × 10⁶) that were mock-infected or infected with IAV at an MOI of 2 for 24 hours. The cells were lysed with 200 µL of 1% SDS-containing lysis buffer and denatured by boiling at 95°C for 5 minutes. This process inhibits cellular ubiquitin hydrolases, preserving Ub-CerS4 conjugates, but disrupts protein-protein interactions. For IP, lysates were diluted fivefold (final volume: 1 mL) in IP lysis buffer supplemented with 1 mM PMSF and incubated by vortexing overnight at 4°C with Protein-G agarose beads (30 µL) preloaded with anti-CerS4 antibodies (Origene). After four rigorous washes to remove nonspecific interactions, the precipitated proteins were analyzed via Western blotting. The experiments were performed twice independently, with similar results. To analyze bound proteins, 20 µL of 2 × sample buffer was added to the beads for elution, followed by immunoblotting with anti-ubiquitin antibodies (Cell Signaling Technology).
## Knockdown by siRNA
Silencer Select siRNA for LASS4 (CerS4) and Silencer Select Negative Control #1 (control scrambled RNA; SCR) were purchased from Thermo Scientific. HEK293T or A549 cells were transfected with 20 nM siRNA using Lipofectamine RNAiMax transfection reagent (Thermo Scientific) according to the manufacturer's instructions.
## Knockdown by lentivirus-derived shRNA
Lentiviral vectors pGFP-CerS4-shLenti #1 and #2 were purchased from Origene (Rock ville, MD). Oligonucleotides encoding different shRNAs were cloned between the U6 promoter and the SV40 promoter. For lentivirus production, 293 FT cells were transfected with pGFP-CerS4-shLenti #1 or #2 along with packaging plasmids (PAX and MD, provided by Dr. Julie Saba, UCSF). Culture supernatants were collected 72 hours post-transfection, centrifuged at 1,500 rpm for 10 minutes at 4°C, and filtered through a 0.45 µm filter. For lentiviral transduction of A549 cells, 1 × 10 4 cells per well were seeded in 24-well plates. After overnight incubation, 0.5 mL of culture supernatants containing sh-CerS4-#1 or a mixture of sh-CerS4 lentiviruses (pGFP-CerS4-shLenti #1 and #2) was added to the wells with 8 µg/mL polybrene (Sigma-Aldrich). The cells were incubated overnight at 37°C and then utilized for subsequent experiments or analyses. After 48 hours of transduction, A549 cells were infected with either mock or IAV pH1N1 at an MOI of 0.1 for 24 hours.
## Real-time PCR
Total cellular RNA was extracted using the DNA-free RNA Isolation Kit (Invitrogen) according to the manufacturer's protocol. The purified RNA was reverse-transcribed into cDNA using random primers (Invitrogen) and reverse transcriptase (Promega) for detection of CerS4 and GAPDH, while primers complementary to viral (-) strand RNA were used for detection of viral (-) strand RNAs of NP or NS1. Using the cDNAs, a separate real-time quantitative PCR (RT-qPCR) was performed with gene-specific primers (56). Targets included human CerS4 (5′-CCC GAC TGG TCC TCT TTC CC -3′ and 5′-GCA GCA ACA TCA GAA GCC CG -3′), viral NP (5′-TAT GTG GCA TCA TTC AGG TTG GA-3′ and 5′-GAC GGA AAG TGG ATG AGA GAA CT -3′), viral NS1 (5′-CTC TGT CGC TTT CAA TCT GTG C-3′ and 5′-TCG CTT GGA GAA ACT GTG ATG A-3′), and human GAPDH (5′-AGC CTC AAG ATC ATC AGC AAT GCC-3′ and 5′-TGT GGT CAT GAG TCC TTC CAC GAT-3′). The qPCRs utilized Power SYBR Green PCR Master Mix (Applied Biosystems) on a StepOne real-time PCR system (Applied Biosystems), with cDNA levels normalized to GAPDH RNA in the same samples.
## Plaque assays
A549 cells were infected with IAV at the specified MOI. At the indicated time points, the supernatants were collected, and virus titers were determined using MDCK cells. For the titration of pH1N1, MDCK cell monolayers were washed with PBS supplemented with 0.3% bovine serum albumin (BSA) and subsequently infected with tenfold serial dilutions of the virus. The infected MDCK cells were overlaid with a mixture containing 0.6% SeaKem LE agarose (Lonza), 2 × L15 (Invitrogen), 0.3% BSA, and 1 µg/mL TPCK-treated trypsin, followed by incubation for 2-3 days. For the titration of WSN, instead of BSA, FBS was used in the absence of trypsin. All infections were performed at 37°C in a 5% CO₂ humidified atmosphere.
## Statistical analysis
Error bars in all figures indicate the mean values ± standard deviations (SD). Significance was determined by running Student's t-test (P < 0.05). Densitometry for Western blots was calculated using ImageJ software. Data shown are representative of at least two independent experimental replicates.
## References
1. Naquin, 'halloran, Ujamaa et al. (2010) "Laboratory-confirmed influenza-associated hospitalizations among children and adults -influenza hospitalization surveillance network" *MMWR Surveill Summ*
2. Mello, Brammer, Blanton et al. (2014) "Update: influenza activity--United States"
3. Simonsen, Clarke, Schonberger et al. (1998) "Pandemic versus epidemic influenza mortality: a pattern of changing age distribution" *J Infect Dis*
4. Meltzer, Cox, Fukuda (1999) "The economic impact of pandemic influenza in the United States: priorities for intervention" *Emerg Infect Dis*
5. Caserta, Frye, Butt et al. (2025) "Edwards Full-Length Text Journal of Virology December"
6. Ee, Gomez, Hitchener et al. (2024) "Spillover of highly pathogenic avian influenza H5N1 virus to dairy cattle" *Nature*
7. Claas, Osterhaus, Van Beek et al. (1998) "Human influenza A H5N1 virus related to a highly pathogenic avian influenza virus" *Lancet*
8. Cowling, Jin, Lau et al. (2013) "Comparative epidemiology of human infections with avian influenza A H7N9 and H5N1 viruses in China: a populationbased study of laboratory-confirmed cases" *Lancet*
9. Fraser, Donnelly, Cauchemez et al. (2009) "Pandemic potential of a strain of influenza A (H1N1): early findings" *Science*
10. Wang, Palese (2009) "Unraveling the mystery of swine influenza virus" *Cell*
11. Cheng, To, Leung et al. (2010) "Oseltamivir-and amantadine-resistant influenza virus A (H1N1)" *Emerg Infect Dis*
12. Dharan, Gubareva, Meyer et al. (2009) "Infections with oseltamivir-resistant influenza A(H1N1) virus in the United States" *JAMA*
13. Marjuki, Mishin, Chesnokov et al. (2015) "Characterization of drugresistant influenza A(H7N9) variants isolated from an oseltamivir-treated patient in Taiwan" *J Infect Dis*
14. Poland, Jacobson, Ovsyannikova (2009) "Influenza virus resistance to antiviral agents: a plea for rational use" *Clin Infect Dis*
15. Jones, Kumar, Barman et al. (2018) "Identification of the I38T PA substitution as a resistance marker for next-generation influenza virus endonuclease inhibitors" *mBio*
16. Kumari, Sharma, Kumar et al. (2023) "Antiviral approaches against influenza virus" *Clin Microbiol Rev*
18. Kageyama-Yahara, Riezman (2006) "Transmembrane topology of ceramide synthase in yeast" *Biochem J*
19. Park, Park, Futerman (2014) "Ceramide synthases as potential targets for therapeutic intervention in human diseases" *Biochim Biophys Acta*
20. Levy, Futerman (2010) "Mammalian ceramide synthases" *IUBMB Life*
21. Mizutani, Kihara, Igarashi (2005) "Mammalian Lass6 and its related family members regulate synthesis of specific ceramides" *Biochem J*
22. Mizutani, Kihara, Igarashi (2006) "LASS3 (longevity assurance homologue 3) is a mainly testis-specific (dihydro)ceramide synthase with relatively broad substrate specificity" *Biochem J*
23. Raichur (2020) "Ceramide synthases are attractive drug targets for treating metabolic diseases" *Front Endocrinol (Lausanne)*
24. Li, Chaurasia, Rahman et al. (2023) "Ceramides increase fatty acid utilization in intestinal progenitors to enhance stemness and increase tumor risk" *Gastroenterology*
25. Hammerschmidt, Brüning (2022) "Contribution of specific ceramides to obesity-associated metabolic diseases" *Cell Mol Life Sci*
26. Raichur, Brunner, Bielohuby et al. (2018) "The role of C16:0 ceramide in the development of obesity and type 2 diabetes: CerS6 inhibition as a novel therapeutic approach" *Mol Metab*
27. Soudani, Hage-Sleiman, Karam et al. (2019) "Ceramide suppresses influenza A virus replication in vitro" *J Virol*
28. Tafesse, Sanyal, Ashour et al. (2013) "Intact sphingomyelin biosynthetic pathway is essential for intracellular transport of influenza virus glycoproteins" *Proc Natl Acad Sci*
29. Hidari, Suzuki, Suzuki (2006) "Suppression of the biosynthesis of cellular sphingolipids results in the inhibition of the maturation of influenza virus particles in MDCK cells" *Biol Pharm Bull*
30. Riebeling, Allegood, Wang et al. (2003) "Two mammalian longevity assurance gene (LAG1) family members, trh1 and trh4, regulate dihydroceramide synthesis using different fatty acyl-CoA donors" *J Biol Chem*
31. Tidhar, Zelnik, Volpert et al. (2018) "Eleven residues determine the acyl chain specificity of ceramide synthases" *J Biol Chem*
32. Laviad, Albee, Pankova-Kholmyansky et al. (2008) "Characterization of ceramide synthase 2: tissue distribution, substrate specificity, and inhibition by sphingosine 1phosphate" *J Biol Chem*
33. Min, Mesika, Sivaguru et al. (2007) "(Dihydro)ceramide synthase 1-regulated sensitivity to cisplatin is associated with the activation of p38 mitogenactivated protein kinase and is abrogated by sphingosine kinase 1" *Mol Cancer Res*
34. Jęśko, Wencel, Wójtowicz et al. (2020) "Fingolimod affects transcription of genes encoding enzymes of ceramide metabolism in animal model of Alzheimer's disease" *Mol Neurobiol*
35. Dou, Revol, Östbye et al. (2018) "Influenza A virus cell entry, replication.virion assembly and movement" *Front Immunol*
36. Carter, Iqbal (2024) "The influenza A virus replication cycle: a comprehensive review" *Viruses*
37. Stith, Velazquez, Obeid (2019) "Advances in determining signaling mechanisms of ceramide and role in disease" *J Lipid Res*
38. Stancevic, Kolesnick (2010) "Ceramide-rich platforms in transmem brane signaling" *FEBS Lett*
39. Kurz, Barthelmes, Blum et al. (2019) "Role of ceramide synthase 2 in G-CSF signaling and G-CSF-R translocation into detergent-resistant membranes" *Sci Rep*
40. Zhang, Ruan, Sheng et al. (2019) "Role of c-Jun terminal kinase (JNK) activation in influenza A virus-induced autophagy and replication" *Virology (Auckl)*
41. Nacken, Ehrhardt, Ludwig (2012) "Small molecule inhibitors of the c-Jun N-terminal kinase (JNK) possess antiviral activity against highly pathogenic avian and human pandemic influenza A viruses" *Biol Chem*
42. Chen, Lin, Chang et al. (2008) "Ceramide induces p38 MAPK and JNK activation through a mechanism involving a thioredoxin-interacting protein-mediated pathway" *Blood*
43. Niaudet, Bonnaud, Guillonneau et al. (2017) "Plasma membrane reorganization links acid sphingomyelinase/ceramide to p38 MAPK pathways in endothelial cells apoptosis" *Cell Signal*
44. Yue, López (2020) "Understanding MAPK signaling pathways in apoptosis" *Int J Mol Sci*
45. (2025) *Full-Length Text Journal of Virology*
46. Basu, Kolesnick (1998) "Stress signals for apoptosis: ceramide and c-Jun kinase" *Oncogene*
47. Ip, Davis (1998) "Signal transduction by the c-Jun N-terminal kinase (JNK) -from inflammation to development" *Curr Opin Cell Biol*
48. Nacken, Anhlan, Hrincius et al. (2014) "Activation of c-jun N-terminal kinase upon influenza A virus (IAV) infection is independent of pathogen-related receptors but dependent on amino acid sequence variations of IAV NS1" *J Virol*
49. Pritzl, Seo, Xia et al. (2015) "A ceramide analogue stimulates dendritic cells to promote T cell responses upon virus infections" *J Immunol*
50. Rodriguez, Pérez-González, Nieto (2007) "Influenza virus infection causes specific degradation of the largest subunit of cellular RNA polymerase II" *J Virol*
51. Wang, Wei, Chen et al. (2014) "Influenza A virus-induced degradation of eukaryotic translation initiation factor 4B contributes to viral replication by suppressing IFITM3 protein expres sion" *J Virol*
52. Jung, Ko, Shin et al. (2019) "Endoplasmic reticulumassociated degradation potentiates the infectivity of influenza A virus by regulating the host redox state" *Free Radic Biol Med*
53. Jung, Pyo, Choi (2018) "Influenza A virus-induced autophagy contributes to enhancement of virus infectivity by SOD1 downregula tion in alveolar epithelial cells" *Biochem Biophys Res Commun*
54. Xia, Vijayan, Pritzl et al. (2015) "Hemagglutinin of influenza A virus antagonizes Type I Interferon (IFN) responses by inducing degradation of Type I IFN receptor 1" *J Virol*
55. Wolf, Xia, Studstill et al. (2021) "Influenza A virus NS1 induces degradation of sphingosine 1-phosphate lyase to obstruct the host innate immune response" *Virology (Auckl)*
56. Xia, Wang, Hahm (2024) "Triggering degradation of host cellular proteins for robust propagation of influenza viruses" *Int J Mol Sci*
57. Xia, Seo, Studstill et al. (2018) "Transient inhibition of sphingosine kinases confers protection to influenza A virus infected mice" *Antiviral Res*
58. Xia, Wolf, Vijayan et al. (2018) "Casein kinase 1α mediates the degradation of receptors for Type I and Type II interferons caused by hemagglutinin of influenza A virus" *J Virol*
59. Seo, Blake, Alexander et al. (2010) "Sphingosine 1-phosphatemetabolizing enzymes control influenza virus propagation and viral cytopathogenicity" *J Virol*
60. Vijayan, Seo, Pritzl et al. (2014) "Sphingosine kinase 1 regulates measles virus replication" *Virology (Auckl)*
61. (2025) *Full-Length Text Journal of Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12645954&blobtype=pdf | # Tissue tropism and functional adaptation of the SARS-CoV-2 spike protein in a fatal case of COVID-19
Katherine Johnson, Sydney Stein, Rita Boateng, Shilpi Jain, Sabrina Ramelli, Trevor Stantliff, Shelly Curran, Marcos Ramos-Benítez, Andrew Platt, Stephanie Banakis, Wei Wang, Stephen Hewitt, Christa Zerbe, Steven Holland, Elizabeth Kang, Manmeet Singh, Emmie De Wit, William Lauer, Eric Rouchka, Melissa Smith, Mehul Suthar, Daniel Chertow, Elodie Ghedin
## Abstract
Systemic spread of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) to extrapulmonary tissues has been observed following acute infections. Autopsy studies further indicate tissue-specific virus diversity, including in immune-priv ileged sites. Questions remain on the viral dynamics leading to the tissue tropism of SARS-CoV-2, including evolutionary trajectories and functional adaptations that could impact persistence and transmission. In this study, we characterized SARS-CoV-2 genomes from 27 distinct tissues collected from an autopsy case where the patient had a primary immune deficiency. We identified tissue-specific virus genotypes, in some instances coexisting within the same sites, with mutations primarily in the receptorbinding domain of the spike protein. Protein simulations and isolation of infectious virus indicate combinations of spike substitutions that would lead to increased protein stability and stronger binding of the virus to host cells. This highlights the importance of studying patients with weakened immune responses where potential tissue reservoirs provide an environment permissive for SARS-CoV-2 evolution and diversification. IMPORTANCE Persistent severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infections in immunocompromised individuals are considered a potential source of new viral variants. Beyond the respiratory tract, the virus can spread within days to organs like the brain, heart, and kidneys, where distinct tissue microenvironments may further drive viral evolution and the emergence of new mutations. In this study, we compared the genetic diversity of SARS-CoV-2 genomic RNA isolated from 27 distinct tissue sites collected from an individual with a weakened immune system. By linking viral population dynamics across these tissue sites, we defined the extent of compartmentali zation during multi-organ spread, highlighting how non-respiratory tissues can impact SARS-CoV-2 diversification. KEYWORDS viral evolution, viral intrahost diversity, tissue tropism, coronavirus S evere acute respiratory syndrome coronavirus 2 (SARS-CoV-2) primarily infects the respiratory epithelium. However, SARS-CoV-2 is not confined to the respiratory tract and has been detected in various tissues, including ocular, cardiac, and neuronal compartments (1-4). These tissues express different levels of SARS-CoV-2′s primary receptor, angiotensin-converting enzyme 2 (ACE2), and its activating protease, trans membrane serine protease 2 (TMPRSS2), which are important for viral entry into cells (5,6). The availability of the ACE2 and TMPRSS2 on the surface of the cells, along with the spatial structure of tissues, whether the tissue is considered immune privileged, and the expression of intracellular cofactors can all shape the types of SARS-CoV-2 variants that enter, replicate, and circulate within a host.
The mechanisms of SARS-CoV-2 transmission to non-respiratory tissues remain unclear, although neuronal, hematogenous, and cellular trafficking transmission routes have all been proposed (7)(8)(9)(10). SARS-CoV-2 can spread to extrapulmonary tissues within days of symptom onset in both healthy individuals and those with comorbidities (1). The presence of SARS-CoV-2 in extrapulmonary tissues following acute infections has led to questions on whether systemic spread of the virus is linked to the extensive set of symptoms associated with post-acute sequelae of COVID-19, also known as "long-COVID, " where individuals report symptoms for weeks to months following infection (11). Typically, infectious SARS-CoV-2 particles are cleared within 9 days, but viral RNA can be detected in the gastrointestinal tract for weeks and sometimes months following infection (12,13). Furthermore, autopsy studies have found that SARS-CoV-2 can also persist in immune-privileged tissues such as the brain (1) and generate tissue-specific diversity, where unique genotypes are observed in the different tissue sites (1)(2)(3), leading to the possibility that non-respiratory tissues may act as reservoirs, allowing SARS-CoV-2 to persist at nearly undetectable levels within individuals.
Persistent SARS-CoV-2 infections may be more common than initially predicted, are associated with a higher risk of long-term symptoms, and have important implications for the evolution and spread of the virus (14,15). Understanding the tissue tropism of SARS-CoV-2 is crucial for characterizing its evolutionary trajectories, persistence, transmission, and disease severity. In this study, we obtained whole-genome sequences of SARS-CoV-2 isolated from 27 distinct tissues collected from an autopsy case. We characterized viral dynamics across the different tissue compartments, including variant richness, evolutionary changes, and shared virus diversity. Additionally, we assessed the functional effects of spike (S) mutations that accumulate during infection using protein simulations and binding and internalization assays with infectious isolates.
## RESULTS
## Multiple SARS-CoV-2 genotypes are identified across tissues in an autopsy case
On 29 September 2021, a 57-year-old male patient presented to a Florida hospital with fever, lethargy, shortness of breath, and cough. His prior medical history was significant for daily immunosuppression with sirolimus following a matched unrelated donor hematopoietic stem cell transplant (HSCT) in January 2021 to treat chronic granuloma tous disease (CGD), a primary immune deficiency, with associated recurrent respiratory infections, severe bronchiectasis, and respiratory insufficiency requiring home oxygen and noninvasive mechanical ventilation. Upon presentation to the emergency room, he tested positive for SARS-CoV-2 7 days following his first COVID-19 vaccine and was admitted for monoclonal antibody and remdesivir therapy. On 4 October, he was discharged following resolution of his acute symptoms. During 21-27 October, he was readmitted with recurrent respiratory symptoms, a new right upper lobe infiltrate, and a negative SARS-CoV-2 test. He was treated with intravenous antibiotics and prednisone 40 mg orally twice daily. Following hospital discharge, the patient moved to New York to be closer to family. On 23 November, he was admitted to a New York hospital with worsening shortness of breath and hypoxia. He again tested positive for SARS-CoV-2. The patient died on 28 November from respiratory failure and on 2 December underwent autopsy at the National Institutes of Health Clinical Center following consent from next of kin.
At autopsy, 73 distinct tissue sites were collected. Of these, 61 had SARS-CoV-2 RNA copies above the detection limit (see Table S1). We selected 32 sites for additional analyses where SARS-CoV-2 genome copies ranged from 1.38 to 5.21 log 10 nucleocapsid (N) gene copies per nanogram of RNA, with an average of 3.89 log 10 N copies per nanogram of RNA. SARS-CoV-2 subgenomic RNA (sgRNA) was detected in 28 of 32 samples, indicating active viral replication occurred in these tissues (Table S1). Nucleo capsid gene copies and sgRNA were positively correlated (r = 0.99, P < 2.2e -16, Pearson) (see Fig. S1A).
Droplet digital PCR (ddPCR) results for each tissue site indicated viral loads similar to those of an acute infection (≤14 days) (1). To determine whether the patient died with a persistent SARS-CoV-2 infection lasting since the initial infection in late September 2021, or an acute reinfection with SARS-CoV-2, we performed phylogenetic analyses using whole-genome sequencing data of SARS-CoV-2 RNA isolated from the 32 distinct tissue sites (Table S1). All 32 samples were sequenced twice using ARTIC V4 primers and the Illumina sequencing platform, and all, except for the tongue and pericardium samples, were sequenced once using long-read PacBio HiFi viral sequencing (16). Although the long-read (PacBio) and short-read (Illumina) data had similar mean and median read depths (Fig. S1B andC), the long-read data had better read-depth evenness across the genome (Fig. S1D), which was not significantly correlated with ddPCR results (r = -0.16, P-value = 0.4, Pearson) (Fig. S1E), indicating that the long-read data were better at capturing low-abundance viral genomes. Therefore, we used the long-read data for variant analyses and confirmed our findings with the short-read data when read coverage was sufficient.
Five samples, including the basilar artery, spleen, dura mater, cervical spinal cord, and thoracic aorta, only had 10× read depths across 10.3%, 35.5%, 43.2%, 58.8%, and 59.4% of the genome, respectively, and were therefore not used for any genomic analyses. The 27 of 32 remaining samples had ≥10× read depths across ≥80% of the genome and were used for consensus and phylogenetic analyses.
At the time of the patient's death, Omicron BA.1 lineages were beginning to displace Delta lineages. Using NextClade, we found that all 27 samples were positive for an AY.119 (Delta, 21J) lineage infection. To determine the location (and general timing) of the infection, we generated a maximum-likelihood phylogenetic tree using IQ-TREE with the 27 SARS-CoV-2 consensus sequences collected from 27 different tissue sites and 6,000 randomly selected, complete AY.119 consensus sequences circulating from 1 August 2021 to 31 January 2022 in the USA (Fig. 1A andB). Root-to-tip distances of the AY.119 lineages increased over time, with an estimated substitution rate of 3.76e -04 substitutions/site/year (Fig. 1C).
The closest sequences were collected from Pennsylvania, Texas, Virginia, and Ohio. Therefore, we were unable to determine the location of infection from the phylogenetic analyses, which limited our ability to infer the length of infection. Samples from the patient were monophyletic but differed in the number of mutations from the shared ancestor, with the estimated number of nucleotide substitutions ranging from 1.19 to 6.83 from the root (Fig. 1D), indicating the coexistence of multiple virus genotypes circulating within the host across the 27 tissues at a single point in time.
## Consensus diversity is primarily in the spike coding region
Across the 27 tissue sites, we identified all SARS-CoV-2 nucleotide positions with a consensus change compared to the AY.119 reference (Fig. 2A). Two mutations were strain-specific (nsp5: C10449T [P132L] and ORF3a: T25485C) and were found to be fixed (>98%) across all 27 tissue sites (see Fig. S2A). Five mutations (nsp2: C2062T [A419A], nsp14: G18612T [E191D], S: C23243T [P561S], S: T24259C [A899A], and N: C28744T [I157I]) reached ≥50% in only one sample. The remaining nine mutations varied in the number of samples in which they were found as a consensus change and were all nonsynonymous mutations in the spike coding region. One substitution (R19T) is in the N-terminal domain (NTD), and eight substitutions (K417R, V445A, G446V, Y453F, G476S, S477N, K478E, and Q493K) are in the receptor-binding domain (RBD) of spike (Fig. 2A). We confirmed that all samples had sufficient coverage to determine the major nucleotide at each variant position, ensuring the differences in the appearance of a consensus mutation are not due to insufficient read depths in the spike region (Fig. S2A). Several consensus variants were minor variants in other tissue sites (Fig. S2A). Interestingly, the presence of specific substitutions was dependent on the tissue site. For example, G446V and A419A (synonymous change) were not identified in lung, tracheal, or cardiac samples above our limits of detection. In contrast, K417S, G476S, K478E, and A899A (synonymous change) were only found in the lung, tracheal, cardiac, and gastrointestinal tissues (Fig. S2A).
To determine if spike mutations were linked, we assembled predicted SARS-CoV-2 spike haplotypes (H) circulating in the tissue sites at frequencies of at least 10% using CliqueSNV (Materials and Methods) (17). Ten unique spike haplotypes (10%-100%) were identified across the tissue sites with different combinations of 11 nonsynonymous amino acid substitutions in the spike (Table 1, Fig. 2B). Four haplotypes (H1-H4) were found in at least two samples: H1 (V445A-S477N-Q493K) in the jejunum, sinus turbinate, rib, stomach, appendix, left optic nerve, inferior lung lobe, kidney, and superior lung lobe, and the right cornea and optic nerve; H2 (G446V) in the right kidney, nasal placode, tongue, right choroid sclera, and right retina; H3 (R19T-V445A-S477N-Q493K) in the right superior lung lobe, left and right bronchus, and right middle lung lobe; and H4 (R19T-V445A-S477N-K478E-Q493K) in the proximal and distal trachea. Four substitutions (R19T, V445A, S477N, and Q493K) were found in at least four distinct haplotypes in different combinations and with other spike mutations (Fig. 2B). predicted to be present at ≥10% are provided. Color indicates the spike haplotype (H1-H10) identified in each tissue site, with the amino acid residues for each haplotype outlined to the right. Only nonsynonymous amino acid substitutions are provided. Asterisks (*) mark haplotypes found in multiple samples. See also Table 1 and Fig. S2 andS3.
## Full-Length Text
The 11 spike substitutions identified in the patient samples were rarely found in other SARS-CoV-2 sequences circulating between January 2021 and January 2024 (cov-spec trum.org). Of these, only 477N (BA.1, BA.2, BA.4, BA.5, XBB.1.5) and 486S (BA.2) are classified as lineage-defining mutations of Omicron strains circulating after AY.119 (Fig. S3A). Two substitutions, G446V and Y453F, were identified as emergent mutations in Delta infections following monoclonal antibody treatment and were found to decrease neutralization efficiency (18). Although we do not observe the exact lineage-defining mutations, later SARS-CoV-2 strains harbor different lineage-specific substitutions at residues 19, 417, 445, 446, 478, and 493 in the spike protein (Fig. S3B), emphasizing that the mutations identified in our autopsy samples occur at mutational hotspots within the spike coding sequence.
## Most combinations of spike substitutions are associated with increased protein stability and higher binding energy to the ACE2 receptor
Considering that most of the substitutions identified were in the receptor-binding domain of the spike protein, we performed protein simulations on the SARS-CoV-2 spike and human ACE2 receptor complex to test how the mutant spike haplotypes impact the stability and function of the spike protein. We first remodeled the spike-ACE2 complex structure (Protein Data Bank [PDB] ID: 7W98) to fill in missing residues. Eight of the 11 substitutions clustered on unstructured loops of the RBD, a region that directly interacts with ACE2 (Fig. 3A). The remaining two were buried in the loops of the NTD (R19T) and sub-domain 1 (SD1) (P561S) (Fig. 3A).
The structural stability of the spike-ACE2 complex for the AY.119 reference and 10 mutant haplotypes (Table 1) was estimated using the global (whole S-ACE2 complex) and domain-based root mean square deviation (RMSD) analysis, which designates the atomic position fluctuations (Fig. 3B andC; Fig. S4). All systems attained convergence after 10 ns, indicating reliable trajectories. A significant difference in RMSD (adj. P-value < 0.0001, Kruskal-Wallis) was established between AY.119 and all mutant haplotypes. Spike sequences for H9 (G446V-F486S, 1.30 nm), H1 (V445A-S477N-Q493K, 1.35 nm), and H3 (R19T-V445A-S477N-Q493K, 1.42 nm) exhibited highly stable or constrained conforma tions compared to AY.119 (1.51 nm). H4 (R19T-V445A-S477N-K478E-Q493K, 1.49 nm) and H7 (R19T-V445A-S477N, 1.51 nm) showed similar structural dynamics to AY.119. Con versely, H8 (R19T-K417R-V445A-Y453F-G476S, 2.61 nm), H5 (V445A-S477N-Q493K-P561S, 2.24 nm), and H6 (K417R-V445A-Y453F-G476S, 1.88 nm) had the highest structural deviation and potential instability, or increased flexibility compared to AY.119. The higher RMSD is due to the increased flexibility of atoms in the subunit 1 region (S1) of the spike, particularly at sites that do not interact with ACE2 (i.e., NTD) (Fig. 3C; Fig. S4A). Individual residue fluctuations were analyzed by calculating root mean square fluctuation (RMSF) to investigate the effect of each substitution on the spike-ACE2 residue interactions. Most residues in the SARS-CoV-2 spike RBD exhibit minimal fluctuations, with RMSF less than 2.0 Å (Fig. S4B). However, relatively high RMSFs are shown across the residues 14-24, 69-77, and 142-149 (in the NTD region) and were prominent in H5 and H8 sequences relative to the other sites and the AY.119 reference.
To determine the impact of the spike substitutions on spike-ACE2 interactions, we calculated Gibbs free binding energy differences between the ACE2 region (atoms 17,458-26,974) and the spike (atoms 1-17,457) (Fig. 3D andE). All 10 mutant haplotypes exhibited significant changes in Gibbs free energy (P-value <0.0001, one-way ANOVA) with respect to AY.119 (-69.4 kcal mol -1 ), except for H1 (-75.4 kcal mol -1 ) and H10 (-70.6 kcal mol -1 ) (Fig. 3D). In AY.119, residue positions K417, G476, F486, and Q493 contributed significantly to the overall energy profile between spike and ACE2 (Fig. 3E). Notably, residue 417 lost its significance in most haplotypes (H1-H5, H7) (Fig. 3E). Across all haplotypes, including AY.119, residues at 486 and 493 showed the highest contribution to the binding of spike to ACE2, indicating their importance in structural stability (Fig. 3E).
## Combinations of spike substitutions impact the viral binding and internaliza tion of infectious isolates
Infectious virus isolates were plaque-purified in VeroE6-TMPRSS2 cells from 6 of the 16 tissue sites tested (Table S2). Depending on the tissue site, SARS-CoV-2 RNA was isolated and sequenced from one to three purified plaques (Table S3). Isolate 1 spike sequence (R19T-V445A-S477N-K478E-Q493K) was found in 9 of 16 plaque isolates (Table S2) and was the dominant spike haplotype in the sequencing data of the proximal and distal trachea (see H4 genotype in Fig. 2B). The spike sequence of isolate 2 (V445A-S477N-Q493K) was the predominant haplotype in the original sequencing data (H1 genotype in Fig. 2B) but was only found in a sinus turbinate and tongue plaque (plaque #6). The remaining three isolates' spike sequences had additional substitutions (E406D, L455F, T573I, G1267R) that were not observed in the consensus sequences of the original sequencing data. Although the E406D, L455F, and T573I substitutions did not reach ≥50% in the original sequencing data, they were present as minor variants (Fig. S5), indicating they may have an advantage in cell lines. Non-significant differences between AY.119 and each mutant spike haplotype are noted with ns, while significant differences are marked with * and indicate a P-value <0.0001 using a one-way ANOVA test. (E) A heatmap of the decomposed energetic interactions for each residue (y-axis) where a consensus variant was found in our data set. AY.119 reference residues include R19, K417, V445, G446, Y453, L455, G476, S477, K478, Q493, and P561. Only variant residues are outlined for each mutant spike haplotype along the x-axis. If not provided, the residue is the reference AY.119 residue. White boxes indicate the residue is significantly involved in the total binding energy. See also Fig. S4. We tested the differences in viral binding and internalization by quantifying SARS-CoV-2 RNA-dependent RNA polymerase (RdRp) gene copies via qPCR for five isolates with distinct spike sequences (Fig. 4A). Isolates 1, 4, and 5 all shared the isolate 2 (V445A-S477N-Q493K) backbone but carried one to two additional substitutions (Table S3). Isolate 3 was unique and only shared the E406D substitution with isolate 4. At a multiplicity of infection (MOI) of 0.1, the cell line minimally affected binding, with isolates 4 and 5 having significantly lower binding in the TMPRSS2 cell line (P = 0.013, P = 0.0045, respectively), suggesting that all isolates were binding efficiently to the endogenously expressed primate ACE2 on the VeroE6-TMPRSS2 cells. In contrast, viral internalization was significantly reduced for isolates 1, 3, 4, and 5 in the TMPRSS2 cells compared to the ACE2-TMPRSS2 cells, especially under low MOI conditions. This result is consistent with The most similar haplotype sequence found in Fig. 2 is provided in parentheses next to each isolate sequence. See also Tables S2 andS3. (B) Distributions of averaged RMSD (y-axis) for the complete spike-ACE2 complex across the spike isolates (1-4 only) and the AY.119 reference (x-axis). Average RMSD values were calculated for 100 individual frames at every nanosecond of an 80 ns molecular dynamics simulation time (n = 8,000). Boxplots represent the median (middle line), first and third quartiles (box), 1.5 * interquartile range (whiskers), and outliers (points). Non-significant differences between AY.119 and each isolate's spike haplotype are noted with ns, while significant differences (P-value <0.0001, one-way ANOVA test) are marked with *. (C) The mean and standard deviation of the change in binding free energy (kcal/mol) (x-axis) between the entire spike structure and ACE2 for each spike sequence (y-axis) using molecular mechanics Poisson-Boltzmann surface area. prior studies reporting improved viral entry and recovery of low-titer virus when using VeroE6 cells that constitutively express both human ACE2 and TMPRSS2 (19). Across the isolates, the cell line and MOI influenced the binding and internalization, with only minimal differences observed between the isolates in the VeroE6-TMPRSS2 cells at both an MOI of 0.1 and 1 (Fig. 4A). Given that isolate 2 was the predominant spike sequence in our sequencing data and had three substitutions found in isolates 1, 4, and 5, we compared changes in viral binding and internalization relative to isolate 2. Compared to isolate 2, isolates 4 and 5 had the largest increases in RdRp gene expression at an MOI of 0.1 for both binding (P = 0.01 and P = 0.003, respectively) and internalization (P = 0.0094 and P = 0.00084) assays, indicating efficient binding and internalization even at low MOIs. Isolates 4 and 5 each carry one additional substitution from the isolate 2 backbone (V445A-S477N-Q493K), with the addition of E406D and G1267R, respectively. The addition of E406D in isolate 4, located in the receptor-binding domain of the spike, led to an increase in the structural and conformational flexibility compared to the AY.119 reference and the isolate 2 spike sequence (Fig. 4B) but did not decrease the estimated change in binding free energy (Fig. 4C). Interestingly, the E406D substitution also emerged in Delta infections following monoclonal antibody treatment and is associated with reduced neutralization by certain monoclonal antibody therapies (18). These data indicate that the change in structural stability caused by E406D may be beneficial for the binding and entry of the virus in the VeroE6 cells expressing human ACE2 and TMPRSS2. Because G1267R is out of range for the protein complex, the spike sequence from isolate 5 could not be simulated. Isolate 3, which also had the E406D substitution but lacked the isolate 2 backbone, also significantly deviated from the AY.119 reference structure (Fig. 4B); however, it had the lowest binding efficiency in the ACE2-TMPRSS2 cells, especially at a low MOI (Fig. 4A). Together, these results further highlight the coexistence of multiple genotypes circulating within the same tissue and how these mutational differences impact the spike's structural stability and function, which is influenced by cell line propagation and multiplicity of infection.
## Tissue compartmentalization and mixing of minor variant populations
Given that several consensus-level mutations were found as minor variants in other samples (Fig. S2A), we calculated pairwise Bray-Curtis dissimilarity indices (BCI) using the single-nucleotide variant (SNV) information to quantify the minor variant diversity that is shared across tissue sites (Fig. S6A). The pairwise BCI scores were hierarchically clustered, which separated the tissue sites into two broad general anatomical locations, with samples from cluster 1 (n = 11) (lung, tracheal, cardiac, and gastrointestinal tissues) extracted from the thoracic and abdominal regions, while cluster 2 (n = 8) was made up of sites mainly from the head, including ocular, nasal, oropharynx regions, and the rib sample (Fig. 5A). Both bone structures extracted (rib and sinus turbinate) fell into cluster 2. The right kidney sample did not cluster with either group, but clustered with samples in cluster 2 when reducing the stringency of SNV cutoffs to 1% and 50×, as most sites fell slightly below the 100× cutoff (Fig. 5A, inset; Fig. S6B). Interestingly, tissues within each cluster often had different consensus sequences (Fig. 1B, Fig. 2A), indicating that even with a high amount of intra-cluster mixing of minor variants, what becomes dominant is dependent on the tissue site. Cluster 1 tissues had lower BCI scores (mean and SD: 0.35 ± 0.22), and more sharing of variants, compared to sites in cluster 2 (mean and SD: 0.55 ± 0.25).
We investigated the differences in minor variant characteristics between tissue sites within each Bray-Curtis cluster (Fig. 5A). Our analysis revealed that all tissue sites varied in the number of minor SNVs, with the tongue (n = 29), left superior lobe (n = 28), rib (n = 25), left optic nerve (n = 24), and appendix (n = 22) each having more than 20 minor SNVs. Tissue sites from cluster 1 exhibited a lower mean (17.5 ± 4.8) and median (16) minor SNV richness compared to cluster 2 samples (mean: 21.5 ± 4.1, median: 20), although this difference was not statistically significant (P-value = 0.056, Mann-Whitney U-test, Fig. 5B). Importantly, minor SNV richness was not significantly correlated with genome copies (r = 0.046, P-value = 0.85, Pearson, Fig. S6C). Cluster 1 had significantly higher nonsynonymous divergence than synonymous divergence (P-value = 0.00077, Mann-Whitney U-test). In contrast, the left optic nerve, rib, and tongue, which had high SNV richness, exhibited higher synonymous divergence than nonsynonymous divergence. The clusters were primarily defined by the presence and frequency of SNVs in the spike coding region (Fig. 5D). While cluster 2 had more unique mutations, these were mainly shared by only two samples, with no SNVs identified in all eight samples in cluster 2. Conversely, SNVs such as S: 22812G, S: 22988A, and S: 24259C were found in all 11 samples of cluster 1 and none of the samples from clusters 2 or 3. This indicates that cluster 1, despite having fewer unique minor SNVs, had mutations that were more consistently shared across its samples, leading to the lower Bray-Curtis index values. Together, these findings suggest that selection pressures differ significantly across tissues.
## DISCUSSION
We characterized the intrahost genetic diversity of SARS-CoV-2 from distinct tissue sites collected from an autopsy case patient with primary immunosuppression. Phyloge netic analyses of 27 assembled SARS-CoV-2 consensus sequences revealed multiple genotypes circulating across the tissues. The sequences are monophyletic, suggesting divergence within the host rather than coinfection with multiple SARS-CoV-2 genomes. Single-nucleotide accumulation estimates ranged from 1.19 to 6.83 substitutions from the shared ancestral node. This range of estimates fits with the estimated intrahost evolutionary rate of ~35.55 (95% CI: 31.56-39.54) substitutions per year (20), indicating that the infection may have persisted since the initial positive COVID-19 test nearly 2 months prior to death.
Although we cannot determine the exact timing of virus dissemination to non-respi ratory tissues, N copy numbers in the respiratory and non-respiratory tissues are similar to those observed in other autopsy cases with acute infections (<14 days) (1). The high copy numbers and sgRNA abundance within tissues, phylogenetic results, presence of Delta-specific monoclonal antibody escape mutants, and high minority SNV richness and diversity indicate potential virus and disease reactivation, leading to the respiratory distress the individual experienced in late November. This aligns with previous observa tions that higher levels of tissue-specific consensus mutations are found in autopsy cases with longer time intervals between symptom onset and death (3). Although we cannot rule out that a large transmission bottleneck event led to the high variant richness without sampling from the earlier positive tests, estimates of bottleneck sizes for SARS-CoV-2 transmission remain small (14,(21)(22)(23). Fluctuations in virus replication of SARS-CoV-2 have been observed in other longitudinal studies of SARS-CoV-2 infections in immunocompromised patients (24), with low levels of virus shedding happening for months before increasing to detectable levels again. Asymptomatic shedding of SARS-CoV-2 RNA may be significantly underestimated, and tissue reservoirs may play an essential role, as individuals with persistent infections are more likely to experience long-COVID systemic symptoms associated with SARS-CoV-2 infections (4,14,25). Tissue reservoirs, or compartments that allow continued virus replication, are observed for other RNA viruses, including Ebola virus and human immunodeficiency virus (26,27). Reactivation of the Ebola virus in reservoirs has also been shown to cause outbreaks within the general population, although the amount of genetic divergence that occurs within reservoirs is limited, likely due to reduced replication of the virus (28,29). Interestingly, in our data set, dominant sequences with the lowest divergence were found in non-respiratory tissues, including the eyes, appendix, kidney, and tongue. Most consensus-defining mutations identified were in the spike coding sequence. We detected nine dominant spike haplotypes, four of which were dominant in more than one tissue site. In contrast to previous autopsy studies, which reported limited high-frequency mutations in the spike and RBD (1-3), the 11 spike mutations that define our spike haplotypes are predominately located in the RBD and specifically at amino acid positions where lineage-defining mutations of SARS-CoV-2 variants occur. Among these, two spike substitutions-477N detected at varying frequencies across all tissue sites and 486S, found in ocular tissues, tongue, and the nasal placode-were classified as lineage-defining mutations in later Omicron lineages. Additionally, three RBD substitu tions (E406D, G446V, and Y453F) were characterized as emergent mutations in Delta strains following monoclonal antibody treatment and led to weakened neutralization (18). The individual in this study was severely immunocompromised. His underlying CGD, characterized by defects in the NADPH oxidase enzyme complex that impair the phagocytic function of innate immune cells, predisposed him to pulmonary infections that resulted in chronic lung disease with impaired pulmonary structure and function. Daily therapy with sirolimus following HSCT and intermittent oral steroids suppressed his adaptive immune response (30,31). Although we lack the exact host response data for the individual in this study, altered lung structure and function from CGD may impact SARS-CoV-2 clearance, while the diminished adaptive immune response applies weakened selective pressure on the virus, leading to the high rates of nonsynonymous mutations we observed in our data, especially in the spike region.
We performed protein simulations to understand the structural and functional impacts of the observed combinations of spike substitutions. Most mutant structures increased the predicted structural stability of the spike region and decreased the binding energy required to bind with the ACE2 structure. The Q493K substitution, a change from glutamine, a polar amino acid, to a positively charged lysine residue, was the most influential mutation in terms of binding energy. Q493K is rarely found in circulating strains, but others have also documented reduced IgG antibody recognition of this variant (32). We also performed viral binding and internalization assays on five infectious isolates with distinct spike sequences at two MOIs (0.1 and 1) using TMPRSS2-ACE2 and TMPRSS2 VeroE6 cell lines. Differences between the isolates in binding and internaliza tion efficiencies were mainly observed at an MOI of 0.1 in the ACE2-TMPRSS2 VeroE6 cells.
With the exceptions of S: E191D (stomach), S: P561S (tongue), and N: I157I (right inf. lobe) mutations, consensus mutations observed in one tissue were present as a minor variant in at least one other tissue. This indicates that multiple genotypes likely coexist within the same tissue sites, as further backed by our isolation of multiple infectious viruses from the nasal placode and tongue tissues. Coinfection with different genotypes has important implications in SARS-CoV-2 evolution, as it allows for recombination to occur (33,34). Interestingly, we rarely observed complete fixation of mutations, which may point to weak bottlenecks or weakened selective pressures within the host that allow for multiple variants to circulate. Except for the tongue, all tissues had three spike substitutions at varying frequencies (V445A, S477N, Q493K). These three substitutions make up the predominant spike haplotype in the sequencing data and were found in four of five infectious isolates. Although we found shared variants across tissue sites, G446V was only in extrapulmonary tissues. Additionally, tracheal, lung, heart, and gastrointestinal tissues all had three nonsynonymous substitutions in the spike (K417R, G476S, K478E) that were not found above our limits of detection in the ocular, nasophar ynx, or oropharynx tissue sites, highlighting the tissue-specific differences in the types of mutations maintained. Furthermore, minor variants differed in their genomic location, frequency, and type of mutation depending on the tissue in which they were identified.
Additional studies are needed to understand what tissue and cellular factors are shaping these observed differences in viral diversity.
Because we did not observe complete compartmentalization of samples, we calculated pairwise Bray-Curtis dissimilarity indices to quantify the extent of shared low-frequency virus diversity across tissue sites. We found that tissues from the same broad anatomical compartments had lower index scores, suggesting increased mixing. Two distinct clusters formed after hierarchical clustering: one with the ocular, rib, naso-, and oropharynx samples, and another with lower respiratory, gastrointestinal, and heart tissues. Except for the rib sample, the clusters distinguished sites collected from the head versus the chest and abdomen. However, the rib clustered with the one other bone structure, the sinus turbinate. Pulmonary tissues were more likely to share their diversity, which likely reflects how the virus is already well adapted to respiratory epithelium but could also point to extensive tissue damage and a breakdown of compartmentalization within the lungs at the point of collection. The fact that anatomical distance appears to influence the number of shared variants between tissue sites raises interesting ques tions regarding the routes of SARS-CoV-2 dissemination. One widely accepted theory is that SARS-CoV-2 spreads via the hematogenous route, where the virus may enter the bloodstream via passage through the gastrointestinal tract or respiratory tract and transmit to other organs (1,8,9). Thus, hematogenous spread could allow for extrapul monary-specific viruses to transit back to the respiratory tract and be transmitted to new hosts. In addition, animal studies have shown that intranasal inoculation can lead to the presence of SARS-CoV-2 in the brain, ocular tissues, and lungs, while intratracheal inoculation did not result in the virus entering the brain or ocular tissues, suggesting that the spread to these areas is neuronal and unidirectional (7,35). SARS-CoV-2 has been identified in the trigeminal neurons in humans and may point to why we observed shared variant populations between the ocular, oropharynx, and nasopharynx tissue sites that were not shared with other tissues. The results presented here offer an important case study, demonstrating high intrahost spike diversity, the coexistence of multiple genotypes within a single tissue site, and tissue-specific differences in virus diversity.
## Limitations of the study
Our study has several limitations, the most notable being that the data were collected from a single individual. However, the inclusion of minor variant data and the extensive tissue set from our study allow us to uniquely bridge observations made by others, including the longitudinal sampling of multiple genotypes from nasopharyngeal swabs during a chronic SARS-CoV-2 infection (20), high spike diversity in immunocompromised patients (34), and tissue compartmentalization of SARS-CoV-2 (1-3). Additionally, we lack host response data, such as antibody titers or gene expression profiles, which would provide insights into the pressures experienced by SARS-CoV-2. Furthermore, virus isolation, binding, and internalization assays were conducted exclusively in VeroE6 cells, which may bias the types of viruses we could isolate and their replication dynamics, particularly those adapted to specific cell and tissue types. Despite these limitations, our data suggest that patients with immunocompromising diseases that weaken adaptive immune responses may create an environment permissive to SARS-CoV-2 evolution and diversification, emphasizing the importance of improving early care strategies for these patients.
## MATERIALS AND METHODS
## Experimental model and study participant details
The study participant was a 57-year-old male patient who died on 28 November 2021 from respiratory failure. An autopsy was performed on 2 December 2021, and tissues were collected as previously described (36) in the National Cancer Institute's Laboratory of Pathology at the National Institutes of Health Clinical Center following consent of the legal next of kin.
Primary isolation of virus from each tissue was done on VeroE6-TMPRSS2-T2A-ACE2 cells; plaque purification was done on Vero E6-TMPRSS2 cells. Viral binding and internalization efficiencies were tested using VeroE6 cells constitutively expressing human ACE2 and TMPRSS2 (VeroE6-TMPRSS2-ACE2) or only human TMPRSS2 (VeroE6-TMPRSS2).
Autopsy, RNA isolation, SARS-CoV-2 RNA, and subgenomic RNA quantifica tion Tissues were collected from the patient as previously described, with adjacent por tions flash-frozen and preserved in RNAlater (Invitrogen) (1). SARS-CoV-2 RNA was isolated using the RNeasy Mini, RNeasy Fibrous Tissue Mini, RNeasy Lipid Tissue Mini, and QIAamp Viral RNA Mini kits (Qiagen) according to the manufacturer's protocols. SARS-CoV-2 RNA quantification was performed using the QX200 AutoDG ddPCR system (Bio-Rad) as detailed in Stein et al. (1). Averages of the N1 and N2 technical replicates are reported in Table S1. SARS-CoV-2-positive samples are those with an average greater than or equal to the manufacturer's limit of detection of 0.1 copies per microliter and ≥2 positive droplets per well. Subgenomic RNA analyses were performed on 32 tissues using 5 µL of sample RNA input into a one-step real-time RT-qPCR assay targeting the envelope (E) gene.
## Amplification, library preparation, and sequencing
SARS-CoV-2 RNA was amplified using the ARTIC V4 primer set and protocol. The ~400 bp amplicons were cleaned using AMPure beads and used as input into the Illumina DNA Prep kit (Illumina, San Diego, CA) according to the manufacturer's protocol. The final concentration and average fragment size of each library were quantified using the Qubit dsDNA HS Assay (Thermo Fisher Scientific, Waltham, MA) and a high-sensitivity D1000 ScreenTape (Agilent, Santa Clara, CA), respectively. Libraries were diluted to be equimolar and pooled at a final concentration of 4 nM before sequencing on the MiSeq (Reagent Kit v.3, 600 cycles, Illumina, San Diego, CA). Amplification, library preparation, and sequencing for all extractions, including vRNA from isolates, were performed twice using vRNA from the same extraction.
Where available, vRNA from the tissue extractions was also sequenced using PacBio HiFi Viral protocol as recommended by the manufacturer (Pacific Biosciences) and outlined previously in Nicot et al. (16). In short, SARS-CoV-2-specific RNA was enriched with a panel of highly specific probes. Following capture, these probes were used to prime molecular inversion PCR to generate tiled amplicons of ~800 bp (~650 bp target, combined ~150 bp for flanking tags) across the SARS-CoV-2 genome. A second round of amplification was performed to introduce sequencing barcodes to each sample to allow for sample multiplexing. Amplified viral genome material from all samples was equimolar pooled and used as input into SMRTbell template preparation, performed as recommended by the manufacturer. Briefly, the amplicon pool underwent DNA damage repair, end repair, and ligation to hairpin sequencing adapters. Endonuclease treatment following adapter ligation was used to remove any unligated amplicon material that would interfere with successful library loading. Libraries were then purified with AMPure PB beads (Pacific Biosciences, Menlo Park, CA), and quality and quantity were evaluated using the Fragment Analyzer (Agilent Biosciences, Santa Clara, CA) and Qubit Fluorome ter 4 (Thermo Fisher Scientific, Waltham, MA), respectively. Sequencing was performed using a single pool for all autopsy samples on the Sequel IIe system using polymer ase v.2.1 and 8 h movies. Circular consensus reads with accuracy greater than 99% (high fidelity, "HiFi" reads) were generated on instrument and used for all downstream analyzes.
## Alignments, filtering, consensus sequences, and variant calling
Primers and low-quality nucleotides were trimmed from the raw sequencing data using Trimmomatic (v.0.39) before aligning the reads to the SARS-CoV-2 reference genome (NC_045512.2, Wuhan-1) using BWA-MEM (v.0.7.17) (37)(38)(39). Duplicate reads were removed from the alignments using PicardTools (v.2.22.2) RemoveDuplicates function (40). Given the sparse read coverage observed in the ARTIC V4 Illumina data, the replicate short-read (average read length: 150 bp, Illumina) data for each sample were combined into one file using Samtools (v.1.14) before calling SNVs using timo (41,42). Sample consensus sequences were generated from the timo outputs for all the long-read (average read length: 650 bp, PacBio) alignments and the short-read (Illumina) alignments for the tongue and pericardium samples. Samples were required to have at least 80% of the genome covered at 10× read depths to be used for consensus analyses (Table S1).
The SARS-CoV-2 consensus sequences were defined as an AY.119 Delta lineage using NextClade (v.3.8.2) (43). Therefore, all single-nucleotide variants identified were compared to an AY.119 reference sequence while maintaining the NC_045512.2 reference genome numbering (i.e., ignoring deletions and insertions). Mutations in the 5´ and 3´ untranslated regions of the viral genome were ignored due to poor sequencing coverage in both the short-read and long-read data. Due to sparse and uneven shortread sequencing, SNV analyses were performed using long-read data and the short-read data for the tongue and pericardium data (Table S1). Consensus, or major, variants were considered those with the highest relative frequency at positions with ≥10× read depths. Minor variants were considered nucleotide variants with the lowest relative frequency at a given position. All variants found in the long-read data at frequencies of 5%-100% and a read depth of ≥100× were kept for analyses and confirmed to be present in the short-read data (≥2%, 200×) when read depths in the short-read libraries were ≥10×. Variants found in both the long-read (2%-5%, 100×) and short-read data (≥2%, 200×) were also kept for analyses.
## Circulating variants and lineage-defining mutations
Proportional data for SARS-CoV-2 lineages and variant positions circulating in the USA from 1 January 2021 to 1 January 2024 were downloaded from CoV-Spectrum (44) (cov-spectrum.org) and visualized in R (45) (v.4.2.3) using ggplot2 (3.4.3) (46).
## Phylogenetic analyses
The 27 SARS-CoV-2 consensus sequences from samples with at least 10× read depths across 80% of the genome were used for phylogenetic analyses. Complete consensus sequences of 6,000 AY.119 strains circulating in humans in the USA from 1 August 2021 to 31 January 2022 were randomly selected and downloaded from GISAID (https:// doi.org/10.55876/gis8.240924yd) (47). All sequences were aligned to the NC_045512.2 (Wuhan-1) reference sequence with the 5´ (1-265) and 3´ (29,675-29,903) removed using MAFFT (v.7.475) and input parameters: --6merpair --keeplength and --addfrag ments (48). Maximum-likelihood trees were constructed with IQ-TREE (v.2.2.0.5) using the generalized time-reversible (GTR+G) model and 1,000 bootstrap replicates (49). Substitution rates were estimated using the time-resolved data from TreeTime (v.0.11.1) with collection dates for the 6,000 GISAID provided as input with the maximum-like lihood tree generated using IQ-TREE (50). Outliers were removed from the tree and root-to-tip distance calculations. Root-to-tip calculations were performed using ape (v.5.7-1), and tree visualization was done in R (4.2.3) using ggtree (3.6.0) and treeio (v.1.22.0) (51)(52)(53).
## Haplotype construction
SARS-CoV-2 spike haplotypes were assembled for each tissue site using CliqueSNV (v.2.0.3). CliqueSNV assembles minority SNVs into predicted haplotypes by calculating the probability that SNV pairs are linked using read counts, relative frequencies, and haplotype length. We limited the haplotype region to the spike coding region (21,384) and set the haplotype frequency and read depth parameters to 10% and 5× to increase confidence (17). All positions were compared to the AY.119 reference with NC_045512.2 (Wuhan-1) numbering.
## Bray-Curtis dissimilarity index
To compare the minor SNV populations across samples, the pairwise BCI were calculated by pulling each sample's relative frequency of SNVs (0%-100%) that were found as a minor variant in at least one sample. Samples that did not have the SNV above our thresholds were marked as having the SNV at 0%. Those found only as a consensus variant (>50%) in tissue sites were not used, as we wanted to see the flow of variants and diversity found at low relative frequencies in tissues. Nucleotide variants were considered independent.
The BCI was calculated between two tissue sites (j and k), where C jk is the sum of the lesser frequencies of each minor variant between tissue sites j and k, and S j and S k are the total sums of minor variant frequencies at each respective tissue site.
$$BC jk = 1 - 2C jk S j + S k$$
## Minor variant richness, divergence, and dN/dS
Richness of minor variants was calculated by counting the total number of single-nucleo tide variants for each sample that passed our required confidence thresholds (outlined above). Divergence per site was calculated by summing the relative frequency of either nonsynonymous or synonymous substitutions and normalizing by the total number of nonsynonymous or synonymous sites in the coding regions of the genome (54). Similarly, dN/dS ratios were calculated by taking the ratio of the number of nonsynony mous mutations normalized by the total number of nonsynonymous sites in the coding regions of the genome to the number of synonymous mutations normalized by the total number of synonymous sites in the coding regions of the genome. For both calculations, deletions, insertions, and non-coding or untranslated sites were not included, and all variant positions were considered independent and unlinked, as confidence in linked sites decreases for low-frequency SNVs.
## Spike structure remodeling and validation
The open spike crystallized structures and protein sequences of the SARS-CoV-2 spike were downloaded from the PDB for Wuhan-01 (PDB ID: 7CAK) and Delta AY.119 (PDB ID: 7W98) (55). Due to missing residues in the three-dimensional (3D) structure, remod eling was done using the SWISS-MODEL server (56). Initially, the template structure was identified, followed by template-target sequences alignment using MAFFT (48) alignment tool. The reliability of the best three resulting models was assessed using PROCHECK (57), VERIFY3D (58), and ProSA-web (59) tools. The best structure was used for subsequent analyses. The consensus evaluation of the remodeled reference structure (PDB ID: 7W98) indicated a reliable and accurate starting structure, with 81.86% of residues scoring ≥0.1 in the 3D/1D profile, an overall quality factor >87.3%, and stereochemistry parameters showing 88.8% of residues in core regions. Mutant residues for each spike haplotype were manually inserted at their specified positions using Discovery Studio Visualizer (60). The protonation state of pH 7.0 was applied to all systems using the H ++ Server (http://biophysics.cs.vt.edu/H++) (61).
## Viral binding and internalization assays
Five plaque-purified isolates (Table S2) were selected to phenotypically test their binding and internalization efficiencies using VeroE6 cells constitutively expressing human ACE2 and TMPRSS2 or only human TMPRSS2 at MOIs of 1 and 0.1. To assess virus binding to the cell surface, cells were incubated with either 0.1 or 1 MOI of relevant viral isolate on ice for 1 h at 37°C before washing three times with cold PBS. After washing, 600 µL of RNA lysis buffer was added directly to the cells. To determine viral internalization into the cells, cells were incubated with the desired MOI (0.1 or 1) of viral isolates on ice for 1 h at 37°C. The cells were washed three times with cold PBS to remove any virus that did not bind to the cells. Pre-warmed 2 mL of serum-free DMEM medium was added to the cells, followed by a 3 h incubation at 37°C. The cells were washed three times with PBS, followed by addition of 600 µL RNA lysis buffer to the cells. The RNA extractions were performed on all harvested samples using Direct-Zol RNA Miniprep Kit (Zymo), and cDNA was prepared using High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) as per the manufacturer's protocol. Viral RNA levels and replication were measured as previously described in Kar et al. (67). Briefly, qRT-PCR was performed using IDT Prime Time gene expression master mix on a QuantStudio5 qPCR system using the cycling conditions recommended by the manufacturer. To measure viral RNA levels, SARS-CoV-2 RdRp-specific forward primer: GTGARATGGTCATGTGTGGCGG; reverse primer: CARATGTTAAASACACTATTAGCATA, and probe 56-6-carboxyfluorescein [FAM]/ CAGGTGGAA/ZEN/CCTCATCAGGAGATGC/3IABkFQ were used. Ct values were normalized to the reference gene GAPDH and represented as a fold change over values from time-matched mock samples.
## Quantification and statistical analysis
All statistical analyses and data visualizations were performed using R and ggplot2. Welch's t-test was used to assess the significance of minor variant comparisons (richness, divergence, and dN/dS) and isolate comparisons. Pearson's correlation coefficient was used to evaluate the strength and significance of correlations between variables. Differences in RMSD values between AY.119 and all mutant spike haplotypes were analyzed using the Kruskal-Wallis test. Changes in Gibbs free energy were compared using one-way ANOVA. A P-value of less than 0.05 was considered significant for all tests, while ns indicates comparisons that were not significant. The specific statistical tests and P-values are detailed in the results text and figure legends. S1 to S3.
## Supplemental Material
## References
1. Stein, Ramelli, Grazioli et al. (2022) "SARS-CoV-2 infection and persistence in the human body and brain at autopsy" *Nature*
2. Van Cleemput, Van Snippenberg, Lambrechts et al. (2021) "Organ-specific genome diversity of replication-competent SARS-CoV-2" *Nat Commun*
3. Normandin, Rudy, Barkas et al. (2023) "High-depth sequencing characterization of viral dynamics across tissues in fatal COVID-19 reveals compartmentalized infection" *Nat Commun*
4. Manrique, Maffia-Bizzozero, Delpino et al. (2024) "Multi-organ spread and intra-host diversity of SARS-CoV-2 support viral persistence, adaptation, and a mechanism that increases evolvability" *J Med Virol*
5. 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*
6. Jackson, Farzan, Chen et al. (2022) "Mechanisms of SARS-CoV-2 entry into cells" *Nat Rev Mol Cell Biol*
7. Jeong, Kwon, Ng et al. (2022) "Ocular tropism of SARS-CoV-2 in animal models with retinal inflammation via neuronal invasion following intranasal inoculation" *Nat Commun*
9. Jacobs, Bain, Naqvi et al. (2022) "Severe acute respiratory syndrome coronavirus 2 viremia is associated with coronavirus disease 2019 severity and predicts clinical outcomes" *Clin Infect Dis*
10. Li, Liu, Yang et al. (2020) "Neurological manifestations of patients with COVID-19: potential routes of SARS-CoV-2 neuroinvasion from the periphery to the brain" *Front Med*
11. Platt, Singh, Stein et al. (2024) "Replication-competent virus detected in blood of a fatal COVID-19 case" *Ann Intern Med*
12. Proal, Vanelzakker, Aleman et al. (2023) "SARS-CoV-2 reservoir in post-acute sequelae of COVID-19 (PASC)" *Nat Immunol*
13. Cevik, Tate, Lloyd et al. (2021) "SARS-CoV-2, SARS-CoV, and MERS-CoV viral load dynamics, duration of viral shedding, and infectiousness: a systematic review and meta-analysis" *Lancet Microbe*
14. Natarajan, Zlitni, Brooks et al. (2022) "Gastrointestinal symptoms and fecal shedding of SARS-CoV-2 RNA suggest prolonged gastrointestinal infection" *Med*
15. Ghafari, Hall, Golubchik et al. (2024) "Prevalence of persistent SARS-CoV-2 in a large community surveillance study" *Nature*
17. Bowe, Xie, Al-Aly (2022) "Acute and postacute sequelae associated with SARS-CoV-2 reinfection" *Nat Med*
18. Nicot, Trémeaux, Latour et al. (2023) "Whole-genome sequencing of SARS-CoV-2: comparison of target capture and amplicon single molecule real-time sequencing protocols" *J Med Virol*
19. Knyazev, Tsyvina, Shankar et al. (2021) "Accurate assembly of minority viral haplotypes from nextgeneration sequencing through efficient noise reduction" *Nucleic Acids Res*
20. Ragonnet-Cronin, Nutalai, Huo et al. (2023) "Generation of SARS-CoV-2 escape mutations by monoclonal antibody therapy" *Nat Commun*
21. Kinoshita, Yamamoto, Kuroda et al. (2024) "Improved efficacy of SARS-CoV-2 isolation from COVID-19 clinical specimens using VeroE6 cells overexpressing TMPRSS2 and human ACE2" *Sci Rep*
22. Chaguza, Hahn, Petrone et al. "Yale SARS-CoV-2 Genomic Surveillance Initiative. 2023. Accelerated SARS-CoV-2 intrahost evolution leading to distinct genotypes during chronic infection" *Cell Rep Med*
23. Bendall, Callear, Getz et al. (2023) "Rapid transmission and tight bottlenecks constrain the evolution of highly transmissible SARS-CoV-2 variants" *Nat Commun*
24. Hannon, Roychoudhury, Xie et al. (2022) "Narrow transmission bottlenecks and limited within-host viral diversity during a SARS-CoV-2 outbreak on a fishing boat" *Virus Evol*
25. Braun, Moreno, Wagner et al. (2021) "Acute SARS-CoV-2 infections harbor limited within-host diversity and transmit via tight transmission bottlenecks" *PLoS Pathog*
26. Nussenblatt, Roder, Das et al. (2025) *Full-Length Text Journal of Virology*
27. Kochenderfer, Ghedin (2022) "Yearlong COVID-19 infection reveals within-host evolution of SARS-CoV-2 in a patient with B-cell depletion" *J Infect Dis*
28. Peluso, Ryder, Flavell et al. (2024) "Tissue-based T cell activation and viral RNA persist for up to 2 years after SARS-CoV-2 infection" *Sci Transl Med*
29. Dufour, Ruiz, Pagliuzza et al. (2023) "Near full-length HIV sequencing in multiple tissues collected postmortem reveals shared clonal expansions across distinct reservoirs during ART" *Cell Rep*
30. Keita, Vidal, Toure et al. (2019) "A 40-month follow-up of ebola virus disease survivors in Guinea (postEbogui) reveals long-term detection of ebola viral ribonucleic acid in semen and breast milk" *Open Forum Infect Dis*
31. Diallo, Sissoko, Loman et al. (2016) "Resurgence of ebola virus disease in Guinea linked to a survivor with virus persistence in seminal fluid for more than 500 days" *Clin Infect Dis*
32. Blackley, Wiley, Ladner et al. (2016) "Reduced evolutionary rate in reemerged ebola virus transmission chains" *Sci Adv*
33. Gardiner, Deffit, Mcletchie et al. (2013) "A role for NADPH oxidase in antigen presentation"
34. Moir, De Ravin, Santich et al. (2012) "Humans with chronic granulomatous disease maintain humoral immunologic memory despite low frequencies of circulating memory B cells" *Blood*
35. Clark, Clark, Pan et al. (2021) "SARS-CoV-2 evolution in an immunocompromised host reveals shared neutralization escape mechanisms" *Cell*
36. Jackson, Boni, Bull et al. (2021) "Generation and transmission of interlineage recombinants in the SARS-CoV-2 pandemic" *Cell*
37. Ko, Radecki, Belinky et al. (2024) "Rapid intra-host diversification and evolution of SARS-CoV-2 in advanced HIV infection" *Nat Commun*
38. Sen, Vannella, Wang et al. (2023) "Histopathology and SARS-CoV-2 cellular localization in eye tissues of COVID-19 autopsies" *Am J Pathol*
39. Huang, Pérez, Kato et al. (2021) "SARS-CoV-2 infection of the oral cavity and saliva" *Nat Med*
40. Li, Durbin (2009) "Fast and accurate short read alignment with burrows-wheeler transform" *Bioinformatics*
41. Li (2013) "Aligning sequence reads, clone sequences and assembly contigs with BWA"
42. Bolger, Lohse, Usadel (2014) "Trimmomatic: a flexible trimmer for Illumina sequence data" *Bioinformatics*
43. Picard Toolkit (2019) "Broad Institute, GitHub Repository"
44. Danecek, Bonfield, Liddle et al. (2021) "Twelve years of SAMtools and BCFtools. Gigascience 10"
45. Roder, Johnson, Knoll et al. (2023) "Optimized quantification of intra-host viral diversity in SARS-CoV-2 and influenza virus sequence data"
46. Aksamentov, Roemer, Hodcroft et al. (2021) "Nextclade: clade assignment, mutation calling and quality control for viral genomes" *JOSS*
47. Chen, Nadeau, Yared et al. (2022) "CoV-Spectrum: analysis of globally shared SARS-CoV-2 data to identify and characterize new variants" *Bioinformatics*
48. (2023) "R: a language and environment for statistical computing"
49. Wickham (2016) "Ggplot2: elegant graphics for data analysis"
50. Elbe, Buckland-Merrett (2017) "Data, disease and diplomacy: GISAID's innovative contribution to global health" *Glob Chall*
51. Katoh, Standley (2013) "MAFFT multiple sequence alignment software version 7: improvements in performance and usability" *Mol Biol Evol*
52. Nguyen, Schmidt, Haeseler et al. (2015) "IQ-TREE: A Fast and Effective Stochastic Algorithm for Estimating Maximum-Likelihood Phylogenies" *Mol Biol Evol*
53. Sagulenko, Puller, Neher (2018) "TreeTime: maximum-likelihood phylodynamic analysis" *Virus Evol*
54. Paradis, Schliep (2019) "Ape 5.0: an environment for modern phylogenetics and evolutionary analyses in R" *Bioinformatics*
55. 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*
56. 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*
57. Xue, Bloom (2020) "Linking influenza virus evolution within and between human hosts" *Virus Evol*
58. Berman, Westbrook, Feng et al. (2000) "The protein data bank" *Nucleic Acids Res*
59. Waterhouse, Bertoni, Bienert et al. (2018) "SWISS-MODEL: homology modelling of protein structures and complexes" *Nucleic Acids Res*
60. Laskowski, Macarthur, Moss et al. (1993) "PROCHECK: a program to check the stereochemical quality of protein structures" *J Appl Crystallogr*
61. Eisenberg, Lüthy, Bowie (1997) "VERIFY3D: assessment of protein models with three-dimensional profiles" *Methods Enzymol*
62. Wiederstein, Sippl (2007) "ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins" *Nucleic Acids Res*
63. Biovia (2019)
64. Gordon, Myers, Folta et al. (2005) "H++: a server for estimating pKas and adding missing hydrogens to macromo lecules" *Nucleic Acids Res*
65. (2025) *Full-Length Text Journal of Virology*
66. Abraham, Murtola, Schulz et al. (2015) "GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers" *SoftwareX*
67. Mark, Nilsson (2001) "Structure and dynamics of the TIP3P, SPC, and SPC/E water models at 298 K" *J Phys Chem A*
68. Petersen (1995) "Accuracy and efficiency of the particle mesh Ewald method" *Chem Phys*
69. Humphrey, Dalke, Schulten (1996) "VMD: visual molecular dynamics" *J Mol Graph*
70. Valdés-Tresanco, Valdés-Tresanco, Valiente et al. (2021) "gmx_MMPBSA: a new tool to perform end-state free energy calculations with GROMACS" *J Chem Theory Comput*
71. Kar (2024) "CD4(+) and CD8(+) T cells are required to prevent SARS-CoV-2 persistence in the nasal compartment" *Sci Adv* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12451622&blobtype=pdf | # Summary of taxonomy changes ratified by the International Committee on Taxonomy of Viruses (ICTV) from the Animal DNA Viruses and Retroviruses Subcommittee, 2025
Arvind Varsani, Adly Abd-Alla, Niklas Arnberg, Kelly Bateman, Mária Benkő, Annie Bézier, Philippe Biagini, Jamie Bojko, Anamarija Butkovic, Marta Canuti, Vladimír Celer, Jean-Michel Drezen, Laszlo Egyed, Matthias Fischer, Sarah François, Benjamin Guinet, Balázs Harrach, Robert Harrison, Elisabeth Herniou, Michael Hess, Jia Hu, Johannes Jehle, Győző Kaján, Adrianna Kajon, Eugene Koonin, Simona Kraberger, Peter Krell, Mart Krupovic, Jens Kuhn, Chengfeng Lei, Matthieu Leobold, Fabrizio Maggi, Suresh Mittal, Hiroaki Okamoto, Tanja Opriessnig, Xiaowei Peng, Judit Pénzes, Iva Podgorski, Thomas Postler, Bergmann Ribeiro, Carmen San Martín, Maria Söderlund- Venermo, Xiulian Sun, András Surján, Zoltán Tarján, Julien Varaldi, Márton Vidovszky, Göran Wadell, Hidemi Watanabe, Natalya Yutin, Monique Van Oers, Ictv Taxonomy, Summary Consortium
## INTRODUCTION
The Animal DNA Viruses and Retroviruses Subcommittee of the International Committee on Taxonomy of Viruses (ICTV) [1] includes Study Groups for the families Adenoviridae, Anelloviridae, Ascoviridae, Asfarviridae, Baculoviridae, Nudiviridae, Bidnaviridae, Circoviridae, Hepadnaviridae, Herpesvirales, Hytrosaviridae, Iridoviridae, Nimaviridae, Papillomaviridae, Parvoviridae, Polydnaviriformidae, Polyomaviridae, Poxviridae and Retroviridae. These Study Groups include virologists who are experts in viruses classified within these families. As with the ICTV Executive Committee, the 165 members of the Animal DNA Viruses and Retroviruses Subcommittee volunteer their time to advance virus taxonomy [1]. The classified viruses and viriforms in the families represented by the Animal DNA Viruses and Retroviruses Subcommittee are classified into 15 hierarchical ranks for virus classification [2].
Taking various aspects of virus taxonomy into account [2][3][4][5], coupled with the binomial species nomenclature (genus +epithet) [6][7][8], 13 proposals were submitted as part of the Animal DNA Viruses and Retroviruses Subcommittee in 2024; and 11 were
## Abstract
The International Committee on Taxonomy of Viruses (ICTV) holds a ratification vote annually after review of newly proposed taxa by ICTV Study Groups and members of the virology community. In March 2025, the vote outcome of the 11 proposals within the mandate of the Animal DNA Viruses and Retroviruses Subcommittee was made public. Here, we provide a summary of the newly accepted proposals. These include reorganization of taxa in the realm Varidnaviria, classification of the 'polinton-like' viruses into a new family (Phypoliviridae) within a new order Archintovirales; establishment of a new phylum (Commensaviricota) in the kingdom Shotokuvirae; the establishment of a new family called Filamentoviridae with two new genera and three new species; the addition of four new genera in the family Anelloviridae with 70 new species; and the addition of 85 new species in the families Adenoviridae (n=16), Baculoviridae (n=5), Circoviridae (n=5), Parvoviridae (n=55) and Polyomaviridae (n=4). Also, in the family Belpaoviridae, 11 species were renamed to comply with the binomial requirement for species names.
accepted by the ICTV Executive Committee for ratification by the broader ICTV membership. Here, we summarize the changes occurring as a result of the ratification of these 11 proposals.
A significant change is the reorganization of taxa in the realm Varidnaviria, including movement of the kingdom Helvetiavirae into a new realm named Singelaviria based on the independent origins of the major capsid proteins encoded by the viruses in the kingdoms Helvetiavirae and Bamfordvirae [9,10]. Also, as part of this reorganization, a new kingdom named Abadenavirae was established as a result of the evolutionary analysis of the protein-primed family B DNA polymerase or its derivatives encoded by tectivirids [11,12], adenovirids [13], adintovirids [14], maveriviricetes and previously unclassified 'polinton-like' viruses [11]. The kingdom Abadenavirae now includes all bacterial and archaeal viruses with double jelly-roll major capsid proteins, with the exception of tectivirids, which remain in the kingdom Bamfordvirae along with all the evolutionarily related eukaryotic viruses. The phylum Preplasmiviricota also underwent notable refinements with the establishment of two new subphyla, two new orders and one new family. The previously unassigned family Yaraviridae [15] was moved into a new class Mriyaviricetes in the phylum Nucleocytoviricota [3,16]; the family Adintoviridae was renamed Eupolintoviridae; the first representative of an extensive group of viruses, broadly known as 'polinton-like' viruses [17,18], was classified into a new family, Phypoliviridae, within a new order Archintovirales [9]. Within the realm Varidnaviria, other changes include renaming the species African swine fever virus in the family Asfarviridae [19] to Asfivirus haemorrhagiae to comply with the binomial name rule [6,7] and establishing 16 new species in the family Adenoviridae [13].
Anelloviridae [20,21] was the only family of eukaryotic single-stranded DNA viruses not assigned to the realm Monodnaviria [3]. Although anellovirids do not encode a homologue of the replication-associated endonuclease of the HUH superfamily, the signature of Monodnaviria, recent results have shown that they encode capsid protein orthologs with a jelly-roll fold typical of cressdnaviricot capsid proteins, establishing an evolutionary link to other eukaryotic ssDNA viruses, specifically, circovirids [22]. Thus, the family Anelloviridae is now classified in the order Sanitavirales, class Cardeaviricetes, phylum Commensaviricota, kingdom Shotokuvirae and realm Monodnaviria. Furthermore, 4 new genera and 70 new species have been established in the family Anelloviridae to classify anellovirids with diverse hosts ranging from marine mammals to terrestrial mammals and avians of various species [22][23][24][25][26][27][28]. Other changes within the phylum Cossaviricota in the kingdom Shotokuvirae include the establishment of 55 new species in the family Parvoviridae (order Piccovirales, class Quintoviricetes) [29] and four species in the family Polyomaviridae (order Sepolyvirales, class Papovaviricetes) [30]. In the phylum Cressdnaviricota, five new species were established in the family Circoviridae (order Cirlivirales, class Arfiviricetes) [31].
The family Filamentoviridae (order Lefavirales, class Naldaviricetes), with two genera (Alphafilamentovirus and Betafilamentovirus) and three species, was established to classify filamentous DNA viruses identified in insects [32][33][34][35][36][37]. In addition, in the order Lefavirales [38], class Naldaviricetes, five new species were established in the family Baculoviridae [39], and one species (Alphabaculovirus altermaconfiguratae) was abolished.
Within the realm Riboviria, in the family Belpaoviridae (order Ortervirales, class Revtraviricetes, phylum Artverviricota, kingdom Pararnavirae) [40], all 11 species were renamed to comply with binomial name rules.
A file including all the Tables of taxonomic changes below is available as a supplementary file to this article.
## MAIN TEXT CONTENTS
## 2024.001D.Alphabaculovirus_1nsp
Title: Create the new species Alphabaculovirus alterhycuneae in the genus Alphabaculovirus (Lefavirales: Baculoviridae) Authors: Peng X-W, Lei C-F, Hu J, Sun XL ( sunxl@ wh. iov. cn) Summary Taxonomic rank(s) affected: Species Description of current taxonomy: In the genus Alphabaculovirus (family Baculoviridae), there are 65 species.
## Proposed taxonomic change(s):
Add one (1) new species to genus Alphabaculovirus.
## Justification:
The genome of the virus Hypantria cuneae nulceopolyhedrovirus B was fully sequenced using a high-throughput method. The divergence in the phylogenetic tree and the K2P distances based on the 38 core-gene concatenated alignment revealed that this virus belongs to a novel species of Alphabaculovirus. For this new species, the species name Alphabaculovirus alterhycuneae is suggested, following the binomial naming proposal as submitted in 2022 and ratified by the ICTV in April 2023 [41]. Submitted: 05/04/23
## Justification:
Based on genome organization and phylogenetic analyses, the establishment of five new species in the genus Circovirus is proposed. The species demarcation is based on the genome-wide pairwise identity between circovirids (less than 80% identity as established species demarcation criterion) [31,42]. Submitted: 21/06/24
## 2024.007D.Filamentoviridae_1nf_2ngen_3nsp
Title: Create a new virus family in the Lefavirales order named Filamentoviridae with two genera Alphafilamentovirus and Betafilamentovirus and three species.
Authors: Bézier A ( annie. bezier@ univ-tours. fr), Leobold M, Guinet B, Drezen J-M, Herniou EA, Varaldi J
## Summary Taxonomic rank(s) affected:
Establishment of a new highly diverse viral family within the order Lefavirales in the class Naldaviricetes, the Filamentoviridae, comprising two genera: Alphafilamentovirus, with the species Alphafilamentovirus leboulardi, and Betafilamentovirus, with the species Betafilamentovirus cocongregatae and Betafilamentovirus altercocongregatae.
## Description of current taxonomy:
The class Naldaviricetes currently includes four families: Baculoviridae, Nudiviridae, Hytrosaviridae and Nimaviridae, the first three belonging to the order of Lefavirales.
## Proposed taxonomic change(s):
Create Filamentoviridae, a new family in the order Lefavirales within the class Naldaviricetes, with two (2) genera (Alphafilamentovirus and Betafilamentovirus) and three species.
## Justification:
New large arthropod-specific dsDNA viruses, which have been described as filamentous particles since the 1970s, have recently been characterized at the genomic level [33]. These viruses share signatures of members of the class Naldaviricetes and order Lefavirales, while encoding specific core genes that distinguish them from the established families of this order. Phylogenetic tree reconstruction indicates that these filamentous viruses form a monophyletic clade distinct from that of their closest relatives, Hytrosaviridae, and supports the creation of a new family, that we propose to name Filamentoviridae. These viruses appear to be preferentially associated with hymenopteran insects with a parasitoid lifestyle [33]. The effects of filamentous viruses on their hosts are still poorly understood compared with other members of the Naldaviricetes. So far, only the Leptopilina boulardi filamentous virus has been studied for its effect and is described as inducing a behavioural manipulation of wasp oviposition decisions and benefiting from the vertical and horizontal transmission. Submitted: 04/06/24; Revised: 23/10/24
## 2024.012D.Shotokuvirae_newphylum
## References
1. "Joint FAO/IAEA Programme of Nuclear Techniques in Food and Agriculture"
2. "France; 7 Equipe Biologie des Groupes Sanguins, UMR 7268 ADES"
3. (1946) "CNRS UMR6047"
4. De, Evolutive "Yakushiji, Shimotsuke-shi" *Computational Biology Branch*
5. Madrid, Spain "Aquintoviricetes; Archintovirales; Ascaris lumbricoides Tas virus; Asfivirus haemorrhagiae; Aveparvovirus anseriform1; Aveparvovirus avian1; Aveparvovirus galliform2; Aveparvovirus passeriform2; Aveparvovirus passeriform3; Aveparvovirus psittacine1; Aveparvovirus psittacine2; Aviadenovirus cerasi; Aviadenovirus orioli; Aviadenovirus oti; Aviadenovirus phalacrocoracidae; Aviadenovirus roseae; Ayintorquevirus; Ayintorquevirus ursid28; Bamfordvirae; Barthadenovirus gerygones; Barthadenovirus varani; Barthadenovirus zootherae; Barthadenovirus zootocae; Belvinaviricetes; Betabaculovirus psincretae; Betafilamentovirus; Betafilamentovirus altercocongregatae; Betafilamentovirus cocongregatae; Betapolyomavirus hipposideri; Blattambidensovirus incertum5; Bocaparvovirus chiropteran6; Bocaparvovirus incertum2; Bocaparvovirus incertum3; Bocaparvovirus incertum4; Bocaparvovirus ungulate10; Bombyx mori Pao virus; Caenorhabditis elegans Cer13 virus; Cardeaviricetes; Circovirus baizhenhe; Circovirus dever; Circovirus patkany; Circovirus python"
6. Adriaenssens, Alfenas-Zerbini, Aylward et al. (2023) "Funding information This work was supported in part through Laulima Government Solutions, LLC prime contract with the U.S. National Institute of Allergy and Infectious Diseases under Contract No. HHSN272201800013C. J.H.K. performed this work as an employee of Tunnell Government Services, a subcontractor of Laulima Government Solutions, LLC, under Contract No. HHSN272201800013C. The views and conclusions contained in this document are those of the authors and should not be interpreted as necessarily representing the official policies, either expressed or implied" *of the U.S. Department of References*
7. (2020) "International Committee on Taxonomy of Viruses Executive Committee. The new scope of virus taxonomy: partitioning the virosphere into 15 hierarchical ranks" *Nat Microbiol*
8. Koonin, Dolja, Krupovic et al. (2020) "Global organization and proposed megataxonomy of the virus world" *Microbiol Mol Biol Rev*
9. Simmonds, Adams, Benkő et al. (2017) "Consensus statement: virus taxonomy in the age of metagenomics" *Nat Rev Microbiol*
10. Simmonds, Adriaenssens, Zerbini et al. (2023) "Four principles to establish a universal virus taxonomy" *PLoS Biol*
11. Siddell, Walker, Lefkowitz et al. (2020) "Binomial nomenclature for virus species: a consultation" *Arch Virol*
12. 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" *Arch Virol*
13. Postler, Rubino, Adriaenssens et al. (2022) "Guidance for creating individual and batch latinized binomial virus species names" *J Gen Virol*
14. Koonin, Fischer, Kuhn et al. (2024) "The polinton-like supergroup of viruses: evolution, molecular biology, and taxonomy" *Microbiol Mol Biol Rev*
15. Krupovic, Makarova, Koonin (2022) "Cellular homologs of the double jelly-roll major capsid proteins clarify the origins of an ancient virus kingdom" *Proc Natl Acad Sci U S A*
16. Krupovic, Kuhn, Fischer et al. (2024) "Natural history of eukaryotic DNA viruses with double jelly-roll major capsid proteins" *Proc Natl Acad Sci U S A*
17. Redrejo-Rodríguez, Ordóñez, Berjón-Otero et al. (2017) "Primer-independent DNA synthesis by a family B DNA polymerase from self-replicating mobile genetic elements" *Cell Rep*
18. Benkő, Aoki, Arnberg et al. (2022) "ICTV Virus Taxonomy Profile: Adenoviridae 2022" *J Gen Virol*
19. Starrett, Tisza, Welch et al. (2021) "Adintoviruses: a proposed animal-tropic family of midsize eukaryotic linear dsDNA (MELD) viruses" *Virus Evol*
20. De, Boratto, Oliveira et al. (2022) "Yaraviridae": a proposed new family of viruses infecting Acanthamoeba castellanii" *Arch Virol*
21. Yutin, Mutz, Krupovic et al. (2024) "Mriyaviruses: small relatives of giant viruses" *mBio*
22. Bellas, Hackl, Plakolb et al. (2023) "Largescale invasion of unicellular eukaryotic genomes by integrating DNA viruses" *Proc Natl Acad Sci U S A*
23. Yutin, Shevchenko, Kapitonov et al. (2015) "A novel group of diverse Polinton-like viruses discovered by metagenome analysis" *BMC Biol*
24. Alonso, Borca, Dixon et al. (2018) "ICTV Virus Taxonomy Profile: Asfarviridae" *J Gen Virol*
25. Varsani, Kraberger, Opriessnig et al. (2023) "Anelloviridae taxonomy update 2023" *Arch Virol*
26. Varsani, Opriessnig, Celer et al. (2021) "Taxonomic update for mammalian anelloviruses (family Anelloviridae)" *Arch Virol*
27. Butkovic, Kraberger, Smeele et al. (2023) "Evolution of anelloviruses from a circovirus-like ancestor through gradual augmentation of the jelly-roll capsid protein" *Virus Evol*
28. Koch, Krupovic, Fielding et al. (2025) "Novel lineage of anelloviruses with large genomes identified in dolphins" *J Virol*
29. Kraberger, Serieys, Richet et al. (2021) "Complex evolutionary history of felid anelloviruses" *Virology*
30. Liu, Sun, Tu et al. (2022) "Virome profiling of an Amur leopard cat reveals multiple anelloviruses and a bocaparvovirus" *Vet Sci*
31. Zhao, Yue, Yang et al. (2022) "Viral metagenomics unveiled extensive communications of viruses within giant pandas and their associated organisms in the same ecosystem" *Sci Total Environ*
32. Goldberg, Clyde, Gendron-Fitzpatrick et al. (2018) "Severe neurologic disease and chick mortality in crested screamers (Chauna torquata) infected with a novel gyrovirus" *Virology*
33. Cibulski, Weber, De et al. (2020) "Viral metagenomics in Brazilian Pekin ducks identifies two gyrovirus, including a new species, and the potentially pathogenic duck circovirus" *Virology*
34. Cotmore, Agbandje-Mckenna, Canuti et al. (2019) "ICTV Virus Taxonomy Profile: Parvoviridae" *J Gen Virol*
35. Moens, Calvignac-Spencer, Lauber et al. (2017) "ICTV Virus Taxonomy Profile: Polyomaviridae" *J Gen Virol*
36. Breitbart, Delwart, Rosario et al. (2017) "ICTV Virus Taxonomy Profile: Circoviridae" *J Gen Virol*
37. Gauthier, Boulain, Van Vugt et al. (2021) "Chromosomal scale assembly of parasitic wasp genome reveals symbiotic virus colonization" *Commun Biol*
38. Guinet, Leobold, Herniou et al. (2024) "A novel and diverse family of filamentous DNA viruses associated with parasitic wasps" *Virus Evol*
39. Lepetit, Gillet, Hughes et al. (2016) "Genome sequencing of the behavior manipulating virus LbFV reveals a possible new virus family" *Genome Biol Evol*
40. Wallace, Coffman, Gilbert et al. (2021) "The discovery, distribution, and diversity of DNA viruses associated with Drosophila melanogaster in Europe" *Virus Evol*
41. Inwood, Skelly, Guhlin et al. (2023) "Chromosome-level genome assemblies of two parasitoid biocontrol wasps reveal the parthenogenesis mechanism and an associated novel virus" *BMC Genom*
42. Varaldi, Fouillet, Ravallec et al. (1930) "Developments in the classification and nomenclature of arthropod-infecting large DNA viruses that contain pif genes" *Arch Virol*
43. Harrison, Herniou, Jehle et al. (2018) "ICTV Virus Taxonomy Profile: Baculoviridae" *J Gen Virol*
44. Soriano, Krupovic, Llorens (2021) "ICTV Virus Taxonomy Profile: Belpaoviridae 2021" *J Gen Virol*
45. Zerbini, Siddell, Lefkowitz et al. (2023) "Changes to virus taxonomy and the ICTV Statutes ratified by the International Committee on Taxonomy of Viruses" *Arch Virol*
46. Rosario, Breitbart, Harrach et al. (2017) "Revisiting the taxonomy of the family Circoviridae: establishment of the genus Cyclovirus and removal of the genus Gyrovirus" *Arch Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12628552&blobtype=pdf | # Cancer Cell International †
Xiaochuan Wang, Tingrui Li, Chunguang Li, Yichao Jin, Jingjing Chen, Xinyang, Zhen Guan, Mei Jin, Jingxian Zhang, Liangheng Xu, Sizhen Tao, Chunping Ao
## Abstract
Background Human papillomavirus (HPV) infection is associated with an increased risk of cutaneous squamous cell carcinoma (CSCC). A comprehensive understanding of the cellular heterogeneity of HPV-positive and -negative CSCC is crucial for improving diagnosis and preventing tumor progression.
MethodsWe conducted an integrated analysis of single-cell RNA and spatial transcriptomic data from different skin tissue sources to map the cellular landscape of the tumor microenvironment (TME) in both HPV positive and negative CSCC. Results were validated through multiplex immunohistochemistry (mIHC) and in vitro experiments.
ResultsWe identified 10 major cell types in CSCC and normal skin samples, including epithelial cells, myeloid cells, T cells, fibroblasts, endothelial cells, B cells, smooth muscle cells, mast cells, melanocytes, and hair follicle cells. Notably, fibroblasts were found to be associated with tumor progression in CSCC with or without HPV infected. We further identified eight major CAF subtypes in CSCC, with iCAFs-CXCL2 promoting tumor progression, while iCAFs-PLA2G2A acted to suppress tumor growth. The MDK-ITGA6 pair was found to mediate interactions between fibroblasts and epithelial cells in CSCC. mIHC analysis confirmed elevated expression of MDK and ITGA6 in CSCC samples. Additionally, cell co-culture experiments confirmed that MDK-mediated CAFs were shown to enhance tumor cell migration and invasion in CSCC.
ConclusionOur findings provide a comprehensive cellular atlas of CSCC, highlighting the association of CAFs in HPV infection and tumor progression of CSCC. These results also offer potential diagnostic and prognostic biomarkers for CSCC patients.
## Introduction
Cutaneous squamous cell carcinoma (CSCC) is a malignant non-melanoma skin cancers (NMSCs) and is the second most common cancer [1,2]. CSCC arises from the malignant proliferation of keratinocytes and clinically presents as an indurated, crusted lesion [3]. Although most cases of CSCC can be effectively treated with complete excision, leading to favorable outcomes for majority of patients, its incidence continues to rise globally [4,5]. Notably, CSCC is characterized by high recurrence rates, perineural invasion, and locoregional metastasis, which contribute to its elevated mortality [6,7]. The risk factors for CSCC include advanced age, male gender, ultraviolet radiation (UVR), ionizing radiation, immunosuppressive agents and antifungal drugs used in organ transplant recipients (OTR), β-human papillomavirus (HPV) infection, smoking, genetic predispositions (such as fair skin and genetic syndromes), and immunosuppression [8,9].
Previous studies have proposed a "hit-and-run" theory, in which β-HPV facilitates the initiation of UV-driven CSCC but is absents in the tumor maintenance [10,11], suggesting a role of β-HPV in the early stages of CSCC development. A prospective observational study provides evidence that β-HPV infection is associated with the carcinogenesis of CSCC in organ transplant recipients (OTRs) [12], supporting the idea that immunosuppression increases the risk of β-HPV-associated CSCC [13]. Additionally, recent research has shown that β-HPV infection is also linked to an increased risk of CSCC in individuals without immunosuppression [14], indicating that β-HPV may involve not only immune suppression but also other regulatory mechanisms.
Recent advancements in high-throughput sequencing techniques, such as single-cell RNA sequencing (scRNA-seq) and spatial transcriptomic sequencing (stRNA-seq), have enabled detailed characterization of the cellular landscape and interactions within the tumor microenvironment (TME) of HPV-positive and negative cancers [15][16][17]. Notably, scRNA-seq analysis has provided the first insights into the cellular heterogeneity of HPV-induced CSCC and paracancerous CSCC (pc-CSCC) tissues [18]. Furthermore, accumulating evidence from scRNA-seq and stRNA-seq studies underscores the pivotal roles of cancer-associated fibroblasts (CAFs) in HPV-related cancers. For instance, CAFs have been shown to drive malignant progression in HPV-associated head and neck squamous cell carcinoma (HNSCC) [19], and extracellular matrix-related myCAF signatures are associated with therapeutic resistance in HPV-HNSCC [20]. In addition, CAFs have been implicated in both protumorigenic and pro-inflammatory signaling in HPVassociated cervical cancer [21]. However, the cellular heterogeneity and characteristics of the TME in CSCC with and without HPV infection remain poorly understood, as do the specific roles of CAFs in HPV-associated CSCC.
In this study, we conducted a comprehensive analysis of scRNA-seq data from both HPV + and HPV-CSCC skin samples, along with normal skin samples, to characterize the cellular landscape and interactions within the tumor microenvironment (TME). Additionally, we combined scRNA-seq data with experimental validation to confirm key cellular interactions in HPV-CSCC.
## Materials and methods
## Data collection
The gene expression profiles of 9 CSCC and 7 unmatched healthy controls were from the Gene Expression Omnibus (GEO) under the accession GSE139505. The scRNAseq datasets included GSE144236 (10 CSCC skin samples and 10 matched normal skin samples) [22], GSE218170 (5 CSCC skin samples) [23], GSE262947 (1 HPV-induced CSCC skin sample from a patient diagnosed via pathological biopsy as HPV16/18-negative and HPV6/11-positive, along with 1 paracancerous CSCC skin samples, pc-CSCC) [18], GSE193304 (3 CSCC skin samples and 1 carcinoma in situ, SCCIS) [24], and the spital transcriptomics dataset GSE144239 (4 CSCC skin samples) [22], all downloaded from the GEO database.
## Single-cell RNA-sequencing (scRNA-seq) data processing and analysis
In the present study, scRNA-seq data processing and analysis were conducted using "Seurat" version 4.4.0 in R. Low-quality cells filtered out based on the following criteria: cells with less than 300 or over 7500 expressed genes (nFeature_RNA >200 and nFeature_RNA < 7500), those with more than 30% mitochondrial transcripts, or more than 50% mitochondrial transcripts for HPVinfected samples. After filtering, a total of 169,421 cells were retained for further analysis. All scRNA-seq data were merged, then normalized and scaled using the "NormalizeData" and "Scaledata" functions. Batch effects across samples were corrected using the "harmony" package [25], after which the normalized scRNA-seq data were transformed into Seurat objects. Subsequently, "FindVariableFeatures" function was conducted to select the top 2000 highly variable genes (HVGs). Principal component analysis (PCA) was used to reduce the dimensionality for scRNA-seq data using the "RunPCA" function, and the significant principal components (PCs) were identified using "ElbowPlot" function. The appropriate PCs for cell clustering were selected based on the proportion of standard variance. Clusters were determined using the "FindClusters" function with a resolution = 0.5. Marker genes for each cluster were identified using the "FindMarkers" function. Cell types were annotated according to the gene markers from the Cellmarker 2.0 [26] and Annotation of Cell Types (ACT, ( h t t p : / / x t e a m . x b i o . t o p / A C T / or h t t p : / / b i o c c . h r b m u . e d u . c n / A C T /) [27]. Finally, cell visualization was conducted using the uniform manifold approximation and projection (UMAP). The subclusters of cancer associated fibroblasts (CAFs) also identified as previous cell annotation methods.
## Pathway enrichment analysis
Differentially expressed genes (DEGs) were selected in each subcluster using the "FindAllMarkers" function with the following parameters: only.pos = TRUE, min. pct = 0.25, and logfc.threshold = 0.25. Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment was archived using "Clusterprofiler" package, with significant pathways were identified at an adjusted p-value (adj. P) < 0.05.
Additionally, the pathway activity was inferred using the PROGENy method, which evaluates the scores of gene sets comprising the top 100 most responsive genes associated with 14 signaling pathways(Androgen, Estrogen, EGFR, Hypoxia, JAK-STAT, MAPK, NFκB, PI3K, p53, TGFβ, TNFα, TRAIL, VEGF, and WNT) [28]. The analysis was performed using the "progeny" R package version 1.22.0.
The cytokine signaling activity also was assessed using Cytokine Signaling Analyzer (CytoSig; h t t p s : / / c y t o s i g . c c r . c a n c e r . g o v /), which provides target genes of cytokines and predicative mode of cytokine signaling cascades based on transcriptomic profiles [29]. The Z-scores for gene sets were obtained from the Cytokine Signaling Analyzer.
## Cell trajectory and C
Cell trajectory analysis of CAFs in CSCC was performed using "Monocle" version 2.24.0 [30]. And the differentiation states of CAFs in CSCC was predicted using Cellular (Cyto) Trajectory Reconstruction Analysis using gene Counts and Expression (CytoTRACE) [31]. Additionally, the gene expression patterns implicated in cell development along the cell trajectory were identified using the branch expression analysis modeling (BEAM) algorithm. Furthermore, we also visualized the expression dynamics of the branch-specific genes along the differentiation trajectory.
## Cell-to-cell interaction analysis
Cell-to-cell communication and ligand-receptor networks were constructed using "CellChat" package version 2.1.2 based on the "cellchatDB'' ligand-receptor database, which included paracrine/autocrine signaling interactions, extracellular matrix (ECM)-receptor interactions, and cell-cell contact interactions [32,33]. With the default parameters of cellchatDB, the signal strength and the relative weight of each ligand-receptor pair within each signal were calculated. The significant ligand-receptor pairs with a p value < 0.05 and a mean value ≥ 1 were identified for subsequently analysis.
## Spatial transcriptome RNA sequencing (stRNA-seq) data analysis
StRNA-seq data for GSM4565823, GSM4565824, GSM4565825, and GSM4565826 from GSE144239 dataset were analyzed using "Seurat" package. Quality control for stRNA-seq was conducted according to the following parameters, genes with fewer than 10 read counts or expressed in fewer than two spots were removed, and spots with the gene counts less than 200 genes were filtered [22]. As with scRNA-seq analysis, the stRNA-seq data were normalized and scaled using the "Normalize-Data" and "Scaledata" functions. Top 4000 HVGs were identified using "FindVariableFeatures" function. Batch effects across samples were corrected using "harmony" package. Dimensionality reduction was performed by PCA using the "RunPCA" function, and the significant PCs were identified using "ElbowPlot" function. Clusters were determined using the "FindClusters" function with a resolution = 0.5. These clusters were then mapped to the histological sections. Furthermore, the scRNA-seq and stRNA-seq data were integrated by mapping using the "FindTransferAnchors" function. Additionally, cellto-cell communication analysis for stRNA-seq data were performed using "NICHES" package, version 1.1.0 [34].
## Human tissue microarray
A human skin cancer tissue microarray (HSkiC060pt01) was obtained from the Shanghai Outdo Biotech Co., Ltd. The skin cancer tissue microarray contains 43 CSCC tumor tissues, 7 adjacent normal tissues, 8 basal cell carcinoma (BCC), and 2 non-tumorous normal skin tissues. The experimental protocol has been approved by ethics committee of Shanghai Outdo Biotech Co., Ltd (Ethics Code: YB M-05-02).
## Multiplex immunohistochemistry (mIHC) analysis
The histological diagnosis of each sample was reconfirmed using hematoxylin and eosin (H&E) staining. Quantitative multiplex immunofluorescence was conducted to assess the expression of MDK and ITGA6 in human skin tumor tissues using the Opal Multiplex Immunostaining Kit (PerkinElmer; Waltham, MA) according to the manufacturer's protocol.
Briefly, paraffin-embedded tissue microarray sections were deparaffinized with xylenes and graded ethanol, followed by antigen retrieval in sodium citrate solution and blocking with 3% bovine serum albumin (BSA). The sections were then incubated overnight at 4 °C with primary antibodies against MDK (ProMab, 1:500, P34640) and ITGA6 (ProMab, 1:100, P30566). Following primary antibody incubation, the sections were incubated with goat anti-mouse/rabbit-peroxidase-conjugated secondary antibody at 37 °C for 1 h. Tyramide signal amplification (TSA) was applied, with antigen retrieval and antibody stripping performed after each TSA cycle. Finally, nuclei were stained with DAPI (Beyotime, C1006). The stained sections were scanned using the Panoramic Scan (MIDI II, 3D Histech Ltd., Budapest, Hungary). Integrated optical density of the single or double-stained proteins was semi-quantitatively analyzed using Aipthwell software (Servicebio, Wuhan, China) [35].
## Clinical specimens
Four tumor tissues were obtained from CSCC patients who underwent surgery at the First People's Hospital of Yunnan Province. This study protocol was approved by the Ethics Committee of the First People's Hospital of Yunnan Province (Ethics Code: KHLL2024-KY292), and written informed consent was obtained from all participants prior to inclusion in the study. The tissues were either immediately processed for subsequent analysis or rapidly frozen in liquid nitrogen and stored at -80 °C.
## Cell lines and cell culture
Fresh human CSCC tumor tissues were obtained, cut into 1 mm³ pieces, and suspended in DMEM/F12 medium (Gibco, NY, USA) containing type I collagenase (Gibco, NY, USA). The suspension was then subjected to a shaking water bath at 37 °C for 2 h. Following this, the cell suspension was filtered through a 70 μm filter and centrifuged at 1500 rpm for 5 min. The resulting pellet was collected and resuspended in DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS, Gibco, NY, USA) and 1% penicillin-streptomycin (Gibco, NY, USA).
The human CSCC cell line SCL1 was obtained from the American Type Culture Collection (ATCC, VA, USA) and cultured in RPMI 1640 medium (Servicebio, China) supplemented with 10% FBS at 37 °C in a 5% CO₂ incubator.
## Cell transfection
CAFs were transfected in 6-well plates with MDK-specific siRNA using Lipofectamine 3000 (Invitrogen, CA, USA) according to the manufacturer's instructions. The expression of MDK was examined 48 h after transfection using qRT-PCR.
## Cell co-culture
CAFs and SCL1 cells were co-cultured using Transwell with a 0.4 μm pore size. Briefly, CAF were plated in 6-well plates at concertation of 2 × 10 5 cells/well, and SCL1 cells were plated in the upper chamber at a concertation of 3 × 10 5 cells/well. Afte co-cultured 72 h, the SCL1 cells were collected for subsequent experimental analysis.
## Quantitative PCR
Total RNA was extracted from the collected SCL1 cells using TRIzol reagent (Lifetech, China). Complementary DNAs (cDNAs) were synthesized using FastKing RT Kit (With gDNase) FastKing cDNA (TIANGEN, China) according to the manufacturer's instruction. The cDNA samples were amplified using Taq Pro Universal SYBR qPCR Master Mix (Vazyme, China), according to the manufacture's instruction. PCR amplification program was set as the following: 95˚C for 10 min, followed by 40 cycles of 95˚C for 15 s and 60˚C for 30 s, followed by incubation at 95 °C for 15 s. The primers used were as follows: MDK forward, 5
## '-T A C A A T G C T C A G T G C C A G-3' and reverse, 5'-T C C T T T C C C T T C C C T T T C-3' . ITGA6 forward, 5'-C T G A T G T T G C T G T T G G T T-3' and reverse, 5'-T T A G G A G T T A C T G T G A T G G T-3' . GAPDH forward, 5'-T T G C C C T C A A C G A C C A C T T T-3' and reverse, 5'-T G G T C C A G G G G T C T T A C T C C-3' . GAPDH was used for normalization the expression levels of MDK and ITGA6.
Relative expression was quantified by the 2 -ΔΔCT method.
## Western blot analysis
The collected SCL1 cells were lysed in RIPA buffer (Beyotime, China). The concentration and purity of the protein were measured using the NanoDrop system. Protein samples of 40 µg per lane were separated on SDS-PAGE gels (10%) and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, MA, USA). The membranes were blocked for 2 h in 5% BSA at room temperature and subsequently incubated with primary antibodies against MDK (Proteintech, 11009-1-AP, 1:1000), ITGA6 (Affinity, DF8323, 1:3000), and β-actin (ZSGB-BIO, TA-09, 1:4000) overnight at 4˚C. The membranes were then washed with TBST three times, incubated with horseradish peroxidase-labeled secondary antibodies (Abmart, M21001L and M21002L, 1:4000) for 2 h at room temperature, and further washed with TBST three times. Subsequently, the blots were detected using the enhanced chemiluminescence kit (Millipore, MA, USA) and analyzed using ImageJ software (National Institutes of Health, MD, USA).
## Wound healing assay
The collected SCL1 cells were plated in 6-well plates at a concentration of 5 × 10 4 cells/well and cultured until converged as a monolayer. Subsequently, a wound was scratched using a culture-insert 4-well µ-Dish 35 mm (IBIDI 80466). Images of each well at 0 and 24 h after scratching were taken under an inverted micro-scope (Olympus, Tokyo, Japan). The wound healing areas were calculated using the ImageJ software (National Institutes of Health, MD, USA).
## Transwell assay
The collected SCL1 cells were plated in the upper chamber with an 8 μm pore size pre-coated with Matrigel matrix gel (Corning, NY, USA) in serum-free medium, add full culture medium to the lower chamber. The cells were incubated at 37 °C, 5% CO 2 for 48 h. The invaded cells were fixed in 4% paraformaldehyde (PFA) and then stained with 1% crystal violet for 15 min. The counted under an inverted micro-scope (Olympus, Tokyo, Japan).
## Statistical analysis
Experimental results are presented as the means ± standard deviation (SD) from at least 3 independent experiments. Images in this study are quantified by Image J (Version 1.8.0, SA. USA). Statistical analyses were performed for all experiments with the GraphPad Prism Software (Version 8.3.0, CA, USA). The statistical differences were calculated by one-way ANOVA analysis of variance with Tukey's multiple comparisons test. P < 0.05 was considered to indicate a statistically significant difference.
## Results
## Identification of the distinct cell types in CSCC
To elucidate the landscape of the tumor microenvironment (TME) in CSCC with or without HPV infection, we collected the scRNA-seq expression profiles of 4 datasets and the spital transcriptomic RNA-seq data. The analysis process is illustrated in Fig. 1A.
The scRNA-seq expression profiles used in this study included 18 CSCC skin samples, 10 normal tissues, 1 SCCIS skin sample, 1 HPV + CSCC and 1 pc-CSCC skin samples. After quality control and dimensionality reduction, a total of 169,421 cells were incorporated for further analyses (Figure S1A-D). Additionally, Harmony algorithm was employed to correct the batch effects between different samples (Figure S2A). Top 2000 HVGs were selected, and these cells were distributed into 11 distinct clusters (Figure S2B-E), which were then identified as 10 major cell types based on classical markers (Fig. 1B,C, Figure S3), including epithelial cells (KRT1 and KRT10), myeloid cells (HLA-DRA and LYZ), T cells (CD2 and CD3D), fibroblasts (COL1A1 and DCN), endothelial cells (VWF and RAMP2), B cells (MZB1 and CD79A), smooth muscle cells (TAGLN and ACTA2), mast cells (CPA3 and TPSAB1), melanocytes (DCT and PMEL), and hair follicle cells (GJB2 and GJB6). Although all cell types were presented in all tissues, the distinction in cell proportion was observed. As shown in Fig. 1D,E, the proportions of epithelial cells and fibroblasts were significantly changed in different originated tissues. Remarkably, the proportion of epithelial cells was significantly reduced, while the proportion of fibroblasts appeared to be increased in HPV-infected skin tissues compared to non-HPV-infected skin tissues. Furthermore, compared to both HPV-negative tumor tissues and normal nontumor tissues, as well as HPV-positive tumor tissues and their corresponding non-tumor tissues, the proportion of fibroblasts was significantly higher in tumor tissues. These findings suggest that the distinct cellular composition of CSCC tissues and indicate that fibroblast enrichment is a characteristic feature of the CSCC tumor microenvironment.
## Identification of eight CAF subtypes in CSCC
We further deeply explored the heterogeneity of fibroblasts in tumor progression of CSCC. We extracted the fibroblasts from all tissues to identify the subtypes of fibroblasts (Figure S4A-B). Top 2000 HVGs were selected and fibroblasts were categorized into 8 distinct clusters (Figure S4D). These 8 major CAF subtypes were identified (Fig. 2A, B, and2D, and Figure S5), including contractile matrix CAFs (mCAFs) (SPON2 and MMP11), extracellular matrix-CAFs (ECM-CAFs) (MFAP5 and TNN), vascular CAFs (vCAFs) (MCAM and ACTA2), dividing CAFs (dCAFs) (MKI67 and TUBA1B), immunomodulatory CAFs (iCAFs)-CXCL2 (IL6 and CXCL2), iCAFs-PLA2G2A (C3 and PLA2G2A), antigen presenting (apCAFs) (HLA-DRA and CD74), and epithelialmesenchymal transition-like CAFs (EMT-like CAFs) (COMP and TNC). Notably, iCAFs-PLA2G2A were significantly enriched, while iCAFs-CXCL2 were reduced in HPV-negative skin tissues compared to HPV-positive skin tissues (Fig. 2C). iCAFs-PLA2G2A were more abundant in normal skin tissues than in tumor tissues, while the proportion of iCAFs-CXCL2 was higher in tumor tissues compared to normal skin tissues. These results suggest that iCAFs-PLA2G2A and iCAFs-CXCL2 play opposing roles in tumor progression. Specifically, iCAFs-PLA2G2A may have a potential tumor-suppressive effect, whereas iCAFs-CXCL2 may promote tumor progression.
## Pathway enrichment analysis of CAF subtypes in CSCC
Next, we explored the biological functions of the CAF subtypes through pathway enrichment analysis. KEGG pathway results indicated that ECM-receptor interaction was significantly enriched in both mCAFs and EMT-like CAFs. And protein digestion and absorption were significantly enriched in mCAFs. Oxidative phosphorylation was significantly enriched in EMC-CAFs, vCAFs, and EMT-like CAFs. The cell cycle pathway was enriched in dCAFs, while Kaposi sarcoma-associated herpesvirus infection, IL-17 and TNF signaling pathways were enriched in iCAFs-CXCL2. iCAFs-PLA2G2A were primarily involved in complement and coagulation cascades, and apCAFs were mainly related to antigen processing and presentation and phagosome (Fig. 2E,F). Furthermore, the activity of the EGFR and MAPK pathways was consistently observed, alongside the significant activation of the TNF-α pathway in iCAFs-CXCL2 (Fig. 2G). Cyto-Sig tool analysis revealed a significant upregulation of the NO cytokine signature in iCAFs-CXCL2, as well as an upregulation of OSM and MCSF in iCAFs-PLA2G2A (Fig. 2H). These findings underscore the potential role of CAFs in the tumorigenesis of CSCC, particularly highlighting the contributions of iCAFs-CXCL2 and iCAFs-PLA2G2A.
## Cell trajectory of different CAF subtypes
To further investigate the differentiation, developmental trajectories, and dynamics of CAFs, we performed cell trajectory analysis using the Monocle 2 and CytoTRACE algorithms. As shown in Fig. 3A, a total of three distinct cell states were identified in CAFs. iCAFs-CXCL2 and iCAFs-PLA2G2A were assigned to different branches (Fig. 3B).
Using the CytoTRACE algorithm, we assessed the developmental trajectories of CAFs. dCAFs, vCAFs, and iCAFs-CXCL2 exhibited higher CytoTRACE scores, indicating their early differentiation stages, whereas iCAFs-PLA2G2A and mCAFs displayed lower CytoTRACE scores, suggesting their terminal differentiation stages (Fig. 3C,D). Furthermore, we inferred that dCAFs and vCAFs represent the initiation state, evolving into iCAFs-CXCL2 (Fig. 3E-F).
Additionally, we examined gene expression patterns associated with the two differentiation branches linked to iCAFs-CXCL2 and iCAFs-PLA2G2A transitions (Fig. 3G,H). Branch 1, representing the initiation state (pre-branch), exhibited distinct gene expression patterns in cell fate 1 (iCAFs-CXCL2) and cell fate 2 (iCAFs-PLA2G2A) (Fig. 3G). As shown in Fig. 3H, in the early differentiation branch of iCAFs-CXCL2, higher expression levels of GNB1, PTP4A2 (cluster 5), and ID3 (cluster 1) were observed. In contrast, the early differentiation branch of iCAFs-PLA2G2A showed elevated expression of WASF2 and GRP153 (cluster 3).
We also investigated differentially expressed genes along both branches (Fig. 3I). We observed that IFI6 expression was higher in the early stages of iCAFs-CXCL2 differentiation compared to iCAFs-PLA2G2A and gradually increased in both branches. Additionally, ISG15 expression was more prominent in the interim stage of iCAFs-CXCL2 compared to iCAFs-PLA2G2A and also increased along both branches. These results suggest that these genes play a significant role in the progression of CSCC.
## Cell communication analysis in HPV-positive and HPVnegative tumor tissues in CSCC
We next investigated the cell communication between the 10 major cell types in the tumor microenvironment (TME) of CSCC, with or without HPV infection. All samples were divided into three groups: normal (10 normal skin samples), HPV-positive CSCC (1 HPV + CSCC skin sample), and HPV-negative CSCC (1 SCCIS and 18 CSCC skin samples). Compared to the normal group, CSCC groups exhibited increased numbers of interactions and stronger interaction strengths (Fig. 4A-B). Notably, the HPV-negative CSCC samples had more interactions than the HPV-positive sample, while the interaction strength was lower in the HPV-negative CSCC samples compared to the HPV-positive sample.
Remarkably, interactions among fibroblasts, smooth muscle cells, endothelial cells, hair follicle cells, and epithelial cells were stronger in CSCC groups (Fig. 4A,B). Fibroblasts emitted the strongest signals, while epithelial cells received more signals than other cell types (Figure S6A,B). To further explore the relative information flow of ligand-receptor (L-R) interactions, we observed stronger signaling in the MK, CD99, FN1, collagen, and laminin pathways (Fig. 3C). Interestingly, the MK and CD99 pathways had the highest proportion of interactions in normal samples, followed by HPV-negative CSCC samples, with the lowest in HPV-positive sample. Conversely, the FN1, collagen, and laminin pathways showed the opposite trend. The MK pathway was identified as the primary pathway involved in communication between fibroblasts and epithelial cells (Fig. 3D).
We also identified MDK-NCL and MDK-ITGA6 as the top L-R pairs involved in this communication (Fig. 3E,F). Upon examining the expression of L-R pairs in the MK pathways in the GSE139505 dataset, we found higher expression of MDK and ITGA6 genes in CSCC tumor samples compared to normal samples (Figure S6C). These findings prompted us to focus on the role of these L-R pairs in the progression of CSCC.
## Cell communication in Spatial tissues in CSCC
We further explored cell communication in spatial regions using NICHES. Similar to the scRNA-seq data analysis, after quality control, a total of 7,902 spots were included for subsequent analysis. The Harmony algorithm was applied to correct batch effects across different samples (Fig. 7A). Top 4000 HVGs were selected, and the spots were categorized into 17 distinct clusters (Figure S7B-C). By integrating scRNA-seq and spatial transcriptomic (stRNA-seq) data, we mapped 9 major cell types to each CSCC tissue slice (Fig. 5A-B). Epithelial cells were identified as the predominant cell type, with fibroblasts and B cells also prominently distributed across the CSCC tissue slices (Figure S7D). Subsequently, we identified differentially expressed L-R pairs in the spatial niches, with the MDK-ITGA6 pair being prominently expressed (Fig. 5C). These results suggest that the MDK-ITGA6 pair plays a significant role in cell communication and tumor progression in CSCC.
Additionally, mIHC staining was employed to explore the relationship between MDK and ITGA6 in CSCC tumor tissues (Fig. 6A,B). We observed an increased number of MDK-and ITGA6-positive cells in CSCC tumor tissues compared to normal tissues (Fig. 6A andD-E, and6G, H). Notably, we also observed a clear colocalization of MDK and ITGA6 in CSCC tumor tissues, in contrast to normal tissues (Fig. 6F andI). These findings indicate the co-expression of MDK and ITGA6 in CSCC.
## MDK mediated CAF promotes the tumor progression of CSCC in vitro
To investigate whether MDK-mediated CAFs influence CSCC progression, we employed a cell co-culture system (Fig. 7A). qPCR and western blot analyses revealed that silencing MDK expression in CAFs, followed by co-culture with SCL-1 cells (Fig. 7B), significantly inhibited both the mRNA (Fig. 7C,D) and protein expression levels of MDK and ITGA6 in SCL-1 cells (Fig. 7E-G). Moreover, CAFs co-cultured with SCL-1 cells promoted migration and invasion of the SCL-1 cells, while CAFs treated with MDK-siRNA and co-cultured with SCL-1 cells blocked these processes (Fig. 7H-K). These findings suggest that MDK-mediated CAFs promote tumor cell migration and invasion in CSCC.
## Discussion
Although HPV infection is known to be associated with an increased risk of CSCC, however, the association between HPV infection and CSCC progression remains unclear. In our study, we integrated publicly available scRNA-seq and stRNA-seq data from CSCC patients with or without HPV infection to investigate cell heterogeneity and their interactions. These findings were further validated using mIHC staining and in vitro experiments. Initially, we identified 10 major cell types in both CSCC and normal skin samples, including epithelial cells, myeloid cells, T cells, CAFs, endothelial cells, B cells, smooth muscle cells, mast cells, melanocytes, and hair follicle cells. Notably, we observed a reduction in epithelial cells and an increase in fibroblasts in HPV-positive skin tissues compared to HPV-negative skin tissues. Additionally, the number of fibroblasts was higher in tumor samples than in non-tumor normal samples, with this trend remaining consistent in both HPV-positive and HPV-negative tissues. These findings underscore the role of CAFs in CSCC, irrespective of HPV status. Our results are in agreement with a previous study that reported an increased number of CAFs in HPV-positive CSCC tissues compared to matched normal tissues and normal skin samples [18].
CAFs represent a major population within the TME of cancers, its plays key roles in CSCC initiation, development, and metastasis [36]. Their heterogeneity and plasticity confer diverse functions [37], and CAFs further drive oncogenic and malignant progression [19,38]. In this study, we identified eight CAF subtypes, including mCAFs, ECM-CAFs, vCAFs, dCAFs, iCAFs-CXCL2, iCAFs-PLA2G2A, apCAFs, and EMT-like CAFs. Notably, iCAFs-PLA2G2A were decreased, while iCAFs-CXCL2 were increased in HPV-positive versus HPV-negative skin, and in tumor versus normal skin, suggesting opposing roles, iCAFs-PLA2G2A are tumor-suppressive, whereas iCAFs-CXCL2 promote tumor progression and associate with HPV infection.
iCAFs exhibit a secretory phenotype, acting as tumor suppressors or promoters by secreting immunomodulatory signals to recruit and activate other cells [36]. In particular, iCAFs-CXCL2 has been linked to immune suppression, cachexia, and chemoresistance in solid tumors [39]. In our study, we investigated the functions of iCAFs-CXCL2 using KEGG, PROGENy, and CytoSig analyses, and found that IL-17, TNF-α, EGFR, MAPK, and NF-κB pathways were significantly activated in iCAFs-CXCL2. Pro-inflammatory cytokines such as IFN-γ, TNF-α, and IL-17 are elevated in HPV-infected breast cancer cells [40], and IL-17 induction may drive persistent inflammation during HPV infection [41]. In CSCC, IL-17 signaling promotes immune exclusion via collagen deposition programs [42], while aberrant EGFR [43,44], MAPK [44], NF-кB pathways contributes to tumor progression and immune suppression [45]. Targeting EGFR is a key systemic therapeutic strategies for CSCC [46]. Notably, BRAF inhibition (BRAFi) therapy, such as vemurafenib, can cooperates with HPV-mediated MAPK activation to tumorigenesis, with β-HPV-associated CSCC showing accelerated progression and frequent RAS, PIK3CA, CKIT, ALK, and EGFR mutations [47,48]. These findings emphasize the tumorpromoting role of iCAFs-CXCL2 in CSCC, particularly in HPV-infected cases.
Secretory calcium-dependent phospholipase A2 (PLA2G2A) acts as a ligand for integrins and plays a role in inflammation, host defense, and tissue regeneration by targeting extracellular phospholipids [49,50]. PLA2G2A exhibits dual roles in tumors, it has been associated with poor survival in pancreatic ductal adenocarcinoma (PDAC) [51] and a subset of PLA2G2A + metabolic cancer-associated fibroblasts (meCAFs) promotes tumor immune escape [52]. whereas in cholangiocarcinoma (CCA) it acts as a tumor suppressor by inhibiting proliferation, migration, and invasion [53]. In the present study, we identified iCAFs-PLA2G2A as tumor-suppressive in CSCC, with enrichment of complement and coagulation cascades, estrogen, OSM, and MCSF signaling pathways.
We analyzed cell trajectories to investigate CAF differentiation and dynamics. Our findings revealed that dCAFs and vCAFs represent the initiation state, while iCAFs-CXCL2, iCAFs-PLA2G2A, and mCAFs represent the terminal stages of differentiation. Using scRNA-seq and stRNA-seq data, we examined CAF interactions within the TME and found increased numbers and strengths of interactions in CSCC compared to normal tissue. Between HPV-positive and HPV-negative CSCC, interaction numbers were higher in HPV-positive cases, while interaction strength was reduced in HPV-negative CSCC. Notably, in both HPV-positive and HPV-negative CSCC samples, MDK-ITGA6 mediated strong interactions between fibroblasts and epithelial cells. Midkine (MDK) is a heparin-binding growth factor that plays a pivotal role in tumor cell growth, metastasis, invasion, angiogenesis, poor survival, and drug resistance [54,55]. MDK maintains immune suppression in aggressive melanomas [56], and CAF-derived MDK contributes to drug resistance in gastric cancer [57] and oral squamous cell carcinoma (OSCC) [58]. High expression of MDK in fibroblasts also enhances intercellular interactions in primary CSCC [59].
Our study highlights the role of CAF-derived MDK in mediating intercellular communication in CSCC. Using mIHC staining and in vitro experiments, we confirmed that MDK-mediated CAFs promote tumor cell migration
## Conclusion
In conclusion, our study presents a comprehensive cell atlas of the TME in both HPV-positive and HPV-negative CSCC, offering in-depth insights into the key cell types involved in each subtype. Our findings highlight the mechanisms by which the MDK-ITGA6 pair mediates CAF interactions, contributing to tumor progression within the TME.
## References
1. Chang, Azin, Demehri (2022) "Cutaneous squamous cell carcinoma: frontier cancer immunoprevention" *Annu Rev Pathol*
2. Que, Zwald, Schmults (2018) "Cutaneous squamous cell carcinoma: incidence, risk factors, diagnosis, and staging" *J Am Acad Dermatol*
3. Hedberg (2022) "Molecular mechanisms of cutaneous squamous cell carcinoma" *Int J Mol Sci*
4. Leiter (2017) "Incidence, mortality, and trends of nonmelanoma skin cancer in Germany" *J Invest Dermatol*
5. Burton, Ashack, Khachemoune (2016) "Cutaneous squamous cell carcinoma: a review of high-risk and metastatic disease" *Am J Clin Dermatol*
6. Desai (2023) "Aggressive cutaneous squamous cell carcinoma of the head and neck: a review" *Curr Oncol*
7. Dessinioti, Pitoulias, Stratigos (2022) "Epidemiology of advanced cutaneous squamous cell carcinoma" *J Eur Acad Dermatol Venereol*
8. Waldman, Schmults (2019) "Cutaneous squamous cell carcinoma" *Hematol Oncol Clin North Am*
9. Mortaja, Demehri (2023) "Skin cancer prevention -recent advances and unmet challenges" *Cancer Lett*
10. Chockalingam, Downing, Tyring (2015) "Cutaneous squamous cell carcinomas in organ transplant recipients" *J Clin Med*
11. Hufbauer, Akgül (2017) "Molecular mechanisms of human papillomavirus induced skin carcinogenesis" *Viruses*
12. Bavinck (2018) "Human papillomavirus and posttransplantation cutaneous squamous cell carcinoma: a multicenter, prospective cohort study" *Am J Transplant*
13. Strickley (2019) "Immunity to commensal papillomaviruses protects against skin cancer" *Nature*
14. Rollison (2021) "Cutaneous human papillomaviruses and the risk of keratinocyte carcinomas" *Cancer Res*
15. Guo (2023) "Spatiotemporally deciphering the mysterious mechanism of persistent HPV-induced malignant transition and immune remodelling from HPV-infected normal cervix, precancer to cervical cancer: integrating single-cell RNA-sequencing and spatial transcriptome" *Clin Transl Med*
16. Lee (2024) "Deciphering head and neck cancer microenvironment: Singlecell and spatial transcriptomics reveals human papillomavirus-associated differences" *J Med Virol*
17. Cha (1136) "Single-cell analysis reveals cellular and molecular factors counteracting HPV-positive oropharyngeal cancer immunotherapy outcomes" *J Immunother Cancer*
18. Ding (2024) "Single cell transcriptomics reveals dysregulated immnue homeostasis in different stages in HPV-induced cutaneous squamous cell carcinoma" *Exp Dermatol*
19. Xue (2025) "Cancer-Associated Fibroblast-Driven progression of HPV + Head and neck cancer with organoid models" *J Dent Res*
20. Llerena (2025) "Extracellular matrix-myCAF signatures correlate with resistance to neoadjuvant aPD-L1 immune checkpoint Inhibition with durvalumab + metformin in HPV + HNSCC" *Clin Cancer Res*
21. Kumar (2016) "Role of estrogen receptor alpha in human cervical cancer-associated fibroblasts: a transcriptomic study" *Tumour Biol*
22. Ji (2020) "Multimodal analysis of composition and Spatial architecture in human squamous cell carcinoma" *Cell*
23. Schütz (2023) "Functionally distinct cancer-associated fibroblast subpopulations establish a tumor promoting environment in squamous cell carcinoma" *Nat Commun*
24. Zou (2023) "Single-cell sequencing highlights heterogeneity and malignant progression in actinic keratosis and cutaneous squamous cell carcinoma. Elife"
25. Cao (2023) "Single-cell RNA sequencing reveals the vascular smooth muscle cell phenotypic landscape in aortic aneurysm" *Cell Commun Signal*
26. Hu (2023) "CellMarker 2.0: an updated database of manually curated cell markers in human/mouse and web tools based on scRNA-seq data" *Nucleic Acids Res*
27. Quan (2023) "Annotation of cell types (ACT): a convenient web server for cell type annotation" *Genome Med*
28. Holland (2020) "Robustness and applicability of transcription factor and pathway analysis tools on single-cell RNA-seq data" *Genome Biol*
29. Jiang (2021) "Systematic investigation of cytokine signaling activity at the tissue and single-cell levels" *Nat Methods*
30. Qiu (2017) "Reversed graph embedding resolves complex single-cell trajectories" *Nat Methods*
31. Gulati (2020) "Single-cell transcriptional diversity is a hallmark of developmental potential" *Science*
32. Jin, Plikus, Nie (2025) "Cellchat for systematic analysis of cell-cell communication from single-cell transcriptomics" *Nat Protoc*
33. Jin (2021) "Inference and analysis of cell-cell communication using cellchat" *Nat Commun*
34. Raredon (2023) "Comprehensive visualization of cell-cell interactions in single-cell and spatial transcriptomics with NICHES" *Bioinformatics*
35. Mei (2022) "Integrin β3-PKM2 pathway-mediated aerobic glycolysis contributes to mechanical ventilation-induced pulmonary fibrosis" *Theranostics*
36. Van Hove, Hoste (2022) "Activation of fibroblasts in skin cancer" *J Invest Dermatol*
37. Knuutila (2024) "Cancer-associated fibroblast activation predicts progression, metastasis, and prognosis of cutaneous squamous cell carcinoma" *Int J Cancer*
38. James (2025) "Fibroblasts regulate the transcriptional signature of human papillomavirus-positive keratinocytes" *Tumour Virus Research*
39. Mao (2021) "Crosstalk between cancer-associated fibroblasts and immune cells in the tumor microenvironment: new findings and future perspectives" *Mol Cancer*
40. Santos (2024) "Immunological response against breast lineage cells transfected with human papillomavirus (HPV)"
41. Macinas (2025) "Comparative analysis of vaginal inflammatory cytokines among HPV-Positive and HPV-Negative women in Manila, Philippines" *Am J Reprod Immunol*
42. Chen (2022) "IL-17-induced HIF1α drives resistance to anti-PD-L1 via fibroblast-mediated immune exclusion" *J Exp Med*
43. Ascierto, Schadendorf (1136) "Update in the treatment of non-melanoma skin cancers: the use of PD-1 inhibitors in basal cell carcinoma and cutaneous squamous-cell carcinoma" *J Immunother Cancer*
44. Huang (2022) "LPCAT1 promotes cutaneous squamous cell carcinoma via EGFR-Mediated protein kinase B/p38MAPK signaling pathways" *J Invest Dermatol*
45. Cañueto (2017) "Epidermal growth factor receptor expression is associated with poor outcome in cutaneous squamous cell carcinoma" *Br J Dermatol*
46. Corchado-Cobos (2010) "Cutaneous squamous cell carcinoma: from biology to therapy" *Int J Mol Sci*
47. Holderfield (2014) "Vemurafenib cooperates with HPV to promote initiation of cutaneous tumors" *Cancer Res*
48. Cohen (2015) "Contribution of beta-HPV infection and UV damage to rapidonset cutaneous squamous cell carcinoma during BRAF-inhibition therapy" *Clin Cancer Res*
49. Dore, Boilard (2019) "Roles of secreted phospholipase a(2) group IIA in inflammation and host defense" *Biochimica et Biophysica Acta (BBA)*
50. Van Hensbergen (2020) "Type IIA secreted phospholipase A2 in host defense against bacterial infections" *Trends Immunol*
51. Zhang (2022) "PLA2G2A phospholipase promotes fatty acid synthesis and energy metabolism in pancreatic cancer cells with K-ras mutation" *Int J Mol Sci*
52. Ge (2023) "PLA2G2A(+) cancer-associated fibroblasts mediate pancreatic cancer immune escape via impeding antitumor immune response of CD8(+) cytotoxic T cells" *Cancer Lett*
53. Chen (2024) "STUB1-mediated K63-linked ubiquitination of UHRF1 promotes the progression of cholangiocarcinoma by maintaining DNA hypermethylation of PLA2G2A" *J Exp Clin Cancer Res*
54. Filippou, Karagiannis, Midkine (2020) "MDK) growth factor: a key player in cancer progression and a promising therapeutic target. Oncogene"
55. Saikia (2010) "Role of midkine in cancer drug resistance: regulators of its expression and its molecular targeting" *Int J Mol Sci*
56. Cerezo-Wallis (2020) "Midkine rewires the melanoma microenvironment toward a tolerogenic and immune-resistant state" *Nat Med*
57. Yang (2022) "CAF-derived midkine promotes EMT and cisplatin resistance by upregulating LncRNA ST7-AS1 in gastric cancer" *Mol Cell Biochem*
58. Zhang (2017) "Midkine derived from cancer-associated fibroblasts promotes cisplatin-resistance via up-regulation of the expression of LncRNA ANRIL in tumour cells" *Sci Rep*
59. Li (2022) "Signatures of EMT, immunosuppression, and inflammation in primary and recurrent human cutaneous squamous cell carcinoma at singlecell resolution" *Theranostics* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548422&blobtype=pdf | # Cytoplasmic translocation of tripartite motif-containing 28 is critical for PRRSV-induced autophagy through promoting Vps34-Beclin1 complex formation
Meng Chen, Yuna Zhao, Hui An, Qingbing Han, Chenchen Cui, Jun Peng, Yihong Xiao, Gang Wang, Yingli Shang
## Abstract
Autophagy, as a highly conserved cellular metabolic regulation mechanism, is a double-edged sword and plays multiple roles in viral infections processes. As a member of the Arteriviridae family within the order Nidovirales, the porcine reproductive and respiratory syndrome virus (PRRSV) induces cell autophagy both in vitro and in vivo. However, the direct or indirect causation of autophagy by PRRSV remains unclear. Identified as an autophagy-related factor, tripartite motif-containing 28 (TRIM28) shows an undefined relationship with autophagy during PRRSV infection. This study investi gates the dynamic changes in autophagy and TRIM28 during PRRSV infection, revealing that PRRSV Nsp4 is identified as a key component responsible for the nuclear export of TRIM28 via a CRM1-dependent pathway, promoting the formation of the Vps34-Beclin1 complex and ultimately initiating autophagy. As a host protein, TRIM28 has exerted a certain antiviral effect, but the mechanism is not yet clear. This study provides detailed insight into the mechanism of PRRSV-mediated autophagy for the first time, offering valuable information for understanding the pathogenesis of porcine reproductive and respiratory syndrome.IMPORTANCE PRRS is one of the major diseases affecting the global swine industry. Infection with PRRSV can cause respiratory disease in pigs of all ages and reproductive disorders in sows. Therefore, understanding the interaction between PRRSV and host factors may help to develop new antiviral strategies against PRRSV. We found that PRRSV Nsp4 was important for nuclear export of TRIM28 in a CRM1-dependent manner during PRRSV infection. TRIM28 in the cytoplasm increases the formation of VPS34-Beclin1 complex by interacting with Vps34, further initiating autophagy. Hence, our study reveals a novel mechanism of PRRSV-mediated autophagy and provides valuable information for further understanding the pathogenesis of PRRS, which might contribute to the development of novel antiviral drugs.
in the PRRSV life cycle, including their involvement in PRRSV-induced autophagy, have been examined (13)(14)(15). However, the specific role of host proteins in PRRSV-induced autophagy remains unclear.
Autophagy, an evolutionarily conserved and highly regulated catabolic process, involves delivering cytoplasmic components to lysosomes for clearance and recycling (16). The formation of an isolation membrane is initiated by Class III phosphatidylinositol 3-kinase, also known as vacuolar protein sorting 34 (Vps34), which converts phospha tidylinositol into phosphatidylinositol-3-phosphate and forms a complex with Vps34/ Beclin1 (17). The elongation of the autophagic membrane requires processing by two ubiquitin-like protein-conjugation systems: Atg5-Atg12 and LC3 (18). Autophagy plays diverse physiological roles, including stress adaptation, development, lipid metabolism, degenerative diseases, protection against inflammation, and defense against intracellular pathogens (17). Host-cell/pathogen co-evolution has led to the selection of microorgan isms capable of evading or exploiting autophagy. As a positive-sense RNA virus, studies have reported that PRRSV-2 induces autophagy in vitro and in vivo (13)(14)(15). However, the mechanisms by which PRRSV infection induces autophagy are not well understood, necessitating further investigation to elucidate the molecular mechanisms involved.
Tripartite motif-containing 28 (TRIM28), also known as KAP1 or TIF1, was initially identified as a nuclear co-repressor for KRAB domain-containing zinc finger proteins (19). It interacts with these proteins to target specific genomic regions and modulates transcription through interactions with HP1 isoforms (20). Additionally, TRIM28 regulates the initiation and elongation of RNA polymerase II (Pol II)-dependent transcription and participates in cellular activities such as cell differentiation, DNA damage response, tumorigenesis, cytokine production, viral replication, stem cell pluripotency, embryonic development, and autophagy (21). Further evidence suggests TRIM28 also functions independently of gene regulation by serving as a signaling scaffold protein, mediating signal transduction through multiple post-translational modifications (PTMs), includ ing serine/tyrosine phosphorylation, SUMOylation, and acetylation, which coordinately regulate its function and protein abundance (22). As an E3 ubiquitin ligase, TRIM28 promotes PRRSV replication by inhibiting viral protein GP4 ubiquitination (23), enhan ces SARS-CoV-2 virulence by increasing nucleocapsid protein SUMOylation (24), and promotes NLRP3 inflammasome activation (25). TRIM28 also forms a repressor complex containing heterochromatin protein 1 (HP1), with importin α playing a crucial role in its nuclear delivery and interaction with HP1. The 462-494 amino acid region of TRIM28 serves as a nuclear localization signal overlapping with its HP1-binding site, known as the HP1 box (19). TRIM28 is considered a critical transcriptional co-repressor for autophagy due to its binding to the conserved KRAB repression domain of many transcription factors (26,27). It has been suggested that TRIM28 promotes porcine epidemic diarrhea virus replication through the inhibition of the JAK-STAT1 pathway mediated by autophagy (28). However, the role of TRIM28 in PRRSV infection-induced autophagy remains unclear and requires further study.
This study investigates the dynamic changes in autophagy and TRIM28 in response to PRRSV infection. The results demonstrate that PRRSV infection induces TRIM28 translocation from the nucleus to the cytoplasm, with PRRSV Nsp4 playing a critical role in the nuclear export of TRIM28 in a CRM1-dependent manner. In the cytoplasm, TRIM28 promotes the formation of the Vps34-Beclin1 complex, initiating the autoph agy process. This study reveals a novel mechanism of PRRSV-mediated autophagy and provides valuable insights into the pathogenesis of PRRS.
## RESULTS
## PRRSV infection induced autophagy accompanied by the increase of cytoplasmic TRIM28
PRRSV infection induces autophagy both in vitro and in vivo (29), as evidenced by increased LC3 puncta due to the conversion of LC3I to LC3II, detectable via immunofluor escence staining or Western blot. Verification of PRRSV-induced autophagy employed Marc-145 and 3D4/21-CD163 cell lines. Immunofluorescence observations revealed a significant increase in LC3 puncta in TA-12 and VR2332 strain-infected Marc-145 cells (Fig. 1A; Fig. S1) and 3D4/21-CD163 cells (Fig. 1B; Fig. S1B) at 36 h post-infection (hpi), indicating autophagy induction. The conversion of LC3I to LC3II was confirmed by Western blot from 12 hpi in TA-12-infected Marc-145 cells (Fig. 1E), 3D4/21-CD163 cells (Fig. 1F) and PAM cells (Fig. 1G), as well as VR2332-infected Marc-145 cells (Fig. S1C) and 3D4/21-CD163 cells (Fig. S1D). These results further confirm autophagy in PRRSV-infected cells. TRIM28's involvement in autophagy (30) prompted an investigation of its dynamic changes. The relationship between TRIM28 and autophagy was confirmed by co-trans fecting Marc-145 cells with GFP-LC3 and TRIM28 plasmids. After 24 h, GFP-tagged LC3 puncta increased in co-transfected cells (Fig. 1C), along with LC3I to LC3II conversion (Fig. 1D), demonstrating TRIM28-induced autophagy.
PRRSV-induced TRIM28 changes were detected in Marc-145, 3D4/21-CD163 cells, and porcine alveolar macrophages (PAMs) during TA-12 and VR2332 isolate infections. Samples collected at various time points were treated with cytoplasmic lysate NP-40 and analyzed by Western blot. Both TA-12 and VR2332 infections increased cytoplasmic TRIM28 in Marc-145 cells (Fig. 1E; Fig. S1C), 3D4/21-CD163 cells (Fig. 1F; Fig. S1D), and PAMs (Fig. 1G). Control infections with SEV and EMCV did not increase cytoplasmic TRIM28 (Fig. S1E andF). These cumulative results indicate PRRSV infection induces both autophagy and increased cytoplasmic TRIM28.
## PRRSV infection causes TRIM28 relocalization from the nucleus to the cytoplasm
TRIM28, known as a transcription regulator localized in the nucleus (31), increases in the cytoplasm following PRRSV infection. To clarify the origin of cytoplasmic TRIM28 during PRRSV infection, mRNA levels of TRIM28 were analyzed. Total RNA extracted at various time points was analyzed by qRT-PCR, showing no significant changes in TRIM28 gene levels in Marc-145 and 3D4/21-CD163 cells infected with TA-12 (Fig. 2A) or VR2332 (Fig. S2A) compared to the control group. This suggests that PRRSV infection affects TRIM28 localization rather than its quantity. Using a sodium dodecyl sulfate (SDS) lysis buffer, the total TRIM28 protein content in PRRSV-infected cells was examined. Western blot analysis revealed no changes in total TRIM28 protein content during TA-12 (Fig. 2B) and VR2332 (Fig. S2B) infections, indicating that total TRIM28 levels remain unchanged during PRRSV infection.
The localization of TRIM28 in PRRSV-infected cells was investigated using nucleo cytoplasmic isolation assays and immunofluorescence analysis. Western blot analysis demonstrated that both TA-12 and VR2332 (MOI=0.1) infections decreased nuclear TRIM28 protein and increased cytoplasmic TRIM28 protein in Marc-145 cells (Fig. 2; Fig. S2C), 3D4/21-CD163 cells (Fig. 2D), and PAMs (Fig. 2E). Immunofluorescence staining confirmed TRIM28 relocalization from the nucleus to the cytoplasm in PRRSV-infected Marc-145 cells (Fig. 2F) and PAMs (Fig. 2G). These findings demonstrate PRRSV-induced TRIM28 relocalization from the nucleus to the cytoplasm in vitro.
## Nsp4 contributes to the TRIM28 redistribution mediated by the CRM1 pathway
During PRRSV infection, Nsps facilitate the nuclear translocation of host cell proteins. For instance, Nsp9 interacts with pRb, relocating it from the nucleus to the cytoplasm (32), while Nsp2 and Nsp10 bind with DDX18, similarly transferring it to the cytoplasm (33). To investigate the role of Nsps in TRIM28 relocalization, plasmids encoding PRRSV Nsps were constructed (Fig. S3A andB). Constructs such as pEGFP-C1-TRIM28 and HA-tag ged Nsp-expressing pCAGGS (Nsp1α, Nsp1β, Nsp3, Nsp4, Nsp5, Nsp6, Nsp7α, Nsp7β, Nsp8, Nsp9, Nsp10, Nsp11, and Nsp12), or pCDNA3.0-Flag-TRIM28 and pEGFP-C1-Nsp2 were co-transfected into Marc-145 or Hela cells and analyzed via confocal microscopy 24 h post-transfection. Results indicated that only Nsp4 facilitated the cytoplasmic relocalization of TRIM28 in both Marc-145 (Fig. 3A) and HeLa cells (Fig. S3C), with TRIM28 colocalizing with Nsp4. Polyclonal antibodies (pAb) against PRRSV Nsp4 were used to detect the localization of Nsp4 after TA-12 virus infected Marc-145 cells (34). Immunofluorescence staining showed that Nsp4 could be localized in both nucleus and cytoplasm (Fig. 3B), and at the same time, Nsp4 and TRIM28 were colocalized in the cytoplasm in TA-12 infected cells overexpressing TRIM28 (Fig. 3C). This interaction was confirmed in HEK-293T cells through co-transfection, Co-IP using GFP antibodies 24 h post-transfection, and subsequent Western blotting, revealing the presence of both Nsp4 and TRIM28 in the Co-IP product (Fig. 3D). Furthermore, the interaction between Nsp4 and endogenous TRIM28 was corroborated by Nsp4 transfection alone in HEK-293T cells (Fig. 3E). Collectively, these findings suggest that the Nsp4-TRIM28 interaction redistrib utes TRIM28 from the nucleus to the cytoplasm, decreasing its nuclear abundance.
Next, the dependence of TRIM28 transport on the nuclear export receptor CRM1 was investigated. As a key nuclear export receptor, CRM1 mediates protein nucleation, and its inhibitor, Leptomycin B (LMB), binds to CRM1, causing its redistribution to the cytoplasm and inhibiting nuclear import (35,36). The role of CRM1 during PRRSV infection was examined in Marc-145 cells treated with LMB. Western blotting and immunofluorescence Full-Length Text analyses revealed that LMB treatment prevented TRIM28 translocation from the nucleus to the cytoplasm in Marc-145 cells (Fig. 4A andB), indicating that CRM1 is involved in PRRSV-induced TRIM28 transfer. To further confirm the interaction between CRM1 and TRIM28, Flag-CRM1 plasmids were transfected into HEK-293T cells, and Co-IP using Flag antibodies demonstrated the presence of both CRM1 and TRIM28 in the Co-IP product (Fig. 4C), suggesting an interaction. Additionally, dynamic changes in CRM1 and TRIM28 interaction post-TA-12 infection were analyzed. Flag-CRM1 transfection followed by PRRSV inoculation in Marc-145 cells showed upregulation of Flag-CRM1 and TRIM28 (Fig. 4D). However, LMB treatment resulted in the absence of TRIM28 despite CRM1 still being present in the Co-IP product post-PRRSV infection (Fig. 4E). These findings confirm that PRRSV-induced TRIM28 redistribution is CRM1-dependent. Given Nsp4's role in TRIM28 redistribution, its involvement in PRRSV-induced TRIM28 relocation via the CRM1 pathway was further examined. HA-Nsp4, GFP-TRIM28, and Flag-CRM1 plasmids were co-transfected into HEK-293T cells to determine Nsp4's contribution to the TRIM28-CRM1 interaction. Co-immunoprecipitation (Co-IP) with HA antibodies and subsequent Western blotting revealed the co-presence of Nsp4, TRIM28, and CRM1 in the Co-IP product, indicating that Nsp4 enhances the interaction between TRIM28 and CRM1 (Fig. 4F). To identify the functional domain responsible for this interaction, plasmids encoding full-length or truncated TRIM28 were generated. Co-IP with Flag antibodies and Western blotting showed that Flag-TRIM28 co-precipitated with CRM1, with the RBCC domain identified as key to this interaction (Fig. 4G). Collectively, these data demonstrate that PRRSV Nsp4-induced transfer of TRIM28 from the nucleus to the cytoplasm is CRM1-dependent, with the RBCC domain playing a crucial role.
## PRRSV infection-induced autophagy depends on TRIM28
PRRSV infection-induced autophagy, coupled with TRIM28 transport, suggests a role for TRIM28 in PRRSV-induced autophagy. To determine whether PRRSV-induced autophagy relies on TRIM28, TRIM28 was knocked down in Marc-145 and 3D4/21 cells using siRNA targeting TRIM28, which effectively inhibited its transcription and expression (Fig. 5A). These cells, along with control siRNA cells, were inoculated with TA-12 at 0.1 MOI. In TRIM28 knockdown cells, PRRSV infection did not induce the conversion of LC3I to LC3II at 12, 24, and 36 h post-infection in both Marc-145 and 3D4/21 cells, as shown by Western blot analysis (Fig. 5B andC). Immunofluorescence analysis also revealed a significant reduction in LC3 spots in Marc-145 cells following TRIM28 knockdown (Fig. 5D). These findings suggest that inhibiting TRIM28 expression impedes PRRSV-induced autophagy. Collectively, this evidence demonstrates that PRRSV-induced autophagy is dependent on TRIM28 participation.
## TRIM28 mediates PRRSV-induced autophagy by interacting with VPS34.
Given that PRRSV infection-induced autophagy depends on TRIM28, the role of TRIM28 as an autophagy regulator interacting with autophagy-associated proteins was further investigated. Vps34, also known as PI3K catalytic subunit type III (PI3KC3), is a central autophagy protein that forms a complex with Beclin1, Vps15, and ATG14L (37)(38)(39). During TA-12 infection in Marc-145 and 3D4/21 cells, Western blot analysis showed upregulation of Vps34 and Beclin1 (Fig. 6A). To determine if PRRSV infection affects the interaction between TRIM28 and Vps34, Co-IP using TRIM28 pAb antibodies at 24 h post-infection revealed PRRSV-induced upregulation of Vps34. Similarly, GFP-TRIM28 transfected Marc-145 cells inoculated with TA-12 at 0.1 MOI showed increased Vps34 levels via Co-IP with GFP antibodies and subsequent Western blotting (Fig. 6B). These findings suggest that PRRSV infection enhances the interaction between TRIM28 and Vps34. Additionally, the effect of CRM1-mediated TRIM28 nucleation on the TRIM28-Vps34 interaction was examined. GFP-TRIM28 transfected Marc-145 cells were inoculated with TA-12 at 0.1 MOI or treated with LMB before TA-12 inoculation. Co-IP with anti-GFP antibodies showed that PRRSV infection enhanced the interaction between GFP-TRIM28/ endogenous TRIM28 and Vps34, while LMB treatment weakened this interaction (Fig. 6C).
Given the importance of the Vps34-Beclin1 complex in autophagy induction (40), the interaction between Vps34 and Beclin1 during PRRSV infection was also investigated. The interaction between Vps34 and Beclin1 was confirmed through transfection of Vps34 in Marc-145 cells, followed by Co-IP using antibodies against Beclin1 to identify the interaction between endogenous Vps34 and Beclin1. The results indicated that the interaction between Vps34 and Beclin1 was significantly enhanced after PRRSV infection (Fig. 6D andE). Additionally, Co-IP assays demonstrated that Vps34 and Beclin1 interact weakly following TRIM28 knockdown in Marc-145 or 3D4/21 cells (Fig. 6F). These data reveal that TRIM28 directly interacts with Vps34 and modulates the Vps34-Beclin1 interaction to induce autophagy.
## TRIM28 regulates PRRSV replication
Autophagy is closely related to viral replication, and whether TRIM28-mediated autophagy induction affects viral replication was also examined. TRIM28, a known transcriptional corepressor for KRAB domain-containing zinc finger transcription factors (41), was targeted via knockdown at Marc-145 cells. The results showed knockdown of TRIM28 significantly increased PRRSV copy numbers after PRRSV infection, particu larly at 48 h (Fig. 7A). Consistently, immunofluorescence revealed elevated levels of the viral N protein at 36 hpi following TRIM28 knockdown (Fig. 7B), and confirmed by western blotting (Fig. 7C). Conversely, TRIM28 overexpression reduced viral copy numbers at 24 and 48 hpi (most notably at 48 h; Fig. 7D) and decreased N protein levels (Fig. 7E). To determine whether TRIM28's nuclear translocation regulates PRRSV replication, we treated Marc145 cells with LMB, a CRM1 inhibitor that blocks nuclear export. LMB treatment increased viral copy numbers (especially at 36 hpi; Fig. 7F) and enhanced N protein expression (Fig. 7G), indicating that nuclear retention of TRIM28 suppresses PRRSV. Collectively, these data demonstrate that TRIM28 expression and nuclear localization inhibit PRRSV replication.
## DISCUSSION
PRRSV infection-induced autophagy has been studied both in vitro and in vivo (29,42,43), yet the mechanisms of virus-host interactions remain incompletely understood. This study found that during PRRSV infection, Nsp4 induces TRIM28 redistribution from the nucleus to the cytoplasm via the CRM1 pathway, forming a complex with VPS34 and Beclin1 to trigger autophagy. The output of this TRIM28 is likely a host defense mechanism. This study first elucidates the potential mechanism of autophagy induction by PRRSV infection and the dynamic changes of TRIM28, providing an understanding of the host's antiviral mechanisms.
TRIM28 has been proven to be an autophagy-related factor (30), relocating from the nucleus to the cytoplasm during PRRSV infection. PRRSV Nsps are known to mediate the nuclear translocation of host cell proteins during infection (32,33). Building on these studies, it was hypothesized that Nsps contribute to TRIM28 translocation during PRRSV infection. Plasmids for PRRSV Nsps and TRIM28 were constructed, and Nsp4 was identified as the only Nsp interacting with TRIM28 in the cytoplasm. Nsp4, the primary proteinase in EAV, possesses chymotrypsin-like serine protease activity (44,45). During PRRSV replication, Nsp4 localizes in both the cytoplasm and nucleus (6,46), which supports its colocalization with TRIM28 in the cytoplasm. These results suggest that Nsp4 may facilitate nuclear translocation, warranting further investigation. The export of large nuclear proteins (>50 kDa) relies on the export adapter pro tein Nmd3, which provides a nuclear export signal (NES). The leucine-rich NES is recognized by the export receptor CRM1, facilitating passage through nuclear pore complexes in the nuclear membrane (36,47,48). It has not been previously reported whether TRIM28 export requires transport receptors. In this study, Western blotting and immunofluorescence results showed that TRIM28 translocation from the nucleus to the cytoplasm occurs alongside up-regulated CRM1 during PRRSV infection, suggest ing a CRM1-dependent nuclear export pathway. To further investigate CRM1's role in TRIM28 translocation after PRRSV infection, the CRM1 inhibitor LMB was used. LMB, a 540 Da polyketide, binds with CRM1, causing its redistribution from the nucleus to the cytoplasm (35,36). Marc-145 cells treated with LMB 3 h before PRRSV infection exhibited CRM1 redistribution from the nucleus to the cytoplasm. The results indicated that LMB treatment retained TRIM28 almost entirely in the nucleus, confirming TRIM28 translocation via the CRM1-dependent nuclear export pathway during PRRSV infection.
Another interesting observation is the simultaneous occurrence of TRIM28 trans location and autophagy in PRRSV-infected Marc-145 cells. TRIM28's participation in autophagy has been reported in glioblastoma (49), where ubiquitination and degrada tion of AMP-activated protein kinase (AMPK) by the cancerspecific MAGE-A3/6-TRIM28 ubiquitin ligase inhibit autophagy in cancer cells (30). These observations raise the question of whether autophagy is induced directly by PRRSV infection or as a result of TRIM28 export. To investigate the relationship among PRRSV infection, TRIM28 translocation, and autophagy, siRNA targeting TRIM28 was used in Marc-145 and 3D4/21 cells. The conversion of LC3I to LC3II did not occur in these cells, demonstrating that PRRSV-induced TRIM28 translocation contributes to autophagy. During autophagy, the formation of double-membrane autophagosomes requires a complex formed by Vps34 bound to Beclin1, Vps15, and ATG14L (37)(38)(39). PRRSV infection enhanced the interac tion between Vps34 and Beclin1, which was weakened by knocking down TRIM28 in Marc-145 or 3D4/21 cells and by LMB treatment. This finding provides evidence that PRRSV-induced autophagy depends on cytoplasmic TRIM28 participation.
TRIM family members can restrict viral replication either directly or by activating innate immune responses (50). Conversely, viruses may encode antagonists to subvert TRIM proteins, thereby enhancing viral replication (23,51,52). TRIM28 has long been recognized as a transcriptional repressor, and its inhibitory role in gene expression has profound implications for viral transcription and replication (53,54). Accumulat ing evidence indicates that TRIM28 effectively suppresses the transcription of various herpesviruses, including KSHV, HCMV, and EBV (55)(56)(57). Notably, a recent study revealed that TRIM28 functions as a restriction factor for prototype foamy virus replication through two distinct mechanisms: mediating H3K9 trimethylation and promoting degradation of the viral transactivator Tas (58). In this study, we demonstrated that TRIM28 overexpression suppresses PRRSV replication, whereas TRIM28 knockdown or inhibition of its nuclear export (via LMB treatment) enhances viral replication. Given its well-established role in transcriptional suppression, we propose that TRIM28 likely modulates PRRSV infection by inhibiting viral gene expression.
In conclusion, this study elucidates that PRRSV infection induces autophagy through the translocation of TRIM28 from the nucleus to the cytoplasm, where it forms a complex with VPS34 and Beclin1. In addition, PRRSV Nsp4 was identified as a key factor that facilitates the nuclear export of TRIM28 via a CRM1-dependent pathway (Fig. 8). As a member of the Arteriviridae family within the Nidovirales order, PRRSV is the primary causative agent of PRRS, which poses a significant challenge in veterinary medicine. This investigation is the first to detail the mechanism of PRRSV-mediated autophagy, offering valuable insights into the pathogenesis of PRRS and potentially other Nidovir ales members, such as SARS-CoV-2, porcine epidemic diarrhea virus, and infectious bronchitis virus.
## MATERIALS AND METHODS
## Cells and viruses
Porcine alveolar macrophage (PAM) cells were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10% (V/V) fetal bovine serum (FBS, Gibco). The 3D4/21-CD163 PAM cell line, HEK-293T cells, HeLa cells, and Marc-145 cells were maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco) with 10% fetal bovine serum (FBS, BI), 100 U/ml penicillin, and 100 µg/ml streptomycin sulfate in a humidified incubator at 37°C and 5% CO 2 . The highly pathogenic PRRSV TA-12 strain (GenBank accession no. HQ416720) and the classical PRRSV VR2332 strain (GenBank accession no. EF536003.1) served as viral inoculum, with titers of 10 7.67 TCID 50 /mL and 10 6.37 TCID 50 /mL, respectively. EMCV and SeV were utilized as controls.
## Plasmid construction and transient transfections
The pcDNA6.2/N-EmGFP-DEST-TRIM28 plasmid was acquired from Addgene. The pCDNA3.0-Flag-TRIM28 plasmid was engineered into the pCDNA3.0-Flag expression vector using HindIII and XhoI recognition sequences. TRIM28 mutant plasmids (R487E, V488E) and truncated plasmids (TRIM28-C, TRIM28-N) were generated with the Q5 Site-Directed Mutagenesis Kit (NEB, E0552S) and stored in our laboratory. cDNA encoding the full-length Nsp from the HP-PRRSV strain was subcloned into the pCAGGS-HA expression vector. Primers for all constructs are listed in Table 1. E. coli strain DH5α was employed for transformation. Plasmid overexpression was achieved using Lipofecta mine 2000 reagent (Invitrogen, 11668-019) following the manufacturer's instructions. for 1.5 h at room temperature and washed again. The cell nuclei were labeled with 4' ,6-diamidino-2-phenylindole (C1006, Beyotime) for 5 min and washed with PBS. Cells were examined using a confocal microscope (Nikon), and images were captured at 63× and 10× magnifications.
## Western blot analysis
Cells were washed twice with cold PBS and lysed in NP-40 buffer (Lysis Buffer 2×: NP-40: 1%; Hepes: 25 mM, pH 7.4; EDTA: 5 mM) containing a protease inhibitor mix (EDTA: 1 mM, pH 8.0; EGTA: 8 mM, pH 8.0; Na 3 VO 4 : 1 mM; NaF: 250 mM; PMSF: 100 µM). One part lysis buffer was mixed with one part protease inhibitor mix and 1 mM dithiothreitol (DTT) for cytoplasmic extracts, or cells were lysed in SDS lysis buffer (Tris-HCl: 100 mM, pH 6.8; 4% SDS; 0.2% bromophenol blue; 20% glycerol; 200 mM DTT) for total cell extracts. Cell lysates were boiled in a 6× loading buffer for 10 min, and equal amounts of samples were loaded onto 12% (wt/vol) SDS-PAGE gels. The separated proteins were transferred to methanol-activated polyvinylidene fluoride membranes (Millipore). Membranes were blocked with 5% BSA (Sigma) in PBS with Tween 20 (PBS-T) at room temperature for 1 h and incubated with various primary antibodies (Table 2) overnight at 4°C. After washing, the membranes were incubated with the appropriate secondary antibodies at room temperature for 1 h. Immunoreactive bands were visualized using an enhanced chemiluminescence (ECL) detection system (Tanon, 5100).
## Fractionation of nuclear and cytoplasmic extracts
First, 4×10 6 -8×10 6 cells were harvested into 200 µL buffer A (Hepes: 10 mM; KCl: 10 mM; EDTA: 0.1 mM; EGTA: 0.1 mM) with protease inhibitor and DTT by gentle pipetting, and incubated on ice for 15 min. Four microliters of 10% NP-40 was added, followed by vortexing and incubation on ice for 2 min. The mixture was centrifuged at 10,000 rpm for 1.5 min at 4°C, and the supernatant, representing the cytoplasmic extract, was collected and boiled in buffer for 10 min. The pellet was lysed in 100 µL 1× SDS lysis buffer with 2.5% β-mercaptoethanol and heated at 95°C for 5 min, and used for subsequent protein level detection. The nuclear protein extraction Kit (A10039) is also used for the separation of nuclear proteins and cytoplasmic proteins. The separately collected nuclear protein and cytoplasmic protein samples are mixed with 5× SDS and heated at 95°C for 5 min for subsequent protein level detection. The abundance of TRIM28 in the nucleus and cytoplasm was detected by Western blotting.
## TaqMan fluorescent quantitative real-time PCR
Total RNA was extracted from cells using the Eastep Super Total RNA Extraction Kit (Promega, LS1040) following the manufacturer's instructions. Reverse transcription was conducted with the M-MLV reverse transcriptase (TaKaRa, #2641A). TaqMan fluores cent RT-qPCR was performed on the ABI sequence detector system (StepOnePlus, Life Technologies Holdings Pte Ltd) in a final volume of 10 µL, including 2 µL cDNA (~100 ng/ µL), 5 µL 2× UltraSYBR Mixture (High ROX) (CWBIO), 0.5 µL primer pairs, and 2 µL water. The PCR conditions were: denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 10 s, annealing at 55°C for 30 s, and extension at 72 ° C for 45 s. All primers used for quantitative real-time PCR are listed in Table 3. siRNA knockdown: Small interfering RNA (siRNA) sequences targeting TRIM28 (siTRIM28) and a negative control RNA (NC) were designed and synthesized by GenePharma (Table 4, Shanghai, China) using the Rosetta algorithm for siRNA design and NCBI BLAST for offtarget analysis. Briefly, Marc145 cells in 12-well plates (60-80% confluence) were transfected with 15 µM siTRIM28 or NC using Lipofectamine RNAiMax (Invitrogen) for 24 h. Subsequently, the cells were infected with 0.1 MOI PRRSV and harvested at the indicated times. TRIM28 gene expression levels were confirmed by quantitative real-time PCR (qRT-PCR), and protein levels were verified by Western blotting.
## Co-immunoprecipitations (Co-IP)
Marc145 and HEK293T cells cultivated in 10 cm dishes were transfected with the appropriate expression plasmid and harvested 24 h post-transfection. Following washing with cold PBS (pH 7.4), the cell pellet was lysed with NP-40 buffer (20 mM Tris, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% NP-40, 2.5 mM sodium pyrophosphate, and 1 mM Na3VO4) containing 100 µM PMSF. Cell lysates were incubated with 2 µg antibody overnight at 4°C with gentle agitation. Subsequently, 20 µL of protein A/G PLUS-Agarose beads (Santa, sc-2003), pre-washed three times with cold PBS (pH 7.4), were added and incubated at 4°C for 4-6 h. The beads were washed three times with Co-IP lysis buffer, and the precipitates were eluted in 60 µL lysis buffer and 6× loading buffer by boiling at 95°C for 8 min. The supernatants were subjected to SDS-PAGE.
## Statistical analysis
Statistical significance was determined using a two-tailed Student's t-test, with P<0.05 considered statistically significant.
## References
1. Allende, Laegreid, Kutish et al. (2000) "Porcine reproductive and respiratory syndrome virus: description of persistence in individual pigs upon experimental infection" *J Virol*
2. Fu, Quan, Zhang et al. (2012) "Porcine reproductive and respiratory syndrome virus induces interleukin-15 through the NF-κB signaling pathway" *J Virol*
3. Pejsak, Stadejek, Daniel (1997) "Clinical signs and economic losses caused by porcine reproductive and respiratory syndrome virus in a large breeding farm" *Vet Microbiol*
4. Adams, Lefkowitz, King et al. (2017) "Changes to taxonomy and the International Code of Virus Classification and Nomenclature ratified by the International Committee on Taxonomy of Viruses"
5. Johnson, Griggs, Gnanandarajah et al. (2011) "Novel structural protein in porcine reproductive and respiratory syndrome virus encoded by an alternative ORF5 present in all arteriviruses" *J Gen Virol*
6. Beura, Sarkar, Kwon et al. (2010) "Porcine reproductive and respiratory syndrome virus nonstructural protein 1beta modulates host innate immune response by antagonizing IRF3 activation" *J Virol*
7. (2025) *Full-Length Text Journal of Virology*
8. Den Boon, Faaberg, Meulenberg et al. (1995) "Processing and evolution of the Nterminal region of the arterivirus replicase ORF1a protein: identification of two papainlike cysteine proteases" *J Virol*
9. Fang, Treffers, Li et al. (2012) "Efficient -2 frameshifting by mammalian ribosomes to synthesize an additional arterivirus protein" *Proc Natl Acad Sci*
10. Li, Zhou, Zhang et al. (2014) "Nsp9 and Nsp10 contribute to the fatal virulence of highly pathogenic porcine reproductive and respiratory syndrome virus emerging in China" *PLoS Pathog*
11. Snijder, Wassenaar, Spaan (1994) "Proteolytic processing of the replicase ORF1a protein of equine arteritis virus" *J Virol*
12. Van Dinten, Wassenaar, Gorbalenya et al. (1996) "Processing of the equine arteritis virus replicase ORF1b protein: identification of cleavage products containing the putative viral polymerase and helicase domains" *J Virol*
13. Wassenaar, Spaan, Gorbalenya et al. (1997) "Alternative proteolytic processing of the arterivirus replicase ORF1a polyprotein: evidence that NSP2 acts as a cofactor for the NSP4 serine protease" *J Virol*
14. Huang, Zhang, Guo et al. (2014) "Porcine reproductive and respiratory syndrome virus nonstructural protein 4 antagonizes beta interferon expression by targeting the NF-κB essential modulator" *J Virol*
15. Li, Chen, He et al. (2019) "E3 ligase ASB8 promotes porcine reproductive and respiratory syndrome virus proliferation by stabilizing the viral Nsp1α protein and degrading host IKKβ kinase" *Virology (Auckl)*
17. Luo, Xiao, Jiang et al. (2008) "Porcine reproductive and respiratory syndrome virus (PRRSV) suppresses interferon-beta production by interfering with the RIG-I signaling pathway" *Mol Immunol*
18. Kroemer, Mariño, Levine (2010) "Autophagy and the integrated stress response" *Mol Cell*
19. Grégoire, Richetta, Meyniel-Schicklin et al. (2011) "IRGM is a common target of RNA viruses that subvert the autophagy network" *PLoS Pathog*
20. Geng, Klionsky (2008) "Quantitative regulation of vesicle formation in yeast nonspecific autophagy" *Autophagy*
21. Moriyama, Sangel, Yamaguchi et al. (2015) "Identification and characterization of a nuclear localization signal of TRIM28 that overlaps with the HP1 box" *Biochem Biophys Res Commun*
22. Gehrmann, Burbage, Zueva et al. (2019) "Critical role for TRIM28 and HP1β/γ in the epigenetic control of T cell metabolic reprograming and effector differentiation" *Proc Natl Acad Sci*
23. Czerwińska, Mazurek, Wiznerowicz (2017) "The complexity of TRIM28 contribution to cancer" *J Biomed Sci*
24. Cheng, Kuo, Dk (2014) "KAPtain in charge of multiple missions: emerging roles of KAP1" *World J Biol Chem*
25. Cui, Zhou, Zhao et al. (2023) "The host E3ubiquitin ligase TRIM28 impedes viral protein GP4 ubiquitination and promotes PRRSV replication" *Int J Mol Sci*
26. Ren, Wang, Zong et al. (2024) "TRIM28-mediated nucleocapsid protein SUMOylation enhances SARS-CoV-2 virulence" *Nat Commun*
27. Qin, Li, Liang et al. (2021) "TRIM28 SUMOylates and stabilizes NLRP3 to facilitate inflammasome activation" *Nat Commun*
28. Bunch, Calderwood (2015) "TRIM28 as a novel transcriptional elongation factor" *BMC Mol Biol*
29. Chang, Hwang, Kim et al. (2021) "TRIM28 functions as a negative regulator of aggresome formation" *Autophagy*
30. Li, Yan, Ma et al. (2024) "TRIM28 promotes porcine epidemic diarrhea virus replication by mitophagy-mediated inhibition of the JAK-STAT1 pathway" *Int J Biol Macromol*
31. Sun, Huang, Wang et al. (2012) "Porcine reproductive and respiratory syndrome virus induces autophagy to promote virus replication" *Autophagy*
32. Pineda, Ramanathan, Tacer et al. (2015) "Degradation of AMPK by a cancerspecific ubiquitin ligase" *Cell*
33. Ryan, Schultz, Ayyanathan et al. (1999) "KAP-1 corepressor protein interacts and colocalizes with heterochromatic and euchromatic HP1 proteins: a potential role for krüppel-associated box-zinc finger proteins in heterochromatin-mediated gene silencing" *Mol Cell Biol*
34. Dong, Zhang, Ge et al. (2014) "The interaction of nonstructural protein 9 with retinoblastoma protein benefits the replication of genotype 2 porcine reproductive and respiratory syndrome virus in vitro" *Virology (Auckl)*
35. Jin, Zhou, Ge et al. (2017) "Cellular DEAD-box RNA helicase 18 (DDX18) promotes the PRRSV replication via interaction with virus nsp2 and nsp10" *Virus Res*
36. Jiao, Li, Liu et al. (2023) "Porcine reproductive and respiratory syndrome virus infection inhibits NF-κB signaling pathway through cleavage of IKKβ by Nsp4" *Vet Microbiol*
37. Rahmani, Dean (2017) "Leptomycin B alters the subcellular distribution of CRM1 (Exportin 1)" *Biochem Biophys Res Commun*
38. Yu, Liu, Cao et al. (2012) "Identification and characterization of three novel nuclear export signals in the influenza A virus nucleoprotein" *J Virol*
39. Simonsen, Tooze (2009) "Coordination of membrane events during autophagy by multiple class III PI3-kinase complexes" *J Cell Biol*
40. Xu, Chen, Huang et al. (2012) "Development of multiplex PCR for simultaneous detection of six swine DNA and RNA viruses" *J Virol Methods*
41. Yang, Klionsky (2010) "Mammalian autophagy: core molecular machinery and signaling regulation" *Curr Opin Cell Biol*
42. Funderburk, Wang, Yue (2010) "The Beclin 1-VPS34 complex--at the crossroads of autophagy and beyond" *Trends Cell Biol*
43. Friedman, Fredericks, Jensen et al. (1996) "KAP-1, a novel corepressor for the highly conserved KRAB repression domain" *Genes Dev*
44. Chen, Fang, Wang et al. (2012) "Induction of autophagy enhances porcine reproductive and respiratory syndrome virus replication" *Virus Res*
45. Wang, Yu, Tu et al. (2015) "Highly pathogenic porcine reproductive and respiratory syndrome virus infection induced apoptosis and autophagy Full-Length Text Journal of Virology October"
46. "in thymi of infected piglets" *PLoS One*
47. Snijder, Wassenaar, Van Dinten et al. (1996) "The arterivirus nsp4 protease is the prototype of a novel group of chymotrypsin-like enzymes, the 3C-like serine proteases" *J Biol Chem*
48. Van Dinten, Rensen, Gorbalenya et al. (1999) "Proteolytic processing of the open reading frame 1b-encoded part of arterivirus replicase is mediated by nsp4 serine protease and Is essential for virus replication" *J Virol*
49. Chen, Li, He et al. (2014) "The amino acid at residue 155 in nonstructural protein 4 of porcine reproductive and respiratory syndrome virus contributes to its inhibitory effect for interferon-β transcription in vitro" *Virus Res*
50. Nakielny, Dreyfuss (1999) "Transport of proteins and RNAs in and out of the nucleus" *Cell*
51. Zhao, Jin, Wieslander (2004) "CRM1 and Ran are present but a NES-CRM1-RanGTP complex is not required in Balbiani ring mRNP particles from the gene to the cytoplasm" *J Cell Sci*
52. Peng, Zhang, Jiang et al. (2019) "TRIM28 activates autophagy and promotes cell proliferation in glioblastoma" *Onco Targets Ther*
53. Yang, Wang, Fu et al. (2025) "TRIM28 functions as SUMO ligase to SUMOylate TRAF6 and regulate NF-κB activation in HBV-replicating cells" *Hepatol Int*
54. Sharma, Tyagi, Khumallambam et al. (1950) "Cocaine-induced DNA-dependent protein kinase relieves RNAP II pausing by promoting TRIM28 phosphorylation and RNAP II hyperphosphorylation to enhance HIV transcription" *Cells*
55. Van Tol, Hage, Giraldo et al. (2017) "The TRIMen dous role of TRIMs in virus-host interactions" *Vaccines (Basel)*
56. Sushma (2011) "KAP1 protein: an enigmatic master regulator of the genome" *J Biol Chem*
57. Imbeault, Helleboid, Trono (2017) "KRAB zincfinger proteins contribute to the evolution of gene regulatory networks" *Nature*
58. Benjamin, Min, Marco et al. (2015) "Release of human cytomegalovirus from latency by a KAP1/TRIM28 phosphorylation switch"
59. Zhang, Zhu, Guo et al. (2014) "Inhibition of KAP1 enhances hypoxia-induced Kaposi's sarcoma-associated herpesvirus reactivation through RBP-Jκ" *J Virol*
60. Li, Burton, Bhaduri-Mcintosh (2017) "Chloroquine triggers Epstein-Barr virus replication through phosphorylation of KAP1/TRIM28 in Burkitt lymphoma cells" *PLoS Pathog*
61. Yuan, Yan, Wang et al. (2021) "Trim28 acts as restriction factor of prototype foamy virus replication by modulating H3K9me3 marks and destabilizing the viral transactivator Tas" *Retrovirology (Auckl)* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12817477&blobtype=pdf | # Protease-mediated maturation of M-PMV reverse transcriptase into a functional heterodimer
Marina Kapisheva, Petra Junkov, Ondřej Van Ěk, Zuzana Jalovcov, Ivana Křížov, | Alžb Ěta Dost, Michaela Rumlov
## Abstract
Reverse transcriptase (RT) of retroviruses orchestrates viral replication, yet its structural diversity remains poorly understood. Well-studied RTs, such as those from HIV-1, murine leukemia virus, and avian myeloblastosis virus, were characterized decades ago, but less prominent retroviruses have escaped detailed analysis. Despite being discovered alongside HIV-1, the RT of Mason-Pfizer monkey virus (M-PMV) has resisted recombinant expression, leaving its properties unresolved. Here, we report the first detailed analysis of M-PMV RT, a betaretroviral enzyme previously thought challenging to obtain recombinantly. Using baculovirus-based expression in insect cells, we produced soluble full-length RT that, upon proteolytic maturation by the M-PMV protease, yielded a heterodimer composed of p65 and p51 subunits. Mass spectrometry, N-terminal sequencing, and analytical ultracentrifugation demonstrated that full-length RT forms a homodimer, which converts into a stable and more enzymatically active heterodimer following proteolytic removal of the C-terminal RNase H domain from one subunit. Functional assays revealed that heterodimer formation enhances polymerase activity while preserving RNase H function, directly linking proteolytic maturation to enzymatic activation. Notably, this heterodimeric architecture is uncommon among betaretroviruses and resembles the wellcharacterized lentiviral HIV-1 RT. These results broaden the evolutionary perspective on RT heterodimerisation by revealing that this architecture extends into betaretroviruses.
## 1 | INTRODUCTION
Retroviruses replicate by converting their singlestranded RNA genome into a stably integrated DNA copy, a process catalyzed by viral reverse transcriptase (RT). This multifunctional RNA-dependent DNA polymerase synthesizes complementary DNA and degrades the original RNA template (Huber et al., 2023). Discovered more than 50 years ago (Baltimore, 1970;Temin & Mizutani, 1970), RT paved the way for new experimental approaches and has become an essential tool in molecular biology, facilitating cDNA synthesis and transcriptome analysis. Despite this long-standing interest, some reverse transcriptases are still not fully characterized, and the structural and biochemical diversity of RTs across retroviral genera remains poorly understood. In particular, the reverse transcriptase of Mason-Pfizer monkey virus (M-PMV), a betaretrovirus, has resisted recombinant production, and its structural organization and enzymatic characterization are largely unexplained. Here, we address this gap by combining recombinant expression with biochemical and biophysical approaches to define the structural and functional properties of M-PMV RT.
In all seven genera of retroviruses: Alpharetrovirus, Betaretrovirus, Gammaretrovirus, Deltaretrovirus, Epsilonretrovirus, Spumavirus, and Lentivirus (Coffin et al., 1997), the reverse transcriptase is encoded within the pol gene and expressed as part of a Gag-Pol polyprotein precursor through ribosomal frameshifting. The details of this strategy differ among retroviruses. In lentiviruses such as HIV-1, a single frameshift produces Gag and Gag-Pol polyproteins (Menéndez-Arias et al., 2017), whereas the betaretrovirus M-PMV employs two ribosomal frameshifts: the first near the 3 0 end of gag yields Gag-Pro and the second near the 3 0 end of pro generates Gag-Pro-Pol (Marx, 2008). RT is released from these precursors by the viral protease during viral maturation. A direct consequence of this dual frameshift mechanism is that M-PMV produces significantly lower levels of RT, roughly 10-fold less than HIV-1 (Kohoutov a et al., 2009). Such limited expression suggests that M-PMV RT may require intrinsically higher enzymatic activity to ensure efficient reverse transcription.
Despite their sequence diversity, all studied retroviral RTs share a conserved modular architecture consisting of two spatially separated domains: the DNA polymerase domain linked to the C-terminal RNase H domain. The DNA polymerase domain forms a structure resembling a right-hand conformation comprising the fingers, palm, and thumb subdomains, which form the nucleic acid-binding cleft (Hizi & Herschhorn, 2008). The palm contains the catalytic site of the polymerase, which is essential for DNA synthesis and is formed by highly conserved amino acid residues. The fingers guide incoming nucleotides toward the template strand, while the thumb stabilizes the DNA/RNA hybrid during elongation. The connection subdomain links the polymerase core to the C-terminal RNase H domain. The RNase H domain is responsible for degrading the RNA strand of RNA/DNA hybrids. Structurally, it adopts a fold of five mixed β-sheets surrounded by four αhelices. The catalytic center includes conserved acidic residues (Asp, Glu) that coordinate Mg 2+ ions essential for cleavage (Herschhorn & Hizi, 2010).
Despite this conserved framework, RTs exhibit striking variability in their quaternary structure, as illustrated by the presence of either monomeric or dimeric organization. Lentiviral HIV-1 RT, for example, is a wellcharacterized RT heterodimer composed of p66 and p51 subunits (London, 2016;London, 2019;Zheng et al., 2015). The p66 subunit contains the active sites for both DNA polymerase and RNase H, whereas the p51 subunit primarily serves as a structural scaffold (Xavier Ruiz & Arnold, 2020). A similar heterodimeric organization is found in alpharetroviral RTs, such as avian leukosis virus (ALV). The larger 94 kDa β-subunit of ALV RT contains polymerase, RNase H, and integrase domains; the smaller 63 kDa α-subunit lacks the integrase domain (Hizi & Herschhorn, 2008;Werner & Wöhrl, 1999). By contrast, RTs from deltaretroviruses (e.g., bovine leukemia virus, BLV) and gammaretroviruses (e.g., murine leukemia virus, MLV) function as monomers (Das & Georgiadis, 2004;Perach & Hizi, 1999). Also, betaretroviral mouse mammary tumor virus (MMTV) RT has been characterized as a monomer (Taube et al., 1998). In contrast, the structural characterization of M-PMV RT, another betaretroviral enzyme, remains limited.
The main obstacle to the detailed structural and biochemical characterization of M-PMV RT has been the failure of previous attempts to produce the enzyme in bacterial systems using recombinant techniques. M-PMV RT characterization has therefore been limited to measurement of its polymerase activity (Křízov a et al., 2012). More recently, however, we have overcome the RT production problems and generated a specific anti-M-PMV RT antibody that enabled the detection of partially processed RT within virions (Dost alkov a et al., 2024). The detection of two distinct RT-derived proteins in mature M-PMV virions suggested that, in contrast to the monomeric RT of related betaretrovirus MMTV, M-PMV may employ a different mode of organization. In this study, we address this issue by establishing a recombinant expression system for M-PMV RT and applying complementary biochemical approaches to characterize its organization and activity. Our results demonstrate that, despite its classification within betaretroviruses, which otherwise contain predominantly monomeric RTs, M-PMV RT undergoes protease-mediated maturation to form a heterodimeric enzyme. This unexpected finding reveals that the structural and functional properties of M-PMV RT more closely resemble those of HIV-1 RT.
## 2 | RESULTS
## 2.1 | Sequence comparison of retroviral reverse transcriptases
Reverse transcriptases across different retroviral genera share conserved features, retaining the structural motifs essential for their catalytic function, yet exhibit notable differences in both sequence and structural organization. A multiple sequence alignment of the fingers and palm subdomains from six representative RTs, M-PMV and MMTV (Betaretrovirus), BLV (Deltaretrovirus), HIV-1 (Lentivirus), ALV (Alpharetrovirus), and MLV (Gammaretrovirus) revealed several conserved regions (Figure 1a, blue). Among them, the canonical YXDD motif (highlighted in red) is a highly conserved sequence within the polymerase active site critical for catalytic activity. Pairwise sequence comparisons showed that M-PMV RT is most related to MMTV RT (with 62.4% sequence identity), consistent with their close taxonomic placement within the betaretrovirus genus. Moderate similarity was observed with ALV (48.2%) and BLV (37.2%) RTs, while the lowest identity was observed with the more distantly related RTs of HIV-1 (32.4%) and MLV (32.1%).
Despite these conserved elements, retroviral RTs exhibit distinct structural organization (Figure 1b). Alpharetroviral RTs (e.g., ALV) form heterodimers, with a 95 kDa β-subunit that includes the polymerase, RNase H, and integrase domains and a 63 kDa α-subunit lacking the integrase. Lentiviral RTs (e.g., HIV-1) also form heterodimers, composed of a 66 kDa subunit (p66) containing both polymerase and RNase H domains and a 51 kDa subunit (p51) lacking the RNase H domain. In contrast, RTs characterized from gammaretroviruses (MLV, 71 kDa), deltaretroviruses (BLV, 65 kDa), and betaretroviruses (MMTV, 66 kDa) are monomeric.
## 2.2 | Purification and proteolytic processing of recombinant M-PMV reverse transcriptase
Initial attempts to express M-PMV RT recombinantly in E. coli were unsuccessful. Despite testing multiple bacterial strains, codon-optimized RT constructs and variants with modified N-and C-termini, none of these approaches yielded a soluble full-length protein.
These limitations prompted us to use the Bac-to-Bac baculovirus expression system. The gene encoding full-length M-PMV RT (Figure 2a, upper panel) was inserted into the pFASTBac vector and expressed in Sf9 insect cells. Recombinant RT was then purified using immobilized metal affinity chromatography (IMAC), followed by size-exclusion chromatography. To enhance solubility, the full-length construct contained C-and N-terminal tags for maltose-binding protein (MBP) and 6 Â His, respectively, each separated from RT by a TEV cleavage site (His-TEV-RT-TEV-MBP). The solubility and stability of RT were also improved by the addition of trehalose to all purification buffers. In the absence of trehalose, the binding of His-TEV-RT-TEV-MBP to the HisTrap HP column was insufficient, likely due to steric hindrance of the His-tag.
Approximately 1 mg of purified full-length RT with >90% purity was obtained from a 500 mL Sf9 suspension culture, as confirmed by SDS-PAGE and Coomassie staining (Figure 2a, lower panel). The purified full-length 65 kDa RT was subsequently used to produce a polyclonal guinea pig antibody, which enabled RT detection in M-PMV virions. Western blot analysis of virions produced in HEK293 cells revealed two distinct RT-specific bands at 65 kDa (p65) and 51 kDa (p51) (Figure 2b). While the full-length product was expected, the presence of a smaller band was unanticipated. This difference indicated the proteolytic removal of the 14 kDa RNase H domain.
To test proteolytic processing directly, purified fulllength RT was incubated with M-PMV PR13 protease at 4 C, 20 C, and 37 C for time intervals ranging from 30 min to 24 h (Figure 2c). No cleavage was observed at 4 C, and only faint cleavage products appeared at 20 C, even after prolonged incubation at this temperature. In contrast, incubation at 37 C resulted in efficient proteolysis, with two distinct bands corresponding to p65 and truncated p51. After 24 h, both products accumulated to similar levels.
Subunit characterization was performed using N-terminal sequencing and electrospray ionization mass spectrometry (ESI-MS). Edman degradation of the excised p65 and p51 bands yielded the same N-terminal sequence [I-D-I-L-A-P-Q], indicating that they originated from the same precursor (Figure 2d). The molecular masses measured by ESI-MS were 64,875 Da and 51,209 Da for p65 and p51, respectively, which is consistent with the cleavage site at the [NNAL#LVFT] site between the polymerase and RNase H domains (Figure 2d).
To further assess heterodimeric formation, we generated a truncated RT construct carrying RT lacking the RNase H domain (RTΔH) and co-expressed this form with full-length RT (His-TEV-RT-TEV-MBP) in Sf9 cells. Purification via the N-terminal His tag, which is present only in full-length RT, co-isolated both the p65 and p51 subunits. These were further purified by SEC, yielding a sample that contained both p65/p51 subunits (Figure 2e, second lane). These results demonstrate that M-PMV RT undergoes viral protease-mediated processing to generate p65 and p51 subunits, which likely form heterodimeric p65/p51 complexes.
To assess the structural stability of RT and RTΔH, we analyzed both forms using nano-differential scanning fluorimetry (nanoDSF) combined with backreflection turbidimetric measurements. The melting temperature (T M ) indicates a point at which 50% of the protein population is unfolded, whereas the temperature of the aggregation onset (T Agg ) corresponds to the temperature at which detectable protein aggregation begins. Full-length RT exhibited a slightly higher T M (by $1 C) but a lower T Agg (by $3 C) compared to RTΔH (Figure 2f). Notably, the reduced T Agg observed for full-length RT likely reflects the presence of aggregation-prone species rather than the intrinsic stability of the native protein. This interpretation is supported by the unfolding and turbidity profiles shown in Figure S1, which indicate early aggregation preceding complete unfolding in the homodimeric RT.
## 2.3 | Oligomerization of M-PMV reverse transcriptase
Our data thus far have shown that the proteolytic processing of full-length RT generates two distinct subunits, p65 and p51. Because the co-expressed p65 and p51 proteins co-purified as a complex, we hypothesized they assemble into a heterodimeric RT. To test this directly, we analyzed the oligomeric state of both fulllength RT and RT cleaved by viral PR13 (Figure 3a) using sedimentation velocity analytical ultracentrifugation. Full-length RT alone (purple) sedimented as a single, symmetric peak at 6.06 S ($117 kDa), consistent with a homodimer (Figure 3b). Protease alone (dark blue) showed a sharp peak at 1.47 S ($13 kDa), corresponding to its monomeric form. In the RT + PR sample (light blue), the RT peak shifted to 5.59 S ($107 kDa), and the original $6.1 S homodimer peak was absent, indicating complete conversion. The other peaks observed in this sample corresponded to the added PR ($1.6 S), cleaved RNase H domain ($1.0 S), and probably also a residual amount of the cleaved MBP tag ($3.2 S), visible also on the SDS-PAGE gel at $40 kDa (Figure 3a). No free RT subunits were detected, confirming their complete dimeric assembly.
The molecular weights predicted from the measured sedimentation coefficients did not precisely match the theoretical molecular weights of either the 130 kDa homodimer or the 116 kDa heterodimer; however, their relative difference of $10 kDa points toward cleavage of a single RNase H domain only, not both. Also, no signals corresponding to monomeric RT (65 kDa) or RTΔH (51 kDa) were detected, thus both cleaved and uncleaved RTs must exist exclusively in dimeric form. The full-length uncleaved RT forms a homodimer composed of two p65 subunits that sediment at $6.1 S. Upon protease cleavage, RT is converted into a heterodimer consisting of one p65 and one p51 subunit. This interpretation is supported by the apparent shift in the sedimentation coefficient distribution profile: cleaved RT displayed a single, distinct peak at $5.6 S, indicating the formation of a homogeneous species. Together with the ESI-MS data, these findings strongly support the existence of a stable p65/p51 heterodimer. Importantly, analysis of virion-derived RT revealed two distinct bands corresponding to p65 and p51, confirming that protease-mediated processing and heterodimer formation also occur during viral maturation.
$$F I G U R E 3 (a)$$
## 2.4 | Enzymatic activity of M-PMV reverse transcriptase
The enzymatic activity of M-PMV RT was initially assessed using a quantitative PCR (qPCR)-based assay with a fixed amount of RNA template and SYTO-9 detection, which allowed quantification of cDNA synthesis. All reactions were performed with equal amounts of RNA, with a non-template control and in-house murine RT as a positive control. To test the direct impact of proteolytic maturation on RT polymerase activity, full-length RT was incubated with M-PMV PR13, and the reaction was monitored for 24 h. Proteolytic removal of one RNase H domain significantly increased the reverse transcriptase polymerase activity. The cleaved RT showed enhanced reverse transcription activity, with 38.2 ± 18.9% of the initial signal after 2 h and 295.9 ± 28.3% after 24 h, compared to the non-cleaved form, as calculated by the relative quantification method (Figure 4a, left panel). No RT activity was detected in the PR13 protease sample, with Cq values comparable to the non-template control, indicating that neither the protease itself nor any proteins potentially present in the sample contributed to reverse transcription.
In contrast, the addition of PR13 to the RT + RTΔH complex modestly increased RT polymerase activity. This construct retained six additional residues at the N-terminus and nine at the C-terminus, following tag removal during purification. These residues were likely cleaved by the M-PMV protease, which enhanced RT's activity. The relative activity of the RT + RTΔH complex increased by approximately 42.0 ± 7.1% after 24 h of incubation with PR13 compared to that of the untreated control (Figure 4a, right panel).
The polymerase activity of recombinant RT and RT + RTΔH was further assessed using a colorimetric assay based on DNA synthesis on a poly(A)Á(dT)₁₅ template/ primer by incorporating digoxigenin-and biotin-labeled nucleotides. The full-length RT exhibited an activity of 4.9 ± 0.8 U μg À1 , whereas RT + RTΔH showed slightly higher activity at 6.1 ± 0.6 U μg À1 (Figure 4b). For comparison, the HIV-1 RT provided with the assay kit displayed an activity of >5 U μg À1 . These results confirm that both M-PMV RT variants are catalytically active, with RT + RTΔH exhibiting moderately increased activity compared to the full-length enzyme.
In addition to polymerase activity, we also measured RNase H activity. As the full-length RT homodimer contains two RNase H domains, whereas proteolytic maturation yields a heterodimer with only one, we investigated whether the loss of one RNase H domain affected endoribonuclease activity. Michaelis-Menten analysis of RNase H activity using an 18-nt RNA/DNA duplex revealed a K M of 0.48 μM (95% CI: 0.37-0.63 μM) and V max of 0.135 μMÁs À1 (95% CI: 0.117-0.159) for RT (Figure 4c). RT + RTΔH yielded a similar K M estimate of 0.45 μM (95% CI: 0.23-1.11) with a V max of 0.109 μMÁs À1 (95% CI: 0.076-0.200). The RT + RTΔH enzyme was assayed for Mg 2+ dependence, showing minimal activity in the absence of Mg 2+ and a dose-dependent increase with an optimum at 10-12.5 mM concentration (Figure 4d).
Together, these results demonstrate that proteasemediated maturation of M-PMV RT enhances reverse transcription activity, whereas both full-length and truncated enzymes retain robust polymerase and RNase H activities.
## 2.5 | Effect of cleavage-site mutations between the polymerase and RNase H domains on RT processing, activity, and M-PMV infectivity
To investigate whether the proteolytic processing of RT could be impaired, we introduced mutations into the cleavage site between the polymerase and RNase H domains. As retroviral proteases preferentially recognize substrates containing hydrophobic residues at the scissile bond (represented in M-PMV by two leucines, L447 and L448, Figure 2d), we designed six substitutions targeting only the P1 and P1' positions. The mutations L447I and L447P were selected based on previous observations that HIV-1 protease disfavors β-branched amino acids and proline immediately upstream of the cleavage site (Tözsér, 2010). Substitutions L447Q, L447R, and L448T introduced polar residues, which are generally unfavorable for protease recognition. Finally, L448G was designed to introduce a small residue at P1', as proteases generally avoid small side chains at positions directly adjacent to the cleavage site. All mutations were introduced into the full-length proviral pSARM construct containing the complete M-PMV genome, as well as into the pSARM-GFP vector, which lacks envelope proteins but expresses GFP enabling quantitative assessment of infectivity by flow cytometry.
Western blot analysis of M-PMV capsid protein (CA) levels in HEK 293 cell lysates and corresponding virions demonstrated comparable CA abundance across all variants (Figure 5a, upper and middle panels), indicating that neither viral polyproteins nor particle assembly and release were substantially affected by any of the introduced mutations. Analysis of RT processing revealed that none of the mutations fully blocked protease-mediated cleavage (Figure 5a, lower panel). The L447I, L447Q, and L448T mutants exhibited processing patterns similar to the wild-type (wt) RT. In contrast, the L447P and L447R variants exhibited a shifted p51 band, suggesting that cleavage at the canonical site was impaired and that alternative cleavage occurred at a nearby position. For the L448G mutant, the p65 precursor was barely detectable, with p51 predominating, indicating altered processing.
Functional assessment of RT activity of released mutant viruses normalized by ELISA showed a pronounced reduction for nearly all mutants except L447I and L448T (Figure 5b). The most severe defect was observed in the L448G mutant, which retained <1% of wt RT activity. Measurements of relative M-PMV infectivity correlated closely with the RT activity data: L447I and L448T displayed minimal reductions, whereas all other mutants, particularly L447P and L448G, exhibited near-complete loss of infectivity.
## 3 | DISCUSSION
Retroviral reverse transcriptases exhibit notable diversity in their oligomeric states and structural organization. Among the retroviral genera, HIV-1 RT is the most extensively studied and is well-known for its heterodimeric structure consisting of p66 and p51 subunits. However, the oligomeric states and maturation mechanisms of RT from other genera remain poorly characterized. Betaretroviruses, such as Mason-Pfizer monkey virus, are considered to possess monomeric RTs, largely because of limited structural and biochemical data. Understanding the structure and maturation of M-PMV RT is important not only for basic virology but also for expanding the scope of comparative retrovirology. In this study, we aimed to bridge this gap by providing detailed structural and biochemical characterizations of M-PMV RT, focusing on its oligomeric state, proteolytic processing, and enzymatic function.
We successfully expressed full-length recombinant M-PMV RT using the baculovirus expression system in Sf9 insect cells (Figure 2a, lower panel). Expression attempts in E. coli, even with codon optimization, failed due to extensive degradation, suggesting that M-PMV RT may require eukaryotic chaperones or folding machinery for proper stability. This contrasts with HIV-1 RT, where both p66 and p51 subunits are efficiently produced and reconstituted in E. coli (Müller et al., 1989;Stahlhut et al., 1994). Although our initial aim was to express full-length RT alone, based on the detection of two distinct RT-derived proteins in M-PMV virions, we performed its co-expression with the RTΔH variant (Figure 2e), during which we observed markedly improved stability of the truncated protein. When expressed individually, RTΔH was highly prone to degradation; however, co-expression with full-length RT led to protein stabilization. Notably, the RTΔH variant carried no affinity tag, yet it co-purified with His-tagged full-length RT using IMAC, suggesting that the two proteins form a stable complex during expression. Moreover, after tag cleavage, full-length RT and RTΔH were readily separated from the MBP tag by size-exclusion chromatography, despite the slight size difference between RTΔH ($51 kDa) and MBP ($40 kDa). This unexpectedly efficient separation further suggests that RTΔH is incorporated into a complex rather than existing as a free monomer.
These data correlate nicely with the detection of two, p65/p51 distinct RT-derived bands in M-PMV viral particles (Figure 2b), suggesting the presence of a heterodimer. To our knowledge, no previous study has directly demonstrated heterodimer formation in betaretroviruses. Although Křížov a et al. (Křízov a et al., 2012) mentioned a possibility of heterodimeric RT, their assumptions were based on the presence of a 50 kDa construct reactive to G-patch antibody and not on the reverse transcriptase itself. However, our ESI-MS data showed that neither p65 nor p51 retains the N-terminal G-patch domain after proteolytic cleavage in vitro. Notably, these results are derived from recombinant protein processing, and the presence or absence of G-patch in virion-derived RT remains uncertain and difficult to compare directly to HIV-1 RT, which lacks this domain entirely. Detection of two RT subunits in M-PMV virions provided direct evidence of proteolytic processing during viral maturation.
To further analyze whether M-PMV RT undergoes a protease-mediated maturation process that parallels the heterodimer formation seen in HIV-1 RT, we performed in vitro processing of RT by the viral protease. Proteolytic cleavage using recombinant M-PMV protease (PR13) generated two major products: a $65 kDa full-length subunit (p65) and a $51 kDa truncated subunit (p51) (Figure 2c) sharing the same N-terminal sequence (Figure 2d), resembling the p66/p51 profile seen in HIV-1 RT (Schulze et al., 1991). Even after 24 h, cleavage was incomplete, indicating that a conformational constraint may prevent M-PMV protease from processing the remaining RNase H domain. Additionally, the cleavage site between polymerase and RNase H domains [NNAL#LVFT] revealed by ESI-MS aligns with known retroviral protease specificity, which typically targets two types of sites: one featuring a proline immediately upstream of the scissile bond, and another characterized by two large hydrophobic residues flanking the cleavage site (Tözsér, 2010). In the case of M-PMV RT, the latter motif applies, suggesting that the protease recognizes a conserved structural motif shared with other retroviral substrates.
Sedimentation velocity analysis provides direct biophysical evidence for the structural transition of M-PMV RT upon proteolytic cleavage. Analytical ultracentrifugation revealed a shift from a homodimer to a heterodimer following cleavage with M-PMV protease, with no evidence of free monomers in the post-cleavage sample (Figure 3b). While the molecular weights predicted from the measured sedimentation coefficients were slightly underestimated compared to the theoretical masses, SDS-PAGE analysis of the same samples (Figure 3a) confirmed the presence of both cleaved and uncleaved forms in roughly equal proportions. These results support a model in which proteolytic processing converts a homodimeric precursor into a stable heterodimer, with one subunit lacking the RNase H domain. Notably, while a study by Müller et al., 1989 indicated incomplete dimerization of p66, the analytical ultracentrifugation data presented here showed complete dimerization of homodimeric RT, with no detectable monomeric subunits. This difference may reflect the expression system used, as M-PMV RT was produced in insect cells, whereas early studies of HIV-1 RT were performed using E. coli.
Proteolytic maturation increases the structural stability of M-PMV RT. Comparison of the homodimeric RT and the heterodimeric RT + RTΔH using nano-differential scanning fluorimetry combined with turbidity assays revealed differences in their thermal behavior. The heterodimer displayed a higher aggregation temperature relative to the RT homodimer, indicating increased resistance to aggregation (Figure 2f). Although the apparent melting temperature of the homodimeric RT was slightly higher, the unfolding curves showed that this signal was dominated by the unfolding of aggregated species rather than the native protein itself, underscoring that aggregation precedes unfolding in the homodimeric form.
Proteolytic cleavage significantly enhances the enzymatic activity of M-PMV RT. Activity assays revealed an approximately 3-fold increase in the activity of the cleaved RT compared to the uncleaved form after 24 h (Figure 4a, left panel). The RT + RTΔH variant also showed an increase, although to a lesser extent (Figure 4a, right panel). One possible explanation is the presence of additional residues at the N-and C-termini that remained after tag removal during purification and reduced activity upon cleavage by M-PMV protease. A more likely explanation is that the recombinantly prepared heterodimer does not strictly consist of 50% full-length enzyme and 50% RTΔH; even a small proportion of homodimer could affect the measured activity. Previous studies on HIV-1 RT reported 2.5-fold (Müller et al., 1989) and 4-fold (Stahlhut et al., 1994) increases in activity for the heterodimeric p66/p51 form compared to the homodimeric p66, which is in line with our observations. Comparison of enzyme activities using the colorimetric assay revealed that RTΔH was $25% more active than full-length RT (Figure 4b). However, this assay uses a different substrate and is primarily intended for quantifying RT concentration in samples (e.g., virions), so variability in the results is expected.
To further characterize enzymatic activities, we performed an RNase H fluorescence-based assay. Kinetic analysis showed that M-PMV RT cleaves RNA/DNA duplex substrate with a K m of 0.48 μM, closely matching the value obtained for RT + RTΔH (0.45 μM) (Figure 4c). These similar affinities may indicate that only one RNase H domain is catalytically active in both the heterodimer and homodimer. Reported K m values for HIV-1 RNase H vary considerably, ranging from $25 ± 3 nM under similar conditions using a fluorescence-based assay (Parniak et al., 2003) to values in the micromolar range in studies using a gel-based assay (Sevilya et al., 2003). M-PMV RT also showed a strong dependence on Mg 2+ , with negligible activity in the absence of Mg 2+ and maximal activity at 10-12.5 mM Mg 2+ (Figure 4d).
The accurate proteolytic release of RT from the Gag-Pro-Pol polyprotein precursor and the correct cleavage at the polymerase-RNase H junction are both required for the functionality of RT derived from the virus. Our results demonstrate that mutational disruption of the polymerase-RNase H cleavage site in M-PMV RT does not prevent protease-mediated processing. This finding is consistent with the previous report on HIV-1 RT cleavage-site mutants (Abram & Parniak, 2005). In that study, extensive substitutions throughout the polymerase-RNase H cleavage junction also failed to block processing. Rather than accumulating as p66 homodimers, these substitutions led to populations of virions containing unstable RT, dominated by p51 or degraded products. Consistent with these findings, none of our P1 or P1' mutations fully blocked the polymerase-RNase H cleavage (Figure 5a, lower panel). While substitutions L447I, L447Q, and L448T preserved wild-type-like cleavage patterns, the altered electrophoretic mobility of the p51 subunit in the L447P and L447R mutants indicates that disruption of the canonical scissile bond can redirect proteolysis to alternative nearby cleavage sites. In the case of the L448G mutant, this substitution appears to substantially disturb proper RT folding, likely exposing both cleavage sites on the p65 subunits within the homodimeric precursor. Although RT-derived products with molecular weights similar to p65 and p51 were still detected, most cleavage-site mutants exhibited severe reductions in enzymatic activity and viral infectivity (Figure 5b). These results suggest that proper enzyme activity requires precise folding in conjunction with accurate, site-specific processing at the polymerase-RNase H junction.
Sequence similarity alone does not predict the oligomeric architecture of retroviral reverse transcriptases. Remarkably, M-PMV RT shares over 62% sequence identity with monomeric MMTV RT in fingers and palm subdomains yet adopts a dimeric structure more similar to HIV-1 RT, with which it shares only $32% identity (Figure 1a). This disconnect suggests that heterodimeric architecture may have evolved independently in different retroviral lineages, driven by similar functional pressures such as the need for asymmetric active sites or subunit stability. This structural similarity across distant viruses points to shared functional constraints beyond phylogenetic relationships.
One might ask why structure prediction tools such as AlphaFold were not used to resolve the oligomeric architecture of M-PMV RT? In fact, we did attempt in silico modeling using AlphaFold; however, viral proteins remain underrepresented in the training data, and the algorithm tended to return conformations resembling known structures, most notably HIV-1 RT. One of the predicted models of the M-PMV RT heterodimer was nearly identical to the p66/p51 heterodimer of HIV-1 RT. Moreover, depending on the random seed, Alpha-Fold produced multiple divergent models, indicating low prediction confidence. These inconsistencies illustrate the limitations of current AI-based prediction tools when applied to structurally uncharacterized viral proteins and further underscore the need for high-resolution structural methods, such as X-ray crystallography or cryo-EM, to determine the authentic architecture of M-PMV RT.
Taken together, our findings establish that M-PMV reverse transcriptase undergoes proteolytic maturation to form a stable, enzymatically active heterodimer. This mechanism was previously considered restricted only to lentiviral and alpharetroviral RTs. This contrasts with the prevailing view that betaretroviral RTs, such as MMTV, exist exclusively as monomers. Our results challenge long-held assumptions about RT architecture and suggest that heterodimeric maturation may be a broader, conserved feature among retroviruses. Highresolution structural studies of M-PMV RT will be essential to define its architecture and uncover lineagespecific features that could advance our understanding of retroviral diversity and evolution.
## 4 | CONCLUSION
Our study provides the first biochemical evidence that the reverse transcriptase of Mason-Pfizer monkey virus undergoes proteolytic maturation to form a functional heterodimer, similar to the well-characterized HIV-1 RT. We demonstrate that the cleaved heterodimeric form is enzymatically more active and that its formation depends on specific conditions, including protease activity at physiological temperature. Mass spectrometry, N-terminal sequencing, and analytical ultracentrifugation collectively support a model in which a homodimer is processed into a stable heterodimer. Furthermore, we have shown that precise folding combined with correct, site-specific processing is required for proper enzyme function. These results broaden our understanding of reverse transcriptase maturation across retroviral genera and offer new insights into the evolutionary relationships between distantly related RTs. By revealing unexpected structural parallels, this study contributes to a more complete view of retroviral diversity and may support the development of alternative RT enzymes for molecular biology applications.
## 5 | MATERIALS AND METHODS
## 5.1 | Amino acid sequence alignment
To compare the reverse transcriptases from different retroviral genera, amino acid sequences corresponding to the polymerase domain (fingers and palm subdomains) were retrieved from the UniProt database (Consortium 2024). Specifically, the following entries were used: HIV-1 (P04585, residues 1-235), MMTV (P11283, residues 1-234), BLV (P25059, residues 1-233), ALV (Q7SQ98, residues 1-189), and MLV (P03356, residues 1-189). The M-PMV RT sequence was obtained from the pSARM vector containing the M-PMV genome (Kohoutov a et al., 2009). All sequences were aligned using the MAFFT algorithm via Jalview v2.11.2.5 with default parameters (MAFFT [web service]), and pairwise sequence identities were calculated within Jalview (Troshin et al., 2011;Troshin et al., 2018).
## 5.2 | Preparation of expression vectors for in vitro protein production
The expression vectors encoding full-length RT (6 Â His-TEV-RT-TEV-MBP) and RT variant lacking the RNase H domain (RTΔH) were constructed using the Bac-to-Bac™ baculovirus expression system (Invitrogen, USA). The RT coding sequence for both full-length RT and RTΔH was cloned from the pSARM vector into the pFastBac HT vector. Used primers: 5 0 Xho His TEV (AAACTCG AGATGTCGTACTACCATCACCATCACC), 3 0 RTstop KpnI (TTTGGTACCTTAAGCCACGATTTTAGTTGCCAA GTCAGCCCGTTGGTTGCCTTGAGCTATGGG) and FOR RT dH pACM4 NdeI (TTTTCATATGATTGACATAC TTGCAC), REV RT dH pACMV4 KpnI (TTTGGTACC TCATAAGGCATTGTTT) for RTΔH. The pFastBac vectors containing RT sequences were transformed into Escherichia coli DH10α. The isolated bacmid DNA was then transfected into Sf9 insect cells to generate an infectious viral stock. To prepare M-PMV PR13 protease, the pSIT vector (Veverka et al., 2003) was used. All constructs were verified by Sanger sequencing.
## 5.3 | Protein production and purification
a. Sf9 cells were kept at 27 C in Sf-900 II medium (Invitrogen, USA). For protein expression, two strategies were used: (i) expression of 6 Â His-TEV-RT-TEV-MBP alone, and (ii) co-expression of full-length RT with RTΔH to mimic heterodimer formation. In both cases, Sf9 cells were transfected with bacmid DNA, and baculovirus-containing supernatants were harvested 5 days post-transfection. The baculoviral stock was subsequently amplified to achieve optimal infectivity. Suspensions of Sf9 cells (3 Â 10 6 cells ml À1 ) were infected with the amplified baculovirus at a 1:100 v/v ratio. After 48 h, cells were harvested by centrifugation (3000Âg, 15 min).
The harvested cell pellets were resuspended in lysis buffer (25 mM Na₂HPO₄, pH 7.4; 500 mM NaCl; 250 mM trehalose; 2 mM MgCl₂; 20 mM imidazole; 2 mM β-mercaptoethanol) and homogenized using a One Shot Cell Disruptor (Constant Systems, United Kingdom) at 1.0 kbar. The lysates were clarified by centrifugation at 50,000Âg for 30 min, and the supernatants were loaded onto a HisTrap™ HP column (Cytiva, USA) pre-equilibrated with the same lysis buffer. For the full-length RT sample, the column was additionally washed with high-salt buffer (25 mM Na₂HPO₄, pH 7.4; 1 M NaCl; 0.5 M urea; 250 mM trehalose; 2 mM MgCl₂; 20 mM imidazole; 2 mM β-mercaptoethanol) prior to elution. In both cases, the proteins were eluted using a 30-500 mM imidazole gradient.
To remove the 6 Â His-TEV and TEV-MBP tags, TEV protease was added at a 1:20 w/w ratio, and the eluate was dialyzed overnight at 4 C against 2 L of dialysis buffer (25 mM HEPES, pH 7.4; 200 mM NaCl; 50 mM trehalose; 2 mM MgCl₂; 2 mM β-mercaptoethanol). Following dialysis, samples were centrifuged at 20,000Âg for 20 min, concentrated, and subjected to size-exclusion chromatography (HiLoad 26/600 Superdex 200 pg., Cytiva, USA) in SEC buffer (25 mM Na-HEPES, pH 7.4; 200 mM NaCl; 25 mM trehalose; 2 mM MgCl₂; 5% glycerol; 1 mM TCEP). Fractions containing purified RT were pooled, concentrated, and stored at À80 C. b. The vector encoding the active M-PMV PR13 protease was expressed in E. coli BL21 (DE3) in LB medium supplemented with 100 μg ml À1 ampicillin.
Protein expression was induced with 1 mM IPTG when cultures reached an OD 600 of 0.9-1.1, followed by a 2-h expression at 37 C. Cells were harvested by centrifugation, and PR13 was isolated from inclusion bodies similarly to the previously described method (Rumlov a et al., 2014). The cell pellet was resuspended in buffer A (50 mM Tris, pH 8; 50 mM NaCl; 1 mM EDTA), followed by two washes with buffer B (50 mM Tris, pH 8; 1 M NaCl; 1 mM EDTA) and a final wash with buffer A. After each resuspension, samples were sonicated and centrifuged (10,000Âg, 10 min). The final pellet was solubilized in 2 mL of 67% (v/v) acetic acid in aqueous solution at room temperature for 60 min, followed by sonication. The solution was then slowly dripped into 10 mL of cold water. The mixture was dialyzed against water for a maximum of 120 min, and then overnight against dialysis buffer (50 mM phosphate, pH 5.8; 1 mM EDTA; 0.05% β-mercaptoethanol). The solution was filtered through a 0.22 μm filter, concentrated, and subjected to size-exclusion chromatography using a HiLoad 26/600 Superdex 200 pg. (Cytiva, USA) column equilibrated with the same buffer. Fractions containing purified PR13 were pooled, concentrated, and stored at À80 C.
## 5.4 | Proteolytic cleavage
Recombinant M-PMV RT (final concentration 5.5 μM) was incubated with M-PMV PR13 protease in cleavage buffer (25 mM Na-HEPES, pH 7.4; 200 mM NaCl; 25 mM trehalose; 2 mM MgCl₂; 2 mM β-mercaptoethanol) at a 1:1 molar ratio. Reactions were incubated at 37 C, 20 C, or 4 C for 30 min, 1 h, 2 h, 4 h, or 24 h.
Proteolytic cleavage products were analyzed by SDS-PAGE.
## 5.5 | Thermal unfolding assay
RT and RTΔH were diluted to a concentration of 0.5 mg ml À1 . Thermal unfolding parameters were then determined using a combination of nano-differential scanning fluorimetry (nanoDSF) and turbidimetry using Prometheus Panta (Nanotemper, Germany). Protein solutions were heated at 1.
## 5.6 | Protein sequencing
Purified reverse transcriptase was incubated with the M-PMV PR13 protease overnight at 37 C in the cleavage buffer described in the proteolytic cleavage. Cleavage products were separated on SDS-PAGE and electroblotted onto a PVDF membrane. N-terminal amino acid sequences were determined by Edman degradation using the Procise Protein Sequencing System (Applied Biosystems, 491 HT or 494 cLC Protein Sequencer, USA). In this method, the free amino group of the N-terminal amino acid reacts specifically with phenylisothiocyanate. The derivatized amino acid is then selectively cleaved from the polypeptide and converted into a stable phenylthiohydantoin (PTH) derivative, while the rest of the peptide remains intact. Each cycle removes one N-terminal residue. The resulting PTH-amino acids are sequentially analyzed by reversed-phase HPLC, allowing determination of the N-terminal amino acid sequence of the protein or peptide.
## 5.7 | ESI MS
Samples including full-length RT, cleaved RT (RT + PR13), and PR13 protease were diluted with 0.1% formic acid to a 0.2 mg ml À1 concentration. 10 μL of the sample was injected onto a desalting column (MassPREP desalting, Waters, USA) and desalted by fast gradient (4 min) of acetonitrile in water with 0.1% formic acid. The separation was carried out by an LC system (I-class, Waters, USA) coupled online to a mass spectrometer (Synapt G2, Waters, USA) to acquire the mass of the protein by electrospray ionization. The raw spectrum was subtracted and deconvoluted (MaxEnt 1, Waters, USA) to produce the final spectrum.
## 5.8 | Analytical ultracentrifugation
Sedimentation analysis, used to monitor the oligomeric state and associative behavior of biomacromolecules (Bl aha et al., 2022;Skořepa et al., 2020), was performed on the analytical ultracentrifuge Optima AUC (Beckman Coulter, USA) using an An50-Ti rotor and double-sector cells equipped with 12 mm Epon centerpieces (Beckman Coulter, USA). Samples in 25 mM HEPES, pH 7.2; 200 mM NaCl; 2 mM MgCl 2 ; 1 mM TCEP buffer were analyzed at 20 C and 42,000 rpm. The sedimentation velocity experiment was recorded as absorbance at 280 nm with 300 scans in 3-min steps. Suitable initial sample concentrations were calculated using Lambert-Beer's law, assuming a 1.2 cm optical path length and 0.2-1.0 desired absorbance value (M-PMV RT, 6.9 μM; M-PMV PR, 6.9 μM; and RT + PR equimolar, 1:1 (v/v) mixture). Buffer density, protein partial specific volumes, and particle dimensions were estimated in Sednterp (Philo, 2023). Data were analyzed with Sedfit (Schuck, 2000) using the c(s) continuous sedimentation coefficient distribution model and visualized in GUSSI (Brautigam, 2015).
## 5.9 | qPCR
Reverse transcription activity of cleaved RT was determined by RT-qPCR using an artificial target SLA RNA (5 0 -AGU UGU UAG UCU ACG UGG ACC GAC AAA GAC AGA UUC UUU GAG GGA GCU AAG CUC AAC GUA GUU CUA ACA GUU UUU U-3 0 ), where the recombinant protein served as the reverse transcriptase. The SLA RNA template was used at a final concentration of 100 nM, and the reverse transcriptase at a final concentration of 20 nM. The primers used were SLA primer (5 0 -AGGGGGGGGGGTAAAAAACTGTTAGAACTA-3 0 ) for the initial reverse transcription to produce cDNA, and primers SLA qPCR F (5 0 -TTGTTAGTCTACGTG-GACCGA-3 0 ) and SLA qPCR R (5 0 -AACTGTTAGAAC-TACGTT-3 0 ) to amplify the cDNA. qPCR reactions were conducted in 96-well plates on a QuantStudio™ 5 Real-Time PCR System (Applied Biosystems™, USA) using a Taq DNA polymerase-based master mix (Top-Bio SYTO-9 qPCR 2x Master Mix) under the following conditions: 15 min at 95 C; 40 cycles of 30 s at 95 C, 1 min at 60 C, and 2 min at 72 C. Measurement was performed three times in duplicates. RT activity corresponded to relative cDNA levels quantified using the comparative Cq method (ΔCq method) calculated as 2^À ΔCq . The activity of the non-cleaved RT was set to 100%, and the activities of other samples were expressed relative to this value.
## 5.10 | Colorimetric RT assay
Reverse transcriptase activity was quantified by colorimetric assay (Roche Applied Science, Cat.
No. 11468120910) with exogenous poly(A)/oligo(dT) as the substrate/primer, according to the manufacturer's instructions. Reverse transcription was carried out for 1 h at 37 C. The reaction product was detected by measuring absorbance at 405 nm. All samples, including standards and controls, were assayed in three replicates. Negative controls without RT and blank wells containing only substrate were included to account for background signal.
## 5.11 | Fluorescence-based RNase H activity assay
RNase H activity was quantified using a fluorescence-based assay in black 96-well plates, with fluorescence measured on a SpectraMax iD5 microplate reader (Molecular Devices, USA). The substrate was a synthetic 18-nt RNA/DNA hybrid composed of a 3 0 -FAM-labeled RNA strand (5 0 -CACCAG-CUCCGUAGUAGC-[FAM]-3 0 ) annealed to a complementary 5 0 -BHQ1-labeled DNA strand (5 0 -[BHQ1]-CGATGATGCCTCGACCAC-3 0 ). Oligonucleotides were annealed in annealing buffer (50 mM HEPES, pH 8.0; 125 mM NaCl; 12.5 mM EDTA) by heating to 93 C for 3 min, followed by gradual cooling to room temperature. The resulting RNA/DNA substrate was aliquoted and stored at À80 C until use. Reactions (100 μL) were carried out in FL assay buffer (50 mM HEPES, pH 7.5; 60 mM KCl; 10 mM MgCl₂; 2.5 mM DTT) with 1 nM recombinant M-PMV RT (either RT or RT + RTΔH) and RNA/DNA hybrid at final concentrations ranging from 15 to 500 nM. Fluorescence was recorded every 4 min at 37 C for 1 h. Subsequently, an excess of enzyme was added to determine the maximum fluorescence (F max ). Initial velocities (v ₀ ) were obtained from the linear portion of the fluorescence increase (ΔF/Δt) for each substrate concentration. The relative slopes were normalized to the total fluorescence change (ΔF = F max À F o ) and multiplied by the respective substrate concentration to yield apparent rates in concentration units (μMÁs À1 ). These values were fitted to the Michaelis-Menten equation to estimate K m and V max .
## 5.12 | Preparation of expression vectors for M-PMV virion production in mammalian cells
The proviral M-PMV pSARM vector, as well as the pSARM-EGFP construct, was used to introduce singlepoint mutations into the cleavage site between the polymerase and RNase H domains. Based on the precisely defined sequence of the cleavage site [NNAL#LVFT], six mutants were generated: four containing amino-acid substitutions at the P1 position (L447I, L447P, L447Q, L447R) and two at the P1 0 position (L448G, L448T) (Table S1). The corresponding mutagenic oligonucleotide primers used for construct generation are listed in Table S2. Single-point mutations were introduced using the previously described EMILI method (Füzik et al., 2014). All constructs were verified by whole-genome sequencing.
## 5.13 | Cell cultures
HEK 293 cells were cultured in Dulbecco's modified Eagle medium (DMEM; Sigma, USA) supplemented with 10% fetal bovine serum (Sigma, USA) and 1% L-glutamine (Sigma, USA) at 37 C in a humidified atmosphere containing 5% CO₂. One day prior to transfection, cells were seeded at a density of 3 Â 10 5 cells mL À1 . The following day, cells were transfected with the M-PMV proviral pSARM vector (wt or mutant constructs) using polyethylenimine at a DNAto-transfection reagent ratio of 1:2 (w/v). After transfection, cells were incubated for an additional 48 h and subsequently used for downstream analyses.
## 5.14 | M-PMV virion production
At 48 h post-transfection, virions released into the culture medium were harvested by ultracentrifugation through a 20% sucrose cushion at 39,000 rpm for 1.5 h at 10 C in an SW 40 Ti rotor (Beckman Coulter, USA). Viral pellets were resuspended in 60 μL of virion lysis buffer (50 mM Tris, pH 7.8; 80 mM KCl; 2.5 mM DTT; 0.75 mM EDTA; 0.5% Triton X-100) and reverse transcriptase activity was quantified as described in the qPCR methods section. Samples were later combined in a 1:1 (v/v) ratio in loading buffer (0.4 M Tris; 4% SDS; 10% β-mercaptoethanol; 24% glycerol; bromophenol blue) and analyzed using SDS-PAGE. The procedure was performed as described previously (Dost alkov a et al., 2024).
## 5.15 | Single-round infectivity assay M-PMV
The infectivity was determined as described earlier (Dost alkov a et al., 2020). Briefly, 48 h post-transfection, the culture media from HEK 293 cells transfected with pSARM-GFP (wt or mutant constructs) and pTMO vectors at a (w/v) ratio 1:1 were collected and filtered through a 0.45-μm membrane filter. M-PMV capsid protein content was determined by ELISA. The freshly seeded HEK 293 cells were infected with ELISA-normalized amounts of M-PMV particles and incubated for 48 h. The cells were fixed with 2% paraformaldehyde and transferred to a FACS tube. Quantification of GFP-positive cells was performed using a BD FACS Aria III flow cytometer (BD Biosciences, USA).
## 5.16 | Assessment of RT activity from released mutant viruses
Reverse transcriptase activity from the released virions was assessed using a previously described RT-qPCRbased assay as detailed in this study. Viral particles were collected by centrifugation and resuspended in 60 μL of virion lysis buffer. From each lysate, 1 μL was used as input for the RT-qPCR reaction. RT activity was quantified as described above. Measurements were performed using two independent biological replicates. To account for differences in viral particle production, the obtained RT activity values were normalized to the M-PMV capsid protein levels determined by ELISA.
## 5.17 | Flow cytometry
The infected cells were analyzed with a BD FACS Aria III flow cytometer (BD Biosciences, USA) with excitation at 488 nm and emission separated by a 530/30 band pass filter, as described earlier (Dost alkov a et al., 2020). The obtained data were analyzed with Diva 8 software (BD Biosciences, USA). Measurements were performed using two independent biological replicates.
AUTHOR CONTRIBUTIONS Marina Kapisheva: Conceptualization; writingoriginal draft; formal analysis; investigation; visualization. Petra Junkov a: Investigation; formal analysis. Ondřej Van ěk: Investigation; writingreview and editing. Zuzana Jalovcov a: Investigation. Ivana Křížov a: Methodology; investigation; formal analysis. Alžb ěta Dost alkov a: Methodology; writingreview and editing. Michaela Rumlov a: Conceptualization; supervision; funding acquisition; project administration; writingreview and editing.
## References
1. Abram, Parniak (2005) "Virion instability of human immunodeficiency virus type 1 reverse transcriptase (RT) mutated in the protease cleavage site between RT p51 and the RT RNase H domain" *J Virol*
2. Baltimore (1970) "Viral RNA-dependent DNA polymerase: RNA-dependent DNA polymerase in virions of RNA tumour viruses" *Nature*
3. Bl Aha, Skořepa, Cmunt et al. (2022) "Structure of the human NK cell NKR-P1: LLT1 receptor:ligand complex reveals clustering in the immune synapse" *Nat Commun*
4. Brautigam (2015) "Chapter five: calculations and publication-quality illustrations for analytical ultracentrifugation data" *Methods in Enzymology*
5. Coffin, Hughes, Varmus et al. (1997)
6. Das, Georgiadis (2004) "The crystal structure of the monomeric reverse transcriptase from Moloney murine leukemia virus" *Structure*
7. Dost, Křížov A I, Kapisheva et al. (2024) "Unveiling the DHX15-G-patch interplay in retroviral RNA packaging" *Proc Natl Acad Sci U S A*
8. Dost, Škach, Kaufman et al. (2020) "PF74 and its novel derivatives stabilize Hexameric lattice of HIV-1 mature-like particles" *Molecules*
9. Füzik, Ulbrich, Ruml (2014) "Efficient mutagenesis independent of ligation (EMILI)" *J Microbiol Methods*
10. Herschhorn, Hizi (2010) "Retroviral reverse transcriptases" *Cell Mol Life Sci*
11. Hizi, Herschhorn (2008) "Retroviral reverse transcriptases (other than those of HIV-1 and murine leukemia virus): a comparison of their molecular and biochemical properties" *Virus Res*
12. Huber, Betz, Marx (2023) "Reverse Transcriptases: from discovery and applications to xenobiology" *Chembiochem*
13. Kohoutov A, Sakalian, Hunter et al. (2009) "The impact of altered polyprotein ratios on the assembly and infectivity of Mason-Pfizer monkey virus" *Virology*
14. Křízov A, Hadravov A R, Doležal et al. (2012) "The G-patch domain of Mason-Pfizer monkey virus is a part of reverse transcriptase" *J Virol*
15. London (2016) "Structural maturation of HIV-1 reverse transcriptase-a metamorphic solution to genomic instability" *Viruses*
16. London (2019) "HIV-1 reverse transcriptase: a metamorphic protein with three stable states" *Structure*
17. Marx, Simian Retrovirus (2008) "Encyclopedia of virology. 3rd ed"
18. Menéndez-Arias, Sebasti An-Martín, Álvarez (2017) "Viral reverse transcriptases" *Virus Res*
19. Müller, Restle, Weiss et al. (1989) "Coexpression of the subunits of the heterodimer of HIV-1 reverse transcriptase in Escherichia coli" *J Biol Chem*
20. Parniak, Budihas, Le Grice et al. (2003) "A fluorescence-based high-throughput screening assay for inhibitors of human immunodeficiency virus-1 reverse transcriptaseassociated ribonuclease H activity" *Anal Biochem*
21. Perach, Hizi (1999) "Catalytic features of the recombinant reverse transcriptase of bovine leukemia virus expressed in bacteria" *Virology*
22. Philo (2023) "SEDNTERP: a calculation and database utility to aid interpretation of analytical ultracentrifugation and light scattering data" *Eur Biophys J*
23. Rumlov A, Křížov A I, Keprov A A, Hadravov A R, Doležal, Strohalmov A K (2014) "HIV-1 protease-induced apoptosis" *Retrovirology*
24. Schuck (2000) "Size-distribution analysis of macromolecules by sedimentation velocity ultracentrifugation and Lamm equation modeling" *Biophys J*
25. Schulze, Nawrath, Moelling (1991) "Cleavage of the HIV-1 p66 reverse transcriptase/RNase H by the p9 protease in vitro generates active p15 RNase H" *Arch Virol*
26. Sevilya, Loya, Adir et al. (2003) "The ribonuclease H activity of the reverse transcriptases of human immunodeficiency viruses type 1 and type 2 is modulated by residue 294 of the small subunit" *Nucleic Acids Res*
27. Skořepa, Pazicky, Kalouskov A B et al. (1998) "Natural killer cell activation receptor NKp30 oligomerization depends on its N-glycosylation" *Cancers (Basel)*
28. Stahlhut, Li, Condra et al. (1994) "Purification and characterization of HIV-1 reverse transcriptase having a 1:1 ratio of p66 and p51 subunits" *Protein Expr Purif*
29. Taube, Loya, Avidan et al. (1998) "Reverse transcriptase of mouse mammary tumour virus: expression in bacteria, purification and biochemical characterization" *Biochem J*
30. Temin, Mizutani (1970) "Viral RNA-dependent DNA polymerase: RNAdependent DNA polymerase in virions of Rous sarcoma virus" *Nature*
31. Tözsér (2010) "Comparative studies on retroviral proteases: substrate specificity" *Viruses*
32. Troshin, Procter, Barton (2011) "Java bioinformatics analysis web services for multiple sequence alignment-JABAWS:MSA. Bioinformatics"
33. Troshin, Procter, Sherstnev et al. (2018) "JABAWS 2.2 distributed web services for bioinformatics: protein disorder, conservation and RNA secondary structure" *Bioinformatics*
34. Veverka, Bauerov A H, Lang et al. (2003) "Three-dimensional structure of a monomeric form of a retroviral protease" *J Mol Biol*
35. Werner, Wöhrl (1999) "Soluble Rous sarcoma virus reverse transcriptases alpha, alphabeta, and beta purified from insect cells are processive DNA polymerases that lack an RNase H 3 0 ! 5 0 directed processing activity" *J Biol Chem*
36. Ruiz (2019) "Evolving understanding of HIV-1 reverse transcriptase structure, function, inhibition, and resistance" *Curr Opin Struct Biol*
37. Zheng, Perera, Mueller et al. (2015) "Asymmetric conformational maturation of HIV-1 reverse transcriptase" *Elife* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793510&blobtype=pdf | Joshua Foster-Tucker, Arnold Monto, Amy Callear, Rachel Truscon, Matthew Smith, Elie-Tino Godonou, Emily Martin
Background. Human parainfluenza viruses (HPIVs) cause significant annual acute respiratory illness (ARI) burdens in the U.S. Yet, HPIV circulation and ARI characteristics are not fully understood, especially for HPIV-4, which may be an underrecognized cause of moderate and severe ARI. We describe the circulation and ARI features of HPIVs 1-4 in the Household Influenza Vaccine Evaluation (HIVE) cohort over 12 years.
Methods. The HIVE study has followed a household cohort in Ann Arbor, MI, since fall 2010, with active study during fall-winter influenza seasons until July 2015 and year-round thereafter. HPIV ARIs may have been under-detected before July 2015, especially HPIV-3, which peaks in summer. This analysis used HIVE data from October 2010 to December 2022. Participants provided swabs for ARIs meeting our case definition of ≥2 ARI symptoms (e.g., cough & fever) with onset during the prior week, which were tested for HPIVs 1-4. Ill participants reported symptoms, medication use, and healthcare sought, with statistical comparisons adjusted for onset age.
Results. During follow-up, 13758 ARIs were reported, with 616 (4.5%) positive for ≥1 HPIV. HPIV-3 was identified in 39% of swabs, HPIV-2 in 23.9%, HPIV-1 in 20.1%, and HPIV-4 in 17.9%. Median ages at HPIV-1 (5.6 y) and -3 (5.8 y) ARI were younger than HPIV-2 (7.3 y) and -4 ARI (7.6 y). Children under age 12 accounted for 69.1% of all HPIV ARIs. Infection frequencies by sex and race were similar. Pre-pandemic circulation aligned with established seasonal patterns; all four HPIVs resumed circulation by fall 2022. The Table shows significant variability in medically attended ARI frequency across HPIVs (range, 17.3% HPIV-4 to 29.8% HPIV-1; p = .04). Medication use during HPIV-4 ARI was frequent. Significant variations in decongestant, acetaminophen, & other fever/pain reliever usage and coinfection occurrence were noted across HPIVs (all p ≤ .01), though major symptom group frequencies were similar (Table ).
Conclusion. During HIVE surveillance from October 2010 to December 2022, HPIVs were etiologic in 4.5% of ARIs. Symptoms were consistent across HPIVs, but healthcare-seeking, medication use, and coinfection frequencies differed significantly. Further study of HPIV ARI features in the community is needed as vaccine candidates enter late-stage trials.
Disclosures. Arnold Monto, MD, Roche: Advisor/Consultant S1336 • OFID 2026:13 (Suppl 1) • Poster Abstracts |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12514659&blobtype=pdf | # Assessing Influenza and SARS-CoV-2 Coinfection in Brazil: A Comprehensive Study of Patient Outcomes From 2020 to 2023
L Gaklik, | Carneiro, B Carneiro
## Abstract
Influenza and SARS-CoV-2 are major respiratory pathogens that have impacted global health, sharing similar transmission routes and clinical symptoms. The COVID-19 pandemic brought attention to coinfection with these viruses, which has been associated with worse clinical outcomes, but the full extent of this impact remains underexplored. As both viruses circulate together during seasonal outbreaks, understanding their coinfection dynamics is crucial for public health response. This retrospective observational study analyzed data from over 30 000 hospitalized patients sourced from the Brazilian Epidemiological Surveillance System (SIVEP-Gripe). Patients were classified into two groups: influenza mono-infection and influenza-SARS-CoV-2 coinfection. Descriptive statistics and multivariate logistic regression were performed to evaluate associations with primary (mortality) and secondary (ICU admission) outcomes. Among approximately 3.7 million reported severe acute respiratory syndrome cases, 35 831 were influenza-infected, with 1763 (4.9%) coinfected with SARS-CoV-2. Coinfected patients exhibited nearly double the risk of death (aOR: 1.87, 95% CI: 1.52-2.30) and a significantly higher likelihood of ICU admission (aOR: 1.27, 95% CI: 1.07-1.52), compared to those with influenza alone. Coinfected patients also presented with more severe respiratory symptoms and longer hospital stays. Coinfection with influenza and SARS-CoV-2 is associated with significantly worse clinical outcomes, including higher mortality and increased need for intensive care. Early identification and tailored management strategies for coinfected patients are essential to improving patient outcomes, particularly for those with underlying comorbidities.
| IntroductionViral infections are significant causes of respiratory tract diseases and, throughout history, have had a profound impact on the socioeconomic status and health of populations. For the past 30 years, infectious respiratory diseases have remained among the leading causes of mortality worldwide [1]. Until 2019, the influenza virus, responsible for the flu, was one of the leading causes of mortality from respiratory diseases in Brazil [2].In 2020, we experienced the COVID-19 pandemic, a disease caused by the SARS-CoV-2 respiratory virus initially identified in China in December 2019 [3]. Although they are phylogenetically distinct viruses, from a clinical point of view, the diseases caused by these two agents are similar. In the early stages of the pandemic, health teams faced challenges in distinguishing between influenza and COVID-19 based solely on symptoms. Accurate differentiation of the two infections required careful evaluation of their distinct characteristics, while also considering the potential for coinfection with both pathogens [4].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.
Since the first cases reported in 2019 and the subsequent pandemics, SARS-CoV-2 has continued to circulate in different variant forms, with a seasonality similar to that of other respiratory viruses. It is now widely accepted that SARS-CoV-2 is considered an endemic virus in the global landscape [5,6]. In the context of respiratory viruses, it is well established that multiple viruses co-circulate during certain seasons, often leading to cases of coinfection [7]. Depending on the specific viruses involved, these coinfections can lead to a worsening of patient outcomes when compared to single infections [8][9][10]. As SARS-CoV-2 and influenza are currently two of the most widely circulating viruses, the likelihood of co-infection between them is high.
In the context of influenza and SARS-CoV-2 co-infection, a study by Dao et al. [11] revealed that the proportion of coinfection in critically ill COVID-19 patients was greater than that in patients in the general population. This finding highlighted the importance of screening respiratory pathogens to detect coinfections. On the other hand, a study by Guan et al. [12] demonstrated the opposite findings: coinfection with these two viruses had no effect on overall mortality, and the risk of critical outcomes was lower in patients with co-infection. Additionally, different associations were found in different regions, which drew attention to the order in which the viruses coinfected patients.
Coinfections also represent an additional challenge during respiratory disease outbreaks, as they can overwhelm health systems, especially during peaks in the incidence of each disease, impacting public policies. Despite increasing reports of coinfections, the literature currently presents contradictory findings regarding the severity and clinical outcomes of co-infection between influenza and SARS-CoV-2, particularly in settings where both viruses are circulating simultaneously. This study, therefore, aims to explicitly compare the clinical outcomes of influenza mono-infection and influenza-SARS-CoV-2 co-infection. Additionally, it seeks to describe the epidemiological and clinical characteristics of the patients included in the study, addressing the need for a more consistent understanding of how these coinfections influence disease progression and patient management, with a large cohort of patients that may contribute to clarifying the conflicting results in the literature.
## 2 | Materials and Methods
## 2.1 | Study Design & Participants
This study is a retrospective observational analysis using the public and anonymized Severe Acute Respiratory Syndrome (SARS) database from the Influenza Epidemiological Surveillance Information System (SIVEP-Gripe). The study included patients who were registered in the SIVEP-Gripe database, with a confirmed laboratory diagnosis (molecular or serological) of influenza or SARS-CoV-2 infection. Inclusion criteria were: (a) a confirmed positive laboratory result for either influenza or SARS-CoV-2 and (b) symptom onset between January 2020 and December 2023. Patients who were only clinically or radiologically diagnosed, those who were not hospitalized, or those with missing data on variables required for determining inclusion or exclusion were excluded from the study.
## 2.2 | Data Source
The data for this study were sourced from the Influenza Epidemiological Surveillance Information System (SIVEP-Gripe, Sistema de Informação da Vigilância Epidemiológica da Gripe), a national surveillance database maintained by the Brazilian Ministry of Health. SIVEP-Gripe serves as the primary instrument for monitoring Severe Acute Respiratory Infections (SARI) throughout Brazil. Notification of all individuals hospitalized with SARI is mandatory for all public and private healthcare facilities, including clinics and hospitals across all complexity levels. This comprehensive and compulsory reporting framework ensures broad national coverage, making SIVEP-Gripe a robust and representative data source for analyzing SARI trends and outcomes in the country.
Within this system, a SARI case is defined as a patient presenting with at least two of the following symptoms: fever (even if referred), chills, sore throat, headache, cough, runny nose, olfactory or taste disturbances, and at least one severe symptom: dyspnea/respiratory distress, persistent chest pressure, O 2 saturation less than 95% on room air, or bluish discoloration of the face. For each notified case, the database includes sociodemographic data, comorbidities, clinical signs, risk factors, and the suspected etiological agent.
## 2.3 | Covariables
The SARS notification form includes more than 150 data fields covering various patient dimensions. For this study, we focused on the collection of sociodemographic information, comorbidities, symptoms, and risk factors. The information was organized into blocks according to this classification and analyzed according to this division. Sociodemographic information included variables such as sex, age, ethnicity, pregnancy status, prior influenza vaccination, and antiviral influenza treatment. Age was categorized into five groups: less than 1 year, 1-5 years, 5-16 years, 16-65 years, and older than 65 years. Pregnancy status was classified as yes or no at the time of hospitalization. Influenza vaccination status referred to whether the patient had received the vaccine in the previous vaccination campaign (within the last 12 months). Antiviral treatment was recorded if the patient had received influenza antiviral medications (Oseltamivir or Zanamivir) immediately before or during hospitalization.
The presence of any comorbidity was recorded, with individual categories including postpartum, heart disease, hematologic disease, Down syndrome, liver disease, asthma, diabetes, neurological disorders, chronic pulmonary disease, immunosuppression, kidney disease, and obesity. Symptoms collected included fever, cough, sore throat, dyspnea, respiratory distress, oxygen saturation < 95%, diarrhea, vomiting, abdominal pain, fatigue, anosmia, and ageusia.
Risk factors analyzed included time to hospitalization (the time between symptom onset and hospitalization), length of hospital stay (duration between hospitalization and case closure), time to case closure (discharge/recovery or death), intensive care unit (ICU) admission (whether the patient required ICU care), length of ICU stay (calculated as the time from ICU admission to case closure), type of ventilatory support (whether invasive or noninvasive support was required), and case closure (categorized as death, recovery, or death by other causes).
## 2.4 | Study Outcomes
The study assessed two primary clinical endpoints. The primary outcome was in-hospital mortality, defined as a case closure record indicating death directly attributed to the acute respiratory infection during the hospitalization period. The secondary outcome was the need for ICU admission, a binary variable indicating whether a patient required ICU care at any point during their hospitalization.
## 2.5 | Data Processing and Analysis
Data were extracted from annual raw files that recorded all SARS notifications during the study period and are available in the Datasus database (https://opendatasus.saude.gov.br/). Owing to the large volume of data, the reading and preliminary analysis of the data were carried out through a script developed in Python, using the pandas module for sorting and organizing the information.
The data were initially filtered according to previously established inclusion criteria and were classified as monoinfection or coinfection. The selected variables were subsequently summarized.
For the identification of duplicate data, two complementary approaches were employed. The first utilized the duplicated method from the pandas library, which compares data rows sequentially to identify entries that are identical across all columns. The second approach applied a hashing procedure using the SHA-256 algorithm via Python's hashlib library. Each record was encoded and processed with SHA-256 to generate a unique 64-character hexadecimal hash. Identical records produced identical hashes, while even minimal differences in the data resulted in distinct outputs. Both methods were executed to ensure comprehensive detection of duplicate entries.
Given that most of our data consists of binary categories, we assessed outliers only for the continuous variables. To detect outliers, we used the interquartile range (IQR) method based on the median, and values were considered acceptable if they were within 1.5 times the IQR above and below the quartiles (25th and 75th). Additionally, for the variables based on dates (time to hospitalization, length of hospital stay, and ICU length of stay), we checked for impossible values (e.g., outside the study period) and potential typographical errors. If any errors were detected in any variable, the corresponding entry was excluded from median calculations and univariate analysis for that category.
We chose not to exclude any entries after applying the inclusion and exclusion criteria. To address the inconsistencies and missing data in the SIVEP database, missing or inconsistent values for binary variables were classified as "Unknown." For continuous variables, missing or inconsistent values were handled as described above. Only for the multivariate analysis were rows with incomplete data for any of the variables included in the analysis excluded.
## 2.6 | Statistical Analysis
To assess the impact of coinfection with influenza and SARS-CoV-2, patients were divided into two groups: monoinfected and coinfected. Initially, a descriptive analysis of the data, including absolute (n) and relative (%) frequencies of categorical variables and means, medians, and IQR of continuous variables, was performed. The distribution of the data was verified via the Kolmogorov-Smirnov test. The Mann-Whitney test was used to compare continuous variables, whereas Pearson's chi-square test was used to compare categorical variables between patients diagnosed with simple influenza virus infection and patients coinfected with SARS-CoV-2. For all analyses, the level of statistical significance used was 5%. The analyses were conducted via the Python scipy library.
To assess the impact of the variables on the primary (death) and secondary (ICU admission) outcomes, multivariate logistic regression was performed. Variables that had a significant result (p < 0.05) in the univariate analysis, had data completeness greater than 70%, and had clinical relevance to the study objective were included in both models. The variables were included in steps and added incrementally, analyzing the impact on the AIC (Akaike information criterion) value.
The data from the multivariate analysis were summarized and used to assemble forest graphs, which represent the different impacts of the variables included in the model on the primary and secondary outcomes of the research. The representations were created via the forestplot and matplotlib modules of the Python language.
## 2.7 | Ethical Considerations
This is a publicly available database that has been provided in an anonymous form, with no possibility of individual identification. Therefore, in accordance with national legislation, registration of this project with the Research Ethics Committee is not required.
## 3 | Results
In this study, we analyzed the effect of coinfection between the influenza virus and SARS-CoV-2 based on public data made available by the Influenza Epidemiological Surveillance System (SIVEP-Gripe). The analysis revealed that coinfection with influenza and SARS-CoV-2 significantly increased the risk of mortality and ICU admission compared to influenza monoinfection. During the period analyzed, approximately 3.7 million notifications of SARS were registered. Among these notifications, the influenza virus (FLU) was identified in 34 566 cases (< 1%), being the exclusive etiological agent in more than 91% of these cases (32 803/35 831). In the context of coinfections, SARS-CoV-2 was the most common combination, occurring in 1763 notified patients (58.2%), followed by respiratory syncytial virus (19.2%, 582/3028) and human rhinovirus (3.7%, 111/3028).
In terms of sociodemographic characteristics (Table 1), most individuals in the coinfected group were male, whereas most of the monoinfected individuals were women, indicating a significant variation in the patient profile between the groups. With respect to age, the most frequent category in our data set was individuals older than 65 years, representing 35.2% of the cases. However, a difference was observed in relation to coinfection status: most coinfected patients were aged 16-65 years, whereas single infections were more prevalent among individuals over 65 years of age.
Regarding influenza vaccination in the previous year, only approximately 10% of hospitalized patients in both groups had been vaccinated, with no significant variation between the groups. Notably, however, there was a high rate of incomplete data (> 60%) for this variable. Also, approximately one quarter of the patients who were monoinfected with influenza used antivirals during hospitalization, while among the coinfected patients, this percentage was significantly lower.
In terms of race, more than 40% of the patients in both groups declared themselves to be white, with no statistically significant differences between the groups. Among women, less than 10% of the cases in both groups involved pregnant women, with similar frequencies between monoinfected and coinfected individuals.
Between the patients included in the study, the most common symptoms were cough, fever, dyspnea, and respiratory distress, all of which were observed in more than 50% patients in both groups. A significant difference was observed in the presence of some symptoms between the groups. Fever, cough, and vomiting were more frequently observed in individuals infected with only the influenza virus. In contrast, more severe symptoms, such as dyspnea, respiratory distress, and low oxygen saturation, were more prevalent in coinfected patients, suggesting a possibly greater severity of the disease in this group (Supporting Information S1: Table 1).
The greatest difference between the groups was observed in the symptoms of anosmia and ageusia, which are characteristic of SARS-CoV-2 infection. The frequency of these symptoms in coinfected patients was more than twice as high as that in the monoinfected group, which suggests that coinfection can result in a sum of the typical symptoms of each infection (Supporting Information S1: Table 1).
With respect to risk factors, at least half of the patients in both groups had at least one factor, and this frequency was greater among coinfected patients. Among the most observed factors were heart disease, which was present in almost a quarter of the patients in both groups, and diabetes, which was observed on average in 15% of the individuals included in this study. Notably, among the risk factors significantly different between the groups, only asthma was more common in the simple infection group. In contrast, conditions such as heart disease, diabetes, immunosuppression, and obesity were more frequently observed among coinfected patients (Supporting Information S1: Table 2).
With respect to the variables associated with the main morbidity and mortality factors in this study, a statistically significant difference was observed in most of the variables compared. The median interval between the onset of the first symptoms and hospitalization was only 3 days for coinfected patients, whereas for monoinfected patients, this interval was almost 50% longer (5 days, p = 0.000). The length of hospitalization, measured as the difference between the day of admission and the outcome (cure or death), was also greater in coinfected patients (Table 2).
In addition, the need for admission to ICUs and ventilatory support was more common among coinfected patients, and the greater demand for invasive ventilatory support in this group was especially noteworthy. Finally, the death rate was almost double that of coinfected patients, indicating greater disease severity in these individuals (Table 2).
To assess the impact of additional variables on the outcome of death, a multivariate logistic regression analysis was performed.
The results revealed that coinfected patients were almost twice as likely to die as patients with isolated influenza virus infection were. This analysis also revealed that other variables, such as the presence of risk factors, age over 65 years, dyspnea, and O₂ saturation < 95%, also increased the risk of death by nearly two times or more for coinfected patients (Figure 1). In the same model, the analysis of the secondary outcome-the need for admission to the ICU-indicated that coinfected patients were almost 20% more likely to be admitted to the ICU than those infected with only the influenza virus. In addition to coinfection, other variables also increased the likelihood of ICU admission, including the presence of risk factors, respiratory distress, and O₂ saturation < 95% (Figure 2).
The results of this study clearly demonstrated that hospitalized patients coinfected with influenza and SARS-CoV-2 viruses had greater clinical severity, a higher incidence of associated risk factors, and worse clinical outcomes than those monoinfected with influenza. These findings suggest a synergistic interaction between the viruses studied, contributing to the deterioration of the clinical picture and increasing the risk of ICU admission and mortality.
## 4 | Discussion
In this study, we analyzed the clinical-epidemiological profile and outcomes of hospitalized patients with simple influenza virus infection compared with those of patients coinfected with influenza and SARS-CoV-2. Coinfection was clearly associated with a significant increase in the severity of cases, corroborating previous findings reported in the literature [13][14][15][16]. Specifically, coinfected patients had a greater demand for invasive ventilatory support, prolonged hospitalizations, and worse clinical outcomes than noninfected patients did, confirming a possible synergistic interaction between the viruses.
An important observation of this study was the shorter interval between the onset of symptoms and hospitalization of the coinfected patients. This rapid clinical progression suggests that coinfection induces a more intense and rapid inflammatory response, as demonstrated by Kinoshita et al. [17], who reported a significant increase in serum IL-6 levels in patients coinfected with influenza A and SARS-CoV-2. This elevation of inflammatory cytokines may explain the rapid clinical deterioration and greater severity of respiratory symptoms observed in this group.
With respect to the immunological mechanisms involved, Kim et al. [15] reported that coinfection promotes severe lymphopenia, significantly impairing the adaptive immune response, especially the CD4 + T-cell response and the production of neutralizing antibodies. These immunological factors seem to contribute directly to the severity of coinfection, hindering the clinical recovery of patients and increasing the risk of mortality, corroborating the findings of this study.
A comparison of the symptoms observed in this study with those reported in other investigations revealed that specific clinical manifestations in influenza-infected patients, such as anosmia and ageusia, are highly suggestive of coinfection, especially due to the characteristic involvement of SARS-CoV-2. Gregianini et al. [18] reported that initial symptoms of coinfection can overlap with those of monoinfections, making immediate differential diagnosis difficult and increasing the importance of simultaneous testing and rigorous epidemiological surveillance. Despite the relatively low prevalence of coinfections described in several studies [19,21], the consensus is that coinfections significantly increase the risk of severe clinical outcomes. Yan et al. [16], through meta-analysis, demonstrated that coinfected patients are more than twice as likely to require invasive mechanical ventilation and admission to ICU than monoinfected patients are, results that are fully compatible with the conclusions obtained in this study.
In addition, observations in animal models and systematic reviews suggest that influenza and SARS-CoV-2 viruses can exert independent pathological effects, even when coinfecting the same organism. Kinoshita et al. [17] reported that although both viruses can replicate simultaneously in different lung areas without direct interference, coinfection exacerbates lung inflammation and prolongs the pathological process, contributing to more extensive tissue damage and slower clinical recovery.
Additional studies noted that the severity of coinfection also depends on age and pre-existing comorbidities [20,21]. In our study, the increased prevalence of comorbidities, such as diabetes, heart disease, obesity, and immunosuppression, among coinfected patients was notable, reinforcing the idea that these conditions play crucial roles in the susceptibility to and clinical severity of coinfection. These findings are especially relevant for clinical management, indicating that patients with these conditions require more aggressive monitoring and interventions early in the course of the disease.
Although patients older than 65 years demonstrated significantly higher adjusted odds of death, they had lower odds of ICU admission. This finding may reflect clinical decision-making processes where elderly patients, often with multiple comorbidities or frailty, are less likely to receive intensive care interventions due to considerations of prognosis, patient preferences, or resource allocation [22][23][24]. Consequently, a proportion of deaths in this age group may occur outside of the ICU setting.
The variability in clinical severity associated with the presence of comorbidities was also already reported [19], who highlighted a greater severity of respiratory and systemic symptoms in coinfected patients with risk factors such as advanced age and obesity. Although their study revealed that coinfected patients were generally younger and had fewer severe comorbidities, the importance of monitoring vulnerable subgroups with greater care remains evident.
Taken together, our results highlight the clinical significance of coinfections in hospitalized patients. The implementation of more comprehensive and sensitive diagnostic tests for viral and bacterial coinfections is urgently needed, as our findings demonstrate that patients coinfected with influenza and SARS-CoV-2 experience worse clinical outcomes compared to those infected with a single virus. Furthermore, these results support the reinforcement and targeted implementation of vaccination campaigns against both influenza and SARS-CoV-2, particularly within critical seasonal windows to maximize protection. Additionally, our study underscores the importance of early identification and risk stratification of coinfected patients, which could enable healthcare providers to prioritize intensive monitoring and tailored therapeutic interventions, ultimately improving patient prognosis and resource allocation in hospital settings.
A major limitation of this study was the considerable proportion of incomplete or missing data in the notification system, which is a frequent issue in retrospective studies using secondary data sources [25,26]. Additionally, molecular testing availability fluctuated throughout the period from 2020 to 2023, potentially impacting case identification. It is also likely that influenza case numbers were underestimated, given that many laboratories prioritized testing for SARS-CoV-2, especially at the beginning of the pandemic [27,28]. Moreover, influenza testing varied widely across Brazilian′s regions, which limited our ability and led to the decision not to conduct a geographic stratification analysis [29].
Another important limitation relates to the absence of information on our database about circulating viral strains and variants-for instance, SARS-CoV-2 variants such as Delta, Omicron, and their sublineages, as well as influenza A subtypes (H1N1, H3N2) and influenza B lineages-which could have influenced disease severity and patient outcomes. Key clinical variables such as vaccination coverage, antiviral or other treatment regimens, and the burden on healthcare resources were not comprehensively available and therefore could not be adjusted for in the analyses. Collectively, these factors may limit the generalizability of our findings and highlight the need for future studies to implement rigorous and standardized protocols for data collection and reporting, enabling more robust and representative analyses.
This study reinforces the perception that coinfection with influenza and SARS-CoV-2 in hospitalized patients is associated with greater clinical severity and a significant increase in hospital mortality. Strategies aimed at the early diagnosis of coinfections, the strengthening of vaccination campaigns, and the appropriate management of comorbidities are essential to mitigate the clinical and epidemiological impact of these concomitant viral infections.
## References
1. Naghavi, Vollset, Ikuta (2024) "Global Burden of Bacterial Antimicrobial Resistance 1990-2021: A Systematic Analysis With Forecasts to 2050" *Lancet*
2. Alexandrino, De Queiroz Xavier, Batista De Oliveira et al. (2022) "Morbimortalidade Por Doenças Do Aparelho Respiratório No Brasil: Um Estudo Ecológico" *Revista Ciência Plural*
3. Huang, Wang, Li (2020) "Clinical Features of Patients Infected With 2019 Novel Coronavirus in Wuhan"
4. Bai, Tao (2021) "Comparison of COVID-19 and Influenza Characteristics" *Journal of Zhejiang University-Science B*
5. Otto, Macpherson, Colijn (2024) "Endemic Does Not Mean Constant as SARS-CoV-2 Continues to Evolve" *Evolution; International Journal of Organic Evolution*
6. Naik, Avula, Palleti (2023) "From Emergence to Endemicity: A Comprehensive Review of COVID-19" *Cureus*
7. Diniz, Dias, Oliveira (2024) "Outcomes of SARS-CoV-2 and Seasonal Viruses Among 2 Million Adults Hospitalized for Severe Acute Respiratory Infection During the COVID-19 Pandemic in Brazil" *Journal of Infectious Diseases*
8. Li, Yu, Wang (2024) "Cocirculation and Coinfection of Multiple Respiratory Viruses During Autumn and Winter Seasons of 2023 in Beijing, China: A Retrospective Study" *Journal of Medical Virology*
9. Babawale, Guerrero-Plata (2024) "Respiratory Viral Coinfections: Insights into Epidemiology, Immune Response, Pathology, and Clinical Outcomes" *Pathogens*
10. Dias, Diniz, Oliveira (2024) "Outcomes of SARS-CoV-2 and Seasonal Viruses Among Children Hospitalized in Brazil" *Pediatrics*
11. Dao, Hoang, Colson et al. (2021) "Co-Infection of SARS-CoV-2 and Influenza Viruses: A Systematic Review and Meta-Analysis" *Journal of Clinical Virology Plus*
12. Guan, Chen, Li (2021) "Impact of Coinfection With SARS-CoV-2 and Influenza on Disease Severity: A Systematic Review and Meta-Analysis" *Frontiers in Public Health*
13. Swets, Russell, Harrison (2022) "SARS-CoV-2 Co-Infection With Influenza Viruses, Respiratory Syncytial Virus, or Adenoviruses" *Lancet*
14. 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*
15. Kim, Nguyen, Casel (2022) "Coinfection With SARS-CoV-2 and Influenza A Virus Increases Disease Severity and Impairs Neutralizing Antibody and CD4+ T Cell Responses" *Journal of Virology*
16. Yan, Li, Lei et al. (2023) "Prevalence and Associated Outcomes of Coinfection Between SARS-CoV-2 and Influenza: A Systematic Review and Meta-Analysis" *International Journal of Infectious Diseases*
17. Kinoshita, Watanabe, Sakurai et al. (2021) "Co-Infection of SARS-CoV-2 and Influenza Virus Causes More Severe and Prolonged Pneumonia in Hamsters" *Scientific Reports*
18. Gregianini, Salvato, Baethgen (2023) ") Infection Followed by Separate COVID-19 Infection" *Revista Panamericana de Salud Pública*
19. Pawlowski, Silvert, O'horo (2022) "SARS-CoV-2 and Influenza Coinfection Throughout the COVID-19 Pandemic: An Assessment of Coinfection Rates, Cohort Characteristics, and Clinical Outcomes" *PNAS Nexus*
20. Liang, Wang, Lin (2024) "Influenza and COVID-19 Co-Infection and Vaccine Effectiveness Against Severe Cases: A Mathematical Modeling Study" *Frontiers in Cellular and Infection Microbiology*
21. Golpour, Jalali, Alizadeh-Navaei et al. (2025) "Co-Infection of SARS-CoV-2 and Influenza A/B Among Patients With COVID-19: A Systematic Review and Meta-Analysis" *BMC Infectious Diseases*
22. Adeyemi, Hill, Siman et al. (2025) "Acute Care Use and Prognosis in Older Adults Presenting to the Emergency Department" *Journal of Pain and Symptom Management*
23. Ding, Lian, Wang (2021) "Management of Very Old Patients in Intensive Care Units" *Aging and Disease*
24. Sinuff, Kahnamoui, Cook et al. (2004) "Rationing Critical Care Beds: A Systematic Review" *Critical Care Medicine*
25. Correia, Padilha, Vasconcelos (2014) "Métodos Para Avaliar a Completitude Dos Dados Dos Sistemas De Informação Em Saúde Do Brasil: Uma Revisão Sistemática" *Ciência & Saúde Coletiva*
26. Ribas, Custódio, Toledo et al. (2020) "Completeness of Notifications of Severe Acute Respiratory Syndrome at the National Level and in a Regional Health Care in the State of Minas Gerais, During the COVID-19 Pandemic" *Epidemiologia e Servicos De Saude: Revista Do Sistema Unico De Saude Do Brasil*
27. Dhanasekaran, Sullivan, Edwards (2022) "Human Seasonal Influenza Under COVID-19 and the Potential Consequences of Influenza Lineage Elimination" *Nature Communications*
28. Staadegaard, Del Riccio, Wiegersma (2021) "The Impact of the SARS-CoV-2 Pandemic on Global Influenza Surveillance: Insights From 18 National Influenza Centers Based on a Survey Conducted Between" *Influenza and Other Respiratory Viruses*
29. Barberia, Boing, Gusmão (2023) "An Assessment of the Public Health Surveillance Strategy Based on Molecular Testing During Three Major Pandemic Waves of COVID-19 in Brazil" *PLOS Global Public Health* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12208009&blobtype=pdf | # An Epidemic of Respiratory and Ocular Infections Caused by the Reemergence of a Recombinant Human Adenovirus, the Novel Type HAdV-B114 (P7H3F3)
Tina Ganzenmueller, Magnus Wolf, | Wolfram, | Charikleia Gkioule, Lars Steinbrück, Albert Heim
## Abstract
Human adenoviruses of species B (HAdV-B) can cause upper respiratory tract infections and conjunctivitis, but also severe lower respiratory tract infections (LRTI). Although HAdV-associated LRTI are non-notifiable in Germany, typing data of our Adenovirus Reference Laboratory indicated an HAdV-B3 epidemic in 2023, with 67 samples initially typed as HAdV-B3 compared to < 10/year in the previous years. Circulation of a novel, highly virulent HAdV-B3 strain was suspected and complete viral genomic sequencing performed, revealing a recombinant phylogeny of the penton gene (P), which originated from HAdV-B7, whereas hexon (H) and fiber (F) genes originated from HAdV-B3. Therefore, this virus was acknowledged by the Adenovirus Working Group as the novel recombinant genotype 114 (P7H3F3). Interestingly, BLAST search of the HAdV-B114 prototype sequence showed 99.91% identity to the old HAdV-B genome type 3a. Additionally, multiple complete adenovirus genomic sequences labeled as HAdV-B3 during the last two decades had > 99.8% identity, suggesting long-term circulation of HAdV-B114 although the recombinant phylogeny of its penton region had not been recognized. This detailed analysis of an HAdV epidemic associated with ocular and respiratory infections, including severe LRTI, led to the discovery of a novel genotype HAdV-B114, which is rather a neglected, re-emergent than an emerging virus.
| BackgroundHuman adenoviruses (HAdV) are non-enveloped DNA viruses and belong to the genus Mastadenovirus within the family Adenoviridae. HAdV types (up to type 51) were initially defined and typed by neutralization testing ("serotypes"). Sanger sequencing of the hexon gene (imputed serotyping by partial sequencing of the region encoding the neutralization epitope of the hexon capsid protein) is still an effective method in many diagnostic laboratories, but cannot correctly type novel, recombinant genotypes [1][2][3].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.
The multitude of HAdV types is classified into species HAdV A to G, initially based on hemagglutination properties and subsequently on molecular phylogeny. An important mechanism of adenovirus evolution is homologous recombination between types of the same HAdV species, which can lead to the emergence of novel HAdV types ("genotypes") [4,5]. With the advent of high-throughput sequencing (HTS), adenovirus genotyping has been established based on three gene regions encoding the major viral capsid proteins: penton (P), hexon (H), and fiber (F) [2,6]. This approach can be used for the discovery of new HAdV types (genotypes) that require confirmation by the Human Adenovirus Working Group, as well as for accurate diagnostic typing of all serotypes and genotypes.
HAdV species B includes (among others) HAdV-3, -7, -14, -21, -55, and -66 sharing a tropism for the respiratory tract. These viruses cause infections of the upper respiratory tract infections (URTI), such as common cold and pharyngoconjunctival fever (PCF), but also more severe lower respiratory tract infections (LRTI), for example, pneumonia and acute respiratory distress syndrome (ARDS) even in immunocompetent patients [7][8][9]. HAdV-B55 and HAdV-B66 were associated with recent outbreaks of severe LRTIs but had been typed initially as HAdV-B11 or HAdV-B7, respectively, due to their recombinant phylogeny with neutralization epitope sequences of the latter viruses [1,[10][11][12][13][14]. Therefore, a recent epidemic of LRTIs and ocular infections in Germany caused by the rather endemic type HAdV-B3, raised the suspicion of the emergence of a novel HAdV-B genotype retaining a HAdV-B3 neutralization epitope sequence in a recombinant genome which may result in enhanced virulence and transmissibility.
Here we describe clinical and phylogenetic features of this virus, which has been accepted by the Human Adenovirus Working Group (http://hadvwg.gmu.edu/) as the novel recombinant genotype HAdV-B114.
## 2 | Methods
## 2.1 | Patients, Samples, and Study Design
In 2023, 79 HAdV-B-DNA-positive specimens (including basic clinical information such as LRTI, PCF, or conjunctivitis) were provided by diagnostic laboratories from all over Germany for HAdV typing in our Adenovirus Reference Laboratory. In total, 85% (67/79) of these samples were typed by imputed serotyping as HAdV-B3. These were investigated in more detail by complete genomic sequencing (or Sanger sequencing of fiber and penton genes if complete genomic sequencing failed). Specimen types included secretions/swabs from the upper or lower respiratory tract, bronchoalveolar lavages, eye swabs, blood, and stool. Medical data from a subset of LRTI and conjunctivitis patients infected with HAdV-B114 (initially typed as HAdV-B3) and treated at the University Hospital Tuebingen were collected to characterize the clinical features more detailed. In addition, archived DNA samples available from 31 of 48 HAdV-B3 positive diagnostic specimens sampled between 2007 and 2021 were sequenced to determine whether HAdV-B114 circulated before 2023.
This study was approved by the Ethics Committee of the University Hospital Tuebingen (No. 774/2023BO2).
## 2.2 | Genotyping by Partial Sequencing of the Penton, Hexon, and Fiber Gene
Routine diagnostic molecular typing (imputed serotyping) was performed by partial hexon gene sequencing [15]. Briefly, HAdV-DNA was amplified by conventional generic PCR targeting the sequence of the immunogenic loop 2 of the ε determinant of the hexon gene [16]. Subsequently, fiber and penton genes were sequenced as previously described [15,17], if complete genomic sequencing was not feasible. Cycle sequencing (Applied Biosystems) was performed in both directions using the same primers as for PCR [17,18].
## 2.3 | Viral Genome Sequencing and De Novo Assembly
HTS of adenoviral genomes was performed as previously described [19]. Briefly, DNA was extracted from HAdV-positive clinical specimens or cell culture isolates on A549 cells (Supporting Information S1: Table S1) on a QIAcube (Qiagen Blood Kit) and sequencing libraries were prepared using the NEBNext Ultra II FS DNA Library Prep Kit for sequencing on an Illumina MiSeq. After quality control of raw data, human reads were removed, the remaining reads were trimmed with fastp and de novo assembled with SPAdes, Minia3, or GATB tools [19]. After scaffolding of the genomes and several correction steps (Pilon and GATK), genome annotation was done using Geneious Prime. A more methodological description can be found in Supporting Information S1: Supplementary Data.
## 2.4 | Phylogenetic Analysis
A multiple alignment of whole genome sequences was constructed by using fast Fourier transforms (MAFFT) (https:// www.ebi.ac.uk/Tools/msa/mafft/) with default gap parameters. Pairwise alignment, comparisons, and visualization of genomes were performed on BioEdit version 7.2.0 (http://www.mbio. ncsu.edu/BioEdit/page2.html). Bootstrapped, maximum likelihood phylogenetic trees with 500 replicates were constructed using RAxML version 8.2.11 with the command line options: -m GTRGAMMA, -f a, -x 1, -N 500, -p 1. MEGA v12 software was used to visualize the trees. Similarity plots and recombination detection (bootscan approach) were performed using Simplot software (version 3.5.1) with a window size of 1000 bp and a step size of 200 bp. In addition to the novel HAdV-B114 sequence, complete genomic nucleotide sequences representing all prototypes of the HAdV-B species, genome type HAdV-B3a and HAdV-B7 vaccine strain were used in the analysis (GenBank accession numbers: OR853835, PQ189748, PQ189750, KF268123, JX423381, AY599834, JN860676, AY594255, MN936178, AY601636, JN860678, AY163756, KT970441, LC177352, ON393912, AY803294, FJ643676, AY601633, KF633445, AY737797, AY128640, KF268328, AY737798).
Virtual restriction fragment length polymorphism (RFLP) patterns for the complete genomic sequences of HAdV-B114 and genome type 3A were generated using the online tool Restriction Analyzer (https://molbiotools.com/restrictionanalyzer.php).
## 3 | Results
## 3.1 | HAdV-B3 Infections From 2012 to 2023 in Germany
HAdV-associated respiratory infections are a non-notifiable disease in Germany but surveillance is voluntarily performed by diagnostic laboratories, which send samples of severe cases to the HAdV Reference Laboratory for HAdV typing by imputed serotyping. These results from 2012 to 2023 are provided in Figure 1. The observed "epidemic" of HAdV-B3 (and to a far lesser extent of HAdV-B7) in 2023 was associated with symptoms of severe LRTI, URTI, conjunctivitis and/or PCF as indicated on the typing requests. In addition, HAdV-DNAemia and shedding in stool samples, suggesting disseminated infection, were observed in several patients (see also details of typical cases below). All specimens initially typed as HAdV-B3 were subsequently re-typed as HAdV-B114 (see below) either by complete genomic sequencing using HTS or by additional Sanger sequencing of the fiber and penton genes.
## 3.2 | Complete Genomic Sequencing
Complete viral genome sequences were generated for 22 of the 67 diagnostic specimens that had initially been typed as HAdV-B3 (imputed serotyping by partial hexon sequencing). HTS from the original samples (n = 6) or HAdV isolates (n = 16) resulted in 4.4 × 10 6 (median, range 1.4 × 10 6 -8.3 × 10 6 ) sequence reads per sample with an average sequence read coverage of 1153 (median, range 19-19 660) per sample; please see Supporting Information S1: Table S1 for more details. Sequenced specimens originated from all over Germany: five from patients from Tuebingen (GenBank #OR853835, selected as representative HAdV-B114 prototype sequence, #PQ189738, #PQ189747, #PQ189748 (variant 1), #PQ189749), five from Hannover (#PQ189753, #PQ189754, #PQ189752, #PQ189755, #PQ189756), four from Ulm (#PQ189741, #PQ189742, #PQ189743, #PQ189744), two from Munich (#PQ189740, #PQ189746), and single samples from Bonn (#PQ189751), Heidelberg (#PQ189739), Landsberg (#PQ189737), Oberhausen (#PQ189736), Regensburg (#PQ189745), and Stuttgart (#PQ189750, variant 2). All these sequences were almost identical to the HAdV-B114 prototype sequence with the highest pairwise distance (0.004) found in variant 2 (#PQ189750).
## 3.3 | Phylogenetic Analysis Revealed the Novel Recombinant HAdV Type B114
Phylogenetic trees of complete genomic sequences, as well as of the hexon, fiber, and penton genes (Figure 2) depicted the clustering of the "2023 epidemic HAdV-B3" sequences with the HAdV-B3 prototype in the hexon and fiber genes, but in the penton gene with the HAdV-B7 prototype (most closely related serotype) and HAdV-B66 (most closely related genotype). Clustering of the HAdV-B114 penton sequence with the HAdV-B7 prototype and HAdV-B66 was confirmed by constructing and bootstrap testing of UPGMA, Neighbor-Joining, Minimum Evolution, and Maximum Parsimony trees (data not shown) in addition to the Maximum Likelihood tree presented in Figure 2B. Nucleic acid identities of the penton gene between the HAdV-B114 and the HAdV-B7 and HAdV-B66 prototype sequences were 1625/1635 (99.39%) and 1630/1635 (99.69%), respectively, but the HAdV-B114 sequence compared to the HAdV-B3 prototype sequence had only 1611/1635 (98.53%) identities. Bootscan analysis confirmed the observed recombinant phylogeny (Figure 3). Both the bootscan and the penton trees rather indicated a recombination with HAdV-B66 than with the HAdV-B7 prototype, but HAdV-B66 itself is a recombinant genotype (P7H7F3) owing a penton gene derived from FIGURE 1 | Overview of HAdV types detected in samples submitted to the German Adenovirus Reference Laboratory from 2012 to 2023. All typing results (based on partial HAdV hexon gene sequencing) have been summarized for each year. HAdV-B3 and -B7 types are highlighted in blue and orange, respectively. All HAdV-B3 typing results of 2023 have to be revised as HAdV-B114 after penton base gene sequencing or complete genomic sequencing.
HAdV-B7. Consequently, the recent HAdV-B3 isolates were designated as the novel recombinant genotype HAdV-B114 (P7H3F3) by the Human Adenovirus Working Group (http:// hadvwg.gmu.edu/). We also analyzed in detail the phylogeny of the early gene regions E1-E4, which encode nonstructural genes. With the exception of the E2 genes, HAdV-B114 sequences were also most closely related to HAdV-B66 (Supporting Information S1: Figure S1).
## 3.4 | Identification of Previous HAdV-B114 Isolates (Prior to 2023) and Identity to Genome Type 3a
An initial NCBI Blast analysis of the HAdV-B114 prototype sequence #OR853835 revealed 103 complete genomic sequences from the last two decades with > 99.8% identity in GenBank (Supporting Information S2: Supplementary Document 1). Most of these originated from clinical isolates from the USA, China, and Japan and had been labeled as HAdV-B3, according to the then valid taxonomy. All these sequences clustered with HAdV-B7 in the penton gene (see highlighted hits in Supporting Information S3: Supplementary Document 2) and can therefore be considered as HAdV-B114 sequences, although a few were erroneously labeled as P3H3F3 instead of P7H3F3. This shows that HAdV-B114 circulated in China, Japan, and the USA prior to 2023.
GenBank sequence #JX423381.1 (99.91% identity to the HAdV-B114 prototype sequence #OR853835) had a more detailed description as variant (genome type) 3a and eight other GenBank sequences had "3a" in their strain designation, although a more detailed description was missing (#KF268123.1, #KF268133.1, #KF268120.1, #JX423380.1, #JX423381.1, #JX423382.1, #OR753121.1, #KF268133.1). This suggested that HAdV-B114 is identical to the previously described genome type 3a. As genome types (not to be confused with genotypes) had been defined by band patterns in RFLP analysis, we performed in silico RFLP genome typing for the HAdV-B114 prototype sequence (#OR853835), which confirmed a band pattern almost identical to genome type 3a (Supporting Information S1: Figure S2).
As no complete genomic sequence data of German HAdV-B3 isolates prior to 2023 were available in GenBank, we (partially) sequenced the penton gene region from archived DNA extracts of diagnostic samples which had previously been typed as HAdV-B3 by imputed serotyping. Archived DNA from 31 of 48 diagnostic samples from 2007 to 2021 was still available for penton gene sequencing. All of these samples had a hexon sequence clustering with HAdV-B3 and a penton sequence clustering with HAdV-B7. Thus, it can be considered that they should be re-typed as HAdV-B114.
## 3.5 | Amino Acid Substitutions in the Penton
Base Gene Compared to the HAdV-B3 Prototype
The deduced HAdV-B114 amino acid sequence of the RGD loop in the penton base revealed a complete identity with both the HAdV-B7 and HAdV-B66 sequences but not with HAdV-B3 (Figure 4). This loop of the penton base binds with its RGD sequence motif to the secondary cellular receptors, integrins, and thus the sequence of the RGD loop may influence the affinity and tropism to different integrins [7]. Of note, very close to the RGD motif at positions 329-331, HAdV-B3 has a negatively charged glutamate (D) at position 326, whereas the positively charged asparagine (N) is found at this position in HAdV-B114, -B66, and -B7. In addition to this crucial amino acid substitution, a multiple alignment of the complete penton base prototype sequences of HAdV-B114, HAdV-B3, HAdV-B7, and HAdV-B66 (deduced amino acid sequences) revealed only a few other substitutions. Both HAdV-B114 and HAdV-B66 had a common I32L substitution compared to HAdV-B7 and HAdV-B3, which corresponded well to the phylogenetic analysis on the nucleic acid sequence level (see above). All HAdV-B114 sequences (except PQ189750) featured a unique A462T FIGURE 3 | Bootscan analysis of the complete genomes of the putative HAdV-B3 (i.e., HAdV-B114) isolate revealed its recombinant phylogeny (P7H3F3). The penton gene region of the novel HAdV-B114 originates from HAdV-B7 (green colored line). Bootscan indicated HAdV-B66 (gray colored line), but HAdV-B66 is itself a recombinant and has a penton gene derived from HAdV-B7. The hexon gene derives from HAdV-B3 (blue colored line). The bootscan is not conclusive in the fiber gene region, because HAdV-B3 and -B66 share an almost identical fiber sequence. substitution compared to the sequences of HAdV-B3, HAdV-B7, and HAdV-B66, but this substitution is distant from the RGD motif. Furthermore, a subset of HAdV-B114 sequences (#PQ189748, #PQ189753, #PQ189752, #PQ189755, #PQ189741) had a unique E319K substitution.
## 3.6 | Clinical and Virological Characteristics of Patients With Severe Disease Associated With HAdV-B114
At the University Hospital Tübingen, three patients with systemic HAdV-B114 disease had been treated in spring 2023: two with rather severe LRTI and one case of adenovirus-associated thrombocytopenia with cerebral venous sinus thrombosis (clinical details summarized in Table 1). Both pneumonia patients were immunocompromised due to allogeneic hematopoietic stem cell transplantation. Although not requiring mechanical ventilation, they presented with fever, dyspnea, and high viral loads in the lower airways (Ct-value 20-24 upon HAdV real-time PCR) and significant DNAemia (5.2 × 10 3 and 7.6 × 10 4 copies/mL plasma, respectively), typical for a disseminated adenoviral disease. Both shed HAdV into the feces and presented with leukopenia. One as well had elevated liver enzymes, potentially indicating an adenoviral hepatitis. Noteworthy, the third case of severe systemic HAdV-B114 disease presented as adenovirus-associated thrombocytopenia and thrombotic events in a 7-year-old child. This case has recently been reported in detail by others [20]; however, without typing and recognizing its association to HAdV-114. Note: Viral loads (copies/mL or Ct-value) were assessed using the Adenovirus R-gene (Biomerieux) or the Panther Fusion AdV/hMPV/RV Assay (Hologic) real-time PCR assays for plasma and stool samples or respiratory specimens, respectively. The case of HAdV-B114-associated thrombocytopenia and thrombotic events in a 7-year-old child has recently been reported in detail by others [20]; however, the recombinant phylogeny of HAdV-114 had not been recognized at that time, therefore, it was included in this table. Abbreviations: ALAT, alanine aminotransferase; BAL, bronchoalveolar lavage; ICU, intensive care unit.
## 3.7 | Characteristics of Patients With
HAdV-B114-Associated Eye Infections and Accompanying URTI Supporting Information S1: Table S2 shows the clinical data of 16 patients with conjunctivitis who were treated in the ophthalmologic out-patient clinic of the University Hospital Tuebingen between February 2023 and June 2023. With one exception (suspected mild keratitis), no involvement of the corneal structure was observed by the treating ophthalmologists. The most frequently reported symptoms were conjunctival redness, foreign body sensation, pruritus, and mild ocular pain. Three patients reported concurrent symptoms of URTI or fever, although this information was not routinely assessed and may be underestimated. It is suspected that these patients suffered from PCF. Noteworthy, 9 of 16 patients reported a recent history of similar cases of conjunctivitis and/or common cold with eye involvement in their household. We detected high HAdV loads in the conjunctival swabs by PCR (2.4 × 10 5 to up to > 1.0 × 10 7 copies HAdV-DNA/mL). HAdV-B114 infection was confirmed by genotyping (sequencing of penton, hexon, and fiber genes), and in three cases, the complete HAdV genome was sequenced. In summary, these data indicate an epidemic of PCF-like illness with relatively frequent intrafamilial spread in Southern Germany at that time.
## 4 | Discussion
Increased circulation of adenovirus was observed worldwide in 2022 and 2023 [21][22][23][24][25]. Noteworthy, the majority of respiratory adenovirus infections during this time were published to be caused by HAdV-B3 [21,26,27], although the vast majority of these were probably HAdV-B114 infections according to the published penton sequences (Supporting Information S3: Supplementary Document 2).
Previous SARS-CoV-2 pandemic measures (e.g., the closing down of kindergartens) also limited transmission of endemic adenoviruses and had probably resulted in an exceptionally large group of immune-naïve infants. Similar to other viruses [28], the postpandemic reemergence of adenoviruses, such as epidemics of HAdV-F41, and other HAdV types (e.g., HAdV-B114), was highly likely and has been indeed observed [21][22][23][24]29]. Adenoviruses may then have spread to older immune-naïve or otherwise highly susceptible patients, for example, immunosuppressed persons. In addition, increased awareness and behavioral changes in the general population after the pandemic may have led to increased rates of diagnosis. Although precise epidemiologic data on respiratory adenoviruses in Germany are lacking, as these are non-notifiable, the increase in initial HAdV-B3 typing results (later retyped as HAdV-B114) was impressive (Figure 1).
To a lesser extent, increased numbers of the closely related type HAdV-B7 were also detected in 2023 (Figure 1). Despite early reports on severe LRTI cases and eye infections caused by HAdV-B3 in the 1950s-1970s [30][31][32][33], HAdV-B7 was considered to be associated with more severe LRTI and outbreaks in military barracks, and was therefore included in the only available adenovirus vaccine, developed in the 1970s, whereas HAdV-B3
was not [34,35]. Subsequently, genomic subtyping of clinically or epidemiologically relevant HAdV-B7 and -B3 isolates by RFLP was introduced and more frequently applied to HAdV-B7 due to its superior clinical significance [36][37][38][39]. For example, genome types 7a-7g were identified in isolates originating from 1958 to 1981 in addition to the prototype (7p) originating from 1954 [36]. Genome type 7h was identified for the first time in 1984 and was associated with severe and even fatal LRTIs in South America [13,14,40]. By 1990, 7h had completely replaced the previously predominant genome type 7c (and other genome types) in South America [14]. Both its association with severe disease and the strain displacement in the late 1980s suggested that 7h was a novel, emerging strain of HAdV-B7. Subsequent sequence analysis of genome type 7h revealed its recombinant molecular phylogeny with its fiber sequence deriving from HAdV-B3 [41]. Therefore, HAdV-B7 genome type 7h was relabeled as the recombinant genotype HAdV-B66 (P7H7F3) in 2012, 28 years after its first isolation. The penton base sequence of HAdV-B66 was found to be slightly evolved, but still was considered as type 7-like [1].
In case of the recent epidemic of what was initially considered to be caused by a strain of HAdV-B3, analysis of the isolates was performed by complete genomic sequencing, and retyping as genotype HAdV-B114 (P7H3F3) was achieved within less than a year. However, initial NCBI Blast analysis of these complete genomic sequences gave somewhat misleading results, showing a very high sequence identity (> 99.8%) to a multitude of complete genome sequences from the USA and China labeled as HAdV-B3 (some of these even erroneously labeled as P3H3F3) which have to be considered as HAdV-B114 (P7H3F3) (Supporting Information S2 and S3: Supplementary Documents 1 and 2). Mislabeling of genotypes in the databases is a general problem that is not easy to mitigate. Common causes for annotation errors in public databases are contamination of biological samples, incorrect metadata submission by users, or errors related to bioinformatic/computational tools depending on applied settings [42]. In case of the sequences erroneously labeled as P3H3F3 instead of P7H3F3, errors related to the analysis of their recombinant phylogeny seem to be probable. Although most of the genomic sequences were recent clinical isolates originating from China and the USA suggesting significant HAdV-B114 circulation in these countries, a few were labeled as the old genome type 3a: #KF268123 with 99.97% identity and #JX423381.1 with 99.91% identity to HAdV-B114 (for comparison, the identity of the HAdV-B114 sequence to the HAdV-B3 prototype sequence #AY599834.1 was only 98.44%). Since genome type 3a was originally defined by its RFLP patterns [37], we performed in silico RFLP testing on our HAdV-B114 sequence data, which confirmed its identity with the genome type HAdV-B3a (Supporting Information S1: Figure S2). Therefore, HAdV-B114 was not considered to be a novel, emerging virus (like HAdV-B66 in the 1980s), but rather a reemergent virus, previously labeled as genome type 3a.
According to the usual taxonomy procedures for genome types, the HAdV-B3 prototype from 1953 was labeled as genome type 3p [37,38] and the oldest HAdV-B3 strains with a different RFLP pattern isolated as early as 1962 in Japan, China, and the USA (but not in Europe and South America) were classified as genome type 3a [38,43]. Very closely related strains were labeled as 3a1-3a7, and subsequent strains discerned by RFLP were labeled as 3b, 3c, and so on [37,38,43,44]. Interestingly, genome type 3a was associated with rather severe disease compared to other HAdV-B3 genome types. In addition to URTI, its clinical manifestations included LRTI, eye infections, and even a few disseminated and central nervous system infections [37,38,43]. These clinical presentations of HAdV-B3a infections appear to be identical to the diseases caused by HAdV-B114 in 2023. As genomic sequencing was not yet feasible when HAdV-B3a was isolated, the emergence of the more virulent genome type 3a was then considered to be a result of positive selection of point mutations rather than a result of recombinant phylogeny and thus HAdV-B114 was not discovered until 2023 [37].
Although the hexon contains the main neutralization determinant, fiber and penton base proteins bind to primary and secondary cellular receptors and therefore rather influence tropism and virulence. It is intriguing that HAdV-B114 (previously HAdV-B3a) and HAdV-B66 (previously HAdV-B7h) share almost identical penton and fiber base sequences (P7 and F3) and both HAdV-B3a and -B7h were considered to be more virulent than their corresponding prototypes [13,14,37,38,40,43]. Moreover, HAdV-B114 and HAdV-B66 were closely related in several early genes, which may influence virulence and pathogenicity (Supporting Information S1: Figure S1). Furthermore, if we assume that recent reports on HAdV-B3 cases represent unrecognized HAdV-B114 infections, our results on circulation and clinical severity are in line with recent studies on an increased circulation of HAdV in the USA in 2023, with 71.9% being typed as HAdV-B3 (based on HAdV hexon sequencing, thus a differentiation of HAdV-B3 and -B114 was not possible) [21]. These HAdV-B3-infected patients were frequently found in the age group between 3 and 5 years [21], but no increased rates of hospitalization were associated with HAdV-B3 infections. Older age and immunosuppression were the top two variables associated with an increased likelihood of hospital admissions with HAdV in the latter study, similar to our observations.
The limitations of this study are the following: first, our study does not indicate that HAdV-B114 evolved by a single recombination event that replaced the penton gene of HAdV-B3 with an HAdV-B7 penton sequence. This may be erroneously suggested by the taxonomic labeling as P7H3F3. HAdV-B114 probably evolved through several recombination events involving HAdV-B66 (with a genomic backbone evolved from HAdV-B7) and HAdV-B3 as suggested by the phylogenetic trees of its early and late gene regions as well as by bootscan analysis. Possible further steps that could be taken in the future include conducting a phylogenetic analysis to determine the possible order of events (e.g., molecular clock or detailed recombination breakpoint analyses). Second, although biological experiments to investigate the potentially enhanced transmission or pathogenicity of HAdV-B114 could be very useful, they are beyond the scope of this study.
In conclusion, retrospective NCBI BLAST analysis, penton base gene sequencing of older German HAdV-B3 isolates and comparison of RFLP patterns to results of studies from the 1980s unequivocally indicate that HAdV-B114 is not a novel emerging virus, but has been circulating already for decades. Nevertheless, the reemergence of HAdV-B114, which can be considered to be more pathogenic than the closely related HAdV-B3 prototype, caused a significant disease burden, which eventually led to its identification as a recombinant HAdV genotype.
## References
1. Dehghan, Liu, Seto (2012) "Five Genome Sequences of Subspecies B1 Human Adenoviruses Associated With Acute Respiratory Disease" *Journal of Virology*
2. Seto, Chodosh, Brister et al. (2011) "Using the Whole-Genome Sequence to Characterize and Name Human Adenoviruses" *Journal of Virology*
3. Seto, Jones, Dyer et al. (2013) "Characterizing, Typing, and Naming Human Adenovirus Type 55 in the Era of Whole Genome Data" *Journal of Clinical Virology*
4. Hage, Gerd Liebert, Bergs et al. (2015) "Human Mastadenovirus Type 70: A Novel, Multiple Recombinant Species D Mastadenovirus Isolated From Diarrhoeal Faeces of a Haematopoietic Stem Cell Transplantation Recipient" *Journal of General Virology*
5. Ismail, Cui, Dommaraju (2018) "Genomic Analysis of a Large Set of Currently-and Historically-Important Human Adenovirus Pathogens" *Emerging Microbes & Infections*
6. Aoki, Benkö, Davison (2011) "Toward an Integrated Human Adenovirus Designation System That Utilizes Molecular and Serological Data and Serves Both Clinical and Fundamental Virology" *Journal of Virology*
7. Hage, Huzly, Ganzenmueller et al. (2014) "A Human Adenovirus Species B Subtype 21a Associated With Severe Pneumonia" *Journal of Infection*
8. Kajon, Lu, Erdman (2010) "Molecular Epidemiology and Brief History of Emerging Adenovirus 14-Associated Respiratory Disease in the United States" *Journal of Infectious Diseases*
9. Potter, Cantrell, Mallak et al. (1999) "Adenovirus-Associated Deaths in US Military During Postvaccination Period" *Emerging Infectious Diseases*
10. Gu, Liu, Li (2012) "Severe Community-Acquired Pneumonia Caused by Adenovirus Type 11 in Immunocompetent Adults in Beijing" *Journal of Clinical Virology*
11. Zhang, Seto, Cao et al. (2012) "Genome Sequence of Human Adenovirus Type 55, a Re-Emergent Acute Respiratory Disease Pathogen in China" *Journal of Virology*
12. Cao, Huang, Pu (2014) "Emergence of Community-Acquired Adenovirus Type 55 as a Cause of Community-Onset Pneumonia" *Chest*
13. Carballal, Videla, Misirlian et al. (2002) "Adenovirus Type 7 Associated With Severe and Fatal Acute Lower Respiratory Infections in Argentine Children" *BMC Pediatrics*
14. Kajon, Wadell (1994) "Genome Analysis of South American Adenovirus Strains of Serotype 7 Collected Over a 7-Year Period" *Journal of Clinical Microbiology*
15. Madisch, Wölfel, Harste et al. (2006) "Molecular Identification of Adenovirus Sequences: A Rapid Scheme for Early Typing of Human Adenoviruses in Diagnostic Samples of Immunocompetent and Immunodeficient Patients" *Journal of Medical Virology*
16. Crawford-Miksza, Schnurr (1996) "Adenovirus Serotype Evolution Is Driven by Illegitimate Recombination in the Hypervariable Regions of the Hexon Protein" *Virology*
17. Madisch, Hofmayer, Moritz (2007) "Phylogenetic Analysis and Structural Predictions of Human Adenovirus Penton Proteins as a Basis for Tissue-Specific Adenovirus Vector Design" *Journal of Virology*
18. Madisch, Harste, Pommer et al. (2005) "Phylogenetic Analysis of the Main Neutralization and Hemagglutination Determinants of All Human Adenovirus Prototypes as a Basis for Molecular Classification and Taxonomy" *Journal of Virology*
19. Götting, Baier, Panagiota et al. (2022) "High Genetic Stability of Co-Circulating Human Adenovirus Type 31 Lineages Over 59 Years" *Virus Evolution*
20. Uzun, Zlamal, Althaus (2024) "Cerebral Venous Sinus Thrombosis and Thrombocytopenia Due to Heparin-Independent Anti-PF4 Antibodies After Adenovirus Infection" *Haematologica*
21. Abdullah, Fall, Klein et al. (2024) "Increased Circulation of Human Adenovirus in 2023: An Investigation of the Circulating Genotypes, Upper Respiratory Viral Loads, and Hospital Admissions in a Large Academic Medical Center" *Journal of Clinical Microbiology*
22. Martin, Gonzalez, Reynolds (2023) "Adeno-Associated Virus 2 and Human Adenovirus F41 in Wastewater During Outbreak of Severe Acute Hepatitis in Children, Ireland" *Emerging Infectious Diseases*
23. Peters, Kim, Mourya (2023) "Recent Increase in Incidence of Severe Acute Hepatitis of Unknown Etiology in Children Is Associated With Infection With Adenovirus and Other Nonhepatotropic Viruses" *Journal of Pediatrics*
24. Martin, Gonzalez, Reynolds (2023) "Adeno-Associated Virus 2 and Human Adenovirus F41 in Wastewater During Outbreak of Severe Acute Hepatitis in Children, Ireland" *Emerging Infectious Diseases*
25. Wyler, Lauber, Manukyan (2024) "Pathogen Dynamics and Discovery of Novel Viruses and Enzymes by Deep Nucleic Acid Sequencing of Wastewater" *Environment International*
26. Mahmood, Ahmed, Farooq (2024) "Molecular Characterization of Human Adenoviruses Associated With Pediatric Respiratory Infections in Karachi, Pakistan" *BMC Infectious Diseases*
27. Wang, De, Han (2024) "High-Frequency Recombination of Human Adenovirus in Children With Acute Respiratory Tract Infections in Beijing, China" *Viruses*
28. Humières, Fouillet, Verdurme (2024) "An Unusual Outbreak of Parvovirus B19 Infections, France, 2023 to 2024" *Eurosurveillance*
29. Okitsu, Khamrin, Hikita (2024) "Remarkable Increase in the Detection and Molecular Characterization of Adenovirus F41 in Children With Acute Gastroenteritis in Japan, 2017-2023" *Journal of Medical Virology*
30. Barr, Kjellén, Svedmyr (1958) "Hospital Outbreak of Adenovirus Type 3 Infections: A Clinical and Virologic Study on 38 Patients Partly Involved in a Nosocomial Outbreak" *Acta Paediatrica*
31. Germanis, Jeansson (1973) "Ocular Illness in Association With Adenovirus Type 3 Infection: An Epidemic of Conjunctivitis and Keratoconjunctivitis in Stockholm 1969" *Scandinavian Journal of Infectious Diseases*
32. Herbert, Wilkinson, Burchak et al. (1977) "Adenovirus Type 3 Pneumonia Causing Lung Damage in Childhood" *Canadian Medical Association Journal*
33. Wright, Beckwith, Gwinn (1964) "A Fatal Case of Inclusion Body Pneumonia in an Infant Infected With Adenovirus Type 3" *Journal of Pediatrics*
34. Top, Jr, Grossman et al. (1971) "Immunization With Live Types 7 and 4 Adenovirus Vaccines. I. Safety, Infectivity, Antigenicity, and Potency of Adenovirus Type 7 Vaccine in Humans" *Journal of Infectious Diseases*
35. Kuschner, Russell, Abuja (2013) "A Phase 3, Randomized, Double-Blind, Placebo-Controlled Study of the Safety and Efficacy of the Live, Oral Adenovirus Type 4 and Type 7 Vaccine, in U.S. Military Recruits" *Vaccine*
36. Li, Wadell (1986) "Analysis of 15 Different Genome Types of Adenovirus Type 7 Isolated on Five Continents" *Journal of Virology*
37. Li, Wadell (1988) "Comparison of 17 Genome Types of Adenovirus Type 3 Identified Among Strains Recovered From Six Continents" *Journal of Clinical Microbiology*
38. Wadell (1984) "Molecular Epidemiology of Human Adenoviruses"
39. Adrian, Becker, Hierholzer et al. (1989) "Molecular Epidemiology and Restriction Site Mapping of Adenovirus 7 Genome Types"
40. Murtagh, Cerqueiro, Halac et al. (1993) "Adenovirus Type 7h Respiratory Infections: A Report of 29 Cases of Acute Lower Respiratory Disease" *Acta Paediatrica*
41. Kajon, Wadell (1996) "Sequence Analysis of the E3 Region and Fiber Gene of Human Adenovirus Genome Type 7h" *Virology*
42. Bagheri, Severin, Rajan (2020) "Detecting and Correcting Misclassified Sequences in the Large-Scale Public Databases" *Bioinformatics*
43. Faulkner, Van Rooyen (1962) "Adenoviruses Types 3 and 5 Isolated From the Cerebrospinal Fluid of Children" *Canadian Medical Association Journal*
44. Kajon, Murtagh, Franco et al. (1990) "A New Genome Type of Adenovirus 3 Associated With Severe Lower Acute Respiratory Infection in Children" *Journal of Medical Virology* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12831674&blobtype=pdf | # Virus-like particle (VLP)-based indirect ELISA (iELISA) for the detection of beak and feather disease virus (BFDV) antibodies
Pangkaj Dhar, Tridi Das, Ba Nath, Prabal Chowdhu, Andrew Peters, Jade Forwood, Shane Raidal, Shubhagata Das
## Abstract
Beak and feather disease virus (BFDV) poses a significant threat to avian biodiversity and global aviculture. Reliable serodiagnostic tools are critical for assessing host immune status and guiding disease management. The haemagglutination inhibition (HI) assay, although historically regarded as the gold standard, is limited by technical complexity and its reliance on seropositive Galah erythrocytes, restricting broader application. This study describes the development and validation of a recombinant BFDV capsid protein-based indirect enzyme-linked immunosorbent assay (iELISA) for detecting anti-BFDV antibodies using dried blood spots from multiple psittacine species. The assay demonstrated high sensitivity and strong analytical performance, employing virus-like particles (VLPs) recombinantly expressed in Escherichia coli. Optimisation of the diagnostic cut-off by two-graph receiver operating characteristic (TG-ROC) analysis established an OD threshold of 1.73, achieving 96.5% sensitivity with a Youden's index of 0.74. Discriminative capacity was further supported by receiver operating characteristic (ROC) analysis, yielding an area under the curve (AUC) of 0.896. Agreement with the HI assay was very strong (Gwet's Agreement Coefficient 1= 0.843). This iELISA represents a scalable and universal serodiagnostic tool, supporting clinical diagnosis, enabling large-scale epidemiological investigations, and advancing conservation-focused BFDV surveillance.
Key points• Efficient detection of anti-BFDV antibodies in psittacines using recombinant capsid protein as coating antigen.• Use of TG-ROC curve to determine cut-off value supported by Gwet's Agreement Coefficient 1.• Easily adaptable method with very high sensitivity in detecting anti-BFDV antibodies.
## Introduction
Psittacine beak and feather disease (PBFD) is a widespread viral infection afflicting psittacine birds globally, caused by the beak and feather disease virus (BFDV), a member of the Circoviridae family (Pass and Perry 1984;Raidal and Cross 1995;Ritchie et al. 1989). The virus possesses a circular, single-stranded DNA (ssDNA) genome of 1.9-2.0 kb in length, encoding two major proteins: a capsid (Cap) protein, which plays a role in viral replication, attachment, and cellular entry and can self-assemble into icosahedral viruslike particles (VLPs) of 60 capsid protein molecules, and a replicase initiator (Rep) protein, which helps with viral replication (Bassami et al. 2001;Patterson et al. 2012;Ritchie et al. 1990;Sarker et al. 2016). BFDV disease exhibits a variety of clinical manifestations, ranging from acute infections in juveniles to chronic feather dystrophy in adult birds (Doneley 2003;Schoemaker et al. 2000).
BFDV Cap protomer's agility to assemble as stable macromolecular structures when recombinantly expressed exalted it to form VLPs, which are remarkably compatible with serological diagnostic assays, including haemagglutination inhibition (HI). Despite lacking any established cell culture system, these recombinantly produced VLPs provide essential synthetic antigens in high concentration and purity for further downstream applications, including immunodiagnostics (Wang et al. 2020).
Existing diagnostic modalities for BFDV predominantly rely on molecular techniques such as polymerase chain reaction (PCR) and quantitative real-time PCR (qPCR), (Katoh et al. 2008;Shearer et al. 2009b;Ypelaar et al. 1999). Other molecular methods, like nested-PCR (Kiatipattanasakul-Banlunara et al. 2002), duplex shuttle PCR (Ogawa et al. 2005), and rolling circle amplification (Varsani et al. 2011), can be used in both diagnostic and research laboratories, offering valuable insights into viral load, strain variation, and infection dynamics. While these methods are sensitive for detecting active viremia, they cannot depict insight into the host's immunological status or previous exposure.
Haemagglutination inhibition (HI) test, on the contrary, has emerged as the only reliable serological tool to assess antibody responses to BFDV in wild and captive birds, which is limited to certain laboratories (Raidal and Cross 1994;Raidal et al. 1993). HI antibody titers have been considered a strong negative predictor of PBFD in cockatoos (Khalesi et al. 2005;Ritchie et al. 1991), hence an active or persistent BFDV infection in other host species may exhibit low and fluctuating anti-BFDV HI titers. This may be the result of bursal or thymus damage and/or an effect of persistent infection in macrophages (Latimer et al. 1990). Moreover, HI assays may lack standardisation when native antigen is used, and therefore are impractical for high-throughput screening. Despite being considered the current gold standard, challenges associated with the availability of hemagglutinating Galah erythrocytes, and variable erythrocyte performance from individual birds have hindered the adoption of HI testing from being an easily accessible, universal diagnostic tool (Sanada and Sanada 2000;Shearer et al. 2009a).
These identified factors underscore the need for a standardised, sensitive, and reproducible diagnostic alternative. Enzyme-linked immunosorbent test (ELISA) based on virus-like particles (VLPs) have become a strong candidate in veterinary diagnostics and surveillance of many infectious diseases and are also used to assess immunological responses to vaccination (Binns et al. 2007;Lequin 2005;Tang et al. 2013). Indirect ELISA (iELISA) offers a scalable, reproducible, and less labour-intensive alternative with the potential to detect past exposure and assess population-level immunity. This technique has been recommended as a standard tool for population-based serological studies (Wright et al. 1993). Use of dried blood spots (DBS) for serological analysis is well-established in serological diagnostics (Holroyd et al. 2022;Samsonova et al. 2022). DBS requires smaller blood volumes, simplifies storage and transportation without special treatment, and reduces biohazard risks (Malsagova et al. 2020;Sharma et al. 2014). This technique has been successfully applied in various fields, including pharmacokinetics, therapeutic drug monitoring, and disease diagnosis (Sharma et al. 2014). The existing pioneer in serodiagnosis of BFDV, HI test, also relies on DBS and produces excellent interpretation (Raidal et al. 1993). Moreover, all BFDV lineages and genogroups from diverse psittacine species exhibit similar behaviour in the existing HI assay, with no evidence of serotype variation (Raidal et al. 1993). All these channelled the hypothesis of using DBS as a source of primary antibody in developing an indirect ELISA.
The present study was aimed at developing a robust, broadly applicable indirect ELISA (iELISA) using recombinantly expressed BFDV capsid VLPs, while prioritising validation of its diagnostic performance against the existing HI test. The assay demonstrated high sensitivity, specificity, and efficiency, thus presented as a reliable tool for clinical diagnosis, as well as for routine monitoring and surveillance.
## Materials and methods
## Blood samples
The study comprised 248 samples from four groups of psittacine birds: eclectus parrots, lorikeets, cockatoos, and orange-bellied parrots. All the samples had been tested as duplicates, and the mean OD value was counted for further analysis. Archived dried blood spots were retrieved randomly from the Veterinary Diagnostic Laboratory (VDL), Charles Sturt University, spanning 2023 and 2024. All of them were sent to VDL, aimed at tests other than iELISA. One punch of blood spot (6 mm) was chopped aseptically, rinsed in 400 µL PBS, vortexed, and incubated overnight at 4° C. The next day, serum was extracted by centrifuging at 10,000 RPM for 10 min. The resulting supernatant was carefully collected and stored at -20 °C until further analysis. A known anti-BFDV chicken serum sample was used as a positive control (Raidal et al. 1993), and a known chicken serum without any BFDV exposures was used as a negative control.
## Expression and purification of recombinant BFDV Cap protein
BFDV capsid gene was cloned into the Sspl cloning site of the pMCSG21 vector, followed by transformation into Escherichia coli BL21 (DE3) Rosetta 2 cells (Novagen, Darmstadt, Germany) for recombinant expression (Sarker et al. 2015b). Starter culture was grown using these transformed colonies, and expression media was grown at 37 °C until an optimum OD 600 of 0.4-0.6 was gained, then the temperature was shifted to 25 °C, and protein synthesis was induced by the addition of 0.5 mM isopropyl-β-Dthiogalactoside (IPTG) for up to 24 h of growth in total. The cells were then harvested through centrifugation in an Avanti JXN-26 (Beckman Coulter, Mt Waverley, VIC, Australia) at 6000 RPM for 30 min, and the cell pellets were resuspended in CAPS buffer A (20 mM N-cyclohexyl-3-aminopropanesulfonic acid) (pH 10.5) and stored at -20°C.
By using FastBreak™ Cell Lysis Buffer (Promega, Alexandria, NSW, Australia) along with 20 mg lysozyme (Sigma-Aldrich, St. Louis, MO, USA) and 0.5 mg of DNase (Invitrogen, Mt Waverley, VIC, USA), bacterial cells were lysed and incubated on ice for 30 min and underwent centrifugation in an Avanti Mt Waverley,VIC,Australia) for 30 min at 15,000 RPM at 16° C to remove the cellular debris, and the resulting supernatant was filtered through a 0.45 µm low protein-binding filter (Millipore, Merck, Darmstadt, Germany). Afterwards, this supernatant was injected into a 5 mL Ni 2+ column (HisTrap HP, GE Healthcare, Chicago, IL, USA) on an AKTA pure FPLC in CAPS buffer A & B. The column was washed extensively with 10 column volumes of buffer to remove the contaminating endogenous proteins and eluted using a gradient elution buffer containing 500 mM imidazole. Finally, this protein was purified by size exclusion chromatography using a Superdex 200 column (GE Healthcare, Chicago, IL, USA) in glutathione S-transferase (GST) buffer A containing 50 mM Tris and 125 mM NaCl (pH 8.0). Proteins that showed peak fractions were pooled and concentrated by an Amicon ultrafiltration device (Milipore, Merck, Darmstadt, Germany), and the purity of this Cap protein at different stages was assessed by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) (Fairbanks et al. 1971) stained with Coomassie blue for 5 min at room temperature (RT) with gentle shaking and de-stained again in a solution containing 10% ethanol and 10% glacial acetic acid for at least 30 min at RT with gentle shaking. Finally, purified BFDV Cap protein was stored at -80° C at 1 mg/mL concentration in GST A buffer (pH = 8.0) until further downstream applications.
## Antigenicity test of Cap protein
Antigenicity of the purified Cap protein was confirmed by performing a western blot following the methods described by Patterson et al. (2013). Simultaneously, after SDS-PAGE, the gel was transferred to a nitrocellulose membrane (Bio-Rad, South Granville, NSW, Australia) using freshly prepared transfer buffer (containing 11.64 g Tris Base, 5.86 g glycine, 750 μL 10% SDS, 400 mL methanol, and water to a final volume of 2 L) and a Criterion Blotter apparatus (Bio-Rad, South Granville, NSW, Australia) at 80 V for 60 min. Afterwards, the membrane was blocked with Tris-buffered saline solution containing 0.05% Tween 20 (TBST) (Dako, Santa Clara, CA, USA) and 5% skim milk with gentle rocking for one hour at RT. The membrane was washed three times for 5 min each using TBST. It was then incubated for 2 h at RT, with similar gentle rocking as previously, with an anti-BFDV monoclonal antibody (MAb), 3F8-1 (1 mg/mL) (Ab Solutions, Perth, WA, Australia) (Shearer et al. 2008) diluted (1:800) in 5% skim milk-TBST solution. After that, the membrane was again washed thrice like before and incubated further for one hour, merged with horseradish peroxidase (HRP)-conjugated goat anti-bird IgY (Sigma-Aldrich, St. Louis, MO, USA) diluted (1:1000) in 5% skim milk-TBST solution. The membrane was then washed three times and visualised by using 3,3′,5,5′-tetramethylbenzidine (TMB) Liquid Substrate System for membranes (Sigma-Aldrich, St. Louis, MO, USA) for 5 min. Finally, the colour development was stopped by washing the membrane with deionised water for 1 min.
## Selection of optimum antigen dilution
Before using the antigen, hemagglutination assays (HA) were performed using a twofold serial dilution of the purified BFDV Cap protein and BFDV VLPs using galah (Eolophus roseicapillus) erythrocytes according to the protocol developed by Raidal et al. (1993). Before proceeding to further downstream experiments, optimal concentrations of the recombinant Cap protein as well as the dilution of the DBS were determined by testing through the criss-cross (checkerboard) serial dilution (Fig. 2). Both positive and negative samples were assessed in different dilutions of dried blood spots and multiple concentrations of Cap protein. The dilution that showed the highest positive to negative ratio (PN ratio) has been selected for iELISA testing.
## HI test
Hemagglutination inhibition (HI) test of the samples has been done commercially in the veterinary diagnostic laboratory, Charles Sturt University. It has been operated simultaneously using recombinant BFDV Cap protein expressed in E. coli BL 21 (DE3) (Sarker et al. 2015b), according to the protocol developed by (Raidal and Cross 1994). HI titres are expressed as the reciprocal of the highest serum dilution that completely inhibits hemagglutination, indicating the presence of virus-specific antibodies. Titres of < 1:20 were considered undetected or negative, while positive results were recorded as serial two-fold dilutions (1:20, 1:40, 1:80, etc.), with higher titres indicating stronger antibody responses.
## Inter-and intra-assay precision
The assay's reproducibility was assessed using HI positive and HI negative sera on different days under similar conditions. Intra-assay variability was evaluated by testing each sample in triplicate on a single plate. Each sample was analysed on three distinct plates to assess inter-assay variability. ELISA's overall precision was defined by evaluating the variation between wells in the same test (intra-assay) or between different runs (inter-assay) by counting the mean optical density (OD), standard deviation (SD), and coefficient of variation (CV) as a ratio of the SD versus the mean OD value.
## iELISA procedure
This iELISA was developed and validated in compliance with the Office International des Epizootiques (O.I.E.) guidelines described in Jacobson (1998) to develop serological assays to diagnose infectious diseases. An in-house recombinant BFDV Cap protein (Sarker et al. 2015b) was used as the primary antigen to coat the 96-well Nunc-Immuno Micro-Well plates (Sigma-Aldrich, St. Louis, MO, USA). The iELISA was operated following a slight modification of the previously described protocol (Das et al. 2026;Neef et al. 2024). One hundred microliters of the Cap protein, diluted with 0.05 M carbonate/bicarbonate binding buffer, was applied to coat each well, followed by an overnight incubation at 4 °C. The next day, the antigen solution was aspirated from each well, and wells were washed five times with trisbuffered saline containing 0.05% (v/v) Tween-20 (TBST). Wells were blocked with 100 µL TBST containing 5% skim milk powder for 1 h at 37 °C. This blocking was washed five times with TBST before adding 100 µL serum (primary antibody) and incubating at 37 °C for 1 h. After incubation, the primary antibody (Ab) was removed, and wells were again washed five times with TBST. 100 µL of Goat Antibird IgY H&L (Ab112773) conjugated with horseradish peroxidase (HRP) (Abcam, Waltham, MA, USA) diluted with TBST containing 5% skim milk (1:2000) (v/v) was then added and incubated at 37° C for 1 h, followed by removal of the secondary Ab solution and washing the wells five times with TBST. To develop the colourimetric reaction, 100 µL of HRP substrate solution (BioRad, South Granville, NSW, Australia) was added to the wells and incubated for 12 min at RT. This colour development is stopped by adding 100 µL of 0.18 M H 2 SO 4 . Optical density (OD) was recorded at 450 nm with a microplate reader (BMG Labtech, Mornington, VIC, Australia) using 630 nm as a reference. Four types of controls were used during each test: HI positive controls, HI negative controls, no primary antibody control, and no secondary antibody control.
## Data analysis
The diagnostic performance of iELISA was evaluated by comparing results with HI assay outcomes. Samples with HI tests having titres of < 1:20 were considered negative, and samples that showed HI titres > 1:20 (1:20, 1:40, 1:80, 1:160, 1:320, etc.) were considered positive. The mean OD value of each sample was counted for the analysis. To determine the cut-off points, both the HI and iELISA test results were analysed by a two-graph receiver operating characteristic (TG-ROC) (Greiner 1995;Greiner et al. 1995). The results were also plotted as a receiver operating characteristic (ROC) curve, and the area under the curve (AUC) was calculated to determine overall assay performance (Gardner and Greiner 2006). Gwet's agreement was calculated to check the concordance between the iELISA and HI test (Gwet 2014). Data were analysed using RStudio version 4.3.2 (Team RC 2010).
## Results
## Antigenicity of BFDV Cap protein
BFDV Cap protein was purified from other non-specific bacterial proteins using HIS and size exclusion chromatography, and subsequently visualised by SDS-PAGE analysis (Fig. 1A). In a concurrent test, the antigenicity of the yielded BFDV Cap protein was tested by western blot assay against a monoclonal BFDV antibody. Strong antigenicity of the BFDV Cap protein (Fig. 1B) in the test assures its potential to develop an iELISA using this antigen.
## Selection of optimum antigen dilution
Determination of the optimal concentration of the recombinant BFDV Cap protein to coat the microwell plate was a crucial step. We have optimised the concentration of this Cap protein along with the dilution of dried blood spots by using checkerboard dilution and concluded with the optimum Cap antigen concentration of 62.5 μg/mL, while the maximum dilution of the dried blood spots that showed the highest PN (Positive sample OD value/Negative sample OD value) ratio was 1:400 (Fig. 2); this means each blood spot was diluted in 400 μL of sterile PBS for serum extraction. In such concentrations of antigen and dilution of DBS, the PN ratio was 15.33, which was higher than any other combination. The optimised concentration of secondary antibody
## Inter-and intra-assay precision
The repeatability (intra-assay) of the developed protocol was verified using positive and negative samples with three replicates on the same plate. Reproducibility (inter-assay) was tested by using both positive and negative samples in three independent iELISA assays across three different batches. The inter-and intra-assay coefficients of variation (CV) for six control sera evaluated with the ELISA were all below 10% and within acceptable limits (Table 3). The mean interassay CV% was 3.97 (range 5.62-11.90), and the median intra-assay CV% was 2.81 (range 1.20-7.41) (Table 1). These were all within acceptable limits, indicating that the results were reproducible.
## iELISA and HI correlation
Distribution of iELISA OD-values for four tested bird species has been shown in a violin plot (Fig. 3). Variability in OD values was observed between HI-positive and negative samples (Fig. 3B) as well as among the species (Fig. 3B), with Lorikeet and Orange-bellied parrot (OBP) exhibiting higher median values, suggesting a potentially stronger serological response to BFDV exposure. Outliers may be attributed to individual differences in immune response or variations in sample quality.
The differences highlight species-specific serological responses that warrant further investigation to optimise diagnostic accuracy across various avian hosts. The narrower interquartile range (IQR) in the positive sample plot in Fig. 3B indicates consistency in antibody detection among seropositive samples.
Conversely, HI negative samples showed lower iELISA OD values with a wider distribution of IQR, suggesting the presence of potential borderline cases or assay background reactivity, which indicates the presence of variable anti-BFDV antibody levels in those HI negative samples that cannot be detected by the HI test but can be differentiated well via the iELISA test. Diagnostic sensitivity and specificity of iELISA compared to HI test results have been charted in Table 2. The table shows that 85 samples out of 88 HI positive (96.59%) were also positive in the iELISA test, which reflects the highly sensitive property of the developed iELISA. A more interesting finding is that 36 samples out of 160 HI negative also became positive in the iELISA test, which may indicate early antibody development, detectable by iELISA but not by HI.
## iELISA cut-off point selection
HI results were counted as the parameter to produce a TG-ROC curve. In HI testing, results are expressed as the reciprocal of the maximum serum dilution that completely inhibits hemagglutination, indicating the presence of virus-specific antibodies. In this study, the proportion of immune birds (HI titre ≥ 1:20) was 35.38%. TG-ROC curve is developed by plotting sensitivity (Se) and specificity (Sp) in the same plot to determine the cut-off value (Fig. 4). Equal Se and Sp (83%) can be found when the cut-off OD value is set to 1.88, but maximum Youden's index depicts that the optimum cut-off point of this TG-ROC curve has to be 1.73. At this point, 96.5% sensitivity can be obtained. But if the sensitivity is the utmost priority, then the cut-off OD value can be shifted far left to 1.55 to ensure 100% sensitivity. Otherwise, different evaluation indices based on different cut-off values are shown in Table 3.
## ROC evaluation
The ROC curve evaluates the overall diagnostic ability of the test, plotting true positive rate (sensitivity) against false positive rate (1 -specificity). The curve in Fig. 5 demonstrates a high diagnostic accuracy, with an area under the curve (AUC) of 89.6%, with a 95% confidence interval (CI) ranging from 0.855 to 0.934. This indicates that the iELISA test has strong discriminative ability in differentiating between positive and negative cases and detecting PBFD antibodies.
## Inter-assay agreement
To optimise the diagnostic threshold for the iELISA assay, Gwet's AC1 (Agreement Coefficient 1) agreement coefficient and overall test efficiency (interpreted as diagnostic accuracy) were evaluated across a range of OD cut-off values. The concordance between the iELISA and the HI test was quantitatively assessed by calculating Gwet's AC1 coefficient (Fig. 6), which serves as a robust measure of interrater agreement. The analysis yielded a percent agreement of 84.3%. The AC1 was calculated at 0.69 (Standard error 0.04), with a 95% confidence interval ranging from 0.602 to 0.783 and a p-value = 0, implying a highly statistically significant outcome. The AC1 coefficient's magnitude demonstrates a substantial level of agreement, highlighting the consistency between the ELISA-derived classification and the reference HI results. The results suggest that the indirect ELISA, when analysed with the chosen optical density threshold, demonstrates significant diagnostic reliability and inter-method coherence.
## Discussion
ELISA is a well-established technique in both human and veterinary medicine for detecting viral infections and assessing immune status at the individual or flock level (Aydin 2015;Greiner and Gardner 2000;Matefo et al. 2022;Wright et al. 1993). Because BFDV infection in susceptible birds can manifest across a wide clinical spectrum, including asymptomatic shedders, there is a critical need for early and reliable diagnosis to limit transmission (Kim et al. 2024). We developed an iELISA using recombinant capsid VLPs expressed in E. coli to detect anti-BFDV antibodies, addressing the limitations of the existing HI assay, such as its dependence on specialised Galah erythrocytes (Raidal and Cross 1994;Raidal et al. 1993;Shearer et al. 2009a), which restricts broader laboratory usability. We also validated our iELISA against HI standards to establish it as a reliable, scalable, and standardised alternative for antibody detection in psittacine birds using DBS.
Previous ELISA methods for detecting BFDV antibodies have faced notable limitations; for example, Johne et al. (2004) used a truncated BFDV Cap and a secondary antibody specific for psittacine IgY, but their study included only 11 serum samples, limiting its representativeness. Besides, they used African grey parrots to produce the antipsittacine IgY, which introduced uncertainty, as negative results in other species could reflect either a true absence of antibodies or insufficient cross-reactivity of the secondary antibody (Khalesi et al. 2005). These challenges underscore the need for a standardised, species-independent serological assay. We assessed our iELISA in at least 4 diverse groups of parrots and cockatoos, including eclectus parrots, lorikeets, sulphur-crested cockatoos, and orange-bellied parrots, thus providing a robust framework to assess the cross-species applicability of the developed iELISA. 4 TG-ROC analysis of PBFD iELISA using the sensitivity (Se) and specificity (Sp) values. The optimal Youden's index (J) was found to be at a cut-off of 1.73, with sensitivities and specificities of 96.5% and 77.5%, respectively. If the sensitivity is more important, the cutoff can be moved left to 1.55, where 100% sensitivity can be achieved The BFDV capsid (Cap) is the sole structural protein of the virion, serving multiple functions including forming the protective shell, binding viral DNA, and mediating nuclear transport essential for replication (Nath et al. 2021;Sarker et al. 2016). Importantly, the Cap protomers self-assemble into virus-like particles (VLPs), mimicking the biological structure of the wild-type virion as a T = 1 icosahedral particle. The retention of structural congruence across hostadapted BFDV genotype clusters, despite surface variations, suggests strong evolutionary constraints on capsid architecture (Das et al. 2019), which may explain the limited serotype diversity observed in extant parrots and underpins the feasibility of developing broadly applicable serological diagnostics. VLPs offer distinct advantages as antigens, as their repetitive and ordered surface array displays all relevant antigenic determinants, enhancing serodiagnostic sensitivity. In addition, their inherent thermostability and extended shelf-life compared to truncated proteins (Das 2018;Sarker et al. 2015b) further strengthen their suitability for reliable and scalable diagnostic applications. Collectively, these properties establish the BFDV Cap as an excellent antigen candidate for serological assays such as ELISA. Recombinant expression systems, including baculovirus and E. coli, have consistently demonstrated strong immunogenicity and antigenicity, with the E. coli system further optimised to produce stable, soluble, and antigenically valid protein (Johne et al. 2004;Khalesi 2007;Sarker et al. 2015b).
The optimisation of DBS dilution and BFDV Cap antigen concentration in this study was comparable to approaches reported in other contemporary works (Deng et al. 2018;Ge et al. 2018). Checkerboard titrations were employed to establish the optimal working conditions (Jacobson 1998), with results demonstrating that a DBS dilution of 1:400 combined with an antigen coating concentration of 62.5 µg/mL yielded the highest PN ratio (15.33%) (Fig. 2). The selected dilution aligns with values used in other avian serology studies (Deng et al. 2018;Ge et al. 2018;Liu et al. 2010;Neef et al. 2024;Wang et al. 2020;Xu et al. 2025) and reflects the need to balance adequate signal strength with minimal background reactivity. Lower dilutions often produced stronger optical density readings but were associated with increased background absorbance, whereas higher dilutions reduced nonspecific binding but risked signal loss in low-titre samples. The chosen 1:400 dilution, therefore, represents a robust compromise, maximising sensitivity while retaining assay specificity. Consistent with best practice, efficient elution of antibodies was enhanced by chopping DBS punches into smaller pieces prior to immersion in PBS, increasing surface exposure and facilitating the release of immunoglobulins.
The optimal coating antigen concentration was determined to be 62.5 µg/mL, a value comparable to recent ELISA studies employing recombinant circovirus VLPs (Mercatali et al. 2018;Neef et al. 2024). Use of recombinantly expressed antigen in E. coli increases the chances of bacteria-derived protein contamination in the downstream applications. This protein can lead to generate false positive results, especially when an anti-bird antibody is used as a secondary antibody source. To avoid such consequences, we used E. coli BL21 (DE3) during the transformation of BFDV cap and maintained a two-step purification of the recombinantly expressed protein, HIS chromatography followed by size exclusion chromatography. Robichon et al. (2011) showed that using E. coli BL21 (DE3) substantially minimises E. coli-derived protein contamination. A secondary antibody was used at a 1:2000 dilution, which provided strong signal intensity while minimising background (Das et al. 2026). To simplify production and broaden species coverage, a polyclonal Goat Anti-bird secondary antibody was selected. Previous work has shown that commercially available anti-chicken IgY antibodies exhibit substantial cross-reactivity with immunoglobulins from diverse avian taxa, including psittacine birds (Cray and Villar 2008). This broad reactivity underpins the suitability of anti-bird antibodies for wildlife serology, where species-specific reagents are often unavailable. Furthermore, anti-bird antibodies have been successfully applied in indirect ELISAs, where the selection of an appropriate cut-off is critical for interpreting ELISA results, as it establishes the threshold for classifying samples as positive, negative, or inconclusive (Matefo et al. 2022). Traditional approaches, such as setting the cut-off at two or three standard deviations above or below the mean, assume a normal distribution of test values in the target population, which may not always be valid. To overcome this limitation, TG-ROC analysis was employed to objectively identify the OD cut-off for the detection of several Fig. 6 Gwet's agreement coefficients against the OD values show a percent agreement of 0.843%, with the AC1 coefficient calculated at 0.69 avian viral infections, including flaviviruses (Hofmeister et al. 2016), alphaviruses (Fassbinder-Orth et al. 2016), and poxviruses (Ellison et al. 2014;Ha et al. 2013), further supporting their diagnostic reliability and versatility.
The violin plots presented in Fig. 3 illustrate distinct patterns of iELISA OD value distribution both across species and in relation to HI serostatus. Among species (Fig. 3A), lorikeets and OBPs exhibited consistently elevated OD values, whereas eclectus showed marginal and cockatoos demonstrated a comparatively attenuated and compressed distribution. These interspecific differences likely reflect variation in historical exposure, host susceptibility, and immunological responsiveness to BFDV, as previously documented in psittacine populations (Das et al. 2019;Sarker et al. 2015a). Figure 3B demonstrates a pronounced separation between HI positive and HI negative samples, with HI positive individuals displaying uniformly higher OD values and limited overlap with the negative cohort. The broader distribution of the HI-negative samples may indicate low-level or past exposure not detectable by HI, consistent with the higher analytical sensitivity typically attributed to ELISA-based antibody assays (Aydin et al. 2025;Shah and Maghsoudlou 2016). These findings underscore the capability of the iELISA to discriminate reliably between seropositive and seronegative individuals across diverse host species and support its suitability as a robust serological tool for BFDV surveillance and diagnostic applications.
Considering the HI titre as the parameter, the TG-ROC curve was developed to generate the cut-off value. TG-ROC analysis offers an adequate array of cut-off values to achieve Se and Sp for both parametric and nonparametric approaches. TG-ROC curve facilitates a quantitative serodiagnostic test when estimates of costs associated with falsepositive and false-negative results are available (Greiner 1996). For both parametric and nonparametric methods, it provides high sensitivity and specificity, as demonstrated in a study on Newcastle disease virus detection in quail serum (Oliveira et al. 2007). While conventional methods like Mean ± 2SD are commonly used, they can be arbitrary. ROC curve analysis is widely accepted for optimising cutoff values and comparing diagnostic test accuracy (Sharma and Jain 2014). It allows for improved test performance by adjusting cut-off values based on clinical conditions. Alternative approaches, such as one based on the area under the ROC curve, have been proposed to define optimal cut-off values (Unal 2017). These methods aim to maximise sensitivity and specificity, crucial for accurate serological diagnosis of infectious diseases.
The diagnostic performance of the iELISA was evaluated using multiple complementary approaches to provide a robust assessment of sensitivity, specificity, and overall reliability. TG-ROC curve analysis enabled visualisation of potential cut-off values, capturing the trade-offs between sensitivity and specificity and revealing the intermediate range of OD values that are not discernible from conventional ROC or AUC plots. Youden's index (J) was used to determine the optimum threshold for cut-off in the TG-ROC curve (Li et al. 2025). Defined as (Sensitivity + Specificity -1), this index identifies the cut-off point that maximises the difference between the true positive rate and the false positive rate across all potential values. Selecting the point of maximum J ensures the selection of a single threshold, providing the best possible combined balance of test sensitivity and specificity for interpretation (Nahm 2022). At the optimum Youden's index, the iELISA achieved high sensitivity (96.5%) with a cut-off of 1.73 and an assay efficiency of 74%, while specificity was comparatively lower (77.5%) (Fig. 4). This pattern likely reflects limitations of the HI reference test, including the presence of false negatives, as suggested by the wider distribution of OD values for HInegative samples in the violin plots. Adjusting the cut-off for specific applications allows flexibility: prioritising maximal sensitivity (100%) for detecting all exposed birds requires lowering the cut-off to 1.55, whereas ensuring balanced sensitivity and specificity shifts the cut-off slightly higher to 1.88. The OD range of 1.55-1.73 can thus be considered an intermediate range, where antibody detection should be interpreted alongside clinical history and signalment, which is particularly critical for managing BFDV in vulnerable or endangered psittacine populations (Khalesi et al. 2005).
Complementary ROC analysis confirmed the overall discriminative power of the assay, with an AUC of 89.6% (95% CI 0.855-0.934) (Fig. 5), reflecting excellent ability to distinguish positive from negative cases (Greiner et al. 2000;Hajian-Tilaki 2013;Swets 1988). Concordance between iELISA and the HI test was further assessed using Gwet's AC1 coefficient. Gwet's AC1 is a chancecorrected measure of inter-rater agreement specifically used for nominal data classifications (e.g., Positive/Negative), which overcomes the prevalence-dependent paradoxes associated with Cohen's kappa (Wongpakaran et al. 2013). The analysis demonstrated 84.3% agreement. The coefficient was estimated with a Standard Error (SE) of 0.04. The SE quantifies the precision of the AC1 estimate, reflecting the variability expected from sampling, and is used to construct the resulting 95% confidence interval (0.602 to 0.783). The p-value (= 0) confirms a highly statistically significant level of agreement, indicating strong agreement beyond chance and confirming the reliability of the iELISA classification (Fig. 6). Together, these analyses highlight the consistency and diagnostic reliability of iEL-ISA as a highly sensitive and broadly applicable diagnostic tool, with clearly defined cut-offs and interpretive guidance for intermediate results. The combined use of TG-ROC, ROC/AUC, and AC1 metrics provides a comprehensive evaluation framework, ensuring accurate detection Page 12 of 15 of anti-BFDV antibodies and supporting its application in surveillance and management programs across diverse psittacine species (Ohyama 2021;Xu and Lorber 2014).
Selection of an appropriate cut-off is critical for interpreting ELISA results, as it establishes the threshold for classifying samples as positive, negative, or inconclusive (Matefo et al. 2022). Traditional approaches, such as setting the cutoff at two or three standard deviations above or below the mean, assume a normal distribution of test values in the target population, which may not always be valid. To overcome this limitation, TG-ROC analysis was employed to objectively identify the OD cut-off that provides the most diagnostically balanced threshold (Greiner 1995). We identified an optimal cut-off value of OD 1.73 for the BFDV iELISA, which corresponded to a sensitivity of 96.5% and the highest diagnostic efficiency (84%) as determined by Youden's index. Adjustments to the cut-off can be made depending on whether higher sensitivity or specificity is desired. OD values between 1.55 and 1.73 define an intermediate range (IR), within which results are considered suspect; values below and above this range are interpreted as negative and positive, respectively (Xu et al. 1997).
While the VLP-based iELISA showed strong concordance with the gold-standard HI assay, with high sensitivity and diagnostic efficiency, several inherent limitations of the iELISA format should be acknowledged. As indirect ELISAs rely on secondary antibody for signal detection, assay performance might vary due to species-dependent cross-reactivity, variation in total IgY concentrations, and background reactivity. These factors may alter OD values independently of true serostatus and may produce weaker signals in species with inherently low immunoglobulin levels (Walker and Crowther 2009).
BFDV infects a wide range of psittacine species, and spillover infections have also been detected in Passeriformes, Coraciiformes, and Strigiformes (Das et al. 2019;MacColl et al. 2024;Peters et al. 2014). Our validation included samples from only four psittacine species. Although the test performed robustly within this group, small but consistent differences in cut-off OD values were observed across species (Fig. 3A), and these differences may be more pronounced in untested hosts. This strongly suggests that a universal cut-off value is unlikely to be suitable for all species. Therefore, caution is warranted when interpreting results from new host taxa. Species-specific cut-off values must be established using a panel of known negative birds, as well as clinically positive and viraemic individuals of the same species, before applying the assay for diagnostic purpose. Such validation should be undertaken alongside the HI assay, which, despite its known limitations, has anecdotally provided reliable results across a wide range of host species in laboratories equipped with an HI test setup (Raidal and Cross 1994;Raidal et al. 1993).
The secondary antibody used in this iELISA is a commercially available anti-bird polyclonal antibody with known cross-reactivity across multiple avian taxa (Escandon et al. 2019;Fassbinder-Orth et al. 2016). However, the relative intensity of cross-reactivity likely varies among species and can influence observed OD values. Comprehensive benchmarking of this anti-IgY reagent across different host species was beyond the scope of this study. Accordingly, establishing accurate cut-off values for any new species requires testing a sufficient number of HI-validated true negative and true positive samples, ideally as part of ongoing laboratorylevel assay optimisation.
Finally, newly developed iELISA demonstrates clear advantages over existing serological methods for BFDV detection, including haemagglutination inhibition (HI) and blocking ELISA (bELISA). The current standard, the HI test, relies on the availability of unique haemagglutinating Galah erythrocytes (Raidal and Cross 1994;Raidal et al. 1993), is labour-intensive due to the need to maintain donor flocks, and presents additional challenges for large-scale application due to ethical considerations. In contrast, the iELISA is simple, cost-effective, and broadly applicable through the use of species-independent secondary antibodies, making it particularly suitable for wildlife and mixed-species surveillance. Compared to bELISA, which requires complex assay design, costly monoclonal antibodies, and inversely proportional signal interpretation, the iELISA utilises recombinant capsid VLPs that preserve the native antigenic structure and employs a pan-avian secondary antibody, allowing broader species applicability, high-throughput processing, and scalable deployment. Collectively, these features establish the validated iELISA as a scientifically robust and operationally feasible platform for broad-scale BFDV serosurveillance.
## Conclusion
The indirect ELISA developed using the recombinant capsid protein of BFDV demonstrates high accuracy, reliability, and reproducibility for detecting anti-BFDV antibodies across multiple psittacine species. This assay represents a viable and effective alternative to the existing haemagglutination inhibition test, providing a universal serological tool for routine diagnosis, surveillance, and monitoring of BFDV in endemic populations. Beyond diagnostics, the iELISA will facilitate vaccine development by enabling post-vaccination serosurveillance, distinguishing immunostatus between vaccinated and unvaccinated birds, and supporting the evaluation of immune responses. Additionally, it offers valuable insights into BFDV pathophysiology, including temporal changes in antibody levels, maternal antibody transfer, and their influence on infection and immunity. In the absence of a universal serodiagnostic tool for psittacine BFDV, this iELISA has the potential to become a cornerstone in BFDV serodiagnosis, driving a paradigm shift in research, clinical management, and broad-scale serosurveillance of beak and feather disease virus.
## References
1. Aydin (2015) "A short history, principles, and types of ELISA, and our laboratory experience with peptide/protein analyses using ELISA" *Peptides*
2. Aydin, Emre, Ugur et al. (2025) "An overview of ELISA: a review and update on best laboratory practices for quantifying peptides and proteins in biological fluids" *J Int Med Res*
3. Bassami, Ypelaar, Berryman et al. (2001) "Genetic diversity of beak and feather disease virus detected in psittacine species in Australia" *Virology*
4. Binns, Green, Bailey (2007) "Development and validation of an ELISA to detect antibodies to Corynebacterium pseudotuberculosis in ovine sera" *Vet Microbiol*
5. Cray, Villar (2008) "Cross-reactivity of anti-chicken IgY antibody with immunoglobulins of exotic avian species" *Vet Clin Pathol*
6. Das (2018) "Insights into the molecular, evolutionary and structural biology of circoviruses"
7. Das, Smith, Sarker et al. (2019) "Assessing circovirus gene flow in multiple spill-over events" *Virus Genes*
8. Das, Nath, Dhar et al. (2026) "Safety and immunogenicity of a novel psittacine beak and feather disease vaccine and optimisation of a thermostable spraydried formulation" *Vaccine*
9. Deng, Li, Zheng et al. "Tian K (2018) Establishment and application of an indirect ELISA for porcine circovirus 3" *Arch Virol*
10. Doneley (2003) "Acute beak and feather disease in juvenile African Grey parrots-an uncommon presentation of a common disease" *Aust Vet J*
11. Ellison, Hofmeister, Ribic et al. (2014) "Relatively high prevalence of pox-like lesions in Henslow's sparrow (Ammodrammushenslowii) among nine species of migratory grassland passerines in Wisconsin, USA" *J Wildl Dis*
12. Escandon, Heatley, Berghman et al. (2019) "Comparison of four anti-avian IgY secondary antibodies used in western blot and dot-blot ELISA to detect avian bornavirus antibodies in four different bird species" *Vet Med (Praha)*
13. Fairbanks, Steck, Wallach (1971) "Electrophoretic analysis of the major polypeptides of the human erythrocyte membrane" *Biochemistry*
14. Fassbinder-Orth, Wilcoxen, Tran et al. (2016) "Immunoglobulin detection in wild birds: effectiveness of three secondary anti-avian I g Y antibodies in direct ELISA s in 41 avian species" *Methods Ecol Evol*
15. Gardner, Greiner (2006) "Receiver-operating characteristic curves and likelihood ratios: improvements over traditional methods for the evaluation and application of veterinary clinical pathology tests" *Vet Clin Pathol*
16. Ge, Cui, Shi et al. (2018) "Development and application of an indirect enzyme-linked immunosorbent assay based on recombinant capsid protein for the detection of mink circovirus infection" *BMC Vet Res*
17. Greiner (1995) "Two-graph receiver operating characteristic (TG-ROC): a Microsoft-EXCEL template for the selection of cut-off values in diagnostic tests" *J Immunol Methods*
18. Greiner (1996) "Two-graph receiver operating characteristic (TG-ROC): update version supports optimisation of cut-off values that minimise overall misclassification costs" *J Immunol Methods*
19. Greiner, Gardner (2000) "Application of diagnostic tests in veterinary epidemiologic studies" *Prev Vet Med*
20. Greiner, Sohr, Göbel (1995) "A modified ROC analysis for the selection of cut-off values and the definition of intermediate results of serodiagnostic tests" *J Immunol Methods*
21. Greiner, Pfeiffer, Smith (2000) "Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests" *Prev Vet Med*
22. Gwet (2014) "Handbook of inter-rater reliability: the definitive guide to measuring the extent of agreement among raters" *Advanced Analytics*
23. Ha, Alley, Howe et al. (2013) "Evaluation of the pathogenicity of avipoxvirus strains isolated from wild birds in New Zealand and the efficacy of a fowlpox vaccine in passerines" *Vet Microbiol*
24. Hajian-Tilaki (2013) "Receiver operating characteristic (ROC) curve analysis for medical diagnostic test evaluation" *Caspian J Intern Med*
25. Hofmeister, Dusek, Fassbinder-Orth et al. (2016) "Susceptibility and antibody response of vesper sparrows (Pooecetes gramineus) to west Nile virus: a potential amplification host in sagebrush-grassland habitat" *J Wildl Dis*
26. Holroyd, Schiaffino, Chang et al. (2022) "Diagnostic accuracy of dried blood spots for serology of vaccine-preventable diseases: a systematic review" *Expert Rev Vaccines*
27. Jacobson (1998) "Validation of serological assays for diagnosis of infectious diseases" *Rev Sci Tech*
28. Johne, Raue, Grund et al. (2004) "Recombinant expression of a truncated capsid protein of beak and feather disease virus and its application in serological tests" *Avian Pathol*
29. Katoh, Ohya, Fukushi (2008) "Development of novel real-time PCR assays for detecting DNA virus infections in psittaciform birds" *J Virol Methods*
30. Khalesi (2007) "Studies of beak and feather disease virus infection"
31. Khalesi, Bonne, Stewart et al. (2005) "A comparison of haemagglutination, haemagglutination inhibition and PCR for the detection of psittacine beak and feather disease virus infection and a comparison of isolates obtained from loriids" *J Gen Virol*
32. Kiatipattanasakul-Banlunara, Tantileartcharoen, Katayama et al. (2002) "Psittacine beak and feather disease in three captive sulphurcrested cockatoos (Cacatua galerita) in Thailand" *J Vet Med Sci*
33. Kim, Park, Baek et al. (2024) "A duplex TaqMan probe-based real-time PCR assay for simultaneous detection of Psittacine beak and feather disease virus and Aves polyomavirus 1 from psittacine birds in Korea" *Korean J Vet Serv*
34. Latimer, Rakich, Kircher et al. (1990) "Extracutaneous viral inclusions in psittacine beak and feather disease" *J Vet Diagn Invest*
35. Lequin (2005) "Enzyme immunoassay (EIA)/enzyme-linked immunosorbent assay (ELISA)" *Clin Chem*
36. Li, Deng, Wang et al. (2025) "Indirect ELISA using multi-antigenic dominants of 3AB and 3C recombinant protein to detect antibodies against Senecavirus A in pigs" *Vet Sci*
37. Liu, Zhang, Zou et al. (2010) "Development of an indirect ELISA for the detection of duck circovirus infection in duck flocks" *Vet Microbiol*
38. Maccoll, Watson, Leseberg et al. (2024) "Beak and feather disease virus detected in the endangered Red Goshawk (Erythrotriorchis radiatus)" *Sci Rep*
39. Malsagova, Kopylov, Stepanov et al. (2020) "Dried blood spot in laboratory: directions and prospects" *Diagnostics*
40. Matefo, Armand, Samantha et al. (2022) "Validation of laboratory developed serology assays for detection of IgG antibody to severe acute respiratory syndrome coronavirus 2 in the South African population" *J Virol Methods*
41. Mercatali, Serra, Miserocchi et al. (2018) "Dried blood and serum spots as a useful tool for sample storage to evaluate cancer biomarkers" *J Vis Exp*
42. Nahm (2022) "Receiver operating characteristic curve: overview and practical use for clinicians" *Korean J Anesthesiol*
43. Nath, Das, Roby et al. (2021) "Structural perspectives of beak and feather disease virus and porcine circovirus proteins" *Viral Immunol*
44. Neef, Nath, Das et al. (2024) "Recombinantly expressed virus-like particles (VLPs) of canine circovirus for development of an indirect ELISA" *Vet Res Commun*
45. Ogawa, Yamaguchi, Fukushi (2005) "Duplex shuttle PCR for differential diagnosis of budgerigar fledgling disease and psittacine beak and feather disease" *Microbiol Immunol*
46. Ohyama (2021) "Statistical inference of Gwet's AC1 coefficient for multiple raters and binary outcomes" *Commun Stat Methods*
47. Oliveira, Folgueras-Flatschart, Flatschart et al. (2007) "Indirect ELISA for the detection of IgG specific to Newcastle disease virus in quail serum" *Arq Bras Med Vet Zootec*
48. Pass, Perry (1984) "The pathology of psittacine beak and feather disease" *Aust Vet J*
49. Patterson, Forwood, Raidal (2012) "Homology modelling and structural comparisons of capsid-associated proteins from circoviruses reveal important virus-specific surface antigens" *Crystal Struct Theory Appl*
50. Patterson, Swarbrick, Roman et al. (2013) "Differential expression of two isolates of beak and feather disease virus capsid protein in Escherichia coli" *J Virol Methods*
51. Peters, Patterson, Baker et al. (2014) "Evidence of psittacine beak and feather disease virus spillover into wild critically endangered orange-bellied parrots (Neophema chrysogaster)" *J Wildl Dis*
52. Raidal, Cross (1994) "The haemagglutination spectrum of psittacine beak and feather disease virus" *Avian Pathol*
53. Raidal, Cross (1995) "Acute necrotizing hepatitis caused by experimental infection with psittacine beak and feather disease virus" *J Avian Med Surg*
54. Raidal, Sabine, Cross (1993) "Laboratory diagnosis of psittacine beak and feather disease by haemagglutination and haemagglutination inhibition" *Aust Vet J*
55. Ritchie, Niagro, Lukert et al. (1989) "Characterization of a new virus from cockatoos with psittacine beak and feather disease" *Virology*
56. Ritchie, Niagro, Latimer et al. (1990) "Ultrastructural, protein composition, and antigenic comparison of psittacine beak and feather disease virus purified from four genera of psittacine birds" *J Wildl Dis*
57. Ritchie, Niagro, Latimer et al. (1991) "Hemagglutination by psittacine beak and feather disease virus and use of hemagglutination inhibition for detection of antibodies against the virus" *Am J Vet Res*
58. Robichon, Luo, Causey et al. (2011) "Engineering Escherichia coli BL21 (DE3) derivative strains to minimize E. coli protein contamination after purification by immobilized metal affinity chromatography" *Appl Environ Microbiol*
59. Samsonova, Saushkin, Osipov (2022) "Dried blood spots technology for veterinary applications and biological investigations: technical aspects, retrospective analysis, ongoing status and future perspectives" *Vet Res Commun*
60. Sanada, Sanada (2000) "The sensitivities of various erythrocytes in a haemagglutination assay for the detection of psittacine beak and feather disease virus" *J Vet Med*
61. Sarker, Forwood, Ghorashi et al. (2015) "Beak and feather disease virus genotypes in Australian parrots reveal flexible host-switching" *Aust Vet J*
62. Sarker, Ghorashi, Swarbrick et al. (2015) "An efficient approach for recombinant expression and purification of the viral capsid protein from beak and feather disease virus (BFDV) in Escherichia coli" *J Virol Methods*
63. Sarker, Terrón, Khandokar et al. (2016) "Structural insights into the assembly and regulation of distinct viral capsid complexes" *Nat Commun*
64. Schoemaker, Dorrestein, Latimer et al. (2000) "Severe leukopenia and liver necrosis in young African grey parrots (Psittacus erithacus erithacus) infected with psittacine circovirus" *Avian Dis*
65. Shah, Maghsoudlou (2016) "Enzyme-linked immunosorbent assay (ELISA): the basics" *Br J Hosp Med*
66. Sharma, Jain (2014) "Right choice of a method for determination of cut-off values: a statistical tool for a diagnostic test" *Asian J Med Sci*
67. Sharma, Jaiswal, Shukla et al. (2014) "Dried blood spots: concepts, present status, and future perspectives in bioanalysis" *Drug Test Anal*
68. Shearer, Bonne, Clark et al. (2008) "Development and applications of a monoclonal antibody to a recombinant beak and feather disease virus (BFDV) capsid protein" *J Virol Methods*
69. Shearer, Sharp, Bonne et al. (2009) "A blocking ELISA for the detection of antibodies to psittacine beak and feather disease virus (BFDV)" *J Virol Methods*
70. Shearer, Sharp, Bonne et al. (2009) "A quantitative, real-time polymerase chain reaction assay for beak and feather disease virus" *J Virol Methods*
71. Swets (1988) "Measuring the accuracy of diagnostic systems" *Science*
72. Tang, Stratton, Tang (2013) "Advanced techniques in diagnostic microbiology"
73. Team (2010) "R: A language and environment for statistical computing"
74. Unal (2017) "Defining an optimal cut-point value in ROC analysis: an alternative approach" *Comput Math Methods Med*
75. Varsani, Regnard, Bragg et al. (2011) "Global genetic diversity and geographical and host-species distribution of beak and feather disease virus isolates" *J Gen Virol*
76. Walker, Crowther (2009) *Methods Mol Biol*
77. Wang, Wang, Duan et al. (2020) "Self-assembly into virus-like particles of the recombinant capsid protein of porcine circovirus type 3 and its application on antibodies detection" *AMB Express*
78. Wongpakaran, Wongpakaran, Wedding et al. (2013) "A comparison of Cohen's Kappa and Gwet's AC1 when calculating inter-rater reliability coefficients: a study conducted with personality disorder samples" *BMC Med Res Methodol*
79. Wright, Nilsson, Van Rooij et al. (1993) "Standardisation and validation of enzyme-linked immunosorbent assay techniques for the detection of antibody in infectious disease diagnosis" *Rev Sci Tech*
80. Xu, Lorber (2014) "Interrater agreement statistics with skewed data: evaluation of alternatives to Cohen's kappa" *J Consult Clin Psychol*
81. Xu, Lohr, Greiner (1997) "The selection of ELISA cut-off points for testing antibody to Newcastle disease by two-graph receiver operating characteristic (TG-ROC) analysis" *J Immunol Methods*
82. Xu, Zhang, Gao et al. (2025) "Expression of chicken circovirus Cap protein and establishment of ELISA method for antibody detection" *Vet Res Commun*
83. Ypelaar, Bassami, Wilcox et al. (1999) "A universal polymerase chain reaction for the detection of psittacine beak and feather disease virus" *Vet Microbiol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12793669&blobtype=pdf | # Abstract citation ID: ofaf695.2366 P-2203. Cytomegalovirus Reactivation in Patients Treated with Bispecific Antibodies (BsAbs)
Tori Pravato, Nina Orsini, Jungwook Kang, Krishna Shah, Yuxuan Li, Susan Seo, Genovefa Papanicolaou
Background. Use of BsAbs for acute lymphocytic leukemia and lymphoma has been associated with cytomegalovirus (CMV) viremia and end-organ disease (EOD), with increased risk due to patients' underlying malignancies, prior chemotherapies, treatment-related complications (e.g., cytokine release syndrome), and prolonged cytopenias. There is no standard for CMV monitoring or prophylaxis after BsAb therapy. The primary objective of the study was to estimate the rate of CMV viremia following treatment with BsAbs.
Methods. A total of 68 patients treated with blinatumomab, epcoritamab, glofitamab, or mosunetuzumab for leukemia and lymphoma were screened for CMV IgG from 09/2024 to 04/2025. Twenty-six CMV-seropositive adults were included and monitored prospectively for CMV by a quantitative PCR in the plasma (lower limit of detection 34.5 IU/mL) at least monthly from first BsAb infusion through 6 months, receipt of allogeneic hematopoietic stem cell transplantation (HSCT) or chimeric antigen receptor T-cell (CAR-T) therapy, or death, whichever occurred first. Demographics and BsAb type and doses were extracted from patient records. The dose and duration of concomitant steroids and antiviral treatment were collected. CMV viral loads, CMV-directed treatment, and outcomes were captured.
Results. Of 26 patients analyzed, the median follow-up after first BsAb infusion was 3 months (IQR, 2-4). Seven (27%) patients received steroids at doses > 20 mg of prednisone equivalent per day, and one (4%) patient had a history of prior allogeneic HSCT. Ten (38%) patients developed detectable CMV viremia at a median 18.5 days (IQR,(15)(16)(17)(18)(19)(20)(21)(22) from first BsAb infusion. Subsequent PCR tests indicated clearance or stable detection < 1000 IU/mL. Two (8%) patients who received high dose steroids with no previous HSCT had maximum CMV PCR levels above 1000 IU/mL, and only one (4%) required treatment with (val)ganciclovir for CMV colitis for a total of 26 days. There was no CMV-related mortality.
Conclusion. One-third of patients had transient low level CMV viremia that resolved spontaneously. One (4%) patient had CMV-related EOD. Despite increasing BsAb use, low rates of CMV reactivation were observed. Further research is ongoing. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12584705&blobtype=pdf | # Comparison of respiratory pathogen infections in hospitalized patients before and during the COVID-19 pandemic in Shanghai, China
Xiaoxiao Li, Jie Wang, Junhua Dai, Fenfen Xiang, Zixi Chen, Mengzhe Zhang, Jiawen Qian, Rong Wu
## Abstract
This study aimed to assess the impact of COVID-19 on the prevalence of respiratory pathogens among hospitalized patients with respiratory tract infections in Shanghai, China. Patients with acute respiratory infections (ARIs) hospitalized at Putuo Hospital, Shanghai University of Traditional Chinese Medicine from January 2017 to December 2022 were collected. Indirect immunofluorescence assay (IFA) was used to detect the IgM antibody of nine common pathogens, including Chlamydia pneumoniae (CP), Mycoplasma pneumoniae (MP), parainfluenza virus (PIV), respiratory syncytial virus (RSV), influenza virus A (FluA), influenza virus B (FluB), adenovirus (ADV), Legionella pneumophila (Lp), and Coxiella burnetii (Cb). A total of 24,933 eligible patients were enrolled. The overall detection rate of respiratory pathogens in the pre-outbreak group (2017-2019, 32.05%) was significantly higher than that in the outbreak group (2020-2022, 11.48%, P < 0.001). MP (16.03%), Lp (2.43%), FluB (1.73%), and PIV (1.64%) were the main pathogens. Positivity for MP, FluB, and PIV declined significantly, whereas the detection rate of Lp was increased (P < 0.05) during the 2020-2022 period com pared with the 2017-2019 period. Children (73.21%) experienced a significantly higher infection rate than adolescents (65.35%), adults (25.20%), and older adults (15.40%). MP+PIV was the most common co-detection pattern. The detection of CP, MP, PIV, RSV, FluB, ADV, and Lp differed significantly between the two periods. From 2020 to 2022, the detection rates of CP, MP, PIV, RSV, FluB, and ADV decreased, whereas the detection of Lp increased. Knowledge of common pathogens' dynamics could serve as a reference for further prevention and control of ARIs. IMPORTANCE This study leverages the changes in the common respiratory spectrum pre-pandemic and during the COVID-19 pandemic in hospitalized patients in Shanghai. These data may serve as a scientific foundation for the prevention and management of ARIs. Doctors and policymakers should pay attention to the changes in the epidemic trends and types of respiratory pathogens and maintain monitoring of respiratory pathogens to better control the prevalence of respiratory pathogens. KEYWORDS respiratory pathogens, COVID-19, acute respiratory infections, hospitaliza tion R espiratory tract infections (RTIs) impose a significant health burden, contributing to morbidity and mortality risks across all age groups (1). RTIs are caused by a diversity of pathogens, including bacteria, viruses, and fungi (2). Mycoplasma pneu moniae (MP), respiratory syncytial virus (RSV), and influenza (Flu) are recognized as significant pathogens in RTI patients, and the prevalence of those pathogens varies by countries, regions, seasons, and periods (3,4). Previous studies have shown that non-pharmaceutical interventions (NPIs), including wearing masks, hand hygiene, and
social distancing, may have had a significant impact on the transmission of common respiratory pathogens during the COVID-19 pandemic (5)(6)(7). However, the changes in the spectrum of common respiratory pathogens before and during the COVID-19 pandemic among hospitalized patients in Shanghai remain unclear.
To comprehensively assess the infection rates of common respiratory pathogens before and during the COVID-19 pandemic, we retrospectively analyzed the results of respiratory pathogen detection among hospitalized patients from January 2017 to December 2022 using a multiple indirect immunofluorescence assay (IFA) kit. This study reveals dynamic variations before and after the outbreak of COVID-19, clarifies the impact of NPIs on non-COVID respiratory pathogens across all age groups, and provides valuable insights into the treatment and prevention of RTI.
## MATERIALS AND METHODS
## Study participants
From January 2017 to December 2022, patients with ARIs admitted to the Department of Respiratory Medicine and Pediatrics of Putuo Hospital, Shanghai, China, were enrolled, including cases of both upper and lower respiratory tract infections. Inclusion criteria for individuals were as follows: (i) at least one manifestation of acute infection (fever [≥37.5°C], chills, or abnormal white blood cell differential) and (ii) at least one of the listed respiratory tract clinical manifestations (rhinorrhea, cough, sputum, shortness of breath, lung auscultation abnormality, or chest pain). Finally, a total of 24,933 patients with results of multiple indirect immunofluorescence assay (IFA) testing were enrolled in this study. All of the enrolled patients were divided into four age groups: children (≤5 years), adolescents (6-17 years), adults (18-60 years), and older adults (>60 years). The period between January 2017 and December 2019 was classified as before the COVID-19 pandemic, whereas the period between January 2020 and December 2022 was classified as during the COVID-19 pandemic. All methods were performed in accordance with the relevant guidelines and regulations. This study was approved by the Ethics Committee of Putuo Hospital, Shanghai University of Traditional Chinese Medicine. As the retrospective analysis was based on anonymized data, the need for individual informed consent was waived by the Institutional Review Board of Shanghai Putuo District Central Hospital.
## Sample collection
Three milliliters of venous blood were drawn from each patient. The samples were centrifuged at 2,000 × g for 10 min at 4°C. The serum was separated and stored at -20°C until assayed with the Pneumoslide IgM test.
## Pneumoslide IgM test (Vircell, Granada, Spain)
Atypical pathogens and respiratory viruses, including Chlamydia pneumoniae (CP), Mycoplasma pneumoniae (MP), parainfluenza virus (PIV), respiratory syncytial virus (RSV), influenza virus A (FluA), influenza virus B (FluB), adenovirus (ADV), Legionella pneumo phila (Lp), and Coxiella burnetii (Cb) in the serum were detected using the Pneumoslide IgM kit (Vircell, Granada, Spain) in accordance with the standard operating procedures. IFAs for all nine pathogens were performed uniformly for each enrolled patient during the acute phase of illness, based on the inclusion criteria of the study. Each slide has 10 wells, with each well containing one of the above pathogen antigens and a cell control. Serum samples were diluted 1:1 with phosphate-buffered saline (PBS) and treated with anti-human IgG sorbent. The sorbent-treated diluted serum was added to every well and incubated for 90 min at 37°C, and then, the slide was washed twice with PBS and dried. The fluorescent IgM secondary antibody was added to the wells and incubated at 37°C for 30 min. The slide was washed twice with PBS, and the fluorescent signal was detected under a fluorescence microscope (EUROStar III Plus). Apple-green fluorescence was observed in the nucleus, cytoplasm, and/or periphery in 1%-15% of the cells for positive samples with ADV, FluA, FluB, RSV, or PIV (with colored syncytial cells observed simultaneously in PIV and RSV). All bacteria in the case of Lp, CP, or Cb exhibit apple-green fluorescence. Apple-green fluorescence was observed in the periphery of the cell for positive samples for MP. A negative sample showed no fluorescence for Lp, CP, and Cb, and a red cellular pattern for MP, ADV, FluA, FluB, RSV, and PIV (8).
## Statistical analysis
Excel 2010 and SPSS 22.0 statistical software were used for data processing and analysis. Bubble plots were created with the ggplot2 and reshape2 packages in R (version 4.1.2). The categorical variables were summarized as frequencies and proportions. Chi-square tests were used to compare the positive detection rates of various pathogens in the respiratory tract among different groups. The linear-by-linear association and gamma values were used to evaluate the trend in pathogen prevalence over 6 years, and a P value < 0.05 was considered statistically significant for all analyses.
## RESULTS
## Demographic characteristics
In this study, we collected a total of 24,933 ARI cases from 2017 to 2022 in Shanghai, China, including 13,374 (53.64%) males and 11,559 (46.36%) females. The mean age of patients was 64.88 ± 25.49 years, with 59.99 ± 28.67 years for patients before the COVID-19 pandemic (2017-2019) and 70.86 ± 19.24 years for patients during the COVID-19 pandemic (2020-2022) (Table 1). The common respiratory pathogens were MP (16.03%), Lp (2.32%), FluB (1.73%), and PIV (1.64%). The detection rates of all pathogens significantly differed between years (P < 0.05) (Table 2). Changes in pathogen detection rates in each year are shown in Fig. 1A.
## Comparison of the detection rates of respiratory pathogens before and during the COVID-19 pandemic
Before the COVID-19 pandemic, the total detection rate of respiratory pathogens was 32.05% (2017-2019), and after the implementation of control measures due to the COVID-19 pandemic, the detection rate decreased to 11.48% (2020-2022) (P < 0.001).
Overall, the detection rates of CP, MP, PIV, RSV, FluB, and ADV decreased during the COVID-19 pandemic, whereas that of Lp increased (P = 0.003). There were no significant differences in the detection rates of FluA and Cb (P > 0.05) (Table 1). More specifically, before the COVID-19 pandemic, the MP, FluB, and PIV ranked in the top three, with detection rates of 22.82%, 3.01%, and 2.59%, respectively. However, during the COVID-19 pandemic, the detection rate of FluB decreased significantly (P < 0.001). Lp replaced FluB and became the top two pathogens in the detection rate. Finally, the three pathogens with the highest detection rates were MP, Lp, and PIV, with 7.73%, 2.63%, and 0.49%, respectively (Table 2).
## Age-specific distribution and positivity rates
Among patients with ARIs tested for all the nine pathogens, the highest rate of pathogen detection was seen in children (aged ≤5 years, 73.21% [1,066/1,456]), followed by 65.35% (775/1,186) in adolescents aged 6-17 years, 25.20% (1,062/4,215) in adults aged 18-60 3). Changes in pathogen detection rates in each year group are shown in Fig. 1B. Positive detection rates in different age groups were compared between the pre-pandemic (2017-2019) and the pandemic (2020-2022) periods (Fig. 2). Before the COVID-19 pandemic, MP, FluB, and PIV ranked among the top three in the detection of pathogens in the children and adolescents groups. MP and Lp were the top two in the detection of pathogens in adults.
During the COVID-19 pandemic, MP, Lp, and PIV were the top three in the detection of pathogens in all age groups (Fig. 2).
## Temporal distribution of respiratory pathogens
The monthly distributions of each pathogen before and during the COVID-19 pandemic are shown in
## Co-detection pattern of pathogens
Co-detections, in which more than one pathogen tested positive, were observed in 807 specimens, with a detection rate of 3.24% of all specimens. Among the 807 co-infection cases, double infections were identified in 726 (89.96%) cases, including 227 (28.13%) cases of MP + PIV, 226 (28.00%) cases of MP + FluB, and 154 (19.08%) cases of MP + Lp. Triple infections were detected in 71 (8.80%) cases, and quadruple infections were detected in 10 (1.24%) cases (Table 4). The commonly encountered co-infection patterns were MP and FluB, MP and PIV, and MP and Lp, as shown in Figure 4.
## DISCUSSION
In this study, we analyzed the detection of respiratory pathogens in ARI among hospitalized patients in Shanghai from 2017 to 2022. The total detection rate was 22.81% in Shanghai, which was similar to that of a previous study in Gansu Province (29.2%) (9), and higher than that reported in Shenzhen (14.55%) (10), Beijing (5.64%) (11), and north China (7.6%) (12), but lower than that in Shaanxi Province (36.01%) (13), Shandong Province (35.75%) ( 14), and Xiamen (56.36%) (15). These differences may be affected by multiple factors, including geographic location, investigated period, climate conditions, study population, and methodological approaches to pathogen detection.
Our study showed that the overall detection rates of respiratory pathogens were droplet and contact precautions (face mask use and increased hand hygiene), societal restrictions (school closures and reduced workplace attendance), isolation of infected individuals, and vaccination, were implemented to curb the spread of COVID-19, which greatly reduced the prevalence of the common respiratory pathogens. Furthermore, other confounding factors, such as reduced healthcare access and changes in patient populations, may also have contributed to the reduction in overall pathogen detection rates.
The results showed that MP was the predominant pathogen with the highest detection rate, followed by Lp, FluB, and PIV. Previous data also showed that MP was the most common atypical bacterium in other studies (16,17). During the COVID-19 pandemic (2020-2022), a significant decrease in the detection rates of CP, MP, PIV, RSV, FluB, and ADV was observed in comparison to the pre-pandemic period (2017-2019). Interestingly, we also noticed a significant increase in the positive rate of Lp, indicating that not all pathogens were restricted by positive prevention (18). This may be explained by the fact that Lp infections are directly contracted from environmental sources and can be transmitted in healthcare or senior-living settings (19). Lp is mostly spread by inhaling infected aerosols or dust aspiration from contaminated soil. In general, it is accepted that water stagnation and poor maintenance of the water system in buildings are risk factors for Lp growth. In the context of the COVID-19 pandemic, many public health institutions have been severely affected by "stay-at-home" orders. All non-urgent hospital activities were suspended, and some wards were closed, with a consequent reduction in the use of the water system, the formation of stagnant water, and dimin ished disinfection in hospital water networks and cisterns. These conditions may have increased the risk of hospitalized patients' exposure to waterborne pathogens, including Lp, thereby raising the Lp infection of patients (20). A previous study revealed that the hospital water network of the three examined wards closed for 3 months because of the COVID-19 emergency showed a higher Lp contamination after the lockdown period (21). Another study conducted in Spain has demonstrated that hotels that suffered the longest prolonged closures (>3 months) could have carried a higher risk of exposure to Lp in the domestic hot water system (22). Furthermore, the incidence of Lp has been high since the onset of the COVID-19 pandemic in Japan (19).
FluA and FluB are major contributors to seasonal epidemics. A recent study showed that NPIs had a strong suppressive effect on FluA and FluB, with the highest cumula tive positivity rate of FluA + FluB in 2023 (31.9%) and the lowest rate in 2021 (2.0%) (23). In our study, the highest positivity rate of FluA + FluB was observed in 2018 (5.7%) and the lowest rate in 2022 (0.1%), which were significantly lower than those reported by the Chinese National Influenza Center (CNIC) and hospital-based data in Chengdu (23,24). Moreover, it has been reported that FluA and FluB nearly disappeared during the COVID-19 phase among children in Guangzhou, China, consistent with our results (25). However, we should also note that pre-pandemic, pandemic, and post-pan demic comparative analysis could be influenced by multiple factors, such as changes in healthcare access (more severe hospitalized and tested patients) and diagnostic practices for pathogens in different phases.
It has been reported that children were notably susceptible to respiratory pathogens both before and during the COVID-19 pandemic (26). The reason might be explained by the lower innate immunity response of children than that of adults (27). Our results also showed that young children aged ≤5 years (73.21%) exhibited a significantly higher positive rate of common respiratory pathogen infections compared with adults (25.20%) and the elderly (15.40%), consistent with a recently published study (28). Therefore, it should be emphasized that young children are at higher risk of being infected with respiratory pathogens. During the COVID-19 pandemic (2020-2022), the detection rates of predominant pathogens among children in Shanghai were MP, Lp, and PIV. In comparison, in the pre-pandemic period (2017-2019), the top three detected pathogens were MP, FluB, and PIV. The pathogen spectrum of children has been impacted by the COVID-19 pandemic. In a national data covering the 2009-2019 period for children, RSV, FluA + B, human rhinovirus (HRV), MP, and PIV were the top five detected pathogens (29).
MP infections usually occur in winter and spring but can happen throughout the year, which was consistent with the seasonal pattern of MP during 2017-2019. Nota bly, the positive rate of MP markedly declined in 2020-2022 and showed no obvious seasonality. Additionally, an obvious detection peak of ADV was observed in the spring of the pre-pandemic. No detection peak was observed during the pandemic, with low incidence throughout the year. These findings were similar to the results reported by Xu et al. (30). The detection peak of PIV was observed in spring and summer in 2017-2019, which was consistent with seasonality reported in hospitalized children with lower respiratory tract infections (LRTIs) (31). However, PIV was detectable at a low rate throughout the year during the pandemic. Many studies have shown that NPIs associated with reduced transmission of COVID-19 have also reduced influenza (32,33). Indeed, the common prevalent seasonal pathogens, such as RSV and Flu, with a few cases, were observed during COVID-19, and other studies also showed similar findings (34,35). Lp infections were mainly observed in summer and autumn in Shaanxi Province, northwest China (13). We discovered that Lp was more common in the hot or cold seasons during COVID-19.
In this study, co-infections with at least two pathogens were only detected in 3.24% of the patients, which appears lower than those in some previous studies reported by Zhao et (13). This may be explained by variation in different diagnostic sensitivities, demographic characteristics, bacterial and viral types, different regions, and the investigated period. It has been reported that bacterial and viral co-infection may present more severe clinical outcomes (37), and MP + RSV was the most common type of viral-atypical bacterial co-infections (38). In this study, the most common combination was MP + PIV, which was consistent with that reported in Lanzhou, China (39). A previous study also indicated MP was the most frequently detected pathogen in co-infections with PIV type 3 (PIV3) (40). Due to the lack of information on its clinical severity, we are unable to determine the significance of co-infection in this study. Some studies have demonstrated that multiple infections were associated with prolonged hospital stay, admission to intensive care units, long-term mechanical ventilation support, and death (41).
There are some limitations to our study. First, this study was performed in a single center in a restricted geographic area, potentially impacting the generalizability of the findings to other populations or regions. Although the sample size was sufficient for a preliminary analysis, it may not have been large enough to detect minor effects or to fully explore the interactions between pathogens and demographic factors. Second, only the Pneumoslide IgM test was employed: factors such as disease progression, nutrition, and immune status can affect the production of antibodies, which may underestimate the positivity rate of certain pathogens. Moreover, the testing kit does not encompass all pathogens; it cannot cover all possible co-infecting viral, bacterial, or fungal respiratory pathogens. Our study did not include testing for emerging pathogens such as SARS-CoV-2, which limits the comprehensiveness of the results. Subsequent studies should broaden the detection spectrum of respiratory pathogens, enriching the epidemiologi cal profile of ARIs. Third, the lack of comprehensive clinical data, including symptoms, laboratory results, and treatments, impeded the assessment of the association between these factors and pathogen positivity or disease severity. Finally, emphasizing the need for caution when extrapolating causality (e.g., NPI impacts) in this observational study, as other unanalyzed factors (such as environmental and health-seeking behavior changes, viral competition, or pathogen evolution) may have contributed to the observed shifts.
## Conclusion
In conclusion, this study showed that during the COVID-19 pandemic (2020-2022), the overall pathogen detection rate has significantly decreased, and the seasonal patterns of certain pathogens have also changed. An unknown number of variables, including NPIs, might be responsible for these changes. MP, Lp, and FluB were the most common respiratory pathogens, with children experiencing significantly higher infection rates. Strengthening vaccination coverage and implementing region-specific public health strategies that account for local environmental and social factors will be essential for mitigating the burden of respiratory pathogen infections in the post-pandemic period. Furthermore, hospitals and policymakers should continuously monitor the epidemiolog ical and evolutionary dynamics of multiple respiratory pathogens to inform targeted intervention strategies and vaccination programs, thereby facilitating the effective management of acute respiratory infections.
## References
1. Li, Wang, Blau et al. (2022) "Global, regional, and national disease burden estimates of acute lower respiratory infections due to respiratory syncytial virus in children younger than 5 years in 2019: a systematic analysis" *The Lancet*
2. Kozinska, Wegrzynska, Komiazyk et al. (2022) "Viral etiological agent(s) of respiratory tract infections in symptomatic individuals during the second wave of COVID-19 pandemic: a single drive-thru mobile collection site study" *Pathogens*
3. Kruizinga, Peeters, Van Veen et al. (2021) "The impact of lockdown on pediatric ED visits and hospital admissions during the COVID19 pandemic: a multicenter analysis and review of the literature" *Eur J Pediatr*
4. Haixu, Haibin (2016) "Detection of 20 respiratory viruses and bacteria by influenza-like illness surveillance in Beijing" *J Infect*
5. Fukuda, Tsugawa, Nagaoka et al. (2019) "2021 Surveillance in hospitalized children with infectious diseases in Japan: pre-and post-coronavirus disease" *J Infect Chemother*
6. Huang, Wood, Jelley et al. (2021) "Impact of the COVID-19 nonpharmaceutical interventions on influenza and other respiratory viral infections in New Zealand" *Nat Commun*
7. Redlberger-Fritz, Kundi, Aberle "Puchhammer-Stöckl E. 2021. Significant impact of nationwide SARS-CoV-2 lockdown measures on the circulation of other respiratory virus infections in Austria" *J Clin Virol*
8. Qin, Zhang, Chen et al. (2020) "Antibodies against atypical pathogens and respiratory viruses detected by pneumoslide IgM test in adults with community-acquired pneumonia in Guangzhou City" *J Clin Lab Anal*
9. Li, Li, Meng et al. (2011) "Viral etiologies and epidemiology of patients with acute respiratory infections based on sentinel hospitals in Gansu Province" *J Med Virol*
10. Wang, Zheng, Deng et al. (2016) "Prevalence of respiratory viruses among children hospitalized from respiratory infections in Shenzhen, China" *Virol J*
12. Yao, Wang, Wei et al. (2017) "Human adenovirus among hospitalized children with respiratory tract infections in Beijing" *Virol J*
13. Yu, Xie, Zhang et al. (2012) "Comparison of the prevalence of respiratory viruses in patients with acute respiratory infections at different hospital settings in north China" *BMC Infect Dis*
14. Liu, Li, Yang et al. (2023) "Viral and atypical pathogen's epidemiology of a large cohort of patients with acute respiratory tract infections in Shaanxi province, northwest China" *Int J Gen Med*
15. Liu, Li, Zhang et al. (2015) "Viral etiology of acute respiratory tract infections in hospitalized children and adults in Shandong province, China" *Virol J*
16. Hong, Li, Wei et al. (2023) "Epidemiology of respiratory pathogens in patients with acute respiratory tract infection in Xiamen"
17. Yan, Sun, Ji et al. (2023) "High incidence of the virus among respiratory pathogens in children with lower respiratory tract infection in northwestern China" *J Med Virol*
18. Kumar (2018) "Mycoplasma pneumoniae: a significant but underrated pathogen in paediatric community-acquired lower respiratory tract infections" *Indian J Med Res*
19. Geng, Hao, Xu et al. (2022) "Clinical features and viral etiology of acute respiratory infection in an outpatient fever clinic during COVID-19 pandemic in a tertiary hospital in Nanjing" *China. J Clin Lab Anal*
20. Yamasue, Komiya, Kinjo et al. (2023) "Rebound mortality rate of Legionella pneumonia in Japan" *Respir Investig*
21. Montagna, Giglio, Napoli et al. (2018) "Control and prevention measures for legionellosis in hospitals: a cross-sectional survey in Italy" *Environ Res*
22. De Giglio, Diella, Lopuzzo et al. (2020) "Impact of lockdown on the microbiological status of the hospital water network during COVID-19 pandemic" *Environ Res*
23. Molina, Bennassar, Palacio et al. (2023) "Impact of prolonged hotel closures during the COVID-19 pandemic on Legionella infection risks" *Front Microbiol*
24. Li, Yang, Chen et al. (2019) "Epidemiological trends of influenza A and B in one hospital in Chengdu and national surveillance data" *Front Cell Infect Microbiol*
25. Li, Ma, Li et al. (2025) "One-year epidemiological patterns of respiratory pathogens across age, gender, and seasons in Chengdu during the post-COVID era" *Sci Rep*
26. Liang, Guo, Zhu et al. (2025) "Changes in the epidemic patterns of respiratory pathogens of children in Guangzhou, China during the COVID-19 pandemic" *BMC Infect Dis*
27. Wu, Zhang, Mo (2023) "Analysis of respiratory virus detection in hospitalized children with acute respiratory infection during the COVID-19 pandemic" *Virol J*
28. Levy (2007) "Innate immunity of the newborn: basic mechanisms and clinical correlates" *Nat Rev Immunol*
29. Wei, Li, Lu et al. (2024) "Changing respiratory pathogens infection patterns after COVID-19 pandemic in Shanghai, China" *J Med Virol*
30. Li, Zhang, Ren et al. (2021) "Etiological and epidemiological features of acute respiratory infections in China" *Nat Commun*
31. Xu, Liu, Su et al. (2022) "Comparison of respiratory pathogens in children with lower respiratory tract infections before and during the COVID-19 pandemic in Shanghai" *China. Front Pediatr*
32. Xiao, Duan, Xie et al. (2016) "Human parainfluenza virus types 1-4 in hospitalized children with acute lower respiratory infections in China" *J Med Virol*
33. Cowling, Ali, Ng et al. (2020) "Impact assessment of non-pharmaceutical interventions against coronavirus disease 2019 and influenza in Hong Kong: an observational study" *Lancet Public Health*
34. Li, Wang, Liu et al. (2021) "Comparison of 11 respiratory pathogens among hospitalized children before and during the COVID-19 epidemic in Shenzhen, China" *Virol J*
35. Bhardwaj, Choudhary, Jadhav et al. "2022 A retrospective analysis of respiratory virus transmission before and during the COVID-19 pandemic in Pune the western region of India" *Front Public Health*
36. Tang, Bialasiewicz, Dwyer et al. (2021) "Where have all the viruses gone? disappearance of seasonal respiratory viruses during the COVID-19 pandemic" *J Med Virol*
37. Zhao, Zhang, Wang et al. (2024) "Epidemiology of respiratory pathogens in patients with acute respiratory infections during the COVID-19 pandemic and after easing of COVID-19 restrictions" *Microbiol Spectr*
38. Lim, Kweon, Kim et al. (2018) "Impact of bacterial and viral coinfection in community-acquired pneumonia in adults" *Diagn Microbiol Infect Dis*
39. Liu, Xu, He et al. (2013) "Epidemiol ogy of respiratory pathogens in children with lower respiratory tract infections" *Jpn J Infect Dis*
40. Wang, Pan, Gao et al. (2019) "Analysis of respiratory pathogens in pediatric acute respiratory infections in Lanzhou, northwest China" *Front Cell Infect Microbiol*
41. Lu, Cheng, Hu (2023) "Parainfluenza virus type 3 co-infection with other respiratory pathogens among hospitalized children with acute respiratory infections in Wuhan, China" *Jundishapur J Microbiol*
42. Chauhan, Slamon (2017) "The impact of multiple viral respiratory infections on outcomes for critically Ill children" *Pediatr Crit Care Med* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12607657&blobtype=pdf | # Stoichiometry of HIV-1 envelope glycoprotein protomers with changes that stabilize or destabilize the pretriggered conformation
Zhiqing Zhang, Saumya Anang, Qian Wang, Hanh Nguyen, Hung-Ching Chen, Ta-Jung Chiu, Derek Yang, Amos Smith, Joseph Sodroski
## Abstract
During human immunodeficiency virus (HIV-1) entry into host cells, binding to the receptors, CD4 and CCR5/CXCR4, triggers conformational changes in the metastable envelope glycoprotein (Env) trimer ((gp120-gp41) 3 ). CD4 binding induces Env to make transitions from its pretriggered conformation (PTC) to more "open" conformations that are sensitive to inhibition by antibodies, CD4-mimetic compounds (CD4mcs), and exposure to cold. Changes in functional membrane Envs that either stabilize or destabilize the PTC have been identified. Here, we investigate the stoi chiometric requirements for the PTC-stabilizing and -destabilizing changes in the Env protomers. To this end, we generated viruses bearing Envs with mixed protomers exhibiting different degrees of PTC stability and determined the sensitivity of the viruses to cold (0°C) and, in some cases, to a CD4mc. The number of stabilized Env protomers required to achieve stabilization of the PTC was inversely related to the degree of PTC stabilization that resulted from the introduced Env change. For strongly stabilizing Env changes, modification of a single protomer was sufficient to achieve PTC stabilization; apparently, with adequate stability, the modified protomer can constrain the conforma tion of the other two protomers to maintain the PTC. Weakly stabilizing Env changes needed to be present in all three protomers to achieve efficient stabilization of the PTC. In many cases, the PTC was disrupted when destabilizing changes were present in only a single protomer. These complementary results suggest that conformational symmetry among the protomers of the functional Env trimer is conducive to the integrity of the PTC.
IMPORTANCEThe human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer consists of three protomers. In response to receptor binding, the flexible Env changes its conformation to mediate virus entry into host cells. The shape-shifting ability of Env also contributes to HIV-1's capacity to evade the host immune system. The pretriggered (state 1) conformation (PTC) of Env is an important target for virus entry inhibitors and host antibodies but is unstable and therefore incompletely characterized. Changes in Env amino acids that either stabilize or destabilize the PTC have been identified. Here, we define how many Env protomers need to be modified by these changes to achieve stabilization or destabilization of the PTC. These results can guide the placement of changes in the HIV-1 Env protomers to control the movement of the Env trimer from the PTC, allowing better characterization of this elusive conformation and testing of its utility in vaccines.
T he human immunodeficiency virus (HIV-1) envelope glycoprotein (Env) trimer is composed of three protomers, each of which consists of a gp120 exterior Env and a gp41 transmembrane Env (1)(2)(3)(4). During HIV-1 entry into host cells, gp120 binds the receptors, CD4 and CCR5/CXCR4, and gp41 mediates the fusion of the viral and target cell membranes (1,2,(5)(6)(7)(8)(9)(10). The Env trimer is metastable, and CD4 binding induces conformational changes from the pretriggered (state 1) conformation (PTC) to more "open" default intermediate (state 2) and full CD4-bound (state 3) conformations (11)(12)(13). CCR5/CXCR4 binding to the full CD4-bound state triggers additional conformational changes that lead to the formation of a very stable gp41 six-helix bundle (1,2,(14)(15)(16). The free energy difference between the metastable PTC and the six-helix bundle is used to drive the fusion of the viral and target cell membranes (1,2,16).
Many HIV-1 biological properties are determined by Env triggerability or reactivity, defined as the propensity of Env to undergo transitions from the PTC to states 2/3 (12,17,18). The metastable pretriggered Env resides in a local energy well, and the height of the activation energy barrier separating the PTC and the default intermediate state is inversely related to Env triggerability (12,17,18). Env triggerability governs the virus requirements for levels of CD4 on target cells, sensitivity to soluble CD4 and CD4-mim etic compounds (CD4mcs), and susceptibility to inactivation by prolonged exposure to cold (0°C) (12,(17)(18)(19). Primary HIV-1 strains exhibit a range of Env triggerabilities (12,17,18). Some natural polymorphisms in Env amino acid residues stabilize or destabilize the PTC (12,17,(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30). In several cases, combinations of individual PTC-stabilizing amino acid changes in Env result in additive viral phenotypes (20,24,25,28,29). This additive property has allowed the creation of HIV-1 Env variants that are stabilized in the PTC to a degree beyond that of natural virus strains (25). These extreme examples exhibit very stable Env trimers that are resistant to activation/inactivation by sCD4 and CD4mcs and are less efficient at supporting cell-cell fusion and virus entry (25).
Here, we address the following question: How many Env protomers must be modified by PTC-stabilizing or -destabilizing changes for the functional Env trimer to achieve stabilization or destabilization, respectively, of the PTC? By evaluating the phenotypes of viruses with mixed Env trimers, we find that the number of protomers that need to be modified to achieve PTC stabilization depends upon the degree of PTC stability achieved by the introduced Env change(s). Strongly PTC-stabilizing Env changes need to be present in only a single protomer to stabilize the pretriggered Env trimer, reveal ing interprotomer cooperativity in regulating PTC stability. On the other hand, weakly PTC-stabilizing changes must be present in all three protomers to achieve stabilization of the PTC. The presence of most destabilizing Env changes in a single protomer was sufficient to disrupt the functional pretriggered Env. These complementary results provide valuable insights into the symmetric nature of the functional pretriggered Env conformation and the interprotomer relationships that govern its maintenance and conversion into downstream conformations.
## RESULTS
## HIV-1 AD8 Env mutants with alterations in the stability of the PTC
HIV-1 Env triggerability/reactivity is a continuous variable, inversely related to the activation energy required to move Env from its metastable pretriggered conformation (PTC) (12,17,18). We assembled a panel of well-characterized HIV-1 AD8 Env mutants exhibiting a range of triggerabilities, including those with unusually low triggerability (high PTC stability) and those with high triggerability (low PTC stability) (Table 1) (20,22,24,25). In this set of closely matched HIV-1 AD8 mutants, virus resistance to cold (0°C) exposure and to CD4mcs has been shown to correlate closely with the degree of PTC stabilization (20,24,25). Cold resistance is a particularly useful surrogate measure of PTC stability for these stoichiometric studies as the assay avoids the use of an Env ligand whose binding could potentially be affected by the introduced Env amino acid changes (see Fig. 1A for example). The HIV-1 AD8 Env variants selected for study exhibit a wide range of sensitivities to cold inactivation and inhibition by a CD4mc, BNM-III-170 (Table 1) (31,32). The pretriggered Env stability index, which correlates directly with PTC stability and inversely with Env triggerability (20,24,25), is the product of the half-life of the viral infectivity at 0°C and the IC 50 values for BNM-III-170. In Table 1, the PTC-stabi lized and PTC-destabilized Env variants are ranked according to their pretriggered Env stability indices, which exhibit >1,500-fold variation in this panel. The diversity of this panel of HIV-1 AD8 Env variants will allow evaluation of the impact of differences in the pretriggered Env stability index on Env protomer stoichiometry.
Most of the Env mutants selected for this study were expressed and processed at least as well as the wild-type HIV-1 AD8 Env (Table 1; Fig. 1B). Some of the Envs (Tri, Tri FPPR, N332T, and Q3-N332T) were processed more efficiently than the wild-type HIV-1 AD8 Env. Only one mutant (FPPR N136E/D325Q) was processed less efficiently than the wild-type HIV-1 AD8 Env. The potential impact of differences in Env processing will be considered in the interpretation of the results (see below).
## Infectivity of recombinant HIV-1 with mixed Env trimers
Our general approach measures the infectivity of single-round recombinant HIV-1 expressing luciferase and pseudotyped with mixed Env trimers. The predicted infectivity of viruses with mixed Envs can potentially be influenced by the number of Env trimers (T) required for infection, i.e., the number of Env trimers in the functional unit of virus entry. To obtain information about the entry stoichiometry of the Envs in our study, we measured the infectivities of pseudoviruses containing mixtures of selected HIV-1 AD8 Envs that vary in PTC stability and a dominant-negative, cleavage-defective Env mutant, R508S/R511S (34)(35)(36)(37). In addition to the wild-type HIV-1 AD8 Env, we selected two PTCstabilized Envs (Tri and Q114E) and two PTC-destabilized Envs (N301E and S546D) for study. The Env content of the viruses with mixed Envs is shown in Fig. 2A. The amounts of a The HIV-1 AD8 Env variants used in this study are listed, with the Env residues numbered according to the current convention (33). b To ensure an accurate ranking of these mutants, the half-lives of pseudovirus infectivity on ice and the IC 50 values of the CD4mc BNM-III-170 were determined in side-by-side assays, as described in Materials and Methods. c The pretriggered Env stability index is the product of the virus half-life at 0°C and the BNM-III-170 IC 50 . Where a range of cold half-lives is reported, an average value was used to calculate the pretriggered stability index. The PTC-stabilized and PTC-destabilized HIV-1 AD8 Env variants are ranked according to their degree of PTC stabilization or destabilization, respectively. d Env processing was evaluated in HOS cells, as described in Materials and Methods. The level of processing, relative to that of the wild-type Env, is designated as follows: +++, wild-type level; ++, moderate decrease; ++++, more efficient processing.
Env were comparable in most of the virus preparations. As previously observed (20,25), the processing of the Tri Env was more efficient than that of the other Envs.
The infectivities of the viruses with mixed Envs for TZM-bl target cells are shown in Fig. 2B. The inclusion of increasing amounts of the dominant-negative R508S/R511S Env resulted in decreases in infection for all the viruses. However, differences in the decreases in infectivity were observed for the viruses depending on the Env paired with the R508S/R511S Env. As the fraction (f) of the R508S/R511S Env in the mixture increased, the infectivity of the viruses decreased in the following order: S546D > N301E > wild type AD8 > Q114E > Tri. The susceptibility of these Env partners to the inclusion of the and mutant HIV-1 AD8 Envs. Seventy-two hours later, the cells were lysed, and the gp120 glycoprotein in the cell medium was captured on Galanthus nivalis lectin (GNL)-beads. The clarified cell lysates and GNL precipitates were Western blotted with a polyclonal goat anti-gp120 antibody (Invitrogen). dominant-negative R508S/R511S mutant in the virus mixture is suggestive of an inverse relationship with their pretriggered Env stability indices (Fig. 2C).
By making several assumptions, the above results can provide information about the entry stoichiometry of the Env variants with altered PTC stability. We assume the following: (i) random mixing of the dominant-negative Env and partner Envs and (ii) complete inhibition of the function of mixed Env trimers containing at least one R508S/ R511S protomer. These assumptions are reasonable based on experimental observations (34)(35)(36)(37)(38)(39). Accurate determination of T requires additional difficult-to-prove assumptions about virion Env functionality, distribution, mobility, and ability to participate in productive Env clusters (35)(36)(37)(40)(41)(42)(43). These parameters are herein encompassed by the variable n t , which generally describes the number of functional Env trimers on virions that are capable of participating in the virus entry process. Some analyses, recognizing that retroviral preparations exhibit very low infectiv ity:particle ratios (35,(44)(45)(46), have assumed that n t = T on most infectious particles. In this case, inactivation of Env trimer function by the dominant-negative Env is linearly related to decreases in viral infectivity; the residual infectivity is related to the fraction (f) of the dominant-negative Env by the following equation:
Using these assumptions, the f-infectivity curves for the viruses with mixed wild-type AD8 and R508S/R511S Envs suggested T values of 2-3 (Fig. 2C andD). Viruses with the PTC-stabilized Envs (Tri and Q114E) exhibited f-infectivity curves suggestive of T values of 1-2. The f-infectivity curves of the PTC-destabilized Envs suggested higher T values (3-4 for N301E and 6 for S546D).
We also considered scenarios where n t > T, i.e., where more functional Env trimers are present on most virions than are required for virus entry. In the case where n t = T + 1, the f-infectivity curves assume sigmoidal shapes (Fig. 2D). The T values predicted by fitting the experimental curves to the theoretical f-infectivity curves when n t = T + 1 are greater than the T values predicted when n t = T (Fig. 2C).
Uncertainty about n t precludes a definitive determination of T. Nonetheless, the pattern of resistance to the dominant-negative Env suggests that Env PTC stability is associated with an increase in the n t /T ratio. This ratio is an indicator of the redundancy of functional Envs on the virion particle. PTC stabilization could potentially increase n t , lower T, or both.
The results of the dominant-negative Env mixing experiments established boundaries for the T and n t values of our Env mutants. We consider the effects of these estimates of T and n t in our analyses of the protomer stoichiometry of PTC stabilization and destabilization below.
$$(1) Residual infectivity = (1 -f) 3T$$
## Protomer stoichiometry of Envs with PTC-stabilizing changes
To estimate the number of protomers (N p ) required for PTC stabilization, we generated single-round recombinant viruses pseudotyped with mixtures of the wild-type HIV-1 AD8 Env and PTC-stabilized Envs in varying proportions. The phenotypes of the viruses with 100% PTC-stabilized Env mutants define the maximum degrees of PTC stabilization achievable by the viruses with mixed Envs. Relative to this maximum, the percentage (p) of Env trimers in the population that achieve a stabilized PTC is a function of the fraction (f) of PTC-stabilized mutant Envs in the mixture, the number of protomers (N p ) required for PTC stabilization, and the number of Env trimers (T) in the functional unit. As discussed above, an accurate determination of T depends upon currently unknown parameters. Therefore, we considered a range of T values in our evaluation of the Env protomer stoichiometry associated with PTC stabilization/destabilization. Figure 3 shows the theoretical curves governing PTC stabilization for T values of 1, 2, and 3. The predicted values of p determine the expected infectivities of viruses with mixed Env trimers, after cold exposure, according to the equation:
(2) Predicted infectivity = infectivity wild-type Env + p(infectivity mutant Env -infectivity wild-type Env )
Our theoretical analyses of PTC stabilization assume the following: (i) a random mixing of wild-type and mutant Envs according to a binomial distribution, (ii) all-ornone Env trimer function, and (iii) a direct relationship between the percentage (p) of functional Env trimers in the population and virus infectivity. The last assumption is valid when n t = T; we consider the effects of higher n t values on the estimation of N p below. Given these assumptions, Fig. 3 shows the ideal curves describing the relationship between f and p. The shapes of these ideal curves qualitatively differ for N p = 1, 2, and 3. Although increasing the value of T shifts the curves and adds slightly to their sigmoidicity, the N p = 1, N p = 2, and N p = 3 curves form distinct groups and can easily be distinguished at all T values. Thus, plotting the empirically observed viral infectivities after cold exposure as a function of f and comparison with the theoretical curves predicted by equation 2 should allow us to deduce the N p values.
Recombinant viruses with mixtures of the wild-type HIV-1 AD8 Env and PTC-stabi lized variant Envs were generated. The infectivity of these viruses was measured after incubation on ice (0°C) for a period of time that resulted in maximal differences between the infectivities of the viruses with 100% wild-type HIV-1 AD8 Env and 100% mutant Env. Plots of the residual virus infectivity after cold exposure as a function of f (the fraction of mutant Envs) revealed three patterns (Fig. 4). Convex upward f-infectivity curves were observed for three Env variants, Q114E, Tri, and Tri FPPR. The Tri and Tri FPPR Envs have multiple PTC-stabilizing changes, one of which is Q114E (20,24,25). The shapes of these f-infectivity curves are consistent with N p = 1, indicating that the presence of these PTC-stabilizing changes in a single protomer is sufficient to achieve cold resistance. For three of the Env variants (FPPR-N136E/D325Q, A582T, and FPPR), the f-infectivity curves were consistent with N p = 2, suggesting that the presence of these PTC-stabilizing changes in two protomers is sufficient to confer cold resistance. A third pattern of concave upward f-infectivity curves was observed for half (6 out of 12) of the PTC-stabilizing changes. These f-infectivity curves indicate that the PTC-stabilizing changes need to be present in all three Env protomers to achieve cold resistance (N p = 3). These results indicate that the number of Env protomers that need to be modified to achieve cold stability of the virus differs for various PTC-stabilizing Env changes.
High levels of expression and processing of mutant Envs relative to those of wild-type Env could hypothetically lower our estimate of N p . As some PTC-stabilized Env mutants like Tri and Tri FPPR exhibit relatively efficient processing and low estimated N p values (20, 25) (Fig. 1B; Table 1), we wished to evaluate whether mixing these mutants with wild-type HIV-1 AD8 Env might have biased the outcome. We examined the Envs in pseudoviruses produced by titrating increasing amounts of Tri and Tri FPPR Envs into the wild-type HIV-1 AD8 Env (Fig. 5A). The levels and processing of the mixed Envs with increasing proportions of Tri and Tri FPPR Envs were similar, particularly when these values were normalized to the levels of Gag p24 capsid protein in the virions. These observations argue against systematic alterations of virion Env levels or processing as a result of mixing the Tri and Tri FPPR Envs with the wild-type HIV-1 AD8 Env.
We also considered how differences in n t might affect the outcome of the mixing experiments. For virions that are potentially infectious, the value of n t must be greater than or equal to T. For any given T, as n t increases, there is an increased probability of forming a group of stabilized Env trimers for entry. Therefore, at a given value of In the models shown, n t = T. We assume that all T Env trimers within the functional unit of virus entry must be stabilized for that functional unit to achieve PTC stability. For each value of N p , the Env trimer combinations that are considered to result in PTC stability are boxed. The variable p represents the sum of probabilities for the Env protomer/trimer configurations that result in PTC stability. Note that the N p = 1, N p = 2, and N p = 3 curves in the f-p graphs form distinct, readily distinguishable groups. The predicted infectivity of the virus after cold exposure is determined by p and by the infectivities of the viruses with 100% wild-type Env and 100% mutant Env, according to the equation shown.
f, an increase in n t results in an increase in p and greater Env stabilization (Fig. 5B). A comparison of the bottom panel of Fig. 3 with Fig. 5B shows that increases in T and n t exert opposite effects on the f-p relationship. These opposing effects may diminish the impact of large T values on the estimation of N p . Finally, we note that the greater-thanexpected positive stabilizing effects of PTC-stabilized Envs like Tri and Tri FPPR (Fig. 4, upper panel) may be a result of high n t values (Fig. 5B).
Notably, for our panel of PTC-stabilizing Env changes, the N p values deduced from the analysis of the f-infectivity curves inversely correlated with the pretriggered Env stability indices (Spearman r S = -0.865, two-tailed P = 0.0006) (Fig. 6). Thus, the number of Env protomers that need to be modified to achieve stabilization of the PTC is inversely related to the degree of PTC stabilization that results from the Env change. Weakly and moderately PTC-stabilizing changes need to be present in multiple protomers (N p = 3 or 2, respectively) to achieve cold resistance. Strongly PTC-stabilizing Env changes efficiently stabilize the functional PTC against the effects of cold exposure when present in a single protomer.
## Protomer stoichiometry of Envs with PTC-destabilizing changes
For mixtures of the wild-type HIV-1 AD8 Env and mutant Envs with PTC-destabilizing changes, we generated theoretical curves describing the relationship between f (the fraction of mutant Envs) and the percentage (p) of maximal PTC destabilization achieved (Fig. 7). The assumptions (random Env mixing and all-or-none Env trimer function) underlying these calculations are similar to those described above for the mixtures of wild-type and PTC-stabilized Envs. In addition, for T > 1 and n t = T, we assumed that destabilization of any one of the Env trimers in the functional unit of virus entry results in PTC destabilization of that functional unit. As seen for the mixtures with the PTC-stabilized Envs, distinct shapes of the f-p curves were associated with the different N p values. At higher T values, which may apply to the PTC-destabilized Env mutants (Fig. 2C), the f-p curves shift but do so uniformly, retaining the separation into distinct N p FIG 6 Correlation between the number of protomers (N p ) required for PTC stabilization and the pretriggered Env stability index. The inverse relationship between the N p values deduced from the shapes of the f-infectivity curves in Fig. 4 and the pretriggered Env stability indices (Table 1) is shown. The Spearman rank correlation coefficient (r S ) and two-tailed P value are shown. Note that the N p = 1, N p = 2, and N p = 3 curves in the f-p graphs form distinct, readily distinguishable groups. The predicted infectivity of the virus after cold exposure is determined by p and by the infectivities of the viruses with 100% wild-type Env and 100% mutant Env, according to the equation shown. groups (Fig. 7). For a given f value, as T increases, the efficiency of Env PTC destabilization is expected to increase; in contrast, increases in n t should decrease the efficiency of Env PTC destabilization. Thus, increases in T and n t may partially cancel each other's effect on our estimation of N p . However, at high actual T values, we could potentially underesti mate N p , a point that is discussed further below.
Recombinant viruses with mixtures of wild-type HIV-1 AD8 Env and PTC-destabilized Env mutants were generated and tested for cold sensitivity. In Fig. 8, we show the observed f-infectivity curves, allowing comparison with the theoretical curves (in the lower right panel). For most of the Env mutants, the presence of the destabilizing changes in a single protomer was apparently sufficient to destabilize the PTC and render the viruses sensitive to incubation at 0°C. A protomer stoichiometry of N p = 1 was deduced from the cold sensitivity plots of the Q3alt, N301E, F317W, and N332T Env mutants. Of interest, three of these PTC-destabilizing changes (Q3alt, N301E, and N332T) involve the loss or shift of an N-linked glycosylation site in gp120 (24). The F317W change alters the tip of the gp120 V3 loop (30).
The f-infectivity curve associated with another PTC-destabilizing change, S546D, suggested that its presence in at least two Env protomers is necessary to destabilize the Env trimer and render the virus sensitive to cold (Fig. 8). Serine 546 is located in the gp41 heptad repeat (HR1 N ) region, and the S546D change does not alter a potential N-linked glycosylation site. Thus, the S546D change differs from the other destabilizing Env mutants tested in the subunit location in the Env trimer, protomeric stoichiometry, and, in some cases, effect on glycosylation.
To evaluate potential relationships between the degree of PTC destabilization and the protomeric stoichiometry (N p ) of the S546D and the other tested Env variants, we compared the sensitivity of viruses with these Envs to cold, CD4mcs, sCD4-Ig, and antibodies. The results are shown in Tables 1 and2, where the PTC-destabilized Env mutants are arranged in order of decreasing pretriggered Env stability indices. The sensitivity of viruses with these Envs to sCD4-Ig exhibits the same rank order (Table 2); PTC destabilization is expected to increase Env triggerability and inactivation by sCD4-Ig (12,17,19,(23)(24)(25). The F317W virus exhibited a twofold increase in sensitivity to sCD4-Ig, but was neutralized by the antibodies tested comparably to the wild-type HIV-1 AD8 virus. Viruses with the other PTC-destabilized Envs (N332T, N301E, and S546D) were more sensitive than the wild-type virus to CD4BS bNAbs, the 10E8.v4 MPER bNAb, and the 447-52D anti-V3 pNAb. The S546D virus was not only more sensitive to these antibodies than the N332T and N301E viruses but also was neutralized by the 19b anti-V3 and 17b CD4i pNAbs. Thus, compared with the other PTC-destabilizing Env changes, the S546D change results in the highest degree of PTC destabilization, based on the viral pretriggered Env stability indices and sensitivities to antibody/sCD4-Ig neutralization. In this set of PTC-destabilized Env variants, greater PTC destabilization is apparently associated with a higher N p requirement.
The deduced protomer stoichiometries (N p values) and the relationship of the observed phenotypes to the degree of PTC stabilization/destabilization for the panel of studied Env variants is summarized in Fig. 9.
## Combination of PTC-stabilizing and -destabilizing changes in Env trimers
We wished to examine the effect of PTC-destabilizing Env changes in the context of an Env trimer that has PTC-stabilizing changes in some or all of its protomers. Several variations were tried. i. Q3 + Q3 N332T: in the background of the wild-type HIV-1 AD8 Env, the N332T change in one protomer (N p = 1) is sufficient for destabilization of the PTC and increased cold sensitivity (Fig. 8). The magnitude of the cold sensitivity phenotype resulting from the N332T change was reduced when the N332T change was introduced into an Env where all three protomers had the PTC-stabilizing Q3 change (Fig. 10A). Nonetheless, the shape of the f-infectivity curve indicated that the protomer stoichiometry remained at N p = 1.
ii. Tri FPPR + AD8-N301E: in the wild-type HIV-1 AD8 background, the N301E change destabilizes the PTC when present in a single protomer (Fig. 8). Likewise, when the AD8-N301E Env was mixed with the Tri FPPR Env, one protomer of AD8-N301E was sufficient to destabilize the PTC of the Env trimer (Fig. 10A). iii. Tri FPPR + AD8-Q3alt: in the wild-type HIV-1 AD8 background, the Q3alt change in a single protomer is sufficient to destabilize the PTC (Fig. 8). This was also the case when the AD8-Q3alt Env was mixed with the Tri FPPR Env (Fig. 10A). iv. Tri FPPR + AD8-S546D: in the wild-type HIV-1 AD8 background, two S546D protomers are required to destabilize the PTC of Env (Fig. 8). However, when the AD8-S546D Env was mixed with the Tri FPPR Env, one AD8-S546D protomer was apparently able to bring about PTC destabilization (Fig. 10A).
In all four examples above, the presence of a single protomer with a PTC-destabiliz ing change apparently resulted in PTC destabilization of Env trimers with at least two protomers containing PTC-stabilizing changes. This is not surprising in the case where the PTC-stabilizing change is Q3, which even when present in all three protomers only weakly stabilizes the Env PTC. However, it is unexpected in the three instances with the strongly PTC-stabilizing Tri FPPR changes that, when mixed with wild-type HIV-1 AD8 protomers, stabilize the PTC when present in a single protomer (N p = 1) (Fig. 10B). Instead, when Tri FPPR protomers are mixed with Env protomers with PTC-destabilizing changes, Env trimers with one or two Tri FPPR protomers are cold-sensitive. Apparently, the N p = 1 stoichiometry of the PTC-destabilizing changes is dominant over the N p = 1 stoichiometry of the PTC-stabilizing Tri FPPR change. The above examples also demon strate that the Env background in which PTC-destabilizing changes are introduced may preserve or alter the protomer stoichiometry (N p ).
## Sensitivity of Env chimeras to CD4mcs
The measurement of viral cold sensitivity affords a ligand-free means of evaluating the HIV-1 Env conformational state, thus providing insight into the spontaneous triggerabil ity of Env. Virus sensitivity to CD4mc inhibition provides a second surrogate for the stability of the PTC, allowing an assessment of induced triggerability. Measuring the protomer stoichiometry of Env changes in the latter case is more technically challenging, as multiple CD4mc concentrations must be empirically evaluated to achieve detectable but different levels of infection for the wild-type and mutant viruses. We determined the protomer stoichiometry of three PTC-stabilizing Env changes (Q114E, Tri FPPR, and FPPR-N136E/D325Q), using virus inhibition by the CD4mc BNM-III-170 as a phenotypic readout. None of these PTC-stabilizing changes overlaps with the known gp120 binding site of the CD4mcs (31,32,47). We mixed HIV-1 AD8 Envs containing each of these changes with the wild-type HIV-1 AD8 Env and tested the susceptibility of the viruses to inhibition by different concentrations of BNM-III-170. As expected (20,25), compared with viruses with the wild-type HIV-1 AD8 Env, the viruses with 100% PTC-stabilized Envs were relatively resistant to BNM-III-170 inhibition. For the strongly PTC-stabilized Q114E and Tri FPPR Envs, the f-infectivity curves exhibited mild shape changes as the BNM-III-170 concentration varied; these may reflect the effects of CD4mc occupancy of the gp120 subunits on the stoichiometric requirements for the PTC-stabilizing changes (Fig. 11A andB). However, the f-infectivity curves for both Envs deviated only modestly from the N p = 2 theoretical curve at all CD4mc concentrations. An N p = 2 stoichiome try was also supported by the f-infectivity curve observed for the PTC-stabilizing Tri FPPR changes when a more potent CD4mc, CJF-III-288 (47), was used (Fig. 11C). For the moderately PTC-stabilizing FPPR-N136E/D325Q Env, the f-infectivity curves were consistent with N p = 3 at multiple BNM-III-170 concentrations (Fig. 11D). Thus, for CD4mc resistance, we assign N p values of 2, 2, and 3 for the Q114E, Tri FPPR, and FPPR-N136E/ D325Q Envs, respectively. The respective N p values for the cold resistance of these mutants were 1, 1, and 2. Apparently, more protomers require PTC-stabilizing changes to achieve resistance to BNM-III-170 than to increase resistance to cold. It may be easier for PTC-stabilizing changes to resist spontaneous Env transitions from the PTC than the transitions induced by CD4mcs. We also wished to evaluate how the presence of a PTC-destabilizing change in one or more Env protomers would affect virus sensitivity to BNM-III-170. For this purpose, we mixed HIV-1 AD8 Envs containing the destabilizing N301E change with the wild-type HIV-1 AD8 Env and tested the susceptibility of the viruses to inhibition by BNM-III-170. As expected (24), viruses with 100% AD8-N301E Envs were inhibited to a greater extent by BNM-III-170, consistent with their increased triggerability. At low concentrations of BNM-III-170, the viruses with two or more AD8-N301E protomers were inhibited (Fig. 11E). At higher BNM-III-170 concentrations, Envs with one protomer of the AD8-N301E Env were efficiently inhibited. Apparently, the CD4mc-induced inactivation of Env is more efficient when two protomers contain the PTC-destabilizing N301E change. Fewer Env protomers with this PTC-destabilizing change are required for virus inactivation when CD4mc concentrations are higher.
## DISCUSSION
In this work, we investigated the number of Env protomers (N p ) that must be modified by particular Env changes that stabilize or destabilize the pretriggered conformation (PTC) to achieve the viral phenotypes of increased resistance or sensitivity, respectively, to cold (0°C) or CD4mc exposure. Cold exposure and incubation with CD4mcs drive HIV-1 Env from the metastable PTC by different means: cold inactivation is ligand-free and spontaneous, whereas CD4mcs bind gp120 and induce changes in Env similar to those triggered by CD4 binding (11,12,19,23,26,31,32,(47)(48)(49)(50). Increases or decreases in the stability of the Env PTC result in increased HIV-1 resistance or sensitivity, respectively, to both cold and CD4mcs (12,(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)48). The product of the virus half-life on ice and the CD4mc IC 50 is the pretriggered Env stability index, which correlates with the stability of the PTC better than either factor on its own (25). We used the pretriggered Env stability indices to rank the HIV-1 AD8 Env mutants in this study with respect to the degree of PTC stabilization or destabilization achieved.
As the Env trimer stoichiometry (T) associated with virus entry can hypothetically influence our estimation of N p , we studied the infectivity of viruses with mixtures of the wild-type HIV-1 AD8 Env and a dominant-negative, cleavage-defective Env (R508S/R511S) (34)(35)(36)(37). The R508S/R511S mutant inhibited the infectivity of viruses with PTC-destabi lized Envs more efficiently than the infectivity of viruses with PTC-stabilized Envs. PTC stabilization could lower T (the number of Env trimers required for virus entry) or increase n t (the number of functional Env trimers on virions that can participate in virus entry) or both. Lower T values have been suggested for the Envs of primary HIV-1 than for the Envs of laboratory-adapted HIV-1, which typically exhibit lower PTC stability (18,37). Determination of actual T values for HIV-1 is complicated by uncertainty about virion Env numbers, distribution, mobility, and ability to contribute to productive entry events, all of which are encompassed in the variable n t . For the purpose of this work, we identified a range of T and n t values that we evaluated for the potential impact on our estimation of N p . The relatively low T and n t values associated with the PTC-stabilized Envs were not predicted to confound our ability to deduce N p values from the f-infectivity curves for these Env mutants. Indeed, in the experiments involving the PTC-stabilized Envs, we observed all three shapes of the f-infectivity curves predicted for the three N p values. Moreover, we observed a strong inverse correlation between the deduced N p values and the pretriggered Env stability indices of the PTC-stabilized Envs. These results argue against severe violations of the assumptions underlying our theoretical analysis of the PTC-stabilized Envs.
The ligand-free cold inactivation assay allowed us to evaluate the cold sensitivity of viruses pseudotyped with mixtures of a great variety of Envs and to estimate the protomer stoichiometry, N p . With respect to cold resistance of the viruses, the number of Env protomers that must be modified to achieve stabilization of the PTC was found to correlate inversely with the degree of PTC stabilization that results from the introduced Env changes (Fig. 6). For strongly PTC-stabilizing Env changes, modification of a single protomer appears to be sufficient to achieve PTC stabilization (N p = 1). This observation suggests that given adequate stability, the PTC-stabilized protomer can influence the conformation of the other two protomers to maintain the PTC. The efficacy of such cross-protomer cooperativity is dependent on the Env partner that is mixed with the strongly PTC-stabilized Env. Thus, for the strongly PTC-stabilizing Tri FPPR changes, N p = 1 when mixed with the wild-type HIV-1 AD8 Env, but N p = 3 when the Env partner contains PTC-destabilizing changes (N301E, Q3alt, or S546D) (Fig. 10B). In the latter cases, the presence of even a single protomer with the PTC-destabilizing changes completely nullified the ability of the Tri FPPR changes to render the mixed Env trimers cold-resist ant. Apparently, disruption of the PTC of a single Env protomer is not readily compensa ted even by strongly PTC-stabilizing changes in the other protomers. In this situation, the protomer stoichiometries of PTC-destabilizing Env changes are dominant.
For weakly PTC-stabilizing Env changes, all three protomers required modification to achieve cold resistance of the virus (N p = 3). Thus, symmetrical placement of weakly PTC-stabilizing changes in three Env protomers is apparently conducive to the mainte nance of the PTC (Fig. 9). A complementary observation is that the presence of weakly PTC-destabilizing Env changes in one protomer was sufficient to increase the cold sensitivity of the virus (N p = 1). The PTC of the wild-type HIV-1 AD8 Env can be disrupted by the asymmetric introduction of a PTC-destabilizing Env change into a single protomer. This PTC-destabilizing N p = 1 stoichiometry dominated, even when strongly PTC-stabiliz ing changes were present in the other protomers. In current Env structures (51)(52)(53)(54)(55)(56), several of the strongly PTC-stabilizing changes involve amino acid residues that are near subunit interfaces. However, PTC-destabilizing changes often involve changes in N-linked glycosylation sites and are located on the periphery of the Env trimer (see below). Thus, direct interactions between Env structures involved in PTC stabilization and destabiliza tion are unlikely explanations for our observations. A more attractive model explaining these observations is that the PTC represents a C3-symmetric trimer. In this model, Env changes that stabilize the PTC need to maintain conformational symmetry among the protomers. Weakly PTC-stabilizing Env changes, lacking the ability of the strongly PTC-stabilizing changes to act across protomers, must be present in all three protomers (N p = 3). Conversely, PTC-destabilizing Env changes in one protomer can lead to a significant loss of conformational symmetry among the protomers, thereby destabilizing the PTC, with consequent increases in virus sensitivity to cold inactivation and CD4mc inhibition. Supporting this model are the many examples of weakly PTC-stabilizing Env changes where N p = 3 and weakly PTC-destabilizing Env changes where N p = 1 (Fig. 9).
Several of the weakly PTC-stabilizing changes (N136E, T138A, N136E/D325Q, and Q3) involve the loss of a gp120 V1 glycan at Asn 136, which presumably exerts its effects at the Env surface (24,25). Other weakly PTC-stabilizing changes (Q567K and D325Q) do not alter a potential Env glycosylation site. Apparently, these changes most effectively stabilize the PTC when they are present in all three Env protomers, consistent with the stabilization of a symmetric Env trimer.
For the group of strongly PTC-stabilizing Env mutants (Q114E, Tri, and Tri FPPR), all of which share the Q114E change, alteration of a single Env protomer is apparently sufficient to achieve cold resistance. In available Env trimer models (51-56), Gln 114 is located in the α1 helix of the gp120 inner domain, near the interface with gp41. The mechanism whereby a change in Gln 114 to glutamic acid in one protomer would stabilize the PTC of the entire Env trimer is not apparent in currently available struc tures. Detailed structures of the membrane Env PTC may shed light on this gap in our understanding.
For most of the PTC-destabilizing Env changes, alteration of a single Env protomer (N p = 1) was apparently sufficient to sensitize the virus to cold inactivation. Loss of a glycan in one protomer (N p = 1) destabilized the PTC in the case of the N301E and N332T changes, which remove glycans that occupy the interprotomer angles of the Env trimer (24). An N p value of 1 was also deduced for the F317W change in the gp120 V3 tip, which does not alter an N-linked glycosylation site. The protomer stoichiometry of these Env variants supports the hypothesis that C3 symmetry contributes to the maintenance of the pretriggered (state 1) Env trimer conformation (Fig. 9). Recent structural studies suggest that the default intermediate (state 2) Env is an asymmetric trimer, which presumably derives from a symmetric pretriggered (state 1) Env (51, 52, 57-59).
One PTC-destabilizing change, S546D, needed to be present in two or more protomers (N p = 2) to render the virus cold-sensitive. The sensitivity of viruses with the S546D Env to inhibition by the dominant-negative, cleavage-defective Env raises the possibility that, due to higher T, we underestimated N p and the actual N p = 3. In either case, the S546D N p estimates are greater than those for other PTC-destabilizing Env changes. An obvious structural explanation is lacking for the higher N p requirements of this mutant. Serine 546 is located in the highly dynamic gp41 HR1 N region, near the trimeric coiled coil formed by HR1 C (51)(52)(53)(54)(55)(56)58). The higher N p value of S546D, relative to those of other PTC-destabilizing changes, may reflect the magnitude of PTC destabilization achieved; the global sensitivity of the S546D mutant to antibody neutralization suggests that significant redistribution into state 3-like Envs results from this change.
To achieve HIV-1 AD8 resistance to CD4mcs, the PTC-stabilizing changes were required in more Env protomers than for cold resistance. Apparently, it is more difficult to resist the direct induction by CD4mcs of entry-related transitions from the PTC than to withstand the more generally disruptive effects of ice formation on Env trimer integrity associated with cold exposure (49,50). With respect to CD4mc sensitivity, a PTC-destabi lizing change, N301E, exhibited different protomer stoichiometries at low (N p = 2) and high (N p = 1) concentrations of the CD4mc. The increased occupancy of Env at higher CD4mc concentrations itself promotes Env transitions from the PTC and likely accounts for the lower protomer requirements for the PTC-destabilizing change.
Our studies of Env protomer stoichiometry suggest the importance of conformational correspondence among the protomers of functional Env trimers. Future efforts will evaluate the hypothesized contribution of Env structural symmetry to PTC integrity and should facilitate further stabilization and characterization of the pretriggered (state 1) Env conformation.
## MATERIALS AND METHODS
## HIV-1 env mutants
The wild-type HIV-1 AD8 env cloned in the pSVIIIenv expression plasmid was used as a template to construct HIV-1 Env mutants in this study (20). The signal peptide/N-termi nus (residues 1-33) and the cytoplasmic tail C-terminus (residues 751-856) of this Env are derived from the HIV-1 HXBc2 Env. The dominant-negative, cleavage-defective Env mutant has two changes, R508S/R511S, that eliminate proteolytic processing of the Env precursor (34)(35)(36). "Tri" indicates the Q114E/Q567K/A582T changes, and "FPPR" indicates the A532V/I535M/L543Q changes (20,24,25). The Q3 and Q3 alt Env mutants were previously reported as Q3 (V1) and Q3 (V1alt), respectively, in reference 24. Env mutants with specific changes were generated by using the QuikChange Lightning site-directed mutagenesis kit (Agilent Technologies). All the Envs contain a His 6 tag at the carboxyl terminus. The presence of the desired mutations was confirmed by DNA sequencing.
## Cell lines
HEK293T, TZM-bl, and HOS cells (ATCC) were cultured in Dulbecco modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 IU penicillin and 100 μg/mL streptomycin (Life Technologies).
## Env expression
To evaluate Env processing and subunit association, 3 × 10 5 HOS cells were seeded in 6-well plates. After 24 hours of incubation, they were transfected with plasmids encoding His 6 -tagged Env variants and Tat at a ratio of 8:1, using the Lipofecta mine 3000 transfection reagent (Life Technologies) according to the manufacturer's instructions. Seventy-two hours after transfection, the cells were lysed in PBS buffer containing 1.0% NP-40 and protease inhibitor (Sigma-Aldrich). Clarified lysates were harvested, boiled, and analyzed by Western blotting with 1:2,500 goat anti-gp120 antibody (Invitrogen) and 1:2,500 HRP-conjugated rabbit anti-goat antibody (Invitrogen). The intensities of the gp120 and gp160 bands from unsaturated Western blots were quantified by using ImageJ software. The Env processing index was calculated by dividing gp120 by gp160 in the cell lysate samples. The processing indices of Env mutants in this study were normalized to those of the wild-type HIV-1 AD8 Env.
Seventy-two hours after transfection, the supernatants of the transfected HOS cells were collected and incubated with Galanthus Nivalis Lectin (GNL)-agarose beads (Vector Laboratories) for 1.5 h at room temperature. The beads were washed three times with PBS containing 0.1% NP-40 and processed for Western blotting with 1:2,500 goat anti-gp120 antibody (Invitrogen), as described above. The subunit association index was calculated by dividing gp120 in the cell lysate samples by the gp120 in the GNL precipitates. The subunit association indices of the Env mutants were normalized to those of the wild-type HIV-1 AD8 Env.
## Virus infectivity
To produce pseudoviruses, HEK293T cells were cotransfected with Env-expressing pSVIIIenv plasmids, a Tat-encoding plasmid and the luciferase-encoding pNL4-3.Luc.R-E-vector (NIH HIV Reagent Program) at a 1:1:3 ratio using polyethyleneimine (PEI, Polysciences). To investigate Env trimer stoichiometry (T), the pSVIIIenv plasmids encoded mixtures of selected PTC-stabilized and PTC-destabilized Envs with varying proportions of the dominant-negative, cleavage-defective R508S/R511S Env mutant (34)(35)(36). To investigate Env protomer stoichiometry (N p ), the pSVIIIenv plasmids encoded mixtures of the wild-type HIV-1 AD8 Env and either PTC-stabilized or PTC-destabilized Env mutants in varying proportions. After 8 hours, the medium was replaced with fresh medium. Seventy-two hours later, the pseudoviruses in the supernatants were harvested and centrifuged (3,500 rpm for 5 min), aliquoted, and either used directly to measure pseudovirus infectivity or stored at -80°C until further use. To evaluate the infectivity of the variants, pseudoviruses were diluted in 96-well plates and incubated with TZM-bl cells in the presence of 20 µg/mL DEAE-dextran. After a 48 hour incubation at 37°C in 5% CO 2 , the TZM-bl cells were lysed, and the luciferase activity was measured using a luminometer.
## Virus sensitivity to cold inactivation
To evaluate the sensitivity of viruses with mixed Envs to cold inactivation, pseudoviruses were incubated on ice (0°C) for a period of time. This period of time was chosen to maximize the infectivity difference between viruses with wild-type HIV-1 AD8 Env and PTC-stabilized/-destabilized mutant Envs. After cold incubation, the infectivity of the pseudoviruses was measured as described above. The level of infectivity of each virus following incubation on ice was normalized to that of the same virus preparation that had not been incubated on ice.
## Virus inhibition by a CD4mc
The CD4mcs BNM-III-170 and CJF-III-288 (31,32,47) were serially diluted in triplicate wells in 96-well plates. Then approximately 100 to 200 TCID 50 (50% tissue culture infectious dose) of pseudoviruses was added and incubated at 37°C for 1 h. Subse quently, approximately 2 × 10 4 TZM-bl cells with 20 µg/mL DEAE-dextran in the medium were added to each well, and the mixture was incubated at 37°C/5% CO 2 for 48 hours. Then, luciferase activity was measured, as described above. The level of infectivity of each virus with mixed Envs was normalized to that of the same virus preparation that had not been incubated with CD4mcs.
## Analysis of Env on virus particles
Approximately 1 mL of the clarified cell supernatant containing pseudovirus was centrifuged at 14,000 × g for 1 h at 4°C. The pelleted virus particles were resuspen ded in 1 × PBS. Equal volumes of the virus suspensions were then lysed in PBS buf fer/1.0% NP-40/protease inhibitor cocktail and analyzed by Western blotting. Western blots were developed with 1:2,500 goat anti-gp120 polyclonal antibody (Invitrogen), 1:2,500 4E10 anti-gp41 antibody, and 1:5,000 rabbit polyclonal antibody against Gag p55/p24/p17 (Abcam). The HRP-conjugated secondary antibodies were 1:2,500 rabbit anti-goat antibody (Invitrogen), 1:2,500 goat anti-human antibody (Invitrogen), and 1:5,000 goat anti-rabbit antibody (Sigma-Aldrich), respectively.
## References
1. Wyatt, Sodroski (1998) "The HIV-1 envelope glycoproteins: fusogens, antigens, and immunogens" *Science*
2. Chen (2019) "Molecular mechanism of HIV-1 entry" *Trends Microbiol*
3. Allan, Coligan, Barin et al. (1985) "Major glycoprotein antigens that induce antibodies in AIDS patients are encoded by HTLV-III" *Science*
4. Robey, Safai, Oroszlan et al. (1985) "Characterization of envelope and core structural gene products of HTLV-III with sera from AIDS patients" *Science*
5. Klatzmann, Champagne, Chamaret et al. (1984) "T-lymphocyte T4 molecule behaves as the receptor for human retrovirus LAV" *Nature*
6. Dalgleish, Beverley, Clapham et al. (1984) "The CD4 (T4) antigen is an essential component of the receptor for the AIDS retrovirus" *Nature*
7. Wu, Gerard, Wyatt et al. (1996) "CD4induced interaction of primary HIV-1 gp120 glycoproteins with the chemokine receptor CCR-5" *Nature*
8. (2025)
9. Trkola, Dragic, Arthos et al. (1996) "CD4-dependent, antibody-sensitive interactions between HIV-1 and its co-receptor CCR-5" *Nature*
10. Feng, Broder, Kennedy et al. (1996) "HIV-1 entry cofactor: functional cDNA cloning of a seven-transmembrane, G protein-coupled receptor" *Science*
11. Alkhatib, Combadiere, Broder et al. (1996) "CC CKR5: a RANTES, MIP-1alpha, MIP-1beta receptor as a fusion cofactor for macrophage-tropic HIV-1" *Science*
12. Munro, Gorman, Ma et al. (2014) "Conforma tional dynamics of single HIV-1 envelope trimers on the surface of native virions" *Science*
13. Herschhorn, Ma, Gu et al. (2016) "Release of gp120 restraints leads to an entrycompetent intermediate state of the HIV-1 envelope glycoproteins" *mBio*
14. Ma, Lu, Gorman et al. (2018) "HIV-1 Env trimer opens through an asymmetric intermediate in which individual protomers adopt distinct conformations"
15. Chan, Fass, Berger et al. (1997) "Core structure of gp41 from the HIV envelope glycoprotein"
16. Weissenhorn, Dessen, Harrison et al. (1997) "Atomic structure of the ectodomain from HIV-1 gp41" *Nature*
17. Melikyan, Markosyan, Hemmati et al. (2000) "Evidence that the transition of HIV-1 gp41 into a sixhelix bundle, not the bundle configuration, induces membrane fusion" *J Cell Biol*
18. Herschhorn, Gu, Moraca et al. (2017) "The β20-β21 of gp120 is a regulatory switch for HIV-1 Env conformational transitions" *Nat Commun*
19. Haim, Strack, Kassa et al. (2011) "Contribution of intrinsic reactivity of the HIV-1 envelope glycoproteins to CD4-independent infection and global inhibitor sensitivity" *PLoS Pathog*
20. Madani, Princiotto, Zhao et al. (2017) "Activation and inactivation of primary human immunodeficiency virus envelope glycoprotein trimers by CD4-mimetic compounds" *J Virol*
21. Nguyen, Qualizza, Anang et al. (2022) "Functional and highly cross-linkable HIV-1 envelope glycoproteins enriched in a pretriggered conformation" *J Virol*
22. Pacheco, Alsahafi, Debbeche et al. (2017) "Residues in the gp41 ectodomain regulate HIV-1 envelope glycoprotein conformational transitions induced by gp120-directed inhibitors" *J Virol*
23. Nguyen, Wang, Anang et al. (2023) "Characterization of the human immunodeficiency virus (HIV-1) envelope glycoprotein conformational states on infectious virus particles" *J Virol*
24. Kassa, Madani, Schön et al. (2009) "Transitions to and from the CD4-bound conformation are modulated by a single-residue change in the human immunodeficiency virus type 1 gp120 inner domain" *J Virol*
25. Zhang, Wang, Nguyen et al. (2023) "Alterations in gp120 glycans or the gp41 fusion peptide-proximal region modulate the stability of the human immunodeficiency virus (HIV-1) envelope glycoprotein pretriggered conformation" *J Virol*
26. Zhang, Anang, Nguyen et al. (2024) "Membrane HIV-1 envelope glycoproteins stabilized more strongly in a pretriggered conformation than natural virus Envs"
27. Kassa, Finzi, Pancera et al. (2009) "Identification of a human immunodeficiency virus type 1 envelope glycoprotein variant resistant to cold inactivation" *J Virol*
28. Kesavardhana, Varadarajan (2014) "Stabilizing the native trimer of HIV-1 Env by destabilizing the heterodimeric interface of the gp41 postfusion six-helix bundle" *J Virol*
29. Leaman, Zwick (2013) "Increased functional stability and homogeneity of viral envelope spikes through directed evolution" *PLoS Pathog*
30. Gift, Leaman, Zhang et al. (2017) "Functional stability of HIV-1 envelope trimer affects accessibility to broadly neutralizing antibodies at its apex" *J Virol*
32. Anang, Zhang, Fritschi et al. (2023) "V3 tip determinants of susceptibility to inhibition by CD4-mimetic compounds in natural clade A human immunodeficiency virus (HIV-1) envelope glycoproteins" *J Virol*
33. Melillo, Liang, Park et al. (2016) "Small-molecule CD4-mimics: structure-based optimization of HIV-1 entry inhibition" *ACS Med Chem Lett*
34. Courter, Madani, Sodroski et al. (2014) "Structurebased design, synthesis and validation of CD4-mimetic small molecule inhibitors of HIV-1 entry: conversion of a viral entry agonist to an antagonist" *Acc Chem Res*
35. Korber, Foley, Kuiken et al. (1998) "Numbering positions in HIV relative to HXB2CG"
36. Iwatani, Kawano, Ueno et al. (2001) "Analysis of dominant-negative effects of mutant env proteins of human immunodeficiency virus type 1" *Virology (Auckland)*
37. Yang, Kurteva, Ren et al. (2005) "Stoichiometry of envelope glycoprotein trimers in the entry of human immunodeficiency virus type 1" *J Virol*
38. Herrera, Klasse, Kibler et al. (2006) "Dominant-negative effect of hetero-oligomerization on the function of the human immunodeficiency virus type 1 envelope glycoprotein complex" *Virology (Auckland)*
39. Brandenberg, Magnus, Rusert et al. (2015) "Different infectivity of HIV-1 strains is linked to number of envelope trimers required for entry" *PLoS Pathog*
40. Salzwedel, Berger (2000) "Cooperative subunit interactions within the oligomeric envelope glycoprotein of HIV-1: functional complemen tation of specific defects in gp120 and gp41" *Proc Natl Acad Sci*
41. Rusert, Krarup, Magnus et al. (2011) "Interaction of the gp120 V1V2 loop with a neighboring gp120 unit shields the HIV envelope trimer against cross-neutralizing antibodies" *J Exp Med*
42. Klasse (2007) "Modeling how many envelope glycoprotein trimers per virion participate in human immunodeficiency virus infectivity and its neutralization by antibody" *Virology (Auckland)*
43. Magnus, Rusert, Bonhoeffer et al. (2009) "Estimating the stoichiometry of human immunodeficiency virus entry" *J Virol*
44. Zarr, Siliciano (2015) "Stoichiometric parameters of HIV-1 entry" *Virology (Auckland)*
45. Katte, Xu, Han et al. (2025) "Inter-protomer opening cooperativity of envelope trimers positively correlates with HIV-1 entry stoichiometry" *mBio*
46. Bourinbaiar (1994) "The ratio of defective HIV-1 particles to replica tion-competent infectious virions" *Acta Virol*
47. Layne, Merges, Dembo et al. (1992) "Factors underlying spontaneous inactivation and susceptibility to neutralization of human immunodeficiency virus" *Virology (Auckland)*
48. Parrish, Gao, Li et al. (2013) "Indoline CD4-mimetic compounds mediate potent and broad HIV-1 inhibition and sensitization to antibody-dependent cellular cytotoxicity" *Proc Natl Acad Sci*
49. Fritschi, Anang, Gong et al. (2023) "Indoline CD4-mimetic compounds mediate potent and broad HIV-1 inhibition and sensitization to antibody-dependent cellular cytotoxicity" *Proc Natl Acad Sci*
50. Mcgee, Haim, Korioth-Schmitz et al. (2014) "The selection of low envelope glycoprotein reactivity to soluble CD4 and cold during simian-human immunodeficiency virus infection of rhesus macaques" *J Virol*
51. Privalov (1990) "Cold denaturation of proteins" *Crit Rev Biochem Mol Biol*
52. Gulevsky, Relina (2013) "Molecular and genetic aspects of protein cold denaturation" *Cryo Lett*
53. Zhang, Wang, Wang et al. (2021) "Asymmetric structures and conformational plasticity of the uncleaved full-length human immunodeficiency virus envelope glycoprotein trimer" *J Virol*
54. Wang, Zhang, Go et al. (2023) "Asymmetric conformations of cleaved HIV-1 envelope glycoprotein trimers in styrene-maleic acid lipid nanoparticles" *Commun Biol*
55. Julien, Cupo, Sok et al. (2013) "Crystal structure of a soluble cleaved HIV-1 envelope trimer" *Science*
56. Lyumkis, Julien, De Val et al. (2013) "Cryo-EM structure of a fully glycosylated soluble cleaved HIV-1 envelope trimer" *Science*
57. Pancera, Zhou, Druz et al. (2014) "Structure and immune recognition of trimeric pre-fusion HIV-1 Env" *Nature*
58. Lee, Ozorowski, Ward (2016) "Cryo-EM structure of a native, fully glycosylated, cleaved HIV-1 envelope trimer" *Science*
59. Zhou, Zhang, Nguyen et al. (2023) "Conformations of human immunodeficiency virus envelope glycoproteins in detergents and styrene-maleic acid lipid particles" *J Virol*
60. Qi, Zhang, Wang et al. (2025) "The membrane-proximal external region of human immunodeficiency virus (HIV-1) envelope glycoprotein trimers in A18-lipid nanodiscs" *Commun Biol*
61. Zhang, Anang, Zhang et al. (2024) "Conformations of membrane human immunodeficiency virus (HIV-1) envelope glycoproteins solubilized in amphipol A18 lipidnanodiscs" *J Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12125796&blobtype=pdf | # Correction: Nucleos(t)ide analogs continuation is not associated with a lower risk of HBsAg seroreversion following PEG-IFN-induced HBsAg loss
Qiyi Zhao, Na Gao, Haishi Wu, Bin Li, Huiying Yu, Lili Wu, Jing Zhang, Nan Zhang, Bingliang Lin, Zhiliang Gao
The corrected sentence is "Additionally, we observed that 26 of 43 patients with HBsAg antigenemia did not have viremia during follow-up, suggesting that HBsAg antigenemia might mainly originate from transcriptionally active HBV integration instead of cccDNA. "
The original article has been corrected. |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12548406&blobtype=pdf | # Multimodal characterization of the responsiveness of eight hepatitis D virus genotype isolates to interferon-alpha treatment
Yibo Ding, Qiudi Li, Yunlu Sha, Ruilin Si, Chengqian Feng, Mei Liu, Zhanghao Feng, Xutong Ding, Ying Li, Huiyuan Fu, Shiquan Liang, Qili Yao, Zhenfeng Zhang, Feng Li, Stephan Urban, Hongbo Guo, Wenshi Wang
## Abstract
Chronic hepatitis delta virus (HDV) infection causes the most severe form of viral hepatitis. Although humans produce 12 subtypes of interferon-alpha (IFN-α), IFN-α2a has been the only commonly used treatment against HDV. Previously, we characterized eight HDV genotype isolates with varying replication kinetics. Herein, we systematically investigated the antiviral efficacy of IFN-α2a and other IFN-α subtypes against HDV genotypes 1-8 during de novo infection, cell mitosis, and in quiescent cells. Our findings revealed that IFN-α2a exhibits potent but varied efficiency against HDV 1-8 isolates upon de novo infection and cell mitosis. Conversely, HDVs in resting cells are resistant to IFN-α2a treatment and the IFN-containing cytokine cocktail collected from peripheral blood mononuclear cells stimulated with TLR7/8 agonist. Mechanistically, both ADAR1 p110 and p150 promote L-HDAg production and inhibit HDV replication. ADAR1 p150, rather than p110, enhances the anti-HDV efficacy of IFN-α during de novo infection and cell mitosis, but not in resting cells. Moreover, different subtypes of IFN-α exhibit varying anti-HDV activities in both de novo infection and cell mitosis, due to their disparity in activating interferon responses. Among these, IFN-α2a, IFN-α10, and IFN-α14 exhibit the strongest anti-HDV activity and synergize with bulevirtide in suppressing HDV replication. In conclusion, the anti-HDV efficacy of IFN-α depends on multiple factors, including HDV genotypes, ADAR1 p150 level, IFN-α subtypes, and HDV's different survival strategies. These findings provide valuable implications for the development and optimization of IFN-based therapies. IMPORTANCE Chronic hepatitis delta virus (HDV) infection represents the most severe form of viral hepatitis. This study comprehensively evaluated the antiviral efficacy of interferon-alpha (IFN-α) subtypes across eight HDV genotypes during de novo infection, cell mitosis, or in quiescent cells. Herein, we found that IFN-α exhibits potent but varied efficiency against HDV 1-8 isolates upon de novo infection and cell mitosis. Conversely, HDVs in resting cells are resistant to IFN-α subtypes, regardless of the cellular ADAR1 levels. Among different subtypes, IFN-α2a, IFN-α10, and IFN-α14 exhibit the strongest anti-HDV activity and synergize with bulevirtide in suppressing HDV replication. These findings provide crucial insights into the optimization of IFN-based monotherapy and combinational therapy against chronic HDV infection. KEYWORDS HDV genotypes, interferon-alpha, cell division-mediated spread, de novo infection, resting cells A s a satellite virus of hepatitis B virus (HBV), hepatitis delta virus (HDV) is considered the human pathogen with the smallest known genome (~1.7 kb). HDV and HBV co-infection causes the most severe form of chronic viral hepatitis and significantly accelerates the progression of liver fibrosis, cirrhosis, and hepatocellular carcinoma (1). Despite the promotion of the HBV vaccination program, recent studies indicate that the
prevalence of HDV has either remained stable or increased globally (2)(3)(4)(5). It is estimated that 12-72 million people are chronically infected with HDV, causing substantial global morbidity and mortality (2)(3)(4)(5).
HDV virion comprises a ribonucleoprotein (RNP) complex and an HBV envelope. Upon de novo infection, HDV RNP was released in the cytoplasm and further translocated to the nucleus to initiate viral replication (6)(7)(8). In the nucleus, the genomic RNA serves as the template for the synthesis of the HDV mRNAs, allowing the production of S-HDAg. During replication, antigenomic RNA is edited by host adenosine deaminase acting on RNA 1 (ADAR1), introducing an A/G mutation in the amber stop codon of the S-HDAg open reading frame. Consequently, a second mRNA is produced coding for the L-HDAg with a C-terminal extension of 19 or 20 amino acids. In contrast to S-HDAg, required to initiate and promote HDV replication, the C-terminal extension confers L-HDAg dual distinct functions: inhibiting HDV RNA replication and supporting virion production (9). Besides HBV envelope-dependent de novo infection, recent studies indicated that HBV-independent cell division-mediated spread and long-lasting viral replication in resting cells collectively contribute to HDV persistence (10)(11)(12)(13).
The exploration of HDV virology has led to the development of novel therapeutic agents (14). For instance, the entry inhibitor bulevirtide (Hepcludex/Myrcludex-B, BLV) received conditional approval from the European Medicines Agency in 2020 as the first HDV-specific drug (15), and the farnesyl transferase inhibitor lonafarnib is being tested in clinical trials. Besides these advances, off-label interferon-alpha (IFN-α) (subtype 2a) has been the only commonly used treatment against HDV for decades (16). IFN-α2a or peg-IFN-α2a monotherapy is only partially effective and not curative for most chronic hepatitis D patients (17,18). Nevertheless, recent clinical studies have demonstrated that combinations of IFN-α with BLV and drugs under development (e.g., lonafarnib) exhibited additive or even strong synergistic antiviral effects in terms of faster and more profound reductions in serum HDV RNA, higher off-treatment responses, and lower relapse rates (11,(19)(20)(21)(22)(23)(24)(25)(26)(27)(28). These observations highlight the essential role of IFN-α in combating HDV infection, whether through IFN monotherapy or IFN-based combination therapies.
HDV is composed of eight genotypes (HDV-1 to HDV-8) defined by an intergenotype similarity >85% or >80%, according to the partial or full-length genome sequence, respectively (29). Eight HDV genotypes exhibited a specific phylogeographic distribu tion pattern worldwide (HDV-1: worldwide distribution; HDV-2: Southeast and East Asia; HDV-3: South America; HDV-4: Japan and China; and HDV-5 to HDV-8: Africa) (3,29). Previously, we cloned and characterized eight HDV 1-8 isolates, exhibiting marked differences in replication efficacy and virus production kinetics (30). Intriguingly, following IFN-α2a treatment, varied treatment responsiveness and prognoses have been reported among patients infected with different HDV genotypes. For instance, a London retrospective study found that patients infected with HDV-5 appeared to have a better prognosis with fewer episodes of hepatic decompensation and better response to peg-IFN treatment than patients infected with HDV-1 (31). In addition, a single Brazilian study reported an unusually high response rate (>95%) to a peg-IFN-entecavir combina tion treatment in non-European HDV-3-infected patients, suggesting that HDV-3 might represent an "easy-to-treat" variant compared to HDV-1 (32). These clinical discoveries underscore the importance of conducting the comparative investigation of IFN-α against HDV 1-8 isolates.
The cellular innate immune system plays a crucial role in virus suppression through the initiation of direct antiviral responses. As the most well-known antiviral cytokine, IFN-α induces hundreds of interferon-stimulated genes (ISGs) via the Janus kinase signal transducer and activator of transcription (JAK-STAT) signaling pathway, thereby triggering an active cellular antiviral state (33). Interestingly, humans express 12 IFN-α subtypes that share 75%-99% amino acid sequence identity (34,35). Although they bind to the same cellular receptor IFNAR1/IFNAR2 and eventually induce ISG expres sion via the JAK-STAT pathway, different subtypes exhibit distinct antiviral capabilities against different viruses. For instance, IFN-α14 exhibited the strongest antiviral activity against both HBV and HIV-1 (36,37), whereas IFN-α5 was superior against severe acute respiratory syndrome coronavirus 2 infection (38). Moreover, IFN-α16, IFN-α5, and IFN-α4 suppressed influenza A H3N2 replication with up to 230-fold greater efficiency than IFN-α2 (39). However, IFN-α2 outperformed other subtypes against the hepatitis E virus (40). To date, only the subtype IFN-α2 is widely used for HDV treatment. The therapeutic potential of other subtypes remains largely unknown.
Herein, we systematically analyzed the anti-HDV potency of IFN-α2a against HDV 1-8 isolates across three distinct conditions: (i) HDV de novo infection, (ii) HDV replication in resting cells, and (iii) HDV transmission in dividing cells. Moreover, a mixture of IFNcontaining cytokines was obtained from peripheral blood mononuclear cells (PBMCs) stimulated with TLR7/8 agonist, and its effectiveness against HDV was subsequently evaluated in these three different HDV surviving conditions. To investigate mechanisms influencing IFN-α efficacy, we explored the role of the host enzyme ADAR1 in modulating the antiviral response. Furthermore, the anti-HDV capabilities of different IFN-α subtypes were compared in three conditions as well. The underlying mechanisms that contribute to the anti-HDV disparity of IFN-α subtypes were elucidated accordingly. Moreover, the antiviral effect of the most potent IFN-α subtypes in combination with BLV was evaluated. This multimodal characterization of the antiviral effects of IFN-α subtypes against HDV 1-8 offers valuable insights for the development and optimization of IFN monotherapy and combinational therapies.
## MATERIALS AND METHODS
## Plasmids
Plasmids pcDNA3.1-HDV-1 to pcDNA3.1-HDV-8 containing 1.1 copy antigenome of HDV genotypes 1-8 (Table S1) and pLX304-HB2.7-B (NCBI accession number: MN645904) were generated as described previously (30). The pcDNA3.1-G418-HDV-1 to pcDNA3.1-G418-HDV-8 plasmids were generated by inserting the 1.1 copy antigenome of HDV genotypes 1-8 into pcDNA3.1 (+) vector harboring the neomycin resistance gene, respectively. The plasmid pJC126 (9, 41) harbored a 1.1-fold copy of the HDV (geno type 1) antigenome and a neomycin resistance gene, designated as pcDNA3.1-G418-HDV-1T. Plasmids pWPI-puro-ADAR1 p150 and pWPI-puro-ADAR1 p110 were generated by inserting human double-stranded RNA adenosine deaminase 1 (ADAR1) p150 (NCBI accession number: U10439) or spliced p110 into the lentiviral pWPI vector. For knocking down ADAR1, short-hairpin RNAs (shRNAs) were designed and inserted into the lentiviral pLVX vector (pLVX-puro-shADAR1). We also designed and cloned the sgRNA sequences into the pLentiCRISPR v.2 vector (Addgene vector #52961) for the knockout of the ADAR1 gene by the CRISPR/Cas9 system.
## Cells
Human NTCP overexpressing HepG2 NTCP and HuH7 NTCP cells were generated as described previously (30,(42)(43)(44). HepaRG NTCP cells were kept in AG Urban's lab (University Hospital Heidelberg). For the generation of HuH7-HDV-1T and HuH7-HDV-1 to -HDV-8 cells, HuH7 cells were transfected with pcDNA3.1-G418-HDV-1T and pcDNA3.1-G418-HDV-1 to pcDNA3.1-G418-HDV-8 plasmids, respectively. Positive cells were selected with G418. For the generation of ADAR1 shRNA knockdown or CRISPR knockout cells (shADAR1 or CRISPR-ADAR1 cells), HuH7 NTCP cells were transduced with lentiviruses encoding shRNA or sgRNA or their respective controls (shCTR and CRISPR-CTR), followed by puromy cin selection. For the generation of ADAR1 overexpression cells, HuH7 NTCP cells were transduced with lentiviruses encoding ADAR1 p150 or ADAR1 p110 (OE-ADAR1 p150 or OE-ADAR1 p110) or overexpression control (OE-CTR) and followed by puromycin selection. The technical support for preparing the primary hepatocytes/non-parenchy mal cells was provided by Liver Biotech (Shenzhen, China).
## Virus production and infection
The HDV virus particles (HDV-1T and HDV-1 to HDV-8) were produced by co-transfec tion of HuH7 cells with the corresponding HDV plasmids (pJC126, pcDNA3.1-HDV-1 to pcDNA3.1-HDV-8) and pLX304-HB2.7-B as previously described (9,30). For infection assays, hepatoma cell lines and primary human hepatocyte (PHH) cells were inoculated with 1/5 to 2/5 of the cell culture supernatant (if not mentioned otherwise) in a medium containing 4% polyethylene glycol (PEG) and 1.5% dimethyl sulfoxide (DMSO). After 16 hours of inoculation, the cells were washed twice with phosphate-buffered saline (PBS) or PHH maintenance medium, and the fresh DMEM medium or PHH maintenance medium containing 1.5% DMSO was provided and replaced every second day until the end of the experiment. For HDV cell division spreading assays, HDV-infected cells were trypsinized at 5-day post-infection (p.i.) and passaged at the 1:12 dilution to undergo cell division-mediated HDV spread (10). The HBV virus particles were produced by collecting the supernatant of the HepAD38 cells as previously described (45).
## De novo infection, dividing-cell (spread) model, and resting-cell model
HDV de novo infection and HDV spread in dividing cells and resting cells were performed as previously described (10,46).
## De novo infection
Cells were seeded to ∼100% confluence and maintained in 1.5% DMSO-containing medium. The next day, cells were inoculated with HDV for 16 hours, then washed and treated with IFN-α2a. On day 5 post-infection (when HDV replication was fully established) (10,46), samples were harvested to assess the inhibitory effect of IFN on HDV (10,46).
## Dividing-cell (spread) model and resting-cell model
Cells were seeded to ∼100% confluence and maintained in 1.5% DMSO-containing medium. On day 5 post-infection (when HDV replication was fully established) (10,46), cells were passaged at a 1:12 split and treated with IFN-α2a from day 5 to day 9. This condition allows quantification of IFN-α2a's impact on HDV propagation through cell division (10,46). This is the so-called dividing-cell (spread) model. In parallel, on day 5 post-infection, cells were kept in 1.5% DMSO-containing medium without passaging. Under these non-dividing conditions, HDV replication occurs without new rounds of de novo infection and cell division-mediated spread (10,46). This is the so-called resting-cell model.
## Preparation of recombinant IFN-I subtypes
The human IFN-α2a was purchased from PBL Assay Science (11101-1). Thirteen recombinant human IFN-I subtypes were produced as previously described (36,37,47). The human IgG1-linked type I interferon gene was inserted into the corresponding expression vector pCMV3. For transient transfection, plasmids were combined with sinofection transfection reagent (Catalog #STF02, Sinobiological) and introduced into HEK293 cells cultured in a serum-free medium. The cells were maintained in Erlenmeyer flasks on an orbital shaker or in a bioreactor with appropriate stirring at 37°C for 6 days. Subsequently, cell culture supernatants were collected and processed using affinity purification columns. The purified proteins were then analyzed via SDS-PAGE and quantified by UV-visible spectrophotometer.
## Production of IFN-containing cytokine mixture
The PBMCs were isolated from whole blood by Ficoll-Paque PLUS density gradient media (GE, 17-1440-02) under the guidance of the Ethical Committee. The TLR7/8 agonist resiquimod (R848) was purchased from Selleck (S8133) and diluted with DMSO to a final stock concentration of 10 mM. PBMCs were supplemented with culture medium (RPMI1640, 10% fetal bovine serum [FBS]) with 200 nM R848 or DMSO. The cell culture supernatants of PBMCs were collected after 24 hours post-incubation with R848 or DMSO.
## Indirect immunofluorescence
Immunofluorescence (IF) was performed as previously described (9,30). In short, cells were fixed with 4% paraformaldehyde (PFA) for 15 minutes at room temperature, followed by 30 minutes of incubation in permeabilization buffer (PBS, 0.25% Triton X-100). Then, cells were incubated with rabbit anti-HDAg polyclonal antibody diluted in 2% bovine serum albumin (BSA, 1:4,000, 4°C, 12 hours). CoraLite 594-conjugated goat anti-rabbit secondary antibody (Proteintech, SA00013-4, 1:500, 37°C, 1 hour) and Hoechst 33342 dye were used. Images were taken with an Olympus IX51 inverted microscope, and image analysis and quantification were performed using ImageJ software.
## RT-qPCR
Reverse transcription quantitative PCR (RT-qPCR) was performed as previously described (9,30). Briefly, total RNA was extracted from cells using the TRIzol RNA extraction kit (Invitrogen, 15596026) according to the protocol provided. The RNA secondary structure was melted by incubation of RNA at 95°C for 5 minutes, followed by immediate cooling down to -80°C. Then RNA was reverse transcribed using the HyperScript First-Strand cDNA Synthesis kit (APExBIO, K1071). Quantitative HDV RNA was performed on a Roche LightCycler 480 system using Luna Universal Probe qPCR Master Mix (New England Biolabs, M3004X) with the primers depicted in Table S2. SYBR Green Supermix (APExBIO, K1070) was used for detecting ISGs mRNA, HBV total RNA, and HBV pregenomic RNA with primers detailed in Table S2.
## In-cell enzyme-linked immunosorbent assay
In-cell enzyme-linked immunosorbent assay was performed as previously described (9,30). Cells were cultured in white, non-transparent 96-well plates and fixed in 4% PFA for 30 minutes, followed by 30 minutes of incubation in permeabilization buffer (PBS, 0.25% Triton X-100). After 30 minutes of incubation in blocking buffer (PBS, 0.05% Tween-20, 3% BSA), cells were incubated for 2 hours with the anti-HDAg monoclonal antibody. After extensive washing, endogenous peroxidases were blocked with a 3% hydrogen peroxide solution (10 minutes). Cells were incubated with a secondary goat anti-rabbit HRP (Jackson Immuno Research) antibody for 1 hour. After extensive washing, the chemiluminescence substrate (Advansta ELISABright) was added to the wells, and luminescence was measured on a plate reader.
## Western blot
Cell lysates were applied to SDS-PAGE on 10% SDS gels. Proteins were transferred onto nitrocellulose membranes by semidry transfer and incubated with primary antibodies as indicated below: rabbit anti-HDAg polyclonal antibody, mouse anti-β-actin (Sigma, A1978), mouse anti-GAPDH (Proteintech, 60004-1-lg), rabbit anti-ADAR1 (Cell Sig naling, 14175), STAT1 monoclonal antibody (Proteintech, 66545-1-lg), STAT2 polyclo nal antibody (Proteintech, 16674-1-AP), p-STAT1: phospho-STAT1 (Tyr701) polyclonal antibody (Proteintech, 28979-1-AP), and phospho-STAT2 (Tyr690) antibody (Affinity, AF3342). After overnight incubation at 4°C, membranes were washed with Tris-buf fered saline with Tween-20 (TBST), and incubated with fluorescent-labeled secondary antibodies (LI-COR Biosciences) for 1 hour at room temperature, washed again, and imaged on a LI-COR Odyssey imaging system.
## Cell viability assay
WST-1 reagent from the cell proliferation and viability detection kit (KeyGen Biotech, KGA316) was added to the cells in a 96-well plate. Cells were maintained at 37°C with 5% CO 2 for 2 hours. The absorbance of each well was read on the microplate absorbance readers (Bio-Rad, USA) at the wavelength of 450 nm.
## Interferon-stimulated response element luciferase reporter assays
HuH7 NTCP and HepG2 NTCP cells were transiently transfected with pGL4.45-luc2P-ISRE-Hygro and pRL-TK (a plasmid expressing the Renilla luciferase protein was used as an internal control) plasmids in 96-well cell culture plates. Eight hours post-treatment with IFN-α, the luciferase values were measured using the duo-lite luciferase reporter assay system (Vazyme, DL101-01) according to the manufacturer's instructions. Data were normalized by determining the ratios of firefly luciferase activities to that of Renilla luciferase (48).
## Statistical analysis
Statistical significance between two groups was determined using unpaired two-tailed Student's t-test, and one-way ANOVA analysis was used to compare the means of three or more groups in the software GraphPad Prism 9 (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, NS: not significant, P > 0.05). All error bars throughout the study represent the standard error of the mean.
## RESULTS
## IFN-α2a exhibited potent but varied antiviral efficacy against HDV 1-8 isolates upon de novo infection
Patients infected with different HDV genotypes showed varied treatment responsiveness and prognoses to IFN-α treatment (31,32,49). Given that multiple factors, including HDV genotypes and individual differences of patients, may contribute to treatment response in clinical settings, studies on the role of HDV genotype in response to IFN-α using more simplified in vitro infection systems were optimal. Therefore, cell culture-derived patient HDV 1-8 isolates were generated in vitro as described previously (30). HuH7 NTCP cells were inoculated with HDV 1-8 and treated with different concentrations of IFN-α2a (Fig. 1A). As shown in Fig. 1B through J, IFN-α2a dose-dependently inhibited HDV 1-8 upon de novo infection, with no cytotoxicity (Fig. S1A). Specifically, the IC 50 values varied in HuH7 NTCP cells (IC 50 : HDV-2, HDV-8 < HDV-5, HDV-4, HDV-7, HDV-3 < HDV-6, HDV-1), whereas different HDV infection rates exerted no significant effect on IC 50 values (Fig. S2). Consistently, similar genotype-specific responsiveness was observed in HepaRG NTCP (IC 50 : HDV-8, HDV-2 < HDV-4, HDV-7, HDV-5, HDV-3 < HDV-1, HDV-6, Fig. 1B; Fig. S3), further highlighting the genotype-specific susceptiveness of HDV to IFN-α2a.
IFN-α exerted its cellular antiviral activity by binding to the cellular receptor IFNAR1 and IFNAR2 to activate the JAK-STAT pathway, thus inducing hundreds of ISGs (Fig. 1K). Interestingly, despite similar expression levels of IFNAR1 and IFNAR2 in HuH7 NTCP and HepG2 NTCP cells (Fig. S4), IFN-α2a showed a weakened antiviral effect against HDV 1-8 in HepG2 NTCP cells (Fig. 1B through J), with no cytotoxicity (Fig. S1B). Notably, IFN-α2a induced higher levels of phosphorylated STAT1 and STAT2 in HuH7 NTCP cells compared to HepG2 NTCP cells (Fig. 1L). Consistently, in comparison with HepG2 NTCP cells, IFN-α2a treatment elicited stronger transactivation of interferon-stimulated response elements (ISREs) driven luciferase activity in HuH7 NTCP cells (Fig. 1M). Moreover, IFN-α2a provoked higher levels of the most well-studied antiviral ISGs in HuH7 NTCP cells (Fig. 1N andO; Fig. S5). These data clarified the differences in how IFN-α2a functioned in different cell models, indicating that the cell's intrinsic IFN response played a crucial role in determin ing the anti-HDV efficacy of IFN-α. Collectively, both HDV genotypes and the cell-intrinsic IFN response contributed to the differential sensitivity to IFN-α treatment during de novo HDV infection. In contrast to the minor anti-HDV effect in resting cells, IFN-α2a manifested efficient but varying antiviral potency against HDV 1-8 isolates during cell division
Besides de novo infection, HDV employed two additional strategies to ensure its persistence: cell division-mediated HDV spread and the maintenance of HDV in resting cells (11). Previously, based on the HDV-1 derivative (HDV-1T, pJC126), it was reported that in contrast to de novo infection, HDV replication was insensitive to IFN treatment once the infection was established (at days 5-6 p.i.) (46). Consistently, our indirect IF assay showed that IFN-α2a significantly inhibited HDV-1T and HDV-6 in the early treatment group (treated with IFN-α2a upon HDV de novo infection, days 0-6), whereas its antiviral effect was significantly diminished in the late treatment group (treated with IFN-α2a after the establishment of infection, days 6-12) (Fig. 2A through E). This observation was further confirmed in the HBV-HDV co-infection model (Fig. S6). These data prompted us to investigate whether such resistance is a pan-genotypic characteris tic of HDV. Indeed, although IFN-α2a elicited comparable ISG levels in early and late treatment groups (Fig. S7), it exhibited a limited antiviral activity against HDV 1-8 isolates in the late treatment group, contrasting with the robust and dose-dependent effects observed in the early treatment group (Fig. 2F; Fig. S8). Interestingly, HDV-5 and HDV-8 isolates exhibited relatively better responsiveness in the late treatment group compared to other genotypes (Fig. 2F; Fig. S8).
Next, based on HuH7 cells, HuH7-HDV-1 to HuH7-HDV-8 cells were generated as described previously (9,50). These eight cell lines continually supported HDV replication of HDV 1-8 isolates, whereas no de novo HDV infection occurred due to the lack of NTCP and HBV envelope proteins. Thus, these eight cell lines represented an optimal in vitro model to investigate the responsiveness of IFN-α against HDV 1-8 in cell resting or dividing state (Fig. 2G). According to previous studies, we chose day 5 p.i. in time for the passage of HDV-positive cells (10,46). Consistent with the late treatment group in Fig. 2F, HDV 1-8 isolates were refractory to IFN-α2a treatment in the cell resting state (P0), albeit with HDV-5 and HDV-8 showing relatively better response compared to other genotypes (Fig. 2H andI). Conversely, IFN-α2a exhibited a much stronger and dose-dependent antiviral effect against cell division-mediated HDV 1-8 spread (P1), with IC 50 values varying among different genotypes (IC 50 : HDV-8, HDV-4 < HDV-5, HDV-3, HDV-2, HDV-7, < HDV-6, HDV-1) (Fig. 2H andI). In sum, compared with the limited anti-HDV effect in resting cells, IFN-α2a potently inhibited HDV 1-8 in de novo infected and propagated via cell division, with efficacy varying among different genotypes.
## HDV in resting cells was refractory even with the treatment of IFN-containing cytokine cocktails
Currently, TLR7/8 agonists are in clinical development for the treatment of chronic hepatitis B (51)(52)(53). Mechanistically, TLR7/8 agonists potently activate PBMCs to promote the release of a mixture of cytokines (e.g., tumor necrosis factor-α [TNF-α] and IFNs) to exert a collective antiviral effect (54)(55)(56). Herein, TLR-7/8 agonist resiquimod (R848) was used to investigate whether an IFN-containing cytokine cocktail can efficiently inhibit HDV, especially in resting cells. This condition was more biologically relevant, as cells in vivo usually encounter multi-cytokines, rather than one cytokine alone. Thus, we stimulated PBMCs with R848 and harvested the supernatant (Fig. 3A), which included multiple cytokines, e.g., IFN-α2a, IFN-β, IFN-γ, IFN-λ1 and TNF-α (Fig. 3B; Fig. S9). Next, we evaluated the anti-HDV effect of the cytokine cocktail in three different conditions: de novo infection (Fig. 3C and D, early), HDV in resting cells (Fig. 3C andD, late, and Fig. 3E and F, P0), and cell division (Fig. 3E and F, P1). Similar to IFN-α2a treatment, R848-induced cytokine cocktail significantly inhibited HDV upon de novo infection or cell mitosis, whereas no effect in resting cells (Fig. 3D andF). Interestingly, although HDV-5 and HDV-8 showed a relatively better response with IFN-α2a treatment (P0), their response to R848-stimulated PBMCs cocktail (P0) was much weaker (Fig. 2I and3F). This was probably due to the fact that the cytokine composition and concentrations were not comparable in these two different settings. In sum, these results indicated that HDV in resting cells exhibited strong resistance with the presence of a mixture of IFN-containing cytokines.
## IFN-inducible ADAR1 p150 contributed to the anti-HDV efficacy of IFN-α during HDV de novo infection and cell mitosis, rather than in resting cells
The ADAR1 gene generated two isoforms via alternative splicing events: the full-length p150 and the spliced p110 (57). IFN induced the expression of full-length p150 (Fig. S10A andB) that could efficiently edit the amber/W site (58). HDV replication was sensitive to ADAR1 editing levels due to the determination of L-HDAg/S-HDAg ratios (57,59). Both shRNA and CRISPR/Cas9 methods were used to inhibit the expression of both forms of ADAR1 (p110 and p150) (Fig. S10C andG). As expected, the ratios of L-HDAg/S-HDAg in shADAR1 and CRISPR/Cas9-ADAR1 samples decreased dramatically compared to their control samples (Fig. S10C andG). For instance, at day 6, compared with shCTR and CRISPR-CTR cells, the L/S ratios in shADAR1 and CRISPR-ADAR1 cells decreased by 67% (Fig. S10C) and 93% (Fig. S10G), respectively. Notably, both the shADAR1 and CRISPR/ Cas9-ADAR1 samples showed significantly increased levels of HDV replication (Fig. S10D through F and H through J). Conversely, overexpression of p150 or p110 increased the ratios of L-HDAg/S-HDAg and inhibited HDV replication during de novo infection (Fig. S10K through N). For instance, at day 6, compared with OE-CTR cells, the L/S ratios in OE-ADAR1 p150 and OE-ADAR1 p110 cells increased 4.23-fold (Fig. S10K) and 3.65-fold (Fig. S10M), respectively. Hence, both p150 and p110 edit the amber/W site to facilitate the production of L-HDAg, thus increasing the ratio of L-HDAg/S-HDAg and inhibiting HDV replication in consequence. This prompted us to investigate the possible roles of ADAR1 in the responsiveness of HDV to IFN-α treatment in three different settings. Therefore, shADAR1 cells, CRISPR/ Cas9-ADAR1 cells (CRISPR-ADAR1), and their respective control cells were subjected to mock, early IFN-α2a treatment, or late IFN-α2a treatment (Fig. 4A). Overall, in both shADAR1 and control cells (shCTR), IFN-α2a significantly inhibited HDV in the early treatment group, whereas no effect was observed in the late treatment group (Fig. 4B andC; Fig. S11B). Consistently, similar results were observed in CRISPR-ADAR1 and control cells (Fig. 4D andE; Fig. S11C). Notably, ADAR1 knockdown decreased the anti-HDV effect of early IFN-α2a treatment compared with shCTR (Fig. 4B andC, columns 2 and 5; Fig. S11B), whereas no significant difference was observed in the late treatment groups (Fig. 4B andC, columns 3 and 6; Fig. S11B). This observation was further validated in CRISPR-CTR and CRISPR-ADAR1 cells by both IF and RT-qPCR assays (Fig. 4D andE; Fig. S11C). Conversely, in the setting of early IFN treatment, p150 overexpression (OE-ADAR1 p150) enhanced the anti-HDV effect of IFN-α2a (Fig. 4F andG, columns 2 and 5; Fig. S11D), whereas p110 overexpression (OE-ADAR1 p110) resulted in no significant differences (Fig. 4H andI, columns 2 and 5; Fig. S11E). Mechanistically, IFN-α2a treatment further upregulates the levels of ADAR1 p150 in OE-CTR (2.1-fold) cells, OE-ADAR1 p150 (2.6-fold) cells, and OE-ADAR1 p110 (2.3-fold) cells, while having no effect on ADAR1 p110 levels (Fig. S12). This observation supported that IFN-inducible ADAR1 p150 enhanced the anti-HDV effect during IFN-α2a treatment, whereas ADAR1 p110 did not. Next, we evaluated the efficacy of IFN-α2a in blocking HDV spread via cell division in ADAR1-overexpressing or ADAR1-knockout cell lines, with parallel assessments of HDAg positivity rates and cluster sizes (Fig. 4J; Fig. S11F andS13A). Consistent with the early IFN-α2a treatment groups (Fig. 4B through E), the downregulation of ADAR1 by shRNA or CRISPR/Cas9 methods slightly but significantly inhibited the anti-HDV efficacy of IFN-α2a in cell-dividing state (P1) (Fig. 4K andL; Fig. S11G, H, S13B andC, P1). Nevertheless, HDV replication in the cell resting state (P0) remained insensitive to IFN-α2a regardless of the levels of p110 and p150 (Fig. 4K through M; Fig. S11G through I,P0). Conversely, the overexpression of p150, rather than p110, promoted the anti-HDV efficacy of IFN-α2a in P1 cells (Fig. 4M, columns 4, 8, and 12; Fig. S13D, columns 2, 4, and6). To examine whether the varied antiviral efficacy of IFN-α2a across HDV 1-8 (Fig. 1 and2) was due to the differential expression levels of ADAR1, ADAR1 was quantified post-IFN-α2a treat ment with the presence of different HDV genotypes. However, IFN-α2a treatment induced comparable levels of ADAR1 p150 (p110 remained unchanged) across cells infected with different HDV genotypes (Fig. S14A through D). Moreover, no significant differences in ADAR1 p150 and p110 were observed among cells infected with HDV 1-8 (Fig. S14E through N), indicating that genotype-dependent IFN sensitivity was not explained by differential ADAR1 induction either. In conclusion, acting as an anti-HDV ISG, IFN-α-inducible ADAR1 p150 contributed to, but was not solely responsible for, IFNα-mediated anti-HDV activity during de novo infection and cell division. However, in resting cells, its role in IFN-α-mediated anti-HDV activity was negligible.
## IFN-α subtypes exhibited dose-dependent and varied anti-HDV activities during de novo infection and cell mitosis but were not effective in resting cells
Humans express 12 IFN-α subtypes with distinct antiviral capabilities against different viruses (31,32,38,49). Currently, only IFN-α2a is widely used to treat chronic HDV infection, leaving the anti-HDV potential of other IFN-α subtypes largely unexplored. Therefore, IFN-α subtypes and two type I IFN subtypes (IFN-β and IFN-ω) were cloned and expressed in HEK293 cells. Their anti-HDV potency was comparatively investigated in three conditions: HDV de novo infection (Fig. 5A), HDV in resting cells (Fig. 5D,P0), or in dividing cells (Fig. 5D,P1). During HDV de novo infection, all tested IFN-I subtypes significantly inhibited HDV replication in a dose-dependent manner (Fig. 5B andC; Fig. S15 andS16). Notably, varied IC 50 values were observed against HDV-1T (IC 50 : 10, 2a, 14 < 8, 5, β, 6, 17, ω, 4 < 16, 21, 1, 7) (Fig. 5B; Fig. S15C). Similar results were also obtained in HDV-6 (IC 50 : 10, 14, 2a< ω, 8, 5, β, 21, 17, 4 < 7, 1, 6, 16) (Fig. 5C; Fig. S16C). Of note, IFN-α10, IFN-α2a, and IFN-α14 were the most potent subtypes, whereas IFN-α1, IFN-α16, and IFN-α7 were the least (Fig. 5B andC; Fig. S15C andS16C). Specifically, the IC 50 values of IFN-α10, IFN-α2a, and IFN-α14 were approximately 100-fold lower than those of IFN-α1, IFN-α16, and IFN-α7 (Fig. 5B andC; Fig. S15C andS16C). Consistently, similar genotypespecific responsiveness was observed in PHH cells (Fig. S17). Subsequently, IFN-α10, IFN-α2a, and IFN-α7 were selected to further evaluate their anti-HDV effects in resting and dividing cells (Fig. 5D). In line with the results of HDV de novo infection (Fig. 5B andC), IFN-α10 and IFN-α2a displayed a much stronger anti-HDV effect in dividing cells (P1) compared to IFN-α7 (Fig. 5G andH). Nevertheless, in resting cells (P0), IFN-α10, IFN-α2a, (D) HuH7 NTCP cells were infected with HDV-1T. The anti-HDV efficacy of IFN-containing cytokine cocktail (supernatant, 5%, vol/vol) was detected by IF in the setting of both early treatment and late treatment (n = 6). (E) Schematic of the experimental setting. (F) HuH7-HDV-1 to HuH7-HDV-8 cells were either passaged at a 1:12 dilution (P1, dividing cells) or left without passaging (P0, resting cells). Next, cells were treated with IFN-containing cytokine cocktail (supernatant, 5%, vol/vol) for 5 days. HDAg-positive cells were quantified at day 5 post-treatment (n = 4). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, NS: not significant, P > 0.05. and IFN-α7 were all ineffective against HDV-1T (Fig. 5E andF). Similar results were observed across HDV 1-8 (Fig. 5I; Fig. S18), with the IC 50 values of IFN-α10 (P1) approxi mately 13.8 (HDV-7) to 890-fold (HDV-3) lower than those of IFN-α7 (Fig. 5I). In terms of mechanism, IFN-α10 and IFN-α2a triggered higher levels of STAT1 phosphorylation (Fig. S19A) and ISRE activation (Fig. S19B) compared to IFN-α7. Consequently, IFN-α10 and IFN-α2a were significantly more potent than IFN-α7 in the induction of the representative ISGs (Fig. S19C through X). These findings showed that enhanced IFN response by IFN-α10 and IFN-α2a correlated with their superior inhibition of HDV replication, consistent with the central role of JAK/STAT signaling in mediating IFN-I antiviral activity (46,60). This association supported a functional link between IFN-triggered JAK/STAT activation and anti-HDV efficacy.
## IFN-α subtypes synergize with bulevirtide against HDV
In clinical practice, combination therapies were widely employed to enhance antiviral efficacy and mitigate the risk of drug resistance. Previous studies demonstrated that combination therapies using IFN-α2a and BLV exhibited strong synergistic anti-HDV effects, leading to faster and more profound reductions in serum HDV RNA levels during treatment (11,(19)(20)(21)(22)(23)(24)(25)(26)61). Building on this foundation, we further investigated the synergistic potential of additional IFN-α subtypes, specifically IFN-α10 and IFN-α14, in combination with BLV (Fig. 6A). The results revealed that IFN-α10 and IFN-α14, similar to IFN-α2a (Fig. 6B andC), all displayed significant synergy with BLV against HDV (Fig. 6D through G). Notably, most synergistic dose windows were observed at 0.003-1 nM IFN-α2a (Fig. 6C), 0.0005-0.2 nM IFN-α10 (Fig. 6E), and 0.0016-0.73 nM IFN-α14 (Fig. 6G) when combined with 0.7-8 nM BLV, respectively. Consistently, similar results were observed in HuH7 NTCP cells infected with HDV-6 (Fig. S20) and HDV-8 (Fig. S21). These results highlight the potential of combining IFN-α10 and IFN-α14 with BLV as a promising therapeutic approach against HDV infection.
## DISCUSSION
Chronic HDV infection results in the most severe form of viral hepatitis. Although HDV was discovered more than 40 years ago, IFN-α2a is currently the only available treatment for HDV infection in most parts of the world (11,62). Intriguingly, HDV infection is complicated by three different survival strategies: HBV envelope-dependent de novo infection, HBV-independent persistence in quiescent cells, and cell division-mediated spread (10,11). Therefore, it is crucial to thoroughly investigate the effectiveness of IFN-α against HDV in these three different conditions. We found that, unlike the minor effect in resting cells, IFN-α2a showed efficient but varying antiviral potencies against HDV 1-8 isolates during de novo infection and cell division. Specifically, upon de novo infection, IFN-α2a is most effective against HDV-2 and HDV-8, while the least effective against HDV-6 and HDV-1 (Fig. 1B). In cell-dividing conditions, HDV-8 and HDV-4 are most sensitive to IFN-α2a treatment, whereas HDV-6 and HDV-1 are the least sensitive (Fig. 2H). Interestingly, in resting cells, although all HDV 1-8 isolates were resistant to IFN-α2a, HDV-5 and HDV-8 showed a relatively better response compared to other genotypes (Fig. 2F andH). Thus, the sensitivity of HDV to IFN-α treatment is highly heterogeneous with respect to HDV genotypes and HDV surviving conditions. Notably, the HDV-8 isolate is much more sensitive to IFN-α2a compared to other genotypes in all three different conditions. This in vitro finding implies that patients infected with HDV-8 may respond better to IFN-α treatment, but more clinical studies are needed to confirm this observation. Nevertheless, HDV-5 is much more susceptible to IFN-α2a compared to HDV-1 in all three different HDV surviving conditions (Fig. 1B and2H). Our in vitro observation aligns with the finding of a retrospective study in London, which suggests that patients infected with HDV-5 appeared to have a better prognosis and better response to peg-IFN-α2a treatment than patients infected with HDV-1 (31). Similarly, HDV-3 showed a relatively better response to IFN-α2a in comparison to HDV-1, which correlated with a study in Brazil reporting an unusually high response rate (>95%) to peg-IFN-entecavir combination treatment in non-European HDV-3-infected patients (32).
We look forward to more clinical retrospective studies to further validate the treatment responsiveness and prognoses of patients with different HDV genotypes following IFN-α2a treatment. We found that HDV persistence in resting cells was refractory to treatment with IFNcontaining cytokine cocktails (Fig. 3D through F). Currently, the underlying mechanisms remain unknown. By contrast, the combinations of IFN-α with extracellular spread inhibitors (e.g., bulevirtide and lonafarnib) could efficiently block both HDV de novo infection and cell division-mediated spread (11,(19)(20)(21)(22)(23)(24)(25)(26)(27)(28). However, there are currently no available antivirals that efficiently inhibit or even eliminate HDV in resting cells. Under standing the underlying mechanism of HDV resistance in quiescent cells and identifying potent antivirals represents the crucial final stage in eradicating HDV persistence and mitigating post-treatment relapses.
HDV replication is closely related to the editing activity of ADAR1, as it determines the ratios of L-HDAg/S-HDAg (57,59). Both forms of ADAR1 (p110 and p150) edited the amber/W site to promote the production of L-HDAg, which in turn inhibited HDV replication during de novo infection (Fig. S10). Therefore, it is important to thoroughly investigate the connections between IFN-α treatment and ADAR1 protein levels in three different surviving conditions of HDV. We found that HDV in quiescent cells remained resistant to IFN-α treatment, regardless of the levels of p110 and p150 (Fig. 4K through M). Nonetheless, IFN-inducible p150, rather than p110, slightly but significantly contrib uted to the anti-HDV effect of IFN-α during de novo infection and cell mitosis, although the levels of p150 remained lower than the levels of constitutively expressed p110 in the p150 overexpression cells (Fig. S10K). Moreover, IFN-α2a further increased the protein levels of p150 in both OE-ADAR1 p110 and OE-ADAR1 p150 cells, but had no effect on the levels of p110 (Fig. S12). These data highlighted that ADAR1 p150 is one of the important antiviral ISGs (effectors) that contribute to IFN-α2a's anti-HDV effect.
Among 12 human IFN-α subtypes, IFN-α2a has been extensively studied as a treat ment for various viral infections, including chronic HDV infection. In this study, to minimize technical variability, all IFN-α subtypes were produced using a standardized expression and purification protocol, with the same production batch consistently applied across all assays (37,47,63). This unified workflow ensures that the observed differences in antiviral activity reflect the intrinsic biological properties of each IFN subtype rather than experimental artifacts. Based on this platform, we found that IFN-α10, IFN-2a, and IFN-14 were the most potent subtypes against HDV, whereas IFN-α1, IFN-16, and IFN-7 were the least during de novo infection. We further selected three representative subtypes (IFN-α10, IFN-2a, and IFN-7) from each group and tested their anti-HDV effects in both actively dividing cells and resting cells. Consistent with de novo infection, IFN-α10 and IFN-α2a were significantly more potent than IFN-α7 in inhibiting HDV in actively dividing cells, which correlated with their ability to induce ISGs. There fore, the varied ability in ISG induction is a key factor in determining the effectiveness of different IFN-α subtypes against HDV. In addition, different affinities and/or interaction interfaces within the IFNAR receptor, as well as the difference in the number and spectrum of IFN subtype-specific regulated ISGs all play a role in determining the antiviral potencies of IFN-α subtypes against different viruses (36,38,64,65). Moreover, in clinical practice, combining multiple drugs is a common strategy to enhance antiviral efficacy and mitigate the risk of drug resistance. Our study verified that the most potent subtypes IFN-α10, IFN-α2a, and IFN-α14 all displayed significant synergy in combination with bulevirtide (Fig. 6; Fig. S20 andS21). These results highlight that among all human IFN-α subtypes, IFN-α2a, which is commonly used in clinical settings, is superior against HDV infection. Furthermore, IFN-α10 and IFN-14 hold great potential for the develop ment of alternative IFN-based therapeutic approaches.
Taken together, we demonstrated that the effectiveness of IFN-α in fighting against HDV depends on multiple factors, including HDV genotypes, the levels of IFN-inducible ADAR1 p150 protein, subtypes of IFN-α, and three different HDV surviving strategies. Our study highlights the necessity to understand the mechanisms of HDV resistance in quiescent cells and to develop potent antivirals that specifically target HDV in these resting cells. These efforts are essential in eradicating HDV persistence and reducing the risk of post-treatment relapses.
## References
1. Farci, Niro (2012) "Clinical features of hepatitis D" *Semin Liver Dis*
2. Chen, Shen, Han et al. (2019) "Prevalence and burden of hepatitis D virus infection in the global population: a systematic review and metaanalysis" *Gut*
3. Ding, Guo, Hong et al. (2024) "The distinct spatiotemporal evolutionary landscape of HBV and HDV largely determines the unique epidemic features of HDV globally" *Mol Phylogenet Evol*
4. Miao, Zhang, Ou et al. (2020) "Estimating the global prevalence, disease progression, and clinical outcome of hepatitis delta virus infection" *J Infect Dis*
5. Stockdale, Kreuels, Henrion et al. (2020) "The global prevalence of hepatitis D virus infection: systematic review and meta-analysis" *J Hepatol*
6. Lucifora, Delphin (2020) "Current knowledge on hepatitis delta virus replication" *Antiviral Res*
7. Lempp, Ni, Urban (2016) "Hepatitis delta virus: insights into a peculiar pathogen and novel treatment options" *Nat Rev Gastroenterol Hepatol*
8. Sureau, Negro (2016) "The hepatitis delta virus: replication and pathogenesis" *J Hepatol*
9. Guo, Li, Li et al. (2024) "Molecular determinants within the C-termini of L-HDAg that regulate hepatitis D virus replication and assembly" *JHEP Rep*
10. Zhang, Ni, Lempp et al. (2022) "Hepatitis D virus-induced interferon response and administered interferons control cell division-mediated virus spread" *J Hepatol*
11. Zhang, Urban (2021) "New insights into HDV persistence: the role of interferon response and implications for upcoming novel therapies" *J Hepatol*
12. Giersch, Bhadra, Volz et al. (2019) "Hepatitis delta virus persists during liver regeneration and is amplified through cell division both in vitro and in vivo" *Gut*
13. Gnouamozi, Zhang, Prasad et al. (2024) "Analysis of replication, cell division-mediated spread, and HBV envelope protein-dependent pseudotyping of three mammalian delta-like agents" *Viruses*
14. Majeed, Zehnder, Koh et al. (2023) "Hepatitis delta: epidemiology to recent advances in therapeutic agents" *Hepatology*
15. Urban, Neumann-Haefelin, Lampertico (2021) "Hepatitis D virus in 2021: virology, immunology and new treatment approaches for a difficult-to-treat disease" *Gut*
16. Sandmann, Wedemeyer (2023) "Interferon-based treatment of chronic hepatitis D" *Liver Int*
17. Yurdaydin, Keskin, Kalkan et al. (2018) "Interferon treatment duration in patients with chronic delta hepatitis and its effect on the natural course of the disease" *J Infect Dis*
18. Wedemeyer, Yurdaydìn, Dalekos et al. (2011) "Peginterferon plus adefovir versus either drug alone for hepatitis delta" *N Engl J Med*
19. Asselah, Lampertico, Aleman et al. (2025) "Bulevirtide monotherapy is safe and well tolerated in chronic hepatitis delta: an integrated safety analysis of bulevirtide clinical trials at Week 48" *Liver Int*
20. Asselah, Chulanov, Lampertico et al. (2024) "Bulevirtide combined with pegylated interferon for chronic hepatitis D" *N Engl J Med*
21. El Messaoudi, Brichler, Fougerou-Leurent et al. (2024) "Effect of Peg-IFN on the viral kinetics of patients with HDV infection treated with bulevirtide" *JHEP Rep*
22. Jachs, Panzer, Hartl et al. (2023) "Long-term follow-up of patients discontinuing bulevirtide treatment upon long-term HDV-RNA suppression" *JHEP Rep*
23. Lampertico, Degasperi, Sandmann et al. (2023) "Hepatitis D virus infection: pathophysiology, epidemiology and treatment" *JHEP Rep*
24. Lampertico, Roulot, Wedemeyer (2022) "Bulevirtide with or without pegIFNα for patients with compensated chronic hepatitis delta: from clinical trials to real-world studies" *J Hepatol*
25. Wedemeyer, Schöneweis, Bogomolov et al. (2023) "Safety and efficacy of bulevirtide in combination with tenofovir disoproxil fumarate in patients with hepatitis B virus and hepatitis D virus coinfection (MYR202): a multicentre, randomised, parallel-group, open-label, phase 2 trial" *Lancet Infect Dis*
26. Wedemeyer, Schöneweis, Bogomolov et al. (2020) "48 weeks of high dose (10 mg) bulevirtide as monotherapy or with peginterferon alfa-2a in patients with chronic HBV/HDV co-infection" *J Hepatol*
27. Yurdaydin, Keskin, Kalkan et al. (2018) "Optimizing lonafarnib treatment for the management of chronic delta hepatitis: the LOWR HDV-1 study" *Hepatology*
28. Yurdaydin, Keskin, Yurdcu et al. (2022) "A phase 2 dose-finding study of lonafarnib and ritonavir with or without interferon alpha for chronic delta hepatitis" *Hepatology*
29. Gal, Brichler, Drugan et al. (2017) "Genetic diversity and worldwide distribution of the deltavirus genus: a study of 2,152 clinical strains" *Hepatology*
30. (2025) *Full-Length Text Journal of Virology*
31. Wang, Lempp, Schlund et al. (2021) "Assembly and infection efficacy of hepatitis B virus surface protein exchanges in 8 hepatitis D virus genotype isolates" *J Hepatol*
32. Spaan, Carey, Bruce et al. (2020) "Hepatitis delta genotype 5 is associated with favourable disease outcome and better response to treatment compared to genotype 1" *J Hepatol*
33. Borzacov, De Figueiredo Nicolete, Souza et al. (2016) "Treatment of hepatitis delta virus genotype 3 infection with peg-interferon and entecavir" *Int J Infect Dis*
34. Wang, Xu, Su et al. (2017) "Transcriptional regulation of antiviral interferon-stimulated genes" *Trends Microbiol*
35. Hardy, Owczarek, Jermiin et al. (2004) "Characterization of the type I interferon locus and identification of novel genes" *Genomics*
36. Wittling, Cahalan, Levenson et al. (2020) "Shared and unique features of human interferon-beta and interferon-alpha subtypes" *Front Immunol*
37. Chen, Li, Lai et al. (2021) "Functional comparison of interferon-α subtypes reveals potent hepatitis B virus suppression by a concerted action of interferon-α and interferon-γ signaling" *Hepatology*
38. Lavender, Gibbert, Peterson et al. (2016) "Interferon alpha subtypespecific suppression of HIV-1 infection in vivo" *J Virol*
39. Schuhenn, Meister, Todt et al. (2022) "Differential interferon-α subtype induced immune signatures are associated with suppression of SARS-CoV-2 infection" *Proc Natl Acad Sci*
40. Matos A Da R, Wunderlich, Schloer et al. (2019) "Antiviral potential of human IFN-α subtypes against influenza A H3N2 infection in human lung explants reveals subtype-specific activities" *Emerg Microbes Infect*
41. Todt, François, Behrendt et al. (2016) "Antiviral activities of different interferon types and subtypes against hepatitis E virus replication" *Antimicrob Agents Chemother*
42. Kuo, Goldberg, Coates et al. (1988) "Molecular cloning of hepatitis delta virus RNA from an infected woodchuck liver: sequence, structure, and applications" *J Virol*
43. Ni, Zhang, Engelskircher et al. (2019) "Generation and characterization of a stable cell line persistently replicating and secreting the human hepatitis delta virus" *Sci Rep*
44. Yi, Lempp, Mehrle et al. (2014) "Hepatitis B and D viruses exploit sodium taurocholate co-transporting polypeptide for species-specific entry into hepatocytes" *Gastroenterology*
45. Lempp, Schlund, Rieble et al. (2019) "Recapitulation of HDV infection in a fully permissive hepatoma cell line allows efficient drug evaluation" *Nat Commun*
46. Ni, Urban (2017) "Hepatitis B virus infection of HepaRG cells, HepaRG-hNTCP cells, and primary human hepatocytes" *Methods Mol Biol*
47. Zhang, Filzmayer, Ni et al. (2018) "Hepatitis D virus replication is sensed by MDA5 and induces IFN-β/λ responses in hepatocytes" *J Hepatol*
48. Shen, Li, Wang et al. (2018) "Hepatitis B virus sensitivity to interferon-α in hepatocytes is more associated with cellular interferon response than with viral genotype" *Hepatology*
49. Zhang, Li, Hou et al. (2024) "miR-26a exerts broad-spectrum antiviral effects via the enhancement of RIG-I-mediated type I interferon response by targeting USP15" *Microbiol Spectr*
50. Roulot, Brichler, Layese et al. "Deltavir study group. 2020. Origin, HDV genotype and persistent viremia determine outcome and treatment response in patients with chronic hepatitis delta" *J Hepatol*
51. Ni, Zhang, Engelskircher et al. (2019) "Generation and characterization of a stable cell line persistently replicating and secreting the human hepatitis delta virus" *Sci Rep*
52. Gane, Dunbar, Brooks et al. (2023) "Safety and efficacy of the oral TLR8 agonist selgantolimod in individuals with chronic hepatitis B under viral suppression" *J Hepatol*
53. Amin, Colbeck, Daffis et al. (2021) "Therapeutic potential of TLR8 agonist GS-9688 (selgantolimod) in chronic hepatitis B: remodeling of antiviral and regulatory mediators" *Hepatology*
54. Yuen, Balabanska, Cottreel et al. (2023) "TLR7 agonist RO7020531 versus placebo in healthy volunteers and patients with chronic hepatitis B virus infection: a randomised, observer-blind, placebo-controlled, phase 1 trial" *Lancet Infect Dis*
55. Hofmann, Vanwalscappel, Bloch et al. (2016) "TLR7/8 agonist induces a post-entry SAMHD1-independent block to HIV-1 infection of monocytes" *Retrovirology (Auckl)*
56. Krug, Kiefer, Koelle et al. (2021) "TLR7/8 regulates type I and type III interferon signalling in rhinovirus 1binduced allergic asthma" *Eur Respir J*
57. Thomas, Laxton, Rodman et al. (2007) "Investigating toll-like receptor agonists for potential to treat hepatitis C virus infection" *Antimicrob Agents Chemother*
58. Casey (2012) "Control of ADAR1 editing of hepatitis delta virus RNAs" *Curr Top Microbiol Immunol*
59. Patterson, Samuel (1995) "Expression and regulation by interferon of a double-stranded-RNA-specific adenosine deaminase from human cells: evidence for two forms of the deaminase" *Mol Cell Biol*
60. Hartwig, Schütte, Warnecke et al. (2006) "The large form of ADAR 1 is responsible for enhanced hepatitis delta virus RNA editing in interferon-alpha-stimulated host cells" *J Viral Hepat*
61. Darnell, Kerr, Stark (1994) "Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins" *Science*
62. Wedemeyer, Schöneweis, Bogomolov et al. (2019) "GS-13-final results of a multicenter, open-label phase 2 clinical trial (MYR203) to assess safety and efficacy of myrcludex B in cwith PEG-interferon Alpha 2a in patients with chronic HBV/HDV co-infection" *J Hepatol*
63. (2025) *Full-Length Text Journal of Virology*
64. Ghany, Buti, Lampertico et al. (2023) "Guidance on treatment endpoints and study design for clinical trials aiming to achieve cure in chronic hepatitis B and D: report from the 2022 AASLD-EASL HBV-HDV Treatment Endpoints Conference" *Hepatology*
65. Chen, Li, Lai et al. (2021) "Functional comparison of interferon-α subtypes reveals potent hepatitis B virus suppression by a concerted action of interferon-α and interferonγ signaling" *Hepatology*
66. Jaks, Gavutis, Uzé et al. (2007) "Differential receptor subunit affinities of type I interferons govern differential signal activation" *J Mol Biol*
67. Lavoie, Crisafulli-Cabatu, Abramovich et al. (2011) "Binding and activity of all human alpha interferon subtypes" *Cytokine* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12791715&blobtype=pdf | # 292. Outcomes of Oral Beta-Lactams versus IV Antibiotics for Gram-Negative Bacteremia in Solid Organ Transplant Patients
Milen Thomas, Wesley Hoffmann, William Musick, Shemual Tsai, Masayuki Nigo
Background. Solid organ transplant recipients (SOT) are at high risk for infections, including gram-negative bacteremia caused by Enterobacterales. While intravenous (IV) antibiotics (abx) are often used, recent studies in immunocompetent patients show that oral abx can be equally effective with fewer adverse events. Data in solid organ transplant recipients is limited, particularly regarding oral beta-lactams (BL).
Methods. This retrospective cohort study included adult SOT recipients admitted to the Houston Methodist Hospital System between June 2016 and September 2023 with a first episode of Enterobacterales bacteremia. Patients discharged on either oral BL or IV antibiotics after initial IV therapy were included. Those with deep-seated infections requiring prolonged antibiotic treatment were excluded. The primary outcome was a composite endpoint of all-cause mortality, recurrence of infection, reinitiation of antibiotics, or unplanned healthcare visit within 30 days of antibiotic completion.
Oral Abstracts • OFID 2026:13 (Suppl 1) • S63
Results. A total of 864 bacteremic patients were identified, and 182 were included: 105 in the oral BL arm and 77 in the IV arm. The median patient age was 59.5 years, and 104 of the patients were female. Urinary tract infections were the most common source with 131 cases, and Escherichia coli was the predominant organism with 135 cases. Patients in the BL arm received a median of 6 days of IV therapy before being transitioned. Of these, 37 (35%) in the BL group, 34 (44%) in the IV group met the primary composite endpoint (p=0.22). There were no significant differences in 30-day bacteremia recurrence, 30-day source recurrence, unplanned healthcare visits, or reinitiation of antibiotics between the oral BL group and the IV group. Adverse events and Clostridioides difficile infection rates were low across all groups without significant differences.
Conclusion. Among SOT recipients treated for gram-negative bacteremia, step down therapy with oral BL were not associated with worse outcomes compared to IV abx. These findings suggest oral BL may be a reasonable step-down option in transplant patients with gram-negative bacteremia without deep-seated infections and further prospective studies are warranted to confirm these results.
Disclosures. All Authors: No reported disclosures 1 1 2 1 1 1 Mayo Clinic, Rochester, MN 2 Mayo Clinic Rochester, Rochester, Minnesota 1 . Most of these patients require multiple prophylactic antimicrobials to prevent opportunistic infections during the post-transplant course 2 , including central nervous system (CNS) infections 3 . Brain abscesses are a form of CNS infection with a high incidence of morbidity and mortality 4 S64 • OFID 2026:13 (Suppl 1) • Oral Abstracts |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12177958&blobtype=pdf | # Retraction Note: Interleukin-27 ameliorates coxsackievirus-B3-induced viral myocarditis by inhibiting Th17 cells
Ping Liu, Hengshan Zhu, Chuang Lou
## References
1. Gary, Vanasse, Winn et al. (2004) "Bcl-2 overexpression leads to increases in suppressor of cytokine Signaling-3 expression in B cells and de Novo follicular lymphoma" *Mol Cancer Res*
2. Hu, Dong, Yue (0209) "In vivo delivery of interleukin-35 relieves coxsackievirus-B3-induced viral myocarditis by inhibiting Th17 cells" *Arch Virol* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12271755&blobtype=pdf | # Prevalence of Chlamydia trachomatis in eye specimens of patients suspected of having viral keratitis: a cross-sectional study
Arash Letafati, Parsa Ghafari, Niloofar Mobarezpour, Mohammad Haddadi, Mersedeh Arbabinia, Zahra Rostami, Yasamin Meamarzadegan, Aniseh Dadgar, Zahra Tayebi, Patricia Schlagenhauf
## Abstract
Background: Chlamydia trachomatis (C. trachomatis) is a major global health concern, recognized among the leading bacterial causes of sexually transmitted infections and implicated in ocular diseases. Its association with chronic follicular conjunctivitis and severe papillary inflammation underscores the importance of accurate identification in diagnosing trachoma. This study evaluated the prevalence of C. trachomatis in patients suspected to viral keratitis referred to the lab and comparing four different eye specimen types. Methods: This cross-sectional study (2020-2022) involved 161 suspected to viral keratitis patients referred to thet lab and checked for viral and bacterial infections (49.1 % female, 50.9 % male) at Tehran University's Clinical Virology Research Center. Tear fluid, corneal epithelium, and aqueous/vitreous humor samples were analyzed using the Qiagen Mini Blood Kit for DNA extraction and Multiplex Real-Time PCR with the Fast-track diagnostics/SIEMENS eye kit. C. trachomatis was detected in 14 patients (8.7 %), who showed clinical features such as follicular conjunctivitis, corneal neovascularization, epithelial erosions, and conjunctival scarring. Details on pathology, disease course, treatments, and outcomes are provided in. Conclusions: This study highlights the prevalence of C. trachomatis in suspected keratitis cases, offering a comparative view across different eye specimen types. Accurate detection using molecular assays supports timely intervention and targeted treatment, improving diagnostic precision and patient outcomes.
## 1. Introduction
Chlamydia trachomatis (C. trachomatis) is a major sexually transmitted pathogen and one of the most prevalent bacterial infections globally. Of its 15 serovars, four (A, B, Ba, and C) can cause ocular infections, including trachoma and inclusion conjunctivitis, underscoring their significant impact on public health [1].
Trachoma, a leading cause of infectious blindness worldwide, is caused by C. trachomatis and marked by follicle formation from lymphocyte infiltration beneath the eye surface. It remains endemic in resource-limited regions of Africa, Asia, and South America, especially in countries like India, Ethiopia, and Nigeria. [2,3]. In Iran, trachoma is relatively rare, and most keratitis cases are not attributed to C. trachomatis. Following the World Health Organization (WHO) through the year 2022, 125 million people were settled in trachoma endemic areas and were threatened by blindness [4].
C. trachomatis is a known cause of ocular infections, with repeated conjunctival infections leading to chronic follicular conjunctivitis (TF) on the upper eyelid. Severe cases show intense papillary inflammation (TI), and active trachoma is defined by the presence of both TF and TI [5].
Advancements in molecular biology, particularly molecular assays, have transformed the diagnosis of many infectious agents across various clinical samples. In trachoma, accurate pathogen identification is essential to prevent complications and ensure effective treatment.
The objective of the study was to evaluate the prevalence of C. trachomatis in patients suspected to keratitis with comparative view of four different eye specimen types.
## 2. Material and Methods
## 2.1. Study population
This prospective cross-sectional study (2020-2022) was conducted at Tehran University's Clinical Virology Research Center. Based on an estimated C. trachomatis prevalence of 10 % in suspected keratitis cases, a sample size of 138 was calculated (95 % confidence, 5 % margin of error). To ensure adequate power, 161 patients (79 females, 82 males) were recruited and examined by ophthalmologists.
Suspected keratitis cases were identified based on symptoms like eye pain, photophobia, vision loss, redness, and tearing, along with clinical signs such as corneal opacity, epithelial defects, or stromal infiltration. Ophthalmologists also reviewed histories of ocular trauma, contact lens use, prior eye infections, or related systemic infections to confirm eligibility and proceed with specimen collection.
## 2.2. Sample collection and genome extraction
Samples were collected from the affected eye, including tear fluid, corneal epithelium, aqueous humor, and vitreous humor, under sterile conditions. A total of 87 right eyes and 74 left eyes were sampled. Ophthalmologists performed deep corneal scrapings using sterile stainless-steel blades under a slit lamp. Tear fluid was obtained by rinsing the ocular surface with 500 μl of saline three times. For epithelial keratitis, corneal epithelium was scraped from the ulcer's edge with plastic swabs. Aqueous humor (0.1 μl) was collected via anterior chamber paracentesis with a 27-gauge needle for stromal or endothelial keratitis. Vitreous humor was also sampled. The specimens, stored in 2 mL of Phosphate Buffer Saline (PBS), were transported to the lab.
Using the Qiagen Mini Blood Kit (Qiagen, Hilden, Germany) and following the manufacturer's guidelines, bacterial nucleic acids were extracted from 100 μl of each specimen. Additionally, the concentration of the extracted genome was evaluated using nanodrop at A 260/280, aiming for an A ratio of approximately 1.8.
## 2.3. Real-time PCR
Real-time PCR was performed using the Fast-track diagnostics/ SIEMENS eye kit (Esch-sur-Alzette, Luxembourg) to detect C. trachomatis. The multiplex PCR targets the 16S rRNA gene, which is conserved across all Chlamydia species for reliable detection. The assay amplifies a 120 base pair segment using the following primers: Forward 5′-AGCAGGAGGAGTGAAGGTG-3′ and Reverse 5′-GCCAGCA-GAGTGGTGTTAC-3'. PCR reactions were performed in a 5 μL volume with thermal cycling: initial denaturation at 95 • C for 3 min, followed by 40 cycles of denaturation at 95 • C for 15 s, annealing at 60 • C for 30 s, and extention at 72 • C for 30 s. Quantification of the PCR product was assessed by measuring the quantification cycle (Cq) value. A Cq value < 35 was considered positive, while >35 was negative.
The primer set targeting the 16S rRNA gene is specific to C. trachomatis, minimizing cross-reactivity and improving diagnostic accuracy. Validated by multiple studies and manufacturer data, the assay has a sensitivity of 96-100 % and specificity of 98-100 %, making it reliable for detecting C. trachomatis in ocular samples.
## 2.4. Statistical analysis
Data were analyzed using STATA (ver. 17) to assess Chlamydia infection, comparing positive test rates across sex and age groups with independent t-tests. Odds Ratio (OR) was calculated to compare group effects, with a significance level of 0.05. After statistical description, the reliability and accuracy of the findings were evaluated.
## 3. Results
## 3.1. Demographic data
The study included 161 patients (82 men, 79 women), with 33 (20.5 %) under 18, 50 (31.1 %) aged 18-49, and 78 (48.4 %) over 50. Of these, 14 (8.7 %) tested positive for C. trachomatis, and 147 (91.3 %) tested negative. The highest detection rate was in tear fluid (92.9 %). In terms of eye laterality, 8 infections were from the right eye and 6 from the left. No reinfections were noted. Alternative diagnoses for negative cases included Staphylococcus aureus (38), Pseudomonas aeruginosa (22), fungal keratitis (17), and other bacterial or idiopathic keratitis (70).
## 3.2. Prevalence of C. trachomatis
We analyzed the prevalence of C. trachomatis infection across the entire cohort, as well as separately for men and women. Subsequently, we conducted detailed statistical analyses, as summarized in Tables 1 and2.
In the male subgroup, 5 out of 82 (6 %) tested positive, while 9 out of 79 women (11.4 %) tested positive. The infection rate was 5.4 % higher in women, but this difference was not statistically significant (P value = 0.116). The Odds Ratio (OR) of infection between women and men was 1.98, indicating women are nearly twice as likely to be infected, though this difference was not statistically significant (P value = 0.24).
In the age group under 18, 5 out of 33 (15 %) tested positive, while 28 (85 %) tested negative. The group aged 18-49 had the lowest infection rate, serving as the baseline. Compared to this group, the under-18 group showed a 0.11 higher infection rate, statistically significant (P value = 0.038). The over-50 group had a 0.05 higher rate, but this difference was not significant (P value = 0.140). Detailed results are in Tables 2 and3.
## 3.3. Detection in different samples
We assessed the effectiveness of three sampling techniques for detection, with results in Table 3. Corneal epithelium identified 21.4 % of positive cases, while tear fluid detected 92.9 %. The 71.5 % difference in detection efficacy was statistically significant (P value = 0.0001).See Table 4
## 3.4. Risk factors among patients
Among the 161 patients, 47 (29.2 %) used contact lenses, 18 (11.2 %) had recent ocular trauma, and 23 (14.3 %) had used topical corticosteroids in the past month. Other risk factors included prior ocular infection (19 patients, 11.8 %) and systemic diseases with ocular involvement (11 patients, 6.8 %). These factors were present in both C. trachomatis-positive and -negative patients, with no statistically significant associations found.
## 4. Discussion
The analysis showed tear fluid samples had the highest positive infection rate, followed by corneal epithelium. This suggests tear fluid is
## Table 1
The proportions and the test of the association between infection and gender. 4, patients with C. trachomatis keratitis exhibited diverse clinical features, including follicular reactions, chronic conjunctivitis, and pannus formation, with some having bilateral involvement or a relapsing course, indicating the chronic nature of chlamydial ocular infections. These findings align with previous data [6], highlighting the importance of PCR testing to differentiate C. trachomatis from other microbial causes, especially in atypical or treatment-resistant cases.
Although C. trachomatis is traditionally associated with ocular infections, its role in keratitis may be limited. Only 8.7 % of our cohort tested positive, a figure consistent with literature suggesting C. trachomatis is not the leading cause of keratitis [7]Our findings reinforce that in Iran, common etiologies include Staphylococcus aureus, Pseudomonas aeruginosa, and fungi. Results from 161 subjects of varying ages indicated a higher prevalence in individuals over 50 and under 18, with noticeably higher infection rates in females.
A review by [12]. Compared to our study, we showed higher infection rate. Our study included older patients which showed high infection rate and that may be a reason for its elevated prevalence compared to this study.
Although the study had a statistically adequate population, further research with a larger sample is recommended. Limitations include the relatively small sample size, absence of comprehensive microbial testing in cases, and lack of long-term follow-up to assess reinfection or treatment outcomes.
## 5. Conclusion
The study found a low prevalence of C. trachomatis in keratitissuspected patients, suggesting a re-evaluation of its role in keratitis and ocular infections, particularly in non-endemic regions like Iran. The findings highlight varying infection rates across age groups, with a significant difference in those under 18. While gender and older age group differences were not statistically significant, tear fluid showed heightened sensitivity in detecting C. trachomatis. Future research should expand diagnostic panels and focus on age-specific interventions for better management of C. trachomatis infections.
## References
1. Faris, Andersen, Mccullough et al. (2019) "Chlamydia trachomatis serovars drive differential production of proinflammatory cytokines and chemokines depending on the type of cell infected" *Front Cell Infect Microbiol*
2. Hu, Harding-Esch, Burton et al. (2010) "Epidemiology and control of trachoma: systematic review" *Trop Med Int Health*
3. Wolle, West (2019) "Ocular Chlamydia trachomatis infection: elimination with mass drug administration" *Expert Rev Anti Infect Ther*
4. (2023) ":~:text=Trachoma%20is%20a% 20blinding%20diseasepiurc"
5. Ramadhani, Derrick, Holland et al. (2016) "Blinding trachoma: systematic review of rates and risk factors for progressive disease" *PLoS Neglected Trop Dis*
6. Solomon, Burton, Gower et al. (2022) *Nat Rev Dis Primers*
7. Cabrera-Aguas, Khoo, Watson (2022) "Infectious keratitis: a review" *Clin Exp Ophthalmol*
8. Teweldemedhin, Gebreyesus, Atsbaha et al. (2017) "Bacterial profile of ocular infections: a systematic review" *BMC Ophthalmol*
9. Sharma, Satpathy, Nayak et al. (2012) "Ocular Chlamydia trachomatis infections in patients attending a tertiary eye care hospital in north India: a twelve year study" *Indian J Med Res*
10. Bobo, Viscidi, Quinn et al. (1991) "Diagnosis of Chlamydia trachomatis eye infection in Tanzania by polymerase chain reaction/enzyme immunoassay" *The Lancet*
11. Macleod, Butcher, Mudaliar et al. (2016) "Low prevalence of ocular Chlamydia trachomatis infection and active trachoma in the Western Division of Fiji" *PLoS Neglected Trop Dis*
12. Nash, Chernet, Weiss et al. (2023) "Prevalence of ocular Chlamydia trachomatis infection in Amhara region, Ethiopia, after 8 years of trachoma control interventions" *Am J Trop Med Hyg* |
biology | europe-pmc | https://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC12777802&blobtype=pdf | # Detection of the BK virus in urban wastewater inlets into the Karun River
Parvaneh Bahrami, Afshin Takdastan, Manoochehr Makvandi, Sahand Jorfi, Mohammad Karimi, Baba Ahmadi, Abdolkazem Neisi
## Abstract
Background and Objectives:The presence of the BK virus in wastewater indicates pollution, as it is shed in the urine of infected individuals. Over fifty percent of the human population remains asymptomatic for the BK virus. Reactivation of the BK virus in immunosuppressed individuals can lead to serious health issues, including cystic hemorrhagic, nephritis, and kidney graft rejection. The BK virus is associated with various cancers, such as head and neck, prostate, bladder, and colorectal cancers. The urban wastewater inlets into the Karun River cause the river contamination. This study focused on detecting the BK virus in wastewater inlet in Karun River, Ahvaz city, Iran. Materials and Methods: Sixty raw wastewater samples were collected from diverse urban sources and concentrated using polyethylene glycol 6000 to isolate BK virus. The BK virus isolates were analyzed for genotypes and the non-coding control region (NCCR). Results: The Nested PCR results indicated that thirty-two out of sixty samples (53.33%) were positive for the BK virus. The phylogenetic analysis revealed the dominance of BK virus genotype Ib2, followed by genotype 4. The BK virus non-coding control region (NCCR) analysis identified an archetype strain.
Conclusion:The sewage plant treatment should be implemented to remove pathogenic viruses specially BK virus and to curb the circulating of BK virus in human and environment.
## INTRODUCTION
The presence of BK virus (BKV) in wastewater is a significant indicator of contamination. BKV, released in the urine of infected individuals who are the carriers, enters the wastewater system and is ultimately discharged into rivers (1). Rivers constitute a primary source of potable water and can be contaminated with BK virus (2). The presence of the BK virus in drinking water leads to the establishment of the BK virus as a persistent and asymptomatic infection in various organs of the body during the early childhood developmental stage (3). Reactivation of latent BKV has been observed in patients with immunocompromised conditions associated with BKV infection, manifesting in ailments such as encephalitis, nephritis, hemorrhagic cystitis, retinitis, and pneumonia; this phenomenon has also been documented in individuals infected with HIV-1 (4). BK virus is characterized as a non-enveloped double-stranded DNA virus that falls under the classification of the Polyomaviridae family (5). The global seroprevalence rate of BK virus has been reported to be 80% (5). The dissemination of BK virus occurs via various mechanisms, encompassing the respiratory system, blood transfusions, seminal fluid, oral-fecal transmission, organ transplantation, as well as excretion through urine and feces (6). BK-PyV is classified into four genotypes, I-IV. Genotype I is the most prevalent worldwide (80%), followed by genotype IV (15%) (5,7). Subtype II and III viruses are rarely detected. Four subgroups of subtype I have been recognized (Ia, Ib-1, Ib-2, and Ic), and six subgroups of subtype IV have been recognized (IVa-1, IVa-2, IVb-1, IVb-2, IVc-1, and IVc-2) (7). Polyomaviruses (PyVs) exhibit significant oncogenic potential. BK Polyomavirus, JC Polyomavirus, and SV40 possess viral oncoproteins, recognized as large T (Tag) and small t (tag) antigens, which are implicated in cellular transformation and oncogenic processes. (6). BK virus DNA has been identified in various malignancies, including those of the urological system, prostate, and brain cancers (8).
Molecular characterization. Two distinct variants of BKV NCCR exist: the archetype NCCR variant, found in the urine of healthy individuals, and the second, a rearranged NCCR variant, found in infected tissue samples (1). The archetypal virus is postulated to represent the transmissible variant of the virus, as it has been identified in both asymptomatic individuals and those afflicted with disease, whereas rearranged variants are predominantly obtained from patients suffering from BKV pathology (1). The BKV archetype encompasses a 376 bp linear O, P, Q, R, S boxes, where "O" denotes replication and P, Q, R, S boxes play the promoters and regulatory domains of the early viral gene region (EVGR) and the late viral gene region (LVGR) (9). The "O" box comprises 142 bp, the "P" box 68 bp, the "Q" box 39 bp, the "R" box 63 bp, and the "S" box 63 bp. The "O" box harbors the origin of DNA repli-cation, and each box possesses a unique configuration of transcription factor binding sites (TFBS) (9). The viral genome is partitioned into three principal components: the early region, the late region, and the non-coding control region (NCCR). Alterations within the NCCR can facilitate the transition from the archetype strain to the rearranged strain, with NCCR rearrangements playing a crucial role in viral pathogenesis (10). The presence of BK virus in wastewater as an indicator of water pollution is well-established (1). The Karun river is one of the sources of the drinking water. This study aimed to detect BK virus genotypes and analyze the BK virus NCCR in wastewater from the inlet to the Karun River in Ahvaz city, Iran.
## MATERIALS AND METHODS
The Karun River is situated in the southwestern region of Iran, it extends approximately 850 km, originating from Zard-Kuh Bakhtiari in the Zagros Mountains and flowing into the international waters of the Persian Gulf. The river traverses the Ahvaz metropolitan zone and bifurcates the city into western and eastern segments. The quality and quantity of the river's water are experiencing a significant decline, primarily attributable to population growth, and the release of wastewater to the river. In this investigation, 60 raw wastewater samples were collected from different locations where untreated wastewater is directly discharged into the Karun River from March to August 2023. The Personal Protective Equipment (PPE), Goggles, Protective face mask or splash-proof face shield, Liquid-repellent coveralls, Waterproof gloves, Rubber boots were provided for the worker, since experts handling raw sewage may be exposed to infectious or oncogenic viruses.
Concentration of wastewater samples. Various techniques such as Polyethylene glycol (PEG), ultracentrifugation, electronegative membrane filtration, and ultrafiltration have been utilized for virus concentration in different water environments (11). The World Health Organization (WHO) has endorsed the PEG-based separation technique (12). The PEG method is favored for its efficiency, cost-effectiveness, reliability, and non-destructiveness to viruses (11), thus the Amdiouni method using PEG was chosen for this study (13).
http://ijm.tums.ac.ir
The concentration methodology was executed in accordance with the procedures delineated by Amdiouni (13). A total of 500 ml of wastewater sample was collected, put on the ice, and sent to the virology laboratory. Then the sample underwent clarification via centrifugation for a duration of 30 minutes at a gravitational force of 1000 g, at 4ºC, after which the pellet was reconstituted in 10 ml of the supernatant. The residual supernatant was retained for further analysis. Chloroform was introduced to the resuspended sample to achieve a concentration of 10%, followed by thorough mixing, and the resultant mixture was subjected to a secondary centrifugation for 5 minutes at 1000 g. The first and second supernatants were amalgamated, and the total volume was quantified. The consolidated supernatants were augmented with NaCl and polyethylene glycol (PEG) to reach final concentrations of 2.2% (w⁄w) NaCl (Sigma) and 7% (w⁄w) PEG 6000 (Fluka, Steinheim, Germany). The mixture was agitated at a temperature of 4°C overnight and subsequently centrifuged for 2 hours at 2000 g at the same temperature. The supernatant was discarded, and the pellet was reconstituted in phosphate buffer at a dilution of 1 ⁄ 100 of the initial volume and kept at -20ºC until DNA extraction.
DNA extraction. The deoxyribonucleic acid (DNA) was isolated from concentrated samples utilizing a viral nucleic acid extraction kit (High Pure Viral Nucleic Acid Kit, Roche, Germany) in accordance with the guidelines prescribed by the manufacturer. The concentration of the isolated viral DNA was assessed employing a Nano-Drop spectrophotometer produced by Thermo Fisher Scientific, United States. The ultimate concentration of the purified DNA derived from the 60 raw wastewater specimens exhibited a range between 10 and 75 ng/µl. The purified DNA samples were preserved at -20ºC to enable future analytical assessments.
## BK virus genotyping. Nested Polymerase Chain
Reaction (PCR) was utilized for identification of the BK virus VP1 region, employing an outer primer pair VP1-7/VP1-2R to amplify a fragment measuring 579 bp, subsequently followed by the application of the inner primer pair 327-1/2 for amplification of a 327 bp sequence ( 14) (Table 1). The genomic segment from 1630 to 1937 (327 bp) of BK viruses is characterized as hypervariable and serves as a critical determinant for the classification of BK virus genotypes (14). The PCR mixture encompassed 10 µL of a 2X master mix (Amplicon, Denmark), 1 µL (10 pmol/µL) of each primer (Table 1), 500 ng of the DNA template, and distilled water adjusted to a final volume of 25 µL. In each PCR assay, both negative and positive controls (the extracted DNA from the urine sample of positive for BK virus DNA) were incorporated, utilizing the following thermal cycling protocol: an initial denaturation at 94°C for 5 minutes, succeeded by 35 cycles consisting of denaturation at 94°C for 30 seconds, annealing at 59°C for 30 seconds, and extension at 72°C for 60 seconds, culminating in a final extension at 72°C for 10 minutes. The second amplification round was conducted utilizing 1 µL of the product obtained from the initial amplification, alongside 1 µL (10 pmol/µL) of each primer, and 10 µL of a 2x PCR master mix, with the volume adjusted to 25 µL using distilled water. The thermal cycling parameters implemented were as follows: an initial denaturation at 94°C for 5 minutes, followed by 32 cycles consisting of denaturation at 94°C for 30 seconds, annealing at 54°C for 30 seconds, and an extension phase at 72°C for 60 seconds. The resultant PCR product, measuring 327 base pairs, yielded a positive result (14).
Gel electrophoresis. Gel electrophoresis was performed utilizing a 2% agarose gel in conjunction with a 100-bp DNA ladder to facilitate the separation and detection of PCR products within a timeframe of 20 minutes. Positive samples were subsequently subjected to Sanger sequencing for further analysis. The duplicate PCR assays were repeated for the positive tests.
## Sequencing and phylogenic analyses of VP1.
For the purposes of verification, the sequencing outcomes pertaining to the partial "VP1" region of the BK virus genome from the isolated samples were systematically aligned utilizing the NCBI BK virus database (https://blast.ncbi.nlm.nih.gov). The isolated VP1 BK virus sequences were further aligned with the reference sequence of the VP1 BK virus through the application of SnapGene software (version 3.2.1). To ascertain the genotyping of the BK virus, a phylogenetic tree was constructed employing the Maximum Likelihood method for each isolated partial VP1 region of the BK virus genome, adhering to the Kimura 2-parameter distance model with 1000 boot- BK virus NCCR assessment. Nested Polymerase Chain Reaction (PCR) was utilized for identification of the BK virus NCCR region. The PCR mixture encompassed 10 µL of a 2X master mix (Amplicon, Denmark), 1 µL (10 pmol/µL) of each primer (table 1), 500 ng of the DNA template, and distilled water adjusted to a final volume of 25 µL. In each PCR assay, both negative and positive controls were incorporated, utilizing the following thermal cycling protocol: an initial denaturation at 94°C for 5 minutes, succeeded by 35 cycles consisting of denaturation at 94°C for 45 seconds, annealing at 53°C for 45 seconds, and extension at 72°C for 45 seconds, culminating in a final extension at 72°C for 10 minutes. The second amplification round was conducted utilizing 1 µL of the product obtained from the initial amplification, alongside 1 µL (10 pmol/µL) of each primer, and 10 µL of a 2x PCR master mix, with the volume adjusted to 25 µL using distilled water. The thermal cycling parameters implemented were as follows: an initial denaturation at 94°C for 5 minutes, followed by 35 cycles consisting of denaturation at 94°C for 45 seconds, annealing at 56°C for 45 seconds, and an extension phase at 72°C for 45 seconds, culminating in a final extension at 72°C for 10 minutes. The resultant PCR product, measuring 375 bp, yielded a positive result (15).
## Gel electrophoresis.
Gel electrophoresis was performed utilizing a 2% agarose gel in conjunction with a 100-bp DNA ladder to facilitate the separation and detection of PCR products within a timeframe of 20 minutes. Positive samples were subsequently subjected to Sanger sequencing for further analysis. The duplicate PCR assays were repeated for the positive tests. The sequences corresponding to the NCCR region of the BK virus, along with their respective accession numbers, were randomly selected from five wastewater samples and aligned with a consensus sequence of the NCCR region of the BK virus archetype.
## RESULTS
The identification of BK DNA was substantiated in 32 out of 60 (53.33%) wastewater specimens. A subset of 10 samples was randomly selected, comprising 5 samples that tested positive VP1 BK virus with accession numbers OR723817 -OR723821, and 5 samples that tested positive NCCR BK virus with accession numbers OR933745-OR933749, all of which were recorded in GenBank. The VP1 BK virus isolates OR723817 -OR723821 from Ahvaz, Iran, exhibited a phylogenetic clustering with the BK virus genotype 1b2 (HE650868) isolate from Kuwait, alongside (OR113380, OR113382, and OR113383) isolates derived from the colon specimens of patients in Ahvaz, Iran. Conversely, the BKV isolates obtained from Ahvaz, specifically OR723818 and OR723820, clustered with genotype IV (JX195569) isolated in Spain, as well as with HE650847 isolated in Kuwait, and OQ129438 and OR509396 isolated in Iran. The phylogenetic tree results are illustrated in Fig. 1.
The results of five BK virus NCCR isolates from Fig. 1. The phylogenetic tree was constructed utilizing the Maximum Likelihood Method for VP1 sequences derived from BK virus genome isolates obtained from wastewater samples. The BK Virus isolates from Ahvaz (OR723817, OR723819, OR723820) were found to cluster with the BK genotype 1b2, alongside the BK virus genotype 1b2 (HE650868) isolate from Kuwait, as well as the OR113380, OR113382, and OR113383 isolates from Ahvaz, Iran. Conversely, the BKV isolates from Ahvaz (OR723818, OR723821) were observed to cluster with genotype IV (JX195569), which was isolated in Spain, along with HE650847 isolated in Kuwait, and OQ129438 and OR509396 isolated in Iran. The Tamura-Nei model was utilized for the Maximum Likelihood method, incorporating 1000 bootstrap replicates for enhanced statistical reliability. The scale bar was established at 0.01.
wastewater samples revealed that all five NCCR isolates show the archetype with few nucleotides substitution in some blocks of NCCR, when compared with consensus BK virus archetype (AY628236) isolate from urine of a healthy control individual in USA (Fig. 2). The results of BK virus NCCR alignment exhibit 100% homology with the consensus BK virus archetype (AY628236) for wastewater samples isolates OR933745, OR933747, and OR933748, while wastewater samples isolate OR933746 and iso-late OR933749 show 99.57% and 97.85% homology with the BK virus archetype strain (AY628236), respectively. Of the 5 isolates archetype variants, the sequences and transcription factors of the three BK virus NCCR boxes, P, Q, R, S, (OR933745, OR933747 and OR933748) found to be 100% homology with the consensus BK virus archetype (AY628236). The nucleotide substitution at position C6G was observed in Q box of BKV NCCR isolate (OR933746). The R box of the isolate OR933749 demonstrates three nucleotide substitutions: G9A, A11C, and G46T. The first two substitutions were not found in the transcription factors, whereas G46T was observed in transcription factor NFI-3. Moreover, the S box of isolate OR933749 reveals two nucleotide substitutions: G19A, observed in transcription factor p53, and G25A, which was not found in the transcription factor region (Fig. 2).
## DISCUSSION
The surveillance of wastewater has emerged as an essential approach for tracking the propagation of viruses within populations, as well as for anticipating potential outbreaks of viral illnesses.
The monitoring of wastewater is especially crucial for viruses that cause subclinical infections (16), considering the examination of JCV and BKV in wastewater may serve as a valuable method for detecting human-derived fecal contamination (1).
The global seroprevalence rate of BK virus has been reported to be 80% (9). BKPyV can be classified into four predominant genotypes (BKPyV I, II, III, and IV) predicated upon neutralization assays and the distinct sequences within the principal BK-PyV capsid gene VP1 (14). Nonetheless, contemporary genotyping methodologies predominantly utilize VP1-and LTAG-sequences to delineate BKPyV subgroups Ia, Ib1, Ib2, Ic, II, III, IVa1, IVa2, IVb1, IVb2, IVc1, and IVc2 (11).
The prevalence of BKV DNA in the urine samples of healthy individuals were reported in different regions of the world. the frequency of BKV DNA in urine samples was in Ahvaz city, Iran (41.8%) (17), Brasilia (12.5%) (18), Tunisia (6%) (19). BKV has been mainly interlinked renal failure with BKV-associated nephropathy (BKVAN) in kidney-transplant recipients (8) and hemorrhagic cystitis (HC) (4) in hematopoietic stem cell transplant recipients (HSCTRs) (20,21).
## Fig. 2.
Illustrates that isolates OR933745, OR933747, and OR933748 exhibited 100% homology with the consensus BK virus archetype (AY628236) isolated in the USA. The Q box of isolate OR933746 presents a nucleotide substitution at position C6G; this mutation was not observed in the transcription factor region of the Q box. The R box of isolate OR933749 demonstrates three nucleotide substitutions: G9A, A11C, and G46T. The G46T substitution was observed in transcription factor NFI-3, while the others were not found in the transcription factors. The S box of isolate OR933749 reveals two nucleotide substitutions: G19A, which was observed in transcription factor p53, and G25A, which was not found in the transcription factor regions of the S box.
BKV has been correlated with kidney and bladder cancer (8), head and neck cancers (22), BK virus association in HIV-associated salivary gland disease (HIVSGD) (9), and colorectal cancer (23).
In the absence of effective treatments, management focuses on adjusting immunosuppressive therapies during viral reactivation or graft injury. Early detection of viral reactivation is vital to reduce the risk of allograft rejection, especially soon after kidney transplantation. Current screening relies on quantifying plasma BKPyV DNA via PCR. The International Consensus 2024 guidelines advocate for monthly BKPyV DNA screening in kidney transplant recipients for nine months, followed by every three months screenings until the second year (or third year for pediatric cases) (24).
In the present investigation, a high rate of BK virus DNA was identified in 53.33% of urban wastewater samples. The relative prevalence of BK virus across various wastewater samples reinforces its potential for transmission through the fecal-oral route (6). BK virus has been identified in urban wastewater across diverse geographical regions, including Egypt (Ahmed et al. 2019) (25), and Pakistan (Ijaz et al.
## 2023) (26).
It is remarkable that fecal matter and wastewater encompass a multitude of pharmaceuticals, industrial discharge, heavy metals, reagents, complex polysaccharides, lipids, proteins, metal ions, and RNases, which can inhibit PCR amplification by several mechanisms. These mechanisms include the inhibition of DNA polymerase activity, degradation or sequestration of target nucleic acids, fluorescent signaling, and chelation of essential metal ions necessary http://ijm.tums.ac.ir for amplification (27,28). We diluted each positive BKV sample (e.g., 1:10) to confirm that inhibition does not mask true positives (29,30).
The phylogenetic tree analyses revealed that the BK virus isolated from Ahvaz corresponds to the sequences, specifically OR723817, OR723819, and OR723820 were found to cluster with the BK virus genotype 1b2 isolate from Kuwait (HE650868), and the OR113380, OR113382, and OR113383 isolates from Ahvaz, Iran. Conversely, the BKV isolates from Ahvaz (OR723818, OR723821) were observed to cluster with genotype IV (JX195569), which was isolated in Spain, along with HE650847 isolated in Kuwait, and OQ129438 and OR509396 isolated in Iran. Within this geographical context, the BK genotype 1b2 was identified as the predominant strain, followed by genotype IV. It is noteworthy that both BK genotypes 1b2 and IV were identified in urine and tumor tissues of patients diagnosed with colorectal cancer within this region (23).
The findings of the BK virus NCCR alignment reveal that the wastewater sample isolates OR933745, OR933747, and OR933748 demonstrate complete homology (100%) with the consensus BK virus archetype (AY628236) isolated in USA, whereas the wastewater sample isolates OR933746 and OR933749 exhibit homologies of 99.57% and 97.85% with the BK virus archetype strain (AY628236), respectively. The nucleotide substitution at position C6G was observed in Q box of BKV NCCR isolate (OR933746). The R box of isolate OR933749 exhibits three nucleotide substitutions: G9A, A11C, and G46T. While G9A and A11C were not found in transcription factors, G46T was observed in transcription factor NFI-3. The S box of isolate OR933749 reveals two nucleotide substitutions: G19A, found in transcription factor p53, and G25A, which was not found in the transcription factor region. The transcription factor binding sites (TFBSs) affect viral replication; a mutation in a Sp1 site that can hinder binding and modify viral gene expression, enabling the viral strain to adopt features of a BKV strain with a rearranged NCCR (31,32). The O block harbors the origin of replication, yet various TFBSs, including Sp1, NF1, NF-kB, and Ets-1, are distributed throughout the NCCR and regulate BKV gene transcription (31,32). Six Nuclear Factor I (NFI) binding sites (NFI-1 to NFI-6), acting as viral enhancers, occur in sequences flanking the late side of the origin. Their mutation in BKV NCCRs boxes reduced BKV DNA (33). NFI-3 has been associated with the regulation of early-late transcription (33). The role of BK virus p53 within the NCCR's S box is influenced by interactions with other transcription factors, like Sp-1 and NF-1, which may modify its transactivation or repression capabilities in infected cells (34). Numerous studies have indicated that nucleotide substitutions or rearrangements within the BK virus NCCR box result in enhanced replication of the BK virus genome and play a pivotal role in the pathogenesis associated with BK virus (9). The activation of BK virus (BKV) is typically associated with the rearrangement of the Non-Coding Control Region (NCCR), wherein the deletion or insertion of Transcription Factor Binding Sites (TFBS) modulates the expression of Early Viral Gene Regulatory (EVGR) and Late Viral Gene Regulatory (LVGR) elements (33,34).
Cloning of 10 rr-NCCRs demonstrated various duplications or deletions that enhanced gene expression, replication capacity, and cytopathology of recombinant BKV in vitro (35). The BKV NCCRs from throat wash samples of individuals with HIV-associated systemic disease (HIVSGD) displayed a specific block arrangement termed "OPQPQQS" in immunosuppressed patients (9). The expression of Large T Ag (TAg) was found to inactivate P53 and retinoblastoma, contributing to oncogenesis (8). Additionally, BKV encodes two miRNAs, 5p-miRNA and 3p-miRNA, that are perfectly complementary to the mRNA of Large T Ag (Tag). BKPyV-infected bladder cancer cells exhibited enhanced proliferation alongside increased expression of miR-B1-3p and -5p (36).
The BK virus and JC virus are categorized as potentially carcinogenic infectious agents for humans (8). Indeed, the classified BKPyV and JCPyV as 2B are regarded as potential oncogenic viruses (37). Consequently, the eradication of BK virus should be prioritized as a primary consideration by environmental authorities in the context of wastewater treatment facilities.
Various methodologies, including skimmed milk flocculation, polyethylene glycol precipitation (PEG), glass wool filtration, ultrafiltration, virus adsorption, and elution have been documented for the concentration of viruses (11,12). We implemented techniques for viral concentration utilizing PEG, which is straightforward, economically viable, and effective, as previously reported (13).
## CONCLUSION
The results of the present investigation demonstrated that 32 (53.33%) wastewater specimens tested positive for the BK virus. The phylogenetic assessment reveals a predominance of BK virus genotype Ib2, followed by genotype 4. These findings pertaining to the BK virus non-coding control region (NCCR) indicated that three BK virus isolates were identified as archetype strains, whereas the Q box of isolate OR933746 exhibits a nucleotide substitution at position C6G; concurrently, the R box of isolate OR933749 displays multiple nucleotide substitutions at positions G9A, A11C, and G46T, while the S box of the same isolate OR933749 shows nucleotide substitutions at positions G19A and G25A, which cannot significantly contribute to the pathogenesis of the BK virus. Consequently, the eradication of BK virus should be prioritized as a primary consideration by environmental authorities in the context of wastewater treatment facilities.
Membrane technology is recognized for its efficacy in eliminating emerging viruses and antimicrobial-resistant genes from wastewater. Microfiltration is widely utilized in commercial membrane filtration to remove protozoa and bacteria (38). Sequential UV-chlorine disinfection enhances virus inactivation synergistically (39). Ultraviolet C radiation efficiently eradicates microorganisms within the 200-280 nm range. The UV disinfection mechanism involves the formation of pyrimidine dimers, which obstruct RNA and DNA replication. Chlorine disinfection operates by disrupting enzyme structures essential for bacterial and viral survival. It is notably effective in deactivating both bacteria and viruses (40). It is notable, the sewage plant treatment should be implemented to remove pathogenic infectious agents especially BK virus, and to restrain the circulation of BK virus in humans and the environment.
## References
1. Iran, Microbiol (2025)
2. Shaheen, Ahmed, Badr et al. (2024) "Detection and quantification of adenovirus, polyomavirus, and papillomavirus in urban sewage" *J Water Health*
3. Rashed, El-Senousy, Sayed (2022) "AlKhazindar M. Infectious pepper mild mottle virus and human adenoviruses as viral indices in sewage and water samples" *Food Environ Virol*
4. Ambalathingal, Francis, Smyth et al. (2017) "BK polyomavirus: clinical aspects, immune regulation, and emerging therapies" *Clin Microbiol Rev*
5. Alcendor (2019) "BK polyomavirus virus glomerular tropism: implications for virus reactivation from latency and amplification during immunosuppression" *J Clin Med*
6. Kotla, Kadambi, Hendricks et al. (2021) "BK polyomavirus-pathogen, paradigm and puzzle" *Nephrol Dial Transplant*
7. Krajewski, Kamińska, Poterek et al. (2020) "Pathogenicity of BK virus on the urinary system" *Cent European J Urol*
8. Morel, Martin, François et al. (2017) "A simple and reliable strategy for BK virus subtyping and subgrouping" *J Clin Microbiol*
9. Zhou, Zhu, Li (2023) "BK polyomavirus: latency, reactivation, diseases and tumorigenesis" *Front Cell Infect Microbiol*
10. Burger-Calderon, Ramsey, Hall et al. (2016) "Distinct BK polyomavirus non-coding control region (NCCR) variants in oral fluids of HIV-associated Salivary Gland Disease patients" *Virology*
11. Sahragard, Yaghobi, Mohammadi et al. (2024) "Impact of BK Polyomavirus NCCR variations in post kidney transplant outcomes" *Gene*
12. Lu, Huang, Luo et al. (2020) "Primary concentration-the critical step in implementing the wastewater-based epidemiology for the COVID-19 pandemic: a mini-review" *Sci Total Environ*
13. Who, In (2003) "Guidelines for Environmental Surveillance of Poliovirus Circulation"
14. Amdiouni, Faouzi, Fariat et al. (2012) "Detection and molecular identification of human adenoviruses and enteroviruses in wastewater from Morocco" *Lett Appl Microbiol*
15. Anzivino, Bellizzi, Mitterhofer et al. (2011) "Early monitoring of the human polyomavirus BK replication and sequencing analysis in a cohort of adult kidney transplant patients treated with basiliximab" *Virol J*
16. Egli, Infanti, Dumoulin et al. (2009) "Prevalence of polyomavirus BK and JC infection and replication in 400 healthy blood donors" *J Infect Dis*
17. Santos, Silva, Reis Teixeira et al. (2024) "Wastewater surveillance for viral pathogens: a tool for public health" *Heliyon*
18. Kaydani, Makvandi, Samarbafzadeh et al. (2015) "Prevalence and distribution of BK virus subtypes in renal transplant recipients referred to Golestan Hospital in Ahvaz, Iran" *Jundishapur J Microbiol*
19. Machado, Fink, Pannuti et al. (2011) "Human polyomaviruses JC and BK in the urine of Brazilian children and adolescents vertically infected by HIV" *Mem Inst Oswaldo Cruz*
20. Boukoum, Nahdi, Sahtout et al. (2016) "BK and JC virus infections in healthy patients compared to kidney transplant recipients in Tunisia" *Microb Pathog*
21. Laskin, Denburg, Furth et al. (2020) "The natural history of BK polyomavirus and the host immune response after stem cell transplantation" *Clin Infect Dis*
22. Peras, Bilić, Mareković (2025) "Recent Insights into the Pathogenesis, Diagnostics, and Treatment of BK Virus Infections in Children After Hematopoietic Stem Cell Transplantation" *Pathogens*
23. Mody, Rocco, Yom et al. (2021) "Head and neck cancer" *Lancet*
24. Javadi, Kaydani, Pirmoradi et al. (2025) "Frequency of BK virus genotypes in patients with colorectal cancer" *Iran J Microbiol*
25. Nourie, Boueri, Minh et al. (2024) "BK Polyomavirus Infection in Kidney Transplantation: A Comprehensive Review of Current Challenges and Future Directions" *Int J Mol Sci*
26. Ahmed, Elmahdy, Allayh et al. (2019) "Prevalence of human polyomavirus and papillomavirus in wastewater and in stool of Egyptian patients" *Egypt J Aquat Biol Fish*
27. Ijaz, Shahzad, Ul Haque "Detection of BK and JC polyomaviruses in sewage water of the urban areas of Lahore" *Biologia (Bratisl)*
28. Kumblathan, Liu, Uppal et al. (2021) "Wastewater-Based Epidemiology for Community Monitoring of SARS-CoV-2: Progress and Challenges" *ACS Environ Au*
29. Sims, Kasprzyk-Hordern (2020) "Future perspectives of wastewater-based epidemiology: Monitoring infectious disease spread and resistance to the community level" *Environ Int*
30. Yan, Toohey-Kurth, Crossley et al. (2020) "Inhibition monitoring in veterinary molecular testing" *J Vet Diagn Invest*
31. Schrader, Schielke, Ellerbroek et al. (2012) "PCR inhibitors -occurrence, properties and removal" *J Appl Microbiol*
32. Bethge, Ajuh, Hirsch (2016) "Imperfect symmetry of Sp1 and core promoter sequences regulates early and late virus gene expression of the bidirectional BK polyomavirus noncoding control region" *J Virol*
33. Walder, Odegard, Meeds et al. (2025) "Single-nucleotide polymorphisms within the BK polyomavirus non-coding control region are genotype-associated" *Microbiol Spectr*
34. Liang, Tikhanovich, Nasheuer et al. (2012) "Stimulation of BK virus DNA replication by NFI family transcription factors" *J Virol*
35. Bethge, Hachemi, Manzetti et al. (2015) "Sp1 sites in the noncoding control region of BK polyomavirus are key regulators of bidirectional viral early and late gene expression" *J Virol*
36. Gosert, Rinaldo, Funk et al. (2008) "Polyomavirus BK with rearranged noncoding control region emerge in vivo in renal transplant patients and increase viral replication and cytopathology" *J Exp Med*
37. Li, Chen, Wang et al. (2025) "Suppression of cisplatin induced ATF3 expression and apoptosis by BK polyomavirus and its encoded mi-croRNA in bladder cancer cells" *Biomed Pharmacother*
38. Galati, Chiantore, Marinaro et al.
39. (2024) "man oncogenic viruses: characteristics and prevention Strategies-Lessons learned from human papillomaviruses" *J. MICROBIOL*
40. Wang, Xiao, Furusho et al. (2025) "Enhancing Adsorptive Removal of Viruses and Extracellular Antimicrobial Resistance Genes in Microfiltration: Impacts of Membrane Materials and Solution Chemistry" *ACS ES T Water*
41. Guerrero-Latorre, Gonzales-Gustavson, Hundesa et al. (2016) "UV disinfection and flocculation-chlorination sachets to reduce hepatitis E virus in drinking water" *Int J Hyg Environ Health*
42. Shi, Chen, Liu et al. (2021) "Efficient synergistic disinfection by ozone, ultraviolet irradiation and chlorine in secondary effluents" *Sci Total Environ* |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.