Coxsackievirus B infections are common in Cystic Fibrosis and experimental evidence supports protection by vaccination

Summary Viral respiratory tract infections exacerbate airway disease and facilitate life-threatening bacterial colonization in cystic fibrosis (CF). Annual influenza vaccination is recommended and vaccines against other common respiratory viruses may further reduce pulmonary morbidity risk. Enteroviruses have been found in nasopharyngeal samples from CF patients experiencing pulmonary exacerbations. Using serology tests, we found that infections by a group of enteroviruses, Coxsackievirus Bs (CVBs), are prevalent in CF. We next showed that a CVB vaccine, currently undergoing clinical development, prevents infection and CVB-instigated lung damage in a murine model of CF. Finally, we demonstrate that individuals with CF have normal vaccine responses to a similar, commonly used enterovirus vaccine (inactivated poliovirus vaccine). Our study demonstrates that CVB infections are common in CF and provides experimental evidence indicating that CVB vaccines could be efficacious in the CF population. The role of CVB infections in contributing to pulmonary exacerbations in CF should be further studied.


INTRODUCTION
Cystic fibrosis (CF) is caused by mutations in the cystic fibrosis transmembrane regulator (CFTR) gene that encodes a chloride channel. Of 2,000 + CFTR mutations identified, the most common is F508del (the deletion of phenylalanine at position 508). CFTR is expressed in most organs with a secretory epithelium including the sinuses, respiratory system, pancreas, and reproductive organs, as well as immune cells (Elborn, 2016;Marson et al., 2016).
Acute and chronic airway infections are the major cause of mortality and morbidity in CF. Most patients are intermittently or chronically colonized by pathogenic airway bacteria that cause chronic airway inflammation and successive decline in lung function (Elborn, 2016). Why individuals with CF suffer from these infections remains unknown. Numerous observations suggest respiratory virus infections are a major cause of pulmonary exacerbations in chronic lung disease and predispose the CF airway to bacterial colonization (Asner et al., 2012;de Almeida et al., 2010;Goffard et al., 2014;Johansen and Hoiby, 1992;Petersen et al., 1981;Singanayagam et al., 2012;Wat et al., 2008). Such exacerbations lead to progressive worsening of CF lung function and culminate in terminal lung disease, the leading cause of death in CF.
Viral pathogens commonly associated with pulmonary exacerbations in CF include enteroviruses, influenza, and respiratory syncytial virus (Asner et al., 2012;de Almeida et al., 2010;Eymery et al., 2019). The frequency and seasonality of respiratory virus infections in CF populations do not seemingly differ from healthy individuals; however, the symptom duration is longer in children with CF (Dijkema et al., 2016;van Ewijk et al., 2008). Individuals with virus-associated pulmonary exacerbations do, however, require prolonged antibiotic treatment and they respond inadequately to standard therapeutic interventions (Etherington et al., 2014). Therefore, preventing common viral infections is likely to be of significant benefit to lung function and survival in CF.
Alterations in the innate and adaptive arms of the immune system have been described in CF, which may affect vaccination responses (Giacalone et al., 2020;Lara-Reyna et al., 2020). Divergent humoral responses have been reported including occasional hypogammaglobulinemia (Matthews et al., 1980), impaired antibody responses to polysaccharide antigens (Browning et al., 2014;Moss et al., 1987), and IgG subclass deficiency (Browning et al., 2014;Garside et al., 2005Garside et al., , 2007. We have also reported that mice harboring the F508del mutation have an impaired immune response to CVB3 infection, with defective viral clearance linked to a delay in virus-neutralizing antibody (nAB) production (Svedin et al., 2017). Despite these observations, few studies have been conducted assessing vaccine responses and virus nAB durability in CF (Browning et al., 2014;Hostetler et al., 2021;Launay et al., 2014;Lucidi et al., 1996).
In this study, we examined the prevalence of CVB infections in CF and compared this with a cohort of healthy controls. Our findings suggest that CVB infections are common in CF. We then performed proof-of-concept studies in a pre-clinical CF model examining protection against CVB infection and CVB-induced tissue pathology after vaccination with a CVB vaccine. Finally, we assessed immunity against poliovirus in CF and healthy control cohorts to gain insight into nAB responses to a similar, commonly used enterovirus vaccine.

CVB infections are frequent in CF
To assess whether CVB infections are common in CF and compare their frequency with a healthy population we used serum samples collected from individuals that attended the Stockholm CF Center, Stockholm, Sweden (Table 1). The control group consisted of healthy volunteers who participated in separate vaccine studies at the Karolinska Institute, Stockholm, Sweden (Table 1). Individuals with CF (n = 65) were on average younger than the healthy controls (n = 33) (p < 0.05), but the female-to-male ratio was not significantly different. All individuals with CF had a classic CF phenotype (defined as one or more clinical phenotype characteristic(s) and a sweat chloride concentration of >60 mmol/L) and 60% were homozygous for the F508del mutation.
Analysis of serum samples for CVB1-6 nABs revealed the serotypes participants were previously infected by. Most individuals with CF (89%) and all healthy controls (100%) were seropositive for at least one CVB serotype. On average, individuals with CF had encountered 2.1 G 1.5 (2) (mean G SD (median)) CVB serotypes compared with 2.8 G 1.4 (3) for healthy individuals. CVB5 was the most common serotype iScience Article in individuals with CF, whilst CVB2 and CVB5 had the highest frequencies in the healthy controls (Figures 1 and S1). As the individuals with CF were on average younger than the healthy control group (Table 1) and the latter consisted solely of individuals aged 18 or above, we next grouped the individuals with CF into those aged below 18 years (n = 30, 12.2 G 3.6 years) and those 18 years old or above (n = 35; 31.4 G 8.4 years). The average number of CVB serotypes that these individuals had been exposed to were (mean G SD) for age group <18 years, 1.8 G 1.4, and for age group R18 years, 2.3 G 1.5. These results suggest that CVB infections are common in CF with a similar prevalence to healthy subjects.

CVB vaccines induce virus nABs in an experimental mouse model of CF
Vaccination against CVB viruses could constitute an attractive complementary therapy to reduce the respiratory virus infection burden in CF, therefore we next explored the feasibility of this concept. Various CF animal models exist (Semaniakou et al., 2018), including the Cftr tm1EUR mouse model that is on a C57Bl/6J background and harbors the most common mutation, F508del (hereon referred to as F508del mice) (van Doorninck et al., 1995). Previously we performed side-by-side comparisons of infected wild-type and F508del mice and reported that F508del mice have a delayed nAB response to live CVB3 virus, which was linked in part to a defective antibody response to T cell-dependent antigens (Svedin et al., 2017). Therefore, we investigated whether the nAB response to CVB vaccination is T cell-dependent using mice lacking TCRab T-cells (TCRa knock-out, ko, representative plot Figure S2) and their wild-type (wt) counterparts. Mice were vaccinated twice (days 0 and 14) with a monovalent CVB3 vaccine (field isolate strain, ; Figure 2A). The CVB3 vaccine was well tolerated ( Figure S3A, data not shown) and from day 14, nAB titers were significantly lower in TCRa-ko mice compared with the wt animals (p < 0.001; Figures 2B and 2C) indicating that nAB responses to a CVB vaccine are in part T cell-dependent.
We next tested whether F508del mice produce virus nABs in response to CVB vaccination. A two-dose vaccination schedule (days 0 and 14; Figure 2D) was employed and F508del and wild-type littermate control mice (wt; C57Bl/6J) were monitored until day 28 after the initial vaccination. Most mice tolerated the CVB3 vaccine well as indicated by good health status scores throughout the study (data not shown). However, after the first vaccination, two F508del mice were removed after losing more weight than was allowed by our ethical permit (>10%; data not shown; these animals were excluded from the analysis in Figure 2). We have not observed such weight loss in our previous studies where we vaccinated mice with varying genetic backgrounds and non-human primates (Hankaniemi et al., 2017;Larsson et al., 2015;Stone et al., 2018Stone et al., , 2020. Occasional deaths have been reported by others that use F508del mice and the expected survival to maturity with this model is 90%. Similar or lower survival rates are seen in other murine models of CF (Semaniakou et al., 2018). Based on this we do not believe that the weight loss of the two described animals was a vaccine-related event.
No further adverse consequences on weight were recorded for mice in the study ( Figure S3B). Wt and F508del mice raised CVB3 nABs after the first vaccination with no statistical differences in nAB titers on days 4 and 5 ( Figures 2E and 2F). By day 14, F508del mice had lower CVB3 nAB titers compared with wt mice although the difference was not statistically significant ( Figures 2E and 2F). After the day 14 boost vaccination, CVB3 nAB titers were equivalent between wt and F508del mice (Figures 2E and 2F). To confirm these studies, F508del mice were vaccinated with another CVB3 vaccine that was produced in an identical manner to the initial vaccine but utilized a different CVB3 virus strain (the Nancy strain; Figures 3G and 3H).  . CVB1-6 seropositivity in CF and healthy control cohorts Serum was extracted from blood-samples collected at yearly check-up appointments in the CF cohort (n = 65) or from samples taken from healthy individuals (n = 33) included in other studies at the Karolinska Institute. CVB1-6 seropositivity was examined by measuring the presence of neutralizing antibodies against the CVB1-6 viruses, using a standard plaque reduction assay. The cut-off titer for seropositivity was set to R1:16 for all six serotypes. The pie charts show the fraction of individuals that were positive for each CVB serotype in the two groups (blue color) and the white numbers show the exact percentage of positive cases in each group. See also Figure S1. We have previously shown that CVB vaccines protect C57Bl/6J mice and other mouse strains from CVB infections (Stone et al., 2018(Stone et al., , 2021. Next, we examined whether CVB vaccines also protect F508del mice against CVB infections. A CVB1 vaccine was also included and the vaccination strategy optimized to a three-dose vaccination strategy to increase the duration of the virus nAB response ( Figure 3A; days 0, 21 and 35) (Hankaniemi et al., 2017;Larsson et al., 2015;Stone et al., 2018Stone et al., , 2020 4B). Infection also caused viraemia (replicating virus in the blood) on days 3 and 4 p.i. in 83% (5/6) and 100% (4/4) of the control (untreated or mock-vaccinated) F508del mice infected with CVB1 or CVB3, respectively ( Figures 4C and 4D). Both vaccines completely prevented systemic infections at the same time points p.i. in all CVB1 (6/6) and CVB3-(8/8) vaccinated animals infected with their respective CVBs ( Figure 4D). Viral dissemination to peripheral organs was also assessed on day 4 p.i.. Most unvaccinated F508del mice infected with CVB1 (5/6; 83%; Figure

CVB vaccines prevent CVB-induced damage to lungs and other organs in F508del mice
Organs were histologically assessed for signs of virus-mediated damage. The pancreas is one of the first organs to incur CVB-mediated damage (e.g. Flodstrom et al., 2001). Exocrine tissue damage and pancreatitis (depicted as immune cell infiltration) were present in 83 and 75% of the CVB1-and CVB3-infected control mice, respectively. Contrastingly, all pancreas collected from vaccinated-infected mice were healthy in appearance ( Figure 5A, representative images Figures 5E-5H, serotype breakdown Figures S5A and S5E). Female and male F508del mice were left untreated (n = 1, CVB3 study; black circles), mock-vaccinated with vaccine buffer (n = 6, CVB1 study; n = 3 CVB3 study; black circles), or vaccinated with CVB1 vaccine (A, C, and E; n = 6; light teal triangles) or CVB3 vaccine (B, D, and F; n = 8; dark teal triangles) as depicted in the schematic in Figure 3A. On day 63 after the prime vaccination, the animals were infected with CVB1 (A, C, and E) or CVB3 (B, D, and F) and blood was collected on days 3 and 4 post-infection (p.i.  5I, and S5D). Deposits were found in the extracellular space and within macrophage-like cells in at least half of the unvaccinated animals but were rarely seen in vaccinated mice, with the former having a significantly higher mean hemosiderin deposit score after CVB iScience Article infection (p < 0.04; Figure 5D). Most hemosiderin deposits were found in CVB1-infected animals rather than those challenged with CVB3 ( Figures S5D and S5H).
Liver morphology also differed between the vaccinated and control groups after infection. Ballooning degeneration (Flodstrom et al., 2001) was seen in 50% of the control CVB3-infected animals ( Figure S6C) but was absent in vaccinated mice ( Figure S6D; data not shown). Furthermore, mild steatosis was detected in 83 and 75% of control mice infected with CVB1 and CVB3 respectively but was absent in the majority of vaccinated-infected mice (1/8 CVB3 vaccine + CVB3-challenge and 0/6 CVB1 vaccine + CVB1-challenge; representative image Figure S6A). Histological assessment of spleens did not reveal any noticeable morphological differences between the groups (data not shown).
Together, these results demonstrate that CVB vaccines protect F508del mice from acute CVB infections, viral dissemination to organs, and virus-mediated tissue damage, including in the lung.

Most individuals with CF respond to poliovirus vaccination
CF is associated with altered immune functions (Giacalone et al., 2020;Lara-Reyna et al., 2020) and reports exist describing weakened vaccine responses in individuals with CF (Browning et al., 2014;Hostetler et al., 2021;Launay et al., 2014;Lucidi et al., 1996). A weak response to the CVB vaccine could limit the efficacy of future vaccine efforts in this group. One of the few enterovirus vaccines available, the inactivated poliovirus vaccine (IPV), is commonly used world-wide including in the Swedish childhood vaccination program. As poliovirus infections are most common in childhood, the vaccine is given early in life (one vaccine dose at each of the following ages: 3, 5, and 12 months), and after immunization, most children develop life-long protective virus nABs. IPV is based on inactivated whole virus and is produced in the same way as the newly developed CVB vaccines (Hankaniemi et al., 2017;Stone et al., 2018Stone et al., , 2020. As such, we hypothesized that the measurement of nAB titers to poliovirus 1 and 3 in our Swedish CF patient cohort and healthy controls would provide insight into how well individuals with CF respond to similar enterovirus vaccines. Polio nAB titers were measured in the same cohorts used to assess CVB infection frequency. Titers of polio 1 and 3 nABs that correlated with protective immunity (seropositivity at a R1:8 dilution) were detected in most individuals with CF and healthy controls ; Figures 6A and 6B). Individuals with CF tended to have lower nAB titers against both poliovirus serotypes but this was not statistically significant ( Figures 6A and 6B). A small fraction of individuals with CF had titers below 1:8 (poliovirus 1: 2/65; poliovirus 3: 3/65). These data show that most individuals with CF respond well to a formalin-inactivated enterovirus vaccine.   Goffard et al., 2014;Hamed et al., 2022;Ong et al., 1989;van Ewijk et al., 2008). Moreover, a recent cross-sectional observational study showed a significant relationship between sputum-positive bacterial culture and enterovirus positivity. A positive test for enterovirus or influenza virus but not rhino-, adeno-, Boca, or human metapneumovirus, was associated with a high risk of hospital admission and need for oxygen therapy (Hamed et al., 2022). These observations suggest that vaccination against enteroviruses would be a viable strategy to lower the burden of respiratory infections in CF.

DISCUSSION
The present study shows that infections by a group of enteroviruses, the CVBs, are common in CF. Little was known about CVB infection frequencies in individuals with CF and therefore serological assays were used to examine this using serum samples from individuals attending the Stockholm CF Center and healthy controls from the Stockholm region. The analyses revealed that CVB infections did not occur more frequently in CF patients compared with healthy individuals but rather we noted a slightly lower CVB seroprevalence in those with CF. This could be because those with CF often employ extra precautions to avoid respiratory infections. Alternatively, individuals with CF may have a weaker nAB response to CVB infection, causing some infections to be undetectable in this cohort. Nevertheless, in our study individuals had on average encountered at least two of the six CVB serotypes, suggesting that CVB infections are common in CF.

A CVB vaccine is currently in phase I clinical trials (2021). A superior method to determine vaccine efficacy and provide evidence of vaccine functionality involves examining protection against infection in vivo.
Studies were therefore undertaken to assess the efficacy of CVB vaccines in a pre-clinical model of CF. CVB vaccination of F508del mice protected against virus-induced weight loss, viremia, and virus dissemination to peripheral organs. Moreover, histological tissue analyses revealed vaccine-mediated protection against CVB-induced pancreas and liver tissue damage. Most importantly, analysis of lung sections showed that vaccinated F508del mice had less damage compared with control F508del animals, particularly when assessing general health and assessing hemosiderin deposits that were more prevalent in control animals. CVB infections are in general associated with mild upper respiratory tract infections and only few studies have documented pulmonary involvement in humans and animal models following infection (e.g. Jahn et al., 1964;Wang et al., 2014). The mechanisms behind the pathophysiology seen in the lungs of the CVB1-infected F508del mice could be investigated in future studies with an aim to understand how mutations in CFTR heighten the risk of pulmonary disease. Collectively, our current data showed that vaccination protected F508del mice from virus-induced lung damage. Alterations in immune system functions (reviewed in Giacalone et al., 2020;Lara-Reyna et al., 2020) and impaired vaccine responses (Browning et al., 2014;Hostetler et al., 2021;Launay et al., 2014;Lucidi et al., 1996) have been described in CF. Measuring nAB titers against polioviruses provides a reasonable surrogate to determine whether individuals with CF have an impaired immune response to enterovirus vaccines. In our study most individuals with CF had a strong, persistent nAB response to poliovirus vaccination (protective titers R1:8), however, a few individuals had low nAB titers (<1:8) to poliovirus 1 (2/65) and poliovirus 3 (3/65). This is consistent with the one existing study that examined immunity to the oral poliovirus vaccine in CF (Lucidi et al., 1996). Notably, it was observed that children >8 years old had weaker nAB responses to poliovirus 2, although titers always reached protective levels (Lucidi et al., 1996). Furthermore, another recent study described low or undetectable immunity to measles, mumps, and varicella zoster in a subgroup of individuals with CF (Hostetler et al., 2021). As CVB infections are common in childhood, immunizations with CVB vaccines would likely follow the same vaccination schedule as IPV (i.e. given in the first year of life). Our study shows that F508del mice generate similar CVB nAB titers to wt mice after a booster CVB vaccine dose, despite an initial weaker response, and that most individuals with CF responded adequately to IPV in early life. Taken together, one can hypothesize that the majority of individuals with CF should respond well to a CVB vaccine but tests confirming immunization-induced immunity should be routinely performed in this group.
Based on strong pre-clinical data including robust nAB responses in non-human primates , the development of a polyvalent formalin-inactivated CVB vaccine for clinical use has recently commenced. Phase I clinical trials began in December 2020 testing the safety and efficacy of this vaccine in humans (NCT04690426). Data indicates that the vaccine met primary safety endpoints and secondary ll OPEN ACCESS iScience 25, 105070, October 21, 2022 iScience Article efficacy endpoints (Provention Bio, 2021). If approved, this vaccine will initially be tested for the prevention of acute CVB infections and the delay or prevention of type 1 diabetes and celiac disease. Given that CVBs cause respiratory illness, the vaccine could also be relevant in CF populations.

Limitations of the study
Finally, we acknowledge some limitations of our studies. First, we only examined the prevalence of CVB infections in a Swedish CF cohort and our observations must be confirmed in other CF populations. However, given the global prevalence of CVBs, one can assume that CVB infections are common in other CF populations. Secondly, we did not have access to childhood poliovirus vaccination records for the study participants, but during the last few decades in Sweden, poliovirus vaccine coverage has been around 97-98% therefore it can be assumed that most individuals were vaccinated in childhood. Whether individuals received a booster dose is unknown. It remains reassuring that most individuals had virus nAB titers that correlate with protection against infection. Thirdly, our proof-of-concept studies were limited to examining one CF-causing mutation, F508del, and our study involved an insufficient number of individuals to address whether different CFTR mutations affect poliovirus nAB responses. Fourthly, our pre-clinical testing of the CVB vaccines consisted of vaccinations with monovalent vaccines rather than a hexavalent CVB1-6 vaccine. In previous studies, we have found that mice respond equally well to monovalent CVB vaccines as to the polyvalent CVB vaccine Stone et al., 2018Stone et al., , 2020. It is therefore likely that a multivalent vaccine would also be efficacious in the F508del model.

Conclusions
In summary, this study shows that CVB infections are common in the CF population and we provide preclinical evidence that CVB infections are preventable through vaccination. As respiratory virus infections are strongly linked to exacerbated pulmonary morbidity in CF and vaccines preventing such infections are advocated in CF, we propose that the role of CVB infections in pulmonary exacerbations should be further studied. If additional evidence supports the role for CVB infections in facilitating bacterial colonization and/or accelerating pulmonary morbidity in CF, the newly developed CVB vaccine could be a candidate for future prophylactic intervention in this disease.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: iScience Article ACKNOWLEDGMENTS We would like to thank the research nurses at the Stockholm CF Center, staff at the pre-clinical laboratory (PKL) at the Karolinska Institute, and staff at the Faculty of Medicine and Health Technology (University of Tampere) for assistance. We also thank Dr. B. Scholte, University of Rotterdam, The Netherlands, for sharing the Cftr tm1EUR mouse model. Drs K. Blom, J. Michaelsson, and J.K. Sandberg. Karolinska Institutet are greatly acknowledged for sharing serum samples from healthy control individuals. We also thank S. Parvin, Karolinska Institute, for her technical help with processing and staining histological samples and the members of the Flodströ m-Tullberg group for discussion and critical feedback on the study. We acknowledge Biocenter Finland and the Karolinska University Hospital for infrastructure support. Finally, we would like to thank the healthy controls and individuals with CF involved in the study for their participation. The graphical abstract was created using BioRender.com. This work has been supported by grants awarded by The Swedish Heart-Lung Foundation (  iScience Article with the male animals; subsequently embryos were collected and implanted into foster mothers to allow for embryo rederivation ensuring a specific pathogen-free state. The resulting pups were genotyped using the following primers (5 0 to 3 0 ): mutant reverse -CTA CCC GCT TCC ATT GCT C; common -TGA CTC CCA AAT CAA TGT GC; wild type reverse GGT GAG ATG ACC CAA AGC AG according to the Jackson Laboratories protocol. Mice were bred as homozygotes (e.g. wt males with wt females or knock-out males with knock-out females) and genotyped as before. Female ko mice and wt controls between 9 and 20 weeks of age were included in the experiment.
The Cftr tm1EUR mouse model (Rotterdam, the Netherlands), carrying the homozygous F508del mutation in the Cftr gene on a C57Bl/6J background was used (van Doorninck et al., 1995). Female and male transgenic (F508del) mice and their wt littermate controls aged between 4 and 22 weeks were included in the experiments. Mice were embryo-derived before being transferred into the animal facility, bred as heterozygotes and genotyped as previously described (van Doorninck et al., 1995).
The animal studies were approved by the Stockholm Southern Animal Ethics Board. The studies were performed in accordance with national and institutional guidelines.

Human subjects
Serum from a Swedish cohort of individuals with CF and from healthy controls was collected for the analysis of neutralizing antibodies against the six known CVB serotypes and poliovirus 1 and 3. All individuals with CF (n = 65) were monitored at Stockholm CF Center, serum dating between 1992 and 2010 was obtained from the clinic's biobank. Serum from healthy individuals from the Stockholm region (n = 33) was obtained from two unrelated vaccination studies conducted between 2008 and 2016. Table 1 contains the characteristics (age, gender and CF genotype) of the CF and control patient cohorts. All healthy donors and individuals with CF gave informed consent prior to participation and experiments were conducted according to the Declaration of Helsinki. The studies were approved by the regional ethical review board in Stockholm, Sweden and were performed in accordance with national and institutional guidelines.

Microbe strains
For vaccine production, CVB3 (wild-type strain from Finland; ), CVB3 Nancy (kindly provided by Dr. G. Frisk, Uppsala University) and CVB1-10796 (wild -type strain from Argentina ) were propagated in Vero cells (National Institute for Health and Welfare, Finland) and recovered from supernatant as described in Hankaniemi et al. (2019). In infection studies, animals were challenged with CVB1-10796 or 10 5 PFU CVB3-Nancy (both propagated in HeLa cells).

Flow cytometry analysis
Flow cytometry analysis was performed to assess the presence of T-and B-cells in splenocytes from the wt and TCRa knock-out mice. Briefly, spleens were harvested, homogenised and then the homogenate was collected and pelleted. Red blood cells in the pellet were lysed and the cells were washed, then resuspended in FACS buffer (PBS +2% fetal bovine serum + 2mM EDTA) and counted. Cells (10 6 cell/ml) were stained with anti-CD4, anti-TCRb or anti-B220 antibodies (all from Biolegend; 1:400 dilution) and then analysed by flow cytometry using an Accuri Flow Cytometer (BD). 10,000 events were assessed per sample and the data was analysed using FlowJo software.

Neutralizing antibody measurements
Measurements of serum neutralizing antibodies towards the six CVB serotypes and polio 1 and 3 in human serum samples and CVB1 in mouse samples were performed at Tampere University, Finland, using a standard plaque reduction assay