Molecular epidemiology of enteroviruses in young children at increased risk of type 1 diabetes

RESEARCH ARTICLE Molecular epidemiology of enteroviruses in young children at increased risk of type 1 diabetes Amir-Babak Sioofy-Khojine 1 *, Sami Oikarinen 1,2 , Hanna Honkanen 1 , Heini Huhtala 3 , Jussi P. Lehtonen 1 , Thomas Briese 4 , Heikki Hyo ¨ty 1,2 , on behalf of the TEDDY Study Group ¶ 1Department of Virology, Faculty of Medicine and Life Sciences, University of Tampere, Tampere, Finland, 2Fimlab Laboratories, Pirkanmaa Hospital District, Tampere, Finland, 3Faculty of Social Sciences, University of Tampere, Tampere, Finland, 4Center for Infection and Immunity, Mailman School of Public Health, Columbia University, New York, New York ¶ The complete membership of the TEDDY Study Group can be found in the Acknowledgments. *Amirbabak.Sioofy.K[email protected] Abstract Background Young children are susceptible to enterovirus (EV) infections, which cause significant morbidity in this age group. However, the current knowledge regarding the epidemiology of EVs and the circulating virus strains is mostly based on viruses detected in children with severe diseases leading to contact with the health care system, while the vast reservoir of EVs that circulate in the general population is less characterized. Methodology The present study investigates the types and the prevalence of EVs circulating in the young children of the background population in Georgia, Colorado, and Washington State in the USA, and Germany, Sweden, and Finland in Europe. A total of 4018 stool samples, collected monthly from 300 healthy and non-hospitalized children at the age of 3–18 months in 2005–2009, were analyzed for the presence of EVs using RT-PCR, followed by sequencing of the VP1-2A region of the viral genome to type the EV(s) present. All of the children carried type HLA-DQ2 or -DQ8 alleles associated with type 1 diabetes. Principal findings Altogether 201 children (67%) were found to be EV positive. The prevalence was much lower in Finnish children (26%) than in the children of the other counties combined (75%). Infections increased by age and showed a nadir during the winter months. Children who carried both the HLA-DQ2 and -DQ8 alleles had less infections than children who were homozygous for these alleles. Coxsackieviruses type A were the most frequently detected viruses in all geographical regions. Coxsackievirus type A4, Echovirus type 18, and Echovirus type 25 were shed for longer time periods than the other EV types. PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 1 / 18 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Sioofy-Khojine A-B, Oikarinen S, Honkanen H, Huhtala H, Lehtonen JP, Briese T, et al. (2018) Molecular epidemiology of enteroviruses in young children at increased risk of type 1 diabetes. PLoS ONE 13(9): e0201959. https://doi. org/10.1371/journal.pone.0201959 Editor: Zheng Xing, University of Minnesota College of Veterinary Medicine, UNITED STATES Received: February 6, 2018 Accepted: July 25, 2018 Published: September 7, 2018 Copyright: ©2018 Sioofy-Khojine et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: The datasets generated and analyzed during the current study have been made available in the NIDDK Central Repository at https://www.niddkrepository.org/ studies/teddy. Due to ethical commitments based on the TEDDY study’s informed consent, deidentified datasets will be available by request from the NIDDK Central Repository in accordance with the NIDDK’s controlled-access authorization process. Conclusions Compared to prevalence data from symptomatic patients requiring medical attention, this study provides a better view of EVs circulating in young children in the USA and in Europe. The observations may prove useful for the selection of strategies for designing EV vaccines in the future. The study also confirms our previous serological findings suggesting that EV infections are relatively rare in Finland. 1. Background Enteroviruses (EVs) are among the most common human viruses infecting nearly one billion people annually worldwide [1]. Even though most EV infections are asymptomatic or mild, EVs are responsible for a vast number of severe infections every year, including paralysis, encephalitis, aseptic meningitis, pleurodynia, pericarditis, myocarditis, hand-foot-and-mouth disease (HFMD), herpangina, severe neonatal systemic disease, and acute haemorrhagic conjunctivitis [2]. Young infants are susceptible to severe EV diseases, possibly due to their immature immune system. EVs belong to the family Picornaviridae consisting of 15 species including Enterovirus A-L, and Rhinovirus A-C [3]. The EV species A, B, C and D infect humans, and they are traditionally divided into Poliovirus (PV) consisting of three serotypes, Coxsackievirus type A (CV-A) with 23 serotypes, Coxsackievirus type B (CV-B) with 6 serotypes, and Echovirus (E) consisting of 28 serotypes [4], as well as some more recently discovered numbered serotypes. EVs show both endemic and epidemic patterns [5]. Epidemics are typically caused by certain EV types such as E-30 (causing aseptic meningitis), EV-D70 (causing acute hemorrhagic conjunctivitis), EV-A71 (causing severe central nervous system disease in the Far East), CV-A6 and CV-A10 (causing severe HFMD), and EV-D68 (causing flu-like symptoms) [1,6]. Recent advances in the molecular typing of EVs by sequencing have greatly improved our ability to study their epidemiology. These methods have also enabled the identification of EVs that are difficult to propagate in cell culture [7–9], and the accurate typing and classification based on genomic sequences [10–12]. A method introduced by Nix et al. [13] has been widely used for molecular typing of EVs. It uses a consensus degenerate hybrid oligonucleotide primer (CODEHOP) approach and amplifies VP1 coding region of all known EV serotypes [14]. Although these methods have advanced epidemiological studies in hospitalized patients, very little is still known about the epidemiology of EVs in the general population, where EV infections remain mainly asymptomatic or mild. These “hidden/subclinical” infections can have important health consequences, since they constitute a reservoir of EVs, which varies constantly due to recombination and a high mutation rate, thus leading to outbreaks and severe diseases in susceptible populations. In addition, some of these viruses may play a role in chronic diseases, such as type 1 diabetes (T1D) and cardiomyopathies, even if they do not cause any other apparent symptom [15,16]. 2. Objective The aim of this study was to characterize the large reservoir of EVs that circulate in young children in the general population of different geographical regions. Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 2 / 18 Funding: Virus analyses were funded by the Juvenile Diabetes Research Foundation (JDRF), Reino Lahtikari Foundation and the Academy of Finland (grants to HH). The TEDDY study has been funded by U01 DK63829, U01 DK63861, U01 DK63821, U01 DK63865, U01 DK63863, U01 DK63836, U01 DK63790, UC4 DK63829, UC4 DK63861, UC4 DK63821, UC4 DK63865, UC4 DK63863, UC4 DK63836, UC4 DK95300, UC4 DK100238, UC4 DK106955, and contract no. HHSN267200700014C from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute of Allergy and Infectious Diseases (NIAID), National Institute of Child Health and Human Development (NICHD), National Institute of Environmental Health Sciences (NIEHS), Juvenile Diabetes Research Foundation (JDRF), and Centers for Disease Control and Prevention (CDC). This work supported in part by the NIH/NCATS Clinical and Translational Science Awards to the University of Florida (UL1 TR000064) and the University of Colorado (UL1 TR001082). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: HH is a shareholder and chairman of the board of Vactech Ltd. (http://www. vactech.fi/en/), which develops vaccines against picornaviruses. There are no marketed products to declare, and this does not alter our adherence to PLOS ONE policies on sharing data and materials. All other authors have declared that no competing interests exist. 3. Materials and methods 3.1. Study population and collection of stool samples The study subjects were selected randomly from the children who participated in a large birthcohort study named “The Environmental Determinants of Diabetes in the Young (TEDDY)” recruiting children in six study centers in the USA and Europe. In the TEDDY study, large numbers of samples including blood and stool specimens are collected from children with genetically increased risk of T1D (19). Participants of the present study represent a randomly selected sub-cohort of TEDDY not enriched for any disease, sex or ethnic group. The participants were recruited from the general population and had no first-degree relatives with T1D. Study subjects carried one of the T1D-associated HLA-DQ genotypes that were screened for at birth (for further information see Hagopian et al. [17]). At the time of selection, all participants were non-diabetic and negative for T1D-associated autoantibodies. A total of 4018 stool samples were collected monthly between the 17 th of January 2005 and the 16 th of March 2009 from 300 children including 147 girls and 153 boys aged 3-18 months, of whom 10 became positive for T1D autoantibodies and one child was eventually diagnosed with T1D (by the end of September 2016). The HLA-DQ genotypes were DQ2/8 (50% of the participants), DQ2/2 (27%) and DQ8/8 (23%). The distribution of the HLA types; however, differed across the study centers (S1 Table). It is worth mentioning that other HLA types, which pose an increased risk for T1D, were not included in this study. Altogether 20 (6.7%) children belonged to an ethnic minority (Hispanic, African American or other ethnic minority in the USA). The sample series included 662 stool samples from 51 children in Colorado (COL), 669 samples from 50 children in Georgia (GEO), 668 samples from 50 children in Washington (WAS), 702 samples from 50 children in Finland (FIN), 645 samples from 49 children in Germany (GER), and 672 samples from 50 children in Sweden (SWE) as shown in Table 1. A median number of 13 samples were collected per child (range 2-16), and 99% of the children were sampled 8–16 times during the study period. Samples were collected evenly throughout the year in all centers (S2 Table) and their number did not differ significantly between the study sites (Table 1). Samples were collected at home by the parents and were sent Table 1. Number of stool samples and children studied, and EV positivity rates in different study centers. Study center Stool Samples Children N Positive N (%) Typed N (%) N# Positive N (%) COL 662 106 (16) 88 (83) 51 31 (61) GEO 669 160 (24) 136 (85) 50 46 (92) WAS 668 97 (15) 75 (77) 50 34 (68) Sub total 1999 368 (18) 299 (81) 151 111 (74) FIN 702 23 (3) 9 (39) 50 13 (26) GER 645 93 (14) 79 (85) 49 36 (73) SWE 672 107 (16) 92 (86) 50 41 (82) Sub total 2019 225 (11) 180 (80) 149 90 (60) Total 4018 586 (15) 479 (82) 300 201 (67) Percent of the samples  Percent of EV positives, which are typed  Percent of the children positive for EV The number of samples and the percentage where the type of EV was identified by sequencing are also shown. Study centers include Colorado (COL), Georgia (GEO), Washington (WAS), Finland (FIN), Germany (GER), and Sweden (SWE). https://doi.org/10.1371/journal.pone.0201959.t001 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 3 / 18 to the TEDDY repository using express courier mail in the USA, or to local TEDDY centers in Europe using regular local mail. They were kept at an ambient temperature for one to three days during the shipment and temperature peaks during shipment were prevented by an icegel pack. Samples were stored at -80˚C on arrival at the first destination. Data on clinical symptoms were collected by parents at home using a diary and were translated to different symptom and disease codes (ICD-10 codes) by study nurses at each study visit. The details of the procedures used for the collection of symptom data have been described previously [18]. 3.2. Ethics statement The TEDDY study was approved by local U.S. Institutional Review Boards and European Ethics Committee Boards in Colorado’s Colorado Multiple Institutional Review Board, Georgia’s Medical College of Georgia Human Assurance Committee (2004–2010), Georgia Health Sciences University Human Assurance Committee (2011–2012), Georgia Regents University Institutional Review Board (2013–2015), Augusta University Institutional Review Board (2015-present), Florida’s University of Florida Health Center Institutional Review Board, Washington state’s Washington State Institutional Review Board (2004–2012) and Western Institutional Review Board (2013-present), Finland’s Ethics Committee of the Hospital District of Southwest Finland, Germany’s Bayerischen Landesa¨rztekammer (Bavarian Medical Association) Ethics Committee, Sweden’s Regional Ethics Board in Lund, Section 2 (2004–2012) and Lund University Committee for Continuing Ethical Review (2013-present). 3.3. Enterovirus detection and typing Stool suspensions (10% w/v) were prepared by thawing the samples and diluting them in Hanks balanced salt solution (HBSS, Sigma) containing 4% fetal calf serum (FCS). Suspensions were clarified by centrifugation at 500xg for 20 minutes at 4˚C. Clarified supernatants were collected in aliquots in sterile cryo-preservation tubes and were stored at -80˚C. EV RNA was extracted using a high-throughput nucleic acid extraction method (MagNa Pure extraction robot, Roche, Applied Science, Germany—using Total Nucleic Acid extraction kit, Roche). The viral genome was then detected by real-time RT-PCR as described elsewhere [19] using primers and probes shown in S3 Table. An average cycle threshold (Ct) from triplicate sample wells was used with a cut-off of 46 for positivity. All positive samples were further investigated by partial sequencing of a stretch of the VP1 coding region as described by Nix et al. [13]. In our study approximately 350-400 nucleotides were identified by sequencing, depending on the virus type. To identify the EV genotype, sequences were analyzed using “Enterovirus Genotyping Tool Version 0.1”; an online tool developed by the National Institute for public Health and Environment (RIVM), Bilthoven, the Netherlands [20]. 3.4. Statistical analyses The SPSS statistical package (version 22) was used for the analyses. The Kruskal-Wallis or Mann-Whitney U test was used to identify the statistical significance of the differences between groups. The Z-test was used to compare two proportions expressed as percentages. Binary logistic regression was used to analyze the probability of being EV positive according to the HLA-type, study center and sex of the child. The statistical significance was set to p= 0.05. The prevalence of infections was defined as the proportion of children with at least one EV positive stool sample (child positivity rate). The rate of sample positivity was defined as the proportion of EV positive samples. Infection episodes were defined by the detection of a solitary EV positive sample or the detection of the same EV strain in consecutive samples Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 4 / 18 (identified by identical sequences). If the sequencing failed, Ct-values were used to identify the episode. An increase in the Ct-values in consecutively positive samples was assumed to represent one episode (indicating a decline in the virus shedding). An algorithm was created to calculate the length of an infection episode (S1 Fig). 4. Results 4.1. Prevalence of enterovirus infections Altogether 14.6% of all samples were EV positive by real-time RT-PCR using a cut-off of Ct = 46 (the majority of them were still EV positive using a lower cut-off of Ct = 42 being positive in 14.1% of total samples, S2 Fig). Sample positivity rate was the lowest in Finland (3.3%) and the highest in Georgia (24%, Fig 1). The mean Ct-value of positive samples (N = 586) was 31.5 (range 15.7–45.6, SD 6.8) showing no difference between the centers (p= 0.54). Fig 1. Prevalence of EVs represented as the child positivity rate and sample positivity rate in each study center. The difference in positivity rates between Finland and all other study centers is statistically significant in each category (p<0.001). Study centers include Georgia (GEO), Washington (WAS), Colorado (COL), Finland (FIN), Sweden (SWE) and Germany (GER). https://doi.org/10.1371/journal.pone.0201959.g001 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 5 / 18 At least one EV positive stool sample was detected in 67% of the study participants. The proportion of EV positive children was much lower in Finland than in the other countries combined (26% vs. 75%, p<0.001; Fig 1). The highest rates were seen in Georgia, where 92% of the children were EV positive. The prevalence of EV infections increased with the age of the child (S3 Fig). The mean number of positive samples per child was 1.98 (SD 2.05, Median of 1, interquartiles of 0, 1, and 3). A clear seasonal pattern was seen in all geographical regions showing a peak during spring, summer, and autumn months (March-October) with some variation between the study centers (Fig 2). Sample positivity rate peaked in certain years. Colorado had a significantly higher number of EVs in 2007 (25%) compared to the average of 9.5% for other years (p<0.001). Georgia had a significantly higher number of EVs in 2008 (30%) compared to 21% for all other years (p= 0.018). 4.2. Prevalence of different enterovirus types Typing was successful in 82% of the EV-positive samples (479 samples; Table 1). The success rate of genotyping depended on the virus load, since the typed samples had significantly lower average Ct-values compared to un-typed samples (mean Ct-value 30 versus 37, p<0.001). The majority (61.5%) of the genotyped viruses belonged to the species Enterovirus A (EV-A), 38% to Enterovirus B (EV-B) and only 0.5% to Enterovirus C (EV-C) (Table 2). Interestingly some human rhinoviruses (HRVs) were also amplified by the sequencing primers, because of their sequence homology with EVs (six samples); these were excluded from the analysis. Fig 2. Seasonal distribution of EV-positive stool samples in different study centers. The year has been devided into 4 seasons from March to February (Spring includes March, April and May). The total number of EV-positive samples in each season is shown for each study center. https://doi.org/10.1371/journal.pone.0201959.g002 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 6 / 18 CV-As were the most frequently detected EVs in all study centers. 37.7% of all children experienced at least one CV-A infection and 61.7% of all genotyped EVs were CV-As (Table 2). The second most common viruses were echoviruses (Es) that were detected in 16.3% of the children and 19% of the genotyped samples, followed by CV-Bs that were detected in 14.7% of the children and 16.1% of the genotyped samples (Table 2). Of the individual EVs CV-A4 was the most common virus in the stool samples (18.2% of all typed EVs) followed by CV-A6 (11.9%), CV-A2 (9.6%), CV-A5 (7.8%), and CV-A10 (5.7%; Table 2), a pattern which showed only minor variations between different centers. Only 3.1% of all typed Table 2. Type of EVs in each study center and the rank of the virus among all typed viruses. Number of typed viruses in stool samples (Ranks) % of typed viruses Species Genotype USA COL GEO WAS Europe FIN GER SWE Total EV-A CV-A2 31 (2) 9 (2) 17 (1) 5 (4) 6 (7) 1 (2) 3 (5) 2 (8) 37 (3) 9.6 CV-A4 48 (1) 22 (1) 12 (2) 14 (1) 22 (1) 3 (1) 9 (1) 10 (1) 70 (1) 18.2 CV-A5 18 (4) 2 (6) 5 (7) 11 (3) 12 (3) 0 5 (3) 7 (3) 30 (4) 7.8 CV-A6 29 (3) 6 (3) 11 (3) 12 (2) 17 (2) 1 (2) 7 (2) 9 (2) 46 (2) 11.9 CV-A8 2 (13) 0 2 (10) 0 2 (11) 0 1 (7) 1 (9) 4 (15) 1.0 CV-A10 12 (5) 1 (7) 7 (5) 4 (5) 10 (4) 0 5 (3) 5 (5) 22 (5) 5.7 CV-A16 10 (7) 2 (6) 6 (6) 2 (7) 5 (8) 0 1 (7) 4 (6) 15 (7) 3.9 EV-A71 3 (12) 0 2 (10) 1 (8) 9 (5) 0 3 (5) 6 (4) 12 (10) 3.1 Sub total 153 42 62 49 83 5 34 44 236 61.5 EV-B CV-A9 5 (10) 0 5 (7) 0 6 (7) 0 2 (6) 4 (6) 11 (11) 2.9 CV-B1 12 (5) 5 (4) 3 (9) 4 (5) 2 (11) 0 2 (6) 0 14 (8) 3.6 CV-B2 8 (8) 1 (8) 6 (6) 1 (8) 2 (11) 0 0 2 (8) 10 (12) 2.6 CV-B3 5 (10) 2 (6) 2 (10) 1 (8) 8 (6) 0 5 (3) 3 (7) 13 (9) 3.4 CV-B4 8 (8) 4 (5) 3 (10) 1 (8) 8 (6) 0 7 (2) 1 (9) 16 (6) 4.2 CV-B5 4 (11) 0 4 (8) 0 5 (8) 0 0 5 (5) 9 (13) 2.3 E-3 4 (11) 0 4 (8) 0 0 0 0 0 4 (15) 1.0 E-6 6 (9) 0 4 (8) 2 (7) 4 (9) 0 1 (7) 3 (7) 10 (12) 2.6 E-9 1 (14) 0 0 1 (8) 3 (10) 0 2 (6) 1 (9) 4 (15) 1.0 E-11 4 (11) 2 (6) 1 (11) 1 (8) 9 (5) 0 5 (3) 4 (6) 13 (9) 3.4 E-13 2 (13) 0 2 (11) 0 6 (7) 0 4 (4) 2 (8) 8(14) 2.1 E-18 11 (6) 5 (4) 3 (9) 3 (6) 1 (12) 0 1 (7) 0 12 (10) 3.1 E-21 0 0 0 0 1 (12) 0 0 1 (9) 1 (16) 0.3 E-25 12 (5) 2 (6) 9 (4) 1 (8) 1 (12) 0 0 1 (9) 13 (9) 3.4 E-30 5 (10) 0 4 (8) 1 (8) 3 0 2 (6) 1 (9) 8 (14) 2.1 Sub total 87 21 50 16 59 0 31 28 146 38 EV-C CV-A1 0 0 0 0 1 (12) 0 0 1 (9) 1 (16) 0.3 CV-A22 1 (14) 0 1 (11) 0 0 0 0 0 1 (16) 0.3 Sub total 1 0 1 0 1 0 0 1 2 0.5 Total 241 63 113 65 143 5 65 73 384 100 Number of typed group of viruses in stool samples (%) All CVA 156 (64.7) 42 (66.7) 66 (58.4) 48 (73.8) 81 (56.6) 5 (100) 33 (50.8) 43 (58.9) 237 (61.7) - All CVB 37 (15.4) 12 (19) 18 (15.9) 7 (10.8) 25 (17.5) 0 14 (21.5) 11 (15.1) 62 (16.1) - All EVs 45 (18.7) 9 (14.3) 27 (23.9) 9 (13.8) 28 (19.6) 0 15 (23.1) 13 (17.8) 73 (19) - Typed viruses are categorized in species EV-A, EV-B, and EV-C. Each child can be positive for a number of EVs. The number of each virus type is shown for each study center. The data from the USA, Europe, and the total numbers are shown separately for comparison. The last column shows the percentage of each virus (and species) among all typed viruses. Numbers in the brackets represent the rank of the virus among all viruses in each category (column). At the bottom 3 rows of the table the total number (and the percentage) of CVA, CVB, and EVs are presented. https://doi.org/10.1371/journal.pone.0201959.t002 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 7 / 18 EVs were EV-A71 and no EV-D68 was found. However, due to it’s biological properties, EV-D68 is hard to detect in stool samples. Among different study centers, Germany had the most frequent CVBs (22% of the samples), followed by Colorado (19%), Georgia (16%), Sweden (15%), and Washington (11%) (Table 2). EVs were detected rarely before the age of 6 months becoming more frequent at older age (S6 Table) and certain CAVs (CVA2, 4, 5, 6) showed an epidemic pattern peaking in certain years (S6 Table). None of the children in this study had meningitis, myocarditis, encephalitis, paralysis or other severe symptoms typical for EV infection. Only 1/95 children with EV-A infection had symptoms compatible with HFMD (commonly caused by CV-A6, A10, A16 or EV-A71) at the time of the infection episode; the child was positive for CV-A6. Duration of virus shedding. A total number of 411 EV episodes were identified using the criteria described in S1 Fig, of these 315 had a positive identification of the virus genotype. Altogether 268 (65.5%) of the episodes included only a single EV positive sample, 112 (27.3%) had two, 25 (6.1%) three, and 6 (1.5%) had four consecutive positive stool samples. At least one infection episode was observed in 164 children (55% of total participants) and 71 (43%) of them had only one infection episode, 51 (31%) had two episodes, 28 (17%) had three episodes, 12 (7%) had four episodes (7.3%), and 2 (1%) had five episodes. Finnish children had significantly less episodes than children in other countries (p<0.001, Fig 3). The mean duration of the infection episodes was 38 days (SD 20.4 days, median 30 days, interquartiles 24, 30, and 50) showing no difference between the study centres (p= 0.08, Fig 3). Among the successfully genotyped EV episodes CV-A4 episodes (N = 52) were characterized by significantly longer virus shedding time compared to all other EV types (mean 51 vs. 36 days, p<0.01; Fig 4). Similarly, E-18 episodes (N = 9) were longer compared to all other EVs (mean 59 vs. 36, p= 0.026), and the same was true for E-25 (N = 9; mean 47 vs. 38 days, p= 0.026; Fig 4). On the other hand, the episodes caused by un-typed EVs were significantly Fig 3. The number of EV infection episodes (A) and the duration of episodes (B) in different study centers. The number of infection episodes is shown by the number of the subjects having 0, 1, 2, or more than 2 infection episodes per child (A) and the duration of the episodes is shown in days (B). The thick horizontal lines in the boxes represent the median for each center with the inter-quartile values; the circles represent the outliers (B). https://doi.org/10.1371/journal.pone.0201959.g003 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 8 / 18 shorter compared to the episodes caused by typed EVs (mean 28 vs. 41 days, p<0.001). This may be a result of timing of sample collection (recovery phase of the infection) or the low levels of the virus shed in the un-typed episodes. Corresponding frequencies of EV episodes caused by different EV genotypes are shown in S4 Table. 4.3. Effect of host factors on enterovirus infections 65.3% of girls and 67.3% of boys were infected at least once with EVs (p= 0.54). The duration of the infection episodes did not differ between them either. However, the HLA genotype correlated with EV positivity, as children who were HLA-DQ2/8 heterozygous were less frequently EV positive compared to children who were homozygous for either HLA-DQ2 or HLA-DQ8 (P= 0.048; Fig 5 and S4 Fig). 5. Discussion To our knowledge, this may be the largest study carried out so far to characterize EVs in otherwise healthy, young children in different countries, by analyzing stool samples which had been collected on a monthly basis and examined for the presence of EV RNA, followed by sequencing of an RT-PCR-amplified region of viral genome to identify the type of these EVs. Fig 4. Duration of EV infection episodes caused by different EV serotypes compared to all other episodes. The duration of CV-A4 infection episodes was significantly longer compared to all other enterovirus types together (p<0.001). E-18 and E-25 also had longer episodes (p= 0.026 and ^ p= 0.026). The three viruses together (CV-A4, E-18, and 25) had significantly longer infection episodes compared to the rest of the viruses ( ^^ p<0.001). The length of infection episodes by un-typed enteroviruses was significantly shorter compared to typed viruses ( “ p<0.001). All 3 a refers to CV-A4, E-18 and E-25 together. https://doi.org/10.1371/journal.pone.0201959.g004 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 9 / 18 Writing – review & editing: Amir-Babak Sioofy-Khojine, Sami Oikarinen, Hanna Honkanen, Heini Huhtala, Jussi P. Lehtonen, Thomas Briese, Heikki Hyo¨ty. References 1. Khetsuriani N, Lamonte-Fowlkes A, Oberst S, Pallansch MA, Centers for Disease Control and Prevention. Enterovirus surveillance-United States, 1970–2005. MMWR Surveill Summ 2006 Sep 15; 55(8):1– 20. 2. Pallansch M, Oberste M, Whitton J. Enteroviruses: polioviruses, coxsakieviruses, echoviruses, and newer enteroviruses. In: Knipe D, Howley P, editors. Fields virology. 6th ed ed.: Lippincott Williams & Wilkins, Philadelphia, PA.; 2013. p. 490–530. 3. Zell R, Delwart E, Gorbalenya AE, Hovi T, King AMQ, Knowles NJ, et al. ICTV Virus Taxonomy Profile: Picornaviridae. J Gen Virol 2017 Oct; 98(10):2421–2422. https://doi.org/10.1099/jgv.0.000911 PMID: 28884666 4. Melnick JL. Enteroviruses. In: Fields BN, Knipe DM, Chanock RM, Hirch MS, Melnick JL, Monath TP, Roizman B, editor. 2nd ed. New York: Raven Press; 1990. p. 590. 5. Pallansch M, Oberste M. Coxsackievirus, echovirus, and other enteroviruses. In: Gorbach S, Bartlett J, Blacklow N, editors. Infectious diseases.: Philadelphia: Lippincott, Williams and Wilkins; 2003. p. 2347–2053. 6. Tan Y, Hassan F, Schuster JE, Simenauer A, Selvarangan R, Halpin RA, et al. Molecular Evolution and Intraclade Recombination of Enterovirus D68 during the 2014 Outbreak in the United States. J Virol 2015 Dec 9; 90(4):1997–2007. https://doi.org/10.1128/JVI.02418-15 PMID: 26656685 7. Lipson SM, Walderman R, Costello P, Szabo K. Sensitivity of rhabdomyosarcoma and guinea pig embryo cell cultures to field isolates of difficult-to-cultivate group A coxsackieviruses. J Clin Microbiol 1988 Jul; 26(7):1298–1303. PMID: 2842370 8. Schmidt NJ, Ho HH, Lennette EH. Propagation and isolation of group A coxsackieviruses in RD cells. J Clin Microbiol 1975 Sep; 2(3):183–185. PMID: 1176625 9. Landry ML, Madore HP, Fong CK, Hsiung GD. Use of guinea pig embryo cell cultures for isolation and propagation of group A coxsackieviruses. J Clin Microbiol 1981 Mar; 13(3):588–593. PMID: 6263943 10. Chiang PS, Huang ML, Luo ST, Lin TY, Tsao KC, Lee MS. Comparing molecular methods for early detection and serotyping of enteroviruses in throat swabs of pediatric patients. PLoS One 2012; 7(10): e48269. https://doi.org/10.1371/journal.pone.0048269 PMID: 23133580 11. Tsao LY, Lin CY, Yu YY, Wang BT. Microchip, reverse transcription-polymerase chain reaction and culture methods to detect enterovirus infection in pediatric patients. Pediatr Int 2006 Feb; 48(1):5–10. https://doi.org/10.1111/j.1442-200X.2006.02157.x PMID: 16490062 12. Vuorinen T, Vainionpaa R, Hyypia T. Five years’ experience of reverse-transcriptase polymerase chain reaction in daily diagnosis of enterovirus and rhinovirus infections. Clin Infect Dis 2003 Aug 1; 37 (3):452–455. https://doi.org/10.1086/376635 PMID: 12884172 13. Nix WA, Oberste MS, Pallansch MA. Sensitive, seminested PCR amplification of VP1 sequences for direct identification of all enterovirus serotypes from original clinical specimens. J Clin Microbiol 2006 Aug; 44(8):2698–2704. https://doi.org/10.1128/JCM.00542-06 PMID: 16891480 14. Nix WA, Oberste MS, Pallansch MA. Sensitive, seminested PCR amplification of VP1 sequences for direct identification of all enterovirus serotypes from original clinical specimens. J Clin Microbiol 2006 Aug; 44(8):2698–2704. https://doi.org/10.1128/JCM.00542-06 PMID: 16891480 15. Hyoty H. Viruses in type 1 diabetes. Pediatr Diabetes 2016 Jul; 17 Suppl 22:56–64. 16. Yeung WC, Rawlinson WD, Craig ME. Enterovirus infection and type 1 diabetes mellitus: systematic review and meta-analysis of observational molecular studies. BMJ 2011 Feb 3; 342:d35. https://doi.org/ 10.1136/bmj.d35 PMID: 21292721 17. Hagopian WA, Erlich H, Lernmark A, Rewers M, Ziegler AG, Simell O, et al. The Environmental Determinants of Diabetes in the Young (TEDDY): genetic criteria and international diabetes risk screening of 421 000 infants. Pediatr Diabetes 2011 Dec; 12(8):733–743. https://doi.org/10.1111/j.1399-5448.2011. 00774.x PMID: 21564455 18. Lonnrot M, Lynch K, Larsson HE, Lernmark A, Rewers M, Hagopian W, et al. A method for reporting and classifying acute infectious diseases in a prospective study of young children: TEDDY. BMC Pediatr 2015 Mar 20; 15:24-015-0333-8. 19. Honkanen H, Oikarinen S, Nurminen N, Laitinen OH, Huhtala H, Lehtonen J, et al. Detection of enteroviruses in stools precedes islet autoimmunity by several months: possible evidence for slowly operating mechanisms in virus-induced autoimmunity. Diabetologia 2017 Mar; 60(3):424–431. https://doi.org/10. 1007/s00125-016-4177-z PMID: 28070615 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 16 / 18 20. Kroneman A, Vennema H, Deforche K, v d Avoort H, Penaranda S, Oberste MS, et al. An automated genotyping tool for enteroviruses and noroviruses. J Clin Virol 2011 Jun; 51(2):121–125. https://doi.org/ 10.1016/j.jcv.2011.03.006 PMID: 21514213 21. Abedi GR, Watson JT, Pham H, Nix WA, Oberste MS, Gerber SI. Enterovirus and Human Parechovirus Surveillance—United States, 2009–2013. MMWR Morb Mortal Wkly Rep 2015 Sep 4; 64(34):940–943. https://doi.org/10.15585/mmwr.mm6434a3 PMID: 26334674 22. Centers for Disease Control and Prevention (CDC). Nonpolio enterovirus and human parechovirus surveillance—United States, 2006–2008. MMWR Morb Mortal Wkly Rep 2010 Dec 10; 59(48):1577–1580. PMID: 21150865 23. CDC U. The National Enterovirus Surveillance System (NESS), USA. March 11, 2016; Available at: https://www.cdc.gov/surveillance/ness/index.html. Accessed 04/16, 2018. 24. Ryu WS, Kang B, Hong J, Hwang S, Kim J, Cheon DS. Clinical and etiological characteristics of enterovirus 71-related diseases during a recent 2-year period in Korea. J Clin Microbiol 2010 Jul; 48(7):2490– 2494. https://doi.org/10.1128/JCM.02369-09 PMID: 20463159 25. Hu YF, Yang F, Du J, Dong J, Zhang T, Wu ZQ, et al. Complete genome analysis of coxsackievirus A2, A4, A5, and A10 strains isolated from hand, foot, and mouth disease patients in China revealing frequent recombination of human enterovirus A. J Clin Microbiol 2011 Jul; 49(7):2426–2434. https://doi. org/10.1128/JCM.00007-11 PMID: 21543560 26. Tryfonos C, Richter J, Koptides D, Yiangou M, Christodoulou CG. Molecular typing and epidemiology of enteroviruses in Cyprus, 2003–2007. J Med Microbiol 2011 Oct; 60(Pt 10):1433–1440. https://doi.org/ 10.1099/jmm.0.029892-0 PMID: 21596905 27. Gopalkrishna V, Patil PR, Patil GP, Chitambar SD. Circulation of multiple enterovirus serotypes causing hand, foot and mouth disease in India. J Med Microbiol 2012 Mar; 61(Pt 3):420–425. https://doi.org/10. 1099/jmm.0.036400-0 PMID: 22052995 28. Duintjer Tebbens RJ, Pallansch MA, Kew OM, Caceres VM, Sutter RW, Thompson KM. A dynamic model of poliomyelitis outbreaks: learning from the past to help inform the future. Am J Epidemiol 2005 Aug 15; 162(4):358–372. https://doi.org/10.1093/aje/kwi206 PMID: 16014773 29. Simonen-Tikka ML, Hiekka AK, Klemola P, Poussa T, Ludvigsson J, Korpela R, et al. Early human enterovirus infections in healthy Swedish children participating in the PRODIA pilot study. J Med Virol 2012 Jun; 84(6):923–930. https://doi.org/10.1002/jmv.23284 PMID: 22499016 30. Simonen-Tikka ML, Pflueger M, Klemola P, Savolainen-Kopra C, Smura T, Hummel S, et al. Human enterovirus infections in children at increased risk for type 1 diabetes: the Babydiet study. Diabetologia 2011 Dec; 54(12):2995–3002. https://doi.org/10.1007/s00125-011-2305-3 PMID: 21932150 31. Tapia G, Cinek O, Rasmussen T, Witso E, Grinde B, Stene LC, et al. Human enterovirus RNA in monthly fecal samples and islet autoimmunity in Norwegian children with high genetic risk for type 1 diabetes: the MIDIA study. Diabetes Care 2011 Jan; 34(1):151–155. https://doi.org/10.2337/dc10-1413 PMID: 20929993 32. Chung PW, Huang YC, Chang LY, Lin TY, Ning HC. Duration of enterovirus shedding in stool. J Microbiol Immunol Infect 2001 Sep; 34(3):167–170. PMID: 11605806 33. Han J, Ma XJ, Wan JF, Liu YH, Han YL, Chen C, et al. Long persistence of EV71 specific nucleotides in respiratory and feces samples of the patients with Hand-Foot-Mouth Disease after recovery. BMC Infect Dis 2010 Jun 18; 10:178-2334-10-178. 34. Viskari H, Ludvigsson J, Uibo R, Salur L, Marciulionyte D, Hermann R, et al. Relationship between the incidence of type 1 diabetes and enterovirus infections in different European populations: results from the EPIVIR project. J Med Virol 2004 Apr; 72(4):610–617. https://doi.org/10.1002/jmv.20033 PMID: 14981763 35. Seiskari T, Kondrashova A, Viskari H, Kaila M, Haapala AM, Aittoniemi J, et al. Allergic sensitization and microbial load—a comparison between Finland and Russian Karelia. Clin Exp Immunol 2007 Apr; 148(1):47–52. https://doi.org/10.1111/j.1365-2249.2007.03333.x PMID: 17302731 36. Pons-Salort M, Oberste MS, Pallansch MA, Abedi GR, Takahashi S, Grenfell BT, et al. The seasonality of nonpolio enteroviruses in the United States: Patterns and drivers. Proc Natl Acad Sci U S A 2018 Mar 20; 115(12):3078–3083. https://doi.org/10.1073/pnas.1721159115 PMID: 29507246 37. Miller E, Gay N. Effect of age on outcome and epidemiology of infectious diseases. Biologicals 1997 Jun; 25(2):137–142. https://doi.org/10.1006/biol.1997.0072 PMID: 9236040 38. Nathanson N, Martin JR. The epidemiology of poliomyelitis: enigmas surrounding its appearance, epidemicity, and disappearance. Am J Epidemiol 1979 Dec; 110(6):672–692. PMID: 400274 39. Viskari HR, Koskela P, Lonnrot M, Luonuansuu S, Reunanen A, Baer M, et al. Can enterovirus infections explain the increasing incidence of type 1 diabetes? Diabetes Care 2000 Mar; 23(3):414–416. PMID: 10868874 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 17 / 18 40. Fellay J, Shianna KV, Ge D, Colombo S, Ledergerber B, Weale M, et al. A whole-genome association study of major determinants for host control of HIV-1. Science 2007 Aug 17; 317(5840):944–947. https://doi.org/10.1126/science.1143767 PMID: 17641165 41. Fellay J, Ge D, Shianna KV, Colombo S, Ledergerber B, Cirulli ET, et al. Common genetic variation and the control of HIV-1 in humans. PLoS Genet 2009 Dec; 5(12):e1000791. https://doi.org/10.1371/ journal.pgen.1000791 PMID: 20041166 42. International HIV Controllers Study, Pereyra F, Jia X, McLaren PJ, Telenti A, de Bakker PI, et al. The major genetic determinants of HIV-1 control affect HLA class I peptide presentation. Science 2010 Dec 10; 330(6010):1551–1557. https://doi.org/10.1126/science.1195271 PMID: 21051598 43. Sadeharju K, Knip M, Hiltunen M, Akerblom HK, Hyoty H. The HLA-DR phenotype modulates the humoral immune response to enterovirus antigens. Diabetologia 2003 Aug; 46(8):1100–1105. https:// doi.org/10.1007/s00125-003-1157-x PMID: 12845430 44. Hraber P, Kuiken C, Yusim K. Evidence for human leukocyte antigen heterozygote advantage against hepatitis C virus infection. Hepatology 2007 Dec; 46(6):1713–1721. https://doi.org/10.1002/hep.21889 PMID: 17935228 45. Magaret A, Dong L, John M, Mallal SA, James I, Warren T, et al. HLA Class I and II alleles, heterozygosity and HLA-KIR interactions are associated with rates of genital HSV shedding and lesions. Genes Immun 2016 Dec; 17(7):412–418. https://doi.org/10.1038/gene.2016.42 PMID: 27853144 Enterovirus molecular epidemiology in young children PLOS ONE | https://doi.org/10.1371/journal.pone.0201959 September 7, 2018 18 / 18