Genetic diversity of hepatitis E virus (HEV) strains derived from humans, swine and wild boars in Croatia from 2010 to 2017

Hepatitis E is an enterically transmitted, mostly self limiting and waterborne acute viral infection in developing countries of Asia, Africa and Central America. In Europe and North America autochthonous infections are sporadic and no cases of HEV transmission through contaminated water have been reported [1‐4]. Even so, HEV RNA was derived from river water [5] therefore it may potentially be a source of infection. Conversely, bottled water is not considered a risk factor for HEV infection since a significantly lower rate of anti-HEV IgG antibodies was found in people that consumed bottled water when compared to the overall IgG seroprevalence [6].

The significance of the disease as a public health problem is arising especially as extra-hepatic manifestations associated with hepatitis E infection [7], as well as cases of chronic, even fulminate infections in immunocompromised patients are recorded [1, 8‐10]. The causative agent, the hepatitis E virus (HEV) is a small (27–34 nm), non-enveloped single-stranded virus with a linear positive-sense RNA genome of approximately 7.2 kb in length [11]. According to its genetic diversity, HEV classifies into the Hepeviridae family consisting of two genera, Orthohepevirus (A, B, C and D) and Piscihepevirus [12]. Members of Orthohepevirus A species are currently divided into seven genotypes [13]. Recently, an eighth genotype has been proposed [14]. HEV-1 and HEV-2 are predominately found in developing countries and are human specific; HEV-3 is distributed worldwide and has been derived from both humans and animals, whereas HEV-4, that also shows a zoonotic potential, is considered to be indigenous to Asia [15]. Even so, cases of infection by HEV-4 have recently also been reported in Europe [16]. Novel genotypes of HEV as well as hosts are consecutively being recognised, however from the epidemiological point of view, only members of the Suidae family are still considered as the only true reservoirs of HEV. Due to the fact that HEV infection remains asymptomatic in animal mammals, therefore in swine as well, and due to a long incubation period reported in patients, only a limited number of human infections are directly linked to the consumption of food of animal origin as the source of infection [17‐19]. Therefore, the comparison of sequences of HEV RNA derived from humans and swine remains the only method of choice for detecting possible routes of transmission [20].

Since 2009 HEV RNA has been continuously detected in swine and wild boars in Croatia [21, 22], and has been followed by the report of the first human case of autochthonous hepatitis E in 2012 [23]. Since then, sporadically infected patients showing signs of hepatic disorders and positive for anti-HEV antibodies and for HEV RNA have regularly been recorded [24]. The seroprevalence in non A-C hepatitis patients has shown to be 10.7% and in HIV positive individuals 1.1% [24]. Likewise, a pilot study revealed a seroprevalence of 5.6% in asymptomatic individuals [25]. A high seroprevalence of 32.9 and 31.1% was detected in swine and wild boar in Croatia, respectively [22]. However, the relation among the reported cases in humans, swine and wild boars has not been investigated. Therefore, this study presents the first results of an epidemiological observation conducted using samples collected between 2010 and 2017 in Croatia based on genotyping and phylogenetic comparison of HEV strains derived from infected humans, swine and wild boars with an aim to reveal potential interspecies transmission or identify mutual sources of infection.

On the basis of genetic analysis carried out in this study, we can confirm that HEV strains derived during 2010–2017 in Croatia are closely related regardless of their geographical origin or host. Our findings indicate interspecies transmission of HEV and/or mutual sources of infection based on the high genetic identity of strains recovered from humans, swine and wild boar. The introduction of HEV infection into Croatia has not been entirely revealed within this study, however due to the result of genetic analyses from 2010 to 2012, we may presume that the virus was introduced by trade or by direct or indirect contact of swine with infected swine and wild boar from Europe. Even though the analyzed sequences rarely indicate a geographical connection, a statistically higher incidence of HEV RNA positive findings in all three investigated species in Northern Croatia where most of the swine farms are located and where the natural habitat of wild boars is recorded, shows that a high density of swine and wild boars increases the risk of viral spread.

HEV RNA was detected in 26 (14.94%) human patients that were IgM or IgM/IgG positive indicating recent infection, whereas IgG positive anti-HEV antibody sera remained HEV RNA negative, as previously described [28]. One positive archive human sera sample from 2010 that remained unrecognized until our study confirms that the introduction of HEV infection into Croatia appeared earlier than the reported case in 2012.

In tested swine, S-HEV was found in nine of the investigated counties with a prevalence of 15.2% (Table 1). Positive faecal samples were detected in six counties regardless of the breeding system, confirming viral shedding in detectable doses via faeces, possibly into the environment. Since farmers, especially within ecological management systems use swine faeces as manure, this finding may be a potential source of viral spread as previously reported [16, 29, 30]. The viral prevalence found in wild boars was 11.5% within the range reported in other European countries. Positive wild boars were detected in six investigated counties with the highest RNA prevalence found in counties with the highest density of wild boars (VP, BP and SM). In all the positive counties, swine with a high RNA prevalence were detected as well. Our results are comparable to other reported European findings [4].

H-HEV, S-HEV and WB-HEV strains from Croatia show to be genetically highly related members of Orthohepevirus A genotype 3. However, differences among them were observed. The derived strains clustered into two general clusters (3abchij and 3efg) [13], and furthermore into four subtypes (3a, 3c, 3e, 3f). Strains identified as members of subtype 3a and 3c are predominant in Croatia and may be considered endemic (Table 1). Only in KK County subtype 3a was not identified probably due to a limited number of tested samples.

H-HEV strains from subtype 3a show to be closely related to strains derived from wild boars, especially strains from SM County indicating a mutual source of infection or interspecies transmission of HEV. The estimated number of wild boars (according to the hunting bag) in Croatia varies from 45,700 to 65,000 per year. Since big game hunting is an important activity in this area, direct or indirect contact between hunters and wild boars is inevitable. Consumption of uncooked wild boar meat may also be considered an additional risk for HEV interspecies transmission in Croatia since H-HEV strains derived during 2012 to 2017 are genetically identical with some WB-HEV strains from 2010, 2015, 2016 and 2017.

S-HEV strains of subtype 3a show a higher genetic identity with strains derived from swine from neighbouring Serbia (HM483381 and HM 483383 from 2010) suggesting possible transmission of HEV by indirect contact, since import of live swine from Serbia is banned, or a mutual source of infection. To understand this finding additional testing of probable sources, including water, must be conducted.

H-HEV strains from subtype 3c appeared in 2010 and 2012 and have shown to be genetically identical to a swine and wild boar strain from Germany (JN415697 from 2009) and a swine strain from The Netherlands (EF372555 from 2005). Members of the same subtype were also derived from swine in 2013 and 2014 and show to be closely related to a human and a swine strain from the Netherlands (JQ929092 from 2010, EF372542 from 2005). We can only speculate the possible routes of infection in both cases. Croatia imports swine and swine products from Western European countries, so this entry could be a result of trade. Likewise, due to classical swine fever outbreaks in 2007–2008, the export of live pigs from Croatia to EU countries was banned or as of 2017 restricted from some regions in Croatia, so it’s less probable that the infection spread from Croatia to EU.

Two genetically identical HEV strains of human and swine origin, both from ZG County showed to be members of subtype 3e even though, they appeared in two different time periods. In swine they were derived in 2010 and 2012, whereas in humans in 2013 and 2014 indicating multiple entries. We can presume that the source of infection in patients is consumption of previously contaminated food of swine origin or the infection could be travel related. However, the source of infection in swine remains unidentified, as well as the reason why HEV 3e spreading or its persistence in the swine population is not recorded. Likewise, there is a lack of information in regards to the differences in virulence, duration of viraemia, host adaptability, as well as viral shedding of different HEV subtypes which may influence the detection of HEV RNA. Therefore, further investigations should be conducted in regards to the infection dynamics of HEV subtypes.

Members of subtype 3f were derived only from two patients, one in 2010 and the other in 2012. They show to be genetically highly related to strains derived from swine from the Netherlands (EF372555 from 2005) and a more distant strain derived from a human from Spain (AF195064, collection date is not mentioned in NCBI). Interestingly, subtype 3f strains have not yet been identified in swine/wild boars from Croatia and have not been detected in humans since 2012 indicating that this HEV subtype was imported.

According to the HEV RNA prevalence, the majority of HEV infected patients originate from ZG County where the largest genetic diversity of HEV strains was found. Since only small swine farms are represented in ZG County, the swine to human transmission was probably realised through consumption of contaminated homemade swine products, or by direct contact. This would correspondent with the fact that seroprevalence rates in Croatia in asymptomatic subjects are significantly higher in residents of suburban and rural areas [25] with a high density of small swine breeding farms. VS County borders with Serbia and Hungary and has the highest number of HEV positive swine and wild boars in Croatia, therefore finding homologous HEV strains derived from swine and wild boars (HM483383 from 2010 from Serbia and EU718646 from 2005 from Hungary) from these neighbouring countries is not surprising. S-HEV and WB-HEV strains are also genetically highly related among themselves confirming transmission of the virus, probably from swine farms to wild boars via contamination of the environment. However, further investigations should be carried out to identify additional HEV sources and transmission routes for a better understanding of HEV transmissibility.

In conclusion, since 2010 HEV is circulating among swine and wild boars in Croatia [21] and from 2012, and possibly even earlier, in humans [24, 25]. Our results confirm that swine and wild boars represent a public health risk as reservoirs of HEV since we confirmed the presence of identical strains of HEV in all three investigated species. However, according to our results, some strains detected in Croatia (subtypes 3a and 3c) are more likely to persist among swine, wild boars and humans, while other strains show a rather sporadic occurrence with tendency of disappearing. To understand these epidemiological differences, further investigation is needed.