Phylogeny of Shrew- and Mole-Borne Hantaviruses in Poland and Ukraine

Earlier, we demonstrated the co-circulation of genetically distinct non-rodent-borne hantaviruses, including Boginia virus (BOGV) in the Eurasian water shrew (Neomys fodiens), Seewis virus (SWSV) in the Eurasian common shrew (Sorex araneus) and Nova virus (NVAV) in the European mole (Talpa europaea), in central Poland. To further investigate the phylogeny of hantaviruses harbored by soricid and talpid reservoir hosts, we analyzed RNAlater®-preserved lung tissues from 320 shrews and 26 moles, both captured during 1990–2017 across Poland, and 10 European moles from Ukraine for hantavirus RNA through RT-PCR and DNA sequencing. SWSV and Altai virus (ALTV) were detected in Sorex araneus and Sorex minutus in Boginia and the Białowieża Forest, respectively, and NVAV was detected in Talpa europaea in Huta Dłutowska, Poland, and in Lviv, Ukraine. Phylogenetic analyses using maximum-likelihood and Bayesian methods showed geography-specific lineages of SWSV in Poland and elsewhere in Eurasia and of NVAV in Poland and Ukraine. The ATLV strain in Sorex minutus from the Białowieża Forest on the Polish–Belarusian border was distantly related to the ATLV strain previously reported in Sorex minutus from Chmiel in southeastern Poland. Overall, the gene phylogenies found support long-standing host-specific adaptation.


Introduction
Formerly grouped in the family Bunyaviridae and the genus Hantavirus, hantaviruses were recently reclassified into a new order (Bunyavirales) and family (Hantaviridae), the latter comprising four subfamilies (Actantavirinae, Agantavirinae, Mammantavirinae and Repantavirinae) [1,2]. The subfamily Mammantavirinae consists of four genera (Loanvirus, Mobatvirus, Orthohantavirus and Thottimvirus) based on DivErsity pArtitioning by hieRarchical Clustering (DEmARC) analysis and using concatenated complete S and M amino acidcoding regions [1]. Hantaviruses possess a single-stranded, negative-sense RNA genome consisting of three segments designated large (L), medium (M) and small (S), which encode

Trap Sites and Specimen Processing
Shrews and moles, captured during 1990-2017 in nine sites across Poland ( Figure 1A) as previously described [22,23], were euthanized by means of cervical dislocation. Carcasses were then stored at -20 • C for months to years before lung tissues were dissected and preserved in RNAlater ® RNA Stabilization Reagent (Qiagen, Valencia, CA, USA). In earlier studies, shrews and moles were collected in Boginia, Chmiel, Huta Dłutowska and Kurowice [22,23]. However, the Global Positioning System coordinates of the trap sites in these sites were different in the present study: Boginia (51 •

Ethics Statement
All trapping and experimental procedures on shrews and moles from Poland were approved by the Łódź Ethical Committee on Animal Testing (

Genetic and Phylogenetic Analyses
Partial S-and L-segment nucleotide sequences were aligned with SWSV, NVAV and other representative hantavirus sequences available on GenBank using the ClustalW in BioEdit [32]. The degree of sequence homology was assessed using pairwise comparisons [13,30]. Phylogenetic trees were constructed using the maximum-likelihood method implemented in PAUP* (Phylogenetic Analysis Using Parsimony, version 4.0a169) and MrBayes 3.1.2 [33] under the GTR+I+Γ model of evolution, as selected by using jMod-elTest version 2.1.7 [34]. Bayesian analysis consisted of 10 million Markov chain Monte Carlo generations to ensure convergence across two runs of six chains each, with average standard deviations of split frequencies less than 0.01 and effective sample sizes well over 100, resulting in consensus trees supported by posterior-node probabilities [13,30]. Each genomic segment was treated separately in phylogenetic analyses. The posterior node probabilities were based on two million generations and estimated sample sizes over 100 (implemented in MrBayes). Maximum-likelihood tress were evaluated via bootstrap analysis of 100 iterations, which was implemented in PAUP*.

Results
In analyzing RNAlater ® -preserved archival lung tissues from 320 shrews (117 Eurasian common shrews, 66 Eurasian pygmy shrews, 65 Mediterranean water shrews and 72 Eurasian water shrews) captured at multiple sites in Poland ( Figure 1A), we detected hantavirus RNA in one Eurasian common shrew from Boginia and one Eurasian pygmy shrew from Białowieża Forest (Table 1). By contrast, of the 26 European moles from Poland and 10 from Ukraine, hantavirus RNA was detected via nested RT-PCR in eight and two of them, respectively. The identities of the hantaviruses were determined via DNA sequencing and genetic and phylogenetic analysis.
Pairwise alignment and comparison of partial S-and/or L-segment sequences demonstrated ALTV strain PL7814LR56 in a Eurasian pygmy shrew from Białowieża Forest, SWSV strain PL7663JH151 in a Eurasian common shrew from Boginia and NVAV strains in European moles from Huta Dłutowska (Poland) and Lviv (Ukraine). The partial 347nucleotide L-segment sequence of SWSV strain PL7663JH151 was >98% and 100% similar at the nucleotide and amino acid level, respectively, to SWSV strains 1107 and 3334, which was previously reported from Boginia and Kurowice (Supplementary Table S1). Analyses of the partial 353-nucleotide L-segment sequences of the NVAV strains showed uniformly high similarities within strains from Poland and the two strains from Ukraine (PL7965LW001TA and PL7970LW006TA), with divergences of approximately 15% at the nucleotide level and <5% at the amino acid level between NVAV strains from both countries (Supplementary Table S1).
The 707-and 698-nucleotide S-segment sequences of NVAV strain PL7712JH226 from Poland and of NVAV strain PL7970LW006TA from Ukraine, respectively, showed 84% and 91% sequence similarity at the nucleotide and amino acid levels, respectively. Despite repeated attempts, the M-segment primers yielded no amplicons.
Phylogenetic analyses, performed by way of the maximum-likelihood method using PAUP*, showed well-supported trees (Figure 2). Similarly, Bayesian methods showed equally robust trees with a tendency toward geographic-specific clustering (Figure 3 and Supplementary Figure S1). Specifically, analyses of partial S-and L-segment sequences of SWSV indicated clustering of strains from Poland (Figures 2 and 3), whereas ALTV strain PL7814LR56 from Białowieża Forest in northeastern Poland was more distantly related to ALTV strain Smin1108 from Chmiel in southeastern Poland. This may be attributed to the non-overlapping short sequences between the two ALTV strains.    In addition, phylogenetic analyses of NVAV strains from Poland and Ukraine, including prototype NVAV from Hungary (MSB95703) and representative NVAV strains from Poland (1129, 2086, 2105, 3328, Te34), France (YA0067 and YA0088) and Belgium (Namur, Vieux-Genappe), showed segregation according to geography, despite the short sequences. The tree topologies were well-supported, with posterior node probabilities of >0.9 for the major rodent-, shrew-and mole-borne hantavirus species (Figure 3). Table 2

Discussion
Previously, we reported BOGV in Eurasian water shrews and SWSV in Eurasian common shrews and Eurasian pygmy shrews from central Poland [22]. Subsequently, we reported the co-circulation of BOGV, SWSV and NVAV in Boginia, Kurowice and Huta Dłutowska [23]. In this study, we failed to detect hantavirus RNA in most of the archival shrew tissues. Because tissues were harvested from whole carcasses stored for prolonged periods under suboptimal temperatures, poorly preserved or degraded RNA may have been contributory. Nevertheless, finding an ATLV strain in a Eurasian pygmy shrew captured in Białowieża Forest in 2004 is significant, as it represents only the second observation of such to date. Located adjacent to the border between Poland and Belarus, Białowieża Forest, particularly the Bialowieza National Park, represents the best preserved remnants of a primeval lowland mixed forest in Europe [35], retaining its primary landscape and striking biological diversity [36]. Białowieża Forest in eastern Poland is more than 400 km from Chmiel in southeastern Poland, where the previous ALTV strain (Smin1108) was detected in a Eurasian pygmy shrew captured in 2010 [24].
The detection of NVAV in European moles from Ukraine is not unexpected, as a greater than 50% prevalence of NVAV infection was reported previously in European moles in Poland [23], France [37] and Belgium [38]. The high prevalence of NVAV infection in European moles suggests efficient virus transmission and a well-established reservoir host-hantavirus relationship similar to the high prevalence of SWSV infection in Eurasian common shrews [29]. To what extent the subterranean lifestyle of European moles also facilitates NVAV transmission is unknown, but NVAV is likely to be widespread throughout the vast geographic range of the European mole, which extends from the United Kingdom to Russia. Studies are warranted to further clarify the genetic diversity and phylogeography of NVAV. In addition, the recent discovery of Academ virus in the Siberian mole (Talpa altaica) in Russia [12] and Asturias virus in the Iberian mole (Talpa occidentalis) in Spain (S.H. Gu ad R. Yanagihara, unpublished observations) would predict that other genetically distinct hantaviruses are harbored by mole species belonging to the genus Talpa. Finally, studies are warranted to map the geographic distribution and genetic diversity of other non-rodentborne hantaviruses in neighboring Ukraine and Belarus. At a minimum, the geographic range of the Eurasian common shrew, Eurasian water shrew and Eurasian pygmy shrew would predict the circulation of SWSV, BOGV and Assikala virus [39], respectively.
The pathogenic potential of NVAV and other non-rodent-borne hantaviruses to cause infection and/or disease in humans is unknown. We previously demonstrated widespread distribution of NVAV RNA in tissues of European moles [23], which is similar to that found in reservoir rodents naturally or experimentally infected with Hantaan virus [40,41] or PUUV [42]. Specifically, the demonstration of NVAV RNA in the kidneys and intestines of wild-caught European moles is consistent with viral shedding in urine and feces. However, there is no evidence of NVAV infection or disease in humans. Nevertheless, heightened awareness about the widespread distribution of NVAV and the occasional contact between humans and moles should prompt physicians to be vigilant for febrile illnesses or unusual clinical syndromes occurring among gardeners, farmers, mole catchers and field biologists, all of whom may be exposed to secretions and/or excretions of European moles or other species of moles known to harbor hantaviruses, such as the Siberian mole [12], Eastern mole (Scalopus aquaticus) [13], American shrew mole (Neurotrichus gibbsii) [43], Japanese shrew mole (Urotrichus talpoides) [44] and long-tailed mole (Scaptonyx fusicaudus) [45].
Studies showed the high prevalence of PUUV and DOBV/BGDV in reservoir rodent hosts in several regions of Poland [25,[46][47][48]. Recently, a highly divergent lineage of PUUV was demonstrated in bank voles from southern Poland [49], and the Dobrava genotype of DOBV/BGDV was isolated from yellow-necked mice captured in the Subcarpathian region [50]. Moreover, molecular evidence of the Kurkino genotype of DOBV/BGDV was found in striped field mice from Lower Silesia in southwestern Poland [25]. Despite mounting evidence of the widespread distribution of disease-causing, rodent-borne hantaviruses in Poland, HFRS is reported infrequently [51][52][53][54][55][56][57]. It is generally agreed, however, that the Subcarpathia province in the southeastern part of Poland, which borders Ukraine, is an endemic area of HFRS [56,57]. Apart from disease, epidemiological surveys show serological evidence of hantavirus infection among mammalogists [58], forestry workers [59] and hunters [60]. It is not clear if HFRS cases are being misdiagnosed and/or being under-reported because of insufficient clinician awareness or of limited access to standardized serodiagnostic tests.
As noted in previous reviews of this paradigm-shifting, medically important group of viruses [5,6], as well as in a recent historical account of Hantaviridae [61], much remains unknown, and future research is needed on many levels. For example, none of the 13 batborne loanviruses and mobatviruses have been isolated in cell culture, and except for Thottapalayam virus [62], Imjin virus [63] and NVAV [64], all of the more than 20 recently described hantaviruses harbored by shrews and moles exist only as partial or full-length sequences. In this regard, in continuing efforts to establish a more rigorous taxonomy of the family Hantaviridae, a proposal was made to require full-length S-, M-and L-segment sequences of all hantaviruses [65]. This will present obvious challenges not only for the dozens of well-known "classic" rodent-borne orthohantaviruses that have yet to be fully sequenced [65], but also for the partial hantaviral S-, M-and/or L-segment sequences detected in archival shrew, mole and bat tissues obtained from museums and field collections. Thus, for many of the non-rodent-borne putative hantaviruses which have yet to be fully sequenced, this will necessitate at least one but more probably repeated trapping expeditions for specific reservoir host species for molecular screening and subsequent whole-genome sequencing. As an obvious case in point, whole genome sequences are currently unavailable for SWSV and BOGV; thus, they would not be considered bona fide hantaviruses until each of their three segments are fully sequenced.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/v15040881/s1, Figure S1: Phylogenetic trees of newfound SWSV, NVAV and ALTV/LENV from Poland and Ukraine; Table S1: Pairwise Comparison of L-Segment Nucleotide Sequences of SWSV Strains; Table S2: GenBank Accession Numbers of Taxa for S and L trees .  Supplementary Table S1. Other presented data are available on request from the corresponding author.