Highly Divergent Genetic Variants of Soricid-Borne Altai Virus (Hantaviridae) in Eurasia Suggest Ancient Host-Switching Events

With the recent discovery of genetically distinct hantaviruses (family Hantaviridae) in shrews (order Eulipotyphla, family Soricidae), the once-conventional view that rodents (order Rodentia) served as the primordial reservoir hosts now appears improbable. The newly identified soricid-borne hantaviruses generally demonstrate well-resolved lineages organized according to host taxa and geographic origin. However, beginning in 2007, we detected sequences that did not conform to the prototypic hantaviruses associated with their soricid host species and/or geographic locations. That is, Eurasian common shrews (Sorex araneus), captured in Hungary and Russia, were found to harbor hantaviruses belonging to two separate and highly divergent lineages. We have since accumulated additional examples of these highly distinctive hantavirus sequences in the Laxmann’s shrew (Sorex caecutiens), flat-skulled shrew (Sorex roboratus) and Eurasian least shrew (Sorex minutissimus), captured at the same time and in the same location in the Sakha Republic in Far Eastern Russia. Pair-wise alignment and phylogenetic analysis of partial and full-length S-, M- and/or L-segment sequences indicate that a distinct hantavirus species related to Altai virus (ALTV), first reported in a Eurasian common shrew from Western Siberia, was being maintained in these closely related syntopic soricine shrew species. These findings suggest that genetic variants of ALTV might have resulted from ancient host-switching events with subsequent diversification within the Soricini tribe in Eurasia.


Introduction
Thottapalayam virus (TPMV), a previously unclassified virus isolated from an Asian house shrew (Suncus murinus), captured in southern India in 1964 [1], predated the discovery of Hantaan virus (HTNV), the prototype virus of hemorrhagic fever with renal syndrome (HFRS), in the striped field mouse (Apodemus agrarius) in Korea by more than a decade [2]. However, this observation went largely unnoticed and the subsequent detection of HFRS antigens in tissues of the Eurasian common shrew (Sorex araneus) and Eurasian water shrew (Neomys fodiens) in European Russia and the former Yugoslavia [3][4][5] similarly failed to incite systematic exploration into the role of shrews in the evolutionary origins of hantaviruses.
Siberian Russia, during July and August 2006. These samples are indicated in bold type in Table 1, and collection sites are shown in Figure 1. Table 1 also summarizes the prevalence of hantavirus RNA previously reported for soricine shrews from selected regions in Finland [23], Hungary [23], Poland [26] and Russia [27]. Field procedures and protocols, including trapping, euthanasia and tissue processing, were performed, following the animal care and use guidelines of the American Society of Mammalogists [33] and were approved by the Institutional Animal Care and Use Committee, of the University of New Mexico (protocol 06UNM026). Standard museum vouchers were prepared, with samples of lung frozen in liquid nitrogen for transport to the Museum of Southwestern Biology, where tissues were archived at −80 • C and associated databases are maintained to foster pathobiology research [34]. Hungary [23], Poland [26] and Russia [27]. Field procedures and protocols, including trapping, euthanasia and tissue processing, were performed, following the animal care and use guidelines of the American Society of Mammalogists [33] and were approved by the Institutional Animal Care and Use Committee, of the University of New Mexico (protocol 06UNM026). Standard museum vouchers were prepared, with samples of lung frozen in liquid nitrogen for transport to the Museum of Southwestern Biology, where tissues were archived at -80 °C and associated databases are maintained to foster pathobiology research [34]. Samples for the present study are shown in bold type.

RNA Extraction and RT-PCR Analysis
Total RNA was extracted from lung tissues, using the PureLink Micro-to-Midi total RNA purification kit (Invitrogen, San Diego, CA), then reverse transcribed, using the SuperScript III First-Strand Synthesis System (Invitrogen) with random hexamers and universal oligonucleotide primer (OSM55, 5'-TAGTAGTAGACTCC-3'), designed from the conserved 3' end of the S and L segments of hantaviruses [15,16,20,26]. Oligonucleotide primers used to amplify the S-, M-and

RNA Extraction and RT-PCR Analysis
Total RNA was extracted from lung tissues, using the PureLink Micro-to-Midi total RNA purification kit (Invitrogen, San Diego, CA, USA), then reverse transcribed, using the SuperScript III First-Strand Synthesis System (Invitrogen) with random hexamers and universal oligonucleotide primer (OSM55, 5 -TAGTAGTAGACTCC-3 ), designed from the conserved 3 end of the S and L segments of hantaviruses [15,16,20,26]. Oligonucleotide primers used to amplify the S-, M-and L-genomic segments are provided in Table S1. For the amplification of hantavirus genes, a two-step PCR was performed in 20-µL reaction mixtures, containing 250 µM dNTP, 2 mM MgCl 2 , 1 U of AmpliTaq polymerase (Roche, Basel, Switzerland) and 0.25 µM of each oligonucleotide primer [14,20,23,27]. Initial denaturation at 94 • C for 5 min was followed by two cycles each of denaturation at 94 • C for 40 s, two-degree step-down annealing from 48 to 38 • C for 40 s, and elongation at 72 • C for 1 min, then 32 cycles of denaturation at 94 • C for 40 s, annealing at 42 • C for 40 s, and elongation at 72 • C for 1 min, in a GeneAmp PCR 9700 thermal cycler (Perkin-Elmer, Waltham, MA, USA). Amplicons were separated by electrophoresis on 1.5% agarose gels and purified using the QIAQuick Gel Extraction Kit (Qiagen, Hilden, Germany). DNA was sequenced directly using an ABI Prism 377XL Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Phylogenetic Analysis
The maximum likelihood and Bayesian methods, implemented in RAxML Blackbox webserver [40] and MrBayes 3.1 [41], under the best-fit general time reversible model of nucleotide evolution with gamma-distributed rate heterogeneity and invariable sites (GTR+I+Γ) [42] and jModelTest version 0.1 [43], were used to generate phylogenetic trees. Two replicate Bayesian Metropolis-Hastings Markov Chain Monte Carlo (MCMC) runs, each consisting of six chains of 10 million generations sampled every 100 generations with a burn-in of 25,000 (25%), resulted in 150,000 trees overall. The S and L segments were treated separately in phylogenetic analyses. Topologies were evaluated by bootstrap analysis of 1000 iterations, and posterior node probabilities were based on 2 million generations and estimated sample sizes over 100 (implemented in MrBayes).

Mitochondrial DNA (mtDNA) Host Phylogeny
The taxonomic identity of the hantavirus-infected shrews was verified in genomic DNA extracted from tissues using the QIAamp DNA Mini Kit (Qiagen) and their phylogenetic relationships were studied by analysis of the complete 1140-nucleotide cytochrome b gene, amplified by PCR using well-tested primers (forward: 5 -CGAAGCTTGATATGAAAAACCAT CGTTG-3 ; and reverse: 5 -CTGGTTTACAAGACCAGAGTAAT-3 ) [44]. Host phylogenies based on mtDNA cytochrome b sequences, along with published sequences for shrews and moles for this gene region, were generated, using the maximum-likelihood and Bayesian methods described previously [45,46]. The tree was based on 3,000,000 MCMC generations, sampled every 100 generation and burn-in after 10,000 trees.

Virus Isolation
Using the previously described methods [6,47], 1% and 10% (w/v) homogenates of lung tissues from Sorex shrews, confirmed as infected with hantavirus by RT-PCR and sequencing, were inoculated onto subconfluent monolayers of Vero E6 cells (ATCC C1008 CRL-1586, American Type Culture Collection, Manassas, VA), grown in 25-cm 2 flasks and maintained with Dulbecco's minimum essential medium containing 5% fetal bovine serum. Cells were subcultured at two-to four-week intervals, at which time aliquots of cells were examined for hantavirus RNA by RT-PCR. Blind passages of cells were conducted for more than 100 days.

RT-PCR Detection of Hantavirus RNA
Hantavirus RNA was detected by RT-PCR and confirmed by DNA sequencing in lung tissues collected from 15 of 49 Laxmann's shrews, four of 12 flat-skulled shrews, one of five Eurasian least shrews, and in none of five tundra shrews and four Siberian large-toothed shrews ( Table 1). The majority of captured shrews were male (35 of 49 Laxman's shrews, eight of 12 flat-skulled shrews, and five of five Eurasian least shrews), and 17 of the 20 hantaviruses were found in male shrews, but the difference was not statistically significant (Fisher exact test value, 0.2287; p > 0.05). Of the 20 Sorex shrew-borne hantaviruses from the Sakha Republic, 14 resembled ALTV, instead of the expected host-specific hantavirus (SWSV and KKMV) ( Table 2). several Sorex caecutiens as ARTV. However, recent analysis has shown that ARTV is a genetic variant of SWSV [31]. Accordingly, SWSV is used in Table 2 and Figures 2, 3 and 4. Table 2 summarizes the S-, M-and L-segment hantavirus sequences obtained for each of the 20 hantavirus-infected Sorex shrews captured at three localities in the Sakha Republic. As noted earlier, hantavirus RNA was found predominantly in male shrews, with evidence of infection in only three female shrews. Pair-wise alignment and comparison of the full-length and partial S-, M-and L-segment sequences from Laxmann's shrews and flat-skulled shrews showed three genetically distinct hantaviruses. The first was Kenkeme virus (KKMV) strain MSB148794, previously reported from the flat-skulled shrew [14]. The second was Artybash virus (ARTV), harbored by the Laxmann's shrew [20]. However, based on recent analysis, ARTV should be called SWSV [31]. In addition, the third was a newfound hantavirus, differing by approximately 40% at both the nucleotide and amino acid levels from SWSV and KKMV, and resembling most closely ALTV, previously detected in a Eurasian common shrew from Western Siberia [27].

Genetic Analysis
The full-length 1287-nucleotide S-genomic segment of SWSV and KKMV from the Laxmann's shrew and flat-skulled shrew, respectively, encoded a 428-amino acid nucleocapsid (N) protein and lacked the additional open reading frame encoding a nonstructural NSs protein, as determined by sequence alignment with cricetid rodent-borne orthohantaviruses. Nearly full-length S-segment sequences of 1164 to 1209 nucleotides, obtained for ALTV-like hantaviruses from three Laxmann's shrews and one flat-skulled shrew, showed 51.1-53.9% and 46.4-47.9% sequence similarity at the nucleotide and amino acid levels, respectively, with KKMV and SWSV ( Figure 2). By contrast, these ALTV-like hantaviruses exhibited amino acid sequence similarity of 98.8-100% (numbers in red) among themselves and approximately 95% similarity with prototype ALTV Telet-Sa302 ( Figure 2). Partial M-segment sequences, amplified from three Laxmann's shrews (Sca370/MSB148580, Sca377/MSB148458 and Sca402/MSB148793) and a flat-skulled shrew (Sr424/MSB148679), also showed low level nucleotide sequence similarity of 61.3-65.6% with KKMV. By contrast, they exhibited 90.0-91.1% nucleotide sequence similarity and LENV Khekhtsir-Sc67 (which served as a surrogate for prototype ALTV Telet-Sa302 because M segment sequences were unavailable).  90% similarity with prototype ALTV Telet-Sa302 ( Figure 3). Thus, there was overall congruence for each genomic segment, suggesting the maintenance and co-circulation of a separate genetic lineage of hantavirus in Sorex shrews in the Sakha Republic.

Recombination Analysis
RDP4 Beta 4.36 failed to disclose any consistent evidence of recombination in the S-and L-genomic segments. Although separate regions of potential recombination were found in a few instances, there was no consistency or concordance between the detection methods, calling into question the validity of the identified sequences or the biological significance of recombination versus general heterogeneity in sequence evolutionary rates.

Phylogenetic Analysis
Phylogenetic trees, based on the coding regions of the full-length and partial S and L segments, and partial M segment, revealed similar topologies using the maximum-likelihood and Bayesian methods. Hantavirus sequences from five Laxmann's shrews (Sca371/MSB148558, Sca372/MSB148559, Sca375/MSB148436, Sca376/MSB148457 and Sca383/MSB148347) and one flat-skulled shrew (Sr422/MSB148794) segregated into separate clades with SWSV and KKMV, respectively, while the other full-length and partial S-, M-and/or L-genomic sequences from 10 Laxmann's shrews (Sca363/MSB146482, Sca370/MSB148580, Sca377/MSB148458, Sca380/MSB148573, Sca381/MSB148574, Sca382/MSB148575, Sca393/MSB148840, Sca396/MSB148745, Sca402/MSB148793    Table 2 summarizes the S-, M-and L-segment hantavirus sequences obtained for each of the 20 hantavirus-infected Sorex shrews captured at three localities in the Sakha Republic. As noted earlier, hantavirus RNA was found predominantly in male shrews, with evidence of infection in only three female shrews. Pair-wise alignment and comparison of the full-length and partial S-, M-and L-segment sequences from Laxmann's shrews and flat-skulled shrews showed three genetically distinct hantaviruses. The first was Kenkeme virus (KKMV) strain MSB148794, previously reported from the flat-skulled shrew [14]. The second was Artybash virus (ARTV), harbored by the Laxmann's shrew [20]. However, based on recent analysis, ARTV should be called SWSV [31]. In addition, the third was a newfound hantavirus, differing by approximately 40% at both the nucleotide and amino acid levels from SWSV and KKMV, and resembling most closely ALTV, previously detected in a Eurasian common shrew from Western Siberia [27].

Recombination Analysis
RDP4 Beta 4.36 failed to disclose any consistent evidence of recombination in the S-and L-genomic segments. Although separate regions of potential recombination were found in a few instances, there was no consistency or concordance between the detection methods, calling into question the validity of the identified sequences or the biological significance of recombination versus general heterogeneity in sequence evolutionary rates.

Host Phylogeny Analysis
The identity of each hantavirus-infected Sorex shrew was molecularly confirmed by amplification and sequencing of the full-length 1140-nucleotide cytochrome b gene. Phylogenetic trees revealed well-supported lineages according to species.

Virus Isolation
Despite using well-established protocols involving multiple blind passages in Vero E6 cells for 100 days or longer, repeated attempts to isolate these highly divergent lineages of hantaviruses from archival frozen shrew tissues were unsuccessful.

Discussion
In 2007, we first detected highly divergent hantavirus sequences in Eurasian common shrews captured in Russia and Hungary. A partial L-segment sequence, named ALTV, was deposited in GenBank, but we were unwilling to report this until an expanded database was assembled. In this report, we demonstrate through genetic and phylogenetic analyses of full-length and near full-length sequences of the L-and S-genomic segments, as well as partial M-segment sequences, that genetic variants of ALTV are harbored and are being maintained in Laxmann's shrews, flat-skulled shrews and Eurasian least shrews in the Sakha Republic in Far Eastern Russia, as well as in Eurasian common shrews in Western Siberia and Hungary. Importantly, these multiple strains of ALTV-like hantavirus are circulating simultaneously with the prototypic host-specific soricine shrew-borne orthohantaviruses. That is, Eurasian common shrews and Laxmann's shrews harbor SWSV and flat-skulled shrews host KKMV, as well as ALTV-related hantaviruses in the same locality and at the same time.
Apart from the Eurasian common shrew, the Laxmann's shrew and the flat-skulled shrew, ALTV-related hantavirus sequences were also found in a Eurasian pygmy shrew, captured in Chmiel, in southeastern Poland. The Eurasian pygmy shrew, which is the principal reservoir host of Asikkala virus in Finland [24] and the Czech Republic [48], has also been demonstrated to harbor SWSV in Germany [22], the Czech Republic [22], Finland [24] and Poland [17,26]. This is yet another example of a soricid species serving as the host of more than one hantavirus species. Moreover, the hosting of the same hantavirus species by multiple closely related rodent, soricid, talpid and bat species represents a stark departure from the previously held notion of one-rodent species harboring a single hantavirus species. Thus, the primary reservoir host of ALTV, which has been detected in Eurasian common shrews, Laxmann's shrews and flat-skulled shrews at high frequency, would be merely speculative. However, the Laxmann's shrew would be a tentative guess, based on its vast geographic range, extending from Sweden, Finland and Belarus across western Siberia and Mongolia to Far Eastern Russia and China.
Although the gender-specific prevalence of hantavirus infection was not statistically significant among soricine shrews in the Sakha Republic, there was a tendency toward a male predominance. This observation is consistent with the reported overrepresentation of hantavirus infection in male shrews and rodents, including the Ussuri white-toothed shrew [6], Norway rat [49], deer mouse [50,51], and marsh rice rat [52].
Our overall findings confirm those of Ling and colleagues, who previously reported the co-existence of two genetically distinct hantavirus species-SWSV and ALTV-like hantavirus-circulating simultaneously in a single host species, the Eurasian common shrew, in Finland [24]. Their comparative analysis, showing two ALTV-like hantavirus strains (designated Uurainen/63L and Lohja/EWS10L) amplified from Eurasian common shrews, was inferred from the 340-nucleotide L segment of ALTV strain Telet-Sa302 we originally deposited in January 2008 (GenBank EU424341). The phylogenetic analyses of 17 ALTV-like hantavirus variants reported here are now based on the near full-length L segment (4997 nucleotides) of ALTV strain Telet-Sa302. Importantly, the collective data emphasize the widespread geographic distribution and host diversity of ALTV-related hantaviruses, further enriching the complexity of hantavirus evolution and phylogeography [8,9].
The changing global landscape of hantaviruses has prompted a re-examination of previously long-held dogma about their host range, evolutionary origins and phylogeography [8,9]. Based on the rapidly expanding literature of newfound hantaviruses, it is likely that many more hantaviruses will be discovered, possibly in hosts belonging to other taxonomic orders and in unanticipated geographic regions [53]. Moreover, textbook chapters on hantaviruses are being revised and re-written, as more information becomes available about the emergence and pathogenic potential of non-rodent-borne hantaviruses [53,54]. However, despite these advances, some of the persistent uncertainties and conundrums in hantavirus research are the direct consequence of the lack of full-length genomes and the dearth of hantavirus isolates. That is, although referred to as novel viruses, nearly all of the hantaviruses newly identified in shrews, moles and bats exist only as viral sequences.
The isolation of hantaviruses is fraught with difficulty, even from freshly collected tissues. Thus, while disappointing, it is not altogether surprising that we failed to isolate ALTV-like hantaviruses from frozen archival tissues stored since 2006. To date, only three hantaviruses have been isolated from non-rodent hosts: TPMV from the Asian house shrew [1]; MJNV from the Ussuri white-toothed shrew [9]; and Nova virus from the European mole [47]. More innovative approaches are urgently needed to isolate hantaviruses, including the establishment of cell lines from tissues of reservoir hosts, the engineering of cells with specific virus-entry receptors, and the development of three-dimensional organoid cultures. Until such time that ALTV, ALTV-like hantaviruses and other non-rodent-borne hantaviruses are isolated and propagated in culture, their biology and pathogenic potential will remain speculative at best. Thus, the road ahead is laden with challenges, but also endless opportunities and unlimited possibilities. Above all, strong partnerships between healthcare providers, public health workers, veterinarians, ecologists, museum curators and pathologists will be vital for the identification and rapid diagnosis of previously unrecognized infectious diseases, caused by newfound hantaviruses.

Conclusions
Eurasian common shrews, Laxmann's shrews and flat-skulled shrews, captured at the same time and in the same location in Hungary and Russia, were each found to harbor hantaviruses belonging to two separate and highly divergent lineages. Pair-wise alignment and phylogenetic analysis of partial and full-length S-, M-and/or L-segment sequences indicated the co-existence and maintenance of two distinct hantavirus lineages in these closely related syntopic soricine shrew species. These findings suggest possible ancient host-switching events from another reservoir. Although ALTV was originally detected in the Eurasian common shrew and ALTV-like hantaviruses have been found in Eurasian common shrews from Finland and Hungary, the primary reservoir host of ALTV is unknown. However, since the vast geographic distribution of ALTV-like hantaviruses coincides with the geographic range of the Laxmann's shrew, this shrew species is the likely candidate. Alternatively, ALTV might represent an ancestral hantavirus lineage that may have subsequently diversified within the Soricini tribe in Eurasia, based on being detected in Sorex araneus, Sorex caecutiens, Sorex minutissimus, Sorex minutus and Sorex roboratus. Co-circulation of hantaviruses in the same host species also raises the distinct possibility of co-infection and reassortment as a mechanism for rapid evolutionary change [25,[55][56][57][58][59]. Our analysis did not show evidence of recombination, but the possibility of reassortment would exist if shrews were infected concurrently with KKMV and ALTV or SWSV and ALTV. Finally, our data did not allow definitive classification of ALTV and ALTV-like hantaviruses into one of the existing four genera of the subfamily Mammantavirinae. Whether this means there might be a fifth genus warrants further intensive investigation.
Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4915/11/9/857/s1. Table S1: Oligonucleotide primers for amplification of SWSV, ARTV, KKMV and ALTV. Figure S1: Phylogenetic tree, based on M-segment sequences of ALTV-like hantaviruses. field work in remote locations in the Sakha Republic. We also acknowledge Janusz Markowski and Janusz Hejduk, who supplied shrew tissues from Poland, reported previously. The funding agencies had no role in study design, data collection and analysis, or preparation of the manuscript.