A previously long-unclassified virus, isolated from an Asian house shrew (Suncus murinus
) captured in Thottapalayam, near Vellore in Tamil Nadu, India, in 1964,1
is now known to be a member of the Hantavirus
genus (Family Bunyaviridae
) by virtue of its morphological characteristics and overall genomic organization as well as its antigenic and evolutionary relationship with well-characterized rodent-borne hantaviruses.2,3
The assumption that Thottapalayam virus (TPMV) represented a spillover event from a rodent reservoir has been thoroughly discredited by full-genome analysis, showing strong support for an ancient non-rodent host origin and an early evolutionary divergence from rodent-associated hantaviruses.3,4
Moreover, the realization that shrews of multiple species (Order Soricomorpha, Family Soricidae
), including the Therese's shrew (Crocidura theresae
Ussuri white-toothed shrew (Crocidura lasiura
northern short-tailed shrew (Blarina brevicauda
Chinese mole shrew (Anourosorex squamipes
Eurasian common shrew (Sorex araneus
masked shrew (Sorex cinereus
dusky shrew (Sorex monticolus
and flat-skulled shrew (Sorex roboratus
harbor hantaviruses that are far more genetically diverse than those viruses carried by rodents suggests that shrews, being evolutionarily more ancient than rodents, may have served as the primordial hosts of ancestral hantaviruses.
Although the isolation of TPMV predates the isolation of all other hantaviruses, including Hantaan virus (HTNV),12
the prototype virus of hemorrhagic fever with renal syndrome (HFRS),13
data about its genetic diversity and phylogeography are unavailable. To address this gap in knowledge, we analyzed tissues originally collected for studies on the prevalence and genetic diversity of bartonellae.14
Archival lung and kidney tissues from 134 Asian house shrews, trapped around densely populated human habitations in Bhaktapur (Pode Tole, Bhelkhel, and Bhelukhel), Kathmandu (Kalimati, Lagan Tole, Sabal Bahal, Teku, Hyumata, and Inakha Tole), and Lalitpur (Bakhar, Bakhar Thati, Lagankhel, and Patan) districts of Nepal () from January 31 to April 10, 1996 and July 22 to September 3, 1996, were tested for hantavirus RNA by reverse transcription polymerase chain reaction (RT-PCR) ().
Map of Nepal showing metropolitan centers (black dots) in Kathmandu, Bhaktapur, and Lalitpur districts where Asian house shrews were captured in 1996.
Prevalence of TPMV infection in Asian house shrews captured in Nepal in 1996
Initially, total RNA was extracted from frozen tissues using the PureLink Micro-to-Midi Total RNA Purification Kit (Invitrogen, San Diego, CA), and cDNA was synthesized using the SuperScript III First-Strand Synthesis System (Invitrogen) and random hexamers as well as a universal primer (OSM55: 5′-TAGTAGTAGACTCC-3′) designed from the genus-specific conserved 3′ end of the S, M, and L segments of all hantaviruses. Oligonucleotide primers designed for amplification and sequencing of hantaviruses are shown in . PCR was performed in 20-μL reaction mixtures containing 250 μM deoxyribonucleotide triphosphate, 2 mM MgCl2
, 1 U AmpliTaq polymerase (Roche, Basel, Switzerland), and 0.25 μM each primer. Initial denaturation was at 94°C for 5 minutes followed by two cycles each of denaturation at 94°C for 40 seconds, 2° step-down annealing from 48°C to 38°C for 40 seconds, elongation at 72°C for 1 minute, 32 cycles of denaturation at 94°C for 40 seconds, annealing at 42°C for 40 seconds, and elongation at 72°C for 1 minute in a GeneAmp PCR 9700 thermal cycler (Perkin- Elmer, Waltham, MA). Amplicons were separated by electrophoresis on 1.5% agarose gels and purified using the QIAQuick Gel Extraction Kit (Qiagen, Hilden, Germany), and DNA was sequenced directly using an ABI Prism 377XL Genetic Analyzer (Applied Biosystems, Foster City, CA).8
Oligonucleotide primers for amplification and sequencing of TPMV
As determined by RT-PCR and sequencing, TPMV RNA was detected in lung tissues of 12 Asian house shrews ( and ). Of these samples, TPMV RNA was detected in kidney tissues of only four shrews, suggesting lower viral burden; 11 of 12 TPMV-infected shrews were captured in the Kathmandu district. Overall prevalence of TPMV infection was essentially the same during the January to April (6/72 or 8.3%) and the July to September (6/62 or 9.7%) time periods in 1996 ().
TPMV strains detected in lung tissues of Asian house shrews captured in Nepal
Although studies comparing the prevalence of hantavirus infection using serologic and genetic assays are few, hantavirus RNA is frequently undetectable in tissues of rodents and shrews, when their sera contain anti-hantavirus immunoglobulin G (IgG) antibody,15,16
possibly because of the insensitivity of RT-PCR or the failure of some rodents and shrews to develop persistent infections. In this study, shrew sera were not available for anti-TPMV antibody testing. Therefore, the frequency of TPMV infection, as determined by RT-PCR, likely represents the minimum prevalence. Conceivably, if serologic evidence of infection had been used instead, the prevalence of TPMV infection might be much higher than the 8.3% or 9.7% noted above.
Pair-wise alignment and analysis of 359–1,199 nucleotides of the nucleocapsid-encoding S segment, 377–404 nucleotides of the envelope glycoprotein-encoding M segment, and 1,436–2,484 nucleotides of the RNA-dependent RNA polymerase-encoding L segment revealed that TPMV strains from Asian house shrews, captured in two adjacent districts in Nepal, exhibited very low nucleotide sequence variation (S = 0–1.7%, M = 0%; L = 0.1–0.7%), but these samples differed in all three segments from the prototype TPMV strain VRC66412 from India by approximately 20% (). However, the amino acid sequences of the encoded proteins were highly conserved, with 2.8–5.3% variation between TPMV strains from Nepal and India. Compared with soricine shrew-borne hantaviruses (shown in ), the genetic distances of TPMV strains from Nepal were approximately 30–50% at the nucleotide and amino acid levels for each segment.
Sequence similarities of the S, M, and L segments of TPMV strain VRC66412 and TPMV strains from Asian house shrews captured in Nepal
Figure 2. Phylogenetic trees generated by the ML and Bayesian methods under the best-fit GTR+I+Γ model of evolution based on partial sequences of the S-, M-, and L-genomic segments of TPMV strains from Nepal and other well-characterized hantaviruses. The (more ...)
To determine the molecular diversity of TPMV strains, DnaSP 5.1017
was used for haplotype diversity, nucleotide diversity, Tajima's D test, and Fu and Li's D and F tests (). The level of diversity in TPMV strains was significantly lower than the diversity of all other hantaviruses in three segments. TPMV strains gave negative values for the Tajima's D and Fu and Li's D and F tests, and most of the values were not statistically significant except for the S segment.
Neutrality tests and haplotype and nucleotide diversity analyses of TPMV
Maximum likelihood (ML) and Bayesian methods, implemented in Phylogenetic Analysis Using Parsimony, 4.0b10 (PAUP*),18
RAxML Blackbox webserver,19
and MrBayes 3.120
under the best-fit GTR+I+Γ model of evolution using jModeltest 0.1.1,21
were used to generate phylogenetic trees. Two complementary and not redundant ML methods were used to gain additional analytical rigor; RAxML is a very useful and powerful method for doing rapid ML bootstrap analysis, but because RAxML is a relatively new method, the topologies were backed up with PAUP*, which is more traditional and well-accepted in the field. Parameters were reestimated during successive rounds of ML heuristic searches using the tree bisection reconnection (TBR) and subtree-pruning-regrafting (SPR) algorithms implemented in PAUP*.22
Nearly identical tree topologies, well-supported by bootstrap analysis (> 70%) and posterior node probabilities (> 0.70), were estimated from analysis of the partial S-, M-, and L-segment sequences of TPMV from Nepal, which formed a monophyletic group with the prototype TPMV strain from India ().
Phylogenetic networks of each genomic segment, as determined by the median-joining algorithm using Network 4.6,23
indicated clear division between TPMV and Imjin virus (MJNV) strains (). Although TPMV strains from Nepal showed closer network relationships with prototype TPMV (VRC66412) from India, they clearly formed separate branches.
Figure 3. Phylogenetic networks based on partial nucleotide (nt) and amino acid (aa) sequences of the S, M, and L segments constructed by the median-joining algorithm using Network 4.6. The numbers on the branches are proportional to the number of allele differences, (more ...)
In this regard, phylogenetic segregation of TPMV sequences according to the geographic origin of the reservoir host species resembled that shown recently for Seewis virus in Sorex araneus24,25
and MJNV in Crocidura lasiura
These data were also consistent with the geographic-specific lineages shown previously for rodent-borne hantaviruses, including HTNV in Apodemus agrarius
Soochong virus in Apodemus peninsulae
Puumala virus in Myodes glareolus
Muju virus in Myodes regulus
and Tula virus in Microtus arvalis
Overall, these data were suggestive of long-standing associations between hantaviruses and their reservoir rodent and soricid hosts, but the basis for these relationships through evolutionary time is unclear.
A distinct weakness of this opportunistic study is that our analysis was based on a very limited number of TPMV strains from a restricted geographic region in Nepal. Phylogeographic studies of TPMV throughout the vast geographic range of the Asian house shrew are necessary to obtain clearer insights into the biogeographic origin and radiation of hantaviruses and their soricid hosts as well as to clarify whether other crocidurine shrews in Asia harbor TPMV or TPMV-related hantaviruses.
Molecular identification of TPMV-infected shrews was determined by amplification and sequencing of the entire 1,140-base pair cytochrome b
gene of mitochondrial DNA using previously described universal primers.22
All TPMV-infected shrews from Nepal were confirmed as Asian house shrews by phylogenetic analysis of cytochrome b sequences of multiple shrew species available in GenBank.32
This seemingly trivial finding firmly established that the Asian house shrew is the natural reservoir of TPMV and dismissed lingering doubts that another shrew or rodent species might be involved. That said, the issue of host sharing with other crocidurine shrew species is now being pursued by testing tissues from the Anderson's shrew (Suncus stoliczkanus
), which is found near watercourses and grassy embankments at lower elevations, occasionally sharing habitats with the Asian house shrew in Nepal.
Besides Nepal, the highly adaptable Asian house shrew is widely distributed throughout Afghanistan, Bangladesh, Bhutan, Brunei Darussalam, Cambodia, China, India, Indonesia, Lao People's Democratic Republic, Malaysia, Myanmar, Pakistan, Singapore, Sri Lanka, Taiwan, Thailand, and Viet Nam. Possibly because their diets consist mostly of insects, Asian house shrews may have been intentionally introduced by humans as a possible means of vector control into Africa (Egypt, Eritrea, Kenya, Republic of Djibouti, Rwanda, Sudan, and Tanzania), the Middle East (Iraq, Kingdom of Bahrain, Kuwait, Saudi Arabia, Sultanate of Oman, and Yemen), islands within the Indian Ocean (Comoros, Madagascar, Republic of Mauritius, and Réunion), and Asia and the Pacific (Japan, Guam, and Philippines). Although typically residing near human habitation, Asian house shrews have been found in diverse environments ranging from natural forests, scrubland, and grasslands to plantations, pasture, cultivated fields, and altitudes exceeding 2,000 m.
Nepal is a landlocked South Asian country sharing borders to the north with the People's Republic of China and to the south, east, and west with India. Of 75 districts, the most densely populated is Kathmandu District, with the population exceeding 1.5 million people. Studies of febrile patients in Kathmandu show an unrelentingly heavy disease burden of enteric fever (caused by Salmonella
serovars Typhi and Paratyphi A), streptococcal pneumonia, dengue fever, murine typhus, scrub typhus, and leptospirosis.33–35
Limited and inadequate diagnostic facilities, however, still leave many febrile illnesses without a definitive microbial etiology. Although human exposure to peridomestic Asian house shrews presumably occurs with some regularity, the frequency of TPMV infection in humans is unknown. The recent detection of IgG antibodies against TPMV by enzyme immunoassay, Western blot analysis, and focus-reduction neutralization test in a febrile Laotian patient hospitalized in Thailand36
is suggestive of past infection, but it falls short of showing disease causation. Intensified efforts, now underway, to test sera from residents in neighborhoods in Kathmandu, where TPMV-infected Asian house shrews have been captured, might clarify if TPMV accounts for undiagnosed febrile diseases.