To the Editor: Hantaviruses (family Bunyaviridae) are transmitted from rodent reservoirs
to humans. These viruses cause life-threatening human diseases: hantavirus cardiopulmonary
syndrome in the Americas and hemorrhagic fever with renal syndrome in Asia and Europe
(
1
). Since 2006, indigenous hantaviruses were reported also from Africa. Sangassou virus
was found in an African wood mouse (Hylomyscus simus) in Guinea (
2
). Discovery of newer African hantaviruses, Tanganya virus and recently Azagny virus,
was even more surprising because they were found in shrews (
3
,
4
).
The detection of hantaviruses in small mammals other than rodents, such as shrews
and also moles (
4
), increasingly raises questions regarding the real hantavirus host range. Bats (order
Chiroptera) are already known to harbor a broad variety of emerging pathogens, including
other bunyaviruses (
5
). Their ability to fly and social life history enable efficient pathogen maintenance,
evolution, and spread. Therefore, we conducted a study on hantaviruses in bats from
Africa.
A total of 525 tissue samples from 417 bats representing 28 genera were tested for
the presence of hantavirus RNA. Samples originated from different regions in western
and central Africa and were collected during 2009 and early 2011. Total RNA was extracted
from tissue samples and reverse transcribed. cDNA was screened by PCR specific for
sequences of the large genomic segment across the genus Hantavirus (
2
).
One sample yielded a product of the expected size and was subjected to cloning and
sequencing. The positive sample (MGB/1209) was obtained from 1 of 18 investigated
slit-faced bats (family Nycteridae). The animal was trapped at the Magboi River within
Gola National Park, Sierra Leone (7°50.194′N, 10°38.626′W), and the identification
as Nycteris hispida has been verified with the voucher specimen (RCJF529). Histologic
examination of organs of the animal showed no obvious pathologic findings.
The obtained 414-nt sequence covers a genomic region, which was found to correspond
to nt position 2,918–3,332 in the large segment open reading frame of prototypic Hantaan
virus. Bioinformatic analysis on the amino acid level showed highest degrees of identity
to shrew- and mole-associated hantaviruses (Thottapalayam virus 73.0%, Altai virus
69.7%, Nova and Imjin virus 69.3%). On the basis of tree topology of a maximum-likelihood
phylogenetic tree, the sequence does not cluster with rodent-associated hantaviruses
but groups with those found in shrews and moles (Figure).
Figure
Maximum-likelihood phylogenetic tree of MGB/1209 virus based on partial large segment
sequence (414 nt) and showing the phylogenetic placement of the novel sequence from
Nycteris spp. bat compared with hantaviruses associated (i) with shrews and moles:
Altai virus (ALTV), Artybash virus (ARTV), Asama virus (ASAV), Ash River virus (ARRV),
Azagny virus (AZGV), Camp Ripley virus (RPLV), Cao Bang virus (CBNV), Imjin virus
(MJNV), Jemez Springs virus (JMSV), Kenkeme virus (KKMV), Nova virus (NVAV), Oxbow
virus (OXBV), Seewis virus (SWSV), Tanganya virus (TGNV), Thottapalayam virus (TPMV),
and (ii) with rodents: Andes virus (ANDV), Choclo virus (CHOV), Dobrava-Belgrade virus
(DOBV), Hantaan virus (HTNV), Laguna Negra virus (LNV), Maporal virus (MAPV), Puumala
virus (PUUV), Rio Mamore virus (RIOMV), Sangassou virus (SANGV), Seoul virus (SEOV),
Serang virus (SERV), Sin Nombre virus (SNV), Soochong virus (SOOV), Tula virus (TULV),
Vladivostok virus (VLAV). The list of the accession numbers used in the analysis is
available from the authors upon request. The tree was computed by using MEGA5 (http://www.megasoftware.net).
The Tamura 3-parameter model with gamma-distributed rate heterogeneity and a proportion
of invariant sites (T92 +G + I) was selected as the best fit evolutionary model according
to the Baeysian information criterion calculated with MEGA5. The values at the tree
branches are the bootstrap support values calculated from 500 replicates. Scale bar
indicates an evolutionary distance of 0.2 substitutions per position in the sequence.
The gray areas indicate association of hantaviruses with reservoir host families.
The MGB/1209 partial sequence of the large genomic segment was deposited in GenBank
under accession no. JN037851.
Considering that bats are more closely related to shrews and moles than to rodents
(
6
), a certain genetic similarity of a putative bat-borne hantavirus with shrew- and
mole-associated hantaviruses seems reasonable. Notably, shrew-associated Thottapalayam
virus (India) and Imjin virus (South Korea) seem to be closer relatives, and African
Tanganya virus (Guinea) and Azagny virus (Côte d’Ivoire) are more distantly related.
Additional sequence data is needed for more conclusive phylogenetic analyses.
Because the new amino acid sequence is at least 22% divergent from those of other
hantaviruses, we conclude that the bat was infected with a newly found hantavirus.
We propose the putative name Magboi virus (MGBV) for the new virus because it was
detected in an animal captured at the Magboi River in Sierra Leone. The MGBV nucleotide
sequence is novel and has not been known or handled before in our laboratory. Before
this study, hantavirus nucleic acid was found in lung and kidney tissues of bats from
the genera Eptesicus and Rhinolophus in South Korea. However, nucleotide sequencing
showed the presence of prototypical Hantaan virus indicating a spillover infection
or laboratory contamination (
7
).
Further screening is necessary to confirm N.
hispida as a natural reservoir host of the virus. Although the presented bat-associated
sequence is obviously distinct from other hantaviruses, which suggests association
with a novel natural host, a spillover infection from another, yet unrecognized host
cannot be ruled out. However, detection of the virus exclusively in 1 organ (lung
but not in liver, kidney, and spleen; data not shown) suggests a persistent infection
that is typically observed in natural hosts of hantaviruses (
8
).
To date, only a few reports exist on cases of hemorrhagic fever with renal syndrome
in Africa (
9
,
10
). However, underreporting must be assumed because the symptoms resemble those of
many other febrile infections. Moreover, in cases of infections by non–rodent-associated
hantaviruses, cross-reactivity with routinely used rodent-borne virus antigens should
be limited and may hamper human serodiagnostics (
1
). The results suggest that bats, which are hosts of many emerging pathogens (
5
), may act as natural reservoirs for hantavirus. The effect of this virus on public
health remains to be determined.