Figure 1

Figure 1. Sites at which Mastomys natalensis rodents were captured in Nigeria during January 2011–March 2013. Red circles represent sites within the Lassa fever–endemic zone; black circles represent sites outside the Lassa fever–endemic...

From January 2011 through March 2013, small mammals were captured in H.B. Sherman live animal traps (https://www.shermantraps.com/) at 8 sites across Lassa fever–endemic and –nonendemic areas in Nigeria (Figure 1). We classified Lassa fever–nonendemic areas as areas where no cases of Lassa fever have been documented (2). Permission to trap rodents in various localities was granted by the Ministry of Environment, Osun State; Gwer West Local Government Council, Benue State; and the Ministry of Health, Taraba State.

Among 782 small mammals, 274 M. natalensis rodents were trapped. Identification of the animals in the field was based on external morphology and later confirmed genetically by cytochrome b gene sequencing. The rodents were euthanized, and biopsy samples (blood, liver, kidneys, spleen) were collected for laboratory analyses. Precautions for working with animals potentially infected with dangerous pathogens were strictly followed (11).

Using a QIAamp Viral RNA Mini Kit (QIAGEN, Valencia, CA, USA), we extracted total RNA from 20 μL of whole blood frozen at −80°C. Extracted RNA was tested with a panarenavirus protocol designed to amplify the L (polymerase) gene (340 nt) (12) and with another reverse transcription PCR specific for LASV, selective for the glycoprotein precursor (GPC) gene (303 nt) (13). We conducted further PCR amplification of the GPC fragment (using primers in Technical Appendix 1 Table) for specimens positive on initial screening. Phylogenies were inferred by use of the Bayesian Markov Chain Monte Carlo method implemented in BEAST version 1.6.2 (http://beast.bio.ed.ac.uk/).

Figure 2

Figure 2. A) Phylogenetic analyses of glycoprotein precursor gene (GPC) of Old World arenaviruses and cytochrome b sequences of 16 arenavirus-positive Mastomys natalensis rodents captured in Nigeria during January 2011–March 2013 (boldface). The...

Of the 274 M. natalensis rodents from the 8 sampled sites, 16 were positive by PCR for arenavirus (Figure 1; Table). Phylogenetic analyses of the GPC and L gene sequences showed that 15 of the viruses were Lassa and 1 was a Mobala-like virus (Figure 2). The LASV sequences from Ekpoma and Eguare-Egoro belonged to lineage II and clustered with strains Nig08-A4, A37, A41, and A47 from patients in Edo State (14). Nucleotide identities between the sequences of LASV from the rodents and those from the human patients were 82%–96% (GPC) and 85%–97% (L), and amino acid identities were 95%–99% (GPC) and 93%–100% (L), respectively (Technical Appendix 2). The sequence from Mayo Ranewo was conversely more distant from those from Mobala and Gairo; nucleotides identities were 69%–73% (GPC) and 77%–80% (L), and amino acid identities were 75%–77% (GPC) and 90%–95% (L), respectively (Technical Appendix 2).

Sequence analysis of the region coding cytochrome b indicated that M. natalensis rodents from Nigeria cluster in 2 clades. The first clade corresponds to phylogroup A-I, which clusters with sequences from rodents from Benin, which is west of Nigeria (Figures 1, 2; Technical Appendix 2). Phylogroup A-I, including sequences from Abagboro, Esira, Kako, Eguare-Egoro, Ekpoma from western Nigeria, extends across the Niger and Benue Rivers into Onmba-Abena in eastern Nigeria.

The second clade corresponds to phylogroup A-II, which clustered with sequences from Cameroon, which is east of Nigeria (Figures 1, 2; Technical Appendix 1 Figure). Phylogroup A-II within Nigeria is represented by M. natalensis rodents from Ngel-Nyaki, Mayo-Ranewo, and Onmba-Abena in eastern Nigeria, but this phylogroup also overlaps the Niger and Benue Rivers westward into Eguare-Egoro and Ekpoma. The contact zone between rodents of phylogroups A-I and A-II in Nigeria was detected at sites relatively close to the Niger and Benue Rivers (Eguare-Egoro, Ekpoma, Onmba-Abena) (Figure 1). The Niger River has been demonstrated to be a natural barrier for some rodents (15) but seems to delimit these 2 phylogroups only to an extent. Human-assisted long-distance migration of commensal rodents could influence their genetic structure, which may be what happened for rodents of the same M. natalensis phylogroup that were detected on opposite banks of the Niger River.

M. natalensis phylogroup A-I rodents were infected with LASV in Eguare-Egoro and Ekpoma but not in Abagboro, Kako, and Esira. Because all rodents from these sites belong to the same phylogroup, some factor other than cytochrome b genetic structure might be responsible for the focal prevalence of LASV. It could be, however, that our study was limited by use of the cytochrome b mitochondrial marker only, which is maternally inherited. Therefore, other biparentally inherited genetic markers, such as microsatellites, should be investigated. Environmental variables such as humidity and temperature could also be considered (2).

M. natalensis phylogroup A-II rodents were infected by LASV and a Mobala-like virus. We did not detect any LASV-positive, phylogroup A-II rodents east of the Niger River, although all the sites sampled in this area lie within the Lassa fever–endemic zone and regularly experience epidemics (2). It is worth exploring the possibility that other small mammals might also host LASV. LASV-positive members of phylogroup A-II, however, were found on the west bank of the Niger River in Eguare-Egoro and Ekpoma (along with LASV-positive members of phylogroup A-I). A crucial implication of these findings is the potential that new, previously naive populations and phylogroups of M. natalensis rodents could become infected with LASV and the disease could emerge in new regions in western Africa.

Detection of the Mobala-like virus in M. natalensis rodents within Mayo-Ranewo in eastern Nigeria deserves further study. We included Mayo-Ranewo among our survey sites because an epidemic of hemorrhagic fever, considered but not confirmed to be Lassa fever, occurred there in 2012. Whether the Mobala-like arenavirus detected in this village has pathogenic properties remains to be determined.