Simian T-Lymphotropic Virus Diversity among Nonhuman Primates, Cameroon

Broad virus diversity warrants active monitoring for cross-species transmission and highlights the risk for human disease.

We used a hunter-based fi eld surveillance approach to investigate STLV diversity among primate bushmeat samples collected from 12 NHP species in different locations in Cameroon. We also sampled NHPs in the surrounding region for the STLV source of the HTLV-4-infected individual. In addition, we examined the utility of using dried blood spots (DBS) in the fi eld for surveillance of crossspecies transmission of retroviruses.

Sample Collection and Preparation
Before the study began, Institutional Animal Care and Use Committee approvals were obtained. Self-identifi ed hunters from 4 study sites in southern Cameroon volunteered to collect DBS from freshly hunted NHP bushmeat ( Figure 1). Hunters were educated about the risks associated with direct contact with NHPs and about appropriate prevention measures. Preliminary identifi cation of hunted species was undertaken by using pictographs of NHPs common in the region (24). Confi rmation of species was performed by analysis of mitochondrial cytochrome oxidase subunit II and/or glucose-6-phosphate dehydrogenase sequences (25,26). Over 2 years, a total of 362 DBS from hunted NHPs were collected on Whatman fi lter paper (Kent, UK), air dried, and stored locally at room temperature in envelopes with dessicant until processed. Nucleic acids were extracted by using the NucliSens nucleic acid isolation kits (bioMérieux, Durham, NC, USA). DNA quality and yield were determined by semiquantitative PCR amplifi cation of the β-actin gene as previously described (27,28). DNA preparation and PCR assays were performed in different laboratories specifi cally outfi tted for processing and testing of NHP samples only, according to established precautions to prevent contamination. Specimens were coded by using a strategy previously described (15).

PTLV Sequence Detection and Sequence Analysis
DNA samples from NHPs were tested for tax sequences by using generic and nested PCR assays capable of detecting viruses from all 4 major PTLV groups (7,19,27). Phylogenetic resolution was achieved by analysis of long terminal repeat (LTR) sequences using PTLV group-specifi c primers (7). PCR amplifi cation of overlapping regions of the 5′ and 3′ STLV-1 LTR (4) and partial STLV-3 LTR (7,19) sequences were performed using primers and conditions reported elsewhere. A PCR-based genome-walking approach (15) was used to obtain partial viral genome fragments of a highly divergent PTLV from monkeys Cmo8699AB and Cni7867AB (Table 1). (NHPs are coded as follows: the fi rst letter of the genus is followed by the fi rst 2 letters of the species name, e.g., Cag, Cercocebus agilis; Cni, C. nictitans; Cmo, C. mona; and Lal, Lophocebus albigena. The last 2 letters in the code indicate the study site, e.g., AB, Abat; MV, Mvangan.) To screen humans for the divergent STLV-3 subtype, we developed a nested PCR assay based on STLV-3 (Cmo8699AB) tax sequences. Similar strategies have been used to screen for the novel HTLV-3 and HTLV-4 viruses in NHP hunters from Cameroon (1,7). DNA for PCR test- STLV-1 (f) HTLV-4 Figure 1. Distribution of primate T-lymphotropic viruses identifi ed in humans and nonhuman primates from rural villages and forests in southern Cameroon. Colored circles and diamonds correspond to human (HTLVs) and simian T-lymphotropic viruses (STLVs) (subtypes), respectively, found at each study site in the current study and reported previously (7). Shaded triangles indicate approximate sampling sites where STLV-3-like strains have been reported by others (9 ing was available from a previous study in which plasma from 63 hunters showed a range of seroreactivity to HTLV antigens by Western blot (WB; Genelabs Diagnostics 2.4 kit [7]). WB profi les were HTLV-1-like (n = 2), HTLV-2-like (n = 4), HTLV-positive but untypeable (n = 8), and HTLVindeterminate (n = 49) (7). New STLV-3 (Cmo8699AB)-tax specifi c primers were designed to screen peripheral blood mononuclear cell DNA from all 63 hunters previously negative for sequences by using generic primers that can detect PTLV-1, PTLV-2, PTLV-3, and PTLV-4 (7).The assay could reliably detect 10 copies of STLV-3 (Cmo8699AB) tax plasmid sequences in a background of human DNA. STLV-3 (Cmo8699AB) tax sequences were not amplifi ed from PTLV-1, PTLV-2, PTLV-3, and HTLV-4 cell line DNA or tax-containing plasmid DNA or from HTLV nonreactive blood donor DNA samples (data not shown), which shows the high sensitivity and specifi city of the assay.

Results
A total of 362 DBS representing 12 NHP and prosimian species were collected (Figure 1), of which 215 (60%) were of adequate quality and quantity for nucleic acid extraction, and 170 (79%) of the 215 yielded adequate amplifi able DNA (Table 2). Blood clots and limited volumes of blood on some DBS accounted for poor DNA yield of some samples.
Because of the limited amount of DBS material available, we used a PCR assay that detects sequences from all 4 major PTLV groups. We observed a broad range of PTLV diversity over a wide geographic distribution. Of the 170 samples screened, 12 (7%) from 4 NHP species were positive for PTLV tax sequences (Table 3). Phylogenetic analysis of the short tax sequences from these 12 samples showed that 7 NHPs (2 Cercocebus agilis and 5 C. nictitans monkeys) were infected with STLV-1 and that 3 (C. agilis, C. nictitans, and Lophocebus albigena monkeys) were infected with STLV-3 ( Figure 2; Table 3). We did not fi nd any evidence of STLV-2, HTLV-4-like STLV, or dual STLV-1 and STLV-3 infections as have been found in C. agilis monkeys in other studies (25).

Phylogenetic Resolution of a Novel PTLV-3 Subtype
The identifi cation of highly divergent STLV-3-like sequences in Cmo8699AB and Cni7867AB was investigated further by additional analyses of a larger tax sequence (1,015 bp). Both tax sequences were nearly identical (99.9%) despite nucleic acid extraction, PCR amplifi cation, and sequencing for both animals all being performed on different days. Analysis of mitochondrial DNA sequences also confi rmed the Cercopithecus species of each monkey and the absence of admixtures of specimens from different NHP species. STLV-3 (Cmo8699AB) tax sequences share 72%-74% nucleotide identity with PTLV-1, PTLV-2, and PTLV-4, but they have the highest nucleotide identity to the PTLV-3 group (82%-84%) in this highly conserved   Cag9812NL Cercopithecus agilis (agile mangabey) Cni8348NY Cni7882AB Lal9589NL Lophocebus albigena (gray-cheeked monkey) Ngoila region where intragroup sequence identity is typically >90%. Phylogenetic analysis of 881-bp tax sequences (Figure 4) from these 2 monkeys with other PTLVs, using bovine leukemia virus as an outgroup, inferred a new lineage with high bootstrap support (99%) from the diversity of other PTLV-3 subtypes (larger tax sequences representing PTLV-3 subtype C were not available for inclusion in this analysis), which suggests a long, independent evolution of this divergent virus. Similar PTLV-3 tree topologies were obtained by analysis of 275-bp LTR sequences ( Figure 5) in which STLV-3 (Cmo8699AB) and STLV-3 (Cni7867AB) had only 70%-74% identity to LTRs from other members of the PTLV-3 group that share >84% nucleotide identity between subtypes A and B (data not shown). LTR sequences from other STLV-3-infected C. agilis and C. nictitans monkeys from Cameroon reported elsewhere were not available from GenBank (9,21,25) and thus were not included in the current phylogenetic analysis. Combined, the phylogenetic analyses of the tax (Figures 3, 4) and LTR ( Figure 5) sequences show that STLV-3 (Cmo8699AB) and STLV-3 (Cni7867AB) each form a distinct cluster with high bootstrap support from the other known PTLV-3 subtypes. On the basis of nomenclature proposed by others (17), our results suggest that these viruses are members of a novel PTLV-3 subtype that we tentatively name as STLV-3 West African subtype D.

Origin of STLV-3 (Cmo8699AB)
To estimate the divergence times of the most recent common ancestor of STLV-3 (Cmo8699AB), we performed additional molecular analyses. We found that the molecular clock hypothesis was not rejected for the 881-bp alignment of PTLV and bovine leukemia virus tax sequences in both PAUP* (http://paup.csit.fsu.edu) and Tree-Puzzle (www. tree-puzzle.de) analyses (p = 0.012 and 0.858, respectively), which is consistent with results obtained recently by others (29). Using a molecular clock model and a tree calibration date estimated for the origin of Melanesian HTLV-1 ≈40,000-60,000 years ago (15,19,29,30), we inferred the evolutionary rate for PTLV to be 9.17 × 10 -7 to 1.38 × 10 -6 substitutions/site/year, which is consistent with rates determined previously both with and without a molecular clock model of evolution (15,17,20,(29)(30)(31). The evolutionary rate for STLV-3 (Cmo8699AB) is estimated to be 2.11 × 10 -6 to 3.16 × 10 -6 , and the most common recent ancestor is inferred to have occurred ≈92,072-138,560 years ago, which suggests an ancient origin and perhaps the identifi cation of one of the oldest viruses in the PTLV-3 group.

Broad STLV-3 Diversity in Wild NHPs
Sequence analysis of the STLV-3 LTR sequences from Cni7882AB, Cag9748NL, and Lal9589NL showed that all  were infected with distinct STLV-3s. LTR sequences (283 bp) from animal Cag9748NL shared the greatest identity (>97%) with those from HTLV-3 (Pyl43) and STLV-3 (Cto604) from a red-capped mangabey from Cameroon (1,20). The 282-bp LTR sequence from Cni7882AB shared the highest nucleotide identity (99%) to STLV-3 (CtoNG409), a red-capped mangabey from neighboring Nigeria (31). The phylogeographic clustering of these sequences supports further the proposed subtype classifi cation of STLV-3 by geographic origin rather than by host species (17,19,20,25,31). In contrast, the 432-bp LTR sequence from L. albigena mangabeys (Lal9589NL) was more divergent; it shared only 10%-16% nucleotide identity with all PTLV-3 LTR sequences. Similar to the phylogenetic relationships inferred with the small tax sequences, the LTR sequence from L. albigena mangabeys (Lal9589NL) formed a new lineage within the diversity of other PTLV-3 sequences from west-central Africa ( Figure 5). Although these results need to be confi rmed with additional LTR sequences from this virus and from other STLV-3-infected L. albigena mangabeys (9), our fi ndings demonstrate a host range and geographic distribution of STLV-3 that is more widespread than previously considered.

Phylogenetic Analysis of STLV-1 Diversity
To investigate further the genetic relationships inferred with the small PTLV-1-like tax sequences, we obtained LTR sequences for 6 of 7 PTLV-1-positive samples by using established primer-pair combinations (3,4,7). Phylogenetic analysis of these sequences, including those identifi ed from our study of infected NHP hunters in Cameroon (7), showed that 4 sequences from C. nictitans mon-   Cercopithecus nictitans (Cni217, Cni227, and Cni3038) reported elsewhere (9,21) and those identifi ed in the current study (in boldface). GenBank accession numbers for the previously reported partial simian T-lymphotropic virus (STLV)-3 tax sequences included in this analysis are AY039033, AF412120, and AM746647-AM746673). Support for the branching order was determined by 1,000 bootstrap replicates; only values >60% are shown. Branch lengths are proportional to the evolutionary distance (scale bar) between the taxa. See Figure 2 legend for abbreviations. keys all clustered in the central African HTLV-1 subtype D clade, consisting of STLV-1 from Mandrillus sphinx and Cercopithecus pogonias monkeys and HTLV-1 sequences from Cameroon ( Figure 6). The STLV-1 (Cni10225NL) LTR sequence was phylogenetically closest to the HTLV-1 (1842LE) strain from an NHP hunter from Cameroon (7) (Figure 6). Similarly, LTR sequences from 2 C. agilis (Cag9812NL and Cag9813NL) monkeys clustered within the HTLV-1F clade ( Figure 6). Combined, these results support further the primate origin of the HTLV-1D and -1F subtypes. We were unable to amplify STLV-1 LTR sequences from DBS samples from a C. nictitans monkey (Cni10026NL) that was positive for STLV-1 tax sequences, possibly because of low viral load in this animal, lower sensitivity of the LTR primers, or genetic variances at the LTR primer binding sites. The absence of STLV-1 LTR sequences in this monkey is not likely to have resulted from infection with an STLV-1/STLV-3 recombinant after dual infection of animal Cni10026NL with both viruses because samples from this animal were repeatedly negative for STLV-3 tax and LTR sequences.

Absence of Novel STLV-3 Subtype Sequences in NHP Hunters
Given the prevalence of the STLV-3 subtype D virus in at least 2 monkey species in Cameroon, we investigated whether this new subtype was also present among NHP hunters in Cameroon. Peripheral blood mononuclear cell DNA samples were available from a previous study of 63 NHP hunters who had a wide range of WB seroreactivity to HTLV (7). HTLV sequences were not previously detected in the DNA of these persons when either generic or groupspecifi c primers were used (7). All 63 NHP hunters were also negative for STLV-3 (Cmo8699AB) tax-specifi c sequences, which suggests the absence of this virus in this subset of persons with broad WB seroreactivity to HTLV.

Discussion
Widespread exposure to a broad range of NHP body fl uids and tissues encountered during hunting, butchering, or keeping primates as pets has been implicated in the emergence of 3 different retrovirus genera: HIV, HTLV, and, more recently, simian foamy virus (2)(3)(4)(5)7,16,28,32). Although little is known about the public health implications of simian foamy virus infection, the social, medical, political, and economic consequences of HIV and HTLV global spread and pathogenicity after cross-species transmission are enormous. The recent discovery of HTLV-3 and HTLV-4 in NHP hunters from Cameroon doubles the number of known deltaretroviruses in humans (7). This   same study also identifi ed novel STLV-1-like infections in NHP hunters (7). These discoveries demonstrate that the diversity of PTLV is far from understood and that zoonotic infection with STLV continues in persons exposed to NHPs (7). Thus, understanding the diversity, prevalence, and geographic range of STLV infection in areas where frequent contact with wild NHPs is common provides useful information about the origin and emergence of HTLV and the risks for exposure to these and possibly other simian viruses.
We demonstrated that monkeys from 3 distant locations in the rain forests of southern Cameroon are infected with a broad range of highly diverse STLV. Our detection of a 7% prevalence of STLV infection among hunted wild monkeys is comparable to the 8%-11% seroreactivity to PTLV recently found in monkey and ape samples collected mostly at urban bushmeat markets in Cameroon (9,25). Through analysis of LTR and larger tax sequences from C. mona and C. nictitans monkeys in our study, we have identifi ed new divergent STLV-3-like strains that form a unique PTLV-3 clade that we designated subtype D. Altogether, these results extend further the range of PTLV diversity and suggest a founder effect for PTLV evolutionary radiation in this region where most PTLV groups have been identifi ed.
Given the propensity of STLV to cross species boundaries, the increased frequency of hunting and demand for primate bushmeat in Africa, and the apparent broad diversity of STLV subtypes in Cameroon (9,21), it is tempting to speculate that human infection with this unique STLV-3 subtype will or may have already occurred. However, PCR testing of DNA samples from Cameroon NHP hunters with broad HTLV WB patterns showed no evidence of STLV-3 (Cmo8699AB)-like infections. Possible explanations for this negative fi nding include the testing of only a limited number of available samples, an unknown sensitivity for serologic detection of this virus with assays used in our study (7), an unknown prevalence and host range of this virus in NHPs, and other factors such as low transmissibility to humans. Nevertheless, the discovery of this novel PTLV-3 subtype in 2 monkey species and an apparent ancient origin of this lineage suggest a possible wider distribution of this variant. Therefore, the ease with which STLVs can cross species barriers and potentially be transmitted during NHP-hunting practices warrants increased surveillance for human infection with this divergent subtype. A similar strategy involving intensifi ed screening of NHP hunters was successful in the discovery of HTLV-3 (1,7) and HTLV-4 (7).
Finding a broad range of STLVs in simian DBS indicates that persons exposed to NHPs from Cameroon are at increased risk for infection with highly diverse STLV. Indeed, phylogenetic analysis of PTLV-1 LTR sequences shows that the new STLV-1 from C. nictitans monkeys identifi ed in the current study is most similar to HTLV-1 from Cameroon NHP hunters (7). Similarly, the clustering of STLV-1 from C. agilis monkeys with LTR sequences obtained from a person from Liberia provides additional support for the primate origin of the HTLV-1F clade (33). Combined, these fi ndings further support the hypothesis of active cross-species transmission of STLV to humans in this region (7).
Moreover, we show that use of DBS collected in the fi eld in collaboration with hunters provides a good tool for surveillance of emerging retroviral infections at the NHPhunter interface. Convenient and cost-effective, this collection strategy provides a unique opportunity to examine zoonotic transmission at the point where pathogen spillover occurs. In conjunction with longitudinal sampling of hunters, these collections have the potential to enable simultaneous documentation of both sides of a cross-species transmission event.  In summary, we found broad diversity of STLV in NHPs from Cameroon and identifi ed a novel STLV-3 subtype. These results provide increasing evidence that the diversity and geographic distribution of PTLVs are much greater than previously thought. Bushmeat hunting, an ancient and common practice in many parts of Africa, is an ideal interface for cross-species transmission of retroviruses between NHPs and humans. Contact with body fl uids and blood during hunting and butchering of NHP bushmeat exposes humans to a plethora of simian retroviruses, as demonstrated here and elsewhere (7,23,25,32,34,35), and increases the likelihood of emerging diseases in humans. To predict and possibly prevent the next retrovirus pandemic, expanded surveillance is needed for these and other retroviruses in their natural host reservoirs and in persons exposed to NHPs (7,36,37).