Evidence for a Second Simian T-Cell Lymphotropic Virus Type 3 in Cercopithecus nictitans from Cameroon

Presently, three types of simian T-cell lymphotropic virus (STLV) are known. STLV-1 has been detected in more than 20 different Asian and African primate species; STLV-2 has only two known representatives, in the species Pan paniscus; and STLV-3, previously named STLV-L, thus far has only one strain, in an Eritrean Papio hamadryas (1–3, 7, 11). In our search for African nonhuman-primate retroviruses, 122 Cameroonian Cercopithecus nictitans were screened for the presence of STLV. Divergent STLVs were discovered in animals Cni217 and Cni227, both of which were killed for the bushmeat trade. High titers of STLV antibodies were detected in whole-blood samples from these animals with the human T-cell lymphotropic virus type I/II (HTLV-I/-II) kit from Abbott Laboratories (Abbott Park, Ill.). The Innolia, a line immunoblot assay (Innogenetics, Ghent, Belgium), showed for both samples a profile most concordant with those of HTLV-2-infected patients and the prototype STLV-3 strain PH969 (Fig. 1). These samples reacted with the generic gag p24-1/2, env gp21-1/2, and env gp46-1/2 peptides, the HTLV-2-specific env gp46 peptides, and, for Cni227, weakly with gag p19. A diagnostic, generic, type-specific tax-rex primate T-cell lymphotropic virus (PTLV) PCR (12) suggested the presence of divergent STLV proviral DNA. A band of the correct length could be amplified from the two C. nictitans samples with the generic PTLV PCR. The four different type-specific inner PCRs, however, discriminating HTLV-1-, HTLV-2-, STLV-2-, and STLV-3-like viruses, were not able to amplify the STLV proviral DNAs of Cni217 and Cni227. The sequenced inner generic tax-rex Cni217 and Cni227 PCR products of 219 bp exhibited sequence similarities of 88.6 and 89.0%, respectively, to STLV-3 strain PH969, detected in an Eritrean Papio hamadryas (3), while nucleotide divergences of 20 to 29% were evident for the prototypic HTLV-1/STLV-1 and HTLV-2/STLV-2 strains. The 219-bp tax-rex fragments of Cni217 and Cni227 were phylogenetically analyzed using PAUP*4.0b5 (9). Neighbor-joining (NJ) and maximum-likelihood (ML) trees were calculated for 22 different taxa by using the Tamura Nei substitution model (6). The Cni217 and Cni227 strains were found to be very closely related, and as expected from the Blast (www.ncbi.nlm.nih.gov/blast; National Center for Biotechnology Information, Bethesda, Md.) search results, they both clustered with PH969 in the NJ tree and the ML tree (100% bootstrap support for NJ; P < 0.01 for ML) (Fig. 2). The clustering was also highly supported by parametric bootstrapping of the ML data through a Monte Carlo simulation (P < 0.01) (5). Although a phylogenetic analysis of a larger sequenced fragment would result in more accurate evolutionary distances, the NJ analysis of this short tax-rex fragment nevertheless revealed that the evolutionary distance between PH969 and Cni217/Cni227 was of the same order of magnitude as the distance between the prototypic HTLV-2a strains and STLV-2 strain PP1664. Different PCRs of gene regions covering the long terminal repeat, env, and 3′ pol-pX, performed using degenerative primers and thus potentially amplifying DNAs of divergent PTLVs, were optimized on reference strains of the four different PTLV types (HTLV-1 MT2, HTLV-2 Cl19, STLV-2 PP1664, and STLV-3 PH969) but failed to amplify the STLV-3 C. nictitans gene fragments. This could be explained in part by the highly divergent character of Cni217 and Cni227 but probably also by the quality of the extracted DNA, derived from frozen whole blood collected from slaughtered simians.

FIG. 1.

Innolia results generated from HTLV-1, HTLV-2-infected-patient plasma, PH969 serum, Cni217 and Cni227 whole blood. Lanes 1 and 2 (from the right) contain the positive control (PC) and negative control (NC) of the kit, respectively. Lanes 3 and 4 show the results for an HTLV-1- and an HTLV-2-infected-patient sample, respectively. Lane 4 contains the STLV-3 PH969 sample. Lanes 5 and 6 contain the Cni217 and Cni227 samples, respectively. The first three control lines contain human immunoglobulin G (IgG) in different concentrations; they are followed by four confirmation lines (2 gag and 2 env HTLV-1/-2 antigen lines) and then three discriminatory lines (2 gag HTLV-1 and 1 env HTLV-2 peptide) at the bottom of the strip.

FIG. 2.

PAUP* NJ tree of a 219-bp tax-rex fragment including sequences from reference strains of each PTLV type and subtype, with the bootstrap values (in percent) and P values (∗∗, P < 0.01; ∗, P < 0.05) noted on the branches. Sequences used were as follows: for STLV-3, Cni217 (GenBank accession no. AY039033), Cni227 (AF412120), and PH969 (Y07616); for STLV-2, PP1664 (Y14570) and PanP (U90557); for HTLV-2, Efe2 (Y14365), GAB (Y13051), G12 (L11456), Gu (X89270), Mo (M10060), NRA (L20734), and SPVW (AF139382); for STLV-1, TE4 (Z46900); and for HTLV-1, ATL-YS (U19949), ATK1 (J02029), BOI (L36905), EL (S74562), HS35 (D13784), Mel5 (L02534), MT2 (L03561), RKI3 (AF042071), and TSP1 (M86840).

Very recently, preliminary results suggested the occurrence of STLV-3 infections in a Cameroonian Cercocebus torquatus (4), an Ethiopian Theropithecus gelada (8), and several Ethiopian Papio hamadryas and Papio hybrids (10). The discovery of STLV-3-like proviral DNA in the simian species C. nictitans (sampled at a great distance from the original Papio hamadryas strain), the only known STLV-3 strain described in detail thus far, and these preliminary reports of still other STLV-3 infections in diverse primate species indicate that this type of virus is not restricted to eastern Africa and raises questions about STLV-3 cross-species transmission and infection in other primate species, including humans.

The fact that the Cni217/Cni227 and PH969 STLV-3 sequences are quite divergent in this generally very conserved gene region, with a distance comparable to the one between HTLV-2a strains and the STLV-2 PP1664 strain, suggests that these viruses have undergone a long and separate evolution within their host species. These results further strengthen the hypothesis of an African origin for STLV-3 and PTLV infections.