Amplification of a complete simian immunodeficiency virus genome from fecal RNA of a wild chimpanzee

Abstract

Current knowledge of the genetic diversity of simian immunodeficiency virus (SIVcpz) infection of wild chimpanzees (Pan troglodytes) is incomplete since few isolates, mostly from captive apes from Cameroon and Gabon, have been characterized; yet this information is critical for understanding the origins of human immunodeficiency virus type 1 (HIV-1) and the circumstances leading to the HIV-1 pandemic. Here, we report the first full-length SIVcpz sequence (TAN1) from a wild chimpanzee (Pan troglodytes schweinfurthii) from Gombe National Park (Tanzania), which was obtained noninvasively by amplification of virion RNA from fecal samples collected under field conditions. Using reverse transcription-PCR and a combination of generic and strain-specific primers, we amplified 13 subgenomic fragments which together spanned the entire TAN1 genome (9,326 bp). Distance and phylogenetic tree analyses identified TAN1 unambiguously as a member of the HIV-1/SIVcpz group of viruses but also revealed an extraordinary degree of divergence from all previously characterized SIVcpz and HIV-1 strains. In Gag, Pol, and Env proteins, TAN1 differed from west-central African SIVcpz and HIV-1 strains on average by 36, 30, and 51% of amino acid sequences, respectively, approaching distance values typically found for SIVs from different primate species. The closest relative was SIVcpzANT, also from a P. t. schweinfurthii ape, which differed by 30, 25, and 44%, respectively, in these same protein sequences but clustered with TAN1 in all major coding regions in a statistically highly significant manner. These data indicate that east African chimpanzees, like those from west-central Africa, are naturally infected by SIVcpz but that their viruses comprise a second, divergent SIVcpz lineage which appears to have evolved in relative isolation for an extended period of time. Our data also demonstrate that noninvasive molecular epidemiological studies of SIVcpz in wild chimpanzees are feasible and that such an approach may prove essential for unraveling the evolutionary history of SIVcpz/HIV-1 as well as that of other pathogens naturally infecting wild primate populations.

FIG. 1.

Amplification of a complete SIVcpzTAN1 sequence from chimpanzee fecal RNA. RT-PCR was used to amplify partially overlapping subgenomic fragments from virion RNA extracted from two independent fecal samples collected on 9 November 2000 and 11 November 2000 from chimpanzee Ch-06 under direct observation. The relative positions of the fragments with respect to the SIVcpzTAN1 genome, along with their order of amplification, are shown. Fragment lengths are drawn to scale (the entire SIVcpzTAN1 sequence is 9,326 bp in length). The first two fragments, which were derived with consensus primers from highly conserved regions of the HIV-SIV genome, are highlighted in red. Sequences of fragments 1, 3, and 4 have been reported previously (32).

FIG. 2.

Genetic diversity of Vpu in SIVcpz from P. t. schweinfurthii. (A) Alignment of the TAN1 and ANT Vpu sequences reveals an extraordinary degree of diversity (identical residues are indicated by asterisks) but demonstrates conservation of two serine residues (denoted by arrows) previously shown to be critical for HIV-1 Vpu-induced CD4 degradation (41). (B) Hydrophobicity profiles of SIVcpzTAN1 and SIVcpzANT Vpu proteins (calculated with the Kyte and Doolittle algorithm [21]) suggest similar secondary structures. Positive scores indicate degrees of hydrophobicity. The x axis shows the number of amino acid residues along the protein.

FIG. 3.

Diversity plots of concatenated protein sequences illustrating the extent of genetic diversity within the HIV-1/SIVcpz group of viruses. The proportion of amino acid sequence differences between SIVcpzTAN1 and SIVcpzGAB1 (green solid line), SIVcpzANT (red solid line), HIV-1/U455 (black solid line), and HIV-1/MVP5180 (blue solid line) is shown, in comparison with the proportion of amino acid sequence differences between SIVagmVER155 and SIVagmGRI-1 (red dotted line) and SIVlhoest7 and SIVsun (blue dotted line), respectively. Predicted protein sequences were concatenated, with C-terminal overlaps being removed and aligned with Clustal W (43) and with minor manual adjustments being made with SEAVIEW (11). Sites that could not be unambiguously aligned, as well as sites containing a gap in any sequence, were excluded from the alignment. Protein sequence differences (number of amino acid sequence mismatches per sequence length) were calculated for windows of 300 amino acids, moved in steps of 20 amino acids across the concatenated alignment. The x axis indicates amino acid positions, with the N termini of Gag, Pol, Vif, Env, and Nef denoted. The y axis denotes the distance between the viral proteins (0.1 = 10% differences).

FIG. 4.

Maximum likelihood tree depicting the relationship of SIVcpzTAN1 to other primate lentiviruses. SIVcpzTAN1 predicted protein sequences were compared to the following HIV and SIV representatives: HIV-1 group M subtype A (U455 [GenBank accession number M62320]) and subtype B (LAI [accession number K02013]), group N (YBF30 [accession number AJ006022] and YBF106 [accession number AJ271370]) and group O (ANT70 [accession number L20587] and MVP5180 [accession number L20571]), and SIVcpz (GAB1 [accession number X52154], US [accession number AF103818], CAM3 [accession number AF115393], CAM5 [accessionnumber AJ271369], and ANT [accession number U42720]). Phylogenetic trees were constructed by both maximum likelihood and neighbor joining methods and rooted by using SIVmndGB1 as an outgroup (M27470). The neighbor joining method was implemented in Clustal W (43) with Kimura's empirical correction (18) and 1,000 bootstrapped replicates. The maximum likelihood method was implemented in PROTML from the MOLPHY package (J. Adachi and M. Hasegawa, Institute of Statistical Mathematics, Tokyo, Japan, 1994) by using a quick search and the Jones, Taylor, and Thornton model for amino acid substitution (16) with amino acid frequencies. SIVcpz strains from P. t. troglodytes are shown in red, and those from P. t. schweinfurthii are shown in blue, with SIVcpzTAN1 highlighted. Horizontal branch lengths are drawn to scale. The scale bar indicates 0.1 substitution per site. Asterisks denote branches (and thus the clades to their right) supported by more than 85% of bootstrapped replicates in trees made by the neighbor joining method.

FIG. 5.

SIVcpz lineage-specific protein signatures. Alignments of SIVcpz and HIV-1 Vif, Nef, Vpr, and gp41 deduced amino acid sequences are shown for selected regions of the proteins. Sequences are compared to SIVcpzTAN1, with dashes denoting sequence identity and dots representing gaps introduced to optimize the alignment. Question marks indicate sites of ambiguous sequence (in SIVcpz) or sites at which fewer than 50% of the viruses contain the same amino acid residue (in HIV-1). Pound signs indicate stop codons. HIV-1 group M, N, and O sequences are consensus sequences obtained from the Los Alamos HIV Sequence Database (http://hiv-web.lanl.gov) (20). For HIV-1 group M, consensus subtype A sequences are shown; however, consensus sequences from the other subtypes yielded the same results (data not shown). Vertical boxes highlight SIVcpz lineage-specific protein signatures in Vif, Vpr, Nef, and gp41. Arrows denote a pair of cysteine residues in the ectodomain of gp41 that is unique to P. t. schweinfurthii viruses (the horizontal line denotes the immunodominant region of the HIV-1 gp41 glycoprotein). Asterisks indicate a highly conserved PPLP motif in Vif, a diacidic β-COP binding motif in Nef, and four C-terminal Arg residues in Vpr (Arg-90 is circled), previously shown to be critical for protein function in HIV-1 (see the text for details). Consistent with their proposed origin (12), HIV-1 groups M, N, and O exhibit protein features characteristic of SIVcpz from P. t. troglodytes.