Parallel germline infiltration of a lentivirus in two Malagasy lemurs

Abstract

Retroviruses normally infect the somatic cells of their host and are transmitted horizontally, i.e., in an exogenous way. Occasionally, however, some retroviruses can also infect and integrate into the genome of germ cells, which may allow for their vertical inheritance and fixation in a given species; a process known as endogenization. Lentiviruses, a group of mammalian retroviruses that includes HIV, are known to infect primates, ruminants, horses, and cats. Unlike many other retroviruses, these viruses have not been demonstrably successful at germline infiltration. Here, we report on the discovery of endogenous lentiviral insertions in seven species of Malagasy lemurs from two different genera -- Cheirogaleus and Microcebus. Combining molecular clock analyses and cross-species screening of orthologous insertions, we show that the presence of this endogenous lentivirus in six species of Microcebus is the result of one endogenization event that occurred about 4.2 million years ago. In addition, we demonstrate that this lentivirus independently infiltrated the germline of Cheirogaleus and that the two endogenization events occurred quasi-simultaneously. Using multiple proviral copies, we derive and characterize an apparently full length and intact consensus for this lentivirus. These results provide evidence that lentiviruses have repeatedly infiltrated the germline of prosimian species and that primates have been exposed to lentiviruses for a much longer time than what can be inferred based on sequence comparison of circulating lentiviruses. The study sets the stage for an unprecedented opportunity to reconstruct an ancestral primate lentivirus and thereby advance our knowledge of host-virus interactions.

Figure 1. Map of the consensus pSIV reconstructed in this study based on one full-length copy of pSIVgml (Microcebus murinus) and between one and three copies of pSIVfdl (Cheirogaleus medius) (the alignment is provided in Dataset S1).

The LTR fragments contained in the ABDC01505939 and ABDC01454290 contigs correspond to the same full length pSIV copy (see Figure S1). The contig ABDC01306160 results from a misassembly between a trace read containing a solo-LTR and a trace read containing the 3′ terminus of the env of the full-length pSIVgml and a fragment of the 3′ LTR (see Figure S1). The different domains of pSIV were identified by comparison with the HIV1-HXB2 reference sequence (see Figure S4 for a more precise map). Closed circles: non-sense frameshifts. Vertical bars: in frame stop codons. Dash lines: missing fragments. The range of pairwise similarity, number of substitutions per site and inferred divergence times between pSIVgml and pSIVfdl sequences are indicated.

Figure 2. Southern blot of digested genomic DNA of various Malagasy lemurs and human using a ∼1 kb probe corresponding to a fragment of pSIVgml env (A) or a ∼300 bp probe corresponding to a fragment of the pSIV LTR (B).

Arrows highlight bands of the same size shared by Microcebus murinus, M. griseorufus and M. ravelobensis but not Cheirogaleus medius. These bands likely correspond to solo-LTRs located at orthologous position in the three Microcebus species. The trees on top of each blot depict the phylogenetic relationships of the species according to ,. See Table S4 for the voucher specimen numbers of the lemur tissue samples used in this study. A picture of the ethidium bromide stained gels used to prepare the blots is shown in Figure S2. The absence of pSIV in Mirza was confirmed by PCR using different sets of primers (Figure S3).

Figure 3. Alignment between a 71 aa region of the RT domain of the primate lentiviruses and the orf2 (46 aa)−env (25 aa) junction of pSIV.

Shading of the different positions represents the level of sequence conservation using the BLOSUM 62 amino acid substitution matrix in BioEdit (Hall, 2004). In addition, each asterisk indicates positions where pSIV amino acids are shared with at least one other primate lentivirus sequence. Accession numbers of the sequences are listed in Table S2.

Figure 4. Unrooted tree of lentiviruses obtained after phylogenetic analysis of an alignment including ∼2350 nucleotides of the gag-pol region.

Numbers associated to internal branches correspond to Bayesian posterior probabilities/bootstrap ML values. Asterisks indicate when an internal branch is supported by posterior probability = 1/bootstrap = 100. Accession numbers of the sequences are listed in Table S2. The alignment used for the analyses is provided in Dataset S2.

Figure 5. PCR screening for presence/absence of orthologous solo LTR in various species of Malagasy lemurs.

For each locus, the larger PCR product indicates presence of the LTR, the smaller product indicates absence. (A) Primers (6160F: 5′-CAG CAK TTT CAT CAG CAA TTT G; 6160R: 5′-GCA AGC TGT GMC ACA TTT ATT BGC) were designed on the regions flanking the solo LTR in contig ABDC01306160. The expected size was ∼670 bp for presence and ∼250 bp for absence. (B) Primers (9233F: 5′-ATC TRT AGT CAA ATC CTG GG; 9233R: 5′-TAA TAC TCA CAA AAA CYT TAC C) were designed on the regions flanking the solo LTR in contig ABDC01159233. The expected size was ∼550 bp for presence and ∼130 bp for absence. (C) Primers (61523F: 5′-AAA TGA GTT TTG TTG CTC TRT YTC; 61523R: 5′-ATG TTR CTT TGG GTA GMT TG) were designed on the regions flanking the solo LTR in contig ABDC01361523. The expected size was ∼585 bp for presence and ∼165 bp for absence. The genus Eulemur and Propithecus belong to the family Lemuridae and Indriidae respectively. All the other species (genera Cheirogaleus, Microcebus and Mirza) belong to the family Cheirogaleidae. The tree depicts the phylogenetic relationships of the species and their divergence times according to ,. See Table S4 for the voucher specimen numbers of the lemur samples used in this study.

Figure 6. Unrooted tree of several LTRs (9

MMU: Microcebus murinus, MRA: M. ravelobensis, MTA: M. tavaratra, MSA: M. sambiranensis, MMY: M. myoxinus, MGR: M. griseorufus, CME: Cheirogaleus medius. See Table S4 for the voucher specimen numbers of the lemur samples used in this study. Numbers associated to internal branched correspond to Bayesian posterior probabilities ≥0.95/bootstrap ML values ≥80. The alignment used for the analyses is provided in Dataset S3.