A stepwise model for asp gene emergence across HIV and SIV lineages: progressive stop-codon loss and regulatory evolution
Background The antisense protein (ASP), encoded by an open reading frame (ORF) overlapping the env gene in human immunodeficiency virus (HIV), was proposed over 30 years ago. ASP is an ~189-amino-acid-long, highly hydrophobic protein with proposed roles in autophagy and viral entry. Although primarily described in pandemic HIV-1 group M, ASP was also identified in a single simian immunodeficiency virus (SIVcpz) isolate, suggesting recent acquisition. However, SIV from more distantly related lineages, including Old World monkeys (OWM), remains largely unexamined, and the broader origin of ASP is unresolved. Results We conducted computational analyses of 91,245 HIV-1, HIV-2, and SIV env sequences to assess the coding potential and evolutionary history of asp across primate lentiviruses. The sequences were classified into seven types based on the asp-frame ORF length and continuity: ASP, S1, and S2 for single continuous ORFs ([≥] 450, 300-449, and 120-299 nucleotides, respectively); F1-F3 for fragmented ORFs of comparable combined length; and Z for sequences lacking a detectable ORF of <120 nucleotides. Many SIV sequences retained a single, uninterrupted ORF of <120 nucleotides in the ASP frame, comparable to HIV-1 ASP-type sequences. A phylogenetic tree of 599 env sequences divided strains into HIV-1/SIVcpz/SIVgor and more distant HIV-2/SIV lineages. Fragmented asp ORFs were concentrated among basal, distantly related strains, suggesting asp emerged from stop-codon-rich ancestral segments. The stop-codon density progressively declined with ORF continuity in both groups ({rho} = -0.63 and -0.77, respectively; both p < 0.001), consistent with the stop-codon-loss mechanism proposed previously for HIV-1 and extended to SIV here. The nonsynonymous/synonymous analysis revealed strong purifying selection within the overlapping region in HIV-1, and a more variable, lineage-dependent pattern in SIV. Motif analysis of the U3/LTR-like regions showed antisense promoter-associated transcription factor binding sites, including NF-{kappa}B and ETS1, were enriched in HIV-1, SIVcpz, and SIVgor. Conclusions These findings support a stepwise model of antisense regulatory evolution, whereby promoter-associated transcription factor motifs accumulated before a continuous asp ORF became fixed in HIV-1. Variation within the pandemic lineage suggests that the asp coding potential remains evolutionarily dynamic, providing new insights into the emergence and stabilization of overlapping genes during lentiviral evolution.