Fitness disadvantage of transitional intermediates contributes to dynamic change in the infecting-virus population during coreceptor switch in R5 simian/human immunodeficiency virus-infected macaques

Abstract

Fitness disadvantage of the transitional intermediates compared to the initial R5 viruses has been suggested to constitute one of the blockades to coreceptor switching, explaining the late appearance of X4 viruses. Using a simian model for human immunodeficiency virus type 1 (HIV-1) coreceptor switching, we demonstrate in this study that similar molecular evolutionary pathways to coreceptor switch occur in more than one R5 simian/human immunodeficiency virus (SHIV)(SF162P3N)-infected macaque. In infected animals where multiple pathways for expansion or switch to CXCR4 coexist, fitness of the transitional intermediates in coreceptor usage efficiency influences their outgrowth and representation in the infecting virus population. Dualtropic and X4 viruses appear at different disease stages, but they have lower entry efficiency than the coexisting R5 strains, which may explain why they do not outcompete the R5 viruses. Similar observations were made in two infected macaques with coreceptor switch, providing in vivo evidence that fitness disadvantage is an obstacle to X4 emergence and expansion.

FIG. 1.

Infecting virus population in macaque CA28 during coreceptor switch. (A) V3 loop sequence comparison of representative SHIV_SF162P3N clones (P3N-1, P3N-2) and week (w)11 and w15 plasma and PBMC viruses in macaque CA28. Dots indicate gaps, dashes stand for identity in sequences, and the net positive charge of this region is shown on the right. Positions 11 and 25 within the V3 loop are indicated by arrows, and the 4-amino-acid deletion is underlined in the reference sequence. The numbers in parentheses represent the numbers of clones matching the indicated sequence per total number of clones sequenced. Amino acid residues predicted to confer CXCR4 usage are boxed. (B) Tracking the emergence of RRW/RRW.A- and Δ22-25-bearing variants in plasma using sequence-specific PCR. wpi, week postinfection; NC, negative control using plasma from an uninfected macaque; M, marker. (C) Distribution of V3 variants in tissue compartments of macaque CA28 at time of necropsy as determined by clonal sequence analysis. Numbers in parentheses indicate numbers of gp120 clones sequenced from each of the tissue sites. Ing, inguinal LN; Mes, mesenteric LN; Axi, axillary LN; Ili, iliac LN; Col, colonic LN; IEL, intraepithelial lymphocyte; LPL, lamina propria lymphocyte from the jejunum.

FIG. 2.

Coreceptor usage (A) and preference (B) of V3 variants in CA28. Relative entry of pseudotyped reporter viruses in U87.CD4.CCR5 and U87.CD4.CXCR4 indicator cells was determined (A), and blocking of entry in TZM-bl cells with 1 μM CCR5-specific (TAK779) and CXCR4-specific (AMD3100) inhibitors was performed to determine coreceptor preference of the reporter viruses (B). Data are the means and standard deviations of results from triplicate wells and are representative of at least two independent experiments. RLU, relative light units.

FIG. 3.

Phylogenetic tree showing the relationship between Env variant sequences (V1 to V5) in macaque CA28. A neighbor-joining tree rooted on four functional sequences of the inoculating virus SHIV_SF162P3N was generated. The scale bar indicates the genetic distance along the branches in nucleotides, and the asterisk at the node represents a bootstrap value of >70%. ⧫, representative clones in the inoculum; ▴, WT variants; •, RRW/RRW.A-bearing variants; ○, Δ22-25-bearing variants.

FIG. 4.

Distribution of V3 variants in plasma and PBMC of macaques DG08 and CA28 over time. Numbers in parentheses indicate numbers of Env clones sequenced at indicated time points.

FIG. 5.

Entry and coreceptor usage efficiencies of V3 variants in macaques DG08 (A) and CA28 (B). Determination of relative entry of pseudotyped reporter viruses in TZM-bl cells and blocking of pseudovirions with increasing concentrations of the CCR5 inhibitors TAK779 and PSC-RANTES and the CXCR4 inhibitor AMD3100 were performed. 50% inhibitory concentrations (IC₅₀) of the inhibitors were determined using Prism 4 software (GraphPad, San Diego, CA). RLU, relative light units. Data shown are means ± standard errors of the mean of results of at least two to four independent experiments.