Transmitted/founder and chronic subtype C HIV-1 use CD4 and CCR5 receptors with equal efficiency and are not inhibited by blocking the integrin α4β7

Abstract

Sexual transmission of human immunodeficiency virus type 1 (HIV-1) most often results from productive infection by a single transmitted/founder (T/F) virus, indicating a stringent mucosal bottleneck. Understanding the viral traits that overcome this bottleneck could have important implications for HIV-1 vaccine design and other prevention strategies. Most T/F viruses use CCR5 to infect target cells and some encode envelope glycoproteins (Envs) that contain fewer potential N-linked glycosylation sites and shorter V1/V2 variable loops than Envs from chronic viruses. Moreover, it has been reported that the gp120 subunits of certain transmitted Envs bind to the gut-homing integrin α4β7, possibly enhancing virus entry and cell-to-cell spread. Here we sought to determine whether subtype C T/F viruses, which are responsible for the majority of new HIV-1 infections worldwide, share biological properties that increase their transmission fitness, including preferential α4β7 engagement. Using single genome amplification, we generated panels of both T/F (n = 20) and chronic (n = 20) Env constructs as well as full-length T/F (n = 6) and chronic (n = 4) infectious molecular clones (IMCs). We found that T/F and chronic control Envs were indistinguishable in the efficiency with which they used CD4 and CCR5. Both groups of Envs also exhibited the same CD4+ T cell subset tropism and showed similar sensitivity to neutralization by CD4 binding site (CD4bs) antibodies. Finally, saturating concentrations of anti-α4β7 antibodies failed to inhibit infection and replication of T/F as well as chronic control viruses, although the growth of the tissue culture-adapted strain SF162 was modestly impaired. These results indicate that the population bottleneck associated with mucosal HIV-1 acquisition is not due to the selection of T/F viruses that use α4β7, CD4 or CCR5 more efficiently.

Figure 1. Clonal virus expansion in a chronically infected subject.

The phylogenetic relationships of SGA-derived 3′ half genome sequences depicting the quasispecies complexity in a chronically infected subject are shown. Sequences highlighted in blue are members of a recently expanded lineage, the consensus sequence of which approximates their most recent common ancestor. The tree was constructed using maximum likelihood methods . Nodes with bootstrap support of greater than 80% are labeled (the scale bar represents 0.02 nucleotide substitutions per site).

Figure 2. T/F and chronic Envs utilize CD4 and CCR5 with similar efficiency.

(A) Efficiency of CD4 usage by T/F and chronic Env pseudoviruses. Luciferase-encoding pseudoviruses were used to infect affinofile cells with maximally-induced CCR5 expression and either low or high CD4 expression. Infection of CD4-low cells was measured based on luciferase activity and normalized to infection of CD4-high cells (y-axis). Data shown are the mean value from three independent experiments for all T/F and chronic Envs, each tested in triplicate; for JR-FL the mean of each of three experiments is plotted. Grey circles indicate poorly infectious Envs for which the CD4 use efficiency is falsely elevated because infection of CD4-low cells was near background, as described in the methods. The bar represents the median value of each group with these Envs excluded. There was no significant difference in the CD4-use efficiency between T/F and chronic envelopes (Mann-Whitney; p = 0.35). (B) CD4 expression levels from a representative flow cytometry experiment depicting CD4 expression on minimally-induced Affinofile cells (low), fully-induced Affinofile cells (high), and primary human CD4+ T cells (CD4+ T). (C) Sensitivity of T/F and chronic Envs to inhibition by maraviroc (MVC). Viral pseudotypes were used to infect NP2/CD4/CCR5 cells in the presence of serial three-fold dilutions of the CCR5 antagonist maraviroc. The concentration required to reduce infection by 50% relative to no drug control (IC₅₀) was calculated for each pseudotype. Log MVC IC₅₀ values are shown in the y-axis. The bar represents the median value of each group. Higher IC₅₀ values indicate that more drug is required to prevent infection corresponding to more efficient CCR5 usage, whereas lower IC₅₀ values correspond to less efficient CCR5 usage. No significant difference in IC₅₀ values was found between T/F and chronic envelope pseudotypes (Mann-Whitney; p = 0.46).

Figure 3. T/F and chronic Env pseudotypes enter primary CD4+ T cell subsets with similar efficiency.

Primary CD4+ T cells were infected with GFP encoding Env pseudotypes to assess the ability of T/F and chronic Envs to mediate entry into different CD4+ T cell subsets. Three days post-infection, cells were stained for viability as well as CD3, CD4, CD45RO and CCR7 expression . In three independent experiments, cells from different donors cells were analyzed by flow cytometry and GFP+ cells were back-gated onto memory markers to evaluate differential infection of subsets; the average percent of infected cells falling into each subset for each Env is plotted on the y-axis. Cells were classified as central memory (T_CM: CCR7+CD45RO+), effector memory (T_EM: CCR7−CD45RO+), effector memory RA (T_EMRA: CCR7−CD45RO−), and naïve (CCR7+CD45RO−). While T_EM cells permitted the most efficient infection by Env pseudoviruses, there was no significant difference in infection efficiency between T/F and chronic Envs in any of the subsets. The bar represents the median percentage of cells of each subset infected by T/F or chronic pseudoviruses.

Figure 4. CD4-use efficiency correlates with CD4 binding site neutralization sensitivity.

The y-axis shows sensitivity to CD4bs mAbs b12 (A) and VRC01 (B) as measured by the maximal percent inhibition (MPI) using 10 µg/ml of each mAb. Envs were divided according to their CD4-use efficiency; those that used CD4 efficiently were more sensitive to CD4bs antibodies than Envs that used CD4 inefficiently. The bar represents the median MPI value and p-values are from two-tailed Mann-Whitney tests.

Figure 5. Blocking α4β7 enhances pseudovirus infection.

Infection of primary CD4+ T cells by Env pseudoviruses is depicted in the presence and absence of Act1 to block α4β7. atRA-treated cells were infected with GFP-expressing pseudotypes containing SF162 and JR-FL Envs whose gp120 proteins do and do not bind α4β7, respectively, in the absence (top two panels) or presence of saturating amounts of the α4β7 specific mAb Act1 (bottom two panels). Infected cells were detected by GFP expression (red overlay). The presence of Act1 resulted in slightly increased infection levels of both pseudotypes. To confirm that saturating levels of Act1 were used, cells were stained with fluorescently labeled Act1 before analysis. The near absence of cells in the upper quadrants of the lower panels shows that binding of labeled Act1 antibody was blocked by unlabeled antibody added before virus infection.

Figure 6. Blocking α4β7 inhibits replication of NL4-3-SF162 and NL4-3-R3A but not YU-2.

CD4+ T cells with or without Act1 pre-treatment were infected at three different multiplicities using CD4+ T cell derived virus stock (1 µl, 10 µl and 100 µl) to initiate a spreading infection. Infections were performed in six replicate wells, each of which was sampled at days three, six and nine. (A-C) Virus production at day six as measured by p24 content in culture supernatants is shown on the y-axis for each of six replicate wells from one of three independent experiments; uncorrected Mann-Whitney p values are shown for comparisons of no antibody (solid symbols) versus Act1-treated (open symbols) replicate wells (bar = mean). (D) Replication kinetics are shown for NL4-3-SF162 (mean of replicates ± SEM is presented) at three different multiplicities of infection. Inhibition of infection was transient and greatest at six days post infection at the lowest viral input; uncorrected Mann-Whitney p-values less than 0.05 comparing no mAb to Act1 are marked by asterisks.

Figure 7. Blocking α4β7 does not inhibit replication of subtype C T/F and chronic IMCs.

CD4+ T cells were infected at three different multiplicities using CD4+ T cell derived virus stock (1 µl, 10 µl and 100 µl). Virus replication was monitored by measuring p24 content in culture supernatants; each p24 measurement was repeated two to four times. The average p24 value in each of six independent wells at six days post-infection is plotted on the y-axis, with the bar representing the mean p24 value. Uncorrected Mann-Whitney p-values comparing no mAb to Act1 pre-treatment are shown for each viral input. Blocking α4β7 with Act1 increased p24 production for some viruses, but had no reproducible inhibitory effect on any T/F (A) or chronic (B) molecular clones tested, under conditions where reproducible inhibition of NL4-3-SF162 was achieved, as previously described .