Enhancement of CD4 Binding, Host Cell Entry, and Sensitivity to CD4bs Antibody Inhibition Conferred by a Natural but Rare Polymorphism in the HIV-1 Envelope

Abstract

A rare but natural polymorphism in the HIV-1 envelope (Env) glycoprotein, lysine at position 425 was selected as a mutation conferring resistance to maraviroc (MVC) in vitro. N425K has not been identified in HIV-infected individuals failing an MVC-based treatment. This study reports that the rare K425 polymorphism in an HIV-1 subtype A Env has increased affinity for CD4, resulting in faster host cell entry kinetics and the ability to scavenge for low cell surface expression of CD4 to mediate entry. Whereas the subtype A wild-type isolate-74 Env (N425) is inhibited by soluble (s) CD4, HIV-1 with K425 A74 Env shows enhanced infection and the ability to infect CCR5+ cells when pretreated with sCD4. Upon adding K425 or N425 HIV-1 to CD4⁺/CCR5+ cells along with RANTES/CCL3, only K425 HIV-1 was able to infect cells when CCR5 recycled/returned to the cell surface at 12 h post-treatment. These findings suggest that upon binding to CD4, K425 Env may maintain a stable State 2 "open" conformation capable of engaging CCR5 for entry. Only K425 was significantly more sensitivity than wild-type N425 A74 to inhibition by the CD4 binding site (bs) compound, BMS-806, the CD4bs antibody, VRC01 and N6, and the single-chain CD4i antibody, SCm9. K425 A74 was also capable of activating B cells expressing the VRC01 surface immunoglobulin. In summary, despite increased replicative fitness, we propose that K425 HIV-1 may be counterselected within infected individuals if K425 HIV-1 is rapidly eliminated by CD4bs-neutralizing antibodies. IMPORTANCE Typically, a natural amino acid polymorphism is found as the wild-type sequence in the HIV-1 population if it provides a selective advantage to the virus. The natural K425 polymorphism in HIV-1 Env results in higher host cell entry efficiency and greater replicative fitness by virtue of its high binding affinity to CD4. The studies presented herein suggest that the rare K425 HIV-1, compared to the common N425 HIV-1, may be more sensitive to inhibition by CD4bs-neutralizing antibodies (i.e., antibodies that bind to the CD4 binding pocket on the HIV-1 envelope glycoprotein). If CD4bs antibodies did emerge in an infected individual, the K425 HIV-1 may be hypersensitive to inhibition, and thus this K425 virus variant may be removed from the HIV-1 swarm despite its higher replication fitness. Studies are now underway to determine whether addition of the K425 polymorphism into the Envelope-based HIV-1 vaccines could enhance protective immunity.

Keywords: CD4 binding; HIV; entry; neutralizing antibodies.

FIG 1

Model of N425K in a subtype A isolate 74 Env modeled into the cryoEM structure of subtype A Bg505 complexed to sCD4 and antibodies 17b and 8ANC195. The subtype A strain A74 with N (A, C) or K425 (B, D) was then modeled onto the BG505 gp120 bound to sCD4 and antibodies 17b and 8ANC195 at 3.5 Å (cryoEM) (19). The Nε of K425 in the structural model superimposed well with a bound water molecule in the binding pocket of F43 from the 1G9M crystal structure. This structural model was also compared to a high resolution structure of gp120 in complex with CD4 and a CD4i antibody 17b (PDB ID 1G9M) (46), where the K425 in a subtype B strain results in a phenylalanine 43 of CD4, forming a new cation-π interaction with the Nε side chain of K425, as well as a new hydrogen bond with the oxygen of S42 CD4 (1). (E) Emergence of A74.K425 mutation in C4 region of HIV gp120 after passage of wild type subtype A HIV-1 isolate in presence of MVC.

FIG 2

Binding affinity of virus-specific N425K Env to sCD4. (A) To access binding affinity of wild-type or N425K A74 Env as a native trimer, sucrose cushion purified chimeric NL4-3 A74.K425.R+ or A74.N425.wt HIV-1 were bound to sCD4-Fc plates (sCD4-Fc bound to anti-IgG Fc-coated plates). HIV-1 removed by the addition of increasing concentrations of sCD4 in separate wells was measured by anti-p24 enzyme-linked immunosorbent assay (ELISA) and reverse transcriptase (RT) activity. (B) Percentage of bound to free HIV-1 at increasing sCD4 concentrations. The virus binding experiments were performed in quadruplicate. Two-way analysis of variance (ANOVA) was used for statistical comparisons for all points and paired two-tailed t tests for individual sCD4 concentrations.

FIG 3

A74.K425.R+ HIV-1 infects cells expressing low cell surface density CD4. (A) Schematic showing how the 293 Affinofile cells were induced to express low levels of CD4 across a range of CCR5 surface expression levels or low levels of CCR5 across a range of CD4 expression. (B, C) Eight cell cultures with ~3,000 CD4 molecules/cell and a range of ~25,000 to 1,200 CCR5 molecules/cell (B) and eight cell cultures with ~1,200 CCR5 molecules/cell and a range of ~150,000 to 3,000 CD4 molecules/cell (C) were infected with NL4-3 pseudotyped with A74.K425.R+ or A74.N425.wt Env. Luciferase activity measured 48 h postinfection. The percentage of infection was calculated based on the luciferase activity measured for the highest induction of both receptors (not shown). The data represent means of triplicates with standard deviations. *, P < 0.01 (one-tailed t test). Dox, doxycyclin.

FIG 4

Time of C34 drug addition to measure viral entry kinetics into host cells. (A) Time-of-drug-addition assays were performed in U87.CD4.CCR5 cells and then exposing to NL4-3Δenv_Luc pseudotyped with A74.K425.R+ or A74.N425.wt Env at 14°C. Warm medium (37°C) was added to synchronize infection with the addition of IC99 concentration of the fusion inhibitor C34 at the indicated time intervals postinfection. Relative infection was measured at 72 h postinfection based on the maximal fusion scored for each virus at 3 h postinfection. (B) The data was plotted as the mean and standard deviations of the percent relative infection measured at day 3 versus the time of C34 addition. The curves were generated using nonlinear regression curve-fitting features of GraphPad Prism 5 software. These experiments were performed in triplicate. Two-way ANOVA was used for statistical comparisons for differences in points on the curve.

FIG 5

Pre-exposure to sCD4 enhances infection mediated by A74.yN425.wt Env. (A) Pseudoviruses were preincubated with sCD4 (0, 20, or 200 nM) and then used to infect 293T cells transfected to express CCR5 but lacking cellular CD4. Luciferase activity was measured and compared with the no-sCD4 condition. The data represent means of triplicates with standard deviations. (B) Schematic representation of the CD4-independent infections described in panel A. (C) Pseudoviruses were treated as described and used to infect 293.CD4 cells transfected to express CCR5. Luciferase activity was measured and compared with the no-sCD4 condition. The data represent means of triplicates with standard deviations. Statistical relevance was measured by two-tailed t tests. (D) Schematic representation of infections in the presence of sCD4 into susceptible cells described in panel C.

FIG 6

Prolonged binding of A74.N425.wt Env to cellular CD4 still maintains ability to utilize CCR5 for host cell entry. (A) Schematic representation of the experimental procedure and possible explanation for experimental outcome shown in panel B. RANTES was added for 2 h, resulting in ~12 h intracellular retention of CCR5 prior to recycling the cell surface, at which time CCR5 may serve as coreceptor for host entry until the timed addition of TAK-779 starting at 12 h. (B) The total infection relative to the addition of TAK779 (at different time points) compared to the same experimental conditions in the absence of TAK779 is plotted for the NL4-3Δenv_Luc HIV-1 pseudotyped with A74.K425.R+ or A74.N425.wt Env. Experiments were performed in triplicate, and the statistics involved two-tailed paired t tests.

FIG 7

HIV-1 pseudotyped with A74.K425.R+ or A74.N425.wt Env inhibited by the broadly neutralizing CD4bs antibody VRC01 and by the CD4bs drug, BMS806. (A to C) HIV-1 pseudotyped with A74.K425.R+ or A74.N425.wt Env was incubated with increasing concentrations of VRC01 (A), N6 antibody (B), or BMS806 (C) in TZM-bl cells, and the luciferase activity (RLU) was used to measure the relative inhibition. All experiments were performed in triplicate, and the estimated EC₅₀ values were calculated and presented for antibodies in Table 1. (D) A model showing the binding of A74.K425.R+ or A74.N425.wt Env to VRC01 is shown and based on the subtype A BG505 SOSIP.664 Env trimer structure complexed to VRC01 (PDB ID 6V8X) (44). (E) Model showing the binding of A74.K425.R+ or A74.N425.wt Env to N6 based on the CRF01_AE 93TH057 gp120 core complexed to N6 (PDB ID 5TE6) (67). (F) Model showing the binding of A74.K425.R+ or A74.N425.wt Env to BMS806 based on the subtype A BG505 SOSIP.664 Env trimer structure bound to BMS806 (BMS-378806) (PDB ID 5U7M) (20).

FIG 8

Activation of VRC01 B cell receptor (BCR) WEHI cells by A74.N425.wt or A74.K425.R+. Doxycyclin was used to induce VRC01 light and heavy chain expression from the Tet on promoter in the WEHI-231 B cell clone. (A) Binding of HIV-1 to the BCR (VRC01 sIgG) can lead to signal transduction through CD76 and expression of mCD69 (20). (B, C) Cell activation and expression of CD69⁺ cells was measured at 72 h by flow cytometry after exposure to A74.N425.wt (B) or A74.K425.R+ (C) at an approximate MOI of ~0.5 with or without 200 nM sCD4. (D) Percent of CD69⁺ WEHI cells with different dilution of A74.N425.wt or A74.K425.R+ with/without sCD4. (E) Experiments were repeated in triplicate and presented as fold change CD69⁺ cells. Statistics involved two-tailed t tests. These experiments were repeated with viruses diluted 3-, 9-, 27-, and 81-fold. FSC, forward scatter; SSC, side scatter.

FIG 9

Inhibition of A74.N425.wt or A74.K425.R+ by the CD4i single-chain monoclonal antibodies SCm16 and SCm9. The tier III viruses A74.N425.wt or A74.K425.R+ were not inhibited by the majority of broadly neutralizing antibodies, including the CD4i antibody, b17 in TZM-bl cells and phytohemagglutinin (PHA)-activated, interleukin (IL)-2-treated human PBMCs (Fig. S1). (A) Inhibition of A74.N425.wt or A74.K425.R+ by CD4i single-chain monoclonal antibodies SCm16 and SCm9 at 12.5 ng/mL. This inhibition was also tested in the absence and presence of preincubation of the viruses with 20 nM sCD4. (B) With evidence of strong inhibition with sCD4 preincubations in panel A, sCD4 (20 nM) preincubated viruses were tested in TZM-bl cells for susceptibility to SCm16 and SCm9 inhibition across of range of 0.0125 to 50 ng/mL SCmAb concentrations. The experiments were performed in triplicate. One-tailed t tests were performed for panel A. The EC₅₀ values were estimated and are provided in Table 1 with statistical analyses.