Indian Long-term Non-Progressors Show Broad ADCC Responses with Preferential Recognition of V3 Region of Envelope and a Region from Tat Protein

Abstract

HIV-specific antibody-dependent cell cytotoxicity (ADCC) is likely to be important in governing protection from human immunodeficiency virus (HIV) and slowing disease progression. Little is known about the ADCC responses to HIV-1 subtype C. We characterized ADCC responses in HIV-1 subtype C-infected Indian subjects with slow disease progression and identified the dominant antigenic regions recognized by these antibodies. ADCC responses were measured in plasma from 34 long-term non-progressors (LTNPs), who were asymptomatic and maintained CD4 count above 500 cells/mm³ for the last 7 years in the absence of antiretroviral therapy (ART), and 58 ART naïve progressors with CD4 count <500 cells/mm³ against overlapping HIV-1 peptides using a flow cytometry-based antibody-dependent natural killer (NK) cell activation assay. The assay measured CD107a expression on NK cells as a marker of antibody-dependent NK cell activation and IFN-γ secretion by NK cells upon activation. The ADCC epitopes were mapped using the matrix of overlapping peptides. Indian LTNPs showed higher and broader ADCC responses compared to the progressors. The Env-C and Tat-specific ADCC responses were associated with lower plasma viral load, whereas the Env-C responses were also associated with higher CD4 counts. Five of 10 LTNP responders targeted epitopes in the V3 region (amino acids 288-330) of Env-C. Additionally, three Tat regions were targeted by ADCC antibodies from LTNPs. ADCC responses were associated with slow HIV progression in Indian subtype C-infected cohort. The frequently recognized peptides from the V3 loop of Env and the novel epitopes from Tat by the LTNPs warrants further study to understand the role of ADCC responses to these regions in control and prevention of HIV-1 infection.

Keywords: ADCC; LTNP; NK cells; Tat; envelope.

Figure 1

Gating strategy for antibody-dependent natural killer (NK) cell activation assay. (A–C) Donor blood (as a source of NK cells) was incubated with the serum/plasma test samples and human immunodeficiency virus (HIV) peptide pools for 5 h and then analyzed by flow cytometry. The lymphocytes were gated using FSC/SSC scatter. The NK cells were identified as CD3−CD56+ cells and assessed for surface CD107a expression and intracellular IFN-γ expression. (D) Representative dot plots for positive control (purified anti-CD16). (E) Unstimulated (DMSO equivalent solution) control. (F,G) Representative displays of ADCC responder (patient-1) and non-responder (patient-2) to HIV-1 Env-C peptide pool.

Figure 2

Characteristics of ADCC responses in Indian long-term non-progressors (LTNPs) and progressors. Vertical scatter plots (error bars—mean with SEM) denote magnitudes of ADCC-mediated natural killer (NK) cell activation in responders from LTNPs and progressors against HIV-1 antigen peptide pools. Magnitude of ADCC response against (A) structural proteins: HIV-1 B Env, HIV-1 C Env, HIV-1 C Gag, and HIV-1 B Pol and (B) accessory proteins: HIV-1 B Tat, Rev, Nef, and Vpu. The magenta colored dots represent ADCC responders (fulfilled both criteria mentioned in Materials and Methods section). The non-specific recognition of the peptides as indicated by the % NK cell activation by HIV-negative plasma after stimulation by same peptide pools is indicated in black dots. (C) Bar diagram represents the frequency of responses in both cohorts against HIV-1 B and C Env, HIV-1 C Gag, HIV-1 B Pol, Tat, Rev, Nef, and Vpu. The frequencies were higher in LTNPs in case of responses against Env subtype C and Gag, whereas a trend of higher tat-specific responses was observed in LTNPs although not significant (Fisher’s exact 2 × 2 test).

Figure 2

Characteristics of ADCC responses in Indian long-term non-progressors (LTNPs) and progressors. Vertical scatter plots (error bars—mean with SEM) denote magnitudes of ADCC-mediated natural killer (NK) cell activation in responders from LTNPs and progressors against HIV-1 antigen peptide pools. Magnitude of ADCC response against (A) structural proteins: HIV-1 B Env, HIV-1 C Env, HIV-1 C Gag, and HIV-1 B Pol and (B) accessory proteins: HIV-1 B Tat, Rev, Nef, and Vpu. The magenta colored dots represent ADCC responders (fulfilled both criteria mentioned in Materials and Methods section). The non-specific recognition of the peptides as indicated by the % NK cell activation by HIV-negative plasma after stimulation by same peptide pools is indicated in black dots. (C) Bar diagram represents the frequency of responses in both cohorts against HIV-1 B and C Env, HIV-1 C Gag, HIV-1 B Pol, Tat, Rev, Nef, and Vpu. The frequencies were higher in LTNPs in case of responses against Env subtype C and Gag, whereas a trend of higher tat-specific responses was observed in LTNPs although not significant (Fisher’s exact 2 × 2 test).

Figure 3

ADCC responses were associated with higher CD4 count and lower plasma viral load. CD4 counts at the testing visit (N = 92) (X axis) with envelope C peptide pool (Y axis) (p = 0.02) (A) and Tat-specific ADCC responses (Y axis) (p = 0.07) (B). (C,D) Association of plasma viral load at the testing visit (N = 92) with ADCC response against Env-C peptide pool and Tat B peptide pool, respectively. (E) Magnitude of envelope C-specific ADCC response in viremic controllers and non-controllers within long-term non-progressors (LTNPs). The magnitude of ADCC responses [% natural killer cell activation] was comparable in between the groups, but the number of responders to envelope C was high in viremic controllers (8 out of 12) (plasma viral load <2,000 copies/ml) as against 4 out of 22 LTNPs with plasma viral load >2,000 copies/ml. Additionally, the preferential recognition of V3 region (aa 288–330) as indicated by blue dot was seen in four out of eight responders from viremic controller LTNPs as against only one out of four responders from non-controllers.

Figure 4

LTNPs showed cross clade as well as broad ADCC responses. (A) Percentage of the responders (Y axis) to both Env-C and -B peptides is significantly higher in LTNPs as compared to the progressors (p = 0.01 in a 2 × 2 Fischer’s exact test). (B) The breadth of the ADCC response. The number of human immunodeficiency virus antigens (Y axis) recognized by ADCC antibodies from LTNPs were significantly higher than the number of antigens recognized by progressors (chi square test, p < 0.0001).