Antibody responses induced by SHIV infection are more focused than those induced by soluble native HIV-1 envelope trimers in non-human primates

Abstract

The development of an effective human immunodeficiency virus (HIV-1) vaccine is a high global health priority. Soluble native-like HIV-1 envelope glycoprotein trimers (Env), including those based on the SOSIP design, have shown promise as vaccine candidates by inducing neutralizing antibody responses against the autologous virus in animal models. However, to overcome HIV-1's extreme diversity a vaccine needs to induce broadly neutralizing antibodies (bNAbs). Such bNAbs can protect non-human primates (NHPs) and humans from infection. The prototypic BG505 SOSIP.664 immunogen is based on the BG505 env sequence isolated from an HIV-1-infected infant from Kenya who developed a bNAb response. Studying bNAb development during natural HIV-1 infection can inform vaccine design, however, it is unclear to what extent vaccine-induced antibody responses to Env are comparable to those induced by natural infection. Here, we compared Env antibody responses in BG505 SOSIP-immunized NHPs with those in BG505 SHIV-infected NHPs, by analyzing monoclonal antibodies (mAbs). We observed three major differences between BG505 SOSIP immunization and BG505 SHIV infection. First, SHIV infection resulted in more clonal expansion and less antibody diversity compared to SOSIP immunization, likely because of higher and/or prolonged antigenic stimulation and increased antigen diversity during infection. Second, while we retrieved comparatively fewer neutralizing mAbs (NAbs) from SOSIP-immunized animals, these NAbs targeted more diverse epitopes compared to NAbs from SHIV-infected animals. However, none of the NAbs, either elicited by vaccination or infection, showed any breadth. Finally, SOSIP immunization elicited antibodies against the base of the trimer, while infection did not, consistent with the base being placed onto the virus membrane in the latter setting. Together these data provide new insights into the antibody response against BG505 Env during infection and immunization and limitations that need to be overcome to induce better responses after vaccination.

Fig 1. Isolated mAbs induced by BG505 SOSIP are more clonally divergent than those induced by BG505 SHIV.

(A) In total, we successfully expressed 32 and 45 mAbs from two BG505 SOSIP (ROp15 and 99–12) and three BG505 SHIV-infected (6453, 43335, and 6446) NHPs, respectively. (B) CDRH3 amino acid length of individual mAbs isolated from the BG505 SOSIP-immunized and BG505 SHIV-infected NHPs. (C) Phylogenetic analysis using the heavy chain VDJ sequences to assess the evolutionary relationship between the mAbs and to compare the sequence variability between the different NHPs. Similar colors are used as panel A and B to distinguish the mAbs per NHP. (D) Variable gene usage for the heavy and light chains of the isolated mAbs. Average amount of somatic hypermutation (SHM) is indicated for each of the NHPs. SHM is determined as the percentage of nucleotide differences with a germline VH, VL, or VK gene segment.

Fig 2. Autologous NAbs isolated from both the BG505 SOSIP-immunized and BG505 SHIV-infected NHPs.

(A) Pie-charts indicating the amount of isolated mAbs able to neutralize the autologous BG505 T332N virus per NHP. We also included the information from mAbs that were previously isolated from two BG505 SOSIP.664-immunized NHPs (rh1987 and rh2011). The inhibitory dilution (ID₅₀) value next to the pie-chart indicates the autologous NAb titer of the serum at week 20 and week 24 post immunization and infection, respectively. (B) The half maximal inhibitory concentration (IC₅₀) values by which the NAbs neutralize the autologous BG505 T332N, closely-related MG505 A2, and heterologous SF162 viruses.

Fig 3. Epitope mapping of mAbs isolated from BG505 SOSIP-immunized and BG505 SHIV-infected NHPs.

(A) Model of BG505 SOSIP without (left) and with (right) glycans. The different epitopes targeted by mAbs induced by BG505 SOSIP are indicated in different colors. (B) Negative stain-electron microscopy was performed to verify various epitope specificities of the isolated mAbs. MAbs are shown in complex with BG505 SOSIP. For mAbs RM12K (orange), RM54B1 (green), RM15E (blue), and RM35A1 (green) 3D reconstructions are displayed. BG505 SOSIP is displayed in grey. For mAbs RM15F and RM35B1 the 2D class averages are shown. (C) ELISA binding of non-neutralizing mAbs from BG505 SOSIP-immunized NHPs to BG505 SOSIP variant (R500A+Q658K) relative to BG505 SOSIP. The BG505 SOSIP variant (R500A+Q658K) diminishes binding of antibodies to the base of the trimer. Value on the y-axis indicates the fold difference in area under the curve (AUC) of binding to BG505 SOSIP variant (R500A+Q658K) relative to the AUC of BG505 SOSIP.

Fig 4. Potency and epitope specificities of isolated mAbs.

Half maximal inhibitory concentration (IC₅₀) for every mAb isolated from the different NHPs. Epitopes targeted by these mAbs are indicated in different colors showing the relationship between neutralization potency and the various epitopes. Antibodies that have a potency higher than 50 μg/mL are classified as non-neutralizing. We also included the information from mAbs that were previously isolated from two BG505 SOSIP.664-immunized NHPs (rh1987 and rh2011).