Report

HIV-1 therapy with monoclonal antibody 3BNC117 elicits host immune responses against HIV-1

See allHide authors and affiliations

Science  20 May 2016:
Vol. 352, Issue 6288, pp. 997-1001
DOI: 10.1126/science.aaf0972

Insights into antibody therapy for HIV-1

Despite the success of antiretroviral therapy, HIV-1-infected individuals still harbor latent virus. Thus, other therapeutic strategies are needed. A single injection of a broad and potent monoclonal antibody targeting the HIV-1 envelope protein reduced viral loads in HIV-1-infected individuals, albeit only transiently. Lu et al. now report that antibody treatment not only blocked free virus from infecting new cells, it also accelerated the clearance of infected cells. Furthermore, Schoofs et al. demonstrate that therapeutic antibody treatment enhanced infected individuals' humoral response against the virus. Thus, neutralizing antibodies may be a promising therapy for HIV-1 because of their potential to reduce the viral reservoir.

Science, this issue pp. 1001 and 997

Abstract

3BNC117 is a broad and potent neutralizing antibody to HIV-1 that targets the CD4 binding site on the viral envelope spike. When administered passively, this antibody can prevent infection in animal models and suppress viremia in HIV-1–infected individuals. Here we report that HIV-1 immunotherapy with a single injection of 3BNC117 affects host antibody responses in viremic individuals. In comparison to untreated controls that showed little change in their neutralizing activity over a 6-month period, 3BNC117 infusion significantly improved neutralizing responses to heterologous tier 2 viruses in nearly all study participants. We conclude that 3BNC117-mediated immunotherapy enhances host humoral immunity to HIV-1.

Development of serum neutralization breadth during HIV-1 infection typically occurs several years after infection and is seen as a continuum, with ~50% of HIV-1–infected individuals developing some level of broad neutralization and a small fraction of individuals acquiring serum neutralizing activity of extraordinary breadth and potency (14). Antibody cloning experiments have revealed that this activity is due to one or more potent broadly neutralizing antibodies (bNAbs) that target one or more epitopes on the viral spike protein gp160 (1, 510).

bNAbs show exceptional breadth and potency in vitro and can protect against or suppress active infection in humanized mice (1113) and macaques (14, 15). Moreover, in a phase I clinical trial, a single injection of 3BNC117, a CD4 binding-site–specific bNAb (6), was safe and effective in suppressing HIV-1 viremia by an average of 1.48 logs (16).

In addition to direct effects on target cells and pathogens, antibody-mediated immunotherapies have the potential to engage the host immune system and induce both innate and adaptive immune responses (17). In particular, the Fc domains of antibodies interact with receptors on innate cells such as natural killer (NK) cells and phagocytes to enhance the clearance of viral particles and the killing of infected cells (18). To test the hypothesis that bNAb-mediated immunotherapy can enhance immunity to HIV-1 in humans, we examined the serologic responses to the virus in individuals who received 3BNC117.

A single 3BNC117 infusion was administered to HIV-1–infected individuals at doses of 1, 3, 10, or 30 mg per kilogram of body weight (mg/kg) (Fig. 1A and table S1A) (16, 19). To determine whether 3BNC117 therapy is associated with changes in viral sensitivity and serologic responses to autologous viruses, we cultured HIV-1 from peripheral blood mononuclear cells (PBMCs) of nine viremic individuals before [day 0 (d0)] and 4 weeks after 3BNC117 infusion (16). On d0, all but one of the cultured viruses were sensitive to 3BNC117, with median inhibitory concentration (IC50) values ranging from 0.09 to 8.8 μg/ml [Fig. 1B and (16)]. At week 4, we found increased resistance to 3BNC117 in most individuals, indicating selection for viral escape variants [Fig. 1B and (16)].

Fig. 1

Virus sensitivity to 3BNC117 and autologous antibody responses. (A) Kinetics of 3BNC117 antibody decay in HIV-1–infected individuals as determined by a validated anti-idiotype enzyme-linked immunosorbent assay (ELISA) (16). Shown are mean values of patients infused in each respective dose group. Each patient sample was measured in duplicates. The gray-shaded area indicates the lower level of accuracy of the assay (2 μg/ml). Red arrows indicate the time points of IgG purification. (B) Autologous virus sensitivity to 3BNC117 before (d0, gray) and 4 weeks after (black) 3BNC117 infusion. The y axis shows IC50’s for 3BNC117 on viral culture supernatants from PBMCs determined by TZM.bl assay. Neutralization assays performed in duplicates. (C) The AUC of the neutralization curves of purified IgGs obtained from sera on d0 (orange) or week 24 (green) against d0 (left) or week 4 (right) autologous viruses. Neutralization assays were performed in duplicates. P values were determined by Wilcoxon signed-rank test.

When the same viral isolates were tested for sensitivity to the matched individual’s immunoglobulins (IgGs) obtained before (d0) and 24 weeks after 3BNC117 infusion (Fig. 1A), we found increased neutralizing activity in the week 24 IgG against both d0 and week 4 autologous viruses (P = 0.0078, Fig. 1C and table S2). Thus, although 3BNC117 infusion selected for 3BNC117-resistant HIV-1 variants, neutralizing antibody responses continued to develop against autologous viruses (20).

To test for changes in heterologous neutralizing activity after 3BNC117 treatment, we assayed individuals’ d0 and week 24 IgG against a panel of tier 1 (n = 1) and tier 2 (n = 12) HIV-1 pseudoviruses that included globally circulating HIV-1 strains (21) (Fig. 2 and tables S1, S3, and S4). Neutralizing activity was compared between the two time points by measuring the area under the neutralization curve (AUC) for patients’ isolated IgG against each virus (table S4B). 15 participants that received 3BNC117 were not on antiretroviral therapy (ART) and had starting viral loads from 640 to 53,470 copies/ml (table S1A). Control IgGs were obtained from 36 viremic individuals who did not receive 3BNC117 and had starting viral loads ranging from 150 to 303,200 copies/ml (Fig. 2 and table S1B).

Fig. 2 Heterologous antibody responses.

(A) The difference in overall AUC (mean AUC change) per individual in TZM.bl assays against 13 heterologous viruses (Fig. 2D) for d0 versus week 24 IgG obtained from 36 untreated viremic controls (mean sampling interval 26.8 weeks), 15 viremic individuals infused with 3BNC117 (mean sampling interval 24.1 weeks), and 12 ART-treated individuals receiving 3BNC117 infusion (mean sampling interval 24.0 weeks) (16). Neutralization assays performed in duplicates. P values were determined by unpaired Wilcoxon test (rank-sum test). (B) The aggregated differences in AUC between d0 and week 24 IgG assayed by TZM.bl for all viruses and all individuals. Each dot represents a single AUC difference for a single virus from one individual displayed in (A). Colored bars represent the mean of all AUCs. Whiskers show standard deviation. P values were determined using generalized estimating equations (38). (C) 3BNC117 antibody levels (ELISA, white) and TZM.bl neutralization titer against tier 2 strain Q769.d22 (green) in patient 2A3. (D) Mean AUCs of IgGs of all individuals at d0 (gray) and week 24 (color of respective group) for each HIV-1 pseudovirus tested. Changes in neutralization of viremic control individuals without 3BNC117 infusion are shown in yellow (left). Change in neutralization of 3BNC117-treated individuals are shown in dark blue (off ART, middle) and light blue (on ART, right). P values were determined by unpaired Wilcoxon test (rank-sum test). Red stars indicate significant P values after Bonferroni correction (threshold P < 0.0038).

During a 6-month observation period, control individuals’ neutralizing activity showed no consistent improvement in either breadth or potency (Fig. 2, A and B, figs. S1A and S2, and tables S4 and S5A) (4, 22). In contrast, all but one of the 15 viremic individuals infused with 3BNC117 showed increased breadth and/or potency against the pseudovirus panel at week 24 (P = 7.1 × 10−7, Fig. 2A, figs. S1B and S2, and tables S4 to S6). The absolute change in neutralizing activity varied between viruses and individuals, ranging from small effects to dramatic increases as observed in patient 2A3 for viral strain Q769.d22 (Fig. 2C and tables S4 to S6). Significant differences were also evident between 3BNC117-treated and control groups regardless of whether sera from all individuals were considered in aggregate or examined against individual viruses (P = 1.9 × 10−9, Fig. 2, B and D).

In addition to viremic patients, we examined 12 individuals that received 3BNC117 while on ART, with no detectable or low-level viremia (<20 to 100 copies/ml). In comparison to viremic patients, the increase in heterologous neutralizing activity was significantly less pronounced in ART-treated individuals (P = 0.037; Fig. 2, A, B, and D; figs. S1B and S2; and tables S3 to S5).

The observed improvement in neutralizing activity could not be explained by confounding factors such as differences in initial viral load or CD4+ T cell levels (fig. S3 and tables S1 and S7). Moreover, we found no correlation between d0 neutralizing activity and neutralization improvement (fig. S4). A comparison of the pattern of neutralization increase with 3BNC117’s neutralization profile ruled out that remaining antibody was responsible for the effect (fig. S5 and table S8). We conclude that 3BNC117 enhances host immunity to heterologous tier 2 HIV-1 viruses irrespective of initial neutralization breadth and potency.

To examine the effects of 3BNC117 immunotherapy on the plasma viral population of treated individuals, we performed single-genome sequencing (SGS) of over 1000 plasma-derived gp160 env genes (gp160) before (d0) and 4 (6), 12, or 24 weeks after infusion (Fig. 3, A and B, figs. S6 to S10, and table S9). With the exception of two individuals who were sexual partners, all other volunteers had epidemiologically unrelated infections (Fig. 3A). On d0, env sequences from patients 2A1, 2A3, and 2C4 comprised multiple lineages, which was reflected in a multimodal distribution of pairwise diversity measurements from these individuals (Fig. 3B and fig. S6). Analysis of env sequences from subsequent time points revealed significant shifts in both nucleotide (six out of nine individuals, Fig. 3B) and amino acid sequence (seven out of nine individuals, fig. S6) diversity. Consistent with the observation that env diversity is associated with neutralization breadth (2325), there was a strong correlation between the initial level of neutralizing activity and the initial diversity of the circulating viral swarm (correlation coefficient = 0.92, Fig. 3C).

Fig. 3 HIV-1 quasispecies diversity before and after 3BNC117 infusion.

(A) Maximum-likelihood phylogenetic tree of single-genome–derived env gene sequences from d0 plasma, before therapy with 3BNC117 (table S9). Asterisks indicate bootstrap values of 100%. Individual viral sequences are color-coded as indicated. (B) Scatter plots depicting pairwise nucleotide sequence diversity of plasma env sequences on d0 and weeks 4 (2E5, week 6), 12, and 24 after infusion. Each dot represents the pairwise genetic difference between two sequences at a given time point. Colored bars indicate median diversity, whereas black bars indicate the interquartile range. P values were determined using a two-sample U statistic–based Z test (3941). (C) Graph shows the relationship between d0 mean heterologous neutralizing AUC against a panel of tier 1 (n = 1) and tier 2 (n = 12) viruses (x axis) and the median pairwise nucleotide diversity for each patient (y axis). R2, correlation coefficient.

We next evaluated viral sequence evolution in each of the 3BNC117-treated patients over time. Shifts in the viral quasispecies were evident regardless of initial 3BNC117 neutralization sensitivity and bNAb dose (Fig. 4 and fig. S7). However, the nature of these shifts differed depending on the individual (Fig. 4 and figs. S7 to S9). For example, in patient 2A1, 15 out of 27 (15/27) d0 sequences fell into a single clade marked “group A” (Fig. 4A and fig. S8). Four weeks after 3BNC117 infusion, group A viruses contracted (2/25 sequences) and group C viruses expanded (16/25). At week 24, the viral quasispecies was primarily composed of group B and D viruses (Fig. 4 and fig. S8). This pattern of “clade shifting” was also seen in patients 2A3 and 2C4 (fig. S7). Patients with lower initial env diversities, such as 2E1, did not harbor distinct viral sublineages at d0 (Figs. 3 and 4A) but continued to accrue mutations, some of which became fixed during the 24-week follow-up (for example, changes in V1/V2 in 2E1, fig. S9).

Fig. 4 Antibody responses to the evolving viral quasispecies.

(A) Maximum-likelihood phylogenetic trees of single-genome–derived env gene sequences from patients 2A1 and 2E1 sampled on d0 and weeks 4, 12, and 24 after 3BNC117 infusion (left). Clades with bootstrap support ≥70% are indicated by a black asterisk and are arbitrarily named groups A to D in the case of patient 2A1. Bar graphs (middle) indicate the time points from which sequences in the tree are derived. Heat maps (right) show the 3BNC117 IC50, d0 IgG IC50, and week 24 IgG IC50 values against autologous pseudoviruses using env sequences as indicated by colored stars. Neutralization assays were performed in duplicates. (B) Sequence logo plots illustrating longitudinal amino acid changes in and around known 3BNC117 contact residues (26, 27) in patients 2A1 and 2E1. Letters indicate deviations from the d0 consensus shown at the top, whereas white spaces indicate agreement with the d0 consensus. Colors indicate basic (dark blue) and acidic (red) residues, and a turquoise “O” is used instead of “N” to indicate a potential N-glycosylation site. Logo plots were generated using LASSIE (28). Plus symbols indicate 3BNC117 contact residues confirmed by two crystal structures (26, 27).

To assess viral sequence changes after 3BNC117 infusion, we generated longitudinal logo plots depicting 3BNC117 contact residues (26, 27) for each patient (Fig. 4B and figs. S7 and S10). Although viruses from all nine patients exhibited mutations within 3BNC117 contact residues relative to the d0 consensus sequence, their number and position varied considerably, as exemplified by patients 2A1 and 2E1 (Fig. 4B and figs. S7 and S10). Using LASSIE (Longitudinal Antigenic Sequences and Sites from Intrahost Evolution) (28), we scanned the entire env protein sequence for sites selected within the 24-week time frame (selection cutoff ≥ 80%) (table S10). Although selected sites were identified in all patients, no consistent mutational pattern was observed (table S10). These data suggest that 3BNC117 immunotherapy is associated with shifts in circulating quasispecies and a number of different env mutations, some of which persist even after the infused antibody levels drop below detection.

To better understand the virus-host interactions that led to the development of enhanced heterologous neutralizing breadth, we performed neutralization assays on 63 pseudoviruses expressing the gp160s found in the circulation on d0 and weeks 4, 12, and 24 from five individuals (Fig. 4, fig. S7, and table S11). The pseudoviruses were tested for sensitivity to the corresponding individual’s IgG obtained on d0 and week 24. In all cases, we were able to identify d0 or week 4 viruses that exhibited greater neutralization sensitivity to week 24 IgG as compared to d0 IgG (Fig. 4, fig. S7, and table S11). For example, all tested 2A1 and 2E1 viruses were 3BNC117-sensitive and exhibited a week 24/d0 fold change of ~1.7 and ~4.8 in IgG IC50, respectively (Fig. 4). On the other hand, all tested 2C4 viruses were 3BNC117-resistant (mean IC50 >20 μg/ml), yet they were ~56.5-fold more sensitive to week 24 IgG versus d0 IgG (fig. S7). In conclusion, viremic individuals receiving 3BNC117 produced antibodies to autologous viruses that were both sensitive and resistant to 3BNC117.

Although exceptional bNAbs to HIV-1 develop only sporadically in a fraction of infected individuals, most HIV-1–infected individuals develop some level of neutralization breadth (14). Here we show that 3BNC117 immunotherapy accelerates this process. This boost in heterologous breadth occurs irrespective of demographic, virologic, or dosage factors and was associated with both transient and lasting changes to the viral quasispecies. It is of note that the neutralization improvements observed were modest in most individuals, potentially owing to the transient nature of therapy with a single antibody as well as the short time frame of observation.

Although the effect of 3BNC117 on neutralizing responses to heterologous HIV-1 viruses may seem surprising, antibodies to HIV-1 have been associated with enhanced immunity in infants born to HIV-1–infected mothers that have circulating antibodies to HIV-1 and macaques treated with monoclonal antibodies or neutralizing serum (2931).

How passively administered antibodies to HIV-1 accelerate the emergence of bNAbs is not completely understood. One possibility is that 3BNC117 infusion selected for viral variants with altered antigenic properties, which in turn stimulated new B cell lineages (2325, 3234). A second possibility is that immune complexes formed by 3BNC117 and circulating viruses act as potent immunogens, a phenomenon that is believed to be responsible for the enhanced CD8+ T cell immunity to tumor antigens in individuals receiving monoclonal antibody–based immunotherapy (3537).

Irrespective of the mechanism(s), the enhanced antibody response found in individuals receiving 3BNC117 therapy indicates that immunotherapy boosts host immunity to HIV-1. Moreover, the finding that antibody responses to heterologous tier 2 viruses develop in nearly all 3BNC117-treated individuals suggests that host genetics or a specific viral envelope sequence do not limit the development of neutralizing antibodies to HIV-1.

Supplementary Materials

www.sciencemag.org/content/352/6288/997/suppl/DC1

Materials and Methods

Figs. S1 to S10

Tables S1 to S11

References (4252)

References and Notes

  1. Materials and methods are available as supplementary materials on Science Online.
Acknowledgments: We thank all study participants, who devoted time to our research. We thank the Rockefeller University Hospital Clinical Research Support Office, the nursing staff for patient care and recruitment, the clinical study group of the Infectious Disease Division at the University Hospital Cologne, and all members of the Nussenzweig lab for helpful discussions. We thank M. Schechter and C. Baro for technical assistance, P. Fast and H. Park for clinical monitoring, E. Gotschlich and B. Coller for input on study design, and P. Hraber for helping with LASSIE analyses. The data reported in this study are tabulated in the main paper and in the supplementary materials. Envelope single-genome sequencing data can be downloaded from GenBank (accession numbers KX027737 to KX028736). This work was supported in part by the Bill and Melinda Gates Foundation Collaboration for AIDS Vaccine Discovery, grants OPP1032144 (M.S.S.) and OPP1092074 and OPP1124068 (M.C.N); the Robertson Foundation to M.C.N.; the NIH Center for HIV/AIDS Vaccine Immunology and Immunogen Discovery, grants 1UM1 AI100663-01 (M.C.N) and 1UM1 AI00645 (B.H.H.); the University of Pennsylvania Center for AIDS Research Single Genome Amplification Service Center P30, grant AI045008 (B.H.H.); and NIH grants UM1AI068618 (M.J.M.), UM1AI069481 (M.J.M.), F30 AI112426 (E.F.K), and HIVRAD P01 AI100148 (P.J.B.). T.S. is supported by a Deutsche Forschungsgemeinschaft postdoctoral fellowship (grant SCHO 1612/1-1). F.K. is supported by the Heisenberg-Program of the Deutsche Forschungsgemeinschaft (grant KL 2389/2-1); the European Research Council (grant ERC-StG639961); and the German Center for Infection Research, partner site Bonn–Cologne, Cologne, Germany. M.B. is supported by the German National Academic Foundation. J.C.C.L. is supported by an award from the Conselho Nacional de Desenvolvimento Científico e Tecnológico “Ciencia sem Fronteiras” (grant 248676/2013-0). M.C.N. is a Howard Hughes Medical Investigator and an inventor on U.S Patent Application no. 14/118,496, filed by Rockefeller University related to 3BNC117.
View Abstract

Navigate This Article