Novel Biopanning Strategy: A Tool to Identify Epitopes Associated with Vaccine Protection

Transcription

1 JVI Accepts, published online ahead of print on 6 February 2013 J. Virol. doi: /jvi Copyright 2013, American Society for Microbiology. All Rights Reserved Novel Biopanning Strategy: A Tool to Identify Epitopes Associated with Vaccine Protection Barbara C. Bachler a,b, Michael Humbert a,c, Brisa Palikuqi a*, Nagadenahalli B. Siddappa a,c, Samir K. Lakhashe a,c, Robert A. Rasmussen a,c, Ruth M. Ruprecht a,c# Running Title (53/54): Phage Display to Identify Protection-linked Epitopes a Department of Cancer Immunology and AIDS, Dana-Farber Cancer Institute, Boston, MA 02115, USA, b VetCore Facility for Research, University of Veterinary Medicine Vienna, 1210 Vienna, Austria, c Harvard Medical School, Boston, MA 02115, USA. * Present address: Weill Cornell Graduate School of Medical Sciences, New York, NY 10065, USA. Present address: Department of Pathology and Laboratory Medicine, Emory University, Atlanta, GA 30329, USA. # Corresponding author: Department of Cancer Immunology and AIDS, Dana- Farber Cancer Institute, 450 Brookline Ave., JFB 809, Boston, MA Phone: Fax: ruth_ruprecht@dfci.harvard.edu Abstract word count: 243/250 Text word count: 5,

2 Identifying immune correlates of protection is important to develop vaccines against infectious diseases. We designed a novel, universally applicable strategy to profile the antibody (Ab) repertoire of protected vaccine recipients, using recombinant phages encoding random peptide libraries. The new approach, termed protection-linked (PL) biopanning, probes the Ab paratopes of protected vaccinees versus those with vaccine failure. As proofof-concept, we screened plasma samples from vaccinated rhesus macaques (RMs) that had completely resisted multiple mucosal challenges with R5-tropic simian-human immunodeficiency viruses (SHIVs). The animals had been immunized with a multi-component vaccine (multimeric HIV-1 gp160, HIV-1 Tat and SIV Gag-Pol particles). After PL-biopanning, we analyzed the phagotopes selected for amino acid homologies; in addition to the expected Env mimotopes, one recurring motif reflected the neutralizing Ab epitope at the N- terminus (NT) of HIV-1 Tat. Subsequent binding and functional assays indicated that anti-tat NT Abs were present only in completely or partially protected RMs; peak viremia of the latter was inversely correlated with anti-tat NT Ab titers. In contrast, highly viremic, unvaccinated controls did not develop detectable Abs against the same epitope. Based upon the protective effect observed in vivo, we suggest that Tat should be included in multi-component HIV-1 vaccines. Our data highlight the power of the new PL-biopanning strategy to identify Ab responses with significant association to vaccine protection regardless of the mechanism(s) or targets of the protective Abs

3 50 51 PL-biopanning is also unbiased with regard to pathogens or disease model, making it a universal tool

4 Introduction The design of a safe, effective vaccine is desirable to control the AIDS pandemic. In order to improve future vaccine candidates, it is important to understand the protective mechanism(s) leading to reduced virus acquisition (1). Thus, correlates of infection risk for the recent RV144 vaccine efficacy trial (2) are the subject of intense study (3, 4). However, additional information can be gained by dissecting immune responses in biologically relevant animal models where immunization induced complete protection from immunodeficiency virus challenge(s) (5, 6). Recent results of active (7, 8) and passive (9) immunization trials in humans and in non-human primates suggested that anti-hiv-1 Env antibodies (Abs) can have protective roles even in the absence of virus neutralization, implying that other Ab-mediated mechanisms may be involved in preventing mucosal virus acquisition. Thus, the ability to profile Ab epitopes associated with protection might help to identify potential targets for future HIV-1 vaccines. Random peptide phage display has been used to dissect Ab responses in the context of primate immunodeficiency virus infections, and several studies have identified HIV-1 or simian-human immunodeficiency virus (SHIV)-specific peptides (10-14). We developed a novel epitope selection strategy, termed protection-linked (PL) biopanning, which was designed to differentiate between protective vaccine-induced Ab responses and those that failed to protect. PLbiopanning can be used in any context where Ab-mediated responses might be involved in a positive vaccine outcome

5 As proof-of-concept, we focused on vaccine efficacy studies in the rhesus macaque (RM)/SHIV model. A recombinant multi-component protein vaccine (containing multimeric HIV-1 clade C (HIV-C) gp160, HIV-1 clade B Tat and SIV Gag-Pol particles) induced protective immune responses in RMs (15). Upon multiple live-virus exposures with a R5-tropic SHIV that encoded a heterologous HIV-C envelope (SHIV-C), we observed complete as well as partial protection. Further analysis identified two parameters that were significantly associated with protection: cellular immunity and neutralizing Ab (nab) titers (15). Based upon these data, we proposed that the nab repertoire in the vaccine-protected RMs differed significantly from that in the RMs with vaccine failure. To probe the paratopes associated with vaccine protection, we applied PL-biopanning. Unexpectedly, we identified a protection-linked epitope that was not related to Env and represented a neutralizing epitope of HIV-1 Tat. This finding demonstrates the power of the new method and its lack of bias regarding the mechanism or target of the Abs associated with protection

6 Materials and Methods Animals We studied Indian-origin RMs (Macaca mulatta) that had been enrolled in three different vaccine/challenge studies described elsewhere (15-17). They were housed at the Yerkes Regional Primate Research Center (Atlanta, GA), and all procedures were approved by the Animal Care and Use Committees of Emory University and the Dana-Farber Cancer Institute (DFCI). Table 1 summarizes the immune status of each animal examined in this study and lists the immunogens and challenge viruses used. The levels of protection are given according to the original publications (15-17), with minor changes for vaccinees RGe-11 and RFo- 11 and control animal RAk-11 (15), since only immune responses during the lowdose challenge phase were analyzed in the present study. For all vaccinees, plasma samples collected at two different time points were evaluated for Ab responses: week 0, after the completion of all immunizations but before the first virus challenge, and week 7, after multiple live-virus exposures. Since monkey RAt-9 had been part of a different immunization/challenge schedule (16), we analyzed similar time points: week -2, day of last immunization (week 0 was not available) and week 30, the day of the second virus challenge (thus after immunization and after the first, but before the second challenge). Biopanning with recombinant phages Plasma samples of protected animals were screened by peptide phagedisplay. Polyclonal RM plasma Abs were captured on paramagnetic beads, - 6 -

7 coated with either rabbit anti-monkey IgG (as published earlier (10)) (Sigma- Aldrich, St. Louis, MO; Dynabeads M-280 tosylactivated; Life Technologies, Grand Island, NY) or protein G (Dynabeads Protein G; Life Technologies). The positive selection used either polyclonal RM plasma in combination with an anti- V3 neutralizing monoclonal Abs (nmabs) (33C6 (18)) (spiked PL-biopanning, experiment #I, Table 2) or the plasma alone (experiments #II-IV, Table 2). For the spiked PL-biopanning, the final concentration of the mabs added was 1% of the total IgG coated on beads. We used this concentration based on the approximate IgG concentration in RM sera (19) and the required amounts of a similar anti-v3 mab leading to complete protection in vivo (20). In the next step, beads were incubated with original phage-display peptide libraries (7mer, cyclic 7mer, 12mer; New England Biolabs, Ipswich MA). After intense washing, bound phages were eluted by ph shift with 0.2 M glycine-hcl, ph 2.2, supplemented with 1 mg/ml bovine serum albumin (BSA) (Sigma-Aldrich) and neutralized with 1 M Tris-HCl ph 9.1 (Sigma-Aldrich). Eluted phages were used in a round of negative selection. Unbound phages were amplified in E. coli (ER2738, New England Biolabs), precipitated overnight at 4 C (20% PEG-8000/2.5 M NaCl; Fisher Scientific, Fair Lawn NJ), and subjected to two more rounds of selection. After the third positive selection, the eluted phages were titered, single clones were picked, and tested by phage ELISA for specific binding. Single-stranded DNA of specific clones was sequenced

8 Computational analysis of mimotopes Peptide sequences were grouped into motifs and either aligned with the published epitope of mab 33C6 (18) or assigned to protein sequences of the immunogens (HIV-C gp160, HIV-1 Tat or SIV Gag-Pol) using the PNNL PepAligner software ( Graphical representation of frequent motifs was generated using WebLogo3 ( (21, 22). The three-dimensional location of the motifs on Tat (PDB-ID: 1JFW (23)) were prepared with Chimera (24). Phage ELISA and cross-reactivity profile of mimotopes Microtiter plates (Greiner-Bio-One GmbH, Frickenhausen, Germany) were coated overnight (4 C) with 1:2,000 diluted plasma samples (100 µl/well, in carbonate-bicarbonate buffer, Sigma-Aldrich) and blocked (2 h, room temperature, 200 µl/well) the next day with 3% casein (Sigma-Aldrich) in phosphate buffered saline (PBS)/0.5% Tween-20 (PBSCT). Single phage clones were amplified overnight and 70 μl of the phage culture were added to 30 μl PBSCT. As negative control, M13KO7 helper phages (New England Biolabs) without peptide insert were included (wildtype phages). The next day, bound phages were detected using an anti-m13 horseradish peroxidase (HRP)- conjugated Ab (1:2,000 in PBSCT; GE Healthcare Biosciences Corp., Piscataway NJ) and o-phenylenediamine dihydrochloride (OPD; Invitrogen, Frederick, MD) + H 2 O 2. After stopping the reaction with 1 N H 2 SO 4, the plates were read at 490/620 nm

9 Tat mimotopes were analyzed for binding activity to polyclonal plasma Abs of different vaccinees. Optical density (OD) signals at least 10x higher than signals detected with the wildtype phages were considered positive and the cross-reactivity profile of each mimotope was expressed as a color-coded heatmap. Reagents The following reagents were obtained through the AIDS Research and Reference Reagent Program (ARRRP), Division of AIDS, NIAID, NIH: (1) HIV consensus subtype B Tat (15-mer) peptides - complete set (#5138); (2) HIV-1 Tat protein, clade B (#2222); (3) monoclonal Ab (mab) to HIV-1 BRU Tat (NT3 2D1.1, #4138); (4) HIV-1 Tat mab 2A4.1 (#4373); (5) HLM1 cells (#2090). The anti-v3 mab 33C6 was produced as previously described (18). MAb HGN194 (25) was a kind gift of Drs. Davide Corti and Antonio Lanzavecchia (Humabs BioMed SA, Bellinzona, Switzerland). mab ELISA Microtiter plates were coated overnight with mab 33C6 (1 µg/ml), HGN194 (1 µg/ml), NT3 2D1.1 (2 µg/ml), 2D4.1 (2 µg/ml) or human IgG1 Herceptin (2 µg/ml) (all 100 µl/well, in carbonate-bicarbonate buffer). Blocking, phage incubation, and detection of bound phages were performed as described above

10 Purification of plasma IgG Plasma samples from vaccinees RBr-11 and RJr-11 (pool of weeks 2 and 3) were used for IgG purification according to the manufacturer s instructions (NAb Protein A/G Spin Kit, Thermo Scientific, Rockford, IL). Eluted IgG was concentrated using Amicon Ultra-4 Centrifuge Filter Devices (100K, Millipore Corporation, Billerica, MA), and its binding specificity to NT Tat peptides was confirmed by ELISA (see below). In parallel, IgG from two naïve RMs was prepared and included as negative control. Tat ELISA Microtiter plates were coated with either 1 µg/ml of a recombinant Tat protein; 50 μl/well carbonate-bicarbonate buffer (Sigma-Aldrich) or 2 µg/ml of overlapping N-terminal Tat peptides (pool of #5113, 5114 and 5115) in 100 μl/well carbonate-bicarbonate buffer). After blocking, plasma samples (in serial dilutions), mabs (1 µg/ml), or purified IgG (different concentrations, starting with 4 µg/ml) were added to the coated peptides/protein (100 μl/well, in PBSCT). The next day, Ab binding was detected using a HRP-conjugated anti-monkey IgG (1 h at room temperature, 1:2,000 in PBSCT; Sigma-Aldrich). The plates were developed as described above. Binding Ab titers were calculated by linear regression and defined as reciprocal plasma dilution with an absorbance 5x higher than reactivity of autologous pre-immune plasma. Binding of the anti-tat mabs (NT3 2D1.1 and 2D4.1) to Tat peptides was compared at OD 490/620 nm. An unrelated human IgG1 mab was included as negative control (Herceptin)

11 Tat neutralization assay HLM1 cells (HeLa-CD4-expressing cell line containing an integrated, Tatdefective HIV-1 provirus) were used for viral rescue assays (26). In this provirus (HXB2), the initiation codon of tat had been mutated to a termination codon. The assay consists of supplying exogenous Tat protein and monitoring viral replication by measuring the viral core antigen (p24) released into cell supernatants (as described previously (27)). Briefly, HLM1 cells were seeded into 96-well plates (20,000 cells/well, in 200 µl of MEM (Sigma-Aldrich)), supplemented with 5% equine serum (Life Technologies). Purified plasma IgG or anti-tat mabs were prepared in serum-free medium (100 µl, in duplicates), starting with a concentration of 600 µg/ml (polyclonal IgG) or 40 µg/ml (mabs), respectively. Five µg/ml of Tat protein was added to each well and incubated at 37 C for 30 min. The mixture was then transferred to the appropriate wells containing cells. After 3 h, the supernatant was removed and the cells were washed 4x to remove residual Tat protein and/or Abs, followed by addition of 200 µl of complete growth medium. The level of viral antigen p24 was determined 72 h later (Advanced Bioscience Laboratories, Kensington, MD). As a negative control, purified naïve RM IgG was included. The percentage of transactivation (based on p24 read-out) was calculated with regard to the p24 read-out in the wells containing cells plus Tat protein only (considered as 100% transactivation)

12 Epitope mapping using overlapping Tat peptides Microtiter plates were coated with 2 µg/ml of overlapping peptides spanning the entire Tat protein (# ) in 100 μl/well carbonatebicarbonate buffer; in duplicates). After blocking, plasma samples from RMs RDo-11, RJr-11 and RBr-11 (collected at week 1 and diluted 1:1,500 in PBSCT) were added and incubated overnight at 4 C. The next day, Abs were detected as described above. As negative control, a scrambled C-terminal gp120 peptide was included (24 AA, GVTKYIPGSIPVEGLKSHKAGSYK, Molecular Biology Core Facilities, DFCI, Boston, MA). OD signals 20x higher than detected with the control peptide were considered positive. Statistical analysis Associations between peak viremia and anti-tat Ab titers were assessed using Spearman correlation analysis (P values <0.05 were considered statistically significant). The percentage of transactivation in the HLM-1 cellbased assay between naïve and immune IgG was compared using an unpaired, two-tailed Student s t-test (including Bonferroni correction: P values <0.025 were considered statistically significant). All statistical analysis was performed using GraphPad Prism 5 for Windows, GraphPad Software

13 Results Background: Vaccinees with different levels of protection In the present study, we examined Ab responses in vaccinated RMs described elsewhere (15-17, 28, 29). After completion of all immunizations (multimeric HIV-1 clade C (HIV-C) gp160, HIV-1 clade B Tat, SIV Gag-Pol particles) and virus challenges (SHIV-1157ipEL-p (30), SHIV-1157ip or SHIV- 1157ipd3N4 (31), respectively), we observed different levels of protection among the vaccinees (Table 1). These results imply varying Ab specificities among the vaccinees. Thus, we asked the question: did the vaccine-induced Ab repertoire in RMs that had resisted multiple low-dose SHIV-C exposures differ from that in animals where the vaccine had failed to protect? If so, what targets were recognized by Abs present in monkeys with vaccine success? To address these issues, we designed a selection strategy based upon random peptide phage-display, which distinguishes between the vaccine-induced Ab repertoires in protected versus non-protected monkeys. This procedure, termed protection-linked (PL) biopanning, consists of three consecutive positive/negative rounds of selection with plasma samples collected at week 0, after immunizations, but before any virus challenge (Fig. 1). We analyzed protected RMs (positive selection) in combination with plasma from a nonprotected vaccinee given the same immunogens (negative selection) and hypothesized that by depleting phages recognizing Abs that only bind but were not associated with protection, we could enrich for phages representing epitopes recognized only by protective Abs (Fig. 1)

14 Spiking plasma of a vaccinated RM with a known nmab: Can PLbiopanning identify the expected epitope? Before dissecting the Ab responses of polyclonal plasma samples, we sought to verify that the proposed strategy can indeed identify Ab epitopes linked to a protective mechanism, such as neutralization. Thus, we performed a control experiment using polyclonal plasma from one of the protected monkeys (RGe- 11) as a positive selector, and spiked it with a known nmab (spiked PLbiopanning, experiment #I, Table 2). The negative counter-selection used plasma from a non-protected vaccinee (RDo-11) that had been enrolled in the same study as monkey RGe-11 that had been used for positive selection (experiment #I, Table 2). For this control experiment, we chose nmab 33C6 (18). Originally, this Ab had been isolated from RMs using a conformation-dependent mimotope (Tc.2) and a novel technology for mimotope-specific isolation of single memory B cells by flow cytometry (18). 33C6 targets the conformational V3 loop crown and cross-neutralizes viruses of different clades. The peptide inserts of specific phages (phagotopes) isolated by spiked PL-biopanning were sequenced and examined for homologies with the published epitope of the 33C6 (18). The phagotopes thus assigned were termed 33C6-specific protection-linked mimotopes (abbreviated: 33C6 PL-mimes). Fig. 2A shows the linear alignment of 33C6 PL-mimes and highlights the homologies with the original mimotope (Tc.2) that had been used for Ab isolation (18). Of note, some of the 33C6 PL-mimes were cross-recognized by a second anti-v3 mab (HGN194). HGN194 is also

15 conformation-dependent and was recently shown to be completely protective in vivo (20) (Fig. 2B). These data are proof-of-concept that spiked PL-biopanning can identify the expected epitope of the nmab added. Further with 33C6 mimes being cross-recognized by protective nmab HGN194, we indirectly verified that PL-biopanning can indeed select mimotopes that are linked to Ab-mediated protection (such as Env-dependent neutralization). PL-biopanning to identify Ab epitopes associated with protection Next, we performed PL-biopannings for three protected vaccinees (RRi- 11, RTr-11 and RGe-11) using the non-protected monkey, RDo-11 as negative selector (Fig. 1 and biopanning #II, Table 2). The phagotopes were sequenced and examined for HIV-1 protein mimicry. Such mimotopes were termed protection-linked mimotopes (abbreviated: PL-mimes). After completion of all immunizations, the vaccinees examined in this study had developed detectable anti-hiv-1 Env Ab responses (15, 17, 29). As expected, each PL-biopanning revealed different Env-specific epitopes that are the topic of a different detailed analysis (data not shown). However, when we analyzed the sequences of the protection-linked (PL) mimotopes isolated using plasma from protected vaccinee RGe-11, not all of them could be aligned to HIV-1 Env. We extended our computational analysis and included sequences of HIV-1 Tat IIIB and SIV Gag- Pol, which had also been used as part of the multi-component vaccine (listed in Table 1) (15-17). This led to the unexpected observation that some of the phagotopes represented a conserved motif mimicking an area at the N-terminus

16 (NT) of HIV-1 Tat. The corresponding phagotopes were termed protection-linked Tat mimotopes (abbreviated: PL-Tat mimes) (biopanning #II, Fig. 3A). Of note, in contrast to the spiked PL-biopanning described above (experiment #I), PL-biopanning using RGe-11 plasma alone (without adding nmab 33C6) did not lead to the selection of any phagotopes mimicking the epitope of nmab 33C6. This confirms again that the spiked PL-biopanning (experiment #I) had identified mimotopes specific for the nmab added. Cross-recognition profile of PL-Tat mimes To investigate whether these mimotopes reflect Ab epitopes specific for several protected vaccinees, we performed a phage-cross ELISA by testing plasma samples of 14 vaccinees (Table 1) for their binding activity to the PL-Tat mimes isolated using RGe-11 (biopanning #II). For each vaccinee, two different time-points were tested: 1) after immunization, but before any challenge (week 0, or week -2 in case of RAt-9, respectively) to evaluate vaccine-induced Ab responses; and 2) after mucosal challenges (week 7, or week 30 in case of RAt- 9, respectively) to examine possible changes in the Ab repertoire due to the livevirus exposures (Fig. 4A). As expected, RGe-11, the positive selector, showed high binding signals for all PL-Tat mimotopes tested, whereas RDo-11, the negative selector, did not recognize any of them. Most of the PL-Tat mimes were only recognized by plasma Abs from vaccinees with evidence of partial or complete protection. The partially protected vaccinee RJr-11 showed the highest binding signals and the

17 broadest cross-reactivity (Fig. 4A). Furthermore, Abs from RAt-9 and RQe-10, which had been enrolled in different studies but had been immunized with the same Tat protein (Table 1), also bound most of the PL-Tat mimes. These data indicate that our selection strategy (biopanning #II) indeed enriched for recombinant phages recognized by Abs specific for protected vaccinees. For RRi-11 and RTr-11, which were aviremic throughout, we did not detect any binding at week 0. However, at week 7, after multiple low-dose exposures to the challenge virus, most of the PL-Tat mimes were recognized. Furthermore, the protected vaccinee RAt-9 showed increased binding signals at week 30, after live-virus challenge. These results are compatible with vaccine-induced priming of anti-tat NT Ab responses that were boosted by the exposure to live virus. In contrast, no boosting was observed for the other three protected animals (RGe- 11, RFo-11 or RQe-10, Fig 4A). Anti-Tat NT Abs in RM RAt-9 with long-term protection Next, we sought to analyze anti-tat Ab responses in monkey RAt-9 given the extensive cross-reactivity to all PL-Tat mimes with plasma from this RM (Fig. 4A). RAt-9 had received three protein immunizations followed by two sequential mucosal challenges with related but distinct R5 SHIV-Cs (16, 28, 29). Although this vaccinee remained aviremic throughout all virus challenges, the anti-tat Ab activity had increased after the first virus challenge (Fig. 4A). Thus, we performed two additional biopannings (biopanning #III and #IV, Table 2) and termed the resulting mimotopes RAt-9 mimes

18 Biopanning #III was designed to examine the RAt-9 Ab repertoire after immunization but before any live-virus challenge. Thus, we used week -2 plasma (week 0 samples were not available) as positive selector and its autologous preimmune plasma as negative selector (conventional biopanning, Table 2). Some of the recombinant phages selected resembled the NT of HIV-1 Tat (Fig. 3A). To examine whether the first live-virus exposure had indeed altered the vaccine-induced Ab repertoire, we performed biopanning #IV with plasma from the day of the second challenge (week 30) versus plasma from week -2 (Table 2). By depleting phages recognizing Abs induced by immunization only, we should enrich for phages presenting epitopes recognized by Abs induced by live virus exposure. Again, some of the mimotopes selected reflected the Tat NT (Fig. 3A), supporting the observation that live-virus exposure had boosted the anti-tat Ab responses. We included the RAt-9 Tat mimes from biopannings #III and IV in the phage cross-elisa (Fig. 4B) and detected strong cross-reactivity with Abs of some of the other protected RMs at week 0 (RGe-11, RJr-11) or week 7 (RRi-11, RTr-11). This suggests that the HIV-1 Tat NT is indeed specifically targeted by Abs present in protected vaccinees. From the three biopannings (#II-IV), we isolated 13 different mimotopes representing the HIV-1 Tat NT; 12 of them showing a central motif (blue box, Fig. 3A). The sequence logo of this motif indicating conserved amino acids among these Tat mimotopes and the location of the motif on the Tat protein structure (PDB-ID: 1JFW; HIV-1 Tat) are shown in Fig. 3B and C, respectively

19 The HIV-1 Tat mimotopes represent a nab epitope To validate Tat specificity of our mimotopes, we used two mouse mabs, which are known to recognize the NT of Tat (NT3 2D1.1 and 2A4.1). MAb NT3 2D1.1 recognized the N-terminal peptides #5113 and #5114, whereas 2A4.1 and a control Ab (human IgG1 mab Herceptin) did not (Fig. 5A). Both anti-tat mabs were then tested in an HLM1 cell-based viral rescue assay (27, 32, 33). The NT- Tat-specific mab NT3 2D1.1 almost completely neutralized Tat transactivation at 40 µg/ml. In contrast, mab 2A4.1 was not able to block Tat-mediated transactivation (Fig. 5B). These two experiments show that NT3 2D1.1 is a neutralizing mab (nmab) targeting the Tat NT. To test if our Tat mimotopes indeed represented the neutralizing Tat NT, we examined the anti-tat mabs for their ability to capture our phage mimotopes. Five out of 13 Tat mimotopes were recognized by nmab NT3 2D1.1. Neither mab 2A4.1 nor Herceptin bound to any of the Tat mimotopes (Fig. 5C). These data confirm the specificity of the mimotopes isolated: they represented the HIV-1 Tat NT. RM IgG with anti-tat neutralizing activity To confirm the presence of neutralizing anti-tat Abs in the plasma of the protected RMs, we sought to purify IgG from the latter and test its neutralization potential in the HLM1 cell-based viral rescue assay. The phage cross-elisas (Fig. 4A, B) indicated that monkey RJr-11 had the highest anti-tat NT binding activity at week 0. Thus, we used samples from this animal for subsequent

20 binding and functional assays. Additionally, we also included samples from the partially protected monkey RBr-11. We first confirmed the ability of the purified polyclonal IgG from vaccinees RJr-11 and RBr-11 to recognize the Tat NT in a peptide ELISA (Fig. 6A, pool of #5113, #5114 and #5115, sequence see insert Fig. 5A); purified IgG from two naïve RMs served as negative control. Next, we tested the polyclonal IgG at different concentrations ( µg/ml) in the HLM1 rescue assay (Fig. 6B). Polyclonal IgG from both vaccinees, RJr-11 and RBr-11, neutralized almost 100% of Tat-induced transactivation at 600 µg/ml. This was statistically significant compared to the transactivation detected with naïve IgG (P=0.002; asterisks illustrate P<0.025, considering Bonferroni correction). At this concentration, the naïve IgGs showed some non-specific inhibition, which disappeared at subsequent dilutions. Importantly, neutralization of Tat transactivation followed a dose-dependent curve for both animals (Fig. 6B). We conclude that both RMs had developed anti-tat IgG able to neutralize Tat transactivation in vitro. Epitope mapping of anti-tat Abs in protected RMs We examined whether the HIV-1 Tat immunogen had induced Ab responses targeting domains other than the NT. Since our immunization regimen contained a clade B Tat (HIV-1 IIIB), we used the homologous consensus clade B Tat overlapping peptide library for epitope mapping. We tested plasma with Tat neutralizing activity from RJr-11 and RBr-11 at a 1:1,

21 dilution for each peptide. As a comparison, we also included plasma from the non-protected RM RDo-11. A pool of pre-immune plasma from all three vaccinees and a scrambled gp120 C-terminal peptide served as negative controls (Fig. 7). Both partially protected animals showed the highest binding against peptides #5113 and #5114 (NT region), whereas plasma Abs from the non-protected RM RDo-11 did not recognize these peptides. This result is in accordance with our phage-cross ELISAs, which showed that the NT of Tat is only strongly recognized by Abs from protected vaccinees. Additionally, we also detected some minor binding to the basic region, the glutamine-rich region as well as to the C-terminus of Tat (Fig. 7). Abs to the Tat NT: Reduced risk of SHIV-C acquisition and virus-induced boosting in aviremic vaccinees To examine the importance of anti-tat NT Abs, we sought to correlate their binding activity with protection observed in vivo. We ranked all vaccinees from our previous study (15) according to increasing peak viral RNA (vrna) loads (Fig. 8A, B). RMs that remained aviremic during the low-dose challenge phase (monkeys RRi-11, RTr-11, RGe-11 and RFo-11, red dots) were assigned vrna loads of 49 copies/ml (assay sensitivity, 50 copies/ml (34)). Only these four completely protected as well as the two partially protected animals, RJr-11 and RBr-11, had developed anti-tat NT binding Abs through vaccination. This resulted in a statistically significant inverse correlation (r= 0.64, P=0.03) between the anti-tat NT Ab titers and peak viremia (Fig. 8A). In contrast, 10 out of 12 vaccinees developed detectable Abs

22 against the full-length HIV-1 Tat protein. However, there was no statistically significant correlation between such Abs and peak viremia (r= 0.53, P=0.07, Fig. 8B). These data show that although most vaccinees mounted Ab responses to the Tat immunogen, only animals with low or undetectable viral loads developed Abs targeting the neutralizing epitope at the Tat NT. In order to examine possible consequences of the live-virus exposures on the vaccine-induced Tat responses, we also measured Ab responses at week 7 (after five low-dose challenges with SHIV-1157ipEL-p). Consistent with the data from our phage ELISAs (Fig. 4A, B), aviremic vaccinees RRi-11, RTr-11 showed increased Ab titers against the Tat NT as well as the full-length Tat protein (Fig. 8A, B). Of note, the same boosting was observed for monkey RAt-9, which has remained aviremic for several years after the virus challenges (data not shown). In contrast, only one out of 11 SHIV-C-exposed controls had developed detectable but low anti-nt Tat titers (RLs-11, Fig. 8C). These data indicate that exposure to the live virus alone did not induce sufficient anti-tat NT binding titers. However, once the animals were primed with the Tat immunogen, the virus was able to boost pre-existing vaccine-induced Abs that were low in some but not all RMs. Discussion In this study, we a) developed protection-linked (PL) biopanning as universal tool to probe Ab paratopes associated with protection; b) showed that a multi-component protein vaccine containing HIV-1 Tat induced Abs targeting the

23 neutralizing NT of Tat; and c) demonstrated that vaccine-specific Abs recognizing the HIV-1 Tat NT are inversely correlated with peak viremia. Taken together, our data implicate that nabs against the Tat NT may have a protective role in vivo, in combination with other immune responses. The interpretation of protective immune correlates against HIV-1 acquisition in trial participants as well as non-human primates is complicated because of several factors such as host genetics (35, 36), differences in innate and adaptive immunity (37, 38), and a variety of possible protein targets for antiviral immune responses (39). Here, we designed a selection strategy that only differentiates between vaccine-induced Abs with a link to protection versus Abs without such a link. We used three random phage-displayed peptide libraries, which are expressing linear or cysteine-constrained peptides (between 7-12 amino acids in length) fused to the piii N-terminus of M13 phage and examined the polyclonal Ab responses of protected vaccinees. Importantly, this strategy has no bias towards either the Ab targets or mechanisms of Abmediated vaccine protection involved. However, there are also some limitations to consider: a) the mimotopes selected represent the internal contour of Ab paratopes. Thus, this approach can give only information on structural mimicry of Ab epitopes, but does not reflect its chemical composition (such as glycan components); b) due to the short peptide size of the phage libraries used, the mimotopes selected might represent only portions of the Ab epitopes. However, our selection strategy is not restricted in terms of the peptide library being used. Thus, we are considering alternative approaches, such as libraries expressing

24 longer peptides or displaying the peptides via the phage s major coat protein (Protein 8), or even a combination of the latter with next-generation sequencing (40). We first performed a spiking experiment and added the known neutralizing and conformation-dependent mab 33C6 (18) to the plasma of a vaccineprotected RM. By this spiked PL-biopanning approach, we successfully isolated nmab 33C6-specific mimotopes. Importantly, some of them were also crossrecognized by a second anti-v3 nmab HGN194, which was recently shown to be completely protective in vivo (20). With this experiment, we indirectly verified that our PL-biopanning approach can indeed identify mimotopes that are linked to Abmediated protection (in this case Env-dependent neutralization). As expected, when dissecting the polyclonal plasma samples, each PLbiopanning identified a variety of different Env-specific epitopes, which are the topic of another detailed analysis (data not shown). However, based on linear alignments, one frequently recurring motif (isolated using plasma from vaccinee RGe-11) could be assigned to the NT of Tat. When we investigated the binding profile of the PL-Tat mimotopes further, we observed broad cross-recognition among protected vaccinees, including our long-term protected monkey, RAt-9. In contrast, non-protected RMs had only low or undetectable binding. Notably, some of the PL-Tat mimotopes were also recognized by an anti-nt nmab, thus confirming our PL-Tat mimotopes as neutralizing anti-tat Ab epitopes. HIV-1 Tat plays an essential role in the viral life cycle (41-43), since it is required for the production of full-length transcripts from the HIV-1 promoter (

25 ). In addition to its major intracellular role as a regulator of transcription, Tat has distinct extracellular functions in viral pathogenesis and replication (45, 47). Therefore, HIV-1 Tat has been targeted by candidate AIDS vaccines (41, 47, 48). Tat consists of amino acids and can be divided into six different functional domains: 1) the NT proline-rich region (residues 1-21), 2) the cysteine-rich region (residues 22-37), 3) the core region (residues 38-48), 4) the basic region (residues 49-59), 5) the glutamine-rich region (residues 60-72), and 6) the C- terminus of Tat (47). In 1988, Krone et al. identified the NT as the major binding site of natural anti-hiv-1 Tat Abs (49). Later, it was shown that Ab responses targeting this conserved epitope (50) are able to neutralize Tat-dependent transactivation in vitro (51-53). Overall, Tat-specific immune responses in humans were linked to lower viral loads and delayed disease progression (54-57). To further investigate the protective potential of Tat-specific immune responses, biologically active Tat protein or the chemically inactivated Tat toxoid were used as immunogens in macaques leading to different outcomes (32, 53, 58-63). In 1999, Cafaro et al. showed that biologically active HIV-1 Tat IIIB elicited both humoral and cellular immune responses in cynomolgus macaques (CMs), and no viral RNA was ever detectable after intravenous (i.v.) challenge with SHIV89.6P in five out of seven vaccinees (32). Pauza et al. immunized RMs with chemically inactivated Tat toxoid and showed lowered viral RNA loads after intrarectal (i.r.) challenge with SHIV89.6PD, but no complete prevention from virus acquisition (59). RMs immunized with either 2 exon full-length 102 amino

26 acid Tat (89.6P Tat) or a truncated 86 amino acid Tat (IIIB Tat) developed high anti-tat serum IgG and robust cellular immune responses. Nevertheless, no prevention of acquisition or lowered viral loads was observed following challenge with SHIV89.6P (58, 60). The variable outcomes of these studies may be explained by differences in macaque species, challenge routes and/or virus strains. In our previous vaccine/challenge studies, HIV-1 Tat IIIB was part of a multi-component protein vaccine designed to protect against mucosal challenges with biologically relevant, CCR5-tropic SHIV-Cs (15-17). We chose this immunization approach to improve vaccine efficacy due to an increased number of vaccine targets. Thus, instead of inducing immune responses against only one viral target, such as the Tat only approaches discussed, we sought to induce broader antiviral activity by focusing on viral targets that are active at different steps in the life-cycle to generate synergistic or additive immunity. Indeed, Lakhashe et al. described both cellular and neutralizing humoral immune responses as vaccine correlates of protection, suggesting a balanced set of immune responses due to the multi-component vaccine (15). In terms of Tatspecific immunity, strong cellular responses had been observed in five vaccinees that had been either completely or partially protected (monkeys RRi-11, RTr-11, RGe-11, RJr-11, and RBr-11). In the present study, we also discovered Tatspecific humoral immune responses in the same animals. This confirms that our multi-component protein vaccine successfully induced balanced anti-hiv-1 Tat immunity, which might have supported complete or partial protection against

27 SHIV-C challenges. We observed a significant inverse correlation of Abs targeting the neutralizing NT of Tat with peak viremia. However, this does not prove a causative link to protection in vivo. To establish the latter, a passive immunization study using Abs targeting this protection-linked epitope of Tat would be necessary. Yet, our new method of probing the protection-linked Ab paratopes is an important first step towards defining Ab-mediated vaccineprotection. Our biopannings selected recombinant phages representing the HIV-1 Tat NT. Epitope mapping in partially protected vaccinees confirmed that the Tatspecific Ab response indeed targeted mainly this area of the Tat molecule. However, consistent with the results of other groups (58, 60, 62), we also detected minor responses targeting the basic region, the glutamine-rich region and the C-terminus. The fact that we did not select mimotopes reflecting any of the other Tat areas is likely due to probability; i.e., the presence of significantly higher Ab titers targeting an immunodominant region; in this case the neutralizing Tat NT. During our Ab epitope analysis, we observed increased anti-tat Ab titers after live-virus exposure(s) in vaccinees RRi-11, RTr-11 and RAt-9, although no plasma viremia was ever detected for these animals. For RTr-11 and RAt-9, this may be explained by cryptic infection of target cells (15, 29) and their subsequent lysis producing adequate amounts of Tat that may have induced a boosting of pre-existing, low-level Ab titers. However, for monkey RRi-11, we did not find any evidence for cryptic infection when we investigated antiviral cellular

28 immunity, consistent with the definition of sterilizing immunity (15). For this animal, we propose that Tat might have been present in the challenge stock, probably bound to the virion surface and thus transported through the mucosa. The basic region of HIV-1 Tat is highly charged, and the protein is known to undergo electrostatic interactions with viral or cell-specific proteins (64-68). Since the virus challenges alone, without preceding immunization, could not induce detectable anti-tat responses, we suggest that the Tat immunogen had primed Ab responses in RRi-11, undetectable with our phage ELISAs, which were then boosted by Tat present in the virus stock. Recently, Monini et al. (69) described potential mechanisms, which parallel our own observations described above. First, Monini et al. (69) showed that extracellular HIV-1 Tat can form a molecular complex with trimeric Env, which provides a possible explanation for Tat being present in the virus stock and boosting pre-existing anti-tat Ab responses even in the absence of systemic infection. Moreover, Monini et al. (69) observed that such Env-bound Tat can favor virus infection of certain immune cells, whereas the presence of anti-tat Abs can restore and increase HIV neutralization, which would support the idea of anti-tat Ab-mediated protection. In summary, we described PL-biopanning as a novel, universal tool that can be used to dissect Ab responses linked to protection. As such, we identified anti-tat NT nabs as vaccine-specific responses that significantly correlated with complete or partial protection. Together, our data argue for including HIV-1 Tat in a multi-component vaccine that seeks to induce humoral as well as cellular immunity

29 Acknowledgements This work was supported by National Institutes of Health grants P01 AI048240, R37 AI034266, and R01 AI to R.M.R., and Harvard University C.F.A.R grant P30 AI The following reagents were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: (1) HIV consensus subtype B Tat (15-mer) peptides - complete set; (2) HIV-1 Tat protein; (3) mab to HIV-1 Tat (NT3 2D1.1) from Dr. Jonathan Karn, courtesy of the NIBSC Centralised Facility for AIDS Reagents; (4) HIV-1 Tat mab 2A4.1 from Dr. Jon Karn; (5) HLM30 (HLM1) cells from Dr. Reza Sadaie. The PepAligner program was downloaded from the Pacific Northwest National Laboratory (PNNL) homepage (Biological MS Data and Software Distribution Center: OMICS.PNL.GOV). COMPETING INTERESTS The Dana-Farber Cancer Institute is applying for patent protection of the PLmimotopes

30 References 1. McElrath MJ, Haynes BF Induction of immunity to human immunodeficiency virus type-1 by vaccination. Immunity 33: Rerks-Ngarm S, Pitisuttithum P, Nitayaphan S, Kaewkungwal J, Chiu J, Paris R, Premsri N, Namwat C, de Souza M, Adams E, Benenson M, Gurunathan S, Tartaglia J, McNeil JG, Francis DP, Stablein D, Birx DL, Chunsuttiwat S, Khamboonruang C, Thongcharoen P, Robb ML, Michael NL, Kunasol P, Kim JH Vaccination with ALVAC and AIDSVAX to prevent HIV-1 infection in Thailand. N. Engl. J. Med. 361: McMichael AJ, Haynes BF Lessons learned from HIV-1 vaccine trials: new priorities and directions. Nat. Immunol. 13: Haynes BF, Gilbert PB, McElrath MJ, Zolla-Pazner S, Tomaras GD, Alam SM, Evans DT, Montefiori DC, Karnasuta C, Sutthent R, Liao HX, DeVico AL, Lewis GK, Williams C, Pinter A, Fong Y, Janes H, DeCamp A, Huang Y, Rao M, Billings E, Karasavvas N, Robb ML, Ngauy V, de Souza MS, Paris R, Ferrari G, Bailer RT, Soderberg KA, Andrews C, Berman PW, Frahm N, De Rosa SC, Alpert MD, Yates NL, Shen X, Koup RA, Pitisuttithum P, Kaewkungwal J, Nitayaphan S, Rerks-Ngarm S, Michael NL, Kim JH Immune-correlates analysis of an HIV-1 vaccine efficacy trial. N. Engl. J. Med. 366: Lakhashe SK, Silvestri G, Ruprecht RM No acquisition: a new ambition for HIV vaccine development? Curr Opin Virol 1:

31 Lifson JD, Haigwood NL Lessons in nonhuman primate models for AIDS vaccine research: from minefields to milestones. Cold Spring Harb Perspect Med 2:a Zhao J, Lai L, Amara RR, Montefiori DC, Villinger F, Chennareddi L, Wyatt LS, Moss B, Robinson HL Preclinical studies of human immunodeficiency virus/aids vaccines: inverse correlation between avidity of anti-env antibodies and peak postchallenge viremia. J. Virol. 83: Xiao P, Zhao J, Patterson LJ, Brocca-Cofano E, Venzon D, Kozlowski PA, Hidajat R, Demberg T, Robert-Guroff M Multiple vaccineelicited nonneutralizing antienvelope antibody activities contribute to protective efficacy by reducing both acute and chronic viremia following simian/human immunodeficiency virus SHIV89.6P challenge in rhesus macaques. J. Virol. 84: Burton DR, Hessell AJ, Keele BF, Klasse PJ, Ketas TA, Moldt B, Dunlop DC, Poignard P, Doyle LA, Cavacini L, Veazey RS, Moore JP Limited or no protection by weakly or nonneutralizing antibodies against vaginal SHIV challenge of macaques compared with a strongly neutralizing antibody. Proc. Natl. Acad. Sci. U.S.A. 108: Humbert M, Rasmussen RA, Ong H, Kaiser FM, Hu SL, Ruprecht RM Inducing cross-clade neutralizing antibodies against HIV-1 by immunofocusing. PLoS ONE 3:e

32 Humbert M, Antoni S, Brill B, Landersz M, Rodes B, Soriano V, Wintergerst U, Knechten H, Staszewski S, von Laer D, Dittmar MT, Dietrich U Mimotopes selected with antibodies from HIV-1- neutralizing long-term non-progressor plasma. Eur. J. Immunol. 37: Scala G, Chen X, Liu W, Telles JN, Cohen OJ, Vaccarezza M, Igarashi T, Fauci AS Selection of HIV-specific immunogenic epitopes by screening random peptide libraries with HIV-1-positive sera. J. Immunol. 162: Dieltjens T, Willems B, Coppens S, Van Nieuwenhove L, Humbert M, Dietrich U, Heyndrickx L, Vanham G, Janssens W Unravelling the antigenic landscape of the HIV-1 subtype A envelope of an individual with broad cross-neutralizing antibodies using phage display peptide libraries. J. Virol. Methods 169: Enshell-Seijffers D, Smelyanski L, Vardinon N, Yust I, Gershoni JM Dissection of the humoral immune response toward an immunodominant epitope of HIV: a model for the analysis of antibody diversity in HIV+ individuals. FASEB J. 15: Lakhashe SK, Wang W, Siddappa NB, Hemashettar G, Polacino P, Hu SL, Villinger F, Else JG, Novembre FJ, Yoon JK, Lee SJ, Montefiori DC, Ruprecht RM, Rasmussen RA Vaccination against heterologous R5 clade C SHIV: prevention of infection and correlates of protection. PLoS ONE 6:e

33 Rasmussen RA, Ong H, Song R, Chenine AL, Ayash-Rashkovsky M, Hu SL, Polacino P, Else JG, Novembre FJ, Ruprecht RM Efficacy of a multigenic protein vaccine containing multimeric HIV gp160 against heterologous SHIV clade C challenges. AIDS 21: Lakhashe SK, Velu V, Sciaranghella G, Siddappa NB, Dipasquale JM, Hemashettar G, Yoon JK, Rasmussen RA, Yang F, Lee SJ, Montefiori DC, Novembre FJ, Villinger F, Amara RR, Kahn M, Hu SL, Li S, Li Z, Frankel FR, Robert-Guroff M, Johnson WE, Lieberman J, Ruprecht RM Prime-boost vaccination with heterologous live vectors encoding SIV gag and multimeric HIV-1 gp160 protein: efficacy against repeated mucosal R5 clade C SHIV challenges. Vaccine 29: Sholukh AM, Mukhtar MM, Humbert M, Essono SS, Watkins JD, Vyas HK, Shanmuganathan V, Hemashettar G, Kahn M, Hu SL, Montefiori DC, Polonis VR, Schur PH, Ruprecht RM Isolation of monoclonal antibodies with predetermined conformational epitope specificity. PLoS ONE 7:e Lu X, Miller CJ Concentration of IgG in the sera of normal rhesus macaques as determined by a species-specific radial immunodiffusion assay. Journal of immunological methods 197: Watkins JD, Siddappa NB, Lakhashe SK, Humbert M, Sholukh A, Hemashettar G, Wong YL, Yoon JK, Wang W, Novembre FJ, Villinger F, Ibegbu C, Patel K, Corti D, Agatic G, Vanzetta F, Bianchi S, Heeney JL, Sallusto F, Lanzavecchia A, Ruprecht RM An anti-hiv-1 V3-33 -

34 loop antibody fully protects cross-clade and elicits T-cell immunity in macaques mucosally challenged with an R5 clade C SHIV. PLoS ONE 6:e Crooks GE, Hon G, Chandonia JM, Brenner SE WebLogo: a sequence logo generator. Genome Res. 14: Schneider TD, Stephens RM Sequence logos: a new way to display consensus sequences. Nucleic Acids Res. 18: Peloponese JM, Jr., Gregoire C, Opi S, Esquieu D, Sturgis J, Lebrun E, Meurs E, Collette Y, Olive D, Aubertin AM, Witvrow M, Pannecouque C, De Clercq E, Bailly C, Lebreton J, Loret EP H-13C nuclear magnetic resonance assignment and structural characterization of HIV-1 Tat protein. C. R. Acad. Sci. III, Sci. Vie 323: Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25: Corti D, Langedijk JP, Hinz A, Seaman MS, Vanzetta F, Fernandez- Rodriguez BM, Silacci C, Pinna D, Jarrossay D, Balla-Jhagjhoorsingh S, Willems B, Zekveld MJ, Dreja H, O'Sullivan E, Pade C, Orkin C, Jeffs SA, Montefiori DC, Davis D, Weissenhorn W, McKnight A, Heeney JL, Sallusto F, Sattentau QJ, Weiss RA, Lanzavecchia A Analysis of memory B cell responses and isolation of novel

35 monoclonal antibodies with neutralizing breadth from HIV-1-infected individuals. PLoS ONE 5:e Sadaie MR, Tschachler E, Valerie K, Rosenberg M, Felber BK, Pavlakis GN, Klotman ME, Wong-Staal F Activation of tatdefective human immunodeficiency virus by ultraviolet light. New Biol. 2: Kashi VP, Jacob RA, Paul S, Nayak K, Satish B, Swaminathan S, Satish KS, Ranga U HIV-1 Tat-specific IgG antibodies in highresponders target a B-cell epitope in the cysteine-rich domain and block extracellular Tat efficiently. Vaccine 27: Rasmussen RA, Lakhashe SK, Ruprecht RM Bimodal AIDS vaccine approach: induction of cellular as well as humoral immunity can protect from systemic infection. Vaccine 28 Suppl 2:B Rasmussen RA, Siddappa NB, Lakhashe SK, Watkins J, Villinger F, Ibegbu C, Florese RH, Robert-Guroff M, Montefiori DC, Forthal DN, O'Connor D, Ruprecht RM High-level, lasting antiviral immunity induced by a bimodal AIDS vaccine and boosted by live-virus exposure: prevention of viremia. AIDS 26: Siddappa NB, Hemashettar G, Wong YL, Lakhashe S, Rasmussen RA, Watkins JD, Novembre FJ, Villinger F, Else JG, Montefiori DC, Ruprecht RM Development of a tier 1 R5 clade C simian-human immunodeficiency virus as a tool to test neutralizing antibody-based immunoprophylaxis. J. Med. Primatol. 40:

36 Song RJ, Chenine AL, Rasmussen RA, Ruprecht CR, Mirshahidi S, Grisson RD, Xu W, Whitney JB, Goins LM, Ong H, Li PL, Shai-Kobiler E, Wang T, McCann CM, Zhang H, Wood C, Kankasa C, Secor WE, McClure HM, Strobert E, Else JG, Ruprecht RM Molecularly cloned SHIV-1157ipd3N4: a highly replication- competent, mucosally transmissible R5 simian-human immunodeficiency virus encoding HIV clade C Env. J. Virol. 80: Cafaro A, Caputo A, Fracasso C, Maggiorella MT, Goletti D, Baroncelli S, Pace M, Sernicola L, Koanga-Mogtomo ML, Betti M, Borsetti A, Belli R, Akerblom L, Corrias F, Butto S, Heeney J, Verani P, Titti F, Ensoli B Control of SHIV-89.6P-infection of cynomolgus monkeys by HIV-1 Tat protein vaccine. Nat. Med. 5: Ensoli B, Buonaguro L, Barillari G, Fiorelli V, Gendelman R, Morgan RA, Wingfield P, Gallo RC Release, uptake, and effects of extracellular human immunodeficiency virus type 1 Tat protein on cell growth and viral transactivation. J. Virol. 67: Hofmann-Lehmann R, Swenerton RK, Liska V, Leutenegger CM, Lutz H, McClure HM, Ruprecht RM Sensitive and robust one-tube realtime reverse transcriptase-polymerase chain reaction to quantify SIV RNA load: comparison of one- versus two-enzyme systems. AIDS Res Hum Retroviruses 16:

37 Sobieszczyk ME, Lingappa JR, McElrath MJ Host genetic polymorphisms associated with innate immune factors and HIV-1. Curr Opin HIV AIDS 6: An P, Winkler CA Host genes associated with HIV/AIDS: advances in gene discovery. Trends Genet. 26: McMichael AJ, Borrow P, Tomaras GD, Goonetilleke N, Haynes BF The immune response during acute HIV-1 infection: clues for vaccine development. Nat. Rev. Immunol. 10: Ackerman ME, Dugast AS, Alter G Emerging concepts on the role of innate immunity in the prevention and control of HIV infection. Annu. Rev. Med. 63: Li L, Li HS, Pauza CD, Bukrinsky M, Zhao RY Roles of HIV-1 auxiliary proteins in viral pathogenesis and host-pathogen interactions. Cell Res. 15: Ryvkin A, Ashkenazy H, Smelyanski L, Kaplan G, Penn O, Weiss- Ottolenghi Y, Privman E, Ngam PB, Woodward JE, May GD, Bell C, Pupko T, Gershoni JM Deep Panning: steps towards probing the IgOme. PLoS ONE 7:e Gallo RC Tat as one key to HIV-induced immune pathogenesis and Tat (correction of Pat) toxoid as an important component of a vaccine. Proc. Natl. Acad. Sci. U.S.A. 96:

38 Haseltine W, Sodroski J, Rosen C Replication and pathogenesis of the human T-cell leukemia/lymphotropic retroviruses. Haematol Blood Transfus 31: Fisher AG, Feinberg MB, Josephs SF, Harper ME, Marselle LM, Reyes G, Gonda MA, Aldovini A, Debouk C, Gallo RC, et al The transactivator gene of HTLV-III is essential for virus replication. Nature 320: Berkhout B, Gatignol A, Rabson AB, Jeang KT TARindependent activation of the HIV-1 LTR: evidence that tat requires specific regions of the promoter. Cell 62: Jeang KT, Xiao H, Rich EA Multifaceted activities of the HIV-1 transactivator of transcription, Tat. J. Biol. Chem. 274: Cullen BR Regulation of HIV-1 gene expression. FASEB J. 5: Campbell GR, Loret EP What does the structure-function relationship of the HIV-1 Tat protein teach us about developing an AIDS vaccine? Retrovirology 6: Caputo A, Gavioli R, Bellino S, Longo O, Tripiciano A, Francavilla V, Sgadari C, Paniccia G, Titti F, Cafaro A, Ferrantelli F, Monini P, Ensoli F, Ensoli B HIV-1 Tat-based vaccines: an overview and perspectives in the field of HIV/AIDS vaccine development. Int Rev Immunol 28:

39 Krone WJ, Debouck C, Epstein LG, Heutink P, Meloen R, Goudsmit J Natural antibodies to HIV-tat epitopes and expression of HIV-1 genes in vivo. J. Med. Virol. 26: Goldstein G, Tribbick G, Manson K Two B cell epitopes of HIV-1 Tat protein have limited antigenic polymorphism in geographically diverse HIV-1 strains. Vaccine 19: Moreau E, Hoebeke J, Zagury D, Muller S, Desgranges C Generation and characterization of neutralizing human monoclonal antibodies against human immunodeficiency virus type 1 Tat antigen. J. Virol. 78: Ruckwardt TJ, Tikhonov I, Berg S, Hatfield GS, Chandra A, Chandra P, Gilliam B, Redfield RR, Gallo RC, Pauza CD Sequence variation within the dominant amino terminus epitope affects antibody binding and neutralization of human immunodeficiency virus type 1 Tat protein. J. Virol. 78: Moreau E, Belliard G, Partidos CD, Pradezinsky F, Le Buanec H, Muller S, Desgranges C Important B-cell epitopes for neutralization of human immunodeficiency virus type 1 Tat in serum samples of humans and different animal species immunized with Tat protein or peptides. J. Gen. Virol. 85: Re MC, Vignoli M, Furlini G, Gibellini D, Colangeli V, Vitone F, La Placa M Antibodies against full-length Tat protein and some low

40 molecular-weight Tat-peptides correlate with low or undetectable viral load in HIV-1 seropositive patients. J Clin Virol 21: Zagury JF, Sill A, Blattner W, Lachgar A, Le Buanec H, Richardson M, Rappaport J, Hendel H, Bizzini B, Gringeri A, Carcagno M, Criscuolo M, Burny A, Gallo RC, Zagury D Antibodies to the HIV-1 Tat protein correlated with nonprogression to AIDS: a rationale for the use of Tat toxoid as an HIV-1 vaccine. J. Hum. Virol. 1: van Baalen CA, Pontesilli O, Huisman RC, Geretti AM, Klein MR, de Wolf F, Miedema F, Gruters RA, Osterhaus AD Human immunodeficiency virus type 1 Rev- and Tat-specific cytotoxic T lymphocyte frequencies inversely correlate with rapid progression to AIDS. J. Gen. Virol. 78 (Pt 8): Rezza G, Fiorelli V, Dorrucci M, Ciccozzi M, Tripiciano A, Scoglio A, Collacchi B, Ruiz-Alvarez M, Giannetto C, Caputo A, Tomasoni L, Castelli F, Sciandra M, Sinicco A, Ensoli F, Butto S, Ensoli B The presence of anti-tat antibodies is predictive of long-term nonprogression to AIDS or severe immunodeficiency: findings in a cohort of HIV-1 seroconverters. J. Infect. Dis. 191: Richardson MW, Mirchandani J, Silvera P, Regulier EG, Capini C, Bojczuk PM, Hu J, Gracely EJ, Boyer JD, Khalili K, Zagury JF, Lewis MG, Rappaport J Immunogenicity of HIV-1 IIIB and SHIV 89.6P Tat and Tat toxoids in rhesus macaques: induction of humoral and cellular immune responses. DNA Cell Biol. 21:

41 Pauza CD, Trivedi P, Wallace M, Ruckwardt TJ, Le Buanec H, Lu W, Bizzini B, Burny A, Zagury D, Gallo RC Vaccination with tat toxoid attenuates disease in simian/hiv-challenged macaques. Proc. Natl. Acad. Sci. U.S.A. 97: Silvera P, Richardson MW, Greenhouse J, Yalley-Ogunro J, Shaw N, Mirchandani J, Khalili K, Zagury JF, Lewis MG, Rappaport J Outcome of simian-human immunodeficiency virus strain 89.6p challenge following vaccination of rhesus macaques with human immunodeficiency virus Tat protein. J. Virol. 76: Watkins JD, Lancelot S, Campbell GR, Esquieu D, de Mareuil J, Opi S, Annappa S, Salles JP, Loret EP Reservoir cells no longer detectable after a heterologous SHIV challenge with the synthetic HIV-1 Tat Oyi vaccine. Retrovirology 3: Tikhonov I, Ruckwardt TJ, Hatfield GS, Pauza CD Tatneutralizing antibodies in vaccinated macaques. J. Virol. 77: Maggiorella MT, Baroncelli S, Michelini Z, Fanales-Belasio E, Moretti S, Sernicola L, Cara A, Negri DR, Butto S, Fiorelli V, Tripiciano A, Scoglio A, Caputo A, Borsetti A, Ridolfi B, Bona R, ten Haaft P, Macchia I, Leone P, Pavone-Cossut MR, Nappi F, Ciccozzi M, Heeney J, Titti F, Cafaro A, Ensoli B Long-term protection against SHIV89.6P replication in HIV-1 Tat vaccinated cynomolgus monkeys. Vaccine 22:

42 Nappi F, Chiozzini C, Bordignon V, Borsetti A, Bellino S, Cippitelli M, Barillari G, Caputo A, Tyagi M, Giacca M, Ensoli B Immobilized HIV-1 Tat protein promotes gene transfer via a transactivationindependent mechanism which requires binding of Tat to viral particles. J Gene Med 11: Marchio S, Alfano M, Primo L, Gramaglia D, Butini L, Gennero L, De Vivo E, Arap W, Giacca M, Pasqualini R, Bussolino F Cell surface-associated Tat modulates HIV-1 infection and spreading through a specific interaction with gp120 viral envelope protein. Blood 105: Chang HC, Samaniego F, Nair BC, Buonaguro L, Ensoli B HIV- 1 Tat protein exits from cells via a leaderless secretory pathway and binds to extracellular matrix-associated heparan sulfate proteoglycans through its basic region. AIDS 11: Barillari G, Gendelman R, Gallo RC, Ensoli B The Tat protein of human immunodeficiency virus type 1, a growth factor for AIDS Kaposi sarcoma and cytokine-activated vascular cells, induces adhesion of the same cell types by using integrin receptors recognizing the RGD amino acid sequence. Proc. Natl. Acad. Sci. U.S.A. 90: Albini A, Ferrini S, Benelli R, Sforzini S, Giunciuglio D, Aluigi MG, Proudfoot AE, Alouani S, Wells TN, Mariani G, Rabin RL, Farber JM, Noonan DM HIV-1 Tat protein mimicry of chemokines. Proc. Natl. Acad. Sci. U.S.A. 95:

43 Monini P, Cafaro A, Srivastava IK, Moretti S, Sharma VA, Andreini C, Chiozzini C, Ferrantelli F, Cossut MR, Tripiciano A, Nappi F, Longo O, Bellino S, Picconi O, Fanales-Belasio E, Borsetti A, Toschi E, Schiavoni I, Bacigalupo I, Kan E, Sernicola L, Maggiorella MT, Montin K, Porcu M, Leone P, Collacchi B, Palladino C, Ridolfi B, Falchi M, Macchia I, Ulmer JB, Butto S, Sgadari C, Magnani M, Federico MP, Titti F, Banci L, Dallocchio F, Rappuoli R, Ensoli F, Barnett SW, Garaci E, Ensoli B HIV-1 Tat Promotes Integrin-Mediated HIV Transmission to Dendritic Cells by Binding Env Spikes and Competes Neutralization by Anti-HIV Antibodies. PLoS ONE 7:e

44 FIG 1 Protection-linked (PL) biopanning to identify Ab epitopes associated with protection. Plasma samples from three vaccine-protected rhesus macaques (RMs) (RRi-11, RTr-11 and RGe-11) (15) were screened for Ab responses specific for protection using a peptide-phage display-based approach (protection-linked (PL) biopanning). Each PL-biopanning round consisted of (1) positive selection, (2) negative selection and (3) amplification of phages selected. Paramagnetic beads (in brown) were used to capture anti-rm IgGs. Red, the Ab Fc portion; light gray, the Fab domain of anti-rm IgGs immobilized on beads via Fc. Positive selection used plasma from one of the three protected animals (week 0). Dark blue, Abs induced by immunization and corresponding phages bound to Abs linked to protection. Positively selected phages were counterselected with plasma from a non-protected, highly viremic vaccinee (RDo-11 (15), enrolled in same study as monkeys RRi-11, RTr-11 and RGe-11 that had been used as positive selectors, same time point). In yellow, non-protective Abs from this negative selector and corresponding bound phages. Purple or red phages, unspecific phages bound to anti-rm Ab or beads, respectively. Unspecific phages and phages recognized by Abs that were not linked to protection (in yellow) are depleted during the negative selection, whereas phages representing epitopes of Abs linked to protection are enriched during the positive selection (in blue). Phages eluted after the third round of positive selection (PLmimotopes) were sequenced and grouped according to their peptide motifs

45 FIG 2 Spiking plasma of a vaccinated RM with a known nmab: Can PLbiopanning identify the expected epitope? (A) In light gray, the sequence of the consensus clade C V3 crown. In dark gray, the sequence of the 33C6- specific mimotope used for Ab isolation (Tc.2). The linear alignment shows the sequences of eight unique 33C6-specific mimotopes (33C6 PL-mimes) isolated by spiked PL-biopanning (experiment #I, Table 2). Homologies between them and Tc.2 are shaded in dark gray and additional homologies to the consensus clade C V3 crown shaded in light gray. (B) The eight 33C6 PL-mimes from (A) were tested in a phage ELISA for binding with mab 33C6, a second anti-v3 Ab, HGN194 (20), the spiked plasma (RGe-11 + mab 33C6) and the two plasma samples used for PL-biopanning (positive selector, RGe-11 and negative selector, RDo-11). Positive control, the original 33C6-specific mimotope Tc.2. Negative control, wildtype phage (WT). One plasma Ab-specific PL-mimotope (RGe-11 PL-mime) isolated with experiment #I was included as well. Dashed line, OD signals 10x higher than signals detected with WT. Fig 3 Sequence alignment and mimotope location on HIV-1 Tat. (A) The sequence of the immunogen (HIV-IIIB) is illustrated in gray in the header. Homologies between mimotopes and the parental strain are shaded in gray. Alignment shows sequences for five protection-linked Tat mimotopes isolated from RGe-11 (PL-Tat mimes) and eight Tat mimotopes isolated using plasma from RAt-9 (RAt-9 Tat mimes). The three different biopannings are indicated (#II- IV, Table 2). (B) Sequence logo of 12 mimotopes (bold, panel A) representing

46 the NT of Tat using WebLogo3. Bits represent the relative frequency of amino acids. (C) Three-dimensional location of the NT epitope on Tat protein (PDB-ID: 1JFW). The five conserved amino acid residues are highlighted in blue (LEPWK). Figure was prepared with Chimera (24). FIG 4 Cross-reactivity profile of Tat mimotopes. (A, B) For each animal, a time-point before the first live-virus encounter (week 0 or week -2, respectively) and a time-point after SHIV-C exposures (week 7 or week 30, respectively) were tested; numbers below monkey names indicate weeks tested. Binding patterns are shown in form of a heat-map. Yellow to dark red squares, OD signals 10x higher than signals detected with the wildtype phage control. White squares, binding signals below this cut-off. Negative control (N), pre-immune plasma pool of vaccinees. The animals were ranked from left to right according to increasing peak vrna loads. (A) Binding profile for protection linked Tat mimotopes (PL-Tat mimes) isolated using the PL-biopanning approach. (B) Binding profile for Tat mimotopes isolated using plasma from RAt-9 (RAt-9 Tat mimes). FIG 5 Epitope mapping using anti-tat mabs. Two mouse anti-tat mabs (2A4-1 and NT3 2D1.1) were tested for binding to linear NT peptides, anti-tat neutralization activity as well as their potential to recognize the Tat mimotopes. (A) Binding ELISA results using the anti-tat mabs. The sequences of the three overlapping peptides containing the same conserved NT motif as the Tat mimotopes are shown (# ). Negative controls, human IgG1 mab

47 Herceptin and a scrambled C-terminal gp120 peptide (control; sequence shown in insert). (B) HLM1 cell-based viral rescue assay to determine the potential of anti-tat mabs to neutralize Tat transactivation. Both mabs were used at 40 µg/ml. The height of the bars illustrates the percentage of transactivation compared to the transactivation measured without any mabs added (cells + Tat, considered as 100% transactivation). (C) Both mabs were tested for their reactivity with the Tat mimotopes. Negative controls, mab Herceptin and helper phage without peptide insert (WT, wild type). For all three panels, the height of each bar represents the result from at least two independent assays. The error bars represent the standard error of the mean (SEM). FIG 6 Purified IgG from vaccinees neutralizes Tat transactivation in vitro. IgG was purified from RJr-11 and RBr-11 (plasma pool from weeks 2 and 3) and tested for binding to linear NT Tat peptides as well as for Tat neutralization. Control IgG purified from two naïve RMs was included in all assays (naïve IgGs). (A) Binding ELISA results using linear Tat peptides (pool of # ). (B) Purified IgG was tested in HLM1 cell-based viral rescue assay. The nmab NT3 2D1.1 was included as positive control (data not shown). The height of the bars represents the average results of two independent assays, testing each IgG fraction in triplicates. Error bars represent the SEM. The percentage of transactivation between naïve and immune IgG was compared using an unpaired, two-tailed Student s t-test (significant P-values after Bonferroni correction, P<0.025 are shown)

48 FIG 7 Epitope mapping using a consensus B Tat peptide library. Polyclonal plasma taken at week 1 (vaccinees RDo-11, RJr-11 and RBr-11) was tested for binding to peptides of an overlapping HIV-1 clade B Tat library (# ). Negative controls, a scrambled C-terminal gp120 peptide (control; sequence shown in insert Fig. 5A) and a pool of the pre-immune plasma of the same three vaccinees. A signal 20x higher than the signal detected with the control peptide was considered positive (dashed line). The height of the bars represents the average results of two independent assays, testing each peptide in duplicates. Error bars represent the SEM. For better orientation, the overlapping peptides were organized into six protein domains. FIG 8 Abs targeting HIV-1 Tat NT are linked to a reduced risk of SHIV-C acquisition. The anti-tat binding titers of polyclonal plasma Abs in 12 vaccinated and 10 control RMs (all derived from the same immunization study (15)) were determined. (A, B) Vaccinees were ranked in ascending order of peak viremia after the five low-dose challenges. Aviremic RMs (RRi-11, RTr-11, RGe- 11 and RFo-11) were assigned a vrna load of 49 copies/ml (34) (red dots, peak vrna loads). All four animals received the high-dose challenge, after which RGe-11 and RFo-11 became viremic. One of the control animals from this immunization study (RAk-11) remained aviremic during the low-dose challenges (15), but became viremic after the high-dose challenge (asterisk and red dot, right Y-axis). X mark, mean peak plasma viremia of unvaccinated controls. W, week of peak viremia. The binding titers were determined at week 0 (two weeks

49 after the last immunization, but before virus challenges, striped bars) and week 7 (after virus challenges, black bars). (A) Binding Ab titers against Tat NT peptides (pool of # ). Inverse correlation of binding titers with peak viremia was assessed using Spearman correlation analysis. The Spearman's rank correlation coefficient (r) and the P-value are shown for the vaccine-induced Ab titers (week 0, striped bars). (B) Binding Ab titers against full-length Tat protein. (C) Binding Ab titers against Tat NT peptides in 10 unvaccinated but SHIV-1157ipEL-p - challenged control animals (15). In red, aviremic control animal RAk-11. Downloaded from on December 16, 2018 by guest

50 Level of protection TABLE 1 Different levels of protection in 14 vaccinees Complete (N=6) Partial (N=3) None (N=5) Animal name Virological outcome Interpretation Immunogens in multicomponent vaccine RRi-11 Aviremic Sterilizing immunity HIV-1 gp160 (clade C) RTr-11 Aviremic Cryptic infection + SIV Gag-Pol particles + HIV-1 Tat (clade B) RGe-11 RFo-11 Aviremic during low-dose challenges Aviremic during low-dose challenges 2 low-level blips (<10 4 copies/ml) Chronic systemic infection after highdose challenge RAt-9 Aviremic Cryptic infection HIV-1 gp160 (clade C) + SIV Gag-Pol particles + HIV-1 Tat (clade B) RQe-10 Aviremic Sterilizing immunity Listeria monocytogenes expressing SIV gag + Ad5hr encoding SIV gag + HIV-1 gp160 (clade C) + HIV-1 Tat (clade B) RJr-11 RBr-11 RKm-11 RUt-11 RDo-11 RAr-11 RDk-11 RPn-11 Lower peak viremia No protection Chronic systemic infection HIV-1 gp160 (clade C) + SIV Gag-Pol particles + HIV-1 Tat (clade B) Challenge viruses Multiple low-doses and one high-dose with SHIV-1157ipEL-p (30) Single low-dose with SHIV-1157ip and single high-dose with SHIV-1157ipd3N4 (31) Multiple low-doses with SHIV-1157ipEL-p (30) Multiple low-doses and one high-dose with SHIV-1157ipEL-p (30) Enrolled in study (15) (16, 28, 29) (17) (15)

51 TABLE 2 Selection strategies and characteristics of vaccinated RMs Biopanning # a RGe-11 was always aviremic during multiple low-dose challenges, but showed 2 low-level blips after high-dose challenge (<10 4 copies/ml) Selection strategy I. Plasma from protected vaccinee + nmab 33C6 (18) vs. plasma from nonprotected vaccine (spiked PL-biopanning) II. III. IV. Protected vs. nonprotected vaccinees (PL-biopanning) Vaccinated vs. naive (conventional biopanning) Vaccinee after vs. before live-virus exposure Yield of mimotopes Mimotopes reflecting the nmab 33C6 epitope and protection-linked mimotopes (PLmimes) Protection-linked mimotopes (PL-mimes) Mimotopes isolated using RAt-9 (RAt-9 mimes) Positive selection Monkey/ mab name RGe-11 a + nmab 33C6 RRi-11 RTr-11 RGe-11 a RAt-9 RAt Monkey status Aviremic during lowdose challenges Always aviremic Always aviremic Aviremic during lowdose challenges Always aviremic Always aviremic Time (week) Negative selection Monkey name Monkey status 0 RDo-11 Not protected 0 RDo-11 Not protected 0 RDo-11 Not protected 0 RDo-11 Not protected -2 RAt-9 Always aviremic 30 RAt-9 Always aviremic Time (week) Naïve -2

52

53

54

55