Human Immunodeficiency Virus (HIV) Vaccine Trials:
http://www.100md.com
病菌学杂志 2006年第5期
Division of Viral Products
Core Facility, Center for Biologics Evaluation and Research (CBER), FDA, Bethesda, Maryland 20892
Office of AIDS Research, NIH, Bethesda, Maryland 20892
Vaccine Clinical Research Branch, Division of AIDS, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, Maryland 20892
Vaccine Research Center, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, Maryland 20892
U.S. Army Medical Component-Armed Forces Research Institute of Medical Sciences (USAMC-AFRIMS), APO AP 96546
U.S. Military HIV Research Program, Rockville, Maryland 20850
National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia 2010
VaxGen, Inc., Brisbane, California 94005
ABSTRACT
All current human immunodeficiency virus (HIV) vaccine candidates contain multiple viral components and elicit antibodies that react positively in licensed HIV diagnostic tests, which contain similar viral products. Thus, vaccine trial participants could be falsely diagnosed as infected with HIV. Additionally, uninfected, seropositive vaccinees may encounter long-term social and economic harms. Moreover, this also interferes with early detection of true HIV infections during preventive HIV vaccine trials. An HIV-seropositive test result among uninfected vaccine trial participants is a major public health concern for volunteers who want to participate in future HIV vaccine trials. Based on the increased number of HIV vaccines being tested globally, it is essential to differentiate vaccine- from virus-induced antibodies. Using a whole-HIV-genome phage display library, we identified conserved sequences in Env-gp41 and Gag-p6 which are recognized soon after infection, do not contain protective epitopes, and are not part of most current HIV vaccines. We established a new HIV serodetection assay based on these peptides. To date, this assay, termed HIV-SELECTEST, demonstrates >99% specificity and sensitivity. Importantly, in testing of plasma samples from multiple HIV vaccine trials, uninfected trial participants scored negative, while all intercurrent infections were detected within 1 to 3 months of HIV infection. The new HIV-SELECTEST is a simple but robust diagnostic tool for easy implementation in HIV vaccine trials and blood banks worldwide.
INTRODUCTION
Since 1987, more than 25,000 individuals have participated in clinical trials with preventive human immunodeficiency virus (HIV) vaccines. Most of the current HIV vaccine candidates are complex products containing multiple viral genes or proteins, many of which are included in licensed HIV serodetection kits, including recently licensed rapid tests. Consequently, sera from vaccine recipients often react in licensed HIV serodetection assays, generating patterns indistinguishable from those for HIV-infected individuals (1, 4, 11, 12). This will complicate future prophylactic vaccine trials, in which early detection of HIV infections is of paramount importance. Furthermore, long-term HIV seropositivity will exclude vaccine trial participants from the pool of blood and plasma donors and may contribute to a range of socioeconomic harms such as denied employment, health insurance, travel, immigration, and recruitment to the armed forces (2, 3). The prospect of seroconversion without a clear differential diagnosis algorithm could deter potential trial participants and severely curtail recruitment into large-scale trials around the globe (8, 9, 13).
Currently, there is no HIV serodiagnostic assay that differentiates between vaccine-generated antibodies and those produced after true HIV infection. Our goal was to develop an antibody-based HIV-1 detection assay in which vaccine-generated antibodies will score negative, whereas virus-induced antibodies can be detected soon after HIV infection. The selection criteria for peptides to be used in such an assay were as follows: (i) epitopes that are not included in HIV vaccines since they do not appear to contribute to protective immunity, (ii) epitopes recognized by antibodies made soon after HIV infection, and (iii) epitopes highly conserved among HIV clades and subtypes.
To identify such sequences, a gene fragment phage display library (GFPDL) was constructed from the entire HIV type 1 (HIV-1) genome and used to screen sera from HIV-infected individuals near the time of seroconversion. This strategy led to the discovery of three novel epitopes, one in Gag p6 and two in the envelope gp41 cytoplasmic tail. Herein, we describe the development of a new HIV-specific enzyme-linked immunosorbent assay (ELISA), termed HIV-SELECTEST, which distinguishes between HIV-infected individuals and uninfected vaccine recipients. The HIV-SELECTEST is a low-cost, high-throughput assay that could be implemented in clinical sites and blood collection centers worldwide and serve as an add-on diagnostic tool in future HIV vaccine trials.
MATERIALS AND METHODS
Construction of complete HIV genome-based gene fragment phage display library. Plasmid pNL4-3, containing the complete HIV-1 NL4-3 proviral DNA, was obtained from the NIH AIDS Research and Reference Reagent Program (McKesson BioServices Corp., Rockville, MD). The full-length HIV-1 genome was PCR amplified from pNL4-3 DNA with an Expand long-template polymerase preparation (Roche Diagnostics, Indianapolis, IN) and primers spanning the Lys tRNA primer binding site (MSF12 [5'-AAAAATCTCTAGCAGTGGCGCCCGAACAG-3']) and the poly(A) signal region of the 3' long terminal repeat (MSR5 [5'-AAGCACTCAAGGCAAGCTTTATTGAGGCT-3']), which amplifies the entire HIV-1 genome except for 75 bp in the unique 5' region of the long terminal repeat. The purified amplified DNA product was digested with DNase I using a DNase shotgun cleavage kit (Novagen, Madison, WI), and fragments between 50 and 300 bp were isolated by preparative gel electrophoresis, treated with T4 DNA polymerase to generate blunt ends, and dephosphorylated using calf intestinal alkaline phosphatase (Roche Diagnostics, Indianapolis, IN). DNAs were purified again using a nucleotide removal kit (QIAGEN Inc., Valencia, CA) and ligated in the presence of the SrfI enzyme into the SmaI site of the M13-derived phage vector for expression as gIIIp fusion proteins, followed by electroporation into Escherichia coli TG1 cells. Tet-resistant transformants were harvested and expanded in liquid culture (2x YT) at 37°C. The cell-free phage supernatant was isolated by centrifugation, and the phage titer was determined and expressed in Tetr transduction units. Ninety-six individual clones were isolated, and DNA inserts were amplified by standard PCR and sequenced to determine the insert size distribution and library diversity.
Selection of phages reactive with HIV antibodies from early-phase infected individuals. Seven plasma samples constituting the HIV-1 seroconversion panel PRB-910 from SeraCare BioServices (Gaithersburg, MD) were used for panning of the HIV-1 GFPDL. For the removal of plasma components, which could nonspecifically interact with phage proteins, a fivefold-diluted plasma was preadsorbed three times to sterile polystyrene petri dishes (35-mm diameter) coated with 1013 UV-killed VCSM13 phage. For biopanning, microtiter strips (Nunc Inc., Naperville, IL) were coated with a mixture of 500 ng each of goat anti-human immunoglobulin G (IgG) Fc- and goat anti-human IgM Fc-specific antibodies in phosphate-buffered saline (PBS), pH 7.4. After three washings with PBST (20 mM PBS containing 0.1% Tween 20), Dulbecco's modified Eagle's medium (DMEM) containing 5% fetal bovine serum (FBS) (blocking solution) was added to wells to block the unoccupied reactive sites. HIV-1-infected human plasma preadsorbed to VCSM13 was added to the wells and incubated for 1 h at room temperature (RT). Wells were washed thrice with PBST, and 1010 phage per well of the HIV-1 GFPDL, diluted in blocking solution, were added for 2 h at RT. Unbound phage were removed by 12 washes with PBST followed by 3 washes with PBS. Bound phage were eluted by the addition of 0.1 N HCl containing bovine serum albumin (1 mg/ml) for 10 min at RT and were neutralized by the addition of 8 μl of 2 M Tris solution per 100 μl eluate. Four rounds of affinity selection were carried out with each individual serum sample from HIV seroconversion panel PRB-910.
Analysis of affinity-selected phage clones. Twenty-two phage clones enriched after four rounds of biopanning on each PRB-910 plasma sample were further screened for specific recognition by HIV-seropositive sera and the absence of reactivity with seronegative sera in an affinity-capture phage-specific ELISA. The wells of ELISA plates (Immulon 2HB; Thermo Labsystems, Franklin, MA) were coated with 100 ng/well of anti-phage antibody (GE Healthcare, Piscataway, NJ) and blocked with DMEM-5% FBS. Subsequently, 1010 phage of the selected clones were added per well and incubated for 1 h at RT. Serially diluted sera (in DMEM-5% FBS) were added to the 96-well plates in duplicate and incubated at RT for 1 h. The bound antibodies were probed with horseradish peroxidase-conjugated goat anti-human IgG-IgM antibodies, and the reactions were developed with O-phenylenediamine substrate solution (Pierce Biotechnology, Rockford, IL). The clones demonstrating the best differential reactivities with HIV-1-seropositive sera were expanded, and the inserts were sequenced and mapped to individual HIV-1 genes. Several inserts were selected for synthetic peptide synthesis and development of the HIV-SELECTEST.
Peptides used in HIV-SELECTEST. Peptide sequences from Gag p6 (452-SRPEPTAPPAESFRFGEEITPTPSQKQEPKDKELYPPLASLRSLFGNDPSSQ-502) and the gp41 cytoplasmic region (SK1 [784-LIAARIVELLGHSSLKGLRRGWEALKYLWNLLQYWGQELKNSAISL-829] and SK2 [836-AVAEGTDRVIEVVQRVCRAILNIPRRIRQGFERALL-871]) were chemically synthesized (amino acid residues are numbered based on the CON-OF-CONS alignment sequence in the Los Alamos database). All peptides were synthesized at the Facility for Biotechnology Resources, CBER, FDA, on Applied Biosystems peptide synthesizer models 431 and 433 (Foster City, California) by standard 9-fluorenyl methoxycarbonyl chemistry. Peptides were purified by reverse-phase high-performance liquid chromatography and characterized by matrix-assisted laser desorption ionization-time of flight mass spectrometry.
HIV-SELECTEST. Based on preliminary screening of HIV-seronegative and -seropositive sera, the optimal conditions for the p6 and gp41 ELISAs were determined. The p6 peptide was used to coat Immulon-2HB plates at 30 ng/100 μl/well, while the gp41 peptides (SK1 and SK2) were used at 150 ng/100 μl/well (each; total, 300 ng/well). After three washes with PBST (20 mM PBS, 0.1% Tween 20), the unoccupied reactive sites were blocked by PBST containing 2% whole milk (2% WMPBST). All specimens (serum or plasma) were diluted 1:100 in 2% WMPBST, added to peptide-coated wells, and incubated for 1 h at RT. The plates were then washed six times with PBST, and 100 μl/well of horseradish peroxidase-conjugated goat anti-human IgG Fc-specific antibody (Jackson ImmunoResearch, West Grove, PA), diluted 1:10,000 in 2% WMPBST, was added. The reactions were quantified using O-phenylenediamine substrate.
Based on the results for 1,000 seronegative samples, cutoff (CO) values were determined individually for the p6 and gp41 peptides. The cutoff values used were the average absorbance values for negative sera (at a 1:100 dilution) plus 5 standard deviations (SD; for each peptide). Specimens with absorbance/cutoff ratios of 1 were considered HIV-1 seropositive, and those with ratios of <1 were considered HIV-1 seronegative.
HIV seroconversion panels and vaccine trial samples. HIV-1 seroconversion panels PRB-910, PRB-924, PRB-927, PRB-928, PRB-929, and PRB-931 and the mixed-titer panel PRB-204 were purchased from SeraCare BioServices (Gaithersburg, MD). A seroconversion panel consists of plasma samples collected serially early after HIV-1 infection, and the virological and immunological profiles for these plasma samples, as assessed by commercial diagnostic kits, were provided by SeraCare BioServices. Additionally, 28 seroconversion panels were provided by the University of New South Wales (PHAEDRA Inventory, Sydney, Australia). HIV-negative serum samples were obtained from the National Institutes of Health Blood Bank and the Vaccine Research Center (VRC, NIAID, NIH, Bethesda, MD).
Serum/plasma samples from the following HIV vaccine trials were tested: HVTN 203 (246 vaccinees and 78 patients receiving placebos; conducted by the HIV Vaccine Trial Network [HVTN]), RV124 (conducted by the Walter Reed Army Institute of Research), VRC 004 (40 vaccinees and 10 patients receiving placebos), VRC 006 (30 vaccinees and 6 patients receiving placebos), VRC 009 (9 vaccinees and no patients receiving placebos), and VRC 010 were conducted by the Vaccine Research Center (NIAID, NIH), and VAX 003 and VAX 004 were conducted by VaxGen Inc. The HIV infection status of a given sample was provided by the collaborating groups and also determined by in-house testing using a Bio-Rad HIV-1/2 Plus O kit (Bio-Rad Laboratories, Woodinwille, WA). Samples obtained from the VRC 009 and VRC 010 trials were also tested with Capillus HIV-1/HIV-2 (not licensed by the U.S. FDA) and Uni-Gold HIV (Trinity Biotech, NY) rapid tests along with the Bio-Rad HIV-1/2 Plus O kit.
All studies were conducted under approval from the Research Involving Human Subjects Committee (RIHSC exemption no. 04-050B) at the Center for Biologics Evaluation and Research.
RESULTS
Identification of HIV sequences recognized by early-phase seroconversion sera using GFPDL. To identify all the HIV sequences recognized by antibodies generated soon after HIV infection, we constructed a GFPDL spanning the entire HIV-1 open reading frame of NL4-3. The HIV-1 GFPDL contained >107 independent transformants. PCR-based analysis and sequencing of the inserts confirmed that the library consisted of 100% recombinants, with insert sizes of 50 to 300 bp and random distribution across the HIV genome.
Seven plasma samples constituting seroconversion panel PRB-910 (obtained from acutely HIV-1-infected individuals; SeraCare BioServices, Gaithersburg, MD) were used for affinity selection of phages displaying HIV-1 peptides. After four rounds of affinity selection, 22 clones (for each plasma sample) were selected for insert sequencing and were reanalyzed by phage ELISA with other panels of HIV-positive and -negative sera to confirm the specificity of reactivity. An alignment of inserts with the HIV-1 genome led to the identification of 12 immunodominant epitopes, mapping to Gag (p24 and p6), Pol, envelope (gp120 and gp41), and Nef. Interestingly, phages displaying sequences from the intracytoplasmic tail of gp41 (amino acids 784 to 871) were repeatedly recognized by antibodies from both acutely (1 to 6 months) and chronically HIV-infected individuals. The cytoplasmic tail of gp41 was selected as the primary candidate for the differential assay because it is unlikely to be targeted by HIV-neutralizing antibodies and is not included in most HIV vaccines currently under development. In addition, a p6 sequence was selected, even though it was included in some early HIV vaccines, since it contains very few HLA-restricted cytotoxic T-lymphocyte epitopes (10) (Los Alamos database [http://hiv-web.lanl.gov]). Importantly, the selected gp41 (amino acids 784 to 829 [SK1] and 836 to 871 [SK2]) and p6 (amino acids 452 to 502) sequences are highly conserved among all HIV-1 M subtypes.
Establishment of HIV-SELECTEST. The p6- and gp41-derived peptides were chemically synthesized and used for the development of the new assay. Peptides were designed based on the consensus sequences representing HIV-1 group M subtypes (Los Alamos HIV sequence database) to encompass the genetic variability among HIV-1 clades. Initially, each peptide was evaluated individually to determine its specificity and to establish cutoff values. Since both gp41 peptides (SK1 and SK2) displayed similar very low reactivities with HIV-seronegative samples, the two were combined. Multiple ELISA conditions were tested, and after screening of 1,000 seronegative samples, cutoff (CO) values for the gp41 (CO = 0.03) and p6 (CO = 0.15) peptides were determined. Each CO value represents the average absorbance value for negative sera (at a 1:100 dilution) plus 5 standard deviations. Additional serum panels containing high, intermediate, and low HIV-specific antibody titers were used to determine the dynamic range of the assay. Panels a and b in Fig. 1 demonstrate the binding of a serially diluted representative HIV-1-positive plasma, PRB-204-06 (from SeraCare BioServices), in the p6 and gp41 ELISAs, respectively. Titrations with additional samples demonstrated a higher maximum reactivity with the gp41 peptides and a broader dynamic range than those with the p6 peptide. Based on these analyses, all subsequent ELISA testing was conducted with a 1:100 dilution of serum or plasma. The HIV infection status of a given sample was determined with licensed detection kits conducted in-house and/or by other laboratories. Assay specificities of 100% for the gp41 peptides and 99.4% for the p6 peptide were established after screening of >2,500 samples either from uninfected individuals or from individuals infected with diverse HIV-1 clades. The combined sensitivity of the gp41 and p6 peptides was 99.3% for the detection of recent and chronic HIV infections in multiple geographical sites with clades A, B, C, D, E, F, and J and circulating recombinants (manuscript in preparation).
Assay robustness and statistical analysis. The reproducibility of the assay was determined by repeatedly testing nine HIV-seropositive and three HIV-seronegative samples from SeraCare BioServices. The distributions of the results obtained on multiple dates were evaluated for normality, and the appropriate P values were calculated using SigmaPlot. Representative plots are shown for one individual for the p6 and gp41 peptides (Fig. 1c and d, respectively). The upper and lower limits (±2 SD) represent the 95% confidence intervals. Interassay variability was 10%, and intraassay variability was 5% for all samples tested.
Acute infections are detected with HIV-SELECTEST. To determine how soon postinfection HIV-specific antibodies are detected with the HIV-SELECTEST, several well-characterized seroconversion panels were obtained from SeraCare BioServices containing sequential bleeds taken within 10 to 40 days of estimated exposure dates. As shown in Table 1, the p6 peptide reacted positively with PBR-910 on collection day 26, in agreement with results obtained using licensed HIV antibody detection kits. The gp41 peptides were reactive with the day 32 sample from the same individual. For PRB-929, day 25 and day 28 samples reacted with the p6 and gp41 peptides, respectively (Table 1). For that individual, infection was confirmed by PCR on day 14, and the Abbott HIV Ag test was positive on day 18. Similar results were obtained with additional seroconversion panels from SeraCare BioServices (data not shown) and demonstrated that HIV infection could be detected by the HIV-SELECTEST within 2 to 4 weeks following HIV-1 RNA detection by PCR, concurrent with the sensitivity limits of licensed HIV diagnostic tests. In addition, we evaluated 28 HIV seroconversion panels from Australia spanning 6 to 18 months postinfection (Table 2 and data not shown). With these panels, p6 showed variable reactivities at later times postinfection, whereas anti-gp41 reactivity increased over time and was maintained at high levels in most individuals, indicating that the kinetics and avidity of the antibody responses against the p6 and gp41 epitopes were not linked.
Evaluation of samples from HIV vaccine trials. The main proof of concept in support of the HIV-SELECTEST should come from evaluating the reactivities of vaccine-induced antibodies in the course of prophylactic vaccine trials. To that end, six blinded panels from completed vaccine trials (502 vaccinees), including HVTN 203 (conducted by the HIV Vaccine Trial Network), RV124 (conducted by the Walter Reed Army Institute of Research), VRC 004, VRC 006, VRC 009, and VRC 010 (conducted by the Vaccine Research Center, NIAID, NIH), were tested. A description of the vaccine constructs used in the various trials and a summary of the results obtained with the HIV-SELECTEST appear in Table 3. Data are shown for samples tested 2 to 4 weeks after the final immunization in various trials by both the HIV-SELECTEST and other HIV serodetection kits. Canarypox virus vaccine constructs used in the RV124 and HVTN 203 trials contained the p6 epitope used in the new assay. Additionally, the protein booster in the RV124 trial was gp160. In contrast, the vaccine constructs used in the VRC 004 and VRC 006 trials lacked the peptide sequences used in the HIV-SELECTEST.
The RV124 trial represents the worst-case scenario, wherein all the peptide sequences used in the HIV-SELECTEST were part of either the priming or boosting immunogens. After the last boost (day 182), 80% of vaccinees strongly seroconverted according to commercial HIV-1 detection kits, even though none were HIV infected (Table 3 and data not shown). In contrast, only two individuals scored positive in the p6 ELISA, and none were positive in the gp41 ELISA. These findings suggest that the epitopes used in the HIV-SELECTEST were not very immunogenic in the context of the RV124 vaccine constructs.
HVTN 203 specimens included coded samples obtained from 324 trial participants prevaccination and 4 and 6 months after the first vaccination. For this panel, 30% of vaccinees seroconverted according to licensed HIV detection assays, while only 12% reacted with the p6 peptide in the HIV-SELECTEST (Table 3). This finding was not surprising since the canarypox virus/HIV priming step (vCP1452) contained p6. Two specimens were also repeatedly reactive with the gp41 sequences that were not in the vaccine constructs. However, after decoding, it was confirmed that both samples were obtained from trial participants who became infected during this phase II trial.
The VRC phase I trials VRC 004, VRC 006, VRC 009, and VRC 010 were conducted in 2002-2005. The DNA plasmids (VRC 004) and nonreplicating recombinant adenovirus serotype 5 vector (rAd5) (VRC 006) used express Gag-Pol-Nef (in VRC 004) or Gag-Pol (in VRC 006) and multiclade (A, B, and C) envelope genes (gp145 in the DNA vaccine and gp140 in the rAd5 vaccine). Among the 50 participants in the VRC 004 trial, 38% (15/40) of vaccinated individuals seroconverted according to licensed HIV diagnostic kits (Table 3). Unexpectedly, two samples were positive in the gp41 ELISA, one of which also reacted with p6 in the HIV-SELECTEST (Table 3). Upon decoding, it was determined that the two individuals (both in the placebo arm) became infected during the VRC 004 trial (B. S. Graham et al., submitted for publication). In the VRC 006 trial (Ad5/HIV), no intercurrent HIV infections were identified, yet 60% of vaccine recipients (18/30) tested positive in licensed HIV detection tests (Graham et al., unpublished data). In contrast, none of the vaccinees reacted with either the p6 or gp41 peptides in the HIV-SELECTEST (Table 3). In the VRC 009 and VRC 010 trials, subsets of DNA-vaccinated individuals (from the VRC 004 and VRC 007 trials, respectively) were boosted with the rAd5/HIV vaccine. Samples from 4 weeks postboost demonstrated a very significant increase in total HIV-specific antibodies (data not shown) and 100% seroconversion using two rapid tests (Capillus HIV-1/HIV-2 and Uni-Gold HIV tests; Trinity Biotech, NY). Importantly, all vaccinees in these trials tested negative in the HIV-SELECTEST (Table 3).
Detection of intercurrent HIV infections during vaccine trials. The data obtained with the coded panels from the HIV vaccine trials indicate that vaccine-generated antibodies are unlikely to react in the HIV-SELECTEST, especially if the vaccines do not contain the p6 sequence. Importantly, the new test detected all intercurrent infections in the blinded samples. To further determine the sensitivity of the new assay at detecting acute HIV infections in the course of vaccine trials, we tested sequential samples from HIV infections in completed phase I, phase II, and phase III trials conducted by HVTN (10), VRC, and VaxGen (VAX 003/VAX 004 efficacy trials) (6).
As shown in Table 4 and Fig. 2a (also data not shown), sequential samples obtained from 22 trial participants infected with HIV during the HVTN trials and the VRC 004 trial reacted positively in the HIV-SELECTEST at early time points after the estimated infection dates. Importantly, no reactivity in the HIV-SELECTEST was observed prior to HIV infection of trial participants, although they had been immunized with complex vaccine products.
Sequential samples taken soon after the first confirmed PCR-positive visit were also obtained for 65 HIV infections during the VAX 003 (AIDSVAX gp120 B/E) trial and for 81 HIV infections during the VAX 004 (AIDSVAX gp120 B/B') trial conducted by VaxGen. The dates of PCR positivity and seroconversion by licensed HIV tests were provided by VaxGen. Table 5 contains an analysis of two representative HIV infections in the VAX 003 and VAX 004 trials that developed strong reactivities to the p6 and gp41 peptides. Furthermore, the HIV-SELECTEST identified all intercurrent HIV infections within 90 days of PCR confirmation (Fig. 2b and c for VAX 003 and VAX 004, respectively; data not shown).
It was also possible to compare the performance of the HIV-SELECTEST with results obtained with the FDA-licensed kits provided by VaxGen (Fig. 3). In most cases, the earliest positive results were observed with the same samples for the licensed diagnostics and the HIV-SELECTEST (dots falling on the diagonal lines). Surprisingly, 24 intercurrent HIV infections in the VAX 003 trial and 25 infections in the VAX 004 trial were detected earlier with the HIV-SELECTEST than with the licensed kits (dots under the diagonal lines in Fig. 3a and b), displaying the efficacy of the HIV-SELECTEST in the early diagnosis of HIV infection. Therefore, the new assay could be part of an algorithm that will provide an important differential diagnostic tool during future phase III prophylactic vaccine trials and for testing of blood and tissue donors.
DISCUSSION
The HIV pandemic continues to take its global toll, with more than 16,000 reported infections and 8,500 deaths occurring daily. Concerted efforts are under way to develop preventative HIV vaccines that will be both efficacious and economical. In the wake of unsuccessful efficacy trials conducted with vaccines containing the gp120 envelope alone (5, 7), the new generation of vaccine candidates are complex products, containing multiple HIV genes or proteins and using diverse delivery systems and new adjuvants. It is anticipated that within a few years, several additional vaccine candidates will progress to large-scale efficacy trials, particularly in countries with high infection rates. It is hoped that this next generation of vaccines will offer at least partial protection against new infections and possibly reduce viral loads and delay disease progression in infected vaccinated individuals. To achieve the statistical power needed to demonstrate partial efficacy, it will be necessary to recruit thousands of volunteers into future phase III HIV vaccine trials. Many of these volunteers will react positively in licensed HIV detection tests. Hence, further improvements in HIV diagnosis are urgently required.
One of the critical determinations during ongoing trials with high-risk populations is the HIV infection status of trial participants. Intercurrent infections must be detected as soon as possible in order to stop vaccination and monitor infected individuals for viral load, immune status, and disease progression. Treatment and public health measures depend on timely diagnostic information. Currently, vaccine trials use an algorithm of HIV detection that incorporates antibody- or antigen-based kits, followed by Western blots and, finally, confirmatory PCR-based assays. Unfortunately, many vaccine trial participants, irrespective of their HIV infection status, seroconvert according to all licensed antibody detection kits, including the recently licensed rapid tests (1, 4, 11, 12), because the vaccine components are the same as those detected by the diagnostic kits. Therefore, the recruitment of volunteers into future trials may be impeded if forms for informed consent need to state that volunteers are likely to seroconvert in licensed detection kits and may remain seropositive for a long time. In published surveys, it has been shown that positive HIV serodiagnosis is the most important concern for volunteers willing to participate in HIV clinical trials (8). Thus, there is an immediate need to develop a simple and inexpensive assay that does not indicate that uninfected vaccine recipients are infected but provides the necessary specificity and sensitivity to detect true HIV infections in the presence of vaccine-induced antibodies.
The use of a GFPDL to clone and express the entire open reading frame of HIV afforded us the opportunity to identify all of the epitopes that are recognized by antibodies shortly after HIV infection. Affinity selection of the phage display library using recent seroconversion panels led to the identification of epitopes in gp41 and p6 that were selected to develop a new differential diagnostic test.
Our studies demonstrated that vaccine-generated antibodies scored either negative or weakly positive in the HIV-SELECTEST even when the p6 or gp41 sequences were part of the vaccine constructs (i.e., in the RV124 and HVTN 203 trials). Furthermore, the HIV-SELECTEST detected all intercurrent HIV infections. It should be noted that while all intercurrent infections in the VAX 004 trial (conducted in the United States and The Netherlands) were with clade B viruses, all of the HIV infections in the VAX 003 trial (conducted in Thailand) were with clade E variants, demonstrating the feasibility of using the HIV-SELECTEST outside the United States in a multiclade scenario, which is a prerequisite for global vaccine trials.
Together, these data provide strong proof of concept for the specificities and sensitivities of the new p6 and gp41 peptide-based ELISAs. They further suggest that if future vaccine candidates do not contain these epitopes, then all uninfected vaccinees are expected to score negative in the new assay. In contrast, antibodies generated following intercurrent infections in the course of HIV vaccine trials or at later times should be detected by the HIV-SELECTEST soon after infection.
This inexpensive and high-throughput assay could be added to the algorithm of detection tests used at clinical sites and in blood and plasma collection centers. As such, this assay will be highly relevant for the early diagnosis of intercurrent HIV infections in future vaccine trials. This is particularly needed for HIV vaccines that, while not able to prevent infection, may reduce viral loads after acquisition. Importantly, the HIV-SELECTEST should help to alleviate the concerns regarding social and economic harms due to long-term seroconversion of uninfected participants in preventive HIV vaccine trials.
ACKNOWLEDGMENTS
This project was supported in part by an NIH Bench-to-Bedside grant in 2003-2004.
We thank the volunteers and the numerous staff members of the clinical trial organizations who made these studies possible. We thank Keith Peden, Basil Golding, Indira Hewlett, and Elliot Cowan for their thorough reviews of the manuscript.
REFERENCES
Ackers, M. L., B. Parekh, T. G. Evans, P. Berman, S. Phillips, M. Allen, and J. S. McDougal. 2003. Human immunodeficiency virus (HIV) seropositivity among uninfected HIV vaccine recipients. J. Infect. Dis. 187:879-886.
Allen, M., H. Israel, K. Rybczyk, M. A. Pugliese, K. Loughran, L. Wagner, and S. Erb. 2001. Trial-related discrimination in HIV vaccine clinical trials. AIDS Res. Hum. Retrovir. 17:667-674.
Belshe, R. B., M. L. Clements, M. C. Keefer, B. S. Graham, L. Corey, R. Sposto, S. Wescott, and D. Lawrence. 1994. Interpreting HIV serodiagnostic test results in the 1990s: social risks of HIV vaccine studies in uninfected volunteers. NIAID AIDS Vaccine Clinical Trials Group. Ann. Intern. Med. 121:584-589.
Chuenchitra, T., C. Wasi, S. Louisirirojchanakul, S. Nitayaphan, R. Sutthent, J. H. Cox, M. S. De Souza, A. E. Brown, D. L. Birx, and V. R. Polonis. 2003. Longitudinal study of humoral immune responses in HIV type 1 subtype CRF01_AE (E)-infected Thai patients with different rates of disease progression. AIDS Res. Hum. Retrovir. 19:293-305.
Francis, D. P., W. L. Heyward, V. Popovic, P. Orozco-Cronin, K. Orelind, C. Gee, A. Hirsch, T. Ippolito, A. Luck, M. Longhi, V. Gulati, N. Winslow, M. Gurwith, F. Sinangil, and P. W. Berman. 2003. Candidate HIV/AIDS vaccines: lessons learned from the world's first phase III efficacy trials. AIDS 17:147-156.
Gilbert, P. B., M. L. Peterson, D. Follmann, M. G. Hudgens, D. P. Francis, M. Gurwith, W. L. Heyward, D. V. Jobes, V. Popovic, S. G. Self, F. Sinangil, D. Burke, and P. W. Berman. 2005. Correlation between immunologic responses to a recombinant glycoprotein 120 vaccine and incidence of HIV-1 infection in a phase 3 HIV-1 preventive vaccine trial. J. Infect. Dis. 191:666-677.
Graham, B. S., and J. R. Mascola. 2005. Lessons from failure—preparing for future HIV-1 vaccine efficacy trials. J. Infect. Dis. 191:647-649.
Gross, M., G. R. Seage, K. H. Mayer, R. S. Goldstein, E. Losina, and C. Wold. 1996. Interest among gay/bisexual men in greater Boston in participating in clinical trials of preventive HIV vaccines. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 12:406-412.
Koblin, B. A., P. Heagerty, A. Sheon, S. Buchbinder, C. Celum, J. M. Douglas, M. Gross, M. Marmor, K. Mayer, D. Metzger, and G. Seage. 1998. Readiness of high-risk populations in the HIV Network for Prevention Trials to participate in HIV vaccine efficacy trials in the United States. AIDS 12:785-793.
Lee, D., B. S. Graham, Y. L. Chiu, P. B. Gilbert, M. J. McElrath, R. B. Belshe, S. P. Buchbinder, H. W. Sheppard, B. A. Koblin, K. H. Mayer, M. C. Keefer, M. J. Mulligan, and C. L. Celum. 2004. Breakthrough infections during phase 1 and 2 prime-boost HIV-1 vaccine trials with canarypox vectors (ALVAC) and booster dose of recombinant gp120 or gp160. J. Infect. Dis. 190:903-907.
Pitisuttithum, P., S. Nitayaphan, P. Thongcharoen, C. Khamboonruang, J. Kim, M. de Souza, T. Chuenchitra, R. P. Garner, D. Thapinta, V. Polonis, S. Ratto-Kim, P. Chanbancherd, J. Chiu, D. L. Birx, A. M. Duliege, J. G. McNeil, and A. E. Brown. 2003. Safety and immunogenicity of combinations of recombinant subtype E and B human immunodeficiency virus type 1 envelope glycoprotein 120 vaccines in healthy Thai adults. J. Infect. Dis. 188:219-227.
Schwartz, D. H., A. Mazumdar, S. Winston, and S. Harkonen. 1995. Utility of various commercially available human immunodeficiency virus (HIV) antibody diagnostic kits for use in conjunction with efficacy trials of HIV type 1 vaccines. Clin. Diagn. Lab. Immunol. 2:268-271.
Sheon, A. R., L. Wagner, M. J. McElrath, M. C. Keefer, E. Zimmerman, H. Israel, D. Berger, and P. Fast. 1998. Preventing discrimination against volunteers in prophylactic HIV vaccine trials: lessons from a phase II trial. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 19:519-526.(Surender Khurana, James N)
Core Facility, Center for Biologics Evaluation and Research (CBER), FDA, Bethesda, Maryland 20892
Office of AIDS Research, NIH, Bethesda, Maryland 20892
Vaccine Clinical Research Branch, Division of AIDS, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, Maryland 20892
Vaccine Research Center, National Institute of Allergy and Infectious Diseases, NIH, Bethesda, Maryland 20892
U.S. Army Medical Component-Armed Forces Research Institute of Medical Sciences (USAMC-AFRIMS), APO AP 96546
U.S. Military HIV Research Program, Rockville, Maryland 20850
National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia 2010
VaxGen, Inc., Brisbane, California 94005
ABSTRACT
All current human immunodeficiency virus (HIV) vaccine candidates contain multiple viral components and elicit antibodies that react positively in licensed HIV diagnostic tests, which contain similar viral products. Thus, vaccine trial participants could be falsely diagnosed as infected with HIV. Additionally, uninfected, seropositive vaccinees may encounter long-term social and economic harms. Moreover, this also interferes with early detection of true HIV infections during preventive HIV vaccine trials. An HIV-seropositive test result among uninfected vaccine trial participants is a major public health concern for volunteers who want to participate in future HIV vaccine trials. Based on the increased number of HIV vaccines being tested globally, it is essential to differentiate vaccine- from virus-induced antibodies. Using a whole-HIV-genome phage display library, we identified conserved sequences in Env-gp41 and Gag-p6 which are recognized soon after infection, do not contain protective epitopes, and are not part of most current HIV vaccines. We established a new HIV serodetection assay based on these peptides. To date, this assay, termed HIV-SELECTEST, demonstrates >99% specificity and sensitivity. Importantly, in testing of plasma samples from multiple HIV vaccine trials, uninfected trial participants scored negative, while all intercurrent infections were detected within 1 to 3 months of HIV infection. The new HIV-SELECTEST is a simple but robust diagnostic tool for easy implementation in HIV vaccine trials and blood banks worldwide.
INTRODUCTION
Since 1987, more than 25,000 individuals have participated in clinical trials with preventive human immunodeficiency virus (HIV) vaccines. Most of the current HIV vaccine candidates are complex products containing multiple viral genes or proteins, many of which are included in licensed HIV serodetection kits, including recently licensed rapid tests. Consequently, sera from vaccine recipients often react in licensed HIV serodetection assays, generating patterns indistinguishable from those for HIV-infected individuals (1, 4, 11, 12). This will complicate future prophylactic vaccine trials, in which early detection of HIV infections is of paramount importance. Furthermore, long-term HIV seropositivity will exclude vaccine trial participants from the pool of blood and plasma donors and may contribute to a range of socioeconomic harms such as denied employment, health insurance, travel, immigration, and recruitment to the armed forces (2, 3). The prospect of seroconversion without a clear differential diagnosis algorithm could deter potential trial participants and severely curtail recruitment into large-scale trials around the globe (8, 9, 13).
Currently, there is no HIV serodiagnostic assay that differentiates between vaccine-generated antibodies and those produced after true HIV infection. Our goal was to develop an antibody-based HIV-1 detection assay in which vaccine-generated antibodies will score negative, whereas virus-induced antibodies can be detected soon after HIV infection. The selection criteria for peptides to be used in such an assay were as follows: (i) epitopes that are not included in HIV vaccines since they do not appear to contribute to protective immunity, (ii) epitopes recognized by antibodies made soon after HIV infection, and (iii) epitopes highly conserved among HIV clades and subtypes.
To identify such sequences, a gene fragment phage display library (GFPDL) was constructed from the entire HIV type 1 (HIV-1) genome and used to screen sera from HIV-infected individuals near the time of seroconversion. This strategy led to the discovery of three novel epitopes, one in Gag p6 and two in the envelope gp41 cytoplasmic tail. Herein, we describe the development of a new HIV-specific enzyme-linked immunosorbent assay (ELISA), termed HIV-SELECTEST, which distinguishes between HIV-infected individuals and uninfected vaccine recipients. The HIV-SELECTEST is a low-cost, high-throughput assay that could be implemented in clinical sites and blood collection centers worldwide and serve as an add-on diagnostic tool in future HIV vaccine trials.
MATERIALS AND METHODS
Construction of complete HIV genome-based gene fragment phage display library. Plasmid pNL4-3, containing the complete HIV-1 NL4-3 proviral DNA, was obtained from the NIH AIDS Research and Reference Reagent Program (McKesson BioServices Corp., Rockville, MD). The full-length HIV-1 genome was PCR amplified from pNL4-3 DNA with an Expand long-template polymerase preparation (Roche Diagnostics, Indianapolis, IN) and primers spanning the Lys tRNA primer binding site (MSF12 [5'-AAAAATCTCTAGCAGTGGCGCCCGAACAG-3']) and the poly(A) signal region of the 3' long terminal repeat (MSR5 [5'-AAGCACTCAAGGCAAGCTTTATTGAGGCT-3']), which amplifies the entire HIV-1 genome except for 75 bp in the unique 5' region of the long terminal repeat. The purified amplified DNA product was digested with DNase I using a DNase shotgun cleavage kit (Novagen, Madison, WI), and fragments between 50 and 300 bp were isolated by preparative gel electrophoresis, treated with T4 DNA polymerase to generate blunt ends, and dephosphorylated using calf intestinal alkaline phosphatase (Roche Diagnostics, Indianapolis, IN). DNAs were purified again using a nucleotide removal kit (QIAGEN Inc., Valencia, CA) and ligated in the presence of the SrfI enzyme into the SmaI site of the M13-derived phage vector for expression as gIIIp fusion proteins, followed by electroporation into Escherichia coli TG1 cells. Tet-resistant transformants were harvested and expanded in liquid culture (2x YT) at 37°C. The cell-free phage supernatant was isolated by centrifugation, and the phage titer was determined and expressed in Tetr transduction units. Ninety-six individual clones were isolated, and DNA inserts were amplified by standard PCR and sequenced to determine the insert size distribution and library diversity.
Selection of phages reactive with HIV antibodies from early-phase infected individuals. Seven plasma samples constituting the HIV-1 seroconversion panel PRB-910 from SeraCare BioServices (Gaithersburg, MD) were used for panning of the HIV-1 GFPDL. For the removal of plasma components, which could nonspecifically interact with phage proteins, a fivefold-diluted plasma was preadsorbed three times to sterile polystyrene petri dishes (35-mm diameter) coated with 1013 UV-killed VCSM13 phage. For biopanning, microtiter strips (Nunc Inc., Naperville, IL) were coated with a mixture of 500 ng each of goat anti-human immunoglobulin G (IgG) Fc- and goat anti-human IgM Fc-specific antibodies in phosphate-buffered saline (PBS), pH 7.4. After three washings with PBST (20 mM PBS containing 0.1% Tween 20), Dulbecco's modified Eagle's medium (DMEM) containing 5% fetal bovine serum (FBS) (blocking solution) was added to wells to block the unoccupied reactive sites. HIV-1-infected human plasma preadsorbed to VCSM13 was added to the wells and incubated for 1 h at room temperature (RT). Wells were washed thrice with PBST, and 1010 phage per well of the HIV-1 GFPDL, diluted in blocking solution, were added for 2 h at RT. Unbound phage were removed by 12 washes with PBST followed by 3 washes with PBS. Bound phage were eluted by the addition of 0.1 N HCl containing bovine serum albumin (1 mg/ml) for 10 min at RT and were neutralized by the addition of 8 μl of 2 M Tris solution per 100 μl eluate. Four rounds of affinity selection were carried out with each individual serum sample from HIV seroconversion panel PRB-910.
Analysis of affinity-selected phage clones. Twenty-two phage clones enriched after four rounds of biopanning on each PRB-910 plasma sample were further screened for specific recognition by HIV-seropositive sera and the absence of reactivity with seronegative sera in an affinity-capture phage-specific ELISA. The wells of ELISA plates (Immulon 2HB; Thermo Labsystems, Franklin, MA) were coated with 100 ng/well of anti-phage antibody (GE Healthcare, Piscataway, NJ) and blocked with DMEM-5% FBS. Subsequently, 1010 phage of the selected clones were added per well and incubated for 1 h at RT. Serially diluted sera (in DMEM-5% FBS) were added to the 96-well plates in duplicate and incubated at RT for 1 h. The bound antibodies were probed with horseradish peroxidase-conjugated goat anti-human IgG-IgM antibodies, and the reactions were developed with O-phenylenediamine substrate solution (Pierce Biotechnology, Rockford, IL). The clones demonstrating the best differential reactivities with HIV-1-seropositive sera were expanded, and the inserts were sequenced and mapped to individual HIV-1 genes. Several inserts were selected for synthetic peptide synthesis and development of the HIV-SELECTEST.
Peptides used in HIV-SELECTEST. Peptide sequences from Gag p6 (452-SRPEPTAPPAESFRFGEEITPTPSQKQEPKDKELYPPLASLRSLFGNDPSSQ-502) and the gp41 cytoplasmic region (SK1 [784-LIAARIVELLGHSSLKGLRRGWEALKYLWNLLQYWGQELKNSAISL-829] and SK2 [836-AVAEGTDRVIEVVQRVCRAILNIPRRIRQGFERALL-871]) were chemically synthesized (amino acid residues are numbered based on the CON-OF-CONS alignment sequence in the Los Alamos database). All peptides were synthesized at the Facility for Biotechnology Resources, CBER, FDA, on Applied Biosystems peptide synthesizer models 431 and 433 (Foster City, California) by standard 9-fluorenyl methoxycarbonyl chemistry. Peptides were purified by reverse-phase high-performance liquid chromatography and characterized by matrix-assisted laser desorption ionization-time of flight mass spectrometry.
HIV-SELECTEST. Based on preliminary screening of HIV-seronegative and -seropositive sera, the optimal conditions for the p6 and gp41 ELISAs were determined. The p6 peptide was used to coat Immulon-2HB plates at 30 ng/100 μl/well, while the gp41 peptides (SK1 and SK2) were used at 150 ng/100 μl/well (each; total, 300 ng/well). After three washes with PBST (20 mM PBS, 0.1% Tween 20), the unoccupied reactive sites were blocked by PBST containing 2% whole milk (2% WMPBST). All specimens (serum or plasma) were diluted 1:100 in 2% WMPBST, added to peptide-coated wells, and incubated for 1 h at RT. The plates were then washed six times with PBST, and 100 μl/well of horseradish peroxidase-conjugated goat anti-human IgG Fc-specific antibody (Jackson ImmunoResearch, West Grove, PA), diluted 1:10,000 in 2% WMPBST, was added. The reactions were quantified using O-phenylenediamine substrate.
Based on the results for 1,000 seronegative samples, cutoff (CO) values were determined individually for the p6 and gp41 peptides. The cutoff values used were the average absorbance values for negative sera (at a 1:100 dilution) plus 5 standard deviations (SD; for each peptide). Specimens with absorbance/cutoff ratios of 1 were considered HIV-1 seropositive, and those with ratios of <1 were considered HIV-1 seronegative.
HIV seroconversion panels and vaccine trial samples. HIV-1 seroconversion panels PRB-910, PRB-924, PRB-927, PRB-928, PRB-929, and PRB-931 and the mixed-titer panel PRB-204 were purchased from SeraCare BioServices (Gaithersburg, MD). A seroconversion panel consists of plasma samples collected serially early after HIV-1 infection, and the virological and immunological profiles for these plasma samples, as assessed by commercial diagnostic kits, were provided by SeraCare BioServices. Additionally, 28 seroconversion panels were provided by the University of New South Wales (PHAEDRA Inventory, Sydney, Australia). HIV-negative serum samples were obtained from the National Institutes of Health Blood Bank and the Vaccine Research Center (VRC, NIAID, NIH, Bethesda, MD).
Serum/plasma samples from the following HIV vaccine trials were tested: HVTN 203 (246 vaccinees and 78 patients receiving placebos; conducted by the HIV Vaccine Trial Network [HVTN]), RV124 (conducted by the Walter Reed Army Institute of Research), VRC 004 (40 vaccinees and 10 patients receiving placebos), VRC 006 (30 vaccinees and 6 patients receiving placebos), VRC 009 (9 vaccinees and no patients receiving placebos), and VRC 010 were conducted by the Vaccine Research Center (NIAID, NIH), and VAX 003 and VAX 004 were conducted by VaxGen Inc. The HIV infection status of a given sample was provided by the collaborating groups and also determined by in-house testing using a Bio-Rad HIV-1/2 Plus O kit (Bio-Rad Laboratories, Woodinwille, WA). Samples obtained from the VRC 009 and VRC 010 trials were also tested with Capillus HIV-1/HIV-2 (not licensed by the U.S. FDA) and Uni-Gold HIV (Trinity Biotech, NY) rapid tests along with the Bio-Rad HIV-1/2 Plus O kit.
All studies were conducted under approval from the Research Involving Human Subjects Committee (RIHSC exemption no. 04-050B) at the Center for Biologics Evaluation and Research.
RESULTS
Identification of HIV sequences recognized by early-phase seroconversion sera using GFPDL. To identify all the HIV sequences recognized by antibodies generated soon after HIV infection, we constructed a GFPDL spanning the entire HIV-1 open reading frame of NL4-3. The HIV-1 GFPDL contained >107 independent transformants. PCR-based analysis and sequencing of the inserts confirmed that the library consisted of 100% recombinants, with insert sizes of 50 to 300 bp and random distribution across the HIV genome.
Seven plasma samples constituting seroconversion panel PRB-910 (obtained from acutely HIV-1-infected individuals; SeraCare BioServices, Gaithersburg, MD) were used for affinity selection of phages displaying HIV-1 peptides. After four rounds of affinity selection, 22 clones (for each plasma sample) were selected for insert sequencing and were reanalyzed by phage ELISA with other panels of HIV-positive and -negative sera to confirm the specificity of reactivity. An alignment of inserts with the HIV-1 genome led to the identification of 12 immunodominant epitopes, mapping to Gag (p24 and p6), Pol, envelope (gp120 and gp41), and Nef. Interestingly, phages displaying sequences from the intracytoplasmic tail of gp41 (amino acids 784 to 871) were repeatedly recognized by antibodies from both acutely (1 to 6 months) and chronically HIV-infected individuals. The cytoplasmic tail of gp41 was selected as the primary candidate for the differential assay because it is unlikely to be targeted by HIV-neutralizing antibodies and is not included in most HIV vaccines currently under development. In addition, a p6 sequence was selected, even though it was included in some early HIV vaccines, since it contains very few HLA-restricted cytotoxic T-lymphocyte epitopes (10) (Los Alamos database [http://hiv-web.lanl.gov]). Importantly, the selected gp41 (amino acids 784 to 829 [SK1] and 836 to 871 [SK2]) and p6 (amino acids 452 to 502) sequences are highly conserved among all HIV-1 M subtypes.
Establishment of HIV-SELECTEST. The p6- and gp41-derived peptides were chemically synthesized and used for the development of the new assay. Peptides were designed based on the consensus sequences representing HIV-1 group M subtypes (Los Alamos HIV sequence database) to encompass the genetic variability among HIV-1 clades. Initially, each peptide was evaluated individually to determine its specificity and to establish cutoff values. Since both gp41 peptides (SK1 and SK2) displayed similar very low reactivities with HIV-seronegative samples, the two were combined. Multiple ELISA conditions were tested, and after screening of 1,000 seronegative samples, cutoff (CO) values for the gp41 (CO = 0.03) and p6 (CO = 0.15) peptides were determined. Each CO value represents the average absorbance value for negative sera (at a 1:100 dilution) plus 5 standard deviations. Additional serum panels containing high, intermediate, and low HIV-specific antibody titers were used to determine the dynamic range of the assay. Panels a and b in Fig. 1 demonstrate the binding of a serially diluted representative HIV-1-positive plasma, PRB-204-06 (from SeraCare BioServices), in the p6 and gp41 ELISAs, respectively. Titrations with additional samples demonstrated a higher maximum reactivity with the gp41 peptides and a broader dynamic range than those with the p6 peptide. Based on these analyses, all subsequent ELISA testing was conducted with a 1:100 dilution of serum or plasma. The HIV infection status of a given sample was determined with licensed detection kits conducted in-house and/or by other laboratories. Assay specificities of 100% for the gp41 peptides and 99.4% for the p6 peptide were established after screening of >2,500 samples either from uninfected individuals or from individuals infected with diverse HIV-1 clades. The combined sensitivity of the gp41 and p6 peptides was 99.3% for the detection of recent and chronic HIV infections in multiple geographical sites with clades A, B, C, D, E, F, and J and circulating recombinants (manuscript in preparation).
Assay robustness and statistical analysis. The reproducibility of the assay was determined by repeatedly testing nine HIV-seropositive and three HIV-seronegative samples from SeraCare BioServices. The distributions of the results obtained on multiple dates were evaluated for normality, and the appropriate P values were calculated using SigmaPlot. Representative plots are shown for one individual for the p6 and gp41 peptides (Fig. 1c and d, respectively). The upper and lower limits (±2 SD) represent the 95% confidence intervals. Interassay variability was 10%, and intraassay variability was 5% for all samples tested.
Acute infections are detected with HIV-SELECTEST. To determine how soon postinfection HIV-specific antibodies are detected with the HIV-SELECTEST, several well-characterized seroconversion panels were obtained from SeraCare BioServices containing sequential bleeds taken within 10 to 40 days of estimated exposure dates. As shown in Table 1, the p6 peptide reacted positively with PBR-910 on collection day 26, in agreement with results obtained using licensed HIV antibody detection kits. The gp41 peptides were reactive with the day 32 sample from the same individual. For PRB-929, day 25 and day 28 samples reacted with the p6 and gp41 peptides, respectively (Table 1). For that individual, infection was confirmed by PCR on day 14, and the Abbott HIV Ag test was positive on day 18. Similar results were obtained with additional seroconversion panels from SeraCare BioServices (data not shown) and demonstrated that HIV infection could be detected by the HIV-SELECTEST within 2 to 4 weeks following HIV-1 RNA detection by PCR, concurrent with the sensitivity limits of licensed HIV diagnostic tests. In addition, we evaluated 28 HIV seroconversion panels from Australia spanning 6 to 18 months postinfection (Table 2 and data not shown). With these panels, p6 showed variable reactivities at later times postinfection, whereas anti-gp41 reactivity increased over time and was maintained at high levels in most individuals, indicating that the kinetics and avidity of the antibody responses against the p6 and gp41 epitopes were not linked.
Evaluation of samples from HIV vaccine trials. The main proof of concept in support of the HIV-SELECTEST should come from evaluating the reactivities of vaccine-induced antibodies in the course of prophylactic vaccine trials. To that end, six blinded panels from completed vaccine trials (502 vaccinees), including HVTN 203 (conducted by the HIV Vaccine Trial Network), RV124 (conducted by the Walter Reed Army Institute of Research), VRC 004, VRC 006, VRC 009, and VRC 010 (conducted by the Vaccine Research Center, NIAID, NIH), were tested. A description of the vaccine constructs used in the various trials and a summary of the results obtained with the HIV-SELECTEST appear in Table 3. Data are shown for samples tested 2 to 4 weeks after the final immunization in various trials by both the HIV-SELECTEST and other HIV serodetection kits. Canarypox virus vaccine constructs used in the RV124 and HVTN 203 trials contained the p6 epitope used in the new assay. Additionally, the protein booster in the RV124 trial was gp160. In contrast, the vaccine constructs used in the VRC 004 and VRC 006 trials lacked the peptide sequences used in the HIV-SELECTEST.
The RV124 trial represents the worst-case scenario, wherein all the peptide sequences used in the HIV-SELECTEST were part of either the priming or boosting immunogens. After the last boost (day 182), 80% of vaccinees strongly seroconverted according to commercial HIV-1 detection kits, even though none were HIV infected (Table 3 and data not shown). In contrast, only two individuals scored positive in the p6 ELISA, and none were positive in the gp41 ELISA. These findings suggest that the epitopes used in the HIV-SELECTEST were not very immunogenic in the context of the RV124 vaccine constructs.
HVTN 203 specimens included coded samples obtained from 324 trial participants prevaccination and 4 and 6 months after the first vaccination. For this panel, 30% of vaccinees seroconverted according to licensed HIV detection assays, while only 12% reacted with the p6 peptide in the HIV-SELECTEST (Table 3). This finding was not surprising since the canarypox virus/HIV priming step (vCP1452) contained p6. Two specimens were also repeatedly reactive with the gp41 sequences that were not in the vaccine constructs. However, after decoding, it was confirmed that both samples were obtained from trial participants who became infected during this phase II trial.
The VRC phase I trials VRC 004, VRC 006, VRC 009, and VRC 010 were conducted in 2002-2005. The DNA plasmids (VRC 004) and nonreplicating recombinant adenovirus serotype 5 vector (rAd5) (VRC 006) used express Gag-Pol-Nef (in VRC 004) or Gag-Pol (in VRC 006) and multiclade (A, B, and C) envelope genes (gp145 in the DNA vaccine and gp140 in the rAd5 vaccine). Among the 50 participants in the VRC 004 trial, 38% (15/40) of vaccinated individuals seroconverted according to licensed HIV diagnostic kits (Table 3). Unexpectedly, two samples were positive in the gp41 ELISA, one of which also reacted with p6 in the HIV-SELECTEST (Table 3). Upon decoding, it was determined that the two individuals (both in the placebo arm) became infected during the VRC 004 trial (B. S. Graham et al., submitted for publication). In the VRC 006 trial (Ad5/HIV), no intercurrent HIV infections were identified, yet 60% of vaccine recipients (18/30) tested positive in licensed HIV detection tests (Graham et al., unpublished data). In contrast, none of the vaccinees reacted with either the p6 or gp41 peptides in the HIV-SELECTEST (Table 3). In the VRC 009 and VRC 010 trials, subsets of DNA-vaccinated individuals (from the VRC 004 and VRC 007 trials, respectively) were boosted with the rAd5/HIV vaccine. Samples from 4 weeks postboost demonstrated a very significant increase in total HIV-specific antibodies (data not shown) and 100% seroconversion using two rapid tests (Capillus HIV-1/HIV-2 and Uni-Gold HIV tests; Trinity Biotech, NY). Importantly, all vaccinees in these trials tested negative in the HIV-SELECTEST (Table 3).
Detection of intercurrent HIV infections during vaccine trials. The data obtained with the coded panels from the HIV vaccine trials indicate that vaccine-generated antibodies are unlikely to react in the HIV-SELECTEST, especially if the vaccines do not contain the p6 sequence. Importantly, the new test detected all intercurrent infections in the blinded samples. To further determine the sensitivity of the new assay at detecting acute HIV infections in the course of vaccine trials, we tested sequential samples from HIV infections in completed phase I, phase II, and phase III trials conducted by HVTN (10), VRC, and VaxGen (VAX 003/VAX 004 efficacy trials) (6).
As shown in Table 4 and Fig. 2a (also data not shown), sequential samples obtained from 22 trial participants infected with HIV during the HVTN trials and the VRC 004 trial reacted positively in the HIV-SELECTEST at early time points after the estimated infection dates. Importantly, no reactivity in the HIV-SELECTEST was observed prior to HIV infection of trial participants, although they had been immunized with complex vaccine products.
Sequential samples taken soon after the first confirmed PCR-positive visit were also obtained for 65 HIV infections during the VAX 003 (AIDSVAX gp120 B/E) trial and for 81 HIV infections during the VAX 004 (AIDSVAX gp120 B/B') trial conducted by VaxGen. The dates of PCR positivity and seroconversion by licensed HIV tests were provided by VaxGen. Table 5 contains an analysis of two representative HIV infections in the VAX 003 and VAX 004 trials that developed strong reactivities to the p6 and gp41 peptides. Furthermore, the HIV-SELECTEST identified all intercurrent HIV infections within 90 days of PCR confirmation (Fig. 2b and c for VAX 003 and VAX 004, respectively; data not shown).
It was also possible to compare the performance of the HIV-SELECTEST with results obtained with the FDA-licensed kits provided by VaxGen (Fig. 3). In most cases, the earliest positive results were observed with the same samples for the licensed diagnostics and the HIV-SELECTEST (dots falling on the diagonal lines). Surprisingly, 24 intercurrent HIV infections in the VAX 003 trial and 25 infections in the VAX 004 trial were detected earlier with the HIV-SELECTEST than with the licensed kits (dots under the diagonal lines in Fig. 3a and b), displaying the efficacy of the HIV-SELECTEST in the early diagnosis of HIV infection. Therefore, the new assay could be part of an algorithm that will provide an important differential diagnostic tool during future phase III prophylactic vaccine trials and for testing of blood and tissue donors.
DISCUSSION
The HIV pandemic continues to take its global toll, with more than 16,000 reported infections and 8,500 deaths occurring daily. Concerted efforts are under way to develop preventative HIV vaccines that will be both efficacious and economical. In the wake of unsuccessful efficacy trials conducted with vaccines containing the gp120 envelope alone (5, 7), the new generation of vaccine candidates are complex products, containing multiple HIV genes or proteins and using diverse delivery systems and new adjuvants. It is anticipated that within a few years, several additional vaccine candidates will progress to large-scale efficacy trials, particularly in countries with high infection rates. It is hoped that this next generation of vaccines will offer at least partial protection against new infections and possibly reduce viral loads and delay disease progression in infected vaccinated individuals. To achieve the statistical power needed to demonstrate partial efficacy, it will be necessary to recruit thousands of volunteers into future phase III HIV vaccine trials. Many of these volunteers will react positively in licensed HIV detection tests. Hence, further improvements in HIV diagnosis are urgently required.
One of the critical determinations during ongoing trials with high-risk populations is the HIV infection status of trial participants. Intercurrent infections must be detected as soon as possible in order to stop vaccination and monitor infected individuals for viral load, immune status, and disease progression. Treatment and public health measures depend on timely diagnostic information. Currently, vaccine trials use an algorithm of HIV detection that incorporates antibody- or antigen-based kits, followed by Western blots and, finally, confirmatory PCR-based assays. Unfortunately, many vaccine trial participants, irrespective of their HIV infection status, seroconvert according to all licensed antibody detection kits, including the recently licensed rapid tests (1, 4, 11, 12), because the vaccine components are the same as those detected by the diagnostic kits. Therefore, the recruitment of volunteers into future trials may be impeded if forms for informed consent need to state that volunteers are likely to seroconvert in licensed detection kits and may remain seropositive for a long time. In published surveys, it has been shown that positive HIV serodiagnosis is the most important concern for volunteers willing to participate in HIV clinical trials (8). Thus, there is an immediate need to develop a simple and inexpensive assay that does not indicate that uninfected vaccine recipients are infected but provides the necessary specificity and sensitivity to detect true HIV infections in the presence of vaccine-induced antibodies.
The use of a GFPDL to clone and express the entire open reading frame of HIV afforded us the opportunity to identify all of the epitopes that are recognized by antibodies shortly after HIV infection. Affinity selection of the phage display library using recent seroconversion panels led to the identification of epitopes in gp41 and p6 that were selected to develop a new differential diagnostic test.
Our studies demonstrated that vaccine-generated antibodies scored either negative or weakly positive in the HIV-SELECTEST even when the p6 or gp41 sequences were part of the vaccine constructs (i.e., in the RV124 and HVTN 203 trials). Furthermore, the HIV-SELECTEST detected all intercurrent HIV infections. It should be noted that while all intercurrent infections in the VAX 004 trial (conducted in the United States and The Netherlands) were with clade B viruses, all of the HIV infections in the VAX 003 trial (conducted in Thailand) were with clade E variants, demonstrating the feasibility of using the HIV-SELECTEST outside the United States in a multiclade scenario, which is a prerequisite for global vaccine trials.
Together, these data provide strong proof of concept for the specificities and sensitivities of the new p6 and gp41 peptide-based ELISAs. They further suggest that if future vaccine candidates do not contain these epitopes, then all uninfected vaccinees are expected to score negative in the new assay. In contrast, antibodies generated following intercurrent infections in the course of HIV vaccine trials or at later times should be detected by the HIV-SELECTEST soon after infection.
This inexpensive and high-throughput assay could be added to the algorithm of detection tests used at clinical sites and in blood and plasma collection centers. As such, this assay will be highly relevant for the early diagnosis of intercurrent HIV infections in future vaccine trials. This is particularly needed for HIV vaccines that, while not able to prevent infection, may reduce viral loads after acquisition. Importantly, the HIV-SELECTEST should help to alleviate the concerns regarding social and economic harms due to long-term seroconversion of uninfected participants in preventive HIV vaccine trials.
ACKNOWLEDGMENTS
This project was supported in part by an NIH Bench-to-Bedside grant in 2003-2004.
We thank the volunteers and the numerous staff members of the clinical trial organizations who made these studies possible. We thank Keith Peden, Basil Golding, Indira Hewlett, and Elliot Cowan for their thorough reviews of the manuscript.
REFERENCES
Ackers, M. L., B. Parekh, T. G. Evans, P. Berman, S. Phillips, M. Allen, and J. S. McDougal. 2003. Human immunodeficiency virus (HIV) seropositivity among uninfected HIV vaccine recipients. J. Infect. Dis. 187:879-886.
Allen, M., H. Israel, K. Rybczyk, M. A. Pugliese, K. Loughran, L. Wagner, and S. Erb. 2001. Trial-related discrimination in HIV vaccine clinical trials. AIDS Res. Hum. Retrovir. 17:667-674.
Belshe, R. B., M. L. Clements, M. C. Keefer, B. S. Graham, L. Corey, R. Sposto, S. Wescott, and D. Lawrence. 1994. Interpreting HIV serodiagnostic test results in the 1990s: social risks of HIV vaccine studies in uninfected volunteers. NIAID AIDS Vaccine Clinical Trials Group. Ann. Intern. Med. 121:584-589.
Chuenchitra, T., C. Wasi, S. Louisirirojchanakul, S. Nitayaphan, R. Sutthent, J. H. Cox, M. S. De Souza, A. E. Brown, D. L. Birx, and V. R. Polonis. 2003. Longitudinal study of humoral immune responses in HIV type 1 subtype CRF01_AE (E)-infected Thai patients with different rates of disease progression. AIDS Res. Hum. Retrovir. 19:293-305.
Francis, D. P., W. L. Heyward, V. Popovic, P. Orozco-Cronin, K. Orelind, C. Gee, A. Hirsch, T. Ippolito, A. Luck, M. Longhi, V. Gulati, N. Winslow, M. Gurwith, F. Sinangil, and P. W. Berman. 2003. Candidate HIV/AIDS vaccines: lessons learned from the world's first phase III efficacy trials. AIDS 17:147-156.
Gilbert, P. B., M. L. Peterson, D. Follmann, M. G. Hudgens, D. P. Francis, M. Gurwith, W. L. Heyward, D. V. Jobes, V. Popovic, S. G. Self, F. Sinangil, D. Burke, and P. W. Berman. 2005. Correlation between immunologic responses to a recombinant glycoprotein 120 vaccine and incidence of HIV-1 infection in a phase 3 HIV-1 preventive vaccine trial. J. Infect. Dis. 191:666-677.
Graham, B. S., and J. R. Mascola. 2005. Lessons from failure—preparing for future HIV-1 vaccine efficacy trials. J. Infect. Dis. 191:647-649.
Gross, M., G. R. Seage, K. H. Mayer, R. S. Goldstein, E. Losina, and C. Wold. 1996. Interest among gay/bisexual men in greater Boston in participating in clinical trials of preventive HIV vaccines. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 12:406-412.
Koblin, B. A., P. Heagerty, A. Sheon, S. Buchbinder, C. Celum, J. M. Douglas, M. Gross, M. Marmor, K. Mayer, D. Metzger, and G. Seage. 1998. Readiness of high-risk populations in the HIV Network for Prevention Trials to participate in HIV vaccine efficacy trials in the United States. AIDS 12:785-793.
Lee, D., B. S. Graham, Y. L. Chiu, P. B. Gilbert, M. J. McElrath, R. B. Belshe, S. P. Buchbinder, H. W. Sheppard, B. A. Koblin, K. H. Mayer, M. C. Keefer, M. J. Mulligan, and C. L. Celum. 2004. Breakthrough infections during phase 1 and 2 prime-boost HIV-1 vaccine trials with canarypox vectors (ALVAC) and booster dose of recombinant gp120 or gp160. J. Infect. Dis. 190:903-907.
Pitisuttithum, P., S. Nitayaphan, P. Thongcharoen, C. Khamboonruang, J. Kim, M. de Souza, T. Chuenchitra, R. P. Garner, D. Thapinta, V. Polonis, S. Ratto-Kim, P. Chanbancherd, J. Chiu, D. L. Birx, A. M. Duliege, J. G. McNeil, and A. E. Brown. 2003. Safety and immunogenicity of combinations of recombinant subtype E and B human immunodeficiency virus type 1 envelope glycoprotein 120 vaccines in healthy Thai adults. J. Infect. Dis. 188:219-227.
Schwartz, D. H., A. Mazumdar, S. Winston, and S. Harkonen. 1995. Utility of various commercially available human immunodeficiency virus (HIV) antibody diagnostic kits for use in conjunction with efficacy trials of HIV type 1 vaccines. Clin. Diagn. Lab. Immunol. 2:268-271.
Sheon, A. R., L. Wagner, M. J. McElrath, M. C. Keefer, E. Zimmerman, H. Israel, D. Berger, and P. Fast. 1998. Preventing discrimination against volunteers in prophylactic HIV vaccine trials: lessons from a phase II trial. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 19:519-526.(Surender Khurana, James N)