Downloaded from at uspto on August 28, 2013

Transcription

1 340 Immunization with Envelope Subunit Vaccine Products Elicits Neutralizing Antibodies against Laboratory-Adapted but Not Primary Isolates of Human Immunodeficiency Virus Type 1 John R. Mascola, Stuart W. Snyder, Owen S. Weislow, Seifu M. Belay, Robert B. Belshe, David H. Schwartz, Mary Lou Clements, Raphael Dolin, Barney S. Graham, Geoffrey J. Gorse, Michael C. Keefer, M. Juliana McElrath, Mary Clare Walker, Kenneth F. Wagner, John G. McNeil, Francine E. McCutchan, and Donald S. Burke for the National Institute of Allergy and Infectious Diseases AIDS Vaccine Evaluation Group Division ofretrovirology, Walter Reed Army Institute ofresearch, Henry M. Jackson Foundation for the Advancement ofmilitary Medicine. and SRA Technologies. Inc., Rockville, Department of Infectious Diseases, Naval Medical Research Institute. and National Institute ofallergy and Infectious Diseases, National Institutes of Health, Bethesda, and Johns Hopkins School ofhygiene and Public Health and School ofmedicine, Baltimore, Maryland; Department of Medicine, Saint Louis University School ofmedicine, St. Louis, Missouri; Department ofmedicine, University ofrochester School of Medicine and Dentistry, New York, New York; Department ofmedicine. Vanderbilt University School ofmedicine, Nashville, Tennessee; Department of Medicine, University of Washington School ofmedicine. Seattle. Washington Phase I studies of volunteersnot infectedwith human immunodeficiency virus type 1(HIV-1)have shown that immunization with envelopesubunit vaccine products elicits antibodies that neutralize laboratory-adapted (prototype) HIV-l strains in vitro. Prototype strains are adapted to grow in continuous (neoplastic)celllines and are more susceptibleto neutralization than are primary isolates cultured in human peripheral blood mononuclear cells. In this study, 50 sera from nine phase I vaccinetrials and 16 from HIV-1-infected personswere evaluated for neutralizing antibody activity against 3 laboratory-adapted and 5 primary HIV-l isolates. Of 50 sera, 49 neutralized at least 1 of the prototype strains; however, none displayed neutralizing activity against primary isolatesof HIV 1. Serum from most HIV-l-infected persons neutralized both laboratory-adapted and primary HIV 1 isolates. These data demonstrate a qualitative, or large quantitative, differencein the neutralizing antibody response induced by envelope subunit vaccination and natural HIV-1 infection. In several viral diseases, neutralizing antibodies are thought to be an important component ofa protective immune response [1-8]. Initial preclinical human immunodeficiency virus type 1 (HIY-1) vaccine studies demonstrated that antisera elicited by immunization with HIY-l envelope glycoproteins gpl20 or gp160 or by peptides derived from the third variable region (V3) ofgp120 could neutralize HIY-1 in vitro [9-11]. In addition, chimpanzees vaccinated with recombinant envelope glycoprotein (rgp) vaccine products or passively immunized with a neutralizing monoclonal antibody to the V3 region of gp120 Received 14 June 1995; revised 18 August Presented in part: conference on AIDS Vaccine Development, 6th Annual National Cooperative Vaccine Development Groups for AIDS meeting, Alexandria, Virginia, 31 October These studies were approved by the institutional review boards at each of the universities conducting the phase I clinical trials. Written informed consent was obtained from all patients according to guidelines from the US Department of Health and Human Services and those of the authors' institutions. The views presented in this article are those of the authors and do not purport to reflect those of the US Departments of the Army, Navy, or Defense. Financial support: US Army Medical Research and Materiel Command; National Institutes of Health (AI-05061, , , , , and ). Reprints or correspondence: Dr. John R. Mascola, Division ofretrovirology, Walter Reed Army Institute of Research, 13 Taft Ct., Rockville, MD The Journal of Infectious Diseases 1996;173: by The Universityof Chicago. All rights reserved / $01.00 were protected from infection when challenged with HIY-l [12-14]. The current focus by vaccine manufacturers on envelope subunit vaccines is due, in large part, to the successful chimpanzee experiments and to the ability of subunit vaccines to induce high levels of neutralizing antibodies [15]. In most in vitro HIV-1 neutralization studies and in the chimpanzee protection experiments described above, neutralization assays used laboratory-adapted strains of HIY-1 that have been cultured and propagated in human neoplastic cell lines. In contrast, primary HIY-1 isolates are cultured only in uninfected human peripheral blood mononuclear cells (PBMC). HIY-1 enters cells via binding to the CD4 receptor, and a soluble form of CD4 (scd4) can neutralize HIV-l in vitro. In 1990, it was demonstrated that neutralization ofhiy-1 primary isolates required ~ 100- to 1000-fold higher concentrations of scd4 than did neutralization of laboratory-adapted (prototype) strains [16, 17]. Subsequent studies have shown that, compared with primary HIY-1 isolates, laboratory-adapted strains are markedly more sensitive to neutralization by polyclonal sera from HIY-1-infected persons and by neutralizing monoclonal antibodies [18-21]. The underlying mechanism for the relative neutralization sensitivity of prototype strains is not well established [22-25], although a similar phenomenon has been described for other retroviruses [26, 27]. Numerous phase I studies using HIY-1-seronegative human volunteers have demonstrated that immunization with rgp160, rgp 120, and Y3 peptide products can elicit neutralizing antibod-

2 JID 1996; 173 (February) Candidate AIDS Vaccines: Neutralizing Antibody 341 ies to laboratory-adapted strains of HlV-I [28-35]. In these studies, primary viruses were not tested. Because of the established disparity in neutralization susceptibility, selected sera from nine phase I trials of mv-1 envelope subunit candidate vaccine products were evaluated for neutralizing antibody activity against a panel of 3 laboratory-adapted and 5 primary isolates of Hlv-I. Materials and Methods Subjects. vaccine preparations, and study protocols. Study volunteers were healthy adults between 18 and 60 years of age who were at low risk for HIV infection and were negative for HIV I by ELISA. Volunteers were enrolled into one of nine multicenter safety and immunogenicity trials of candidate HIV-1 envelopederived subunit vaccines conducted by the National Institute of Allergy and Infectious Diseases-sponsoredAIDS Vaccine Evaluation Group (AVEG); the trials have been described in detail elsewhere [29, 31, 33, 34, 36-40]. Table I summarizes each candidate vaccine, including information on manufacturers, virus strain, dose, adjuvant, and immunization regimen. Sera were collected and frozen throughout the trials for HIV-I-specific serologic assays, including neutralizing antibody assays. Sera were tested at the AVEG central immunology laboratory at Duke University for neutralizing antibody activity against the homologous vaccine strain. Several sera demonstrating activity from each trial were selected for further evaluation. A blinded panel of 51 sera, including I negative control from a placebo recipient, was sent to the Table 1. Description of mv-1 candidate vaccines and immunization regimens. Walter Reed Army Institute of Research laboratories for evaluation of neutralizing antibody activity against a panel of primary and laboratory-adapted HIV-1 isolates. In addition, sera and heparinized whole blood were obtained from a convenience sample of 16 clinically asymptomatic HIV I-seropositive volunteers attending the HIV clinic of the National Naval Medical Center (Bethesda, MD). All volunteers were HIV I seroconverters; the amount of time each had been infected with HIV-I was estimated as the midpoint between the last negative and first positive HIV-l serology. These volunteers and their sera were referred to as USI-USI6. With one exception (USI5), subjects had CD4 T lymphocyte counts>500/f-tl at the time of sample collection. The 5 primary viruses (US1-US5) used in neutralization assays were isolated from volunteers US1-US5. The sera obtained from US1-US16 were not blinded. Virus isolation. Virus was isolated by cocultivation of ficollseparated PBMC from HfV-1-infected subjects with phytohemagglutinin (PHA)-stimulated uninfected donor PBMC as previously described [41]. For expansion to high-titered virus stocks, culture supernatant from the initial virus isolation was used to infect 2 X 10 7 fresh PHA-stimulated donor PBMe. Cultures were followed using an EIA (Coulter, Hialeah, FL) for production of HIV-I p24 antigen, and supernatant collected at a single time point (between days 6 and 10) was stored in l-rnl. aliquots in the vapor phase of liquid nitrogen. None of these primary isolates were passaged through neoplastic cell lines. To prepare laboratory-adapted HIV strains (MN, IlIB, and SF2), virus supernatant from chronically infected H9 or HUT-78 cells was used to infect human PBMC as described above. Serum code Adjuvant or Dose schedule no. Vaccine product Dose Source of antigen delivery system (months) Time point (weeks)' 1-6 MicraGeneSys rgp [60 IlIB 640 IJ.g Insect cells Alum 0, 1,6, after 3rd or 4th immunization 7-11 Immuno rgp 160 IlIB 50 IJ.g Vera cells Alum/DOC O. 1,6, 12 4 after 3rd immunization Genentech rgp 120 liib 300 IJ.g CHO cells Alum 0,1,6,12 2 after3rdimmunization Genentech rgp 120 liib 300 IJ.g CHO cells Alum 0, 1.8, 13, 19 2 after 5th immunization + MN' Genentech rgp 120 IlIB 300 IJ.g CHO cells Alum 0, 1,6, 12 2 after4th immunization +MW Biocine Env IJ.g Yeast MF59/MTp PE 0, 1,6 2-4 after 2nd or 3rd immunization' Biocine rgp 120 SF2 15 or 50 IJ.g CHO cells MF59/MTP PE 0, 1,6 4 after 3rd immunization Genentech rgp 120 I1lB 300 IJ.g CHO cells Alum 0, 1,6, 12 2 after 3rd immunization + MN: , Bristol-Myers Squibb HIVAC 2 HIVAC, Vaccinia, CHO cells Vaccinia virus, 0,4,8 4 after 3rd immunization rgp 120 boosti 1 rgp 120 alum, or MF United Biomedical octameric 20, 100, or Synthetic MAPS 0, 1,6 4 after3rd immunization V3 peptide (MN) 500 IJ.g 118 Alum 0, 1,6 2 after3rdimmunization NOTE. Vero cells, monkey kidney cells; DOC, deoxycholate; MF59, microfluidized oil-in-water emulsion; MTP-PE, muramyl tripeptide-phosphatidylethanolamine; MAPS, multiple antigen peptides presented on oligo-lysine backbone; rgp, recombinant envelope glycoprotein. Supplier locations: MicroGeneSys, West Haven, CT; Genentech, South San Francisco; Biocine, Emeryville, CA; Bristol-Myers Squibb, Evansville, IN: United Biomedical, Hauppauge, NY; Immuno, Vienna. Time sera obtained for measurement of neutralizing antibody. t 3 immunizations with rgpl20 IIIB followed by 2 with rgpl20 MN. t 300 f.lg each, given together. Samples 104 and 105 were obtained I month after 2nd immunization. imvac = recombinant vaccinia virus expressing rgpl60 (IIIB strain). Boosts: 119, rgp120 SF2; 120 and 123, rgp120 IIIB.

3 342 Mascola et al. lid 1996; 173 (February) Standard neutralization assay. Unless otherwise specified, serum neutralizing antibody activity was assessed as previously described, with minor modifications [20]. Virus growth kinetics and TCID so were determined within the assay format. The same donor cryopreserved PBMC were used in the titration and neutralization assays. To facilitate evaluation of a large number of sera, each serum sample (heat-inactivated at 56 C for 45 min) was initially assayed at a single dilution of I:5, and selected sera were titered by serial 5-fold dilutions. In brief, 25 j-ll of test serum per well was aliquoted in quadruplicate wells of a 96-well microtiter plate. Sextuplicate wells containing pooled, heat-inactivated normal human serum (NHS) served as the control for baseline virus growth. An equal volume of virus stock (25 j-ll), representing 100 TCIDso, was added to each well. After 30 min at 37 C, 2 X io' PHA-stimulated PBMC was added and incubated overnight at 37 C. Cells were then washed extensively to remove p24 antigen and plasma anti-p24 antibody and transferred to a 96-well microtiter plate with culture media containing 10% interleukin-2. Inhibition ofpbmc infection was assessed by quantitative p24 measurement of cell supernatants during the logarithmic growth phase (days 4-6 for viruses in this study). Average p24 antigen output in sextuplicate control (NHS) wells was usually ng/ml (coefficients of variation range, 7%-44%; average, 18%). For sera tested at a single dilution (1:5), results were reported as percent neutralization (decrease in p24 level in test serum wells compared with NHS control wells). The final serum dilution was defined as the dilution of serum in the presence of virus (i.e., 1:10) and before addition of PBMC. To maximize sensitivity, a 50% or greater reduction in p24 antigen production was defined as a positive result. Infectivity reduction assay. Selected sera were also evaluated in a neutralization assay measuring effect of serum on virus endpoint titer. Virus TCID so was determined in a microtiter plate assay using PHA-stimulated PBMC as target cells. Serial 5-fold dilutions of virus stock were preincubated in quadruplicate wells with test sera or NHS diluted 1:10. After 30 min at 37 C, los PHA-stimulated PBMC were added to each well. Cells were washed after 24 h and maintained with culture media containing 10% interleukin 2. Extracellular p24 antigen was measured on day 8, and wells containing>100 pg/ml were scored as positive. TCID so and 95% confidence intervals were calculated with computer software using the statistical method of Spearman-Karber [42], and the TCID so values for virus preincubated with NHS and test sera were compared. Prior experiments with multiple replicates ofnhs demonstrated that TCID so values usually were within a 5-fold range, and for TCID so values differing by IO-fold, calculated 95% confidence intervals did not overlap. Thus, a TCID so reduction of ;do-fold was defined as evidence of serum-mediated neutralization. Most study samples had been unblinded before the infectivity reduction assays were done. MT-2 phenotype assay. The ability of primary viruses US1 US5 to induce syncytia in the MT-2 cell line was evaluated according to the AIDS Clinical Trials Group consensus protocol [43]. Syncytium-inducing viruses had ;;.3 multinucleated giant cells/well. Results A total of50 blinded serum samples from candidate vaccine recipients was evaluated for neutralizing antibody activity against 5 primary and 3 laboratory-adapted strains of HIV-I. For comparison, 16 serum samples from HIV-I-infected persons were assayed against the same viruses. Vaccinee sera were obtained from nine phase I clinical trials of seven different HIV-1 subunit vaccine products and one recombinant virus (vaccinia-gpi60). Vaccine product, dose, adjuvant, and immunization regimen are summarized in table 1. Sera selected for the neutralization assays were from time points after the third or fourth immunization, except for the Biocine Env 2-3 product, for which 2 of 5 serum samples were collected after the second immunization time point. Table 2 summarizes the neutralizing antibody activity of vaccinee sera against 5 primary HIV-1 isolates and 3 laboratory-adapted strains. All primary viruses were previously characterized as HIV-1 genetic subtype B by gp160 envelope gene DNA sequence (US1-US4) [20] or by heteroduplex mobility assay (US5, data not shown) [44]. The laboratory-adapted strains correspond to the 3 vaccine strains used to produce the subunit vaccine products. The results represent an average of two independent experiments (in a few cases, a third assay was done to resolve minor discrepancies). For most of the vaccine products studied, the magnitude and breadthof serum neutralizing antibodies against laboratory-adaptedstrains ofhiv-l have been summarized [31-37, 39, 40]. The current study was designed primarily to evaluate neutralizing antibody activity against primary HIV-1 isolates. Results in table 2 confirm that immunization with each of the vaccine products elicited serum neutralizing antibodies against the homologous HIV-l vaccine strain. Of 50 sera, 49 neutralized at least 1 of the laboratory-adapted HIV-1 strains and in many cases, sera also neutralized heterologous laboratory-adapted strains. At the 1:10 dilution, some vaccinee sera essentially prevented infection ofpbmc by 100 TCID so ofthe homologous virus (>99% neutralization; figure 1). Several of these strong neutralizing sera were titered by serial 5-fold dilutions against the 3 laboratory-adapted strains and were compared with serum from an HIV-I-infected subject (US2) that has strong neutralizing activity against primary HIV-I isolates [20]. Neutralization curves of serum from US2 and selected vaccine recipients were similar against prototype strains, especially MN and SF2 (figure 1). However, despite levels of neutralizing antibody against laboratory-adapted strains that were occasionally similar to those induced by natural HIV-1 infection, none of the vaccinee sera demonstrated neutralizing activity against any of the 5 primary isolates of HIV-1 (table 2, figure 1). Thus, antisera elicited by all ofthe vaccine products tested could neutralize laboratory-adapted but not primary isolates ofhiv-i. The MT-2 cell phenotype of the primary viruses did not affect this result. Of the 5 viruses, 2 (USI and US2) were syncytia-inducing and 3 were non-syncytia-inducing in the MT-2 phenotype assay. In contrast to sera from the vaccine recipients, sera from clinically asymptomatic HIV-1-infected persons could often significantly neutralize primary isolates (table 3, figure 1). At

4 JID 1996; 173 (February) Candidate AIDS Vaccines: Neutralizing Antibody 343 Table 2. HIV-I neutralizing antibody activity in sera from vaccine recipients. Neutralizing antibody activity Laboratory strains Primary isolates Serum code no. Vaccine product MN SF2 IlIB USI US2 US3 US4 US5 1-6 MicroGeneSys rgp 160 IlIB 016 «50) 0/6 «50) 5/6 (57) 0/5 0/5 0/5 0/5 0/ Immuno rgp 160 IIIB 0/5 «50) 1/5 «50) 5/5 (80) 0/5 0/5 0/5 0/5 0/ Genentech rgp 120 IlIB 3/5 (53) 4/5 (61) 5/5 (83) 0/5 0/5 0/5 0/5 0/ Genentech rgp 120 IIIB + MN 5/5 (93) 5/5 (83) 3/5 (55) 0/5 0/5 0/5 0/5 0/5 22~26 Genentech rgp 120 IlIB + MN 5/5 (96) 5/5 (92) 5/5 (80) 0/5 0/5 0/5 0/5 0/5 101~1O5 Biocine Env 2-3, SF2 1/6 «50) 616 (87) 016 «50) 0/5 0/5 0/5 0/5 0/ Biocine rgp 120 SF2 6/6 (82) 6/6 (92) 2/6 «50) / Genentech rgp 120 IlIB + MN 5/5 (91) 5/5 (95) 4/5 (69) 0/5 0/5 0/5 0/5 0/5 119, 120, 123 Bristol-Myers Squibb HIVAC 3/3 (78) 1/3 (54) 3/3 (98) 0/3 0/3 0/3 0/3 0/3 + rgpl20 boost" United Biomedical octameric 5/5 (95) 0/5 «50) 0/5 «50) 0/5 0/5 0/5 0/5 0/5 V3 (MN) (alum) Oil 011 Oil Oil Oil Oil NOTE. Data are no. positive/no. tested (mean % neutralization, which is arithmetic mean of sera tested against each virus). All sera were tested at I: 10 dilution; positive result defined as ~50% reduction in p24 antigen production (results are average of 2 or 3 independent experiments). See table I for supplier locations. * HIVAC = recombinant vaccinia virus expressing rgpl60 (IlIB strain). Boosts: 119, rgpl20 SF2; 120 and 123, rgpl20 IlIB. a 1:10 dilution, neutralization values for the 16 sera against the 5 primary isolates ranged from 4% to 99%, and each of the 5 viruses was strongly neutralized (>90%) by at least several sera. Ofnote, the average neutralization for the 16 sera against each of the 3 laboratory-adapted strains ranged from 93% to 98% compared with 64%-78% against the 5 primary viruses. These data are consistent with those from earlier reports showing primary HIV-1 isolates to be less susceptible than laboratory strains to neutralizing antibody yet fully susceptible to neutralization by sera from some HIV-1-infected persons [19, 20, 45]. Primary isolates are also relatively resistant to neutralization by scd4, although this resistance can be overcome by prolonged preincubation of virus and scd4 at 4 C [46,47]. Thus, using primary isolates US1 and US2 and 5 vaccinee sera with strong neutralizing activity against laboratory strains (samples 10, 22, 24, 108, and 110), sera and virus were preincubated for 8 h at 4 C prior to addition of cells. Using these same 2 viruses and 5 sera, several other experimental modifications of the standard neutralization assay were done. To test for a possible prozone effect, vaccinee sera were evaluated undiluted and diluted 1:10, 1:100, and 1:1000. To assess the influence of complement proteins, aliquots of these sera, which had never been heat-inactivated, were again tested at dilutions of 1:10, 1:100, and 1:1000. Results of each of these experiments were similar: Viruses US2 and US4 were neutralized with antisera from HIV-1-infected subjects but not with antisera from envelope subunit vaccine recipients (data not shown). In the previous experiments, neutralization assays were done with a high (100 TCID so ) virus inoculum. To assess the possibility that vaccinee sera would demonstrate neutralizing activity at a lower virus inoculum, selected sera were studied in an infectivity reduction assay. As described in Methods, a TCID so reduction of lo-fold or greater by test sera (compared with NHS) was defined as evidence of serum-mediated neutralization. Two strong neutralizing serum samples from HIV-1 infected volunteers and 8 selected samples from vaccine recipients were tested against viruses US1 and US2. The vaccinee serum samples each strongly neutralized the homologous vaccine strain (table 4). Virus TCID so was not significantly altered by preincubation with sera from vaccine recipients. In contrast, serum samples from HIV-l-infected subjects US12 and US16 reduced virus TCID so by 17- to 84-fold. For vaccinee serum sample 119, a repeat experiment using virus USI again showed a nonsignificant reduction (2-fold) in virus titer compared with that in the control. Discussion Our data confirm that immunization with HIV-1 envelope subunit candidate vaccines can elicit high levels of functional (neutralizing) antibodies against laboratory-adapted strains of HIV-1. In a few selected cases, the level was similar in magnitude to that found in natural infection (figure 1). Yet, none of the sera tested had detectable neutralizing activity against 5 primary HIV-1 isolates. This result was obtained in two different neutralization assays and persisted despite attempts to increase assay sensitivity. Others have similarly reported that sera from HIV-1 subunit vaccine recipients displayed neutralizing activity against prototype but not primary isolates of HIV-l [24, 48, 49]. The current study was not designed to compare the magnitude ofneutralization elicited by each product against

5 344 Mascola et al. JID 1996;173 (February)... = = 100 HIV MN 100 HIV IIIB.:: !:: -; := ell Ẓ..... =ell CJ ell = Serum Dilution Serum Dilution 100 HIV SF2 100 HIV US4 Figure 1. Comparison of neutralization curves for several se- = lected vaccinee sera and serum.::... sample from HIV-1~ infected.!:: = subject (US2) with strong neu- -; tralizing activity. Sera were run... := against HIV-1 laboratory-adapted ell strains MN, IIIB, and SF2 and HIV-l primary isolate US4. Ẓ..... ell =... =ell CJ Serum Dilution Serum Dilution HIY-I infected (US2) --<>-- Immuno AG rgpl60 IIIB; #8 and displayed 90%-99% neutralization at a 1:10 dilution (table 3). This contrasts with the inability of even undiluted vaccinee sera to inhibit p24 antigen production by 50%. It is possible that vaccinee sera contain low levels of primary isolate neutralizing antibodies that were undetectable in our assays. However, if the difference in the neutralizing antibody response induced by subunit vaccination and natural infection is quantitatively different, it appears to be large in magnitude. The well-described disparity in neutralization sensitivity between laboratory-adapted and primary isolates suggests fundamental differences in neutralization epitopes or their accessibil- V Genentech rgpl20 IIIB; #16 l:j. Genentech rgpl20 IIIB+MN; # Biocine rgpl20 SF2; #112 laboratory-adapted strains; thus, for most sera, neutralization ofprototype strains was assessed at only a 1:10 serum dilution, and conclusions based on the percent neutralization at a single dilution are limited. As previously mentioned, primary isolates are markedly less sensitive to neutralization by scd4 and polyclonal antibody than are laboratory-adapted strains. Nonetheless, each of the 5 primary isolates evaluated was neutralized by some of the sera from HlV-1-infected subjects in our panel. Several of the stronger neutralizing sera exhibited 50% titers in the range of 1:100-1:800 against primary isolates [20] (data not shown)

6 JID 1996; 173 (February) Candidate AIDS Vaccines: Neutralizing Antibody 345 Table 3. Neutralization of primary isolates by sera from mv-1-infected persons. Primary virus isolates Time Average,' Serum code no. infected* USI US2 US3 US4 US5 USI-US5 USI 24 68' US US t US US t 94 US US US US USIO USll US US US US US NOTE. US1-US 16 are 16 clinically asymptomatic H1V-I-infected persons. Values represent % reduction in p24 expression by phytohemagglutinin-stimulated peripheral blood mononuclear cells (as described in Materials and Methods). Strong neutralizing activity (=290% neutralization) is indicated in bold. * Estimated no. of months subject infected (as described in Materials and Methods). t Arithmetic mean of % neutralization values for each serum against 5 viruses. 1 Neutralization with autologous serum obtained at same time virus was isolated. This result was omitted from calculated average % neutralization value for each serum. ity [23, 25, 50, 51]. It is possible that immunization with current HIV-1 envelope subunit vaccine products does not elicit antibodies to epitopes that are essential for the neutralization of primary isolates. Others have demonstrated that the V3 region Table 4. HIV-I infectivity reduction assay results. of gpl20 is the principal neutralizing determinant for laboratory-adapted strains [11, 32, 52]. In serum from envelope subunit vaccine recipients, the presence of fairly high levels of neutralizing antibodies directed against the V3 region [32, 53, USI virus US2 virus % neutralization Serum vs. (vaccine Infectivity Infectivity code no. Vaccine product strain)* TClO,o reduction rem; reduction NHS pool II Immuno rgp 160 (IlIB) 90 (liib) Genentech rgp 120 (IIIB + MN) 86 (IlIB), 97 (MN) 3125 I Genentech rgp 120 (liib + MN) 86 (IlIB), 96 (MN) Biocine rgp 120 SF2 82 (SF2) Biocine rgp 120 SF2 99 (SF2) Genentech rgp 120 (IIIB + MN) 82 (liib), 98 (MN) I 119 Bristol-Myers Squibb HIVAC (IlIB) 96 (IlIB), 77 (SF2) I + rgp120 SF2 128 United Biomedical octameric V3 94 (MN) peptide (MN) (alum) 14(MN) I USl2 HIV-l infected 98 (MN) USI6 HIV-l infected 99 (MN) NOTE. End-point titer (TClO so) for virus preincubated with pooled normal human sera (NHS) was compared with TCIDso for virus preincubated with test sera. Infectivity reduction is virus TClO,o in NHS/TCIDso in test serum. Data are results from 1 set of experiments. See table 1 for supplier locations. * % reduction in p24 antigen at I: 10 serum dilution (as described in Materials and Methods for standard neutralization assay).

7 346 Mascola et al. JID 1996; 173 (February) 54] suggests that either the V3 epitope of primary isolates is inaccessible or that V3 antibodies alone are insufficient for neutralization ofprimary isolates. Thus, studies that define the epitope or epitopes responsible for neutralization of primary isolates may be an essential part ofthe rational design offuture vaccine immunogens. Recent data showing that some human monoclonal antibodies can neutralize primary isolates ofhiv 1 should also aid in the development of vaccines [55, 56]. Of note, there were considerable differences in neutralizing activity among the 16 sera from asymptomatic HIV-1-infected volunteers. The 3 sera that exhibited an average of <50% neutralization against the 5 primary isolates were from persons infected for ~18 months (US6, US7, and USIO; table 3). In contrast, the 5 sera demonstrating 90% or greater neutralization against the primary isolates were from persons infected for ~5-7 years (US2, US4, US5, USl2, and USI6). While this study was not designed to correlate time of HIV-1 infection with serum neutralizing antibody activity, these data suggest a general trend of low neutralizing antibody levels early in HIV infection, with an increase in magnitude and breadth over several years in asymptomatic persons. Data in table 3 indicate that most of the sera from HIV-I-infected subjects had 50% neutralizing activity against the 5 primary isolates. However, if a more stringent criterion of 90% neutralization were used, only about one-third of the sera would be active against each virus. This is consistent with the results of others, which indicate that not all sera from HIV-1-infected persons have strong neutralizing activity [19, 21, 45,57]. Our data do not exclude the possibility that vaccinee sera contain low levels of neutralizing antibody against primary HIV-1 isolates. Assays for functional antibody may not be particularly sensitive [58-61], and it is possible that serum neutralizing activity was below the level of detection in our assays. Our limited data showing that neutralizing activity of sera is weak during the first 18 months ofinfection are consistent with those from several prior studies showing that neutralizing antibodies are often not detectable for weeks to months after acute HIV-I infection [57, 62, 63]. In contrast, gp120 binding antibodies can be detected within several days of clinical presentation with acute HIV infection [63]. While the cellular arm of the immune system may be completely responsible for abrogating the high-level plasma viremia ofacute infection [64], it is equally possible that undetectable levels of functional antibody play a role. As in primary infection, the absence of measurable neutralizing antibody does not prove that the vaccinee sera are completely deficient in antibodies with functional activity against primary viruses. Since the magnitude and breadth ofthe neutralizing antibody response seem to increase over time in natural HIV-I infection, it is feasible that a detectable neutralizing antibody response could be elicited in vaccinees with booster immunizations over time or with better adjuvants and immunization regimens. As currently formulated, HIV-I envelope subunit immunogens clearly elicit a lesser magnitude and breadth of humoral immunity than those found in natural HIV-I infection. In the absence of established correlates of protection for HIV-1, it is reasonable and rational to attempt to improve the functional humoral responses induced by these vaccine products. As we do not know what aspects of the humoral and cellular immune response will confer protection against HIV-I infection or disease, the inability to detect neutralizing antibodies to primary viruses in this in vitro assay system should not be presumed to predict the ultimate success or failure of envelope subunit products in human phase III clinical trials. Indeed, there are several examples of effective retroviral vaccines that do not elicit detectable levels of neutralizing antibodies [65-67]. The validation of any postulated in vitro correlates of protection for HIV-l must await in vivo studies showing vaccine efficacy. Acknowledgments We thank Thomas Mathews (Duke University), Deborah O'Neill, Leah Johnston, and staff (NationalNaval Medical Center HIV Clinic), Douglas Mayers, Deborah Birx, Thomas VanCott, and AndrewArtenstein (WalterReed Army Institute of Research), Mark Louder and Sonja Surman(Henry M. Jackson Foundation), and Meredith Yeager (SRA Technologies). References 1. Hammon W McD, Coriel LL, Wehrle PF. Evaluation of Red Cross gamma globulin as a prophylactic agent for poliomyelitis. lama 1953; 151: Halstead SB. Pathogenesis of dengue: challenges to molecular biology. Science 1988;239: de Noronha F, Schafer W, Essex M, Bolognesi DP. Influence of antisera to oncornavirus glycoprotein (gp71) on infections of cats with feline leukemia virus. Virology 1978;85: Schwartz H, Fischinger PJ, Ihle IN, et al. Properties of mouse leukemia viruses, XVI: suppression of spontaneous fatal leukemias in AKR mice by treatment with broadly reacting antibody against the viral glycoprotein g71. Virology 1979;93: Chen RT, Markowitz LE, Albrecht P, et al. Measles antibody: reevaluation of protective titers. J Infect Dis 1990; 162: Suss J, Sinnecker H. Immune reactions against rabies viruses-infection and vaccination. Exp Pathol 1991;42: Smithburn KC, Mahaffy AF. Immunization against yellow fever. Studies on the time of development and the duration of induced immunity. Am 1 Trop Med Hyg 1945;25: Frank AL, Taber LH, Glezen PW, Geiger EA, McIlwain S, Paredes A. Influenza B virus infections in the community and the family: the epidemics of and in Houston, Texas. Am J EpidemioI1983;118: Lasky LA, Groopman JE, Fennie CW, et al. Neutralization of the AIDS retrovirus by antibodies to a recombinant envelope glycoprotein. Science 1986;233: Steimer KS, Haigwood NL. Immunization of primates with native, recombinant HIV-SF2 gpl20 generates broadly effective neutralizing antibodies directed to conformational epitopes. AIDS Res Hum Retroviruses 1992;8:1391. II. Javaherian K, Langlois AJ, McDanal C, et al. Principal neutralizing domain of the human immunodeficiency virus type I envelope protein. Proc Natl Acad Sci USA 1989;86:

8 JID 1996; 173 (February) Candidate AIDS Vaccines: Neutralizing Antibody Berman PW, Gregory Tf, Riddle L, et al. Protection of chimpanzees from infection by HIV-I after vaccination with recombinant glycoprotein gpl20 but not gp160. Nature 1990;345: Girard M, Kieny MP, Pinter A, et al. Immunization of chimpanzees confers protection against challenge with human immunodeficiency virus. Proc Nat! Acad Sci USA 1991;88: Emini EA, Schleif WA, Nunberg JH, et al. Prevention of HIV-I infection in chimpanzees by gpl20 V3 domain-specific monoclonal antibody. Nature 1992; 355: Mascola JR, McNeil JG, Burke DS. AIDS vaccines: are we ready for human efficacy trials? [editorial comment]. JAMA 1994;272: Gomatos PJ, Stamatos NM, Gendelman HE, et al. Relative inefficiency of soluble recombinant CD4 for inhibition of infection by monocytetropic HIV in monocytes and T cells. J Immunol 1990; 144: Daar ES, Li XL, Moudgil T, Ho DD. High concentrations of recombinant soluble CD4 are required to neutralize primary human immunodeficiency virus type 1 isolates. Proc Nat! Acad Sci USA 1990; 87: Robb ML, Polonis V, Vahey M, et al. HIV neutralization assay using polymerase chain reaction-derived molecularsignals. J Acquir Immune Defic Syndr 1992;5: Sawyer LSW, Wrin MT, Crawford-Miksza L, et al. Neutralizationsensitivity of human immunodeficiency virus type I is determined in part by the cell in which the virus is propagated. J Virol 1994;68: Mascola JR, Louwagie J, McCutchan FE, et al. Two antigenically distinct subtypes of human immunodeficiency virus type I: viral genotype predicts neutralization serotype. J Infect Dis 1994; 169: Moore JP, Cao Y, Qing L, et al. Primary isolates of human immunodeficiency virus type I are relatively resistant to neutralization by monoclonal antibodies to gp120, and their neutralization is not predicted by studies with monomeric gp120. J ViroI1995;69: Ashkenazi A, Smith DH, Marsters SA, et al. Resistance of primary isolates of human immunodeficiency virus type I to soluble CD4 is independent of CD4-rgp120 binding affinity. Proc Natl Acad Sci USA 1991; 88: Moore JP, Burkly LC, Connor RI, et al. Adaptation oftwo primary human immunodeficiency virus type I isolates to growth in transformed T cell lines correlates with alterations in the responses of their envelope glycoproteins to soluble CD4. AIDS Res Hum Retroviruses 1993; 9: Wrin T, Loh TP, Vennari JC, Schuitemaker H, Nunberg JH. Adaptation to persistent growth in the H9 cell line renders a primary isolate of human immunodeficiency virus type 1 sensitive to neutralization by vaccinee sera. J Virol 1995; 69: Moore JP, McKeating JA, Huang Y, Ashkenazi A, Ho DD. Virions of primary human immunodeficiency virus type I isolates resistant to soluble CD4 (scd4) neutralization differ in scd4 binding and glycoprotein gpl20 retention from scd4-sensitive isolates. J ViroI1992;66: Baldinotti F, Matteucci D, Mazzetti P, et al. Serum neutralization offeline immunodeficiency virus is markedly dependent on passage history of the virus and host system. J ViroI1994;68: Cook RF, Berger SL, Rushlow KE, et al. Enhanced sensitivity to neutralizing antibodies in a variant of equine infectious anemia virus is linked to amino acid substitutions in the surface unit envelope glycoprotein. J Virol 1995; 69: Dolin R, Graham BS, Greenberg SB, et al. The safety and immunogenicity of a human immunodeficiency virus type I (HIV-I) recombinant gp160 candidate vaccine in humans. Ann Intern Med 1991; 114: Graham BS, Matthews TJ, Belshe RB, et al. Augmentation of human immunodeficiency virus type I neutralizing antibody by priming with gp 160 recombinant vaccinia and boosting with rgp 160 in vaccinia-naive adults. J Infect Dis 1993; 167: Kovacs JA, Vasudevachari MB, Easter M, et al. Induction of humoral and cell-mediated anti-hiv responses in HIV seronegative volunteers by immunization with recombinant gp160. J Clin Invest 1993;92: Belshe RB, Clements ML, Dolin R, et al. Safety and immunogenicity of a fully glycosylated recombinant gp 160 human immunodeficiency virus type I vaccine in subjects at low risk of infection. J Infect Dis 1993; 168: Montefiori DC, Graham BS, Zhou J, et al. V3-specificneutralizing antibodies in sera from HIV-I gpl60-immunized volunteers block virus fusion and act synergistically with human monoclonal antibody to the conformation-dependent CD4 binding site of gp120. J Clin Invest 1993; 92: Belshe RB, Graham BS, Keefer MC, et al. Neutralizing antibodies to HIV I in seronegative volunteers immunized with recombinant gpl20 from the MN strain of HIV-1. JAMA 1994;272: Schwartz DH, Gorse G, Clements ML, et al. Induction ofhiv- I-neutralising and syncytium-inhibiting antibodies in uninfected recipients ofhiv IIIlB rgpl20 subunit vaccine. Lancet 1993;342: Kahn 10, Sinangil F, Baenziger J, et al. Clinical and immunologic responses to human immunodeficiency virus (HIV) type ISF2 gp 120 subunit vaccine combined with MF59 adjuvant with or without muramyl tripeptide dipalmitoyl phosphatidylethanolamine in non-hiv-infected human volunteers. J Infect Dis 1994; 170: Keefer MC, Graham BS, Belshe RB, et al. Studies of high doses of a human immunodeficiency virus type I recombinant glycoprotein 160 candidate vaccine in HIV type I-seronegative humans. AIDS Res Hum Retroviruses 1994; 10: Gorse GJ, Keefer M, Weinhold K, et al. Phase I study of HIV-I MN octameric peptide vaccine [abstract PA0375]. In: Program and abstracts: Xth International Conference on AIDSlInternational Conference on STD (Yokohama, Japan). Tokyo: Japanese Foundation for AIDS Prevention, Walker MC, Fast PE. Clinical trials of candidate AIDS vaccines. AIDS 1994;8(suppll):S Keefer M, Graham B, McElrath J, Matthews T, Chernoff D, Dolin R. Phase I trial of an HIV-I vaccine candidate, Env 2-3, in combination with MTP-PE/MF59 [abstract PO-A ]. In: Program and abstracts: IXth International Conference on AIDS/IV STD World Congress (Berlin). London: Wellcome Foundation, Graham B, Keefer M, McElrath J, et al. Phase I trial of native HIV ISF-2 rgpl20 candidate vaccine. NIAID AIDS Vaccine Clinical Trials Network [abstract PO-A ]. In Program and abstracts: IXth International Conference on AIDS/IV STD World Congress (Berlin). London: Wellcome Foundation, Burke DS, Fowler AK, Redfield RR, et al. Isolation of HIV-I from the blood of seropositive adults: patient stage of illness and sample inoculum size are major determinants ofa positive culture. J Acquir Immune Defic Syndr 1990;3: National Center for Biotechnology Information, National Library ofmedicine, National Institutes of Health. Bethesda, Maryland: Spouge ID-50 software. Version 2.0, Japour AJ, Fiscus SA, Arduino lm, Mayers DL, Reichelderfer PS, Kuritzkes DR. Standardized microtiter assay for determination of syncytiuminducing phenotypes of clinical human immunodeficiency virus type I isolates. J Clin Microbiol 1994; 32: Delwart EL, Shpaer EG, Louwagie J, et al. Genetic relationships determined by a DNA heteroduplex mobility assay: analysis of HIV-I env genes. Science 1993;262: Cao Y, Limo Q, Zhang L, Safrit J, Ho DD. Virologic and immunologic characterization of long-term survivors of human immunodeficiency virus type I infection. N Engl J Med 1995;332: Moore JP, Klasse P1. Thermodynamic and kinetic analysis of scd4 binding to HIV-I virions and ofgpl20 dissociation. AIDS Res Hum Retroviruses 1992;8: O'Brien WA, Mao SH, Cao Y, Moore JP. Macrophage-tropic and T-cell line-adapted chimeric strains of human immunodeficiency virus type I differ in their susceptibilities to neutralization by soluble CD4 at different temperatures. J Virol 1994; 68:

9 348 Mascola et al. JID 1996; 173 (February) 48. Matthews T, McDanal C, Greenwell T, et al. Serological reactivity from HIV-I vaccine recipients participating in the AIDS vaccine clinical trials network [abstract]. AIDS Res Hum Retroviruses 1994; 10:S Steimer KS, Sinangil F, Kahn J, et al. Primary isolate neutralizing activity of human antibodies directed to recombinant native HIV-SF2 gpl20 [abstract]. AIDS Res Hum Retroviruses 1994; 1O:S VanCott TC, Polonis VR, Loomis LD, Michael NL, Redfield RR, Birx DL. Differential role of V3-specific antibodies in neutralization of primary versus laboratory-adapted HIV-I [abstract 1242]. In: Program and abstracts: 34th Interscience Conference on Antimicrobial Agents and Chemotherapy (Orlando). Washington DC: American Society for Microbiology, VanCott TC, Bethke FR, Burke DS, et al. Lack of induction of antibodies specific for conserved, discontinuous epitopes on HIV-I glycoprotein by candidate AIDS vaccines. J Immuno11995; 155: Prof)' AT, Salinas PA, Eckler LI, Dunlop NM, Nara PL, Putney SD. Epitopes recognized by the neutralizing antibodies of an HIV-I-infected individual. J Immunol 1990; 144: Stephens DM, Eichberg JW, Haigwood NL, Steimer KS, Davis D, Lachmann PJ. Antibodies are produced to the variable regions of the external envelope glycoprotein of human immunodeficiency virus type I in chimpanzees infected with the virus and baboons immunized with a candidate recombinant vaccine. J Gen Viroll992;73: Pincus SH, Messer KG, Schwartz DH, et al. Differences in the antibody response to human immunodeficiency virus-i envelope glycoprotein (gpi60) in infected laboratory workers and vaccinees. J Clin Invest 1993; 91: Conley AJ, Kessler JA II, Boots LJ, et al. Neutralization of divergent human immunodeficiency virus type I variants and primary isolates by IAM-41-2F5, an anti-gp41 human monoclonal antibody. Proc Nat! Acad Sci USA 1994;91: Burton DR, Pyati J, Koduri R, et al. Efficient neutralization of primary isolates of HIV-I by a recombinant human monoclonal antibody. Science 1994;266: Wrin T, Crawford L, Sawyer L, Weber P, Sheppard HW, Hanson CV. Neutralizing antibody responses to autologous and heterologous isolates of human immunodeficiency virus. J Acquir Immune Defic Syndr 1994;7: Dimmock NJ. Neutralization of animal viruses. CUIT Top Microbiol ImmunoI1993;183: Layne SP, Merges MJ, Dembo M, et al. Factors underlying spontaneous inactivation and susceptibility to neutralization of human immunodeficiency virus. Virology 1992; 189: Berman PW, Nakamura GR. Adhesion mediated by intercellular adhesion molecule I attenuates the potency of antibodies that block HlV-I gp160 dependent syncytium formation. AIDS Res Hum Retroviruses 1994; 10: Golding H, D'Souza MP, Bradac J, Mathieson B, Fast P. Neutralization of HIV-1. AIDS Res Hum Retroviruses 1994; 10: Albert J, Abrahamsson B, Nagy K, et al. Rapid development of isolatespecific neutralizing antibodies after primary HIV-I infection and consequent emergence of virus variants which resist neutralization by autologous sera. AIDS 1990;4: Moore JP, Cao Y, Ho DD, Koup RA. Development of the anti-gpl20 antibody response during seroconversion to human immunodeficiency virus type I. J ViroI1994;68: Koup RA, Safrit JT, Cao Y, et al. Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type I syndrome. J Virol 1994; 68: IsseI CJ, Horohov DW, Lea DF, et al. Efficacy of inactivated wholevirus and subunit vaccines in preventing infection and disease caused by equine infectious anemia virus. J Viro11992; 66: York SM, York CJ. Development of a whole killed feline leukemia virus vaccine. J Am Vet Med Assoc 1991;199: Hoover EA, Perigo NA, Quackenbush SL, et al. Protection against feline leukemia virus infection by use of an inactivated virus vaccine. J Am Vet Med Assoc 1991; 199: