Vaccine 18 (2000) 1793±1801
www.elsevier.com/locate/vaccine
Microparticles in MF59, a potent adjuvant combination for a recombinant protein vaccine against HIV-1 D.T. O'Hagan a,*, M. Ugozzoli a, J. Barackman a, M. Singh a, J. Kazzaz a, K. Higgins a, T.C. Vancott b, G. Ott a a Chiron Corporation, 4560 Horton Street, Emeryville, CA 94608, USA Henry M. Jackson Foundation, 13 Taft Court, Suite 200, Rockville, MD 20850, USA
b
Received 29 July 1999; accepted 15 October 1999
Abstract Novel adjuvant formulations involving PLG microparticles with entrapped recombinant protein antigens (env gp120 and p24 gag) from human immunode®ciency virus type-1 (HIV-1), dispersed in the emulsion adjuvant MF59 were evaluated as potential HIV-1 vaccine candidates in mice and baboons. In mice, the adjuvant combination induced signi®cantly enhanced antibody responses in comparison to either adjuvant used alone. In addition, the polylactide co-glycolide polymer (PLG) microparticles and MF59 combination induced CTL activity against HIV-1 p24 gag. In baboons, the adjuvant combination induced signi®cantly enhanced antibody titers after a single dose of gp120, but the responses were comparable to gp120 in MF59 alone after boosting. Both MF59+gp120 alone and PLG/gp120 in MF59 induced neutralizing antibodies against a T cell line-adapted (TCLA) strain and a primary isolate of HIV-1. In contrast to the observations with gp120, immunization in baboons with PLG/ p24 in MF59 induced signi®cantly enhanced antibody responses after boosting, in comparison to immunization with MF59 alone+p24. # 2000 Elsevier Science Ltd. All rights reserved.
1. Introduction AIDS, which is caused by infection with the human immunode®ciency virus type-1 (HIV-1), is a large and growing problem worldwide, causing enormous morbidity and mortality. Therefore, a safe and eective vaccine to reduce the transmission of HIV-1 infection or to prevent the disease progression are desperately needed, particularly in developing countries where >90% of infections occur [1]. Although it is not currently known which components of the virus need to be included to allow the development of an ecacious HIV-1 vaccine, and the immune correlates for protection have not yet been established, many vaccine candidates * Corresponding author. Tel.: +1-510-923-7662; fax: +1-510-9232586. E-mail address:
[email protected] (D.T. O'Hagan).
have already been extensively evaluated. The majority of vaccines evaluated have included some form of the HIV-1 envelope glycoprotein gp120 or gp160 [2]. These vaccines have induced serum neutralizing antibodies against homologous T-cell line adapted HIV-1 and in some cases heterologous HIV-1 strains [3]. However, the early enthusiasm engendered by the induction of HIV-1 neutralizing antibodies in humans soon waned when it became clear that the `principal neutralizing determinant', the V3 loop of gp120, was a poor target for neutralization of primary isolates of HIV-1 and none of the candidate vaccines elicited neutralizing activity against primary isolates [4]. In addition, env based vaccines did not induce signi®cant levels of HIV-1 speci®c CD8+ mediated T-lymphocyte mediated cytotoxicity (CTL). It has become increasingly clear that an eective HIV-1 vaccine will be required to induce broad immune responses, including both humoral and cell mediated immunity. The poten-
0264-410X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved. PII: S 0 2 6 4 - 4 1 0 X ( 9 9 ) 0 0 5 2 2 - 8
1794
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
tial importance of neutralizing antibodies to env proteins has been recently highlighted in studies in primates, which showed that neutralizing antibodies clear infectious virus from sera following virus challenge [5], and oer protection against challenge with a chimeric SHIV virus [6,7]. However, it should be noted that these studies were performed under `idealized' experimental situations, which are very dierent from the real situations in which infections occur in human subjects. In early studies, it was noted that lack of infection in HIV-1 exposed individuals was correlated with the presence of a CTL response in peripheral blood [8±10]. In addition, CTLs have also been associated with initial control of acute infection [11,12] and in the control of viremia in chronically infected individuals [13]. Alternative studies have shown that viral load was inversely correlated with HIV-1 speci®c CD4+ proliferation in infected individuals, further suggesting that T cells are responsible for controlling infection [14]. In a separate study, a correlation was also found between lack of infection with HIV-1 in active sexual partners of seropositive individuals, and the presence of proliferative T cell responses [15]. In experimental studies in primates, virus burden post-challenge with SIV has been shown to be inversely related to the precursor CTL frequency [16]. Hence accumulating evidence indicates an important role for cellular immunity, particularly CTLs, in controlling virus infection and spread. While CD4+ responses are necessary to help maintenance of both CTL and humoral responses, CD8+ T cells are directly cytolytic on infected cells and also release HIV-1 antiviral factors [17] and b-chemokines [18]. In previous studies in mice, PLG microparticles have been shown to be potent adjuvants for the induction of humoral immunity to entrapped antigens [19± 21]. Cleland et al. [22] prepared microparticles with entrapped gp120 and showed that the protein was unaected by the encapsulation process and was released intact from the microparticles in vitro. In addition, controlled release of the antigen in vivo was capable of inducing potent long term immune responses, including neutralizing antibodies [23]. Potent long term immunity to HIV-1 has also been induced with a peptide immunogen representing the V3 loop of gp120 in microparticles [24]. PLG microparticles with entrapped gp120 have also been shown to induce CTL activity in mice, following mucosal delivery [25]. In the current studies, microparticles were evaluated as an adjuvant for entrapped gp120 and p24 from HIV-1 in mice. In addition, the ability of an emulsion adjuvant, MF59, to further enhance immunity was evaluated. Following encouraging observations in mice, the adjuvanticity of the microparticle formulations of gp120 and p24 were evaluated in baboons.
2. Materials and methods 2.1. Antigens HIV-1SF2 rgp120 was produced in chinese hamster ovary cells and HIV-1SF2 p24 gag was produced in yeast cells at Chiron Corporation (Emeryville, CA). 2.2. Animals Female Balb/c mice, aged 6±8 weeks, were obtained from Charles River Laboratories (Wilmington, MA) and were used for immunization studies. Baboon studies were undertaken at Southwest Foundation for Biomedical research, San Antonio, TX. Young adult baboons were used (5±10 kg). 2.3. Adjuvant preparation and characterization Microparticles (mean size 1mm) with entrapped HIV-1 antigens were prepared essentially as previously described [26,27], using poly lactide co-glycolide polymers (PLG Ð ratio 50/50) obtained from Boehringer Ingelheim. Brie¯y, microparticles with a 1% w/w loading level were prepared by emulsifying 1 ml antigen solution (5 mg/ml of gp120) with 10 ml of a 5% w/v polymer solution in methylene chloride. The primary emulsion was homogenized for 3 min at 12,000 rpm and then 40 ml of a 10% w/v solution of polyvinyl alcohol in distilled water was added, and the multiple emulsion formed was homogenized at 12,000 rpm for 3 min. The w/o/w emulsion was then stirred at 1000 rpm on a magnetic stirrer overnight at room temperature. This resulted in complete evaporation of the solvent and formation of PLG microparticles. The resulting microparticles were washed twice with distilled water and freeze-dried. The size distribution of the microparticles was determined using a particle size analyzer (Malvern Instruments, UK). The rate of antigen release and the loading of antigen was determined following hydrolysis of the microparticles as previously described [28,29]. The antigen load of microparticles was about 1% w/w, with about 20% burst release in the ®rst day in vitro, followed by release of the remaining protein over about 50±60 days. The MF59 adjuvant used in the studies described was prepared and characterized as previously described [30]. The integrity of the initially released gp120 was evaluated by CD4 binding in an established assay. Test samples of gp120 were assayed for CD4 binding activity by adding a known quantity of recombinant soluble CD4 to each test sample then separating free from complexed CD4 using size exclusion HPLC (SEHPLC). The quantity of CD4 added to each sample was pre-
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
determined to be in molar excess to active gp120 content. The free CD4 content was determined against a CD4 standard curve and used to calculate the quantity of bound CD4. An equivalent volume of the test sample without added CD4 was assayed in parallel and used to quantitate gp120 content determined against a gp120 standard. The percent CD4 binding activity for each test sample was then calculated as the percent molar ratio of bound CD4 to gp120 content. Repetitive assays of gp120 bulk antigen determined it to have an average CD4 binding activity of 57% and was assigned a normalization factor of 1. CD4 binding reproducibility was determined from reference standard injections made on consecutive assays (total n = 8) and was found to have a %RSD equal to 7.1%, with a 95.5% prediction interval of 0.80±1.07 of normalized activity. Reference gp120 samples were assayed for comparison purposes as time zero samples. Results from injections of test samples without added CD4 were used to assess changes in gp120 aggregation state or appearance of low molecular weight substances (substances eluting at lesser or greater elution volumes than that of gp120) as compared to reference samples. Although nonquantitative, these assays indicated that the gp120 initially released from microparticles retained some ability to bind to CD4.
2.4. Immunization protocols For antibody induction against gp120, groups of 30 mice were immunized intramuscularly (IM) on three occasions at weeks 0, 4 and 8 with 10 mg of gp120. In three dierent groups of mice, soluble antigen was compared with antigen entrapped in PLG, and with PLG in MF59. For CTL induction, groups of ®ve mice were immunized IM on days 0, 7 and 14 with 10 mg of p24 in PLG microparticles and PLG in MF59; these formulations were compared to a dose of 50 mg of soluble p24 IM. For comparison, an Iscom p24 formulation, which was prepared as previously described by Bror Morein et al. [31] was also included in the studies at a dose of 1 mg of p24. This dose of p24 was limited by the toxicity of the Iscom's formulation. In addition, another group of mice were immunized with vaccinia gag/ pol, to provide an assay positive control group. For baboon studies, the animals were immunized at weeks 0, 4 and 26. In one study, groups of ®ve baboons were immunized with gp120 in MF59, in PLG/gp120, and in a combination of PLG/gp120 in MF59. In a second study, groups of six baboons were immunized with p24 in MF59 alone, or with PLG/p24 in MF59.
1795
2.5. Determination of gp120-speci®c antibody responses in mice An enzyme-linked immunosorbent assay (ELISA) designed to measure HIV-1SF2 gp120-speci®c antibody was performed on mice sera at week 6 and 10. Puri®ed HIV-1SF2 gp120 protein was coated onto Nunc Maxisorp U bottom plates at 2 mg/ml. Sera were tested at 1:20 and 1:200 dilutions followed by serial three-fold dilutions. Horseradish peroxidase conjugated goat anti mouse IgG(G+A+M) (Gibco; diluted 1:6000) was used as a second antibody. After the 1-h incubation at 378C, plates were washed to remove unbound antibody. TMB substrate was used to develop the plates and the color reaction was blocked after 10 min by the addition of 2 N HCL. The titers reported are the reciprocal of the serum dilutions that gave an optical density at 450 nm of 0.5 ELISA absorbency units. 2.6. Determination of gp120 speci®c serum antibody responses in baboons Serum from individual baboons were assayed for IgG antibody responses using ELISA. ELISA plates (Nunc Maxisorb U96) were coated with puri®ed HIV1SF2 gp120 antigen in 50 mM sodium borate buer, pH 9.0. Sera samples were added to the plates initially diluted 1:50, 1:200, 1:500, or 1:1000 fold with diluting solution (100 mM sodium phosphate, 0.5 M sodium chloride, 1 mM EDTA, 0.01% thimerosal, pH 7.5) followed by serial two-fold dilution with the same buer. Horseradish peroxidase conjugated goat anti-monkey IgG (Cappel Organon Teknike Corp., West Chester, PA) was used as detection antibody. Plates were incubated for 1 h at 378C followed by washing with 0.05% Triton-X100 between each step. The data for each sample was ®t using curve ®tting software to a spline function and the resulting titers were reported as the reciprocal dilution required to achieve 0.5 OD units at 450 nm after developing with TMB substrate. 2.7. Evaluation of neutralizing antibodies in baboons These assays were performed in the laboratory of Tom Van Cott at the Henry M. Jackson Foundation, Rockville, MD. For the 2 week post third immunization sera from baboons, assessment of neutralizing antibody activity of serum against a T-cell line adapted HIV-1IIIB and a primary isolate HIV-1BZ167 was performed as previously described [32±35]. HIV-1BZ167 is a CXCR4-using primary isolate found to be sensitive to antibody-mediated neutralization [32]. Phytohemagglutinin (PHA)-stimulated peripheral blood mononuclear cells (PBMC) were used as target cells for both primary and T cell line adapted (TCLA) strains. All baboon sera were heat-inactivated at 568C for 40 min
1796
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
prior to use. Sera were run at a single dilution of 1:4 in the presence of virus and 1:8 in the presence of virus and target cells. Inhibition of PBMC infection was assessed by quantitative p24 measurement of cell supernatants during the early virus growth phase (day 3 for BZ167 and day 4 for IIIB). The results are reported as percent neutralization (decrease in p24 antigen level in immune sera wells compared to control sera). Neutralization was considered signi®cant if there was a ®ve-fold (80%) reduction of virus growth in the presence of immune sera as compared to the preimmune serum control.
2.8. Evaluation of CTL responses in mice Spleens from immunized mice were harvested 2 weeks following the 3rd immunization and used as pools of ®ve. Spleen cells from immunized mice were cultured in a 24-well dish at 5 106 cells per well. Of these cells, 1 106 were sensitized with synthetic p7 g peptide (amino acids 194±213) at a concentration of 10 mM for 1 h at 378C and then washed and co-cultured with the remaining 4 106 untreated cells. The cells were stimulated as a bulk culture in 2 ml of Splenocyte culture medium: RPMI 1640 with 100 mM Lglutamine (Gibco, Grand Island, New York, USA)/aMem (Minimum Essential Medium Alpha Medium with L-glutamine, deoxyribonucleosides or ribonucleosides) (1:1) supplemented with 10% heat-inactivated fetal calf serum (Hyclone, Logan, Utah, USA), inactivated in a 568C water bath for 30 min, 100 U/ml penicillin, 100 mg/ml streptomycin, 10 ml/l of 100 mM sodium pyruvate and 50 mM 2-mercaptoethanol. In addition, 5% Rat T-Stim IL2 (Rat T-Stim; Collaborative Biomedical Products, Bedford, Massachusetts, USA) was used as a source of IL2 and is added to the culture media just before the cells are to be cultured.
2.9.
51
Chromium assay
After a stimulation period of 6±7 days, the cells were collected and used as eectors in a chromium release assay. Approximately 1 106 SV/Balb target cells were incubated in 200 ml of medium containing 50 mCi of 51Cr and with the correct peptide p7 g at a concentration of 1 mM for 60 min and then washed, or in the absence of peptide as the negative control. Eector cells were cultured with 5 103 target cells at various eector to target ratios in 200 ml of culture medium in 96-well tissue culture plates (round or v-bottom) for 4 h. The average cpm from duplicate wells was used to calculate percent speci®c release as previously described [34].
2.10. Statistics Analysis of variance was calculated by using the StatView program for Macintosh computers. Dierences among groups of animals at signi®cance levels of 95% were calculated by analysis using Fisher's protected least-signi®cant-dierence test. 3. Results 3.1. Antibody responses in mice PLG/gp120 induced potent serum IgG responses in mice, which were signi®cantly better than the responses induced by MF59 ( p = 0.0002). In addition, the levels of IgG antibodies induced were signi®cantly enhanced over PLG alone, when the PLG/gp120 microparticles were combined with MF59 ( p = 0.0009; Fig. 1). In addition, it was also shown that PLG microparticles induced high levels of the IgG2a isotype, while MF59 induced mainly the IgG1 isotype (data not shown). In a separate study in mice, the antibody responses to PLG/p24 and p24 alone were determined in mice (a titer of 26,000 was induced for PLG/p24, while the response to the p24 protein alone was not detectable). In a separate study, the titers to PLG/p24 and MF59+p24 were also compared in mice and were shown to be not statistically dierent (data not shown).
Table 1 CTL activity in splenocytes of mice following immunization at weeks 0, 1 and 2 with p24 entrapped in PLG microparticles, in PLG mixed with MF59, incorporated into an Iscom formulation, or in saline alone. Vaccinia gag/pol was used as a positive control group in this assay. The table shows cytolysis at dierent eector to target cell ratios on SV/Balb cells in the presence of the relevent epitopic peptide (SV/p7 g) and in the absence of peptide (SV/0) Eector
E:T ratio
SV/O
SV/p7 g
PLG/p24 10 mg IM
60:1 12:1 2.4:1 60:1 12:1 2.4:1 60:1 12:1 2.4:1 60:1 12:1 2.4:1 60:1 12:1 2.4:1
7 3 2 4 5 3 10 5 3 8 5 2 <1 2 2
27 1 0 35 5 <1 44 11 2 3 0 <1 58 50 20
PLG/p24+MF59 10 mg IM p24 ISCOM 1mg IM soluble/p24 protein 50 mg IM vaccinia gag pol 1 107 pfu IP
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
1797
Fig. 1. Serum IgG responses (mean2SEM) in mice following IM immunization at weeks 0, 4 and 8 with 10 mg of gp120 in saline, entrapped in PLG microparticles, or entrapped in PLG microparticles which were co-administered with MF59 adjuvant.
3.2. CTL responses in mice PLG/p24 and PLG/p24 in MF59 both induced CTL activity in mice. In addition, although CTL assays based on bulk cultures of splenocytes are not considered to be quantitative, the level of activity induced by the PLG microparticles appeared to be comparable to the Iscom positive control adjuvant group. In contrast, p24 protein alone at a much higher dose was not CTL active following IM immunization (Table 1).
dierences between the groups (Fig. 2). In contrast to the observations with gp120, PLG/p24 in MF59 showed a signi®cantly enhanced response over MF59 alone+p24 at week 28, following three immunizations (Fig. 3). 3.4. Neutralizing antibody responses in baboons
After a single immunization, the PLG/gp120 in MF59 group showed a signi®cantly enhanced serum IgG antibody response in comparison to the other groups, PLG/gp120 and gp120 in MF59 (Fig. 2). However, after a boost, the dierences between the groups were no longer signi®cant. Moreover, after an additional boost at 26 weeks, there were no signi®cant
Sera from MF59+gp120 (5/5) immunized animals and from PLG/gp120+MF59 (3/5) immunized animals neutralized (>80% inhibition of viral growth) HIV-1 (IIIB); the value of % neutralization titer as a mean of pre-immune sera is shown in Table 2. The mean ELISA titer at this time point as determined in the laboratory of Tom VanCott is also included in Table 2 for comparison. In contrast, 0/5 animals immunized with PLG/gp120 alone showed neutralization titers against IIIB (>80% inhibition). Zero out of ®ve animals (>80% inhibition) showed neutralization against HIV-1 BZ167 from the MF59+gp120, but 2/5 from
Fig. 2. Serum IgG responses (mean 2 SEM) in baboons following IM immunization at weeks 0, 4 and 26 with 50 mg gp120 in MF59, entrapped in PLG microparticles, or entrapped in PLG microparticles which were co-administered with MF59 adjuvant.
Fig. 3. Serum IgG responses (mean 2 SEM) in baboons at week 28, following IM immunization at weeks 0, 4 and 26 with 50 mg of p24 in MF59, or entrapped in PLG microparticles which were co-administered with MF59 adjuvant.
3.3. Antibody responses in baboons
1798
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
Table 2 ELISA and neutralizing antibody responses in baboons following IM immunization at weeks 0, 4 and 26 with 50 mg gp120 in MF59, entrapped in PLG microparticles, or entrapped in PLG microparticles which were co-administered with MF59 adjuvant. Neutralization titers are reported as the percent neutralization for individual baboon sera in comparison to pre-immune control sera Group
Treatment
Bleed
Mean ELISA titer
HIV-1 (IIIB) % Neuts
HIV-1 (BZ167) % Neuts
1
gp120+MF159
2wp3
58,813
2
PLG/gp120
2wp3
22,286
3
PLG/gp120+MF59
2wp3
67,559
95 99 90 100 100 < 50 < 50 < 50 < 50 59 99 < 50 100 89 67
78 72 58 < 50 53 < 50 < 50 < 50 < 50 < 50 93 < 50 96 64 < 50
the PLG/gp120+MF59 group showed neutralizing titers against this primary isolate (Table 2). In contrast, no animals from the PLG/gp120 alone group showed neutralizing titers against HIV-1 BZ167.
4. Discussion The work described here highlights the potential of PLG microparticles as an adjuvant for a recombinant protein vaccine against HIV-1. PLG microparticles with entrapped gp120 induced potent antibody responses in mice, as previously reported [25]. In addition, the adjuvant eect of microparticles was enhanced by administration in MF59, a potent emulsion adjuvant. Although in the current study, PLG microparticles alone were signi®cantly better than MF59 for induction of antibodies against gp120, in a previous study in smaller groups of mice (n = 10), the two adjuvants were shown to induce comparable responses (titers 4 weeks post three immunizations, PLG 11,250 vs MF59 18,250). Although we do not know the reason for the relatively poor performance of MF59 alone in the current study, the variable response to MF59 alone and gp120 in the two studies does not detract from the main observation reported here, that PLG+MF59 is signi®cantly more potent than either adjuvant used alone. PLG microparticles with a much larger mean size have previously been used to develop a single dose controlled release vaccine against HIV-1 [22,23]. However, in the current studies, smaller microparticles were used for optimal adjuvant eect for both antibody and CTL induction [19,20,37]. In studies previously described involving microencapsulation of gp120, QS21 was added to the large microparticles
to promote CTL induction [22,23]. The current studies demonstrated for the ®rst time that PLG microparticles were an eective adjuvant for CTL induction against p24 gag, following IM immunization in mice. PLG has previously been reported to be an eective adjuvant for CTL induction with alternative proteins [25,38]. Importantly, the CTL activity of PLG was not diminished by co-administration of MF59. The emulsion adjuvant MF59 has previously been shown to be a safe and eective adjuvant for the induction of potent antibody responses [30]. However, MF59 is not considered to be an eective adjuvant for the induction of CTL responses following IM immunization in mice. In previous studies, it has been shown that CTL induction against p24 by presentation on a yeast derived virus-like particle was eliminated by co-administration with adjuvants which are not eective for CTL induction [39]. Therefore, it was an important observation that the PLG/p24 and MF59 combination was eective for CTL induction and had similar potency to Iscom's, a potent adjuvant for CTL induction [31]. Iscom's and PLG microparticles could not be evaluated at an equivalent dose of protein, due to the toxicity of Iscoms, which has proven lethal to Balb/c mice at a 10 mg dose of p24 protein (unpublished observations). Since we do not consider CTL assays based on bulk cultures of splenocytes to be quantitative, we cannot dierentiate between the levels of activity induced by the Iscoms or the PLG adjuvants. All that we can say is that both adjuvant approaches were positive for CTL activity in splenocytes, in comparison to immunization with protein alone, which did not induce cytolysis (Table 1). The signi®cant bene®t of the PLG/MF59 adjuvant combination for the induction of enhanced immune responses was more clearly
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
demonstrated through the induction of antibodies against gp120 in mice. Disappointingly, the observations concerning the enhanced potency of the PLG and MF59 adjuvant combination were not reliably reproduced in primates. However, this may partly be due to the use of only small groups of primates (®ve or six baboons), in contrast to the use of much larger groups of mice (30 animals). Nevertheless, following a single immunization with PLG/gp120 in MF59 in baboons, this group of animals showed a signi®cantly enhanced serum IgG response in comparison to the use of either adjuvant alone. However, the signi®cantly enhanced response was not maintained after boosting, and there was no signi®cant dierence between the adjuvant groups after two or three doses. Nevertheless, in contrast, the baboons immunized with PLG/p24 in MF59 did show signi®cantly enhanced serum IgG responses in comparison to immunization with MF59 alone after three doses. Although we do not have an explanation for the dierence in responses between the two proteins, the combined primate data is supportive of a signi®cant bene®t in terms of enhanced immunogenicity, through the combination of PLG microparticles in MF59. The ability of PLG microparticles to induce CTL responses against HIV-1 gag was not evaluated in the current study, but will be reported in a subsequent paper. Encouragingly, both MF59+gp120 and PLG/gp120 in MF59 induced antibodies capable of neutralizing a T cell-adapted line strain of HIV-1. However, only the PLG/gp120+MF59 adjuvant combination was capable of inducing signi®cant levels of neutralization against a primary isolate of HIV-1 (two out of ®ve baboons). The inability of PLG/gp120 alone to induce signi®cant neutralizing titers was a disappointment and may re¯ect a common problem with antigens entrapped in PLG microparticles, in that the antigens are often subject to denaturation [40]. Although gp120 has previously been successfully entrapped in PLG microparticles and has induced long term high levels of neutralizing antibodies [22,23], the neutralization titers from the primate studies using only PLG/gp120 alone suggested that the protein had been denatured during the formulation process. This assertion was supported by V3 loop epitope mapping studies from baboon sera, which indicated that the gp120 in microparticles was structurally altered and V3 reactivity was reduced (data not shown). The apparent instability of gp120 in the current microparticle formulation, in contrast to that described previously [22,23], may be a consequence of the high level of shear applied to the current formulation to generate small microparticles. While the use of small microparticles may be bene®cial for the induction of CTL responses [25,38] and may be optimal for antibody induction [19,20], the high level
1799
of shear involved in microparticle preparation may be disadvantageous in terms of protein stability in the formulation. Although the PLG/MF59 adjuvant combination was eective for the induction of neutralizing antibodies, it appears likely that only the protein released initially from the microparticles contributed to the neutralizing titer, which was adjuvanted by MF59. During the formulation studies described here, the integrity of gp120 was only assessed for protein released initially, since it had been reported previously that gp120 was stable in PLG microparticles [22,23]. The available data would appear to suggest that while the initially released protein was intact, protein released at later time points may have been degraded and was not capable of inducing neutralizing antibodies. To develop an optimal adjuvant formulation with microencapsulated gp120, it seems likely that further formulation development is needed to optimize protein stability in PLG. However, it appears that monomeric gp120 is not the optimal immunogen to use in future studies for the induction of broadly cross-neutralizing antibodies against primary isolates of HIV-1 [35,36,41]. Subsequent studies will evaluate the ability of PLG microparticles and additional adjuvant combinations to perform as eective adjuvants for alternative forms of env from HIV-1. In summary, we have shown here that PLG microparticles are a potent adjuvant for the induction of antibody and CTL responses in mice against recombinant HIV-1 protein antigens. In addition, signi®cantly enhanced responses were obtained in mice when the microparticles were administered in MF59 emulsion. Limited studies in primates with small groups of animals provided support for these observations and con®rmed that the PLG/MF59 combination was a potent adjuvant for the induction of serum IgG antibody responses. Nevertheless, gp120 stability in microparticles may be a problem and probably restricted the ability of the combined formulation to induce optimally enhanced levels of neutralizing antibodies in primates.
Acknowledgements We are grateful to Gary Van Nest and Susan Barnett for advice and encouragement during the performance of these studies. We are also grateful to Bror Morein for the preparation of the p24 iscom formulation. References [1] Global overview: a powerful HIV/AIDS pandemic. AIDS in the world. In: Mann J, Tarantola D, editors. Status and Trends of
1800
[2] [3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801 the HIV/AIDS Pandemic. New York: Oxford University Press, 1996. Dolin R. Human studies in the development of human immunode®ciency virus vaccines. J Infect Dis 1995;172:1175±83. Kahn JO, Sinangil F, Baenziger J, Murcar N, Wynne D, Coleman R, Steimer K, Dekker C, Cherno D. Clinical and immunologic responses to human immunode®ciency virus (HIV) type 1SF2 gp120 subunit vaccine combined with MF59 adjuvant with or without muramyl tripeptide dipalmitoyl phosphatidylethanolamine in non-HIV-infected human volunteers. J Infect Dis 1994;170:1288±91. Mascola JR, Snyder SW, Weislow OS, Belay S, Belshe R, Schwartz D, Clements M, Dolin R, Graham B, Gorse G, Keefer M, McElrath M, Walker M, Wagner K, McNeil J, McCutchan F, Burke D. Immunization with envelope subunit vaccine products elicits neutralizing antibodies against laboratory-adapted but not primary isolates of human immunode®ciency virus type 1. J Infect Dis 1996;173:340±8. Igarashi T, Brown C, Azadegan A, Haigwood N, Dimitrov D, Martin MA, Shibata R. Human immunode®ciency virus type 1 neutralizing antibodies accelerate clearance of cell-free virions from blood plasma. Nat Med 1999;5:211±6. Shibata R, Igarashi T, Haigwood N, Buckler-White A, Ogert R, Ross W, Willey R, Cho MW, Martin MA. Neutralizing antibody directed against the HIV-1 envelope glycoprotein can completely block HIV-1/SIV chimeric virus infections of macaque monkeys. Nat Med 1999;5:204±10. Mascola Jr, Lewis ML, Stiegler G, Haris D, Vancott TC, Hayes D, Louder MK, Brown CR, Sapan CV, Frankel SS, Lu Y, Robb ML, Katinger H, Birx DL. Protection of macaques against pathogenic simian/human immunode®ciency virus 89.6PD by passive transfer of neutralizing antibodies. J Virol 1999;73:4009±18. Rowland-Jones SL, Nixon DF, Aldhous MC, Gotch F, Ariyoshi K, Hallam N, Kroll JS, Froebel K, McMichael A. HIV-speci®c cytotoxic T cell activity in an HIV exposed but uninfected infant. Lancet 1993;341:860±1. Beyrer C, Artenstein AW, Rugpao S, Stephens H, Vancott TC, Rob ML, Rinkaew M, Birx DL, Khamboonruang C, Zimmerman PA, Nelson KE, Natpratan C. Epidemiologic and biologic characterization of a cohort of human immunode®ciency virus type 1 highly exposed, persistently seronegative female sex workers in northern Thailand. J Infect Dis 1999;179:59±67. Clerici M, Giorgi JV, Chou C-C, Gudeman VK, Zack JA, Gupta P, Ho H-N, Nishanian PG, Berzofsky JA, Shearer GM. Cell-mediated immune response to human immunode®ciency virus (HIV-1) type 1 in seronegative homosexual men with recent sexual exposure to HIV-1. J Infect Dis 1992;165:1012±9. Koup RA, Safrit JT, Cao Y, Andrews CA, McLeod G, Borkowsky W, Farthing C, Ho DD. Temporal association of cellular immune responses with the initial control of viremia in primary human immunode®ciency virus type 1 syndrome. J Virol 1994;68:4650±5. Borrow P, Lewicki H, Hahn BH, Shaw GM, Oldstone MB. Virus-speci®c CD8+ cytotoxic T-lymphocyte activity associated with control of viremia in primary human immunode®ciency virus type 1 infection. J Virol 1994;68:6103±10. Ogg GS, Jin X, Bonhoeer S, Dunbar PR, Nowak MA, Monard S, Segal JP, Cao Y, Rowland-Jones SL, Cerundolo V, Hurley A, Markowitz M, Ho DD, Nixon DF, McMichael AJ. Quantitation of HIV-1 speci®c cytotoxic lymphocytes and plasma load of viral RNA. Science 1998;279:2103±6. Rosenberg ES, Billingsley JM, Caliendo AM, Boswell SL, Sax PE, Kalams SA, Walker BD. Vigorous HIV-1-speci®c CD4+ T cell responses associated with control of viremia. Science 1997;278:1447±50.
[15] Mazzoli S, Trabattoni D, Lo Caputo S, Piconi S, Ble C, Meacci F, Ruzzante S, Salvi A, Semplici F, Longhi R, Fusi ML, Tofani N, Biasin M, Villa ML, Mazzotta F, Clerici M. HIV-speci®c mucosal and cellular immunity in HIV-seronegative partners of HIV-seropositive individuals. Nat Med 1997;3:1250±7. [16] Gallimore A, Cranage M, Cook N, Almond N, Bootman J, Rud E, Silvera P, Dennis M, Corcoran T, Stott J, McMichael A, Gotch F. Early suppression of SIV replication by CD8+ nefspeci®c cytotoxic T cells in vaccinated macaques. Nat Med 1995;1:1167±73. [17] Walker CM, Moody DJ, Stites DP, Levy JA. CD8+ lymphocytes can control HIV infection in vitro by suppressing virus replication. Science 1986;234:1563±6. [18] Cocchi F, DeVico AL, Garzino-Demo A, Arya KS, Gallo RC, Lusso P. Identi®cation of RANTES, MIP-q and MIP 1q as the major HIV-suppressive factors produced by CD8+ T cells. Science 1995;270:1811±5. [19] Eldridge JH, Staas JK, Meulbroek JA, Tice TR, Gilley RM. Biodegradable and biocompatible poly (DL-lactide-co-glycolide) microspheres as an adjuvant for staphylococcal enterotoxin B toxoid which enhances the level of toxin-neutralizing antibodies. Infect Immun 1991;59:2978±86. [20] O'Hagan DT, Rahman D, McGee JP, Jeery H, Davies MC, Williams P, Davis SS, Challacombe SJ. Biodegradable microparticles as controlled release antigen delivery systems. Immunol 1991;73:239±42. [21] Vordermeir HM, Coombes AGA, Jenkins P, McGee JP, O'Hagan DT, Davis SS, Singh M. Synthetic delivery systems for tuberculosis vaccines: immunological evaluation of the M. tuberculosis 38 kDa protein entrapped in biodegradable microparticles. Vaccine 1995;13:1576±82. [22] Cleland JL, Lim A, Barron L, Duenas ET, Powell MF. Development of a single-shot subunit vaccine for HIV-1: Part 4. Optimizing microencapsulation and pulsatile release of MN rgp120 from biodegradable microspheres. J Cont Rel 1997;47:135±50. [23] Cleland Jl, Lim A, Daugherty A, Barron L, Desjardin N, Duenas ET, Eastman DJ, Vennari JC, Wrin T, Beman P, Murthy KK, Powell MF. Development of a single-shot subunit vaccine for HIV-1: Part 5. Programmable in vivo autoboost and long lasting neutralizing response. J Pharm Sci 1998;87:1489±95. [24] O'Hagan DT, McGee JP, Boyle R, Gumaer D, Li X-M, Potts B, Wang CY, Ko WC. The preparation, characterization and pre-clinical evaluation of an orally administered HIV-1 vaccine, consisting of a branched synthetic peptide immunogen entrapped in controlled release microparticles. J Cont Rel 1995;36:75±84. [25] Moore A, McGuirk P, Adams S, Jones WC, McGee P, O'Hagan DT, Mills KHG. Immunization with a soluble recombinant HIV protein entrapped in biodegradable microparticles induces HIV-speci®c CD8+ cytotoxic T lymphocytes and CD4+ Th1 cells. Vaccine 1995;113:1741±9. [26] Jeery H, Davis SS, O'Hagan DT. The preparation and characterization of poly(lactide-co-glycolide) microparticles. II. The entrapment of a model protein using a (water-in-oil)-in-water emulsion solvent evaporation technique. Pharm Res 1993;10:362±8. [27] McGee JP, Singh M, Li XM, Qiu H, O'Hagan DT. The encapsulation of a model protein in poly (D, L-lactide-co-glycolide) microparticles of various sizes: an evaluation of process reproducibility. J Microencap 1997;14:197±210. [28] O'Hagan DT, Jeery H, Davis SS. The preparation and characterization of poly(lactide-co-glycolide) microparticles. III: Microparticle/polymer degradation rates and the in vitro release of a model protein. Int J Pharm 1994;103:37±45. [29] Barackman JD, Singh M, Ugozzoli M, Ott GS, O'Hagan DT. Oral immunization with poly(lactide-co-glycolide) microparticles
D.T. O'Hagan et al. / Vaccine 18 (2000) 1793±1801
[30]
[31]
[32]
[33]
[34]
containing an entrapped recombinant glycoprotein (gd2) from herpes simplex type 2 virus. STP Pharma Sci 1998;8:41±6. Ott G, Barchfeld GL, Cherno D, Radhakrishnan R, Hoogevest PV, Van Nest G. Design and evaluation of a safe and potent adjuvant for human vaccines. In: Newman JM, Powell MF, editors. Vaccine Design: The Subunit and Adjuvant Approach. New York: Plenum Press, 1995. p. 277±95. Wilson AD, Lovgren-Bengtsson K, Villacres-Ericsson M, Morein B, Morgan AJ. The major Epstein±Barr virus (EBV) envelope glycoprotein gp340 when incorporated into iscoms primes cytotoxic T-cell responses directed against EBV lymphoblastoid cell lines, Vaccine 1999;17:1282±1290. Vancott TC, Mascola JR, Kaminski RW, Kalyanaraman V, Hallberg PL, Burnett PR, Ulrich JT, Rechtman DJ, Birx DL. Antibodies with speci®city to native gp120 and neutralization activity against primary human immunode®ciency virus type 1 isolates elicited by immunization with oligomeric gp160. J Virol 1997;71:4319±30. Mascola J, Weislow O, Snyder S, Belay S, Yeager M, McCutchan F, McNeil J, Burke D, Walker MC. Neutralizing antibody activity in sera from human immunode®ciency virus type-1 vaccine recipients form the AIDS vaccine clinical trials network. AIDS Res Hum Retroviruses 1994;10:S55. Mascola JR, Snyder SW, Weislow OS, Belay SM, Belshe RB, Schwartz DH, Clements ML, Dolin R, Graham BS, Gorse GJ, Keefer MC, McElrath MJ, Walker MC, Wagner KF, McNeil JG, McCutchan FE, Burke DS. Immunization with envelope subunit vaccine products elicits neutralizing antibodies against laboratory-adapted but not primary isolates of human immuno-
[35]
[36]
[37]
[38]
[39]
[40] [41]
1801
de®ciency virus type 1. The National Institute of Allergy and Infectious Diseases AIDS Vaccines Evaluation Group. J Infect Dis 1996;173:340±8. Vancott TC, Polonis VR, Loomis LD, Michael NL, Nara PL, Birx DL. Dierential role of V3-speci®c antibodies in neutralization assays involving primary and laboratory-adapted isolates of HIV type 1. AIDS Res Hum Retroviruses 1995;11:1379±91. Doe B, Steimer KS, Walker CM. Induction of HIV-1 envelope (gp120)-speci®c cytotoxic T lymphocyte responses in mice by recombinant CHO cell-derived gp120 is enhanced by enzymatic removal of N-linked glycans. Euro J Immunol 1994;24:2369±76. O'Hagan DT, Jeery H, Davis SS. Long-term antibody responses in mice following subcutaneous immunization with ovalbumin entrapped in biodegradable microparticles. Vaccine 1993;11:965±9. Maloy KJ, Donachie DT, O'Hagan DT, Mowat MA. Induction of mucosal and systemic immune responses by immunization with ovalbumin entrapped in poly(lactide-co-glycolide) microparticles. Immunol 1994;81:661±7. Harris SJ, Woodrow SA, Gearing AJH, Adams SE, Kingsman AJ, Layton GT. The eects of adjuvants on CLT induction by V3: Ty-virus-like particles (V3-VLPs) in mice. Vaccine 1996;14:971±6. O'Hagan DT, Singh M, Gupta RK. Poly(lactide-co-glycolide) microparticles for the development of single-dose controlledrelease vaccines. Adv Drug Del Rev 1998;32:225±46. LaCasse RA, Follis KE, Trahey M, Scarborough JD, Littman DR, Nunberg JH. Fusion-competent vaccines: broad neutralization of primary isolates of HIV. Science 1999;283:357±62.