Effect of Carrier Priming on Immunogenicity of Saccharide-Protein Conjugate Vaccines CARLA C

Home , Priming (immunology)

INFECTION AND IMMUNITY, Oct. 1991, p. 3504-3510 Vol. 59, No. 10 0019-9567/91/103504-07$02.00/0 Copyright C 1991, American Society for Microbiology

Effect of Carrier Priming on Immunogenicity of Saccharide-Protein Conjugate Vaccines CARLA C. A. M. PEETERS,lt* ANNE-MARIE TENBERGEN-MEEKES,1 JAN T. POOLMAN,2 MICHEL BEURRET,3 BEN J. M. ZEGERS,1 AND GER T. RIJKERS' Department ofImmnunology, University Hospital for Children and Youth "Het Wilhelmina Kinderziekenhuis," P.O. Box 18009, 3501 CA Utrecht,1 and Unit of Bacterial Vaccine Development and Pathogenesis Research, National Institute of Public Health and Environmental Protection, Bilthoven,2 The Netherlands, and Praxis Biologics, Rochester, New York3 Received 19 March 1991/Accepted 15 July 1991

Previous studies with saccharide-protein conjugates have demonstrated that antibody responses to the saccharide can be improved by the preexistence of carrier immunity. Here we report that prior exposure to the carrier prptein can either enhance or suppress antibody response to polysaccharides administered in saccharide-protein conjugates. A dose-dependent role for carrier priming in the antisaccharide antibody response to three saccharide-protein conjugate vaccines, i.e., a Streptococcus pneumoniae type 4 polysaccharide-tetanus toxoid (TT) conjugate (PS4TT), a Neisseria meningitidis group C polysaccharide-TT conjugate (MenCTT), and a N. meningitWidis group C oligosaccharide-diphtheria mutant toxin conjugate (MenCCRM), was investigated. The results showed that an increase in the antipolysaccharide antibody response could be obtained for both PS4TT and MenCTT but not for MenCCRM with low-dose carrier priming (0.025 to 0.25 ,ug). However, suppression of the antipolysaccharide antibody response was observed with the PS4TT and MenCTT vaccines wi,h high-dose (25-,ug) carrier priming. There was no suppression effect with MenCCRM. The increase in the antipolysaccharide antibody response was shown to be restricted to the immunoglobulin Gl (IgGl) subclass, whereas suppression with high-dose carrier prining affected all antipolysaccharide subclass antibodies induced by PS4TT (IgGl, IgG2b, and IgG3) and only two of the four subclass antibodies induced by MenCTT (IgG2a and IgG2b). The increase in the antipolysaccharide antibody response was also present at the antipolysaccharide IgM antibody level but was not observed at the anti-carrier IgG antibody level.

The enhancement of the immunogenicity of polysaccha- The application of saccharide-protein conjugates is of rides by coupling polysaccharides or oligosaccharides to primary importance in inducing protective immunity against proteins has been amply demonstrated (1, 2, 4, 7, 9, 14). In infection with several encapsulated bacteria in infants and in most cases, saccharides have been coupled to large immu- elderly and immunodeficient patients. Therefore, we inves- nogenic proteins such as tetanus toxoid (TT) or diphtheria tigated in an animal model the role of carrier priming on the toxoid (DT). These proteins are chosen in most studies antibody response to a saccharide-protein conjugate. Since because both have been employed for human vaccination for different saccharides can be coupled to TT, it is important to many years without untoward side effects. Moreover, since know whether the influence of carrier priming is similar with most individuals are immunized with TT and DT, the re- different saccharides when TT is used as a carrier. We found sponse against a hapten coupled to carriers like TT and DT that preimmunization with low doses of TT enhanced the potentially can be improved by the preexistence of anticar- antibody response to Streptococcus pneumoniae type 4 rier immunity. However, the antibody response to a hapten polysaccharide (PS4) and meningococcal group C (MenC) coupled to the carrier protein can also be inhibited when the polysaccharide upon vaccination with TT conjugates of recipient has been previously immunized with the unmodi- these polysaccharides. High doses of carrier priming, how- fied protein. This phenomenon has been termed carrier- ever, inhibited a conjugate-induced antipolysaccharide anti- induced epitope suppression (17) and was recently demon- body response. Carrier priming with DT did not affect an strated to occur with a number of synthetic peptide-protein antibody response to a MenC oligosaccharide-mutant DT conjugates (10, 19, 20, 27, 28, 31). In humans, carrier- (CRM197) conjugate (MenCCRM197). induced epitope suppression has been described for synthetic peptides coupled to TT (10, 14). Several investigators suggested that epitope suppression could also occur upon MATERIALS AND METHODS vaccination with saccharide-protein conjugate vaccines in Animals. One-week-old male and female and 8- to 12- human adults (7, 8, 16), but experimental evidence was not week-old female NIH/RIVM mice (random outbred strain) given. Carrier priming with one dose of carrier protein has were used. They were bred and kept at the National Institute been shown to have a positive effect on the polysaccharide of Public Health and Environmental Protection. response (8, 16). Antigens. In this study, three different saccharide-protein conjugates were used. A polysaccharide-TT conjugate of S. * Corresponding author. pneumoniae type 4 (PS4TT) with a polysaccharide protein t Present address: Unit of Bacterial Vaccine Development and ratio of 0.7:1 (wt/wt) was prepared by using the carbodiimide Pathogenesis Research, National Institute of Public Health and coupling procedure as described earlier (21). In brief, PS4 Environmental Protection, Antonie van Leeuwenhoeklaan 9, 3720 was activated with cyanogen bromide and subsequently BA Bilthoven, The Netherlands. coupled to a spacer 6-aminohexanoic acid. The purified 3504 VOL. 59, 1991 CARRIER PRIMING AFFECTS POLYSACCHARIDE ANTIBODY LEVELS 3505 product of this reaction was coupled to TT by using 1- To exclude interference by antipneumococcal cell wall ethyl-3(3-dimethylaminopropyl)carbodiimide. Polysaccharide- polysaccharide antibodies in the determination of IgM anti- protein conjugates were purified by gel filtration on a Seph- PS4 antibodies, all sera were adsorbed with soluble cell wall arose CL-4B column. The second polysaccharide-protein polysaccharide (CPs; State Serum Institute). To that end, conjugate used was a conjugate of MenC and TT (MenCTT) soluble cell wall polysaccharide at a concentration of 250 with a polysaccharide/protein ratio of 1:1 (wt/wt). The con- ,ug/ml was added to the dilution buffer during serum incuba- jugate was prepared by using methods similar to those tion as described in detail elsewhere (22). described above for PS4TT (4). Statistical analysis. Results are expressed as logarithmic In addition, an experimental oligosaccharide-mutant DT mean titers of n independent observations + standard devi- conjugate (CRM197) vaccine of Neisseria meningitidis group ation. Geometric mean (antilog) titers are expressed in C (MenCCRM) was used. Oligosaccharides of MenC poly- parentheses. Significance was tested by a one-way or two- saccharide, obtained by mild periodate oxidation, were way analysis of variance. Dunnett's correction (12) was used coupled to CRM197 with a saccharide-to-protein ratio of to compare the different treatment groups with the control 1:2.5 by using reductive amination (2). group. TT and DT were purchased from the National Institute of Public Health and Environmental Protection, Bilthoven, The RESULTS Netherlands. Immunization scheme. Groups of 8 to 10 mice were immu- Effect of preimmunization with carrier on antibody re- nized subcutaneously (s.c.) with 0.1-ml solutions of variable sponses to polysaccharide-protein conjugates in adult mice. In doses (ranging from 0.025 to 25 pig) of TT or DT in 0.9% order to investigate the dose dependency as well as the NaCl. A s.c. second immunization with the saccharide- specificity of carrier priming, animals were immunized with protein conjugate containing 0.5 pLg of saccharide was given various amounts (0.025 to 25 pig) of carrier protein (TT or 42 days after carrier priming. Neonatal mice were primed DT). Six weeks later, animals were immunized with 0.5 pig of with variable doses of TT at 1 week of age and were a saccharide-protein conjugate. Anti-polysaccharide IgG an- immunized with 0.5 ,ug of conjugate (containing 0.5 p,g of tibodies as measured on day 21 after immunization with the saccharide) at days 28 and 56 after carrier priming. Blood conjugate are shown in Table 1. Mice which were preimmu- samples were obtained from the tail veins at various times nized with a low dose of carrier protein (0.025 pig of TT) after immunization with the conjugate. Sera were stored at before administration of the polysaccharide-protein conju- -20°C until use. gates (MenCTT or PS4TT) had antipolysaccharide antibod- Antibody titers. Antibody titers against MenC polysaccha- ies higher than those of mice injected with saline or an ride, PS4, TT, and DT were determined by enzyme-linked irrelevant carrier protein (DT). This effect was statistically immunosorbent assay as described below. significant for mice immunized with MenCTT (P < 0.01) Wells of highly activated immunoassay microplates (Flow, after carrier priming but not for mice immunized with PS4TT Irvine, United Kingdom) were coated with rabbit anti-PS4 after carrier priming. When mice were primed with a high antibodies (State Serum Institute, Copenhagen, Denmark) in dose of carrier protein (25 pig) before immunization with the 0.05 M carbonate buffer (pH 9.6). After incubation for 3 h at polysaccharide-protein conjugates, the antipolysaccharide 37°C, plates were washed and incubated overnight at 4°C antibody response was suppressed, unlike the response in with 1 pig of PS4 (American Type Culture Collection, Rock- mice primed with a high dose of irrelevant protein or saline. ville, Md.) per ml in 0.9% NaCl. For the detection of In mice immunized with PS4TT, the lower anti-PS4 antibody anti-MenC polysaccharide antibodies, microplates were response was statistically significant (P < 0.01). Mice immu- coated with sheep antibodies to MenC polysaccharide (10 nized with MenCTT after being primed with 25 pig of TT pig/ml) (3) in 0.01 M phosphate-buffered saline (PBS; pH 7.2) showed a tetidency to a lower antipolysaccharide antibody overnight at room temperature, washed, and incubated with response; however, differences from their respective con- MenC polysaccharide (0.1 pig/ml) (National Institute of trols were not significant. In contrast to results with the Public Health and Environmental Protection) in PBS-Tween polysaccharide-TT conjugates, no effect of carrier priming 80 (0.05%)-bovine serum albumin (0.5%) for 2 h at 37°C. For on the antipolysaccharide IgG antibody response induced by the determination of anticarrier antibody responses, mi- the oligosaccharide-protein conjugate vaccine MenCCRM croplates were coated with TT or DT (1 pig/ml) in carbonate could be observed after priming at any dose level of DT. buffer (0.05 M, pH 9.6) overnight at room temperature. Anti-TT IgG antibody responses in mice primed with 25 pig Subsequently, all coated plates were washed and then incu- of TT and reimmunized with PS4TT did not differ from those bated for 2 h at 37°C with threefold serial dilutions of serum of mice primed with 2.5 pig of TT (Fig. 1). Either low- or samples in PBS containing 0.05% (vol/vol) Tween 20 and 1% high-dose carrier priming enhanced the anti-PS4 IgM anti- (wtlvol) bovine serum albumin. Plates were washed again body response (P < 0.01) but did not affect anti-MenC and incubated for 2 h at 37°C with alkaline phosphatase- polysaccharide IgM antibody responses (Table 2). The in- conjugated goat anti-mouse immunoglobulin (Ig) antibodies hibiting effect of high-dose (25-pig) carrier priming was thus (Southern Biotechnology Associates, Birmningham, Ala.), limited to the antipolysaccharide IgG antibody response. which detect IgG, IgM, IgGl, IgG2a, IgG2b, and IgG3. Effect of preimmunization with carrier on antibody re- Plates were washed and incubated with the phosphatase sponses to a polysaccharide-protein conjugate in neonatal substrate p-nitrophenylphosphate (Sigma, St. Louis, Mo.) at mice. Of interest is whether the effects in neonatal mice of a concentration of 1 mg/ml in 10% diethanolamine buffer (pH carrier priming on the polysaccharide antibody response, 9.8). After 30 to 60 min of incubation at room temperature, when the polysaccharide is administered as part of a the reaction was stopped by the addition of 50 pil of 2.4 N polysaccharide-protein conjugate, are similar to the effects NaOH and the A405 was read by using a Titertek Multiscan observed in adult mice. Neonatal mice were immunized with ELISA reader (Flow). All antibody titers were expressed as variable doses of TT at 1 week of age and then with 0.5 pig of percentages of a hyperimmune serum either for PS4 or PS4TT at 5 weeks of age. As in adult mice, low-dose carrier MenC polysaccharide or for TT or DT. priming (0.025 pig of TT) of neonatal mice enhanced a 3506 PEETERS ET AL. INFECT. IMMUN.

TABLE 1. Dose-dependent effects of carrier priming on 10,000 antipolysaccharide IgG antibody level to saccharide-protein conjugates in adult mice

0 Primary No. of 1,000 immunization Secondary Mean log10 titer responding * and amt immunizationb' SD (GMT)C micentota 4LO, no. of mice tested ._-o Saline PS4TT 2.079 ± 0.856 (120) 8/9 to 100 TT I- 0.025 PS4TT 2.735 ± 0.442 (544) 10/10 0.25 PS4TT 2.730 ± 0.444 (537) 10/10 I.- 2.5 PS4TT 1.568 ± 0.582 (37) 7/10 C 25 PS4TT 1.000 ± 0.723 (10) 5110d (!9 10 DT 0.025 PS4TT 2.225 ± 0.746 (168) 10/10 0.25 PS4TT 2.393 ± 0.804 (247) 9/10 2.5 PS4TT 2.097 ± 0.833 (125) 9/10 25 PS4TT 2.143 ± 0.698 (139) 9/10 1 J,&r=. 0 .025 .25 2.5 25 Saline MenCTT 2.519 ± 0.413 (416) 10/10 TT 1ag tetanus toxoid 0.025 MenCTT 3.412 ± 0.250 (2,583) 0ll1od 0.25 MenCTT 3.145 ± 0.396 (1,395) 1ollod FIG. 1. Dose-dependent influence of TT priming on the anti-TT 2.5 MenCTT 2.491 ± 0.290 (310) 10/10 IgG antibody response to a polysaccharide-protein conjugate in 25 MenCTT 2.238 ± 0.279 (173) 10/10 adult mice. Groups of 10 mice were immunized with various DT amounts of TT (O to 25 F.g). Anti-TT IgG antibodies were analyzed 0.025 MenCTT 2.677 ± 0.276 (475) 10/10 at day 28 after immunization (A). All mice were reimmunized with 0.25 MenCTT 2.599 ± 0.294 (397) 10/10 PS4TT, containing 0.5 ,ug of saccharide, at day 28 after immuniza- 2.5 MenCTT 2.816 ± 0.427 (655) 10/10 tion with the carrier. Anti-TT IgG antibodies were also analyzed at 25 MenCTT 2.413 ± 0.436 (259) 10/10 day 21 after injection with the conjugate (A). Anti-TT IgG antibody titers are shown as geometric means ± 1 standard deviation. Saline MenCCRM 2.117 ± 0.648 (131) 10/10 TT 0.025 MenCCRM 2.164 ± 0.364 (146) 10/10 0.25 MenCCRM 1.851 ± 0.326 (71) 10/10 2.5 MenCCRM 1.832 ± 0.348 (68) 10/10 from both neonatal and adult mice showed a significant 25 MenCCRM 1.881 ± 0.393 (76) 10/10 increase in the anti-PS4 IgG antibody response (P < 0.01) of DT low-dose-carrier-primed mice reimmunized with the conju- 0.025 MenCCRM 1.875 ± 0.419 (75) 10/10 gate and a significant decrease (P < 0.05) in the anti-PS4 IgG 0.25 MenCCRM 2.428 ± 0.396 (268) 10/10 response of high-dose-carrier-primed mice reimmunized 2.5 MenCCRM 2.382 ± 0.253 (241) 10/10 25 MenCCRM 1.996 ± 0.499 (99) 10/10 with the conjugate. Influence of TT priming on IgG subclass distribution of a Groups of 8 to 10 mice were immunized s.c. antipolysaccharide antibodies. Preimmunization with TT may b Mice were immunized s.c. with 0.5 ,ug of saccharide-protein conjugate on day 42 after immunization with the carrier. also influence the IgG subclass distribution of the anti- ' Expressed as percentage of a reference serum at day 21 after immuniza- polysaccharide antibody response. Results in Table 3 show tion with the conjugate vaccine. GMT, geometric mean titer (antilog). that upon low-dose carrier priming, the increased levels of d p < 0.01. P value was based on comparisons with controls not immunized were with TT or DT prior to vaccination with the conjugate vaccine. anti-MenC polysaccharide IgG antibodies mainly due to an increase in the IgGl subclass (P < 0.01). The IgG2b subclass increased to a lesser extent (P < 0.05), whereas the IgG3 antibody level was significantly decreased (P < 0.05). conjugate-induced anti-PS4 antibody response by both the No effect on the IgG2a antibody level was observed with IgM and IgG isotypes (Fig. 2). High-dose carrier priming low-dose carrier priming. The increase in IgGl antibodies resulted in suppression of the conjugate-induced anti-PS4 was also observed with PS4TT but not with the oligosaccha- IgG antibody response. The phenomena of enhancement and ride-protein conjugate MenCCRM. High-dose carrier prim- inhibition of anti-PS4 antibody formation after carrier priming showed suppression of all IgG subclasses (IgGl, IgG2b, ing therefore were operative in both adult and neonatal mice. and IgG3) (P < 0.05) with the PS4TT vaccine, whereas Neonatal mice were immunized a second time with PS4TT at suppression of the antipolysaccharide immune response to 9 weeks of age, and this immunization resulted in an 8.4-fold MenCTT (which induces IgGl, IgG2a, IgG2b, and IgG3 increase in anti-PS4 IgG antibodies in low-dose (0.025-,ug)- antibodies) was evident only for the IgG2a (P < 0.01) and carrier-primed mice. In nonprimed animals, the anti-PS4 IgG IgG2b antibodies. No effect of high-dose carrier priming was antibody response increased only 3.6-fold upon a second observed with the oligosaccharide-protein conjugate vaccine immunization with PS4TT (Fig. 2). In high-dose-carrier- which induced predominantly antipolysaccharide antibodies primed mice, no increase in either PS4 IgM or PS4 IgG of the IgGl subclass. antibodies could be observed upon secondary immunization In neonatal mice, the effects of low- and high-dose carrier with the conjugate (Fig. 2). The consequences of carrier priming on the antipolysaccharide IgG subclass distribution priming thus remained clear even after repeated immuniza- upon immunization with the PS4TT vaccine were similar to tion with the conjugate vaccines. Statistical analysis of data those found in adult mice (data not shown). VOL. 59, 1991 CARRIER PRIMING AFFECTS POLYSACCHARIDE ANTIBODY LEVELS 3507

TABLE 2. Effect of carrier priming on antipolysaccharide IgM 10,000 antibody level to saccharide-protein conjugates in adult mice 2 2 Primary No. of la 1.00 immunization Secondary Mean log1o titer responding 0 M- and amt immunizationb' SD (GMT)c micentota c:0 100 (jjg)a mice tested ;C, I* 1 1 (A Saline PS4TT 0.954 ± 0.279 (9) 8/9 (L CL .1 10- TT c 0.025 PS4TT 1.740 ± 0.265 (55) 6/6d a ac 0.25 PS4TT 1.663 ± 0.188 (46) 0l/lod 2.5 PS4TT 1.491 ± 0.252 (31) jollod 25 PS4TT 1.568 ± 0.344 (37) 8/8d 0 .025 .25 2.5 25 0 .025 .25 2.5 25 DT priming dose (jg tat. tox.J priming dose (pg tat. tox. 0.025 PS4TT 1.000 ± 0.338 (10) 7/8 FIG. 2. Dose-dependent influence of carrier priming on anti-PS4 0.25 PS4TT 1.322 ± 0.217 (21) 10/10 IgG and IgM antibody levels after two vaccinations with PS4TT in 2.5 PS4TT 1.113 ± 0.399 (13) 7/10 neonatal mice. Groups of 8 to 10 1-week-old mice were immunized 25 PS4TT 0.954 ± 0.431 (9) 8/10 with various amounts of TT (O to 25 ,ug per animal). Four weeks after immunization with the carrier, all mice were vaccinated with Saline MenCTT 2.364 ± 0.207 (231) 9/9 0.5 ,ug of PS4TT. Anti-PS4 IgM (left) was analyzed in individual TT serum samples at day 8 following immunization, and anti-PS4 IgG 0.025 MenCTT 2.472 ± 0.236 (297) 10/10 antibodies (right) were analyzed at day 28 after immunization (open 0.25 MenCTT 2.399 ± 0.143 (251) 10/10 bars). A second vaccination with the conjugate was given 4 weeks 2.5 MenCTT 1.857 ± 0.207 (72) 8/8 after primary immunization with the conjugate. Anti-PS4 IgM and 25 MenCTT 1.991 ± 0.201 (98) 10/10 anti-PS4 IgG antibodies were analyzed in individual serum samples DT on days 8 and 21, respectively, after the second vaccination with the 0.025 MenCTT 2.272 ± 0.127 (187) 10/10 conjugate (shaded bars). 0.25 MenCTT 2.410 ± 0.292 (257) 10/10 2.5 MenCTT 2.332 ± 0.301 (215) 10/10 25 MenCTT 2.232 ± 0.326 (171) 10/10 thetic peptides coupled to the carrier protein but not for Saline MenCCRM 1.397 ± 0.250 (25) 9/9 polysaccharides or oligosaccharides until now. Our data TT indicate that apart from enhancement of the antisaccharide 0.025 MenCCRM 1.176 ± 0.312 (15) 8/10 0.25 MenCCRM 1.114 ± 0.265 (13) 8/10 antibody response, suppression can also occur by carrier 2.5 MenCCRM 1.230 ± 0.250 (17) 9/10 priming when high-molecular-weight capsular polysaccha- 25 MenCCRM 1.041 ± 0.190 (11) 8/10 rides are coupled to proteins. In a number of other systems, DT it has been demonstrated that concomitant immunization 0.025 MenCCRM 1.362 ± 0.538 (23) 8/10 with carrier protein and conjugate vaccine does lead to a 0.25 MenCCRM 1.875 ± 0.378 (75) 10/10 higher antipolysaccharide response (6, 25). Enhancement of 2.5 MenCCRM 1.633 ± 0.449 (43) 10/10 the antipolysaccharide antibody response was observed in 25 MenCCRM 1.255 ± 0.350 (18) 9/10 our studies by low-dose carrier priming for PS4TT and a Groups of 8 to 10 mice were immunized s.c. MenCTT. It was observed that low-dose carrier priming b Mice were immunized s.c. with saccharide-protein conjugate, containing especially increased IgGl and IgG2b antipolysaccharide 0.5 ,ug of saccharide, at day 42 after immunization with the carrier. antibodies. Since both IgGl and IgG2b (IgGl > > IgG2b) are c Expressed as percentage of a reference serum at day 21 after immunization with the conjugate vaccine. GMT, geometric mean titer (antilog). subclasses restricted to a typical T-cell-dependent antibody d p < 0.01. P values were based on comparisons with controls not response in mice (29), it might be suggested that an increase immunized with TT or DT prior to vaccination with the conjugate vaccine. in carrier-specific T-cell help resulted in a more T-cell- dependent antibody response. Antibodies of the IgG3 subclass are predominantly induced in an immune response to DISCUSSION T-cell-independent type 2 antigens (such as polysaccharides) and were found in minor quantities upon vaccination with This study was performed to investigate whether carrier PS4TT and MenCTT. Antibodies of this subclass were priming affects the antibody response to polysaccharides observed to be decreased after low-dose carrier priming and when DT or TT is used as a carrier protein in saccharide- subsequent vaccination with MenCTT. This might indicate protein conjugate vaccines. It is known from earlier studies that the antipolysaccharide antibody response indeed with hapten-carrier and synthetic peptide-carrier conjugate switched to a more T-cell-dependent character. Remarkable vaccines that the antibody response to the hapten conjugated was that the doses of TT used for priming that caused to a protein can be either suppressed or enhanced by prior enhancement of the antipolysaccharide response induced by immunization with the carrier protein alone. Previous obser- TT-conjugated polysaccharide did not induce detectable vations in mice have shown that preimmunization with the anti-TT antibody levels after primary immunization (Fig. 1). carrier protein and a Haemophilus influenzae type b polysac- Increasing the priming dose of carrier to 2.5 ,ug per animal, charide-bovine serum albumin conjugate enhances the anti- which results in detectable anti-TT antibodies, abrogated the polysaccharide antibody response (24). Similar observations positive effect of carrier priming on the antipolysaccharide have been made by other investigators who used the same antibody response. In contrast to earlier observations, we saccharide and the pneumococcal serotype 6A capsular were able to show that using a high dose of TT (25 jig per polysaccharide coupled to other carrier proteins (1, 3, 6, 25, animal) administered as PS4TT or MenCTT for a priming 26). The phenomenon of epitope suppression has been dose suppressed the antipolysaccharide IgG antibody re- described for haptens, small peptides, and polymeric syn- sponse. In agreement with data in the literature on hapten- 3508 PEETERS ET AL. INFECT. IMMUN.

TABLE 3. Effect of carrier priming on IgG subclass distribution of antipolysaccharide antibodies Primary Mean log1o titer + SD (GMT) forc: immunization Secondary and amt immunization' (jj.g)a IgGl IgG2a IgG2b IgG3 Saline PS4TT 2.076 ± 0.857 (119) 0.778 ± 1.109 (6) 1.255 ± 0.728 (18) TT 0.025 PS4TT 2.759 ± 0.493 (574) 1.204 + 0.469 (16) 1.114 + 0.474 (13) 0.25 PS4TT 2.740 ± 0.480 (550) 1.415 ± 0.469 (26) 1.146 ± 0.393 (14) 2.5 PS4TT 1.462 ± 0.588 (29) 0 (1) 0.301 ± 0.401 (2) 25 PS4TT 0.778 ± 0.709 (6) 0 (1) 0.477 ± 0.487 (3) DT 0.025 PS4TT 2.149 ± 0.833 (141) 1.079 ± 0.865 (12) 1.041 ± 0.478 (11) 0.25 PS4TT 2.283 ± 1.276 (192) 1.255 + 0.582 (18) 1.146 ± 0.506 (14) 2.5 PS4TT 1.919 ± 1.273 (83) 1.041 ± 0.677 (11) 1.000 ± 0.906 (10) 25 PS4TT 2.021 ± 0.741 (105) 0.778 ± 0.792 (6) 1.362 ± 0.602 (23) Saline MenCTT 2.584 ± 0.459 (384) 1.301 ± 0.468 (20) 1.230 ± 0.418 (17) 2.152 ± 0.334 (142) TT 0.025 MenCTT 3.420 ± 0.340 (2,632) 1.114 ± 0.441 (13) 1.785 ± 0.382 (61) 1.380 ± 0.333 (24) 0.25 MenCTT 3.157 ± 0.371 (1,436) 0.778 + 0.558 (6) 1.230 ± 0.283 (17) 1.342 ± 0.290 (22) 2.5 MenCTT 2.393 ± 0.272 (247) 1.000 + 0.792 (10) 1.301 ± 0.559 (20) 1.813 ± 0.380 (65) 25 MenCTT 2.267 ± 0.322 (185) 0 (1) 0.301 + 1.123 (2) 1.612 ± 0.464 (41) DT 0.025 MenCTT 2.694 ± 0.299 (494) 1.380 ± 0.401 (24) 1.322 ± 0.742 (21) 2.053 ± 0.238 (113) 0.25 MenCTT 2.619 ± 0.344 (416) 1.079 ± 0.800 (12) 1.079 ± 0.497 (12) 1.839 ± 0.238 (69) 2.5 MenCTT 2.829 ± 0.525 (674) 1.643 ± 0.462 (44) 1.256 ± 0.621 (18) 2.100 ± 0.210 (126) 25 MenCTT 2.459 ± 0.429 (288) 0.778 ± 0.722 (2) 1.000 + 0.436 (10) 1.778 ± 0.410 (60) Saline MenCCRM 2.152 ± 0.613 (142) TT 0.025 MenCCRM 2.217 ± 0.367 (165) 25 MenCCRM 1.886 ± 0.344 (77) 2.5 MenCCRM 1.892 ± 0.328 (78) 25 MenCCRM 1.899 ± 0.386 (79) DT 0.025 MenCCRM 1.908 ± 0.410 (81) 0.25 MenCCRM 2.507 ± 0.420 (321) 2.5 MenCCRM 2.407 ± 0.246 (255) 25 MenCCRM 2.068 ± 0.545 (117) a Individual serum samples from 10 mice per experimental group were analyzed for antipolysaccharide IgG antibodies 3 weeks after immunization with the saccharide-protein conjugates. Mice were primed with various amounts of TT or DT (0 to 25 p.g per animal). b Six weeks after immunization with the carrier, mice were immunized with saccharide-protein conjugates, containing 0.5 ,ug of saccharide. c Given as percentage of a hyperimmune serum. GMT, geometric mean titer (antilog). Anti-PS4 IgG antibodies of the IgG2a subclass were not increased compared with antibodies in nonimmune sera when mice were immunized with PS4TT after carrier priming. Only anti-MenC polysaccharide antibodies of the IgGl subclass were enhanced compared with antibodies in nonimmune sera when mice were immunized with MenCCRM after carrier priming. carrier and peptide-carrier conjugate studies, we could not enced by high-dose carrier priming. No indication for either observe an influence of high-dose carrier priming on the low-dose-carrier-induced enhancement or high-dose-carrier- antisaccharide IgM antibody level. It has been suggested (8) induced inhibition of the antipolysaccharide antibody re- that the use of high-dose carrier priming led to antihapten sponse could be observed when a DT-MenCCRM combina- antibody responses with a T-cell-independent character. tion was used. In contrast to results in the study by Vella and Since most polysaccharides are T-cell-independent type 2 Ellis (30), no DT priming was necessary for the generation of antigens, the T-cell-independent character of the anti- an antipolysaccharide antibody response with an oligosac- polysaccharide antibody response under epitope-suppressed charide-mutant DT conjugate vaccine. CRM197 differs from conditions should be very clear in this study. However, DT at a single amino acid position (15). It has, however, analysis of the antipolysaccharide IgG subclass distribution been demonstrated that there is a high degree of antibody showed that all three subclasses (IgGl, IgG2b, and IgG3) cross-reaction (23) and T-cell cross-reactivity (5) between which could be induced by PS4TT were suppressed upon CRM197 and DT. The lack of epitope suppression in the high-dose carrier priming. With MenCTT, only IgG2a and DT-MenCCRM combination is therefore unlikely to be due IgG2b of the four contributing subclasses (IgGl, IgG2a, to minor antigenic differences between the protein used for IgG2b, and IgG3) were significantly suppressed. The obser- priming and the carrier protein of the conjugate. The absence vation that only IgG2a and IgG2b were suppressed by of epitope suppression might be due to the lower immuno- high-dose carrier priming was reported earlier by Herzen- genicity of DT compared with the immunogenicity of TT in berg et al. (17, 18). Our data do not support the hypothesis NIH/RIVM mice; after primary immunization of mice with that the hapten behaves as a T-cell-independent antigen, DT, we could hardly detect antibodies against DT even with since the IgG3 subclass, the antibody subclass preferentially the highest dose (25 R.g) of DT used (geometric mean titer, induced by T-cell-independent antigens, would not be influ- <1). Another explanation might be the difference in confor- VOL. 59, 1991 CARRIER PRIMING AFFECTS POLYSACCHARIDE ANTIBODY LEVELS 3509 mation and/or saccharide size of MenCCRM compared with immunized only with TT, whereas in humans, TT is admin- those of MenCTT and PS41T, which are prepared according istered together with diphtheria, pertussis, and polio vac- to a similar coupling procedure and which contain the whole cines in a combination vaccine. Two studies in mice have capsular polysaccharide. To our knowledge, there are two shown that the use of Bordetella pertussis vaccine and its other studies in which the influence of high-dose DT priming adjuvant components can modulate immune suppression on the antibody response to a peptide-DT or a saccharide- (17, 31). Thus, it is possible that the pertussis component in protein conjugate was reported (1, 14). In one study, DT also the combination vaccine diminishes the potential for TT to failed to induce epitope suppression, whereas when TT was cause epitope suppression in humans. Clearly, more studies used as a carrier protein to the same peptide, suppression are needed to address the effect of carrier priming on the was induced (14). In the study by Anderson (1), an improved outcome of saccharide-protein conjugate vaccinations as antipolysaccharide antibody response occurred upon vacci- well as to develop different ways of circumventing epitope nation with the H. influenzae type b oligosaccharide-mutant suppression. DT conjugate in rabbits primed with 100 ,ug of carrier protein on aluminum phosphate adjuvant compared with nonprimed ACKNOWLEDGMENTS animals. We are greatly indebted to B. van Hilten for his invaluable In this study, we have used two TT-polysaccharide con- biotechnical assistance and to I. van der Tweel, Department of jugates and one CRM197-oligosaccharide conjugate, which Biostatistics, University School of Medicine, Utrecht, The Nether- is apart from carrier protein and is also constructed differ- lands, for performing the statistical analysis. ently from the two TT conjugates. Therefore, variables such This study was supported by grant 28-1246 from the Prevention as the nature of the polysaccharide, the protein/polysaccha- Fund. ride ratio, and unidentified factors related to the particular construction and shape may have contributed to the inhibi- REFERENCES tory effects observed. No general rule on the effects of 1. Anderson, P. 1983. Antibody responses to Haemophilus influ- carrier priming can therefore be deduced from these exper- enzae type b and diphtheria toxin induced by conjugate of iments. The data do indicate, however, that high-dose car- oligosaccharides of the type b capsule with the non-toxic protein CRM197. Infect. Immun. 39:233-238. rier priming may cause significant suppressive effects on the 2. Anderson, P. W., M. E. Pichichero, R. A. Insel, R. Betts, R. Eby, antipolysaccharide IgG antibody response induced by con- and D. H. Smith. 1986. Vaccines consisting of periodate cleaved jugated polysaccharide. oligosaccharides from the capsule of H. influenzae type b The mechanism responsible for suppression of the anti- coupled to a protein carrier: structural and temporal require- hapten antibody response to hapten-carrier conjugates upon ments for priming in the human infant. J. Immunol. 137:1181- high-dose carrier priming is still unclear. Involvement of 1186. both T suppressor cells (13, 17-19, 27, 28) and carrier- 3. Barington, T., K. Kristensen, J. Henrichsen, and C. Heilmann. specific B memory cells (13, 19, 27, 28) has been reported. 1991. Influence of prevaccination immunity on the human B Herzenberg et al. (17, 18) postulated activation of T suppres- lymphocyte response to a Haemophilus influenzae type b conjugate vaccine. Infect. Immun. 59:1057-1064. sor cells. Schutze et al. (27, 28) showed the involvement of 4. Beuvery, E. C., F. van Rossum, and J. Nagel. 1982. Comparison B memory cells in the phenomenon of epitope suppression. of the induction of immunoglobulin M and G antibodies in mice They argue that because of prior immunization with the with purified pneumococcal type 3 and meningococcal group C carrier protein, a predominance of carrier-specific B-cell polysaccharides and their protein conjugates. Infect. Immun. clones exists that competitively decreases the interaction 37:15-22. between hapten-specific B cells and the hapten at the time of 5. Bixler, G. S., and S. Pillai. 1989. The cellular basis of the conjugate immunization. Another possible explanation immune response by conjugate vaccines. Contrib. Microbiol. might be that circulating antibodies against the carrier pro- Immunol. 10:18-47. tein are responsible for scavenging of the antigen or the 6. Chu, C., R. Schneerson, J. B. Robbins, and S. C. Rastogi. 1983. formation of immune complexes which inhibit Further studies on the immunogenicity of Haemophilus influ- the immune enzae type b and pneumococcal type 6A polysaccharide-protein response. conjugate. Infect. Immun. 40:245-256. Recent studies with H. influenzae type b polysaccharide- 7. Cryz, S. J., J. C. Sadoff, E. Furer, and R. Germanier. 1986. or oligosaccharide-protein conjugates show that these con- Pseudomonas aeruginosa polysaccharide-tetanus toxoid conju- jugate vaccines are highly efficacious in preventing H. influ- gate vaccine: safety and immunogenicity in humans. J. Infect. enzae type b disease in infants and young children (32). Dis. 154:682-688. These conjugate vaccines as well as new conjugate vaccines 8. Di John, D., J. R. Torres, J. Murillo, E. A. Herrington, S. S. under development will therefore be used on a wide scale in Wasserman, M. J. Lasonsky, D. Sturker, and M. M. Levine. the near future in order to protect children against infections 1989. Effect of priming with carrier on response to conjugate by several encapsulated bacteria. In most vaccine. Lancet ii:1415-1418. cases, DT or TT 9. Einhorn, M. S., G. A. Weinberg, E. L. Anderson, G. A. will be used as the carrier protein in the conjugate vaccine. Weinberg, P. D. Granoff, and D. M. Granoff. 1986. Immunoge- Because of the growing emphasis on TT immunizations in nicity in infants of H. influenzae type b polysaccharide in a different health programs, it is important to be aware of a conjugate vaccine with Neisseria meningitidis outer-membrane potential negative effect of the vaccination status of humans protein. Lancet ii:299-302. in their response to those conjugate vaccines. However, 10. Etlinger, H. M., A. M. Felix, D. Lillessen, E. D. Haemer, M. extrapolation of our observed data to the human situation Just, J. R. L. Pink, F. Sinigaglia, D. Sturchier, B. Tacaks, A. should be made cautiously for several reasons. First, the Trzeciak, and H. Matile. 1988. Assessment in humans of a induction of epitope suppression in mice required relatively synthetic peptide-based vaccine against the sporozoite stage of high doses of carrier the human malaria parasite, Plasmodium falciparum. J. Immu- protein relative to body weight (25 ,ug nol. 140:626-633. per animal), since only 10 to 15 ,ug is used for immunizations 11. Etlinger, H. M., D. Giliessen, H. W. Lahm, H. Matile, H. J. in humans. However, we do not know whether sensitization Schonfeld, and A. Trzeciak. 1990. Use of prior vaccinations for by repeated injections with lower doses of TT can also the development of new vaccines. Science 249:423-425. induce epitope suppression. Second, in our study, mice were 12. Fleiss, J. L. 1986. The design and analysis of clinical experi- 3510 PEETERS ET AL. INFECT. IMMUN.

ments. John Wiley & Sons, Chicester, England. 24. Schneerson, R., 0. Barrera, A. Sutton, and J. B. Robbins. 1980. 13. GalIeli, A., and B. Chariot. 1990. Clonal anergy of memory B Preparation, characterization and immunogenicity of Haemoph- cells in epitope specific regulation. J. Immunol. 145:2397-2405. ilus influenzae type b polysaccharide-protein conjugates. J. 14. Gaur, A., K. Arunan, 0. Singh, and G. P. Talwar. 1990. Bypass Exp. Med. 152:360-376. by an alternate carrier of acquired unresponsiveness to hCG 25. Schneerson, R., J. B. Robbins, C. Chu, A. Sutton, W. Vann, upon repeated immunization with tetanus-conjugated vaccine. J. C. Vickers, W. T. London, B. Curfman, M. C. Hardegree, J. Int. Immunol. 2:151-155. Shiloach, and S. C. Rastogi. 1984. Serum antibody responses of 15. Giannini, G., R. Rappuoli, and G. Ratti. 1984. The amino juvenile and infant rhesus monkeys injected with Haemophilus sequence of two non-toxic mutants of diphtheria toxin: CRM45 influenzae type b and pneumococcus type 6A capsular polysac- and CRM197. Nucleic Acids Res. 12:4063-4069. charide-protein conjugates. Infect. Immun. 45:582-591. 16. Granoff, D. M., E. G. Boies, and R. S. Munson. 1984. Immuno- 26. Schneerson, R., J. B. Robbins, J. C. Parke, C. Bell, J. J. genicity of H. influenzae type b polysaccharide diphtheria Schiesselman, A. Sutton, Z. Wang, G. Schiffman, A. Karpas, and conjugate vaccines in adults. J. Pediatr. 105:22-27. J. Shiloach. 1986. Quantitative and qualitative analysis of serum 17. Herzenberg, L. A., and T. Tokuhisa. 1982. Epitope-specific antibodies elicited in adults by H. influenzae type b and pneu- regulation. I. Carrier-specific induction of suppression for IgG mococcus 6A capsular polysaccharide-TT conjugates. Infect. anti-hapten antibody responses. J. Exp. Med. 155:1730-1740. Immun. 52:519-528. 18. Herzenberg, L. A., T. Tokuhisa, and K. Hayakawa. 1983. 27. Schutze, M. P., E. Deriaud, G. Przewlocki, and C. LeClerc. Epitope-specific regulation. Annu. Rev. Immunol. 1:609-632. 1989. Carrier-induced epitopic suppression is initiated through 19. LeClerc, C., M. P. Schutze, E. Deriaud, and G. Przewlocki. clonal dominance. J. Immunol. 142:2635-2640. 1990. The in vivo elimination of CD4+ T cells prevents the 28. Schutze, M. P., C. LeClerc, M. Jolivet, F. Audibert, and L. induction but not the expression of carrier-induced epitopic Chedid. 1985. Carrier-induced epitopic suppression, a major suppression. J. Immunol. 145:1343-1349. issue for synthetic vaccines. J. Immunol. 135:2319-2322. 20. Lise, L. D., D. Mazier, M. Jolivet, F. Audibert, L. Chedid, and 29. Slack, J., G. P. Der Balian, M. Nahm, and J. M. Davie. 1980. D. Schlesinger. 1987. Enhanced epitopic response to a synthetic Subclass restriction of murine antibodies. II. The IgG plaque human malarial peptide by preimmunization with tetanus toxoid forming cell response to thymus independent type 1 and type 2 carrier. Infect. Immun. 55:2658-2661. antigens in normal mice and mice expressing an X-linked 21. Peeters, C. C. A. M., A. J. Tenbergen-Meekes, D. E. Evenberg, immunodeficiency. J. Exp. Med. 151:853-862. J. T. Poolman, B. J. M. Zegers, and G. T. Rikers. 1991. A 30. Vella, P. P., and R. W. Ellis. 1991. Immunogenicity of Hae- comparative study of the immunogenicity of pneumococcal type mophilus influenzae type b conjugate vaccines in infant rhesus 4 polysaccharide- and oligosaccharide-tetanus toxoid conju- monkeys. Pediatr. Res. 29:10-13. gates in adult mice. J. Immunol. 146:4308-4314. 31. Vogel, F. R., C. LeClerc, M. P. Schutze, M. Jolivet, F. Audibet, 22. Peeters, C. C. A. M., A. J. Tenbergen-Meekes, J. T. Poolman, T. W. Klein, and L. Chedid. 1987. Modulation of carrier-induced B. J. M. Zegers, and G. T. Rijkers. Submitted for publication. epitopic suppression by Bordetella pertussis components and 23. Porro, M., M. Saletti, L. Neucioni, L. Tagliaferi, and L. Marsili. muramyl peptide. Cell. Immunol. 107:40-51. 1980. I. Immunogenic correlation between cross-reacting mate- 32. Ward, J., and S. Cochi. 1988. Haemophilus influenzae, p. rial (CRM197) produced by a mutant C. diphtheria and diphthe- 300-332. In S. A. Plotkin and E. A. Mortimer (ed.), Vaccines. ria toxoid. J. Infect. Dis. 142:716-724. The W. B. Saunders Co., Philadelphia.