Regulatory T Cells in the Antibody Response to Haemophilus ...

2 downloads 0 Views 79KB Size Report
zyme-linked spot-forming cell assay described by Sedgwick and Holt (19). Poly- ..... Douglas, R. M., J. C. Paton, S. J. Duncan, and D. J. Hansman. 1983. Antibody ... R. Bell and G. Torrigiani, (ed.), Towards better carbohydrate vaccines. John.
INFECTION AND IMMUNITY, Feb. 1999, p. 789–793 0019-9567/99/$04.0010 Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Vol. 67, No. 2

Regulatory T Cells in the Antibody Response to Haemophilus influenzae Type b Polysaccharide MIJKE A. BREUKELS, GER T. RIJKERS,* MARLEEN M. VOORHORST-OGINK, BEN J. M. ZEGERS Department of Immunology, Wilhelmina Children’s Hospital, Utrecht, The Netherlands Received 12 August 1998/Returned for modification 8 October 1998/Accepted 16 November 1998

An in vitro culture system for the induction of an antipolysaccharide response was used to study the cellular interactions which determine the magnitude and nature of this B-lymphocyte response. Healthy adult volunteers were vaccinated with the Haemophilus influenzae type b polysaccharide (PRP)-tetanus toxoid (TT) conjugate vaccine. Optimal in vitro anti-PRP and anti-TT antibody responses were obtained when B cells were cultured with equal amounts of T cells. The in vitro response is antigen dependent and antigen specific. Culturing with PRP mixed with TT in the presence of T cells induces the highest number of anti-PRP antibody-secreting cells (ASC) (128.4 3/4 15.9 [geometric mean 3/4 standard deviation] immunoglobulin M [IgM] anti-PRP ASC/106 cells; 9.3 3/4 7.6 IgG anti-PRP ASC/106 cells). Culturing without T cells induced no anti-PRP ASC; culturing with only PRP, in the presence of T cells, yielded low numbers of anti-PRP ASC (3.7 3/4 5.2 IgM anti-PRP ASC/106 cells and 1.2 3/4 2.2 IgG anti-PRP ASC/106 cells). Transwell studies showed that the requirements for the antibody response against the polysaccharide are different from those of an antiprotein response. Cytokines formed as a consequence of contact between protein-specific B and T cells were on their own not sufficient to activate TT-specific B cells (8.4 3/4 1.4 anti-TT ASC/106 cells); direct contact between T and B cells appeared to be an absolute requirement. However, physical contact between B and T cells in one compartment of the Transwell system resulted in the release of soluble factors able to stimulate B cells in the other compartment to secrete antipolysaccharide antibodies (164 3/4 1.6 anti-PRP ASC/106 cells).

molecules. The peptide-MHC class II complex on B cells is then able to activate specific T cells. In this interaction, CD40CD40L functions as an essential ligand-receptor pair which provides a second activation signal (13). Because polysaccharide processing does not occur (1), this model is not valid for TI-2 antigens. In vivo, the first step in B-cell activation by polysaccharides occurs via ligation and cross-linking of mIg. A second activation signal is probably provided by coligation of complement receptor 2 (CR2, CD21). Polysaccharide-C3d complexes, formed by complement activation through the alternative pathway, have the ability to bind to CD21 (9). The mechanism of coligation of mIg and CD21 may account for the fact that antigen-specific T cells are not strictly required for induction of an antipolysaccharide B-cell response. While the in vitro B-cell response against TI-2 antigens can be induced in the absence of T cells, the presence of T cells augments the magnitude of the response (17). The T cells that mediate this function have been termed amplifier cells, to distinguish them from helper T cells in the TD antibody response to protein antigens (3). The in vivo antipolysaccharide antibody response induced by polysaccharide-protein conjugates exhibits the characteristics of a TD antibody response. In such an antibody response, the role of T helper cells and the specificity of these cells are still unclear. To investigate the cellular interactions which determine the magnitude and nature of the antipolysaccharide antibody response, we used a previously described in vitro culture system for restimulation of in vivo primed human B lymphocytes (15). Using this system, we showed that an in vitro anti-PRP antibody response can be induced in human B cells derived from in vivo-primed individuals. Vaccination with H. influenzae type b polysaccharide covalently linked to tetanus toxoid (PRPTT) is required to obtain a positive in vitro anti-PRP antibody re-

The defense against infections with encapsulated bacteria such as Haemophilus influenzae type b and Streptococcus pneumoniae depends primarily on the ability to produce antibodies against the capsular polysaccharides of these microorganisms. The immune response against these antigens (categorized as T-cell-independent type 2 [TI-2]) has several characteristics. There is no memory formation (11), the isotypes used are preferentially immunoglobulin M (IgM) and IgG2, and idiotype use of anti-TI-2 antibodies is restricted (10). Furthermore, responsiveness to TI-2 antigens develops relatively late in life (5, 6, 11), implying that children up to the age of 18 to 24 months generally are less able to produce antipolysaccharide antibodies and thus are more susceptible to infections with these encapsulated bacteria. Polysaccharide-based vaccines are not effective in this age group (2, 11). Coupling of polysaccharides to carrier proteins converts the antipolysaccharide response to a response with a T-cell-dependent (TD) character. Polysaccharide-protein conjugate vaccines are able to induce antipolysaccharide antibodies in 2- to 3-month-old children. Moreover, H. influenzae type b polysaccharide (polyribosyl ribitol phosphate [PRP])-protein conjugate vaccines have proven to be clinically effective during infancy, virtually eliminating invasive H. influenzae type b disease (4, 7). The mechanism by which these conjugate vaccines induce T- and B-lymphocyte activation remains a matter of debate. TD protein antigens are bound and internalized by the antigen receptor on B cells (mIg) and reexpressed as processed peptides in major histocompatibility complex (MHC) class II * Corresponding author. Mailing address: Dept. of Immunology, Wilhelmina Children’s Hospital, Nieuwe Gracht 137, 3512 LK Utrecht, The Netherlands. Phone: 31-30-2320911. Fax: 31-30-2320712. E-mail: [email protected]. 789

790

BREUKELS ET AL.

INFECT. IMMUN.

sponse (15). The in vitro generation of anti-PRP antibodysecreting cells (ASC) was shown to be T-cell dependent, antigen dependent, and antigen specific. In the present study, we further investigated the cellular interactions between B and T cells which determine the magnitude and nature of the antipolysaccharide antibody response. We therefore modified the culture system, using the Transwell system to physically separate the cell populations. The results show that T- and B-cellderived soluble factors are able to stimulate antigen-primed B cells to secrete antipolysaccharide antibodies and that physical contact between T and B cells is not absolutely required for antipolysaccharide B-cell differentiation. MATERIALS AND METHODS Immunization. Healthy adult volunteers (median age, 30 years; range, 22 to 57 years) were given an intramuscular immunization with a full dose of PRPTT (Act-Hib; Pasteur Me´rieux, Lyon, France) which contains 10 mg of the H. influenzae type b polysaccharide covalently linked to tetanus toxoid (TT; 624 mg). Blood samples were collected 3 to 4 weeks after immunization. In The Netherlands, children are routinely immunized at the ages of 3, 4, 5, and 12 months with diphtheria-pertussis-tetanus-poliomyelitis vaccine. Booster immunizations with diphtheria-tetanus-poliomyelitis vaccine are given at the ages of 4 and 9 years. All adult donors used in this study were vaccinated according to this schedule during childhood. Lymphocyte preparations. Peripheral blood mononuclear cells were isolated from heparinized blood (100 ml) from healthy adult donors by density gradient centrifugation on Ficoll-Isopaque (1.077 g/cm3; Pharmacia, Uppsala, Sweden) at 1,000 3 g for 20 min. The cell suspension was washed twice with MEM-Tris (Tris-buffered minimal essential medium; Gibco, Grand Island, N.Y.). T cells were separated from non-T cells by rosetting with 2-aminoethylisothiouronium bromide (Sigma Chemical Co., St. Louis, Mo.)-treated sheep erythrocytes. The non-T-cell fraction was then depleted of monocytes by treatment with iron carbonyl. After these procedures, the monocyte-depleted non-T-cell fraction contained an average of 40% B cells (25 to 72%) and less than 1% T cells. A fixed concentration of 5% monocytes was used in all experiments by mixing appropriate numbers of monocyte-depleted non-T cells and non-T cells not depleted for monocytes. Culture conditions. Cultures were set up to consist of a 1:1 ratio of T cells to non-T cells, and each culture was supplemented with 5% monocytes (15). A total of 2 3 106 cells in a volume of 2 ml was cultured at 37°C, 100% humidity, and 5% CO2. The medium consisted of RPMI 1640 (Flow Laboratories, Irvine, England) supplemented with 200 mg of glutamine, 100 IU of penicillin, and 100 mg of streptomycin per ml. 1025 M 2-mercaptoethanol, and 10% heat-inactivated pooled (five donors) human AB serum. After culturing for 6 days, cells were washed once in MEM-Tris supplemented with 1% bovine serum albumin (BSA; Organon Teknika, Oss, The Netherlands). Cells were then resuspended in 1.0 ml RPMI 1640 supplemented with 200 mg of glutamine per ml, 100 IU of penicillin per ml, 100 mg of streptomycin per ml, and 10% heat-inactivated fetal calf serum (Flow Laboratories). In some experiments the monocyte-depleted non-T-cell fraction was physically separated from the T-cell–5% monocyte fraction by the use of the Transwell system wells (Costar, Cambridge Mass.), 6.5 mm wide with pore size of 0.4 mm. In preliminary experiments, it was demonstrated that the Transwell system allowed for transfer of biologically active cytokines across the membrane (data not shown). After 6 days of culture, cells from the various compartments were analyzed separately for specific ASC. Determination of anti-TT and anti-PRP ASC. B cells secreting anti-TT or anti-PRP IgG or IgM antibody were enumerated by a modification of the enzyme-linked spot-forming cell assay described by Sedgwick and Holt (19). Polyvinyl chloride 96-well microtiter plates (Flow) were coated overnight at 4°C with 100 ml of tyramine-coupled PRP (5 mg/ml) in 0.9% NaCl or TT (10 mg/ml) in bicarbonate buffer, pH 9.6. Plates were also prepared for the detection of all IgGor IgM-secreting lymphocytes by coating with goat anti-human IgG or IgM (Tago Inc., Birmingham, Ala.) at a 1:1,000 dilution in bicarbonate buffer (pH 9.6) overnight at 4°C. All plates were then blocked following three washes with phosphate-buffered saline (PBS) with 1% BSA in MEM for 30 min at 37°C. The cultured cells were added to the coated wells in serial dilution starting with 200,000 cells per well and the plates were incubated for 3 to 4 h at 37°C, 100% humidity, and 5% CO2. The wells were then washed with PBS to remove all cells and incubated with 100 ml of alkaline phosphatase-conjugated goat anti-human IgM or IgG (Tago) at a dilution of 1:1,000 in PBS–1% BSA for 2 h at 37°C. After extensive washing with PBS plus 0.05% (wt/vol) Tween 20, 100 ml of the alkaline phosphatase-conjugated substrate 5-bromo-4-chloro-3-indolylphosphate (0.1 mg/ ml) in 1 M 2-amino-2-methyl-1-propanol buffer, pH 10.25 (Sigma), containing 5 mM MgCl2, 0.01% Triton X405 (Sigma), 0.01% sodium azide, and 0.6% 36°C gelling agarose (Sigma) was added to each well. Blue spots of ASC were counted after overnight incubation at room temperature in an inverted microscope (Zeiss, Oberkochen, Germany). ASC were expressed as the number per 106

FIG. 1. In vitro anti-PRP response of blood B cells obtained 3 to 4 weeks after vaccination with PRPTT. B cells (0.5 3 106/ml) were cultured with an equal number of irradiated T cells and 5% monocytes in medium containing 10% heat-inactivated pooled human AB serum and various antigens as indicated. After 6 days of culture, IgG and IgM anti-PRP ASC were determined with a spot-forming cell assay and are expressed per 106 lymphocytes. Final concentrations: PRP, 5 ng/ml; TT, 1.5 mg/ml; PRPTT, 0.5 mg of PRP and 1.2 mg of TT/ml. The data are from 21 independent experiments with different donors. Not all antigens were included in all experiments.

input cells. The specificity of the ASC was assessed by addition of excess soluble antigen (50 mg/ml) during incubation of cultured cells on antigen-coated plates. This procedure resulted in a .80% reduction of the number of spots, whereas remaining spots were significantly smaller than in noninhibited conditions. Antigens. The antigens used for in vitro cultures were PRP (final concentration, 5 ng/ml; Pasteur Me´rieux, Lyon, France) (15, 16), TT (1.5 mg/ml; National Institute of Public Health and Environment, Bilthoven, The Netherlands), and PRPTT conjugate at final PRP and TT concentrations of 0.5 and 1.2 mg/ml, respectively. Two TT peptides encompassing T-helper epitopes were synthesized (14): 156 (corresponding to TT positions 830 to 843, QYIKANSKFIGITE) and 158 (corresponding to TT positions 947 to 967, FNNFTVSFWLRVPKVS ASHLE). These peptides were used at a concentration of 7.5 mg/ml. Statistics. Data are presented as geometric means 6 standard deviation. Differences in numbers of ASC generated under various culture conditions were calculated by nonparametric, paired, two-sided t test. P values of ,0.05 are considered significant.

RESULTS In vitro induction of the anti-PRP (and anti-TT) antibody response. Adult volunteers were given a single dose of the PRPTT vaccine (Act-Hib, Pasteur Me´rieux); 3 to 4 weeks later, peripheral blood mononuclear cells were isolated and purified B cells, mixed with equal numbers of T lymphocytes, were cultured and stimulated with various forms of PRP (Fig. 1). In vitro culturing with PRP alone induced low numbers of IgM (3.7 3/4 5.2) and IgG (1.2 3/4 2.2) anti-PRP ASC/106 cells, comparable to ASC numbers detected in cells cultured in medium only (2.8 3/4 3.8 and 2.0 3/4 2.7, respectively). Restimulation with PRPTT induced 23.3 3/4 13.9 IgM anti-PRP ASC/106 cells and 4.9 3/4 7.3 IgG anti-PRP ASC/106 cells. Comparable ASC responses were obtained in cultures stimulated with PRP plus TT (TT in the same amounts as present in the conjugate) (128.4 3/4 15.9 IgM anti-PRP ASC/106 cells and 9.3 3/4 7.6 IgG anti-PRP ASC/106 cells; not significantly different [P 5 0.2783 and P 5 0.3594, respectively, in Wilcoxon matched-pairs signed-ranks test]) from PRPTT cultures (Fig. 1). From cells cultured with TT only, we obtained variable numbers of IgM anti-PRP ASC (73 3/4 1.8 anti-PRP ASC/106 cells) and fewer IgG anti-PRP ASC (2.4 3/4 4.9 anti-PRP ASC/106 cells). Anti-TT ASC were obtained only after culturing with TT, either free, mixed with PRP, or conjugated to PRP. The response to TT consisted predominantly (87, 78, and 90%, respectively) of IgG ASC (Fig. 2). It thus can be concluded that an optimal in vitro antibody

VOL. 67, 1999

T CELLS IN THE ANTIPOLYSACCHARIDE ANTIBODY RESPONSE

791

FIG. 2. In vitro IgG anti-TT response of blood B cells obtained 3 to 4 weeks after vaccination with PRPTT. See the legend to Fig. 1 for details.

response to PRP also requires stimulation with TT. There is, however, no absolute need for polysaccharide and protein to be physically coupled. In addition to PRP and TT, T cells are required for an optimal in vitro anti-PRP B-cell response (reference 15 and Fig. 3). Relationship between the in vitro-induced anti-PRP response and T-cell proliferation. PRPTT vaccination in adults actually is a booster vaccination with respect to the TT component. The in vitro proliferative T-cell response induced by TT (Table 1) is comparable in magnitude to that induced by a T-cell mitogen such as pokeweed mitogen. Because of the magnitude of this T-cell response, the question arose as to whether the T cells that augment the anti-PRP B-cell response are antigen (TT) specific, or if the T-cell proliferation is due to a nonspecific activation process. To address this question, TT peptides 156 and 158, encompassing two different universal T-helper epitopes, were synthesized (14). Although these peptides are claimed to be universally antigenic irrespective of MHC haplotype, both induced T-cell proliferation above background values in only 50% of vaccinated donors tested (Table 1). An association was found between the ability of a given TT peptide to induce T-cell proliferation and to support the in vitro anti-PRP antibody response (Table 2). These data show that TT peptide-specific T-cell activation is able to support the anti-PRP B-cell response. Interaction between T and B lymphocytes in the anti-PRP antibody response. To characterize in more detail the T-cell help necessary for an anti-PRP B-cell response, Transwell studies were set up. In this system, it is possible to physically separate different cell populations without disrupting diffusion of antigen and soluble mediators. Culturing of B lymphocytes with only PRP induces no or minimal IgM anti-PRP ASC (data not shown). Also, no anti-

FIG. 3. In vitro anti-PRP (E) and anti-TT (F) responses of blood B cells under various culture conditions in the Transwell system. Lymphocytes were obtained 3 to 4 weeks after vaccination with PRPTT. In culture a, B cells (0.7 3 106) were grown in medium containing 10% heat-inactivated pooled human AB serum and PRP plus TT. In the control experiment (culture b), B cells (0.7 3 106) were grown with an equal number of irradiated T cells and 5% monocytes in medium containing 10% heat-inactivated pooled human AB serum and PRP plus TT. By means of a Transwell system, B cells (0.7 3 106) were also grown separately from T cells (0.7 3 106) with 5% monocytes (culture c). d and e, results of culturing B cells (0.7 3 106) separated by the Transwell system from the lower compartment containing B cells (0.7 3 106) and an equal number of irradiated T cells and 5% monocytes. After 6 days of culture, anti-PRP as well as anti-TT ASC were determined with a spot-forming cell assay and are expressed per 106 lymphocytes. All cultures were stimulated with PRP (5 ng/ml) plus TT (1.5 mg/ml). *, compartment from which the ASC are depicted. The data represent six experiments conducted separately with six different donors. Note that not all combinations were included in every experiment.

PRP antibody response is seen when the B cells (lower compartment) are cultured with PRP plus TT in the absence of T cells in the upper compartment (Fig. 3, culture a). Culturing of B and T lymphocytes together with 5% antigen-presenting cells in one compartment, with only PRP as the antigen, induces a few IgM anti-PRP ASC (Fig. 1). In vitro culture of B and T lymphocytes in the lower compartment with PRP and TT as antigens induces both anti-PRP (geometric mean, 771 3/4 2.1 ASC/106 cells) and anti-TT ASC (601 3/4 1.7) (Fig. 3, culture b). Culturing the B cells physically separated from the T cells and the monocytes, however, elicits no anti-TT ASC (6.7 3/4 1.2) and very few anti-PRP ASC (22 3/4 1.9) (Fig. 3, culture c). When B and T cells are cultured together with 5% monocytes in the lower compartment and B cells only in the upper compartment of the Transwell system, striking results are obtained. The lower compartment shows, as expected, both antiPRP (1,922 3/4 2.7) and anti-TT ASC (1,373 3/4 1.3) (Fig. 3,

TABLE 1. T-cell proliferative responses upon stimulation with TT and TT peptides Antigen

Antigen concn (mg/ml)

None (medium) Pokeweed mitogen Peptide 156 Peptide 158 TT

5 7.5 7.5 1.5

Mean [3H]thymidine incorporation (1,000 cpm; n 5 4) for donor: 1

2

3

4

5

0.6 21 9.7 1.8 61

0.2 42 0.5 6.2 57

1.2 22 2.0 1.8 43

0.3 15 3.1 2.8 70

0.5 24 0.6 0.7 25

6

0.4 35 0.8 1.1 101

7

8

9

0.3 49 0.3 2.5 82

0.3 14 0.3 0.3 12

0.5 57 4.2 28 78

792

BREUKELS ET AL.

INFECT. IMMUN.

TABLE 2. Relationship between the in vitro anti-PRP response of blood B cells obtained 3 to 4 weeks after vaccination with PRPTT and in vitro-induced T-cell proliferationa Antigen

Antigen concn (mg/ml)

None (medium) PRP PRPTT PRP 1 TT PRP 1 peptide 156 PRP 1 peptide 158 TT

0.5 0.5 0.5 1 1.2 0.5 1 7.5 0.5 1 7.5 1.5

a

Anti-PRP

Anti-TT

T-cell proliferation mean [3H]thymidine incorporation (1,000 cpm; n 5 4)

5 ,5 88 165 114 ,5 221

,5 ,5 316 497 ,5 ,5 1,340

1.8 ,1 13.5 48.5 23 ,1 58

Response (ASC/106 cells)

Data are from one representative donor out of four independent experiments.

culture d). Assessment of ASC from the upper compartment (B cells only) shows that significant higher numbers of antiPRP ASC are generated as in culture c, but the number of anti-TT ASC remains low (Fig. 3, culture e). In a series of 11 independent experiments, 51 3/4 1.7 anti-PRP ASC and 8.1 3/4 1.3 anti-TT ASC (P 5 0.0381, two-tailed Mann-Whitney test) were generated. This result suggests that soluble mediators derived from T cells can support an antipolysaccharide antibody response by B lymphocytes, while an anti-TT antibody response is not elicited under these circumstances. It can be concluded that optimal in vitro B-cell activation by polysaccharides requires T-lymphocyte activation. However, there seems to be a distinct difference between the T-cell help necessary for B-cell activation by polysaccharides and the Tcell help needed in an antiprotein response. In the latter situation, physical interaction between T and B lymphocytes is required; for a polysaccharide response, soluble mediators are sufficient. DISCUSSION The proposed cellular interaction mechanism for proteinconjugated polysaccharides combines the models for TI-2 and TD B-cell activation. The B-cell receptor of a polysaccharidespecific B cell recognizes the polysaccharide component and binds and internalizes the polysaccharide-protein conjugate. Intracellularly, the protein is processed and peptide fragments are subsequently presented in MHC class II molecules. Peptide-specific T cells become activated and produce B-cell-stimulating cytokines. In this way, the polysaccharide-specific B cell receives T-cell help from the peptide (protein)-specific T cell (18, 20). It is not known whether it is sufficient to have activated (peptide-specific) T cells, producing B-cell stimulatory cytokines, in the proximity of polysaccharide-specific B cells or, alternatively, whether physical contact between the two cell types is necessary. From our experiments, it can be concluded that the presence of TT-activated T cells is needed for the formation of antipolysaccharide ASC. This is in accordance with the above-described model for the working mechanism of conjugate vaccines: the activated TT peptide-specific T cells produce cytokines that stimulate polysaccharide-specific B cells. As we have shown, this T-cell help can be induced by culturing with PRP plus TT, with PRPTT, and with PRP mixed with TT peptides. The number of anti-PRP ASC generated correlates with the magnitude of the T-cell proliferative response to these antigens (Table 2). The highest numbers of anti-PRP ASC are obtained after culturing with PRP mixed with TT or conjugated to TT (Fig. 1). We do find IgM anti-PRP ASC when culturing with TT

only, even though at the time of sampling (3 to 4 weeks after vaccination with PRPTT) no anti-PRP (or anti-TT) ASC are found in a direct ASC assay (data not shown). We presume that at the time of sampling, the peripheral blood of the vaccinees contains cells that are committed to become anti-PRP ASC. They apparently need only activation provided by cytokines secreted by activated (anti-TT) T cells. In vivo, admixture of polysaccharide and protein does not change the nature of the antipolysaccharide antibody response. For an effective conjugate vaccine, the polysaccharide and the protein must be physically coupled (8). In vitro, however, mixtures of polysaccharide and protein are as effective as a conjugate in inducing B-cell activation in donors immunized with the conjugate vaccine (16). These data are therefore compatible with Mosier’s postulate (12) that the function of a conjugate is merely be to ensure localization of protein antigens (required for T-cell activation) at the site of polysaccharide homing. In vitro, protein and polysaccharide are present in close proximity to B as well as T cells, and so physical coupling may not be necessary. Throughout all of our experiments, PRP conjugated to TT induced anti-PRP responses equal to those seen after restimulation with mixture of PRP and TT. This finding confirms the prior observations mentioned above. It should be noted that all of our experiments were performed with adults who were vaccinated with the PRPTT conjugate vaccine. There may be differences with infants, the population at greatest risk for H. influenzae type b infections, in terms of both maturation of the immune system and degree of natural exposition to these encapsulated bacteria. Direct assessment of this issue is precluded, however, by the high numbers of B lymphocytes required for these types of in vitro studies. In our in vitro system, the presence of T cells is necessary for an optimal antiprotein as well as antipolysaccharide ASC response (Fig. 3). The requirements for the antibody response against the polysaccharide, however, are different from those in the response against proteins. In agreement with the model for activation of protein-specific B cells described in the introduction, direct physical contact between B cells and (activated) T cells is necessary for induction of an antiprotein antibody response (Fig. 3). When T cells and antigen-presenting cells were physically separated from the B cells by the Transwell membrane, no anti-TT ASC were found. Also, when contact between B and T cells was possible in one compartment of the Transwell system, the release of soluble factors due to contact between T and B cells in that compartment did not result in an anti-TT antibody response in the other compartment. Apparently, cytokines formed as a consequence of contact between protein-specific B and T cells are in themselves not sufficient to activate proteinspecific B cells; direct T-cell–B-cell contact is required.

VOL. 67, 1999

T CELLS IN THE ANTIPOLYSACCHARIDE ANTIBODY RESPONSE

The Transwell studies also indicate that contact between B and T cells is necessary to induce an optimal anti-PRP antibody response (Fig. 3). When T cells and antigen-presenting cells were physically separated from the B cells by the Transwell membrane, hardly any anti-PRP specific ASC were found (Fig. 3, culture c). The presence of low levels of anti-PRP ASC when the B-cell fraction was cultured with PRP plus TT may be explained by T-cell activation in the upper compartment by TT. These activated T cells may secrete cytokines which augment differentiation of PRP-activated B cells in the lower compartment. The potential for contact between B and T cells in one compartment of the Transwell system apparently resulted in the release of soluble factors able to stimulate B cells in the other compartment to secrete anti-PRP antibodies (Fig. 3, culture e). In both culture c and culture e, T lymphocytes are cultured with antigen-presenting cells (monocytes in culture c; B cells and monocytes in 3 culture e) and TT (Fig. 3). These T lymphocytes can produce cytokines which can diffuse to the other compartment. Yet this does not result in anti-PRP ASC formation in culture c, while anti-PRP ASC are generated in culture e (Fig. 3). The difference between cultures c and e is the presence of B lymphocytes in the latter. Our results suggest that cytokines are produced by B cells that can activate polysaccharide-specific B cells to make antibodies. These results are partly in accordance with the proposed model for the working mechanism of a conjugate vaccine: the polysaccharide-specific B cells internalize the conjugate, process the protein part, and express the peptides in their MHC molecules to activate peptide-specific T cells. Direct physical contact between B and T cells is necessary in this part of the process. Activated T cells then secrete cytokines that activate the polysaccharide-specific B cells to become ASC. We now have shown that in this part of the process, the activated B cells may secrete cytokines that are able to activate other polysaccharide-specific B cells to become ASC. The nature of these soluble mediators is currently under investigation. Preliminary data indicate that interleukin-4 is one of the cytokines which can support the differentiation of B cells into antipolysaccharide ASC. ACKNOWLEDGMENTS This work was supported by grant 920-02-091 from the NWO Medical Research Foundation. We thank J. T. Poolman for synthesis of the TT peptides during his tenure at the National Institute of Public Health and Environment, Bilthoven, The Netherlands. REFERENCES 1. Amlot, P. L., D. Grennan, and J. H. Humphrey. 1985. Splenic dependence of the antibody response to thymus independent (TI-2) antigens. Eur. J. Immunol. 15:508–512.

Editor: R. N. Moore

793

2. Austrian, R. 1989. Pneumococcal polysaccharide vaccines. Rev. Infect. Dis. 11(Suppl. 3):S598–S602. 3. Baker, P. J. 1992. T cell regulation of the antibody response to bacterial polysaccharide antigens: an examination of some general characteristics and their implications. J. Infect. Dis. 165(Suppl. 1):S44–S48. 4. Black, S. B., H. R. Shinefield, B. Fireman, R. Hiatt, M. Polen, and E. Vittinghoff. 1991. Efficacy in infancy of oligosaccharide Haemophilus influenzae type b (HbOC) vaccine in a United States population of 61,080 children. Pediatr. Infect. Dis. 10:97–104. 5. Cowan, M. J., A. J. Ammann, D. W. Wara, V. M. Howie, L. Schultz, N. Doyle, and M. Kaplan. 1978. Pneumococcal polysaccharide immunization in infants and children. Pediatrics 62:721–727. 6. Douglas, R. M., J. C. Paton, S. J. Duncan, and D. J. Hansman. 1983. Antibody response to pneumococcal vaccination in children younger than five years of age. J. Infect. Dis. 148:131–137. 7. Eskola, J., H. Ka ¨yhty, A. K. Takala, H. Peltola, P. Ro¨nnberg, E. Kela, E. Pekkanen, P. H. McVerry, and P. H. Ma ¨kela ¨. 1990. A randomized prospective field trial of a conjugate vaccine in the protection of infants and young children against invasive Haemophilus influenzae type B disease. N. Engl. J. Med. 323:1381–1387. 8. Eskola, J., H. Peltola, P. Makela, et al. 1985. Antibody levels achieved in infants by course of Haemophilus influenzae type b/diphtheria toxoid conjugate vaccine. Lancet i:1184–1186. 9. Griffioen, A. W., G. T. Rijkers, P. Janssens-Korpela, and B. J. M. Zegers. 1991. Pneumococcal polysaccharides complexed with C3d bind to human B lymphocytes via complement receptor type 2. Infect. Immun. 59:1839–1845. 10. Insel, R. A., A. Kittelberger, and P. Anderson. 1985. Isoelectric focusing of human antibody to the Haemophilus influenzae b capsular polysaccharide: restricted and identical spectrotypes in adults. J. Immunol. 135:2810–2816. 11. Ka ¨yhty, H., V. Karanko, H. Peltola, and P. H. Ma ¨kela ¨. 1984. Serum antibodies after vaccination with Haemophilus influenzae type b capsular polysaccharide and responses to reimmunization: no evidence of immunologic tolerance or memory. Pediatrics 74:857–865. 12. Mosier, D. E., and A. J. Feeney. 1986. The physiology of B lymphocytes capable of generating antipolysaccharide antibody responses, p. 243–249. In R. Bell and G. Torrigiani, (ed.), Towards better carbohydrate vaccines. John Wiley & Sons, New York, N.Y. 13. Noelle, R. J., J. A. Ledbetter, and A. Aruffo. 1992. CD40 and its ligand, an essential ligand-receptor pair for thymus-dependent B-cell activation. Immunol. Today 13:431–433. 14. Panina-Bordignon, P., A. Tan, S. Termijtelen, S. Demotz, G. Corradin, and A. Lanzavecchia. 1989. Universally immunogenic T cell epitopes: promiscuous binding to human MHC class II and promiscuous recognition by T cells. Eur. J. Immunol. 19:2237. 15. Peeters, C. C., A. M. Tenbergen Meekes, C. J. Heijnen, J. T. Poolman, B. J. Zegers, and G. T. Rijkers. 1991. In vitro induction of a Haemophilus influenzae type b polysaccharide-specific antibody response in human peripheral blood lymphocytes of individuals recently vaccinated with an oligosaccharide-protein conjugate. J. Immunol. 147:4192–4199. 16. Peeters, C. C. A. M., A. M. Tenbergen-Meekes, J. T. Poolman, B. J. M. Zegers, and G. T. Rijkers. 1993. In vitro antibody response of human lymphocytes to the Haemophilus influenzae type b capsular polysaccharide. Springer Semin. Immunopathol. 15:247–258. 17. Rijkers, G. T., and D. E. Mosier. 1985. Pneumococcal polysaccharides induce antibody formation by human B-lymphocytes in vitro. J. Immunol. 135:1–4. 18. Rijkers, G. T., E. A. M. Sanders, M. A. Breukels, and B. J. M. Zegers. 1996. Responsiveness of infants to capsular polysaccharides: impact for vaccine development. Rev. Med. Microbiol. 7:3–12. 19. Sedgwick, J. D., and P. G. Holt. 1983. A solid phase immunoenzymatic technique for the enumeration of specific antibody secreting cells. J. Immunol. Methods 57:301–309. 20. Siber, G. R. 1994. Pneumococcal disease; prospects for a new generation of vaccines. Science 265:1385–1387.