Materials and Methods

Reagents.nPT and mutated nontoxic PT-9K/129G (dPT), a component of antipertussis acellular vaccine, were donated by Chiron-Biocine. Endotoxin contamination, as determined by limulus amebocyte lysate (LAL) (E-Toxate; Sigma Chemicals), was <0.2 EU/mg of protein (<0.02 ng/mg of protein). Biologic properties and purification procedures of these toxins are detailed elsewhere [6]. LPS from Escherichia coli were obtained from Sigma. Anti–IL-12 (C8.6) monoclonal antibody (MAb) was donated by F. Malavasi (Università di Torino, Italy). Human recombinant (r) granulocyte-macrophage colony-stimulating factor (GM-CSF) and human rIL-4 were obtained from Novartis Pharma; alternatively, culture supernatants (4% vol/vol) of cell line transfected with human IL-4 (IL-4-62) were also used as source of IL-4 [19]. Phytohemagglutinin (PHA; HA16) was obtained from Wellcome Diagnostics

Purification and culture of DCs.DCs were obtained from peripheral blood of 30 healthy blood donors (courtesy of G. Girelli, Centro Trasfusionale dell’Università, Rome, Italy). Monocytes were purified by Percoll gradient (density, 1.124 g/mL; Biochrom) from peripheral blood mononuclear cells (PBMC) obtained after ficoll gradient (lympholyte-H; Cedarlane). Monocyte purification was assessed by cytofluorometric analysis. CD14⁺ cells were further purified by positive sorting through CD14 MAb-conjugated magnetic microbeads (Miltenyi). Alternatively, CD14⁺ cells were obtained by positive CD14 cell sorting (Miltenyi), starting directly from PBMC [20]. Cytofluorometric analysis showed that the purification procedure yielded >97% pure CD14⁺ cells. CD14⁺ cells were cultured at 4×10⁵ cells/mL in RPMI 1640 medium (ICN-Flow) supplemented with heat-inactivated 10% LPS-screened fetal calf serum (LAL, <1 ng/mL), 1 mM sodium pyruvate, 0.1 mM nonessential amino acids, 2 mM l-glutamine, 25 mM HEPES, 100 U/mL penicillin, 100 μg/mL streptomycin (all from Hyclone Laboratories), and 0.05 mM 2-mercaptoethanol (Sigma; hereafter referred to as complete medium) at 37°C in 5% CO₂, in the presence of 1000 U/mL rIL-4 and 50 ng/mL rGM-CSF (Novartis-Pharma) [21]

After 6 days, immature DCs were washed and analyzed by cytofluorometric analysis for CD1a, CD14, and CD3 expression. The mean percentage (±SD) of CD1a⁺, CD14⁺, and CD3⁺cells was 93.3±4.7, 1.8±1.3, and 1.6±1.2, respectively. In some experiments, CD14⁺ cell preparations were further deprived of CD3⁺ cells by use of appropriate magnetic microbeads (Miltenyi). DCs were further cultured (10⁶/mL) in the presence of nPT or dPT at a concentration of 5 μg/mL or LPS (100 ng/mL), unless otherwise specified, or their association at the respective concentration for 48 h to induce maturation. Mature DCs were analyzed by flow cytometry for the expression of maturation markers, including CD80, CD83, CD86, and HLA-DR. Heat-inactivated (1 h at 96°C) nPT or dPT was also used as DC stimulant in some experiments

Cytofluorometric analysis.Monocytes and DCs were washed and resuspended in PBS containing 1% human serum and incubated with a series of fluorochrome-conjugated MAbs to human antigens for 30 min at 4°C. We used the following MAbs: anti-CD1a and anti-CD86 (Serotec), anti-CD83 (PharMingen), anti-CD80, anti-CD14, and anti–HLA-DR (Becton Dickinson). Cells were analyzed by flow cytometry. Data were collected and analyzed by flow cytometer (FACScan; Becton Dickinson), and analysis was performed with CELLQuest software (Becton Dickinson). Cells were gated according to light scatter properties to exclude cell debris. Fluorescence data are reported as a percentage of CD83⁺ cells, because treatment induced the expression of the marker in cells that were negative. Median fluorescence intensity was used when treatment increased the expression of the marker in cells that were already positive for CD80, CD86, and HLA-DR markers

Surface and intracellular staining.Intracellular staining for cytokine production was done by culturing DCs or T cells with the indicated stimuli for 6 h. The intracellular staining was performed from the beginning of the experiment in the presence of brefeldin A (10 μg/mL; Sigma), a compound that blocks proteins in the endoplasmic reticulum, thus inhibiting cellular secretion and avoiding binding of secreted cytokines to cell surfaces. Cells without brefeldin A were also seeded to measure cytokine secretion by ELISA. Cells were stained for cell surface antigens by using pretitrated fluorochrome-conjugated MAb specific for cell surface, as specified above. Cells were then fixed, permeabilized according to the manufacturer’s instructions (Cytofix-Cytoperm and Perm/Wash protocol; BD PharMingen), and stained with a predetermined optimal concentration of fluorochrome-conjugated anti-cytokine MAb or appropriate isotype MAb control. After incubation for 30 min at 4°C, cells were washed and analyzed by cytofluorometric analysis. The sensitivity of intracellular staining assay was tested by adding, before staining, scalar number of untreated DCs to LPS plus PT-treated DCs. The percentage of cytokine-stained cells decreased linearly with the increased number of untreated DCs (data not shown)

Cytokine measurement by ELISA or cytometric bead array (CBA) assay.To measure cytokine production, DCs (10⁶ cells/mL) were cultured in the presence of the indicated stimuli in 0.5 mL of complete medium in 24-well plates (Costar) or in 5-mL tubes (Falcon; Becton Dickinson) at 37°C in a 5% CO₂ atmosphere. Unless otherwise specified, supernatants were collected after 48 h and used to measure IL-1β, tumor necrosis factor (TNF)–1α, IL-2, IL-6, IL-10, IL-12 p70, and interferon (IFN)–γ by an immunoenzymatic assay (ELISA) (Quantikine; R&D Systems). This assay has a sensitivity (in pg/mL) of 1 for IL-1β, 4.4 for TNF-α, 0.7 for IL-6, 3.9 for IL-10, 5 for IL-12, and 8 for IFN-γ. Optical density was read at 450 nm with a UV microplate reader (model 3550; BioRad)

Cytokines in the supernatants of polarized naive T cells were also assayed by using human Th1/Th2 CBA (BD PharMingen). In total, we used 6 bead populations with distinct fluorescence intensities, coated with capture antibody specific for IL-2, IL-4, IL-5, IL-10, TNF-α, and IFN-γ proteins. Samples were incubated with human cytokine capture beads and stained with phycoerythrin detection reagent. After incubation for 3 h at room temperature, samples were washed and analyzed by FACScan (Becton Dickinson). Data were analyzed by use of Cell Quest and Cytometric Bead Array (version 1.1; both by Becton Dickinson)

mRNA cytokine expression by TaqMan real-time quantitative reverse-transcription polymerase chain reaction (RT-PCR) analysis.To measure cytokine mRNA expression, we used TaqMan real-time quantitative RT-PCR analysis (Perkin Elmer/Applied Biosystems). RNA extraction and RT were performed as described elsewhere [4]. TaqMan assays were done per the manufacturer’s instructions with an ABI 7700 thermocycler (Perkin Elmer/Applied Biosystems). Duplex PCR was done by amplifying in the same tube the target cDNA (IFN-γ, IL-5, or IL-12 p40 transcripts) and β-actin cDNA as endogenous control. The probe, labeled at the 5′ end with a fluorescent reporter and at the 3′ end with a quencher annealing the amplicon, was added to the PCR. During PCR, the 5′–3′ nuclease activity of TaqMan polymerase cleaves the probe, resulting in the displacement of the quencher from the reporter, which thus releases a fluorescent signal. Specific paired primers and probes were obtained from Perkin Elmer/Applied Biosystems. Data were analyzed with Sequence Detector (version 1.7) and Relative Quantification (version 1.1; both from Perkin Elmer). Specific mRNA transcript levels were expressed as fold increase compared with basal condition

DC–T cell cocultures.Allogenic T cells were purified from PBMC by immunomagnetic negative selection (Dynal) with anti-CD14, CD16, CD19, and HLA-DR MAb sorting, as described elsewhere [22]. DCs were cultured with the indicated stimuli for 48 h, washed extensively, and cultured in different amounts in round-bottom 96-well plates (Costar) with 5×10⁴ T cells/well in triplicate. Plates were tested for [³H]thymidine incorporation on day 5 [5]. Data are shown as the mean (±SD) of the counts per minute×10³

Th1/Th2 polarization experiments.To evaluate the capacity of DC-conditioned medium (CM) to polarize T lymphocytes, experiments were done with naive T cells from cord blood. We enriched T cells by depriving CD14⁺ cells by use of CD14 MAb–conjugated magnetic microbeads (Miltenyi). The polarization procedure has been described elsewhere [23]. In brief, 2×10⁶ cord blood T cells were cultured in complete medium in 24-well plates (Costar) in the presence of PHA (1.5 μg/mL) and CM (diluted to 50% vol/vol) obtained from untreated (NT) DCs or DCs cultured for 48 h in the presence of LPS, nPT, dPT, or LPS plus either dPT or nPT. dPT plus LPS CM-treated T cells were also cultured in the presence of anti–IL-12 MAb (50 μg/mL). T cells cultured in the presence of rIL-12 (100 U/mL; Serotec) or rIL-4 (100 U/mL; Novartis) plus anti–IL-12 MAb (50 μg/mL) were used as control Th1- or Th2-polarizing conditions, respectively. At day 3, each T cell culture was divided into 2, and rIL-2 (50 U/mL; Roche) was added. At day 5, the cells were extensively washed and stimulated with PHA for 24 h for cytokine secretion measurement by CBA assay or for 6 h for intracellular cytokine staining in the presence of brefeldin A as described above

Statistics.Descriptive statistical analyses were done with the SPSS statistical package. Differences between mean values were assessed by paired Student’s t test

Results

Phenotypic analysis of PT-treated DCs.Immature monocyte-derived DCs are characterized by low expression of costimulatory molecules (e.g., CD80 and CD86) and absence of CD83, a protein expressed only on mature DCs [24]. As shown in table 1, treatment of monocyte-derived DCs with nPT markedly increased the expression of CD86, CD80, and HLA-DR antigens and caused the appearance of CD83⁺ cells as well. This effect was comparable to or only slightly inferior to that shown by LPS. By the same criteria, dPT was also capable of inducing DC maturation (table 1), thus indicating that this event did not require the enzymatic activity of nPT [6]. On the other hand, heat inactivation (96°C, 1 h), abolished the capacity of both nPT and dPT to induce DC maturation (data not shown), demonstrating that the effect of PT preparations was not due to contaminating LPS. No apparent influence was exerted by either dPT or nPT preparation on LPS-induced costimulatory molecules and HLA-DR expression, although the values obtained with nPT (but not with dPT) plus LPS were in general lower than those obtained with LPS alone (table 1)

Induction of cytokine production.DCs are powerful producers of key inflammatory and immunoregulatory cytokines, especially when coactivated by antigens and adjuvants of microbial pathogens [15]. Thus, we tested the ability of PTs to promote cytokine gene expression and cytokine production by DCs. Figure 1 shows the levels of IL-1β, TNF-α, IL-6, IL-10, IL-12, and IFN-γ in supernatants of 48-h activated DCs following stimulation with LPS, nPT, or dPT (with or without LPS), as detected in 5–23 independent experiments (depending on the cytokine assessed) with different monocyte donors. Overall, the 2 PT preparations stimulated comparable levels of cytokine production by DCs. In general, LPS was a more potent cytokine stimulant than nPT or dPT, except IFN-γ (see below). The combination of either dPT or nPT preparation with LPS did not consistently influence the amount of cytokine produced (with the notable exception of IL-12, see below), although nPT, but not dPT, appeared to proportionally increase LPS effects on IL-1β and TNF-α production (figure 1)

IL-12 and IFN-γ production.Our research focused on critical Th1-driving cytokines, which play an essential role in the protection against B. pertussis and other bacterial pathogens [16, 17, 25]. IFN-γ was minimally produced, if at all, by LPS-stimulated DCs while being produced to elevated amounts (on average, ∼800 pg in 10 independent DC cultures) after DC stimulation by nPT or dPT (figure 1)

As noted, IL-12 had a particular behavior. Low levels of IL-12 (mean, 50 pg/mL) were produced by nPT or dPT-treated DCs, whereas LPS alone, on average, produced ∼300 pg/mL (figure 1). However, the combination of LPS with nPT or dPT caused a substantial increase in the quantity of cytokines produced by the cells (nPT plus LPS, 726±101 pg/mL [mean±SE] of 20 experiments; dPT plus LPS, 1276±270 pg/mL [21 experiments] with DCs from different donors; P<.0001). Figure 2A shows that the stimulatory effect of PT plus LPS was PT dose-dependent and occurred even at dPT doses as low as 0.1 μg/mL when no IL-12 was measurable after stimulation by PT alone. This clearly indicates a synergistic activity between the 2 stimulants. This observation was strengthened (figure 2B ) by the finding that ⩾0.1 μg/mL LPS is needed to induce the synergistic action of PTs plus LPS in IL-12 secretion

To rule out that the synergism between LPSs and PTs on IL-12 production by DCs could be because of differences in kinetics of cytokine synthesis and release following the 2 different stimulations, we assessed the kinetics of IL-12 production by measuring the IL-12 p40 mRNA expression by TaqMan quantitative PCR assay and the amount of IL-12 p70 biologic active protein dimer by ELISA. Figure 3 shows that LPS induced a rapid IL-12 p40 mRNA expression, which peaked at 5 h and fell at 20 h. In contrast, IL-12 p40 gene transcription was measurable to relatively low levels only after >20 h of stimulation by nPT or dPT. The addition of either dPT or nPT preparation to LPS dramatically increased IL-12 p40 mRNA expression, with kinetics similar to that induced by LPS alone (peak at 5 h and rapid fall at 20 h) followed by stabilization at lower quantitative level. The measurement of IL-12 p70 was fully in keeping with these kinetics, showing that the cytokine message was translated into the protein starting near its peak production. IL-12 in the supernatant remained at a substantially high steady-state level for the duration of measurement (20–48 h; data not shown). Overall, these results indicated that the synergism between LPS and nPT or dPT was due to a real increase of the IL-12 message and its translation into the biologically active IL-12 p70 protein and not to a selection of a particular time of cytokine measurement in the culture medium

We next examined whether the enhanced capacity of DCs to secrete IL-12 after costimulation with PT and LPS was caused by an expanded proportion of IL-12–secreting DCs or to an enhanced rate of IL-12 production and secretion per cell. To rule out that the IFN-γ detected in our cultures was secreted by the low number of T cells that normally contaminate DC preparations, we also examined whether human monocyte–derived DCs could secrete IFN-γ. Thus, the number of IL-12– and IFN-γ–producing cells was enumerated by using the cytokine intracellular staining technique. Figure 4A shows that roughly an equal percentage (5.4%–5.8%) of IL-12–positive DCs was induced by PT plus LPS or by LPS alone after 6 h of stimulation. This suggests that the costimulation of DCs by LPS and PT induced increased rates of IL-12 gene transcription and biologic active cytokine secretion compatible with the results of IL-12 mRNA expression kinetics (figure 3) but not an increased number of cells able to produce IL-12. Figure 4B clearly indicates that, after activation by PT or PT plus LPS, some DCs acquire the capacity of directly producing IFN-γ. Of note, the number of IL-12– and IFN-γ–producing cells was quite low, compared with the protein levels of these cytokines, detected in culture supernatants (figure 1) but, in the latter case, the accumulated protein was measured after 48 h of culture

PT and LPS enhance the allostimulatory capacity of DCs. Since critical aspects of DC function in vivo are antigen presentation and T cell activation, we evaluated the capacity of DCs to stimulate alloreactive T cells after DC maturation in the presence of PT, LPS, or both. Figure 5A shows that DCs obtained after culture with LPS or dPT were able to increase the relatively low degree of proliferation of allogenic T cells achievable with unstimulated DCs. However, treatment of DCs with dPT-LPS combination greatly enhanced the proliferative ability of T cells. We also examined whether IL-12 induced by PT plus LPS-stimulated DCs was a factor in the enhanced capacity of alloreactive T cells to proliferate. We treated DCs with an anti–IL-12 MAb that caused ∼two-thirds reduction of IL-12 (figure 5B ). Of note, treatment of DCs with anti–IL-12 MAb also reduced the IFN-γ production induced by PT or by PT plus LPS-stimulated DCs (figure 5B ). The addition of the anti–IL-12 MAb also markedly affected the potentiation exerted by nPT plus LPS on T cell-alloreactive proliferation (figure 5C ), thus indicating that IL-12 and IFN-γ were probably involved in the functional activation of PT plus LPS-stimulated DCs

CM from PT and LPS-stimulated DCs polarizes naive T cells for Th1 function.The data in the previous sections point to the capacity of PT preparations to synergize with LPS in IL-12 production by maturing human DCs and, consequently, for these cells to act functionally in antigen presentation. Because IL-12 is a critical cytokine for Th1 cell polarization and antipertussis protective responses [16, 17], we evaluated whether the CM of DCs stimulated with PT or LPS or their combination contained factors capable of inducing polarization of naive T cells from cord blood

Cord blood T cells were stimulated with PHA and 50% (vol/vol) diluted CM obtained from NT CM, LPS (LPS CM), and either dPT CM or nPT CM, or PTs plus LPS CM-treated DCs and assayed for IFN-γ (Th1) or IL-5 (Th2) message and cytokine secretion. In parallel cultures, Th1 (human rIL-12)– or Th2 (human rIL-4 plus neutralizing anti–IL-12 MAb)–polarizing conditions were used as controls. As shown in figure 6, the CM from nPT, dPT, or LPS-treated DCs, but mostly the CM from nPT plus LPS or dPT plus LPS-treated DCs, induced marked production of IFN-γ but not of IL-5. This was observed both as gene transcription (figure 6A ) and cytokine secretion (figure 6B ). IFN-γ production by naive T cells stimulated by nPT (or dPT) plus LPS CM was comparable to the positive control stimulation, IL-12. In all cases, IFN-γ production was almost totally abolished by anti–IL-12 MAb (figure 6)

We also assessed the intracellular cytokine production. As shown in figure 7, the percentage of IFN-γ–secreting cells was even higher (13%–15.6%) in T cells from cord blood treated with LPS CM and LPS plus dPT CM than in those induced by the standard Th1-polarizing condition (9.2%). The Th1-polarizing effect of the CM from LPS plus PT-treated DCs was again substantially reduced by treatment with the anti–IL-12 MAb

Any CM, regardless of the DC stimulant, had no or minimal effect on IL-5 production (figure 6). This was confirmed by determining the number of IL-4–producing naive T cells incubated with CM from PT or PT plus LPS-treated DCs (data not shown). In preliminary experiments, PT plus LPS-treated DCs were seen to drive Th1 polarization by cord blood T cells, confirming the observations made with the CM (data not shown). Altogether, the experimental evidence supports the notion that nPT, dPT, and LPS (and especially their combinations) are Th1-polarizing stimulants and that IL-12 production by DCs is critical for this polarization

Discussion

Despite clear evidence of the high efficacy of pertussis vaccines, the specific immune response elicited by these vaccines that contribute to the induction and persistence of protection remains undefined [1, 2]. There is no doubt that detoxified preparations of PT, a major virulence factor of B. pertussis are a necessary constituent of all acellular pertussis (aP) vaccines and that a substantial degree of the protection achieved in multicomponent aP vaccines is due to the immune response against this major antigenic constituent [1, 2, 7]. Although this is usually interpreted because of PT acting as a highly immunogenic protein against which protective antibody and/or cell-mediated responses are elicited, it is important to consider that PT is a truly complex molecule with a set of immunomodulatory properties that directly or indirectly influence quality and magnitude of adaptive immune responses

In the aP vaccine formulation, PT is rendered nontoxic either by chemical or genetic detoxification, both of which eliminate ADP ribosylation activity [6, 26]. However, genetic detoxification appears to leave intact most of the physiologic and immunologic effects mediated by the B oligomer [10, 27 –30]. The adjuvanticity of PT and the essential participation of the B subunit in this property have been emphasized in several studies. In mice, Ryan et al. [28] showed that PT and its nontoxic mutant induce T cell proliferation, Th1/Th2 cytokine secretion to coadministered antigen, and enhance the expression of the costimulatory molecule CD28. The antigen-specific cytokine production induced by PT is due to clonal expansion of T cells [31]. Roberts et al. [29] showed that dPT may act as a strong mucosal adjuvant, whereas He et al. [32] demonstrated that inoculation of PT in mice enhances the capacity of splenocytes to produce IL-12 in response to both microbial and nonmicrobial stimuli. Of more importance, these authors described the induction by PT of a healing phenotype in mice infected with Leishmania major which is known to induce a Th1-dependent response [33]

In this study, we found that both nPT and dPT can induce maturation of monocyte-derived human DCs, that is, the cells that are essential for initiation and regulation of immune response [14, 15, 34]. Of importance, the nPT and dPT capability of enhancing the expression of costimulatory molecules (CD80 and CD86) and elevating the percentage of CD83⁺ cells was comparable to that induced by optimal doses of LPS from E. coli a potent signal for DC maturation. Thus, our data indicate that PT is a strong stimulant of DC maturation, allowing these cells to functionally enter the process of efficient antigen presentation and T cell activation. The fact that dPT is as capable as nPT of inducing DC maturation suggests that this event is probably attributable to the B oligomer. Moreover, heat-inactivated nPT or dPT were unable to induce DC maturation, suggesting that the immunomodulatory activity equally exerted by nPT and dPT (and abolished by heat inactivation) may be important in inducing DC maturation. These findings concomitantly rule out that PT activities are mediated by contaminating LPS as also demonstrated by the divergent effects of LPS and PT on IL-10 and IFN-γ production by DCs, as discussed below

PT-treated DCs produced cytokines capable of activating and regulating immune responses. We studied the production of IFN-γ and IL-12 for the critical property of these cytokines for most protective antiinfectious responses. To our knowledge, our findings are the first to show that not only native and genetically detoxified PT can induce the production of IL-12 by mature DCs, but also that they potently synergize with LPS in production of IL-12. The synergism between PT and LPS in IL-12 production was clearly dose dependent, not due to any LPS contamination of PT, and was likely mediated by up-modulated transcription of IL-12 p40 mRNA. In fact, as shown by IL-12 intracellular staining, no more cells were involved in IL-12 production in the PT plus LPS-stimulated DCs than in LPS-stimulated DCs; however, more transcript was produced and probably stabilized in the presence of PTs. The synergism between LPS and PT for IL-12 p70 production was not paralleled by any synergism for DC maturation, recalling previous observations by Lapointe et al. [35], who used CD40 ligand as additional stimulant to LPS. As in our study, they observed no correlation between extent of phenotypic maturation of DC and IL-12 production, which was synergistically augmented by the combined stimulation by CD40L and LPS [35]. Thus, our data support the notion that multiple maturation signals promote the best efficiency of DCs in the production of IL-12. Along this line, future studies on the binding of PTs to receptors of natural immunity, in particular the toll-like receptors [36], would be of special importance

We show that by strongly potentiating IL-12 production by LPS-stimulated DCs, PT enhances the capacity of LPS to induce Th1 polarization by DCs [37]. This is remarkable in view of the report by Roberts et al. [29] that dPT mostly elicited an antitetanus toxin antibody profile of a marked Th2 type in mice. Although the 2 findings come from very different approaches, our data rule out that dPT would act exclusively as Th2-promoting signal, at least in ex vivo human cells. Rather, they indicate that PT, especially after interaction with LPS, is a clear promoter of Th1 cell polarization

It is possible that the differences observed in polarization of immune response depend on the different origin of DCs. Monocyte-derived DCs have greater potential to differentiate into Th1-promoting DC1, whereas DCs of other origin (e.g., plasmacytoid DCs) may induce Th2 differentiation [38]. Thus, it is possible that DCs of different origin and different cytokine profile can be activated by PT in vivo

The anti-inflammatory cytokine IL-10 counteracts Th1 type cytokines (e.g., IL-12 and IFN-γ) and inhibits DC maturation. Corinti et al. [39] showed that production of IL-12 by DCs stimulated with microbial products is inhibited by IL-10. In our experiments, IL-10 was not or was minimally produced by nPT- or dPT-stimulated DCs and the amount of this cytokine produced by LPS-stimulated DCs was not significantly enhanced by the addition of either form of PT

In contrast, nPT and dPT (with or without LPS) caused a marked production of IFN-γ in our DC cultures. Moreover, IFN-γ production by human DCs was documented by intracytoplasmic staining technique. Recently, high levels of IFN-γ production by activated macrophages and DCs were reported in mice [40]. IFN-γ is an antagonist of IL-10 [41] that promotes IL-12 production by DCs, mostly when helped by signals from CD40L, without significantly altering IL-10 production [41, 42]. Thus, the capacity of nPT and dPT to induce high IFN-γ levels is a likely factor in fostering IL-12 production by DCs. Theoretically, it can be envisaged that the synergism between PT and LPS could be caused, among other factors, by the capacity of PT (but not LPS) to induce IFN-γ, thus counteracting the down-regulatory effect of IL-10 on IL-12 production, as shown by Corinti et al. [39]. In this context, the high ability of PT to induce IFN-γ, coupled with the scarce ability to induce IL-10, speaks again in favor of a Th1-driven protective immune response and strengthens the antipertussis protective role of IFN-γ, already strongly suggested by the results of B. pertussis experimental infection [1]

The Th1 promoting ability of PT deserves attention for the mechanism of antipertussis protection. The aP vaccines consistently induce a mixed Th1-Th2 cytokine profile in children [4, 43], although the balance between the 2 profiles was not equal in recipients of different aP vaccines and was not independent of the antigenic specificity of cell-mediated immune response [4]. Our findings suggest that among the antigens contained in aP vaccines, PT, in particular its genetically detoxified mutant, substantially contributes to the Th1-type component of the cytokine profile [44]. All of these factors likely contribute to the high level of protection achievable with the use of PT-based aP vaccines