TEXT

Top Abstract Text References

Most infectious agents gain entry into the host via mucosal surfaces, particularly through those of the respiratory and gastrointestinal tract. It is well documented that protection against such agents requires activation of the local mucosal immune system in addition to systemic immunity, something that is best achieved through mucosal immunization. Mucosal immunization induces antigen-specific responses in both the mucosal and systemic immune compartments, while systemic immunization generally results in the induction of only systemic immune responses (20, 23). However, the poor immunogenicity of mucosally administered proteins has been a major obstacle to the development of efficient oral and nasal vaccines. Also, induction of oral tolerance can occur concurrently with the development of specific antibody responses at mucosal sites (24, 26).

Dendritic cells (DC) have been termed nature's own adjuvant due to their capacity to induce cellular and humoral responses against particulate and soluble antigens in the absence of an external adjuvant. They reside in most tissues and sense the environment by capturing and processing antigens. Once activated by inflammatory stimuli and infectious agents, DC migrate to the draining lymphoid organs to interact with antigen-specific lymphocytes. During migration, they acquire professional antigen-presenting capacity, up-regulate major histocompatibility complex (MHC) and costimulatory molecules, and become competent to activate both T cells (7, 11) and B cells (28).

Vaccination with antigen-pulsed DC has proven to be efficient in inducing immune responses and protection against intracellular bacteria (4, 14), virus (1, 25), and tumors (13, 21, 29). However, the ability of antigen-pulsed DC to induce protective immunity against extracellular mucosal pathogens has not been explored. We therefore investigated the protective efficacy of DC vaccination of mice against infection with the extracellular mucosal pathogen Bordetella pertussis. We also explored the mucosal and systemic immune responses mediated by vaccination with DC.

To this end, DC were cultured from bone marrow of C57BL/6 or BALB/c mice as described previously (8). The DC populations obtained were heterogenous with 30 to 35% of the cells expressing high levels of MHC class II, 25% CD40, 55 to 60% CD80, 40 to 50% CD86, and 35% CD11c as determined by fluorescence-activated cell sorter analysis. At day 6, DC were pulsed with 10⁷ heat-killed B. pertussis bacteria (prepared as described previously; see reference 15) per 10⁶ DC overnight. After extensive washing of the cells, autologous (C57BL/6-derived) B. pertussis-pulsed DC were administered either intravenously (i.v.) or intranasally (i.n.) to C57BL/6 mice. Alternatively, heterologous (BALB/c-derived) DC were administered i.v. to C57BL/6 mice. Three doses of 5 × 10⁵ to 1 × 10⁶ DC per dose were given at weekly intervals. Control mice were given mock-pulsed DC or 10² heat-killed B. pertussis bacteria as described above. Two weeks after the last administration of DC, the mice were challenged with 5 × 10⁶ B. pertussis bacteria i.n. (16). The animals were killed 1 week after challenge and assayed for B. pertussis-specific immunoglobulin G (IgG) levels in serum, bacterial load in the lungs, and B. pertussis-specific antibody-secreting cells (ASC) in the lungs.

Efforts were also made to track whole or fragmented DC in vivo by administering 10^{6 51}Cr-labeled DC (1.53 × 10⁶cpm/10⁶ cells) i.n. or i.v. Recipient mice were sacrificed after 24 h, and organs and blood were collected and measured for  emission.

For the detection of phagocytosed B. pertussis, the protocol of Drevets et al. was followed (5). Essentially, DC were incubated with fluorescein isothiocyanate (FITC)-labeled heat-killed B. pertussis at 37°C for 20 min and then washed extensively. To distinguish between ingested and extracellular bacteria, the cell suspension was mixed with ethidium bromide at a final concentration of 5 µg/ml and analyzed under the fluorescence microscope.

For the determination of the bacterial load, the lungs were removed aseptically and homogenized in 5 ml of phosphate-buffered saline. Serially diluted homogenates from individual lungs were plated onto Bordet-Gengou agar, and the number of CFU was determined after 5 to 6 days of incubation at 37°C. Results are expressed as numbers of viable bacteria per lung.

For the enumeration of ASC in the lungs, individual lungs were cut into small pieces, suspended in an enzyme solution consisting of Hanks balanced salt solution with 1 mg of collagenase-Dispase (Boehringer Mannheim)/ml and 0.25 mg of DNAse 1 type IV (Sigma)/ml, incubated at 37°C for 30 min, filtered through a 150-µm-pore-size nylon mesh, and analyzed by enzyme-linked immunospot assay (2) for numbers of IgG and IgA ASC specific for B. pertussis, on nitrocellulose plates coated with bacterial sonicate (prepared as described previously; see reference 17).

Statistical analyses were done by Student's t test with the Bonferroni correction for multiple analyses.

i.v. but not i.n. administration of antigen-pulsed DC leads to protection against intranasal challenge with live B. pertussis. i.v. administration of B. pertussis-pulsed DC resulted in a significant reduction of levels of B. pertussis in the lungs of infected mice (Table 1) compared to results for mice that had received mock-treated DC. In two of three experiments the bacteria were completely eradicated. Mice treated i.n. with B. pertussis-pulsed DC showed no differences in bacterial load compared to results for untreated mice (Table 1, experiment 1).

Tissue distribution of ⁵¹Cr after i.v. and i.n. delivery of Cr-labeled DC. Since i.v. delivery of B. pertussis-pulsed DC induced protective immunity against infection and i.n. administration did not, we compared the localization of DC 24 h after i.v. or i.n. delivery of ⁵¹Cr-labeled DC. After i.v. ⁵¹Cr-DC injection, a considerable proportion of the radioactivity was found in the liver. High levels of radioactivity could also be detected in the kidneys, lungs, and spleen. Following i.n. delivery of ⁵¹Cr-DC, the majority of radioactivity was found in the lungs. Almost no ⁵¹Cr could be detected in the blood or in any internal tissues, including secondary lymphoid organs such as the lymph nodes and spleen, indicating that neither DC nor their contents reached the appropriate sites for induction of an immune response (Table 2).

Protection against B. pertussis infection correlates with IgG antibodies in serum. To determine whether DC vaccination could induce a systemic antibody response, titers of B. pertussis-specific IgG were measured in serum following administration of DC. Using correlation analysis, we noted that serum IgG responses correlated with protection against infection (P < 0.01). The levels of IgG in serum after three i.v. vaccinations with B. pertussis-pulsed DC were very strong, indeed so strong that no booster effect was observed following the B. pertussis challenge (Fig. 1). Since isotype switching to IgG requires cytokines secreted from CD4⁺ T cells (6), this suggests that antigen-specific T-cell priming has occurred.

View larger version (17K): [in this window] [in a new window]   FIG. 1.   B. pertussis-specific IgG titers in serum before (filled bars) and after (empty bars) challenge with live B. pertussis. Data are expressed as means + standard deviations. Triple asterisk, P < 0.001; double asterisk, P < 0.01, nontreated versus treated mice. Heterol., heterologous.

Protected animals are primed for local antibody production in the lungs. Protection against B. pertussis infection involves mainly CD4⁺ T-cell function, although B cells are also required for full protection and cannot be replaced by immune serum (12, 18). Since local antibody production is of importance in protection against mucosal pathogens (19, 27), we investigated whether treatment with B. pertussis-pulsed DC was able to prime for local production of antibodies in the lungs. We found that animals that were pretreated with antigen-pulsed DC (n = 3) exhibited high numbers of B. pertussis-specific IgG ASC (365 ± 115 ASC/10⁶ MNC; P < 0.05) following a challenge with live bacteria and lower numbers of specific IgA ASC (26 ± 4 ASC/10⁶ MNC; P < 0.001). Animals that were not pretreated with antigen-pulsed DC failed to show any B. pertussis-specific ASC. The local antibody production observed might therefore be an important contributor to the protection against B. pertussis infection. The majority of ASC detected in the lungs following DC vaccination and bacterial challenge consisted of IgG-producing cells. Since IgG constitutes the predominant antibody isotype in the lower respiratory tract and IgA predominates in the nasal and upper tracheal secretions (3), our finding fits with the general consensus.

Heat-killed B. pertussis is ingested by DC and presented to T cells. To establish that the B. pertussis-specific systemic and mucosal antibody production observed was not due to unprocessed bacteria that were attached to the surface of DC and transferred to the host animals, we confirmed that B. pertussis cells were indeed phagocytosed by the DC. For this purpose, heat-killed FITC-labeled bacteria were incubated with DC, washed, and mixed with ethidium bromide to distinguish between extracellular and intracellular bacteria. When mixed with ethidium bromide, external bacteria fluoresced in orange while internalized bacteria were protected by the cellular plasma membrane and kept their green color intact. Fluorescence microscope analysis showed that DC ingested heat-killed bacteria and that virtually no bacteria were found attached to the DC cell surface (Fig. 2). Incubation with the phagocytosis-inhibiting substance cytochalasin B inhibited the ingestion of FITC-labeled B. pertussis by DC (data not shown).

View larger version (167K): [in this window] [in a new window]   FIG. 2.   Overlay of fluorescence and light-microscopic image of DC pulsed with FITC-labeled B. pertussis and then counterstained with ethidium bromide. Ingested bacteria remain green, whereas extracellular bacteria become orange.

View larger version (8K): [in this window] [in a new window]   FIG. 3.   IFN- production by B. pertussis-specific T cells cocultured with B. pertussis-pulsed DC (black bars) or with unpulsed DC (white bars).

In vitro pulsing of DC with heat-killed DC gives rise to production of interleukin 12 (IL-12) but not of IL-10. Unfortunately, vaccination with DC, besides priming for B. pertussis-specific responses, also primed for fetal calf serum components, thereby impeding us in performing in vitro T-cell analyses. However, in vitro pulsing of DC with heat-killed B. pertussis gave rise to high levels of IL-12 production (>240 pg/ml versus 30 pg/ml in mock-pulsed DC supernatants; results not shown), which is known to prime for Th1 responses in vivo (22), but no detectable levels of IL-10. Furthermore, overnight pulsing of DC with heat-killed B. pertussis led to the up-regulation of MHC and costimulatory molecules (the frequency of DC expressing MHC class II increased by 7 to 10%, CD80 by 6 to 8%, CD86 by 13 to 15%, and CD40 by 5 to 8%) compared to results with mock-pulsed DC, indicating that B. pertussis pulsing of DC specifically up-regulates MHC and costimulatory molecules that are required for the induction of a CD4⁺ T-cell response in vivo.

Autologous as well as heterologous DC protect mice against B. pertussis infection. Even though vaccination with DC has been utilized with both animals and humans against intracellular infections and tumors, it is not clear if it is indeed the injected DC that present the antigen to the immune system or if the antigen is passed on to other antigen-presenting cells through either cell-to-cell contact (28), exosomes (10), or phagocytosis of the injected DC (9). In an attempt to investigate this matter, we injected mice with antigen-pulsed, MHC-mismatched DC, which presumably cannot by themselves activate the host's T cells against peptide antigens, and measured the level of protection and the subsequent serum antibody response in recipient mice. Surprisingly, heterologous DC vaccination gave rise to protection against B. pertussis, evidenced by a significant reduction in bacterial load in the lungs (4.4 ± 0.35 log bacteria per lung in heterologous DC-treated mice, compared to 6.28 ± 0.28 in untreated mice) and also induced an antibody response comparable in magnitude to that obtained with autologous DC (Fig. 1). The strong antibody response after vaccination with DC suggests that unprocessed antigen is passed from heterologous to autologous DC, since B cells require native antigens for activation. This finding has obvious implications for the development of immunomodulating therapies. In summary, we have shown that DC vaccination can prime for a mucosal response and protect against a presumed strictly noninvasive mucosal pathogen. Furthermore, heterologous as well as autologous DC could be utilized for this purpose.