|Home | About | Journals | Submit | Contact Us | Français|
The use of inactivated vaccines associated with suitable adjuvants has been demonstrated to confer a good level of protection against Chlamydophila abortus. However, the basis of the immune protective response induced by these vaccines has been poorly studied. B cells act as an immune regulatory population during primary infection by C. abortus. Thus, it was considered of interest to study the role of B cells in an infection after immunization with a killed vaccine. For this, C57BL/6 and B-cell-deficient mice were immunized with a killed vaccine against C. abortus using QS-21 as the adjuvant. After challenge, the course of infection was established by analysis of morbidity, C. abortus burden in the liver, and histopathological changes. The immune response induced was studied by real-time PCR techniques. Experiments involving transfer of immune serum from vaccinated or previously infected mice were also carried out. The lack of B cells reduced the protection conferred by the QS-21 adjuvant vaccine. Vaccinated B-cell-deficient mice showed a 1,000-fold-greater bacterial burden in the liver than their wild-type counterparts. Obvious differences existed in the liver, where a severe neutrophilic reaction and extended areas of necrosis were observed with vaccinated B-cell-deficient mice. An analysis of the immune response pointed to a significant increase in inflammatory cytokines and chemokines and the deficient production of transforming growth factor beta. The transfer of antibodies restored the level of protection. This study demonstrates that B cells play a crucial role in controlling C. abortus multiplication and prevent an exacerbated inflammatory response.
Chlamydophila abortus is a gram-negative obligate intracellular bacterium responsible for enzootic abortion in small ruminants. The disease caused by the bacteria is the most commonly diagnosed cause of ovine abortion in a number of western countries, where it is responsible for severe economic loss. C. abortus can also induce abortion in pregnant women as a result of contact with aborting sheep and goats (27).
Mouse models have been widely used in studies of the pathogenesis and immune response induced by members of the family Chlamydiaceae (6, 15, 19, 37). The use of these models in C. abortus infection has underlined the importance of innate immunity controlling the infection (23), the role of the cellular immune response in its clearance (20), and several aspects of the pathogenesis of C. abortus-induced abortion (4, 8). These experimental studies demonstrated that after systemic spreading of C. abortus, the establishment of an effective immune response can eliminate the infection in all organs except the placenta of pregnant animals, where multiplication of the bacteria induces abortion (12). The effective immune response observed for the spleen and liver involves neutrophils and NK cells that act as a first line of defense (9) and in the recruitment of other leukocyte subpopulations (35). Also, a strong type 1 specific immune response is rapidly induced, involving the production of proinflammatory cytokines, especially gamma interferon (IFN-γ) (33), and the activation of T cells, with CD8+ T cells playing an important role in the resolution of the infection (14). The role of the humoral immune response during the C. abortus infection has been less studied, but an immunomodulatory role suggested for B cells in the early events of the primary infection would protect mice against an exaggerated inflammatory response (7). Furthermore, it has been reported that C. abortus infection can induce the production of neutralizing antibodies that provide protection after a passive transfer (13, 21).
Research into an effective vaccine against C. abortus has been carried out in several laboratories. A temperature-sensitive mutant strain of C. abortus (strain 1B) was developed as a live vaccine (44) providing a good level of protection. However, concerns still remain over the safety of using live attenuated vaccines, particularly as regards the possibility of colonization of the human placenta or of the attenuated strain reverting to virulence, with the consequent potential to cause disease and abortion in the vaccinated animal (28). Killed vaccines against C. abortus are safer and have been demonstrated to confer effective protection when they are used with a suitable adjuvant (16). The immune mechanisms involved in the protection conferred by the killed vaccines against C. abortus are poorly understood, and since B cells play a substantial role in the development of the response against the primary infection, it would be valuable to analyze the role of B cells in the immune response induced by a killed vaccine in order to develop more effective and safer vaccines.
Eight-week-old female C57BL/6J (wild-type [WT]) mice and B-cell-deficient (KOB) mice, presenting a disruption of the transmembrane portion of the μ chain gene, in the same genotype were used in this study. WT mice were purchased from Harlan UK Limited (Blackthorn, United Kingdom) or Charles River (Barcelona, Spain). No differences were found between animals from the two sources. KOB mice were obtained from Jackson Laboratory (Bar Harbor, ME). They were free of common viral and bacterial pathogens, according to the results of routine screening procedures performed by the suppliers.
An inactivated vaccine with a purified derivate of saponin (QS-21; Antigenics, Inc., Framingham, MA) as an adjuvant, which provided optimal protection in results of previous studies (10, 16), was used in this study. The vaccine was prepared as described by Caro et al. (16). Briefly, the AB7 strain of C. abortus (45) was cultured in McCoy cell monolayers and subsequently purified in a Renografin 76 (Schering, Lys-Lez-Lannoys, France) gradient as described by Buendía et al. (11). The obtained purified C. abortus was inactivated by treatment with binary ethylenimine (Sigma, Madrid, Spain). After inactivation, the bacteria were mixed with the QS-21 adjuvant. Each dose of vaccine contained 15 μg of protein from binary ethylenimine-inactivated C. abortus in a 0.2-ml volume and was inoculated subcutaneously. Mice received booster injections 12 days after the first vaccination in the same conditions.
The AB7 strain of C. abortus was used in the challenge. The bacteria were propagated in the yolk sacs of developing chick embryos and titrated by counting inclusion-forming units (IFUs) on McCoy cells as described by Buendía et al. (9). Standardized aliquots were frozen at −80°C until use.
WT and KOB mice vaccinated with QS-21 as the adjuvant (QS-WT and QS-KOB mice, respectively) were challenged 8 days after the booster injection of the vaccine by intraperitoneal (i.p.) injection with 106 IFUs of C. abortus in 0.2 ml of phosphate-buffered saline, pH 7.2. Control nonvaccinated WT and KOB mice (C-WT and C-KOB mice, respectively) were infected with the same strain and dose. After infection, mice were caged individually and weighed daily to establish the morbidity of the infection. At day 4 postinfection (p.i.), five mice per each group were killed, and samples of organs and serum were collected under aseptic conditions. All the experiments were approved by the Bioethical Committee of the University of Murcia.
The course of infection was also evaluated by counting IFUs from the liver after isolation of C. abortus on McCoy cell monolayers. The caudate lobe of the liver was examined, and the number of IFUs/g was calculated. The detection limit was 2.6 log IFUs per sample.
To measure the liver-associated enzyme alanine transaminase (ALT) in serum, a protocol modified from a commercial kit (Sigma) was employed. Briefly, in a 96-well plate, 4 μl of serum was added to 20 μl of dl-alanine (0.2 mol/liter) and α-ketoglutaric acid (1.8 mmol/liter). The plate was shaken, and the mixture was incubated for 30 m at 37°C; 20 μl of 2,4-dinitrophenylhydrazine was then added, and the was mixture incubated a further 20 min at room temperature. The reaction was halted by the addition of 200 μl 0.4 N NaOH, and sample absorbances were measured at 492 nm after 5 min.
Livers were collected at the time of necropsy and fixed in formalin for histopathological analysis or zinc fixative (BD Pharmingen, San Diego, CA) for immunohistochemistry. After being dehydrated and embedded in paraffin at 56°C, sections (5 μm) were cut, stained with hematoxylin and eosin, and analyzed for histopathological changes. Immunohistochemistry was carried out using a polyclonal anti-CD3 antibody (Dako, Carpinteria, CA) or a monoclonal rat antibody against murine neutrophils (clone NIMP-R14; Santa Cruz Biotechnology, Inc., Santa Cruz, CA). The reaction was performed using the avidin-biotin-peroxidase complex method according to the instructions of the manufacturer (Vector Laboratories, Burlingame, CA). The incidence and location of neutrophils and T cells in the inflammatory infiltrate of the liver were ascertained by counting the positive cells in 20 fields (magnification, ×400) from sections of each mouse.
A sandwich enzyme-linked immunosorbent assay (ELISA) was used for measuring this cytokine. The capture antibody was R4-6A2, and the biotinylated detection antibody was XMG1.2 (BD Pharmingen). Biotin-conjugated antibody was detected with a horseradish peroxidase-streptavidin conjugate (BD Pharmingen) and a soluble substrate, ABTS (2,2′-azinobis[3-ethylbenzthiazoline-6-sulfonate]), obtained from Sigma. The optical density was read at 414 nm.
In order to test the protection conferred by antibodies, experiments involving serum transfer to nonvaccinated WT or KOB mice were carried out. The different sera used in the experiments were (i) immune serum obtained from WT mice infected i.p. with 2 × 105 IFUs of the AB7 strain of C. abortus and killed 30 days p.i.; (ii) serum from WT mice immunized with the QS-21 adjuvant vaccine as described above and killed 30 days after booster inoculation; (iii) nonimmune serum obtained from noninfected and nonvaccinated WT mice, serving as a negative control for protection; (iv) the previously described (21) neutralizing monoclonal antibody FA2H10, directed against the oligomeric form of the major outer membrane protein (MOMP) of C. abortus, serving as a positive control of protection. Sera from several mice were pooled and heated at 56°C for 30 min for complement inactivation. In order to assess whether the immune sera had a comparable amount of C. abortus-specific antibodies, the titers of samples from the pooled sera were determined by an ELISA using purified elementary bodies of C. abortus as the antigen. The serum transfers were carried out i.p. with 200 μl of undiluted serum for each WT or KOB mouse. In the case of FA2H10, a dose of 0.5 mg of the purified antibody in 200 μl of phosphate-buffered saline was inoculated i.p. Serum was transferred to five mice in each group. Mice were challenged 6 h after the serum transfer, as described in Materials and Methods. Groups of vaccinated and nonvaccinated WT and KOB mice were also infected and used as nontransferred controls.
Liver tissue pieces were homogenized, and total RNA was extracted using an RNeasy kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions. RNA was eluted with 30 μl of RNase-free water and quantified by spectrophotometry. Total RNA was reverse transcribed to cDNA with random hexamers using a High Capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA) following the manufacturer's instructions. Briefly, approximately 1 μg of total RNA was reverse transcribed in a final volume of 20 μl after incubation at 37°C for 120 min. The resulting cDNA was stored frozen (−20°C) until being assayed by real-time PCR.
Expression of proinflammatory (IFN-γ and tumor necrosis factor alpha [TNF-α]) and anti-inflammatory (interleukin-10 [IL-10] and transforming growth factor beta [TGF-β]) cytokines and two chemokines (CCL2 and CXCL10) produced early during the chlamydial infection (1) were analyzed for each in each group of mice. For each target gene, different primer concentrations (300 nM to 900 nM) were tested to optimize the PCR amplification, and efficiencies of target and endogenous controls proved to be approximately equal. Aliquots (2.5 μl) of each primer were added to a 25-μl reaction as specified below. β-Actin was used as an endogenous control. Primers for the murine IFN-γ, CCL2, and CXCL10 were designed using Primer Express software (Applied Biosystems), while primer sequences for IL-10, TNF-α, and TGF-β had been described elsewhere (40) Sequences and concentrations of the primers used are shown in Table Table1.1. Forty cycles of amplification were performed for cDNA samples; all reactions were performed in triplicate, with cycle parameters as follows: 15 s at 95°C and 1 min at 60°C. PCR and data analysis were performed with an ABI PRISM 7500 sequence detection system (Applied Biosystems). Relative quantification of gene expression was calculated using liver samples from an uninfected nonvaccinated animal from each mouse strain as the calibrator sample. The calculation method used for relative quantification was the comparative CT method (ΔΔCT).
Data were expressed as means ± standard deviations (SD). The significance of the differences in the results for the different parameters between groups was determined by two-tailed Student's t test. A P value of <0.05 was considered significant. All the experiments were carried out twice in order to assess the repeatability of the techniques.
After challenge, all the mouse groups except the QS-WT group showed clinical signs (lethargy, ruffled fur) starting on day 2 p.i. When morbidity, expressed as weight loss, was analyzed (Fig. (Fig.1a),1a), a substantial loss was observed for the three groups that showed signs of illness, while no weight loss could be observed for the QS-WT group, in agreement with its healthy appearance. The greatest weight loss was observed for the QS-KOB group, which showed significant differences from their WT counterparts. The QS-KOB group also showed increased morbidity compared to the nonvaccinated C-KOB group, although the difference was not significant (P = 0.076).
After necropsy on day 4 p.i., no macroscopic lesions could be observed for vaccinated or nonvaccinated WT mice, although macroscopic hepatic lesions were observed for the B-cell-deficient mice. In the liver samples from the C-KOB group, very small whitish areas less than 1 mm in diameter could be observed, while the QS-KOB mice displayed large yellowish-white lesions (several millimeters in diameter) resembling necrotic areas.
A histopathological analysis of the liver lesions showed differences between the groups of mice; these differences were related to the presence of necrosis, number and size of inflammatory foci, and cellular composition of the foci. Numerous inflammatory foci were observed for the C-WT group (Fig. (Fig.2a).2a). The inflammatory foci in the C-KOB mice were larger, and areas of coagulative necrosis that affected the hepatocytes could be observed (Fig. (Fig.2b).2b). Among the vaccinated groups, QS-WT mice had scarce and very small inflammatory foci (Fig. (Fig.2c),2c), while QS-KOB mice showed larger inflammatory foci and the distinctive presence of extensive areas of coagulate necrosis (Fig. (Fig.2c).2c). Thrombotic lesions in the vessels close to the necrotic areas were observed. In relation to the composition of the inflammatory foci, these foci were composed mainly of macrophages and neutrophils, with scarce lymphocytes in the C-WT group (Fig. (Fig.2e);2e); the number of neutrophils in the foci was higher in the C-KOB group. The small foci observed for the QS-WT group were formed almost exclusively by lymphocytes. The foci in the QS-KOB were composed of numerous lymphocytes but also a substantial number of neutrophils and some macrophages (Fig. (Fig.2f);2f); lymphocytes could also be observed scattered in the liver parenchyma in this group.
The grade of severity of the liver lesions was confirmed by the presence of the enzyme ALT in serum. An analysis of this parameter (Fig. (Fig.1b)1b) showed that QS-KOB mice had significantly increased liver damage compared with that for QS-WT (P = 0.008) and C-KOB (P = 0.037) mice.
An analysis of the C. abortus burden in the liver (Fig. (Fig.1c)1c) demonstrated that B cells play a key role in the protection conferred by the QS-21 adjuvant vaccine, since the QS-KOB group showed a significant increase (1,000-fold) in the number of C. abortus IFUs in the liver compared with the number for the QS-WT group. The chlamydial burden in the liver of the QS-KOB group was very similar to that of the immunocompetent nonvaccinated mice (C-WT group). The presence of IFN-γ in serum is closely related to the C. abortus organic burden, as has been reported for previous experimental murine models (8, 16). In the present study, B-cell-deficient mice showed significantly increased values with respect to their WT counterparts (Fig. (Fig.1d),1d), the differences being especially evident in the case of the QS-KOB and QS-WT groups. No significant differences in relation to the chlamydial liver burden or the presence of serum IFN-γ could be found between C-KOB and QS-KOB groups to explain the increased morbidity and severity in the lesions described above.
An analysis of the immunohistochemical staining confirmed that neutrophils were the predominant leukocyte subpopulation in the inflammatory foci of the nonvaccinated groups (Fig. (Fig.3a),3a), with the C-KOB mice showing a significant degree of neutrophilia compared with C-WT mice. A substantial number of neutrophils in the liver of the QS-KOB group was observed, contrasting with the lack of these leukocytes in QS-WT mice. A characterization of the immune infiltrate in the liver also showed a significantly higher number of T cells in QS-KOB mice compared with the other groups (Fig. (Fig.3b).3b). T cells were found in the inflammatory foci but also were disseminated among the hepatocyte layers. In the nonvaccinated groups, T cells appeared in the periphery of the inflammatory foci and surrounding blood vessels, while in the QS-WT group, T cells were observed only in the inflammatory foci. The sum of both leukocyte populations showed that their number in the liver of QS-KOB mice was 2.3- and 1.8-fold higher than in the nonvaccinated C-WT and C-KOB groups, respectively, and was 7.8-fold higher than that in the QS-WT group.
The level of expression of proinflammatory cytokines (IFN-γ, TNF-α), anti-inflammatory cytokines (IL-10 and TGF-β), and selected chemokines (CCL2 and CXCL10) was studied by real-time PCR (RT-PCR) (Fig. (Fig.4)4) to determine whether the KOB groups had deficiencies in their response to C. abortus. Increased proinflammatory cytokine and chemokine expression was detected in the QS-KOB group, with significant differences in relation to the QS-WT group, except in the case of CXCL10 (P = 0.066). Both groups of vaccinated mice also showed differences in the expression of anti-inflammatory cytokines, with a significantly higher level of IL-10 production in the QS-KOB group and a higher level of production of TGF-β in the QS-WT group. An analysis of the RT-PCR results obtained for the nonvaccinated mice showed statistically significant differences between C-WT and C-KOB in the expression of CXCL10 (higher in C-WT) and IL-10 (higher in C-KOB); furthermore, C-WT showed a higher, but not significantly higher, level of IFN-γ than that for the C-KOB group.
Experiments carried out by transfer of immune serum against C. abortus demonstrated that antibodies alone were able to confer a similar level of protection to that obtained through immunization with the QS-21 adjuvant vaccine. Heat-treated sera obtained from previously infected or vaccinated mice, containing similar titers of C. abortus-specific antibodies (Fig. (Fig.5a),5a), significantly reduced the C. abortus burden in the liver, although to a lesser extent than did the use of a purified neutralizing monoclonal antibody (Fig. (Fig.5b),5b), while groups of mice which had undergone nonimmune serum transfer showed no changes in relation to C-WT or C-KOB mice. A histopathological analysis of the liver (Fig. 6a and b) confirmed the presence of reduced pathology (similar to results for the QS-WT group) in the immune serum transfer groups. Other studied parameters, such as weight loss (morbidity) or presence of IFN-γ in serum, were found to be also similar in mice transferred with immune serum (previously infected or QS-vaccinated mice) and in the QS-WT group (data not shown).
The results of the present study point to a substantial role for the humoral immune response against C. abortus in vaccinated mice. First, B cells seem to be a key subpopulation in the control of bacterial multiplication in previously immunized animals, since a 1,000-fold increase in the liver chlamydial burden was observed in the absence of B cells. However, B cells and specific antibodies seemed to be unconnected with any immunity to the species in the family Chlamydiaceae in the first studies that were performed using mouse models for the genital tract (47) or for pulmonary infection (48). The organ-specific location of these previous reports was clearly different than that for the systemic infection with C. abortus described in this study and could explain the discordant results. More-recent studies have suggested an important role for B cells and antibodies in the immune response against C. abortus (7) and in protection against other chlamydial species. In fact, the importance of antibodies in protection against chlamydial challenge has been demonstrated in several recent studies using a mouse model of infection with Chlamydia muridarum (Chlamydia trachomatis MoPn strain) in which the following was demonstrated. (i) Immune serum conferred marked protective immunity against the genital tract reinfection (38). (ii) FcR−/− mice suffered a greater secondary infection, marked by greater multiplication of chlamydiae in the genital tract and, consequently, a more severe ascending disease (36). (iii) The passive transfer of a monoclonal antibody conferred protection against C. muridarum challenge both in immunocompetent and immunodeficient mice (41). Similarly, B cells and antibody-mediated responses contribute to immunity against a number of intracellular microbial infections through a variety of mechanisms (17, 18). Pathogens in which a significant role has been demonstrated for the humoral immune response include viruses (22), bacteria (2, 31), fungi (43), and protozoa (26). The direct action of antibodies may involve one or more of the following mechanisms (36): (i) antibody neutralization of free infectious particles (elementary bodies), (ii) activation of the lytic components of the complement, (iii) antibody augmentation of chlamydial particle ingestion and the destruction by phagocytes following opsonisation of free particles or infected cells; and finally, (iv) the destruction of infected cells expressing chlamydial antigens on the surface by NK cells or macrophages via FcR-dependent cellular cytotoxicity.
A somewhat surprising finding of this study was the increased morbidity and pathology observed for the vaccinated B-cell-deficient mice compared with the nonvaccinated group. The liver damage was accompanied by an increased number of T cells in the lesions. The most likely hypothesis to explain this damage is that in the absence of antibodies to eliminate the organism, the exacerbated C. abortus multiplication would provide sufficient antigen to elicit a strong memory cell-mediated response. Antibody itself downregulated the T-cell response by eliminating the microorganism. The contribution of the T-cell response to the development of the disease is a characteristic feature of other chlamydial infections (24). The analysis of the expression of proinflammatory cytokines and chemokines demonstrated an acute inflammatory response in the absence of antibodies that was not observed for the vaccinated WT mice. This fact fits nicely with the hypothesis that the excessive number of bacteria is the cause of the increased pathology. Furthermore, the presence of some of these proinflammatory chemokines has been recently reported as nonprotective, and even deleterious, in the outcome of the chlamydial infection (39).
Increased pathology in the absence of B cells has been demonstrated in previous studies of the primary response to C. abortus from our laboratory (7) and also in other experimental models of infection with intracellular pathogens (3, 31, 46). The increased pathology in all the models was associated with substantial neutrophilia in the tissues. The persistence of high levels of antigen in the absence of antibodies could maintain the influx of neutrophils to the inflammatory foci as a source of tissue injury. However, C. abortus infection of RAG-2 knockout mice (which lack mature T and B cells) showed no extensive necrotic lesions in the early stages of C. abortus infection (A. J. Buendía, M. R. Caro, and J. Salinas, unpublished data), suggesting that the increased tissue damage observed in this study is, at least in part, T-cell dependent. A similar response involving the recruitment of CD3+ T cells to the lesions in the absence of B cells has recently been reported in a model of primary aerosol infection with Mycobacterium tuberculosis (31). FcR signaling in the regulation of the CD4+ T-cell-mediated inflammatory response has been reported in the case of Schistosoma mansoni (25), since both B-cell-deficient mice and FcR−/− mice showed markedly exacerbated hepatic granuloma formation and failed to downmodulate the pathology during the chronic stage of the disease. An FcR-dependent mechanism to regulate the T-cell response against M. tuberculosis has recently been confirmed (30).
An alternative hypothesis that cannot be ruled out is a direct regulatory role for B cells. The existence of a subset of B cells (regulatory B cells or Bregs) with the capacity to downregulate the T-cell response has been proposed in recent years (5, 34). There is increasing evidence that B cells exert their regulatory role through anti-inflammatory cytokines, such as IL-10 production (32) or TGF-β1 secretion (42), and their regulatory activity was established in the response against microbial pathogens, such as M. tuberculosis (29). In our study, B-cell-deficient mice (naïve and vaccinated) showed a higher level of expression of IL-10 RNA in the liver than their immunocompetent counterparts; in contrast, TGF-β1 RNA expression in the liver was lower in B-cell-deficient mice, especially in the vaccinated group, confirming previous results which demonstrated its low presence in the serum of B-cell-deficient mice after primary infection with C. abortus (7) and suggesting a hypothetical mechanism for downregulation of the T-cell response in immunocompetent mice.
In conclusion, the findings of this study confirm that future vaccines against C. abortus should consider the need to suitably activate the B cells to provide protective effector antibodies. The presence of protective antibodies could control an exacerbated T-cell response that would otherwise have harmful consequences. Furthermore, this model of infection could be a useful tool for future studies of the interactions between humoral and cellular immune responses.
This work was supported partially by a grant from Ministerio de Educación y Ciencia cofinanced with FEDER funds (AGL2004-06571). N. Ortega was the recipient of a postdoctoral grant from Cajamurcia.
We thank Charlotte Kensil from Antigenics, Inc., for providing the QS-21 adjuvant.
Editor: R. P. Morrison
Published ahead of print on 24 August 2009.