Main

Worldwide, an estimated 90 million sexually transmitted Chlamydia trachomatis infections occur each year1. More than two-thirds of these cases occur in the developing world, where diagnostic and treatment services are almost absent. Sub-Saharan Africa and southern and Southeast Asia have particularly high burdens of disease, with an estimated 15 million new cases occurring in Africa and 45 million new cases in southern Asia every year. The prevalence of infection in Asia might be even higher than this estimate, because a recent study in China concluded that 2.5% of people of 20–64 years of age are infected2. Similar prevalence rates (2.1%) have also been documented in a recent population-based study in Britain3. Rates are about twofold higher (4.2%) among a random sample of young adults (18–26 years) in the United States, highlighting a universal epidemiological feature of C. trachomatis — that infection is mainly observed in adolescents and young adults4.

Sexually transmitted C. trachomatis infection is an important public-health concern because of its adverse effects on reproduction1. In women, infection with C. trachomatis causes PELVIC INFLAMMATORY DISEASE (PID) and has long-term consequences — such as infertility, ECTOPIC PREGNANCY and chronic pelvic pain — that are secondary to scarring of the fallopian tubes (caused by SALPINGITIS) and ovaries. In addition, infection with C. trachomatis facilitates the transmission of HIV5 and might be a co-factor in human papilloma virus (HPV)-induced cervical neoplasia6. Because of public-health concerns, programmes to control C. trachomatis have been implemented in many developed countries; these involve the detection of infected individuals through diagnostic testing, which is followed by antimicrobial treatment and tracing of individuals who might have been exposed through sexual contact with the infected person. Although these programmes might control C. trachomatis infection, many regions are now showing an increase in the number of infected individuals7. This increase might reflect, in part, improvements in diagnostic testing and/or changes in sexual behaviour. Alternatively, the administration of antimicrobial agents might be altering the development of natural immunity to C. trachomatis in the population. For example, antimicrobial agents have clearly been shown to blunt the development of immunity to Chlamydia muridarum in mouse models of infection8. Antimicrobial treatment of infected individuals helps to reduce transmission by shortening the average duration of infection. In the absence of antimicrobial therapy, C. trachomatis infections typically last for many months, but they can undergo spontaneous clearance9,10,11, which is associated with increasing age and duration of infection and is presumed to be immune mediated9,12.

Why C. trachomatis infections take so long to clear is not certain, but it might be a consequence of the many immune-evasion strategies of the organism (Box 1). Data from animal models of infection indicate that clearance depends on the recruitment of effector T cells and their clonal expansion to a crucial threshold in the genital tract13, and it might be that reaching this threshold takes many months in humans. Taken together, these observations indicate that, by shortening the average duration of infection, control programmes that involve antimicrobial treatment might be blunting the development of immunity to C. trachomatis and thereby increasing the susceptibility of the population to C. trachomatis transmission.

This hypothesis for why the rates of infection with C. trachomatis increase in the face of control programmes needs to be validated, but if it is correct, it has obvious implications for the need for a vaccine to adequately control this infectious disease. Because C. trachomatis is such an important pathogen from a public-health perspective and because current programmes for the control of C. trachomatis infection are not affordable for much of the developing world and might have an inherent weakness, vaccine development has been identified as essential to controlling infection with C. trachomatis.

In general, a vaccine against C. trachomatis needs to elicit protective T-cell and B-cell immunity in the genital-tract mucosa. Mouse models of genital infection with C. muridarum, which has most of the same genes as the human strains of C. trachomatis14, have provided information on the immune mechanisms of clearance of infection and resistance to re-infection, and these models seem to be useful for analysing immunity to C. trachomatis in humans12,15. However, there are several important differences between C. muridarum and C. trachomatis that might affect the immunobiology of infection. First, C. trachomatis infection in humans is much more prolonged than C. muridarum infection in mice: mice generally resolve infection after 4 weeks, whereas in humans, C. trachomatis infection can last several months before spontaneous clearance9,10,11. Second, immune-evasion strategies also differ such that some strains of C. trachomatis use tryptophan biosynthesis to escape interferon-γ (IFN-γ)-mediated defence mechanisms of the host (Box 1), whereas C. muridarum does not14,16. Last, C. trachomatis shows substantial allelic variation of its dominant surface protein — the major outer-membrane protein (MOMP) — whereas C. muridarum has a single allele14. Because MOMP seems to be an important target of immunity, the specificity of immunity to different serovars (strains) of C. trachomatis12 cannot be studied in the C. muridarum model. Although these differences limit the direct extrapolation of findings from C. muridarum infection to C. trachomatis infection, the mouse model has provided information about the immunobiology of C. trachomatis and is guiding the development of a vaccine against infection with this organism.

Here, we review the data generated from studies of C. muridarum genital-tract infection of mice and the similar observations obtained from studies of human infection with C. trachomatis that have led to our current understanding of the immunology of infection with Chlamydia spp. Understanding the immunological basis of immunity to Chlamydia spp. and identifying correlates of protective immunity will provide a rational foundation for the design of a vaccine against infection with C. trachomatis17. Because T-cell immunity is central to both mouse and human immunity to Chlamydia spp., we describe the antigens derived from C. trachomatis and C. muridarum that are important for eliciting T-cell responses. Last, we describe how these results, together with recent findings from studies of multisubunit vaccines administered to C. muridarum-infected mice, are informing the design of a vaccine against C. trachomatis for use in humans.

Infection process

C. trachomatis is an obligate intracellular bacterium that causes several sexually transmitted diseases in humans18 (Table 1). C. trachomatis normally infects the single-cell columnar layer of the epithelium in the endocervix of women (Fig. 1) and the urethra of men. Inside epithelial cells, Chlamydia spp. undergo a unique developmental cycle that produces infective forms (known as elementary bodies), which then infect neighbouring epithelial cells (Fig. 2). At the site of mucosal infection, intense inflammation that is characterized by redness, oedema and discharge can occur, resulting in the clinical syndrome of MUCOPURULENT CERVICITIS in women and NON-GONOCOCCAL URETHRITIS in men19. However, despite initiating local inflammation, C. trachomatis infection remains subclinical in a high proportion of infected individuals (70–90% of women and 30–50% of men)19. Asymptomatically infected women can show signs of disease: in general, mucopurulent endocervical discharge, HYPERTROPHIC CERVICAL ECTOPY and friability (that is, easily induced bleeding of the cervical epithelium)20. Clinical symptoms include dysuria, abnormal vaginal discharge, abnormal menstrual bleeding, postcoital bleeding and lower abdominal pain19. In some untreated women (20–40%), infection ascends the endometrial epithelium to the fallopian tubes, where C. trachomatis can establish persistent infection and cause PID. Overall, 11% of women with PID develop tubal factor infertility and 9% develop ectopic pregnancies21. Moreover, this risk seems to be higher for those with PID caused by infection with C. trachomatis compared with PID caused by other factors, such as infection with Neisseria gonorrhoeae22.

Table 1 Chlamydia trachomatis serovars and their associated human diseases
Figure 1: Infection of the female genital tract with Chlamydia trachomatis.
figure 1

Chlamydia trachomatis elementary bodies infect the columnar epithelial cells of the cervix, which often causes few or no clinical symptoms. The bacteria can ascend to infect the endometrium and the fallopian tubes, causing pelvic inflammatory disease, tubal inflammation (also known as salpingitis), scarring and occlusion, which can lead to infertility or ectopic pregnancy. The inflammatory reaction is characterized by an influx of macrophages and neutrophils and the formation of immune inductive sites in the submucosa. These inductive sites, which contain B cells, T cells, dendritic cells and macrophages, coordinate the initiation of an acquired immune response, including the deployment of a secretory IgA (sIgA) response. pIgA, polymeric IgA.

Figure 2: The developmental cycle of Chlamydia trachomatis.
figure 2

Chlamydia trachomatis is an obligate intracellular pathogen that resides within a specialized vacuole and has a biphasic developmental cycle140. An infectious, but metabolically inactive, elementary body (EB) is taken up by mucosal epithelial cells. After internalization, the EB is surrounded by an endosomal membrane to form an inclusion — a vacuole formed from normal endosomal-trafficking pathways — which creates a permissive intracellular niche for the replication of C. trachomatis117. Within the inclusion, the EB transforms into a larger metabolically active reticulate body (RB), which divides by binary fission. Within 40–48 hours (hr), the RBs transform back into infective EBs, which are subsequently released from the inclusion vacuole to infect neighbouring cells. In the presence of growth inhibitors, such as interferon-γ, intracellular C. trachomatis bacteria acquire a non-replicating, persistent form, and bacteria in this form differentiate back into infectious forms after removal of the inhibitor.

Immunobiology

Elucidating the immunobiology of infection with Chlamydia spp. is essential for developing a vaccine. A vaccine needs to induce immune responses that are protective and not responses that are associated with persistence of infection or immunopathology. Establishing immune correlates of protection facilitates the identification of protective antigens in animal models of infection and guides Phase I and Phase II trials of immunogenicity in humans. Identification of immune correlates of protection is an important priority in C. trachomatis research, and C. muridarum infection models have begun to shed light on immune correlates of protection against infection with C. trachomatis. The mouse model of vaginal infection (using C. muridarum) has been used to analyse the innate and adaptive responses to infection with C. trachomatis, and it seems to closely mimic acute infection of the genital tract in women12,15.

Cytokines. After infection with Chlamydia spp., epithelial cells produce various pro-inflammatory mediators, including CXC-chemokine ligand 1 (CXCL1), CXCL8 (also known as interleukin-8, IL-8), CXCL16, granulocyte/ monocyte colony-stimulating factor (GM-CSF), IL-1α, IL-6 and tumour-necrosis factor (TNF)23,24. Infected epithelial cells also upregulate expression of the chemokines CC-chemokine ligand 5 (CCL5) and CXCL10, and they secrete cytokines that promote the production of IFN-γ, including IFN-α, IFN-β and IL-12 (Refs 24,25). Infected fibroblasts secrete IFN-α, IFN-β and nitric oxide26, whereas infected macrophages produce TNF and IL-6 (Ref. 27). Most of these are T helper 1 (TH1)-cell cytokines, which have a role in polarizing the immune response to Chlamydia spp. towards a protective TH1-type response24. By contrast, cytokines such as TNF, IL-1α and IL-6 might be involved in the pathology associated with infection with Chlamydia spp.27 Together, these cytokines trigger inflammation and promote the recruitment of immune cells, thereby actively contributing to the development of innate and adaptive immune responses.

Toll-like receptors and dendritic cells. Toll-like receptors (TLRs) detect microbial infection and have an essential role in the induction of innate and adaptive immune responses28. A recent hypothesis states that differential expression and engagement of TLR-family members at the surface of dendritic cells (DCs) influences the type of immune response that is induced by a microbial pathogen28. Infection with C. muridarum has been shown to stimulate DCs to produce IL-12 (a cytokine that polarizes immune responses to TH1-type responses)29,30 and CXCL10 (a chemokine that recruits T cells) and to express CC-chemokine receptor 7 (CCR7; a chemokine receptor that is required for the migration of DCs to local lymph nodes)31. And, although it is not confirmed which particular TLRs expressed by DCs are engaged by Chlamydia spp., TLR2 might have an important role in the activation of DCs by Chlamydophila pneumoniae32. Furthermore, signalling through TLR2, but not TLR4, is associated with increased fallopian-tube pathology in C. muridarum-infected mice27, indicating that engagement of TLR2 is a potential common pathway in both the immunity and immunopathology induced by Chlamydia spp. Given the high level of expression of TLRs by DCs and the ability of DCs to polarize immune responses, the identification of the role of DCs in Chlamydia-specific immune responses is crucial for understanding the type of immune response that is elicited and therefore also for designing a vaccine against infection with C. trachomatis.

DCs have been found in mouse vaginal and cervical mucosae33 and are recruited to the site of inflammation in response to infection with Chlamydia spp.34 Evidence indicates that sampling of microbial antigen across the epithelia of the vagina is accomplished by migratory DCs that carry antigens to peripheral lymph nodes, where antigen is presented to naive T cells35. Mature DCs are highly effective at presenting antigen and priming protective adaptive immune responses. Accordingly, adoptive transfer of DCs pulsed with C. muridarum elementary bodies protects mice against subsequent infection29. Live and inactivated C. muridarum induce different levels of DC maturation, and adoptive transfer of DCs pulsed with live C. muridarum has been shown to be even more effective at providing protective immunity than DCs pulsed with inactivated bacteria36. These observations might help to explain why vaccination with whole inactivated C. trachomatis was only partially protective in human trials37.

Immature DCs and regulatory DCs have also been described to be associated with immune tolerance38 and therefore might have a role in promoting disease pathogenesis, although this has not yet been studied for Chlamydia spp. Studies of DCs that reside in the genital tract will be essential to enable the design of vaccines against infection with C. trachomatis.

Inductive sites. Although the female genital tract (FGT) mucosa lacks the organized lymphoid structures that are found at other mucosal sites (Box 2), such as the PEYER'S PATCHES in the intestine, after infection with Chlamydia spp., immune cells are recruited to the inflammatory site in the FGT in response to chemokines that are secreted by infected epithelial cells. This results in the subsequent accumulation of lymphocytes and other immune cells and the formation of immune inductive sites (Fig. 1), in which naive B and T cells are clonally selected and expanded39. In FGT infection with C. muridarum, these sites form perivascular lymphoid clusters that mainly contain CD4+ T cells40. In women who have a genital-tract infection with C. trachomatis, the inductive sites form lymphoid follicles that mature into germinal centres41. By contrast, in primates with trachoma — an ocular disease caused by C. trachomatis — inductive sites take the form of lymphoid follicles that contain plasma cells, B cells, T cells, DCs, macrophages and neutrophils42. Importantly, systemically circulating lymphocytes also seem to be recruited to the FGT during infection with Chlamydia spp., because the chemokines (CCL5, CCL7 and CXCL10) that attract lymphocytes are abundantly secreted by C. muridarum-infected epithelial cells24 and because the adhesion molecules MADCAM1 (mucosal vascular addressin cell-adhesion molecule 1) and VCAM1 (vascular cell-adhesion molecule 1), which are required for lymphocyte homing from mucosal and systemic inflammatory sites, are highly expressed in fallopian-tube epithelia that are infected with C. trachomatis43. So, it seems that Chlamydia-specific adaptive immune responses occur not only at mucosal immune inductive sites but also at more distant secondary lymphoid structures, such as regional lymph nodes, and immune cells at these sites then migrate to the local inflammatory site41,44,45.

CD4+ and CD8+ T cells. Studies of animal models have clearly established that T cells have a crucial role in the resolution of infection with Chlamydia spp. Accordingly, nude mice cannot control infection, and adoptive transfer of CD4+ or CD8+ Chlamydia-spp.-specific T-cell lines allows these mice to successfully control infection46,47. Specifically, protection in the C. muridarum-infection model seems to be mediated by CD4+ T cells that produce IFN-γ15,48,49, as mice deficient in MHC class II molecules50, CD4 (Refs 50,51), IL-12 (Ref. 52), IFN-γ49 or the IFN-γ receptor53 and mice depleted of C. muridarum-specific CD4+ T cells51 all have a marked inability to control infection. Furthermore, adoptive transfer of C. muridarum-specific CD4+ TH1-cell clones, but not TH2-cell clones, protected nude mice against infection with C. muridarum54.

The role and effector mechanism of Chlamydia-specific CD8+ T cells are less clear. MHC class I peptide presentation to CD8+ T cells is not essential for clearance of infection with Chlamydia spp.: mice deficient in β2-microglobulin resolved infection as efficiently as wild-type mice50,51, and mice deficient in perforin or CD95 (also known as FAS) — which are crucial cytolytic effector molecules of CD8+ T cells — effectively cleared infection with C. muridarum55, implying that CD8+ T cells are not essential for clearance of infection with Chlamydia spp. However, C. muridarum-specific CD8+ T cells efficiently lysed C. muridarum-infected cells when cells were transfected with intercellular adhesion molecule 1 (ICAM1), indicating that, in some situations, CD8+ T cells might be important for the elimination of cells infected with Chlamydia spp.56 Also, adoptive transfer of CD8+ T-cell lines specific for serovar L2 of C. trachomatis protected mice against infection with C. trachomatis through a mechanism involving production of IFN-γ57. So, CD8+ T cells might have a supporting role in limiting infection with Chlamydia spp.

Considerable in vitro and in vivo evidence shows that production of IFN-γ by C. muridarum-specific T cells is essential for clearance of C. muridarum from the genital tract15. Although the effector mechanisms of IFN-γ-mediated control of in vivo infection with C. trachomatis are not completely understood, it is well established that IFN-γ controls the in vitro growth of C. trachomatis through inducing production of the enzyme indoleamine-2,3-dioxygenase (IDO)58. Activation of IDO by IFN-γ leads to the degradation of tryptophan, and lack of this essential amino acid causes the death of C. trachomatis through tryptophan starvation58 (Fig. 3). Recently, it has been shown that genital, but not ocular, serovars of C. trachomatis can use indole as a substrate to synthesize tryptophan in the presence of IFN-γ, which might allow genital strains of C. trachomatis to escape IFN-γ-mediated eradication in the genital tract by using indole provided by the local microbial flora of the FGT16,59. Additional immune effector mechanisms that are induced by IFN-γ include the induction of nitric-oxide production, which inhibits the growth of C. muridarum60, and the promotion of TH1-type protective immune responses, which downregulate non-protective TH2-type responses49.

Figure 3: Inhibition of chlamydial growth by interferon-γ.
figure 3

Interferon-γ (IFN-γ) produced by T cells induces the expression of cellular indoleamine-2,3-dioxygenase (IDO), which degrades tryptophan, thereby resulting in reduced levels of intracellular tryptophan58,123. The lack of tryptophan leads to the death of Chlamydia spp. through tryptophan starvation. However, a population of Chlamydia spp. reticulate bodies (RBs) responds to the lack of tryptophan by acquiring a non-replicating but viable, persistent form. After removal of IFN-γ and replenishment of tryptophan, the persistent forms of Chlamydia spp. bacteria rapidly redifferentiate into infectious elementary bodies (EBs). Some strains of Chlamydia spp. (genital serovars) have a functional tryptophan synthase that can use indole to synthesize tryptophan and therefore bypass this growth-inhibitory mechanism16,59.

B cells. The importance of antibodies in immunity to C. trachomatis was indicated by an early epidemiological observation of an inverse correlation between the amount of IgA in cervical secretions and the amount of C. trachomatis recovered from the cervix of infected women61. In vitro, antibodies specific for C. trachomatis can neutralize infection in tissue culture62. However, high titres of C. trachomatis-specific antibody do not correlate with resolution of infection in humans and, in fact, are more strongly correlated with increased severity of sequelae of infection, such as tubal infertility in women63. Moreover, mice that lack B cells do not show a markedly altered course of primary genital infection with C. muridarum64. By contrast, B cells are probably important for resistance to secondary infection, because mice that have normal numbers of B cells but are depleted of CD4+ and CD8+ T cells successfully resolve secondary infection51,65. Interestingly, mice that lack Fc receptors suffer more severe secondary infection with C. muridarum than wild-type mice, owing in part to impaired cellular immune responses, which indicates that B cells and antibodies might also be important for enhancing protective effector T-cell responses66. These results indicate that, although B cells do not have a decisive role in resolution of primary infections, they might be required to control re-infection. Possible mechanisms for how B cells contribute to immunity to re-infection include antibody-mediated neutralization and opsonization, as well as enhanced antigen presentation to T cells, possibly following Fc-receptor-mediated uptake of antigen–antibody complexes51,67,68.

Overall, these data show that Chlamydia-specific CD4+ TH1 cells, and to a more limited extent CD8+ T cells and B cells, are required to control C. muridarum infection of the genital tract in mice15,69. And, observations from humans infected with C. trachomatis indicate that similar immune effector mechanisms occur in humans12,69. However, despite the mobilization of many immune effectors, infection with C. trachomatis can be recurrent and/or prolonged, which probably reflects the array of immune-evasion mechanisms of this pathogen (Box 1). Immune-avoidance mechanisms might also contribute to pathogenesis and tissue damage, by inducing persistent infection and by enhancing susceptibility to re-infection.

Models of pathogenesis

The pathogenesis of C. trachomatis disease is not completely understood, and mouse models (using infection with C. muridarum) have been less helpful in this area than they have been in elucidating the basis of immunity. Part of this discrepancy might result from the dependence of C. trachomatis pathogenesis on prolonged infection, as C. muridarum does not typically cause long-term infections.

Circumstantial evidence from studies of animals infected with Chlamydia spp. and from observations of humans infected with C. trachomatis repeatedly shows a strong correlation between Chlamydia-specific immune responses, such as antibodies and T cells specific for heat-shock protein 60 (HSP60) from Chlamydia spp., and PID and fallopian-tube pathology70,71. There are two main hypotheses of pathogenesis, and these are not mutually exclusive: first, the immunological hypothesis states that immune responses induce collateral tissue damage that is central to pathogenesis70; and second, the cellular hypothesis states that pro-inflammatory cytokines that are produced by persistently infected cells are the direct cause of tissue damage72.

The immunological hypothesis is supported by the following evidence: first, protective CD4+ TH1 cells preferentially home to the infected fallopian-tube tissue, where they can confer immunity, as well as cause tissue damage25,73,74; second, TH2 cells that are generated in response to infection with Chlamydia spp. might downregulate the protective TH1-type immune responses, thereby promoting persistent infection49,75,76,77; third, host and Chlamydia-derived antigens (such as HSP60) are recognized by autoreactive T and B cells through MOLECULAR MIMICRY78,79,80; and fourth, C. trachomatis-specific CD4+ and CD8+ T-cell epitopes are often identified in C. trachomatis-associated chronic infections, such as reactive arthritis81,82.

By contrast, the cellular hypothesis (based on the deleterious effects of some cytokines) is supported by the finding that pro-inflammatory cytokines, such as transforming growth factor-β, TNF, IL-1α and IL-6, are secreted by cells infected with Chlamydia spp.23,24 So, persistent infection might induce the secretion of pro-inflammatory cytokines, leading to chronic inflammatory cellular responses and tissue damage83,84.

An alternative proposal that might reconcile the two competing hypotheses stems from the observation that IL-10-deficient mice are more resistant to infection with C. muridarum and have a shorter course of infection than wild-type mice76,77. Regulatory T cells produce IL-10 (Refs 85,86) and might be important in the pathogenesis caused by Chlamydia spp. For example, T cells that are reactive to C. trachomatis HSP60 and produce IL-10 have been found in infertile women87 and therefore might be involved in the suppression of C. trachomatis-specific responses, which could contribute to the ability of C. trachomatis to persist. Although T cells with regulatory properties have been described in the mouse FGT88, the effect of regulatory T cells has not been examined in infections with C. trachomatis, and this warrants further investigation.

Vaccine development

Developing a vaccine against C. trachomatis remains a challenge. In part, this results from our poor understanding of the regulation of the immune response in the FGT (which seems to be highly influenced by sex hormones) (Box 3), the lack of adjuvants that target vaccines to the genital mucosa, our limited knowledge of which C. trachomatis antigens induce protective immune responses and the absence of tools to genetically manipulate Chlamydia spp.89 The observation that the immune response is directly or indirectly involved in the pathogenesis of disease caused by Chlamydia spp. also introduces further complexity to the vaccine-development process90. Nonetheless, the substantial progress that has been made in elucidating the immunobiology of C. muridarum infection is greatly facilitating a renewed effort to design a vaccine against infection with C. trachomatis. Selection of defined antigens for a recombinant subunit vaccine that stimulates CD4+ TH1 cells is central to the current design strategy.

Subunit vaccines. Initial human vaccine trials involved intramuscular administration of whole inactivated C. trachomatis elementary bodies37, which led to the development of partial short-lived protection. However, in some individuals, the vaccine seemed to exacerbate disease during re-infection episodes37,90. As a consequence, the focus of C. trachomatis vaccine research has now turned to the production of subunit vaccines that are based on individual C. trachomatis protein antigens, which are administered with adjuvant or other delivery vehicles. As described earlier, T-cell-mediated immune responses are the main requirement for controlling C. trachomatis infection, and several antigens that trigger T-cell responses have been identified in humans and in mice. C. trachomatis proteins that are recognized by CD4+ or CD8+ T cells in various C. trachomatis-related infections57,82,91,92,93,94,95,96,97 are shown in Table 2.

Table 2 Chlamydia trachomatis proteins that are recognized by human or mouse T cells

Selecting antigens. Because immune protection against infection with C. trachomatis is likely to be mediated by immunization with C. trachomatis proteins that are targets of CD4+ and possibly CD8+ T cells, identification of such proteins is particularly important. Considerable progress has been made during the past seven years in the characterization of eight C. trachomatis proteins that are targets for T-cell recognition (Table 2). When the inclusion-membrane-associated protein CrpA (cysteine-rich protein A), which contains C. trachomatis-specific CD8+ T-cell epitopes, is injected into mice, partial protection against challenge with C. trachomatis is observed57. Similarly, Cap1 (class I accessible protein 1) — another C. trachomatis inclusion-membrane-associated protein, which has high homology among the human C. trachomatis serovars — also contains T-cell epitopes, thereby making it a potential vaccine candidate97. However, the most studied and most promising vaccine candidate is C. trachomatis MOMP. MOMP constitutes 60% of the total protein mass of the bacterial outer membrane and is 84–97% identical (at the amino-acid level) between the many C. trachomatis serovars70. MOMP has four variable domains, which contain serovar-specific epitopes, and five constant domains, which are highly conserved between the different serovars and which contain several conserved CD4+ and CD8+ T-cell epitopes98. Another vaccine candidate is C. trachomatis outer-membrane protein 2 (OMP2). OMP2 is also an immunodominant antigen that contains CD4+ and CD8+ T-cell epitopes (Table 2). It is more highly conserved in amino-acid sequence among different C. trachomatis serovars than MOMP99; therefore, in a vaccine, it could provide protection against the different C. trachomatis serovars. Recent experiments have shown that inclusion of OMP2 considerably improves the protective potential of MOMP-based vaccines (discussed later). Other potential protective antigens that contain known T-cell epitopes include HSP60, YopD homologue (homologue of Yersinia pseudotuberculosis YopD), enolase and PmpD (polymorphic membrane protein D) (Table 2). However, whether these T-cell antigens provide immune protection remains to be determined.

Other C. trachomatis T-cell antigens for potential incorporation in a vaccine include secreted components of the C. trachomatis TYPE III SECRETION SYSTEM100 — the principal virulence mechanism of the organism. Because proteins secreted by this system enter the cytosol, they are likely to enter the MHC class I antigen processing and presentation pathway and be targets for recognition by CD8+ T cells. Further candidate vaccine antigens might also be revealed by analysis of peptides eluted from MHC class I and class II molecules expressed by DCs pulsed with Chlamydia spp.36,101

For the design of a vaccine against infection with C. trachomatis, it is also important to consider that some C. trachomatis antigens contain epitopes that might be associated with pathogenic responses that occur through molecular mimicry. For example, C. trachomatis-specific T cells that recognize C. trachomatis OMP2 or HSP60 have been found in patients with reactive arthritis (an autoimmune condition in which HLA-B27-restricted responses are thought to have a role) that was triggered by previous infection with C. trachomatis82 (Table 2). In addition, responses to an OMP2-derived peptide were found to be associated with autoimmune heart disease in a mouse model of C. muridarum-induced myocarditis92. However, this 'pathogenic' epitope is found in the leader peptide of the pro-protein92 and is not likely to be presented to the immune system during natural infection, indicating that an OMP2 protein that lacks the signal sequence might be an acceptable vaccine candidate.

Adjuvants and delivery systems. Because sexually transmitted infections with C. trachomatis are restricted to the genital-tract mucosa, to be effective, a vaccine might need to target the genital mucosal inductive sites or the associated secondary lymphoid tissues. In general, the mucosa-associated lymphoid tissue has been regarded as a compartmentalized immune environment containing inductive sites that interact with effector sites in the same compartment (that is, other mucosae)102. So, mice that were pre-infected intranasally with C. muridarum had enhanced TH1-type protective immunity compared with mice that were infected orally or subcutaneously, and these mice were resistant to re-infection of the genital tract103. By contrast, mice that were immunized intramuscularly with MOMP formulated with the adjuvant ISCOMs (immunostimulating complexes) were found to be better protected against C. muridarum infection of the genital tract than mice that were immunized intranasally with MOMP and ISCOMs104. These results indicate that an appropriate combination of antigen and adjuvant can be successful even if the vaccine is delivered to a non-mucosal site.

Adjuvants and multisubunit vaccines. MOMP has been extensively used in C. trachomatis and C. muridarum vaccination studies, together with a diverse range of adjuvants and vaccine delivery systems, and these studies have shown varying levels of protection105,106,107,108,109,110,111,112. Table 3 lists the results of some C. muridarum vaccination studies. For example, immunization with MOMP DNA has been shown to be highly protective in the lung model of C. muridarum infection113 but not in the genital-tract model110, although at present the reasons for this are not understood. Priming the immune response with MOMP DNA followed by boosting with MOMP protein (formulated with ISCOMs) was found to be highly protective in the lung model of C. muridarum infection108. Although DNA vaccines are a useful experimental tool, their application to human vaccines is uncertain: it has been observed that DNA vaccines are better expressed by transcriptionally active cells of young animals; therefore, they might not be as effective in older humans as in young mice114.

Table 3 Recent results of vaccination trials using Chlamydia muridarum major outer-membrane protein as antigen

The success of a MOMP-based vaccine might depend on several factors, including the presence of the MOMP epitopes in the correct conformation105, the availability of an appropriate vector to deliver the MOMP antigen and the presence of other antigens in addition to MOMP106. This latter possibility is supported by the finding that, although adoptive transfer of DCs pulsed with whole C. muridarum elementary bodies protected mice against infection of the genital tract, adoptive transfer of DCs pulsed with MOMP alone did not29,115. More recently, it was shown that mice immunized with Vibrio cholerae GHOSTS expressing both MOMP and OMP2 were better protected than mice immunized with V. cholerae ghosts expressing MOMP alone106 (Table 3). Mice immunized with both antigens also had a higher frequency of TH1 cells. These results confirm that MOMP alone is probably not sufficient for providing protection, and they support the idea that an effective vaccine is likely to be based on several C. trachomatis antigens106.

Conclusions and future prospects

The protective immune response to infection with Chlamydia spp. is highly dynamic and involves both innate and adaptive immune responses. Infection of mice with C. muridarum has shown that CD4+ T cells, and possibly CD8+ T cells, producing IFN-γ, as well as B cells, are required to clear infection and to prevent re-infection. However, immune responses that are associated with persistent infection with C. trachomatis seem to induce pathology as a result of chronic inflammation and tissue damage. So, a fine balance between protective immunity and immune-associated disease pathogenesis characterizes the host response to infection with C. trachomatis, and this has an impact on the future design of vaccines.

The search for a vaccine against infection with C. trachomatis continues to be a complex task. Nevertheless, progress has been achieved in the past few years and has led to the identification of various protective C. trachomatis antigens as potential vaccine candidates. Although immunization regimens involving priming with DNA vaccines and boosting with protein-based vaccines have been found to be highly protective in mice, their practical application in humans remains unclear. Given that multisubunit protein vaccines seem to be more effective than vaccines based on single antigens, in future, C. trachomatis vaccine candidates are likely to include various antigens.

C. trachomatis vaccine research will continue to focus on the identification of additional C. trachomatis antigens that induce protective T-cell responses and on the mechanisms that promote protective immunity in the FGT, including the role of DCs in antigen uptake and presentation and the role of pro-inflammatory cytokines in influencing the TH1/TH2 response bias. Further data are required to understand the mechanisms that downregulate the immune response in the FGT, including the effects of sex hormones and the menstrual cycle, as well as the possible regulatory effect of particular T-cell populations. Finally, a better definition of human immune-response correlates with C. trachomatis protective immunity and disease pathogenesis needs to remain an important research priority if we are to develop a vaccine against C. trachomatis infection that has protective and not deleterious effects.