Immunopathogenesis of Chlamydia trachomatis infections in women

Transcription

1 FERTILITY AND STERILITY VOL. 79, NO. 6, JUNE 2003 Copyright 2003 American Society for Reproductive Medicine Published by Elsevier Inc. Printed on acid-free paper in U.S.A. SPECIAL CONTRIBUTION Immunopathogenesis of Chlamydia trachomatis infections in women Joseph Debattista, M.Sc., a,b Peter Timms, Ph.D., a John Allan, M.B., B.S., F.R.A.C.O.G., c and Janet Allan, M.B., B.S. c Centre for Molecular Biotechnology, Queensland University of Technology, Brisbane; Sexual Health and AIDS Service, The Prince Charles Hospital Health Service District, Brisbane; and Wesley Hospital, Queensland, Australia Objective: To develop a model of pathogenesis by which Chlamydia trachomatis progresses from acute to chronic infection, and finally serious disease (salpingitis, tubal occlusion). Design: Review of current literature located through web-based Medline searches using key words: Chlamydia trachomatis, immunology, cytokines, heat shock protein, infertility. Result(s): Cell-mediated immune mechanisms appear to be critical in determining whether acute infection is resolved or progresses into chronicity with pathological outcome. What determines the particular immune pathway depends on a range of determinants HLA subtype and human genetics, cytokine profile, infectious load, route of infection, and endocrinology. A clearer picture of the natural history of chlamydial pathology may assist in providing better predictors of those women who may go on to develop significant sequelae after infection. Conclusion(s): Predicting those who may develop serious disease, including infertility, may contribute to improved management of such persons during earlier stages of infection and assist in prevention. (Fertil Steril 2003;79: by American Society for Reproductive Medicine.) Key Words: Chlamydia trachomatis, infertility, immunology, 60-kD heat shock protein Received May 31, 2002; revised and accepted September 27, Reprint requests: Joseph Debattista, M.Sc., Sexual Health and AIDS Service, 270 Roma Street, Brisbane, Queensland 4000, Australia (FAX: ; joedebat@powerup. com.au). a Centre for Molecular Biotechnology. b Sexual Health & AIDS Service. c Department of Reproductive Medicine, Wesley Hospital /03/$30.00 doi: /s (03) GENITAL CHLAMYDIAL DISEASE Chlamydia trachomatis is an obligate intracellular bacteria characterized by a two-phase developmental cycle of replication: an extracellular infectious bacterial form (elementary body), which is endocytosed by eukaryotic cells into a cytoplasmic inclusion, followed by conversion to an intracellular, replicative form (reticulate body), which multiply within the inclusions by binary fission. As the inclusion fills with reticulate bodies (RB), chlamydiae revert back into the metabolically inert elementary bodies (EB) that rupture out of the host cell to infect further host cells. C. trachomatis is now recognized as one of the most common sexually transmissible bacterial infections throughout the world, particularly among persons 25 years of age living in industrialized nations (1). Paralleling this rise in chlamydial infections during recent decades, the prevalence of pelvic inflammatory disease (PID), ectopic pregnancy, and tubal infertility has similarly undergone a steady increase. Estimates of the current infertility rate for Western countries have shown that by the end of 1 year of unprotected sexual intercourse, 10% 15% of couples will fail to conceive (2) with 37% and 85% of infertility in developed and developing countries, respectively, due to tubal factor infertility (3). Chlamydia trachomatis has received significant attention as a primary etiological factor, responsible for approximately 40% 50% of PID and salpingitis (1), 25% of ectopic pregnancies (4), 50% of tubal infertility (2), and 50% of epididymitis in men (5). In a series of Swedish studies (1), 17.2% of women with PID were infertile because of tubal blockage after one or more chlamydial infections. After only one episode of PID, 11.4% became infertile (2) and with each new bout of PID, the infertility rate roughly doubled. In women who experienced one episode of severe PID, fertility was found to be preserved in only 60% of cases. Investigators from at least 13 different cities around the world 1273

2 have documented that tubal blockages are strongly associated with the presence of chlamydial antibody and when these studies are combined, 70% of women with tubal infertility had antibodies to Chlamydia compared with 26% of control women (2). The clinical significance of antichlamydial antibodies was determined in a total of 1,303 subfertile couples by Eggert- Kruse et al. (6). All patients were asymptomatic for genital tract infection yet elevated titers of IgG antichlamydial antibodies were found in 20.8% of all women and 12.6% of men. In women, antichlamydial IgG was associated with tubal infertility. Further studies have strongly supported the link between serological evidence of previous C. trachomatis infections and tubal infertility and ectopic pregnancies (7 10). Seventy-five percent of women with tubal factor infertility (11) have been found to have significantly elevated antibody titers compared to 20% 40% of fertile controls. Antichlamydial antibodies appear to persist for long periods after the infection has cleared and consequently, serology may not distinguish chronic active infection from past infection (7). Persistence of chlamydial IgG and long-term sequelae of PID were studied in 70 women treated for PID 3 6 years previously, 51 due to C. trachomatis infection (12). Chlamydial IgG persisted at stable levels in 43% of women for up to 6 years, 43% of women showed a decrease in IgG titer, and 13% showed an increase. Although chlamydial IgG antibodies may disappear within a few months, persistence at a stable level has been noted for up to 8 years. In this same study, IgA levels in serum correlated with IgG serum levels and with the presence of cervical IgA antibodies. However, although IgG antibodies may denote previous, resolved infection, the presence of IgA in some women has been reported in a number of studies as a marker of chronic infection and linked to pathological sequelae of C. trachomatis infection ascending from the cervix and internalizing within the upper reproductive tract (12 14). Furthermore, C. trachomatis DNA has been detected at sites of inflammation within reproductive tissue in humans with PID and tubal infertility (7, 8, 15 18) and strongly indicates the possibility of chronic persistence within the fallopian tube. CYTOKINES AND THE IMMUNE SYSTEM Immunological responses can be broadly categorized as nonspecific (i.e., directed against a broad range of microorganisms, e.g., neutrophils that phagocytose and kill bacteria) or specific (i.e., directed against a single organism, e.g., antibody response). The control, development, and interrelationship of these immunological responses, and the further differentiation of specific immune responses toward either a cell-mediated or humoral pathway is determined and maintained by a complex network of small intercellular regulatory proteins that mediate a range of immunologic as well as nonimmunologic biological functions. These are cytokines and chemokines, induced by specific stimuli such as particular bacterial products, and responsible for the stimulation and differentiation of various cell types as well as the further production of other cytokines that may enhance or inhibit the synthesis of protein products or biological effects of other cell types and proteins (19). This results in a complex, sensitive, and fine-tuned regulatory network of chemical messengers, enhancing and dampening various arms of the immune system through a series of positive and negative feedback loops. It is this ability or inability to generate certain cytokines or cytokine patterns in response to particular infections that will determine the clinical course of infection and whether this will result in successful clearance of the microorganism or pathology. The respective roles of humoral and cell-mediated immunity (CMI) has received significant attention in determining both the pattern of effective protective response to chlamydial infection and the precise mechanisms by which inflammatory damage, chronic infection, and pathogenesis are facilitated. Stimulated by infection, CD4 T helper cells will respond to follow one of two distinct patterns or subsets of T helper lymphocytes (Th) that can be defined on the basis of immune function, and the distinct patterns of cytokines each produces (20, 21). Th-1 cells produce interleukin (IL)-2, interferon (IFN)-, and tumor necrosis factor (TNF), whereas Th-2 cells secrete IL-4, IL-5, IL-6, IL-10, and IL-13. The Th-1 cells are involved in CMI responsible for the control of CD8 cytotoxic T cells. Their production is stimulated by macrophage IL-12 or IFN-, and is suppressed by IL-4. The Th-2 cells are integral for humoral immunity, and are stimulated by IL-4 and suppressed by IFN-. Humoral immunity is primarily focused on the neutralization and elimination of extracellular pathogens bacteria, fungi, protozoa, and multicellular parasites through the activation of naive B lymphocytes to secrete IgM and other antibody isotypes. Cellular immunity involves the destruction of infected cells by cytotoxic T lymphocytes (CTLs), or the destruction of intracellular pathogens by macrophages activated by Th-1 cells, and is directed at intracellular parasites viruses and intracellular bacteria such as mycobacteria and Chlamydia. However, this relatively simple division of roles across two arms of the immune system may be more complex than first assumed, with a greater overlap across the Th-1/Th-2 divide. The IL-10 is not strictly Th-2 specific, but rather IL-4 appears to be the significant cytokine involved in Th-2 cell development. The IL-4 promotes differentiation of Th-2 cells and inhibits the Th-1 cytokine profile. However, IL-10 does not promote the development of a Th-2 cytokine pat Debattista et al. Immunopathogenesis of Chlamydia trachomatis Vol. 79, No. 6, June 2003

3 tern, but rather inhibits development of Th-1 cells by blocking the synthesis of IFN- (22). Classification of Th-1 as cell-mediated immunity and Th-2 as humoral immunity may be an oversimplification (23). The Th-1 cells not only activate macrophages but IgG2a opsonizing and compliment-fixing antibodies. Therefore, it may be more appropriate to consider Th-1 as effectors of phagocyte-dependent responses. The Th-2 cells not only induce B-cell antibody but inhibit macrophage function, and therefore could be regarded as effectors of phagocyteindependent responses (23). The Th-1 and Th-2 are not the only cytokine patterns possible. The Th-0 cells secrete both patterns within a mixed population of partially differentiated effector cells (23). Cytokine responses can, therefore, remain mixed, or induced to differentiate into either Th-1 or Th-2. A heterogeneity in cytokine synthesis is possible even at the single cell level (23). Naive CD4 T cells are not irreversibly committed to either Th subset, and therefore, can generate either Th subset in response to a given antigen within a particular cytokine environment (24). The Th-2 may also serve a more significant role, not as effectors of humoral immunity, but rather as regulators of Th-1 responses. Although the Th-1 response is considered protective, an ongoing Th-1 response in the context of persistent infection can result in inflammatory tissue injury. A switch to Th-2 can occur if Th-1 does not eradicate organisms, thereby inhibiting activity of macrophages and avoiding or reducing inflammatory tissue damage (23). Furthermore, the observation that Th-1 protects against intracellular parasites and Th-2 eliminates extracellular pathogens may also be an oversimplification, based on limited animal models (25). Although the ideal of Th-1 and Th-2 responses developed out of Leishmania major infections in mice, contrary results have been obtained due to strain differences in virulence, subtleties of experimental protocol, other cell types present, and additional cytokines. Recent studies indicated a role for B cells against intracellular pathogens (26). Secretion of antibody appear to protect against intracellular cryptococcus and Mycobacterium tuberculosis by enhancing cell-mediated immunity. Other studies have not implicated a Th-2 role toward intracellular bacterial pathogens but have suggested an important role for priming protective T-cell responses. For Toxoplasma gondii, both IL-4 and IFN- appear necessary for protection; IL-4 is important in acute infection by limiting inflammation, whereas IFN- is important in later infection and parasite removal (25). The IL4 has been shown to both exacerbate disease and protect the host depending on the levels of infection and the time of infection. Individual cytokines, therefore, can produce opposing effects depending on dose and timing of their participation in the immune response. Ultimately, the Th-1 and Th-2 patterns of immune response appear to act in a balanced, coordinated fashion with no single T-cell subset solely responsible for disease clearance. Both are required for resolution. The generation of key cytokines to induce Th-1 or Th-2 pathways may not be solely dependent on CD4 or CD8 T cells, but rather innate immune cells (NK cells and mast cells) are also capable of initiating such responses (25). IMMUNOLOGY OF C. TRACHOMATIS Given the intracellular developmental cycle of Chlamydia, there is strong evidence for the involvement of cellmediated immunity (Th-1 pathway) and its associated cytokines, IFN-, IL-2, and IL-12, in resolving a chlamydial infection (27 37). Interferon- can also have a direct physiological effect on Chlamydia, inhibiting its growth through the induction of host cell indoleamine 2,3-dioxygenase, a tryptophan-degrading enzyme. This results in the depletion of the host cells tryptophan pool and consequent nutrient deprivation of the chlamydiae (38). Tumor necrosis factor may similarly enhance the activity of indoleamine 2,3-dioxygenase (39). In addition to this cytotoxic activity it also has a proinflammatory activity that promotes the production of other cytokines capable of blocking intracellular replication and may provoke cell death by apoptosis (40). However, response to chlamydial infection can result either in a protective or pathological immune response. Being an intracellular pathogen, chlamydiae possess little intrinsic toxicity. Disease appears to result more from immune recognition to specific antigens expressed directly by the organism or indirectly on the surface of infected host cells (41). Given the primary protective role effected by the cellmediated Th-1 response in resolving infection, inflammatory damage within the upper reproductive tract may be the result of a failed or weak Th-1 action resulting in chronic infection (27, 42 46) or as a result of an exaggerated or overstimulated Th-1 response (47 52). A balance, therefore, exists between the protective and deleterious effects of cell-mediated immunity, represented in Figure 1 as a U-shaped curve. An effective immune response that successfully clears infection occurs in the center, lower region of the curve. Each end of the curve indicates either a too weak or too strong cellular immune response, responsible for pathology consequent to chronic chlamydial infection or hyperinflammation, respectively. A more accurate description of the balance may be one of proinflammatory and anti-inflammatory cytokine production with increasing polarization toward either Th-1 or Th-2 responses responsible for the severity and chronicity of immune responses (53). Not only are naive T cells malleable in their development along Th-1 and Th-2 lineages, but antigen-presenting cells are responsible for the production of both proinflammatory cytokines and anti-inflammatory cytokines (54). FERTILITY & STERILITY 1275

4 FIGURE 1 Relationship between level of cellular immunity (CMI) and inflammatory damage caused by C. trachomatis. Debattista. Immunopathogenesis of C. trachomatis. Fertil Steril Carefully regulated cytokine production is central to successful immune responses to intracellular pathogens. Deficiencies in the production or activity of these proinflammatory cytokines can be associated with failure of protective immunity or harmful inflammation. As an example, the suppression of IFN- in association with the ongoing production of IL-4 has been shown to be responsible for the persistence and progression of tuberculosis in susceptible individuals (53). The model of pathogenesis by which C. trachomatis progresses from acute to chronic infection, and finally serious disease (salpingitis, tubal occlusion) is yet to be fully determined. The pathways by which acute chlamydial infection achieves resolution or alternatively progresses to chronic infection with severe pathology, appears to be varied and dependent on a closely entwined interplay between host and pathogen. A number of interconnected factors appear to play a significant role human genetics (42, 55) and endocrinology (56 58), cytokine profile (21, 34, 45, 59, 60), previous infections (46, 47, 61), pathogen load (62), chlamydial strain (63 66), the presence of other genital infections (67), and route of infection (68 70). The means by which a particular immune pathway humoral or cell-mediated immunity is adopted, strengthened, or shifted away from balance in favor of one, as demonstrated in other infections (71), all under the influence of changing cytokine patterns, and the degree to which each of these effect the outcome of infection is yet to be resolved. A formidable obstacle to understanding the significance of cytokines on infection is the difficulty of extrapolating from animal to human models (25). Cytokines, shown to be significant in animal models, can remain undetected in humans (25, 72). The contribution of IFN- to chlamydial resistance has become less clear due to variations in cytokine susceptibility across different C. trachomatis strains (73) and genetic differences between mouse strains (74). The role of HLA class 1 restricted CD8 CTL responses in human chlamydial infections is poorly understood and although a protective function has been associated with those CTLs producing IFN-, they have in the main been implicated in pathology such as scarring trachoma (72). In contrast, a CD8 CTL response in mice has been demonstrated by the adoptive transfer of CTLs to be protective (75), although other studies appear to suggest the CD8 cells are not important for resolution of infection (26, 74). In another animal model, the rhesus macaque, CD8 T cells responding to C Debattista et al. Immunopathogenesis of Chlamydia trachomatis Vol. 79, No. 6, June 2003

5 FIGURE 2 Flowchart describing the course of C. trachomatis infection as determined by immune pathways. Debattista. Immunopathogenesis of C. trachomatis. Fertil Steril trachomatis were shown to be associated with salpingitis (75). There has been an assumption of the protective role in CD8 CTL acting against C. trachomatis, given its intracellular developmental cycle, and therefore like other intracellular pathogens, antigens are presented through a cytoplasmic processing and presentation pathway to MHC class I molecules (72, 76). However, Chlamydia are confined to endosomes, and consequently, it cannot be assumed that chlamydial antigens are processed into class I MHC (72, 76). Chlamydial antigens might consequently be presented through MHC class II to CD4 cells. Once inflammation is induced, expression of both class I and II antigens would be up-regulated by proinflammatory cytokines, particularly IFN- (72). The flow chart in Figure 2 attempts to provide a simplified model of chlamydial infection and the interplay between host immunity, pathology, and infection. Progression along a particular pathway is determined by particular immune responses, which in turn are determined by a range of host and pathogen factors. At each stage there exists an opportunity for resolution and clearance of organisms, or progression to further serious pathology. Although this diagram attempts to show the interplay between cellular and humoral immunity, and the effects of Th polarization in eliciting pathology either through hyperinflammation or chronicity, an alternate view has been presented by Yang (77). Using a murine lung infection model, they demonstrated that there might be two different functional types of delayed type hypersensitivity (DTH) induced by Chlamydia and based on cytokine patterns a Th-1-type DTH associated with protection and a Th-2-type DTH associated with immunopathology. The DTH responses have long been considered as both potentially protective and pathological and it is the functional differences in Th-1- and Th-2-type DTH responses that account for this dual role in chlamydial immunity and pathology. Chlamydia is able to induce both Th-1 and Th-2 DTH responses and those DTH responses associated with Th-2 cytokines differ from that mediated by Th-1 cytokines with respect to clearance of chlamydial infection and immunopathology. In Th-2 DTH, inflammation does not specifically focus on sites of inclusion but diffuses throughout the lung and other organs. This observation, set within the context of the previously mentioned limitations of animal studies, does present a new model of chlamydial immunopathology one not based on balance between Th-1 and Th-2 responses, but rather where Th-1 cytokines serve a dominant role in protective immunity and Th-2 cytokines are associated with immunopathology. FERTILITY & STERILITY 1277

6 IMMUNE PATHWAYS An acute infection of chlamydial organisms requires colonization of the intended site of infection. At this initial stage, humoral immunity in the form of secretary IgA antibodies targeting the major outer membrane protein (MOMP) of the pathogen, appears to be the significant protective response and the determinant of whether infection develops or is cleared (29, 78 81). This protective immune response to infection is associated with serovar-specific determinants on the MOMP (82). The MOMP is the principle surface protein component and is abundant in both infectious elementary bodies and intracellular reticular bodies. It is antigenically complex of molecular weight 40,000 44,000 Da and constitutes approximately 60% of total surface protein of elementary bodies (EBS). Functional MOMP probably serves as a porin, regulating the chlamydial developmental cycle by the passage of small molecules through the outer membrane (83). Antibodies to surface accessible epitopes on MOMP may neutralize infection and animal studies suggest a role for MOMP in protective immunity (82). Short-lived immunity to C. trachomatis does develop after infection and is serovar specific asdefined by serological variation in MOMP. The MOMP elicits C. trachomatis antibodies that recognize its surface-exposed variable segments, therefore B-cell responses directed at serovar-specific epitopes located in variable segments of MOMP may be critical to immunity. Yet despite this, T helper cells, important for priming and maintaining B-cell responses were primarily stimulated by constant segment peptides and only one variable segments peptide (84). The MOMP contains serovar-specific B-cell epitopes in four variable regions of the molecule whose amino acid sequence varies among serovars (85). Several T-cell epitopes were identified in MOMP and one that stimulated T cells from 80% donors was resolved as 12 amino acids adjacent to variable segment-3. Both CD4 and CD8 T cells were stimulated. Proliferative responses and IFN- production was stimulated (83). There is uncertainty as to whether immunity to constant region epitopes might protect against infection with multiple chlamydial serovars, but reinfection is nonetheless common, and recall responses to MOMP constant segment epitopes might contribute actually to inflammatory damage. However, other C. trachomatis antigens with epitopes shared among different serovars could elicit damage, and MOMP constant segments could indeed provide protective immunity (86). There is some suggestion of partial protection against C. trachomatis reinfection, although such immunity is relatively short lived. Less severe infections of a shorter duration do develop after reinoculation with either the same or different C. trachomatis serotype, indicating cross-protective immunity. In humans, serovars isolated from secondary infections are often different from primary infections, suggesting that neutralizing antibodies responses to variable segment epitopes in the MOMP might be important for protective immunity (75). If organisms are not cleared by the initial humoral response, either through inadequate or deficient immune mechanisms or by too great an inoculum of pathogen, Chlamydia will become established as intracellular parasites within the host s tissue. At this point, antibody responses alone are inadequate to resolve what has now become an acute infection. The host response to this acute infection occurs within 1 2 days and is characterized by inflammation and mucosal infiltration with neutrophils, T-cell lymphocytes, and monocytes. Neutrophils migrate in large numbers to the site of infection and destroy accessible chlamydial EBs, whereas T cells accumulate and control the infection through cell-mediated destruction of infected cells (87). In the first 96 hours of infection, the nature of the immune response and the eventual outcome of infection are determined to a great extent by the initial cytokine profile elicited with the infection (59, 88). It is at this critical juncture of acute infection that a number of pathways are possible. Cell-mediated immunity (Th-1 pathway) mediated through cytokines such as IL-2, IL-12, and IFN- can become the predominant and effective protective response leading to resolution of the infection through either destruction of infected cells by CTLs, phagocytosis of released EBs by activated macrophages (89, 90), or the direct anti-chlamydial action of IFN- and TNF on bacterial metabolism (33). However, it may in fact be a balance between Th-1 and Th-2 responses that provide the optimal protective response. The B and T cells interacting synergestically recognize chlamydial antigens on infected cell surfaces by specific antibody and subsequently lyse the infected cell by antibodydependent cellular cytotoxicity. Alternatively, the production of specific antibody may inactivate free EBs, preventing further infection (26). The CD8 T cells have been shown to be cytolytic for Chlamydia infected cells in vitro. Cytolysis appears to occur late in the developmental cycle. Early in intracellular development ( 24 hours) Chlamydia exist as metabolically active, noninfectious RB. From 24 to 72 hours, RB differentiate into metabolically inert, infectious EBs. Thus, for CTLs to be protective, an early lytic event rather than a late one would be preferable, lest infectious EBs are spread. However, Chlamydia have been shown to suppress IFN inducible MHC class I and II expression. Suppression of MHC class I synthesis circumvents the processing and presenting of chlamydial epitopes on the cell surface and subsequent lysis by CTLs. Chlamydia, therefore, evade host immune recognition and establish persistent infections (26). However, other studies have shown MOMP-specific CTLs to lyse infected ME180 cervical cells relatively early 1278 Debattista et al. Immunopathogenesis of Chlamydia trachomatis Vol. 79, No. 6, June 2003

7 in the chlamydial growth cycle (24 and 48 hours after infection) (75). This suggests that CTLs might be capable of killing infected cells in vivo, while most organisms are replicating as reticular bodies, unable to infect neighboring cells. Lysis of infected epithelial cells by CTLs was shown to be selective and specific. Therefore, direct lysis of infected cells early in the chlamydial growth cycle could provide protection against Chlamydia by clearing infected cells from the genital tract and limiting the replication. Alternately, a hyperinflammatory Th-1 response, resulting from a failure in immunoregulation or amplification of inflammatory cytokines, may lead to significant tissue destruction through the action of CTLs (51, 58, 91, 92). A third pathway is the opposite of this overwhelming cellular immune response (i.e., a weak Th-1 response), leading to inadequate clearance of the infection. In this pathway, the Th-2 response is dominant. A failure to mount a sufficient Th-1 CTL response (28, 35, 43), a premature shift from Th-1 to Th-2 pathways (42), production of higher levels of Th-2-inducing cytokines (55, 93), a defect in IL-2 (45), IL-12 (34), or IFN- secretion (37, 46), or inadequate lymphoproliferation by T cells (27, 43) lead to defective activation of macrophages for bacterial killing and provide conditions for Chlamydia to enter a chronic state. Very early production by chlamydial-infected epithelial cells of large amounts of anti-inflammatory IL-11, combined with weak potency of Chlamydia lipopoly saccharide to induce inflammatory response and a minimal and delayed induction of proinflammatory cytokines may sustain chlamydial infections and delay immune responses (94). Alternately, NK cell-derived IFN may provide a rapid defense against primary infection that dramatically decreases the infectious burden of Chlamydia, but the associated antigen load could also terminate prematurely the induction of an acquired memory T-cell response capable of stopping reinfection. Therefore, repeated or persistent infections may lead to pathology (73). Apoptosis induction occurs in human cells after infection with C. trachomatis. Intracellular chlamydial infection of macrophages can induce T-cell apoptosis through a secretory mechanism. Apoptosis induction by Chlamydia may explain how persistently infected macrophages escape T-cell surveillance and why T-cell responses are diminished during persistent chlamydial infections. The ability of pathogens to induce apoptosis may play a role in the initiation of the infection, the survival of the pathogens, and their escape from the host immune response. Infection of cells not only changes the susceptibility of host cells to apoptosis but also can mediate apoptosis of noninfected cells (95). Infected nonactivated macrophages could consequently provide a secondary site for replication and dissemination of Chlamydia (44, 70). As well, low levels of IFN- secretion may stimulate the presence of persistently infected cells in the genital mucosa with atypical, noninfectious, nonreplicative chlamydial bodies (46). In vitro, these persistent infections correlated with an increased level of expressed 60-kD chlamydial heat shock protein (chsp60) and with a significant reduction in the levels of MOMP (46, 96). Chlamydia HSP60 appears to be expressed early 2 26 hours after infection. Moreover, in cells treated with IFN- and thereby induced into a persistent or chronic state, MOMP expression is decreased and HSP60 increased, perhaps contributing to the delayed hypersensitivity type responses associated with the chronic inflammation and tubal scarring (97). The presence of persistently infected cells in vivo brought on by the localized production of low levels of IFN- could stimulate such chronic inflammation by serving as reservoirs of the immunopathogenic and potentially cross-reactive chsp60 capable of sustaining such an inflammatory response (49, 98). Interferon- (74) and IL-1 (97) produced in humans and mice induces the synthesis of indoleamine deoxygenase, which depletes tryptophan essential for chlamydial replication. Tryptophan depletion is chlamydiastatic, resulting in an incomplete growth cycle, persistent infection, and upregulated heat shock proteins. The result is periodic fluctuations in the shedding of infectious Chlamydia as initial shedding of chlamydial antigen leads to production of IFN that restrains further shedding until cellular immunity wanes. Therefore, chronicity of chlamydial infections, fluctuating between periods of acute replication may be the norm, rather than the exception (74), given that unlike ideal in vitro conditions, significant metabolic stresses exist in vivo causing at least a portion of Chlamydia at any one time to switch to a persistent state (38). Again, as with other immunological studies of C. trachomatis, the contribution of IFN- to chlamydial resistance is based on a variety of animal and in vitro models that show differences due to C. trachomatis strain variation. All human serovars are much more sensitive to the inhibitory actions of IFN than the mouse pneumonitis (MoPn) strain (73). THE ROLE OF chsp60 Evidence implicating chsp60 as the critical antigen responsible for stimulating immune-mediated inflammation and disease is extensive through studies reporting the increasing prevalence of chsp60 antibodies among women with increasing severity of chlamydial disease. Women with visually observed chlamydial PID showed a direct correlation between high chsp60 antibody titers and severe inflammatory manifestations (99). A number of studies investigating serological responses of infertile women to chsp60 have demonstrated a correlation between antibodies and tubal infertility ( ). Fur- FERTILITY & STERILITY 1279

8 ther studies in animals have demonstrated that chsp60 induced delayed hypersensitivity responses in C. trachomatis salpingitis in monkeys (104) and that inoculation of the chsp60 antigen into experimental animals, previously sensitized by ocular chlamydial infection, induces a marked clinical response (105). Antibody to chsp60 significantly correlated with risk factors for PID and occluded fallopian tubes. In a monkey model of chlamydial infection, three different chsp60 antibody responses could be observed after infection: no chsp60 antibody; an initial strong chsp60 antibody response that returns to normal; and a strong chsp60 response that persists (106). The 60-kD heat shock protein belongs to a group of chaperone proteins ranging in molecular weights from 15 to 110 kd, which are produced in response to heat stress, infection, inflammation, toxic chemicals, low oxygen pressure, and other cellular insults (107). Originally called heat shock proteins because they were thought to be effective in preventing the unfolding of proteins only at high temperature, they were later found to be synthesized in response to many other stresses, and therefore were labeled as stress proteins. Their function appears to be essential for cellular survival, preventing protein denaturation or abnormal aggregation during periods of stress by helping to preserve folding of cellular proteins. It is now clear that they are also responsible for the synthesis, aggregation, and packaging of newly synthesized proteins and general housekeeping functions within normal cells. Therefore, they have become known as molecular chaperones. The HSP60 is a phylogenetically conserved protein in both structure and function and found across the entire spectrum of life from bacteria to human. Given that the heat shock response is one of the most universal reactions known, it is reasonable to consider heat shock proteins among the most conserved proteins (108). Chlamydial HSP60 exhibits greater than 80% homology between Chlamydia species, 60% identity with other bacteria, and 50% homology with eukaryotic HSP60 from plants and animals. In Chlamydia it is involved in the assembly of the outer membrane of the elementary body (from where it can be isolated) during the developmental change from the large intracellular reticulating body to the compact infective elementary body. It may also serve a protective role by assisting the pathogen to evade host defenses (109). Necrotic, but not apoptotic, cell death leads to the release of biologically potent HSPs from cells. The released HSPs stimulate macrophages and dendritic cells to secrete cytokines and induce expression of antigen-presenting and costimulatory molecules on the dendritic cells (110). The LPS of gram-negative bacteria have similar properties to HSPs, but HSPs are more ancient and ubiquitous than LPS. Therefore, HSPs, normally intracellular, are released in large quantities into the extracellular milieu and serve as immune signals signifying cellular disintegration consequent to bacterial and viral infections (111). The mechanism by which exposure to chsp60 elicits pathology has been explained in a number of ways: 1. C. trachomatis enters a persistent state within human host cells during which expression of chsp60 is enhanced whereas levels of MOMP are minimized. This provides a continued source of antigenic stimulation (112). 2. Repeated exposure to other bacteria carrying homologous HSPs or repeated chlamydial infections may result in a prolonged antibody response and also prime the cellular immune system such that a subsequent infection may induce a pathogenic cell-mediated immune response (112). This continuous exposure to chlamydial heat shock would promote chronic disease either through direct antigenic stimulation or by signal transducers that result in macrophage activation (113). A third pathological mechanism has been postulated. Human HSP60 epitopes are expressed at the surface of stressed macrophages or tissue cells. Their presence, stimulated by the production of inflammatory cytokines, serves as an indicator of immune activation. The production of reactive oxygen metabolites during the course of inflammation by macrophages may require HSP production as a form of self protection (108). In response to the stress imposed by activated macrophages, pathogens may also induce their own HSP as a means of counterdefense. When entering the host from the environment, pathogens are confronted by several stressful changes such as changes in temperature, ph, and partial pressure of oxygen, as well as the host s natural resistance mechanisms such as phagocytosis. Therefore, in addition to the ongoing inflammatory damage as a consequence of chronic C. trachomatis infection, there exists the possibility of autoimmune-mediated damage to tissue as a result of antigenic cross-reactivity between hyperexpressed chsp60 on the surface of persistently infected cells, and overexpression of human HSP produced secondary to the inflammatory damage induced by infiltrating CTLs targeting chsp60 of the chronic infection. The observation of antibodies to human mitochondrial HSP60 observed in eight women with C. trachomatis-related ectopic pregnancies (114) has been explained as immunopathology due to cross-reactivity between chlamydial and human HSP60. Studies of HSP60 proteins expressed by other significant pathogenic bacteria (e.g., mycobacteria) have provided similar models of T-cell reactivity implicated in the onset of autoimmune conditions (115). Furthermore, human HSP60 is physiologically expressed during the pre- and peri-implantation stages of pregnancy by the embryo and maternal decidua (116). These host HSP60s expressed in early pregnancy may, in the presence of a chronic chlamydial infection, cross-react with lymphocytes previously sensitized to chlamydial HSP60. The activated 1280 Debattista et al. Immunopathogenesis of Chlamydia trachomatis Vol. 79, No. 6, June 2003

9 lymphocytes release pro-inflammatory cytokines that can disturb the immune regulatory mechanisms necessary to implantation and maintenance of the embryo. Circulating antibodies to HSPs may impair embryo development. Consequently, embryos are less protected from adverse environmental conditions and are more likely to degenerate or undergo apoptosis (116). Discussion of HSPs and the cross-reactivity between human and microbial HSP has so far focused on immunopathology. However, cross-reaction can also provide a mechanism for protective immunoregulation, i.e., the recognition of self-hsp could be an important immunological strategy that contributes to the establishment or maintenance of selftolerance and the control and down-modulation of inflammatory responses, irrespective of their origin (117). Thus, the presence of autoantibodies to human HSP cross-reactive with their chlamydial homologs may suggest an alternative to autoimmune disease. The expression of stress up-regulated human HSP on the surface of damaged and inflamed tissue serves a protective modulating function by providing the immune system with a target self-antigen through which potentially deleterious inflammatory reactions can be monitored and controlled (117). The overexpression of human HSP, normally expressed in low numbers, may therefore play an important immunoregulatory role, down-regulating the cell-mediated Th-1 inflammatory response through the action of IL-10 and the promotion of a Th-2 pathway (23, ). In this manner a damaging inflammatory response would eventually be brought under control and resolved. However, extensive irreversible damage may have already occurred by this stage. Thus, evidence for autoantibodies and anti-chsp60 antibodies in women with chlamydial-related salpingitis and infertility may indicate this damping down of hyperinflammation rather than the etiology of tissue destruction. The significance of IL-10 in damping immune responses has been supported by its apparent late production in the inflammatory process, its inhibitory effects on proinflammatory cytokine (IFN-, IL-2) production, and its down-regulation of class II MHC molecules ( ). Repeated infections of women with C. trachomatis may lead to repeated exposure to chsp60, cross-reactive with human HSP, and could induce earlier production of IL-10, thus switching off the Th-1 response before it has effectively cleared the infection. This may suggest that those experiencing multiple chlamydial infections over time may be more vulnerable to pathological sequelae. Kinnunen et al. (125, 126) recently showed that most chsp60-specific T-cell clones in mice and humans produced predominantly IL-10. They postulated that the Th-2 dominant response may replace an initial Th-1 dominant response in chronic C. trachomatis infections. Therefore, salpingeal lymphocytes responding to chronically expressed chsp60 may produce IL-10 and are responsible for reducing the Th-1 response needed to clear infection. This conflicts with the previous antigenic mimicry model of chsp60 triggering an autoimmune response against self-hsp60. The alternate, immunoregulative model suggests that self-hsp60 reactivity observed after C. trachomatis infection may induce Th-2 responses that delay resolution of the infection. Although this maintains the persistence of infection, it also protects the host from overreacting to inflammatory stimuli. Both models could be right and that the stage of equilibrium is determined by as yet undefined factors such as the type of infected cells or the HLA genotype of the host (126). However, as with all immune models, they are often limited by the host and strain of infecting organism. Using other models of chronic disease, such as bacterial reactive arthritis where Th-2 cytokine pattern predominate in the joints of patients, Th-1 cytokines are necessary for elimination of reactive arthritis-associated bacteria. Therefore, Th-2 may contribute to bacterial persistence in the joint resulting in reactive arthritis (127). In contrast to this, a Th-2 response seems necessary for the elimination of Borrelia burgdorferi and the presence of a wrong Th-1 response in the joint seems to contribute to persistence (127). There is some contention as to whether chsp60 is in itself pathogenic, either through molecular mimicry, inducing antibody immune complexes capable of tissue inflammation, or inducing a dominant Th-2 response and lowering cellular immunity. Antibodies to chsp60 may not of themselves be disease specific, but rather are indicators of chronicity (97). Studies on a population of 280 female sex workers prospectively studied over a 33-month period demonstrated that preexisting antibody to chsp60 predicts a two- to threefold increased risk for C. trachomatis PID. Although antibodies to chsp60 are commonly found in women with tubal infertility and ectopic pregnancy, it is unclear whether this association is the result of a persistent upper genital tract infection with C. trachomatis, or whether immune responses to chsp60 antedate infection and possibly mediate tissue inflammation and damage (128). Earlier research in trachoma, which demonstrated that reinfecting previous infected guinea pigs with extracts of chsp60 in the detergent Triton X-100 stimulated local infiltration of monocytes, may be questioned due to the crudeness of the extracts, and the fact that guinea pigs are unusually prone to developing DTH reactions. Well-authenticated clinical examples of pathological autoimmunity due to self- HSP responses are rare (72). Further than just implicating chsp60 as a significant antigen, 163 patients with tubal factor infertility were shown to have significantly higher frequencies and titers for all antibodies (C. trachomatis HSP60, HSP10, common chlamydial lipopolysaccharide) except to Chlamydia pneumoniae. No interaction between C. trachomatis and C. pneumoniae suggesting a synergistic effect was found, although FERTILITY & STERILITY 1281

10 the HSP from these two organisms are immunologically similar (129). A purified native and recombinant 10-kD heat shock protein (HSP) has been found to be a strong stimulator of peripheral blood T-cell proliferation in Mycobacterium leprae (130). Similar strong proliferative responses of PBMC to the 10-kD HSP has been shown in Mycobacterium tuberculosis (131). The detection of anti-chsp10 antibodies has been associated with chronicity of C. trachomatis genital tract infection and does not parallel that of anti-chsp60 IgG antibodies. The chsp10 residues have been found to contain an immunodominant, although not a universal, B epitope. Cross-reactions with C. pneumoniae or Escherichia coli GroES protein are limited but may occur (132). chsp10 has been identified as a potential correlate to the immunopathogenic process in women with tubal infertility, with serological responses to HSP10 more strongly correlated with tubal infertility than were HSP60 and MOMP (133). Women with tubal infertility have recognized HSP10 more frequently than actively infected women, perhaps as a consequence to repeated or persistent chlamdial exposure (113). GENETIC DETERMINANTS Alternatively, women susceptible to serious pathology may represent a subset of the population who are either more vulnerable to reinfection or to developing chronic infection through a predisposing lower capacity to mount a strong Th-1 response. Although a majority of women may be capable of resolving infection as a result of strong cellular immune responses, a minority of women may be incapable of adequate cellular responses, therefore these women could be selected out for chronic infection. Consequently, they present a higher risk for pathological sequelae. Peeling et al. (128) noted that preexisting antibody to chsp60 in 280 female sex workers predicted a two- to threefold increased risk for C. trachomatis PID and that immune responses to chsp60 antedate the chlamydial infection. There was also a link between chsp60 antibodies and weaker antibody responses to MOMP, a target for initial protective immunity. This may suggest that some women are genetically and immunologically selected out toward pathology. Research with other types of C. trachomatis infection support an underlying genetic susceptibility to pathology. In areas of endemic trachoma, those with moderate to severe conjunctival scarring due to trachoma showed a significant association with the class 1 allele HLS-A*6802 (134). Turner et al. (135) demonstrated that familial clustering and family histories of trachoma indicate a possible inheritable susceptibility. Genetic controls may determine the adequacy of Th-1 and Th-2 immune mechanisms at each stage of infection and pathogenesis by controlling for the: level of cytokine secretion by infected mucosal epithelial cells; level of cytokine secretion by various lymphocyte and leukocyte cell subsets (i.e., cytokine receptivity); overstimulation or understimulation of various negative and positive feedback mechanisms resulting in suppression or exaggerated enhancement. Genetic control would similarly determine the sensitivity of recognition and cross-reactivity of CD4 and CD8 T cells for a range chsp60 epitopes as well as the strength of hypersensitivity response to these epitopes. Significant genetic variation across individuals would determine the relative immunogenicity and cross-reactivity of significant epitopes responsible for lymphoproliferation. Antibody responses to chsp60 appear to be genetically determined in mice (136). Studies by Khamesipour et al. (42) showed that induction of chlamydial tubal infertility in two strains of mice and differing outcomes in morbidity was determined by differing immunogenetic responses to HSP60. Cell-mediated responses specific for chlamydial antigens have been noted to be MHC linked (86, 137). Variations in the MHC class 1 epitopes found on leukocytes in Macaque monkeys infected with C. trachomatis indicated an association between susceptibility to PID through the formation of pelvic adhesions and MHC haplotypes (138). In their study of genetically determined differences in IL-10 and IFN- responses to chlamydial infection and consequent clearance of the organisms in the mouse model, Yang et al. (139) demonstrated that genetic susceptibility to infection is expressed through variations in cytokine profile at the time of T-cell stimulation by antigen, and whether a Th-1 or Th-2 response is elicited. OTHER IMMUNE DETERMINANTS Route of Infection Cytokine profiles have been shown to vary according to the route of chlamydial antigen administration (68). A dominant Th-1 response developed in mice given live mouse pneumonitis biovar of C. trachomatis by the intranasal, oral, and vaginal routes; this response was indicated by a predominance of IFN- -secreting cells. In contrast, mice injected subcutaneously produced a Th-2 response with higher levels of IL-4 and IgG serum titers. When challenged with a second infection vaginally, those mice immunized s.c. and producing a Th-2 response cleared the infection more slowly. Rank et al. (69) also indicated that both the route of administration and form of antigen may impact on the disease process. Subcutaneous immunization with HSP60 in guinea pigs appeared to decrease ocular pathogenesis after a challenge with C. trachomatis and resulted in high titers of IgA and IgG antibodies in tears. However, s.c. immunization combined with ocular inoculation by attenuated Salmonella typhimurium-expressed HSP60 antigen had no such protective effect against infection with C. trachomatis Debattista et al. Immunopathogenesis of Chlamydia trachomatis Vol. 79, No. 6, June 2003