Humoral immune response to membrane components of Chlamydia ...

2 downloads 0 Views 64KB Size Report
that their pelvic infection occurred in an asymptomatic or silent form. The prominent role of Chlamydia trachomatis in. Recent evidence suggests that Chlamydia ...
Human Reproduction vol.12 no.5 pp.925–929, 1997

Humoral immune response to membrane components of Chlamydia trachomatis and expression of human 60 kDa heat shock protein in follicular fluid of in-vitro fertilization patients A.Neuer1,2,4, K.-N.Lam2, F.-W.Tiller3, L.Kiesel2 and S.S.Witkin1 1Division

of Immunology and Infectious Diseases, Department of Obstetrics and Gynecology, Cornell University Medical College, 515 East 71st Street, New York, NY 10021, USA, 2University of Tu¨bingen Women’s Hospital, Tu¨bingen, Germany, and 3Medical Immunology Laboratories, Mu¨nchen, Germany 4To

whom correspondence should be addressed at: Division of Immunology and Infectious Diseases, Department of Obstetrics and Gynecology, Cornell University Medical College, 515 East 71st St., New York, NY 10021, USA

Recent evidence suggests that Chlamydia trachomatis can persist in the female upper genital tract in an unculturable state. Since unsuspected C.trachomatis infection has been associated with adverse in-vitro fertilization (IVF) outcome we sought to detect further evidence of C.trachomatis in the genital tracts of women undergoing IVF. The prevalence and distribution of antibodies to the major structural proteins of C.trachomatis in paired follicular fluid and sera of women undergoing IVF were examined. Sera and follicular fluid samples from 149 women were assayed for immunoglobulin (Ig)G and IgA antibodies to two C.trachomatis antigens, the major outer membrane protein (MOMP) and a recombinant lipopolysaccharide (rLPS) fragment. Additionally, the expression of human 60 kDa heat shock protein (hsp 60) in follicular fluid was determined. All cervical and follicular fluid samples were negative for C.trachomatis by polymerase chain reaction, ligase chain reaction and DNA probe. Sera from 60% of the subjects were positive for antichlamydial rLPS IgG; 36% were positive for anti-MOMP IgG. Similarly, rLPS-directed and MOMP-directed IgA were detected in sera of 34 and 14% of the subjects respectively. IgG antibodies to MOMP and rLPS were detected in 42 and 41% of the follicular fluid examined respectively. Anti-MOMP IgA was identified in 8.7% of the follicular fluid while 27.5% were positive for anti-rLPS IgA. Human hsp 60 expression was documented in 11.6% of the follicular fluid tested. IgA antibodies to both MOMP (P J 0.03) and rLPS (P J 0.02) in follicular fluid were associated with a failure to become pregnant after embryo transfer. IgG antibodies in sera and follicular fluid and IgA antibodies in sera were unrelated to IVF outcome. Similarly only anti-MOMP IgA (P 5 0.02) and anti-rLPS IgA (P J 0.04) in follicular fluid were correlated with human hsp 60 expression in follicular fluid. The unique association between IgA antibodies to two chlamydial antigens in follicular fluid and both hsp 60 expression and IVF failure provides further support for the © European Society for Human Reproduction and Embryology

possibility that a persistent upper genital tract chlamydial infection contributes to IVF failure in some women. Key words: Chlamydia trachomatis/follicular fluid/human heat shock protein/infertility/IVF

Introduction Evaluating infertility retrospectively, most women with tubal occlusion do not report a history of acute salpingitis, suggesting that their pelvic infection occurred in an asymptomatic or silent form. The prominent role of Chlamydia trachomatis in these silent infections is supported by the known association between tubal infertility and circulating chlamydial antibodies (Moore et al., 1982; Meikle et al., 1994). C.trachomatis constitute a unique group of inclusion-forming bacteria. Immune responses are most prominent among infected persons, but as obligate intracellular parasites members of the Chlamydiae can evade immune defences and persist within their host for prolonged periods. In fimbrial and peritubal adhesions of patients with tubal infertility, for example, chlamydial DNA can be detected even after antibiotic treatment (Patton et al., 1994). C.trachomatis can ascend to the Fallopian tubes and induce an inflammatory reaction without clinical symptoms. The pathogenic events of chronic chlamydial inflammation are not well understood, but the persistence of this organism in tissue has been identified as a key element in the initiation and maintenance of chronic disease. Viable C.trachomatis can also persist in macrophages and this may provide an alternative target cell for chlamydial replication in a number of infections (La Verda and Byrne, 1994). Since follicular fluid contains tissue macrophages in large amounts (Loukides et al., 1990; Lachapelle et al., 1996) as well as T and B lymphocytes (Castillo et al., 1990), it was of interest to evaluate follicular fluid for evidence of infection and its relation to in-vitro fertilization (IVF) outcome. The human 60 kDa heat shock protein (hsp 60) is a highly conserved protein that shares a 48% amino acid sequence identity with the chlamydial hsp 60 (Cerrone et al., 1991). Under conditions of rapid cell growth or differentiation or following environmental stress such as inflammation, host hsp 60 synthesis is induced. Initiation of heat shock gene transcription results in inhibition of transcription of genes coding for proinflammatory cytokines (Schmidt and Abdulla, 1988). Thus, hsp 60 expression may indicate a down-regulatory response to inflammation. It has also been suggested that, in women previously sensitized to a conserved epitope of a microbial hsp 60, the reactivation of hsp 60-reactive lymphocytes by expression of human hsp 60 may interfere with 925

A.Neuer et al.

implantation or early embryo development (Witkin et al., 1996). To elucidate further the possible cryptic presence of C.trachomatis in IVF patients and its possible contribution to IVF failure we examined the distribution of immunoglobulin (Ig)A and IgG antichlamydial antibodies in follicular fluid and sera of asymptomatic women undergoing IVF. Additionally the expression of human hsp 60 in follicular fluid was determined. Results were compared with the cause of infertility and subsequent pregnancy outcome.

Materials and methods Patients and sample collection A total of 149 women undergoing a cycle of IVF at the Women’s Hospital of the University of Tu¨bingen, Germany between January 1994 and December 1995 participated in this study. Informed written consent was obtained from each individual patient. All women had negative results for cervical C.trachomatis by DNA probe (Genprobe®, San Diego, CA, USA) 2 days prior to oocyte retrieval. Vaginal cultures for microbial pathogens, mycoplasmas and bacterial vaginosis were routinely performed 7 days after onset of stimulation for IVF and were uniformly negative. Follicular fluid and sera were obtained at the time of oocyte retrieval by vaginal aspiration and venipuncture respectively. None of the samples was visibly contaminated with blood. The samples were stored at 4°C and transferred to the laboratory within 6 h. Follicular fluid was centrifuged and the resulting pellet was used for polymerase chain reaction (PCR) and ligase chain reaction (LCR). Tubal occlusion was diagnosed by laparoscopy and hysterosalpingosonography.

water bath for 60 min. The wells were then washed and a 1:500 dilution of rabbit anti-hsp60 polyclonal antibody (SPA 804, StressGen) was added. After 1 h incubation and another 4-fold washing, an alkaline phosphatase (AP)-conjugated goat antibody to rabbit IgG (Kierkegaard and Perry, Gaithersburg, MD, USA; dilution 1:400) in PBS–Tween was added. Following an additional 60 min incubation at 37°C, the wells were washed as above and the colourless AP substrate, p-nitrophenylphosphate, in 10% diethanolamine buffer (Pierce, Rockford, IL, USA) was added. After a 30–60 min room temperature incubation the appearance of a yellow colour in the wells was quantified at 405 nm. Known positive and negative controls were always assayed in parallel to the test samples. Inter- and intra-assay variations were ,10%. A positive sample was defined as one yielding an optical density value that was at least 2SD above the mean value of known negative samples. PCR/LCR PCR was performed employing two sets of plasmid PCR primers (KL1–KL2 and T1–T2) as described by Mahony et al. (1994). Additionally the commercially available Amplicor® system (Roche Molecular Systems, Branchburg, NJ, USA) employing primers specific for the chlamydial cryptic plasmid DNA was used. Positive controls were derived from C.trachomatis grown in McCoy cell cultures. The Abbott–LCx® system (Abbott, Wiesbaden, Germany) was utilized for LCR (Lee et al., 1995). Statistics Fisher’s exact test and Student’s t-test were utilized to examine differences in discrete variables among the groups of subjects. P , 0.05 was considered significant.

Results In-vitro fertilization Ovarian stimulation was performed using a similar protocol for all patients. Gonadotrophin-releasing hormone analogue was continuously administered from the midluteal phase of the previous cycle. After complete down-regulation gonadotrophins [follicle stimulating hormone (FSH) or human menopausal gonadotrophin (HMG)] were given from the second day of the stimulation cycle until human chorionic gonadotrophin (HCG; 10 000 IU) administration. The dosage of gonadotrophins was adapted according to oestradiol concentration and follicle size as determined by transvaginal sonography. When at least three follicles reached 17 mm in diameter by ultrasound, ovulation was induced. Transvaginal ooctye retrieval was performed 34–36 h after ovulation induction and pooled follicular fluid free of visible blood contamination was collected. Immunological analyses Antibodies to chlamydial major outer membrane protein (MOMP) were detected by the Immunocomb® Chlamydia Bivalent (Orgenics, Israel) assay. Chlamydial anti-lipopolysaccharide (LPS) antibodies were determined by enzyme-linked immunosorbent assay (ELISA) using a recombinant Chlamydia-specific LPS fragment (rELISA®, medac Hamburg, Hamburg, Germany). Samples were evaluated according to the instructions of the manufacturer. A titre of 1:200 was considered positive for recombinant (rLPS) IgA in follicular fluid. A recombinant antibody to human hsp 60 (SPA 806, StressGen, Victoria, BC, Canada) was diluted to 10 µl/ml in 0.1 M carbonate buffer, pH 9.8, and 0.1 ml added separately to wells of a microtitre plate. After an overnight incubation at 4°C the wells were washed four times with PBS–Tween. Aliquots (0.1 ml) of follicular fluid diluted 1:4 were added to the wells and the plate floated on a 37°C

926

Distribution of chlamydial antibodies in sera and follicular fluid The distribution between serum and follicular fluid of all antibodies tested is shown in Table I. Sera from 90 (60.4%) and 54 (36.2%) subjects were positive for antichlamydial IgG utilizing the rLPS and the MOMP assays respectively. rLPSdirected and MOMP-directed IgA were detected in sera from 50 (33.6%) and 21 (14.1%) of the subjects respectively. In follicular fluid, IgG antibodies to MOMP were detected in 63 (42.3%) cases while anti-MOMP IgA was present in 13 (8.7%) subjects. IgG anti-rLPS antibodies were present in follicular fluid from 50 (41%) women and 41 (27.5%) women were IgA anti-rLPS positive. There was not a complete concordance between antibodies in serum and follicular fluid. This was most striking for anti-MOMP IgA where 46% of women with this antibody in their follicular fluid were seronegative. Polymerase and ligase chain reaction All cervical samples and follicular fluid samples were negative for C.trachomatis by PCR and LCR. IVF outcome No fertilization occurred with oocytes from 25 (17%) women. Most of these cases were due to severe male infertility. Embryo transfers were performed in 124 (83%) women. Of these, 25 (20%) were delivered of viable term infants, six (5%) ended in spontaneous abortion, four (3%) in ectopic pregnancy and

Follicular immune response to Chlamydia and IVF outcome

Table I. Distribution of chlamydial antibodies in sera and follicular fluids Antibody

Table IV. Relationship between tubal infertility and antichlamydial antibodies in sera

No. positive/no. tested (%) Antibody Sera

MOMP IgA MOMP IgG rLPS IgA rLPS IgG

21/149 54/149 50/149 90/149

No. positive (%)

P value

Follicular fluid (14.1) (36.2) (33.6) (60.4)

13/149 63/149 41/149 50/122

(8.7) (42.3) (27.5) (41.0)

MOMP 5 major outer membrane protein; rLPS 5 recombinant lipopolysaccharide; Ig 5 immunoglobulin.

MOMP IgA MOMP IgG rLPS IgA rLPS IgG

Tubal infertility (n 5 64)

Non-tubal infertility (n 5 106)

14 40 32 49

7 21 26 57

(21.9) (62.5) (50.0) (76.6)

(6.6) (19.8) (24.5) (53.8)

0.006 ,0.0001 0.009 0.003

See Table I for abbreviations.

Table II. Antichlamydial antibodies in sera and pregnancy outcome following embryo transfer Antibody

MOMP IgA MOMP IgG rLPS IgA rLPS IgG

Table V. Relation between tubal infertility and antichlamydial antibodies in follicular fluid

No. positive (%) Pregnant (n 5 35)

Not pregnant (n 5 89)

3 12 13 25

17 38 35 60

(8.5) (34.2) (37.1) (71.4)

(19.1) (42.6) (39.3) (67.4)

See Table I for abbreviations.

Antibody

MOMP IgA MOMP IgG rLPS IgA rLPS IgG

No. positive (%)

P value

Tubal infertility (n 5 64)

Non-tubal infertility (n 5 106)

10 45 23 23

3 34 19 24

(15.6) (70.3) (35.9) (35.9)

(2.8) (32.1) (17.9) (27.6)

0.005 ,0.0001 0.01

See Table I for abbreviations. Table III. Antichlamydial antibodies in follicular fluid and pregnancy outcome after embryo transfer Antibody

MOMP IgA MOMP IgG rLPS IgA rLPS IgG

No. positive (%) Pregnant (n 5 35)

Not pregnant (n 5 89)

0 18 6 10

11 (12.3)a 46 (51.6) 35 (39.3)b 32 (35.9)

(0.0) (51.4) (17.1) (28.5)

5 0.03 vs pregnant; bP 5 0.02 vs pregnant. See Table I for abbreviations. aP

89 (72%) showed no evidence of pregnancy after embryo transfer. There was no significant difference in outcome according to patients’ age. The number of transferred embryos was the same in women with or without antichlamydial antibodies. Antichlamydial antibodies and pregnancy outcome In sera, no association between pregnancy outcome and antichlamydial antibodies could be demonstrated (Table II). In marked contrast, however, the presence of IgA antibodies to both MOMP and rLPS in follicular fluid was associated with a negative pregnancy outcome after embryo transfer (P 5 0.03 and P 5 0.02 respectively, Table III). MOMP IgA was not present in any of the follicular fluid from the 35 women who became pregnant, as opposed to its detection in 11 of 89 (12.3%) women who did not conceive (P 5 0.03). Similarly, rLPS IgA was present in 6 of 35 (17.1%) follicular fluids of women who became pregnant versus 35 of 89 (39.3%) of patients whose IVF cycle did not result in pregnancy (P 5 0.02). There was no relationship between IgG antichlamydial antibodies in follicular fluid and IVF outcome.

Table VI. Relationship between chlamydial antibodies and human 60 kDa heat shock protein expression in follicular fluid Antibody

MOMP IgA MOMP IgA MOMP IgG MOMP IgG rLPS IgA rLPS IgA rLPS IgG rLPS IgG

hsp 60

Antibody No. positive (%)

Result

No. tested

1 – 1 – 1 – 1 –

7 53 7 53 7 53 7 53

3 4 7 34 6 23 6 34

(43) (8) (100) (64) (86) (43) (86) (64)

P value

0.02 0.08 0.04 0.40

1 5 present; – 5 absent. See Table I for abbreviations.

Antichlamydial antibodies and cause of infertility Of our patients, 64 (43%) had occluded Fallopian tubes. The remaining 85 patients (non-tubal infertility group) had multiple causes of infertility, including poor sperm quality (48%), endometriosis (10%), endocrine disorders leading to anovulation (20%), unexplained factors (17%) and miscellaneous other causes (6%). The presence of antichlamydial IgG and IgA MOMP and rLPS antibodies in sera were each associated with a diagnosis of tubal infertility (Table IV). Similarly in follicular fluid, anti-MOMP IgG (P , 0.0001), anti-MOMP IgA (P 5 0.005) and rLPS IgA (P 5 0.01) antibodies were all related to a diagnosis of tubal infertility (Table V). In 13 (20%) patients diagnosed with occluded Fallopian tubes no antichlamydial antibody was present in sera or follicular fluid. 927

A.Neuer et al.

Chlamydial antibodies and human heat shock protein expression in follicular fluid Due to limited sample volumes, human hsp 60 was analysed in only 60 follicular fluid samples. Of these, seven (11.6%) were positive. The presence of MOMP IgA (P 5 0.02) and rLPS IgA (P 5 0.04) in follicular fluid and the expression of human hsp 60 were significantly related (Table VI). Interestingly, all follicular fluid positive for human hsp 60 were derived from patients with tubal infertility who did not become pregnant after embryo transfer. However, due to the small number of cases the relationship between hsp 60 and poor IVF outcome was not statistically significant and further exploration is needed to draw final conclusions. Discussion The overwhelming majority of women who are infertile due to occluded Fallopian tubes have never been diagnosed as having a sexually transmitted disease (Cates et al., 1993). However, the association between tubal factor infertility and antichlamydial antibodies (Jones et al., 1982) strongly implicates asymptomatic infections by this organism as a major cause of this condition. Studies from several laboratories have further provided evidence that C.trachomatis infections can persist for long periods of time in an unculturable state. The presence of chlamydial DNA and/or protein and persistence of localized inflammation, in the absence of a positive chlamydial culture, have been reported for occluded Fallopian tubes (Patton et al., 1994), trachoma (Holland et al., 1992) and reactive arthritis (Keat et al., 1987). Many women with occluded Fallopian tubes now turn to IVF for treatment of infertility. Whether C.trachomatis still persists in some form in the genital tracts of a number of these women, and whether its presence affects IVF outcome, is still under investigation. A relationship between tubal factor infertility combined with hydrosalpinx and poor IVF outcome has been reported (Katz et al., 1996). We have recently shown that women with cervical IgA antibodies to C.trachomatis surface antigens or to the chlamydial hsp 60, but who were negative for C.trachomatis in the cervix by PCR, had an increased prevalence of unsuccessful IVF outcomes (Witkin et al., 1994). The distribution pattern of antibodies to C.trachomatis in paired follicular fluid and sera from IVF patients has not previously been examined. Detection of IgG and IgA antiMOMP and anti-rLPS antibodies in both sera and follicular fluid suggest that these antibodies can enter the follicular fluid by transduction from the circulation. However, the unique relationship between follicular fluid IgA anti-MOMP and antirLPS antibodies and both expression of human hsp 60 and IVF failure after embryo transfer strongly suggest that local production of IgA antibodies in response to C.trachomatis was also being induced within the genital tract. C.trachomatis infections of the male genital tract have also been associated with local IgA antibody production in the absence of systemic anti-chlamydial immunity (Witkin et al., 1995). A recent study identified an association between IgA antibodies, but not IgG antibodies, to the chlamydial hsp 60 928

and chronic salpingitis (Dieterle and Wollenhaupt, 1996). This suggested the predominance of a mucosal, rather than a systemic, immune response to C.trachomatis in women with upper genital tract infections. The absence of C.trachomatis in follicular fluid, as determined by PCR and LCR, indicates that active Chlamydia replication and release of elementary bodies into the extracellular fluid was not occurring at this location. The antibody observations, plus the findings that all women whose follicular fluid was positive for anti-MOMP IgA, anti-rLPS IgA and human hsp60 had laparoscopy-verified tubal occlusion and who were unsuccessful at IVF, suggest the possibility of a persistent chlamydial infection in these patients. In conclusion, this study provides another link between C.trachomatis-associated tubal infertility and subsequent IVF failure. Women with a history of occluded Fallopian tubes who were positive for chlamydial IgA antibodies and human hsp 60 in follicular fluid had a poor IVF outcome. We would suggest that a persistent upper genital tract chlamydial infection might contribute to the IVF failure observed in these women. Acknowledgements The technical assistance of Ulrike Golesorkhi and M.Klehr-Martinelli, and the critical review and expert discussion with Jan Jeremias is gratefully acknowledged. Supported in part by Deutsche Forschungsgemeinschaft grant Ne 602/1-1.

References Castillo, J.A., Sampalo, A., Molina, R. et al. (1990) Mononuclear cell subpopulations in human follicular fluid from stimulated cycles. Am. J. Reprod. Immunol., 22, 127–129. Cates, W. Jr., Joesoef, M.R. and Goldman, B. (1993) Atypical pelvic inflammatory disease: Can we find predictors? Am. J. Obstet. Gynecol., 169, 341–346. Cerrone, M.C., Ma. J.J. and Stephens, R.S. (1991) Cloning and sequence of the gene for heat shock protein 60 from Chlamydia trachomatis and immunological reactivity of the protein. Infect. Immun., 59, 79–90 . Dieterle, S. and Wollenhaupt, J. (1996) Humoral immune response to the chlamydial heat shock proteins hsp60 and hsp70 in Chlamydia-associated chronic salpingitis with tubal occlusion. Hum. Reprod., 11, 1352–1356. Holland, S.M., Hudson, A.P., Bobo, L. et al. (1992) Demonstration of chlamydial RNA and DNA during a culture-negative state. Infect. Immun., 60, 2040–2047. Jones, R.B., Ardery, B.R., Hui S.L. et al. (1982) Correlation between serum chlamydial antibodies and tubal factor infertility as a cause of infertility. Fertil. Steril., 38, 553–558. Katz, E., Akman, M.A., Damewood, M.D. and Jairo, E.G. (1996) Deleterious effect of the presence of hydrosalpinx on implantation and pregnancy rate with in vitro fertilization. Fertil. Steril., 66, 122–125. Keat, A., Dixey, J., Soonex, C. et al. (1987) Chlamydia trachomatis and reactive arthritis: The missing link. Lancet, 1, 72–74. Lachapelle, M-H., Hemmings, R., Roy, D.-C. et al. (1996) Flow cytometric evaluation of leukocyte subpopulations in the follicular fluids of infertile patients. Fertil. Steril., 65, 1135–1140. La Verda, D. and Byrne, G. I. (1994) Interactions between macrophages and chlamydiae. In Zwilling, B. and Eisenstein, T.K. (eds), Macrophage– Pathogen Interactions. Marcel Dekker, New York, pp. 381–399. Lee, H.H., Chernesky, M.A., Burczak, J.D. et al. (1995) Diagnosis of Chlamydia trachomatis genitourinary infection in women by ligase chain reaction assay of urine. Lancet, 345, 213–216. Loukides, J.A., Randall, A., Edwards, R. et al. (1990) Human follicular fluid contains tissue macrophages. J. Clin. Endocrinol. Metab., 71, 1363–1367. Mahony, J.B., Luinstra, K.E., Waner, J. et al. (1994) Interlaboratory agreement study of a double set of PCR plasmid primers for detection of Chlamydia trachomatis in a variety of genitourinary specimens. J. Clin. Microbiol., 32, 87–91. Meikle, S.F., Xiaozheng, Z., Marine, W. et al. (1994) Chlamydia trachomatis

Follicular immune response to Chlamydia and IVF outcome antibody titers and hysterosalpingography in predicting tubal disease in infertility patients. Fertil. Steril., 62, 305–312. Moore, D.E., Foy, H.M., Daling, J.R. et al. (1982) Increased frequency of serum antibodies to Chlamydia trachomatis in infertility due to distal tubal disease. Lancet, 11, 574–577. Patton, D.L., Askienazy-Elbhar M., Henry-Suchet, J. et al. (1994) Detection of Chlamydia trachomatis in fallopian tube tissue in women with postinfectious tubal infertility. Am. J. Obstet. Gynecol., 171, 95–101. Schmidt, J.A. and Abdulla, E. (1988) Down-regulation of IL-1β biosynthesis by inducers of the heat-shock response. J. Immunol., 6, 2027–2034. Witkin, S.S., Sultan, K.M., Neal, G.S. et al. (1994) Unsuspected Chlamydia trachomatis infection and in vitro fertilization outcome. Am. J. Obstet. Gynecol., 171, 1208–1214. Witkin, S.S., Kligman, I. and Bongiovanni, A.M. (1995) Relationship between an asymptomatic male genital tract exposure to Chlamydia trachomatis and an autoimmune response to spermatozoa. Hum. Reprod., 10, 2952–2955. Witkin, S.S., Jeremias, J., Neuer, A. et al. (1996) Immune recognition of the 60 kDa heat shock protein: Implications for subsequent fertility. Infect. Dis. Obstet. Gynecol., 4, 157–158. Received on December 11, 1996; accepted on March 5, 1997

929