Induction of immunoresponse by subclinical male genital tract infection?*

Transcription

1 FERTILITY AND STERILITY Vol. 65, No, 6, June 1996 Copyright 1996 American Society for Reproductive Medicine Printed on acid-free paper in U. S. A. Induction of immunoresponse by subclinical male genital tract infection?* Waltraud Eggert-Kruse, M,D,t+ Stephan Probst, M.D. t Gerhard Rohr, M.D., Ph.D Wolfgang Tilgen, M.D. II Benno Runnebaum, M.D.t Women's University Hospital and University of Heidelberg, Heidelberg, Germany Objective: To determine the relationship of subclinical infection or inflammation of the male genital tract, as evaluated with seminal markers, with local antisperm antibodies as potential parameter of immunoresponse. Patients: One hundred ninety-one randomly chosen males of subfertile couples who were asymptomatic in terms of genital tract infection. Setting: Outpatient Infertility Clinic of the University of Heidelberg, Germany. Main Outcome Measures: Determination of leukocytes rates in semen using an immunocytochemical method for differentiation of round cells and measurement of polymorphonuclear (PMN) granulocyte elastase concentration in seminal plasma in addition to semen cultures as screening for subclinical infection of the male genital tract. Determination of local antisperm antibodies (Ab) with the mixed antiglobulin reaction ([MAR] immunoglobulin [Ig] G and IgA) in aliquots of the same ejaculates. Results: Leukocyte rates of the round cells ranged from 0 to 93, leukocytospermia was found in 6.8. This was not related significantly to the presence of local antisperm antibodies of the IgG or IgA class. There was also no significant association of antisperm Ab with the concentration of PMN granulocyte elastase in seminal plasma and the outcome of semen cultures. Conclusions: The results of this prospective study suggest that when the rate or number of leukocytes or the concentration of PMN elastase in semen are taken as markers for subclinical infection or inflammation of the male genital tract, this is not associated significantly with the production of local antisperm Ab of the IgG or IgA class as indicator of immunoreaction. Fertil Steril 1996;65: Key Words: Immunoresponse, cross-reactivity, antisperm antibodies, seminal leukocytes, semen cultures, mixed antiglobulin reaction, MAR It is often assumed that infections ofthe male genital tract might stimulate the immune system resulting in the formation of antisperm antibodies (Ab) Received August 3, 1995, revised and accepted January 30, * Presented in part at the 11th International Meeting of the European Society of Human Reproduction and Embryology, Hamburg, Germany, June 28 to July 1, t Department of Gynaecological Endocrinology and Reproductive Medicine, Women's University Hospital. :j: Reprint requests: Waltraud Eggert-Kruse, M.D., Department of Gynaecological Endocrinology and Reproductive Medicine, Women's University Hospital, University of Heidelberg, VoBstrasse 9, Heidelberg, Germany (FAX: ). Department Internal Medicine IV, Klinikum Mannheim. II Division of Andrology, Department of Dermatology Eggert-Kruse et al. Induction of immunoresponse (1-3). One potentially triggering mechanism might be cross-reactivity between bacteria and certain sperm epitopes (4). In a subclinical form, genital tract infection or inflammation of the male is difficult to diagnose, but elevated rates of white blood cells (WBe) in semen and/or high concentrations of polymorphonuclear (PMN) granulocyte elastase are considered to be adequate seminal markers (5-7). There is increasing evidence that local antisperm Ab are clinically more relevant than circulating antisperm Ab in serum samples of subfertile couples (8-12). Therefore in the present study, antisperm Ab in semen, differentiated in those of the immunoglobulin (Ig) G or IgA class with the mixed antiglobulin reaction (MAR) (13), were related to semen cultures and leukocyte Fertility and Sterility

2 rates after round cell differentiation as well as to the concentration of PMN elastase as potential parameters of subclinical infection or inflammation. Patients MATERIALS AND METHODS One hundred ninety-one randomly chosen males from couples presenting for infertility investigation and treatment at the Outpatient Reproductive Medicine Unit of the University of Heidelberg were enrolled in this prospective study. Couples' infertility ranged from 1 to 17 years with a median of 4 years. Patients' ages ranged from 22 to 57 years with a median of 33 years, and the age of their female partners from 22 to 41 years with a median of 31 years. Medical history was obtained and physical examinations were performed in both partners. Couples were submitted to a comprehensive evaluation of potential causes of their infertility, as described previously (9, 11). Patients with azoospermia were excluded. None of the patients of this investigation offered clinical signs or symptoms of lower genital tract infection, and none of them was treated with antibiotics, antiphlogistics, or corticosteroids. Andrological medication, which had been administered on the basis of the results of sperm analyses and hormonal evaluations, was continued during the time ofthe study (mostly kallikrein in 11, pentoxiphyllin in 4, tamoxiphen in 4, and hmg and hcg in 0.5). Semen Analysis Ejaculates were obtained in the hospital by masturbation into sterile glass jars after 5 days of sexual abstinence. Standard criteria of the World Health Organization (WHO) (13) were used for sperm analysis including sperm volume, ph, count, progressive motility, morphology, viability, and fructose concentration. Examination and Differentiation of Round Cells Immunocytochemical round cell differentiation was performed as described in detail previously (14). Briefly, the number of round cells was counted using a Neubauer chamber (LHD, Heidelberg, Germany). Monoclonal antibodies (mab) of high specificity for common leukocyte antigen were used (Dako LC; Dakopatts, Glostrup, Denmark) for differentiation, and the streptavidin-biotin system (HistoStain SP kit; Zymed Laboratories, South San Francisco, CA) was used for staining of semen smears. Slides were incubated in a blocking solution to eliminate the nonspecific background. Primary antibody, biotinylated second antibody, enzyme conjugate, and a substrate Vol. 65, No.6, June 1996 chromogene mixture were applied exactly according to the manufacturer's instructions. Next, smears were counterstained with Meyer's hematoxillin and mounted by an aqueous mounting solution (Glycergel; Dako, Hamburg, Germany). Positive (peripheral leukocytes) and negative (phosphate-buffered saline [PBS], ph 7.4) controls were included in each of the test series. One hundred cells were examined, and the percentage of leukocytes (referred to as percentage ofleukocytes or leukocyte rate) ofthe round cells was determined using the high power field (HPF) of a standard light microscope. The mean of triplicate readings was taken for analysis. Differentiation of round cells was performed by the same observer throughout the study. Determination of the Granulocyte Elastase Concentration A commercial assay (PMN Elastase IMAC; Merck, Darmstadt, Germany) was used to determine the granulocyte elastase concentration in seminal plasma samples. Briefly, PMN elastase was measured with this immunoactivation test in seminal plasma obtained after centrifugation of semen for 10 minutes at 600 g and was stored at -70 C until use. According to the manufacturer's recommendations, seminal plasma specimens and duplicate test standards were incubated with the provided antibody conjugate solution, bound elastase arproteinease inhibitor complexes, and then were reacted with a secondary antibody that was linked to alkaline phosphatase. The enzyme reaction was stopped after further substrate incubation with 4-nitrophenylphosphate, the optical density was measured photometrically (500 nm), and by means of a standard curve the concentration (ILg/L) of immunoreactive PMN elastase was determined. The mean of duplicate measurements was used. Microbial Screening For bacterial cultures, 10 ILL of semen were inoculated, directly after liquefaction, in a universal transport medium (Port-a-Cul Universal; Becton Dickinson, Heidelberg, Germany) and additionally in Shepard's medium for culture of mycoplasmas. Standard methods were applied (Department of Microbiology, University of Heidelberg, Heidelberg, Germany) for culture and identification of microorganisms, as described in detail previously (15). An immunofluorescence test (Chlamydia trachoma tis Antigen IF test; Medac, Hamburg, Germany) was used to screen semen specimens for the presence of C. trachoma tis antigen. Simultaneously, a broad microbial screening was performed in patients' female partners, and cervical Eggert-Kruse et al. Induction of immunoresponse 1203

3 swabs were taken for detection of potentially pathogenic micro-oganisms and species of the physiologic flora, which were processed as described for ejaculates. Additional swabs from the fornix posterior vaginalis were cultured for yeasts and were screened for Trichomonas vaginalis (phase-contrast microscope). Endocervical material was obtained for isolation of C. trachoma tis (McCoy cell culture) and Herpes simplex virus (cell culture technique) (Microbiology Laboratories of the Department of Dermatology and Department of Virology, University of Heidelberg, Germany) (15). Sperm Antibody Testing Aliquots of the same ejaculates were screened for the presence of local antisperm antibodies by means of the MAR test (13). The MAR test was performed as described in detail by Jager et a1. (16, 17). Briefly, either IgG- or IgA-coated erythrocytes were used to detect seminal antisperm Ab of the IgG and/or IgA class. For MAR IgG testing erythrocytes sensitized with anti-d serum and prepared by the Blood Grouping Laboratory of the Institute of Serology of the University of Heidelberg were used. The major steps for preparation of IgA erythrocytes were as follows: isolation of the gamma fraction from colostrum, fixing, tanning, and coating. The IgA-coated erythrocytes were stored at -20 C after rapid freezing with liquid nitrogen. For MAR IgG testing a drop offresh native semen, a drop of undiluted monospecific antiserum (Behringwerke, Marburg, Germany), and a drop of the sensitized erythrocyte suspension were mixed thoroughly on a microscopic slide before the reaction was then studied at once under a light microscope. For MAR IgA the reaction was observed in capillary tubes directly after aspiration and after 30 min incubation in a moist atmosphere at 37 C. The results were noted as the percentage of the motile spermatozoa incorporated in the mixed agglutinates, which is described as MAR (16,17). MAR testing was performed by a consistent observer throughout the study, the mean of triplicate readings was taken for analysis. Statistical Analysis Chi square (X 2 ) and Fisher's two-tailed exact tests, Wilcoxon rank sum tests, and Spearman rank correlation were performed for statistical evaluation using the Statistical Analysis System (SAS, Cary, NC). The level of significance was set at P < RESULTS Results of Sperm Analysis Standard semen analysis offered medians of 3.8 (range, 0.5 to 9.5) ml for ejaculate volume, 39 (4 to 1204 Eggert-Kruse et al. Induction of immunoresponse 178) X 1Q 6 /ml for sperm count, with a progressive motility of 40 (range, 10 to 60) after liquefaction, 61 (range, 40 to 78) normal forms (standard morphology [13]), with a viability of 65 (range, 24 to 80), ph of 7.2 (range, 7.2 to 8.0), and fructose concentration of 1,540 (range, 220 to 4,000) J..lg/mL. Oligozoospermia «20 X 1Q 6 /ml) was found in 18.9, reduced sperm count «40 X 1Q 6 /ml) in 50.3, and asthenozoospermia «40) in Although the median progressive motility was lower in MAR IgA-positive samples (30 versus 40), results of sperm analysis did not offer a significant relationship with the outcome of MAR testing (MAR IgG and MAR IgA) within this study population. Leukocyte Rates The percentage of leukocytes after immunocytochemical differentiation of round cells ranged from 0 to 93. No significant direct correlation with local antisperm Ab in the same ejaculates was found (Spearman-rank correlation test). A potential relationship of these parameters was also analyzed with regard to several different cutoffs. Leukocyte rates of :2:3 were found in 34.6 (66/191), :2: 10 in 20.4, and :2: 15 in 17.8 (34/191) of semen samples. Ejaculates with elevated leukocyte rates did not offer evidence of local antisperm Ab more frequently than the other specimens. The outcome of MAR testing is shown in Table 1. The presence of seminal antisperm Ab was analyzed with regard to thresholds at MAR > 10 and MAR >30, based on previous findings (17); furthermore, medians and ranges were compared. Mixed antiglobulin reaction IgG was positive (> 10) in 10.5 of semen samples with low leukocyte rates «10) compared with 10.3 in semen specimens with a leukocyte rate :2: 10. Differences also were not significant when higher cutoffs were taken, with an MAR outcome of >30 in 5.1 of semen samples with :2: 10 leukocytes of the round cells compared with 6.6 in semen specimens with a lower percentage of leukocytes or when median MAR results were analyzed. This lack of association also could be shown with regard to MAR IgA, evaluated in 177 ejaculates (93), with a maximum MAR IgA result of 20 in the group of semen samples with leukocyte rates :2: 10 (see Table 1). When other thresholds for leukocyte rates of the round cells were used for selection of semen specimens, e.g., the median, 2, 3, the 75th percentile, or > 15, considered "leukocyte positive," because of interference with sperm functional capacity in previous studies (14), there was also no significant relationship with local antisperm Ab of the IgG or IgA Fertility and Sterility

4 Table 1 Relationship of Leukocytes in Semen and Local Antisperm Antibodies Leukocyte rates* <10 2:10 Total Pt:j: Antisperm Ab in semen MARlgG > (16/152) > (101152) Median IgG ~ 1.5 (0 to 89) Percentage of patients** 79.6 (152) MARlgA > (6/141) > (/141) Median IgA ~ 0 (0 to 60) Percentage of patients** 78.0 (141) 10.3 (4/39) 5.1 (2/39) 2 (0 to 93) 20.4 (39) 2.8 (1/39) 0 (0/39) 0 (0 to 20) 22.0 (39) 10.5 (20/191) 6.3 (12/191) 1.5 (0 to 93) 100 (91) 4.0 (7/177) 0.6 (11177) 0 (0 to 60) 100 (177) l1 :j: :j: * Immunocytochemically determined percentage of leukocytes of the round cells. t Fisher's two-tailed exact test, compared with corresponding group below the threshold indicated. :j: Wilcoxon rank-sum test. Values are percentages per group (with the number of samples/total number per group in parentheses). II All differences not significant (). ~ Values are medians with the ranges in parentheses. ** Values are total percent (with the number in parentheses). class. For example, positive MAR results (MAR IgG >10) were seen in 11.8 (4/34) in leukocyte-positive samples compared with 10.2 (16/157) in ejaculates with lower leukocyte rates, MAR IgG outcome >30 in 5.9 (2/34) compared with 6.4 (10/157), and MAR IgA-positive samples were found in 3.1 versus 4.1, respectively (not significant [], Fisher's two-tailed exact tests). The outcome of MAR IgG and IgA testing correlated significantly (r = 0.63, P < ), and leukocyte rates correlated significantly with the number of leukocytes per ejaculate (r = 0.97, P < ) (Spearman-rank correlation). Leukocytospermia according to WHO definition (> 1 X lob/ml) was found in 6.8 (13/191) of patients (maximum number of leukocytes 3.2 X lob/ml, median 1.5 X lob/ml, [mean 1.7 X lob/ml], median PMN elastase 535 J-lg/L [mean 436 J-lg/L] in this subgroup). Leukocytospermic samples offered 7.7 (1/ 13) MAR-positive results (MAR IgG > 10) compared with 10.7 (19/178) of the other specimens. None of the leukocytospermic semen samples was positive for MAR IgG >30, and all were negative with regard to MAR IgA (MAR 2::: 10). The presence oflocal antisperm Ab also did not differ significantly when semen samples with leukocyte concentrations below or above the median, or the 75th percentile, were compared; furthermore, there was no significant correlation of seminal antisperm Ab of the IgG or IgA class with leukocyte counts in the whole ejaculate. Relationship of Local Antisperm Ab and PMN Elastase The PMN elastase could be determined in seminal plasma samples of the same ejaculates in 158 pa- tients. No direct correlation of this marker enzyme with local antis perm Ab was found (Spearman rank correlation). Concentrations of 2:::250 J-lglL were found in 39.9 and of 2:::500 J-lglL PMN elastase in 25.9 of samples, which were not related significantly to local antisperm Ab of the IgG or IgA class (see Table 2). This could also be shown with regard to several other thresholds as well as for medians and ranges of PMN elastase concentration. None of the ejaculates with seminal plasma PMN elastase 2::: 500 J-lg/L was positive for antisperm Ab of the IgA class. Differences were also not significant, when PMN elastase levels in seminal plasma were compared in ejaculates with stronger evidence of local antisperm Ab (MAR >30) with concentrations of 2::: 500 J-lglL in 18.2 in samples with MAR IgG >30 compared with 26.5 in semen specimens with MAR results below 30 MAR IgG. Medians were 195 (range, 1 to 880) J-lglL in samples with MAR IgG >30, compared with 145 (range, 36 to 760) J-lglL PMN elastase in ejaculates with lower MAR outcome (, Wilcoxon rank sum test). Local Antisperm Ab and Semen Cultures Microbial screening in ejaculates offered growth of potential pathogenic bacteria (e.g., Escherichia coli, Enterococcus, Streptococcus agalactiae [Group B], Staphylococcus aureus, Proteus, and Klebsiella) in Escherichia coli was cultured in 12.6, enterococci in 29.1, B streptococci in 5.7, and Gardnerella vaginalis in 4.6, with polymicrobial growth for the most part. Additionally the majority of specimens were colonized with species of the physiologic flora (e.g., Staphylococcus epidermidis [40.6], Vol. 65, No.6, June 1996 Eggert-Kruse et al. Induction af immunarespanse 1205

5 ' Table 2 Relationship of PMN Elastase and Seminal Antisperm Antibodies Antisperm antibody testing Negative (.,;10) Postive (>10) Total p*t MAR IgG PMN elastase ():j: 2: 250 IlglL 2:500 llgll 2:750 IlglL Median PMN elastase (llgll)11 Percentage of patients ()~ MARlgA PMN elastase ():j: 2:250 IlglL 2:500 IlglL 2:750 IlgIL Median PMN elastase (llgll)11 Percentage of patients ()~ 40.0 (56/140) 25.7 (36/140) 5.0 (71140) 148 (36 to 760) 88.6 (140) 39.9 (57/143) 27.8 (39/143) 6.3 (91143) 158 (1 to 880) 96.0 (143) 38.9 (7/18) 27.8 (5/18) 11.1 (2/18) 178 (1 to 880) 11.4 (18) 33.3 (2/6) o (016) o (0/6) 123 (22 to 455) 4.0 (6) 39.9 (63/158) 25.9 (41/158) 5.7 (91158) 154 (1 to 880) 100 (158) 39.6 (59/149) 26.2 (39/149) 6.0 (91149) 154 (1 to 880) 100 (149) * Fisher's two-tailed exact test, compared with corresponding group below the threshold indicated. t Wilcoxon rank-sum test. :j: Values are percentages per group (with the number of samplesltotal number per group in parentheses). All differences not significant (). II Values are medians with the ranges in parentheses. ~ Values are total percent (with the number in parentheses). Streptococcus viridans [13.7], Corynebacteria, and Diptheroids). Mycoplasmas were cultured in 15.4 of samples, Mycoplasma hominis in 3.8, and Ureaplasma urealyticum in All semen samples were negative for C. trachomatis, as well as endocervical material of patients' female partners, evaluated by means of McCoy cell culture, and all semen and cervical-vaginal specimens were negative for Neisseria gonorrhoeae and Trichomonas vaginalis. Microbial findings were not related significantly to local sperm Ab of the IgG or IgA class, as can be seen in Table 3. All MAR IgA-positive samples were negative for mycoplasmas. The presence of local antisperm Ab also did not differ significantly when MAR results of >30 were taken as cutoff, e.g., MAR IgG >30 was found in 6.6 in sterile ejaculates or those colonized with species of the physiologic flora compared with 4.1 (MAR IgA in 0) in ejaculates with growth of potentially pathogenic bacteria. Furthermore, there were no significant differences between MAR positive or negative semen samples with regard to the different bacterial species, e.g., E. coli, enterococci, Streptococcus agalactiae, S. viridans, Staphylococcus epidermidis, or Proteus, confirming previous findings in a different group of asymptomatic subfertile patients (17). Antisperm Ab in semen were also not related to microbial findings in the lower genital tract of patients' female partners. The microbial spectrum of the cervix was con- Table 3 Outcome of Semen Cultures and Local Antisperm Antibodies Antisperm antibody testing Negative «10) Positive (> 10) Total p* MAR IgGt Potentially pathogenic bacteria 40.6 (65/160) Sterile cultures or physiologic flora 59.4 (95/160) M. hominis 4.3 (7/162) U. urealyticum 12.3 (20/162) 45.0 (9/20) 55.0 (11/20) 0 (0/20) 10.0 (2/20) 41.1 (74/180) 58.9 (106/180) 3.8 (7/182) 12.1 (22/182) :j: MAR IgA Potentially pathogenic bacteria 42.0 (68/162) Sterile cultures or physiologic flora 58.0 (94/162) M. hominis 4.3 (7/164) U. urealyticum 11.6 (19/164) 20.0 (1/5) 80.0 (4/5) 0 (0/5) 0 (015) 41.3 (69/167) 58.7 (98/167) 4.1 (71169) 11.2 (19/169) * Fisher's two-tailed exact test. t Values are percentages per group (with the number of samplesltotal number per group in parentheses). :j: All differences not significant () Eggert-Kruse et al. Induction of immunoresponse Fertility and Sterility

6 sistent with that reported in a much larger study population (15). DISCUSSION In this prospective study, local antisperm Ab of the IgG or IgA class were related to WBC in semen and the concentration of PMN elastase in seminal plasma as potential markers of subclinical infection or inflammation of the genital tract. The results indicate that in unselected asymptomatic patients, these parameters are not associated significantly with seminal antisperm Ab. All tests were performed in aliquots of the same ejaculates which were also used for semen cultures and therefore allow adequate comparison. The MAR has been proven as a reliable test system (10, 16, 17). The outcome did not correlate significantly with WBC in semen or with the concentration of PMN elastase in seminal plasma. This lack of association could be demonstrated with regard to several different thresholds for antis perm Ab testing, e.g., MAR >10 or >30, or for seminal infection or inflammation markers, used according to previous studies (7, 14, 17), as well as with regard to medians and ranges. Antisperm Ab in semen are an important cause of male subfertility (8,10,12), particularly those of the IgA class. The outcome of MAR testing was related significantly to subsequent fertility in prospective studies (17). However, although clinically relevant, it has to be considered that the formation of sperm antibodies in the genital tract is only one indicator of immunoresponse. The possible induction of other reactions of the immune system by subclinical genital tract infection or inflammation was not evaluated and can not be excluded in the present study. Subclinical infection of the male genital tract is difficult to diagnose. Leukocytes are considered as an established infection and inflammation marker, although the role of seminal leukocytes for male fertility has been the subject of considerable controversy during the last years (18, 19). The use ofmab and, e.g., the streptavidin-biotin system for reading allows clear differentiation of round cells to determine the rate ofleukocytes. A possible interobserver variability was excluded in the present investigation by using one consistent observer throughout the whole study. A number of > 1 X 10 6 /ml leukocytes in semen has been recommended as a critical cutoff for leukocytospermia by the WHO (13), because of its association with infection (5). Leukocytospermia may also be a result of noninfectious inflammatory processes in the male genital tract that are evoked in reponse to the presence of abnormal spermatozoa or chemical Vol. 65, No.6, June 1996 irritants. High leukocyte rates (>15) of the round cells were related significantly to reduced sperm functional capacity in sperm-cervical mucus migration tests (14), and pyospermia interfered with the outcome of the hamster-oocyte assay (20). In the present study, no significant differences were found for the outcome of MAR testing when semen samples with low or elevated leukocyte rates were compared, using several different cutoffs for analyses. Leukocytospermic patients did not offer evidence of local antisperm Ab of the IgG or IgA class in their semen more frequently than the other men. The relatively (6,7) low prevalence ofleukocytospermia in the population of this study is consistent with that reported by others in large groups of males of subfertile couples attending an infertility clinic (2.7 and 7) (18, 19). All patients of the present investigation were asymptomatic in terms of genital tract infection. It has to be pointed out that results might be quite different in males with clinical evidence of infection or inflammation of the genital organs. More information might be obtained by further differentiation of the WBC to evaluate the impact of certain subpopulations (21, 22), and further efforts are necessary to evaluate the relationship with cytokines (23). Furthermore it may be essential to determine the localization of genital tract inflammation. Polymorphonuclear elastase is secreted by activated granulocytes and is therefore regarded as a suitable inflammation marker. The concentration of this enzyme, using an immunological test system, correlated significantly with the rate of WBC in semen (r > 0.5, P < ), confirming other reports (7). Polymorphonuclear elastase levels were not significantly different in antisperm Ab-positive and -negative samples. This could be demonstrated with regard to several thresholds as well as for medians and ranges. However, it has to be kept in mind, that PMN concentrations considered to be "really" high (> 1,000 J1glL) (7), were not found within this asymptomatic study population. The value of bacterial cultures from semen specimens and from the uterine cervix for the detection of subclinical genital infections is controversial, and the relationship of positive microbial findings in asymptomatic patients with subfertility is unclear (5, 15,20). However, genital tract infections in both males and females are thought potentially to trigger immunoresponse with relevance for subsequent fertility, and some studies indicate that inflammations such as urethritis, prostatitis, or epididymitis may facilitate the formation of antisperm Ab (1-3). Common antigenic determinants might be shared by infectious organisms and components of the reproductive system, and molecular mimicry between bac- Eggert Kruse et al. Induction of immunoresponse 1207

7 teria and sperm might be a major factor inducing antisperm immunologic reactions. The likely candidates for cross-reactivity might be either proteins or lipids with carbohydrate terminations (4). Monoclonal anti-human sperm Abs offered widespread cross-reactivity in vitro with, e.g., E. coli, S. viridans, Klebsiella, and S. aureus (4). The screening for colonizing microorganisms in the present study offered, apart from mycoplasmas, positive semen cultures in the majority of semen specimens. Potentially pathogenic bacteria were found in 41, e.g., E. coli in 13, enterococci in 29, mostly polymicrobial growth was observed. The most frequently cultured species of the physiologic flora were Staphylococcus epidermidis in another 41 or S. viridans in 14 of samples, confirming previous studies (15). Microbial findings did not differ in ejaculates with or without evidence of local antisperm Ab. None of the MAR IgA positive semen specimens offered growth of mycoplasmas according to findings of others (24, 25) who used different methods for detection of sperm Ab. All samples evaluated in the present study were free of C. trachoma tis, and there was no evidence of chlamydial infection in patients' partners. The microbial spectrum of the cervix, which confirmed results in a much larger asymptomatic population (15), was not related significantly to antisperm Ab or seminal infection markers of their husbands. It cannot be excluded that the use of more sensitive methods for detection of C. trachoma tis might have offered a higher prevalence of these sexually transmitted pathogens, and for potentially pathogenic anaerobes a much higher rate may be expected if a more specific and differentiated screening with special focus on anaerobic bacteria would have been applied instead ofthe standard procedure for semen cultures used in this study. The findings of this investigation confirm the lack of relationship between microbial colonization of semen samples and formation of antisperm Ab, demonstrated in previous studies, in which several different methods were used for determination of circulating antisperm Ab in serum samples of asymptomatic subfertile couples, e.g., the tray-agglutination test (TAT), RIA, or ELISA (9, 11). In summary, although results might be quite different in patients with evidence of genital tract infection, the results of this prospective study suggest that when the rate or number ofleukocytes in semen or the concentration of PMN elastase are taken as markers for sublinical infection or inflammation of the male genital tract, this is not associated significantly with the production of local antisperm Ab of the IgG or IgA class in semen as indicator of immunoreaction Eggert-Kruse et al. Induction of immunoresponse REFERENCES 1. Witkin SS, Toth A. Relationship between genital tract infections, sperm antibodies in seminal fluid, and infertility. Fertil Steril 1983;40: Ingerslev HJ, Walter S, Anderson JT, BrandenhoffP, Eldrup J, Geerdsen JP, et al. A prospective study of antisperm antibody development in acute epididymitis. J Urol 1986;136: Shahmanesh M, Stedronska J, Hendry WF. Antispermatozoal antibodies in men with urethritis. Fertil Steril 1986;46: Kurpisz M, Alexander NJ. Carbohydrate moieties on sperm surface: physiological relevance. Fertil Steril 1995;63: Comhaire F, Verschraegen G, Vermeulen L. Diagnosis of accessory gland infection and its possible role in male infertility. Int J Androl 1980;3: WolffH, Politch JA, Martinez A, Haimovici F, Hill JA, Anderson DJ. Leukocytospermia is associated with poor semen quality. Fertil Steril 1990;53: Wolff H, Anderson DJ. Evaluation of granulocyte elastase as a seminal plasma marker for leukocytospermia. Fertil Steril 1988;50: Bronson R, Cooper G, Rosenfeld D. Sperm antibodies: their role in infertility. Fertil Steril 1984;42: Eggert-Kruse W, Christmann M, Gerhard I, Pohl S, Klinga K, Runnebaum B. Circulating antisperm antibodies and fertility prognosis: a prospective study. Hum Reprod 1989;4: Kremer J, Jager S. The significance of antisperm antibodies for sperm-cervical mucus interaction. Hum Reprod 1992;7: Eggert-Kruse W, Huber K, Rohr G, Runnebaum B. Determination of antisperm antibodies in serum samples by means of enzyme-linked immunosorbent assay-a procedure to be recommended during infertility investigation? Hum Reprod 1993;8: Marshburn PB, Kutteh WHo The role of antisperm antibodies in infertility. Fertil Steril 1994;61: World Health Organization. Laboratory manual for the examination of human semen and semen-cervical mucus interaction. 3rd ed. Cambridge, The Press Syndicate ofthe University of Cambridge, 1992: Eggert-Kruse W, Bellmann A, Rohr G, Tilgen W, Runnebaum B. Differentiation of round cells in semen by means of monoclonal antibodies and relationship with male fertility. Fertil Steril 1992;58: Eggert-Kruse W, Pohl S, Niiher H, Tilgen W, Runnebaum B. Microbial colonization and sperm-mucus interaction-results in 1000 infertile couples. Human Reprod 1992;7: Jager S, Kremer J, van Schlochteren-Draaisma T. A simple method of screening for antis perm antibodies in the human male. Detection of spermatozoal surface IgG with the direct mixed antiglobulin reaction carried out on untreated fresh human semen. Int J Fertil 1978;23: Eggert-Kruse W, Hofsiii.l A, Haury E, Tilgen W, Gerhard I, Runnebaum B. Relationship between local antisperm antibodies and sperm-mucus interaction in vitro and in vivo. Hum Reprod 1991;6: Tomlinson MJ, Barratt CLR, Cooke ID. Prospective study of leukocytes and leukocyte subpopulations in semen suggests they are not a cause of male infertility. Fertil Steril 1993; 60: Yanushpolsky EH, Politch JA, Hill JA, Anderson DJ. Antibiotic therapy and leukocytospermia: a prospective, randomized, controlled study. Fertil SteriI1995;63: Fertility and Sterility

8 Berger RE, Karp LE, Williamson RA, Koehler J, Moore DE, Holmes KK. The relationship ofpyospermia and seminal fluid bacteriology to sperm function as reflected in the sperm penetration assay. Fertil SteriI1982;37: Barratt CL, Harrison PE, Robinson A, Cooke ID. Antisperm antibodies and lymphocyte subsets in semen-not a simple relationship. Int J Androl 1990;13: Witkin SS, Goldstein H. Reduced levels of suppressor/cytotoxic lymphocytes in semen from vasovasostomized men: relationship to sperm autoantibodies. J Reprod Immunol 1988; 14: Hill JA, Haimovici F, Politch JA, Anderson DJ. Effects of soluble products of activated lymphocytes and macrophages (lymphokines and monokines) on human sperm motion parameters. Fertil SteriI1987;47: Upadhyaya M, Hibbard BM, Walker SM. The effect of Ureaplasma urealyticum on semen characteristics. Fertil Steril 1984;41: Hargreave TB, Torrance M, Young H, Harris AB. Isolation of Ureaplasma urealyticum from seminal plasma in relation to sperm antibody levels and sperm motility. Andrologia 1981;14: Vol. 65, No.6, June 1996 Eggert-Kruse et al. Induction of immunoresponse 1209