The relationship between immunosuppressive activity and immunoregulatory cytokines in seminal plasma: Influence of sperm autoimmunity and seminal leukocytes

Journal of Reproductive Immunology 71 (2006) 57–74

The relationship between immunosuppressive activity and immunoregulatory cytokines in seminal plasma: Influence of sperm autoimmunity and seminal leukocytes Robert Ochsenk¨uhn a,1 , Anne E. O’Connor a , Jonathan J. Hirst b , H.W. Gordon Baker c , David M. de Kretser a , Mark P. Hedger a,∗ a

Monash Institute of Medical Research, Monash University, 27-31 Wright Street, Clayton, Vic. 3168, Australia b Department of Physiology, Monash University, Melbourne, Vic., Australia c Monash IVF, Epworth Hospital, Melbourne, Vic., Australia Received 20 September 2005; received in revised form 4 January 2006; accepted 4 January 2006

Abstract While the contributions of prostasomes, polyamines and prostaglandins to the immunosuppressive activity (ISA) of human seminal plasma have been well-characterised, the contribution of immunoregulatory cytokines found in seminal plasma has received relatively little attention. Semen samples were collected from adult men displaying normospermic parameters, sperm antibodies or substantially elevated seminal leukocytes. Samples were processed through ultracentrifugation and dialysis (<3500 Da) to remove prostasomes, polyamines and prostaglandins, and then assayed for ISA by an in vitro T lymphocyte inhibition assay, as well as by specific immunoassays for transforming growth factor ␤1 (TGF␤1 ), interleukin-10 (IL-10), activin A and the activin-binding protein, follistatin. Seminal plasma from all groups retained substantial ISA following processing. Compared with normospermic men, this ‘large’ molecular weight ISA fraction was significantly increased in a subset of men with sperm antibodies, but was not altered in the group with elevated leukocytes. There was no relationship between ISA and any cytokine examined, and only TGF␤1 was present at levels sufficient to contribute to ISA. Inhibition with a TGF␤-specific antibody reduced ISA in seminal plasma by approximately 50%. Across all patients, TGF␤1 levels were positively correlated with sperm numbers in the ejaculate ∗

Corresponding author. Tel.: +61 3 9594 7003; fax: +61 3 9594 7114. E-mail address: [email protected] (M.P. Hedger). 1 Present address: Department of Obstetrics and Gynecology, Klinikum Großhadern, Ludwig-MaximiliansUniversity, 81377 Munich, Germany. 0165-0378/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.jri.2006.01.002

58

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

and with activin A, but not with follistatin or IL-10. Activin A and IL-10 also displayed a positive relationship, and elevated leukocytes was associated with a significant elevation of IL-10 and activin A, but not TGF␤1 . It is concluded that ‘large’ molecular weight molecules, the most important of which appears to be TGF␤1 , make a significant contribution to immunosuppression by human seminal plasma. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Activin; Autoimmunity; Assays; Inflammation; Interleukin-10; Transforming growth factor ␤

1. Introduction The inhibitory effect of seminal plasma in T lymphocyte functional assays in vitro has been attributed to a number of factors, including prostasomes (Kelly et al., 1991), oxidised polyamines (Allen and Roberts, 1986), prostaglandins of the E series (Skibinski et al., 1992), non-specific lymphocyte-suppressing proteins (Maccioni et al., 2001; Veselsk´y et al., 2002) and immunoregulatory cytokines (Nocera and Chu, 1993; Rajasekaran et al., 1996). In particular, prostaglandins and cytokines are well-characterised regulators of immunity and inflammation, and are most likely to play a significant immunological role in the male and female reproductive tracts. With respect to immunosuppression by seminal plasma, the prostaglandins have received the most study, with less attention paid to the contribution of immunoregulatory cytokines. Cytokines with immunosuppressive activity that have been positively identified in human seminal plasma are transforming growth factor ␤1 (TGF␤1 ) and TGF␤2 (Nocera and Chu, 1993; Srivastava et al., 1996; Loras et al., 1999), interleukin-10 (IL-10) (Rajasekaran et al., 1996; Huleihel et al., 1999; Miller et al., 2002) and activin A (Anderson et al., 1998). While a number of studies have investigated changes in the levels of these particular cytokines in various infertility conditions, their responses to immune events within the male reproductive tract have not been investigated. In general, infection and inflammation lead to up-regulation of cytokines, which contributes to activation of either type 1 (cell-mediated) and type 2 (antibody-mediated) responses of the immune system (Jankovic et al., 2001). The cytokines of the type 2 response, particularly IL-10, and specific immunosuppressive cytokines such as TGF␤1 also play an important role in limiting and ultimately resolving the immune response (MacDonald, 1998). In addition to resolution of the inflammatory/immune response, production of IL10 and the TGF␤s is generally associated with protection against autoimmune disease (Letterio and Roberts, 1998; Volk et al., 2001). Activin A is a member of the TGF␤ family which inhibits both T and B lymphocyte activity in vitro and opposes the action of the key inflammatory cytokines IL-1 and IL-6 (Phillips et al., 2001). In contrast to TGF␤, which is regulated by processing of a biologically inactive (latent) precursor to a mature (active) form, the biological activity of activin A in vivo is regulated by a specific binding protein called follistatin (Phillips et al., 2001). It is likely that immune events within the male reproductive tract are associated with alterations in these cytokines and/or their bio-available levels. In support of this concept we have previously shown that, following removal of prostasomes and inactivation of polyamine activity, there is an inverse relationship between T lympho-

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

59

cyte inhibitory activity in the seminal plasma and autoimmune infertility associated with sperm antibodies (Imade et al., 1997). This observation suggested that immunosuppression by seminal plasma could be an important determinant of autoimmune infertility in the male. The present study was designed to quantify the association between immunosuppression by human seminal plasma and the presence of the immunoregulatory cytokines TGF␤1 , IL10 and activin A as well as the activin-binding protein, follistatin. Moreover, as changes in these parameters might be related to the development of autoimmune infertility, or a marker that such an event has occurred, these were compared in the seminal plasma of men with normal semen parameters, men with sperm auto antibodies and men with evidence of an ongoing inflammatory reaction as indicated by the presence of substantially elevated numbers of seminal leukocytes.

2. Materials and methods 2.1. Patient recruitment and semen analysis Semen samples were obtained with informed consent from patients attending either the Andrology Clinic (Monash Medical Centre, Clayton) or Monash IVF Clinic (Epworth Hospital). All procedures were approved by the appropriate Human Ethics Committees at each institution. Analysis of semen parameters (concentration, % viability, progressive motility, semen volume and leukocyte counts) were carried out according to the appropriate World Health Organization guidelines (WHO, 1999). The presence of sperm antibodies was determined by anti-IgA and anti-IgG immunobead binding test (Kremer and Jager, 1992; Imade et al., 1997). Samples from patients (28–46 years of age) were assigned to the following groups (15 patients per group): men with normal sperm parameters (normospermic group), men with >50% IgG and/or IgA sperm-binding antibodies measured by immunobead binding test (the antibody group) and men with substantially elevated leukocyte numbers in seminal plasma (leukocyte group). The leukocyte group included two patients with leukocyte counts that were slightly less than the arbitrary WHO limit of 1 million/ml for leukocytospermia: 0.78 and 0.94 million/ml, respectively. However, both values were more than 3 S.D.s from the mean leukocyte concentration observed in samples from a larger cohort of patients without defined leukocytospermia (0.07 ± 0.22 million/ml; n = 66 patients) collected over the same time period, and were well above the normal range of leukocyte numbers found in other published studies (Aitken and Baker, 1995). Patient overlap between the antibody group and the leukocyte group was avoided. Although a Kremer sperm–mucus penetration test was not performed to confirm functional infertility in the antibody-positive patients, it has been established that there is a strong direct relationship between impaired sperm function and the percentage of sperm with antibody binding in semen samples (Eggert-Kruse et al., 1991; Kremer and Jager, 1992). All cytokines used in these studies were human recombinant (hr) proteins obtained from R and D Systems, Minneapolis, MN, USA. The activin-binding protein, hr-follistatin 288, was provided by the NHPP/National Institute of Diabetes, Digestive and Kidney Disease, Bethesda, MD, USA.

60

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

2.2. Processing of seminal plasma samples Unless otherwise noted, all seminal plasma samples were ultracentrifuged (100,000 × g, 60 min, 4 ◦ C) in a Beckman L8-M ultracentrifuge to remove prostasomes and other particulates prior to assay, as previously described (Imade et al., 1997). Seminal plasma samples (1.0 ml) were dialysed (molecular weight cut-off 3500 Da) against three changes of 600 ml Dulbeccos’ phosphate-buffered saline (DPBS) every 6–8 h at 4 ◦ C. Following ultracentrifugation and dialysis, samples were sterilised by filtration (0.22 ␮m) and stored at −20 ◦ C. 2.3. In vitro T lymphocyte inhibition assay Immunosuppressive activity (ISA) was measured using a rat T lymphocyte proliferation assay in the presence of 0.1 mM hydroxylamine to inhibit polyamine oxidation, as previously described (Imade et al., 1997). Activity was indicated by the ability of samples to inhibit polyclonal activation of proliferation of thymocytes from adult male Dark Agouti rats (Central Animal House, Monash University, Melbourne, Australia) in a dose-dependent manner. The bioassay standard was a pool of dialysed seminal plasma from normospermic patients, which was assigned an arbitrary potency value of 1.0. Relative bioactivity estimates were obtained by comparison of the standard and sample log dose–response curves over the linear range of inhibitory doses using parallel line bioassay statistics (Imade et al., 1997). For the antibody-blocking experiment, aliquots of hr-TGF␤1 and a pool of seminal plasma from three normospermic men were preincubated (30 min) with various concentrations of a pan-specific TGF␤ antibody that immunoneutrarises TGF␤1 , TGF␤2 and TGF␤3 (R and D Systems) prior to addition to the assay, as previously described (Hedger et al., 1998). 2.4. Prostaglandin E (PGE) assay Seminal plasma samples were methyloximated and assayed for PGE content in duplicate without extraction using a heterologous antibody assay, as previously described (Fowden et al., 1987). Methyloximated PGE2 (Sigma) was used as standard, methyloximated [3 H]PGE2 was used as tracer, and the antiserum was raised in sheep against the methyloxime of PGE2 conjugated to bovine serum albumin (generously supplied by Dr. R.I. Cox, CSIRO, Blacktown, Australia). The assay displays cross-reactivities of 270% with PGE1 , 0.3% with 15-keto-PGE2 and <0.1% with all other cyclo-oxygenase products. Samples were assessed in a single assay, with a sensitivity of 0.4 nM and an intra-assay coefficient of variation of 10.6%. 2.5. Cytokine ELISAs Active TGF␤1 was assayed by commercial two-site ELISA kit (R and D Systems) which uses hr-TGF␤1 as standard (Loras et al., 1999). According to the manufacturer, this assay shows 57% cross-reactivity with human TGF␤2 , 0.15% with TGF␤2 and 0.96% with TGF␤3 . All samples were assayed in serial dilutions against the standard. In order to measure total TGF␤1 levels, latent TGF␤1 in the samples was acid-activated by addition

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

61

of 2.5N acetic acid/10 M urea for 15 min, followed by neutralisation with 2.7N NaOH/1 M HEPES, as per the manufacturer’s instructions. Longer periods of acidification (>30 min) led to a significant decline in recovered activity in the seminal plasma samples (data not shown). The assay sensitivity was 25 pg/ml, and the intra- and inter-assay coefficient of variations were 14 and 22%, respectively (n = 3 assays). Due to the relatively large inter-assay coefficient of variation of this assay, all comparisons of TGF␤1 reported below involve samples measured in the same assay. Seminal plasma IL-10 was assayed by commercial human IL-10 UltraSensitive ELISA (Biosource International, Camarillo, CA, USA) using hrIL-10 as standard. According to the manufacturer, this assay shows no cross-reactivity with a broad range of human cytokines. A pool of samples containing elevated levels of IL-10 was assayed at serial dilutions against the standard for assessment of parallelism, and all individual patient samples were assayed at a single dose in a single assay. The assay sensitivity was 0.06 pg/ml, and the intra-assay coefficient of variation was 9.5%. Activin A was measured using a specific ELISA (Knight et al., 1996) according to the manufacturer’s instructions (Oxford Bio-Innovations, Oxfordshire, UK) with some modifications. The standard used was hr-activin A. Standard and samples were diluted with 5% bovine serum albumin in 0.01 M PBS and treated as per instructions. Duplicates were added to the assay plate, which was incubated overnight at room temperature in a sealed humidified box. The next day, the plates were washed before biotinylated E4 monoclonal antibody was added to each well and incubated for 2 h at room temperature. After washing, alkaline phosphatase linked to streptavidin (Invitrogen Corporation, Carlsbad, CA, USA) was added to the wells and incubated at room temperature for 1 h. After further washes, the alkaline phosphatase activity was detected using an amplification kit (ELISA Amplification System; Invitrogen). The limit of detection of the assay was 10 pg/ml, the average intra-assay coefficient of variation was 5.5% and the inter-assay coefficient of variation was 5.7% (n = 4 assays). A pool of samples was assayed at serial dilutions against the standard for assessment of parallelism. All assay plates were read on a Multiskan RC plate reader (Labsystems, Helsinki, Finland) and data were processed using Genesis Lite EIA software (Labsystems). 2.6. Follistatin RIA Follistatin was measured in seminal plasma by a radioimmunoassay employing a rabbit antiserum to bovine follistatin, human follistatin 288 as standard and 125 I-tracer (O’Connor et al., 1999). This assay incorporates a dissociating reagent to eliminate any interference from binding to activin, thereby allowing measurement of total follistatin. A pool of samples was assayed at serial dilutions against the standard for assessment of parallelism, and all individual patient samples were assayed at a single dose in a single assay. The sensitivity of the assay was 0.78 ng/ml and the average intra-assay coefficient of variation was 7.7%. 2.7. Protein recovery and stability study The recovery and stability of each of the proteins in seminal plasma were assessed by addition of activin A (final concentration 8 ng/ml), follistatin 288 (120 ng/ml), IL-10

62

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

(1.1 ng/ml) and active TGF␤1 (90 ng/ml) to three separate pools of untreated whole seminal plasma (i.e. seminal plasma that had not been subjected to ultracentrifugation or dialysis; four normospermic patients/pool). Samples were maintained at room temperature (23 ◦ C) for 0, 6 and 24 h before centrifugation (10,000 × g, 15 min), freezing and storage at −20 ◦ C, prior to assay. 2.8. Statistical analysis All data were analysed using one-way analysis of variance (ANOVA) following appropriate transformation to normalise data and equalise variance where necessary. Specifically, sperm concentration was square root transformed prior to analysis (Comhaire et al., 1987). Mean values were compared using the Student–Newman–Keuls multiple range test, and analysis of correlation was performed using the Pearson correlation test (Sigmastat Version 1.0; Jandel Scientific Software, San Rafael, CA, USA). Differences between groups were considered statistically significant at the p < 0.05 level. ELISA and RIA dose–response curves were compared by linear regression after log–log transformation of the data. The curves were considered to be parallel if the 95% confidence limits of the slopes overlapped.

3. Results 3.1. Effect of repeated dialysis on prostaglandin E and immunosuppressive activity Repeated dialysis of a pool of seminal plasma collected from normospermic men reduced PGE levels below the limit of detection, but approximately 25% of the total ISA measured in the lymphocyte proliferation suppression assay was retained (Fig. 1). This confirmed the presence of functionally significant immunosuppressive factors larger than 3500 Da in human seminal plasma, representing the ‘large’ molecular weight immunosuppressive fraction. Similar levels of this activity were present in seminal plasma from normal (relative activity 0.98 ± 0.41; mean ± S.E.M.; n = 4 samples) and vasectomised (0.95 ± 0.12; n = 4 samples) men, indicating that the activity is primarily derived from the accessory glands or distal was deferens rather than the testis or epididymis. 3.2. Validation of activin A, follistatin, IL-10 and TGFβ1 assays Seminal plasma diluted in parallel with the standard for the activin A and IL-10 ELISA and follistatin RIAs across the entire assay working range up to and including undiluted samples (data not shown), but there was a significant deviation in the curve for TGF␤1 at higher assay concentrations (Fig. 2). This deviation was observed both in serial dilutions of seminal plasma alone and in seminal plasma dilutions that had been individually spiked with a constant amount (1 ng/ml) of TGF␤1 . This appeared to be due to interference by seminal plasma at dilutions of 1/8 or above, leading to an underestimation of absolute levels. As a result, all seminal plasma samples (activated and non-activated) were assayed for TGF␤1 at concentrations below this dilution.

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

63

Fig. 1. Effect of repeated dialysis (molecular weight cut-off 3500 Da) on immunosuppressive activity (ISA) as measured in the lymphocyte suppression assay (a) and prostaglandin E (PGE) levels (b) in a seminal plasma pool from normospermic men. Values are mean and 95% confidence intervals (ISA), or mean of duplicate estimates (PGE). Broken line indicates the level of ISA that is resistant to dialysis.

The recovery and stability of cytokines in seminal plasma was assessed by addition of exogenous activin A, follistatin, IL-10 and active TGF␤1 to three separate sample pools of seminal plasma 0, 6 and 24 h prior to assay. In all cases, the spiked samples diluted out in parallel with the assay standards, with recoveries (mean ± S.D.; n = 3) of the exogenous cytokines in the 0 h samples of 97 ± 12% for IL-10, 108 ± 31% for follistatin, 74 ± 3% for activin A and 78 ± 6% for TGF␤1 . This indicated that absolute values for activin A and TGF␤1 in seminal plasma tended to be slightly underestimated. Both exogenous and endogenous activin A and follistatin were completely stable in seminal plasma for up to 24 h at room temperature (Fig. 3a and b). IL-10 and active TGF␤1 showed small but significant (<12%) losses over 6 h at room temperature, although more substantial declines were evident at the longer time-point: 24 ± 3% (mean ± S.D.; n = 3) for IL-10 and 42 ± 5% for TGF␤1 (Fig. 3c and d). Concentrations of all four proteins in the pooled samples without exogenous protein were close to the mean levels present in the dialysed samples of seminal plasma, indicating that all four proteins were stable under the dialysis conditions.

64

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

Fig. 2. Validation of the TGF␤1 ELISA for human seminal plasma samples. Log dose–log response curves for the TGF␤1 standard (䊉), seminal plasma alone (
) or seminal plasma dilutions “spiked” with 1000 pg/ml hr-TGF␤1 (). Values are mean of duplicates. Useful assay range: range of doses of human seminal plasma that diluted in parallel with the standard. At higher doses, significant deviation from parallelism with the standard was observed, as well as significant reduction in the apparent concentration of the exogenous TGF␤1 .

3.3. Semen parameters Compared with normospermic controls, antibody-positive men showed a significant reduction (p < 0.05) in progressive motility (50.5 ± 3.0% versus 38.4 ± 4.7%; mean ± S.E.M.; n = 15 patients/group), but no difference in sperm number or viability. Men with elevated leukocytes had significantly reduced sperm density compared with normals (115.7 ± 17.9 million/ml versus 67.1 ± 18.1 million/ml; mean ± S.E.M.; n = 15 patients/group), as determined after appropriate square root transformation of the sperm density data (Comhaire et al., 1987). However, the leukocyte group also included three azoospermic patients and one severely oligospermic (0.1 million/ml) patient. After removal of these patients from the analysis, sperm viability and sperm density in the leukocyte group were not different from the normospermic controls, but progressive motility (40.7 ± 3.3%; mean ± S.E.M.; n = 11 patients/group) was significantly reduced (p < 0.05). 3.4. Immunosuppressive activity, activin A, follistatin, TGFβ1 and IL-10 in seminal plasma Compared with normospermic men, mean ISA levels in the ‘large’ molecular weight fraction were not affected in men with elevated leukocyte numbers, but were significantly elevated by 40% in men with sperm autoimmunity (Fig. 4). Nonetheless, the majority of men with sperm auto immunity displayed ISA levels that fell within the normal range.

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

65

Fig. 3. Stability of exogenous cytokines in seminal plasma. Pools of seminal plasma (SP) from normospermic patients were spiked with a cocktail of activin A (8 ng/ml), follistatin (120 ng/ml), IL-10 (1.1 ng/ml) and active TGF␤1 (90 ng/ml), and incubated at 23 ◦ C. Aliquots were taken at 0, 6 and 24 h for assay of (a) activin A, (b) follistatin, (c) IL-10 and (d) TGF␤1 . Values are mean ± S.E.M. of the three individual sample pools. Different letter superscripts indicate significant differences between groups at p < 0.05.

Activin A, follistatin and TGF␤1 were detectable in the seminal plasma of all normospermic men, and IL-10 was detectable in all but two samples (Fig. 5). Follistatin and active TGF␤1 showed a much wider range of levels (5–85 pg/ml and 0–8 ng/ml, respectively) than either activin A (0.1–0.3 ng/ml), IL-10 (0–10 pg/ml) or total TGF␤1 (80–200 ng/ml). TGF␤1 was by far the most abundant of the cytokines in seminal plasma. Most of the TGF␤1 in seminal plasma was present in the latent form (98%), but active TGF␤1 was also detectable in all but one sample. There was no difference in the levels of any cytokine between normal men and men with clinically significant IgA levels in seminal plasma. However, elevated leukocytes were associated with significantly increased activin A and IL-10 levels (p < 0.05) (Fig. 5). Total TGF␤1 also appeared to be elevated in some men with leukocytes, but this did not achieve significance. Note that one patient in the leukocyte group displayed very high levels (>2 S.D. above the mean) of both activin A and IL-10, although TGF␤1 and follistatin levels were

66

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

Fig. 4. Immunosuppressive activity (ISA) in individual samples from normospermic men (normal), men with sperm autoimmunity (antibodies) and men with elevated leukocytes (leukocytes). Bar in each sample group represents arithmetic mean value. Different letter superscripts indicate significant differences between groups at p < 0.05.

within group limits. Removal of this patient from the analysis, however, did not effectively alter the results obtained. When data from all groups were combined, total TGF␤1 showed a small but significant positive correlation with the square root of the sperm number (r2 = 0.090; p < 0.05), while IL-10 showed a significant negative correlation with sperm viability (r = 0.114; p < 0.001). There was no significant relationship between any other sperm parameter and cytokine, follistatin or ISA levels. There was a positive correlation between activin A and follistatin (Fig. 6a), IL-10 (Fig. 6b) and total TGF␤1 (Fig. 6c). Following removal of the single patient with abnormally high activin A levels from the analysis, activin A was still positively correlated with IL-10 and total TGF␤1 , but not with follistatin (p > 0.05). Active TGF␤1 levels in seminal plasma showed no relationship with total TGF␤1 , IL-10, follistatin or activin A levels (data not shown). There was no correlation between ISA and any of the proteins measured, either within any patient group or within the combined patient data. 3.5. Assessment of the contribution of TGFβ1 to immunosuppressive activity The possibility that the immunosuppression was attributable to TGF␤1 was examined further. Both seminal plasma and TGF␤1 suppressed T cells in a dose-dependent manner, although the dose–response curves deviated slightly (Fig. 7a), indicating the influence of

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

67

Fig. 5. Activin A (a), follistatin (b), IL-10 (c), total TGF␤1 (d) and active TGF␤1 (e) in individual samples from normospermic men (normal), men with sperm autoimmunity (antibodies) and men with elevated leukocytes (leukocytes). The cross-bar in each sample group represents arithmetic mean value. Different letter superscripts indicate significant differences between groups at p < 0.05.

other factors. The mean IC50 (dose at 50% suppression of maximum) for seminal plasma in the assay was 1.5 ␮l, and for TGF␤1 was 1–2 ng/ml (n = 4 assays). Latent TGF␤1 was approximately 100-fold less potent than the mature form. As levels of active TGF␤1 in seminal plasma were <10 ng/ml, this cytokine could not account for the inhibitory activity of the seminal plasma in the assay, unless there was

68

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

Fig. 6. Correlation analysis of follistatin (a), IL-10 (b) and total TGF␤1 (c) with activin A levels in patient samples from all groups combined.

conversion of the pre-existing latent TGF␤1 in the assay itself. Consequently, in order to directly determine the contribution of TGF␤1 in the assay, a pan-specific TGF␤ antibody was added to the assay. Whereas TGF␤1 itself was completely inhibited by the antibody, seminal plasma immunosuppression was only partially inhibited (approximately 50%) even at the highest antibody concentration (Fig. 7b). These data indicate that, while a significant proportion of the ISA in the ‘large’ molecular weight fraction of seminal plasma is due to the presence of TGF␤1 (and the other TGF␤ isoforms), other factors still remain to be

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

69

Fig. 7. Comparison of immunosuppressive activity of a seminal plasma and TGF␤1 . (a) Log dose–response curve for a pool a seminal plasma from normospermic patients, hr-active TGF␤1 and hr-TGF␤1 precursor (latent TGF␤1 ) in the lymphocyte proliferation inhibition assay. (b) Effect of co-incubation with a pan-specific TGF␤ antibody on immunosuppressive activity of a seminal plasma and TGF␤1 (final assay concentration: 2 ng/ml). All values are mean ± S.E.M. of quadruplicate assay wells. The dashed line indicates [3 H]-thymidine incorporation in the absence of seminal plasma or TGF␤1 (control response). The dotted line indicates the IC50 response value of the assay.

determined. This contribution by other factors was consistent with the failure to see a direct relationship between ISA and TGF␤1 levels in the various patient groups.

4. Discussion Although there have been a number of studies on immunosuppression by human seminal plasma and the presence of cytokines in this fluid, previous studies have not attempted to quantify the relative contribution of cytokines to seminal plasma immunosuppression. In order to examine this association, we first validated the procedures and assays used for separation and measurement of ISA and specific immunosuppressive cytokines in human seminal plasma. This included separation of the cytokine-containing fraction of seminal plasma from fractions containing the prostasomes, polyamines and prostaglandins using ultracentrifugation and dialysis through a 3500 Da cut-off membrane. Among the larger molecular weight factors with immunosuppressive activity that have been positively identified in human seminal plasma are the immunosuppressive cytokines TGF␤1 and TGF␤2 (Nocera and Chu,

70

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

1993, 1995; Srivastava et al., 1996; Loras et al., 1999), IL-10 (Rajasekaran et al., 1996; Huleihel et al., 1999; Miller et al., 2002) and activin A (Anderson et al., 1998). It has already been established that in human seminal plasma, levels of TGF␤2 are less than 10% of TGF␤1 (Nocera and Chu, 1995; Srivastava et al., 1996). In the T lymphocyte immunosuppression assay used in the present study, TGF␤1 displays an IC50 of 1–2 ng/ml (Hedger et al., 2000) while activin A and IL-10 are considerably less potent in T lymphocyte function assays (Liu et al., 1997; Hedger et al., 2000). The effective bioactivity of activin A would be further reduced by the presence of relatively high concentrations its binding protein, follistatin, in the seminal plasma. The very low concentrations of IL-10 and activin A relative to TGF␤1 in all of the samples examined clearly implicate TGF␤1 as the only immunoregulatory cytokine that contributes significantly to immunosuppression by human seminal plasma. The concentrations of TGF␤1 in seminal plasma were between 80 and 200 ng/ml, with average levels of the activated cytokine of around 3 ng/ml in all treatment groups, although considerable variability was observed. The relatively low proportion of active to total TGF␤1 in most human seminal plasma samples was consistent with previous observations (Nocera and Chu, 1995; Srivastava et al., 1996; Loras et al., 1999; Robertson et al., 2002). The enzymes necessary to activate latent TGF␤1 are also found in seminal plasma, although conversion is dependent upon relatively low pH conditions (Chu and Kawinski, 1998), and it does not appear that latent TGF␤1 is significantly activated in the seminal plasma (pH 7–8) under normal conditions. It should be noted, nonetheless, that the levels of active TGF␤1 in seminal plasma still lie within the range of physiological significance. Based on the failure of a pan-specific TGF␤ antiserum to completely inhibit ISA in seminal plasma, and lack of a significant relationship between the ‘large’ molecular weight ISA and active or total TGF␤1 levels in the seminal plasma of any patient group, other larger molecular weight factors must also contribute significantly to this activity. Although non-cytokine proteins with immunosuppressive activity have been identified in the seminal plasma or accessory gland secretions of rodents and boars, including prostatic steroidbinding protein and the seminal glycoprotein PSP II (Maccioni et al., 2001; Veselsk´y et al., 2002), the physiological significance of such molecules in human seminal plasma still remains to be established. The data in the present study indicate that, with respect to measurement of IL-10, activin A and TGF␤1 in seminal plasma by ELISAs, reduced recoveries and degradation of the cytokines during collection and subsequent processing, as well as direct interference by other seminal plasma components, must be considered confounding factors. Although both IL-10 and TGF␤1 showed progressive losses of activity in seminal plasma at room temperature, the rate of degradation was relatively slow, linear and consistent. Moreover, all three cytokines diluted in parallel with the assay standard across the assay range, whether in the presence or absence of seminal plasma. Within these limits, therefore, it can be assumed that the impact of these variables on studies of measurement of these cytokines is both relatively minor and systematic. For a subset of men with sperm antibodies at least, there was a significant increase in the ‘large’ molecular weight component of ISA compared with normospermic men. However, there were no differences between these two groups with respect to TGF␤1 , IL-10 or activin A levels. Nor was there a significant relationship between ISA and any cytokine in any patient group. The observed increase in activity in some antibody-positive patients may have been

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

71

due to an increase in other large molecular weight immunosuppressive molecules or to changes in seminal plasma cytokines with stimulatory effects on T lymphocyte proliferation (Gruschwitz et al., 1996; Dousset et al., 1997; Naz and Evans, 1998; Huleihel et al., 1999; Eggert-Kruse et al., 2001; Maegawa et al., 2002). It should be pointed out also that, in our previous study in which prostaglandins were not removed by dialysis prior to assay and polyamine activity was inhibited by hydroxylamine instead, we identified a significant inverse relationship between the immunosuppressive activity in seminal plasma and the presence of immobilising sperm antibodies in infertile men (Imade et al., 1997). Taken together, the data in both studies indicate that sperm autoimmunity is associated with a reduction in small molecular weight immunosuppressive factors (<3500 Da), most likely prostaglandins which constitute approximately 70% of the total immunosuppressive activity, rather than specific immunoregulatory cytokines. Ongoing inflammation as indicated in the present study by the presence of elevated leukocytes, was associated with a significant increase in both IL-10 and activin A, but no change in TGF␤1 or total ISA. This lack of effect on ISA was somewhat surprising as numerous studies have shown that pro-inflammatory cytokines, including those with direct effects on the T lymphocyte assay, are dramatically elevated in patients with inflammation and/or leukocytospermia (Gruschwitz et al., 1996; Eggert-Kruse et al., 2001; Maegawa et al., 2002). The failure of seminal plasma inflammatory cytokines to alter ISA in the patients with elevated seminal leukocytes may be attributable, at least in part, to the coincident rise in IL-10 and activin A levels. Activin A in normal seminal plasma is principally of testicular or epididymal origin (Anderson et al., 1998), while TGF␤1 and follistatin are derived from the distal genital tract, including the seminal vesicles and/or the prostate gland (Nocera and Chu, 1995; Anderson et al., 1998). IL-10 is a product of monocytes and regulatory T lymphocytes, although epithelial cells within the male tract are potential sources as well (de Waal Malefyt and Moore, 1998). At present, it is not certain what is the significance of the correlation between activin A levels and the other three proteins assayed, but this seems worthy of further investigation. Our data showing an increase in IL-10 levels in patients with elevated seminal leukocytes are comparable to the results of Rajasekaran et al. (1996). An increase in IL-10 also has been observed in seminal plasma of men with chronic prostatitis, irrespective of the presence of elevated leukocytes (Miller et al., 2002). These studies contrast with the findings of Huleihel et al. (1999), who reported a significant reduction in IL-10 in men with combined leukocytospermia and abnormal sperm parameters (oligoteratoasthenospermia). Although we observed no relationship between any individual sperm parameter and IL-10 in the leukocyte group, leukocytospermia has a complex aetiology, ranging from ongoing genital tract infection to specific subsets of chronic prostatitis (Aitken and Baker, 1995; Weidner et al., 1999), and a much larger study with more patient groups separated on specific inflammatory criteria is warranted. Moreover, the possibility that elevated IL-10 and activin A concentrations are due to production by increased numbers of seminal plasma monocytes, which are major producers of these two cytokines, must be considered (Er¨amaa et al., 1992; de Waal Malefyt and Moore, 1998). Finally, although the relationship between immunosuppressive molecules in seminal plasma and immunological infertility in men remains unclear, there is no doubt that these

72

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

conditions lead to significant alterations in both inflammatory and immunosuppressive cytokines in seminal plasma. It should be borne in mind that these factors also have direct effects on the epithelia, stroma and leukocytes of the female genital tract (Jones et al., 2002; Robertson et al., 2002). Insemination has been causally linked to activation and expansion of populations of lymphocytes mediating forms of ‘active’ immune tolerance at the implantation site, a process in which seminal plasma TGF␤1 has been especially implicated (Robertson et al., 1997). Elevated cytokine concentrations are significantly associated with increased leukocyte numbers and high levels of leukocytes in semen are associated with lower pregnancy rates, even in couples where the man has normozoospermia. Consequently, changes in immunoregulatory proteins and cytokines may have direct consequences for conception rates and fertility in the partners of men with ongoing immunological events. Acknowledgements We are grateful to Ms. Julie Muir for her expert technical assistance. The expert technical assistance of the staff of the Andrology Clinic in Clayton and the Monash IVF Clinic in Richmond is also gratefully acknowledged. This work was supported by grants from the National Health and Medical Research Council of Australia (grants no. 973218 and 143781) and by Monash IVF, Epworth Hospital, Melbourne, Vic., Australia. Part of these data were presented at the Annual Meeting of the European Society of Human Reproduction and Embryology, Vienna, Austria, July 2002. References Aitken, R.J., Baker, H.W., 1995. Seminal leukocytes: passengers, terrorists or good Samaritans? Hum. Reprod. 10, 1736–1739. Allen, R.D., Roberts, T.K., 1986. The relationship between the immunosuppressive and cytotoxic effects of human seminal plasma. Am. J. Reprod. Immunol. Microbiol. 11, 59–64. Anderson, R.A., Evans, L.W., Irvine, D.S., McIntyre, M.A., Groome, N.P., Riley, S.C., 1998. Follistatin and activin A production by the male reproductive tract. Hum. Reprod. 13, 3319–3325. Chu, T.M., Kawinski, E., 1998. Plasmin, substilisin-like endoproteases, tissue plasminogen activator, and urokinase plasminogen activator are involved in activation of latent TGF-␤1 in human seminal plasma. Biochem. Biophys. Res. Commun. 253, 128–134. Comhaire, F.H., de Kretser, D.M., Farley, T.M.M., Rowe, P.J., 1987. The significance of semen analysis for the evaluation of male fertility. Int. J. Androl. 7, 34–46. de Waal Malefyt, R., Moore, K.W., 1998. Interleukin-10. In: Thomson, A.W. (Ed.), The Cytokine Handbook. Academic Press, San Diego, pp. 333–364. Dousset, B., Hussenet, F., Daudin, M., Bujan, L., Foliguet, B., Nabet, P., 1997. Seminal cytokine concentrations (IL-1␤, IL-2, IL-6, sRIL-2, sR IL-6), semen parameters and blood hormonal status in male infertility. Hum. Reprod. 12, 1476–1479. Eggert-Kruse, W., Hofs¨aß, A., Haury, E., Tilgen, W., Gerhard, I., Runnebaum, B., 1991. Relationship between local anti-sperm antibodies and sperm–mucus interaction in vitro and in vivo. Hum. Reprod. 6, 267–276. Eggert-Kruse, W., Boit, R., Rohr, G., Aufenanger, J., Hund, M., Strowitzki, T., 2001. Relationship of seminal plasma interleukin (IL)-8 and IL-6 with semen quality. Hum. Reprod. 16, 517–528. Er¨amaa, M., Hurme, M., Stenman, U.H., Ritvos, O., 1992. Activin A/erythroid differentiation factor is induced during human monocyte activation. J. Exp. Med. 176, 1449–1452. Fowden, A.L., Harding, R., Ralph, M.M., Thorburn, G.D., 1987. The nutritional regulation of plasma prostaglandin E concentrations in the fetus and pregnant ewe during late gestation. J. Physiol. 394, 1–12.

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

73

Gruschwitz, M.S., Brezinschek, R., Brezinschek, H.P., 1996. Cytokine levels in the seminal plasma of infertile males. J. Androl. 17, 158–163. Hedger, M.P., Phillips, D.J., de Kretser, D.M., 2000. Divergent cell-specific effects of activin-A on thymocyte proliferation stimulated by phytohemagglutinin, and interleukin 1␤ or interleukin 6 in vitro. Cytokine 12, 595–602. Hedger, M.P., Nikolic-Paterson, D.J., Hutchinson, P., Atkins, R.C., de Kretser, D.M., 1998. Immunoregulatory activity in adult rat testicular interstitial fluid: roles of interleukin-1 and transforming growth factor ␤. Biol. Reprod. 58, 927–934. Huleihel, M., Lunenfeld, E., Horowitz, S., Levy, A., Potashnik, G., Mazor, M., Glezerman, M., 1999. Expression of IL-12, IL-10, PGE2, sIL-2R and sIL-6Rin seminal plasma of fertile and infertile men. Andrologia 31, 283–288. Imade, G.E., Baker, H.W., de Kretser, D.M., Hedger, M.P., 1997. Immunosuppressive activities in the seminal plasma of infertile men: relationship to sperm antibodies and autoimmunity. Hum. Reprod. 12, 256–262. Jankovic, D., Liu, Z., Gause, W.C., 2001. Th1- and Th2-cell commitment during infectious disease: asymmetry in divergent pathways. Trends Immunol. 22, 450–457. Jones, R.L., Salamonsen, L.A., Findlay, J.K., 2002. Potential roles for endometrial inhibins, activins and follistatin during human embryo implantation and early pregnancy. Trends Endocrinol. Metab. 13, 144–150. Kelly, R.W., Holland, P., Skibinski, G., Harrison, C., McMillan, L., Hargreave, T., James, K., 1991. Extracellular organelles (prostasomes) are immunosuppressive components of human semen. Clin. Exp. Immunol. 86, 550–556. Knight, P.G., Muttukrishna, S., Groome, N.P., 1996. Development and application of a two-site enzyme immunoassay for the determination of ‘total’ activin-A concentrations in serum and follicular fluid. J. Endocrinol. 148, 267–279. Kremer, J., Jager, S., 1992. The significance of antisperm antibodies for sperm–cervical mucus interaction. Hum. Reprod. 7, 781–784. Letterio, J.J., Roberts, A.B., 1998. Regulation of immune responses by TGF-␤. Annu. Rev. Immunol. 16, 137– 161. Liu, Y., de Waal Malefyt, R., Briere, F., Parham, C., Bridon, J.M., Banchereau, J., Moore, K.W., Xu, J., 1997. The EBV IL-10 homologue is a selective agonist with impaired binding to the IL-10 receptor. J. Immunol. 158, 604–613. Loras, B., V´etel´e, F., El Malki, A., Roller, J., Soufir, J.C., Benahmed, M., 1999. Seminal transforming growth factor-␤ in normal and infertile men. Hum. Reprod. 14, 1534–1539. Maccioni, M., Riera, C.M., Rivero, V.E., 2001. Identification of rat prostatic steroid binding protein (PSBP) as an immunosuppressive factor. J. Reprod. Immunol. 50, 133–149. MacDonald, T.T., 1998. T cell immunity to oral allergens. Curr. Opin. Immunol. 10, 620–627. Maegawa, M., Kamada, M., Irahara, M., Yamamoto, S., Yoshikawa, S., Kasai, Y., Ohmoto, Y., Gima, H., Thaler, C.J., Aono, T., 2002. A repertoire of cytokines in human seminal plasma. J. Reprod. Immunol. 54, 33–42. Miller, L.J., Fischer, K.A., Goralnick, S.J., Litt, M., Burleson, J.A., Albertsen, P., Kreutzer, D.L., 2002. Interleukin10 levels in seminal plasma: implications for chronic prostatitis-chronic pelvic pain syndrome. J. Urol. 167, 753–756. Naz, R.K., Evans, L., 1998. Presence and modulation of interleukin-12 in seminal plasma of fertile and infertile men. J. Androl. 19, 302–307. Nocera, M., Chu, T.M., 1993. Transforming growth factor ␤ as an immunosuppressive protein in human seminal plasma. Am. J. Reprod. Immunol. 30, 1–8. Nocera, M., Chu, T.M., 1995. Characterization of latent transforming growth factor-␤ from human seminal plasma. Am. J. Reprod. Immunol. 33, 282–291. O’Connor, A.E., McFarlane, J.R., Hayward, S., Yohkaichiya, T., Groome, N.P., de Kretser, D.M., 1999. Serum activin A and follistatin concentrations during human pregnancy: a cross-sectional and longitudinal study. Hum. Reprod. 14, 827–832. Phillips, D.J., Jones, K.L., Scheerlinck, J.Y., Hedger, M.P., de Kretser, D.M., 2001. Evidence for activin A and follistatin involvement in the systemic inflammatory response. Mol. Cell. Endocrinol. 180, 155–162. Rajasekaran, M., Hellstrom, W., Sikka, S., 1996. Quantitative assessment of cytokines (GRO␣ and IL-10) in human seminal plasma during genitourinary inflammation. Am. J. Reprod. Immunol. 36, 90–95.

74

R. Ochsenk¨uhn et al. / Journal of Reproductive Immunology 71 (2006) 57–74

Robertson, S.A., Mau, V.J., Hudson, S.N., Tremellen, K.P., 1997. Cytokine-leukocyte networks and the establishment of pregnancy. Am. J. Reprod. Immunol. 37, 438–442. Robertson, S.A., Ingman, W.V., O’Leary, S., Sharkey, D.J., Tremellen, K.P., 2002. Transforming growth factor ␤—a mediator of immune deviation in seminal plasma. J. Reprod. Immunol. 57, 109–128. Skibinski, G., Kelly, R.W., Harrison, C.M., McMillan, L.A., James, K., 1992. Relative immunosuppressive activity of human seminal prostaglandins. J. Reprod. Immunol. 22, 185–195. Srivastava, M.D., Lippes, J., Srivastava, B.I., 1996. Cytokines of the human reproductive tract. Am. J. Reprod. Immunol. 36, 157–166. ˇ Veselsk´y, L., Dost´al, J., Kraus, M., Peknicov´a, J., Hol´an, V., Zaj´ıcov´a, A., Jon´akov´a, V., Zelezn´ a, B., 2002. Reverse effect of indomethacin on the immunosuppressive activity of boar seminal immunosuppressive fraction. Anim. Reprod. Sci. 71, 111–123. Volk, H., Khusru, A., Gallagher, G., Sabat, R., Gr¨utz, G., 2001. IL-10 and its homologs: important immune mediators and emerging immunotherapeutic targets. Trends Immunol. 22, 414–417. Weidner, W., Krause, W., Ludwig, M., 1999. Relevance of male accessory gland infection for subsequent fertility with special focus on prostatitis. Hum. Reprod. Update 5, 421–432. World Health Organisation, 1999. WHO Laboratory Manual for the Examination of Human Semen and Sperm–Cervical Mucus Interaction. Cambridge University Press, Cambridge.