Department of Medical Microbiology and Immunology, Bartholin Building, University of Aarhus, DK-2000 Aarhus, Denmark
Once again the detection and clinical significance of antisperm antibodies (ASA) is being discussed (Helmerhorst et al., 1999). Although it is cautiously concluded that `there is reason to accept antibody-mediated antisperm immunity as a cause of subfertility', the authors list three reasons why the clinicians are not inclined to test for ASA in subfertile couples: (i) lack of a standardized and universally accepted assay for ASA; (ii) lack of consensus on the clinical consequences of ASA; and (iii) absence of evidence for a mechanistic explanation on how ASA impair conception. These three claims each deserve a comment.
The discussion by Helmerhorst et al. (1999) is based strictly on modern immunological concepts, leading to the recommendation of immunoblotting and affinity chromatography as promising in this field. For those who have been in the field for many years, these recommendations may provoke a feeling of deja vu. This was the common strategy ~20 years ago, and many studies have since been made by enzyme-linked immunosorbent assays (ELISA), using spermatozoa or antigen fractions, and by immunoblotting (Lehmann et al., 1985; Mettler et al., 1985
). However, the results have generally been disappointing; the clarity one had hoped for has not been achieved. There are no good explanations for the lack of success, but one aspect of the problem may be that conventional immunological thinking is not enough to achieve progress in reproductive immunology. The physiology of reproduction must also be taken into consideration. A look at the possible in-vivo effects of ASA may illustrate this.
Mechanistic explanations for impairment of conception by ASA?
The claim by Helmerhorst et al. of `an absence of evidence for a mechanistic explanation on how ASA impair conception' is rather surprising, particularly in a Dutch paper, even a paper from Groningen. It was in Groningen that the outstanding group, headed by Jan Kremer, demonstrated that the migration of spermatozoa with ASA of the immunoglobulin (Ig) A class on their surface through cervical mucus was strongly impaired, apparently because the Fc part of the IgA molecule (but not of the IgG molecule) binds effectively to the micelles of the cervical mucus (Jager et al., 1980). In the spermcervical mucus contact test, the percentage of spermatozoa covered with IgA antibodies was proportional to the percentage of spermatozoa revealing the so-called `shaking' phenomenon, i.e. heavy tail movements without forward progression in the cervical mucus. Further evidence for the significance of IgA was found when treatment of IgA-covered spermatozoa with an IgA1 protease was shown to improve sperm migration through cervical mucus (Bronson et al., 1987
).
This means that IgA antibodies to any antigen expressed on the sperm membrane will impair penetration and that, in this situation, the immunoglobulin class of the ASA becomes more important than the antigen specificity. This is in contrast to conventional immunological reactions which may be the reason why the shaking phenomenon is often neglected in immunological discussions. However, it is obviously a very effective mechanism; in experiments on the migration of IgA-covered spermatozoa into cervical mucus in capillary tubes, the spermatozoa from patients with strong immune responses often reached only a few mm into the mucus. The binding of the Fc piece of IgA to the cervical mucus should be seen in an evolutionary perspective. With cervical mucus as the only barrier between the environment and the female genital tract and peritoneal cavity, there must be defence mechanisms in the mucus to prevent micro-organisms from entering the genital tract. IgA antibodies to micro-organisms serve this function by trapping the micro-organisms. In rare cases when a woman has produced IgA antibodies to spermatozoa or, more commonly, when spermatozoa entering the mucus are already covered with IgA, the result will be the same: trapping of the spermatozoa or impairment of migration.
A more conventional immunological mechanism by which ASA might reduce fertility would be through binding to the receptor by which spermatozoa attach to the egg, thereby blocking spermovum interaction. In this situation, the antigen specificity of the ASA will be crucial, and since there is more than one autoantigen in the sperm membrane, not all ASA might have this effect. Over the last two decades, this phenomenon has been extensively studied in connection with in-vitro fertilization (IVF) using spermatozoa from men with ASA. In nearly all studies, a significant reduction in the percentage of fertilized eggs was observed. One of the early studies also indicated that the blocking effect was mainly caused by IgA antibodies (Clarke et al., 1985), but this has not been noted in subsequent studies, and absorption experiments have shown that an inhibitory effect can also be caused by IgG antibodies (Bronson et al., 1982
). A recent study on the binding of spermatozoa to the zona pellucida (Francavilla et al., 1997
) is particularly illustrative. In each experiment, they added equal numbers of spermatozoa from a fertile person and from a patient with strong reactivity in the immunobead test, so that the two kinds of spermatozoa, labelled with different fluorochromes, were allowed to compete for binding to the zona pellucida. Compared with the spermatozoa from the fertile men, spermatozoa from patients with normal semen samples (apart from the presence of ASA) were inhibited in ~50% of cases, enhanced in a few cases and not affected in the remaining cases, fitting with the concept that antibody specificity must play a role. However, in spite of the patients' strong immune responses with high titres of circulating ASA and, in most cases with both IgG and IgA, on nearly all spermatozoa, the inhibition was never complete but some of the antibody-covered spermatozoa were bound to the zona pellucida. This is in agreement with the experience from many other IVF studies that, even though the percentage of fertilized eggs may be reduced by the presence of ASA, some eggs usually will have been fertilized.
The conclusion from this discussion on the possible fertility-reducing effects of ASA is that impairment of sperm migration through cervical mucus by IgA antibodies is a very effective mechanism, while the blocking of spermovum interaction (caused by both IgG and IgA ASA) seems to be less effective, probably rarely preventing but maybe delaying conception. Clinical experiences support these conclusions. In infertility clinics, men with ASA rarely have an isolated IgG response but usually present a mixture of IgG and IgA antibodies. On the other hand, in fertility studies of vasectomized and subsequently vasovasostomized men, immune responses restricted to the IgG class were not rare but these men appeared to have normal fertility whereas the fertility rate was inversely correlated with the IgA response (Meinertz et al., 1990). Therefore it seems that `absence of evidence for a mechanistic explanation' should not keep the clinician from testing patients (at least men) with fertility problems for ASA.
Lack of a standardized and universally accepted assay for ASA?
Although no arguments against this statement by Helmerhorst et al. (1999) can be made, the question is whether it is a fair and necessary requirement for an assay to be used? If so, the broad spectrum of laboratory tests might be dramatically reduced. The essential question seems to be whether certain assays for ASA produce clinically relevant results.
For many years, sperm agglutination tests (either macroscopic gelatin agglutination or microscopic tray agglutination) were the common assays (Rose et al., 1976). More recently, these assays have been discredited and often described as unreliable, particularly in connection with description of new techniques. True, none of them are ideal, both requiring fresh semen samples of optimal quality. In the gelatin agglutination test, large volumes of semen are used, and in the tray agglutination test, experience and caution is neeeded in the reading of the results in order to distinguish true agglutinates from non-specific clumping. Nevertheless, these tests have given clinically relevant results in males. Several studies have shown that high titres in serum or free antibodies in seminal plasma very rarely occur except in patients with a fertility problem (in ~10% of men from couples with unexplained fertility) (e.g. Bronson et al., 1985), and among normospermic men in a follow-up study (Rümke et al., 1974
), increasing serum titres in gelatin agglutination were correlated with decreasing fertility (e.g. only 12.5% of 64 men with titres from 128512 and none of 11 with titres
1024 had induced conception during the observation period of 216 years). Since the ASA detected in serum are mainly of the IgG class, it may seem strange that serum titres are related to subfertility if the anti-fertility effect of ASA is mainly caused by IgA antibodies in semen. The explanation is that IgA antibodies very rarely occur without IgG antibodies, and high IgA concentrations are mainly found in patients who have also high IgG concentrations.
An equally important observation in these studies was that low concentrations of antibodies (titres of 416) were equally common among fertile controls and infertile patients, indicating that such low values apparently did not affect fertility in any significant way.
More recently, the indirect immunobead test has become the most popular (Bronson et al., 1982). Fresh spermatozoa, harvested by swim-up technique, are incubated with serum or seminal plasma, then carefully washed, and immunobeads covered with anti-IgG or anti-IgA (or anti-IgM) are added. Under the microscope, immunobeads are seen adhering to the motile spermatozoa if the sample contains ASA of the proper immunoglobulin class. The use of an anti-immunoglobulin guarantees the specificity of the reaction; it is a very sensitive test, determines the immunoglobulin class of the ASA, and requires only small amounts of spermatozoa, thus making it, apparently, the ideal assay. Nevertheless, there is a problem. In order to keep it practicable, it is usually carried out with only one dilution of serum, usually a low dilution (most commonly 1:4). In such cases, the indirect immunobead test works as a very sensitive screening test, so that even the low antibody levels of no significance for fertility (which is probably the major part of the positive reactions) may induce strong reactivities, at least for IgG (e.g. see Bronson et al., 1985). This may explain why several investigators have questioned the value of testing for ASA and the significance of ASA as a cause of infertility. If the test was carried out with higher dilutions of serum (e.g. 1:50 and 1:250), it might provide better clinical guidance.
The most rational way to test for ASA in males is obviously to determine immunoglubulins (IgG and IgA) on the surface of the patients spermatozoa by means of the direct mixed agglutination reaction (MAR) or immunobead tests. Both tests are quick, easy, and sensitive, but require motile spermatozoa to ensure that the erythrocytes or immunobeads adhere to the sperm membrane. They are excellent screening tests; if the reaction is negative or weakly positive (50% of the motile spermatozoa carrying immunoglobulins) the patient can hardly have an infertility problem caused by ASA and no further immunological tests are needed. On the other hand, if all, or the majority, of spermatozoa have immunoglobulins on their membrane, the consequences for fertility are more difficult to evaluate, first of all because these sensitive tests cannot distinguish between spermatozoa with only minute amounts of immunoglobulin, and spermatozoa which are fully saturated with antibody on all available antigen molecules. Both for the IgA-induced impairment of sperm migration through cervical mucus and for a possible block of spermovum interaction, the number of antibody molecules must be assumed to play a role. In this situation, determination of ASA in serum and seminal plasma may be useful. With high titres in serum or an excess of free ASA in seminal plasma, it would seem likely that the amount of ASA on the spermatozoa would also be high.
That these theoretical considerations are in agreement with the clinical reality was illustrated in the most clearcut way in our follow-up study on vasovasostomized men, mentioned above (Meinertz et al., 1990).
Within the observation period (22111 months, median 44 months), 10 of the 15 men (66.7%) with a negative direct MAR (10% of the motile spermatozoa with adhering erythrocytes) had induced conception but so had 42 of the 79 men (53.2%) with a positive MAR (Fisher's two tailed exact test: P = 0.4995). Thus, a mere positive MAR (>10% for IgG or IgA) gave no information on fertility status. Most of the men with ASA had both IgG and IgA on their spermatozoa but some had only IgG. Pure IgA reactions were not recorded. Looking at the men with immune responses restricted to the IgG class, the conception rate for the total group was 85.7% (18/21), and for those with IgG on all spermatozoa it was equally high (11/13 = 84.6%). Thus, no anti-fertility effect could be recorded for IgG; in fact the conception rate was remarkably high, even higher than for patients without ASA (although the difference was not statistically significant). On the other hand, the men with both IgG and IgA reactivity revealed decreasing fertility with increasing IgA reactivity. For the total group the conception rate was 42.9% (24/56) and for those with IgA on all spermatozoa 21.7% (five out of 23). Compared with the groups with only IgG reactivity, the differences were highly significant. In an attempt to identify the men with particularly strong immune responses, it appeared that none of 10 men with 100% IgA reactivity in the MAR and circulating sperm agglutinins in titres of 256 or greater had induced pregnancy and only one of 13 men with 100% IgA MAR and free sperm agglutinins in seminal plasma had succeeded.
Although at present there is no single assay which can evaluate the fertility status of the individual patient in a clearcut way with regard to immune subfertility or infertility, it seems justified to conclude that, with the proper combination of tests (e.g. MAR or immunobead tests combined with sperm agglutinins in serum or seminal plasma), we can come rather close to this goal. A recent survey of the diagnosis and management of ASA, perfomed by reproductive medicine centres throughout the UK (Krapez et al., 1998), showed that the great majority of centres do in fact follow this strategy. Thus, 44 out of 48 centres tested for ASA on spermatozoa by direct MAR (28 laboratories) or immunobead tests (19 laboratories). In some cases, seminal plasma was also tested and 17 laboratories examined serum from male partners by various tests, mainly the tray agglutination test. This means that nearly all centres identified in an easy way, the men who might have an immunological cause for their subfertility. Although this is the best that can be done at present, it is not optimal. What is needed is a technique that can identify in a more precise way the men who do have an imunological problem. At present, flow cytometry seems a promising technique which may be able to determine the exact amount of IgG and IgA on the individual spermatozoa (Räsanen et al., 1992
; Ke et al., 1995
).
Lack of consensus on clinical consequences of ASA?
At first sight this claim by Helmerhorst et al. may seem correct, but the question is whether investigations with many different techniques and criteria can be compared? Years ago it was generally believed that ASA could cause infertility, but nowadays the picture seems much less clear. There may be several reasons for this change in the view.
Firstly, we have learned that even with the highest levels of ASA (particularly IgA) conception can hardly be excluded. Therefore, the question is rather to what extent ASA may cause subfertility. The studies on male infertility, discussed in this review have mainly been based on various conventional (some might say old-fashioned) techniques. However, even though consensus may be too strong a word to use, the results of such studies are generally in good agreement,indicating that low levels of ASA do little harm, somewhat higher values (particularly IgA) cause subfertility increasing to something close to infertility for very high levels of ASA; a conclusion which obviously presents a serious obstacle for the attempts to produce an anti-fertility vaccine, based on sperm antigens!
Secondly, when new assays for ASA are being introduced, they should be compared with the conventional assays, and if they do not determine the same kind of ASA, one should keep this in mind. When a World Health Organization (WHO) multi-centre study on antibodies to reproductive antigens was performed 15 years ago, it turned out that several of the techniques, including most of the ELISAs, immunoblotting techniques, passive haemagglutination, and cytotoxicity tests, did not detect the ASA recorded in conventional techniques. Furthermore, the results usually had little clinical relevance, showing no significant difference between fertile and infertile groups (Hjort et al., 1985). Unfortunately, this is often forgotten in discussions on ASA.
Thirdly, the quantitative aspect and the immunoglubulin class of the antibodies should be included in the testing. Thus, all patients with reactivity in direct MAR or immunobead test should not be considered as one group, but those with weak reactivity (probably fertile) should be distinguished from those with antibody on nearly all spermatozoa (possibly subfertile), and similarly those with only IgG (probably fertile) should not be mixed up with those with strong reactivities for both IgG and IgA (probably subfertile or infertile).
The recent survey of reproductive medicine centres in the UK (Krapez et al., 1998) indicated that, usually, a distinction between IgG and IgA ASA is not made. This is probably due to the fact that most infertile men with ASA will have a mixture of the two immunoglobulin classes on their spermatozoa. However, the attempts to treat the patients agree well with the conclusions reached from laboratory investigations and `clinical experiments' (studies on vasovasostomized men). Thus, intrauterine insemination is used in most centres to overcome the IgA-induced impairment of sperm penetration through the cervical mucus. Either IVF or intracytoplasmic sperm injection (ICSI) was used for patients with a strong immunity to spermatozoa (>50% reactivity in MAR or immunobead test), but generally with liberal indications in comparison with the conclusions reached above. This obviously reflects the fact that ASA are often not the only cause of infertility in these patients, but low sperm counts, poor motility or problems in the female partner may also play a role. Thus, the anti-fertility effects of ASA are difficult to analyse in groups of infertile patients.
Vasectomized and subsequently vasovasostomized men seem to offer a much better group; in fact a group very similar to experimental animals. Both partners have usually previously proven their fertility so that genetic factors for infertility can be excluded. A group of men who have been vasectomized for a limited period and who have had successful vasovasostomy with an acceptable number of spermatozoa showing good motility in the semen sample, can be selected. Thus if both partners have been healthy since their last conception, infertility factors other than possibly ASA are much less likely than in a group of infertile patients. The big advantage in the vasovasostomy group is indeed that their fertility status in not known from the beginning, and a control group is, therefore, automatically included; i.e. some have no ASA, some have only IgG and some have both IgG and IgA. This seems an ideal group for studying the effects of ASA and for evaluating new techniques, and it is surprising that it has been so scarcely used. The crucial question in this connection is, of course, whether vasectomized men produce antibodies to all the same antigens against which infertile men may react. Further studies are required, but so far we have no evidence that this should not be the case.
This leads to the conclusion that sperm immunology can provide useful information for the clinicians. It may not be so essential now, when the immunological factor together with many other infertility factors can be overcome by assisted reproduction techniques. However, it is still a good principle to reach as exact a diagnosis as possible before a treament is being instituted.
Notes
This debate was previously published on Webtrack 80, August 4, 1999
References
Bronson, R., Cooper, G., Hjort, T. et al. (1985) Anti-sperm antibodies detected by agglutination, immobilization, microcytotoxicity and immunobead-binding assays. J. Reprod. Immunol., 8, 279299.[ISI][Medline]
Bronson, R.A., Cooper, G.W. and Rosenfeld, D.L. (1982) Spermspecific isoantibodies and autoantibodies inhibit the binding of human sperm to the zona pellucida. Fertil. Steril., 38, 724729.[ISI][Medline]
Bronson, R.A., Cooper, G.W., Rosenfeld, D.L. et al. (1987) The effect of an IgA1 protease on immunoglobulins bound to the sperm surface and sperm cervical mucus penetrating ability. Fertil. Steril., 47, 985991.[ISI][Medline]
Clarke, G.N., Lopata, A., McBain, J.C. et al. (1985) Effect of sperm antibodies in males on human in vitro fertilization (IVF). Am. J. Reprod. Immunol. Microbiol., 8, 6266.[Medline]
Francavilla, F., Romano, R., Santucci, R. et al. (1997) Occurrence of the interference of sperm-associated antibodies on sperm fertilizing ability as evaluated by the spermzona pellucida binding test and by the TEST-yolk buffer enhanced sperm penetration assay. Am. J. Reprod. Immunol., 37, 267274.[ISI][Medline]
Helmerhorst, F.M., Finken, M.J.J. and Erwich, J.J. (1999) Detection assays for antisperm antibodies: What do they test? Hum. Reprod., 14, 16691671.
Hjort, T, Johnson, P.M. and Mori, T. (1985) An overview of the WHO international multi-centre study on antibodies to reproductive tract antigens in clinically defined sera. J. Reprod. Immunol., 8, 359362.[ISI][Medline]
Jager, S., Kremer, J., Kuiken, J. and van Slochteren-Draaisma, T. (1980) Immunoglobulin class of antispermatozoal antibodies from infertile men and inhibition of in vitro sperm penetration into cervical mucus. Int. J. Androl. 3, 114.[ISI][Medline]
Ke, R.W., Dockter, M.E., Majumdar, G. et al. (1995) Flow cytometry provides rapid and highly accurate detection of antisperm antibodies. Fertil. Steril. 63, 902906.[ISI][Medline]
Krapez, J.A., Hayden, C.J., Rutherford, A.J. and Baalen, A.H. (1998) Survey of the diagnosis and management of antisperm antibodies. Hum. Reprod., 13, 33633367.[Abstract]
Lehmann, D., Temminck, B., Da Rugna, D. et al. (1985) Blot-immunobinding test for the detection of anti-sperm antibodies. J. Reprod. Immunol., 8, 329336.[ISI][Medline]
Meinertz, H., Linnet, L., Fogh-Andersen, P. and Hjort, T. (1990) Antisperm antibodies and fertility after vasovasostomy: a follow-up study of 216 men. Fertil. Steril., 54, 315321.[ISI][Medline]
Mettler, L., Czuppon, A.B., Alexander, N. et al. (1985) Antibodies to spermatozoa and seminal plasma antigens detected by various enzyme-linked immunosorbent (ELISA) assays. J. Reprod. Immunol., 8, 301312.[ISI][Medline]
Räsanen, M.L., Hovatta, O.L., Penttila, I.M. and Agrawal, Y.P. (1992) Detection and quantitation of sperm-bound antibodies by flow cytometry of human semen. J. Androl. 13, 5564.
Rose, N.R., Hjort, T., Rümke, Ph. et al. (1976) Techniques for detection of iso- and auto-antibodies to human spermatozoa. Clin. Exp. Immunol., 23, 175199.[ISI]
Rümke, Ph., van Amstel, N., Messer, E.N. and Bezemer, P.D. (1974) Prognosis of fertility of men with sperm agglutinins in the serum. Fertil. Steril., 25, 393397.[ISI][Medline]