- Email: [email protected]
Detection of antisperm antibodies in seminal plasma by flow cytometry: comparison with the indirect immunobead binding test
FERTILITYAND STERILITY@ Copyright
a 1997 American
Published
by Elsevier
Society for Reproductive
Vol. 68, No. 6, December Medicine
1997
Printed on acid-free paper in U. S. A.
Science Inc.
Detection of antisperm antibodies in seminal plasma by flow
cytometry: comparison with the indirect immbnobead binding test
Simon C. Nicholson, M.D. Julian N. Robinson, M.D.? Ian L. Sargent, Ph.D. David H. Barlow, M.D. Nuffield Department
of Obstetrics and Gynaecology,
John Radcliffe Maternity Hospital,
Oxford, United Kingdom
Objective: To compare flow cytometry with the established indirect immunobead binding test (IBT) for the detection of antisperm antibodies in seminal plasma. Design: A prospective, comparative study. Setting: University-based andrology unit. Patient(s): One hundred and fifty-eight men with suspected male factor subfertility. Intervention(s): Seminal plasma samples were incubated with antisperm antibody-negative donor sperm. Surface-bound antibody was detected with fluorescence-labeled antihuman antibody in the flow cytometry assay or with immunobead-labeled antihuman antibody in the IBT. Main Outcome Measure(s): The percentage of sperm that tested positive for surface-bound antibody was determined in the two assays. Seminal plasma was antisperm antibody-positive when ~20% of the sperm were antibody-bound, and clinically significant levels were present when 250% of the sperm were antibody-bound. Result(s): Of 71 samples that were negative by the IMT, 66 (93%) also were negative by flow cytometry. Of 63 samples that had 250% immunobead binding, 55 had equivalent results by flow cytometry. Overall statistical analysis showed a good correlation between the two assays. Conclusion(s): There is a good correlation between the indirect IBT and indirect flow cytometry for the detection of antisperm antibodies in seminal plasma. (Fertil Steril@ 1997;68:1114-9. 0 1997 by American Society for Reproductive Medicine.) Key Words: Antisperm antibody, immunobead binding test, flow cytometry, seminal plasma
Antisperm antibodies have a variety of effects on human fertility, including impairment of sperm motility, trapping and phagocytosis of sperm in cervical mucus, and impairment of sperm-oocyte interaction (1). The most commonly used assays to detect antisperm antibodies are the immunobead binding test (IBT) (2) and the mixed antiglobulin reaction (MAR) (3). They can be performed as direct assays, de-
ReceivedMarch 24, 1997; revisedand acceptedJuly 21, 1997. The Jones Institute for Reproductive Medicine, Norfolk, Virginia. Reprint requests and present address: Simon C. Nicholson, M.D., Department of Obstetrics and Gynaecology, Simpson Memorial Maternity Pavilion, Edinburgh, EH3 9YW, United Kingdom (FAX: 44-131-536-7474).
1114
tecting antibodies already bound to the surface of the patient’s sperm, or as indirect assays, in which donor sperm are used to detect the presence of unbound antisperm antibody in seminal plasma, serum, or cervical mucus. The direct assays necessitate the presence of adequate numbers of motile sperm and are not suitable for men with poor sperm parameters (4). Both the IBT and the MAR require that the investigator assess the percentage of motile sperm that are positive for antisperm antibodies by counting a minimum of 100 sperm using light microscopy. This can be both time-consuming and subjective. Antisperm antibody detection by flow cytometry first was described by Haas and Cunningham (5). The principle of the assay is that sperm with bound
0015-0282/97/$17.00 PI1 SOO15-0282(97)00374-9
surface antibody are incubated with a fluorescencelabeled antihuman antibody. The proportion of antibody-positive sperm then is identified by passing thousands of sperm through a fluorescence-activated cell sorter, which is a highly objective assessor. In addition, the number of antibody molecules bound to the surface of the sperm can be calculated using fluorescent calibration beads. This can be of use in assessing methods of reducing antisperm antibodies both in vitro and in vivo (6). A close correlation between flow cytometry and the more established MAR has been described for both direct and indirect versions of the assays (7). Comparisons of flow cytometry with the IBT have produced more variable results using the indirect testing of male serum (8, 9). The detection of antisperm antibodies in seminal plasma by flow cytometry and other methods has been reported recently and suggests a good correlation between indirect flow cytometry and the indirect IBT on seminal plasma in a small series of positive samples (10). In this study, we examined the validity of indirect seminal plasma antisperm antibody assessment by flow cytometry in a much larger series. MATERIALS
AND METHODS
Seminal Plasma Samples
Seminal plasma samples were obtained from men attending the Oxford Andrology Clinic for evaluation of suspected male factor subfertility. A minimum of 200 /JL of seminal fluid was centrifuged at 1,000 x g for 15 minutes in a 1.5mL capped Eppendorf tube (Sigma, Poole, Dorset, United Kingdom). The supernatant then was transferred to a clean tube, stoppered, and heat-inactivated in a water bath at 56°C for 30 minutes to destroy complement. The cell-free seminal plasma then was stored at -20°C until it was used in the indirect IBT or flow cytometry assay. Indirect Immunobead
Binding Test
A semen sample was obtained from an antisperm antibody-negative semen donor and prepared by the swim-up method to obtain a washed sperm population containing 50 X lO?mL of motile sperm (11). Ten-microliter volumes of seminal plasma each were diluted in 40 PL of phosphate-buffered saline (PBS) supplemented with 5% (50 g/L) bovine serum albumin (BSA) in Eppendorf tubes, and 50 PL of the sperm suspension then was added to each tube and gently mixed. The samples were incubated at 37°C for 60 minutes and then were centrifuged at 500 x g for 5 minutes. The supernatants were discarded and
Vol. 68, No. 6, December
1997
the sperm pellets were resuspended gently in 500 PL of PBS supplemented with 0.4% BSA. The tubes again were centrifuged at 500 X g for 5 minutes. The supernatants were discarded and the pellets were resuspended gently in 50 I.ILof fresh PBS plus 5% BSA. On separate clean microscope slides, 5 PL of the washed sperm suspension was mixed with 5 PL of the GAM immunobeads (Biorad, Richmond, CA), containing a mixture of immunobead-bound antihuman immunoglobulin (Ig) G, A, and M. The slides were incubated at ambient temperature in a humidified chamber for 10 minutes and then were examined at x400 magnification by phase-contrast microscopy. At least 200 motile sperm were counted, differentiating between those with and those without immunobeads attached. The localization of any bound beads also was determined (i.e., head, midpiece, tail, tail tip). The percentage of motile sperm with GAM immunobeads attached was calculated. The test result was regarded as positive when 220% of motile sperm had beads attached and as clinically significant when ~50% were coated with the beads (12). Binding restricted to the tail tip is not considered to be clinically significant (13) and was discounted in this study. Known antisperm antibody-positive and antibody-negative control samples always were included in each run of the assay. Flow Cytometry
The seminal plasma samples were thawed at room temperature and vortexed to ensure thorough mixing. A 5O+L aliquot of each seminal plasma sample was added to 40 PL of PBS supplemented with 5% goat serum albumin. Ten microliters of a sperm suspension, prepared by the swim-up method from antibody-negative donor semen, containing 125,000 motile sperm was added to each sample. Known antisperm antibody-positive and -negative seminal plasma samples were included as controls. A blank sample was included in which sperm were processed in the absence of seminal plasma. After incubation at 37°C in a 5% CO2 incubator for 1 hour, the sperm were washed twice to remove unbound antibody. One milliliter of PBS was added and the mixture was agitated. The tubes then were centrifuged at 500 x g for 5 minutes and the supernatants were discarded. The sperm pellet was resuspended with 1 mL of PBS and the wash was repeated. After centrifugation, the pellet was resuspended with 50 PL of a 1:5 solution of fluorescein isothiocyanate conjugate (FIT(J)-labeled goat anti-
Nicholson
et al.
Antisperm antibodies by flow cytometry
1115
human immunoglobulin directed against IgA, IgG, and IgM (Southern Biotechnology Associates, Inc., Birmingham, AL) and was incubated for a further 1 hour at 4°C in the dark. Unbound antibody was removed by washing twice with PBS as described earlier and the sperm were analyzed by flow cytometry to assess the number with bound antibody using a Coulter Epics Elite Flow Cytometer (Coulter Electronics, Luton, United Kingdom). The sperm population was gated using go-degree and forward-angle light scatter to exclude debris and aggregates. A total of 5,000 sperm were analyzed from each sample, recording single-parameter histograms of the log FITC intensity (5). Digital subtraction of the blank sample from each test sample using the Immunod Elite software program (Coulter Electronics) was performed to determine the percentage of sperm in each sample that were positive for antisperm antibody. A result of <20% of the sperm coated by antibody was regarded as negative and ~20% as positive. All flow cytometry assay conditions were chosen after preliminary experiments to determine the optimum separation of antisperm antibody-negative and -positive sperm populations. Comparison of Indirect Immunobead Indirect Flow Cytometry Assays
Binding and
The two assays were compared by determining the strength of agreement at <20%, 20%-49%, and 250% of the sperm with surface-bound antibody. Seventy-one consecutive antisperm antibody-negative seminal plasma samples (<20% immunobead binding) and 87 consecutive positive seminal plasma samples (~20% immunobead binding) were analyzed for the presence of antisperm antibodies by both flow cytometry and IBT. All seminal plasma samples were tested in duplicate and the mean result (+SD) was taken. Statistical analysis to compare this strength of agreement was performed using the K (14). The K is a chance corrected proportional agreement that assesses the agreement between two sets of values in excess of the agreement that would be expected by chance. The maximum value of the K is 1.00. This is the situation when there is total agreement between the two sets of measured variables. A K of 0 indicates no better agreement than would be expected by chance and a value of <0 indicates worse agreement than would be anticipated by chance. For K values between 0 and 1.00, the measure of agreement is as follows: a K value 520 indicates poor strength of agreement, 0.21-0.40 fair agreement, 0.41-0.60 moderate agreement, 0.61-0.80 good
1116
Nicholson
et al.
Antisperm
antibodies
by flow cytometry
I
Fluorescence intensity / log (Fluoresceinisothiocyanate)
Figure 1 Single-parameter histograms of FITC-labeled goat antihuman Ig binding to sperm after initial incubation with buffer (blank), antisperm antibody (A&B&negative seminal plasma, and antisperm antibody-positive seminal plasma.
agreement, and 0.81-1.00 very good strength of agreement between the two assays (15). RESULTS
Seminal plasma samples were tested first for the presence of antisperm antibodies by the IBT. In total, 158 seminal plasma samples were analyzed. Seventy-one were antisperm antibody-negative (~20% immunobead binding), 63 were strongly positive (~50% immunobead binding), and 24 were intermediate (20%-49% immunobead binding). The samples then were tested for the presence of antisperm antibodies by flow cytometry. A histogram of the log FITC intensity after incubation of sperm with antibody-positive and antibody-negative seminal plasma and after incubation with buffer alone is shown in Figure 1. The histogram after incubation with antisperm antibody-negative seminal plasma overlies that for the sperm population incubated in the absence of seminal plasma. Of the 71 samples that were negative by the indirect IBT, 66 (93%) also were negative by flow cytometry (i.e., <20% of the sperm were positive for antibody). Four (6%) of the seminal plasma samples that were negative by the IBT produced intermediate results by the flow cytometry assay (20%-49% of the sperm were bound with antibody) and 1 (1%) produced strongly positive results (i.e., >50% of the sperm were bound with antibody). Sixty-three seminal plasma samples had ~50% immunobead binding. Of these, 55 (87%) had ~50% binding in the flow cytometry assay, 7 (11%) had 20%-49% binding, and 1 (2%) had <20% binding. Of the 24 samples that produced intermediate results by the IBT, 19 (79%) produced similar results
Fertility
and Sterility@
. . . T
20
0
40
60
80
Ikl
Resultsof indirectIBT (percentof sperm positivefor antibody) -I
L
Figure 2 Correlation between indirect immunobead binding test (IBT) and flow cytometry results for the presence of seminal plasma antisperm antibodies (r = 0.924; P < 0.001).
by flow cytometry and 5 (21%) produced negative results. None of these intermediate samples produced a ~50% result with flow cytometry. The comparison between the two assays is demonstrated in Figure 2 and the data are summarized in Table 1. The mean percentage of binding (+SD) was 74% + 28% for the immunobead-positive results and 70% t 29% for the flow cytometry-positive samples. For the negative samples, the mean percentage of binding (*SD) for IBT was 6% + 6% and that for flow cytometry was 8% ? 4%. Statistical analysis of the strength of agreement between the two assays gave a K of 0.82, demonstrating a very good correlation between the indirect flow cytometry assay and the indirect IBT. With a 20% cutoff for a positive antisperm antibody result, the sensitivity of the new assay was 0.94 and the specificity was 0.92. DISCUSSION
This study denmnstrates a good correlation between an indirect flow cytometry assay and the indirect IBT for the detection of antisperm antibodies in seminal plasma. Reliable assays are necessary to direct the appropriate clinical interventions should antisperm antibodies be indicated as a possible cause of infertility. Intermediate-dose cyclical steroid therapy for the male partner with immune subfertility has been demonstrated to produce a significant improvement in the spontaneous pregnancy rate compared with placebo (16). Assisted methods of conception, including IUI with washed sperm (171,
Vol. 68, No. 6, December
1997
GIFT (18), and IVF-ET (19), also have been shown to be of benefit in couples with immune subfertility. Ke et al. (8) previously have demonstrated an equally good correlation between flow cytometry and the IBT for the detection of antisperm antibodies in male serum. They tested 37 men who had undergone vasectomy and 35 fertile control men. Flow cytometry detected all 22 of the IBT IgG-positive samples and 12 of the 14 IgA-positive samples. All the IBTnegative samples were identified correctly as negative by flow cytometry. In an earlier study, Haas et al. (9) reported a specificity of 100% but a sensitivity of only 77% when they compared these two assays for antibody detection in serum. However, in their study, dead cells were not excluded from the analysis. Dead cells can produce false-negative or false-positive results because of denaturation of sperm antigens or exposure of internal antigens (20). In our study, dead sperm were excluded by the swim-up preparation and electronic gating. In the MAR and IBT, nonspecific binding to dead sperm is excluded by only counting viable, motile sperm. Indirect assays for antisperm antibody detection are necessary for the testing of serum, cervical mucus, and seminal plasma when sperm parameters are very poor. An indirect assay also obviates the need for same-day analysis because seminal plasma samples can be frozen and processed at a later date, with other samples, more efficiently. However, in men with good sperm parameters, there is debate as to whether a direct or an indirect assay is more appropriate. Direct assays detect antibodies already bound to the surface of sperm. It has been suggested that antibodies that remain within seminal plasma may not be reflective of the Igs that are bound to the sperm cell surface. In a study by Bronson et al. (21) in which direct and indirect IBTs were compared in 36 antisperm antibody-positive men, a good correlation between the two assays was obtained. All 36 men had positive results with the direct assay and 33 had positive
Table 1 Comparison of the Number of Samples Observed With Varying Percentages of the Sample Showing Positive Sperm Binding for Flow Cytometry and IBT Results of 158 Seminal Plasma Samples Flow cytometry IBT results
Nicholson
20%-49%
250%
7 19 4
55 0 1
250% 20%-49% <20%
et al.
Antisperm
results
arztibodies
by flow cytometry
<20% 1 5 66
1117
results with the indirect assay. Of the original 36 men, 29 had head-directed antisperm antibodies. The indirect assay failed to detect the head-directed antibodies in 18 of these 29 men. Hellstrom et al. (22) assessed interdonor variability by testing a single antisperm antibody-positive serum sample against 27 different donors using the IBT. The mean total percentage of binding was 92%, with a range of 62%- 100% for all donors. The localization of binding also varied. Analysis of variance showed there to be a statistically significant difference between the donors. However, the positive serum tested was confirmed to be positive consistently, at >50% in all 27 donors. Rasanen et al. (10) recently carried out a study comparing direct and indirect assays. They showed a good correlation between the direct MAR and a direct flow cytometry assay, although the percentage of sperm that tested positive with flow cytometry tended to be less than with the MAR. However, of the 11 semen samples that were positive for IgG antisperm antibodies by the direct MAR, 9 also were positive by the indirect MAR, 8 by the indirect IBT, and only 5 by the indirect flow cytometry. The results for IgA were better, with both the indirect MAR and the indirect IBT detecting 10 of 11 direct MAR-positive samples and indirect flow cytometry detecting all the direct MAR IgA-positive samples. When they compared the indirect IBT results with the indirect flow cytometry results of the same 11 men for the presence of IgG, they demonstrated a good correlation, as in our study. These investigators suggested that the large discrepancy in their study between the results of direct and indirect assays may have occurred because tailtip binding of antibody is detected less well by flow cytometry than is head-directed binding. The most frequent type of binding in their IBT- and MARpositive samples was tail-tip. In our analysis, tailtip binding alone was not regarded as a positive result. In assessing the correlation between the two assays, we compared their ability to discriminate between positive results (~-20% of sperm antibodybound) and negative results (<20% of sperm antibody-bound) and to determine results of clinical significance (~50% of sperm antibody-bound) (12). These factors are important for a clinical assay and may have contributed to the close correlation between the assays. Flow cytometry assesses antibody binding on very large numbers of sperm and provides a highly objective assessment. Quantitative information on the number of antibody molecules per sperm can be ob-
1118
Nicholson
et al.
Antisperm antibodies by flow cytometry
tained, which may be of use in the evaluation of antibody-lowering interventions. By performing an indirect assay, seminal plasma samples can be stored frozen and tested as weekly batches. This is of importance in flow cytometry because in most centers, the flow cytometer is located in a different department and access to the machine often is limited. In practice, it would not be possible to process sporadic sperm samples because patients attend a clinic for evaluation and most andrology centers do not possess their own machine. In summary, this study demonstrates a good correlation between the indirect IBT and indirect flow cytometry for the detection of antisperm antibodies in seminal plasma, with high specificity and sensitivity, providing an objective and quantitative method of antibody detection. However, flow cytometry does require the availability of expensive equipment and a skilled operator. Therefore, for most andrology units, assays such as the IBT are likely to continue to have an important clinical role in assessing subfertile couples and flow cytometry will remain a useful research tool at present. REFERENCES 1. Hendry WF. The significance of antisperm antibodies: measurement and management. Clin Endocrinol 1992;36:21921. 2. Bronson R, Cooper G, Hjort T, Ing R, Jones WR, Wang SX, et al. Anti-sperm antibodies, detected by agglutination, immobilization, microcytotoxicity and immunobead-binding assays. J Reprod Immunol 1985;8:279-99. 3. Jager S, Kremer J, Van Slochteren-Draaisma T. A simple method for screening for antisperm antibodies in the human male. Int J Fertil 1978;23:12-8. 4. Meinertz H, Bronson R. Detection of antisperm antibodies on the surface of motile spermatozoa. Comparison of the immunobead binding technique (IBT) and the mixed antiglobulin reaction (MAR). Am J Reprod Immunol Microbial 1988;18: 120-3. 5. Haas GG, Cunningham ME. Identification of antibody-laden sperm by cytofluorometry. Fertil Steril 1984,42:606-13. 6. Rasanen ML, Hovatta OL, Penttila IM, Agrawal YP. Detection and quantitation of sperm-bound antibodies by flow cytometry of human semen. J Androl 1992; 1355-64. 7. Nikolaeva MA, Kulakov VI, Ter AG, Terekhina LN, Pshenichnikova TJ, Sukhikh GT. Detection of antisperm antibodies on the surface of living spermatozoa using flow cytometry: preliminary study. Fertil Steril 1993;59:639-44. 8. Ke RW, Dockter ME, Majumdar G, Buster JE, Carson SA. Flow cytometry provides rapid and highly accurate detection of antisperm antibodies. Fertil Steril 1995;63:902-6. 9. Haas GG, D’Cruz 0, DeBault L. Comparison of the indirect immunobead, radiolabelled and immunofluorescence assays for immunoglobulin G serum antibodies to human sperm. Fertil Steril 1991;55:377-88. 10. Rasanen M, Agrawal YP, Saarikoski S. Seminal fluid antisperm antibodies measured by direct flow cytometry do not correlate with those measured by indirect flow cytometry, the
Fertility and Sterility@
11.
12.
13.
14. 15. 16.
indirect immunobead test, and the indirect mixed antiglobulin reaction. Fertil Steril 1996;65:170-5. Mortimer D. Semen analysis and sperm washing techniques. In: Gagnon C, editor. Controls of sperm motility: biological and clinical aspects. Boca Raton (FL): CRC Press, 1990:263-84. Barratt CLR, Dunphy BC, McLeod I, Cooke ID. The poor prognostic value of low to moderate levels of sperm surfacebound antibodies. Hum Reprod 1992;7:95-8. World Health Organization. Laboratory manual for the examination of human semen and semen-cervical mucus interaction. 3rd ed. New York: Cambridge University Press, 1993: 66-8. Fleis JL. Statistical methods for rates and proportions. Toronto: John Wiley, 1981. Landis RJ, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74. Hendry WF, Hughes L, Scammell G, Pryor J, Hargreave TB. Comparison of prednisolone and placebo in subfertile men with antibodies to spermatozoa. Lancet 1990;335:85-8.
Vol. 68, No. 6, December 1997
17. Yovich JL, Matson PL. The treatment of infertility by the high intrauterine insemination of husband’s washed spermatozoa. Hum Reprod 1988;3:939-43. 18. Van DMJ, Kruger TF, Windt ML, Hulme VA, Menkveld R. Treatment of male sperm autoimmunity by using the gamete intrafallopian transfer procedure with washed spermatozoa. Fertil Steril 1990;53:682-7. 19. Naaktgeboren N, Devroey P, Van SA. Successful in vitro fertilization with sperm cells from a man with immune fertility. Ann N Y Acad Sci 1985;442. 20. Bronson R, Cooper G, Rosenfeld D. Sperm antibodies: their role in infertility. Fertil Steril 1984;42:171-83. 21. Bronson RA, Cooper GW, Rosenfeld DL. Seminal fluid antisperm antibodies do not reflect those present on the sperm surface. Fertil Steril 1987;48:505-6. 22. Hellstrom W, Overstreet JW, Moore SM, Samuels SJ, Chang RJ, Lewis EL. Antisperm antibodies bind with different patterns to sperm of different men. J Urol 1987; 138:895-8.
Nicholson
et al.
Antisperm antibodies by flow cytometry
1119
a 1997 American
Published
by Elsevier
Society for Reproductive
Vol. 68, No. 6, December Medicine
1997
Printed on acid-free paper in U. S. A.
Science Inc.
Detection of antisperm antibodies in seminal plasma by flow
cytometry: comparison with the indirect immbnobead binding test
Simon C. Nicholson, M.D. Julian N. Robinson, M.D.? Ian L. Sargent, Ph.D. David H. Barlow, M.D. Nuffield Department
of Obstetrics and Gynaecology,
John Radcliffe Maternity Hospital,
Oxford, United Kingdom
Objective: To compare flow cytometry with the established indirect immunobead binding test (IBT) for the detection of antisperm antibodies in seminal plasma. Design: A prospective, comparative study. Setting: University-based andrology unit. Patient(s): One hundred and fifty-eight men with suspected male factor subfertility. Intervention(s): Seminal plasma samples were incubated with antisperm antibody-negative donor sperm. Surface-bound antibody was detected with fluorescence-labeled antihuman antibody in the flow cytometry assay or with immunobead-labeled antihuman antibody in the IBT. Main Outcome Measure(s): The percentage of sperm that tested positive for surface-bound antibody was determined in the two assays. Seminal plasma was antisperm antibody-positive when ~20% of the sperm were antibody-bound, and clinically significant levels were present when 250% of the sperm were antibody-bound. Result(s): Of 71 samples that were negative by the IMT, 66 (93%) also were negative by flow cytometry. Of 63 samples that had 250% immunobead binding, 55 had equivalent results by flow cytometry. Overall statistical analysis showed a good correlation between the two assays. Conclusion(s): There is a good correlation between the indirect IBT and indirect flow cytometry for the detection of antisperm antibodies in seminal plasma. (Fertil Steril@ 1997;68:1114-9. 0 1997 by American Society for Reproductive Medicine.) Key Words: Antisperm antibody, immunobead binding test, flow cytometry, seminal plasma
Antisperm antibodies have a variety of effects on human fertility, including impairment of sperm motility, trapping and phagocytosis of sperm in cervical mucus, and impairment of sperm-oocyte interaction (1). The most commonly used assays to detect antisperm antibodies are the immunobead binding test (IBT) (2) and the mixed antiglobulin reaction (MAR) (3). They can be performed as direct assays, de-
ReceivedMarch 24, 1997; revisedand acceptedJuly 21, 1997. The Jones Institute for Reproductive Medicine, Norfolk, Virginia. Reprint requests and present address: Simon C. Nicholson, M.D., Department of Obstetrics and Gynaecology, Simpson Memorial Maternity Pavilion, Edinburgh, EH3 9YW, United Kingdom (FAX: 44-131-536-7474).
1114
tecting antibodies already bound to the surface of the patient’s sperm, or as indirect assays, in which donor sperm are used to detect the presence of unbound antisperm antibody in seminal plasma, serum, or cervical mucus. The direct assays necessitate the presence of adequate numbers of motile sperm and are not suitable for men with poor sperm parameters (4). Both the IBT and the MAR require that the investigator assess the percentage of motile sperm that are positive for antisperm antibodies by counting a minimum of 100 sperm using light microscopy. This can be both time-consuming and subjective. Antisperm antibody detection by flow cytometry first was described by Haas and Cunningham (5). The principle of the assay is that sperm with bound
0015-0282/97/$17.00 PI1 SOO15-0282(97)00374-9
surface antibody are incubated with a fluorescencelabeled antihuman antibody. The proportion of antibody-positive sperm then is identified by passing thousands of sperm through a fluorescence-activated cell sorter, which is a highly objective assessor. In addition, the number of antibody molecules bound to the surface of the sperm can be calculated using fluorescent calibration beads. This can be of use in assessing methods of reducing antisperm antibodies both in vitro and in vivo (6). A close correlation between flow cytometry and the more established MAR has been described for both direct and indirect versions of the assays (7). Comparisons of flow cytometry with the IBT have produced more variable results using the indirect testing of male serum (8, 9). The detection of antisperm antibodies in seminal plasma by flow cytometry and other methods has been reported recently and suggests a good correlation between indirect flow cytometry and the indirect IBT on seminal plasma in a small series of positive samples (10). In this study, we examined the validity of indirect seminal plasma antisperm antibody assessment by flow cytometry in a much larger series. MATERIALS
AND METHODS
Seminal Plasma Samples
Seminal plasma samples were obtained from men attending the Oxford Andrology Clinic for evaluation of suspected male factor subfertility. A minimum of 200 /JL of seminal fluid was centrifuged at 1,000 x g for 15 minutes in a 1.5mL capped Eppendorf tube (Sigma, Poole, Dorset, United Kingdom). The supernatant then was transferred to a clean tube, stoppered, and heat-inactivated in a water bath at 56°C for 30 minutes to destroy complement. The cell-free seminal plasma then was stored at -20°C until it was used in the indirect IBT or flow cytometry assay. Indirect Immunobead
Binding Test
A semen sample was obtained from an antisperm antibody-negative semen donor and prepared by the swim-up method to obtain a washed sperm population containing 50 X lO?mL of motile sperm (11). Ten-microliter volumes of seminal plasma each were diluted in 40 PL of phosphate-buffered saline (PBS) supplemented with 5% (50 g/L) bovine serum albumin (BSA) in Eppendorf tubes, and 50 PL of the sperm suspension then was added to each tube and gently mixed. The samples were incubated at 37°C for 60 minutes and then were centrifuged at 500 x g for 5 minutes. The supernatants were discarded and
Vol. 68, No. 6, December
1997
the sperm pellets were resuspended gently in 500 PL of PBS supplemented with 0.4% BSA. The tubes again were centrifuged at 500 X g for 5 minutes. The supernatants were discarded and the pellets were resuspended gently in 50 I.ILof fresh PBS plus 5% BSA. On separate clean microscope slides, 5 PL of the washed sperm suspension was mixed with 5 PL of the GAM immunobeads (Biorad, Richmond, CA), containing a mixture of immunobead-bound antihuman immunoglobulin (Ig) G, A, and M. The slides were incubated at ambient temperature in a humidified chamber for 10 minutes and then were examined at x400 magnification by phase-contrast microscopy. At least 200 motile sperm were counted, differentiating between those with and those without immunobeads attached. The localization of any bound beads also was determined (i.e., head, midpiece, tail, tail tip). The percentage of motile sperm with GAM immunobeads attached was calculated. The test result was regarded as positive when 220% of motile sperm had beads attached and as clinically significant when ~50% were coated with the beads (12). Binding restricted to the tail tip is not considered to be clinically significant (13) and was discounted in this study. Known antisperm antibody-positive and antibody-negative control samples always were included in each run of the assay. Flow Cytometry
The seminal plasma samples were thawed at room temperature and vortexed to ensure thorough mixing. A 5O+L aliquot of each seminal plasma sample was added to 40 PL of PBS supplemented with 5% goat serum albumin. Ten microliters of a sperm suspension, prepared by the swim-up method from antibody-negative donor semen, containing 125,000 motile sperm was added to each sample. Known antisperm antibody-positive and -negative seminal plasma samples were included as controls. A blank sample was included in which sperm were processed in the absence of seminal plasma. After incubation at 37°C in a 5% CO2 incubator for 1 hour, the sperm were washed twice to remove unbound antibody. One milliliter of PBS was added and the mixture was agitated. The tubes then were centrifuged at 500 x g for 5 minutes and the supernatants were discarded. The sperm pellet was resuspended with 1 mL of PBS and the wash was repeated. After centrifugation, the pellet was resuspended with 50 PL of a 1:5 solution of fluorescein isothiocyanate conjugate (FIT(J)-labeled goat anti-
Nicholson
et al.
Antisperm antibodies by flow cytometry
1115
human immunoglobulin directed against IgA, IgG, and IgM (Southern Biotechnology Associates, Inc., Birmingham, AL) and was incubated for a further 1 hour at 4°C in the dark. Unbound antibody was removed by washing twice with PBS as described earlier and the sperm were analyzed by flow cytometry to assess the number with bound antibody using a Coulter Epics Elite Flow Cytometer (Coulter Electronics, Luton, United Kingdom). The sperm population was gated using go-degree and forward-angle light scatter to exclude debris and aggregates. A total of 5,000 sperm were analyzed from each sample, recording single-parameter histograms of the log FITC intensity (5). Digital subtraction of the blank sample from each test sample using the Immunod Elite software program (Coulter Electronics) was performed to determine the percentage of sperm in each sample that were positive for antisperm antibody. A result of <20% of the sperm coated by antibody was regarded as negative and ~20% as positive. All flow cytometry assay conditions were chosen after preliminary experiments to determine the optimum separation of antisperm antibody-negative and -positive sperm populations. Comparison of Indirect Immunobead Indirect Flow Cytometry Assays
Binding and
The two assays were compared by determining the strength of agreement at <20%, 20%-49%, and 250% of the sperm with surface-bound antibody. Seventy-one consecutive antisperm antibody-negative seminal plasma samples (<20% immunobead binding) and 87 consecutive positive seminal plasma samples (~20% immunobead binding) were analyzed for the presence of antisperm antibodies by both flow cytometry and IBT. All seminal plasma samples were tested in duplicate and the mean result (+SD) was taken. Statistical analysis to compare this strength of agreement was performed using the K (14). The K is a chance corrected proportional agreement that assesses the agreement between two sets of values in excess of the agreement that would be expected by chance. The maximum value of the K is 1.00. This is the situation when there is total agreement between the two sets of measured variables. A K of 0 indicates no better agreement than would be expected by chance and a value of <0 indicates worse agreement than would be anticipated by chance. For K values between 0 and 1.00, the measure of agreement is as follows: a K value 520 indicates poor strength of agreement, 0.21-0.40 fair agreement, 0.41-0.60 moderate agreement, 0.61-0.80 good
1116
Nicholson
et al.
Antisperm
antibodies
by flow cytometry
I
Fluorescence intensity / log (Fluoresceinisothiocyanate)
Figure 1 Single-parameter histograms of FITC-labeled goat antihuman Ig binding to sperm after initial incubation with buffer (blank), antisperm antibody (A&B&negative seminal plasma, and antisperm antibody-positive seminal plasma.
agreement, and 0.81-1.00 very good strength of agreement between the two assays (15). RESULTS
Seminal plasma samples were tested first for the presence of antisperm antibodies by the IBT. In total, 158 seminal plasma samples were analyzed. Seventy-one were antisperm antibody-negative (~20% immunobead binding), 63 were strongly positive (~50% immunobead binding), and 24 were intermediate (20%-49% immunobead binding). The samples then were tested for the presence of antisperm antibodies by flow cytometry. A histogram of the log FITC intensity after incubation of sperm with antibody-positive and antibody-negative seminal plasma and after incubation with buffer alone is shown in Figure 1. The histogram after incubation with antisperm antibody-negative seminal plasma overlies that for the sperm population incubated in the absence of seminal plasma. Of the 71 samples that were negative by the indirect IBT, 66 (93%) also were negative by flow cytometry (i.e., <20% of the sperm were positive for antibody). Four (6%) of the seminal plasma samples that were negative by the IBT produced intermediate results by the flow cytometry assay (20%-49% of the sperm were bound with antibody) and 1 (1%) produced strongly positive results (i.e., >50% of the sperm were bound with antibody). Sixty-three seminal plasma samples had ~50% immunobead binding. Of these, 55 (87%) had ~50% binding in the flow cytometry assay, 7 (11%) had 20%-49% binding, and 1 (2%) had <20% binding. Of the 24 samples that produced intermediate results by the IBT, 19 (79%) produced similar results
Fertility
and Sterility@
. . . T
20
0
40
60
80
Ikl
Resultsof indirectIBT (percentof sperm positivefor antibody) -I
L
Figure 2 Correlation between indirect immunobead binding test (IBT) and flow cytometry results for the presence of seminal plasma antisperm antibodies (r = 0.924; P < 0.001).
by flow cytometry and 5 (21%) produced negative results. None of these intermediate samples produced a ~50% result with flow cytometry. The comparison between the two assays is demonstrated in Figure 2 and the data are summarized in Table 1. The mean percentage of binding (+SD) was 74% + 28% for the immunobead-positive results and 70% t 29% for the flow cytometry-positive samples. For the negative samples, the mean percentage of binding (*SD) for IBT was 6% + 6% and that for flow cytometry was 8% ? 4%. Statistical analysis of the strength of agreement between the two assays gave a K of 0.82, demonstrating a very good correlation between the indirect flow cytometry assay and the indirect IBT. With a 20% cutoff for a positive antisperm antibody result, the sensitivity of the new assay was 0.94 and the specificity was 0.92. DISCUSSION
This study denmnstrates a good correlation between an indirect flow cytometry assay and the indirect IBT for the detection of antisperm antibodies in seminal plasma. Reliable assays are necessary to direct the appropriate clinical interventions should antisperm antibodies be indicated as a possible cause of infertility. Intermediate-dose cyclical steroid therapy for the male partner with immune subfertility has been demonstrated to produce a significant improvement in the spontaneous pregnancy rate compared with placebo (16). Assisted methods of conception, including IUI with washed sperm (171,
Vol. 68, No. 6, December
1997
GIFT (18), and IVF-ET (19), also have been shown to be of benefit in couples with immune subfertility. Ke et al. (8) previously have demonstrated an equally good correlation between flow cytometry and the IBT for the detection of antisperm antibodies in male serum. They tested 37 men who had undergone vasectomy and 35 fertile control men. Flow cytometry detected all 22 of the IBT IgG-positive samples and 12 of the 14 IgA-positive samples. All the IBTnegative samples were identified correctly as negative by flow cytometry. In an earlier study, Haas et al. (9) reported a specificity of 100% but a sensitivity of only 77% when they compared these two assays for antibody detection in serum. However, in their study, dead cells were not excluded from the analysis. Dead cells can produce false-negative or false-positive results because of denaturation of sperm antigens or exposure of internal antigens (20). In our study, dead sperm were excluded by the swim-up preparation and electronic gating. In the MAR and IBT, nonspecific binding to dead sperm is excluded by only counting viable, motile sperm. Indirect assays for antisperm antibody detection are necessary for the testing of serum, cervical mucus, and seminal plasma when sperm parameters are very poor. An indirect assay also obviates the need for same-day analysis because seminal plasma samples can be frozen and processed at a later date, with other samples, more efficiently. However, in men with good sperm parameters, there is debate as to whether a direct or an indirect assay is more appropriate. Direct assays detect antibodies already bound to the surface of sperm. It has been suggested that antibodies that remain within seminal plasma may not be reflective of the Igs that are bound to the sperm cell surface. In a study by Bronson et al. (21) in which direct and indirect IBTs were compared in 36 antisperm antibody-positive men, a good correlation between the two assays was obtained. All 36 men had positive results with the direct assay and 33 had positive
Table 1 Comparison of the Number of Samples Observed With Varying Percentages of the Sample Showing Positive Sperm Binding for Flow Cytometry and IBT Results of 158 Seminal Plasma Samples Flow cytometry IBT results
Nicholson
20%-49%
250%
7 19 4
55 0 1
250% 20%-49% <20%
et al.
Antisperm
results
arztibodies
by flow cytometry
<20% 1 5 66
1117
results with the indirect assay. Of the original 36 men, 29 had head-directed antisperm antibodies. The indirect assay failed to detect the head-directed antibodies in 18 of these 29 men. Hellstrom et al. (22) assessed interdonor variability by testing a single antisperm antibody-positive serum sample against 27 different donors using the IBT. The mean total percentage of binding was 92%, with a range of 62%- 100% for all donors. The localization of binding also varied. Analysis of variance showed there to be a statistically significant difference between the donors. However, the positive serum tested was confirmed to be positive consistently, at >50% in all 27 donors. Rasanen et al. (10) recently carried out a study comparing direct and indirect assays. They showed a good correlation between the direct MAR and a direct flow cytometry assay, although the percentage of sperm that tested positive with flow cytometry tended to be less than with the MAR. However, of the 11 semen samples that were positive for IgG antisperm antibodies by the direct MAR, 9 also were positive by the indirect MAR, 8 by the indirect IBT, and only 5 by the indirect flow cytometry. The results for IgA were better, with both the indirect MAR and the indirect IBT detecting 10 of 11 direct MAR-positive samples and indirect flow cytometry detecting all the direct MAR IgA-positive samples. When they compared the indirect IBT results with the indirect flow cytometry results of the same 11 men for the presence of IgG, they demonstrated a good correlation, as in our study. These investigators suggested that the large discrepancy in their study between the results of direct and indirect assays may have occurred because tailtip binding of antibody is detected less well by flow cytometry than is head-directed binding. The most frequent type of binding in their IBT- and MARpositive samples was tail-tip. In our analysis, tailtip binding alone was not regarded as a positive result. In assessing the correlation between the two assays, we compared their ability to discriminate between positive results (~-20% of sperm antibodybound) and negative results (<20% of sperm antibody-bound) and to determine results of clinical significance (~50% of sperm antibody-bound) (12). These factors are important for a clinical assay and may have contributed to the close correlation between the assays. Flow cytometry assesses antibody binding on very large numbers of sperm and provides a highly objective assessment. Quantitative information on the number of antibody molecules per sperm can be ob-
1118
Nicholson
et al.
Antisperm antibodies by flow cytometry
tained, which may be of use in the evaluation of antibody-lowering interventions. By performing an indirect assay, seminal plasma samples can be stored frozen and tested as weekly batches. This is of importance in flow cytometry because in most centers, the flow cytometer is located in a different department and access to the machine often is limited. In practice, it would not be possible to process sporadic sperm samples because patients attend a clinic for evaluation and most andrology centers do not possess their own machine. In summary, this study demonstrates a good correlation between the indirect IBT and indirect flow cytometry for the detection of antisperm antibodies in seminal plasma, with high specificity and sensitivity, providing an objective and quantitative method of antibody detection. However, flow cytometry does require the availability of expensive equipment and a skilled operator. Therefore, for most andrology units, assays such as the IBT are likely to continue to have an important clinical role in assessing subfertile couples and flow cytometry will remain a useful research tool at present. REFERENCES 1. Hendry WF. The significance of antisperm antibodies: measurement and management. Clin Endocrinol 1992;36:21921. 2. Bronson R, Cooper G, Hjort T, Ing R, Jones WR, Wang SX, et al. Anti-sperm antibodies, detected by agglutination, immobilization, microcytotoxicity and immunobead-binding assays. J Reprod Immunol 1985;8:279-99. 3. Jager S, Kremer J, Van Slochteren-Draaisma T. A simple method for screening for antisperm antibodies in the human male. Int J Fertil 1978;23:12-8. 4. Meinertz H, Bronson R. Detection of antisperm antibodies on the surface of motile spermatozoa. Comparison of the immunobead binding technique (IBT) and the mixed antiglobulin reaction (MAR). Am J Reprod Immunol Microbial 1988;18: 120-3. 5. Haas GG, Cunningham ME. Identification of antibody-laden sperm by cytofluorometry. Fertil Steril 1984,42:606-13. 6. Rasanen ML, Hovatta OL, Penttila IM, Agrawal YP. Detection and quantitation of sperm-bound antibodies by flow cytometry of human semen. J Androl 1992; 1355-64. 7. Nikolaeva MA, Kulakov VI, Ter AG, Terekhina LN, Pshenichnikova TJ, Sukhikh GT. Detection of antisperm antibodies on the surface of living spermatozoa using flow cytometry: preliminary study. Fertil Steril 1993;59:639-44. 8. Ke RW, Dockter ME, Majumdar G, Buster JE, Carson SA. Flow cytometry provides rapid and highly accurate detection of antisperm antibodies. Fertil Steril 1995;63:902-6. 9. Haas GG, D’Cruz 0, DeBault L. Comparison of the indirect immunobead, radiolabelled and immunofluorescence assays for immunoglobulin G serum antibodies to human sperm. Fertil Steril 1991;55:377-88. 10. Rasanen M, Agrawal YP, Saarikoski S. Seminal fluid antisperm antibodies measured by direct flow cytometry do not correlate with those measured by indirect flow cytometry, the
Fertility and Sterility@
11.
12.
13.
14. 15. 16.
indirect immunobead test, and the indirect mixed antiglobulin reaction. Fertil Steril 1996;65:170-5. Mortimer D. Semen analysis and sperm washing techniques. In: Gagnon C, editor. Controls of sperm motility: biological and clinical aspects. Boca Raton (FL): CRC Press, 1990:263-84. Barratt CLR, Dunphy BC, McLeod I, Cooke ID. The poor prognostic value of low to moderate levels of sperm surfacebound antibodies. Hum Reprod 1992;7:95-8. World Health Organization. Laboratory manual for the examination of human semen and semen-cervical mucus interaction. 3rd ed. New York: Cambridge University Press, 1993: 66-8. Fleis JL. Statistical methods for rates and proportions. Toronto: John Wiley, 1981. Landis RJ, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159-74. Hendry WF, Hughes L, Scammell G, Pryor J, Hargreave TB. Comparison of prednisolone and placebo in subfertile men with antibodies to spermatozoa. Lancet 1990;335:85-8.
Vol. 68, No. 6, December 1997
17. Yovich JL, Matson PL. The treatment of infertility by the high intrauterine insemination of husband’s washed spermatozoa. Hum Reprod 1988;3:939-43. 18. Van DMJ, Kruger TF, Windt ML, Hulme VA, Menkveld R. Treatment of male sperm autoimmunity by using the gamete intrafallopian transfer procedure with washed spermatozoa. Fertil Steril 1990;53:682-7. 19. Naaktgeboren N, Devroey P, Van SA. Successful in vitro fertilization with sperm cells from a man with immune fertility. Ann N Y Acad Sci 1985;442. 20. Bronson R, Cooper G, Rosenfeld D. Sperm antibodies: their role in infertility. Fertil Steril 1984;42:171-83. 21. Bronson RA, Cooper GW, Rosenfeld DL. Seminal fluid antisperm antibodies do not reflect those present on the sperm surface. Fertil Steril 1987;48:505-6. 22. Hellstrom W, Overstreet JW, Moore SM, Samuels SJ, Chang RJ, Lewis EL. Antisperm antibodies bind with different patterns to sperm of different men. J Urol 1987; 138:895-8.
Nicholson
et al.
Antisperm antibodies by flow cytometry
1119