High levels of sperm-associated antibodies impair human spermoolemma interaction after subzonal insemination*

Transcription

1 FERTILITY AND STERILITY Vol. 63, No.3, March 1995 Copyright 1995 American Society for Reproductive Medicine Printed on acid-free paper in U. s. A. High levels of sperm-associated antibodies impair human spermoolemma interaction after subzonal insemination* J. Philippe Wolf, M.D., Ph.D.t:j: Daniel Rodrigues, T.L. t Marta De Almeida, M.D., Ph.D. t Pierre Jouannet, M.D.t Beatrice Ducot, Ph.D. University of Paris XI, and INSERM U 292, Hopital Bicetre, Le Kremlin Bicetre, France Objective: To investigate the impact of sperm-associated antibodies in the fertilization process beyond the zona pellucida in human oocytes. Design: Subzonal insemination (SUZI) was performed with antibody-coated spermatozoa from patients with autoimmune infertility. Sperm parameters and antibody binding were analyzed after Percoll selection and were compared with SUZI outcome. Patients: Patients (n = 29) with an immunobead test higher than 60% for immunoglobulin (Ig)G, IgA, or both classes were included in the study (38 attempts). Twenty-six patients had previous IVF failures, and the semen characteristics for three were unsuitable for IVF. Results: Diploid fertilization occurred in 52.6% of cycles and 16.0% of oocytes (n = 263). The fertilization rate was inversely correlated to the proportion of spermatozoa coated with IgG in the Percoll-selected sperm suspension. Three of the 16 attempts in which >90% of the spermatozoa were coated with IgG resulted in a low diploid rate of 4.1 %. Fertilization was obtained in 18 of the 22 attempts in which <90% of spermatozoa were IgG coated, with a diploid fertilization rate of 22.8%. Association of both classes of antibodies (IgG and IgA) further impaired the fertilization rate. Twenty ETs gave four pregnancies: one biochemical, one ongoing, and two giving birth to healthy babies (1 singleton and 2 twins). Conclusions: Subzonal insemination is a potential solution for achieving fertilization in cases of IVF failure because of sperm autoimmunity. However, IgG inhibits in a dose-dependent manner the fusion of gametes' membranes. Though the level of IgA antibodies does not appear as critical as that of IgG antibodies, association of both classes of Igs impairs the probability of fertilization. Fertil Steril 1995;63: Key Words: Antisperm antibody, oolemma, IVF failure, SUZI, human sperm, infertility Received May 25, 1994; revised and accepted September 30, Supported in part by grant EA 1606 from Direction de la Recherche et des Etudes Doctorales, Paris, France. t Laboratoire de Biologie de la Reproduction et du Developpement-Histologie, Embryologie, Cytogenetique. :j: Reprint requests: Jean Philippe Wolf, M.D., Ph.D., Laboratoire de Biologie de la Reproduction et du Developpement, Hopital Biditre, 78, rue du General Leclerc, Le Kremlin Bicetre, France (FAX: ). INSERM U 292, Hopital Bicetre. Antisperm antibodies have been shown to play an important role in male infertility when >60% of the ejaculated spermatozoa are coated with antibodies (1). Because these antibodies generally prevent spermatozoa from migrating through the cervical mucus (2), IVF has been recommended as a method for fertilization for couples with such autoimmune infertility (3-5). Even in vitro, sperm-associated antibodies may interfere with the gamete interaction process and be responsible for low fertilization rates. Indeed in such cases, complete fertilization failure was observed in 35% to 47% of attempts (3, 5). Our laboratory attempted 49 IVF during the last 3 years for 36 couples suffering from male infertility because of high levels of antisperm antibodies on the spermatozoa membranes. The mean fertilization rate was 40%, but 16 attempts were complete 584 Wolf et al. SUZI with antibody-coated spermatozoa Fertility and Sterility

2 failures. These IVF failures were associated with decreased sperm movement values and an incapacity of the antibody-coated spermatozoa to bind to the zona pellucida (ZP) (6). There is also some evidence from the zona-free hamster oocytes penetration tests that sperm-associated antibodies interfere with the membrane fusion process. Human spermatozoa associated in vivo with antibodies have a reduced ability to penetrate hamster oocytes (7), but the degree of impairment is variable. Recently, Zouari et a1. (8) showed that antibodies, eluted from autoimmune ejaculates and transferred in vitro onto normal donor spermatozoa, affected differently their ability to penetrate the human ZP and zona-free hamster oocytes. In experiments using human spare oocytes from an IVF program, Liu et a1. (9) found that antisperm antibodies did not inhibit sperm binding to the plasma membrane of zona-free human oocytes. Apparently, sperm-associated antibodies inhibit sperm-zona binding more consistently than sperm oocyte interaction. If so, subzonal insemination (SUZI) may help achieve fertilization for patients with previous IVF failures because of sperm-associated antibodies. Indeed SUZI has been used successfully for patients following IVF failures, without antisperm autoimmunity but presenting flagellar dyskinesia and for who defective interaction between spermatozoa and the ZP were suspected (10). Whether SUZI is sufficient to allow antibody-coated spermatozoa to fertilize human oocytes is unknown currently. We therefore studied the impact of sperm-associated antibodies on the spermatozoa-oolemma interaction in human oocytes. The criteria were the fertilizing ability of the sperm and early embryo development as assessed by the fertilizing, cleavage, and implantation rates. Results were analyzed according to the degree of immunization and class of sperm -associated immunoglobulins (Igs). Patients included in the SUZI program had previous IVF failures or such poor sperm parameters that IVF insemination was not possible. Informed consent was obtained from all subjects. The program received the agreement of the local ethical committee. Patients MATERIALS AND METHODS Twenty-nine couples were included in the study. They had been primarily infertile for 2 to 10 years (mean, 5.2 years). Most (n = 26) had previous IVF failures with a total of 9 embryos for 291 insemi- nated oocytes (3.1%). In the majority ofthese IVF attempts, fertilization failure was associated with poor sperm binding to the ZP. The number of motile spermatozoa available from the three other patients was too low to perform standard IVF. At the time of SUZI, the male partners were 24 to 66 years old (mean, 35.1 years) and had the following mean (±) semen characteristics: sperm concentration 48.3 ± 56.3 (range, 2.2 to X 106/mL), sperm motility 29.0% ± 11.8% (range, 10% to 55%), and normal forms 39.3% ± 12.4% (range, 13% to 69%). The percentage of antibody-bound motile spermatozoa in the semen, as measured by the direct immunobead test, was higher than 60% for at least one class of Igs in all patients; it varied from 15% to 100% for IgG (mean, 76.2%) and from 3% to 99% for IgA (mean, 63.4%). The female partners were between 23 and 39 years old (mean, 32.1 years). They had normal reproductive profiles with the exceptions of tubal pathology in one patient and ovulatory dysfunction in two others. Sperm Processing Ejaculates were collected by masturbation at the laboratory directly into a volume of B2 medium (Api System, Marcy l'etoile, France) corresponding to twice the mean semen volume observed in previous ejaculates. The diluted samples were passed immediately through capillary Pasteur pipettes to break up the seminal clots and centrifuged at 600 X g for 10 minutes. The sperm pellets were resuspended in 2 ml of B2 medium and layered over a two-step (47.5% and 95%) Percoll gradient (0.5 ml each layer) (11). After 20 minutes of centrifugation at 300 X g, the pellet containing the highly motile sperm fraction was removed and washed twice by centrifugation for 10 minutes at 600 X g with 5 ml of B2 medium. The final sperm pellets were resuspended in 0.2 to 0.5 ml of B2 medium and kept for 3 hours at 37 C and 5% CO 2 in air before use. Sperm Analysis The following analyses were performed on aliquots of Percoll-selected sperm preparations. The percentage of motile spermatozoa was estimated by microscopic observation at X200 magnification. For morphological analysis, 100 spermatozoa were examined at X1,000 magnification on a Shorrstained smear (Shorr Staining; Merck, Darmstadt, Germany) and classified according to the criteria of David et a1. (12). The direct immunobead test was Vol. 63, No.3, March 1995 Wolf et al. SUZI with antibody-coated spermatozoa 585

3 used for the analysis of sperm surface-bound antibodies. Ten microliters of the migrated sperm suspensions were placed on a glass slide, mixed with an equal volume of isospecific anti -IgA or anti -IgG immunobeads prepared as previously described (13), and covered with a 22 X 22-mm coverslip. After 10 minutes of incubation in a moist chamber at room temperature, the percentage of motile spermatozoa coated with the immunobeads was scored, and the site of immunobead binding was observed using phase-contrast optics at X400 magnification. A minimum of 100 spermatozoa per assay were analyzed. The intra-assay variation coefficient was <1 %. The motility of aliquots of Percoll-selected sperm suspensions was analyzed using a computerassisted semen analyzer (HTM 2030 version 7.2; Hamilton Thorn Research, Danvers, MA) 15 minutes after processing and 3 hours after incubation in B2 at 37 C and 5% CO 2 in air. An aliquot of the sperm suspension was diluted to 1 X 10 6 cellsjml with B2 and loaded into a fiat glass microcapillary tube 200 /-Lm deep (Vitro Dynamics, Rockaway, NJ) that was transferred to the analyzer at 37 C. At least 150 motile spermatozoa were scored to obtain mean values for the following movement parameters: curvilinear velocity (VCL), straight line velocity (VSL), amplitude of lateral head displacement (ALH), linearity (LIN = VSLjVCL), and beat cross frequency (BCF) (14, 15). Oocyte Preparation and SUZI Superovulation was performed in the Gynecological Department of Hopital A. Beclere (Clamart, France) and Hopital International de la Cite Universitaire (Paris, France). Follicular growth was stimulated by hmg associated with a GnRH agonist. Ovulation was induced by administration of 10,000 IU of hcg, and the oocytes were harvested 36 hours later by a transvaginal ultrasound procedure (16). The cumuli were washed and individually placed into a 30-/-LL drop ofb2 medium under equilibrated oil. The cumulus cells were removed by exposure of the eggs to 0.1 % hyaluronidase (type III; Sigma Chemical Co., St. Louis, MO) in B2 medium, and nuclear status of the oocytes was checked. Sub zonal insemination was performed as described previously (17) on metaphase II oocytes, 6 to 8 hours after retrieval. Two micromanipulators IM5-B (Narishige, Japan) and one inverted microscope IMT-2 (Olympus, Japan) were used. Between 1 and 16 spermatozoa were chosen among those with forward progressive movement and normal morphology and were microinjected into the perivitelline space. Analysis Criteria The oocytes were checked 16 to 18 hours after SUZI for evidence of fertilization. Oocytes exhibiting two pronuclei were considered normally fertilized, and zygotes were kept in culture for an additional 24 hours. Only embryos that regularly cleaved were transferred using a Frydman catheter (CCD, Paris, France). Attempts were classified according to the percentage of spermatozoa coated with IgG in the sperm suspension used for the SUZI. Fertilization Rate According to the Number of Spermatozoa Microinjected During the first attempts, the number of spermatozoa that were microinjected was assigned randomly to each oocyte, between 1 and 16. When it appeared that very low numbers of spermatozoa almost never achieved fertilization, the minimum number of spermatozoa microinjected was increased to 4. Statistical Analysis Results are expressed as means ± SD and were compared using nonparametric tests. Wilcoxon's paired test was used to compare fertilization rates and percentages of antibody-coated spermatozoa in different groups. The relationship between IVF rates and sperm characteristics was examined by Spearman's rank correlation coefficients. RESULTS For the 38 attempts of SUZI, 263 metaphase II oocytes were micro injected successfully. Fertilization was observed in 57 of these oocytes (21.6%), but 15 of them were polyspermic (5.7%). Of the 42 diploid zygotes (16.0% of microinjected oocytes), 39 (92.9%) cleaved normally. Thirty-three embryos were transferred in 20 cycles. Ten transfers were performed with one embryo, 7 with two embryos, and 3 with three embryos. Six embryos were cryopreserved. In one case, all zygotes were polyspermic. Four pregnancies were obtained: one biochemical and three clinical. Three patients were transferred with one embryo each and the fourth with two embryos. Two of the three clinical pregnancies ended in full term deliveries of healthy 586 Wolf et al. SUZI with antibody-coated spermatozoa Fertility and Sterility

4 100 " II " " 80 "rp " " " " ~ " 60 ~ " " O~~--~~--r--r--~-r~~~~ Fertilized 0 10 W ~ ~ ~ ro m W 00 ~ " Unfertilized IgG(%) Figure 1 Fertilizing and nonfertilizing attempts of SUZI according to the percentage of spermatozoa coated with antibodies from the IgG and IgA classes. D, the attempts without any fertilization., the attempts during which at least one oocyte was fertilized. singleton and twin babies. The third pregnancy is still ongoing. Two of these pregnancies were in patients who previously had not attempted IVF and the other two in patients with previous IVF failures. The implantation rate was 12.1 %. The percentage of spermatozoa coated with IgG in the sperm suspensions obtained after Percoll migration was significantly higher in the cycles with complete fertilization failure (n = 17) than in the cycles during which fertilization occurred (n = 21) (90.2% ± 12.4% versus 75.6% ± 15.2%, P < 0.01). There was a negative and statistically significant correlation (r = -0.54, P < 0.01) between the fertilization rate and the proportion of spermatozoa coated with antibodies of the IgG class. There was a nonsignificant inverse correlation between IgA sperm immunization and the fertilization rate. Most patients included in this study had both IgA and IgG antibodies on their spermatozoa (Fig. 1), but the level of sperm -associated IgA was generally lower and even negligible (::5:20%) in seven attempts (Fig. 1). In these seven attempts (61 oocytes), only IgG were present on 68 % to 94 % of the inseminated spermatozoa. At least one oocyte was fertilized in six cases, and the overall fertilization rate was 29.5%. In the attempts with a percentage of spermatozoa coated with IgG in the same range but associated to IgA antibodies (n = 14,93 oocytes) the fertilization rate was significantly lower (17.2%; P < 0.01). None of the other parameters, including sperm characteristics before and after Percoll selection, were correlated significantly to the fertilization rate. Because the proportion of spermatozoa coated with Igs from the IgG class appeared determinant, the different attempts were classified according to the proportion of IgG-associated spermatozoa in the Percoll-selected sperm suspension actually used for SUZI (Table 1). Four groups were distinguished. In group I with 100% IgG and 100% to 72% IgA (8 cycles, 43 oocytes), only one oocyte was fertilized. In group II with 99% to 90% IgG, IgA ranged between 98% and 10%, and two oocytes out of 53 fertilized (8 cycles). In the other two groups, the proportions of spermatozoa coated with IgG were 80% to 89% (group III: 9 cycles, 63 oocytes) and <80% (group IV: 13 cycles, 104 oocytes) and the overall fertilization rates increased up to 27.0% and 34.6%, respectively (Table 1). The percentage of spermatozoa bound to IgA antibodies in group I was significantly higher than in the other groups (P < 0.01), but it was not different between groups II, III, and IV. Sperm count, motility, and morphology as well as the sperm movement values of ALH, VCL, LIN, and BCF did not differ significantly between the groups. Straight line velocity was lower in groups I and II than in groups III and IV (55.3 ± 15.5 versus 68.3 ± 18.2 Ils-I, P = 0.053) both after the Percoll selection and after 3 hours of incubation. All the sperm characteristics, including movement parameters measured after Percoll selection, did not differ between the attempts with and without fertilization. The localization of antibodies on the sperm surface was not distributed uniformly among the four groups. All but 2 ofthe 24 patients from groups I, II, and III (IgG between 80% and 100%) had antibodies bound to the whole sperm surface. On the contrary, in half of the patients from the fourth group «80% of spermatozoa coated with IgG), the antibodies were localized on the flagella only. Interestingly, the fertilization rates for group IV were higher when the antibodies were localized strictly to the flagella (8 attempts, 73 oocytes, 38.4% fertilized) than when they also were localized on the head surface (5 attempts, 31 oocytes, 25.8% fertilized), but the difference was not significant. The diploid fertilization rates of oocytes were scored according to the number of spermatozoa delivered into the perivitelline space and to their immunization (Fig. 2). When less than four spermatozoa were microinjected, the fertilization rate remained very low at 2.9% (lout of 34 oocytes). When more than four were microinjected, the overall fertilization rate increased up to 24.7% and plateaued around this percentage. The fertilization rates for groups I and II (IgG between 90% and Vol. 63, No.3, March 1995 Wolf et al. SUZI with antibody-coated spermatozoa 587

5 Table 1 Fertilization Rates and Pregnancies Obtained in the SUZI Attempts Grouped According to the Percentage of Spermatozoa Coated With Antibodies of the IgG Class No. of Group* IgGt IgAt cycles 100 ±o 94.0 ± (100) (100 to 72) II 94.3 ± ± (99 to 90) (98 to 10) III 85.3 ± ± (89 to 80) (89 to 0) IV 65.7 ± ± «80) (87 to 0) * See text for the definition of the groups. t Values are means ± SD with ranges in parentheses. No. of No. of oocytes Diploid:j: Polyspermic:j: pregnancies 43 1 (2.3) (5.7) (15.9) 7 (11.1) (26.9) 8 (7.7) 2 :j: Values in parentheses are percents. Significantly different from groups II, III, and IV (P < 0.01). 100%) were low, independent of the number of spermatozoa used. To obtain fertilizations, more than seven spermatozoa from group III (IgG between 80% and 89%) and more than four from the group IV (IgG < 80%) had to be micro injected. The rate of polyspermy increased linearly with the number of microinjected spermatozoa. DISCUSSION Results of the SUZI using antibody-coated spermatozoa show undoubtedly that bypassing the ZP 50 ] 40 II '" B [ 30 'a g, J GI Gil Gill GIV Figure 2 Diploid fertilization rate according to the percentage of spermatozoa coated with antibodies from the IgG class and to the number of sperm micro injected into the perivitelline space. The percentage of fertilized oocytes is given according to the number of spermatozoa microinjected (vertical bars) and to the degree of sperm immunization with the antibody from the IgG class. Group I: 100% IgG; group II: IgG between 90% and 99%; group III: IgG between 80% and 89%; group IV: IgG < 80%., 1 to 3 spermatozoa microinjected. II, 4 to 6 spermatozoa microinjected. II, 7 to 9 spermatozoa microinjected. ~, 10 or more spermatozoa microinjected. Numbers above the bars represent the number of oocytes in each subgroup. allows fertilization for patients with long-lasting infertility and unsuccessful IVF attempts because of the presence of sperm-associated antibodies. However, the presence of antibody on the sperm cell surface reduces its fertilizing ability. Indeed, the fertilization rates with immune sperm are lower than those obtained by SUZI after IVF failure with normal sperm (fertilization rate of 40.0% with 17.0% polyspermy) and compares with those obtained with subnormal sperm (fertilization rate of 17.2% with 4.9% polyspermy) (17). Because in the SUZI attempts analyzed here, fertilization is not related to the patients sperm parameters such as count, motility, and morphology, it definitely suggests the responsibility of the antibodies in the reduced fusiogenic ability of the sperm membrane. The proportions of the spermatozoa associated with either IgG or IgA antibodies were higher in the sperm populations that did not fertilize after SUZI. However, this difference was only statistically significant for IgG antibodies. Furthermore, there was a statistically significant dose-dependent inhibition of the fertilization rate by IgG antibodies. When all the spermatozoa present in the suspension were coated with IgG antibodies, the chance of fertilization was almost nil (1 of 43 oocytes). In contrast, when the proportion of spermatozoa coated with IgG was <80%, the antibodies did not appear to interfere with the sperm-egg fusiogenic process because the fertilization rate in these cases (34.6%) was similar to that obtained after SUZI of antibody-free normal sperm (17). In the majority of the SUZI attempts analyzed here, both classes of antibodies were present on the sperm cells. It is thus difficult to reach conclusions on their respective responsibility in the observed effects, but the level of IgA antibodies was statistically less critical than that of IgG antibodies for the spermatozoa-oocyte membrane fusion. How- 588 Wolf et al. SUZI with antibody-coated spermatozoa Fertility and Sterility

6 ever, our data provide evidence that association of both classes of Igs on the sperm surface decreases the fertilization rates more than IgG alone. One also can speculate that in a sperm suspension, the spermatozoa susceptible to bind to the ooplasm may be those that are not coated effectively with Igs. Previous data from human-hamster oocyte penetration assays (7) showed a significant correlation between the presence of sperm-associated IgG alone or associated with IgA and the failure to penetrate the hamster oocytes; sperm -associated IgA alone did not appear to interfere. But recent experimental evidence suggests that both classes of antibodies must be associated to impair sperm penetration into zona-free hamster oocytes (8). Thirty-five of the 38 attempts of SUZI analyzed here were performed after a previous IVF failure. According to a previous report by Zouari et al. (8), these patients had a narrowed ALH, and antibodies mostly localized on their head. Therefore, little variation between their sperm movement parameters and antibody localization could be expected among them. It is, however, interesting to note that among sperm movement parameters, low VSL in the migrated sperm suspensions is associated with higher percentages ofigg-coated spermatozoa and the lowest fertilization rates. But low VSL has not been shown to be a reason for fertilization failure after SUZI. Amplitude of lateral head displacement, which has been shown to be correlated with the ability of sperm to pass through the egg vestments (6), does not vary with the degree of immunization of the sperm used for SUZI or with the obtained fertilization rate. The localization of the antibody is critical for the ability of spermatozoa to pass through the ZP (8). Because almost all patients had failed to fertilize during previous IVF (26 of 29), most of them had actually antibodies bound on the sperm head (24 of 29). However, in 8 of the 13 SUZI attempts from group IV (with the highest fertilization rate), antibodies were localized on the flagella. In these 8 attempts, the fertilization rate (38.4 %) was slightly higher than in the remaining 5 with head localized antibodies (25.8%). The low number of cycles with antibodies only on the flagella was too low for statistical analysis. The fertilization rate varies more with the degree of sperm immunization than with the number of micro injected spermatozoa. Below four spermatozoa micro injected, the fertilization rate was almost nil, independent of the degree of sperm immunization (1 oocyte out of 34). Above this threshold, the diploid fertilization rate increased up to 25% and plateaued at this level. Increasing the number of microinjected spermatozoa improved the fertilization rate only for patients from groups III and IV. Definitely, to achieve fertilization, a minimum of four spermatozoa should be micro injected for sperm with IgG between 60% and 80% and a minimum of seven for those with IgG between 80% and 89%. Polyspermy appeared only when more than four spermatozoa were microinjected and increased linearly above this number. The overall rate of polyspermy was low (5.7%) compared with that observed when normal sperm is used (17.0%) (17). This is presumably due to the impaired fusiogenic ability of the spermatozoa coated with antibodies. The implantation rate of embryos derived from antibody-coated spermatozoa were lower than that for the group of patients with IVF failure and normal sperm (12.1 % versus 17.2%) (17). It compares well with the implantation rate obtained after SUZI for IVF failures with subnormal sperm (11.8%) (17). There was no evidence that associated with Igs the spermatozoa impaired embryo development. In conclusion, SUZI may be useful to achieve fertilization in cases of antisperm autoimmune infertility after failure of standard IVF attempts. However, there is a dose-dependent inhibition of the fertilization process by antibodies mainly from the IgG class. Above 90% IgG-coated spermatozoa in the Percoll-selected sperm, the fertilization rate is almost nil, but if <80% of the spermatozoa are coated with IgG, the 27% of diploid fertilization observed justifies the use of SUZI as the last therapeutic trial for these infertile couples. Though the level of IgA antibodies does not appear as critical as IgG antibodies for the fertilization rate, association of both classes of Igs reduces the probability of fertilization. According to this result, fertilization rates could be improved by the immunoselection of antibody-free spermatozoa to inject under the ZP. This approach is currently under investigation. REFERENCES 1. Ayvaliotis B, Bronson RA, Rosenfeld DL, Cooper GW. Conception rates in couples where autoimmunity to sperm is detected. Fertil Steril 1985;43: Menge AC, Beitner O. Interrelationships among semen characteristics, antisperm antibodies, and cervical mucus penetration assays in infertile human couples. Fertil Steril 1989;51: Clarke GN, Lopata A, McBain JC, Baker HW, Johnston W JH. Effect of sperm antibodies in males on human in vitro fertilization. Am J Reprod Immunol MicrobioI1985;8: Junk SM, Matson PL, Yovich JL. The fertilization of human oocytes by spermatozoa from men with antispermato- Vol. 63, No.3, March 1995 Wolf et al. S UZI with antibody-coated spermatozoa 589

7 zoa antibodies in semen. J In Vitro Fert Embryo Transf 1986;3: De Almeida M, Gazagne I, Jeulin C, Herry M, Belaisch-Allart J, Frydman R, et al. In vitro processing of sperm with autoantibodies and in vitro fertilization results. Hum Reprod 1989;4: Zouari R, De Almeida M, Rodrigues D, Jouannet P. Localization of antibodies on spermatozoa and sperm movement characteristics are good predictors of in vitro fertilization success in cases of male autoimmune infertility. Fertil Steril 1993;59: Haas GG, Ausmanus M, Culp L, Tureck RW, Blasco L. The effect of immunoglobulins occurring on human sperm in vivo on the human sperm/hamster ova penetration assay. Am J Reprod Immunol MicrobioI1985;7: Zouari R, De Almeida M. Effect of sperm associated antibodies on human sperm-ability to bind to zona pellucida and to penetrate zona-free hamster oocytes. J Reprod Immunol 1993;24: Liu DY, Clarke GN, Baker HWG. Inhibition of human sperm-zona pellucida and sperm-oolemma binding by antisperm antibodies. Fertil Steril 1991;55: Wolf JP, Feneux D, Escalier D, Rodrigues D, Frydman R, Jouannet P. Pregnancy after subzonal insemination with spermatozoa lacking outer dynein arms. J Reprod Fertil 1993;97: McClure RD, Nunes L, Tom R. Semen manipulation: improved sperm recovery and function with a two-layer Percoll gradient. Fertil Steril1989;51: David G, Bisson JP, Czyglik F, Jouannet P, Gernigon P. Anomalies morphologiques du spermatozo"ide humain. I. Proposition pour un systeme de classification. J Gynecol Obstet BioI Reprod 1975;4(1 SuppIJ: De Almeida M, Soumah A, J ouannet P. Incidence of spermassociated immunoglobulins in infertile men suspected of antisperm autoimmunity. Int J AndroI1986;9: RobertsonL, WolfDP, TashJS. Temporalchangesinmotility parameters related to acrosomal status: identification and characterization of populations of hyperactivated human sperm. BioI Reprod 1988;39: Burkman LJ. Discrimination between non hyperactivated and classical hyperactivated motility patterns in human spermatozoa using computerized analysis. Fertil Steril 1991;55: Frydman R, Forman RG, Belaisch-Allart J, Hazout A, Rainhorn JD, Fries N, et al. Improvements in ovarian stimu- 1ation for in vitro fertilization. Ann NY Acad Sci 1988;541: Wolf JPh, Ducot B, Kunstmann JM, Frydman R, Jouannet P. Influence of sperm parameters on outcome of subzonal insemination in the case of previous IVF failure. Hum Reprod 1992;7: Wolf et al. SUZI with antibody-coated spermatozoa Fertility and Sterility