Thinus F. Kruger, M.D.** Zev Rosenwaks, M.D. Gary D. Hodgen, Ph.D. tt

Transcription

1 FRTILITY AND STRILITY Copyright e 1988 The American Fertility Society Vol. 49, No. 4, April 1988 Printed in U.S.A. The hemizona assay (HZA): development of a diagnostic test for the binding of human spermatozoa to the human hemizona pellucida to predict fertilization potential*t* Lani J. Burkman, Ph.D. Charles C. Coddington, M.D. II Daniel R. Franken, Ph.D.~ Thinus F. Kruger, M.D.** Zev Rosenwaks, M.D. Gary D. Hodgen, Ph.D. tt The Jones Institute for Reproductive Medicine, Department of Obstetrics and Gynecology, astern Virginia Medical School, Norfolk, Virginia The authors present their initial results with the hemizona assay (HZA), which was developed to predict the fertilizing potential of spermatozoa. The HZA uses the matching halves of a human zona pellucida from a nonfertilizable and nonliving oocyte, providing an internal control on zona-to-zona variability. Maximal binding of human sperm to the hemizona usually occurred after 4 to 5 hours of coincubation. Sperm from fertile men exhibited significantly higher binding capacity to hemizonae compared with sperm from men who had fertilization failure during in vitro fertilization (IVF) treatment. The HZA index is calculated as follows: (bound sperm from subfertile male) + (bound sperm from fertile male) X 1. These findings demonstrate that the HZA may be a useful diagnostic tool in male infertility evaluations. Fertil Steril 49:688, 1988 Received October 9, 1987; revised and accepted December 7, * Supported in part by a research grant from Serono Laboratories, Inc., Randolph, Massachusetts. t Presented in part at the Fifth World Congress on In Vitro Fertilization and mbryo Transfer, Norfolk, Virginia, April5 to 1, 1987, and at the Forty-Third Annual Meeting of The American Fertility Society, Reno, Nevada, September 26 to 3, :j: Material presented herein are opinions of the authors and do not represent the opinion of the Department of Defense or the Department of the Nayy. The Jones Institute for Reproductive Medicine, Department of Obstetrics and Gynecology, astern Virginia Medical School. II Present address: Department of Obstetrics and Gynecology, Portsmouth Naval Hospital, Portsmouth, Virginia. If Present address: Department of Obstetrics and Gynecology, Infertility Laboratory, University of Orange Free State, Bloemfontein, Republic of South Africa. Supported by the South African Medical Research Council, ** Present address: Infertility Clinic, Tygerberg Hospital and University of Stellenbosch, Parow, Republic of South Africa. tt Reprint requests: Gary D. Hodgen, Ph.D., Scientific Director, Department of Obstetrics and Gynecology, The Jones Institute for Reproductive Medicine, astern Virginia Medical School, P.O. Box 198, Norfolk, Virginia It has been estimated that male reproductive dysfunction is the primary factor in up to 4% of infertile couples. 1 This high incidence of male factor infertility has promoted an intense search for reliable means to predict human sperm fertilizing potential in vivo and in vitro. For ethical reasons, scientists and physicians have frequently hesitated to perform direct diagnostic functional assessments; that is, binding of human spermatozoa to intact viable human oocytes usually has been regarded as an inappropriate test system. 2 However, reliable and discriminating prognostic assays are needed to determine which infertile men are likely to achieve fertilization in vitro or impregnate their wives when artificial insemination by husband (AIH) is used. Tight binding of human spermatozoa to the human zona pellucida is an early critical event in gamete interaction leading to fertilization and activation of development. This binding step may provide unique information predictive of ultimate sperm fertilizing potential. Because of species spec- 688 Burkman et al. Hemizona assay for human spermatozoa Fertility and Sterility

2 ificity, human spermatozoa will bind firmly to no other zonae pellucida except that of the human. 3 The feasibility of tight human sperm binding to the zonae pellucida of nonliving human oocytes was first examined by Overstreet and Hembree. 2 Sperm binding and zona penetration were tested using oocytes recovered from postmortem ovarian tissue and, later, from ovaries removed for surgical indications.4 These studies, however, did not provide a vigorous investigation of the kinetics of zona binding or the specific relationship between the binding event and fertility in the male. We have examined the feasibility of a new diagnostic assay for tight sperm binding to the human hemizona pellucida. We have cut each oocyte in half using a micromanipulation knife. Three test advantages have resulted from the two matched zona hemispheres: (1) the two halves (hemizonae) created are functionally (qualitatively) equal zona surfaces, allowing a controlled comparison of binding; (2) the very limited number of recovered human oocytes is amplified, since an internally controlled test can be carried out on a single oocyte; and (3) ethical objections to possible inadvertent fertilization of a viable oocyte are eliminated by first cutting the egg into halves. Our findings are reported in three sections. In part I, we examined the feasibility of sperm binding to these hemizonae. For part II, attention was given to elucidating the kinetics of tight sperm binding and then to optimizing measurement of sperm binding to hemizonae. Finally, in part III, we performed initial tests using the hemizona assay (HZA) to distinguish functional differences between the sperm from known fertile men versus men whose sperm had failed to fertilize their wives' oocytes in one or more in vitro fertilization (IVF) treatment cycles. In these latter cases, male infertility was believed to be the limiting factor in achieving fertilization. MATRIALS AND MTHODS All oocytes used were nonliving, with no developmental potential. Oocytes were obtained from two sources: (1) ovarian tissue that was collected postmortem, and (2) donated unfertilized oocytes from the IVF treatment program. In the first instance, the ovarian tissue was excised within 24 hours of death and stored at +4 oc in phosphate-buffered saline (PBS; GIBCO, Grand Island, NY). Between 2 and 48 hours later, manual dissection was carried out following the protocol of Overstreet et al.4 After Vol. 49, No. 4, April 1988 careful mincing of the tissue, zona-intact oocytes denuded of granulosa cells were recovered and placed directly into a 2-M solution of dimethylsulfoxide (DMSO) in PBS. One to four oocytes were transferred with 3 ~1 of DMSO solution to the interior of glass microcapillary tubes (1-J.tl volume; Corning Company, Corning, NY). The tube ends were sealed with Critoseal (Fisher Scientific, Springfield, NJ), and immediately frozen at -7 C. The oocytes were either thawed in a three-step4 or in a one-step process, with no apparent difference on later zona function. The thawed oocyte was rinsed several times in PBS, and cut by micromanipulation into nearly equal halves: the matched hemizonae. A smaller number of immature or postmature oocytes was donated by IVF patients (informed consent was obtained). Such oocytes had been collected by follicular aspiration following ovarian stimulation using exogenous gonadotropin for IVF therapy. 5 About half of these eggs were frozen in the DMSO solution on the same day they became available (24 to 48 hours after aspiration); the remaining eggs were stored up to 72 hours under oil ( +4 C) until they were cut by micromanipulation and then immediately used in sperm binding tests. Again, these differences in handling had no discernible effect on hemizona performance in tight sperm binding. Cutting the Oocyte into Hemizonae by Micromanipulation A complete micromanipulation system (Narishige, Tokyo, Japan) was used for bisecting the eggs. An inverted, phase-contrast microscope (Nikon Diaphot, Garden City, NY) was equipped with a pair of Narishige micromanipulators (model MO 12); the connecting tubing was filled with mineral oil. A micropipette puller (model PP83) was used for initial preparation of the egg-holding pipette from thin-walled capillary tubing (inner diameter =.6 mm; outer diameter =.9 mm; Drummond Scientific, Broomall, PA). The pipette tips were fire-polished and partially closed using a Narishige microforge (model MF-79). When completed, the outer diameter of the tip was appro ximately 1 ~m; the diameter of the opening measured 15 to 2 ~m. Two regions of the pipette were heated and bent at right angles so that the tip could be oriented perpendicularly to the cutting blade. A #11 microscalpel blade was glued to the side of a metal bar (12 em long; 3 mm diameter). The bar was bent to give a sigmoidal shape; the tip of the Burkman et al. Hemizona assay for human spermatozoa 689

3 Figure 1 Two matching hemizonae from a newly cut, nonliving oocyte. The degenerated ooplasm has been pulled away from the hemizonae (total magnification, was X2). bar had a flat, vertical face for attachment of the blade. A 1-mm plastic petri dish (Falcon #25382, Falcon Plastics, Oxnard, CA) served as the cutting chamber. Culture medium (Ham's F-1; GIBCO) or PBS was poured into the dish to a depth of 3 to 4 mm. After the egg-holding pipette and the cutting blade were positioned at the center and bottom of the dish, the egg was transferred to the working area of the dish using a finely drawn glass pipette. The egg was held at the tip of the holding pipette by gentle suction while the blade was centered over the egg. Using a total magnification of 2X, the blade was lowered slowly, first partially flattening the egg then finally initiating a midline cut into the zona. A further lowering of the blade, along with 1 to 2 side-to-side excursions, produced two cleanly cut hemizonae (Fig. 1). Repeated practice using frozen-thawed mouse oocytes proved useful in perfecting this technique. The dense ooplasm inside each hemizona was then dislodged by vigorous pipetting. Only one egg was cut at a time to assure that matched hemizonae remained paired for subsequent sperm binding tests. ach hemizona pair was placed in a separate 5-ttl droplet of medium in a petri dish, covered with mineral oil, and stored at +4 C. Semen Sources The studies in part I used discarded portions of semen from men undergoing semen analysis within the Jones Institute for Reproductive Medicine. Semen aliquots were used only when sperm motility exceeded 4% and sperm concentration was ~4 X 1 6 /ml. For these men, fertility status was unknown. For the kinetics experiments in part II, the semen donors had all fathered a child during the 69 Burkman et al. Hemizona assay for human spermatozoa preceding 24 months. For tests in part III, semen was obtained from husbands who had not achieved fertilization of their wives' oocytes on one or more occasions of IVF treatment within the preceding 12 months. As in part II, the control specimens were from proven fertile men. Seminal sperm concentrations exceeded 28 X 1 6 /ml for all specimens used in part III. The basic semen parameters (sperm concentration and percentage of motile cells) usually were assessed with a computerized videoanalysis system (CASA, Cryo Resources, Ltd, New York, NY). 6 Otherwise, the sperm concentration was calculated using a hemacytometer and percentage motility was assessed by scoring 1 sperm per slide. Sperm Preparation An aliquot of semen (.5 ml) was diluted with 1 ml of Ham's F-1 culture medium, and supplemented with 7.5% heat-inactivated, human fetal cord serum. After centrifugation (5 minutes at 3 X g), the pelleted sperm were washed a second time. The second pellet was overlayed with.5 ml of Ham's F-1 medium and incubated (37 C, 5% C2 in air) to effect a "swim-up" separation. 7 After 1 to 2 hours of incubation, the sperm supernatant was recovered and used for hemizona binding testing. xperiments for Part I xperiment 1: Determining the Precision of Zona Cutting Accuracy in producing matched hemizonae of nearly equal size was assessed for 12 cut oocytes. The concave depth of each cup-shaped hemizona was measured with a reticle and the data were compared within each hemizona pair. xperiment 2: Comparing Sperm Binding of Thawed Postmortem Oocytes Versus Never-Frozen Oocytes Discarded from the In Vitro Fertilization Laboratory Six experiments were carried out using sperm from different donors; a total of 13 intact eggs was studied. A portion of the swim-up supernatant was diluted with Ham's F-1 medium to give a motile sperm concentration of 1,/ml. Two sperm drops (5-ttl volume) were placed under oil; one intact oocyte from postmortem tissue was added to the first drop, while an intact oocyte from IVF treatment was placed in the matching drop. The duration of coincubation at 37 C was 2 to 3 hours. The number of sperm tightly bound to the zona pellucida was counted after vigorous pipetting (5X) Fertility and Sterility

4 using a narrow bore pipette and fresh medium. Due to slight variations in the shape of hand-pulled pipettes, the same rinsing pipette was used throughout a given experiment. xperiment 3: Measuring the Sperm Binding Properties of the Zona with and without Cutting via Micromanipulation We wished to test whether potential release of ooplasmic debris during the cutting process would alter sperm binding. In five experiments, a 5-~Jl droplet of swim-up sperm (25, motile sperm/ ml) was placed under oil. For each experiment, one to two intact zonae and two to three hemizona pairs were placed in the sperm droplet. After 2 to 3 hours of coincubation, the total number of spermatozoa associated with the outer zona surface was first counted in the insemination droplet, before disturbing the intact eggs or hemizonae; thus, sperm with either loose association or tight binding were included. This approach seemed preferable since it was not known whether the spherical intact egg versus the hemizona shell could be pipetted with the same shear force to remove loose sperm. For each experiment, therefore, a single pipette was used to rinse the intact eggs and hemizonae vigorously five times. The number of sperm bound to the inner surface of the hemizona was counted after rinsing. For statistical comparisons between intact zona and hemizona, the number of other hemizona attachments was doubled in each experiment. xperiment 4: Determining Whether Matching Hemizonae Have qual Potential for Sperm Binding For these experiments, each test dish had two matching sperm drops (1, motile sperm/ml). One hemizona was added to the first drop, and the matching hemizona was placed in the second drop. ight matched pairs were tested, using sperm of five different men. Coincubation lasted 2 to 3 hours, after which sperm binding was assessed for the outer zona surface, as described in experiment 3. xperiments for Part II xperiment 5a: Kinetics: Optimizing the Conditions of Sperm Binding to Hemizonae The sperm of four known fertile men were coincubated with hemizonae for periods up to 8.5 hours.8 Seminal aliquots were washed twice with Ham's F-1 medium containing serum and pre- Vol. 49, No.4, April1988 pared for a swim-up incubation. The recovered spermatozoa were diluted to give a final motile concentration of 5,/ml. At least ten sperm drops (two to three drops per time interval) were prepared and covered with oil before adding the hemizonae. A minimum of five matched pairs of hemizonae were used per experiment. Two hemizonae were analyzed per time interval, constituting duplicate tests. The hemizonae were assigned so that matching halves were used at different times. This aspect of the protocol permitted comparison of sperm attachment over time for the same egg. All hemizonae began coincubation at the same time. After periods of 1, 2.5, 4., 5.5, and 8.5 hours, the hemizonae were removed to assess sperm binding. They were pipetted five times to dislodge loosely associated sperm; the number of tightly bound sperm on the outer surface was counted. Care was taken to identify matching hemizonae for later statistical evaluation. xperiment 5b: Characterization of the Sperm Binding Pattern During the First 4 Hours of Coincubation The data from experiment 5a indicated the need for more detailed information on the early coincubation period, specifically the 1- to 4-hour interval. Semen specimens were received from five different fertile donors (four of these had been used in experiment 5a). The swim-up supernatants were obtained as described in the Sperm Preparation section. For each experiment, one sperm droplet was prepared under oil (5, motile sperm/ml) and the two matching hemizonae were added. Coincubation was interrupted briefly at the end of 1, 2, 3, 4, and 5.5 hours, with a final reading on tightly bound sperm numbers at 8.5 hours. At these times, the two hemizonae were quickly recovered and placed in a warm droplet of rinsing medium (Ham's F-1) under oil. The sperm dish was immediately returned to the incubator. Hemizonae were rinsed five times to dislodge loosely associated spermatozoa. While counting, the microscopic stage area was warmed (35 C to 37 C) using a stream of air (Arenberg Sage Air Curtain Incubator, model 279, Jamaica Plain, MA). For each hemizona, the number of sperm bound to the outer surface was counted. The matched pair then was transferred back to the original sperm drop for further coincubation. The counting procedure was carried out during the final 8 to 1 minutes of each assessment interval. All rinsing and counting steps were performed by a single investigator. Burkman et al. Hemizona assay for human spermatozoa 691

5 xperiments for Part III xperiment 6: Determining Whether Sperm Binding to the Hemizona is Correlated with Sperm Fertilizing Ability Ten experiments were performed comparing the hemizona binding capacity for the sperm of eight known fertile men versus eight husbands who had not achieved fertilization of their wives' eggs during one or more IVF treatments. The spermatozoa from the eight infertile men exhibited extremely abnormal morphology ("P" pattern), as assessed with a new, highly selective technique. 9 1 The semen from all 16 men had a sperm concentration exceeding 28 X 1 6 /ml. leven had a percentage motility > 37%; five of the unsuccessful IVF patients had motility ::5 37%. Semen aliquots were washed with the standard Ham's F-1 medium. After a 1-hour swim-up, supernatant spermatozoa were diluted with Ham's F-1 medium, and sperm drops were prepared under oil (25, motile sperm/ml). In each experiment, one half of the hemizona pair was placed with the sperm from a fertile man, while the matching hemizona was coincubated with sperm from a previously unsuccessful IVF husband. In two instances, duplicate experiments were performed for the known fertile and infertile men. The hemizonae were removed after 3 to 4 hours of coincubation, rinsed (as described), and the number of sperm tightly bound to the outer surface was counted. Within each experiment, objective comparisons of the matching hemizona data were made possible by calculating the ratio for the HZA index of tightly bound sperm, HZA index: Part I number of infertile sperm bound :----''----:---:- X 1. number of fertile sperm bound xperiment 1 RSULTS Bisecting of the zona by surgical micromanipulation showed small deviations from an exact 5/5 cut. When the sizes of matching hemizona halves were compared, the mean difference in the depth of the concave hemizona shells was 1 ± 2.% (median difference = 8.1%). Interestingly, individual unmatched hemizona varied between 7 and Burkman et al. Hemizona assay for human spermatozoa J,Lm in depth, reflecting the unequal sizes of respective intact oocytes before they were cut. xperiment 2 The use of oocytes from postmortem tissue versus eggs discarded from IVF did not significantly influence sperm-binding potential of the zona pellucida (Fig. 2). After vigorous pipetting, the number (mean ± standard error of the mean [SM]) of firmly bound spermatozoa was not statistically different for the intact postmortem zona (5. ± 1.9, n = 6) versus the intact IVF zona (5.4 ± 2.1, n = 7). xperiment 3 When assessed before pipetting, total sperm association (loose and tight binding) with the outer surface of cut hemizonae was not different from that observed for the intact zonae (Fig. 3). When the total number of associated spermatozoa was doubled for all hemizonae, the means were not statistically different: hemizonae (57 ± 16, n = 12) versus the intact egg (64 ± 21, n = 8). Some tight sperm binding did occur on the inner zona surface of the hemizona. Following rinsing to remove loosely bound sperm, the mean number attached to the outer surface was 1.4 ± 5.1 (mean ± SM) compared with 1.6 ±.8 on the inner surface (n = 8). xperiment 4 Matching hemizonae showed no detectable difference in their capacity for tight sperm binding to the outer zona surface. When the hemizonae with the larger number of bound sperm were assigned to 8 (7) >- (6) += ~Q;6 :;::a. C/) - ' 4 "-Q) Q)_c.(,) «~ 2 :::::1- Ztn Postmortem IVF Figure 2 Comparison of the number of tightly bound (attached) sperm to the intact zona of a postmortem egg versus the intact zona of an IVF egg (experiment 2; mean± SM; number of zonae shown in parentheses). Fertility and Sterility

6 "U Q) -«S u en en ~ O(j; '-a. ~en 4 ::J c: 2 (12) (8) HZ Intact HZ data doubled for better comparison , " c: ::J. 1 ~ g 75 ;;;. ::J z 5 25 Fertile mele whh high-binding Fertile mele with law-binding Figure 3 Comparison of total number of loosely and tightly associated sperm for a hemizona (HZ)* versus an intact zona (experiment 3; mean ± SM; number of zonae shown in parentheses). group 1, and the matching hemizonae assigned to group 2, there was no statistically significant difference between the group means (3.5 ±.6 and 2.6 ±.4, n = 8). Part II xperiment 5a The spermatozoa from four known fertile men showed similar trends for binding to the hemizonae over the 8.5-hour period. The kinetic curves for two representative experiments are depicted in Figure 4, illustrating a high and a low binding pattern. The differences in binding capacity between fertile men, as shown here, may reflect variability in the efficiency of sperm binding to the zona. Uniformly, tightly bound sperm were present on the hemizona surface within 1 hour of coincubation. Here, the number (mean± SM) of bound sperm was 39.1 ± 12 (four men, nine hemizona pairs). In this study, maximal binding occurred at 3.5 hours for one man, 4. hours for two others, and at 5.5 hours for the fourth donor. Interestingly, there was a consistent decline in the number of sperm tightly bound to the hemizonae at the first observation time beyond the binding peak. Thereafter, the numbers for tightly bound sperm remained almost constant through 8.5 hours. The change in sperm binding over time between matching hemizonae also was assessed in this first kinetics study. Paired data were analyzed for three coincubation intervals (1 versus 4 hours, 2.5 versus 5.5 hours, and 4 versus 8.5 hours, with four to six Vol. 49, No.4, April Hours of Coincubation Figure 4 The kinetics for tight binding of sperm from fertile men to the hemizonae of different eggs during coincubation of 1 through 8.5 hours: representative curves for a high-binding male (closed boxes) versus a low-binding male (closed circles) (experiment 5a; mean± SM). hemizona pairs per interval). The paired data are illustrated in Figure 5. During the 1-versus-4-hour interval, all binding slopes rose dramatically. The 215 I 14 ~ 12 Q.., " c: ::J. ~ g ;;; ::J z Duration of Coincubation (hoursl Figure 5 Time-dependent changes in binding of spermatozoa from fertile.men when matched hemizonae pairs were evaluated over three intervals in the kinetics study (experiment 5a). Burkman et al. Hemizona assay for human spermatozoa 693

7 12 1 :. a. "' u c ::> D 8 >- : 6, :;:: i;; 4 D ::> z 2 o Matching HZ lor Male A o Matching HZ lor Male B 6 Matching HZ lor Male C servations made at 1, 2, 3, 4, 5.5, and 8.5 hours, the respective pooled values were 13.9 ± 6, 33.3 ± 11, 63.5 ± 34, 94.9 ± 49, 88.2 ± 42, and 79.6 ± 4. The calculated 95% confidence interval 11 (two-tailed test, degrees of freedom [df] = 11) for the 4-hour data was 6.3 to tightly bound sperm; these values compare well with the matching 4-hour confidence interval from our first kinetics study (75.4 to 122.5; experiment 5a). Part III Hours of Coincubation Figure 6 Tight sperm binding to matching hemizonae pairs for three fertile males. In each assay, the same hemizonae pair was used to evaluate sperm binding at all time points indicated. Open circles, matching HZ for male A; open boxes, matching HZ for male B; open triangles, matching HZ for male C. (experiment 5b; mean± SM). overall increase was 17.5 ± 4.4 sperm/hour (mean ± SM). Between 2.5 versus 5.5 hours, the rise was not sustained; in four of five cases, the slopes were shallow, or declining, coinciding with peak binding during this period (range, -8. to +5.3 sperm/hr). One case had a late peak in binding at 5.5 hours, and displayed the one sharply rising slope during this period. Uniformly, relative sperm binding to matched hemizonae declined between 4 and 8.5 hours (-1.6 ± 4.8 sperm/hr; mean± SM). xperiment 5b Representative sperm binding curves from three of six experiments are shown in Figure 6. In all six experiments, the means for the matched pairs of hemizonae revealed that a consistent peak in sperm binding was reached after approximately 4 hours of coincubation, after which there was a slight decline through 8.5 hours. Importantly, by the use of frequent counts, our observations typically revealed a nearly linear increase in tight sperm binding between 1 and 4 hours. Reproducibility of binding for matched hemizonae was high. In five of the six cases, differences between paired hemizonae ranged from 8% to 2% for the number of sperm bound at 4 hours. The paired curves showed similar slopes for the rising phase of sperm binding. The data then were pooled across all six experiments; the mean number of sperm bound (mean ± SM) at each hour was calculated. For the ob- 8.5 xperiment 6 Using matched hemizona pairs in all ten experiments, the spermatozoa from known fertile men bound to the hemizona more efficiently than sperm from IVF husbands who had not achieved fertilization of their wives' eggs in vitro (Table 1). For this limited patient group, the mean value for the HZA index was 21 ± 8 (mean± SM), and 95% of the individual values fell below 62. The observations reported here are insufficient for the firm identification of an index threshold. However, our preliminary data suggest 62 as a tentative threshold, such that lower HZA indices may be prognostic of very low to nil fertilizing potential. The overall number (mean ± SM) of sperm bound was significantly different for the two groups tested (34. ± 8.1 [fertile] versus 5.9 ± 2.3 [infertile]; P <.1). DISCUSSION The zona pellucida of mammalian oocytes is a critical site for sperm-egg interaction during initiation of the fertilization process. Unless sperm are Table 1 Comparison of Number of Sperm Tightly Bound to Pairs of Matched Hemizonae for Proven Fertile Men Versus Infertile Men After 4 Hours of Coincubation No. of tightly bound sperm HZA index Infertile Fertile Infertile/fertile xperiment no. men men X Burkman et al. Hemizona assay for human spermatozoa Fertility and Sterility

8 capable of binding tightly to the zona pellucida, the subsequent steps of zona penetration, fusion with the oolemma and, ultimately, egg activation are impossible Whereas the zona-free, hamster penetration assay14 15 can demonstrate completion of the human sperm acrosome reaction and sperm/ oolemma membrane fusion, only tests using the human zona pellucida can assess the capacity of human sperm for binding to the intact egg and for penetration of the zona. The potential of zona pellucida tests for evaluating sperm function has been limited by access to human ovaries and the small number of oocytes recovered per ovary. Because of differences in follicular and oocyte maturation, it was anticipated that individual intact zonae would demonstrate significant variability with respect to sperm binding.16 More recently, the availability of extra oocytes from IVF treatment programs has provided a new source of oocytes for laboratory studies. Reports of storage and use of nonliving, unfertilized human oocytes for sperm testing have appeared in the literature for a decade now. Overstreet and Hembree demonstrated the recovery of immature human oocytes from postmortem ovarian tissue, followed by egg storage in a DMSO solution.2 Subsequent studies have tested human sperm binding to human zonae that were obtained from surgically excised ovaries Yanagimachi et atl 9 also recovered oocytes from fresh, surgically derived ovaries; these oocytes were stored in a concentrated salt solution, and later used in a sperm penetration test. These reports, however, did not provide specific studies on: (1) the temporal characteristics of sperm/zona binding in the human, (2) the use of zona binding as a reliable assay endpoint, and (3) the careful assessment of male fertility in relationship to abnormal zona binding. Thus, the present data more comprehensively demonstrate specific temporal and functional characteristics of tight binding of human sperm to the zona pellucida. Here, we have reported the test conditions which demonstrate the initial feasibility of the HZA for predicting the fertilizing potential of human spermatozoa. 2 Having examined the kinetics of sperm binding to the hemizona, in search of an optimal coincubation milieu, we then demonstrated that sperm from known fertile men exhibited significantly (P <.1) higher HZA binding compared to sperm from men seeking treatment by IVF.21 In the initial studies (part I), it was found that accurate bisecting of the human zona pellucida was not difficult. Hence, differences in the numbers of sperm bound to matching hemizonae were not due to dissimilar surface areas. Most probably, the large number of potential sperm binding sites on the zona surface makes the small differences in hemizona size insignificant. Second, it is noteworthy that there were no obvious differences in sperm binding capacity of postmortem oocytes versus extra IVF oocytes matured in vitro. A similar finding has been reported by Cross et al.,22 in that no differences were detected in tight binding of mouse spermatozoa to the mouse zonae pellucida of immature follicular oocytes versus ovulated oocytes. In the third experiment, it was clear that extrusion of the ooplasm during the cutting process did not adversely affect the HZA by coating the hemizona surface with interfering substances. The observation of limited tight sperm binding on the inner surface of the zona was not surprising, in view of the recent report from Ahuja and Bolwell.23 They, too, identified sperm-binding components on the inner surface of hamster zonae pellucida, although at concentrations well below that detected on the outer zona surface. Finally, the results from experiment 4 gave crucial evidence that establishing an assay using hemizona pairs was statistically warranted, as would be required for a reliable assay. Here, the matching hemizona halves did show equivalent capacities for sperm binding, thus confirming our proposed use of the matched hemizona halves as an internal control against the known variable sperm-binding capacities of different eggs. In our first kinetics study (part II), the shape of the sperm binding curve was consistent among known fertile donors. Following 1 hour of the swim-up procedure plus 1 hour of zona coincubation (capacitation time of approximately 2 hours), tight sperm binding to the hemizona had begun, but was not yet maximal. Substantial tight binding within 1 hour of coincubation also has been reported in an ultrastructural study of IVF human oocytes16 and for the in vitro penetration of nonliving intact oocytes.18 In experiment 5, the spermatozoa from fertile men exhibited maximal binding capacity after approximately 4 to 5 hours of coincubation. These findings are consister-;.:; with an earlier kinetics study,8 24 where we showed that human spermatozoa capacitated in Ham's F-1 medium generally exhibited peak hyperactivated motility after 3 to 6 hours of total incubation. These cumulative data support the current opinion that hyperactivated motility may be initiated be- Vol. 49, No.4, April1988 Burkman et al. Hemizona assay for human spermatozoa 695

9 fore the acrosome reaction and concomitant zona binding. 12 Our studies here showed evidence for individual variation in the time course of maximal sperm binding to the human zona pellucida among known fertile men. The hourly observations made in the second study demonstrated the nearly linear increase in sperm binding to hemizonae through 4 hours of coincubation. These results confirmed that maximal sperm binding to the hemizona was achieved after approximately 4 hours. For the purpose of the HZA, therefore, it appeared that 4 hours of coincubation was statistically optimal. These kinetic data may suggest that capacitation and fertilizing ability have peaked for most sperm of fertile men within 5 hours of sperm incubation, consistent with two other reports on human IVF data By internally controlling for inherent zona variability through use of matched hemizona pairs, the sperm binding curves can accurately reflect events that are truly sperm-dependent. In experiment 5b, the serial rinsing and counting at each observation interval may have introduced some small artefact, yet the progressive use of each hemizona uniquely demonstrated that the kinetic curves for matching pairs of hemizona were typically very similar. The data from the two kinetic studies are complimentary to but much more extensive than the information provided in an earlier study/ 7 where human sperm binding to intact zonae was examined. For Singer and co-workers/ 7 the experimental design focused on the ability of progressively aging sperm to initiate new binding to the zona surface during a fixed 4-hour coincubation period. Our kinetics study differed in that progressively aging sperm were examined throughout continuous incubation with the hemizona from 1. to 8.5 hours. Our findings suggest that zona binding of spermatozoa was maximal after 4 to 5 hours of coincubation. The decrease in bound sperm at the subsequent time point may be due to four events: (1) some depletion of the bound population as sperm penetrated through the hemizona; (2) loss of some bound sperm which exhibit vigorous flagellar movement (hyperactivation) and pull free from the zona surface 26 ; (3) a decline in the number of sperm expressing initiation of tight sperm binding; and (4) a decline in the efficiency of the hemizona surface to act cooperatively in tight sperm binding. The plateau in bound sperm numbers observed after approximately 5 hours of coincubation may be due to any or all of these factors acting in concert. Among infertile men, zona binding events may be affected by abnormalities in the sperm itself or interfering constituents of seminal fluid, such as microorganisms or antibodies. Our data and those of Singer et au 7 emphasize that aged sperm (;;;::.7 hours of capacitation), having diminished capacity to bind tightly to the zona pellucida, may not be suitable for HZA evaluation. Similarly, we anticipate that some eggs will have such poor binding characteristics as to have no discriminating power for testing of sperm binding. The data presented must be interpreted cautiously with respect to the limited number of experiments completed. Reproducibility in the HZA is presently being examined in two studies: simultaneous testing of a given semen specimen with three pairs of matching hemizonae, and the repeated testing of a given man over several months. The kinetics of hemizona binding for subfertile men also will be assessed. Finally, a large cohort of men from the Norfolk IVF program is being evaluated using the HZA in order to extend the limited data offered here. Acknowledgments. We thank James Swanson, Ph.D., and his staff in the Andrology Laboratory, for their cooperation in obtaining semen samples from specific patients. RFRNCS 1. Glass RH: Infertility. In Reproductive ndocrinology, Physiology, Pathophysiology and Clinical Management (2nd edition), dited by SC Yen, RB Jaffe. Philadelphia, WB Saunders, 1986, p Overstreet JW, Hembree WC: Penetration of the zona pellucida of nonliving human oocytes by human spermatozoa in vitro. Fertil Steril 27:815, Bedford JM: Sperm/egg interaction: the specificity of human spermatozoa. Anat Rec 188:477, Overstreet JW, Yanagimachi R, Katz DF, Hayashi K, Hanson FW: Penetration of human spermatozoa into the human zona pellucida and the zona-free hamster egg: a study of fertile donors and infertile patients. Fertil Steril 33:534, Rosenwaks Z, Muasher SJ: Recruitment of fertilizable eggs. In In Vitro Fertilization Norfolk, dited by HW Jones Jr, GS Jones, GD Hodgen, Z Rosenwaks. Baltimore, Williams & Wilkins, 1986, p 3 6. Mathur S, Carlton M, Ziegler J, Rust PF, Williamson HO: A computerized sperm motion analysis. Fertil Steril 46:484, McDowell JS: Preparation of spermatozoa for insemination in vitro. In In Vitro Fertilization Norfolk, dited by HW Jones Jr, GS Jones, GD Hodgen, Z Rosenwaks. Baltimore, Williams & Wilkins, 1986, p Burkman LJ: Temporal pattern of hyperactivation-like motility in human spermatozoa. Bioi Reprod 34(suppl 1):226, Kruger TF, Menkveld R, Stander FSH, Lombard CJ, Van 696 Burkman et al. Hemizona assay for human spermatozoa Fertility and Sterility

10 der Merwe JP, van Zyl JA, Smith K: Sperm morphologic features as a prognostic factor in in vitro fertilization. Fertil Steril 46:1118, Kruger TF, Acosta AA, Simmons KF, Swanson RJ, Matta JF, Oehninger S: Predictive value of abnormal sperm morphology in in vitro fertilization. Fertil Steril 49:112, Remington RD, Schork MA: Statistics with Applications to the Biological and Health Sciences. nglewood Cliffs, NJ, Prentice-Hall, 197, p Yanagimachi R: Mechanisms of fertilization in mammals. In Fertilization and mbryonic Development In Vitro, dited by L Mastroianni Jr, JB Biggers. New York, Plenum Press, 1981, p Gould J, Overstreet JW, Yanagimachi H, Yanagimachi R, Katz DF, Hanson FW: What functions of the sperm cell are measured by in vitro fertilization of zona-free hamster eggs? Fertil Steril 4:344, Yanagimachi R, Yanagimachi H, Rogers BJ: The use of zona-free animal ova as a test system for the assessment of the fertilizing capacity of human spermatozoa. Bioi Reprod 15:471, Yanagimachi R: Zona-free hamster eggs: their use in assessing fertilizing capacity and examining chromosomes of human spermatozoa. Gamete Res 1:187, Chen C, Sathananthan AH: arly penetration of human sperm through the vestments of human eggs in vitro. Arch Androl 16:183, Singer SL, LambertH, Overstreet JW, Hanson FW, Yanagimachi R: The kinetics of human sperm binding to the human zona pellucida and zona-free hamster oocyte in vitro. Gamete Res 12:29, LambertH, Overstreet JW, Morales P, Hanson FW, Yanagimachi R: Sperm capacitation in the human female reproductive tract. Fertil Steril 43:325, Yanagimachi R, Lopata A, Odom CB, Bronson RA, Mahi CA, Nicholson GL: Retention of biologic characteristics of zona pellucida in highly concentrated salt solution: the use of salt-stored eggs for assessing the fertilizing capacity of spermatozoa. Fertil Steril 31:562, Coddington CC, Burkman LJ, Hodgen GD: Hemi-zona pellucida for assessing human sperm function (Abstr). Presented at the Fifth World Congress on IVF/T Norfolk, Virginia, April5-1, Published by The American Fertility Society, program supplement p Burkman LJ, Coddington CC, Kruger TF, Hodgen GD: Hemi-zona assay (HZA): is human sperm binding to the zona pellucida predictive of in vitro fertilizing potential? (Abstr). Presented at the 43rd annual meeting of The American Fertility Society, Reno, Nevada, September 28-3, Published by The American Fertility Society, program supplement p Cross NL, Lambert H, Samuels S: Sperm binding activity of the zona pellucida of immature mouse oocytes. Cell Bioi Rep 1:545, Ahuja KK, Bolwell GP: Probable asymmetry in the organization of components of the hamster zona pellucida. J Reprod Fertil 69:49, Burkman L: xperimental approaches to evaluation and enhancement of sperm function. In In Vitro Fertilization Norfolk, dited by HW Jones Jr, GS Jones, GD Hodgen, Z Rosenwaks. Baltimore, Williams & Wilkins, 1986, p Lopata A, McMaster R, McBain J, Johnston 1: In vitro fertilization of preovulatory human eggs. J Reprod Fertil 52:339, Burkman LJ, Coddington CC: Interaction of hyperactivated-like human spermatozoa with the human zona pellucida (Abstr). Presented at the forty-second annual meeting of The American Fertility Society, Toronto, Canada, September 27 to October 2, Published by The American Fertility Society, program supplement p 128 Vol. 49, No.4, April1988 Burkman et al. Hemizona assay for human spermatozoa 697