Efficiency of sex pre-selection of spermatozoa by albumin separation method evaluated by double-labelled fluorescence in-situ hybridization

Transcription

1 Human Reproduction vol.12 no.9 pp , 1997 Efficiency of sex pre-selection of spermatozoa by albumin separation method evaluated by double-labelled fluorescence in-situ hybridization Ming-Jer Chen, Hwa-Fen Guu and have failed to demonstrate the effect of Y sperm enrichment Esther Shih-Chu Ho 1 by albumin separation (Evans et al., 1975; Ross et al., 1975; Brandriff et al., 1986). Department of Obstetrics and Gynecology, Taichung Veterans General Hospital, No. 160, Section 3, Taichung Harbor Road, Evaluation of the real efficacy of various techniques of Taichung, 40705, Taiwan sperm separation has been made feasible since the introduction 1 To whom correspondence should be addressed of the quinacrine staining (F-body) method by Barlow and Vosa (1970). However, quinacrine does not enable evaluation To evaluate the separation efficiency of Ericsson s two- and of the real efficacy of sperm separation procedures because it three-layer albumin separation methods, semen samples is unreliable (Ueda and Yanagimachi, 1987; Van Kooij and from 21 healthy males were studied. Seven patients already van Oost, 1992). Improved identification of Y spermatozoa had two or more sons, another seven had two or more through chromosome evaluation rather than by F-body identidaughters and the other seven had primary infertility due fication is critical to provide more accurate information. to female factors. The semen samples were divided into Previous efforts of karyotypic analysis required the use of the three aliquots: one remained unprocessed initially, the human sperm/hamster egg system which precluded larger other two aliquots went through two- and three-layer scale evaluation (Brandriff et al., 1986). Recently double albumin separation methods respectively. All samples were fluorescence in-situ hybridization (FISH) techniques have been then stained with X Y double staining probes. In each used to detect sex chromosomes in decondensed human sperm group, four or five samples were processed at room temper- nuclei. X- and Y-chromosome probes could be applied simulature, and two or three at body temperature (37 C). The taneously with high labelling efficiency (Han et al., 1993a). labelling efficiency of X Y double staining probe was over This technique could thus facilitate clinical studies on the 99%. The X:Y sperm ratios were even in the original efficiency of sperm separation (Han et al., 1993b; Wang et al., samples. The ratios of the X and Y spermatozoa were 1994a). Cran and Johnson (1996) have recently reviewed the altered slightly but significantly after the two-layer (P pre-determination of sex in various domestic animals and 0.05) or the three-layer (P 0.005) separation. The in the human using sperm samples enriched for X- or Y- alterations occurred only at room temperature. The X chromosome-bearing spermatozoa obtained by flow cytometry spermatozoa increased and the Y spermatozoa decreased, and cell sorting. both to a small degree of difference ( %). Double Historically, four protocols for sperm separation by albumin fluorescence in-situ hybridization analysis therefore columns have been used in the literature (Beernink et al., showed that albumin separation methods do not enrich Y 1993). Protocol 1 is not for sex pre-selection; protocol 2 has spermatozoa. been used for female sex pre-selection in conjunction with Key words: albumin/fish/sex pre-selection/sperm separation clomid citrate treatment of the female; protocol 3 and modified method/y spermatozoa protocol 3 are used for male sex pre-selection. Modified protocol 3 is used only in cases with total sperm counts of It is not known why the three-layer two-step protocols 3 and modified 3 alter the sex ratios to male Introduction preponderance, but the two-layer one-step protocol 2 changes Sex pre-selection is sometimes justified by medical considerations. the ratio to female preponderance. A recent debate relates the The male:female sex ratio at birth is 106:100, and could separation effect to recovery rates of spermatozoa (Ericsson, be affected by demographic factors as well as by other clinical 1994). In this study, we applied the double FISH technique to or laboratory manipulations (Zarutskie et al., 1989). Since the determine the efficacy of sex pre-selection by two of the invention of Ericsson s method of Y sperm enrichment in 1973 albumin methods (protocols 2 and 3) at two incubation (Ericsson et al., 1973), laboratory sperm separation has become temperatures (room temperature and body temperature). clinically relevant. Ensuing methods for sperm separation have been proposed and also reported to effectively alter the natural sex ratio (Steeno et al., 1975; Kaneko et al., 1983; Check and Materials and methods Katsoff, 1993). Of these, modifications of Ericsson s original Source of the semen, semen analysis, and experimental design method for the purpose of Y sperm selection have gained in Semen samples were obtained from 21 normal healthy donors: seven clinical popularity and are claimed to be very effective clinic- from patients who asked for a Y sperm selection due to already ally (Beernink et al., 1993). However, some laboratory studies having two or more daughters without son (ask-for-y group); seven 1920 European Society for Human Reproduction and Embryology

2 Lack of Y-spermatozoa enrichment by albumin separation Table I. Semen parameters of the donors before and after separation Group After separation Before separation Protocol 2 Protocol 3 Count Total count a Motility Count Total count Motility Count Total count Motility (%) (%) (%) Control group 1 RT RT RT RT BT BT BT Ask-for-Y group 1 RT RT RT RT RT BT BT Ask-for-X group 1 RT RT RT RT BT BT BT Mean SD a Initial total counts must be further divided into halves because semen samples were divided to serve two separation procedures. RT: room temperature; BT: body temperature (37 C). from patients who already had two or more sons without daughter centrifuged at 300 g, and resuspended in Tyrode s solution and then (ask-for-x group); and another seven from infertile couples with relayered in 0.5 ml over 1.0 ml 12.5% HSA over 0.5 ml 20% HSA proven female infertility (control group). Donor semen specimens for 1.5 h at room temperature (four or five cases in each group) or were obtained by masturbation after 2 4 days of sexual abstinence body temperature (two or three cases in each group). and studied after complete liquefaction, and within 1hofcollection. At the end of the incubation period, the bottom layers of each Initial semen analyses were carried out according to protocols column were pooled, after removing the top layers, diluted with recommended by the World Health Organization manual (WHO, Tyrode s solution 1:1, and centrifuged for 10 min at 300 g. The 1992). The results are shown in Table I. The semen samples were supernatant was discarded, and the pellet was resuspended in Tyrode s then divided into three aliquots, one little aliquot was used as control solution. The sperm concentration was adjusted to range from 5 to and received the double FISH stain without sperm separation. The motile spermatozoa/ml. The results were as shown in Table other two equal aliquots were treated by either the two-layer one- I. Recovery rates (RR, Table II) were calculated as the percentage of step albumin separation method (protocol 2) or the three-layer twostep the final total motile spermatozoa recovered from the original total method (protocol 3) respectively and then the final sperm motile spermatozoa, i.e. (total count) (% motility) after separation suspensions were prepared for the double FISH analysis. divided by (total count) (% motility) before separation, all divided by two since the original sperm samples were split into halves, each Procedures of the albumin separation method proceeding to one of the two protocols. The initial total counts of The methods as described by Beernink et al. (1993) were followed. samples undergoing each separation procedure were in Briefly, aliquots of semen were first diluted with equal volumes (1:1) 17 of the 21 samples (Table I). of Tyrode s solution (Sigma Chemical Co., St Louis, Missouri, USA) and then centrifuged for 10 min at 300 g. After centrifugation, the Double FISH procedure spermatozoa were resuspended in the same medium at a concentration Pretreatment of sperm (sperm nuclei decondensation) of ~ spermatozoa/ml. Multiple albumin columns were prepared The control and processed spermatozoa were spread on clean slides for each sample. and air-dried. The slides were fixed in methanol:acetic acid (3:1) For protocol 2: 0.5 ml of washed spermatozoa were layered over (Sigma) for 30 min. Then the slides were washed in 2 concentrated 1.0 ml of 7.5% human serum albumin (HSA) (Sigma) over 0.5 ml standard saline citrate solution (SSC; 0.15 M NaCl/0.015 M sodium of 17.5% HSA for 1.5 h at room temperature (four or five cases in citrate) (Vysis Inc., Framingham, MA, USA), and dehydrated through each group) or body temperature (two or three cases in each group). an ethanol series. Sperm nuclei were decondensed by slide incubation For protocol 3: 0.5 ml of washed spermatozoa were layered over at 37 C in 50 mm Tris buffer (ph 7.4) (Sigma) containing 5 mm 1.0 ml of 7.5% HSA for 1 h. The lower HSA layer was aspirated, dithiothreitol (DTT) (Sigma) and 1% Triton X-100 (Sigma) for 60 min. 1921

3 M.-J.Chen, H.-F.Guu and E.S.-C.Ho Table II. Y and X labelling results for the albumin separation methods Group Before separation Protocol 2 Protocol 3 Labelled Y X Labelled Y X RR Labelled Y X RR (%) Control group (n ) 1 RT RT RT RT BT BT BT Ask-for-Y group (n ) 1 RT RT RT RT RT BT BT Ask-for-X group (n ) 1 RT RT RT RT BT BT BT Mean SD RR (recovery rate) percentage of motile spermatozoa recovered from the original total motile spermatozoa; n no. of sperm nuclei scored; RT room temperature; BT body temperature. DTT-treated slides were examined by phase contrast microscope Red fluorescein isothiocyanate (FITC) triple band pass filter. 500 for evidence of head decondensation. Slides were washed in 2 spermatozoa were scored from each slide by one observer concentrated SSC, dehydrated again through an ethanol series, air- (H.-F.G). Randomly selected areas of the slides were screened and dried and were then treated with RNase A100 µg/ml (Stratagene, photographed. The sperm cells were identified by their unique tails St Diego, CA, USA) in 2 SSCfor1hat37 C, and washed four and/or oval nuclei morphology. The scoring criteria were as follows: times in 2 SSC before hybridization (Vidal et al., 1993; Chevret only sperm cells with a distinct single orange or green signal domain et al., 1994). were classified as haploid X- or Y-bearing spermatozoa, respectively; Probe/slide preparation and assay procedure whereas cells that contained more than one signal domain were documented separately as anhaploid cells. In case the situation of The probes used were Spectrum CEP chromosome enumeration split signal domains were encountered, efforts were made to look system: SOX/SGY (Vysis), which consist of chromosome specific carefully for any bridge strands between the signals. Only obvious sequences from highly repeated human satellite DNA directly labelled split signal domains with comparable brightness, a minimal distance with Spectrum Orange fluorophores for the X chromosome and of one signal domain and without connecting strands would enable Spectrum Green fluorophores for the Y chromosome. the classification of the cells as anhaploid (Martin et al., 1995). In brief, slides were denatured in 70% formamide/2 SSC (Vysis) Certain sperm nuclei were eliminated from scoring, such as those for 5 min at 75 C, rinsed through cold 70, 80 and 100% ethanol with confusing overlapping nuclei, non-intact nuclei, and very large solutions serially and then briefly kept on a 45 C slide warmer. For nuclei. Labelling efficiency was calculated as the ratio of sperm cells each slide µl of CEP (chromosome enumeration DNA probes) probe with fluorescent orange and/or green spot signals to the total number was mixed thoroughly with 7 µl CEP hybridization buffer (Vysis) of sperm cells counted in each slide. and 2 µl distilled water. The mixture was denatured at 75 C for 5 min, and then applied within 2 min (of the slide preparation) to each slide under a coverslip and sealed with rubber cement (Avery Statistical analysis Dennison, Framingham, MA, USA). All statistical analyses were performed using SPSS-PC program Hybridization was allowed to proceed for h in a pre-warmed (SPSS Inc. Chicago, IL, US). Non-parametric analysis with humidified box in a 42 C incubator. After removal of rubber cement Wilcoxon s matched-pairs signed-ranks test for checking the signific- and coverslips, the slides were washed at 45 C for 30 min in 50% ance of changes in % X spermatozoa, % Y spermatozoa and ratios formamide/2 SSC, 10 min in 2 SSC and 5 min in 2 SSC/0.1% of the Y and X spermatozoa before and after separation. Kruskal NP-40 (Vysis). Slides were counterstained with 10 µl DAPI II (Sigma). Wallis one-way analysi of variance (ANOVA) for checking the Slides were examined at 1000 with a microscope (Labphot, differences between subgroups (of semen sources and recovery rates). Nikon, Japan) equipped with epifluorescence and a DAPI Texas P 0.05 was considered significant. 1922

4 Lack of Y-spermatozoa enrichment by albumin separation Table III. Comparison of the separation effects by different incubation temperatures Before separation (%) After separation Protocol 2 Protocol 3 (%) P a (%) P b Room temperature group (n 13) Y X RR ( ) ( ) Body temperature group (n 8) Y X RR (range) (9.0 30) ( ) a Comparing results from protocol 2 with original semen. b Comparing results from protocol 3 with original semen. The P-values were calculated on the changes in the Y:X ratios before and after separation. RR recovery rate; all values are presented as mean SD (range). Results in protocol 3. This contributed to a significant change in the The labelling efficiencies and counted ratios of the Y and X Y:X ratio (P 0.04) in this subgroup. Kruskal Wallis onespermatozoa for each individual case is listed in Table II. In way ANOVA test revealed no differences in Y:X ratios between total spermatozoa nuclei were carefully evaluated. The the three sources in the original semen as well as in the post- overall labelling efficiencies were 99% (99.11% in the separation samples. original samples, 99.18% in protocol-2 samples and 99.17% Analysis with regard to the effect of motile spermatozoa in protocol-3 samples). The Y:X sperm ratio in the original recovery upon the separation efficiencies at room temperature samples was even (49.14:49.97), no matter whether the samples group (13 cases) was shown in Table V. There were no were from control, ask-for-y or ask-for-x group (48.69:50.37, significant changes in Y:X ratios before and after separation 50.22:48.93 and 48.51:50.60 respectively). except that a marked reduction in the Y spermatozoa was As shown in Table III, the separation effects were only noted in the moderate recovery rate (5 10%) subgroup (P evident at room temperature (13 cases, sperm nuclei 0.03) in protocol 3. This contributed to a significant change were evaluated). There was a significant change in the Y:X in the Y:X ratio (P 0.03) in this subgroup. A trend toward ratios in protocol 2 (P 0.03) and there was an even more X enrichment was also detected when the recovery rate was significant change in protocol 3 (P 0.002). Both protocols 5% in protocol 3 (P 0.07). Kruskal Wallis one-way showed changes toward Y sperm deprivation and X sperm ANOVA test revealed no differences in Y:X ratios between enrichment. The Y spermatozoa decreased and X spermatozoa the three or four different recovery rate subgroups in the increased significantly, but both only to a small degree (1.4 original semen as well as in the post-separation samples. 3.5%). The effects were not seen in the body temperature The frequencies of aneuploid spermatozoa encountered are group. listed in Table VI. The anhaploid sperm count was 0.21% in The mean recovery rate of motile spermatozoa was 7.0% the original samples, 0.071% in the protocol-2 samples and (range %) in protocol 3 and 9.3% (range %) 0.089% in the protocol-3 samples. in protocol 2. Incubation at body temperature increased the recovery rates to 8.7% (range %) in protocol 3 and 17.5% (range %) in protocol 2. Discussion If we carefully scrutinize the Y:X ratios of the room Since Ericsson s first report in 1973 demonstrated that human temperature cases in Table II, we could find that the ratios Y spermatozoa can be enriched with the use of albumin were evenly distributed before separation (seven of 13 cases columns (Ericsson, 1973), numerous laboratory and clinical above unity and six of 13 cases below unity), but only two studies have been published regarding the separation efficiency (nos. 2 and 4 of the control group) of all 13 cases had ratios of these methods (Evans et al., 1975; Ross et al., 1975; above unity after protocol 2 separation, and all cases had Dmowski et al., 1979; Quinlivan et al., 1982; Brandriff et al., values less than unity after protocol 3 separation. 1986; Beernink et al., 1993; Vidal et al., 1993; Pyrzak et al., Table IV shows further analysis of the effect of semen 1994; Wang et al., 1994a). However, the reported results are source upon the separation efficiencies at room temperature. rather confusing and often contradictory. There were no significant changes in Y:X ratios before and Although the best validation of the separation efficiency is after separation except that a marked reduction in the Y the sex ratio of the offspring, there has been a lack of controlled spermatozoa was noted in the ask-for-y subgroup (P 0.04) studies on the sex ratio of newborns following the use of 1923

5 M.-J.Chen, H.-F.Guu and E.S.-C.Ho Table IV. Comparisons of the separation effects at room temperature by different semen source a Before separation (%) After separation Protocol 2 Protocol 3 (%) P a (%) P b Control group (n 4) Y X RR Ask-for-Y group (n 5) Y X RR Ask-for-X group (n 4) Y X RR a Including only cases from room temperature group. b Comparing results from protocol 2 with original semen. c Comparing results from protocol 3 with original semen. The P-values were calculated on the changes in the Y:X ratios before and after separation. RR recovery rate; all values are presented as mean SD (range). Table V. comparisons of the separation effects at room temperature by different recovery rates (RR) a Protocol 2 Protocol 3 Before (%) After (%) P b Before (%) After (%) P c Low recovery rate group ( 5%) (n 5) ( 5%) (n 4) Y X RR Medium recovery rate group (5 10%) (n 3) (5 10%) (n 6) Y X RR High recovery rate group (10 15%) (n 2) ( 10%) (n 3) Y X RR High recovery rate group ( 15%) (n 3) Y X RR a Including only cases from room temperature group. b Comparing results from protocol 2 with original semen. c Comparing results from protocol 3 with original semen. The P-values were calculated on the changes in the Y:X ratios before and after separation. albumin separation (Martin, 1994). Furthermore clinical fac- 1993a; Flaherty et al., 1996). There have been lots of reports, tors, that probably would affect the sex ratio, such as agents based on these techniques, which question the effectiveness for ovulation induction and timing of insemination, would of laboratory manipulation of sperm separation (Han et al., confound the final sex ratio (James, 1985). Reliable methods 1993b; Vidal et al., 1993; Wang et al., 1994a; Wang et al., to validate efficiency of the laboratory procedures themselves 1994b; Claassens et al., 1995). are thus desirable. Fluorescent in-situ hybridization, specifically In a defensive essay published as a debate, Ericsson (1994) double-labelled FISH, which detects X- and Y spermatozoa stressed the importance of all his methodological details simultaneously, appears to be the method of choice for evaluating including processing at room temperature, preparation medium sex pre-selection procedures at present (Han et al., with Tyrode s solution, the isolation tube dimension and 1924

6 Lack of Y-spermatozoa enrichment by albumin separation in protocol 2. These figures were rather similar to those obtained Table VI. Frequencies of haploid and anhaploid cells before and after by Wang et al. (1994a). Their median (range) recoveries were separation a 8.5% (2.2 23%) in protocol 3 and 10.0% ( %) in Before separation After separation protocol 2. Our recoveries were slightly higher than those (%) obtained by Pyrzak (1994) ( % for protocol 3), but Protocol 2 (%) Protocol 3 (%) were well below the result from the original work of Ericsson YY et al. (1973) (in the range of 36 52%). XX Further analysis of the data was performed based on the XY Total anhaploid different levels of sperm yield (recovery rate of total motile a Including all cases. spermatozoa). However, perhaps due to the small case number (n 2 6 in each category, although the number of sperm nuclei analysed were 3000 to 9000 in each subgroup), we only the number of spermatozoa layered per column. All these obtained a trend toward X enrichment (P 0.07) in low parameters are aimed at reaching two endpoints: a small final recovery rate ( 5%) group and a significant X enrichment sperm yield (3 5%) and a highly motile final fraction (80 (P 0.03) in the moderate recovery rate (5 10%) group in 90%). Some of the reported results inconsistent with his data protocol 3. There was no specific recovery rate which would were criticized as failure to duplicate his original methods. In significantly facilitate X or Y enrichment in protocol 2. our study we intended to follow the procedure faithfully as Although we obtained statistically significant X enrichment, described by Ericsson (1994) except that we tried to find the the absolute values of the separation differences were very effect of temperature and performed the processing at two small (in the range of %). Similarly, in the report that temperatures instead of just room temperature. confirmed Y enrichment (Claassen et al., 1995), only a 3% The labelling efficiencies of the X/Y probe we used were increase in Y spermatozoa could be detected. The value of highly satisfactory ( 99%). This would assure the reliability spermatozoa separation by the albumin method, no matter of our data. Although we expected that the results would whether protocol 2, 3, or modified 3, in the clinical application accord with those reported by Ericsson et al. (1973), just to of sex pre-selection must be questioned critically. the contrary, our staining results revealed a Y-deprivation and Furthermore, our data revealed that the mean recovery rate an X-enrichment tendency in all of the samples in protocol 3 was in the range of %. Whether the low rate of motile and all but two (Y:X 1.02 and 1.11 respectively) of the sperm recovery would compromise the clinical pregnancy rate samples in protocol 2. This X-enrichment phenomenon reached is another important point to consider. In the only controlled statistical significance only when the separation procedure was study for gender selection, Jaffe reported no improvement in performed at room temperature, no matter whether it was the male birth after albumin separation, and he also noticed the two-layer (protocol 2) or three-layer (protocol 3) method. Here low pregnancy rate in each cycle (8, 13 and 9% for the first we confirmed previous observations that a separation effect three cycles) (Jaffe et al., 1991). could not be obtained at body temperature (Vidal et al., 1993). The frequency of anhaploid spermatozoa observed in this The data from the body temperature group were then excluded Oriental series (0.21%) was close to that of the Caucasian from further analysis. series reported by Martin et al. (1995), who documented a The X-enrichment result was consistent with that reported by Wang et al. (1994a), but contrary to that reported by diploid frequency of 0.16%. Our data also showed decreased Claassens et al. (1995). The former authors analysed frequencies of anhaploid spermatozoa after albumin separation sperm nuclei (12 patients) in each protocol with highly efficient methods, although they still had not reached statistical signific- double probes and found that all samples in protocol 3 and all ance. The results of our study also confirmed that the X- but one (Y:X 1.02) sample in protocol 2 were X-enriched. and Y-bearing spermatozoa were equally frequent, no matter The latter authors applied single probes with lower ( 95%) whether the semen samples were from males who had already labelling efficiencies on a smaller number (n 2092, 20 fathered two sons or two daughters. These two points were patients) of sperm nuclei, but they adopted the modified also consistent with the previous report by Wang et al. (1994a). protocol 3 which requires a shorter incubation time than ours. Two recent reviews also summarize the inefficiency of Unfortunately, they did not provide data on their recovery various gradient techniques for clinical use in pre-conception rates, nor analysis on the relationship between recovery rates sex selection, and conclude that only flow cytometry has been and sex ratios of the spermatozoa. shown by FISH to produce a clinically significant enrichment Recovery of total motile spermatozoa was intentionally kept of X and/or Y spermatozoa (Reubinoff, 1996; Flaherty et al., as small as possible but frequently it was rather difficult to 1996). The sorting is based on a 2.8% difference in DNA confine. It varied considerably from sample to sample. We had content of X and Y spermatozoa (Johnson et al., 1993). two cases with very high initial total counts ( ). Furthermore, the fluorochrome used in the staining of the Although we did not apply the modified protocol 3 for them, spermatozoa during flow cytometry sorting Hoechst their recovery rates were rather low and satisfactory. We were (bisbenzimide) has been demonstrated not to affect sperm also able to obtain sperm yields with very high motility motility parameters, nor mutation induction at low concentra- ( 90%) but the mean recovery rates of motile spermatozoa tion (Watkins et al., 1996). The flow cytometry separation is were 7.0% ( %) in protocol 3 and 9.3% ( %) the most promising sex pre-selection technology that probably 1925

7 M.-J.Chen, H.-F.Guu and E.S.-C.Ho would eliminate the use of selective abortion as a means of Quinlivan, W.L.G., Preciado, K., Long, T.L. et al. (1982) Separation of human X and Y spermatozoa by albumin gradients and Sephadex chromatography. decreasing X-linked genetic disorders (Johnson et al., 1993). Fertil. Steril., 37, Reubinoff, B.E. and Schenker, J.G. (1996) New advances in sex preselection. Fertil. Steril., 66, Ross, A., Robinson, J.A., Evans, H.J. (1975) Failure to confirm separation of Acknowledgements X- and Y-bearing human sperm using BSA gradients. Nature, 253, The authors would like to express their sincere thanks to Professor Steeno, O., Adimoelja, A. and Steeno, J. (1975) Separation of X- and Teresa Yang-Feng for her help in checking our FISH quality and Y-bearing human spermatozoa with the sephadex gel-filtration method. Andrologia, 7, Professor Cathy W.S.Chen and Miss Nien Hui-chen for their assistance Ueda, K. and Yanagimachi, R. (1987) Sperm chromosome analysis as a new in the final statistics. This work was based on research grant No. 84- system to test human X- and Y-sperm separation. Gamete Res., 17, from Taichung Veterans General Hospital, Taiwan. Van Kooij, R.J. and van Oost, B.A. (1992) Determination of sex ratio of spermatozoa with a deoxyribonucleic acid-probe and quinacrine staining: a comparison. Fertil. Steril., 58, Vidal F., Moragas M., Catala V. et al. (1993) Sephadex filtration and human References serum a albumin gradients do not select spermatozoa by sex chromosome: Brandriff, B.F., Gordon, L.A., Haendel, S. et al. (1986) Sex chromosome a fluorescent in-situ hybridization study. Hum. Reprod., 8, ratios determined by karyotypic analysis in albumin-isolated human sperm. Watkins, A.-M., Chan, P.J., Kalugdan, T.H. et al. (1996) Analysis of the flow Fertil. Steril., 46, cytometer stain Hoechst on human spermatozoa. Mol. Hum. Reprod., Barlow, P. and Vosa, C.G. (1970) The Y chromosome in human spermatozoa. 2, Nature, 226, Wang, H.X., Flaherty, S.P., Swann, N.J. et al. (1994a) Assessment of the Beernink, F.J., Dmowski, W.P. and Ericsson, R.J. (1993) Sex preselection separation of X- and Y-bearing sperm on albumin gradients using doublethrough albumin separation of sperm. Fertil. Steril., 59, label fluorescence in situ hybridization. Fertil. Steril., 61, Wang H.X., Flaherty S.P., Swann N.J. et al. (1994b) Discontinuous Percoll Check, J.H. and Katsoff, D. (1993) A prospective study to evaluate the gradients enrich X-bearing human spermatozoa: a study using double-label efficacy of modified swim-up preparation for male sex selection. Hum. fluorescence in-situ hybridization. Hum. Reprod., 9, Reprod., 8, World Health Organization (1992) Laboratory Manual for the Examination of Chevret, E., Rousseaux, S., Monteil, M. et al. (1994) Male meiotic segregation Human Semen and Sperm Cervical Mucus Interaction, 3rd edn. Cambridge of genosomes analysed by two colour FISH in human interphase University Press, Cambridge. spermatozoa. Hum. Genet., 94, Zarutskie, P.W., Muller, C.H., Magone, M. et al. (1989) The clinical relevance Claassens, O.E., Franken, D.R., Oosthuizen, C.J.J. et al. (1995) Fluorescent of sex selection techniques. Fertil. Steril., 52, in situ hybridization evaluation of human Y-bearing spermatozoa separated by albumin density gradients. Fertil. Steril., 63, Received on March 5, 1997; accepted on June 26, 1997 Cran, D.G. and Johnson, L.A. (1996) The predetermination of embryonic sex using flow cytometrically separated X and Y spermatozoa. Hum. Reprod. Update, 2, Dmowski, W.P., Gaynor, L., Rao, R. et al. (1979) Use of albumin gradients for X and Y sperm separation and clinical experience with male sex preselection. Fertil. Steril., 31, Ericsson, R. (1994) Sex selection via albumin column: 20 years of results. Hum. Reprod., 9, Ericsson, R.J., Langevin, C.N. and Nishino, M. (1973) Isolation of fractions rich in human Y sperm. Nature, 246, Evans, J.M., Douglas, T.A. and Penton, J.P. (1975) An attempt to separate fractions rich in human Y sperm. Nature, 253, Flaherty, S.P. and Matthews, C.D. (1996) Application of modern molecular techniques to evaluate sperm sex selection methods. Mol. Hum. Reprod., 2, Han, T.L., Ford, J.H., Webb, G.C. et al. (1993a) Simultaneous detection of X- and Y-bearing human sperm by double fluorescence in situ hybridization. Mol. Reprod. Dev., 34, Han, T.L., Flaherty, S.P., Ford, J.H. et al. (1993b) Detection of X- and Y- bearing human spermatozoa after motile sperm isolation by swim-up. Fertil. Steril., 60, Jaffe, S.B., Jewelewicz, R., Wahl, E. et al. (1991) A controlled study for gender selection. Fertil. Steril., 56, James, W.H. (1985) The sex ratio of infants born after hormonal induction of ovulation. Br. J. Obstet. Gynecol., 92, Johnson, L.A., Welch, G.R., Keyvanfar, K. et al. (1993) Gender preselection in humans? Flow cytometric separation of X and Y spermatozoa for the prevention of X-linked diseases. Hum. Reprod., 8, Kaneko, S., Yamaguchi, J., Kobayashi, T. et al. (1983) Separation of human X- and Y-bearing sperm using Percoll density gradient centrifugation. Fertil. Steril., 40, Martin, R.H. (1994) Human sex pre-selection by sperm manipulation. Hum. Reprod., 9, Martin, R.H., Spriggs, E., Ko, E. et al. (1995) The relationship between paternal age, sex ratios, and aneuploidy frequencies in human sperm, as assessed by multicolor FISH. Am. J. Hum. Genet., 57, Pyrzak, R. (1994) Separation of X- and Y-bearing human spermatozoa using albumin gradients. Hum. Reprod., 9,