To determine whether the characteristics of Percoll-selected spermatozoa are more predictive of in-vitro fertilization (IVF) results than are those of native semen, 118 semen samples from patients undergoing an IVF attempt were studied. Motility, using computer-assisted sperm analysis, and morphology were recorded before and after sperm selection on a Percoll gradient. Percoll selection increased the percentage of morphologically normal spermatozoa (58.2 +/- 19.9% versus 47.9 +/- 17.9%; p < 0.0001). This increase concerned almost all abnormalities and especially abnormalities of the midpiece (cytoplasmic droplets and bent tails). However, morphology of spermatozoa in the Percoll fraction had the same predictive value for IVF as did that for whole semen. Concerning motility, all parameters, except linearity, were improved by Percoll preparation, and their predictive value for IVF results was significantly higher in the Percoll fraction than in native semen. Therefore, it is concluded that, even if sperm characteristics are improved dramatically after Percoll selection, only motility analysis then gave more practical information than did analysis of sperm characteristics in native semen. This suggests that impairment of spermiogenesis, which presumably induces sperm abnormalities, is more important than are the actual characteristics of the spermatozoa.

To determine whether variations in spontaneous and induced acrosome reactions are correlated with semen quality, and to identify the inducers of clinical interest, the acrosome reaction, sperm concentration, motility and morphology were recorded in 117 semen samples from patients undergoing an in-vitro fertilization (IVF) attempt. The spontaneous acrosome loss after 24 h incubation in Ménézo B2 medium and after induction by calcium ionophore A23187, progesterone, human follicular fluid, cyclic adenosine 3'-5'-phosphate (cAMP) analogue and phorbol ester (TPA) were measured using the fluorescein isothiocyanate-GB24 antibody. The mean (range) spontaneous acrosome reaction was 3.4 +/- 0.6% (0.0-30.0). Response to the tested inducers was 23.3 +/- 1.6% (0.0-72.0) for calcium ionophore, 5.4 +/- 0.8% (0.0-60.0) for progesterone, 1.0 +/- 0.7% (0.0-24.0) for human follicular fluid, 2.6 +/- 0.7% (0.0-39.0) for the cAMP analogue and 2.3 +/- 0.7% (0.0-31.0) for TPA. The response to calcium ionophore was correlated significantly to sperm concentration, motility and morphology, while the responses to progesterone and TPA were correlated significantly to motility and acrosome morphology. The responses to other inducers were not linked to classic semen parameters. When studying acrosome reaction as a function of IVF results, the responses to calcium ionophore and TPA were discriminant. The results of this study show that the spontaneous acrosome loss and the responses to acrosome reaction inducers are highly variable and partially linked to semen quality. The responses to calcium ionophore and TPA could be of interest in predicting the fertilizing ability in vitro.

A total of 130 semen samples were examined for motility (by computer-assisted sperm analysis), morphology and acrosomal status. A high positive correlation was found between percentages of normal forms and progressive motility in the whole semen (r = 0.539, P < 0. 0001) as well as in the Percoll fraction (r = 0.702, P < 0.0001). Among the specific abnormalities, acrosome defects were most highly correlated with progressive motility (r = -0.492, P < 0.0001, in the Percoll fraction). The percentage of total spontaneously acrosome-reacted spermatozoa in the Percoll fraction was negatively correlated with the progressive motility (r = -0.499, P < 0.0001) and with the percentage of normal forms (r = 0.430, P < 0.0001). Surprisingly, the percentage of total spontaneously acrosome-reacted spermatozoa was poorly linked with head abnormalities but displayed significant positive correlations with the percentages of bent tails (r = 0.359, P < 0.0001) and of coiled tails (r = 0.371, P < 0.0001). These data suggest that sperm defects are often linked together, reflecting spermiogenesis and/or epididymal dysfunctions.

The fertilizing ability of human spermatozoa depends upon numerous functions such as motility, normal morphology, ability to bind to the zona pellucida and to undergo the acrosome reaction. Hence a lot of tests have been developed to try and predict IVF results. In a previous study we had established a scoring method, based on parameters such as sperm morphology, vitality, motility and the acrosome reaction, which was able to predict up to 83% of in-vitro fertilization results. The present study aimed to validate this score on a separate set of sperm samples. The results confirmed those of the first series. The score allowed prediction of fertilization failures with a 56% sensitivity, a 91% specificity, a 56% positive predictive value and a 91% negative predictive value. Therefore, this score could be used routinely to choose between conventional IVF and ICSI.

In order to study the characteristics of human acrosome reacted spermatozoa, we developed a method to select them. Spermatozoa, obtained from 20 fertile volunteers, were selected by using GB24 antibody fixed on magnetic immunobeads, after acrosome reaction induced by 50% follicular fluid. Bead-bound spermatozoa were then detached using sheep anti-mouse IGG F(ab')2 antibody. This method allowed recovery of 170 +/- 48 x 10(3) spermatozoa (n = 20), free of GB24 antibody, as assessed by incubation with FITC-rabbit anti-mouse antibody. The percentage of acrosome reacted spermatozoa in the selected population was 88 +/- 3% versus 32 +/- 6% in the whole sperm population. Concerning sperm morphology, the percentage of head abnormalities was lowered (15 +/- 3% versus 20 +/- 3%). The motility of selected spermatozoa was dramatically reduced (7 +/- 3% versus 53 +/- 7% in the whole population) despite no difference in viability (84 +/- 3% versus 80 +/- 4%). However, the viability after an 18 h incubation was very low (1 +/- 0.5% versus 46 +/- 5%). These results show that acrosome reaction occurs in the most morphologically normal spermatozoa and is followed by a loss in motility and a decrease in longevity.