The Activation of Proteolysis in the Acrosome Reaction of Guinea-Pig Sperm

J. Cell Sci. 32, 153-164 (1978) 153 Printed in Great Britain © Company of Biologists Limited lgyS

THE ACTIVATION OF PROTEOLYSIS IN THE ACROSOME REACTION OF GUINEA-PIG SPERM

D. P. L. GREEN Physiological Laboratory, Cambridge CBz 3EG, England

SUMMARY The divalent metal cation ionophore A23187 rapidly induces a normal acrosome reaction in a population of guinea-pig sperm suspended in calcium medium. In the course of the acrosome reaction, proacrosin, the zymogen precursor of the protease acrosin, is activated. Although the acrosome reaction causes exocytosis of the acrosomal contents, 'soluble' acrosin is not released in significant amounts until well after the sperm population as a whole has undergone an acrosome reaction. This suggests that proacrosin is stored within the acrosome in an insoluble form and that exocytosis of the acrosomal contents in the acrosome reaction is insufficient, by itself, to cause its immediate dissolution. Electron micrographs of sperm undergoing an A23i87-induced acrosome reaction in the presence of the acrosin inhibitors benzamidine, p-amino-benzamidine and phenylmethyl- sulphonyl fluoride show that the acrosome reaction proceeds normally but that dispersal of the acrosomal contents is inhibited. These morphological changes are, for the most part, below the limit of resolution of the light microscope and using light microscopy to assess whether an acrosome reaction has taken place, it can be mistakenly inferred that the reaction itself is inhibited by the acrosin inhibitors. The inhibition of the dispersal of the acrosomal contents by acrosin inhibitors suggests that acrosin activity is important in solubilizing acrosin. These experimental observations, taken with the evidence that the acrosome reaction is a response to an increase in intracellular free calcium, have been taken as the basis of a proposal for the mechanism of proacrosin activation in the acrosome reaction.

INTRODUCTION The acrosomes of mammalian sperm contain a number of enzymes. One of these is acrosin, a protease which is broadly similar to trypsin; it hydrolyses the same synthetic substrates (e.g. benzoyl arginine ethyl ester (BAEE), etc.) and it is inhibited by a wide range of synthetic and naturally occurring trypsin inhibitors (e.g. benzamidine, p-aminobenzamidine, soybean trypsin inhibitor, ovomucoid, etc.). Before the acrosome reaction acrosin exists in sperm almost completely as an inactive zymogen precursor, proacrosin (Meizel & Mukerji, 1976). Proacrosin, as far as is known, can only be activated by proteolytic cleavage (cf. trypsinogen, etc.). Activation can be caused by either trypsin or acrosin. There is no direct evidence that proacrosin is activated during the acrosome reaction, although if acrosin is necessary for sperm penetration of the zona it must be. The fate of proacrosin in the acrosome reaction has been examined using the A23i87-induced reaction of guinea-pig sperm (Green, 1978 a). At the same time, the effect of trypsin inhibitors on the acrosome reaction has been studied in view of a report that some inhibitors prevent the acrosome reaction from taking place (Meizel & Lui, 1976). 154 D. P. L. Green

METHODS Phenyl methyl sulphonyl fluoride (PMSF), ^-aminobenzamidine, soybean trypsin inhibitor (STI) Type I-S, lima bean trypsin inhibitor (LTI), ovomucoid and benzoyl arginine ethyl ester (BAEE) were purchased from Sigma, and benzamidine from Aldrich. The basic experimental procedures were the same as described previously (Green, 1978a). The time course for the loss of the acrosome induced by A23187 in the presence of trypsin inhibitors was estimated from inspection by light microscopy of sperm fixed at suitable intervals after addition of A23187. The acrosome reaction was induced by addition of 50 /il of the stock solution of A23187 in dimethyl sulphoxide (DMSO) to 5 ml of sperm suspension at 37 °C. Samples (0-5 ml) were pipetted into Karnovsky's fixative in which magnesium and 100 fiM EGTA were substituted for calcium. 500 sperm were examined for each time point. With the exception of PMSF, all inhibitors were dissolved directly in calcium medium. For the experiments with PMSF, 20/tl of a 25 mM stock solution in DMSO were added directly to s ml sperm suspension. In one experiment, sperm were incubated with PMSF in magnesium medium for 0-5 h and then washed 5 times in magnesium medium, 10 ml each time, before resuspension in calcium medium and addition of A23187. Acrosin activity was measured after acid extraction of sperm by following the hydrolysis of BAEE at 256 ran. The acrosome reaction was started with addition of A23187 from a stock solution in DMSO. In a typical experiment, o-s-ml samples were removed from 5 ml of sperm suspension, one sample before and the rest after addition of A23187, and pipetted into 0-5 ml of 20 mM HC1. The acid suspension was centrifuged at 14000 g,v after not less than 15 min and the supernatant assayed for acrosin activity. The assay solution contained 300 mM Tris HC1 buffer, pH 8-o, 150 mM CaCl! and 3 mM BAEE. To 0-9 ml of this solution was first added 0-9 ml of 10 mM KOH followed, after mixing, by 09 ml of acid extract. The change in absorbance at 256 nm was measured at 20 °C in a Zeiss PMQ II spectrophotometer. In one experiment the appearance of soluble acrosin discharged from sperm undergoing an A23i87-induced acrosome reaction was compared with the appearance of acid-extractable acrosin under the same conditions. Sperm in calcium medium were induced to undergo the acrosome reaction by addition of A23187 as before. At suitable time intervals, o-6 ml of sperm suspension was transferred to an Eppendorf tube and spun for 15 s at 14000 g,v. The centrifuge was rapidly stopped and 05 ml of the supernatant pipetted into a test tube containing 0-5 ml of 20 mM HC1. The whole procedure takes about 40-45 s: it was started 15 s before each time point and ended 25-30 s after. The acid solution was assayed as described previously for the acid-extracted acrosin.

RESULTS An acrosome reaction is induced in guinea-pig sperm in calcium medium by the divalent cation ionophore A23187 (Green, 1976; Summers et at. 1976; Talbot et al. 1976; Green, 1978 a). The acrosome reaction results in the loss of the acrosome and a substantial change in the shape and size of the head. The time course for the loss of the acrosome is shown in Fig. 3. In the course of the acrosome reaction, the total quantity of acrosin rises (Figs, i, 2). This increase in acrosin reflects the proteolytic activation of proacrosin. Activation is dependent on the concentration of A23187 and external calcium, both in rate and extent (Figs. 1, 2): activation does not occur in magnesium medium after addition of A23187 (Fig. 1) or in magnesium medium containing o-i % Triton X-100. Extraction of sperm with acid removes acrosin and proacrosin (Meizel & Mukerji, 1976) and at the same time dissociates the acrosin inhibitor (Brown, Andani & Hartree, 1975; Schleuning, Hell & Fritz, 1976). The acrosin activity measured therefore represents total acrosin, not free acrosin. Proteolysis in the acrosome reaction 155 The effect of acrosin inhibitors on the loss of the acrosome during the acrosome reaction is shown in Figs. 3-5. The reversible acrosin inhibitors benzamidine and p-aminobenzamidine and the irreversible acrosin inhibitor PMSF delay but do not prevent the loss of the acrosome (Figs. 3, 5), but the naturally occurring protein acrosin inhibitors, soybean trypsin inhibitor, lima bean trypsin inhibitor and ovomucoid are without effect (Fig. 4).

0-2 r

Time, min Figs. 1, 2. Proacrosin activation, induced in guinea-pig sperm in calcium medium by A23187. Fig. 1. Proacrosin activation for a fixed A23187 concentration, 80 fiM, and different external calcium concentrations, inmni: 10, # ; 2, O ; and 0-4, •. Sperm in magnesium show no activation on addition of A23187, ••

When sperm in magnesium medium are incubated with PMSF and then extensively washed before resuspension in calcium medium and addition of A23187, the loss of the acrosome takes place normally (Fig. 5). Electron micrographs of sperm undergoing an acrosome reaction in the presence of either p-aminobenzamidine or PMSF fixed 5 min after addition of A23187 (i.e. when Fig. 3 shows the population of sperm to have undergone almost completely an acrosome reaction in the absence of inhibitors) have the appearance shown in Figs. 6 and 7. Both micrographs show head-to-tail sections through a stack of sperm heads. They 156 D.P.L. Green are to be compared with figs. 1 and 4 of the preceding paper (Green, 1978 a). What they both show is that all the sperm have undergone an acrosome reaction, i.e. membrane fusion is well advanced, but the acrosomal contents have failed to disperse. As a result, all the sperm, irrespective of the sequence in which they underwent the acrosome reaction have been arrested at the same stage.

0-2 r-

0 5 10 Time, min Fig. 2. Proacrosin activation for a fixed external calcium concentration, 10 mM, and different concentrations of A23187, in/iM; 80, •; 16, •; and 3-2, O-

A comparison of the rate of appearance of acid-extractable acrosin with 'soluble' acrosin is shown in Fig. 8. This shows that 'soluble' acrosin does not significantly appear in the supernatant until proacrosin activation is essentially complete. This statement errs on the side of caution because a 15-s spin at 14000 g cannot be con- sidered an adequate test of solubility. It is almost certain, therefore, that release of soluble acrosin is delayed well beyond the period of its formation from proacrosin, if it ever occurs completely at all. Proteolysis in the acrosome reaction

1OOr

Time, min Figs. 3—5. The percentage of guinea-pig sperm which have lost their acrosome in the A23187-induced acrosome reaction as a function of time. Acrosome loss was detected by Nomarski light microscopy. The reaction was initiated at zero time by addition of A23187 to give a final concentration of 40 /tM. Fig. 3. The effect of the 2 synthetic acrosin inhibitors benzamidine and p-aminobenzamidine: no inhibitor, #; 100 /*M benzamidine, A; 100 /tM £-amino-benzamidine, Q.

100 r

Time, min Fig. 4. The effect of 3 protein acrosin inhibitors; no inhibitor, A; Kunitz soybean trypsin inhibitor (1 mg/ml), #; lima bean trypsin inhibitor (1 mg/ml), O; and ovomucoid (1 mg/ml), •• 11 CEL 32 D. P. L. Green

DISCUSSION Acrosin activity appears during the course of the acrosome reaction (Figs, i and 2) and this almost certainly represents activation of the zymogen precursor of acrosin, proacrosin. Zymogens generally can only be activated by proteolytic cleavage (Neurath, 1975), normally at a specific lysine residue, with loss of an iV-terminal peptide. The active site of the enzyme is distorted in the zymogen, but following cleavage it changes shape and activity is acquired. Zymogens themselves have detectable proteolytic

100 r-

$ E <5 o. CO a?

Time, min Fig. 5. The effect of the irreversible acrosin inhibitor phenyl methyl sulphonyl fluoride (PMSF). Sperm were either incubated in 100 fiM PMSF and washed before addition of A23187 (#) or kept in 100 fiM PMSF throughout (•)• activity (Kassell & Kay, 1973) and they react with irreversible inhibitors, although at a much slower rate than the enzymes themselves (Morgan, Robinson, Walsh & Neurath, 1972). Proacrosin can be activated by trypsin (Meizel & Mukerji, 1976) which indicates cleavage at either an arginine or lysine residue. It can also be activated by acrosin, whose properties as a protease are closely similar to trypsin: it hydrolyses the same synthetic substrates and is inhibited by the same synthetic and naturally occurring inhibitors (Polakoski & McRorie, 1973). Proacrosin autoactivation (i.e. activation by acrosin) takes place in the absence of calcium (Schleuning et al. 1976) and is not increased either in rate or extent by calcium at physiological concentrations. This effectively rules out the absence of calcium as a means of preventing proacrosin autoactivation in the presence of acrosin. There is no evidence that an intact acrosome actually contains any acrosin (as opposed to proacrosin, which it does contain). When sperm are extracted with acid, Proteolysis in the acrosome reaction

Figs. 6, 7. Sperm incubated in either 100 fiM />-aminobenzamidinc or 100 /IM PMSK and fixed 5 min after addition of A23187 to a final concentration of 40/(M. Fig. 6. The effect of />-aminobenzamidine. x 33500. Fig. 7. The effect of PMSF. x 32500. Figs. 6 and 7 both show that the acrosome reaction has proceeded normally but that detachment of the acrosomal conten ts has been arrested after cavitation at a time when in the absence of inhibitor they would have become completely detached. i6o D. P. L. Green the extract contains acrosin, but since in any population of sperm there are some that have prematurely undergone an acrosome reaction it could be argued that the acrosin comes wholly from these. Assume, however, that the acrosome does contain some acrosin. The only effective way of preventing autoactivation in situ is to inhibit this acrosin. What then happens in the acrosome reaction depends on the nature of the inhibitor. If it is small and reversible, it could rapidly diffuse away as soon as the acrosomal and plasma membranes fuse, leaving acrosin behind to activate proacrosin.

0-2 i-

0 5 10 15 Time, min Fig. 8. A comparison of the rate of proacrosin activation and the release of 'soluble' acrosin in the A23187-induced acrosome reaction in guinea-pig sperm; acid-extracted, O; 'soluble', «.

In this circumstance, whether a cell undergoes proacrosin autoactivation is simply a function of whether it undergoes an acrosome reaction, i.e. the inhibitor has the opportunity of diffusing away. The incomplete proacrosin activation of Fig. 1 would by this token indicate a division of the sperm population into those which had and those which had not undergone an acrosome reaction. Since the reaction is almost certainly dependent on an increase in cytoplasmic free calcium the incomplete proacrosin activation would suggest that even in the presence of A23187 some cells maintain their cytoplasmic free calcium below the threshold needed for membrane fusion. If, however, the inhibitor is more nearly irreversible a second explanation has to be Proteolysis in the acrosome reaction 161 sought, for opening the acrosome would be of no use in reversing the inhibition of acrosin. In this circumstance, a second protease would be needed to initiate activation of proacrosin. This protease could, itself, be activated in one of two ways: it could be released from inhibition by a rapidly reversible inhibitor in much the same way as proposed for acrosin, or it could be activated by an increase in calcium. There is clear evidence that washed sperm contain a protein acrosin inhibitor which is, for practical purposes, irreversible under normal physiological conditions (Brown et al. 1975; Schleuning et al. 1976) although its precise location is unknown. If the inhibitor is a fail-safe device, i.e. if it is present to mop up any acrosin which might arise either through spontaneous hydrolysis or the proteolytic action of proacrosin, it must be present in an amount sufficient to inhibit all the acrosin normally present before activation; but if acrosin activity is to appear as the result of activation it must be less than the total potential acrosin; and because of this it must be able to discriminate between proacrosin and acrosin and bind much more strongly to the acrosin. To sum- marize this mechanism: the acrosome is assumed to contain proacrosin, acrosin and a soluble acrosin inhibitor; the inhibitor binds strongly to acrosin but only weakly if at all to proacrosin; its quantity is sufficient to inhibit the acrosin normally present in the acrosome but is less than the total potential acrosin; activation is produced by a second protease which is not inhibited by the a crosininhibitor; this protease is activated as a consequence of an increase in intracellular free calcium. This mechanism owes a great deal to that already established for the exocrine pancreas, the only tissue where the synthesis, storage, secretion and activation of zymogens has been studied in any detail (Palade, Siekevitz & Caro, 1962; Neurath 1975). The pancreatic acinar cells secrete the digestive enzymes trypsin, chymo- trypsin, etc., as their zymogen precursors, trypsinogen, chymotrypsinogen, etc. The zymogens are stored prior to secretion in the zymogen granules together with a trypsin inhibitor (Kazal pancreatic trypsin inhibitor) (Greene, Rigbi & Fackre, 1966). This inhibitor is secreted into the lumen with the zymogens as part of the pancreatic juice from which it can be isolated free despite an 80-fold excess of trypsinogen (i.e. at io~6 M, its approximate concentration in pancreatic juice, it does not form a stable complex with trypsinogen). Its amount in pancreatic juice is about 1 % of the total potential trypsin activity. It therefore shows the 2 properties required of the acrosin inhibitor; it distinguishes between zymogen and active enzyme and it is present in sufficient quantity to inhibit all the active enzyme in the zymogen granules without preventing enzyme activity appearing on activation of the zymogen. The acrosin inhibitor in sperm is insufficient to inhibit more than a fraction of the potential acrosin (Brown et al. 1975) but whether it discriminates between proacrosin and acrosin is unknown. It is a prediction of this model that it will. There is one important difference between the pancreatic zymogens and proacrosin. The pancreatic zymogens are activated after secretion by an exogenous second protease, enteropeptidase (at least, trypsinogen is activated by enteropeptidase and the trypsin formed activates the other zymogens and the trypsinogen). There is no protease external to sperm and any activating protease must be intracellular. Activation of proacrosin by a calcium-dependent protease would entail two functions i62 D. P. L. Green for calcium in the acrosome reaction, membrane fusion and activation. This leads to the question of how calcium reaches the acrosomal contents. In the normal acrosome reaction in the absence of A23187, calcium can reach the acrosome after an increase in cytoplasmic free calcium has caused the acrosomal and plasma membranes to fuse or, if the acrosomal membrane is permeable to calcium, as soon as cytoplasmic free calcium rises. A23187 makes it possible for acrosomal calcium to rise irrespective of whether membrane fusion takes place. To return to Fig. 1, the incomplete proacrosin activation still suggests that a threshold has been crossed by only a fraction of the sperm, but whether it is the threshold to membrane fusion (bearing in mind that the fusion automatically exposes the acrosomal contents to the calcium in the external medium) or the threshold for proacrosin activation is a matter for speculation. When the time course for the loss of the acrosome is established in the presence of trypsin inhibitors it is found that the loss is delayed by small, synthetic inhibitors but not by the larger, naturally occurring protein inhibitors. When sperm undergoing an acrosome reaction in any of the synthetic inhibitors are fixed 5 min after addition of A23187, i.e. when, in the absence of inhibitor the bulk of the population would have undergone an acrosome reaction, they appear as in Figs. 6 and 7. These micrographs show that the sperm have undergone a normal acrosome reaction but also that dispersal of the acrosomal contents has been inhibited. This failure of the acrosomal contents to disperse rapidly in the presence of the synthetic inhibitors indicates that proteolytic digestion is normally involved in the dispersal. The changes which sperm undergo during an acrosome reaction induced in the presence of the synthetic inhibitors are initially very difficult to detect by light microscopy because so many of them are below the limit of resolution (e.g. all the vesiculation). It is only as the disintegration of the acrosomal matrix occurs that the changes become clearly recognizable. It is important, therefore, to distinguish between the acrosome reaction, which cannot be detected by light microscopy, and the changes which immediately follow the acrosome reaction, which, by the time they have reached the stage of substantial loss of the acrosomal contents, can be. Although the 3 synthetic inhibitors used all have the same effect, there is no evidence that they all act at precisely the same point in the acrosome reaction. PMSF is a small, uncharged irreversible acrosin inhibitor, soluble in both water and organic solvents and there is no simple mechanism by which it could be excluded from the acrosome. Zymogens are, generally speaking, resistant to rapid inactivation by irreversible inhibitors such as diisopropylfluorophosphate (DFP) and PMSF (Morgan et al. 1972; Kassell & Kay, 1973) and it is likely that only acrosin is inhibited (i.e. no reaction with the proacrosin): if the acrosome reaction is induced in the presence of PMSF, any acrosin that is produced from pro-acrosin activation will be quickly titrated off by the PMSF. Benzamidine and />-aminobenzamidine, on the other hand, both have a guanidinium group with a pK& of about 12-5 and it is much less certain that they enter the acrosome to any substantial extent before membrane fusion has occurred. Lastly, there is the absence of any substantial effect from the naturally occurring acrosin inhibitors. These inhibitors suffer from 3 relative disadvantages when set beside the synthetic inhibitors; they are used at lower concentrations; their onset rate constants Proteolysis in the acrosome reaction 163 for inhibition are certain to be much smaller because of the greater complexity of their interactions; and they diffuse more slowly. The last of these properties is likely to be particularly important in the acrosomal matrix throughout which acrosin is distributed (Green & Hockaday, 1978). It is certain that these inhibitors do not gain access to the acrosome until after membrane fusion has made it directly accessible to the external medium. Since this may also be true for benzamidine and p-s^nino- benzamidine, the difference between the 2 kinds of inhibitor could be purely a kinetic one and the difference in their effects cannot be taken as evidence that the synthetic inhibitors gain access to an intact acrosome. If PMSF is removed from sperm before the acrosome reaction is induced by A23187, the rate of loss of the acrosome is normal. This suggests that proacrosin has activated normally, which it could only have done if it were substantially uninhibited by PMSF treatment. It also suggests that the activating enzyme is itself resistant to PMSF or protected from its action by a reversible inhibitor. Finally, there is the question of the relationship between proacrosin activation and the appearance of acrosin in soluble form. Fig. 8 shows that the 2 events do not occur simultaneously. The simple explanation that proacrosin activation takes place within an intact acrosome and acrosin is released immediately after the acrosome reaction has occurred is inconsistent with the morphological evidence, which clearly indicates that virtually the whole sperm population has undergone an acrosome reaction before any acrosin is released in soluble form. This is not to say that proacrosin activation does not occur within an intact acrosome; it may well do. It is clear, however, that acrosin remains bound to the sperm head immediately after the acrosome reaction. Since it is a constituent of the acrosomal matrix (Green & Hockaday, 1978) and the detachment of the matrix is itself dependent on proteolytic digestion, the evidence suggests that the appearance of soluble acrosin is itself the result of proteolytic digestion, much, if not all of which is autodigestion. The persistence of the matrix after an acrosome reaction has been induced in the presence of the synthetic acrosin inhibitors suggests that the medium per se has little effect in dispersing the acrosomal contents and therefore, by implication, solubilizing acrosin. With this in mind, an alternative explanation can be provided for the experimental observation of Brown & Hartree (1976) that in denuded ram sperm in the absence of calcium, no acrosin appears in the supernatant, whereas in the presence of calcium it does. The conclusion drawn by Brown & Hartree (1976) was that acrosin remains insoluble in the absence of calcium. No reference was made to the fact that the acrosin would have been present as proacrosin. On the basis of the behaviour of guinea-pig sperm, the effect of calcium would have been to activate the proacrosin and the acrosin would subsequently appear as the result of proteolytic digestion. This mechanism receives additional support from the estimates of Brown & Hartree (1976) for the amount of acrosin in denuded sperm: when the sperm have not been exposed to calcium it is about 0-12 units/109 cells but when the sperm are extracted under conditions which would undoubtedly produce proacrosin activation, the amount is 18-22 units/109 cells. This huge difference is not discussed, but much of it must be due to proacrosin activation. 164 D. P. L. Green

REFERENCES BROWN, C. R., ANDANI, Z. & HARTREE, E. F. (1975). Studies on ram acrosin. Biochem.jf. 149, 133-146. BROWN, C. R. & HARTREE, E. F. (1976). Effects of acrosin inhibitors on the soluble and membrane-bound forms of ram acrosin, and a reappraisal of the role of the enzyme in fertilisation. Hoppe-Seyler's Z. pkysiol. Chem. 357, 57-65. GREEN, D. P. L. (1976). Induction of the acrosome reaction in guinea pig spermatozoa in vitro by the Ca ionophore A23187. J. Physiol., Lond. 260, 18P-19P. GREEN, D. P. L. (1978a). The induction of the acrosome reaction in guinea-pig sperm by the divalent metal cation ionophore A23187. jf. Cell Sci. 32, 137-151. GREEN, D. P. L. (19786). The osmotic properties of the acrosome in guinea-pig sperm. J. Cell Sci. 32, 165-176. GREEN, D. P. L. & HOCKADAY, A. R. (1978). The histochemical localization of acrosin in guinea-pig sperm after the acrosome reaction. J. Cell Sci. 32, 177-184. GREENE, L. J., RIGBI, M. & FACKRE, D. S. (1966). Trypsin inhibitor from bovine pancreatic juice. J. 610/. Chem. 241, 5610-5618. KASSELL, B. & KAY, J. (1973). Zymogens of proteolytic enzymes. Science, N. Y. 180, 1022-1027. MEIZEL, S. & Lui, C. W. (1976). Evidence for the role of a trypsin-like enzyme in the hamster sperm acrosome reaction. J. exp. Zool. 196, 137-144. MEIZEL, S. & MUKERJI, S. K. (1976). Biochemical studies of proacrosin and acrosin from hamster cauda epididymal spermatozoa. Biol. Reprod. 14, 444-450. MORGAN, P. H., ROBINSON, N. C, WALSH, K. A. & NEURATH, H. (1972). Inactivation of bovine trypsinogen by di-isopropylfluorophosphate. Proc. natn. Acad. Sci. U.S.A. 69, 3312-3316. NEURATH, H. (1975). Limited proteolysis and zymogen activation. In Cold Spring Harbor Conferences on Cell Proliferation, vol. 2, Proteases and Biological Control (ed. E. Reich, D. B Rifkin & E. Shaw), pp. 51-64. Cold Spring Harbour Laboratory. PALADE, G. E., SIEKEVITZ, P. & CARO, L. G. (1962). Structure, chemistry and function of the pancreatic exocrine cell. In Ciba Fdn Symp. on the Exocrine Pancreas (ed. A. V. S. de Reuck, & M. P. Cameron), pp. 23-49. London: Churchill. POLAKOSKI, K. L. & MCRORIE, R. A. (1973). Boar acrosin. II. Classification, inhibition and specificity studies of a proteinase from sperm acrosomes. J. biol. Chem. 248, 8183-8188. SCHLEUNING, W.-D., HELL, R. & FRITZ, H. (1976). Multiple forms of boar acrosin and their relationship to proenzyme activation. Hoppe-Seyler's Z. physiol. Chem. 357, 207-212. SUMMERS, R. G., TALBOT, P., KEOUGH, E. M., HYLANDER, B. L. & FRANKLIN, L. E. (1976). Ionophore A23187 induces acrosome reactions in sea urchin and guinea pig spermatozoa. J. exp. Zool. 196, 381-386. TALBOT, P., SUMMERS, R. G., HYLANDER, B. L., KEOUGH, E. M. & FRANKLIN, L. E. (1976). The role of calcium in the acrosome reaction: an analysis using ionophore A23187. J. exp. Zool. 198, 383-392- {Received 13 January 1978)