Abstract. During the course of a study of glycoprotein processing mannosidases in the rat epididymis, we have made an intriguing discovery regarding the presence of a novel ot-o-mannosidase on the rat sperm plasma membranes. Unlike the sperm acrosomal "acid" mannosidase which has a pH optimum of 4.4, the newly discovered Ot-D-mannosidase has a pH optimum of 6.2, and 6.5 when assayed in sperm plasma membranes and intact spermatozoa, respectively. In addition, the two enzymes show different substrate specificity. The acrosomal Ot-D-mannosidase is active mainly towards synthetic substrate, p-nitrophenyl a-Dmannopyranoside, whereas the sperm plasma membrane O~-D-mannosidase shows activity mainly towards mannose-containing oligosaccharides. Evidence is presented which suggest that the sperm plasma membrane Ot-D-mannosidase is different from several processing mannosidases previously characterized from the rat liver.The newly discovered a-D-mannosidase appears to be an intrinsic plasma membrane component, since washing of the purified membranes with buffered 0.4 M NaCI did not release the enzyme in soluble form. The enzyme requires nonionic detergent (Triton X-100) for complete solubilization. The enzyme is activated by Co 2+ and Mn 2÷. However, Cu 2÷ and Zn 2÷ are potent inhibitors of the sperm plasma membrane Ot-D-mannosidase. At a concentration of 0.1 mM, these divalent cations caused nearly complete inactivation of the sperm enzyme. In addition methyl-c~-D-mannoside, methyl-ot-D-glucoside, mannose, 2-deoxy-D-glucose, and D-mannosamine are inhibitors of the sperm surface a-D-mannosidase. The physiological role of the newly discovered enzyme is not yet known. Several published reports in three species, including the rat, suggest that the sperm surface c~-o-mannosidase may have a role in binding to mannose-containing saccharides presumably present on the zona pellucida.I X is generally accepted that one step in the fertilization process requires interaction between complementary molecules present on the surface of the spermatozoon and the zona pellucida. The chemical nature of these complementary recognition sites is poorly understood, although there is growing evidence that carbohydrate moieties on surface membrane glycoconjugates are involved in these interactions (1, 34).In mammals, several sperm proteins have been suggested to bind to zona pellucida (28). Some of these macromolecules have enzymatic activity and are thought to form a stable enzyme-substrate complex by binding to the oligosaccharide units present on the zona pellucida glycoproteins. In mice, galactosyltransferase present on the head region of the spermatozoa mediates sperm-egg binding by interacting with its substrate on the zona pellucida (23). There is also evidence that trypsin-like protease present on spermatozoa of mouse (4, 31) initiates sperm-egg binding. Mouse spermatozoa also contain sialyltransferase (13) and fucosyltransferase (30). The latter enzyme has been suggested to be involved in some aspect of sperm-egg recognition (Apter, E M., ...

Previous studies from this laboratory and others have identified several enzymes on the surface of mammalian spermatozoa. Some of these enzymes, namely a galactosyltransferase and a novel alpha-D-mannosidase, are believed to play a ligand-like role in recognizing and binding to the complementary moiety(ies) present on zona pellucida glycoconjugates. However, little or no information is available about the occurrence of these enzymes in human spermatozoa. In the present report, we show that a very small amount of the total galactosyltransferase activity present in human semen is associated with spermatozoa. Moreover, our failure to find a significant amount of the enzyme on sperm plasma membranes suggests that the enzyme is not associated with the sperm surface. Therefore, it is unlikely that galactosyltransferase in humans has the same ligand-like role in zona binding that is demonstrated in mouse sperm. In contrast, nearly 5% of alpha-D-mannosidase activity was repeatedly found in the salt-washed plasma membrane fraction. The recovery and enrichment of the alpha-D-mannosidase was nearly one-half that observed for adenylate cyclase and nearly one-third that for phosphodiesterase I, the two sperm plasma membrane marker enzymes. The differential enrichment and recovery of the sperm surface alpha-D-mannosidase is consistant with our previous studies in rat spermatozoa, and suggests that alpha-D-mannosidase may be localized on morphologically distinct region(s) of the sperm plasma membranes. The properties of human sperm surface alpha-D-mannosidase are quite similar to those reported by us for rat sperm plasma membrane mannosidase, but quite different from human sperm acid alpha-D-mannosidase. In addition, whereas anti-rat epididymal alpha-D-mannosidase antibody (IgG-fraction) cross-reacted with the human sperm acid alpha-D-mannosidase, no cross-reactivity was observed with the sperm surface mannosidase. A small amount of fucosyltransferase (less than 1% of the enzyme originally present on spermatozoa) was found in the salt-washed plasma membrane, but the enrichment of the enzyme was only one-tenth of that observed for adenylate cyclase. The potential ligand-like role of human sperm surface alpha-D-mannosidase and other sperm surface enzymes during fertilization is discussed.

Several glycosidases, purified and characterized from mammalian tissues, have been shown to be optimally active under acidic conditions when p-nitrophenyl (PNP) or 4-methylumbelliferyl glycosides are used as substrates. Although high levels of the glycosidases are present in the epididymal lumen, their physiological role remains uncertain. To be functional, the glycosidases are expected to be enzymatically active at or near the physiological pH of luminal fluid. In this report, we demonstrate that the rat epididymal luminal fluid beta-D-galactosidase, optimally active toward PNP beta-D-galactoside at pH 3.5, shows maximum activity towards a glycoprotein substrate ([Gal-3H]fetuin) at neutral pH. Several lines of evidence, including immunoprecipitation studies using antibody to the acid beta-D-galactosidase, and substrate competition studies, indicate that PNP galactosidase and [3H]Gal galactosidase activities are caused by a single enzyme, and that the two substrates are probably cleaved by the same catalytic site(s). Competition studies with various disaccharides indicate that this enzyme is capable of cleaving a variety of galactose linkages found in both O- and N-linked oligosaccharides. Molecular-sieve column chromatography of the beta-D-galactosidase of luminal fluid under several conditions of buffer and pH show that, whereas the enzyme eluted as a tetramer (apparent M(r) 320,000) under acidic conditions (pH 3.5-4.3), only dimers and monomers (apparent M(r) 180,000 and 92,000 respectively) were observed in neutral conditions (pH 6.8). This aggregation/dissociation phenomenon is reversible. These studies indicate that beta-D-galactosidase is present in the luminal fluid in dissociated forms, and is therefore optimally active towards glycoprotein substrates at physiological pH. The potential role of the enzyme in modification of sperm surface glycoproteins is discussed.

Spermatozoa acquire fertilizing ability during passage through the epididymis. Modification of oligosaccharide moieties on sperm surface glycoproteins are some of the biochemical changes believed to be important in the production of functionally mature spermatozoa during passage through the epididymis. In an attempt to understand the mechanism underlying these modifications, we quantified four glycosyltransferase activities (the enzymes that catalyze the transfer of sugar residues from nucleotide sugar donor to the sugar chains on glycoproteins and glycolipids) of spermatozoa and fluid from various regions of the epididymis. Our results are as follows. (1) Only 10-20% of the total glycosyltransferase activities (sialyltransferase, fucosyltransferase, galactosyltransferase, and N-acetyl glucosaminyltransferase) sedimented with the spermatozoa; the remaining 80-90% of the four enzymes were present in soluble form in the epididymal fluid. (2) When the four transferase activities were expressed per 10(6) spermatozoa, only sialyltransferase and fucosyltransferase activities showed maturation-dependent changes. The former enzyme was significantly higher on the proximal caput spermatozoa and the latter on the distal caput spermatozoa. The higher levels of the two enzymes on caput spermatozoa could be due to their binding to the endogenous sugar acceptor molecules on the sperm surface, and subsequent release following sequential sialylation and fucosylation of the molecules in the proximal and distal caput spermatozoa, respectively. (3) When spermatozoa from the proximal and distal caput, corpus, and proximal and distal cauda were incubated with fucose-labeled nucleotide sugar (GDP[14C]fucose), higher levels of radioactivity were routinely incorporated into the spermatozoa from the distal caput. (4) The [14C]fucose-labeled spermatozoa or sperm plasma membranes, when solubilized, resolved on SDS-PAGE, and visualized by autoradiography, showed that the radioactivity had been incorporated into an endogenous acceptor of 86 kDa (major component) and several minor components. Treatment of the solubilized spermatozoa with N-glycanase suggested that the [14C]fucose is mainly present on N-linked oligosaccharide units. These studies demonstrate that some of the sperm surface components are fucosylated during sperm maturation. The potential significance of the in vitro fucosylation of sperm surface components in the production of functionally mature spermatozoa is discussed.

During transit through the epididymis, spermatozoa acquire fertilizing the cell surface exhibits an altered glycoprotein pattern. Epididymal cells and their secretions contribute to these sperm-surface changes. To examine this process, epithelial cells from rat caput and cauda epididymidis were cultured and examined for the synthesis, processing and secretion of two glycoprotein-modifying enzymes, beta-galactosidase and beta-glucuronidase. Cells were cultured four days, incubated with D-2-[3H] mannose and L-[35S] methionine, and placed in isotope-free media. Levels of both cellular and secreted beta-galactosidase and beta-glucuronidase were determined by immunoprecipitation of cell homogenates or medium, followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and scintillation counting of bands. During a 1-h pulse, both caput and cauda cells synthesize two precursor forms of beta-galactosidase (Mr = 84,000 and 87,000), which are processed to the mature (Mr = 63,000) enzyme during a 24-h chase. Caput cells release a high molecular weight (HMW) form (Mr = 90-100,000) and mature beta-galactosidase into the media, but not the Mr = 84-87,000 precursor. On the other hand, cauda cells release mostly mature beta-galactosidase. Ratios of radiolabeled mannose/methionine demonstrate a 7-fold greater mannose content in the cellular precursor of beta-galactosidase than in total protein. Another glycosidase, beta-glucuronidase, is synthesized as a Mr = 78,000-precursor which is processed to the mature Mr = 72,000 form. Medium in which caput and cauda cells were cultured contains both mature enzyme and a Mr = 94,000 form, but no 78,000-precursor form. Ratios of radiolabeled mannose/methionine in the cellular precursor of beta-glucuronidase are 2-fold greater than ratios in the total glycoprotein. Secretion is the major pathway of turnover for several epididymal glycosidases, since more than 50% of the total is secreted/day. These results indicate that cultured epithelial cells from the epididymis synthesize glycosidases and that processing and release differ, depending on the enzyme and the epididymal segment from which the epithelial cells were isolated.