Pyridinedicarboxylates, the first mechanism-derived inhibitors for prolyl 4-hydroxylase, selectively suppress cellular hydroxyprolyl biosynthesis. Decrease in interstitial collagen and Clq secretion in cell culture.

Peptidyl-dipeptidase A (angiotensin converting enzyme; ACE, EC 3.4.15.1), has been purified from pig kidney and striatum by affinity chromatography employing the selective inhibitor lisinopril as ligand. The inclusion of a 2.8 nm spacer arm improved the yield of the enzyme compared with the 1.4 nm spacer arm described in previous work. Two forms of striatal ACE (Mr 180,000 and 170,000), but only a single form of kidney ACE (Mr 180,000), were isolated by this procedure. Both forms of striatal ACE were recognized by a polyclonal antibody to kidney ACE. No significant differences in substrate specificity or inhibitor sensitivity between kidney and striatal ACE could be detected. In particular, the amidated neuropeptide, substance P, was hydrolysed identically by both preparations and no significant hydrolysis of the related tachykinin peptides neurokinin A and neurokinin B could be detected. After chemical or enzymic deglycosylation, kidney and both forms of striatal ACE migrated identically on sodium dodecyl sulphate/polyacrylamide-gel electrophoresis with an apparent Mr of 150,000. We suggest that the two detectable forms of ACE in pig brain are not isoenzymes but are the result of differential glycosylation in different cell types in the brain. It appears that ACE, unlike endopeptidase-24.11, does not have the general capacity to hydrolyse and inactivate the tachykinin peptides at a significant rate in brain.     

Distribution of the Glucose Transporter in the Mammalian Brain

We used [3H]cytochalasin B as a specific ligand to study the glucose transporter of the following tissue preparations: (a) microvessels derived from the cerebral cortex and cerebellum of the rat and pig, (b) particulate fractions of the cerebral cortex and cerebellum of the rat and pig, (c) lateral, third, and fourth ventricular choroid plexus of the pig, and (d) synaptosomes from the pig cerebral cortex. Specific, D-glucose-displaceable binding of [3H]cytochalasin B was present in all the preparations studied. This binding was saturable and displayed the kinetics of a single class of binding sites, similar to the glucose transporter found in other mammalian tissues. The density of the glucose transporter was much higher in cerebral and cerebellar microvessels and choroid plexus than either in crude particulate fractions of the cerebrum and cerebellum or in cerebral synaptosomes. These findings agree with the physiologic function of brain microvessels that transport glucose, not only for their own use, but also for the much greater mass of the entire brain. In the pig, the density of the glucose transporter in cerebral microvessels was significantly higher than in cerebellar microvessels. Irreversible photoaffinity labeling of the glucose transporter of synaptosomal membranes with [3H]cytochalasin B followed by solubilization and polyacrylamide gel electrophoresis demonstrated a single region of radioactivity that corresponded to a molecular mass of 60,000-64,000 daltons.     

Proteins of the kidney microvillar membrane. Purification and properties of carboxypeptidase P from pig kidneys.

Carboxypeptidase P has been purified by immunoaffinity chromatography from pig kidneys. A single-step assay with Z-Pro-Met (where Z represents benzyloxycarbonyl) as substrate was used, methionine being determined by using L-amino acid oxidase and horseradish peroxidase. The enzyme constitutes about 1.5% of the kidney microvillar proteins. Triton X-100-solubilized and papain-released forms of the enzyme were isolated. The former had an apparent subunit Mr of 135 000, and the latter form contained two polypeptide chains of Mr 128 000 and 95 000. The undenatured forms were dimeric proteins. In common with other microvillar hydrolases, carboxypeptidase P was a glycoprotein and each subunit contained one Zn atom. MnCl2 (1 mM) in the assay was necessary for maximum activity; in its absence, 0.5 mM-ZnSO4 produced a limited activation, but was inhibitory at higher concentrations. The Km for Z-Pro-Met, in the presence of MnCl2, was 4.1 mM, and the kcat. for freshly prepared enzyme was 1230 min-1. The enzyme lost activity during storage at -20 degrees C. In a limited survey of peptides, hydrolysis was observed only with substrates containing a proline, alanine or glycine residue in the P1 position, and these included angiotensins II and III. The best substrate in this series was Val-Ala-Ala-Phe.     

Membrane Localization of Endopeptidase-24.11 and Peptidyl Dipeptidase A (Angiotensin Converting Enzyme) in the Pig Brain: A Study Using Subcellular Fractionation and Electron Microscopic Immunocytochemistry

Brains from piglets were dissected and a block of tissue including the substantia nigra, globus pallidus, and entopeduncular nucleus was homogenized and then fractionated on discontinuous Percoll gradients. Ligand-binding assays using (-)-[3H]nicotine and [3H]quinuclidinyl benzilate served to delineate fractions containing nicotinic and muscarinic acetylcholine receptors. In this system endopeptidase-24.11 exhibited a biphasic distribution, consistent with its presence on both pre- and postsynaptic membranes. Peptidyl dipeptidase A (angiotensin converting enzyme; ACE) was associated with membrane fractions containing muscarinic receptors. An immunoblot of these fractions with an affinity-purified polyclonal antibody to ACE revealed only the neuronal form of ACE (Mr 170,000), the endothelial form (Mr 180,000) being undetectable. Electron microscopic immunoperoxidase staining of the substantia nigra, with an affinity-purified antibody to endopeptidase-24.11 at the preembedding stage, showed this antigen to be confined to the plasma membranes of boutons, axons, and some dendrites. Both pre- and postsynaptic membranes were stained, and occasionally other regions of the dendritic membrane were positive. No staining of synaptic vesicles within the boutons was observed. Thus, two independent approaches indicate that endopeptidase-24.11 is present on both pre- and postsynaptic membranes in the pig substantia nigra. The subcellular fractionation suggests that neuronal ACE is confined to dendritic membranes.     

Cathepsin D-like activity in the release of Gal beta 1-4GlcNAc alpha 2-6sialyltransferase from mouse and guinea pig liver Golgi membranes during the acute phase response

1. Rat Gal beta 1-4GlcNAc alpha 2-6sialyltransferase (E.C. 2.4.99.1) is released from Golgi membranes by cleavage of a portion of the enzyme containing the active site from a membrane anchor; this effect was most dramatic during the acute phase response. The enzyme that cleaved sialyltransferase had the properties of cathepsin D was most active at pH 5.6 and was likely of lysosomal origin (Lammers and Jamieson, 1988). 2. The acute phase response of sialyltransferase in mouse and guinea pig was previously found to differ from that in the rat. Release of sialyltransferase from mouse and guinea pig Golgi membranes has now been studied in order to make a comparison with the rat system. 3. Maximum release of sialyltransferase from mouse and guinea pig Golgi occurred at pH 4.6 and 5.2, respectively; like the rat a cathepsin D-like proteinase was responsible for release of both enzymes. 4. Immunoblot analysis showed that membrane-bound rat and mouse sialyltransferase had Mr 49,000, whereas the guinea pig enzyme had Mr 42,000. The released form of the rat enzyme had Mr 42,000, but released forms of mouse and guinea pig enzymes had Mr 38,000 suggesting a different cleavage site for these two enzymes compared to the rat enzyme.     

Phosphorylated adenosine derivatives as low-affinity adenosine-receptor agonists. Methodological implications for the adenylate cyclase assay.

In cellular systems provided with activatory (Ra-site) receptors for adenosine, such as rat cerebral microvessels and rat liver plasma membranes, the adenosine-receptor antagonist 8-phenyltheophylline (10 microM) significantly decreased adenylate cyclase activity if ATP was the substrate and only if GTP was present. With dATP as substrate, adenylate cyclase activities in both preparations remained unaffected by 8-phenyltheophylline. In rat cerebral-cortical membranes, with inhibitory (Ri-site) receptors for adenosine, 8-phenyltheophylline significantly enhanced adenylate cyclase activity only in the presence of GTP and if ATP was the substrate. In rat cardiac ventricular membranes, which are devoid of any adenylate cyclase-coupled adenosine receptor, the methylxanthine had no GTP-dependent effect, irrespective of the substrate used. All assay systems contained sufficiently high amounts of adenosine deaminase (2.5 units/ml), since no endogenous adenosine, formed from ATP, was found chromatographically. In order to demonstrate a direct influence of phosphorylated adenosine derivatives on adenylate cyclase activity, we investigated AMP in a dATP assay system. AMP was verified chromatographically to remain reasonably stable under the adenylate cyclase assay conditions. In the microvessels, AMP increased enzyme activity in the range 0.03-1.0 mM, an effect competitively antagonized by 8-phenyltheophylline. In the cortical membranes, 0.1 mM-AMP inhibited adenylate cyclase, which was partially reversed by the methylxanthine. The presence of GTP was again necessary for all observations. In the ventricular membranes, AMP had no effect. Since the efficacy of adenosine-receptor agonists and, probably, that of other hormones on adenylate cyclase activity can be more efficiently measured with dATP as the enzyme substrate, this nucleotide seems preferable for adenylate cyclase measurements in systems susceptible to modulation by adenosine.     

Structural studies on membrane-bound and soluble growth-hormone-binding proteins of rabbit liver.

Covalent cross-linking techniques have been used to investigate the structural characteristics of the growth-hormone (GH) receptor in a variety of rabbit liver cell membrane preparations (particulate and soluble). Two classes of GH-binding protein have been identified which differ in their Mr by gel filtration and susceptibility to precipitation with poly(ethylene glycol) (PEG). The first, a PEG-precipitable (Mr approximately 300,000) protein, contained Mr-65,000 and Mr-40,000 binding proteins linked by disulphide bonds. It was present in aqueous extracts derived from microsomal membranes but was not present in cytosol preparations. The second, a PEG-non-precipitable protein (Mr approximately 100,000) was composed of a non-disulphide-linked primary GH-binding subunit of Mr 60,000-66,000. This binding protein was present in all rabbit liver cell fractions and/or preparations. Both binding-protein classes contained intramolecular disulphide bonds. It is not clear whether the Mr-approximately 100,000 form, or perhaps higher-Mr species which have not been identified by cross-linking studies, represents the native, endogenous, form of the GH receptor present in particulate microsomal or plasma membranes. Accordingly, although these data have identified two classes of GH-binding protein, especially a primary GH-binding subunit of Mr 60,000-66,000, they indicate that, unlike studies on the insulin receptor, covalent cross-linking techniques alone are not sufficient to delineate the complete subunit structure of the native and endogenous form of the GH receptor.     

Isolation and characterization of two different forms of inositol phospholipid-specific phospholipase C from rat brain

Two different forms of inositol phospholipid-specific phospholipase C (PLC) have been purified 2810- and 4010-fold, respectively, from a crude extract of rat brain. The purification procedures consisted of chromatography of both enzymes on Affi-Gel blue and cellulose phosphate, followed by three sequential high performance liquid chromatography steps, which were different for the two enzymes. The resultant preparations each contained homogeneous enzyme with a Mr of 85,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. One of these enzymes (PLC-II) was found to hydrolyze phosphatidyl-inositol 4,5-bisphosphate (PIP2) at a rate of 15.3 mumol/min/mg of protein and also phosphatidylinositol 4-monophosphate and phosphatidylinositol (PI) at slower rates. For hydrolysis of PI, this enzyme was activated by an acidic pH and a high concentration of Ca2+ and showed a Vmax value of 19.2 mumol/min/mg of protein. The other enzyme (PLC-III) catalyzed hydrolysis of PIP2 preferentially at a Vmax rate of 12.9 mumol/min/mg of protein and catalyzed that of phosphatidylinositol 4-monophosphate slightly. The rate of PIP2 hydrolysis by this enzyme exceeded that of PI under all conditions tested. Neither of these enzymes had any activity on phosphatidylcholine, phosphatidylethanolamine, phosphatidylserine, or phosphatidic acid. These two enzymes showed not only biochemical but also structural differences. Western blotting showed that antibodies directed against PLC-II did not react with PLC-III. Furthermore, the two enzymes gave different peptide maps after digestion with alpha-chymotrypsin or Staphylococcus aureus V8 protease. These results suggest that these two forms of PLC belong to different families of PLC.     

The metabolism of neuropeptides. Endopeptidase-24.11 in human synaptic membrane preparations hydrolyses substance P.

Synaptic membrane preparations from human striatum and human diencephalon were shown to contain a phosphoramidon-sensitive metalloendopeptidase that appeared identical with endopeptidase-24.11. The activity of endopeptidase-24.11 was determined with an enzymic assay employing [D-Ala2,Leu5]enkephalin as substrate, and its distribution in human brain was similar to that in pig brain, with the striatum containing the highest levels. The choroid plexus and pons also contained substantial activity. A good correlation (r = 0.97) was obtained for the distribution of the endopeptidase in pig brain and pituitary by the enzymic assay and by an immunoradiometric assay specific for pig endopeptidase-24.11. Synaptic membrane preparations from human striatum and diencephalon hydrolysed substance P at the same sites as did preparations of pig striatal synaptic membranes, and hydrolysis was substantially abolished by phosphoramidon. These results suggest that endopeptidase-24.11 is the principal enzyme hydrolysing substance P in human synaptic membrane preparations.     

Control of glycoprotein synthesis. Purification of UDP-N-acetylglucosamine:alpha-D-mannoside beta 1-2 N-acetylglucosaminyltransferase II from rat liver

A new affinity chromatography adsorbant, in which UDP-GlcNAc has been linked to thiopropyl-Sepharose at the 5 position of the uracil via a 5-mercuri mercaptide bond, was utilized to purify UDP-GlcNAc:alpha-D-mannoside beta 1-2 N-acetylglucosaminyltransferase II 60,000-fold from rat liver. After extraction of rat liver membranes with Triton X-100, the enzyme was found to exist in two molecular weight forms of markedly differing size, separable on Sephadex G-200. The low Mr form was separated from the high Mr form on columns of CM-Sephadex and hydroxylapatite, and was further purified by sequential elutions with NaCl, UDP-GlcNAc, and EDTA from the 5-mercuri-UDP-GlcNAc affinity adsorbant. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the purified low Mr form under reducing conditions revealed two protein bands of Mr 48,000 and 43,000. The purified enzyme catalyzes the transfer of N-acetylglucosamine from UDP-GlcNAc to the compound: (Formula: see text) The high Mr form of the enzyme, which eluted in the void volume of Sephadex G-200, was resistant to a number of treatments in attempts to reduce its molecular weight. These results suggest that the high Mr form of the enzyme may represent either a complex which normally exists in Golgi membranes as a result of strong protein-protein interactions or a protein with one or more "anchor" segments.     

Molecular heterogeneity of angiotensin converting enzyme in human cerebrospinal fluid.

Three different molecular forms of angiotensin converting enzyme (ACE) (approximately Mr 150,000, 80,000 and 40,000, respectively), have been recovered from human cerebrospinal fluid. All three enzymes were inhibited by captopril and enalapril and their activity was potentiated by chloride ions. They were capable of degrading Leu-enkephalin-Arg6 and substance -P, but gave no conversion of neurokinin A. In all these aspects, the CSF enzymes were identical with the human pulmonary enzyme. The Mr 40,000 form of ACE is the smallest active form of the enzyme hitherto reported and is likely to represent a fragment of the C-terminal part of native ACE, where its active center is located.     

Receptor-mediated activation of spermatozoan guanylate cyclase

The sea urchin egg peptides speract (Gly-Phe-Asp-Leu-Asn-Gly-Gly-Gly-Val-Gly) and resact (Cys-Val-Thr-Gly-Ala-Pro-Gly-Cys-Val-Gly-Gly-Arg-Leu-NH2) bind to spermatozoa of the homologous species (Lytechinus pictus or Arbacia punctulata, respectively) and cause transient elevations of cyclic GMP concentrations (Hansbrough, J. R., and Garbers, D. L. (1981) J. Biol. Chem. 256, 1447-1452). The addition of these peptides to spermatozoan membrane preparations caused a rapid and dramatic (up to 25-fold) activation of guanylate cyclase. The peptide-induced activation of guanylate cyclase was transient, and the subsequent decline in enzyme activity coincided with conversion of a high Mr (phosphorylated) form of guanylate cyclase to a low Mr (dephosphorylated) form. When membranes were incubated at pH 8.0, the high Mr form was converted to the low Mr form without substantial changes in basal enzyme activity. However, the peptide-stimulated activity of the low Mr form of guanylate cyclase was much less than the peptide-stimulated activity of the high Mr form. Activation of the low Mr form by peptide was not transient and persisted for at least 10 min. In addition, the pH 8.0 treatment that caused the Mr conversion of guanylate cyclase also caused an increase in the peptide-binding capacity of the membranes. We propose a model in which activation of the membrane form of guanylate cyclase is receptor-mediated; the extent of enzyme activation is modulated by its phosphorylation state.     

THE SUBCELLULAR LOCALIZATION OF ACETYLCHOLINESTERASE AND ITS MOLECULAR FORMS IN PIG CEREBRAL CORTEX

Abstract— The distribution of acetylcholinesterase among the subcellular fractions of pig cerebral cortex was determined. The crude mitochondrial and microsomal fractions obtained by differential centrifugation accounted for 75% of the enzyme, with the remainder divided between the crude nuclear and soluble fractions.
The occurrence and distribution of the multiple molecular forms of AChE was the same in all four fractions with the dominant species of molecular weights 350,000, 270,000 and 60,000. Further purification of the mitochondrial fraction by density gradient centrifugation gave a series of membrane fractions with very similar multiple forms. The one possible exception was the fraction containing the purified synaptosomal membranes where one band of mol wt 270,000 predominated, although the other molecular weight entities were present. The electrophoretic pattern of AChE present in the fractionated microsomes was the same as in the crude preparation. The content and pattern of the multiple molecular forms of AChE was therefore the same in all fractions of pig brain, apart from that containing the purified synaptosomal membranes.     

Rat Brain Synaptosomal ATP:AMP-Phosphotransferase Activity

Adenylate kinase activity (ATP:AMP-phosphotransferase; EC 2.7.4.3) was studied in various subcellular fractions of rat brain tissues. Because of the presence of other adenosine nucleotide-utilizing enzymes, adenylate kinase activity was assayed in both the forward and reverse directions by using coupled enzyme systems and by using a specific adenylate kinase inhibitor, P1,P5-di(adenosine-5') pentaphosphate. As expected, the highest specific adenylate kinase activity (2.89 mumol/min/mg of protein) was detected in the cytosolic brain fraction. However, substantial enzyme activity (0.68 mumol/min/mg) was also found in the intact synaptosomal fraction isolated on Percoll/sucrose gradients. The increased specific enzyme activity of purified synaptosomes and the differences found between the kinetic parameters of the membrane-bound and cytosolic enzyme forms suggest that the synaptosomal adenylate kinase activity cannot be attributed to the small amount of contaminating cytosol present in our preparations. The adenylate kinase enzyme adhered to purified synaptic plasma membranes and was not released by washings with isoosmotic sucrose medium. The facts that the adenylate kinase enzyme activity could be measured in intact synaptosomal preparations and that both its substrates and its inhibitors do not cross intact plasma membranes support the possibility that the synaptosomal adenylate kinase is an ecto-enzyme.     

Metabolic stability of the LHRH antagonist antide to cell-surface peptidases

The susceptibility to hydrolysis of LHRH and the decapeptide analogue Antide has been compared. The hydrolysis of LHRH by pig kidney brush border membranes is inhibited by phosphoramidon (I50 = 5.6 nM) implicating endopeptidase-24.11 in the initiation of hydrolysis. Under conditions in which LHRH is fully degraded by brush border membranes, Antide was completely resistant to hydrolysis. Similar results were obtained with purified preparations of both endopeptidase-24.11 and angiotensin converting enzyme. These data confirm that the remarkable duration of action of Antide is due principally to its stability to hydrolysis by cell-surface peptidases.     

Purification of the phosphorylated night form and dephosphorylated day form of phosphoenolpyruvate carboxylase from Bryophyllum fedtschenkoi.

Phosphoenolpyruvate carboxylase of Bryophyllum fedtschenkoi was shown to exist in two forms: a night form, which is phosphorylated and has low sensitivity to inhibition by malate, and a day form, which is dephosphorylated and 10 times more sensitive to malate. The day and night forms of the enzyme were purified retaining their distinct malate sensitivities and phosphorylation states. The purified enzymes contained a major protein (subunit Mr 112,000) and a minor protein (subunit Mr 123,000). The two polypeptides appeared to have closely related amino acid sequences and were present in a similar ratio in extracts that had been prepared rapidly. The phosphate present in the night form of the enzyme was covalently bound to serine. It was not a catalytic intermediate. Alkaline phosphatase removed the phosphate group in vitro and increased the malate sensitivity of the enzyme to that observed for the day form. Both the day and night forms of the enzyme were probably tetramers, and their apparent Mr was lowered by the presence of malate, but was unaffected by Mg2+ ions, EDTA, a rise in pH or a 10-fold change in enzyme concentration. The rapid loss of malate sensitivity, observed in extracts of leaves prepared during the day and at night, was shown to be due to proteolysis of the enzyme. It was slowed in the presence of malate and by phosphorylation of the enzyme.     

Modulation of glucose transporters in rat diaphragm by sodium tungstate

The somatic isoform of angiotensin-converting enzyme (ACE) consists of two homologous domains (N- and C-domains), each bearing a catalytic site. We have used the two-domain ACE form and its individual domains to compare characteristics of different domains and to probe mutual functioning of the two active sites within a bovine ACE molecule. The substrate Cbz-Phe-His-Leu (N-carbobenzoxy-L-phenylalanyl-L-histidyl-L-leucine; from the panel of seven) was hydrolyzed faster by the N-domain, the substrates FA-Phe-Gly-Gly (N-(3-[2-furyl]acryloyl)-L-phenylalanyl-glycyl-glycine) and Hip-His-Leu (N-benzoyl-glycyl-L-histidyl-L-leucine) were hydrolyzed by both domains with equal rates, while other substrates were preferentially hydrolyzed by the C-domain. The inhibitor captopril ((2S)-1-(3-mercapto-2-methylpropionyl)-L-proline) bound to the N-domain more effectively than to the C-domain, whereas lisinopril ((S)-N(alpha)-(1-carboxy-3-phenylpropyl)-L-lysyl-L-proline) bound to equal extent with all ACE forms. However, active site titration with lisinopril assayed by hydrolysis of FA-Phe-Gly-Gly revealed that 1 mol of inhibitor/mol of enzyme abolished the activity of either two-domain or single-domain ACE forms, indicating that a single active site functions in bovine somatic ACE. Neither of the k(cat) values obtained for somatic enzyme was the sum of k(cat) values for individual domains, but in every case the value of the catalytic constant of the hydrolysis of the substrate by the two-domain ACE represented the mean quantity of the values of the corresponding catalytic constants obtained for single-domain forms. The results indicate that the two active sites within bovine somatic ACE exhibit strong negative cooperativity.     

Structural organization of membrane and soluble forms of somatic angiotensin-converting enzyme.

The catalytic activity and quaternary structure of soluble (s) and membrane (m) forms of angiotensin-converting enzyme (ACE) were studied in reversed micelles of ternary system Aerosol OT--water--octane. The profile of the dependence of the catalytic activity of the two enzyme forms on the degree of surfactant hydration (micellar size) had several optima corresponding to the function of various active oligomeric enzyme forms; the curves for the s- and m-forms of ACE were different. Data of sedimentation analysis prove that in reversed micelles, s-ACE can exist as monomers, dimers, or tetramers depending on the hydration degree, and the m-form is present as dimers and tetramers only. The values of the kinetic parameters for the hydrolysis of the substrate furylacryloyl-Phe-Gly-Gly by all the enzyme forms were determined, and the data indicate that the activity of the m-form is enhanced by oligomerization. The ACE activity strongly depends on the medium; it is higher when ACE is in contact with matrix or other enzyme molecules.     

Guinea pig proacrosin is synthesized principally by round spermatids and contains O-linked as well as N-linked oligosaccharide side chains

Proacrosin is the zymogen precursor of acrosin, a sperm protease believed to play an essential role in fertilization. In this study, we used primary cultures of guinea pig spermatogenic cells to examine the temporal appearance and mechanisms of synthesis and processing of proacrosin during acrosome development. Following [35S]methionine incorporation and immunoprecipitation, cultured spermatogenic cells were found to synthesize two forms of proacrosin (Mr 54,000 and 57,000). Proacrosin was synthesized mainly by round spermatids. By immunoblotting, proacrosin became very prominent in round spermatids and persisted throughout spermiogenesis. Pulse-chase experiments demonstrated that the Mr 54,000 form of proacrosin was converted to the Mr 57,000 form, presumably reflecting posttranslational processing of carbohydrate side chains. When spermatogenic cells were cultured in the presence of tunicamycin, the synthesized proacrosin had an Mr of 54,000. However, in vitro translation of mRNA extracted from guinea pig testis followed by immunoprecipitation indicated that the core polypeptide of proacrosin has an Mr of 44,000. Guinea pig spermatogenic cells incorporated glucosamine and fucose into the oligosaccharides of proacrosin. Treatment of guinea pig testis proacrosin with N-glycosidase or O-glycosidase reduced the Mr by 3-7%. These results indicate that proacrosin is synthesized by postmeiotic cells and the enzyme contains N- and O-linked oligosaccharides.