Pepsin-mediated processing of synthetic precursor-like sequence yields neurotensin-like peptide.

A 15 amino acid synthetic peptide, which spanned the dibasic cleavage site C-terminal to neurotensin (NT), in its 170-residue canine precursor, was synthesized by solid-phase methods. Using this substrate in combination with a radioimmunoassay specific for the C-terminal region of NT, a simple assay was developed to monitor protease-mediated cleavage of the Leu8-Lys9 bond in the substrate. Hog pepsin and the related enzymes, rhizopus pepsin, bovine cathepsin D, and mouse renin, were found to be effective in this assay, pepsin cleaving only this bond to liberate the NT-like sequence. The pH dependence of the reaction indicated that pepsin, cathepsin D, and renin exhibited significant activity at pH's characteristic for secretory vesicles (pH 5.5-6.5). In addition, pepsin and cathepsin D were shown to process the native precursor at pH's as high as 5.5. These results, although not proof, are consistent with the idea that endoproteases with pepsin-like specificity may be involved in the processing of the NT precursor in neural/endocrine cells.     

Co-localization of xenopsin and gastrin immunoreactivity in gastric antral G-cells

Studies indicating evidence for the presence of the amphibian octapeptide xenopsin in gastric mucosa of mammals prompted us to investigate the cellular localization of this peptide. Using the peroxidase-antiperoxidase method and a specific antiserum to xenopsin (Xen-7) on paraffin and adjacent semithin sections of gastric antral mucosa from man, dog, and Tupaia belangeri, we found numerous epithelial cells showing a specific positive immunoreaction. These cells were of typical pyramidal shape and could be classified as of the "open" type. Cell quantification in serial sections processed for xenopsin and gastrin immunoreactivity, respectively, revealed an identical number of cells per section and an identical distribution of these cells in the middle zone of the antral mucosa. Furthermore, adjacent semithin sections demonstrated the colocalization of xenopsin and gastrin immunoreactivity within the same G-cell. The xenopsin antiserum could be completely absorbed with synthetic xenopsin but not with gastrin. Preabsorption tests with neurotensin, a xenopsin related peptide, or with somatostatin, glucagon, and enkephalins gave no evidence for crossreactivity of the xenopsin antiserum with these peptides. It is concluded that gastric antral G-cells in addition to gastrin also contain the amphibian peptide xenopsin.     

Xenopsin-related peptide(s) are formed from xenopsin precursor by leukocyte protease(s) and cathepsin D

Lysates of isolated rat polymorphonuclear leukocytes and macrophages were found to generate xenopsin-related peptides when incubated with a liver extract used as a source of precursor. The lysosomal enzyme, cathepsin D, was also shown to display this property and to share with the lysate a similar pH dependence (optimum, approximately pH 3.5) and sensitivity to the acid protease inhibitor, pepstatin A (ID50: lysate, 10 nM; cathepsin D, 30 nM). When subjected to HPLC on mu-Bondapak C-18, the xenopsin-related peptides generated by the lysate eluted near to those formed by cathepsin D and when tested in a radioreceptor assay for neurotensin, they displayed similar cross-reactivities (peak 2, approximately 50%; peak 1, approximately 100%). These results indicate that cathepsin D from lysed granulocytes can process precursor protein(s) to form radioreceptor-active xenopsin-related peptides.     

Isolation and sequence of canine xenopsin and an extended fragment from its precursor

Canine xenopsin and a 27 residue segment of its precursor immediately surrounding the xenopsin moiety were isolated from acidic extracts of stomach. The six C-terminal residues of canine xenopsin, H-Phe-His-Pro-Lys-Arg-Pro-Trp-Ile-Leu-OH, were identical to those in Xenopus xenopsin less than Glu-Gly-Lys-Arg-Pro-Trp-Ile-Leu-OH. The amino acid sequence determined for the segment of the precursor was similar to the corresponding region of Xenopus pro-xenopsin (approximately 33% homology) and to the related Xenopus precursors, pro-levitide, pro-PGLa, pro-magainin and pro-caerulein. These results, indicating evolutionary conservation of xenopsin and a portion of its precursor, suggest that this peptide has important biologic function(s).     

Localization of xenopsin and xenopsin precursor fragment immunoreactivities in the skin and gastrointestinal tract of Xenopus laevis

Summary Xenopsin (Xp) and xenopsin precursor fragment (XPF) are bioactive peptides derived from a single precursor molecule; both were isolated previously from extracts of Xenopus laevis skin. The present immunohistochemical study was undertaken to determine the specific cellular localization of these two peptides in the skin and also in the gastrointestinal tract of adult Xenopus. We report here that Xp-like and XPF-like immuno-reactivities co-exist in the granular glands of the skin and specific granular cells in the lower esophagus and stomach. However, only Xp-like immunoreactivity, not XPF-like immunoreactivity, was detected in tall, thin cells of the duodenum and in club-shaped cells of the large intestine. The immunochemical co-localization of the two peptides in specific cells of the skin, lower esophagus and stomach suggests that the same gene is expressed in each of these cells, and that the precursor molecule undergoes similar post-translational processing. In contrast, the observation that certain cells of the duodenum and large intestine display only one peptide immunoreactivity suggests an alternative phenomenon, possibly involving selective peptide accumulation or expression of a different gene.     

Xenopsin immunoreactivity in antral G-cells may reside in the N-terminus of gastrin 17

The nature of xenopsin immunoreactivity in mammalian antral G-cells has been reassessed. Xenopsin immunostaining was most intense in human antral G-cells, present in those of the dog and pig and not detected in guinea pig or rat tissues. Rigorous specificity controls for ionic binding of immunoglobulins to antral G-cell granules indicated that this mechanism was not responsible for xenopsin immunostaining. Preincubation of the xenopsin antiserum with xenopsin, human gastrin 1-13 and gastrin 2-17 completely abolished immunostaining at similar molar concentrations. Gastrin 34 was ineffective at much higher concentrations. These results infer that xenopsin-immunoreactivity in antral G-cells resides in the N-terminal region of gastrin 17. Examination of the primary structures of xenopsin and the N-terminal regions of some mammalian gastrins reveals a hitherto unrecognized homology.     

Identification of xenin, a xenopsin-related peptide, in the human gastric mucosa and its effect on exocrine pancreatic secretion.

One of the peptides previously discovered in amphibians is the octapeptide xenopsin. As immunohistochemistry has also indicated the presence of xenopsin immunoreactivity in man, we extracted in the present investigation xenopsin-immunoreactive material from human gastric mucosa and purified it to homogeneity with several high performance liquid chromatography (HPLC) reverse phase and ion exchange chromatographic steps. The eluates were monitored with a radioimmunoassay for amphibian xenopsin. Determination of the amino acid sequence revealed a 25-amino acid peptide having 6 C-terminal amino acids in common with amphibian xenopsin. The sequence of this peptide, termed xenin 25, is M-L-T-K-F-E-T-K-S-A-R-V-K-G-L-S-F-H-P-K-R-P-W-I-L. The peptide was custom-synthesized. Mass spectrometry of the synthetic and the extracted peptide revealed identical molecular mass. Purification of 250 ml of human postprandial plasma with Sep-Pak C18 cartridges, reverse phase HPLC, and ion exchange chromatography demonstrated circulating xenin immunoreactivity at a retention time identical to xenin 25. The amount of xenin immunoreactivity at the position of xenin 25 on C18-HPLC increased significantly after a meal. A radioimmunoassay utilizing antibodies to xenin 25 and a 125I-labeled analogue of xenin 25 was used to measure immunoreactive xenin in the plasma of 10 volunteers. There was a significant rise of xenin immunoreactivity in the plasma after a meal. Intravenous infusion of the synthetic peptide in dogs stimulated exocrine pancreatic secretion beginning at a dose of 4 pmol/kg/min. The maximal effect was seen with 64 pmol/kg/min. We have detected, therefore, a new peptide, xenin 25, in human gastric mucosa; we have provided evidence for the presence of this peptide in the human circulation, and have shown a rise of plasma xenin concentrations after a meal. This peptide stimulates exocrine pancreatic secretion. Its physiologic role deserves further investigation.     

Xenopsin immunoreactivity in antral G-cells may reside in the N-terminus of gastrin 17

Summary The nature of xenopsin immunoreactivity in mammalian antral G-cells has been reassessed. Xenopsin immunostaining was most intense in human antral G-cells, present in those of the dog and pig and not detected in guinea pig or rat tissues. Rigorous specificity controls for ionic binding of immunoglobulins to antral G-cell granules indicated that this mechanism was not responsible for xenopsin immunostaining. Preincubation of the xenopsin antiserum with xenopsin, human gastrin 1–13 and gastrin 2–17 completely abolished immunostaining at similar molar concentrations. Gastrin 34 was ineffective at much higher concentrations. These results infer that xenopsin-immunoreactivity in antral G-cells resides in the N-terminal region of gastrin 17. Examination of the primary structures of xenopsin and the N-terminal regions of some mammalian gastrins reveals a hitherto unrecognized homology.     

Occurrence of neurotensin-immunoreactive cells in the digestive tract of lower vertebrates and deuterostomian invertebrates

Summary In the mucosal epithelium of the digestive tract of two marine teleost bony fish, one cartilaginous fish, one cyclostome, and in that of two of three representatives of deuterostomian invertebrates studied, endocrine cells of open type were found, exhibiting immunoreactivity with antisera against C-terminal sequences of mammalian neurotensin and of the structurally closely related amphibian neurohormonal peptide xenopsin.From these observations, and from those of previous studies, it is suggested that neurotensin cells do not occur in the digestive tract mucosa until at the evolutionary level of the more highly developed deuterostomian invertebrates. Three evolutionary stages seem to exist in the distribution pattern. The first stage, characterized by few, widely scattered cells, is found in the uro- and cephalochordates, the cyclostomes, the cartilaginous fish, and the stomachless bony fish. In the second stage, comprising the remaining submammalian classes, including more highly developed bony fish, the typical distribution pattern is that of numerous neurotensin immunoreactive cells in the antrum, pylorus, and duodenum. The final stage of neurotensin evolution is found in higher mammals and is characterized by a great density of neurotensin immunoreactive cells in the ileum.Dedicated to Prof. Dr. J. Staubesand on the occasion of this 60th birthday     

Human cathepsin D precursor is associated with a 60 kDa glycosylated polypeptide.

Human cathepsin D is synthesized as a 53 kDa precursor. Most of it is segregated into lysosomal compartments and subjected to a proteolytic fragmentation. Using a cross-linking reagent we show that a large proportion of the precursor is associated with a distinct protein which--under denaturing and reducing conditions--is characterized as a 60 kDa glycopeptide. Studies on cells cultured in the presence of drugs known to affect the intracellular transport (deoxynojirimycin, brefeldin A and NH4Cl) indicated that the association with cathepsin D precursor occurs early after the synthesis and is at least partially maintained after secretion.     

Peripheral inactivation of neurotensin. Isolation and characterization of a metallopeptidase from rat ileum

A peptidase that inactivated neurotensin by cleaving the peptide at the Pro10-Tyr11 bond, generating the biologically inactive fragments neurotensin(1-10) and neurotensin(11-13) was purified from whole rat ileum homogenate. The purified enzyme behaved as a 70-75-kDa monomer as determined by SDS-PAGE analysis in reducing or non-reducing conditions and gel permeation on Ultrogel AcA34. The peptidase was insensitive to thiol-blocking agents and acidic and serine protease inhibitors but could be strongly inhibited by 1,10-phenanthroline, EDTA, dithiothreitol and heavy metal ions such as zinc, copper and cobalt. Zinc was the only divalent cation able potently to reactivate the apoenzyme. This enzyme could be distinguished from endopeptidases EC 3.4.24.15 and EC 3.4.24.11, angiotensin-converting enzyme, proline endopeptidase, aminopeptidase and pyroglutamyl-peptide hydrolase since it was not affected by micromolar concentrations of their specific inhibitors. The peptidase displayed a high affinity for neurotensin (1.6 microM). Studies concerning the specificity of the enzyme towards the sequence of neurotensin established the following. (a) Neurotensin(9-13) was the shortest partial sequence that fully inhibited tritiated neurotensin degradation; shortening the C-terminal part of the neurotensin molecule led to inactive fragments. (b) Amidation of the C-terminal end of the peptide did not prevent the recognition by the peptidase. (c) There existed a strong stereospecificity of the peptidase for the residues in positions 8, 9 and 11 of the neurotensin molecule. (d) Pro-Xaa dipeptides (where Xaa represented aromatic or hydrophobic residues) were the most potent inhibitors of tritiated neurotensin degradation while all the Xaa-Pro dipeptides tested were totally ineffective. (e) The neurotensin-related peptides: neuromedin N, xenopsin and [Lys8-Asn9]neurotensin(8-13), as well as angiotensins I and II and dynorphins(1-8) and (1-13) were as potent as neurotensin in inhibiting [3H]neurotensin hydrolysis.     

Purification and characterization of a novel neurotensin-degrading peptidase from rat brain synaptic membranes

A peptidase that cleaved neurotensin at the Pro10-Tyr11 peptide bond, leading to the formation of neurotensin-(1-10) and neurotensin-(11-13), was purified nearly to homogeneity from rat brain synaptic membranes. The enzyme appeared to be monomeric with a molecular weight of about 70,000-75,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and high pressure liquid chromatography filtration. Isoelectrofocusing indicated a pI of 5.9-6. The purified peptidase could be classified as a neutral metallopeptidase with respect to its sensitivity to pH and metal chelators. Thiol-blocking agents and acidic and serine protease inhibitors had no effect. Studies with specific peptidase inhibitors clearly indicated that the purified enzyme was distinct from enzymes capable of cleaving neurotensin at the Pro10-Tyr11 bond such as proline endopeptidase and endopeptidase 24-11. The enzyme was also distinct from other neurotensin-degrading peptidases such as angiotensin-converting enzyme and a recently purified rat brain soluble metalloendopeptidase. The peptidase displayed a high affinity for neurotensin (Km = 2.6 microM). Studies on its specificity revealed that neurotensin-(9-13) was the shortest neurotensin partial sequence that was able to fully inhibit [3H]neurotensin degradation. Shortening the C-terminal end of the neurotensin molecule as well as substitutions in positions 8, 9, and 11 by D-amino acids strongly decreased the inhibitory potency of neurotensin. Among 20 natural peptides, only angiotensin I and the neurotensin-related peptides (xenopsin and neuromedin N) were found as potent as unlabeled neurotensin.     

Differential processing of the neurotensin/neuromedin N precursor in the mouse

Posttranslational processing of the neurotensin/neuromedin N (NT/NN) precursor has been investigated in mouse brain and small intestine by means of region-specific radioimmunoassays coupled to chromatographic fractionations. In brain, total NT/NN immunoreactivity measured with a common C-terminal antiserum was 15.72 pmol/g. NT measured with an N-terminal antiserum was 9.74 pmol/g and NN measured with an N-terminal antiserum was 5.98 pmol/g. In small intestine, combined NT/NN immunoreactivity was 108.55 pmol/g, consisting of 66.37 pmol/g NT but only 0.96 pmol/g NN. Gel permeation chromatography and reverse phase HPLC revealed that the large discrepancy in the NT and NN values obtained in small intestinal extracts was due to the presence of a high molecular weight, hydrophobic peptide, which was reactive only with the common C-terminally directed antiserum. Pepsinization of this generated an immunoreactive peptide with similar chromatographic characteristics to NN. In mouse intestine, NN is only partially cleaved from the common NT/NN precursor, resulting in the presence of an N-terminally extended molecular species. This novel molecular species of neuromedin N may be the physiological mediator of certain peripheral biological effects hitherto attributed to neurotensin or neuromedin N.     

Characterisation of cathepsin B and collagenolytic cathepsin from human placenta

1. Human placental cathepsin B and collagenolytic cathepsin were separated by chromatography on columns of Amberlite CG-50. Collagenolytic cathepsin was partially purified by chromatography on DEAE-Sephadex (A-50) and Sephadex G-100. Cathepsin B was purified by chromatography on CM-cellulose and Sephadex G-100. 2. Both enzymes required activation by thiol compounds and were bound to organomercurial-Sepharose-4B. Sulphydryl-blocking reagents were inhibitory, which confirmed an essential thiol group to be present. 3. The enzymes degraded soluble calf skin collagen and insoluble bovine tendon collagen in the telopeptide region at pH 3.5 and 28 degrees C to yield mainly alpha-chain components. 4. In contrast to cathepsin B, collagenolytic cathepsin was found not to hydrolyse any of the low-molecular-weight synthetic substrates that were tested. 5. Leupeptin, a structural analogue of arginine-containing synthetic substrates, and antipain, an inhibitor of papain, were strongly inhibitory to both enzymes. 6. The isoelectric points of the enzymes were similar, being 5.4 for cathepsin B and 5.1 for collagenolytic cathepsin. 7. From chromatography on Sephadex G-100 the molecular weight of cathepsin B was calculated to be 24 500 and that of collagenolytic cathepsin to be 34 600.     

Evaluation of Cathepsins D and G and EC 3.4.24.15 as Candidate β-Secretase Proteases Using Peptide and Amyloid Precursor Protein Substrates

Abstract: No single protease has emerged that possesses all the expected properties for β-secretase, including brain localization, appropriate peptide cleavage specificity, and the ability to cleave amyloid precursor protein exactly at the amino-terminus of β-amyloid peptide. We have isolated and purified a brain-derived activity that cleaves the synthetic peptide substrate SEVKMDAEF between methionine and aspartate residues, as required to generate the amino-terminus of β-amyloid peptide. Its molecular size of 55–60 kDa and inhibitory profile indicate that we have purified the metalloprotease EC 3.4.24.15. We have compared the sequence specificity of EC 3.4.24.15, cathepsin D, and cathepsin G for their ability to cleave the model peptide SEVKMDAEF or related peptides that contain substitutions reported to modulate β-amyloid peptide production. We have also tested the ability of these enzymes to form carboxy-terminal fragments from full-length, membrane-embedded amyloid precursor protein substrate or amyloid precursor protein that contains the Swedish KM → NL mutation. The correct cleavage was tested with an antibody specific for the free amino-terminus of β-amyloid peptide. Our results exclude EC 3.4.24.15 as a candidate β-secretase. Although cathepsin G cleaves the model peptide correctly, it displays poor ability to cleave the Swedish KM → NL peptide and does not generate carboxy-terminal fragments that are immunoreactive with amino-terminal-specific antiserum. Cathepsin D does not cleave the model peptide or show specificity for wild-type amyloid precursor protein; however, it cleaves the Swedish "NL peptide" and "NL precursor" substrates appropriately. Our results suggest that cathepsin D could act as β-secretase in the Swedish type of familial Alzheimer's disease and demonstrate the importance of using full-length substrate to verify the sequence specificity of candidate proteases.     

Stimulatory effect of xenin-8 on insulin and glucagon secretion in the perfused rat pancreas

Xenin is a 25-amino acid peptide of the neurotensin/xenopsin family identified in gastric mucosa as well as in a number of tissues, including the pancreas of various mammals. In healthy subjects, plasma xenin immunoreactivity increases after meals. Infusion of the synthetic peptide in dogs evokes a rise in plasma insulin and glucagon levels and stimulates exocrine pancreatic secretion. The latter effect has also been demonstrated for xenin-8, the C-terminal octapeptide of xenin. We have investigated the effect of xenin-8 on insulin, glucagon and somatostatin secretion in the perfused rat pancreas. Xenin-8 stimulated basal insulin secretion and potentiated the insulin response to glucose in a dose-dependent manner (EC(50)=0.16 nM; R(2)=0.9955). Arginine-induced insulin release was also augmented by xenin-8 (by 40%; p<0.05). Xenin-8 potentiated the glucagon responses to both arginine (by 60%; p<0.05) and carbachol (by 50%; p<0.05) and counteracted the inhibition of glucagon release induced by increasing the glucose concentration. No effect of xenin-8 on somatostatin output was observed. Our observations indicate that the reported increases in plasma insulin and glucagon levels induced by xenin represent a direct influence of this peptide on the pancreatic B and A cells.     

Co-localization of xenopsin and gastrin immunoreactivity in gastric antral G-cells

Summary Studies indicating evidence for the presence of the amphibian octapeptide xenopsin in gastric mucosa of mammals prompted us to investigate the cellular localization of this peptide. Using the peroxidase-antiperoxidase method and a specific antiserum to xenopsin (Xen-7) on paraffin and adjacent semithin sections of gastric antral mucosa from man, dog, and Tupaia belangeri, we found numerous epithelial cells showing a specific positive immunoreaction. These cells were of a typical pyramidal shape and could be classified as of the open type. Cell quantification in serial sections processed for xenopsin and gastrin immunoreactivity, respectively, revealed an identical number of cells per section and an identical distribution of these cells in the middle zone of the antral mucosa. Furthermore, adjacent semithin sections demonstrated the colocalization of xenopsin and gastrin immunoreactivity within the same G-cell. The xenopsin antiserum could be completely absorbed with synthetic xenopsin but not with gastrin. Preabsorption tests with neurotensin, a xenopsin related peptide, or with somatostatin, glucagon, and enkephalins gave no evidence for crossreactivity of the xenopsin antiserum with these peptides.It is concluded that gastric antral G-cells in addition to gastrin also contain the amphibian peptide xenopsin.     

Endocytosis and the recycling of plasma membrane

For study of the time order of glycosylation, formation of complex oligosaccharides and proteolytic maturation as well as the site of proteolytic maturation of cathepsin D, fibroblasts were subjected to pulse-chase labeling, and cathepsin D was isolated from either total cell extracts or subcellular fractions by immune precipitation and analyzed for its molecular forms and sensitivity to endo-beta-N- acetylglucosaminidase H. After a 10-min pulse, cathepsin D was detected in its glycosylated precursor form, indicating an early, probably a cotranslational, N-glycosylation of cathepsin D. Conversion of the high- mannose oligosaccharide side chains into forms resistant to endo-beta-N- acetylglucosaminidase H started after approximately 40 min, indicating that transport of cathepsin D from the endoplasmic reticulum to the trans-Golgi apparatus requires approximately 40 min. Processing of the 53-kdalton precursor polypeptide of cathepsin D to a 47-kdalton intermediate followed about 20 min after the formation of complex oligosaccharides, and, another 30 min later, 31-kdalton mature forms of cathepsin D were detected. Processing of cathepsin D was first observed in light membranes as a partial conversion of the 53-kdalton precursor into the 47-kdalton intermediate. Both the precursor and the intermediate are transferred into the high density-class lysosomes. After 8 h, the processing to the mature 31-kdalton form of cathepsin D is mostly completed.     

Further characterization of a neurotensin-degrading neutral metalloendopeptidase from rat brain

A peptidase inactivating neurotensin at the Pro¹⁰-Tyr¹¹ peptidyl bond, leading to the biologically inactive fragments neurotensin_1–10 and neurotensin_11–13 was purified from rat brain homogenate. The peptidase was characterized as a 70 kDa monomer and could be classified as a metaliopeptidase with respect to its sensitivity to o-phenanthroline, EDTA and divalent cations. The enzyme was also strongly inhibited by dithiothreitol but appeared totally insensitive to thiol-blocking agents, acidic and serine protease inhibitors. Experiments performed with a series of highly specific peptidase inhibitors clearly indicated that the peptidase was a novel enzyme distinct from previously purified cerebral peptidases. The enzyme displayed a rather high affinity for neurotensin (Km = 2.3 itM). Studies on its specificity indicated that: (i) neurotensin_9–13 was the shortest neurotensin fragment with full inhibitory potency of [³H]neurotensin degradation. Shortening the C-terminal end of the neurotensin molecule progressively led to inactive analogs; (ii) the peptidase exhibited a strong stereospecificity towards the residues in positions 8, 9 and 11. By contrast, neither introduction of a steric hindrance in position 11 nor amidation of the C-terminal end of the neurotensin molecule affected the ability of the corresponding analog to inhibit [³H]neurotensin degradation; (iii) Pro-Phe was the most potent dipeptide to compete for [³H]neurotensin degradation; (iv) the peptidase could not be described as an exclusive “neurotensinase” activity since, in addition to the neurotensin natural analogs (neuromedin N and xenopsin), non related natural peptides such as angiotensins I and II, dynorphins 1–8 and 1–13, atriopeptin III and bradykinin potently inhibited [³H]neurotensin degradation. Most of these peptides behaved as substrates for the enzyme.     

Membrane bound pituitary metalloendopeptidase: Apparent identity to enkephalinase

A membrane bound zinc-metalloendopeptidase from bovine pituitaries with a specificity toward bonds on the amino side of hydrophobic amino acids, cleaves Met- and Leu-enkephalin at the Gly-Phe bond, releasing Phe-Met and Phe-Leu respectively. The enzyme also hydrolyzes bonds on the amino side of hydrophobic amino acids in oxytocin, bradykinin, neurotensin and several synthetic substrates. A free carboxyl group on a dipeptide C-terminal to the hydrolyzed bond is not a requirement for activity. The enzyme is also present in brain membrane fractions. The regional distribution of this enzyme in brain, its specificity toward natural and synthetic substrates, and its sensitivity to inhibitors, suggest that the enzyme is identical to an activity referred to as “enkephalinase”, which has been described as dipeptidyl carboxypeptidase. The data show that the enzyme is an endopeptidase with a specificity similar to that of a group of microbial proteases, one of which is thermolysin.