Specific cleavage of beta-LPH and ACTH by tonin: release of an opiate-like peptide beta-LPH (61-78).

Tonin, a rat enzyme capable of cleaving angiotensinogen, the tetradecapeptide renin substrate and angiotensin I directly to antiotensin II is also shown to cleave beta-lipotropin into beta-LPH 1–50, 1–51, 51–60, 52–60, 61–78 and 79–91, thereby selectively releasing the opiate-like segment beta-LPH 61–78. Its action on ACTH was similar, releasing ACTH 1–8, 1–7, 3–8, 3–7 and 9–39. In both situations the cleavages are of a selective tryptic-chymotryptic type at specific arginine, phenylalanine residues. Comparison of the tonin cleavage with those of trypsin, trypsin in combination with citraconylation of the lysine residues of beta-LPH is made. The data presented show that tonin does not cleave Met-enkephalin and can be used as an enzyme to study the presence of endorphin-like sequences in polypeptides.     

Primary structure and morphine-like activity of human beta-endorphin.

The complete amino acid sequence of human beta-endorphin was obtained by automatic sequencing of a sulfonyl isothiocyanate derivative of this peptide, in combination with peptide mapping of a tryptic digest of the native molecule. It was found to be identical with the carboxy-terminal portion 61-91 of human beta-lipotropin (beta-LPH). The morphine-like activity of beta-endorphin is comparable both in the mouse vas deferens bioassay and in the opiate receptor binding assay. However, beta-LPH is not active up to concentrations of 10(-6) M.     

Differential immunostaining of adenohypophysial cells with antisera to ACTH and beta-endorphin

Three antisera, raised in rabbits against human beta-endorphin were denoted Y-10, Y-18 and R-230 and tested for their ability to immunohistochemically stain cells in teh adenohypophysis of the rat and dog. By radioimmunoassay Y-10 was highly specific towards beta-endorphin with minimal cross-reactivity against beta-LPH while Y-18 and R-230 cross-reacted with both beta-endorphin and beta-LPH on an equimolar basis. In the rat pituitary, Y-18 and R-230 antisera stained cells identified as corticotrophs. With the dog, however, beta-endorphin-staining cells in the anterior pituitary and infundibulum did not necessarily co-stain for ACTH-like material. When the beta-endorphin antiserum, Y-10, was applied to rat anterior pituitary tissues, the subsequent positive immunocytochemical staining was associated not only with corticotrophs but also with somatotrophs. The findings are consistent with (a) a differential processing of the 31K pro-opiocortin and (b) the presence in rat somatotrophs of determinants that cross-react immunologically with some beta-endorphin antisera.     

A direct radioimmunoassay for tonin.

A sensitive radioimmunoassay for the measurement of tonin is described. It is based on the use of antibodies produced in rabbits against highly purified tonin obtained from rat submaxillary gland. A radioiodinated enzyme with high specific activity was obtained by the chloramine-T method. Optimal conditions for radioimmunoprecipitation were established and the double-antibody method was used to separate bound and free tonin. The radioimmunoassay may be used to measure tonin in amounts as low as 150 pg. This radioimmunoassay was applied to the measurement of tonin in rat saliva and in homogenates of submaxillary glands. Excellent correlation was found between tonin activity measured fluorometrically and tonin concentration obtained by the radioimmunoassay.     

Pro-opiocortin peptides in rat cerebrospinal fluid

Cerebrospinal fluid (CSF) taken from rats implanted with chronic cisternal cannulae was subjected to gel filtration chromatography on Sephadex G-50. Fractions were monitored using radioimmunoassays for N-terminal pro-opiocortin (N-POC), gamma 3-melanotropin (gamma 3-MSH), C-terminal adrenocorticotropin (C-ACTH), alpha-endorphin, beta-endorphin, gamma-lipotropin (gamma-LPH) and alpha-MSH. Two peaks which corresponded in elution position to rat N-POC (1-74) and gamma 3-MSH were detected. The major C-ACTH-immunoreactive (IR) peak was found to correspond to 14k ACTH. While no alpha-endorphin immunoreactivity was detected in rat CSF, three beta-endorphin-IR peaks were identified in positions expected for beta-LPH, beta-endorphin (1-31) and beta-endorphin (1-27), as well as a major peak of activity with the elution characteristics and cross-reactivity of rat gamma-LPH. HPLC of the alpha-MSH-IR material in rat CSF revealed the presence of a major peak of immunoreactivity whose retention time did not correspond to the known oxidised and reduced forms of alpha-MSH and its desacetylated and diacetylated derivatives. The identity of this peak is unknown.     

Post-translational processing of pro-opiomelanocortin (POMC)-derived peptides during fetal monkey pituitary development. II. beta-Lipotropin (beta-LPH)-related peptides

We have studied the post-translational processing of POMC-derived peptides during fetal monkey pituitary development using immunoassay and reverse-phase high-performance liquid chromatography (RP HPLC). Whole pituitary glands obtained from Day 50 and 55 fetal monkeys and separated lobes From Day 65 to 155 were extracted, fractionated, and analyzed for beta-melanotropin (beta-MSH), midportion beta-endorphin (beta-EP), and acetylated beta-EP immunoactivity. Separated adult pituitary lobes were analyzed for comparison. At Day 50, POMC-containing cells were located in both the anterior and intermediate pituitary lobes by immunofluorescence staining, the majority of these cells were localized in the anterior lobe. The Day 50 and 55 whole pituitaries contained predominantly beta-lipotropin (beta-LPH), gamma-lipotropin (gamma-LPH), beta-EP(1-31), and 2.2-kda beta-MSH. No acetylated products were found in Day 50 whole pituitary extracts. By Day 55, carboxy-shortened and acetylated beta-EPs were barely detectable in whole pituitary extracts. These forms were more apparent in the Day 65 separated neurointermediate lobe (NIL) extracts, and were similar to adult proportions by Day 80. The adult anterior lobe contained predominantly beta-LPH, beta-EP, and gamma-LPH. Adult NILs contained almost exclusively 2.2-kda beta-MSH, alpha-N-acetyl beta-EP(1-31) and alpha-N-acetyl beta-EP(1-27). The production of 2.2-kda beta-LPH in the monkey NIL indicates that monkey beta-LPH is different from rat beta-LPH in that it must contain the paired-basic cleavage site required for the formation of 2.2-kda beta-MSH that is known to be lacking in rat beta-LPH. Another finding was that monkey beta-EP contains a Tyr residue at position 27 as found in human beta-EP but appears to have the rat Gln substitution at position 31. The post-translational processing patterns characteristic of each lobe were well established by midterm fetal development (Day 80).     

Effect of met-enkephalin and (d-Met2,Pro5)-enkephalinamide on the adenylate cyclase activity of rat brain

Among opiatelike peptides, stimulatory as well as inhibitory effects are encountered on adenylate cyclase activity. These actions are dependent not only on the investigated brain region but also on the applied peptide. Met-enkephalin stimulates adenylate cyclase activity in the rat brain stem (D-Met2, Pro5)-enkephalinamide and beta-endorphin inhibited it, whereas all three peptides inhibited the activity of cortex. Naloxone antagonized the effects of the applied peptides in the presence of sodium chloride.     

In vitro processing of the artificial analog of beta-lipotropin

The processing of the recombinant analogue of beta-lipotropin (beta-LHP) having 11 additional N-terminal amino acid residues and separated from the hormone by the processing signal, was investigated using rat adrenal secretory granule lysate as a test system of processing "in vitro". It was found that incubation of the beta-LPH analogue with secretory granule enzymes leads to its limited specific degradation with a release of native beta-endorphin. It is concluded that the additional N-terminal amino acids induced no qualitative changes in beta-LPH processing.     

Chromogenic and fluorogenic glycosylated and acetylglycosylated peptides as substrates for serine, thiol and aspartyl proteases.

We synthesized short chromogenic peptidyl-Arg-p-nitroanilides containing either (Galbeta)Ser or (Glcalpha,beta)Tyr at P2 or P3 sites as well as O-acetylated sugar moieties and studied their hydrolysis by bovine trypsin, papain, human tissue kallikrein and rat tonin. For comparison, the susceptibility to these enzymes of Acetyl-X-Arg-pNa and Acetyl-X-Phe-Arg-pNa series, in which X was Ala, Phe, Gln and Asn were examined. We also synthesized internally quenched fluorescent peptides with the amino acid sequence Phe8-His-Leu-Val-Ile-His-Asn14 of human angiotensinogen, in which [GlcNAcbeta]Asn was introduced before Phe8 and/or after His13 and ortho-aminobenzoic acid (Abz) and N-[2-, 4-dinitrophenyl]-ethylenediamine (EDDnp) were attached at N- and C-terminal ends as a donor/receptor fluorescent pair. These peptides were examined as substrates for human renin, human cathepsin D and porcine pepsin. The chromogenic substrates with hydrophilic sugar moiety increased their susceptibility to trypsin, tissue kallikrein and rat tonin. For papain, the effect of sugar depends on its position in the substrate, namely, at P3 it is unfavorable, in contrast to the P2 position that resulted in increasing affinity, as demonstrated by the higher inhibitory activity of Ac-(Gal3)Ser-Arg-pNa in comparison to Ac-Ser-Arg-pNa, and by the hydrolysis of Ac-(Glcalpha,beta)Tyr-Arg-pNa. On the other hand, the acetylation of sugar hydroxyl groups improved hydrolysis of the susceptible peptides to all enzymes, except tonin. The P'4 glycosylated peptide [Abz-F-H-L-V-I-H-(GIcNAcbeta)N-E-EDDnp], that corresponds to one of the natural glycosylation sites of angiotensinogen, was shown to be the only glycosylated substrate susceptible to human renin, and was hydrolysed with lower K(m) and higher k(cat) values than the same peptide without the sugar moiety. Human cathepsin D and porcine pepsin are more tolerant to substrate glycosylation, hydrolysing both the P'4 and P4 glycosylated substrates.     

Purification and characterization of a serine protease (esterase B) from rat submandibular glands

A new protease has been purified to homogeneity from rat submandibular gland homogenate by using DEAE-Sephadex chromatography, chromatofocusing, aprotinin-Sepharose affinity chromatography, and high-performance liquid chromatography. The enzyme has been named esterase B, since it represents the second major esterolytic peak on DEAE-Sephadex chromatography of submandibular gland homogenate. It is an acidic protein (pI = 4.45) with an apparent molecular weight of 27 000. It is heat-stable and has an optimum pH of 9.5. Esterase B hydrolyzed the synthetic substrates tosyl-L-arginine methyl ester and Val-Leu-Arg-p-nitroanilide (S2266). It also cleaved dog plasma kininogen to produce a kinin, identified as bradykinin on reverse-phase high-performance liquid chromatography. Esterase B, however, is only a weak kininogenase, since it had only 5% of the kininogenase activity of equimolar concentrations of glandular kallikrein and had no effect on rat mean blood pressure or on the isolated rat uterus. Esterase B activated plasminogen and had caseinolytic activity. It was inhibited by aprotinin, soybean trypsin inhibitor, lima bean trypsin inhibitor, phenylmethanesulfonyl fluoride, antipain, leupeptin, and p-tosyl-L-lysine chloromethyl ketone. On double immunodiffusion, when reacted with kallikrein and tonin antisera, esterase B showed partial identity with kallikrein but not with tonin. On immunoelectrophoresis against kallikrein antisera, esterase B formed a precipitin arc at a position different from that of kallikrein. Esterase B appears to be a trypsin-like serine protease having some homology with glandular kallikrein.     

Immunocytologic analysis, in rat hypothalamus, of neurons producing peptides related to endorphin, ACTH, MSH and beta-LPH. Comparison with other peptidergic and monoaminergic neurons]

In rat hypothalamus intraventricularly injected with colchicine, the same neurons of the ventral region are stained with I.S. against alpha and beta-endorphin, (1-24) and (17-39) ACTH, alpha and beta-MSH, and beta-LPH. They are distinct from those producing LH-RH, somatostatin, neurophysin, and dopamine. These results suggest that the same neurons elaborate peptides identical with or immunologically related to endorphins, ACTH, alpha-MSH and beta-LPH, probably issued from a common precursor.     

In vitro secretion of ACTH, beta-endorphin and beta-lipotropin in Cushing's disease and Nelson's syndrome

Tissue from histologically confirmed ACTH cell adenomas in Cushing's disease (CD) and Nelson's syndrome (NS) was gained by transsphenoidal surgery. Combined enzymatic and mechanic agitation of tumor tissue yielded a cell suspension. Aliquots of the cell suspension were transferred to superfusion chambers immediately after isolation and investigated for ACTH and beta-endorphin production. Feedback action of cortisol (CO) and dexamethasone on basal hormone production and on lysine vasopressin (LVP) induced ACTH secretion were studied. Adenomatous tissue and anterior lobe tissue from the same patient in CD could be investigated simultaneously in 4 cases. The paraadenomatous tissue showed depression of basal and LVP-induced ACTH secretion. In all adenomatous tissues investigated there was missing or reduced suppression of basal ACTH secretion by physiological levels of CO. CO not only failed to suppress LVP-induced ACTH secretion but also seemed to enhance LVP stimulation in some experiments. This study confirms former results, that a missing or inversed feedback action or glucocorticoids in adenoma cells is a mechanism involved in the pathological ACTH secretion in CD and NS. Bioassayable and immunoreactive ACTH from media of superfusion and short-term static incubation were compared with beta-endorphin and beta-LPH in an assay detecting these two peptides with equimolar sensitivity. Secretory patterns were basically parallel but great differences showed in quantities of hormones secreted. In addition, Sephadex G-50 gel chromatography was performed to separate beta-endorphin from beta-LPH and to calculate the ratios. These profiles show great variations between different adenomas.     

Identification of serine proteinases with tonin-like activity in the rat submandibular and prostate glands.

Two enzymes with tonin-like activity, designated rSMT3 and rSMT4, were purified from rat submandibular glands and another, rPT1, was obtained from the prostate. The three enzyme fractions hydrolysed angiotensin I, angiotensinogen (AG) and synthetic AG(1-14) to form angiotensin II. With angiotensin I as substrate, pH optima were 6.5 for rSMT3, 6.8 for rSMT4 and 7.5 for rPT1. With AG(1-14), the three enzymes had optimal activity at pH 7.5. The three enzymes had negligible activity upon a kallikrein substrate, Ac-Phe-Arg-Nan. The enzymes were inhibited by aprotinin, soybean trypsin inhibitor and phenylmethanesulfonyl fluoride but not by two angiotensin converting enzyme inhibitors, ethylenediaminetetracetic acid or enalaprilat. N-tosyl-L-phenylalanine chloromethyl ketone (1 mM) inhibited rPT1 and rSMT4 but not rSMT3. Molecular weights (SDS-PAGE) were 31,700 for rSMT3, 29,800 for rSMT4 and 28,100 for rPT1. Total activity in the prostate is 150-times lower than in the submandibular gland, where 92% of the tonin activity is related to rSMT4. Physical and chemical properties suggest that rSMT4 is tonin, whereas rSMT3 and rPT1 are tonin-like enzymes which can generate angiotensin II from different substrates.     

From beta-lipotropin to beta-endorphin and 'pro-opio-melanocortin'.

Studies on the biosynthesis of beta-LPH on the one hand, and of ACTH on the other, have produced a new concept, that of a single precursor form which contains three active molecules. Thus, it is proper to name such a precursor 'pro-opio-melanocortin.' The concept that beta-LPH was a precursor molecule was first put forward in 1967 and was based on both structural forms and biological activities. The discovery that morphine-like substances are part of the C-terminal fragment of beta-LPH brought an additional important biological side product. That, together with the recent demonstration of ACTH as part of a still larger precursor, constitutes an exciting model for the study of peptide hormone biosynthesis. We have shown unambiguously that beta-endorphin is the result of a maturation process from the large precursor, while beta-LPH is an important and transient intermediary. Since it is also present in the brain, our recent results using pars intermedia cells can be applied to study the fabrication and degradation of these molecules in the brain. We expect to see it established that all other neuropeptides are also biosynthesized as larger precursor molecules whose structure at the site of cleavage could well be constituted of two basic amino acids like in the pro-opio-melanocortin.     

A novel serine protease with vasoconstrictor activity coded by the kallikrein gene S3

A fraction separated from rat submandibular gland homogenates was found to contain a potent vasoconstrictor when tested on isolated rabbit aortic rings. The vasoconstrictor was purified by a series of chromatographic steps. The purified compound (2.77 x 10(-9) M) induced 40% of the maximum contractile response to 60 mM KCl. Constriction was slow in onset, long-lasting, rinse-resistant, and unchanged by de-endothelialization; in addition, it was dose-related and inhibited by both EGTA and verapamil, but it was not affected by DUP753, an angiotensin II receptor antagonist. The compound was found to be a protein having a pI of 7.36 and a molecular weight of approximately 29,000 and exhibiting partial immunologic identity to rat glandular kallikrein and rat tonin. After 2-mercaptoethanol treatment, it separated into heavy (approximately 19,900) and light (approximately 10,700) chains having amino-terminal sequences of AY(X)HNNDLMLL and VVGGYN(X)ETNSQ, respectively. We found that they correspond to the amino-terminal and internal sequence of a previously unidentified kallikrein-like serine protease whose mRNA, named S3, has been found in the rat submandibular gland and prostate. The vasoconstrictor is able to hydrolyze t-butoxycarbonyl-valine-proline-arginine-methylcoumarin amide (a thrombin substrate), although its Kcat/Km was only 0.02% that of rat thrombin. Both vasoconstrictor and enzymatic activity on t-butoxycarbonyl-valine-proline-arginine-methylcoumarin amide were completely suppressed by amidinophenylmethylsulfonyl fluoride and soybean trypsin inhibitor; however, they were unaffected by hirudin, a thrombin inhibitor. At pH 6.5, it released angiotensin II when incubated with sheep angiotensinogen, although it had approximately one-tenth the activity of tonin. The submandibular enzymatic vasoconstrictor is a kallikrein-like enzyme, having some properties of both tonin and thrombin. It directly contracts vascular smooth muscle, acting via a mechanism that requires intact enzymatic activity.     

Quantitation of the endopeptidase activity generating gamma-endorphin from beta-endorphin in rat brain synaptic membranes by a radiometric assay

gamma-Endorphin is a naturally occurring biologically active peptide that is produced by an endopeptidase activity cleaving its precursor beta-endorphin. This enzyme was termed gamma-endorphin generating enzyme (gamma-EGE). In order to quantitate gamma-EGE activity by means of a simple and sensitive assay two synthetic peptides derived from the sequence surrounding the gamma-EGE cleavage site in beta-endorphin were tested as substrates. One of these peptides Ac-Val-Thr-Leu-Phe-Lys-NHCH3 fulfilled all criteria for a suitable gamma-EGE substrate. The peptide was exclusively cleaved at the correct bond for gamma-EGE upon incubation with brain synaptic membranes, and this cleavage was inhibited by the naturally occurring substrate beta-endorphin. The peptide was insensitive to cleavage by exopeptidases and cathepsin D. Addition of a 14C-labeled methyl group at the lysine residue of this peptide by reductive methylation did not alter its properties as a substrate for gamma-EGE activity. The use of the 14C-labeled peptide allowed sensitive quantitation of its radioactive products after simple separation by hydrophobic chromatography on minicolumns containing polystyrene beads. gamma-EGE activity increased linearly with a protein concentration and incubation time. This assay can be used for reliable quantitation of gamma-EGE activity and permits investigations on the regulation of gamma-endorphin production.     

Evidence for serotonergic stimulation of pituitary beta-endorphin release: preferential release from the anterior lobe in vivo

Using gel filtration chromatography (Sephadex G-50) and radio-immunoassay for beta-endorphin (beta-END) and beta-lipotropin (beta-LPH) we investigated the site [anterior lobe (AL) vs. intermediate lobe (IL)] for serotonergic control of pituitary beta-END-like immunoreactivity (beta-END-LI) in the rat. Since the secretion of beta-LPH in vitro clearly distinguishes beta-END-LI release by the AL as compared to the IL, we interpreted changes in plasma levels of immunoreactivity resembling beta-LPH to reflect beta-END-LI release from the AL. Following the administration of L-tryptophan (200 mg/kg, 30 min, ip), a serotonin precursor, nearly all of the rise in total plasma beta-END-LI was due to the form of immunoreactivity resembling beta-LPH in molecular size. Similarly, 5-hydroxytryptophan (30 mg/kg, 30 min, ip), a serotonin precursor, and fluoxetine (10 mg/kg, 15 min, ip), a serotonin reuptake blocker, predominantly increased circulating levels of beta-LPH-sized immunoreactivity with little effect on beta-END-sized immunoreactivity. Quipazine (2.5 and 5.0 mg/kg, 30 min, ip), a serotonin receptor agonist, elevated plasma levels of both forms of beta-END-LI; however, the immunoreactive peak coeluting with beta-LPH was primarily affected, being increased 9.5-fold while that resembling beta-END was increased less than 1-fold. Immobilization stress (30 min) dramatically elevated plasma levels of both forms of immunoreactivity, however, a greater relative rise in beta-LPH than beta-END was observed. Intraventricular administration of 5,7-dihydroxytryptamine (75 micrograms, free base, 10 d), a serotonin neurotoxin, lowered plasma levels of both forms of immunoreactivity about equally in stressed animals. Further, dexamethasone, a synthetic glucocorticoid which selectively inhibits AL corticotroph secretion in vitro, attenuated the beta-LPH response to serotonergic activation in vivo. Together, these findings indicate that serotonergic drugs predominantly influence the release of beta-END-LI resembling beta-LPH and further suggest that serotonin neurons preferentially regulate the release of beta-END-LI from AL corticotrophs in vivo.     

Resistance exercise and plasma beta-endorphin/beta-lipotrophin immunoreactivity

Serum cortisol and plasma beta-endorphin/beta-lipotrophic hormone (LPH) immunoreactivity were measured in five males before and after endurance exercise (treadmill) and burst activity resistance exercise (weight lifting). Mean beta-endorphin/beta-LPH immunoactivity increased significantly following treadmill testing (p < .05) and weight training (p < .06). Post-exercise hormonal values were similar for the two activities. The hormonal changes previously reported with endurance activities also occur with burst activity exercise.     

Three high molecular weight protease inhibitors of rat plasma. Reactions with trypsin

We have compared the reactions of trypsin with human alpha 2-macroglobulin (alpha 2M), and three rat plasma protease inhibitors, alpha 1-macroglobulin (alpha 1M), alpha 1-inhibitor III (alpha 1I3), and alpha 2M. All four of these proteins appear to contain reactive thiol esters. The electrophoretic mobility in agarose gels of human and rat alpha 2M is increased by 1 mol of trypsin, while the mobility of alpha 1M and alpha 1I3 is decreased. Treatment with methylamine causes similar mobility changes, except in the case of rat alpha 2M. Titration of human and rat macroglobulins by repeated small additions of trypsin and by assay of liberated SH groups or enhanced ligand fluorescence revealed a stoichiometry of about 1 mol of trypsin/mol of inhibitor. In contrast, addition of macroglobulin to a fixed amount of trypsin and detection of residual amidase or protease activity revealed a stoichiometry of about 2 mol of trypsin for 1 mol of human alpha 2M, about 1.4 mol for rat alpha 1M, and about 1 mol for rat alpha 2M. One mol of trypsin reacted with 2 or more mol of alpha 1I3 by the criteria of SH groups liberated or protease inhibition. Methylamine-treated rat alpha 2M binds a significant amount of trypsin releasing about 2 mol of SH. Radioactive beta-trypsin was covalently bound to subunits of the purified plasma inhibitors. The Mr of the labeled products with rat and human alpha 2M had molecular weights which suggested trypsin was bound to intact as well as cleaved subunit chains and also to multiple chains via cross-linking. Rat alpha 1M also produced a product which may be an intact subunit alpha chain plus trypsin. Greater than 80% of the trypsin was bound covalently to these inhibitors at low molar ratios.     

Isolation, characterization, and localization of antigen gamma, a serine proteinase of the "kallikrein-family" in the rat submandibular gland

A trypsin-like serine proteinase, antigen gamma, immunologically partially identical to glandular kallikrein when run against anti-rat glandular kallikrein antiserum in immunoelectrophoresis, was purified from the rat submandibular gland. The enzyme was purified by a two-step chromatography procedure, ionexchange chromatography followed by gel filtration. The criteria for purity were one band in SDS-polyacrylamide gel electrophoresis and in immunoelectrophoresis, respectively. Antigen gamma had a molecular mass of 25,000 Da and consisted of two polypeptide chains with molecular masses of 14,000 and 11,000 Da. The preparation contained several isoenzymes with pI ranging from 4.1 to 4.5. The enzyme showed high specific enzyme activity against the substrate D-valyl-L-leucyl-L-arginine-4-nitroanilide (S-2266), some trypsin-like and kininogenase activity, but no angiotensin converting enzyme, kininase, or tonin activity. Amidolytic activity was increased and stabilized by the presence of detergent in the assay buffer. The pH-optimum of antigen gamma amidolytic activity was about 10. Antigen gamma was inhibited by SBTI and PMSF, whereas aprotinin had to be added in a more than 100 times higher concentration than for glandular kallikrein. The binding pattern of antigen gamma to plasma proteins was different from that of tonin and glandular kallikrein. Antiserum against antigen gamma was raised in rabbits and characterized against rat submandibular gland homogenate. Immunohistochemistry showed antigen gamma in the secretory granules of the submandibular gland granular tubular cells but only adhering to the luminal cell wall in the striated and main excretory ducts. Antigen gamma was not detected in the sublingual or parotid gland or in the kidney. Antigen gamma was demonstrated by immunoelectrophoresis in rat submandibular gland saliva. The concentration was higher in sympathetically than in parasympathetically induced secretion.