Jocelyn Spragg

Identification of human glandular kallikrein in the beta cell of the pancreas.

To determine the cellular localization of glandular kallikrein in the human pancreas, immunohistochemical studies were performed with a monospecific antibody against the antigenically identical urinary kallikrein (urokallikrein). The localization of glandular pancreatic kallikrein to the beta cells of the islets was the same as that of insulin in normal human pancreas and in two islet-cell tumors. When beta cells were lacking in islet-cell tumors or in the pancreas of a patient with juvenile-onset diabetes, kallikrein antigen was not detectable. Anti-urokallikrein absorbed with purified urinary or pancreatic kallikrein no longer identified a pancreatic antigen, whereas absorption with insulin had no effect. The beta-cell localization of human pancreatic kallikrein, an endopeptidase that, in concert with carboxypeptidase B, converts bovine proinsulin to a polypeptide with the electrophoretic mobility of insulin, suggests that pancreatic kallikrein may be involved in the physiologic activation of proinsulin.

Antigenic separation of a nonkinin-generating TAMe esterase from human urinary kallikrein and immunohistochemical comparison of their localization in the kidney.

Two urinary enzymes that cleave alpha-N-p-tosyl-L-arginine methyl ester (TAMe) have been separated and utilized to elicit monospecific antisera; only one, urinary kallikrein (urokallikrein), had kinin-generating activity. The nonkinin-generating TAMe esterase and urokallikrein were antigenically unrelated. Immunoperoxidase studies of normal human kidney revealed localization of nonkinin-generating TAMe esterase to epithelial cells of the distal tubule, including the ascending thick limb, the macula densa region, and some areas of convoluted tubule. Immunoreactivity for urokallikrein was confined to reabsorption droplets of proximal tubules and to focal segments of the distal convoluted tubules. Electrophoretic, antigenic, and immunohistochemical studies have established that urokallikrein and a nonkinin-generating TAMe esterase represent two distinct renal distal tubule enzymes.

Preparation of human kininogen—III: Enzymatic digestion and modification

Abstract Purified human plasma kininogen of mol. wt 70,000 has been shown to be resistant to carboxypeptidase B treatment and to be a substrate for both plasma and urinary kallikrein. Two kininogen-containing peaks isolated from fresh plasma by stepwise column Chromatography on Sephadex DEAE-A50 were shown to have identical mobility upon disc gel electrophoresis in 8 M urea or after neuraminidase treatment and to have sedimentation coefficients slightly smaller and slightly larger than that of albumin. These data, together with a difference in partial specific volume, suggest a role for carbohydrate content in accounting for the apparent heterogeneity of kininogens.

Purification of single-chain human low-molecular-weight kininogen and demonstration of its cleavage by human urinary kallikrein

Human low-molecular-weight kininogen (LMWK) was purified to apparent physical and functional homogeneity by a six-step procedure consisting of ion-exchange chromatography, reverse ammonium sulfate gradient solubilization, hydrophobic chromatography on phenyl-Sepharose, gel filtration, and removal of contaminating proteins by their affinity for Affi-Gel blue and zinc. The recovery averaged 15.6% (n = 4). Purified LMWK presented as a single stained band on alkaline polyacrylamide gel electrophoresis which corresponded to the region of function in eluates from a duplicate gel. The apparent homogeneity was also observed in sodium dodecyl sulfate (SDS)-gel electrophoresis, where the protein presented as a single band of Mr = 65,000 without reduction and 68,000 with reduction. A mole of substrate released 0.8 mol of kinin in 5 min when cleaved by human urinary kallikrein (HUK), and 0.9 mol after 30 min. Cleavage of the single-chain LMWK released kinin from within a disulfide loop as indicated by the SDS-gel electrophoresis of reduced and unreduced kinin-free LMWK. The heavy chain exhibited an Mr = 62,000, which is similar to the Mr of the amino-terminal chain of human HMWK and is consistent with their antigenic relatedness. In contrast to the Mr = 64,000 procoagulant chain of human HMWK, the small (less than 10,000) carboxy-terminal chain of LMWK has no procoagulant activity and may serve only to protect the kinin moiety in the intact substrate.

Measurement of human urinary kallikrein and evidence for non-kallikrein urinary tame esterases by direct immunoassay and by affinity chromatography

Abstract Kallikrein was separated from other p- tosyl- L -arginine methyl ester (TAMe) esterases in human urine by direct affinity chromatography of concentrated fresh pooled urine. Quantitative analysis of total TAMe esterase activity in pooled, fractionated urine indicated that less than one-third was due to urokallikrein and that the remainder was attributable to one or more esterases which lack kinin-generating activity and fail to react with a monospecific anti-urokallikrein serum. Using a radial immunodiffusion assay for human urokallikrein, recovery of purified urokallikrein added to urine was 95 per cent and the coefficient of variation in replicate analyses was 8.4 per cent. When this method was compared with a kinin-generating and a [ 3 H]TAMe esterase method for determination of kallikrein activity in urine, all three assays were well correlated in 50 urine samples from normal subjects in varying states of salt and water metabolism. However, analysis of the regression line of esterase activity on antigen concentration indicated that at least half of the urinary TAMe esterase activity was due to non-kallikrein esterases. The demonstration by direct assay and by separation techniques that at least one-half of the alkaline TAMe esterase activity of urine is not urokallikrein indicates that changes in urinary esterase activity cannot be equated solely with alterations in urokallikrein. A combination of direct immunological and kinin-generating assays should permit accurate evaluation of urokallikrein concentration and activity.

Kinin-generating and esterolytic activity of purified human urinary kallikrein (urokallikrein).

Abstract Urinary kallikrein (urokallikrein), as defined by its capacity to generate kinin from heat-inactivated plasma or from purified human kininogen, was isolated from fresh concentrated male human urine and shown to be an antigenically unique urinary p -tosyl-L-arginine methyl ester HCl (TAMe) esterase. The isolation procedure achieved a 400- to 576-fold purification of the kinin-generating activity/mg of protein and yielded a product with albumin as the only significant contaminant at the isoelectric focusing step. The purified urokallikrein, defined by its kinin-generating activity, exhibited an isoelectric point with a range from pH 3.9 to 4.2 with charge heterogeneity, an apparent molecular weight of 25,000–40,000 on Sephadex gel filtration, and an anodal mobility on alkaline disc gels. Urokallikrein eluted from disc gels and identified by kinin-generating activity elicited monospecific antiserum in the rabbit. That purified urokallikrein is a TAMe esterase was evident from the concordance of kinin-generating activity, antigenic reactivity with a donkey antipancreas serum shown to recognize urokallikrein and esterolytic capacity as assessed after isoelectric focusing. There was suppression of the esterolytic activity of purified urokallikrein by increasing doses of TAMe or benzoyl-L-arginine methyl ester HCl (BAMe), and analysis of these data with Dixon plots indicated substrate inhibition.

Interaction of Human Low Molecular Weight Kininogen with Human Mast Cell Tryptase

The capacity of purified tryptase, the major neutral tryptic protease of human lung mast cells, to serve as a kininogenase was examined with purified human low molecular weight kininogen (LMWK) as the substrate. Incubating of 25 mug of tryptase with LMWK for 2 to 30 minutes, with or without heparin, yielded no net time-dependent kinin release as determined on the estrous rat uterus. The 0.4 mug of kinin seen represented less than 10% of that released from excess LMWK by 5 mug of human urinary kallikrein in 5 min. Incubation at pH 5.5 with or without heparin did not significantly alter this result. LMWK did not appear by SDS-PAGE to be cleaved by tryptase either in the presence or absence of heparin. In contrast to its action on HMWK, tryptase did not extensively cleave LMWK, or destroy its reactivity with kallikrein.

The antigenic relationship of a contaminant of human urinary kallikrein to Tamm-Horsfall protein.

In the January 1981 issue of the Journal of Histochemistry and Cytochemistry (7) we reported the different renal immunohistochemical localization of human urinary kallikrcin (urokallikrein) and a contaminant protein that cochnomatographed with urokallikrein during conventional purification procedures (4) but was separable from it by alkaline disc gel electrophoresis. Eluates from the gel slices of the urokallikrein region and the contaminant protein region possessed esterolytic activity for a-N-p-tosyl-L-argininc methyl ester (TAMe), but the contaminant protein region had no kinin-generating activity and demonstrated no reactivity with an anti-pancreas serum previously shown to recognize urokallikrein (6). Protein eluted from each region of replicate gels elicited monospecific antisera in separate rabbits. These antisera were used to demonstrate the antigenic unrelatedness of urokallikrein and the contaminant protein (7). The antiserum to the contaminant localized an antigen in the human kidney in the epithelial cells ofthe ascending thick limb, including those in the region of the macula densa, in focal epithelial areas in the distal convoluted tubules, and occasionally and weakly in intratubular amorphous material. Because this localization was similar to that described for Tamm-Horsfall protein in human and other mammalian kidneys (3,8,9), the antigenic relationship of the urinary kalliknein contaminant to Tamm-Honsfall urinary glycoprotein was examined. Purified human Tamm-Honsfall protein and antiserum directed against it were the gift of Dr. John R. Hoyer (3). Antiserum against uromucoid; which is isolated from urine by techniques different from those of Tamm and Horsfall (1) but is apparently identical to TammHorsfall protein (2), was purchased from Cappel Laboratories, Cochnanville, PA. Antiserum against the urokallikrein contaminant (previously designated nonkinin-generating TAMe esterase) was that descnibed (7), while the concentrated effluent from aprotinin-Sepharose affinity chromatography of urine (5) was used as a source of this antigen. It is apparent by Ouchtcrlony immunodiffusion in 1.5% agarose that there is a high degree of antigenic cross-reactivity between human Tamm-Horsfall protein and the urokalliknein contaminant (Figure 1). Antiserum against Tamm-Honsfall protein (3) and the urokallikrein contaminant (7) each give single major immunoprecipitin bands when diffused against Tamm-Honsfall protein, and these bands react in a line of complete identity with the precipitin band formed with the effluent from aprotinin-Sepharose affinity chromatography of urine. The slight spurring noted in the direction of the well containing antiTamm-Horsfall serum indicates that the antiserum directed against the urokallikrein contaminant recognizes more extensive determinants on this antigen. The diffuse minor precipitin bands adjacent to the wells containing the purified Tamm-Horsfall protein most likely represent reactivity with aggregated Tamm-Horsfall protein, since the purified protein was stored in distilled water (10). Immunoelectrophoresis of the concentrated effluent from aprotinin-Sepharose affinity chromatography of urine gave a single immunoprecipitin line ofsimilar mobility with antiserum directed against either Tamm-Honsfall protein or the urokalliknein contaminant (Figure 2). Uromucoid antiserum also recognized an antigen with the same mobility. The more complex precipitin patterns obtained after immunoelectrophoresis of purified Tamm-Horsfall protein again reflect the effects of aggregation but arc similar with all three antisera. When adjacent sections of human kidney were stained by direct immunofluonescence using the antiserum against Tamm-Horsfall protein and by indirect immunofluonescence using the antiserum against the urokallikrein contaminant, similar staining patterns and distnibuiions were obtained. Both antisera reacted with antigen in the epithelial cells of the ascending thick limb and in portions of the distal convoluted tubules. This staining was appreciably diminished when tissue sections were pretreated with unlabeled antiserum against the urokallikrein contaminant prior to staining with fluorcsccinated antiTamm-Horsfall serum. It appears, therefore, that the urinary kallikncin contaminant separabic from kalhikrcin by replicate alkaline disc gel electrophoresis has elicited an antiserum that recognizes Tamm-Horsfall protein. The functional nature of the esterolytic activity associated with the contaminant at the disc gel step and its relationship to Tamm-Horsfall protein remain to be determined. Separation of Tamm-Horsfall protein and urokalhikrein antigens, and elicitation of non-cross-reacting monospecific antisera against each of them permitted the demonstration of different renal distribution patterns by immunohistochcmical studies (7).

A Radio-Immunoassay for Bradykinin

The current bioassay methods for such vasoactive peptides as bradykinin and angiotensin all present problems of specificity, quantitation, reproducibility, and convenience. Haber et al. (1965) have recently reported a new method for the quantitation of angiotensin II, which depends upon the binding of the polypeptide by specific antiangiotensin antibody. This paper describes a similar method for the assay of bradykinin.

On localization of kallikrein in the human pancreas and salivary glands by direct immunofluorescence.

In a recent article by #{216}rstavik et al (5) describing the localization of kalhikrein in the human pancreas and salivary glands by a direct immunofluorescence technique, several comments in the Discussion were directed toward our earlier study which localized kallikrein immunohistochemically to the beta cells of the human pancreas (2). Although the anti-human kalhikrein used by these authors is characterized only by Ouchterlony analysis, they comment that in an earlier study (6) their anti-rat kalhikrein precipitated kinin-generating (kininogenase) activity and is therefore comparable to our anti-human urinary kallikrein serum. We wish to point out that the characterization of our anti-serum was not based upon its capacity to precipitate kinin-generating activity, since that would not have permitted us to differentiate between removal of the enzyme by precipitation and inhibition of the enzyme by antibodies directed at or near the active site. Instead, we prepared monovalent Fab fragments from an IgG fraction of the anti-human urinary kalhikrein serum and examined their inhibitory capacity. Generation ofkinin from the natural protein substrate, kininogen, was competitively inhibited in a dose-dependent fashion. When the small synthetic substrate TAMe (a-N-p-tosyl-L-arginine methyl ester) was used, Dixon analysis of the data obtained showed competitive inhibition at the lower concentrations of Fab fragments with an average inhibitory constant of 2.5 x 10 M (3). The capacity of nonprecipitating Fab fragments to inhibit competitively the action of kalhikrein on the small synthetic substrate indicates antibody specificity at or near the active site of the enzyme. This finding has been independently confirmed by showing that the anti-kahlikrein hgG fraction from which the Fab fragments were prepared can be used in an active site-specific radioimmunoassay for urinary kalhikrein which correlates well (r = 0.91, p < 0.001) with the kinin-generating activity determined in the same samples (7). In addition to demonstrating the specificity of our anti-serum for urinary kalhikrein we have also shown by several criteria that it is monospecific for this antigen ( 1-3). The antiserum recognizes a single antigen with the electrophoretic mobility of kallikrein function as assessed in eluates of sliced alkaline disk gels; it recognizes with a line of complete identity a single antigen in concentrated crude urine, concentrated parotid saliva, pancreatic extract, highly purified urinary kalhikrein, and highly purified pancreatic kalhikrein; and it recognizes a single antigen with the mobility of kinin-generating function when concentrated crude urine is subjected to immunoehectrophoresis. The urinary kalhikrein antigen used for absorption experiments has also been shown to be highly purified by several analytical criteria (4) and has been used to elicit a monospecific antiserum in a rabbit. Thus, by numerous criteria and using a variety of techniques to evaluate both our antiserum and antigen, our data support the localization of active tissue kalhikrein in the beta cells of the human pancreatic islets. In the recent study ofOrstavik et al. (5), the kallikrein preparation was not completely purified as evidenced by the fact that it elicited a polyspecific antiserum. Under these conditions analysis by diffusion in an Ouchterlony plate does not serve to characterize the reactants, since the functional nature of the antigen-antibody precipitin lines is not defined by this technique and since reactants in very high or very low concentrations would escape detection. Because the purified urinary kalhikrein was contaminated with other immunogenic materials, its use as an immunoabsorbent is limited. The use of absorbed immunofluorescent reagents containing soluble complexes requires careful interpretation when positive staining is obtained, since antigen-antibody reactions are reversible and antibody introduced in a soluble complex is capable oftransferring to tissue antigen. The use of immunogen and polyspecific antiserum which have not been rigorously characterized makes it difficult to draw firm conclusions about either the function or tissue localization of the pancreatic kalhikrein that is functionally and antigenically related to kalhikrein purified from urine.