Identification of Tissue Kallikrein in Brain and in the Cell-free Translation Product Encoded by Brain mRNA*

A monoclonal antibody against purified rat urinary kallikrein was coupled to agarose and used to isolate kallikrein from rat brain. The purified enzyme has N"-tosyl-L-arginine methyl esterase activity with a pH optimum at 9.0, kinin-releasing activity from a purified low molecular weight kininogen, and a parallelism with standard curves of rat urinary kallikrein in a direct radioimmunoassay. Brain kallikrein is inhibited by a series of tissue kallikrein inhibitors with IC60 values similar to those for urinary kallikrein. The purified brain enzyme was labeled with ['4C]diisopro-pylphosphorofluoridate and visualized by fluorogra- phy on a sodium dodecyl sulfate-polyacrylamide gel. Electrophoretic mobility of the enzyme was closely similar to that of urinary kallikrein with estimated M, of -38,000. With Western blot analyses using a rabbit anti-kallikrein antibody, both brain and urinary kallikrein were visualized at identical positions by immunoperoxidase staining and by autoradiography with '251-protein A binding. Brain mRNA was used to direct cell-free protein synthesis in wheat germ and rabbit reticulocyte lysate systems.

roducts of tissue kallikrein have been long known to affect ehavior (12), cardiovascular function (13), brain amine levels (14) and peptide release (15). Kininase activities are present in brain (7,16), and recently, both bradykinin-like immunoreactivity (17) and specific kinin-binding sites (18) have also been found in the central nervous system and the spinal cord, respectively. All of this work infers, but does not prove, the existence of an endogenous tissue kallikrein in the central nervous system. The present work using both monoclonal and polyclonal antibody against rat tissue kallikrein establishes that the brain contains and is capable of synthesizing a kallikrein indistinguishable from purified tissue kallikrein.

Purification of Rat Tissue Kallikreins and Polyclonal Antibodies
Rat urinary or submandibular kallikreins were purified to homogeneity and antisera to purified rat urinary kallikrein were generated in a sheep or rabbits as described previously (19). Rat urinary arginine esterase A, a plasminogen activator, was purified as described (20). Antisera titers were determined by the radioimmunoassay developed by Shimamoto et al. (21). Rat submandibular kallikrein (5 mg/ml in 0.1 M MOPS,' pH 7.5) was coupled to Affi-Gel 15. Fifty ml of sheep anti-kallikrein antiserum or 20 ml of rabbit anti-kallikrein antiserum were diluted 4-fold in 0.1 N NaCl, 0.01 M sodium phosphate, pH 7.0, and passed through the kallikrein-agarose affinity column equilibrated with the same buffer. Adsorbed antibodies were then eluted with 0.1 M acetic acid as described previously (20,21). Purified kallikrein, arginine esterase A, or protein A was labeled with lZ5I using the lactoperoxidase method according to Shimamoto et al. (21).

Kallikrein Assays
Kallikrein-like TosArgOMe esterase activity was determined as described previously (22,23). The kininogenase assay of Shimamoto et al. (24) was used to determine the kinin-generating activity of kallikreins. Generated kinins were measured by a kinin radioimmunoassay (25). Purified bovine low molecular weight kininogen was used as substrate. The direct radioimmunoassay procedure for tissue kallikrein was carried out using a sheep anti-rat urinary kallikrein antiserum as described previously (21,26).

Preparation of Brain Extracts
Male or female Sprague-Dawley rats (200-300 g) were used. From 10 to 30 whole brains were used at a time. After excision, the tissue was blotted free of blood, minced, and homogenized in 0.05 M Tris/ HC1, pH 7.5, using a motor-driven homogenizer, and then centrifuged a t 600 X g for 15 min. Deoxycholate was added to the supernatant to a final concentration of 0.5% (w/v), incubated at room temperature for 30 min, and then centrifuged at 10,000 X g for 20 min. This supernatant was centrifuged at 100,000 X g for 60 min and dialyzed overnight a t 4 "C against a buffer containing 0.1 M NaC1, 0.01 M NaHZP04, pH 6.0. This dialysis extract was used for further purification. Alternatively, minced brain tissue was further fractionated using differential centrifugation as described by Powers and Nasjletti (11). Briefly, preparations in 0.01 M Tris/HCl, pH 7.5, containing 0.25 M sucrose were centrifuged at 10,000 X g for 20 min; the supernatant was saved and the pellet was resuspended in the buffer and centrifuged another 20 min at 10,000 X g. The final 10,000 X g pellet was resuspended in the buffer and saved. The supernatant from the first 10,000 X g centrifugation was centrifuged for 60 min at 105,000 X g in a Beckman L5-65 ultracentrifuge. The 105,000 X g supernatant was saved and the pellet was resuspended in the buffer and recentrifuged at 105,000 X g for 60 min; the supernatants were combined and the pellet was resuspended in Tris/HCI buffer and saved. All of the above tissue fractions were stored at -20 "C for measurement of kallikrein-like TosArgOMe esterase activity. Protein concentration of the various fractions was determined by the method of Lowry et al. (27) using bovine serum albumin as standard.

Monoclonal Antibodies against Rat Tissue Kallikrein
A mutant mouse myeloma line SP2/0-AG14, resistant to 8-azaguanine and a nonproducer of immunoglobulin, was provided by Dr. R. Hyman, Salk Institute, San Diego, CA. Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Female BALB/C mice were immunized twice, with intraperitoneal injections of 50 pg of purified rat urinary kallikrein emulsified with complete Freund's adjuvant, with an interval of 3-4 weeks. Three weeks after the second injection, the mice were boosted intravenously with 100 pg of kallikrein dissolved in phosphate-buffered saline, pH 7.3. Three days later, spleen cells were obtained and mixed with the mouse myeloma cells for fusion in 50% (v/v) polyethylene glycol 4000 in serum-free medium as described previously (28). Cells were dispersed into multiwell dishes in a selective medium supplemented with 20% fetal calf serum, hypoxanthine (1 X

M),
aminopterin (4 X M), and thymidine (1.6 X M). Hybrid cells producing antibodies of interest were subcloned twice by limited dilution (0.4 ce11s/0.2 ml) in microtiter plates. Hybridomas secreting antibodies to kallikrein were identified by both radioimmunoassay (26) and by ELISA (29). Monoclonal antibodies secreted by a cloned hybridoma designated V4G6 show specific binding to rat tissue kallikrein but not other kallikrein-like enzymes such as arginine esterase A (20). Its subtype was determined to be IgG, using antisera against specific mouse immunoglobulin subclasses with ELISA assay (29). Antibody was fractionated from the growth medium with 50% ammonium sulfate precipitation at 4 "C and in a concentration of 10 mg/ml, coupled to Affi-Gel 10 as described previously (21). Mice were pretreated with 0.2 ml of Pristane for 2 weeks, then injected intraperitoneally with 1 X lo6 hybrid cells. One to two weeks after the injection, ascites fluid was collected, centrifuged and stored at -70 "C.
Active Site Labeling Rat brain enzyme (-10 pg) or purified urinary kallikrein (-4 pg) was incubated with [14C]dii~~pr~pylpho~phor~fluoridate (0.2 pCi; specific activity, >lo0 mCi/mmol) in a total volume of 30 pl of HZ0 for 1 h a t 37 "C followed by 16 h a t 4 "C. The mixture was then added to 30 pl of buffer containing 0.13 M Tris/HCl, pH 6.8, 30% glycerol, 10% mercaptoethanol, and 5% SDS and heated to 100 "C for 5 min. Electrophoresis was performed in a 7.5 to 17.5% linear gradient. polyacrylamide gel according to the procedures described previously (20,30). After electrophoresis, the gels were stained for 30 min in methano1:acetic acidwater (45:9:45) containing 2.5% Coomassie blue and destained overnight in methanokacetic acid:water. The gels were treated with ENHANCE and then dried on Whatman No. 3MM filter paper. Fluorography was carried out by placing the gels in direct contact with Kodak X-Omat AR film. Films were developed after 1 week of exposure.
Western Blot Analyses (a) Immunoperoxidase Staining Method-Kallikreins were visualized by immunoperoxidase staining as described by Towbin et al. (31). Electrophoresis was performed in a 7.5 to 17.5% linear gradient polyacrylamide gel containing 0.1% SDS as described above. The proteins were then electrophoretically transferred to nitrocellulose paper as described by Burnette (32). Electrophoretic transfer was performed at 15 V for 18 h at 4 "C. The unstained blots were soaked for 1 h ar 40 "C in a buffer containing 5% bovine serum albumin, 0.15 M NaCl, 0.01 M Tris/HCl, pH 7.4, to saturate unoccupied protein binding sites. The blot was incubated for 2 h at room temperature with affinity-purified rabbit anti-rat urinary kallikrein antibody (5 pg/ml) in a buffer containing 0.15 M NaC1, 0.005 M EDTA, 0.05 M Tris/HCl, pH 7.4, 3% bovine serum albumin, and 0.05% Nonidet P- 40. The blots were washed with 0.01 M Tris/HCl, pH 7.4, 0.15 M NaCI, and 0.05% Nonidet P-40 and incubated for 2 h at room temperature with goat anti-rabbit immunoglobulin-conjugated horseradish peroxidase (1:lOOO) (Cappel). The blots were washed as described above and incubated with a substrate solution containing odianisidine (25 pg/ml), 0.01 M Tris/HCl, pH 7.0, and 0.01% hydrogen peroxide. The orange bands corresponding to protein bands appeared within a few minutes, and the reaction was terminated after 20 min by washing with HZO.
(b) Autoradiography with lZ5Z-Protein A Binding-Alternatively, following incubation of the blot with rabbit anti-rat urinary kallikrein antibody, the blot was washed and incubated with 12'I-protein A (lo6 cpm/ml) for 2 h at room temperature. The solution was then aspirated and the nitrocellulose sheet was rinsed with H20, then soaked with shaking in 0.15 M Tris/HCl, pH 7.4, 0.005 M EDTA, 0.15 M NaC1, 0.25% gelatin, 0.5% Triton X-100, and 0.1% SDS at room temperature for 2 h. The sheet was affixed to a sheet of filter paper and exposed to Kodak X-Omat AR film at -70 "C using Cronex lightning plus a intensifying screen for 6-24 h.

Isolation of Poly(A+) mRNA from Brain
Approximately 80 g of rat whole brain tissue were minced and homogenized at 4 "C with a Polytron instrument in 5 volumes of buffer (25 mM Tris/HCl, pH 7.5, 0.25 M sucrose, 25 mM NaCI, 5 mM MgCI,, 0.1% diethylpyrocarbonate, and 0.1 mM cycloheximide). The homogenate was centrifuged at 1000 X g for 10 min and the supernatant was further centrifuged a t 27,000 X g for 15 min. The pellet containing RNA from the 27,000 X g fraction was resuspended in 60 ml of the same buffer, and sodium deoxycholate and Triton X-100 were added at final concentrations of 1% each. The suspension was clarified by centrifugation at 10,000 X g for 10 min. RNA was extracted from the supernatant by SDS-phenol-chloroform-isoamyl alcohol treatment (33). Polyadenylate-containing mRNA was isolated by oligo(dT)-cellulose column chromatography (34).

Cell-free Translation
Wheat germ extract or rabbit reticulocyte lysate was prepared and used as described previously (35,36). The reaction mixture (total volume, 50 p1) contained 20 mM N-2-hydroxyethylpiperazine-N2'ethane sulfonate, pH 7.6, 1.5 mM magnesium acetate, 64 mM potassium acetate, 2 mM ATP, 0.1 mM GTP, 10 mM creatine phosphate, 10 pg/ml of creatine phosphokinase, 0.7 mM spermidine, 20 p M of 19 unlabeled amino acids, 20-30 pCi of [35S]methionine (specific activity, >400 Ci/mol), 60 pg/ml of RNA with the wheat germ extract or 30 pg/ml of RNA with reticulocyte lysate, and 15 pI of wheat germ extract or 10 pl of reticulocyte lysate. After incubation at 25 "C for 120 min, with occasional shaking, tricholoroacetic acid-insoluble radioactivity was determined. The reaction mixture for protein synthesis was also scaled up 5to 15-fold, and the reaction was terminated by placing the samples on ice. The sample was diluted in 3 volumes of buffer A containing 10 mM Na phosphate, pH 7.6,l mM Na EDTA, 1% Nonidet P-40, 1% sodium deoxycholate, 0.3% SDS, and 2 mM dithiothreitol and centrifuged at 12,000 X g for 15 min in a microfuge, and the supernatant was collected. A portion (5 ~1 ) of this supernatant was used to determine trichloroacetic acid-insoluble radioactivity. The supernatant was then immunoprecipitated with 40 pg of rabbit or sheep anti-kallikrein antibody or control IgG, and 2.2 pg of purified urinary kallikrein was added as carrier protein. The mixture was incubated at 4 "C overnight. The resulting immunoprecipitate was washed 3-4 times with buffer A. Total and immunoprecipitated translation products were analyzed by SDS-polyacrylamide gel electrophoresis and radioactive protein bands were visualized by fluorography as described above.

Materials
The following materials were obtained from commercial sources: molecular weight markers, protein A and protein A-Sepharose (Phar-

RESULTS
Specificity of Monoclonal Antibody-The monoclonal antibody secreted by the hybridoma clone designated V4Gs was harvested from culture media or ascites fluid as described under "Experimental Procedures." The antibody titration curves of Fig. 1 show that the polyclonal antibody at a final dilution of 1:4 x lo3 not only binds kallikrein but also purified arginine esterase A (20), while the monoclonal antibody binds only to kallikrein over a broad range of dilution. This monoclonal antibody was, therefore, used to prepare an antibodyaffinity column for brain kallikrein purification.
Purification of Brain Tissue Kallikrein-Kallikrein-like activity was measured by the TosArgOMe esterase radiochemical method. Rat brain (-40 g wet tissue a t a time) was homogenized and centrifuged as described under "Experimental Procedures." After dialysis and DEAE-cellulose column chromatography as previously described (19)(20)(21), TosArgOMe esterase activity binding to the column was eluted between NaCl concentrations of 0.2-0.3 M. The fractions containing activity were pooled and adjusted to 0.2 M NaCl, 0.01 M NaHZPO4, pH 6.0. The pooled fractions were applied to the monoclonal antibody affinity column (1.5 x 15 cm) equilibrated with the same buffer. As shown in Fig. 2, most of the protein flowed through the column and the adsorbed Tos-ArgOMe esterase activity was eluted with 0.1 M acetate buffer, pH 3.5. Brain Kallikrein Assays and Characteristics-Appropriate  dilutions of the affinity column eluate were used to obtain a range of antigen concentrations comparable to the rat urinary kallikrein standard (0.08-20 ng/tube). Fig. 3 shows typical log-logit transformations of radioimmunoassay standard curves of rat urinary kallikrein and serial dilutions of the rat brain esterase. Linear curves which are always parallel to the standard curves were obtained and indicate the immunological identity of brain tissue enzyme and urinary kallikrein. The rat brain enzyme had TosArgOMe esterase activity of 116 e.u./mg of enzyme as calculated from the direct radioimmunoassay. The enzyme also released kinin from purified bovine low molecular weight kininogen substrate at a rate of 3.3 pg of kinin generated/min/mg of kallikrein. Thus, the rat brain esterase has these enzymatic qualities indistinguishable from those of rat urinary kallikrein.
The pH profile of TosArgOMe esterase activity of the rat brain enzyme followed a Gaussian distribution with optima of 9.0, similar to rat urinary (renal) kallikrein (37). A comparison of the effects of various serine proteinase inhibitors on rat brain and urinary kallikrein activities is shown in

TABLE I
Inhibition of rat brain and urinary kallikrein ICs0 is the inhibitor concentration giving 50% inhibition. Twenty pl of rat brain kallikrein or rat urinary kallikrein (2 X 10" e.u.  Table I. Both enzymes were inhibited to a similar extent by antipain, leupeptin, D-PhePheArgCHnC1, pentamidine, benzamidine and aprotinin. Neither enzyme was affected by LBTI or a,-antitrypsin over a broad concentration range. As with  (30,000), soybean trypsin inhibitor (20,100), and a-lactalbumin (14,400). 1, rat urinary kallikrein B; 2, rat brain kallikrein. Right, Western blot analysis of urinary kallikrein and brain kallikrein. Electrophoresis was performed in a 7.5 to 17.5% linear gradient gel containing 0.1% SDS and proteins were electrophoretically transferred to nitrocellulose paper. The blots were incubated with rabbit anti-rat urinary kallikrein antiserum followed by either goat anti-rabbit IgG-horseradish peroxidase conjugate or 9protein A. Horseradish peroxidase conjugate was visualized with a color reaction and "'I-protein A binding was visualized with autoradiography. A, immunoperoxidase staining; B, "'I-protein A binding. 1, rat urinary kallikrein B; 2, rat brain kallikrein.
rat urinary kallikrein (19), the rat brain enzyme is inhibited by high concentrations, but not by low concentrations of SBTI.
Active Site Labeling-Purified brain and urinary kallikreins were prelabeled with ['Cldiisopropylphosphorofluoridate and then electrophoresed in a polyacrylamide slab gel under reducing conditions.
The fluorogram of brain enzyme shows a single band with electrophoretic mobility similar to but slightly slower than that of urinary kallikrein B (Fig. 4, left). The molecular weight of the brain enzyme is approximately 38,000. The data show that the brain enzyme is a serine proteinase as it covalently binds ['VJdiisopropylphosphorofluoridate.
Western Blot Analyses-To demonstrate specific recognition by tissue kallikrein antibody, brain kallikrein was subject to SDS-polyacrylamide gel electrophoresis and then electrophoretically transferred to nitrocellulose paper. After incubation with rabbit anti-kallikrein antibody, both brain and urinary kallikreins were visualized by immunoperoxidase staining (Fig. 4, right, A) or by autoradiography using ""Iprotein A binding (Fig. 4, right, B). These Western blot studies show that the brain and urinary enzymes are closely related, that brain kallikrein is recognized by antibody to urinary kallikrein, and are consistent with the results obtained from active site labeling.
Cell-free Translation of Rat Brain mRNA-Poly(A+) mRNA was obtained from rat brain after chromatography on oligo(dT)-cellulose and the A260/A280 quotient of the mRNA was 2.10. On incubation of the mRNA in the wheat germ system, 2% of [""Slmethionine radioactivity was incorporated into trichloroacetic acid-precipitable protein. This maximal incorporation occurred at an mRNA concentration of approximately 60 pg/ml. Slab gel electrophoresis of rat brain mRNA translation products derived from both wheat germ extract (Fig. 5A) and rabbit reticulocyte lysate (Fig. 5B) is shown. Total translation products or immunoprecipitates in the absence of added brain mRNA were used as controls (lanes 1 and 2). Brain mRNA-directed total products and the product immunoprecipitated by the affinity-purified kallikrein antibody are shown in lanes 3 and 4. The electrophoretic mobilities of kallikrein synthesized in the cell-free systems and immunoprecipitated by the kallikrein antibody are closely similar to that of authentic tissue kallikrein (lane 5).

DISCUSSION
The results of this study demonstrate that tissue kallikrein is present in rat brain. The rat brain enzyme was isolated from deoxycholate-treated brain extracts followed by DEAEcellulose and monoclonal antibody-affinity column chromatography. The brain enzyme is a serine proteinase, In cell-free translation systems using brain mRNA, labeled kallikrein was identified after specific immunoprecipitation.
Other investigators have explored the presence of components of the kallikrein-kinin system in brain. Hori (7) was the first to suggest the presence of kinin-releasing, and destroying, activities and kinin-like peptides in the rabbit brain. However, acetone or acid treatment was required to elicit what Hori considered only slight kinin-forming activity in brain subcellular fractions. Shikimi et al. (8) described the subcellular distribution of a kinin-forming activity and kininogen in the rat brain. This kinin-forming activity was said to be highest in the cerebral cortex with lower amounts measured in the brainstem and cerebellum. In contrast, kininogen content of the cerebral cortex was described as onethird that of the cerebellum or brain stem. Recently, Shisheva et al. (9) reported the presence of a rat brain kinin-forming activity which was inhibited by aprotinin, but not by SBTI or ovomucoid trypsin inhibitor. Finally, another kininogenase activity in the porcine anterior pituitary gland has been characterized by Powers and Nasjletti (11). This pituitary enzyme differs from tissue kallikrein in its sensitivity to inhibitors and substrate specificity, as well as the kinin product formed. This body of preceding work suggested the existence of tissue kallikreins and kallikrein-like enzymes within the mammalian brain, which is now supported by the present study.
The cellular localization of this brain kallikrein is a question of great interest, in light of some recent studies of tissue kallikreins and kallikrein-like enzymatic activities. Nolly and Lama (4) have found a kininogenase activity similar to tissue (and not plasma) kallikrein in mesenteric vasculature. If confirmed, this finding must be considered in order to determine whether brain kallikrein is resident in some portion of the central vasculature.
The availability of both polyclonal and monoclonal antibodies to tissue kallikreins should allow definitive immunohistochemical identification and radioimmunoassay quantitation of kallikrein concentrations in vasculature, specific brain nuclei, tracts or, perhaps, supporting tissue. Specific localization will then allow further consideration of function in relation to specific proteolysis leading not only to local kinin generation, but perhaps to proteolytic functions that are nonkinin related. These latter kallikrein capabilities (38)(39)(40) are of increasing interest since tissue kallikrein is now known to be quite similar to the nerve growth factor y subunit, the fl nerve growth factor endopeptidase, the epidermal growth factor binding protein, and tonin (41,42). In addition, it is reasonable to suggest that some sort of kallikrein-like enzyme may be involved in pro-opiomelanocortin cleavage, since Lys-Arg and Arg-Ser bond cleavage is involved in product generation from this substrate (43, 44) and a serine proteinase inhibitor (t-butoxycarbonyl-D-Phe-ProArgH) blocks the release of these peptide products (45). It would be of interest to know whether purified tissue kallikreins can release active endocrine peptides from this precursor.
Note Added in Proof- Powers and Nasjletti (46) have described recently a tissue kallikrein-like activity in the rat pituitary pars intermedia.