Purification and Characterization of Salivary Kallikrein from an Insectivore (Scalopus aquaticus): Substrate Specificities, Immunoreactivity, and Kinetic Analyses

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 329, No. 1, May 1, pp. 104–112, 1996 Article No. 0197

Purification and Characterization of Salivary Kallikrein from an Insectivore (Scalopus aquaticus): Substrate Specificities, Immunoreactivity, and Kinetic Analyses G. P. Richards,* C. Zintz,† J. Chao,† and L. Chao†,1 *National Marine Fisheries Service, Southeast Fisheries Science Center, Charleston Laboratory, Charleston, South Carolina 29422-2607; and †Medical University of South Carolina, Department of Biochemistry and Molecular Biology, 171 Ashley Avenue, Charleston, South Carolina 29425

Received October 25, 1995 and in revised form February 12, 1996

We report the successful one-step separation of tissue kallikrein from the salivary glands of an insectivore, the Eastern Atlantic mole (Scalopus aquaticus) by perfusion chromatography. Purified mole salivary kallikrein was characterized as a 30-kDa serine proteinase with a pI of 5.3 and a pH optimum of 9.0. It was readily recognized by human tissue kallikrein antibody in immunoblot analyses. It preferentially hydrolyzes fluorogenic peptidyl substrates with arginyl residues, rather than lysyl residues at the P1 substrate recognition site, indicating that it is like other mammalian kallikreins. Mole kallikrein efficiently releases kinin from low molecular weight human, dog, and bovine kininogen substrates with specific activities similar to that of human tissue kallikrein. Steady state kinetics performed with the synthetic tripeptidyl substrates, Phe-Phe-Arg-, Pro-Phe-Arg-, and Val-Leu-Arg7-amino-4-methylcoumarin, gave Km values for mole kallikrein of 3.3, 46.1, and 2.8 mM, respectively, and specificity constants, kcat/Km , of 3818, 165, and 8714 s01 pM01, respectively. Mole kallikrein, when compared with human and rat tissue kallikreins, more closely resembles human kallikrein based on immunoreactivity and kininogenase activity. Mole kallikrein appears to be a member of a single gene or small multigene family. S. aquaticus is recommended for studying the evolution of mammalian proteins and may offer advantages over rodent models for biomedical research. q 1996 Academic Press, Inc. Key Words: characterization; evolution; insectivore; mole; purification; salivary or tissue kallikrein; serine proteinase; kinetics.

1 To whom correspondence should be addressed. Fax: (803) 7924322.

The Eastern Atlantic mole, Scalopus aquaticus, is a common insectivore along the Eastern portion of the United States. It leads a subterranean existence and adults range in weight from 40 to 140 g with a head and body length of 70 to 110 mm (1). This species is the sole member of the genus. Insectivores are believed to hold a unique position in the evolution of placental mammals and may provide important clues on the structure and function of mammalian enzymes. The diversification of mammals is believed to have arisen from a broad, bush-like radiation from primitive insectivores approximately 100 million years ago (2–4). Cladistic models, using palaeontological databases of tooth morphology, suggest that insectivores could be ancestral to a diverse group of mammals, including dogs and cats; primates; horses, cows, and pigs; deer; rats and mice; whales and porpoises; walruses; bats; and anteaters (5). Such a broad radiation from insectivores appears unlikely based on current morphological and molecular information (6), although tooth morphology shows strong similarities between carnivores and insectivores (7). Enzymes found in insectivores are expected to be more primitive than their counterparts in higher mammals and may provide important links concerning enzyme structure and function among the mammals. One family of enzymes of particular interest are the tissue kallikreins. Tissue kallikreins are trypsin-like serine proteinases which, in mammals, have been shown to be members of multigene families. Humans contain a family of only three known kallikrein genes, but rats display an extensive family consisting of at least 13 known genes and perhaps as many as 25 (8, 9). In higher mammals, different patterns of gene expression are mediated by alternative splicing (10–12). Tissue kallikreins serve as processing enzymes for activating growth factors

104

0003-9861/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

/ 6b17$$9405

03-29-96 17:11:47

arca

AP: Archives

CHARACTERIZATION OF SALIVARY KALLIKREIN FROM AN INSECTIVORE

and hormones in mammals (13–16). Perhaps the most studied function of kallikreins, however, is their cleavage of kininogen to produce kinin. Kinin is a vasoactive peptide important in blood pressure homeostasis. It functions to relax vascular smooth muscle, increase vasodilation, and enhance vascular permeability (17). This study (a) details the perfusion chromatographic purification of a novel kallikrein from the salivary glands of the Eastern Atlantic mole, (b) investigates substrate specificities, immunoreactivity, molecular mass, isoelectric point, kinetics, and optimal pH of this novel kallikrein, (c) compares mole salivary kallikrein with human and rat tissue kallikreins, and (d) evaluates whether mole kallikrein is a member of a single or multigene family. MATERIALS AND METHODS

Enzyme Purification and Activity Measurement Animals were obtained from the wild and sacrificed upon arrival at the laboratory. Salivary glands, consisting of submandibular and sublingual glands from two animals, were minced, sonicated on ice in 3 ml of 20 mM Tris–Cl, pH 8.0, microcentrifuged for 3 min at 10,000g, and the supernatant was subjected to anion exchange perfusion chromatography through a 4.6 1 100 mm HQ/M column (PerSeptive Biosystems, Cambridge, MA) on a BioCad Sprint perfusion chromatograph (PerSeptive Biosystems). The column was equilibrated with 10 column volumes (CV)2 of 20 mM Tris/Bis-Tris propane (Sigma), pH 8.0, injected with 500 ml of the salivary supernatant, washed with 5 CV of equilibration buffer and bound proteins eluted over a 25 CV linear NaCl gradient from 0 to 350 mM NaCl. Three peak fractions, identified at a wavelength of 280 nm, were collected and stored on ice. Column washing was performed with 3 CV of 2 M NaCl. Purifications were repeated with 500-ml injections until 3000 ml of the 3150 ml of supernatant had been loaded. Each perfusion chromatographic separation required only 7.8 min, including equilibration time and column cleaning. Screening for enzymatic activity was performed for each fraction using the synthetic, fluorogenic substrate D-valyl-leucyl-arginyl-7amino-4-methylcoumarin (Val-Leu-Arg-AMC, Enzyme Systems Products, Livermore, CA). Reaction mixtures consisted of 2 ml of 20 mM Tris–Cl, pH 8.0, 1 ml of each fraction, and 50 mM of Val-LeuArg-AMC. Activity was measured by the linear production of product at an excitation wavelength of 380 nm (Ex380nm) and an emission wavelength of 460 nm (Em460nm) for 2 min on a LS-5 spectrofluorometer (Perkin–Elmer). Fractions containing high activity were dialyzed and concentrated. Fractions were placed in SpectraPor membranes (Spectrum Medical Industries, Los Angeles, CA; molecular weight cut off of 6000– 8000) and dialyzed overnight at 47C in 3 liters of 20 mM Tris–Cl, pH 8.0. Dialyzed fractions were concentrated and further desalted using Centriprep 10 Microconcentrators (Amicon). Protein levels were estimated before and after concentration by measuring absorbance at 280 nm. Each concentrate was assayed electrophoreti-

2 Abbreviations used: CV, column volume; AMC, 7-amino-4-methylcoumarin; Ex380nm, excitation wavelength of 380 nm; Em460nm, emission wavelength of 460 nm; DTT, dithiothreitol; TBS, Tris-buffered saline; ECL, enhanced chemiluminescence; HPLC, high performance liquid chromatography; RIA, radioimmunoassay; IEF, isoelectric focusing; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis.

/ 6b17$$9405

03-29-96 17:11:47

arca

105

cally for relative purity on a 10% sodium dodecyl sulfate (SDS) polyacrylamide gel followed by Coomassie blue R-250 staining. The purity of concentrates was further verified by reversed phase HPLC perfusion chromatography through a R2/M (4.6 1 100 mm) column (PerSeptive Biosystems). Mobile phase solution was 0.1% trifluoroacetic acid. Elution of bound proteins was detected at 220 nm over a linear acetonitrile gradient. Fractions were partially neutralized as they were collected using 1 M NH4HCO3 . Total protein was determined by Bradford assays (Bio-Rad, Hercules, CA) using bovine serum albumin as a standard.

Characterization of Purified Mole Salivary Kallikrein Electrophoretic migration under reducing and nonreducing conditions. Molecular mass was determined in side-by-side comparisons of mole salivary kallikrein, human urinary (or tissue) kallikrein, and rat tissue kallikrein on a 7.5–15% SDS–PAGE gradient gel (18) under reducing conditions. Five microliters of each enzyme was combined with sample buffer in a total volume of 15 ml to a final concentration of 50 mM Tris–Cl, pH 6.8, 10% glycerol, 2% SDS, 0.1% bromophenol blue, and 100 mM dithiothreitol (DTT). A separate part of the same gel was loaded with nonreduced samples (no DTT) in order to compare migration. Samples were boiled for 5 min and run at 70 mA (for two gels) for 1.25 h in 2 liters of running buffer (2.5 mM Tris– Cl, 19.2 mM glycine, 0.01% SDS, pH 8.3). A duplicate gel was run for Western transfer and immunoblotting. Immunoblots. Proteins were electrotransferred from the gel to a Hybond nitrocellulose membrane (Amersham, Arlington Heights, IL) according to the Western blot technique (19). Immunoblotting was performed by enhanced chemiluminescence (ECL, Amersham) as follows. The membrane was blocked at room temperature for 1 h in 50 ml of BLOTTO (20) consisting of 5% nonfat dry milk powder in 10 mM sodium phosphate, pH 7.4; 0.14 M NaCl; 1 mg/liter thimerosal, 1 mM p-phenylmethanesulfonylfluoride, 200 mg/liter NaN3 ; and 0.01% [w/v] antifoam. The BLOTTO was then drained and replaced with 25 ml of rabbit anti-human urinary kallikrein antibody diluted 1:1000 in BLOTTO and agitated gently on an orbital shaker at room temperature for 1 h. The membrane was thoroughly washed with Tris-buffered saline (TBS, pH 7.6) containing 0.1% Tween 20. Horseradish peroxidase-conjugated donkey anti-rabbit IgG (Amersham) was diluted 1:1000 in TBS and added to the drained membrane for 1 h at room temperature. The antiserum was drained and the membrane was washed with TBS, drained and overlayed with a luminolbased detection reagent (Amersham) for precisely 1 min followed by brief blotting on 3 MM paper (Whatman). Autoradiographs were performed using Kodak XAR film with exposures for up to 10 min. Dot blot analyses. Dot blots were performed on human, mole, and rat kallikreins after twofold serial dilutions in 20 mM Tris–Cl, pH 8.0, containing 1 mg/ml bovine serum albumin. Dots (2 ml) were spotted onto duplicate Hybond nitrocellulose membranes for kallikrein concentrations ranging from 1.0 mg to 7.3 ng/dot and allowed to dry. One membrane was probed with human urinary kallikrein antibody and the other with rat tissue kallikrein antibody according to the enhanced chemiluminescence protocol mentioned above. Isoelectric focusing. Isoelectric focusing (IEF) of the purified mole salivary kallikrein (5 mg), human urinary kallikrein (6 mg), and rat tissue kallikrein (5 mg) was performed on pH 3.5 to 9.5 polyacrylamide slab gels (Pharmacia) according to the manufacturer’s protocol. Gel pHs were determined using pI markers (Integrated Separation Systems, Natick, MA) and confirmed by measuring pHs of 1cm2 strips cut from the edge of the gel along the pH gradient after soaking the strips in 1 ml of dH20 overnight. The focused proteins were fixed and stained with Coomassie brilliant blue R-250 according to the gel manufacturer’s protocol (Pharmacia). Destaining continued overnight with several changes of destain solution and the gel was air dried for 48 h.

AP: Archives

106

RICHARDS ET AL. TABLE I

Purification Table for Mole Salivary Kallikrein

Purification step

Volume (ml)

Homogenate 3.81 Supernatant 3.15 Anion exchange chromatography Fraction 1 12 Fraction 2 38 Fraction 3 15 Total: 65 Dialysis and Centricon concentration Fraction 1 3 Fraction 2 15 Fraction 3 5 Total: 23

Total protein (mg)

Activity (units 1 1003 per ml)

Total activity (units)

167.0 128.0

112.0 106.0

427.0 334.0

2.6 2.6

2.7 16.7 4.3 23.7

1.3 2.7 1.4

15.6 102.6 21.0 139.2

5.8 6.1 4.9 xV Å 5.9

2.1 14.8 3.7 20.6

3.5 5.6 3.2

10.5 84.0 16.0 110.5

5.0 5.7 4.3 xV Å 5.4

Specific activity (units/mg protein)

Yield (%) 100 78

33

26

Note. One unit is defined as 1 mM of 7-amino-4-methylcoumarin (AMC) released from the substrate Val-Leu-Arg-AMC per min at 237C, pH 9.0.

Kininogenase activity. The ability of the mole salivary kallikrein to cleave kininogen, the natural substrate for tissue kallikreins, was evaluated using human, rat, dog, and bovine low molecular weight kininogens. The kininogens were each incubated for 30 min at 377C with 50 ng of mole salivary kallikrein in reaction mixtures containing 3 mg of the designated kininogen in 100 mM sodium phosphate buffer, pH 8.5, in a total volume of 500 ml. The reaction was stopped by boiling and the released kinin measured by radioimmunoassay (RIA) (21). For comparison, the assay was repeated using human and rat tissue kallikreins. Substrate specificities. The mole salivary kallikrein was measured for amidolytic activity toward a variety of fluorogenic kallikrein and trypsin peptidyl substrates. Reaction mixtures consisted of 2 ml of 20 mM Tris–Cl, pH 8.0, 2 ml (about 3.75 ng) of the mole enzyme, and 2 ml of 10 mM substrate dissolved in dimethylformamide (DMF). Kallikrein substrates were D-Val-Leu-Arg-AMC, D-Phe-PheArg-AMC, D-Pro-Phe-Arg-AMC, and Z-Arg-Arg-AMC while trypsin substrates were Z-Lys-AMC (Enzyme System Products), Boc-ValLeu-Lys-AMC and Suc (Meo) Ala-Phe-Lys-AMC (Bachem Biosciences, King of Prussia, PA). A chymotrypsin substrate, Suc Ala-AlaPro-Phe-AMC (Enzyme Systems Products), was also evaluated. Enzymatic activity was measured by the linear production of 7amino-4-methylcoumarin (AMC) at 237C for 2 min at Ex380nm and Em460nm on an LS-5 spectrofluorometer. The amount of AMC released from the substrates by the mole enzyme was determined from a standard curve. For comparison, assays with these same substrates were performed using human and rat tissue kallikreins. Kinetic analyses. Initial velocities of mole, rat, and human kallikreins were determined using twofold dilutions of the proteinases from 200 ng to 6.25 ng in 20 mM Tris–Cl, pH 9.0, and 20 mM of the synthetic kallikrein substrates D-Phe-Phe-Arg-AMC, D-Pro-Phe-ArgAMC, or D-Val-Leu-Arg-AMC. The total volume was 2.0 ml including 20 ml each of substrate and enzyme. As for previous activity measurements, the linear production of AMC was measured at Ex380nm and Em460nm. Kinetics parameters were determined including the Km , Vmax , kcat , and the specificity constant, kcat /Km , using the Lineweaver–Burke equation. All measurements were made at a constant temperature of 257C on a Perkin–Elmer LS-50B spectrofluorometer. Extinction coefficients. The extinction coefficient (E280nm, 0.1%) of HPLC-purified mole, human, and rat tissue kallikreins were calculated from spectrometric measurements taken at 205 and 280 nm according to van Iersel et al. (22) using the formula:

/ 6b17$$9405

03-29-96 17:11:47

arca

0.1% E280 Å

34.14 1 A280 (A205 1 10) 0 0.02

where A is the absorbance. Optimum pH. The pH optimum for mole salivary kallikrein was determined in three pH-overlapping buffer systems: 20 mM maleate buffer (pH 5.0, 5.5, 6.0, 6.5, and 7.0); 20 mM Tris–Cl (pH 7.0, 7.5, 8.0, 8.5, and 9.0); and 20 mM glycine–NaOH buffer (pH 9.0, 9.5, 10.0, and 10.5). The pH optimum was determined by adding an aliquot of a 1:100 dilution of mole kallikrein and 2 ml of 10 mM Val-Leu-ArgAMC (dissolved in DMF) to 2 ml of the appropriate buffer. Enzymatic activity was measured at 217C for 2 min by the linear production of product at Ex380nm and Em460nm. The optimal pH was determined and graphed after normalization of the data. Zooblots. DNA was extracted from tissues of the mole, human, and rat by grinding 0.3–0.6 g of frozen heart or liver tissue in liquid nitrogen, digesting in lysis buffer (Applied Biosystems) for 10 min, followed by proteinase K digestion for 2 h. The samples were extracted twice with phenol:chloroform:isoamyl alcohol (25:24:1) and twice with chloroform. An equal volume of 2-propanol was layered onto the aqueous phase and the DNA spooled onto a 1-ml pipet. The DNA was rinsed with 80% ethanol, dried, and resuspended in TE buffer (23). Five micrograms of DNA from each animal were digested overnight at 377C with BamHI, DraI, and EcoRI. Digests were run on 0.8% 11 TEA agarose gels (23). The DNAs in the gels were partially depurinated with nitric acid and then neutralized. Capillary transfer was performed onto Immobilon-N membranes (Millipore Corp., Bedford, MA) by the method of Southern (24). The membrane was UV cross-linked, prehybridized with single-stranded herring sperm DNA, and hybridized with a nick-translated, full-length human tissue kallikrein cDNA probe overnight at 607C. The membrane was washed to a final stringency of 21 SSPE (23) at 607C. Autoradiography was performed at 0807C for 6–9 days with one intensifying screen. The same blot was then stripped with 0.4 M NaOH, neutralized, and reprobed with a nick-translated, full-length rat tissue kallikrein cDNA.

RESULTS

Chromatographic Purification of Mole Salivary Kallikrein The purification scheme for the isolation of mole salivary kallikrein, the protein recoveries, and the specific

AP: Archives

CHARACTERIZATION OF SALIVARY KALLIKREIN FROM AN INSECTIVORE

107

krein migrated as a single 30-kDa protein under reducing conditions compared with human urinary kallikrein which produced a broad band from 38–48 kDa and rat tissue kallikrein which produced a single band from 40–42 kDa (Fig. 4, right panel). Under nonreducing conditions, human urinary kallikrein shifted slightly with a broad band from 31 to 43 kDa, rat tissue kallikrein produced three bands from 27 to 30 kDa, which represented a major shift from the reduced form, and mole salivary kallikrein formed a somewhat broader band from 26 to 31 kDa (Fig. 4, left panel). Immunological Properties

FIG. 1. Chromatographs of mole salivary kallikrein after purification by anion exchange perfusion chromatography. Peak fractions (Fr 1–3) were collected for subsequent analyses. The elution gradient shown is a measure of conductivity expressed as millisiemen (mS) and represents a gradient from 0–350 mM NaCl. INJ marks the point of sample injection at approximately 1.5 min.

activities are given in Table I. The synthetic substrate Val-Leu-Arg-AMC was employed to detect enzyme activity throughout the purification. Purification by anion exchange perfusion chromatography resulted in three peak fractions (Fig. 1) which, when combined, gave a specific activity of 5.9 units/mg protein (Table I). The yield of purified enzyme was 33% after chromatography, but was reduced to 26% after the dialysis and concentration steps. SDS–PAGE followed by protein staining of the concentrates reveals protein purity by the appearance of a single broad band for fraction 2 and purification to near homogeneity for fraction 1 (Fig. 2). Protein purity of the second peak fraction was assessed by reversed phase HPLC. The resulting single, symmetrical peak (Fig. 3) indicates the successful purification of the mole kallikrein to apparent homogeneity. Molecular Mass and Migration Under Reducing and Nonreducing Conditions Molecular masses were determined and compared for mole, rat, and human tissue kallikreins. Mole kalli-

/ 6b17$$9405

03-29-96 17:11:47

arca

Immunoreactivity of mole kallikrein was determined using rabbit anti-human tissue kallikrein antiserum and rabbit anti-rat tissue kallikrein antiserum. Western blots probed with human kallikrein antiserum gave strong signals; however, rat kallikrein antiserum did not cross-react with the mole kallikrein. Quantitative comparisons of immunoreactivity toward 7.5 to 1000 ng of human, rat, and mole kallikreins were made by dot blot analyses (Fig. 5). Clearly, the human kallikrein antiserum shows weak cross-reactivity toward the mole kallikrein, but no immunoreactivity toward the rat kallikrein. Conversely, the rat antiserum immunoreacts with human kallikrein, but not with mole kallikrein.

FIG. 2. SDS–polyacrylamide gel electrophoresis of mole kallikrein and human tissue or urinary kallikrein (HUK). Fractions 1–3 from the anion exchange perfusion column were run on a 10% polyacrylamide gel under reducing conditions along with human urinary kallikrein and molecular mass markers. Proteins were stained with Coomassie blue and the relative purity of each fraction determined. Fraction 2 shows purification to apparent homogeneity and fraction 1 is nearly homogeneous.

AP: Archives

108

RICHARDS ET AL.

cleaves on the carboxyl side of arginyl residues, but only minimally at lysyl residues (Table III). Cleavage by the mole enzyme is consistent with the results from human and rat tissue kallikreins. The P3 sites appear particularly important in regard to substrate specificities. Kinetics of Tissue Kallikreins

FIG. 3. Reversed phase HPLC of mole kallikrein. A portion of fraction 2 from the HQ anion exchange column, containing 65 mg of protein, was run on reversed phase HPLC to determine purity and the resulting chromatograph is shown. The protein, eluting at 40% acetonitrile in a linear gradient from 25–55% acetonitrile (axis labeled B), illustrates high sample purity. INJ is the point of sample injection.

Isoelectric Point

Results of steady state kinetic studies of mole, rat, and human kallikreins toward synthetic kallikrein substrates are shown in Table IV. The low Km values for all but one of the assays signify high affinity of the enzymes for the designated substrates. Mole kallikrein has lower affinity and specificity for Pro-Phe-Arg-AMC as indicated by the high Km and low specificity constant, kcat /Km , respectively. Concentrations of Phe-PheArg-AMC ú 10 mM inhibited human urinary kallikrein activity, but not the other kallikreins. A plot of substrate concentration versus velocity of AMC release was linear between 1.25 and 10 mM of Phe-Phe-ArgAMC, but between 10 and 40 mM the slope steadily decreased. Extinction Coefficients Mole, rat, and human kallikreins were further characterized by calculation of their molar extinction coefficients (E280nm, 0.1%) from spectrophotometric data. Coefficients were 1.4 for mole salivary kallikrein and 1.5 for both human and rat tissue kallikreins.

The isoelectric point of the mole kallikrein was determined and compared with those of human and rat tissue kallikreins. The mole kallikrein appeared as a single band at pI 5.3, whereas both human and rat tissue kallikreins had multiple bands between pI 3.5 and 4.3 (Fig. 6), signifying the presence of different posttranslationally modified species. Kinin Generation from Kininogen Mole, human, and rat kallikreins were compared for their ability to release kinin from low molecular weight rat, human, dog, and bovine kininogens as determined by a kinin RIA. Results are presented in Table II. Mole kallikrein demonstrated the highest specificity for dog kininogen and the least for rat kininogen. Likewise, human kallikrein showed the same tendencies. Only the rat kallikrein exhibited high specificity for rat kininogen. Specificities of Enzymes Toward Fluorogenic Substrates Specificities toward several kallikrein and trypsin substrates were compared for mole, human, and rat kallikreins. Results indicate that the mole enzyme

/ 6b17$$9405

03-29-96 17:11:47

arca

FIG. 4. Characterization of mole salivary kallikrein and comparison with human and rat tissue kallikreins. Molecular masses were determined under reducing conditions and migration of the kallikreins was determined under both reducing and nonreducing (no dithiothreitol) conditions on a 7.5–15% gradient SDS–polyacrylamide gel after staining with Coomassie blue.

AP: Archives

109

CHARACTERIZATION OF SALIVARY KALLIKREIN FROM AN INSECTIVORE

FIG. 5. Comparison of immunoreactivity of mole, human, and rat tissue kallikreins. Dot blots were performed with mole salivary kallikrein and rat and human tissue kallikreins using twofold serial dilutions of the proteins in 20 mM Tris–Cl, pH 8.0, containing 1 mg/ml bovine serum albumin. The dilutions (2 ml each) containing from 7.5 to 1000 ng of each enzyme were spotted onto nitrocellulose membranes, dried, and probed using either human or rat tissue kallikrein antisera. Detection was by enhanced chemiluminescence followed by autoradiography (see text).

Optimal pH The pH optimum for mole kallikrein was determined over a pH range from 5.0 to 10.5. The enzyme achieved optimal activity at pH 9.0, but was active over a broad range of pHs (Fig. 7). At pHs from 7.5 to 10.0, ú50% of the optimal activity was still present. This high pH optimum for mole kallikrein is consistent with pH optima for rat and human tissue kallikreins, which are § 9.0 (25). Genomic Southern Blots Southern blots of mole, human, and rat DNAs were probed with full-length human and rat tissue kallikrein cDNA probes. The human kallikrein cDNA probe bound strongly to the human DNA and weakly to the mole and rat DNAs (Fig. 8A). On the other hand, the rat kallikrein cDNA probe hybridized strongly with many rat sequences, but only weakly with human and mole kallikrein genes (Fig. 8B). Weak signal for the mole DNA could be caused by low homology between the probe and target genes. DISCUSSION

We succeeded in purifying and characterizing tissue kallikrein from salivary glands of the Eastern Atlantic

/ 6b17$$9405

03-29-96 17:11:47

arca

FIG. 6. Isoelectric points of mole, human, and rat tissue kallikreins. Isoelectric focusing of purified mole and rat tissue kallikreins (5 mg each) and human urinary kallikrein (6 mg) was performed on a pH 3.5 to 9.5 polyacrylamide slab gel. Lane 1 contains pI markers and the pIs were confirmed by measurement of the pH of gel sections cut from the edge of the gel.

mole. Anion exchange perfusion chromatography provided the key for the rapid purification of mole salivary kallikrein. The simplicity of the purification and amount of protein obtained from the salivary glands of only two animals was remarkable. This one-step chromatographic process may provide the solution for the purification of other previously difficult-to-separate proteins. The three peaks obtained by anion exchange perfusion chromatography have similar characteristics and appear to contain the same protein. Fraction 3 (Fig. 1) contains mole kallikrein plus several contaminants which account for the peak on the trailing shoulder of fraction 2. Results obtained from both SDS–PAGE and nondenaturing gels showed that neither size nor charge appears different between the kallikreins of

TABLE II

Release of Kinin from Several Low Molecular Weight Kininogens by Human, Rat, and Mole Tissue Kallikreins mg kinin released/mg enzyme in 30 min at 377C Substrate

Human

Rat

Mole

Rat kininogen Human kininogen Dog kininogen Bovine kininogen

38.7 324.2 479.1 233.7

197.3 304.7 518.4 175.1

53.0 316.2 617.7 229.0

AP: Archives

110

RICHARDS ET AL. TABLE III

Comparison of Specificities of Human, Rat, and Mole Tissue Kallikreins Toward Several Kallikrein and Trypsin Substrates mM AMC released/mg Enzyme/Min at 237C Enzyme

VLR-AMC

FFR-AMC

PFR-AMC

RR-AMC

K-AMC

VLK-AMC

AFK-AMC

AAPF-AMC

Human urinary kallikrein Rat tissue kallikrein Mole salivary kallikrein

1340 1300 1180

õ130 1600 3010

350 500 2280

õ130 õ250 õ130

õ130 õ250 õ130

õ130 õ250 õ130

õ130 õ250 290

õ130 õ250 õ130

Note. AMC, 7-amino-4-methylcoumarin.

fractions 1 and 2 and the reason for their differential separation chromatographically is not clear. However, each of six chromatographic separations of mole salivary tissues provided highly reproducible profiles. The molecular mass of 30 kDa for mole kallikrein is somewhat lower than for human and rat kallikreins, but is within the range of molecular masses (24 – 45 kDa) reported for tissue kallikreins in general (17). Mole kallikrein, like its human and rat counterparts, appears to contain a single subunit since it migrates as a single band under reducing conditions (Fig. 4). Broad bands or multiple closely spaced bands among the kallikreins usually signify proteins containing different carbohydrate moities and, occasionally, partially degraded proteins. All of the kallikreins tested exhibited either broad or multiple closely spaced bands on a nonreducing gel while human urinary kallikrein produced broad bands under both reducing and nonreducing conditions. These characteristics are common to tissue kallikreins which are subjected to many different posttranslational modifications. The mass range for human tissue kallikrein extended somewhat higher than expected, a factor attributed

to a high carbohydrate content. Multiple bands are present also on IEF gels, suggesting different patterns of glycosylation which can influence mass and charge. The pI of mole kallikrein is 5.3 which is higher than the reported range of pIs (3.5 – 4.4) for mammalian kallikreins (26, 27). A comparison of migration under reducing and nonreducing conditions shows some differences for all three enzymes (Fig. 4). Generally faster mobility of the nonreduced samples is attributed to the more compact nature of nonreduced proteins. The reduced proteins, which exhibit less order due to breakage of interchain disulfide bonds and general unravelling of tertiary structure, migrate more slowly. Rat tissue kallikrein appears to undergo substantial unfolding under reducing conditions as evidenced by the difference in migration between the reduced and nonreduced forms (Fig. 4). Mole and human kallikreins do not differ substantially in their migration in the reduced and nonreduced states. This implies that mole and human kallikreins may be more similar than rat kallikreins in their tertiary structure. The immunological evidence suggests that human

TABLE IV

Steady State Kinetic Parameters for Mole, Human, and Rat Tissue Kallikreins Using Three Fluorogenic Tripeptidyl Substrates Substrate

Km (mM)

Vmax (mM/s)

kcat (s01)

kcat/Km (s01 pM01)

Mole

FFR-AMC PFR-AMC VLR-AMC

3.3 46.1 2.8

8.3 25.0 20.0

12.6 7.6 24.4

3818 165 8714

Human

FFR-AMC PFR-AMC VLR-AMC

1.2 3.7 1.8

9.1 5.6 19.7

7.6 2.4 8.6

6333 649 4778

Rat

FFR-AMC PFR-AMC VLR-AMC

4.5 4.8 2.9

8.1 10.1 9.9

13.5 8.4 33.0

3000 1750 11380

Tissue kallikrein

Note. The mean molecular masses of mole, human, and rat kallikreins, as determined from SDS–PAGE gels, were 30.5, 44.0, and 41.5 kDa, respectively, and these values were used for calculating kcat .

/ 6b17$$9405

03-29-96 17:11:47

arca

AP: Archives

CHARACTERIZATION OF SALIVARY KALLIKREIN FROM AN INSECTIVORE

and mole kallikreins are more closely related because they share epitopes not found on the rat enzyme. Conversely, the rat kallikrein antiserum demonstrates moderate immunoreactivity toward human tissue kallikrein, but none toward mole kallikrein, implying a stronger evolutionary linkage between the rat and human enzymes. This paradox cannot be explained at this time. Substrate specificity and kinetic analyses demonstrate similarities in amidolytic activity of the enzymes toward Phe-Phe-Arg-AMC; however, mole salivary kallikrein shows lower specificity for Pro-Phe-Arg-AMC than the other enzymes. The strong specificity of mole kallikrein toward Phe-Phe-Arg-AMC is eliminated by substitution of the strongly hydrophobic phenylalanine with a proline in the P3 position (Table IV). Therefore, the P3 position appears crucial in determining enzyme specificity for mole kallikrein. The specificity constant for human urinary kallikrein is also low for Pro-PheArg-AMC, suggesting some similarities between mole and human kallikreins and the requirement for a hydrophobic residue in the P3 position. Mole, human, and rat kallikreins hydrolyzed Val-Leu-Arg-AMC well, perhaps due to the hydrophobicity imparted by the side chains of valyl- and leucyl-residues. Genomic blots of the mole, human, and rat DNAs indicated that mole salivary kallikrein is a member of a small multigene family, as in humans, or may represent a single gene (Fig. 8). It is recognized that rats contain a large family of at least 13 kallikrein genes and that mice have around 25, but humans carry only three known kallikrein genes (8, 9, 28). This signifies

FIG. 7. Determination of pH optimum. The relative activity of mole salivary kallikrein was determined using three pH-overlapping buffer systems from pH 5.0 to 10.5 and the synthetic kallikrein substrate D-Val-Leu-Arg-7-amino-4-methylcoumarin. Bars represent the standard deviation from three assays.

/ 6b17$$9405

03-29-96 17:11:47

arca

111

FIG. 8. Genomic Southern blot of human, mole, and rat DNAs. Extracted DNAs were cut with restriction enzymes, run on an 0.8% agarose gel, blotted to an Immobilon membrane, and probed with nick-translated human (A) or rat (B) cDNA probes as discussed under Materials and Methods. Exposure of the blot probed with rat kallikrein cDNA (B) was intentionally overexposed to enhance the weak banding present on the mole and human DNAs. Abbreviations: H, human urinary kallikrein; M, mole salivary kallikrein; R, rat tissue kallikrein.

a rapid duplication of the gene family in rodents and places the gene families of humans and moles phylogenetically closer. It does appear that rats may be an outgroup to other common mammalian orders as suggested by Li et al. (29) and Bulmer et al. (30). By comparing mole, human, and rat tissue kallikreins, it is possible to speculate on the relationship between these species. Unlike rat tissue kallikrein, mole and human kallikreins share closer immunological identity (Fig. 5), show reduced ability to release kinin from rat kininogen (Table II), and are apparently small gene families (Fig. 8). These data support previous cladistic models which suggest that primitive insectivores may have been predecessors of modern-day primates (31). The high specificity of mole kallikrein for dog kininogen (Table II) and the strong resemblance of tooth morphologies between insectivores and carnivores (7) provides further evidence that carnivores radiated from insectivores. Insectivores may hold the key for unlocking some of the mysteries surrounding mammalian evolution. They may also provide crucial information on the evolution of serine proteinases, other proteins, and gene families. Insectivores may offer advantages over currently existing rodent models for biomedical research because of their apparently closer link to humans. Although some insectivores, such as S. aquaticus, are not commercially available at this time, their prevalence in the wild could permit opportunities for continued research. ACKNOWLEDGMENTS The authors thank Drs. Jon Ahlquist and JoAnne Simson for their helpful suggestions and critical review of the manuscript. This work

AP: Archives

112

RICHARDS ET AL.

was supported by National Institutes of Health Grants DE 09731 and HL 29397.

REFERENCES 1. Nowak, R. M. (1991) in Walker’s Mammals of the World, Vol. 1, pp. 114–179, The John Hopkins Univ. Press, Baltimore, MD. 2. Gregory, W. K. (1910) Bull. Am. Mus. Nat. Hist. 27, 1–524. 3. Simpson, G. G. (1945) Bull. Am. Mus. Nat. Hist. 85, 1–350. 4. Romer, A. S. (1966) in Vertebrate Paleontology, pp. 207–214, Univ. Chicago Press, Chicago, IL. 5. Benton, M. J. (1991) in The Rise of the Mammals, pp. 40–61, Crescent Books, New York. 6. Novacek, M. J. (1992) Nature 356, 121–125. 7. Novacek, M. J. (1986) Bull. Am. Mus. Nat. Hist. 183, 1–111. 8. MacDonald, R. J., Margolius, H. S., and Erdo¨s, E. G. (1988) Biochem. J. 253, 313–321. 9. Berg, T., Bradshaw, R. A., Carretero, O. A., Chao, J., Chao, L., Clements, J. A., Fahnestock, M., Fritz, H., Gauthier, F., MacDonald, R. J., Margolius, H. S., Morris, B. J., Richards, R. I., and Scicli, A. G. (1992) in Recent Progress on Kinins, pp. 19–25, Birkha¨user Verlag, Basel. 10. Ashley, P. L., and MacDonald, R. J. (1985) Biochemistry 24, 4520–4527. 11. Richards, R. I., Coghlan, J. P., Digby, M., Drinkwater, C. C., Lloyd, C., Lyons, I., and Xiao-Yun, Z. (1989) in The KallikreinKinin System in Health and Disease (Fritz, H., Schmidt, I., and Dietze, G., Eds.), pp. 215–225. Limbach-Verlag, Braunschweig, Germany. 12. Wines, D. R., Brady, J. M., Pritchett, D. B., Roberts, J. L., and MacDonald, R. J. (1989) J. Biol. Chem. 264, 7653–7662. 13. Sealey, J. E., Atlas, S. A., Laragh, J. H., Oza, N. B., and Ryan, J. W. (1978) Nature 275, 144–145. 14. Chao, J. (1983) J. Biol. Chem. 258, 4434–4439.

/ 6b17$$9405

03-29-96 17:11:47

arca

15. Mason, A. J., Evans, B. A., Cox, D. R., Shine, J., and Richards, R. I. (1983) Nature 303, 300–307. 16. Clements, J. A. (1989) Endocrinol. Rev. 10, 393–419. 17. Bhoola, K. D., Figueroa, C. D., and Worthy, K. (1992) Pharmacol. Rev. 44, 1–80. 18. Chao, J., and Margolius, H. S. (1979) Biochem. Pharmacol. 28, 2071–2079. 19. Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl. Acad. Sci. USA 76, 4350–4354. 20. Johnson, D. A., Gautsch, J. W., Sportman, J. R., and Elder, J. H. (1984). Gene Anal. Tech. 1, 3–8. 21. Shimamoto, K., Chao, J., and Margolius, H. S. (1980) J. Clin. Endocrinol. Metab. 51, 840–848. 22. van Iersel, J., Jzn, J. F., and Duine, J. A. (1985) Anal. Biochem. 151, 196–204. 23. Maniatis, T., Fritsch, E., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 24. Southern, E. M. (1975) J. Mol. Biol. 98, 503–517. 25. Richards, G. P., Chang, P., Chao, J., and Chao, L. (1995) Comp. Biochem. Physiol. 111C, 69–82. 26. Pisano, J. J. (1975) in Proteases and Biological Control (Reich, E., Rifkin, D. B., and Shaw, E., Eds.), pp. 199–222, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. 27. Fiedler, F. (1979) in Bradykinin, Kallidin and Kallikrein (Erdo¨s, E. G., Ed.), pp. 103–161, Springer-Verlag, New York. 28. Murray, S. R., Chao, J., Lin, F. K., and Chao, L. (1990) J. Cardiovasc. Pharm. 15, (Suppl. 6) S7–S16. 29. Li, W-H., Gouy, M., Sharp, P. M., O’hUigin, C., and Yang, Y-W. (1990) Proc. Natl. Acad. Sci. USA 87, 6703–6707. 30. Bulmer, M., Wolfe, K. H., and Sharp, P. M. (1991) Proc. Natl. Acad. Sci. USA 88, 5974–5978. 31. Carroll, R. L. (1988) in Vertebrate Paleontology and Evolution, pp. 443–501, Freeman, New York.

AP: Archives