inositol 4,5-bisphosphate (PIP2) at a rate of 15.3 pmol/ min/mg of protein and also phosphatidylinositol 4- monophosphate and phosphatidylinositol (PI) at ...
THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1988 hy The American Society for Biochemistry andMolecular Biology, Inc.
Vol. 263,No. 14, Issue of May 15, pp. 6592-6598.1988 Printed in U.S. A.
Isolation and Characterization of Two Different Forms of Inositol Phospholipid-specific Phospholipase C from Rat Brain* (Received for publication, August 28, 1987)
Yoshimi Homma, Junko Imaki,Osamu Nakanishi, and Tadaomi Takenawa From the Department of Pharmacolotv. Tokvo ”. - Metrooolitan Institute of Gerontology, 35-2 Sakaecho, Ztabashi-ku, Tokyo 173, Japan
Two different forms of inositol phospholipid-specific (PI) yields inositol phosphate and 1,P-diacylglycerol (4). The phospholipase C (PLC) have been purified 2810- and physiological role of inositol phosphate is unknown, but1,24010-fold, respectively, from a crude extract of rat diacylglycerol is thought tobe a second messenger that actibrain. The purification procedures consisted of chro- vatesproteinkinase C (5). On the other hand, the PLCmatography of both enzymeson Affi-Gel blue andcel- mediated breakdownof phosphatidylinositol 4,5-bisphosphate lulose phosphate, followed by three sequential high (PIP,) yields 1,2-diacylglycerol and inositol phosphates inperformance liquid chromatography steps, which were cludinginositol 1,4,5-trisphosphate (6-8). This system, in different for the twoenzymes. The resultant prepara- addition to the generationof a second messenger in the form tions each contained homogeneous enzyme with a M , of 1,2-diacylglycerol, produces other second messengers, inoof 85,000 as determined by sodium dodecyl sulfatesitol polyphosphates. Inositol polyphosphates are reported to polyacrylamide gel electrophoresis. One of these enbe involved in mobilizing Caz+ from both intracellular and zymes (PLC-11) was found to hydrolyze phosphatidylinositol 4,5-bisphosphate(PIP2)at a rate of 15.3 pmol/ extracellular storage sites with consequent elevation of the Recently, Majerus and min/mg of proteinandalso phosphatidylinositol 4- intracellular Ca2+ concentration (1,9). monophosphate and phosphatidylinositol (PI)at slower co-workers (10, 11) obtained evidence for distinct physiologrates. For hydrolysisof PI, this enzyme was activated ical roles of these two pathways of inositolphospholipid by an acidic pH anda high concentration of Ca2+ and hydrolysis. Namely, P I hydrolysis by PLC mainlycauses showed a Vmaxvalue of 19.2 pmol/min/mg of protein. protein kinase C activation, whereas PIP, hydrolysis by PLC The other enzyme (PLC-111) catalyzed hydrolysis of predominantly results in activation of Ca2+-dependentsystwo pathways PIP, preferentially at a V,,, rate of 12.9 pmol/min/mg tems. However, it is still unknown whether the of protein and catalyzed that of phosphatidylinositol are regulated by one enzyme or by several specific ones. 4-monophosphate slightly. The rateof PIP2hydrolysis Although a number of PLC enzymes have been identified by this enzymeexceeded that of PI under all conditions from mammalian tissues (12-28), severallines of evidence tested. Neither of these enzymes had any activity on have suggested a heterogeneity of PLC enzymes in one tissue phosphatidylcholine, phosphatidylethanolamine,phos- (13,15, 24, 25). These findings may support the idea that phatidylserine, or phosphatidic acid. These two eneach PLC enzyme within a cell is stimulated in a different zymes showed not only biochemical but also structural manner and mediates a different physiological pathway for differences. Western blotting showed that antibodies cellular activation. Therefore, for understanding the mechadirected against PLC-I1 did not react with PLC-111. nism of inositol phospholipid hydrolysis induced by external Furthermore, the twoenzymes gave different peptide stimuli,itisimportantto identify andcharacterizePLC maps after digestion with a-chymotrypsinor Staphy- enzymes which hydrolyze inositol phospholipids. lococcus aureus V8 protease. These results suggest that In this study, we resolved three peaks with inositol phosthese two formsof PLC belong to different familiesof pholipid-specific PLCactivity from ratbrainextractand PLC. isolated two of them to homogeneity. The biochemical properties and structural relationships of these two enzymes were examined. The rate-limiting regulation of phosphoinositide metabolism in stimulus-responsecoupling is thought to be hydrolysis of inositol phospholipids by a specific phospholipase C (PLC)’ (1-3). The PLC-mediated hydrolysis of phosphatidylinositol
EXPERIMENTALPROCEDURES
Materials
Ph0sphatidy1[2-~H]inositol(16.3 Ci/mmol) and phosphatidyl[2* This work was supported in part by grants from the Ministry of 3H]inositol 4-monophosphate (1 Ci/mmol) were purchased from Education, Science, and Culture of Japan. The costs of publication Amersham Corp. Ph0sphatidy1[1,2-~H]inositol4,5-bisphosphate (2 of this article were defrayed in part by the payment of page charges. Ci/mmol) was from Du Pont-New England Nuclear. PhosphatidylThis article must therefore be hereby marked “advertisement” in ethanolamine (PE; soybean) and PI (soybean) were from Serdary Research Laboratories (London, Ontario, Canada). Diisopropyl fluoaccordance with 18 U.S.C. Section 1734 solelyto indicate this fact. The abbreviations used are: PLC, phospholipase C; PI, phospha- rophosphate (DFP), phenylmethylsulfonyl fluoride (PMSF), a-chytidylinositol; PIP, phosphatidylinositol 4-monophosphate; PIP2, motrypsin, Stuphylococcusaureus V8 protease, leupeptin, 4-chloro-lnaphthol, and anti-rabbit IgG antibody (peroxidase-conjugated)were phosphatidylinositol 4,5-bisphosphate; PE, phosphatidylethanolamine; PMSF, phenylmethylsulfonyl fluoride; DFP, diisopropyl fluo- from Sigma. Complete and incomplete types of Ribi adjuvant were rophosphate; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic obtained from Ribi ImmunoChem Research, Inc. (Hamilton, MT). A acid; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonicacid; protein assay kit, a silver stain kit, and Affi-Gel blue were from BioMes, 2-(N-morpholino)ethanesulfonicacid; SDS-PAGE, sodium do- Rad.Cellulose phosphate (P11) was from Whatman. Prepacked decyl sulfate-polyacrylamide gel electrophoresis; BSA, bovine serum HPLC columns of TSKgel SP5PW, TSKgel DEAE5PW, and TSKgel G3000SW werefrom Toyo Soda (Tokyo, Japan). A prepacked column albumin; HPLC, high performance liquid chromatography.
6592
Purification of Phospholipases Brain CRat from
6593
Step 3: Cellulose Phosphate-The pooled fraction wasdialyzed against buffer consisting of 10 mM Mes (pH 6.5), 1 mM EDTA, 0.1 mM PMSF, 0.1 mM DFP, and 10% (v/v) glycerol. Then, it was centrifuged (20,000 X g, 30 min), and the supernatant was applied to a column (1.2 X 13 cm) of cellulose phosphate equilibrated with the PLC Assay above buffer. The column was washed with 100 ml of the same buffer, PLC activity was assayed by incubation for 10 min at 37 “C of a and material was eluted with 300 ml of a linear gradient of 0-1.5 M reaction mixture containing 50 mM Mes buffer (pH 6.8), 100 pM NaCl in the same buffer at a flow rate 60 ml/h. EDTA, 100 p~ Ca”, 200 p~ KC1,l mg/ml BSA, 100 or 500 p M PIP,, PLC activity was eluted as two peaks with 0.55 and 0.7 M NaCl, 20,000 dpm of [3H]PIP2 (specific activity, 40 or 200 dpm/nmol), 50 respectively (Fig. 1). Thefractions eluted with 0.5-0.65 and 0.65-0.8 p~ PE, and a sample aliquot in a total volume of 50 pl. The reaction M NaCl were pooled as PLC-I1 and PLC-111, respectively. was terminated by the addition of 2 ml of chloroform/methanol (2:1, Step 4 for PLC-ZZ: TSKgelSP5PW HPLC-Pooled fractions of v/v), and inositol phosphates were extracted with 0.5 ml of 1 N HC1. PLC-I1 from cellulose phosphate were diluted with 2 volumes of the A 0.7-ml aliquot of the upper aqueous phase was removed for meas- same buffer used for chromatography on cellulose phosphate and urement of radioactivity. Before the assay, the reaction buffer was treated with 0.5% (w/v) cholate (pH 6.5) on ice. After centrifugation subjected to ultrasonication to form unilamellar vesicles in which (100,000X g, 30 min) to eliminate aggregated protein, the supernatant substrates were embedded, as reported previously (17). Free Ca2+ was subjected to cation-exchange SP5PW chromatography at a flow concentrations were stabilized using CaEGTA buffers (30) and were rate of 2 ml/min. The column was washed with 20mlof the same confirmed by measurement of the fluorescence of Quin 2 (31). buffer supplemented with 0.5% (w/v) cholate and 0.1 M NaCl and then developed with 80 ml of a linear gradient of 0.1-0.5 M NaCl in Peptide Mapping the same buffer. Only one peak of activity was eluted at 0.20-0.26 M Limited digestion of samples with a-chymotrypsin and S. aurew NaC1, and fractions in this peak were pooled. Step 5 for PLC-II: TSKgelG3OOOSW HPLC-The pooled fractions V8 protease and peptide mapping were carried out by the method of Cleveland et al. (32). Briefly, purified samples (3 pg of PLC-I1 and 1 were concentrated to 10 mlby reverse dialysis against 30% (w/v) pgof PLC-111) were concentrated onSDS-PAGE. The gel wasstained, polyethylene glycol 6000 and dialyzed against the same buffer conand appropriate bands were excised and overlaid with protease on taining 0.1 M NaCl. Then, an aliquot of 2 mlwas applied to a the stacking gel of a second SDS gel. Digestion proceeded directly in TSKgelG3000SW gel filtration column (2.6 X 40 cm). The flow rate the stacking gel during the subsequent electrophoresis, and the re- was 2 ml/min, and only one peak of activity was eluted at 57-62 min (Fig. 2 A ) . sultant bands were located by silver staining (Bio-Rad). Step 6 for PLC-ZI: Mono S HPLC-The fractions from Step 5were Antiserum pooled and subjected to cation-exchange Mono S column chromatogRabbit antiserum against PLC-I1 or PLC-111 was independently raphy in the same buffer. The column was developed with 20 ml of a linear gradient of 0.1-0.3 M NaCl at a flow rate of 1 ml/min. Activity prepared as follows. An aliquot of homogeneously purified PLC-I1 or was eluted as a single peak at 0.22 M NaCl (Fig. 3A), and fractions PLC-111 (10 pgof protein) was mixed with Ribi adjuvant, and the in this peak were pooled and stored in portions at -80 “C. resultant mixture was injected intradermally into multiple sites. The Step 4 for PLC-ZIZ: Hydroxylapatite (HCA) HPLC-Pooled fracrabbit was boosted with 5 or 10 pg of enzyme incomplete type Ribi tions of PLC-I11 from cellulose phosphate were dialyzed against the adjuvant at 4 and 7 weeks after the primary immunization. An aliquot of serum taken 9 weeks after the primary immunization was used for column buffer consisting of 10 mM sodium phosphate (pH 7.2), 0.1 mM PMSF, 0.1 mM DFP, and 10% (v/v) glycerol. The preparation these studies. The antiserum was found to be effective by Western was then centrifugated at 100,000 X g for30 min to eliminate blotting. aggregated protein, and the supernatant was applied to a hydroxylapatite column equilibrated with the same buffer. The column was Purification of PLC-ZZ and PLC-ZZZ washed with 10 ml of 50 mM sodium phosphate and developed with PLC-I1 and PLC-I11 were purified separately to homogeneity as 50 ml of a linear gradient of 50-500 mM sodium phosphate at a flow described below. All procedures were carried out at 4 ‘C. rate of 1 ml/min. Activity was eluted in a peak with 150-200 mM Step 1: Extraction-Rat brains (300 g) were minced in 1.5 liters of sodium phosphate, and fractions in this peak were pooled. homogenization buffer containing 25 mM Tris (pH7.3), 1mM EGTA, Step 5 for PLC-ZIZ: TSKgelG3000SW-Pooled fractions were con1mM EDTA, 0.25 M sucrose, 0.1 mM PMSF, 0.1 mM DFP, and 1pg/ centrated to 6 ml by reverse dialysis against 30% (w/v) polyethylene ml leupeptin and homogenized with a Teflon homogenizer. The glycol 6000 and dialyzed against buffer consisting of 20 mM Tris (pH homogenate was centrifuged at 40,000 X g for 1h, and the supernatant 7.2), 1 mM EDTA, 0.1 mM PMSF, 0.1 mM DFP,and 10% (v/v) was adjusted to pH 7.3 with 1 N NaOH. glycerol. The aggregated proteins were sedimented by centrifugation Step 2: Affi-Gel Blue Chromatography-The supernatant was ap- (100,000 X g, 30 min), and the resulting supernatant (2 ml) was plied to a column (200 ml, 4 X 16 cm) of Affi-Gel blue previously applied to a TSKgelG3000SW gel filtration column (2.6 X 40 cm). equilibrated with buffer consisting of20 mM Tris (pH 7.3), 1 mM The column was developed at a flow rate of 2 ml/min, and activity EGTA, 1mM EDTA, 0.1 mM PMSF, 0.1 mM DFP, 10% (v/v) glycerol, and 0.1 M NaCl. The column was washed with 1.5 liters of this buffer and then developed with a stepwise gradient of 0.3, 0.6, 1.0, and 2.0 M NaCl (500 ml of each) in the same buffer at a flow rate of 150 ml/ h. PLC-I1 and PLC-111 were coeluted with 0.6 M NaCl (Table I) and 1.5 collected for the next step. The other PLC activity was also detected in the washing fraction, but this type of PLC (PLC-I) has not yet been purified to homogeneity. of hydroxylapatite (HCA) was from Mitsui Toatsu (Chiba, Japan). PIP and PIP2 were prepared by the method of Schacht (29) from the Folch’s fraction extracted from bovine spinal cord.
-
TABLE I Stepwise elution of PLC activity from an Affi-Gel blue column PLC activity in each fraction was determined using 500 p~ PIP, as a substrate as described under “Experimental Procedures.” Fraction
Run-through 54.6 Washing 0.3 M NaCl 0.6 M NaCl 1.0 M NaCl 15.8 2.0 M 1.0NaCl
Total protein
PLC activity
w
pmol/rnin
4620 726 62 1 535 310 257
1.5 24.2 90.0
1 *
l’O
0.5
0 0
10
20
30
40
50
Fraction number
FIG. 1. Cellulose phosphatecolumn chromatography. The enzyme fraction (0.6 M NaC1) obtained by Affi-Gel blue chromatography was applied to a cellulose phosphate column (1.2 X 13 cm), and fractions were assayed with 100 PM PIP, in the presence of 100 p~ EDTA and 100 p~ Ca” at pH 6.8 (0).Details were as described under “Experimental Procedures.”
Purification of Phospholipases C from Brain Rat
6594
brain, and treatmentwith detergent caused marked reduction in the activity. Therefore, the soluble fraction was employed as starting material for enzyme purification. When all chromatography fractions were assayedfor PIPz hydrolysis, at 0.02 least three peaksof PLC were separated during thepurification procedure. In the first step of chromatography on an 0.01 Affi-Gel blue column, two major peaks were obtained in the washing and 0.6 M NaCl fractions, respectively. The material in the washing fraction, designated as PLC-I, hasbeen puri0 fied 470-fold, but not tohomogeneity, and has been shown to I r 1 5 be a type of inositol phospholipid-specific PLC with a M , of 135,000. The material in the other peak eluted with 0.6 M NaCl was subsequently separated into two forms, which were purified to homogeneity by the procedures described here. 0.010 For this purification,cellulose resin was unsuitable because the recovery was low, probably due tononspecific interaction. 0.005 0.01 Nevertheless, we used cellulose phosphate as the second purification step and, in this way, obtained significant separation of the two forms of PLC, PLC-I1 and PLC-111. Following 40 50 60 70 80 cellulose phosphate chromatography, we employed an HPLC Fraction n u m b e r system, which reduced the purification time andincreased the FIG. 2. Gel filtration HPLC of PLC-I1 ( A )and PLC-I11 ( B ) . reproducibility. After HPLC of each enzyme, only one peak Fractions in the peak of PLC-I1 activity from TSKgelSP5PW or that of activity was observed in each elution profile, and PLC-I1 of PLC-111activity obtained by hydroxylapatite chromatography were could not be purified by the procedure used for PLC-I11 and purified by TSKgelG3000SW gel filtration. vice versa. Our procedure achieved 2810- and 4010-fold purifications and 3.4 and 1.8%recoveries of PLC-I1 and PLC-111, respectively. The purifications are summarized in Table 11. The purities of the final preparations were confirmed by SDSPAGE, and bothenzymes were found to have an apparentM , of 85,000 (Fig. 4). The biochemical characteristics of the two enzyme preparations were examined. Biochemical Properties of Purified PLC-11 and PLC-IIIAddition of excess EGTA (2 mM) to the reaction mixture entirely inhibited the activitiesof both PLC-I1 and PLC-111, suggesting that bothenzymes require Ca2+ for activity.Thus, we examined the effects of various concentrations of Ca2+on PLC activity. The free Caz+ concentration was stabilized using CaEGTA buffer as described (30), and low concentrations of Ca2+ (below 10"j M) were determined by measurement of Ca2+-inducedfluorescence of Quin 2 (31). As shown in Fig. 5, activation of P I hydrolysis by PLC-I1 was observed to show a 0 10 20 30 40 dose dependence with respect to Caz+ concentrations up to F r a c t l o n number M untilit reacheda plateau a t 2 X or lo-' M Ca2+, FIG. 3. Ion-exchange chromatography of PLC-I1 ( A ) and whereashydrolysis of PIP and PIP, by this enzyme was PLC-I11 ( B ) .Fractions of PLC-I1 activity from gel filtration were M Ca2+ and was gradually reduced by subjected to cation-exchange Mono S HPLC. Those in the peak of maximal at 5 X PLC-111activity from gel filtration were subjected to anion-exchange higher concentrations of Ca2+.PLC-I11 was shown to require more than 1 mM Ca2+ for hydrolysisof PI andwas maximally HPLC. M Ca2+forhydrolysis of PIPand by activated by 5 X M Ca2+for that of PIP,. From theseresults, it is obvious was eluted as a peak at 60-65 min (Fig. 2 B ) . Step 6 for PLC-III: Mono Q HPLC-The pooled fractions were that at Ca2+concentrations of the order of M, both PLCapplied to an anion-exchange Mono Q column equilibrated with the I1 and PLC-I11 can degrade PIP, as well as PIP; whereas at same buffer as described in the previous step at a flow rate of 1 ml/ M,PLC-I1 can hydromin. The column was washed and developed with 20 ml of a linear Ca2+concentrations of the order of gradient of 0.1-0.3 M NaCl in the same buffer. Activity was eluted as lyze PI, but PLC-I11 cannot. The effect of pH on enzyme activity was also studied. Tris a single peak with 0.20-0.24 M NaCl (Fig. 3B), and the fractions in the peak were pooled and stored in portions at -80 "C. malate, Mes, Hepes, and sodium acetate buffers were used to adjust the pH of the reaction mixture. Hydrolysis of P I by Other Methods PLC-I1 was maximally activated in acetatebuffer at pH 5.5Protein was determined by a Bio-Rad protein assay with BSA as 5.8, whereas that of P I P or PIP, by this enzyme was fully a standard. SDS-PAGE was carried out by the method of Laemmli activated in Mes or Hepes buffer at pH 6.8 (Fig. 6). On the (33). Protein bands were visualized by a silver stain kitobtained from other hand, PLC-I11 did not show activity with PI at anyof Bio-Rad. these assay conditions andhydrolyzed PIP and PIPzactively at a neutral pH. Addition of 0.5% (w/v) cholate to PLC-I1 RESULTS shifted the optimal pH for PI to6.0-6.5, but had noeffect on for PIP or PIP, hydrolysis by PLC-I1 or PLCPurification of the Two Forms of PLC-Over 70% of the the optimal pH total PLC activitywas extracted in thesoluble fraction of rat 111. I
0.03 L
A
4
Purification of
Phospholipases Brain C Rat from
6595
TABLEI1 Purification of two distinct formsof inositol phospholipid-specific phospholipase C from 300 g of rat brain Step
Total
Total protein
activity
Specific activity
Purification
Yield
mg
pmolfmin
pmol f minf mg
-fold
%
PLC-I1 1. Crude extract 10,900 59.3" 0.00544 1 100 535 0.0856 15.7 77.2 2. Affi-Gel blue 45.8" 49.1 18.6 0.379 69.7 31.4 3. Cellulose phosphate 13.6 5.04 8.09 1.61 296 4. TSKgelSP5PW 0.910 4.81 5.29 972 8.11 5. TSKgelG3000SW 6. Mono S 0.132 2.01 15.3 2,810 3.39 PLC-111 1. Crude extract 10,900 35.1" 0.00322 1 100 2. Affi-Gel blue 535 25.4" 0.0475 14.8 72.3 10.8 8.21 0.760 236 23.3 3. Cellulose phosphate 7.29 1.03 2.56 2.49 773 4. Hydroxylapatite 0.236 1.23 5.21 1,620 3.50 5. TSKgelG3000SW 6. Mono Q 0.048 0.619 12.9 4,010 1.76 The relative contribution of each enzyme to thetotal amount of activity in the crude extract and in the 0.6 M NaCl fraction obtained by Affi-Gel blue chromatography was estimated from the area of each activity in Table I and Fig. 1.
The effects of substrates on activity were determined under two sets of conditions which induce maximal activation of PLC-I1 on PI and the two enzymes on PIP,, respectively (Fig. 7). Small unilamellar vesicles containing various amounts of substrates were used for the assay. Under the first set of conditions (2 mM Ca2+,100 p~ EDTA at pH 5.5),the apparent K, values of PLC-I1 for PI, PIP, andPIP, were 135,125, and 90 p ~respectively. , The VmaX values of PLC-I1 were 19.2 pmol of PI, 8.4 pmol of PIP, and1.6 pmol of PIP, cleaved per min/ mg of protein. On the other hand, the K, values of PLC-I11 , and the for PIP, and PIP were 130 and 80 p ~ respectively; V,,, values were 1.3 pmol of PIP, and6.0 pmol of PIP cleaved per min/mg of protein. Under these conditions, PIP, was not a good substate for PLC-I11 because this concentration of Ca2+inhibited PLC-I11 activity on PIP, (Fig. 5). Under the second set of conditions (100 pM Ca2+,100 p~ EDTA at pH 6.8), PLC-I1 and PLC-I11 showeda maximal activity on PIP, with Vmaxvalues of 15.3 and 12.9 pmol/min/mg of protein, respectively. PLC-I1 catalyzed hydrolysis of PIP and PI with V,,, values of 3.3 and 2.7 pmol/min/mg of protein. The apparent K, values of PLC-I1 for PIP,, PIP, and PIwere 60, 75, and >200 p ~ respectively. , PLC-111 also catalyzed the hydrolysis of PIP with a Vmaxvalue of 6.5 pmol/min/mg of protein, but this enzyme could not hydrolyze PI under these conditions. The apparent K, values of PLC-I11 for PIP, and , These twoenzyme PIP were 130 and 80 p ~ respectively. preparations were shown not to contain any PLC activity against PE, phosphatidylcholine, phosphatidylserine, or phosphatidic acid. Addition of deoxycholate at 1 mg/ml to the assay mixture containing 100 p M Caz+and 100 p~ EDTA at pH 6.8 reduced the apparent K,,, values of both enzymes by approximately 20% and also decreased the Vmaxvalues by 30%,whereas octyl glucoside (1% ,w/v) increased the rates of hydrolysis of PIP and PIP, mediated by both enzymes by 150 or 300%.Triton X-100 and Nonidet P-40 absolutely inhibited both enzyme activities. Structural Comparison of the Two Forms of PLC-The antigenicities of purified PLC-I1 and PLC-I11 wereexamined with their respective antisera raised in rabbits. The results from Western blot analysis are shown in Fig. 8. Anti-PLC-I1 antiserum reacted with PLC-11,but notwith the same amount of PLC-111, on nitrocellulose paper. On the other hand, antiPLC-I11 antiserum recognized only PLC-111, but not PLC-11. Additionally, these two antisera reacted with the M , 85,000
proteins in the crude extract of rat brain which were apparently corresponding molecules to PLC-I1 and PLC-111, respectively. The two forms of PLC were compared further by proteolytic analysis. As indicated in Fig. 9, both enzymes were sensitive to a-chymotrypsin and S. aureus V8 protease. However, the resultant peptide patterns from the two forms were quite different. These results suggest that these two forms of PLC differ not only in antigenicity but also in primary structure. DISCUSSION
In thiswork, we purified two forms of PI-specific phospholipase C to homogeneity from the soluble fraction of rat brain. One of these, designated PLC-11, hydrolyzed PIP, with a specific activity of approximately 15.1 pmol/min/mg of protein. The other one, designated PLC-111, catalyzed the hydrolysis PIP, with a specific activity of 12.9 pmol/min/mg of protein. The molecular weights of the two enzymes were both 85,000. These two enzymes were separated on a cellulose phosphate column (Fig. 1) and purified further by different procedures which werenot interchangeable. Furthermore, the two enzymes were found to have very different biochemical properties, such as dependencies on Ca2+and pH, and different substrate specificities (Figs. 5-7), and to show different structural features including antigenicities (Figs. 8 and 9). These results strongly suggest that the two enzymes belong to different families of PLC and that one is not derived from the other or from higher molecular weight enzymes by proteolysis during the purification procedures (Fig. 8). As reported previously, PLC-I1 and PLC-111, described here, also require Caz+for their activity. However, their profiles of Caz+dependence are very different depending on the substrate (Fig. 5). PLC-I1 is activated to hydrolyze PI in a dose-deM Ca2+,whereas PLC-I11requires pendent fashion above enormous concentrations of Ca2+(at least 2 X M Ca2+) which are much higher than those that occur physiologically. On the other hand, PLC-I11 shows the greatest activity for hydrolysis of polyphosphoinositides at M Ca2+,whereas PLC-I1 does so at M Ca2+.Thus, PLC-I11 can be easily activated to hydrolyze PIPz in the resting cells (e100 nM Ca2+),whereas PLC-I1 cannot; but PLC-I1 can degrade PI in activated cells in which Caz+concentration is elevated to or M. These results suggest different physiologicalroles of PLC-I1 and PLC-111.
Purification of Phospholipases C from Rat Brain
6596
a
A
b
4 t A
1
200
100
.-m 0
80
3
B
1
L
40
0 0
P
20
I
-3
-5
-2
M
FIG. 5. Effect of Ca2+on rate of hydrolysis of PI ( A ) , PIP ( B ) ,and PIP. (C). PLC-II(0) or PLC-III(0) activity was measured in buffer containing various concentrations of Ca2+and 100 p~ PI, PIP, or PIP? as substrate. Various free Ca2+ concentrations were obtained using CaEGTA buffer (30) and were confirmed by measuring fluorescence of Quin 2 (31).The assay buffer was adjusted to pH 6.8. Where indicated with an arrow (f), no Ca2+and 2 mM EGTA were added.
m
-
-6 -4
-1
logICa''l.
. 60 -
+
.e
-
1
I
I
1
02
0.4
0.6
0.8
4tA
I
Rf
FIG. 4. A, purity of the final enzyme preparations. 0.4 pg of PLC( l a n e b) were subjected to 7.5% SDS-PAGE. The gel was silver-stained (Bio-Rad) to visualize the protein bands. Arrowheads indicate molecular weight standards from the top: myosin (200,000), 8-galactosidase (116,000),phosphorylase b (94,000), BSA (67,000),and ovalbumin (45,000). B , molecular weight estimations ( ~ 1 0 of~ the ) two forms of PLC (0)from rat brain by SDS-PAGE. Electrophoretic mobilities are shown on the abscissa. The markers usedwere myosin, 8-galactosidase, phosphorylase b, BSA, ovalbumin, and carbonic anhydrase (0).
I1 ( l a n e a) and 0.15 pg of PLC-111
During purification procedures, PLC-I and PLC-I1 active peaks wererecognizedby the routine PLC assay (which contains PI as a substrateand or lo-' M Ca2+),butthe third peak was not. The PLC-I11 peak was found when fractions of eluate from cellulose phosphate were assayed in the reaction mixture containing PIP, and a low concentration of Ca2+.Therefore, this enzyme shows a marked preference for polyphosphoinositides. Thus far, anumber of forms of inositol phospholipidspecific phospholipase C have been isolated from the soluble fraction (12, 13, 18-26) as well as the particulatefraction (27, 28) of various mammalian tissues. Hofmann and Majerus (13) have identified two distinct forms of PLC (PLC-I and PLC11) in a soluble fraction of sheep seminal vesicles. Both enzymes were subsequently shown to hydrolyze PIP, and PIP as well as PI, and they appeared to be somewhat similar to ourpreparation of PLC-I1 in molecular weight, substrate specificity and pH dependence; but they were very different
s
1
4rc
4
5
6
7
8
9
pH
FIG.6. Effect of pH on rate of hydrolysisof PI ( A ) ,PIP ( B ) , and PIP2 (C).PLC-I1 (0)or PLC-111 (0)activity was determined a t various pH values. Tris malate (pH 4.0-9.0), Mes (pH 5.5-7.0), Hepes (pH 7.0-8.5), and sodium acetate (pH 4.0-5.5) were used to obtain the indicated pH values. The reaction mixture also contained 100 p~ EDTA, 100 p~ Ca", and 100 p~ substrate.
Purification of Phospholipases Brain CRat from I
a
6597 b
c
d
e
f
g
h
*-
t FIG.9. Peptide mapping analyses ofthe two forms of P L C by the method of Clevelandet d.(32).3 pg of PLC-I1 (lanesa, b, e, and f ) or 1 pg of PLC-I11 (lanesc, d, g, and h) was digested with S. aureus V8 protease or a-chymotrypsin, and theresultant peptides were subjected to SDS-PAGE (15%) and located by silver staining. Locations of authentic proteases are indicated by arrowheads. M,X
E
lo3.
weight of Type I was estimated to be approximately 300,000 by gel filtration, whereas that of PLC-I1 is 85,000;and we did not observe any PLC activity of material with such a high Ibubstratel. I N molecular weightduring purification. This inconsistency may FIG.7. Effects of substrateconcentration on PLC-I1 ( A and C)and PLC-111 ( Band D ) activities. PLC activity was determined be explained in part by the observation that, under nondeunder two conditions which induce activation of PLC-I1 for PI naturing conditions, PLC molecules aggregate to form oligohydrolysis ( A and B ) and that of both enzymes for PIP, hydrolysis mers (21). Ryu et al. (21) also isolated three forms of PLC (Cand D),respectively. The reaction mixtures for the former assay from bovinebrain. They used isoelectricsedimentation at pH contained 50 mM sodium phosphate (pH 5.5), 2 mM Ca2+, 100 pM 5.0 ofa brain extract as a first step finally and obtained nearly EDTA, and increasing amounts of small unilamellar vesicles of [3H] homogeneouspreparations of enzymes with molecular weights PI/PE (O), [3H]PIP/PE (A), or [3H]PIP,/PE (W). Those for the latter assay contained 50 m M Mes (pH 6.8), 100 pM Ca2+,100 pM of 150,000 (PLC-I), 145,000 (PLC-11),and 85,000 (PLC-111). EDTA, and the same substrates as described above. The ratio of We also resolved a PLC peak corresponding to their PLC-I and PLC-I1 on the first column of Affi-Gel blue.This enzyme phosphoinositide to PE in the vesicles was 1:0.4. (PLC-I described in our preparation) was recovered in the washing fraction and was purified 470-fold, and itsmolecular a b c d e f weight was estimated to be 135,000.This enzyme wasprecipitable at pH 5.0, as they reported; whereas the other forms of PLC (PLC-I1 and PLC-111) were not. Therefore, the two D enzymes described here seem to be different from the two forms of PLC that they isolated. It is likely, however,that the D M , 85,000PLC-I11 they isolated from bovine brain is the corresponding molecule of rat brain PLC-I1 reported here because of biochemical similarities of the two enzymes. Additionally, Bennett andCrooke (26) isolated a PLC (PI-PLCI) with a M , of 62,000 from guinea pig uterus. Although too little information exists about the biochemical properties of these enzymes forcomparison, this enzyme might bea different kind of PLC from rat brain PLC-I1 and PLC-I11 because of differences in molecular weight, pH preference, and Ca2+ dependence. Recently, Rebecchi and Rosen (24) have reported heterogeneity of a PLC enzyme in bovine brain tissue and FIG.8. Western blot analyses of PLC-JI and PLC-III. Anti- succeeded in identifying one of them, PLC-111. Although the PLC-I1 and anti-PLC-111antibodies were obtained as describedunder final enzyme preparation is not pure, this enzyme has aM , of "Experimental Procedures." The crude extract andpurified enzymes 88,000 and prefers PIP, as a substrate. Furthermore, hydrolwere subjected to 7.5% SDS-PAGE and immunoblotted with anti- ysis of PIP, by this enzyme is greatly enhanced by octyl PLC-I1 antibody (lanesa-c) or with anti-PLC-I11 antibody (lanes dglucoside. These properties are very similar to the PLC-I11 f). Immunoreaction products were detected with 4-chloro-l-naphthol described here. as a peroxidase substrate. Lanes a and d, crude extract of rat brain Recently, Majerus and co-workers (10,ll) have shownthat (25 and 50 pg of protein, respectively); lanes b and e, purified PLC-I1 (0.5 pg of each); lanes c and 1, purified PLC-111 (0.3 pgof each). diacylglyceride,which is a putative secondmessengerfor Arrowheads indicate molecular weight standards from the top: myosin activation of protein kinase C, is mainly derived fromPI, but (200,000), @-galactosidase(116,000), phosphorylase b (94,000), BSA not from PIP,. This finding suggests that breakdown of PI (67,000),and ovalbumin (45,000). was preferentially associated with activation of protein kinase C, and that PIP, was shown to be more important for actifrom the PLC-I11 described here. Nakanishi et al. (18) also vation of the intracellular Ca"-dependent process than of reported two forms of PLC in rat brain. These forms were protein kinase C. In this study, we isolated two different not purified to homogeneity; but one of them, designated forms of PLC from rat brain which differently hydrolyze Type I, showed similarities to PLC-11. However,the molecular inositol phospholipids with various concentrations of Ca2+as 0
200
600
400
l
800
0
200
400
(00
800
6598
Purification of Phospholipases C from Rat Brain
described above. These results may support the following. At the moment an external stimulus binds to its receptor, PLCI1 is activated to hydrolyze PIP2, which causes generation of inositol 1,4,5-trisphosphate and the resultant increase in intracellular Ca2+concentration. An increase of Ca2+concentration induces activation of intracellular Ca2+-dependent systems and also autoactivation of PLC-I1 to hydrolyze PI as well as PIP. Degradation of PI causes protein kinase C activation through generation of diacylglycerol.On the other hand, PLC-I11 activation by an external stimuluscauses only inositol 1,4,5-trisphosphate production, which is followed by activation of the Ca2+-dependentsystems. Further studies are necessary on the involvement of an individual PLC molecule in cellular responses to external stimuli, including the possible requirement of a nucleotide-binding protein or other components. REFERENCES 1. Berridge, M. J., and Irvine, R. F. (1984) Nature 312,315-321 2. Downes, P., and Michell, R. H. (1985) Mol. Aspects Cell. Regul. 4.2-56 3. Hokin, L. E. (1985) Annu. Rev. Biochem. 54, 205-235 4. Dawson, R.M.C., Freinkel, N., Jungalwala, F. B., and Clarke, N. (1971) Biochem. J. 122,605-607 5. Nishizuka, Y. (1984) Science 233, 305-312 6. Irvine, R. F., Letcher, A. J., Heslop, J. P., and Berridge, M. J. (1986) Nature 320, 631-634 7. Wilson, D. B., Bross, T. E., Sherman, W. R., Berger, R. A., and Majerus, P. W. (1985) Proc. Natl. Acad.Sci. U. S. A . 82,40134017 8. Connolly, T. M., Bansal, V. S., Bross, T. E., Irvine, R. F., and Majerus, P. W. (1987) J. Biol. Chem. 262,2146-2149 9. Houslay, M. D. (1987) Trends Biochem. Sci. 12, 1-2 10. Majerus, P. W., Connolly, T. M., Dickmyn, H., Ross, T. S., Bross, T. E., Ishii, H., Bansal, V. S., and Wilson, D. B. (1986) Science 234,1519-1526 11. Inhorn, R.C., Bansal, V. S., and Majerus, P. W. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 2170-2174
12. Takenawa, T., and Nagai, Y. (1981) J. Bwl. Chem. 2 5 6 , 67696775 13. Hofmann, S. L., and Majerus, P. W. (1982) J. Bwl. Chem. 2 5 7 , 6461-6469 14. Hakata, H., Kambayashi, J., and Kosaki, G. (1982) J. Biochem. (Tokyo) 9 2 , 929-935 15. Hirasawa, K.,Irvine, R. F., and Dawson, R. M. (1982) Bwchem. J. 205,437-442 16. Low, M. G., Carroll, R. C., and Weglicki, W. B. (1984) Biochem. J. 221,813-820 17. Wilson, D. B., Bross, T. E., Hofmann, S. L., and Majerus, P. W. (1984) J. Bwl. Chem. 259,11718-11724 18. Nakanishi, H., Nomura, H., Kikkawa, U., Kishimoto, A., and Nishizuka, Y. (1985) Biochem. Biophys. Res. Commun. 132, 582-590 19. Low,M. G., Carroll, R. C., and Cox, A. C. (1986) Biochem. J. 2 3 7 , 139-145 20. Banno, Y., Nakashima, S., and Nozawa, Y. (1986) Biochem. Biophys. Res. Commun. 136, 713-721 21. Ryu, S. H., Cho, K. S., Lee, K. Y., Suh, P. G., and Rhee, S. G. (1986) Biochem. Biophys. Res. Commun. 141, 137-144 22. Manne, V., and Kung, H. F. (1987) Biochem. J. 2 4 3 , 763-771 23. Ryu, S. H., Cho, K. S., Lee, K. Y., Suh, P. G., and Rhee, S. G. (1987) J. Bwl. Chem. 262,12511-12518 24. Rebecchi, M. J., and Rosen, 0.M. (1987) J. Biol. Chem. 2 6 2 , 12526-12532 25. Ryu, S. H., Suh, P. G., Cho, K. S., Lee, K. Y., and Rhee, S. G. (1987) Proc. Natl. Acad. Sci. U. S. A. 8 4 , 6649-6653 26. Bennett, C. E., and Crooke, S. T. (1987) J. Biol. Chem. 262, 13789-13797 27. Lee, K. Y.,Ryu, S. H., Suh, P. G., Choi, W. C., and Rhee, S. G. (1987) Proc. Natl. Acad. Sci. U. S. A . 8 4 , 5540-5544 28. Katan, M., and Parker, P. J. (1987) Eur. J. Biochem. 168, 413418 29. Schacht, J. (1978) J. Lipid Res. 1 9 , 1063-1067 30. Fabiato, A., and Fabiato, F. (1979) J. Physiol. (Paris) 75, 463505 31. Tsien, R. Y.,Pozzan, T., and Rink, T. J. (1982) J. Cell Biol. 9 4 , 325-334 32. Cleveland, D. W., Fischer, S. G., Kirschner, M. W., and Laemmli, U. K.(1977) J. Bwl. Chem. 2 5 2 , 1102-1106 33. Laemmli, U. K. (1970) Nature 227,680-685
