Purification and Characterization of a Phosphoinositide-specific Phospholipase C from Guinea Pig Uterus PHOSPHORYLATION BY PROTEIN KINASE C IN VIVO*

Two peaks of phosphoinositide-specific phospholi- pase C (PI-PLC) activity were resolved when guinea pig uterus cytosolic proteins were chromatographed on a DEAE-Sepharose column. The first peak of enzyme activity eluting from the DEAE-Sepharose column (PI- PLC I) was further purified to homogeneity, whereas the second peak of enzyme activity was enriched 300- fold. PI-PLC I migrated as a 62-kDa protein on sodium dodecyl sulfate-polyacrylamide gels. Antibodies pre- pared against PI-PLC I failed to react with PI-PLC 11. PI-PLC I hydrolyzed all three phosphoinositides, ex- hibiting a greater Vmax for phosphatidylinositol 4,5-bisphosphate > phosphatidylinositol 4-phosphate > phosphatidylinositol. Hydrolysis of phosphatidylinosi-to1 was calcium-dependent, whereas significant hy- drolysis of phosphatidylinositol 4-phosphate and phos-phatidylinositol4,5-bisphosphate occurred in the pres- ence of 2.5 mM EGTA. At physiological concentrations of calcium, phosphatidylinositol 4-phosphate and phos- phatidylinositol 4,5-bisphosphate were the preferred substrates. Antibodies specific for PI-PLC I reacted with a 62-kDa protein in both the cytosol and membrane frac- tions from guinea pig uterus.


Purification and Characterization of a Phosphoinositide-specific Phospholipase C from Guinea Pig Uterus
Two peaks of phosphoinositide-specific phospholipase C (PI-PLC) activity were resolved when guinea pig uterus cytosolic proteins were chromatographed on a DEAE-Sepharose column. The first peak of enzyme activity eluting from the DEAE-Sepharose column (PI-PLC I) was further purified to homogeneity, whereas the second peak of enzyme activity was enriched 300fold. PI-PLC I migrated as a 62-kDa protein on sodium dodecyl sulfate-polyacrylamide gels. Antibodies prepared against PI-PLC I failed to react with PI-PLC 11. PI-PLC I hydrolyzed all three phosphoinositides, exhibiting a greater Vmax for phosphatidylinositol 4,5bisphosphate > phosphatidylinositol 4-phosphate > phosphatidylinositol. Hydrolysis of phosphatidylinosi-to1 was calcium-dependent, whereas significant hydrolysis of phosphatidylinositol 4-phosphate and phos-phatidylinositol4,5-bisphosphate occurred in the presence of 2.5 mM EGTA. At physiological concentrations of calcium, phosphatidylinositol 4-phosphate and phosphatidylinositol 4,5-bisphosphate were the preferred substrates.
Antibodies specific for PI-PLC I reacted with a 62-kDa protein in both the cytosol and membrane fractions from guinea pig uterus. Quantitation of the immunoblots revealed that 25% of the 62-kDa protein was membrane-associated, whereas only 5% of the total enzyme activity was membrane-associated. Approximately 20% of the membrane-bound phospholipase C activity and immunoreactive material were loosely bound, whereas the remainder required detergent extraction for complete solubilization. The 62-kDa protein associated with the membrane fractions did not bind lectin affinity columns, suggesting that it was not glycosylated. PI-PLC I was identified as a phosphoprotein in [32P]orthophosphate-labeled rat basophilic leukemia (RBL-1) cells by two-dimensional gel electrophoresis followed by immunoblotting. In untreated cells, 32P-labeled PI-PLC I was found in the cytosolic fraction. Treatment of RBL-1 cells with those phorbol esters which are known to activate the Ca2+/ phospholipid-dependent enzyme protein kinase C, resulted in a time-dependent increase in the phosphorylation of both membrane-bound and cytosolic PI-PLC I. Thus, in RBL-1 cells, protein kinase C may play an important role in the regulation of phospholipase C through protein phosphorylation.
PI-specific PLCs are ubiquitous enzymes, present in most mammalian cells, as well as in plants and various microorganisms (reviewed in Ref. 27). Based on separation of multiple peaks of enzyme activity by ion exchange chromatography, gel filtration chromatography, or chromatofocusing, recent reports have suggested the existence of multiple forms of PI-PLC in mammalian cells (27)(28)(29)(30)(31)(32)(33)(34)(35)(36). However, these types of analyses did not discriminate between the existence of different molecular entities and proteolysis or aggregation of a single enzyme species. In fact, some reports suggest that proteolysis may be responsible for generating some of the observed heterogeneity in enzyme activity (30,33,35). More definitive proof for the existence of multiple species of PI-PLC was derived from immunochemical studies, in which two immunologically distinct forms of PI-PLC were demonstrated in sheep seminal vesicle and other tissues (31). PI-specific PLCs have been purified to apparent homogeneity from rat liver (37), sheep seminal vesicles (31), bovine platelets (38), and bovine brain (39). The purified enzymes may be grouped into two categories on the basis of molecular masses. The purified enzymes from sheep seminal vesicles and rat liver 13789 exhibited molecular masses of 65 and 70 kDa, respectively, on SDS-polyacrylamide gels (31,37), whereas the enzymes purified from bovine platelets and bovine brain exhibited molecular masses of approximately 143-150 kDa on SDS-polyacrylamide gels (38,39). In the case of bovine brain, two immunologically distinct forms of PI-PLC were also identified which exhibited molecular masses of 145 and 150 kDa. Thus, within each molecular mass group, several distinct isoenzymes may exist. The initial studies characterized the substrate specificity and divalent ion requirements for the isolated enzymes (31,(37)(38)(39)(40)(41). All four enzymes hydrolyzed phosphatidylinositol, but not phosphatidylcholine, phosphatidylethanolamine, or phosphatidylserine (31,(37)(38)(39)(40)(41). The sheep seminal vesicle enzymes were analyzed in detail for hydrolysis of the polyphosphoinositides (41). It was reported that, at physiological concentrations of calcium, PIP, was the preferred substrate.
Many unanswered questions remain concerning the regulation of PI-PLC by divalent cations, guanine nucleotide regulatory proteins, phospholipids, post-translational modification, etc., as well as the genetic regulation of the enzymes. As a first step in answering these questions, we have purified to homogeneity a PI-specific PLC from guinea pig uterus cytosol. The enzyme exhibits characteristics which are consistent with its being coupled to receptor-mediated phosphoinositide hydrolysis, e.g. substrate preference, divalent cation requirements, and membrane localization. Furthermore, we demonstrate that PI-PLC I is phosphorylated in vivo in response to those agents which are known t,o activate protein kinase C.

RESULTS
Characterization of PI-PLC I from Guinea Pig Uterus-PI-PLC I from guinea pig uterus exhibited similar pH optima and kinetic parameters as sheep seminal vesicle PLC (31). As previously reported (31), deoxycholate markedly enhanced enzyme activity (data not shown); therefore, all subsequent experiments were performed using an optimum concentration of deoxycholate (1 mg/ml). Purified PI-PLC I exhibited a biphasic pH optima using PI as a substrate, showing a minor peak at pH = 5.5, and a major peak of activity between pH = 6.5 and pH = 7.0 ( Fig. 7). Changing the free calcium concentration from 4 @M to 1 mM did not change the biphasic character of the pH profile (data not shown).
The calcium requirements of purified PI-PLC I were analyzed using PI, PIP, and PIP, as substrates. Hydrolysis of PI was calcium-dependent, reaching a maximal rate of hydrolysis of 550 nmol/mg/min, with 10 PM PI, at 1.0 mM calcium (data not shown). The rates of hydrolysis for the three different phosphoinositides in the physiological range of calcium were determined a t neutral pH. In contrast to PI, significant hydrolysis of PIP and PIP, occurred in the absence of calcium (Fig. 8). In the presence of 2.5 mM EGTA, PIP hydrolysis was 25% of the rate observed at 4.0 HM free calcium, and PIP, hydrolysis was 33% (Fig. 8). The calcium-stimulated hydrolysis profiles for PIP and PIP, were very similar. The rate of  The calcium concentration between 0 and 4 PM was determined as described previously (60). For data points without standard error bars, the standard error of the mean was less than the size of the symbol.
hydrolysis for both substrates reached plateau levels at 700 nM free calcium and was half-maximal between 120 and 200 nM (Fig. 8). Increasing the free calcium concentration from 4.0 to 300 PM produced a further increase in the rate of hydrolysis of PIP and PIP, by 35 and 57%, respectively. In contrast to PIP and PIP2, the hydrolysis of PI was only modestly affected by changes in the calcium concentration over the range of 0-4 WM (Fig. 8). Other divalent cations such as cobalt, cadmium, manganese, magnesium, and zinc did not substitute for calcium in stimulating PI hydrolysis (data not shown).
The substrate kinetics of PI-PLC I from guinea pig uterus were also similar to that reported for sheep seminal vesicle PLC I (41). At 1.0 mM calcium and pH = 7.0, the enzyme exhibited a K,,, value between 25 and 35 FM and a V,,, of 3.2 pmol/mg/min (data not shown). The substrate kinetics toward the polyphosphoinositides were examined at 4.0 pM calcium to more closely mimic the intracellular calcium concentration (Fig. 9). The K , values obtained by double-reciprocal plots of the data in Fig. 9 were: PI, 11 FM; PIP, 40 p M ; and PIP2, 100 PM; whereas V,,, values for PI, PIP, and PIP, were 0.7, 3.2, and 7.1 @mol/mg/min, respectively. It should be noted that, at substrate concentrations greater than 5 p M , the rate of hydrolysis of PIP, was 5-10 times greater than the Assays were performed at pH = 7.0 in the presence of 4 p~ calcium for 10 min a t 37 "C. Enzyme concentration was changed for each substrate such that between 5 and 25% of total substrate was hydrolyzed. Each data point was determined in triplicate. The standard error was less than the size of the symbol for those data points without standard error bars. rate of hydrolysis of PI (Fig. 9). Subcellular Distribution of PLC I-The role of cytosolic PI-PLC in mediating agonist-induced phosphoinositide hydrolysis has been controversial (21, 25, 27, [42][43][44]. The substrate preference and calcium requirements of PI-PLC I isolated from guinea pig uterus cytosol are consistent with this enzyme being coupled to agonist-induced phosphoinositide turnover. To determine whether PI-PLC I was membrane-localized, the subcellular distribution of PI-PLC was examined by immunoblotting. Equal quantities of membrane and cytosolic proteins (100 pg) from guinea pig uterus were loaded onto each lane of the gel. The gel was either stained with Coomassie Brillant Blue R-250 ( Fig. 10A) or transferred to nitrocellulose paper and probed with PI-PLC I antibodies ( Fig. 10B) or with preimmune immunoglobulins (Fig. lOC). Antibodies generated against PI-PLC Ib reacted specifically with a 62-kDa protein in cytosolic and membrane fractions from guinea pig uterus (Fig. 10B). These data also demonstrate that PI-PLC I is not a proteolytic product of a higher molecular weight protein. Preimmune immunoglobulins failed to react with any proteins on the nitrocellulose paper (Fig. 1OC). The staining intensity was slightly greater for membrane fractions (Fig.  10B) suggesting that the PI-PLC I was enriched in membranes relative to cytosol. Quantitation of the amount of immunoreactive activity in each fraction by densitometric scanning revealed that 25% of the total PI-PLC I was membrane-bound. In contrast, using either PIPz ( Table 2) or PI as substrates, only 3-5% of the total enzymatic activity was membrane-bound. Thus, measurement of enzymatic activity under the conditions described under "Materials and Methods" may underestimate the total amount of PI-PLC associated with membranes.
The type of interaction between PI-PLC I and cell membranes was determined by sequentially extracting the membranes with low ionic strength buffer (10 mM bis-Tris, pH = 7.0, 10 mM NaC1, 2 mM EGTA) twice, followed by a high ionic strength buffer (1 M KCl), and finally a detergent extraction with 1% sodium cholate. The extracts were ana-    11A) to visualize the total proteins separated on the gel, and the proteins in the second gel were transferred to nitrocellulose paper and probed with PI-PLC I antibodies (Fig. 11B). PI-PLC I was detected in each of the extracts (Fig 11B). Quantitation of the amount of PI-PLC I in each fraction by densitometric scans of the immunoblots revealed that 16% of the immunoreactive material was found in the low salt extracts, 10% in the high salt extracts, 70% in the detergent extract, and 4% in the residual pellet. The distribution of PI-PLC I closely paralleled the distribution of total enzymatic activity. Using PIP, as a substrate, 20% of total enzymatic activity in the membrane was extracted in the two 10 mM NaCl extracts, 10% in the 1 M KC1 extract, and 140% in the 1% sodium cholate extract ( Table 2). Similar results were also obtained using PI as a substrate (data not shown).
T o determine whether the membrane-bound form of PI-PLC I was glycosylated, the 1% sodium cholate extract was applied to a lentil lectin-Sepharose column and bound material eluted with 250 mM a-methylmannoside. No detectable enzymatic activity bound to the column, with the total enzyme activity applied to the column being accounted for in the flowthrough fractions. Furthermore, immunoblot analysis of flowthrough and eluted fractions demonstrate that the 62-kDa protein was in the flow-through fractions (data not shown). Similar results were obtained with wheat germ lectin. Thus, we cannot account for the strong interaction of PI-PLC I with membranes by glycosylation.
Phosphorylation of PI-PLC I in Vivo-Introduction of a net negative charge to PI-PLC I through phosphorylation may affect the interaction between PI-PLC I and membrane components. Therefore, phosphorylation of PI-PLC I was examined as a mechanism for regulating the distribution of PI-PLC I between soluble and membrane-bound forms. T o address this possibility it was necessary to label PI-PLC I in vivo with 32P. Several cell lines were screened for the presence of the 62-kDa protein by immunoblotting. A rat basophilic leukemia cell line (RBL-1) produced the 62-kDa protein, which reacted with the PI-PLC I antibodies, and co-migrated with the corresponding proteins from guinea pig uterus on SDS-polyacrylamide gels (data not shown). Phosphorylation of PI-PLC I was demonstrated to occur in these cells by labeling cells with ["P]orthophosphate (100 pCi/ml for 2.5 h at 37 "C), lysing the cells and separating total cellular proteins by two-dimensional gel electrophoresis (45). The proteins were transferred from the gel to nitrocellulose paper and then exposed for autoradiography. PI-PLC I was identified by immunoblotting using alkaline phosphatase-conjugated-goat anti-rabbit IgG to detect the primary antibody. The autoradiography was superimposed over the immunoblot, thus identifying PI-PLC I as a phosphoprotein. A minor 62-kDa phosphoprotein, which exhibits an apparent PI of 6.7 (Fig. 12a), and protein from lo6 cells (250 pg) was separated by two-dimensional gel electrophoresis (isoelectric focusing/SDS). Proteins were transferred to nitrocellulose paper and exposed for autoradiography. After obtaining a proper exposure, the nitrocellulose sheets were probed with PI-PLC I antibodies. Immunoreaction products were detected with alkaline phosphatase-conjugated goat anti-rabbit IgG as described under "Materials and Methods." a and c represent the autoradiographs from control and PMA-treated cells, respectively. The corresponding immunoblots are shown in b and d. Proteins which consistently showed an increase in phosphorylation in response to PMA treatment are indicated with arrows. PI-PLC I is indicated by arrows pointing to the left.
was identified as PI-PLC I (Fig. 12b). Numerous investigations have demonstrated that prior treatment of cells or tissues with phorbol esters such as PMA or derivatives which activate protein kinase C would attenuate a subsequent agonist-induced PI turnover and/or calcium mobilization (46)(47)(48)(49)(50)(51)(52)(53)(54). These results suggest that protein kinase C phosphorylated either the receptor, the guanine nucleotide-binding protein, phospholipase C, inositol trisphosphate 5'-phosphomonoesterase (54) or a combination of the above. The effects of PMA treatment on phosphorylation of PI-PLC I were examined as described above. Pretreatment of RBL-1 cells for 5 min with 1 p~ PMA prior to cell lysis resulted in enhanced phosphorylation of several proteins, as indicated by arrows (Fig. 12c) compared to control (Fig. 12a). Phosphorylation of the 62-kDa protein that reacts with PI-PLC I antibodies was markedly increased (Fig. 12, C and d ) compared to control levels (Fig. 12a). The specificity of the enhanced phosphorylation was examined using phorbol ester derivatives which are both active and inactive in stimulating protein kinase C. Phosphorylation of the 62-kDa protein was not increased above control levels (Fig. 13A) by the inactive phorbol analogues 4a-phorbol 12,13-didecanoate (Fig. 13B) or 4P-phorbol (data not shown), but was stimulated by 1 p~ PMA (Fig. 13c) and 1 p~ phorbol 12,13-dibutyrate (Fig.  130). Thus, PI-PLC I is phosphorylated in response to treatment by those compounds which activate protein kinase C. PMA-induced phosphorylation of PI-PLC I was time-dependent (Table 3), reaching maximum levels between 15 and 30 min. These data suggest that PI-PLC I may be phosphorylated in vivo by protein kinase C.  Cytosolic and washed membrane fractions from RBL-1 cells labeled with ["P]orthophosphate were analyzed by two-dimensional gel electrophoresis followed by immunoblotting as described under "Materials and Methods" to determine the subcellular localization of phosphorylated PI-PLC I. The phosphorylated form of PI-PLC I was found in the cytosol fraction from control cells (Fig. 14A, arrow), with no detectable phosphoprotein in the membrane fraction (Fig. 14B,  arrow). In contrast, cells pretreated with 1 PM PMA for 5 min (data not shown) and for 15 min exhibited phosphorylated PI-PLC I in both cytosolic (Fig. 14C) and membrane fractions (Fig. 140). There was no detectable change in the distribution of PI-PLC I as judged by the corresponding immunoblots (data not shown). These data demonstrate that PI-PLC I is phosphorylated in RBL-1 cells by a protein kinase activated by treatment with active phorbol esters, probably protein kinase C. Furthermore, pretreatment of cells with PMA results in phosphorylation of primarily the membrane-associated form of PI-PLC I.

DISCUSSION
We have demonstrated that there are at least two immunologically distinct forms of PI-PLC in unsynchronized guinea pig uteri. Form I was purified to apparent homogeneity as determined by SDS-polyacrylamide gel electrophoresis, and is similar in chemical and kinetic properties to PI-PLC I from sheep seminal vesicle (31). However, PI-PLC I isolated from guinea pig uterus and sheep seminal vesicle exhibits slightly different molecular masses on SDS-polyacrylamide gels (62 and 65 kDa, respectively). This difference in molecular masses may be due to different electrophoretic techniques, or may represent species differences. Further characterization of PI-PLC I1 will require purification of the enzyme to homogeneity. It is of interest that the most enriched preparation of form I1 PI-PLC from guinea pig uterus contains three predominant bands exhibiting apparent molecular weights of 87,000, 73,000, and 58,000 which are similar to the molecular weight reported by Wilson et al. (41) for their most enriched preparation of form I1 PI-PLC from sheep seminal vesicle ( M , = 90,000, 62,000, 58,000, and 54,000). Also, both preparations contain proteins that have apparent molecular weights of 75,000-85,000 by gel filtration chromatography suggesting that PI-PLC I1 is not a dimeric enzyme. The relationship between the individual bands visualized by SDS-polyacrylamide gel electrophoresis is currently unknown.
PI-PLC I specifically hydrolyzes the phosphoinositides. At 4.0 FM free calcium, PI-PLC exhibited a greater V,,, toward PIP2, followed by PIP then PI. However the substrate affinity was reversed, e.g. PI > PIP > PIP,. At resting, intracellular calcium levels (-100 nM) PIP and PIP2 would be the preferred substrates. PI-PLC I failed to hydrolyze phosphatidylcholine and phosphatidylethanolamine (data not shown), as was reported for the sheep seminal vesicle enzyme (31), rat liver PI-PLC (37), and the platelet enzyme (38). Furthermore, guinea pig uterus PI-PLC I was inactive in liberating alkaline phosphatase from placental membranes under conditions in which Staphylococcus uureus PI-PLC was a~t i v e .~ Thus, the uterine enzyme does not appear to hydrolyze membrane proteins anchored to the membrane by a specific interaction with the polar head group of phosphatidylinositol (55,56).
The relationship between membrane-bound and cytosolic PI-PLC has been controversial (21, 25, 27, [42][43][44], in that it is not clear whether they are distinct enzymes, or whether the same enzyme is distributed between the two intracellular pools. It is generally believed that only the membrane-bound enzyme is involved in receptor-mediated phosphatidylinositol metabolism. The finding that PI-PLC I is distributed between soluble and membrane-bound pools supports the conclusion that membrane-bound and cytosolic PI-PLC are the same enzyme. These data are also consistent with PI-PLC I being involved in agonist-induced PI hydrolysis. The fact that we and others find 70-95% of enzymatic activity in the cytosol can be explained by artifactual redistribution of the enzyme upon cell homogenization. Alternatively, the differential distribution of the enzyme may represent a mechanism for regulating enzyme activity in a manner analogous to the regulation of protein kinase C (2, 12).
The mechanism by which phospholipase C binds to the membrane is unknown. We found that approximately 20% of enzymatic activity and mass of PI-PLC I was loosely bound to the membrane. Only 10% of total membrane PI-PLC was solubilized with 1 M KCI, suggesting that the interaction with the membrane was not ionic. The remaining enzyme activity, as well as mass of PI-PLC I, was solubilized with 1% sodium cholate, which suggests that it was tightly associated with the membrane. We were unable to detect any glycosylated enzyme which would account for the preferential association with the cell membrane. As discussed below, protein phosphorylation, or lack thereof, may play a role in the association of the enzyme with the membrane. It should be noted that extraction of membranes with detergent resulted in a 2-fold increase in total enzyme activity. This could be explained by the presence of an endogenous inhibitor in membranes. The observation M. G. Low, personal communication.

FIG. 14. Subcellular localization
of phosphorylated PI-PLC I. Rat basophilic leukemic cells were labeled with ["2P]orthophosphate, treated with 1 p~ PMA and fractionated as described under "Materials and Methods." Proteins were separated by two-dimensional gel electrophoresis, transferred to nitrocellulose paper, exposed for autoradiography, and probed with PI-PLC I antibodies.

5.7K
that approximately 25% of the total mass of PI-PLC I was associated with membranes, whereas only 3% of enzyme activity was associated with membrane fractions (of which 30-40% can be accounted for by PI-PLC I) supports this conclusion.
Pretreatment of intact cells with active phorbol esters desensitizes agonist-induced PI turnover (46)(47)(48)(49)(50)(51)(52)(53)(54). This result may be explained by protein kinase C phosphorylating the receptor (57), the GTP-binding protein (58), or PI-PLC. Phosphorylation of the a-adrenergic receptor in response to phorbol ester treatment has been demonstrated to occur in vivo (57), and phosphorylation of the CY subunit of the inhibitory guanine nucleotide-binding protein by kinase C has been demonstrated to occur in vitro (58). Recently, protein kinase C was shown to phosphorylate platelet inositol trisphosphate 5'-phosphomonoesterase (54). Phosphorylation of the phosphatase increases the V,,, of the enzyme which, in turn, would decrease the biological half-life of inositol 1,4,5-trisphosphate (54). Thus, .phosphorylation of the phosphatase would attenuate agonist-induced calcium transients, but not diacylglycerol production. In the present manuscript we demonstrate that phosphorylation of PI-PLC I is enhanced by treating RBL-1 cells with PMA or phorbol 12,13-dibutyrate, but not by the inactive compounds 4a-phorbol 12,13-didecanoate and 4P-phorbol. PMA-enhanced phosphorylation of PI-PLC I was time-dependent. These data suggest that PI-PLC I is phosphorylated in vivo by protein kinase C. We have also demonstrated that purified guinea pig uterus PI-PLC I is a substrate for partially purified protein kinase C in Thus, protein kinase C may potentially attenuate agonistinduced phosphoinositide metabolism though feedback inhibition a t several steps.
In untreated RBL-1 cells, 32P-labeled PI-PLC I was detected only in the soluble fraction. Prolonged exposure of the x-ray film failed to provide evidence for "P-labeled PI-PLC I in membrane fractions, even though PI-PLC I was demonstrated to be present in the membrane fractions by immuno-C. F. Bennett, unpublished observation.
blotting. In contrast to control cells, phosphorylated PI-PLC I was detected in both cytoplasmic and membrane fractions of those cells treated with PMA. The identification of phosphorylated PI-PLC I in membrane fractions following PMA treatment is consistent with the model that PMA activates and promotes the translocation of protein kinase C from the cytoplasm to the cell membrane, resulting in the phosphorylation of membrane proteins (12,59). It is not known what effect phosphorylation of PI-PLC I has on the signal transduction process. An attractive hypothesis would be that phosphorylation of membrane-bound PI-PLC I transforms the enzyme into a state such that it is no longer capable of interacting with the proposed guanine nucleotide-binding protein or another unidentified membrane component. Experiments are currently in progress to answer this question. The r e l a t i v e r a t i o between peaks I and I1 varled frm preparatlon to preparatlon and may reflect the hornanal status of the uterl.
Peak I 11% f u r t h e r p u r l f l e d t o h o m g m e l t y .
detected but were not well resolved (Fig. 2). Peaks Ia and I b were pooled Separately and Peak I was next applled t o an amlnohexyl-Sepharore column. Two peaks Of a c t i v i t y were I (31) guinea p i g u t e r l PLC I & and l b bound t o heparin-SePharore * I t h a higher a f f l n i t Y further jractlonated over a heparin-sepharose column.
I" c o n t r a s t t o sheep seminal vesicle PLC w l t h C m a r l e b r l l l a n t R-250 (Fig. 4 0