The Specificity of Phospholipase A, and Phospholipase C in a Mixed Micellar System*

The activities of cobra venom phospholipase A, (Nuja nuja nuja) and phospholipase C (Bacillus cereus) toward different phospholipids in mixed micelles with Triton X-100 are reported. The nonionic surfactant acts as an inert matrix which solubilizes the phospholipids in similar structures; thus, the observed activities can be compared directly.

The activities of cobra venom phospholipase A, (Nuja nuja nuja) and phospholipase C (Bacillus cereus) toward different phospholipids in mixed micelles with Triton X-100 are reported.
The nonionic surfactant acts as an inert matrix which solubilizes the phospholipids in similar structures; thus, the observed activities can be compared directly. Knowledge of phospholipid specificity is important in interpreting the action of these enzymes on natural membranes. Phospholipase A, prefers phosphatidylcholine as its substrate; it hydrolyzes phosphatidylethanolamine poorly. The enzyme requires Ca"+ for activity and its true specificity toward phosphatidylserine is difficult to interpret because this negatively charged phospholipid forms complexes with Cay+. Sphingomyelin is not hydrolyzed at all. Phospholipase C, which is a Zny+ metalloenzyme, hydrolyzes both phosphatidylcholine and phosphatidylethanolamine at almost equivalent rates; phosphatidylserine is a Z-to 3-fold poorer substrate than these other lipids. Sphingomyelin does not serve as a substrate.
Both enzymes show an effect of fatty acid chain length: as the chain length is reduced, enzyme activity increases. This trend is observed with both phosphatidylcholine and phosphatidylethanolamine and probabIy reflects optimization of the interfacial properties of the phospholipid. NMR studies on phosphatidylcholine in mixed micelles have shown that the cu-methylene groups of the sn-1 and sn-2 fatty acids are chemically shifted from one another and that the ethanolamine moiety alters this chemical shift pattern of the fatty acid cu-methylene groups somewhat from that observed in phosphatidylcholine.
This difference between the two phospholipids may correlate with the specificity results reported here for phospholipase A, catalysis, but does not for phospholipase C catalysis.
* These studies were supported by National Institutes of Health Grant 501 and National Science Foundation Grant PCM 7% 21552. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Postdoctoral Fellow of the National Institutes of Health 190    The phospholipid concentration was 6 rn~ and the Triton X-100 concentration was 48 mM (mole ratio 8:1), 24 mM (mole ratio 411, or 12 mM (mole ratio 21) or it was omitted (mole ratio 0).
- activit.y is comparable to that of the long chain dipalmitoyl phosphat.idylcholine in Triton micelles.

Advantage of Mixed
Micelles for Specificity Studies-In order to determine phospholipid preferences of enzymes which work on aggregated substrates, one needs a well defined matrix which is independent of phospholipid structure. The nonionic surfactant Triton X-100 serves this purpose. At 8:l Tritonlphospholipid, although enzyme activity is much lower than toward 2:l mixed micelles because of surface dilution (9, 141, there is little perturbation of the basic Triton micelle structure (121. We have now found for a monodisperse Triton analogue that all phospholipids, regardless of head group, form similarly sized micelles at this surfactantlphospholipid ratio." Also, the large excess of Triton molecules should serve to decrease any patching or clustering of phospholipid molecules in the micelle surface which might occur. Triton X-100 is also useful for comparing enzyme activity on synthetic short chain phospholipids. When Triton, with its very low critical micelle concentration (0.3 mM), is combined with zwitterionic short chain phosphatidylcholines, the critical micelle concentration of the phospholipid is reduced sharply so that assays can be conducted at phospholipid concentrations in which all of the phospholipid is micellar. This makes it easy to compare enzyme activity toward phospholipids of varying chain length without the complication of different micellar structures, sizes, or residual phospholipid monomers. Phospholipase A, Specificity-Cobra venom phospholipase A, has been used to investigate the sidedness of phospholipid distribution in red blood cells and other natural membrane systems. Using a large excess of enzyme, it was shown that in red blood cells, phosphatidylcholine was hydrolyzed before phosphatidylethanolamine or phosphatidylserine (33, 34). If these results are real indications of membrane asymmetry, one must be sure that the relative activity of phospholipase A, toward the different phospholipids in these membranes is not very different. The results reported here show different apparent specificities for phospholipase A, toward the different phospholipids when each is assayed separately in the Triton model system. In order to generalize to membranes, it will be necessary to examine specificities in mixtures of ' R. J. Robson and E. A. Dennis, manuscript in preparation. phospholipids and in preparations containing other components as these too can conceivably affect the structure of the sub&ate and enzyme activity. ' Earlier work on snake venom specificity showed certain trends: phosphat idylcholine and phosphat idylet.hanolamine were good subst.rates, while phosphatidylserine was hardly hydrolyzed. Dawson (35) used an ethereal solution at 30" with 0.45 mM Ca>+, but found hydrolysis of phosphatidylethanolamine occurred without added CaZ' and also without ether. Ibrahim et al. (3) found similar results using deoxycholate micelles and no added Cad+. These results differ dramatically from t,hose reported in this study. It is likely that the discrepancy results from t.he nonenzymat,ic hydrolysis of phosphatidylet.hanolamine which was apparently not followed in the lat,ter studies and which we have shown is quit.e rapid at pH = 8 in the Triton mixed micelle system. Cobra venom phospholipase A, (Naja naja nuja) was found here to require Ca'+ for phosphatidylethanolamine hydrolysis in mixed micelles. In a more recent study, Salach et al. (8) assayed in the presence of Triton X-100 and found phosphat idylethanolamine to be a poor substrate, but. their experiments were done at 25", a temperature well below the t,hermotropic phase transition temperature (36, 37) of several of the saturated phospholipids they studied. In fact, one of t.he major conclusions of that study was that saturation of t.he fatty acid in t,he sn-2 position results in a large decrease in t.he hydrolysis rate without, an increase in K,,, At. that temperat.ure, dipalmitoyl phosphatidylcholine/Triton X-100 mixtures actually phase separate (12,38) and this apparently decreased the hydrolysis rate (9).
In all studies phosphatidylserine was found to be a poor substrate. The usual explanation given is that a net negatively charged substrate adversely affects the enzyme which is specific for zwitterionic substrates. An alternate reason is that phosphatidylserine, unlike the choline-or ethanolaminecontaining phospholipids, binds Ca'+ tightly, and under the conditions of most phospholipase A, assays, the large excess of Ca'+ forms insoluble complexes with phosphat.idylserine. This insoluble or precipitated material would be inert to phospholipase A, action. It is only as one lowers the Ca'+ concentration so that the phosphatidylserine stays in solution that the inherent enzyme activity can be measured. We have found that the rate is 0.05 to 0.15 that for dipalmitoyl phosphatidylcholine at 0.5 mM Ca'+. This is a lower limit for the relative activity: phospholipase A, has an absolute requirement for Ca'+ and will have to compete with phosphatidylserine for Ca'+ ions. If the phospholipid-Ca'+ affinity were known, one could calculate the true activity of phospholipase A, toward phosphatidylserine in Tritonlphospholipid mixed micelles or other structures.
The results with synthetic short chain phospholipids indicate that the cobra venom phospholipase A, is not very active toward monomers, as is the case for phospholipases from most other sources (6,7). The enzyme greatly prefers micellar substrates, either Trit,on/phospholipid mixed micelles or pure phospholipid micelles (as for dioctanoyl phosphatidylcholine). Furthermore, at a mole ratio of 2:1, the Triton mixed micelles show comparable hydrolysis rates to pure dioctanoyl phosphatidylcholine micelles. Phospholipuse C Specificity-Purified phospholipase C from Bacillus cereus is known to hydrolyze phosphatidylcholine, phosphatidylethanolamine, and phosphatidylserine in natural membranes (39)(40)(41)(42). Studies on red cell ghosts indicat.e that phosphat.idylserine is hydrolyzed more slowly t.han the other lipids (39). As with phospholipase AL, this could be due to a greater availability of phosphat idylcholine and phosphatidylet.hanolamine in the membrane or to a preference of phospholipase C for these two lipids.
With the Tritonphospholipid mixed micelle system, the act.ivity of phospholipase C toward phosphat idylet hanolamine was only slightly less than t.oward phosphatidylcholine and its activit.y toward phosphatidylserine was only 2 to 3 times less than toward phosphatidylcholine.
Unfort,unat.ely, the fat.ty acid composition of the phosphatidylserine is not exactly the same as any of t.he phosphatidylcholines, so a precise evaluation of the specificity toward this phospholipid must be performed on synthetic phosphatidylserines which have low thermotropic phase transition temperat,ures; such phospholipids have not been available. The possibility of Zn'+ chelation by phosphatidylserine could also complicate the interpretation of phosphat.idylserine hydrolysis by phospholipase C in analogy to Ca'+ chelation with phospholipase Ai, alt.hough t.his is less likely with phospholipase C since the Zn" is quite tightly bound to this enzyme (24).
Compa.rison of Phospholipase A, and C Specificities in Mixed Micelle System-These two pure phospholipases show different head group specificities toward individual phospholipids in the well characterized Triton X-100 mixed micelle system. Phospholipase A, prefers phosphatidylcholine; phospholipase C will hydrolyze phosphatidylcholine and phosphatidylethanolamine at, similar rates. Cal+ does not affect phospholipase C activity so that the assay mixture containing mixed micelles and 10 mM Ca'+ was identical for evaluating the different. specificities toward phosphatidylcholine and phosphatidylethanolamine of the two enzymes. Recently, we (13) showed that 'H NMR detects differences in the a-methylene groups of the sn-1 and sn-2 fatty acid chains in phosphatidylcholine and phosphatidylserine.
Phosphatidylethanolamine shows a broadened pattern, indicat.ing a local conformation or environment different from the other two phospholipids. This is of particular interest because phospholipase A, catalyzes the hydrolysis of the fatty acid at the carbonyl carbon of the sn-2 chain. In the case of the two zwitterionic phospholipids, the conformational differences near the site of enzymatic attack may be important for phospholipase AL catalysis. Since this region is quite removed from the site of phospholipase C catalysis, it may not. be important, for t,hat enzyme.
Both enzymes display less activit,y toward phosphatidylserine than toward phosphatidylcholine, but evaluation of this apparent specificity is complicated as discussed above for each enzyme.
The Thus, further work is required to delineate the precise cause of the apparent speciticit,ies reported here.