Hydrolysis of Phosphatidylcholine in Phosphatidylcholine-Cholate ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1988 by The American Society for Biochemiatry and Molecular Biology, Inc.

Vol. 263,No. 24, Ieaue of August 25, pp. 1180&11813,1988 Printed in U.S.A.

Hydrolysis of Phosphatidylcholinein Phosphatidylcholine-Cholate Mixtures by Porcine Pancreatic Phospholipase AB* (Received for publication, September 14, 1987)

Nurit Gheriani-GruszkaS, Shlomo AlmogS, Rodney L. Biltoneng, and Dov LichtenbergSll From the $Departmentof Physiology and Pharmacology, Tel Aviv UniversitySchool of Medicine, Tel Aviv69978,Israel and the §Departments of Biochemistry and Pharmacology, University of Virginia School of Medicine, Charlottesville, Virginia22908

Pancreatic phospholipase A, (PLA,)-catalyzed hyPancreatic phospholipase A2 (PLAZ)’ is a soluble enzyme drolysis of egg yolk phosphatidylcholine (PC) in mixed which hydrolyzesphospholipids in the digestive system. Much PC-cholate systems depends upon composition, struc- effort has been devoted toward understanding the structure ture, andsize of the mixed aggregates. Thehydrolysis and function of this enzyme and other phospholipases using of PC-cholate-mixed micelles made of an equalnumber monomeric, micellar, mixed micellar, and lamellar (bilayer) of PC and cholate molecules is consistent with a K,,, of substrates (1, 2). These studies have clearly shown that, in about 1 mM and a turnover number of about 120 s-‘. general, the activity of the enzyme toward aggregated lipids Increasing the cholate/PC ratio in the micelles results is much higher than toward monomeric lipid substrate (2-4) in a decreased initial velocity. Hydrolysis of cholatecontaining unilamellar vesicles is very sensitive to theand that the character of hydrolysis is determined by the ratio of cholate to PC in the vesicles. The hydrolysisof surface properties of the aggregated lipid (4). vesicles with an effective cholate/PC ratio greater than Synthetic, saturated long-chain phospholipids in the form 0.27 is similar to that of the mixed micelles. The time of lipid vesicles are not hydrolyzedby PLAz unless their by addition of a second component such course of hydrolysis of vesicles with lower effective structure is perturbed ratios is similar tothat exhibited by pure dipalmitoyl- as alcohols and detergents (5),or alteredby an osmotic shock phosphatidylcholine (DPPC) large unilamellar vesicles (6,7), stiff curvature ( 5 ) , or if the substrate is within the gel in the thermotropic phase transition region. In the to liquid crystalline transition temperature range (6-11). Unlatter two cases, the rate of hydrolysis increases with der these conditions the kinetics of phospholipid hydrolysis time until substratedepletion becomes significant. The may be complex. In many cases the rateof hydrolysis increases reaction can be divided phenomenologically into two continuously with time in a manner consistent with an actiphases: a latency phase where the amount of product vation mechanism suggested to involve enzyme dimerization formed is a square function of time ( P ( t )= At2) and a on the bilayer surface (6, 7, 12). This initial phase is often phase distinguished by a sudden increase in activity. followed by a relatively abrupt increase in the rate of hydrolThe parameter A , which describes the activation rate ysis, which has been attributed to acceleration of the reaction of the enzyme during the initial phase in a quantitative by a critical mole fraction of reaction products (13-18). fashion, increases with increasing [PLA,], decreasing Much information is available on the hydrolysis of phos[PC], decreasing vesicle size, and increasing relative phatidylcholine (PC) contained in PC detergent-mixed micholate content of the vesicles. The effect of [PLA,] celles by several soluble phospholipases (19-21) including and [PC] on the hydrolysis reaction is similar to that pancreatic PLA2. A common characteristic of these hydrolysis found with pure DPPC unilamellar vesicles in their reactions is the absence of any latency phase. This also thermotropic phase transition region. The effect of appears to be the case for pancreatic PLA2 hydrolysis of PC cholate on the hydrolysis reactionis similar to thatof temperature variation within the phase transition of from natural sources in the presence of bile salts. The latter DPPC. These results are consistent with ourpreviously system, which is of special relevance to thedigestion of PC in proposed model, which postulates that activation of the gastrointestinal tract, has been investigated by Nalbone PLA, involves dimerization of the enzyme on the sub- and his collaborators (22) over a wide range of bile salt/PC strate surfaceand that the rate of activation is directly mole ratio. Theseauthors showed thatthe reaction rate proportional to themagnitude of lipid structural fluc- increases abruptly when the bile salt concentration is suffituations. It is suggested that large structural fluctua- cient to solubilize the PC as PC-bile salt-mixed micelles. tions, which exist inthe purelipid system in the phase Further increase of the bile salt concentration inhibits the transition range,are introduced intoliquid crystalline hydrolysis, a phenomenon previously observed with other vesicles by the presence of cholate and thus promote detergents such as TritonX-100 (19). activation of the enzyme. In the study by Nalbone et al. (22), the rate of hydrolysis

* This work was supported by a grant from the Israel Academy of Sciences and Humanities (to D. L.),Grant PCM 80-03645 from the National Science Foundation (to R. L. B.) and GrantGM-37658 from the National Institutes of Health (to R. L. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ll To whom correspondence should be addressed.

The abbreviations used are: PLAZ, phospholipase A, from bovine pancreas; PC, phosphatidylcholine (from egg yolk, unless stated 0thenvise); DPPC, dipalmitoylphosphatidylcholine;T,, gel to liquid crystalline phase transition temperature; Re, the effective ratio of cholate to PCin PC-cholate-mixed aggregates (micelles and vesicles); Rh, hydrodynamic radius of vesicles or micelles; T , the lag time (latency) prior to occurrence of the maximal rate of hydrolysis. It is defined as thetime at which the line describing the initial hydrolysis intersects with the line describing the maximal-rate hydrolysis; P,, the fraction of substrate hydrolyzed at time T ; V,, the maximal velocity observed in the time course of any given hydrolysis reaction (in some cases V , = initial velocity).

11808

11809

Phosphatidylcholine Hydrolysis byPLA2

was described in termsof the totalPC hydrolyzed in the first and methanol. Evaporation of the organic solvents followed by dis5 min of t h e reaction. As long as the cholate content in t h e persion of the residue in our standard saltsolution yielded transparent solutions with Re values of 0.85-1.00 (27, 28). These solutions were of t h e PC then further diluted to the desired concentration with our standard dispersion is insufficient for complete solubilization and if the rate of hydrolysis increases with time, the infor- salt solution containing the appropriate amount of cholate. mation obtainedon the hydrolysis duringthe first few minutes The hydrolysis of mixed micelles wasfound to be a function of the is of limited significance. Furthermore, the state of aggrega- composition and micelle size. The effect of these two interrelated tion of t h e substrate was not well characterized, and the factors was studied at a constant PC concentration of 7 mM while details of the dependence of hydrolysis on the actualstate of varying the cholate concentration. Studies of the dependence of hydrolysis on the PC concentration required that the effective ratio lipid aggregation could not be explicitly addressed. Hoffman and the micelle size be maintained constant while varying the PC et al. (23) measured the hydrolysis of PC in better character- concentration upon which the micelle composition and size depend. ized PC-bile salt-mixed micellar solutions(24) and concluded This was done by diluting varying volumes of a mixed micellar stock that the rate of hydrolysis is governed toa large extent by the solution containing 100 mM PC and 105 mM cholate with a solution containing 5.5 mM cholate, 10 mM CaC12,and 135 mM NaCl to a final effectiveratio of cholate to PC i n the mixedaggregates. However, these studies focused only on mixed micelles and volume of 10 ml. This procedure yields solutions of varying [PC] at a constant Re value of 1.00 as described by Equation 2. provided no information on bile salt-containing lamellar PC PC-cholate-mixed vesicles were prepared by two successive dilusystems (25). Our previous studies on the structure of PC- tions of a stock mixed micellar solution with solutions containing cholate-mixed micelles and vesicles (26-28) provide the basis varying cholate concentrations. Vesicles of varying size and constant with which we can attempt to relate the activity of PLAz t o composition were made by first diluting a PC-cholate-mixed micellar the entire rangeof structures of the mixed aggregates.In these system with our standard salt solution to obtain mixtures of various previous studies we characterized the size and composition of concentrations and thus varying Re values. The resultant vesicular dispersions each contained vesicles of a specific size determined by cholate-containing vesicles present in various Ca2+-containing Re. These dispersions were then brought to a constant composition mixtures of egg PC and cholate. Here we present the results by a second step of dilution with the appropriate volume of our of a study of the hydrolysis of PC in such aggregates and standard salt solution. Aswe have previously shown (27, 281, the relate the details of the reaction parameters to the substrate second dilution does not induce any change in vesicle size. Consestructure and composition. These results indicate that the quently, the outlined procedure resulted in vesicles which differed only in size, while having the same composition. mechanism of activation of PLAz by cholate-PC vesicles is To vary Re or the lipid concentration one at a time, while keeping similar to activation bypure PC vesicles in their phasetransthe vesicle size colstant, we first prepared a concentrated dispersion ition region. of large vesicles (Rh = 45 nm) with Re = 0.27. To obtain vesicles of different cholate content, we further diluted aliquots of this “vesicle EXPERIMENTALPROCEDURES stock dispersion” with equal volumes of Ca2+-containingsaline soluMaterials-Egg yolk phosphatidylcholine (PC) was purified chro- tions of varying cholate concentration. This procedure yielded vesicles matographically according to Singleton et al. (29). Its purity was of constant size a t a constant PC concentration but of varying Re. confirmed by thin layer chromatography and its concentrationmeas- The concentration of cholate in the diluting media was chosen so ured according to Stewart (30). The fattyacid content was similar to that the final value of R,, as given by Equation 1, was within the that found by Hertz and Barenholz (31). Cholate (Sigma) was crys- range of0.18 to 0.27. To vary [PC] we diluted the vesicle stock tallized from ethanol. Porcine pancreatic PLA2, Sigma (P-9139), was dispersion with different volumes of our standard salt solution of first diluted 50-fold with a standard salt solution of CaC12 (10 mM) varying cholate concentration chosen according to Equation 1to yield and NaCl (135 mM). This diluted solution, containing 0.2mgof dispersions of different PC concentrations having a common Re. proteinlml, was then dialyzed four times against 200-400-foldexcess a In all experimental procedures that involved dilution-induced of the same medium for approximately 15 h. The pH of the dialyzed changes in Re, the diluted preparations were used only after waiting solution was adjusted to 8.0 and theenzyme concentration determined at least 30 min following the initial dilution step. by light absorption at 280 nm (t = 1.24 OD units/l g/liter (32)).This Characterization of PC-Cholate-mixed Micelles and Vesicles-The solution was stored at 0-4 “C for no longer than 14 days. Just before structure of mixed micelles and vesicles depends in a well-defined use, the pH was readjusted and the enzyme concentration redeterfashion on their composition (28). The dispersions used in the present mined. studies were characterized as described in our previous work (28) Hydrolysis Studies-In all hydrolysis studies the lipid dispersions using the following procedures. (i) The composition of the vesicles (10 ml)were first adjusted to pH8.0 in a pH-stat vessel a t 38 & 1“C. was determined by chemical analysis of PC (30) and cholate (33) 0.05-0.9 ml of a solution containing 0.2-0.3 mg of PLA2/ml was then before and after centrifugation for 8 h at 2 X 10‘ g. (ii) The vesicle’s injected into the stirred lipid dispersion and the extentof hydrolysis mean size were determined by quasielastic light scattering using a recorded continuously with a pH-stat (Radiometer) using 0.01 M Nicomp model HN-5/90 spectrometer equipped with a computer NaOH asthetitrant.In some experiments a probe colorimeter autocorrelator, model6894. These vesicles were also examined by (Brinkmann) was also immersed in the dispersion to follow simulta- electron microscopy as previously described (27). The molar turbidity neous changes in turbidity at 450 or 570 nm. of various vesicle dispersions was found to depend linearly on the Preparation of Cholate-PC-mixed Micelles and Vesicles-We have mean hydrodynamic radius as determined by quasielastic light scatshown previously (31,32) that the state of aggregation of the lipid in tering. Molar absorbance of each dispersion was therefore measured cholate-PC mixtures is a function of the cholate/PC ratio in the routinely immediately before its use for studies of enzymatic hydrolmixed aggregates (Re).At Re values less than 0.30, the lipid exists in ysis. vesicular form and

R, = [cholate]/([PC] + 1/K)

(1)

where K is the partition coefficient of cholate between the vesicles and the aqueous medium. If Re > 0.40, the system is mixed micellar, and theeffective ratio is given by

Re = ([cholate] - CMC)/[PC]. (2) In 10 mM ca2+, 1/K= 16 mM and the critical micellar concentration (CMC) = 5.5 mM (28). In the range of Re values between 0.30 and 0.40 vesicles and micelles coexist. All mixed micelle and vesicle dispersions were made from mixedmicellar stock solutions containing 100 mM PC and 90.5-105.5 mM cholate. These stock solutions were prepared from solutions of the appropriate amount of cholate and PC in a mixture of chloroform

RESULTS P L A 2 Hydrolysis of Micelles-Hydrolysis of PC in mixed micelles is observed immediately upon addition of PLAz to the substratesolution. The initial rate of hydrolysis is a monotonicallyincreasingfunction of [PLAz] and [PC] as shown in Fig.1,A and B. The results in Fig. 1B are consistent with a K,,, of about 1 mM and a turnover number of about 120 s-’. The initial rate of hydrolysis appears t o be a decreasing function of the effective ratio of cholate to PC. For example the initial rateobtained at Re = 2.50 is onlya third of the rate obtained at Re values in the range of 0.60 t o 1.00 (Fig. IC). PLA2 Hydrolysis of Vesicular Substrates-The rate of PLA,-

11810

Phosphatidylcholine Hydrolysis by PLAz

20

40

60

80

Time (min) 5 IO 15 PLAz] (pg/ml)

5 IO [PC] (mM)

2.0

1.0

R.

FIG. 1. The initial rate of hydrolysis of PC contained in 10 ml of PC-cholate-mixed micelles as a function of [PLA2] (paneZA), [PC] (panel B ) , and Re(panel C). [PLAp] in B and C was 3 pg/ml, [PC] in A and C was 7.0 mM and Re in A and B was 1.00. The initial rates of hydrolysis are described in terms of pmol of PC hydrolyzed/min. 60

1

40

20

5

IO

15

20

25

Time (min)

FIG. 2. The time courses of the hydrolysis of vesicular P C (4 mM) by PLA2 (20 pg/ml in curues A and C; 7.6 I,rg/ml in

cumes B and D ) . Re = 0.18 in the experiments described by the solid curves C and D, and Re = 0.27 in those described by broken curves A and B.

catalyzed hydrolysis of pure egg phosphatidylcholine vesicles in the liquid crystalline state is extremely slow and appears to be independent of time. However, vesiclesof PC-containing cholate induce apparent activation of the enzyme as shown in Fig. 2. At Re = 0.27 instantaneous hydrolysis is observed at a rate proportional to theenzyme concentration (curves A and B ) . At Re =; 0.18, the time course of the reaction is complex. The rateof hydrolysis increases monotonically with time until a maximum rate is achieved (curves C and D). Thereafter the rate decreases until the reaction is complete. Increasing the enzyme concentration results inan increase in the apparent rate of activation and an earlier onset of rapid hydrolysis (compare curves C and D). This type of behavior is similar to that observed for the hydrolysis of pure dipalmitoylphosphatidylcholine (DPPC) vesicles in theirgel-liquid crystalline phase transition region (6-8). The results in Fig. 2 demonstrate that the rate of activation, using the time to achieve the maximal rate of hydrolysis as an index, is a function of both enzyme concentration and Re. The possibility exists that during the time course of the hydrolysis reaction the cholate-containing vesicles transform into micelles. To address this possibility the optical density of the solution and the amount of hydrolysis were simultaneously monitored as shown in Fig. 3. During the early portion of the reaction, the amount of product formed and theturbidity increase in a parallel fashion. Upon onset of rapid hydrolysis, the turbidity began to decrease followedby a large increase in turbidity? The large decrease in turbidity follow-

’

This large increase of turbidity is probably due to precipitation of calcium salts of free fatty acids. Addition of CaC12 (10 mM) to a mixture of PC, lyso PC, oleic acid (2 mM each), and cholate (3.6 mM)

FIG. 3. The time course of the hydrolysis of vesicular P C (4 mM) by PLA2 (20 pg/ml). These large vesicles ( R h = 45 nm) contained about 15 mol % cholate (Re= 0.18).The cumulative percent hydrolysis is given by the broken line A (scale on the left). The accompanyingchanges in the turbidity of the dispersion, as measured simultaneously by the probe colorimeter, is given by solid line B (scale on the right).

ing the onset of rapid hydrolysis is very likely the result of micelle formation. The initial phase of the reaction accompanied by increased turbidity thus appears to be inconsistent with the possibility that the marked increase in the rate of hydrolysis is due to micelle formation. The ability to vary vesicle size and Re independently (27, 28) allowed us to investigate the dependence of PLAz activation by lamellar substrates on a number of new experimental variables. The following parameters were selected to monitor the reaction: V,, the maximal velocity of the hydrolytic reaction; 7 , the length of the latency phase defined as thatpoint in time at which a straight line describing the rate of initial hydrolysis and aline describing the maximal rate of hydrolysis intersect; and P,, the fraction of substrate hydrolyzed at time T. V,,, is an obvious parameter describing enzymatic activity. The latency T has been introduced previously to describe the activation and its dependence on enzyme and substrate concentration and appears to be inversely proportional to the actual rate of initial activation (6-8). P , was selected as a parameter of interest because of previous suggestions that a “critical fraction” of product is required for activation (1318). The variation of T,P,, and V , with [PLAz]is shown in Fig. 4 at Re values of 0.18 and 0.27. V , appears to reach a limiting value at both Re values, although the limitingvalue of V, and the [PLA2] at which it is reached are different in the two cases. P, appears to be a decreasing function of [PLAZ] in both cases. T decreases with [PLA2], a result which is qualitatively consistent with either activation being related to the formation of a critical fraction of product or an activation process involving enzyme-enzymeinteraction on the substrate surface (8) or both. According to either of these possibilities, T should be an increasing function of [PC]. Unfortunately, the range of [PC] which could be used in a studyof the present system is limited. Therefore, ratherthan vary [PC]ina systematic fashion, six time course experiments at two PC concentrations were carried out to answer this question. In these experiments [PLAz] = 7.5 mg/ml, Re = 0.18, and the mean hydrodynamic radius of the vesicles was 45 nm. For [PC] = 2.0 mM, the measured latency varied from 0.6 to 2.5 made by a 25-fold dilution of a mixed micellar system resulted in a similar turbidity increase due to calcium oleate precipitation (Lichtenberg, D., Werker, E., Bor, A., Almog, S., and Nir, S. (1988) C h m . Phys. Lipids,in press).The kinetic details of this precipitation reaction were complex, and under variousconditions it occurred only after a latency phase which was extremely sensitive to the concentration of the free fatty acid. The increase in turbidity observed well after hydrolysis approaches its maximal rate is probably a reflection of the complex kinetics of this precipitation reaction.

Phosphatidylcholine by Hydrolysis

PLAz

11811

continuous rather than a stepwise function of vesicle size. Thus, the conclusion that gel state vesicles below a critical size can activate butlarger vesicles cannot activatePLA2 (12) is incorrect. The most intriguing result of thecurrent study is the dependence of PLA, activation on the presence of cholate in the vesicles. Fig.6A describes the latency as afunction of the effective ratio at two different [PLA2]. It is clear from this figure that at both PLA, concentrations studied, the length of the latency phase is a decreasing function of the cholate content. This acceleration of the activation process is also accompanied by a decrease in the fraction of PC hydrolyzed prior to the onset of rapid hydrolysis (Fig. 6B). At relatively low [PLA,], the maximal rate of hydrolysis observed after the latency phase is an increasing function of the cholate/PC ratio in the vesicles (Fig. 6C, solid symbols) while at higher enzyme concentration, the rate is independent of the cholate content of the vesicles (Fig. 6C, open symbols). The early time data were also analyzed in terms of the function P ( t ) = At'. The product formed at any time t ( P ( t ) was ) found to fit such a t 2 dependence up to 20-50% hydrolysis. The A parameter thus derived was a linear function of [PLA,]', as previously found for the activation of the enzyme by pure DPPC large unilamellar vesicles (8). In Fig. 7, the parameter A is shown as a function of Re. The rather abrupt increase in A at Re = 0.24 is reminiscent of the dramatic change in the latency observed with DPPC large unilamellar vesicles as the gelliquid crystalline transition temperature is approached.

4L

0

C

l

0

0

,

I

20

IO

[PLA,]

.

I

,

30

L 40

(pg/ml)

FIG.4. The dependence of the hydrolysis of vesicular PC (4.0 mM) on [PLA,]. The hydrolysis is characterized by the latency (panel A ) , the product at time T , P, (panel B ) , and the maximal rate ( V , in pmollmin, panel C). In all the experiments described in this figure, large unilamellar vesicles (E,,= 45 nm) containing cholate (Re= 0.18, solid symbols or Re = 0.27, open symbols) were used. T

-

4t 0

20

40

FIG.5. The dependence of the hydrolysis of vesicular PC (3.0mM, R. = 0.16) by PLAa (20 pg/ml) on the vesicle size( R h in nm, as determined in our previous work (27,28)). Panel A describes the size dependence of the latency; panel B describes the size dependence of the product formed at time T .

DISCUSSION

Pancreatic phospholipase AP is a water-soluble enzyme responsible for the hydrolysis of water-insoluble phospholipids that exist as micelles, bilesalt-containing lipid bilayers, or triglyceride-containing emulsion particles in the gastrointestinal tract. The present study focuses on the activation of PLA2 in mixedmicelles and vesicles containing egg yolk lecithin and cholate. The results presented are consistentwith previous studies on the hydrolysis of mixed micellar PC. New information on the hydrolysis of cholate-containing PCvesicles which contribute to our understanding of the activation process of PLA2 on substrate surfaces is also provided.

min. For [PC] = 4.0 mM the measured latency varied from 2.7 to 6.0 min. Similar differences in 7 were observed at higher enzyme concentrations and higher Re. From these results itis clear that thelatency increases with [PC]. It should be noted that no dependence of either V , or P, on [PC] was observed in the experiments described above. The investigation of the dependence of the hydrolysis reaction on vesicle composition and size has provided new information regarding the PLA, activation process or processes. We have found that if the reaction was carried out with vesicles of constant composition but varying size at constant enzyme and substrate concentrations, the latency increases monotonically with vesicle size as shown in Fig. 5A. P , also appears to increase slightly with increasing vesicle size (Fig. 0.18 0.21 0.24 027 5 B ) while the maximal rate of hydrolysis remains unchanged Re (turnover number = 35 rf: 4 s-'). These results are consistent FIG.6. The dependence of the hydrolysis of vesicular PC with the observation that gel state small unilamellar vesicles R h = 4 5 nm) by PLA2 (7.5 pg/ml, solid symbols; 2 0 (7.0 (E,, = 11 nm) made of DPPC activate PLA2 very rapidly rg/ml,mM, open symbols) on the effective cholate to PC ratio (Re). whereas large unilamellar vesicles ( R h = 35 nm) in the gel The hydrolysis reaction is characterized in terms of the latency T state do not activate PLAz within any reasonable time. These (panel A ) , the product a t time T (panel B ) , and the maximal rate of new results thus suggest thattherate of activation is a hydrolysis (panel C).

11812

Phosphatidylcholine Hydrolysis byPLAz

At2 until substrate depletion begins to dominate. The very abrupt increase of enzymatic activity following the latency phase has not been investigated in detail, although its onset time described interms of T is highly, if not completely, XI correlated to theA parameter of the above model proposed to A I describe initial activation. The parameter A is directly proI I portional to [PLAZ]’ and a decreasing function of [PC]. As we have argued previously, such results indicate that enzyme dimerization on the bilayer surface of DPPC large unilamellar vesicles is an essential part of activation (8).Mechanisms in which the enzyme is activated as amonomer on the substrate surface or dimerizes in the solution are inconsistent with the experimental results. With pure DPPC vesicles 7 can change by almost two orders 018 0.22 0.26 of magnitude within a temperature range of 1°C with the maximal rate of activation being achieved very near or at the Re (6,7). This gel-liquid crystalline phase transition temperature FIG. 7. Dependence of A on Re.The coefficient A (%/s2) de- result suggested that the rate of initial activation is directly scribes the variation in product concentration (in % of PC) with the proportional to the magnitude of dynamic structural fluctuasquare of time ( P ( t )= At’. The results shown are from the experiments describedin Fig. 6: [PLAZ] = 7.5 pg/ml (solid symbols) and tions of the membrane substrate.With PC-cholate-mixed of activation, [PLAz] = 20 pg/ml (X-X-X). The values of A were obtained from vesicle systems, we have now found that the rate least squares estimates of the slope of P Versus t 2 during the initial as reflected in the A parameter (Fig. 7), is a function of the phase of the hydrolysis reaction ( P ( t )< 20%). cholate/PC ratio and undergoes a dramatic increase at Re = 0.24. We suggest that structural fluctuations similar to those The maximal rate of PLA2-catalyzed hydrolysis is instan- occurring in the gel-liquid crystalline transition region of pure taneously achieved when the effective cholate/lipid ratio in DPPC vesicles exist in themixed PC-cholate vesicles and are the mixed system is higher than 0.27, regardless of whether dominantin the PLA2 activation process. Inthis respect the substrate is in micellar or vesicular form. With both types variation of Re in the PC-cholate-mixed vesicular system can of substrates, the hydrolysis exhibits the general features of be regarded as being analogous to temperature variation in Michaelis-Menten kinetics. The initial velocity obtained with pure vesicular systems. Our proposal implies that the primary effect of cholate, as mixed micellar substrates at constant [ E ]and [SI is a decreasing function of Re. Similar behavior has been observed for the it is manifested in the activation of PLAp, is on the dynamic hydrolysis of PC contained in mixed micellesmade with other structure of the aggregated substrate. Activation of the endetergents (19-21). This decrease in the initial velocity has zyme by monomeric cholate is possible, but we think this is been attributed to substrate dilution on the micellar surface unlikely because such an explanation makes it difficult to and interpreted in terms of dual binding of PC to the “acti- rationalize the very pronounced decrease of the latency phase vator” and “catalytic” sites of the enzyme (2, 34). However, (increase in the parameter A ) observed upon increasing the this decrease in velocity would also be expected if the rate- cholate/PC effective ratio from 0.22 to 0.24 (Figs. 6 and 7) as limiting step changed from surface diffusion of PC at low Re being the result of a change in the concentration of monomeric to release of enzyme from the surface of the smaller micelles cholate from 3.52 to 3.84 mM. It is also possible that memwhich exist at large Re. It could also be a result of direct brane-associated cholate promotes activation by increasing inhibition of the enzyme by cholate. More studies are needed the binding ofPLA’ to the membrane. This appears to be unlikely because the apparent rate of activation decreases to clarify the situation. The behavior of cholate-PC aggregates of Re > 0.27 are with increasing PC concentration. Furthermore, such an exsimilar to DPPC-small unilamellar vesicles in thegel state in planation would also have to rationalize the abrupt change in the sense that maximal activity is instantaneously observed. the activation parameter at Re = 0.24 and its strong dependWe suggest that in all cases where maximal activity is ob- ence on the hydrodynamic radius of the vesicle. The cause of the rapid increase in activity following the served at zero time activation of the enzyme also occurs but is simply too fastto be detected. This statementis based upon initial phase of activation is not yet clear. Jain andco-workers the facts that the rate of initial activation of PC by vesicles (13-18) have proposed that this dramatic increase in hydroof constant size increases continuously as Re increases and lytic activity is the result of a critical fraction of product being might be related in some manner to that the rate of activation by vesiclesof constant composition formed and that it increases continuously as vesiclesize decreases. In neither enhanced binding. This latter suggestion is supported by the case does there appear to be a discontinuous change from findings that negatively charged lipids and high concentravesicles which slowly activate the enzyme to vesicles which tions of product increase enzyme binding (21, 34). While enzyme binding to the lipid surface is obviously necessary, it are instantaneously hydrolyzed. Unilamellar vesicles formed at cholate to lipid effective is not sufficient for activation. For example PLA2 binds to ratios below 0.27 generally exhibit a lag or activation period DPPC large unilamellar vesicles in the gel state,but no prior to achievement of maximal activity. This behavior is hydrolysis is observed (6). Vesicle fusion of partially hydrosimilar to thatobserved with DPPC large unilamellar vesicles lyzed, negatively charged vesicles could possibly contribute to in their gel-liquid crystalline phase transitionregion (6). With the apparent “erruption” in activity. Previous studies of the both substrate systems, product formation at early times was fusion of PC-PS vesicles show that these negatively charged vesicles fuse when the mole fraction of the negatively charged found to obey the relation P ( t ) = At’. This representation applies quantitatively only to theearly phase of the reaction- phosphatidylserine exceeds a critical value (35). It is possible time profile. In some cases this may only be 20% hydrolysis that upon formation of a critical fraction of product fusion is although in other cases the time course can be described by enhanced in cholate-PCvesicles. The small, but reproducible, 3.0

Phosphatidylch.oline Hydrolysis byPLAz increase in turbidity observed just prior to the erruption of hydrolytic activity (Fig. 3) may be an indication of fusion occurring. However, the existence of such phenomenon does not challenge the validity of the previously proposed enzyme activation model which applies to the initial stage of hydrolysis. The actual role of cooperative structural fluctuationsin the activation of PLA, is not yet clear. Such fluctuations could be required for "penetration" of the enzyme into the bilayer, a phenomenon suggested to be important by Verger and de Hass (4). They could also play a role in the conformational coupling between the enzyme's "activation site" and catalytic site as proposed by Dennis and co-workers (2,19-21, 34) in their dual-site activationmodel. Regardless of the mechanism of activation such fluctuations appear to play a central role in thePLA2 activation process. We suggest that the physical process responsible for activation inthe mixed vesicular systems is a critical phenomenon similar to percolation (36). Percolation is, in its simplest terms, adynamic equilibrium between small clusters andvery large clusters of one of the lipid components. At concentrations below a critical value (percolation point X J , the minor component exists primarily in small clusters whereas at concentrations above X , it exists in very large clusters. At and near X,cooperative fluctuations occur between these clustersize extremes. It is these cooperative fluctuations which we think mediate activation of the enzyme. X , is determined largely by the coordination number of the two-dimensional lattice (34) and the possibility of strong complexes being formed by the two components. For apurely random mixture (no interaction) in a hexagonal lattice X c = 0.5 whereas if a strong complex is formed X , < 0.5. For example, in mixed cholesterol-DPPC bilayers the lipid/cholesterol ratio of the complex is 2:l andX , = 0.25, a value consistent with observed dramatic changes in the lateral diffusion constant (37). Analysis of differential scanning calorimetry data is also in accord with this value of X , (38). Preliminary experiments indicate that at temperatures outside the phase transition region activation of PLA, in cholesterol-DPPC and cholesterol-dimyristoylphosphatidylcholine vesicles is enhanced compared to pure PC vesicle^.^ Taken together, the latterresults andthose of the present study suggest that critical phenomena similar to percolation can play a majorrole in PLA, activation, although much work is still needed to substantiate this hypothesis. In conclusion the results we have presented on PLA2 hydrolysis of PC-cholate-mixed unilamellar vesicles are consistent with a previously proposed model for activation of PLA, by pure DPPC large unilamellar vesicles in or near its phase transition. The activation of the enzyme probably involves dimerization and requires large structural fluctuations of the substrate. The effective cholate/lipid ratio appears to be a variable analogous to temperature. At a critical ratio, the rate of initial activation increases dramatically in a mannersimilar A. Yotsukura and R. L. Biltonen, unpublished results.

11813

to that observed in pure lipid systems as the transition temperature is approached. We suggest that a percolation phenomenon in the mixed system is the basis for this activation of PLA,. Such events may also be important in other membrane-mediated biological processes. REFERENCES 1. Verheij, H. M., Slotboom, A. J., and de Hass, G. H. (1982) Reu. Physiol. Biochem. Phurmarol. 91,91-106 2. Dennis, E. A. (1983) in The Enzymes, X V i (Boyer, P. D., ed)pp. 308-357, Academic Press, New York 3. Pieterson, W. A,, Vidal, J. C. C., Volwerk, J. J., and de Haas, G. H. (1974) 1455-1460 Biochemistrv 13. ~, - ~ ~ - 4. Verger, R . , and de Haas, G. H. (1976) Annu. Rev. Biophys. Bioeng. 6,77~~~

117

5. Ja&;M.C., Leeson, J., Upreti, A., and Upreti, G. C. (1978) Biochim. Biophys. Acta 6 0 9 , l - 8 6. Lichtenberg, D., Romero, G., Menashe, M., and Biltonen, R. L. (1986) J. Biol. Chem. 261,5334-5340 7. Romero, G., Thompson, K., and Biltonen, R. L. (1987) J. Biol. Chem. 2 6 2 , 13476-13482 8. Menashe, M., Romero, G., Biltonen, R. L., and Lichtenberg, D. (1986) J. Biol. Chem. 261.5328-533.7 9. Op-den $amp, J. A. F.,-Kaue&, M. T., and van Deenen, L. L. M. (1975) Biochrm. Bwphys. Acta 406,169-177 10. Op den Kamp, J. A. F., de Gier, J., and van Deenen, L. L. M. (1974) Biochim. Bwphys. Acta 346,253-256 11. Goormghtigh, E., van Campehoud, M., and Ruysschaert, J. M. (1981) Biochem. Biophys. Res. Commun. 401,1400-1418 12. Lichtenberg, D., Menashe, M., and Biltonen, R. L. (1985) Colloids and Surfaces 14,293-301 13. Jain, M. K., Egmond, M. R., Verheij, H. M., Apitz-Castro, R., Dijkman, R., and de Haas, G. H. (1982) Biochim. Biophys. Acta 688,341-348 14. Jain, M. K., and de Haas, G. H. (1983) Biochirn. Biophys. Acta 7 3 6 , 157162 15. Jain, M. K., and Vahagidar, D. V. (1985) Biochim. Biophys.Acta 814,313318 16. Jain, M. K., Rogers, J., Jahagirdar, D. V., Marecek, J. F., and Ramirez, F. (1986) Biochrrn. Bwphys. Acta 860,435-447 17. Jain, M. K., Rogers, J., Marecek, J. F., Ramirez, F., and Eibl, H.(1986) Biochim. Bio hys Acta 860,462-474 18. Jain, M. K., de fiaas, G. H., Marecek, J. F., and Ramirez, F. (1986) Biochim. Biophys. Acta 860,475-483 19. Deems, R. A., Eaton, B. R., and Dennis, E. A. (1975) J. Biol. Chem. 2 6 0 , sn1 x-anm 20. Roberts, M. F. Deems, R.A., and Dennis, E. A. (1977) Proc. Natl. Acad. Sci. LI. S. A.'74,1950-1954 21. Pluckthun, A., and Dennis, E. A. (1985) J. Biol. Chem. 2 6 0 , 11099-11106 22. Nalhone, G., Lairon, D., Charbonnier-Augeire, M., Vigne, J. L., Leonardi, J., Chahert, C., Hauton, J., and Verger, R. (1980) Biochim. Biophys. Acta 620. fil2-625 ~, 23. Hoffman, W. J., Vahey, M., and Hajdu, J. (1983) Arch. Biochem. Biophys. 221,361-370 24. Shankland, W.(1970) Chem. Phys. Lipids 4 , 109-130 25. Stark, R. E., Gosselin, G. J., and Roberts, M. F. (1985) Surfactants in Solutions, 5th InternationalSymposium (Bothorel, P., and Mittal,K. L., e&) pp. 286-298, Plenum Press, New York 26. Lichtenberg, D. (1985) Biochim. Biophys. Acta 821,470-478 27. Almog, S., Kushnir, T., Nir, S., and Lichtenberg, D. (1986) Biochemistry 26,2597-2605 28. Almog, S., and Lichtenberg, D.(1988) Biochemistry 27,873-880 29. Singleton, W. S., Gray, M. S., Brown, M. L., and White, J. L. (1965) J , Am. Oil Chem. SOC. 42.53-56 30. Stewart, J. C. (1980) Anal. Biochem. 104, 10-14 31. Hertz, R., and Barenholz, Y. (1975) Chem. Phys. Lipids 16,138-156 32. Nievwenhuizeh, W.,Kunze, H., and de Haas, G. H. (1974) Methods Enzymol. 30, 147-154 33. Turnberg, L. A,, and Anthony-Mote, A. (1969) Clin. Chim. Acta 2 4 , 253259, 34. Dennis, E. A., and Pluckthun, A. (1986) in Enzymes ofLipid Metabolism I1 (Freysz, L., Dreyjus, H., Massarelli, R., and Gatt, S., eds) pp. 121-132, Plenum Publishing Corporation, New York 35. Nir, S., Bentz, J., Wilschut, J., and Duzgunes, N. (1983) Prog. Surface Sci. "" " "

"

_"

"~

"_

2 . -1-124 _, 36. Ziman, J. M. (1979) in Models of Disorder, pp. 370-379, Cambridge University Press, Cambridge 37. Rubenstein, J. L. R., Smith, B. A,, and McConnell, H. M. (1979) Proc. Natl. Acad. Sci. U. S. A. 7 6 , 15-18 38. Snyder, B., and Friere, E. (1980) Proc. Natl. A c d . Sci. U. S. A. 77.40554059