Kinetics of hydrolysis of dispersions of saturated phosphatidylcholinesby Crofalusatrox phospholipase A,

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

DAVID0. TINKER,A. DAVIDPUWDON, JANEWEI, A N D EILEEN MASON Dc2pcirtmernt ofBiochcr~rl'stry,Unir-ersity c,fToronto, Toronto, OM., Cailucia M5S I A 8 Received January 18, 1978 This pa6rc.r is cdedicubed to the memory oftlie labe Dr. 6;. Malcolm Bro~un Tinker, D. O., W~rdon,A. D., Wei, J. & Mason, E. (1978) Kinetics of hydrolysis of dispersions of saturated phosphatidylcholiraes by Crotalus atrox phospholipase A,. Can. J . Bicachcm. 56, 552-558 Dispersions of lan-aellarphase dipallmitoyl phosphatidylcholine (DDPC) and dirnyristoyl phosphatidylcholine (DMPC) in 0.01 M CaCl, were subjected to hydrolysis by phospholipase A2 (EC 3.1.1.4) from Croraius afrox venom. The reaction was followed continuo~aslyby titrating the released fatty acids. For hydrolysis of gel phase phosphatides, the steady-state initial velocities were hyperbot ic functions of lpi~lklipid concentrations. At the "re-transition' temperature (34°C for DPPC, 15°C for DMPC), there was a large increase in the Michaelis parameter V,,, but no change in the parameter K,. A model was devised to account for these observations, in which the enzyme desorks from the lipid sunface after hydrolysis. 'Phe desorption rate constant is postulated to increase abc~vethe pretransition temperature. For hydrolysis of liquid crystalline phosphatides, the reaction consisted of a short initial burst crf hydrolysis, a Bong 'lag' period of very slow reaction, followed by a dramatic increase in the reaction rate. Addition sf 10 msl% lysolecithin or fatty acid abolished the 'lag' period. It was postulated that the enzyme adsorbs irreversibly to the srarface of the liquid crystalline phase. Reaction products are postulated to stimulate desorption of enzyme from the su~face.Thus, temperature-dependent changes in the rate of hyclrolysis of dispersed phosphatidylcholines are attributed to changes in the rate ofciesorgtion of the enzyme from the lipid surface.

affected by the physical properties of the lipid-water interface upon which catalysis occurs (4, 7, 15, 16). Unfortunately, the natalre of this effect is not always easy to understand. Verger and co-workers (17-20) have studied the hydrolysis of phosph~atidemonolayers, in which the interfacial properties can be controlled, and have devised a kinetic scheme which incorporates these properties. Deasnis (21, 22) and Deems et ul. (23) utilized a similar scheme to describe hydrolysis of micellar substrates. Relatively few studies have been carried out on the phospholipase A, catalyzed hydrolysis of aqueous dispersions of long-chain phosphatid ylcholines, which form bimolecular Hamellae :as opposed to globular rnicelles or monolayers. Op den Kr~nmpet al. (24, 25) studied hydrolysis of saturated phosphatidylcholines by pancreatic phospholipase A2 and found a remarkable temperature dependence of the reaction. The velocity of hydrolysis exhibited :a sharp maximum near the gel-liquid transition temperature of the Iipid (videitflra), and it was suggested that the enzyme can only penetrate (absorb to) the interface when the two phases coexist. While the extension of monolayer kinetics to such disperse systems has been suggested (171, formulation of a kinetic model incorporating this temperature effect has not been achieved. We chose to study hydrolysis of the saturated lipids, ABBREVIATIONS: DMPCI dimyristoyl phosphatidylcholine: DPPC, dipalmitoyl phosphatidylchoiine; DSPC, distearoyl DMPC, and DPPC. The phase diagrams for these lipids phosphatidyicho[ine; phosphatidylcholine, 1,2-di-O-acy]-sn- have been described (26-29). In excess water, these glyce~~~-)-~-phosph~~ry~chOl~ne; palmitoyl lysolecithin, 1-0- lipids may exist in one of three phases; two of these are designated as 'gel' phases, in which the hydrocarbon hex;tdecanoyl-sn-glycerol-3-0-phosphorylc holine. chains form a rigid, semicrystalline array, while the third l s a a ~ ~ o r t ebv d grant MT 2378 from the Medical Research is the liquid crystalline phase, In which the chains are ~ o u n coi f ~ a n a b a '

Introduction Phospholipase A2 (EC 3.1.1.4) of Crotabus utrox venom catalyzes the hydrolysis of the 2-ester bond of phosphatidylcholine and related phospholipids (1). The enzyme has been purified to homogeneity and characterized by physico-chemical methods (2, 3). The protein exists at neutral pH as a dimer of two apparently identical polypeptide chains; the molecular weight of the dimer is 29 000. The enzyme requires calciilm ion for its activity. While the mechanism of catalysis in the C . alrox enzyme has not been extensively studied, Well and coworkers (4- 12) have thoroughly characte~-izedthe catalytic mechanism of phospholipase '4, from the venom of the closely related snake, 6. adc~mrinteus(13). This enzyme has many features in common with the 6. atr0.v enzyme (2, 3) and it is reasonable to suppose that both have similar catalytic mechanisms. It has been established that only the dimeric form of the C. adurnatateus enzyme is catalytically active (4, 141, and that calcium ion and substrate add to the enzyme in an ordered bi-ter mechanism (4). Participation of a Iysine E-aminogroup in the catalysis has been suggested (5,6, $). It has become evident that the activity of phospholipase A, and other lipolytic enzynses is markedly -

TINKER ET AL.

disordered. All the phases are lamellar. For the purpose of this study, the three phases are designated asp', p , and a. The phase equilibria may be described schematically by the following:

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

where 6' is the low-temperature form. The transition temperature for the p-cr equilibrium is designated T , and is 23°C for DMPC and 41°C for DPPC. The pr-p transition is designated the 'pre-transition' and occurs at 15°C for DMPC and 33°C for DPPC. The purpose of the present work was to study the kinetics and temperature dependence of hydrolysis of long-chain saturated phosphatidylcholines by C . atrox phospholipase A,. We will show that the reaction is profoundly influenced by changes in the surface properties of the lipid phase and will attempt to formulate a kinetic description of hydrolysis in such heterogeneous systems. Materials and methods

Materiuls DPPC, DMPC, palmitoyl lysolecithin, and palmitic acid were purchased from Serdary Research Laboratories, London, Ont., Canada. The phosphatidylcholines were rechromatographed on silicic acid columns (30), dissolved in a minimurn of warm ethanol, precipitated with ether at -20°C, and dried in vncr4o. Thin-layer chromatographic examination showed that each phosphatidylcholine was free from By solechithin and fatty acids after purification; anafysis of the fatty acid composition (33) showed that the DPPC contained 98% hexadecanoic acid, while DMPC contained 95% tetradecanoic acid. Palmitic acid and lysolecithin were used as supplied; each was chromatographically pure. The phosphatidylcholines were also subjected to phospholipase .4, hydrolysis in the ether system of Wells and Hanahan (13); 100% hydrolysis to lysolecithin plus free fatty acids was achieved, providing evidence that the lipids had not been racemized. Phospholipase A, was prepared as previously desciibed (3) and stored as a lyophilized powder. Other chemicals were of the highest purity available. Lipid Di,spersions Stock lipid dispersions containing 2.5 mg of phosphatidylcholine per millilitre were made up in 12- to 20-ml volumes of 0.01 M CaCI, as required, and utilized the same day. This concentration of CaCl, was 10-fold greater than the Michaelis constant Kc,. (3). When pure phosphatidyicholines were to be dispersed, the appropriate amount was weighed into a flask. Some of the CaCl, solution was added to the flask containing dry lipid, and the contents suspended using a Vortex mixer. The suspension was transferred to a water-jacketed vessel and subjected to ultrasonication for 8 min using an MSE 100-W ultrasonic disintegnator equipped with an exponentially tapered titanium probe, at a frequency of 20 kilocycles per second. Samples were blanketed by a nitrogen atmosphere, anci the temperature was maintained at a value well below the transition temperature of the lipid: DPPC was sonicated at 20°@,DMPC at IO°C. (Dispersions sonicated on ice tended to precipitate after an hour or so.) After sonication, the dispersions were centrifuged for 10 min in a clinicai centrifuge in the cold room, to remove a small amount of titanium dust, and stored on ice. In appearance, the dispersions were turbid but there was no macroprecipitate. Dispersions were routinely examined by thin-layer chromatography to ensure that no hydrolysis occurred during preparation. They were

553

also examined by electron microscopy after negative staining (31). The dispersions contained numerous multilamellar structures which tended to clump together on the grids; only a very small fraction of apparently unilamellar vesicles (32) were observed. The micrographs were very similar to previously published photographs of lamellar-phase lipids (31). No detailed morphologic analysis was attempted. Enzyme Solutions A stock solution of enzyme having an A,,, of 0.5 to 1.0 was prepared as required in 0.01 M Tris-HCI buffer pH 7.5, containing 0.01 M CaCI,. and kept on ice. Protein concentration was calculated assuming a value for E,,,~% of 10, and the n~olarities calculated using a molecular weight of 29000. When reactions were to be curried out, a control was first run by pipetting the appropriate amount of enzyme solution into a voluine of 0.01 M CaC1, pH 7.5 equal to the voiume of lipid dispersion that was to be used. Sometimes there was a small drop in the pH (probably arising from the temperature dependence of the pk', of Tlis), resulting in a smdl addition of base from the automatic titrator; in such cases, the pH of the enzyme stock was adjusted slightly upward to eliminate this spurious effect. Recordirrg Hydrolysis Cl4n.e~ Release of fatty acid in reaction mixtures was recorded by continuous titration with NaOH solution (standardized against potassium acid phthalate). The reaction was followed using a titrigraph SBR2C - titrator 11 - pH meter 26 combi~aationobtained from Radiometer (Copenhagen). The titrigraph wasoperated in pH-stat mode, and the pH was maintained at 7.5. The reaction mixture was contained in a standard water-.jacketed Radiometer titration cup, and temperature was maintained to within k0.05"C with a circulating constant temperature water bath mai~ufacturedby Neslab Inc., Portsnaouth, NH. Temperature was measured in the titration cup using an electronic thermometer (model 46-TUC, Yellow Springs Instruments, Yellow Springs, OH). The stock lipid dispersion was diluted to the appropriate concentration with 0.01 M CaCI, and the required amount pipetted into the titration cup. The dispersion was incubated at the required temperature for 5 min, and the pH was then brought to 7.5 with NaOH solution (the pH was6.5-6.8 before adjustment). The reaction was initiated by injecting 20 to 30p1 of stock enzyme solution into the lipid dispersion, using a Hamilton syringe. The final volunae of the solution was either 5.0 or 6.0 ml. The reaction mixture was stirred in an atmosphere of watersaturated nitrogen. The pH was maintained at 7.5 using NaOH solutions (of varying concentrations (0.002, 0.005, or 0.010 N ) depending on the reaction rate) delivered from a calibiated AgIa micrometer syringe assembly (total voiume. 0.5 ml). pH was measured using Radiometer electrodes K4112 and 622B, calibrated at pH 7.5 at the operating temperature with standard buffers. Nr4merical Arzulysis Curve fitting to experimental data was carried out using the on-fine mathematical modeling program MLAB (34). through the courtesy of the Division of Computer Research and 'Technology, National Institutes of Health. Bethesda, MD.

Results Hydrolysis of DPPC In preliminary experiments using DPPC dispersions prepared by mechanical agitation on a Vortex mixer, we found that hydrolysis rates were quite low and not reproducible even when the same stock dispersion was used.

554

CAN. J. BIOCHEhI. VOL. 56, 1938

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

Since Wu and Tinker (1) reported that ultrasonication of egg lecithin dispersicsns greatly enhances the susceptibility to enzymic attack, we used this method to produce lipid particles having a greater sudace:volhmme ratio. The method described yielded lipid dispersions that were stable over at least an 8-h period and were rapidly hydrolysed. Kinetic parameters measured on a single dispersion at various times on the day of preparation were reproducible to within 5-1096. While different dispersions yielded results that were qualitatively similar, there were often large quantitative variations hetween such pra~aietersas initial velocities and length of 'lag' periods. Thus, the extent of dispersion influences the reaction profoundly. Figure Icr shows a set of progress curves for hydrolysis of DPPC at several temperatures below the transition 10 20 30 40 50 TEMPERATURE ( " C ) temperature, '$, (which is 41°C for this lipid). The initial velocity is seen to increase with temperature. These FIG.2. Dependence on temperature of initial velocity of hycurves were all obtained using the same stock lipid dis- dredysi5 of DPPC dispersions by phospholipase A?. Ordinate persion. In Fig. Ih, we show a set of progress curves expressed as mole$ substrate hydrolysed per lnole enzyme per obtained near and above T , using a different stock lipid second 4s- '). Conditions in all experiments: lipid concentration, dispersion. At temperatures of 4 1°C and higher, the 0.553 mM; enzyme concentration, 1.1 1 x 10- M ; total volume, ml. Inset, Arrheniras plot of data for the range from 20 to 40°C. curves exhihit three phases: a short initial burst of hy- 5The transition temperature defined by the intersection of the two drolysis, theta a period of very slow hydrolysis, folIowed tines corresponds to 34'C. The Arrhenius activation energies are by a dramatic increase in the reaction rate. In some of the the following: T < 34"C, E , = 1.43 t 64.28 kcaI mol '; T > 34"C, experiments (not shown), the period of slow hydrolysis E , = 42.5 2 4.98 kcal mol-I. (lag period) lasted as long as an hour. It was impossible to measure the velocity during the initial burst; therefore, centrations of DPPC (about four times the value of K,). the velocltyduring the lag period was taken as the 'initial' The velocities were expressed as moies product hyvelocity above 41°C. drolysed per mole enzyme per second (s-I); in separate In order to obtain a profile initial velocity versus temp- experiments, it was shown that the initial velocity was erature and also to determine the substrate dependence linearly proportional to enzyme concentration in the on the initial velocity, a very large number of initial range reported in Fig. 2. velocity measurements were performed on dilutions of a A sharp rise in the initial velocity is observed to occur single stock DPPC cjispersion at various temperatures. between 30 and 4WC. An Arrhenius plot of the data in this Progress curves were recorded for about 5 min in order to range indicates that this velocity increase can be interobtain the initial velocities. In Pig. 2 , we show the initial preted as arising from a transition of the system between velocities determined in the presence of saturating con- two states with a transition temperature of 34°C. Since this is the ' pre-transition9 temperature for DPPC, we tentatively attribute the rise invelocity to theP1-/3 transition. The sharp decrease in velocity at 41°C can likewise be attributed to the p-cr transition. The initial velocity ( v ) was determined as a function of DPPC concentration [S], at each temperature shown in Fig. 2. At all temperatures up to 42"C, there appeared to be a hyperbolic dependence of v upon [ S ] with some scatter evident. In Table 1 are shown the MichaelisMenten parameters V,,, and K , determined at temperatures up to 42°C. It is evident that the increase in velocity accompanying the 6'-/3 transition is caused by an increase in the parameter Vm,, only. The mean value of K , in this range was 1.41 + 0.28 x M , and no depenT ~ m e(rnin) dence on temperature was evident. At temperatures of 42°C and above, v seenned to be FIG. 1. Reaction progress curves for hydrolysis of DPPC dispersions by phospholipase '4,( u )at and below the P-a phase independent of BPPC concentration, but the values were transition and ( h )at and above this transition. Conditions: lipid so small that this could not be precisely established. concentration, 64.556 m M ; total volume, 6 ml; temperature sf reactions ("C) as indicated on progress curves. Enzyme concen- Hydro!ysis of D W C Figure 3 shows hydrolysis progress curves recorded tration was 1.10 x l W 7 M in (a) and 2.89 x 1W7M in ( b ) ; different stock lipid dispersions were used in experiments ( a 1 for a single dispersion of DMPC at various temperatures. As in the case of DPPC, these curves are characterized and 66).

555

TINKER ET AL.

TABLE 1. Temperature dependence of MichaelisMenten parameters for hydro1ysis of gel-phase DPPC" Temperature,

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

.e:

Kn, ,

vma,

M x 104

s- I

x

defined a characteristic time, r, as the abscissa of the intersection of the two tangents. v , , v2. and 7 are regarded asempirical parameters characterizing the initial position of the hydrolysis progress curves. Figlare 4 shows the temperatt~redependence of the parameters 11, and L,,. It is to be noted that the temperature variation of the initial velocity is not correlated with known phase transitions of DMPC; however, the velocity v , rises to a sharp maximum near the p-a transition ternperatlnre of DMPC which is reported to be 23°C (27)The dependence of and v , upon substrate concentration was studied at 20 and 3CB°C, temperatures respectivel y below and above the p-a transition temperature. -4s shown in Fig. 5, v , appeared to be independent of \?,

11,

'Conditions as in Fig. 2. Values expressed as least-squares estimate k standard deviation, obtained by fitting initial velocity to the fa1Iowing equation: t 3 K,,, [ S ] / ( K , $ [ S ] ) ,where [ S ] is the bulk substrate concentration (hi). (11)

-

0

10

20

30

48

50

TEMPERATURE ( " C )

TIME jmin)

FIG.4. Dependence of the velocities I . , and t7, upon temperature for the hydroiysis of DMPC dispel-sions by phnspholipase -4,. Conditions as in Fig. 3. A-A, t l , ; -t- - - +, i.,.

FIG. 3. Reaction progress callves for hycirolysis of DMPC dispersions by phospholipase -4,. Conditions: lipid concentration, 0.60 naM; enzyme concentration, 1.17 x lop7M ; total volume, 6 ml; temperature sf reactions ("0as indicated on progress curve. The empirical velocities I?, and 1*, were obtained by drawing tangents at zero time and at the inflection point as indicated for the c u n e measured at 30°C. 'The time .r defined by the abscissa of the intersection of these two tangents is disclassed in the text.

by a rapid initial reaction, followed by a slower phase and then an acceleration at later times; these features are most pronounced in the curve recorded at 30°C. For reasons discussed below, we defined empirical velocities v , and v, as the slopes of tangents drawn to the progress curves at zero time and at the inflection point respectively. This was made possibie by the discovery that the initial par3 of the progress curves could be desct-ibed very exactly by a cubic function,

wherep is the extent of hydrolysis, t is the time, and a I , a,, and a 3 are positive constants. Equations of this type were fitted by a least squares procedure to 10-20 points measured on the experimental progress curves of Fig. 3. (The inflection point is defined by the condition that . ) also d2p8db2= 0; thus, v = a , and v, = a , - a 2 2 / 3 ~ 3We

FIG.5 . Dependence of the velocities 11, and I?, (see Fig. 3) upon DMPC concentration at 20 and 30°C. Conditions: enzyme concentration, 1.12 x 10 M; total volume, 6 ml. Velocities expressed as moles substrate hydroIysed per mole enzyme per second (s-I)).Note different ordinate scalesfor a s , and 11,. Key to points: A , al, at 30°C; , v , at 20°C: , v2 at 20°C. Solid line through t7, points is the least squares Michaelis-Menten cilrve (V-, = 0.510 k 0.047s-I, K , = 1.987 _9 8.566 x IW4M).

5 56

CAN. J. BIOCKEM. VOL. 56. 1978

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

DMPC concentration at both temperatures, although there was considerable scatter at 30°C. At 20°C, the parameter aJ2 appeared to exhibit a hyperbolic dependence upon DNBPC concentration. The measui-ed K , value was 1.987 k 0.857 x lOA4M;this was clc~seto the K , for gel-phase BPPC. The data in Figs. 4 and 5 indicate that in the case of DMPC the dependence of the parameter a?, upon temperature and substrate concentration is similar to that of the initial velocity in the case of DPPC. In the case of BMPC, however, the progress curves suggest that a presteadystate 'burstq of hydrolysis occurs, as has been observed for other hydrolytic enzymes (35, 36) and that v, is a presteady-state velocity. Presteady-state kinetics of this type rare predicted to exhibit the foliowing features (36): (i) the velocity 11, is independent of substrate concentration: (ii) the characteristic relaxation time, T, is independent of the concentration of enzyme; (iii) the velocity v, is linearly propol-tional to enzyme concealtration; and (iv) as a coro8lary of (ii) and (iii), the extent of the presteadystate "urst' is linearly proportional to enzyme concentration. These predictions were tested for the case of hydrolysis of a-phase DMPC :at 38)"C.Feature (i) above has been verified. En Table 2, we show the dependence of v,, tr2,and T (designated as the 'burst time') upon enzyme concentration. These data show that the initial part of the progress curves appear to follow presteady-state 'burst' kinetics. A similar initial burst, though much less pronounced, was also observed in the hydrolysis of a-phase DPPC (Fig. 1). lZfl~ct~$Re~(cfion PI'O~UC~S Hydrolysis of liquid crystalline DMPC and BPPC is characterized by a Iong 'lag period' followed by a rapid increase of reaction velocity, to which we apply the term 'explosion.' Et seemed reasonable to suppose that the explosion was due to the accumulation of reaction products: accordingly, we investigated the effect of adding reaction products to the initial lipid mixture. These experiments were carried out using DPPC at 44°C. We defined the length of the lag period by the abscissa of the intersection of tangents to the progress curve in the slow and explosion phases. Figure 6 shows the effects of varying initial amounts of each of the products (palmitic acid TABLE 2. Dependence of initial velocities and burst time upon enzyme concentration for hydrolysis of liquid crystalline DMPCa Enzyme concentration, bt .%if

~'1.

1'2 s

nmo%*rnin-'nmolmin-'

T, Burst size, min %hydrolysis

"To 6.01~11 of a 0.60 mlM D M P C dispersion at 3OrC,here added 10, 20, 30 40 or 50 P I of a stock enzyme soli~tlon,concentration 2.95 ;, 10-§ I I f . I . , : c2,'and T were measured on the progress curves as shown in Fig. 5. The burst size is ( ~ ' ~ by 7 ) definition. The mean value of T is 1.46 k 0.15 min.

PRODUCT. SUBSTRATE MOLAR RATI 0

FIG.6 Dependence of length of 'lag period' upon initial product concentration for hydrolysis of DPPC dispersions by ghospholipase A2 at 44°C. Either palmitic acid or palnlitoyl lysolecithin was rnixed with the substrate :at the molar ratios shown. Conditions: substrate concentration, 0.556 anM; entotal volume, 6 ml. zyme concentration, 2.67 x 10 7 M ;

4

8

1.2

1.6

2

PRODUCT SUBSTRATE MOLAR RATIO

FIG. 7. Dependence of initial velocity lapon initial product concentration for hydrolysis of DPPC dispersions by phospholipase A, at 44°C. Conditic~nsas in Fig. 6. and palmitoyl lysolecithin) on the lag period. Evidently, the lag period is abolished when the initial mole fraction of either product is 8.1. For both products, there was a stimulation of the initial velocity of the reaction at mole fractions higher than those required to abolish the lag period. In the case of lysolecithin, the velocity attained a maximum at a mole fraction of0.2, then decreased, while in the case sf palmitic acid, the velocity increased continuously with mole fraction up to a mole fraction of 0.4, then remained constant. The data is shown in Fig. 7. Discussion Hydrc~lysis of phosphatides catalyzed by phospholipase Pa2 is an example of heterogeneous catalysis, since the reaction takes place at the surface of an insolu-

557

'TINKER ET ,4L.

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

ble lipid phase. It has been shown in the preceding section that the catalytic action of the enzyme is nlarkedly affected by phase changes of the lipid and by accurnulation of products. It is thus obvious that the characteristics of the lipid surface profoundly influence the action of the enzyme, as has indeed been noted by many other authors. The problem is to devise a kinetic scheme that takes surface effects into account and that quantitatively describes the phenomena seen in the present study. One such kinetic model has been developed by De Haas and co-workers (17, 18) to account for the hydrolysis catalyzed by pancreatic phospholipase A,. In effect, this is an 'adsorption-limited' model, in which catalysis occurs on the lipid surface after preliminary adsorption of the enzyme; changes in the rate of hydrolysis are associated with changes in the rate of adsorption or 'penetration' of the enzyme onto the lipid sufiace as well as with changes in the rate of reaction on the surface. As mentioned earlier, this model has been invoked to accouilt for the hyclrolysis of phosphatidplcholine dispersions by pancreatic phospholipase A, (24, 25). The increase in velocity near the phase transition temperature, T,. was attributed to an increased rate of adsorption of the enzyme onto the lipid surface. For several reasons, we do not feel this model can account for the present data obtained with the C . ntrox enzyme. In the first place, according to equation IX of Ref. 17, changes in the adsorption or desorption rate constants would affect only the Michaelis constant K , describing hydrolysis in the bulk dispersion; in fact, we have shown in the case of DPPC that the increased rate of hydrolysis of the p-phase is associated only with a change in V.',,. Again according to equation IX 4171, changes in the bulk parameters V,,, and K , can be related to changes in the surfice Michaelis-Menten parameters, designated kc,, and K g in Ref. 17; but the only way to account for an increase in V,,, is to assume a specific increase in the catalytic constant kc,,,. It seems unlikely that kcat would be affected by a change in orientation of the lipid molecules. A further argument against this model is that hydrolysis is predicted to exhibit a lag phase in the presteady state. In fact, we have observed a presteady-state burst of hydrolysis for a-phase DMPC and DPPC. This phenomenon cannot be accouilted for in the 'adsorptionlimited' model. Finally, it is difficult to see how this model can account for the low steady-state velocity of hydrolysis of a-phase lipids compared with the gel phase. One would expect u priori that adsorption and surface diffusion of the enzyme would be favoured in the more loosely packed a-phase, resulting in an increased rate of hydrolysis. - We have considered a number of alternate models in an effort to understand the kinetic data obtained in this study. One model which does adequately describe the data is shown in Fig. 8. Afull mathematical description of this model will be presented elsewhere (manuscript in preparation), but at this time, a qualitative description of its properties may be given. The key feature of the model is that the enzyme can

0 7 aapATH

1w

"PATH 2"

se E

S

B

F ~ 8. ~~ . i ~model ~ for ~ hydrolysis t i ~ of phosphatidylcholine aggregates by C . catvox phospholipase A,. See text for description.

only bind to the lipid by forming an ES complex. This implies that a conformational change occurs in the enzyme when it forwas such a complex, resulting in exposure of hydrophobic sites that can penetrate the lipid surface. Once hydrolysis has occurred, the enzyme can either desorb from the surface (path 1) or diffuse along the surflace to an adjacent substrate molecule (path 2). We consider that path f describes hydrolysis of the gel phases, i.e., in this case, k , >> k , , whereas for the a-phase, k , % k , and hydrolysis proceeds mainly by path 2 after a11 the enzyme is adsorbed. This seems intuitively reasonable, since it would be thought that adsorption and surpace diffusion of the enzyme would be favoured in the more loosely packed a-phase. However, surface diffusion is much slower than diffusion of the enzyme in water (since the viscosity of the lipid surfxe is high), hence hydrolysis by path 2 would be much slower than by path 1. This explains qualitatively the lower rate of hydrolysis of the a-phase. Moreover, the model predicts a presteady-state burst of hydrolysis which occurs as the enzyme settles down on the surface. This phenomenon was observed for hydrolysis of a-phase DPPC and DMPC, although it is much more pronounced in the latter case. The data are in accord with the predicted behaviour. The apparent differences in the behaviour of the enzyme with DMPG and DPPC can be resolved if we assume the initial velocity (v,) is a presteady-state velocity in the case of DMPC hydrolysis. For the gel phases, we assume hydrolysis proceeds mainly by path 1, since penetration of the tightly packed lipid surface is unfavourable. If this is so, we can obtain the following expressions for the steady-state Michaelis-Menten parameters: Vmax =

(4)

k3k4 CEO] and k , + k4

B - k, k - 1 + k 3 K, =ek3+k4 kl '

where E is the fraction of the total lipid initially available to the enzyme. According to this scheme,,,,fI is ahought

558

CAN. 9. BICPCHEM. VOI-. 56. 1978

Can. J. Biochem. Downloaded from www.nrcresearchpress.com by Depository Services Program on 12/06/14 For personal use only.

to depend on the desorption rate constant k,. while K , is :actually an adsorption equilibrium constant. Now depending on the behaviour of k - , , k , , and A,. it is possible for significant changes to occur in Vmaxwithout large changes in Km. We would expect k , (and hence V,,,) to increase if the lipid molecules became more tightly packed. This is what in fact occurs at the P-P' transition (28, %9),so the increase in La,accompanying this transition is as expected. An attractive feature of the model is that it can account for the stirnulatory effects of the prod~mctsof the reaction. We assume that the probability of an enzyme molecule remaining adsorbed to the lipid surface depends on the time required to 'find' another substrate molecule by diffusing along the surface. If it does not find a substrate, it is likely to desorb and undergo a conformationa8 change to the form in which the hydrophobic sites are masked. Thus. accumulatioia of products on the surface stimulates desorption of the enzyme, with conconnitant increase in the rate of hydrolysis. Our model is at present purely hypothetical, although it provides a unifying and relatively simple explanation for the variety of complex phenomena observed in this study. Its major attraction as a hypothesis is that it can be independently tested. If the model is correct, we must be able to show that the enzyme adsorbs to the surface of hydrolysable lipid phases, that it does not adsorb to lipid surfaces formed by nonsubstrates (e.g., sphingomyelin), that the adsorption eqa~ilibaiumis sensitive to phase changes in the lipid. that desorption is stim~ilatedby lysolecithin and fatty acid, and that there is a detectable conformational chringe in the enzyme upon adsorption. These experiments are currently in progress in this 1;alscsr:ttory.

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

One of the authors (D.O.T.) wishes to thank Dr. George Weiss. Chief, Physical Sciences Laboratory, Division of Computer Research and Technology, National Institutes of Health, for his hospitality and to acknowledge many stimulating discussions with Drs. V. A. Parsegian, N. L. Gershfeld, I. G. Darvey, and R. Shrager during an extended stay at the National Institutes of Health. 1 . Wu, T.-W. & Tinker, D . 0. (1969) Biocizernistry 8 , 1558- 1568 2. Hachimori, Y . , Wells, M. A . & Hanahan, D . J. (1971) Br'ochenristiy 10,4084-4889 3. hrcion, A. D., 'Tinker, D. 0. & Spel-o, t. (1977) C Q P IJ.. Wioc-izetn.55, 205-214 4. Wells, M. A. (1972)Biocl~eniistry11. 1830-1041 5. Wells, M . -4. (1973) Biochenaistgv 12, 1088- 1085

28. 29. 30.

31. 32. 33. 34. 35. 36.

Wells, M. A. (1973) Biochemistry 12, 1086- 1093 Wells, M. A. (1974) Bic~cltemistry13.2248-2257 Wells, M. A. (1974) Bio(*izesnistry13, 2258-2264 Wells. W-ll. A. (1974) Bioche~lzistry13,2265-2268 Wells, M . A . ( 1974) Biochemistry 63,4937-4942 Misiorowski, R. L. 8r Wells, M. A. (1974)BiochQniis~ry13, 492 1-4927 P m n , P. H . $r. Wells, M. A. (1974) Biochemistry 13, 4928-4934 Wells, M. A . & Hanahan, IT. J. (1969) Biochenli.g.try 8 , 414-424 Shen.B.W.,Tsao.F.H.C..Eaw,J.H.&Kezdy,F.J. (1975) J. Am. Chem. Soc. 97, 1205-1208 De Haas. 6. H . , Bonsen, B. P. M . , Pieterson, W. A. $r. van Deenen, L . E. M. (1971) Bbochirn. Bioplzys. Actu 239. 252-266 Rohslt, 0. A. $r. Schlarnowitz, M. (1961) Arch. Biochrt7i. Biophy~;. 94,364-379 Verger, R . , Mieras. M. C 9 E. & de Haas, G . H . (1973) J. Biol. Chem. 248,4023-4034 Verger, R. & de Haas, G . H . (1976) Allrl~c.Re\!. Biophys. Bioeng. .5,77- 1 17 Slotboom. A. J., Verger, R . , Verheij. PI. M . , Baartrnans. P. H. M . . van Deenen, L . L . M. & d e Haas, G . H . (1976) Chenl. P6zys. Eipicis 17, 128-147 Verger, R., Rietsch, J., van Dam-Mier-as, M. C . E. & de Haas, G . H. (1976) J. Biol. Cllem. 251,3128-3133 Dennis, E. A . (1973) Arch. Biochern. Bic~phys. 158, 485-493 Dennis, E. A . (1973) J. LipidRes. 14, 152-159 Deems. R . A . , Eaton, B. R. & Dennis. E. A. (1975) J . Biob. &72eva.250,9013-9020 Op den Karnp, 9. A. F . , d e Gier. J. & van Deenen, E. L . M. (1974)Biochhrre. Biophys. Actu 345,233-256 Op den Kamp. J . A. F . , Kaiierz, M. T. & van Deenen, L. L. M. (1975)Biochirn. Biopizys. Acts 406, 169- 177 Phillips, M . C., Ladbrooke, B. D . & Chapman, D. (1970) Biochim. Biophyr. Actu 196,35-44 Tardieu, A . , Luazati, V. & Reman, F. C. (1973) J. Mol. Bioi. 75,711-733 Rand, R. P., Chapman, D. & Larsson, K . (1975)Biopily.s.J. 15, 1 1 17-1124 Janiak, M. J., Small. D . M. & Shipley . G . 6 . (1976) Biochemistry 15,4575-4580 Tinker, D. 0. B Pinteric, L . (19'71) Biocl~emistgp, 10, 860-865 Tinker, D. O.,Pinkeric, I,., Hsia,J. C. & Wand,R. P. (1976) Ccan. J. Biochenr. 54,209-218 Huang, C. (1949) Biochsmistry 8 , 3 4 4 - 3 2 K u k ~ i sA., , Marai. I,.,Breckenridge, W. C.,Gornatl, D. A. & Stachnyk, 0. (1969) Can. J. PizysioB. P / I u ~ I ~ N L 46, .(P/ 5 1 1-524 Mnott, 6 , D, d Shrager, R. I. (1972) Covrap~~t. Grcry/tic's6 (4), 138-158 Segel, I. H. (1975) Ezzynze Kinetics, pp. 621-623, John Wiley and Sons, New York, NY, U . S . A . Scrirngesur, K . G . (19771 Chemistry undgbntro/c$En:vme Re~rciiotls,p. 158 and p. 318, Acatfernic Press, London, England

Kinetics of hydrolysis of dispersions of saturated phosphatidylcholines by Crotalus atrox phospholipase A2.

Kinetics of hydrolysis of dispersions of saturated phosphatidylcholinesby Crofalusatrox phospholipase A, Can. J. Biochem. Downloaded from www.nrcrese...
731KB Sizes 0 Downloads 0 Views