Kinetic analysis and substrate specificity of a lysophospholipase from the macrophage-like cell line P388D1

43

Biockimica et Biopkysica Acta, I167 (1993) 43-48 Q 1993 Elsevier Science Publishers B.V. All rights reserved 0005-2760/93/$06.00

BBALIP 54115

Kinetic analysis and substrate specificity of a lysophospholipase from the ~acrophage-like cell line P388D, Richard E. Stafford, Ying-Yi Zhang, Raymond A. Deems and Edward A. Dennis Department

ofCkemistry, Uni~~ersityof California, San Diego. La .iolia, CA (USA) (Received 17 August 1992)

Key words: Lysophospholipase; P388D, cell; Interfacial activation; Substrate specificity; Kinetic analysis

The kinetics of the Iysophospholipase purified from the P388D, macrophage-like cell line (Zhang and Dennis (1988) J. Biol Chem. 263, 996.5-9972) have been explored. Three different lysophospholipids were used in these studies: l-hexadecanoyllysophosphatidyicholine, l-tetradecanoyllysophosphatidylchoiine, and l-hexadecanoyllysophosphatidylglyceroi. Since all of the substrate dependence data for these substrates fit a Hi11model, the enzyme’s activity appears to be cooperative requiring at least two lipid molecules for full enzymatic activity. The enzyme did not show a preference for any of these substrates since their k,,, ranged from 1.2 to 1.5 ymoi min-’ rng.--’ and their half-maximal activities ([S&,1 were achieved at substrate concentrations between 15 and 33 FM. Enzymatic activity also appeared to be independent of the aggregation state of the substrate. No dramatic changes in rate could be associated with substrate aggregation at the critical micelle concentration. This is in marked contrast to some phospholipases A, that exhibit dramatic activations when their substrates aggregate. In very dilute solutions, less than 0.5 pg/ml protein, the lysophospholipase loses activity irreversibly within minutes. This effect of low protein concentrations can be overcome by maintaining the enzyme in the presence of greater than 10 PM lysophospholipid. This inactivation can affect kinetic studies, since the [S],,,s for this enzyme are usually in this range. We have found that increasing the protein concentration with a ‘non-specific protein’, e.g., cytochrome c, can protect the enzyme without affecting activity and, thus, allow valid kinetic data to be obtained over the full substrate concentration range.

Introduction

Lysophosphoiipids are amphipathic compounds that are important lipid components of the cell and are potent biological agents due to their detergent-like and membrane lytic characteristics (for a review see Ref. 1). Their levels in the cell are tightly regulated at very low levels. This control is achieved by balancing the activities of at least three enzymes. Lysophospholipids are produced by the action of phospholipases on phospholipids. They are removed by reacyiation to phospholipid via an acyl transferase or by further hydrolysis to glycerol phosphate and fatty acid via a lysophospholipase (for a review see Ref. 2). These pathways not only control the level of lysophospholipid but also which portion of the lysophospholipid is shunted into remodeling and which is degraded further. Thus, the levels of lysophospholipase and the modulation of its activity could be important factors in

controlling the flow of lysophospholipid through these two pathways. Despite the possible importance of this enzyme, relatively few lysophospholipases have been purified, and fewer still have been characterized kinetically. As part of our continuing study of phospholipid metabolism and eicosanoid production in the P388D, macrophage-like cell line 13-51, several lysophospholipases have been purified to homogeneity and characterized in our laboratory 16471. We now present a kinetic analysis of one of these lysophospholipases. We have explored the enzyme’s kinetics, its polar head group specificity, and whether it exhibits the surface activation phenomenon that is characteristic of several of the extracellular phospholipases A,. In the course of these studies, we have delineated conditions under which the enzyme is stable and under which it can be studied kinetically. Materials

Correspondenceto: E.A. Dennis, Department of Chemistry, University of California, San Diego, La Jolla, CA 92093-0601, USA. Abbreviations: cmc, critical micelle concentration; IysoPC, l-acy-snglycero-3-phospho~lchoiine; IysoPG, 1 acyl-sn-glycero-3-phosphorylglycerol; PC, l,Z-diacyl-~~-glycero-3-phospho~lchoiine.

and Methods

Ma teriaIs

All non-radiolabeled lipids were purchased from Avanti Polar Lipids, Birmingham, AL. The radiolabeled lipids were purchased from Amersham, New

44 England Nuclear, or prepared as described below. The assay solvents were reagent grade, and the TLC plates were silica gel G (Merck). The solvents used in the synthesis were Gold Label (Aldrich Chemical) and Optima Grade (Fisher) which were further purified by standard methods [S]. The enzyme was purified and stored as described previously [6]. Ly.~ophospho~~paseassay

The standard assay mixture contained 150 PM lysophospholipid dissolved in 100 mM Tris-HCI buffer, (pH 8.01, in a final volume of 0.5 ml. The assays were initiated by adding a known amount of enzyme. The reactions were carried out in siliconized test tubes at 40°C for 60 min. The reactions were quenched and the products and reactants separated by either the TLC or the Dole extraction procedures. In the substrate dependence studies, the assay time and enzyme concentration were varied to maintain the extent of hydrolysis at appropriate levels for the different substrates. TLC separation

The assay was quenched with 1 ml of CHCl,/ MeOH/ HOAc (2 : 4 : 1, v/v). Then, 0.5 ml of CHCl 3 and 0.5 ml H,O were added, and each tube was vortexed for at least 15 s. The mixtures were centrifuged (3000 x g for 3 min) and the CHCI, phase removed. The CHCI, was evaporated under vacuum at 4O”C, the lipids brought up in 30 ~1 of CHCl,/ MeOH/HOAc (2: 4: 1, v/v), and spotted onto TLC plates. The TLC plates were run in CHCl,/MeOH/ HOAc/H,O t 15 : 25 : 4 : 2, v/v). The products and reactants were visualized with I,, scraped, and counted.

For assays containing lysoPG, the Dole extraction procedure was used because 1ysoPG did not extract completely into any organic phase. In the Dole extraction, fatty acids are completely extracted into a heptane phase along with a varying amount of the IysoPG. The latter is quantitatively removed with silica gel. The heptane phase can then be counted to determine the amount of fatty acid produced during the assay. In this procedure 191,the reaction was stopped by the addition of 2.5 ml of 2-propanol/heptane/O.5 M H$O, (20: 5 : 1, v/v). About 0.1 g of silica gel (Bio-Sil A, 100-200 mesh, Bio-Rad) was added and vortexed immediately. Then 1.5 ml of heptane and 1.5 ml of deionized water were added and vortexed for at least 10 s. 1 ml of the upper heptane phase was mixed with 5 ml of scintillation fluid and counted. Enzyme stabilization

The stability of the lysophospholipase was tested by diluting the stock enzyme, 10-2.5 pg/mI, to 360 ng/ml in a total volume of 350 ~1 with a preincubation buffer

which contained I mM dithiothreitol, 0.5 mM EDTA, 10 mM Tris-HCl (pH 8.0), and various amounts of the potential protecting agent. This was incubated at 40°C for 60 min. At the end of the incubation 150 ~1 of substrate were added. The final assay contained 125 PM lysophospholipid and 0.1 M Tris-HCI (pH X.01. This assay was incubated at 40°C for 30 min, whereupon the products were analyzed by either the Dole or the TLC procedures. Preparation

of

I-/l- 14C~tetradecanoyi-sn-gly~ero~phos-

p~~o~ic~~oline l-[l-‘4C]Tetradecanoyl-s~-gIycerolphosphorylcholine cl-tetradecanoyl IysoPC) was prepared by enzymatic hydrolysis of the 1,2-[1-‘4C]ditetradecanoyl PC by phospholipase A, (5 pg/ml) in CHCI,/O.OS M borate buffer (pH 7.0) (1: 11, 10 mM CaCl,, and 25 mM NaCl at 40°C. The reaction was generally complete after 60 min. The reaction mixture was extracted by the method of Bligh and Dyer [lo]. The CHCl, phase was recovered, dried, and then brought up in a minimal volume of methanol. The lipid was then further purified by precipitation in cold diethyl ether as described previously [l 11. Preparation of I-[l- “Cjhexadecanoyl-sn-glycerolphosphorylglycerol I,2 di-[ l- ‘JC]Hexadecanoyl-sn-glycerolphosphoryl-

glycerol was prepared by transphosphatidylation of 1,2di-[ I- ‘“Clhexadecanoyl-m-glycerolphosphorylcholine in the presence of 60% glycerol as described previously by Comfurius and Zwaal [ 12f. The I-[ l- 14C]hexadecanoyl-~~-gIycerolphosphorylglycerol (I-hexadecanoyl IysoPG) was prepared from this by phospholipase A, treatment as described by Stafford et al. [13]. Results Enzyme stability

The purified lysophospholipase, when stored at a protein concentration of lo-25 pg/ml in the presence of 60-70% glycerol and 10 PM P-mercaptoethanol, was stable for months at -20°C or for 24 h at room temperature. The enzyme gave tinear time courses under standard assay conditions using I-hexadecanoyl 1ysoPC as the substrate [G]. However, at low substrate concentrations the time courses were not linear. We investigated the cause of this non-linearity by preincubating the enzyme with 3 PM and 10 PM l-hexadecanoyl 1ysoPC. Aliquots were taken at various times and assayed in a standard assay containing 150 PM substrate. The enzyme that was incubated with 3 PM substrate lost 65% of its activity within 30 min (Fig. 1). In contrast, little activity is lost in 30 min if the preincubation buffer contained 10 FM substrate. 1-Tetradecanoyl 1ysoPC stabiiizcd the enzyme below its cmc (70

45

yM) while l-hexadecanoyl 1ysoPC stabilized the enzyme above its cmc (7 FM). Over the course of the preincubations, between 8 and 20% of the 1ysoPC is hydrolyzed. We cannot, therefore, rule out the possibility that the reaction products, fatty acid or glycerol phosphorylcholine, are contributing to the stabilization of the enzyme. This lack of stability at Iow substrate concentrations could be a serious impediment to conducting kinetic studies on this enzyme, since the substrate concentration that yielded half maximal activity, [ES],,, was in this region. Thus, a significant portion of any kinetic study must be done at low substrate concentrations in a region where the enzyme activity is also being affected by the enzyme’s instability. We tested two methods of stabilizing the enzyme to overcome this problem. Since it was possible that the 1ysoPC was stabilizing the enzyme by offering it a membrane like environment, we checked to see if phospholipid could be substituted for lysophospholipid in protecting the enzyme. Up to 1000 PM phosphatidylchoIine vesicles were found not to protect the enzyme. Our second approach was based on the fact that many proteins are susceptible to denaturation when very dilute and that this denaturation can be prevented, at least in some cases, by adding nonspecific protein to the solution to increase the overall protein concentration. We have tried several proteins in the hopes of finding one that does not affect the enzyme or the assay but that still protects the enzyme from inactivation. Cytochrome c seems to fill these criteria. It neither activates nor inhibits the enzyme nor does it interfere with the lipid extractions. Its protection shows a dose dependence and 100% protection is

7 ;

r

0.4

.’ E

0.0

1.6

0 s

I .2

x .c ., z 4 .u L : ::

0.8 0.4 0.0 1.6 1.2 0.8 0.4

-

0.0 0

50

-

100

150

200

Substrate

250

300

350

(FM)

Fig. 2. Lysophospholipase activity toward (A) I-hexadecanoyl IysoPC, (B) I-tetradecanoyl IysoPC, and (Cl I-hexadecanoyl IysoPG as a function of concentration. The cmc of each substrate is indicated. The lines represent non-linear regressions of the data fit to the Hill model.

afforded when the cytochrome greater than 500 pg/ml.

c concentration

is

K~~~t~~ analyze and substrate ~~~~~~n~~

The substate concentration dependence of the lysophospholipase activity was determined for three different lysophospholipids. In an attempt to keep from adding a possible perturbant to the kinetic assays, these experiments were not done in the presence of cytochrome c. Assay times were kept as short as. possible to keep enzyme inactivation to a minimum. The

=Oc

TABLE

I

Kitzetic parameters of ~y~u~~ospho~ipuse Substrate

cmc (r~,Ml

ISI”, &Ml

11

1.5

17 15

1.4 1.7

1.3 1.3

5 33

k,, (pmol

mini’

mg-‘1

0

10

20

30 Preincubation

40 Time

50

60

70

(min)

Fig 1. Activity of lysophospholipase as a function of preincubation time. The enzyme (64 nmoll was incubated in 0.1 M Tris-HCI (pH 8.0), in a total volume of 0.5 ml, at 40°C with either 3 FM l-hexadecanoyl IysoPC (0) or 10 PM I-hexadecanoyl IysoPC (0). At each time point, an aliquot was removed and assayed for enzymatic activity in the standard assay with 150 FM I-hexadecanoyl IysoPC. Each data point was assayed in quadruplicate.

I-Tetradecanoyl IysoPC I-Hexadecanoyl IysoPC I-Hexadecanoyl IysoPC + cytochrome c I-hexadecanoyl IysoPG

a These kinetic parameters

70 7

20

1.2

2.3

-

1.5

were obtained from the nonlinear regression analysis [15] using a Hill model. The standard errors in the estimates of k,,, and n were +7% while those for [S],,s were i 18%.

46 data for 1-hexadecanoyl IysoPC, 1-tetradecanoyl IysoPC, and 1-hexadecanoyl IysoPG are shown in Fig. 2. The cmc values [13] of each substrate are indicated. Various kinetic models were tested by fitting the model to the data via non-linear regression [5,14]. In each case, the data did not fit a simple Michealis-Menten model. Analysis of the [Sl,,,,/[S],,,, ratios of this data, as suggested by Segel [153, indicated the presence of a significant amount of sigmoidicity. We, therefore, tried to fit the data to two types of sigmoidal kinetic model. The first was a simple Hill model and the second was the boundary lipid model proposed by Sanderman et al. [16,17]. We chose to test the latter model because of the possibility that the stabilizing effect of lysophospholipid on the enzyme is due to the substrate forming a boundary lipid layer around the enzyme, or that the enzyme forms some type of premicellar aggregate with the substrate. Non-linear regression indicated that only the Hill model, with a Hill coefficient between 1.4 and 1.7, fit the data. The boundary lipid models assume a larger number of lipids are binding to the enzyme, but these data clearly indicated that just two molecules are required. The kinetic parameters obtained from the nonlinear regression are presented in Table I. Interestingly, the micellization of the substrate does not seem to correlate with any consistant change in enzymatic activity. The [S],,,, for l-tetradecanoyl IysoPC is below the cmc of the substrate, while the [S],,, values of I-hexadecanoyl IysoPC and 1-hexadecanoyl 1ysoPG are very close to the cmcs of these substrates 1131. In the latter two cases, there are no obvious changes in the curves at the cmcs. In light of the fact that the [S],,,s for these three substates were in the lo-30 PM range, the substrate dependence curves had to be extended below the limit required for enzyme stability to obtain sufficient data for kinetic analysis. There is the possibility that the

t

4

50

f

.__

13”

100 Hexadecanoyl

LysoPC

^^^

L””

($4)

Fig. 3. Lysophospholipase activity toward 1-hexadecanoyl 1ysoPC in the presence of 500 wg/ml cytochrome c. The line represents a non-line& regression of the data fit to the Hill model.

sigmoidicity is in some way due to the inactivation of the enzyme at the lower substrate concentrations. To ascertain whether this is occurring, we repeated the 1-hexadecanoyl 1ysoPC experiment with cytochrome c (500 pg/ml) present in the assay to stabilize the enzyme. These data are shown in Fig. 3. Again, the data could be fit only to the Hill model with a Hill coefficient of 2.3. While the maximum velocity was comparable to the others, the [S],,,, (5 PM) was one-third of that obtained for the same substrate in the absence of cytochrome c. Discussion

Lysophospholipase stability The purified P3880, lysophospholipase was very unstable. We have, however, found conditions under which the enzyme’s stability is significantly increased. The enzyme can be stored at -20°C in 60-70% glycerol in the presence of P-mercaptoethanol for several months without losing activity. The protein concentration should be kept above 10 pg/ml during storage. Upon dilution, the enzyme again becomes unstable. At assay concentrations the enzyme loses activity with time and nonlinear time courses are obtained. This particular inactivation can be prevented by maintaining the lysophospholipid concentration in the assay above lo-20 PM. In the standard assay, which contains 150 PM lysophospholipid, the enzyme is quite stable and linear time courses are obtained. Since both monomeric 1-tetradecanoyl 1ysoPC and micellar I-hexadecanoyl IysoPC stabilized the enzyme, there appears to be no connection between enzyme stability and the presence or absence of an interface or lysophospholipid aggregate. The idea that stability is not conferred by the presence of an interface is bolstered by the fact that phospholipid vesicles were unable to stabilize the enzyme. The stabilizing effect of lysophospholipid on the enzyme seems to be due to a direct interaction between the enzyme and individual lysophospholipid molecules. The instability of the enzyme at low lysophospholipid concentrations poses a problem for kinetic studies. Because the [S],,,, values for the enzyme are on the order of 15 PM, substate concentrations down to l-2 PM must be used to obtain good kinetic data that covers an appropriate range of substrate concentrations. The rates obtained at these concentrations will be artificially lowered due to enzyme inactivation. This introduces an additional factor into the kinetic analysis; one that is difficult to deal with. In an attempt to stabilize the enzyme under these conditions, we have found that adding exogenous protein to the enzyme solution protects the lysophospholipase from inactivation; the best results were obtained with cytochrome c. Cytochrome c was the only protein that protected the

47 lysophospholipase without affecting activity and without affecting the assay extractions. Since several unrelated proteins can protect and since significant amounts of protein are required (500 bg/ml) to stabilize the enzyme, the protection is probably due to non-specific interactions. One is always leery of adding extraneous proteins to an assay even though they appear to be inert. In the case of low substrate concentrations with this enzyme, the benefits may outweigh the potential disadvantages. Enzyme kinetics

There have been few kinetic characterizations of lysophospholipases and even fewer still that have looked at substrates other than 1-hexadecanoyl 1ysoPC [l,lS]. One salient conclusion that can be drawn from these kinetic studies is that all of these enzymes exhibit complex kinetics. The non-hyperbolic plots that we have obtained for the P388D, lysophospholipase are typical of these complex kinetics. We have probed the nature of this phenomenon with the P388D, lysophosphoiipase and a series of related lysophospholipids. We have also explored the possibility that this lysophospholipase exhibits interfacial activation by determining whether its activity increases dramatically when the substrate aggregates into micelles. Our first concern in analyzing the kinetic data was to determine if the enzyme’s instability was responsible for the complex kinetics that were observed. To this end, we examined the kinetics of l-hexadecanoyl 1ysoPC hydrolysis both in the presence and absence of cytochrome c and, thus, in the presence and absence of enzyme inactivation. While the kinetics were not identical, - k,,, was the same but the [S]0,5~were different - the basic characteristics of the kinetics were the same. Both sets of data exhibited cooperative kinetics, and the data fit a Hill model. The cooperativity did seem to be greater in the presence of the cytochrome c. The Hill number went from 1.7 for the non-protected enzyme to 2.3 for the cytochrome c protected one. Thus, the inactivation of the enzyme at low substrate concentrations produced a small quantitative effect on the kinetics but was not the cause of the compIex, cooperative kinetics of this enzyme. If this enzyme’s role in the cell is to clear lysophospholipid, the steepness of the activity-substrate dependence produced by this cooperativity would aid the enzyme in keeping these deleterious compounds below certain critical levels. Since the complex cooperative nature of lysophospholipase kinetics appears to reflect a true characteristic of the enzyme and not an artifact of enzyme instability, we explored the possible correlation of this complex behavior to the aggregation state of the substrate. Lysophospholipids undergo a concentration dependent aggregation at their cmcs when monomers of

lysophospholipid aggregate into large micellar structures. Many phospholipases exhibit dramatic changes in activity when their substrates undergo such changes [19]. We have employed hexadecanoyl lysoPC, cmc 7 PM, hexadecanoyl IysoPG, cmc 20 PM, and tetradecanoyl lysoPC, cmc 70 ,uM, to determine whether there is any correlation between the substrates cmc and the kinetic parameters. As Table I clearly indicates, no such correlation exists. The maximal activities, kfat, are all within experimental error. While the [SJ,, for lhexadecanoyl 1ysoPG was twice that of the other two lysophospholipids, there was no apparent correlation of this with the cmcs. Thus, the aggregation state of the substrate does not appear to affect P388D, lysophospholipase activity. Sandermann, Fleischer and co-workers [l&17,20] have shown that cooperative behavior can be caused by other phenomena. If an enzyme requires a boundary layer of lipid around it to acheive maximal activity, cooperative kinetic behavior can be observed under certain circumstances. We have tested several of their models but have found that their models did not fit our kinetic data. In the end, it appears that the P388D, lysophospholipase requires only two or three lysophospholipid molecules and that it interacts with these in a cooperative fashion. The cooperative nature of its kinetics is indeed a characteristic of the enzyme and is not an artifact of enzyme instability, nor is it related to substrate aggregation.

This lysophospholipase exhibits little substrate specificity for chain length and hydrolyzes 1ysoPG as well as it does 1ysoPC. The maximal activities were all within experimental error and the [S],,, values varied by only 2-fold for the non-cytochrome c experiments. While we have previously shown that this enzyme is not a general esterase [61 and is specific for lysophospholipids, it appears to show little specificity within this class of compounds.

These studies were supported by National Institute of Health Grants GM 20,501 and HD 26,171. References Stafford, R.E. and Dennis, E.A. (1988) Colloids Surfaces 30, 47-64. Roberts, M.F. and Dennis, E.A. (1989) in Phosphatidylcholine Metabolism (Vance, D., ed.), pp. 121-142, CRC Press, Boca Raton. Ross, MI., Deems, R.A., Jesaitis, A.J., Dennis, E.A. and Uievitch, R.J. (1985) Arch. Biochem. Biophys. 238, 247-258. Ulevitch, R.J., Sane, M, Watanabe, Y., Lister, M., Deems, R.A. and Dennis, E.A. (19881 J. Biol. Chem. 263, 3079-3085.

48 5 Listcr, M.D., Deems, R.A., Watanabe, Y.. Ulevitch, R.J and Dennis, E.A. (1988) J. Biol. Chem. 263, 75067513. 6 Zhang. Y.Y. and Dennis, E.A. (1988) J. Biol. Chem. 263, 99659972. 7 Zhang, Y.Y.. Deems, R.A. and Dennis, EA. (1991) Methods Enzymol. 197, 4566468, Academic Press. Orlando. 8 Gordon, A.J. and Ford. R.A. (1972) The Chemists’s Companion. A Handbook of Practical Data, Techniques and References, p. 429, John Wiley & Sons, New York. 9 Ibrahim, S.A. (1967) Biochim. Biophys. Acta 137, 413. 10 Bligh, E.G. and Dyer, W.J. (1959) Can. J. Biochem. Physiol. 31, 911-917. 11 Pliickthun, A. and Dennis, EA. (1981) J. Phys. Chem. 85. 678683. 12 Comfurius, P. and Zwaal, R.F.A. (1977) Biochim. Biophys. Acta 488, 36-42. I3 Stafford, R.E., Fanni, T. and Dennis, EA. (1989) Biochemistry 2x. 5 I 13-s 120.

14 Press, W.H., Flannety, B.P., Tuekolsky, S.A. and Vetterling, W.T. (1986) Numerical Recipes. pp. 498-546, Cambridge University Press. London. I5 Segel, I.H. (1975) Enzyme Kinetics: Behavior and Analysis of Rapid Equilibrium and Steady-State Enzyme Systems. pp. 346 367, John Wiley R: Sons, New York. 16 Sandermann, H. (1982) Eur. J. Biochem. 127,123-l%. 17 Sandermann, H. Jr., McIntyre, J. and Fleischer, S. (1986) J. Biol. Chem. 261, 6201-620X. 1X Stafford, R.E. (19X9) Ph.D. Dissertation, Univeristy of California, San Diego. 19 De Haas, G.H.. Bonsen, P.P.M., Pieterson, W.A. and Van Deenen, L.L.M. f 1971) Biochim. Biophys. Acta. 239, 2.52-266. 20 Cortese, J.D.. Vidal, J.C.. Churchill, P.. McIntyre. J.O. and Fleischer. S. (19823 Biochemistry 2t. 3X99-3908.