266
Biochimica et Biophysics Acta, 573 (1979) 266-275 0 Elsevier/North-Holland Biomedical Press
BBA 57369
PURIFIED PHOSPHOLIPASE MEMBRANE
A2 FROM S3EEP
PREFERENTIAL HYDROLYSIS 2-ACYL CHAINS
JUAN JIMENO-ABENDA~O Institute of Biochemistry,
ACCORDING
ERYTHROCYTE
TO POLAR GROUPS AND
and PETER ZAHLER *
University of Berne, Freiestrasse 3, Berne (Switzerland)
(Received November 24th, 1978)
Key words: Phosphol~pase A,; Polar group hydrolysis; Acyl chain; ~Pur~f~cat~on,Sheep ery throcy te membrane)
Summary Hydrolysis of natural phospholipids by pure erythrocyte membrane phospholipase AZ was compared to the reaction catalyzed by the soluble pancreatic enzyme. Fatty acids liberated during both types of reaction were quantitatively analyzed by gas liquid chromatography. We confirm for the pancreatic enzyme lack of specificity with respect to the sn-2 acyl chain of the phospholipids and preference for negatively charged polar head groups. Conversely, the membranous enzyme preferentially attacks uncharged phospholipids and within one class of phospholipid preferentially splits long chain unsaturated fatty acids in the sn-2 position. The significance of such differences between pancreatic and sheep erythrocyte enzyme is discussed in relation to the possible physiological role of the latter enzyme.
Introduction Phospholipase AZ (EC 3.1.1.4) activities have been found in several organisms and tissues. The ones characterized best are the soluble enzymes purified from pancreas f 1] or snake and bee venoms [2,3]. We have reported the existence of a membr~e-bound phospholip~e A, in ruminant erythrocytes [4, 51 which has been purified to electrophoretic homogeneity [6]. The function of ruminant phospholipase A2 is not clearly understood as is the case for most * To whom
correspondence shouldbe addressed.
267
of the membrane-bound phospholipases. One conspicuous fact about the structure of ruminant erythrocyte membrane is its low phosphatidylcholine content. In lipid extracts of sheep red cells, only trace amounts of this phospholipid are found beside a relatively high amount of sphingomyelin. This is in contrast to the majority of mammalian red cell membranes and other sheep tissues including serum and bone marrow, where phosphatidylcholine is a major component of membrane phospholipids [ 7,8]. As both facts, low phosphatidylcholine content and presence of phospholipase A, activity, are peculiar to ruminant red cells, one is tempted to connect them in search for the enzyme function. Therefore, we have previously proposed [4] that the role of the ruminant phospholipase might be to inhibit the increase of phosphatidylcholine concentration in the red cell membrane. We have indeed reported several lines of evidence showing that sheep red cells increase their phosphatidylcholine content when incubated with sheep or human serum provided its phospholipase is inhibited or inactivated [ 59 1. Also we have been able to show that the zwitterionic phosphatidylcholine is the preferred substrate in contrast to most of the soluble enzymes preferring the negatively charged phospholipids [4]. However, also phosphatidylethanolamine is split, but this is of little consequence due to the fact that the phospholipase is located on the outside of the sheep red cell membrane [ll], whereas most of the phosphatidyleth~olamine is in the inner half side of the lipid bilayer [lo]. In order to know more about the possible function of this membrane-bound enzyme, we have further studied in detail the problem of specificity, e.g. class of phospholipid and length and unsaturation of the split fatty acids in comparison with the pancreatic phospholipase, which was shown to split off the fatty acids independent of their length and unsaturation [ 17 1. Materials and Methods Products Crystalline porcine pancreatic phospllolipase was a generous gift of Doctor A.J. Slotboom. Egg phosphatidylcholine, egg phosphatidyleth~olam~e, phosphatidylglycerol and phosphatidylserine were purchased from Lipid Products, South Nutfield NR Redhill SY, U.K. Margaric acid, margaric methyl ester and magnesium silicate were purchased from Merck Laboratories. Other standards for gas chromatography were provided by Supelco Inc., Bellafonte, PA, U.S.A. All the other chemicals were analytical grade. Phospholipase purification Sheep red cell membranes were prepared as previously described [5] from carefully washed erythrocytes. After lauryl sulfate solubilization of the stroma, the enzyme was purified more than 1000 times by gel filtration and affinity chromatography as reported in Ref. 6. Phosphalipase assay Enough phospholipid (egg lecithin, egg phosphatidylethanolamine, bovine phosphatidylserine, bovine phosphatidylglycerol) to reach a final concentration of 700 nmol per ml was dried under nitrogen, then suspended in 10% sodium
268
cholate (final concentration 0.5%). Glycylglycine/NaOH buffer 10 mM, pH 8, CaCl, 1 mM was then added up to a final volume comprised between 0.5 and 20 ml. The reaction was started by adding purified enzyme to a final concentration of 1 fig of protein per ml for pancreatic phospholipase and 0.5 r.tgfor sheep phospholipase. The reaction tubes were incubated at 37°C with constant shaking. Aliquots were treated with the extraction mixture to stop the reaction after different incubation times. Extraction and analysis of futty acids The fatty acids liberated during the enzymatic hydrolysis were extracted according to [ 121 by diisopropyl ether/methanol (99 : 1) (4 ml of solvent per ml incubation mixture + 400 mg magnesium silicate) 75% of the organic phase was evaporated under nitrogen, dessicated under vacuum and the free fatty acids were methylated in presence of freshly prepared diazomethane [ 13,141. After evaporating the solvent, the methyl esters were solub~ised in a small volume of benzene and injected in a Perking-Elmer 990 Gas Chromatograph equipped with a glass column (chromadsorb W. AW-DMCS 80-100 mesh, diethylglycerolsuccinate 5%) under nitrogen flow. Initial and final temperatures were 150 and 210°C respectively (8 min initial temperature, 6 min final temperature, 8°C per min between initial and final temperatures), Each peak was identified by calibration with pure methyl ester’s standards: Cl2 : 0, Cl4 : 0, Cl6 : 0, Cl6 : 1, Cl7 : 0. Cl8 : 0, Cl8 : 1, Cl8 : 2, Cl8 : 3, c20 : 0, c20 : 1, c20 : 2, c20 : 3, C20 : 4, C22 : 1, C24 : 1. Margaric acid (Cl7 : 0) added to the extraction solvent (75 nmol per ml) was used as an internal standard for the quantitation of the peaks corresponding to the individual fatty acids on the chromatograms. The background was obtained by extracting and chromato~aphying the incubation mixture according to the same procedure before adding the enzyme. Absolute rates of hydrolysis are reported as nmol of liberated fatty acids per min per kg protein. Relative or normalized rates of hydrolysis for each class of 2-acyl chains are expressed as the measured rate divided by the corresponding fatty acid content of the phospholipid in moles per cent. Hydrogenation of fatty acid methyl esters was performed in ethyl acetate under hydrogen flow with palladium as catalyst. Sheep erythrocyte phospholipids were extracted, separated by two-dimensional thinlayer chromatography and quantified as previously described [ 51. Results 2-Acyl chains in egg lecithin and eggphosphatidylethanolamine The fatty acid composition of pure egg lecithin and egg phosphatidylethanolamine reported in the literature [lS] is varied enough to make them suitable substrates for our studies. Figs. 1 and 2 illustrate the pattern of liberated fatty acids upon hydrolysis of both substrates by the sheep erythrocyte phospholipase AZ or by the pancreatic enzyme respectively. During the 120 min incubation time, pancreatic phospholipase had hydrolyzed 80--90% of the substrate while the erythrocyte enzyme hydrolyzed 60% of lecithin and 7% of phosphatidylethanolamine. It is clear that the major classes of fatty acids present in
369
1
.
c. :* .:: :: .*
i 4 :: .. .* ::
:: ::
A
8
4
-
6
15 210
r.
time
1 . .
8 Olmin temperature
5
10
(mint
150 *---j
(%I
Fig. 1. Gas liquid chromatography analysis of phosphatjdylcholine hydrolysis by sheep erythrocyte phospholipase (A) and pancreatic phospholipase (B). The enzymes, as described in Material and Methods, were incubated 120 min in presence of the substrate. Upon extraction and methylation of the fatty acids they or hydrogenated and then injected (- - - - - -). In each were injected into the gas chromatograph: ( -) case, background was ubstracted according to the chromatographic recording of the extracted reaction mixture without enzymatic action.
position 2 of egg lecithin or egg phosphatidylethanolamine are oleic, hnoleie, arachidonic, and two unidentified long chain fatty acids that we label X1 and X,. Further characterization of the peaks can be obtained by hydrogenating the mixture of methyl esters. We see that upon hydrogenation, oleate, linoleate and arachidonate peaks shift to the respective positions of stearate and arachidate as expected. Peaks X, and X, disappear while one peak the are of which is roughfy the sum of X, and X, appears at the level of behenate (C22 : 0). This strongly suggests that both peaks correspond to unsaturated fatty acid of 22 carbon atoms length chain in good agreement with the reported composition of egg phospholipids [ 151. hydrolysis kinetics of egg lecithin and egg phosp~atidylet~a~o~a~~ne according to the 2-acyl chain In the chromatograms of Figs. 1 and 2 one can see that the relative amount
4
I 16
10
(I
0
tlme(mln) Pig. 2. Gas liquid chromatographic amlvsjs of phosphatidylethanolamine hydrolysis by sheep erythrocyte phospholipase (A) and pancreatic phospholipase (B). The enzymes. as described in Material and Methods, were incubated 120 min in Presence of the substrate. Upon extraction and methylation of the fatty acids (---they were injected into the gas chromatograph: ) or hydrogenated and then injected (- - - -. -). In each case, background was substracted according to the chromatographic recording of the extracted reaction mixture without enzymatic action,
of the C22 unsaturated fatty acids seems higher when the hydrolysis of lecithin or phosphatidyI-ethanolamine is catalyzed by the sheep phospholipase. This difference suggests a preference of this enzyme for the egg phospholipids carrying a long chain unsaturated fatty acid in position 2 and can be specified by measuring the rate of liberation of each fatty acid during the enzymatic hydrolysis of the corresponding phospholipid. Figs. 3 and 4 illustrate such kinetics with phosphatidylcholine and phosphatidylethanolamine as substrates. As the different fatty acids are present in different amounts for one class of phospholipid, absolute amounts of reaction products are not adequate for expressing
271
c20:4
50
.
time
(min)
100
Fig. 3. Kinetics of phosphatidylcholine hydrolysis by sheep erythrocyte phospholipase (A) CJI pancreatic phospholipase (B). Absolute quantities of the fatty acids liberated for each species were determined by measuring the area of the corresponding peak and normalizing the results according to the area of the internal standard margaric acid (Cl7 : 0).
any preference of the enzyme in relation with the 2-acyl chain. In order to take into account the composition of the phospholipid used as a substrate, we have divided, for every incubation time, the number of nanomoles of fatty acid produced per ml by its contribution (in per cent) to the phospholipid composition determined after total hydrolysis by pancreatic phospholipase. The composition of both phospholipids and the normalized rate of hydrolysis for each fatty acid and each enzyme is given in Table I. The figures confirm reported lack of specificity for the pancreatic phospholipase [17] which liberates all the fatty acids at the same normalized rate. Oleic and linoleic acids are also liberated by the ruminant phospholipase at comparatively the same rate. But arachi-
212
50
time
(min)
Fig. 4. Kinetics of phosphatidylethanolamine pancreatic phospholipase (B).
100
hydrolysis
by
sheep
erythrocyte
phospholipase
(A)
or
donic acid seems to be liberated by this enzyme at a faster rate, and when coming to the 22 carbon atoms fatty acids, a clear preference is shown: they are liberated ten times faster than the 18 carbon atoms chains. This preference is observed equally well with phosphytidylcholine as with phosphatidylethanolamine. Rate of hydrolysis according to the polar head of the substrate Overal rate of hydrolysis for one phospholipid class can be obtained after gas liquid chromatography analysis of the time course reaction by adding up the
273 TABLE
I
NORMALIZED OF
THE
RATES
OF
HYDROLYSIS
ACCORDING
TO
THE
SnP-ACYL
CHAIN
COMPOSITION
SUBSTRATE
Phospholipid
2-acyl
chain
COmpO-
Normalized
sition
fig-‘.
velocities
nm
. mine1 .
(mol%)-’
(mol%) Pancreatic
Sheep
phospholipase
phospholipase
(X102) Egg
phosphatidylcholine
Egg phosphatidylethanolamine
Cl8
1
56.3
5.6
5.3
lo”
Cl8
2
28.2
5.6
4.2
lo-2
c20
4
8.0
5.6
c22
Xl
3.5
5.6
42.0
lo-2
c22
x2
3.5
5.6
42.0
l(T2
Cl8
1
30.0
9.0
2.7
lcr-3
Cl8
2
14.0
9.0
2.8
lo-3
c20
4
36.0
9.0
5.0
lo-3
c22
Xl
10.8
9.0
33.0
lo-3
c22
x2
9.5
9.0
33.0
l(r3
6.5
lU*
partial rates of hydrolysis of all 2-acyl chains. Such rates are summarized in Table II for different phospholipids treated by pancreatic or by sheep erythrocyte phospholipase. Although with our incubations conditions the specific activity of pancreatic phospholipase is lower than the usually reported activities [ 161, its activity is particularly high on phosphatidylserine, lower on phosphatidylglycerol and phosphatidylethanolamine and still lower on phosphatidylcholine. Conversely, sheep erythrocyte phospholipase activity is about ten times more active on phosphatidylcholine than on phosphatidylethanolamine or phosphatidylglycerol. No activity on phosphatidylserine was observed after three hours’ incubation time. Phosphatidylserine is present in sizable amounts in sheep erythrocyte membrane. This is shown in Table III, where we report the phospholipid composition of the stroma used for the purification of sheep erythrocyte phospholipase. The figures are in good agreement with reports of others [5].
TABLE
II
HYDROLYSIS
OF
DIFFERENT
Phospholipid
SUBSTRATES Velocity
BY (nmols
SHEEP
. min-’
AND
.
pg-‘)
Pancreatic
Sheep
phospholipase
phospholipaae
Phosphatidylcholine
5.5
8.0
Phosphatidylethanolamine
9.0
1.0
Phosphatidylglycerol
8.0
1.3
20.0
0.0
Phosphatidylserine
PANCREATIC
erythrocyte
PHOSPHOLIPASES
214
TABLE
III
PHOSPHOLIPID 400
COMPOSITION
pg phospholipid/mg
OF
SHEEP
ERYTHROCYTE
MEMBRANES
protein
Sphingomyelin
60.0
Phosphatidylethanolamine
24.0
Phosphatidylserine
10.0
LYSO phosphatidylethanolamine
3.5
Phosphatidylcholine
>0.05
Phosphatidylinositol
0.5
Others
2.0
Discussion We have shown that pure sheep erythrocyte phospholipase hydrolyzes phosphatidylcholine at a faster rate than phosphatidylethanolamine, phosphatidylglycerol or phosphatidylserine. At pH 8 which is the pH of the incubating medium, all these phospholipids bear a net negative electric charge [18], specially phosphatidylserine on which the enzyme is not active. Our observations with the purified enzyme confirm our previously reported preference of membrane-bound hydrolytic activity for phosphatidylcholine. This preference of the enzyme for the uncharged choline moiety considerably distinguishes it form other phospholipases. Pancreatic phospholipase, the well characterized mammal enzyme, has been reported to be specific for acidic phospholipids, and it hydrolyzes phosphatidylserine or phosphatidylethanolamine faster than phosphatidylcholine, as shown in our report and with synthetic lipids [ 171. On the other hand, no clear preference has been reported for phosphatidylethanolamine or phosphatidylcholine with the enzyme purified from different varieties of bee venom [ 31. The enzyme from rat liver mitochondria is similar to our enzyme in that it is membrane-located, although it can be purified in soluble form without the help of detergents [19,20]. But this enzyme hydrolyzes preferentially phosphatidylserine and phosphatidylethanolamine and more slowly phosphatidic acid and phosphatidylcholine, endogenous phosphatidylethanolamine being preferred to other phospholipids [21]. Snake venom phospholipase seems to be more similar to sheep erythrocyte phospholipase as it also seems to prefer phosphatidylcholine to other phospholipids [ 211. In addition, as far as the specificity for the liberated acyl chain is concerned, pancreatic phospholipase has been reported to be insensitive to chain length or degree of unsaturation with synthetic phospholipids [17], and we confirm this result with egg lecithin phospholipids. The snake enzyme preferentially liberates saturated fatty acid [22], while conflicting reports are published for the mitochondrial enzyme [19,20]. A recent report on phospholipase A2 in human fetal membranes suggests preferential breakdown of arachidonic containing phosphatidylethanolamine [23]. Nevertheless, the faster rates of hydrolysis we observe for phospholipids carrying 22 carbon chain unsaturated fatty acids in position 2 seems to be a peculiarity of the sheep erythrocyte phospholipase.
275
We have to keep in mind, however, that the environment of the purified enzyme during the assay is very different from its natural environment and that the variations in rate of hydrolysis could be due to detergent effects or to variations in the micellar state of the phospholipids according to their polar head. In this respect the length of the 2-acyl chain and the degree of its unsaturation are also relevant factors which chould affect the micellar packing of some phospholipid molecules in a way proper to facilitate its attack by the ruminant phospholipase. On the other hand, the longest carbon chain which has been reported to be in the phospholipids of ruminant serum or erythrocyte membranes is that of arachidonic acid (C20 : 4) [8], while in human serum and red cell membranes poly-unsaturated C22 chains have been reported in sizeable amounts [8]. This difference points out to the possibility of the function of ruminant enzyme being to hydrolyze preferentially the phosphatidylcholine species carrying a C22 fatty acid in position 2 or a related essential fatty acid. References Slotboom,
A.A.J.,
Joubert.
F.J.
Shipolini.
R.A.
Kramer,
R..
Zwaal,
Pieterson,
(1975) (1971)
Jungi,
R.F.A.,
W.A..
Biochim. Eur.
B. and
Volwerk.
Biophys. J. Biochem.
Zahler,
Fliickiger,
379,
20,
de Haas,
Biochim.
S.
G.H.
(1976)
Lipids
1, 99-107
329-344
459-468
P. (1974) Moser,
R..
J.J. and
Acta
and
Biophys.
Zahler,
P.
Acta
(1974)
373,
404.-415
Biochim.
Biophys.
Acta
373.
416-
424 6
Kramer,
R.M.,
Wuthrich,
C.,
Bollier.
Ch.,
Allegrini.
P.R.
and
Zabler.
P. (1978)
Biochim.
Biophys.
Acta
507,381-384 7
Mulder,
8
White.
E.,
R.M.C.. 9
de Gier,
D.A.
J. and
(1973)
eds.)
Borochov,
Vol.
H..
Form
van Deenen, and
3, PP. 441482,
Zahler,
P..
L.L.M.
Function
of
BBA
Wilbrandt,
(1962)
library,
W.
Biochim.
Phospholipids,
and
Biophys.
(Ansell,
Elsevier,
Shinitzky,
G.B.,
Acta
59.
502-504
Hawthorne,
J.N.
and
Dawson,
Amsterdam M. (1977)
Biochim.
Biophys.
Acta
470,
382-
388 10
Colley,
C.M.,
Zwaal.
R.F.A.,
Roelofsen.
B. and
van
Deenen,
Acta.
550.
L.L.M.
(1973)
Biochim.
Biophys.
Acta
307.74-82 11
Frei,
12
Smith,
E. and J.B.,
13
Kate,
M.
Zahler.
P. (1978)
Silver, (1972)
M.J. in
Biochim.
and Webster.
Techniques
Biophys. G.R.
of
(1973)
Biochem.
Lipidology,
PP.
450-463 J. 131.
526--527,
615418
North-Holland
Publishing
Company,
Amsterdam 14
Wieland, Co.,
H.
(1952)
J.C.
(1959)
in
Die
Praxis
des
organ&hen
Chemikers,
pp.
234.-238,
Walter
de
Gmyter
and
Berlin
15
Hawke,
16
de
Haas.
Acta
G.H.,
159,
Biochem.
Postema.
588-592
Wenhuizen.
W.N.
and
van
Deenen,
L.L.M.
(1968)
Biocbim.
Biophys.
103-117
17
van Deenen,
18
Bangham.
A.D.
19
Nachbaur.
J., Colbeau,
20
Waite,
21
Hanaban,
22
Smith,
23
Okazaki, 432-438
J. 71.
N.M.,
L.L.M.
M. and
S&on.
D.J.. A.D.. T.,
and de Haas,
(1968)
Progr. A.
Brockerhoff.
GUI, Okita.
and
P. (1971)
S. and J.R.,
G.H.
(1963)
Biochim.
Biophys.
Mol.
Biol.
Vignais,
P.M.
(1972)
Biochemistry H. and
Thomson, MacDonald,
Barron. R.H.S. P.C.
18,
Biophys.
Acta
70,
538-553
31
Biochim.
Biophys.
Acta
274,
426-446
10,2377-2383 E.J. (1972) and
(1960)
J. Biol.
Biochim. Johnston,
Chem.
Biophys. M. (1978)
235, Acta Am.
1917-1923 289,
147-157
J. Obstet.
GYIHXO~.
130
(4).
Biochimica et Biophysics Acta, 573 (1979) 266-275 0 Elsevier/North-Holland Biomedical Press
BBA 57369
PURIFIED PHOSPHOLIPASE MEMBRANE
A2 FROM S3EEP
PREFERENTIAL HYDROLYSIS 2-ACYL CHAINS
JUAN JIMENO-ABENDA~O Institute of Biochemistry,
ACCORDING
ERYTHROCYTE
TO POLAR GROUPS AND
and PETER ZAHLER *
University of Berne, Freiestrasse 3, Berne (Switzerland)
(Received November 24th, 1978)
Key words: Phosphol~pase A,; Polar group hydrolysis; Acyl chain; ~Pur~f~cat~on,Sheep ery throcy te membrane)
Summary Hydrolysis of natural phospholipids by pure erythrocyte membrane phospholipase AZ was compared to the reaction catalyzed by the soluble pancreatic enzyme. Fatty acids liberated during both types of reaction were quantitatively analyzed by gas liquid chromatography. We confirm for the pancreatic enzyme lack of specificity with respect to the sn-2 acyl chain of the phospholipids and preference for negatively charged polar head groups. Conversely, the membranous enzyme preferentially attacks uncharged phospholipids and within one class of phospholipid preferentially splits long chain unsaturated fatty acids in the sn-2 position. The significance of such differences between pancreatic and sheep erythrocyte enzyme is discussed in relation to the possible physiological role of the latter enzyme.
Introduction Phospholipase AZ (EC 3.1.1.4) activities have been found in several organisms and tissues. The ones characterized best are the soluble enzymes purified from pancreas f 1] or snake and bee venoms [2,3]. We have reported the existence of a membr~e-bound phospholip~e A, in ruminant erythrocytes [4, 51 which has been purified to electrophoretic homogeneity [6]. The function of ruminant phospholipase A2 is not clearly understood as is the case for most * To whom
correspondence shouldbe addressed.
267
of the membrane-bound phospholipases. One conspicuous fact about the structure of ruminant erythrocyte membrane is its low phosphatidylcholine content. In lipid extracts of sheep red cells, only trace amounts of this phospholipid are found beside a relatively high amount of sphingomyelin. This is in contrast to the majority of mammalian red cell membranes and other sheep tissues including serum and bone marrow, where phosphatidylcholine is a major component of membrane phospholipids [ 7,8]. As both facts, low phosphatidylcholine content and presence of phospholipase A, activity, are peculiar to ruminant red cells, one is tempted to connect them in search for the enzyme function. Therefore, we have previously proposed [4] that the role of the ruminant phospholipase might be to inhibit the increase of phosphatidylcholine concentration in the red cell membrane. We have indeed reported several lines of evidence showing that sheep red cells increase their phosphatidylcholine content when incubated with sheep or human serum provided its phospholipase is inhibited or inactivated [ 59 1. Also we have been able to show that the zwitterionic phosphatidylcholine is the preferred substrate in contrast to most of the soluble enzymes preferring the negatively charged phospholipids [4]. However, also phosphatidylethanolamine is split, but this is of little consequence due to the fact that the phospholipase is located on the outside of the sheep red cell membrane [ll], whereas most of the phosphatidyleth~olamine is in the inner half side of the lipid bilayer [lo]. In order to know more about the possible function of this membrane-bound enzyme, we have further studied in detail the problem of specificity, e.g. class of phospholipid and length and unsaturation of the split fatty acids in comparison with the pancreatic phospholipase, which was shown to split off the fatty acids independent of their length and unsaturation [ 17 1. Materials and Methods Products Crystalline porcine pancreatic phospllolipase was a generous gift of Doctor A.J. Slotboom. Egg phosphatidylcholine, egg phosphatidyleth~olam~e, phosphatidylglycerol and phosphatidylserine were purchased from Lipid Products, South Nutfield NR Redhill SY, U.K. Margaric acid, margaric methyl ester and magnesium silicate were purchased from Merck Laboratories. Other standards for gas chromatography were provided by Supelco Inc., Bellafonte, PA, U.S.A. All the other chemicals were analytical grade. Phospholipase purification Sheep red cell membranes were prepared as previously described [5] from carefully washed erythrocytes. After lauryl sulfate solubilization of the stroma, the enzyme was purified more than 1000 times by gel filtration and affinity chromatography as reported in Ref. 6. Phosphalipase assay Enough phospholipid (egg lecithin, egg phosphatidylethanolamine, bovine phosphatidylserine, bovine phosphatidylglycerol) to reach a final concentration of 700 nmol per ml was dried under nitrogen, then suspended in 10% sodium
268
cholate (final concentration 0.5%). Glycylglycine/NaOH buffer 10 mM, pH 8, CaCl, 1 mM was then added up to a final volume comprised between 0.5 and 20 ml. The reaction was started by adding purified enzyme to a final concentration of 1 fig of protein per ml for pancreatic phospholipase and 0.5 r.tgfor sheep phospholipase. The reaction tubes were incubated at 37°C with constant shaking. Aliquots were treated with the extraction mixture to stop the reaction after different incubation times. Extraction and analysis of futty acids The fatty acids liberated during the enzymatic hydrolysis were extracted according to [ 121 by diisopropyl ether/methanol (99 : 1) (4 ml of solvent per ml incubation mixture + 400 mg magnesium silicate) 75% of the organic phase was evaporated under nitrogen, dessicated under vacuum and the free fatty acids were methylated in presence of freshly prepared diazomethane [ 13,141. After evaporating the solvent, the methyl esters were solub~ised in a small volume of benzene and injected in a Perking-Elmer 990 Gas Chromatograph equipped with a glass column (chromadsorb W. AW-DMCS 80-100 mesh, diethylglycerolsuccinate 5%) under nitrogen flow. Initial and final temperatures were 150 and 210°C respectively (8 min initial temperature, 6 min final temperature, 8°C per min between initial and final temperatures), Each peak was identified by calibration with pure methyl ester’s standards: Cl2 : 0, Cl4 : 0, Cl6 : 0, Cl6 : 1, Cl7 : 0. Cl8 : 0, Cl8 : 1, Cl8 : 2, Cl8 : 3, c20 : 0, c20 : 1, c20 : 2, c20 : 3, C20 : 4, C22 : 1, C24 : 1. Margaric acid (Cl7 : 0) added to the extraction solvent (75 nmol per ml) was used as an internal standard for the quantitation of the peaks corresponding to the individual fatty acids on the chromatograms. The background was obtained by extracting and chromato~aphying the incubation mixture according to the same procedure before adding the enzyme. Absolute rates of hydrolysis are reported as nmol of liberated fatty acids per min per kg protein. Relative or normalized rates of hydrolysis for each class of 2-acyl chains are expressed as the measured rate divided by the corresponding fatty acid content of the phospholipid in moles per cent. Hydrogenation of fatty acid methyl esters was performed in ethyl acetate under hydrogen flow with palladium as catalyst. Sheep erythrocyte phospholipids were extracted, separated by two-dimensional thinlayer chromatography and quantified as previously described [ 51. Results 2-Acyl chains in egg lecithin and eggphosphatidylethanolamine The fatty acid composition of pure egg lecithin and egg phosphatidylethanolamine reported in the literature [lS] is varied enough to make them suitable substrates for our studies. Figs. 1 and 2 illustrate the pattern of liberated fatty acids upon hydrolysis of both substrates by the sheep erythrocyte phospholipase AZ or by the pancreatic enzyme respectively. During the 120 min incubation time, pancreatic phospholipase had hydrolyzed 80--90% of the substrate while the erythrocyte enzyme hydrolyzed 60% of lecithin and 7% of phosphatidylethanolamine. It is clear that the major classes of fatty acids present in
369
1
.
c. :* .:: :: .*
i 4 :: .. .* ::
:: ::
A
8
4
-
6
15 210
r.
time
1 . .
8 Olmin temperature
5
10
(mint
150 *---j
(%I
Fig. 1. Gas liquid chromatography analysis of phosphatjdylcholine hydrolysis by sheep erythrocyte phospholipase (A) and pancreatic phospholipase (B). The enzymes, as described in Material and Methods, were incubated 120 min in presence of the substrate. Upon extraction and methylation of the fatty acids they or hydrogenated and then injected (- - - - - -). In each were injected into the gas chromatograph: ( -) case, background was ubstracted according to the chromatographic recording of the extracted reaction mixture without enzymatic action.
position 2 of egg lecithin or egg phosphatidylethanolamine are oleic, hnoleie, arachidonic, and two unidentified long chain fatty acids that we label X1 and X,. Further characterization of the peaks can be obtained by hydrogenating the mixture of methyl esters. We see that upon hydrogenation, oleate, linoleate and arachidonate peaks shift to the respective positions of stearate and arachidate as expected. Peaks X, and X, disappear while one peak the are of which is roughfy the sum of X, and X, appears at the level of behenate (C22 : 0). This strongly suggests that both peaks correspond to unsaturated fatty acid of 22 carbon atoms length chain in good agreement with the reported composition of egg phospholipids [ 151. hydrolysis kinetics of egg lecithin and egg phosp~atidylet~a~o~a~~ne according to the 2-acyl chain In the chromatograms of Figs. 1 and 2 one can see that the relative amount
4
I 16
10
(I
0
tlme(mln) Pig. 2. Gas liquid chromatographic amlvsjs of phosphatidylethanolamine hydrolysis by sheep erythrocyte phospholipase (A) and pancreatic phospholipase (B). The enzymes. as described in Material and Methods, were incubated 120 min in Presence of the substrate. Upon extraction and methylation of the fatty acids (---they were injected into the gas chromatograph: ) or hydrogenated and then injected (- - - -. -). In each case, background was substracted according to the chromatographic recording of the extracted reaction mixture without enzymatic action,
of the C22 unsaturated fatty acids seems higher when the hydrolysis of lecithin or phosphatidyI-ethanolamine is catalyzed by the sheep phospholipase. This difference suggests a preference of this enzyme for the egg phospholipids carrying a long chain unsaturated fatty acid in position 2 and can be specified by measuring the rate of liberation of each fatty acid during the enzymatic hydrolysis of the corresponding phospholipid. Figs. 3 and 4 illustrate such kinetics with phosphatidylcholine and phosphatidylethanolamine as substrates. As the different fatty acids are present in different amounts for one class of phospholipid, absolute amounts of reaction products are not adequate for expressing
271
c20:4
50
.
time
(min)
100
Fig. 3. Kinetics of phosphatidylcholine hydrolysis by sheep erythrocyte phospholipase (A) CJI pancreatic phospholipase (B). Absolute quantities of the fatty acids liberated for each species were determined by measuring the area of the corresponding peak and normalizing the results according to the area of the internal standard margaric acid (Cl7 : 0).
any preference of the enzyme in relation with the 2-acyl chain. In order to take into account the composition of the phospholipid used as a substrate, we have divided, for every incubation time, the number of nanomoles of fatty acid produced per ml by its contribution (in per cent) to the phospholipid composition determined after total hydrolysis by pancreatic phospholipase. The composition of both phospholipids and the normalized rate of hydrolysis for each fatty acid and each enzyme is given in Table I. The figures confirm reported lack of specificity for the pancreatic phospholipase [17] which liberates all the fatty acids at the same normalized rate. Oleic and linoleic acids are also liberated by the ruminant phospholipase at comparatively the same rate. But arachi-
212
50
time
(min)
Fig. 4. Kinetics of phosphatidylethanolamine pancreatic phospholipase (B).
100
hydrolysis
by
sheep
erythrocyte
phospholipase
(A)
or
donic acid seems to be liberated by this enzyme at a faster rate, and when coming to the 22 carbon atoms fatty acids, a clear preference is shown: they are liberated ten times faster than the 18 carbon atoms chains. This preference is observed equally well with phosphytidylcholine as with phosphatidylethanolamine. Rate of hydrolysis according to the polar head of the substrate Overal rate of hydrolysis for one phospholipid class can be obtained after gas liquid chromatography analysis of the time course reaction by adding up the
273 TABLE
I
NORMALIZED OF
THE
RATES
OF
HYDROLYSIS
ACCORDING
TO
THE
SnP-ACYL
CHAIN
COMPOSITION
SUBSTRATE
Phospholipid
2-acyl
chain
COmpO-
Normalized
sition
fig-‘.
velocities
nm
. mine1 .
(mol%)-’
(mol%) Pancreatic
Sheep
phospholipase
phospholipase
(X102) Egg
phosphatidylcholine
Egg phosphatidylethanolamine
Cl8
1
56.3
5.6
5.3
lo”
Cl8
2
28.2
5.6
4.2
lo-2
c20
4
8.0
5.6
c22
Xl
3.5
5.6
42.0
lo-2
c22
x2
3.5
5.6
42.0
l(T2
Cl8
1
30.0
9.0
2.7
lcr-3
Cl8
2
14.0
9.0
2.8
lo-3
c20
4
36.0
9.0
5.0
lo-3
c22
Xl
10.8
9.0
33.0
lo-3
c22
x2
9.5
9.0
33.0
l(r3
6.5
lU*
partial rates of hydrolysis of all 2-acyl chains. Such rates are summarized in Table II for different phospholipids treated by pancreatic or by sheep erythrocyte phospholipase. Although with our incubations conditions the specific activity of pancreatic phospholipase is lower than the usually reported activities [ 161, its activity is particularly high on phosphatidylserine, lower on phosphatidylglycerol and phosphatidylethanolamine and still lower on phosphatidylcholine. Conversely, sheep erythrocyte phospholipase activity is about ten times more active on phosphatidylcholine than on phosphatidylethanolamine or phosphatidylglycerol. No activity on phosphatidylserine was observed after three hours’ incubation time. Phosphatidylserine is present in sizable amounts in sheep erythrocyte membrane. This is shown in Table III, where we report the phospholipid composition of the stroma used for the purification of sheep erythrocyte phospholipase. The figures are in good agreement with reports of others [5].
TABLE
II
HYDROLYSIS
OF
DIFFERENT
Phospholipid
SUBSTRATES Velocity
BY (nmols
SHEEP
. min-’
AND
.
pg-‘)
Pancreatic
Sheep
phospholipase
phospholipaae
Phosphatidylcholine
5.5
8.0
Phosphatidylethanolamine
9.0
1.0
Phosphatidylglycerol
8.0
1.3
20.0
0.0
Phosphatidylserine
PANCREATIC
erythrocyte
PHOSPHOLIPASES
214
TABLE
III
PHOSPHOLIPID 400
COMPOSITION
pg phospholipid/mg
OF
SHEEP
ERYTHROCYTE
MEMBRANES
protein
Sphingomyelin
60.0
Phosphatidylethanolamine
24.0
Phosphatidylserine
10.0
LYSO phosphatidylethanolamine
3.5
Phosphatidylcholine
>0.05
Phosphatidylinositol
0.5
Others
2.0
Discussion We have shown that pure sheep erythrocyte phospholipase hydrolyzes phosphatidylcholine at a faster rate than phosphatidylethanolamine, phosphatidylglycerol or phosphatidylserine. At pH 8 which is the pH of the incubating medium, all these phospholipids bear a net negative electric charge [18], specially phosphatidylserine on which the enzyme is not active. Our observations with the purified enzyme confirm our previously reported preference of membrane-bound hydrolytic activity for phosphatidylcholine. This preference of the enzyme for the uncharged choline moiety considerably distinguishes it form other phospholipases. Pancreatic phospholipase, the well characterized mammal enzyme, has been reported to be specific for acidic phospholipids, and it hydrolyzes phosphatidylserine or phosphatidylethanolamine faster than phosphatidylcholine, as shown in our report and with synthetic lipids [ 171. On the other hand, no clear preference has been reported for phosphatidylethanolamine or phosphatidylcholine with the enzyme purified from different varieties of bee venom [ 31. The enzyme from rat liver mitochondria is similar to our enzyme in that it is membrane-located, although it can be purified in soluble form without the help of detergents [19,20]. But this enzyme hydrolyzes preferentially phosphatidylserine and phosphatidylethanolamine and more slowly phosphatidic acid and phosphatidylcholine, endogenous phosphatidylethanolamine being preferred to other phospholipids [21]. Snake venom phospholipase seems to be more similar to sheep erythrocyte phospholipase as it also seems to prefer phosphatidylcholine to other phospholipids [ 211. In addition, as far as the specificity for the liberated acyl chain is concerned, pancreatic phospholipase has been reported to be insensitive to chain length or degree of unsaturation with synthetic phospholipids [17], and we confirm this result with egg lecithin phospholipids. The snake enzyme preferentially liberates saturated fatty acid [22], while conflicting reports are published for the mitochondrial enzyme [19,20]. A recent report on phospholipase A2 in human fetal membranes suggests preferential breakdown of arachidonic containing phosphatidylethanolamine [23]. Nevertheless, the faster rates of hydrolysis we observe for phospholipids carrying 22 carbon chain unsaturated fatty acids in position 2 seems to be a peculiarity of the sheep erythrocyte phospholipase.
275
We have to keep in mind, however, that the environment of the purified enzyme during the assay is very different from its natural environment and that the variations in rate of hydrolysis could be due to detergent effects or to variations in the micellar state of the phospholipids according to their polar head. In this respect the length of the 2-acyl chain and the degree of its unsaturation are also relevant factors which chould affect the micellar packing of some phospholipid molecules in a way proper to facilitate its attack by the ruminant phospholipase. On the other hand, the longest carbon chain which has been reported to be in the phospholipids of ruminant serum or erythrocyte membranes is that of arachidonic acid (C20 : 4) [8], while in human serum and red cell membranes poly-unsaturated C22 chains have been reported in sizeable amounts [8]. This difference points out to the possibility of the function of ruminant enzyme being to hydrolyze preferentially the phosphatidylcholine species carrying a C22 fatty acid in position 2 or a related essential fatty acid. References Slotboom,
A.A.J.,
Joubert.
F.J.
Shipolini.
R.A.
Kramer,
R..
Zwaal,
Pieterson,
(1975) (1971)
Jungi,
R.F.A.,
W.A..
Biochim. Eur.
B. and
Volwerk.
Biophys. J. Biochem.
Zahler,
Fliickiger,
379,
20,
de Haas,
Biochim.
S.
G.H.
(1976)
Lipids
1, 99-107
329-344
459-468
P. (1974) Moser,
R..
J.J. and
Acta
and
Biophys.
Zahler,
P.
Acta
(1974)
373,
404.-415
Biochim.
Biophys.
Acta
373.
416-
424 6
Kramer,
R.M.,
Wuthrich,
C.,
Bollier.
Ch.,
Allegrini.
P.R.
and
Zabler.
P. (1978)
Biochim.
Biophys.
Acta
507,381-384 7
Mulder,
8
White.
E.,
R.M.C.. 9
de Gier,
D.A.
J. and
(1973)
eds.)
Borochov,
Vol.
H..
Form
van Deenen, and
3, PP. 441482,
Zahler,
P..
L.L.M.
Function
of
BBA
Wilbrandt,
(1962)
library,
W.
Biochim.
Phospholipids,
and
Biophys.
(Ansell,
Elsevier,
Shinitzky,
G.B.,
Acta
59.
502-504
Hawthorne,
J.N.
and
Dawson,
Amsterdam M. (1977)
Biochim.
Biophys.
Acta
470,
382-
388 10
Colley,
C.M.,
Zwaal.
R.F.A.,
Roelofsen.
B. and
van
Deenen,
Acta.
550.
L.L.M.
(1973)
Biochim.
Biophys.
Acta
307.74-82 11
Frei,
12
Smith,
E. and J.B.,
13
Kate,
M.
Zahler.
P. (1978)
Silver, (1972)
M.J. in
Biochim.
and Webster.
Techniques
Biophys. G.R.
of
(1973)
Biochem.
Lipidology,
PP.
450-463 J. 131.
526--527,
615418
North-Holland
Publishing
Company,
Amsterdam 14
Wieland, Co.,
H.
(1952)
J.C.
(1959)
in
Die
Praxis
des
organ&hen
Chemikers,
pp.
234.-238,
Walter
de
Gmyter
and
Berlin
15
Hawke,
16
de
Haas.
Acta
G.H.,
159,
Biochem.
Postema.
588-592
Wenhuizen.
W.N.
and
van
Deenen,
L.L.M.
(1968)
Biocbim.
Biophys.
103-117
17
van Deenen,
18
Bangham.
A.D.
19
Nachbaur.
J., Colbeau,
20
Waite,
21
Hanaban,
22
Smith,
23
Okazaki, 432-438
J. 71.
N.M.,
L.L.M.
M. and
S&on.
D.J.. A.D.. T.,
and de Haas,
(1968)
Progr. A.
Brockerhoff.
GUI, Okita.
and
P. (1971)
S. and J.R.,
G.H.
(1963)
Biochim.
Biophys.
Mol.
Biol.
Vignais,
P.M.
(1972)
Biochemistry H. and
Thomson, MacDonald,
Barron. R.H.S. P.C.
18,
Biophys.
Acta
70,
538-553
31
Biochim.
Biophys.
Acta
274,
426-446
10,2377-2383 E.J. (1972) and
(1960)
J. Biol.
Biochim. Johnston,
Chem.
Biophys. M. (1978)
235, Acta Am.
1917-1923 289,
147-157
J. Obstet.
GYIHXO~.
130
(4).
