183

et Biophysics Acta, 424 (1976) 183-194 @ Elsevier Scientific Publishing Company, Amsterdam - Printed in The Netherlands Biochimica

BBA 56717

PHOSPHOLIPASES Al and A2 IN BOVINE THYROID

M. DE WOLF, A. LAGROU, H.J. HILDERSON and W. DIERICK RUCA-Laboratory

for Human

Biochemistry,

University

ofAntwerp

(Belgium)

(Received June lOth, 1975)

Summary

In both supernatant and sediment of thyroid tissue homogenate phospholipase and lysophospholipase activities were demonstrated. In the supematant, using l-acyl-2[ 1-l 4 C] linoleoyl-sn-glycero-3-phosphorocholine in the presence of sodium taurocholate, phospholipase Ai activity with pH optima at 3.6 and 4.8 and phospholipase AZ activity with pH optima at 3.6 and 5.7 were found. The sediment showed mainly phospholipase AZ activity with a pH optimum at pH 6.5. Lysophospholipase activity (optimum pH F-8), using l[ 9,10- 3H] stearyl-sn-glycero-3-phosphorocholine as a substrate was present in both supernatant and sediment. Enzyme assays performed on subcellular fractions suggest the soluble phospholipases to be of lysosomal origin and the solubilized phospholipase AZ activity of homogenate sediment to be of microsomal origin. Incubations with 3H- ’ 4 C mixed labelled phosphatidylcholine further confirmed the above observations.

Introduction

1-acylhydrolase EC The presence of phospholipase A, (phosphatidate 3.1.1.32) and phospholipase AZ (phosphatide 2-acylhydrolase EC 3.1.1.4) activities have been reported in various organs and tissues using both endogenous and exogenous phospholipids as substrate [l-17]. Phospholipases could play an important role in both normal and pathological thyroid glands. Levis et al. [18] did find relatively high amounts of lysophosphatidylethanolamine in cold nodules from human thyroid tissue. Changes in endogenous phospholipase activity were thought to be important in this disorder. Haye et al. [19-201 studying the biosynthesis of protaglandins suggest that a phospholipase is the rate limiting enzyme in the thyroidal system. They also found that thyroid stimulating hormone produced by the pituitary gland is the regulating factor. Although phospholipids and phospholipid metabolism appear to be involved in maintaining the thyroid gland in full functional activity, reports on the phos-

184

pholipases of this organ are rather scarce. Dawson [21,22] mentions the occurrence of a lysophospholipase in sheep thyroid and of a phospholipase (type C; attacking phosphatidylinosito~) in pig thyroid. This lack of knowledge prompted us to initiate a study on phospholipase activities in bovine thyroid. Material and Methods Phospholipase AZ ex Vipera russellii venom (5 units/mg protein) was obtained from Koch-Light Laboratories Ltd. (4703t). Lipase was purified from porcine pancreas as described by Sarda et al. [23] and assayed according to the method of Mahadevan et al. [24] using olive oil (highly refined; low acidity; Sigma O-1500) as a substrate. Bovine albumin (fraction V powder) was obtained from Sigma (puriss. A-4503); sodium taurocholate (puriss. 5059h) from Koch-Light Laboratories Ltd. Stearic acid-[9,10-3H] (specific activity 490 mCi/mmol) and linoleic acid-[l-’ 4C] (specific activity 58 mCi/mmol) were purchased from the Radiochemical Center (Amersham, G.B.). All other solvents and reagents were analytical grade and used without further purification unless stated otherwise. Proteins were determined by an adapted method of Lowry et al. [25,26] with bovine serum albumin as a standard. Phospholipid phosphorus was assayed according to Rouser [ 271,

Preparation ofsubstrates Unlabelled phosphatidylcholine was purified from commercially available egg lecithin (95-lOO%, BDH 29053) [ 28 J . Phosphatidylethanolamine was obtained from egg yolks by the method of Lea et al. [29]. Lysophosphatidylcholine was prepared enzymatic~ly by the action of snake venom phospholip~e AZ (Ti. russetlii) on phosphatidylcholine [ 301. The gelatinous precipitate was spun down and dissolved in a minimal volume of absolute ethanol. Lysophosphatidylcholine was precipitated from this solution by addition of 10 vol. ether, recovered by centrifugation and redissolved in ethanol. This step was performed twice. Finally the purified lyso-derivative was stored in chloroform~meth~ol 1 : 1 (v/v) (at -20°C). The purity of these substrates was checked by thin-layer chromatography. 2-[ ’ 4 C] phosphatidylcholine and phosphatidylethanolamine were prepared by sonicating 3 pmol l-acyl-sn-glycero-3-phosphorylcholine/ethanolamine in the presence of 3 pmol [l- 1 4C] linoleic acid (? 50 yCi), 20 pmol adenosine t~phosphate and 0.2 pmol coenzyme A (0.44 M potassium phosphate buffer pH 7.4, 0.136 M NaF). 1 ml suspension of rat-liver microsomal fraction (311 was added to the mixture (1.3 ml) and incubated for 2 h at 37°C [32]. After incubation the lipids were extracted according to Bligh and Dyer [ 331 and the 2-[ ’ 4C] phosphatidylcholine/phosphatidylethanolamine isolated by preparative thin-layer chromatography. The labelled phosphatidylcholine was eluted from the silica gel with 300 ml chloroform/methanol 4 : 1 (v/v), resulting in a 99% recovery 1341. Phosphatidylethanolam~e was eluted with 75 ml chloroform/methanol 9 : 1 (v/v) 1341. Hydrolysis of the 2-[ ’ “C] phosphatidylcholine/phosphatidylethanolamine by snake venom phospholipase A2 followed by isolation of the breakdown products showed 99% of the [l-’ 4 C] linoleic acid to be localised at the 2-position of phosphatidylcholine. In phosphatidyletha-

185

nolamine 95% was found at that position. The specific activities of those labelled substrates were ?4 * lo6 dpm/mg phospholipid. l-[ 9,10-3 H] stearyl-2-acyl-sn-glycero-3-phosphorylcholine was prepared in the same way as the [ ’ 4C] compound with the exception that a 2-acyl lysoderivative was used as acyl receptor in the presence of f 6 mCi [ 3H] stearic acid. This 2-acyl lysocompound was obtained by the action of pancreas lipase [ 231 on purified egg phosphatidylcholine [ 351. Checking the intramolecular distribution of [9,10-3H] stearic acid with phospholipase AZ showed 70--80% of the fatty acid in the l-position of phosphatidylcholine. The positional specificity was increased up to 95% by hydrolysis of the labelled compound with phospholipase AZ followed by the reacylation of the 2-position with unlabelled oleic acid by rat liver microsomes. The specific activity of this substrate was +6 * 10’ d.p.m./mg phospholipid. For preparation of 1-[9,10-3H] stearyl-sn-glycero-3-phosphorylcholine the corresponding labelled phosphatidylcholine was treated with phospholipase AZ and the tritium labelled lyso-derivative isolated as described for unlabelled lysophosphatidylcholine. The labelled lyso-derivative showed a specific activity of +25 * lo6 d.p.m./mg phospholipid. All radioactive lipids were stored as solutions in chloroform/methanol 1 : 1 (v/v) at -20” C and for use in enzyme assay diluted with carrier lipids resulting 2- [ ’ 4 C] phosphatidylcholine 5 . lo4 in the following specific activities: d.p.m./2 mg; 2-[’ 4C]phosphatidylethanolamine 5 * lo4 d.p.m./2 mg; l[ 3H] phosphatidylcholine 1.7 * lo5 d.p.m./2 mg; l-[ 3H] lysophosphatidylcholine 1.2 - 10’ ‘d.p.m./mg. Doubly labelled phosphatidylcholine was prepared by mixing 2-l 4C- and 1-3H-labelled samples in order to give an 3H/14C ratio (d.p.m.) of +7. This corresponds to +5 * lo5 d.p.m. 3H and +7 - lo4 d.p.m. ’ 4C/2 mg phosphatidylcholine. Preparation of enzyme Bovine thyroids were obtained fresh from the slaughterhouse. After anatomic preparation the tissue was washed several times until free of blood and homogenized in 0.14 M NaCl (Virtis Homogenizer: max speed, 3 X 1 min, 4°C) [36]. The homogenate was centrifuged for 2 h at 75 000 X g in an angular rotor (+4”(Z). The upper lipid layer was removed by filtration of the supernatant through a paper filter Whatman No. 1 resulting in a supernatant fraction homogenate-supernatant. The sediment was delipidated by a slightly modified procedure of Morton [37] : (a) homogenization of the sediment (Waring Blendor, 10 min, 10 vol. acetone) resulting in a homogenate that was filtered through Whatman No. 1 on a Buchner filter; (b) homogenization of the residue (10 vol. n-butanol, 10 min), filtration, resuspension (10 vol. acetone, 3 min), acetone washings (3 to 5 times for removing traces of n-butanol); (c) extraction (5 vol. peroxide-free diethyl ether, 15 min, occasional stirring); (d) drying of the residue (exposing it first free in the air for 15 min and finally over P2 O5 under vacuum during 1 h). The dried powder was kept at -20°C. All extractions except for diethyl ether (room temperature) were performed at a temperature below 0” C. For assaying the enzyme activity the powder was extracted with 0.14 M

186

NaCl by stirring for 1 homogenate-sediment. -20°C) on storage of powder were also stable

h at 4” C. This enzyme preparation was defined as the Phospholipase A activity was quite stable (3 months, acetone-dried powder. The extracts prepared from the (3 months at -20” C).

Preparation of subcellular fractions 0.25 M sucrose thyroid homogenate was subjected to subcellular fractionation as described by Dierick and Hilderson [38] yielding five fractions*. Marker enzymes assayed in each fraction were: acid phosphatase [39] and /3-glucosidase [40] (lysosomes), cytochrome oxidase [41] and monoamine oxidase [ 421 (mitochondria), glucose-6-phosphatase [ 431 and rotenone-insensitive NADPH-cytochrome reductase [ 441 (microsomes), AMPase [ 431, alkaline phosphatase [ 391 and (Na’-K’)-activated ouabain inhibited ATPase [45] (plasma membranes). Assay of phospholipase A activity: radioactive procedure After evaporation of the solvent the labelled substrate and carrier were dispersed ultrasonically (Braun Sonic 300, setting 60, 3 min, 0°C) in 0.14 M NaCl in the presence of sodium taurocholate unless indicated otherwise. 0.1 ml of this phospholipid preparation (containing 2 mg phospholipid and 2 mg taurocholate) was incubated with 0.2 ml enzyme extract or subcellular fraction and 0.2 ml buffer solution (final volume 0.5 ml). In each series of experiments blanks were included. All enzyme assays were performed in screwcapped tubes at 37” C under continuous shaking. The reaction was stopped by the addition of 1.25 ml methanol and the lipids were extracted according to the method of Bligh and Dyer [33] (extraction time 15 min) followed by two additional extractions of the water/methanol layer with chloroform (3 ml). In the water/methanol layer no remaining radioactivity was found. In single labelling experiments the reaction products were separated by thin-layer chromatography (silica gel H, chloroform/methanol/water 60 : 25 : 4, v/v/v). In double labelling experiments a two-step elution procedure was used: (1) chloroform/petroleum benzine (boiling point 60--8O”C)/acetic acid (65 : 25 : 2, v/v/v); (2) chloroform/methanol/water (65 : 25 : 4, v/v/v). The spots were visualised with iodine vapors and the lipid fractions eluted into counting vials. Fatty acids were eluted directly from the silica gel (100% recovery) with Buhler scintillation solution or methanol [46]. Lyso-derivatives were eluted with methanol/water (24 : 1, v/v) (*90% recovery), whereafter the solvent was evaporated under a stream of Nz and the lyso-derivative redissolved in 15 ml Buhler solution. In later experiments (except for doubly labelled substrates where the compounds were always eluted from the silica gel) the silica gel spots containing lysocompound or fatty acid were scrapped directly from the thinlayer chromatographic plate into the counting vials, and aquasol added [47] . In double labelling experiments the eluate was evaporated, 1 ml of water added and the doubly labelled substrate and its breakdown products counted in 10 ml aquasol. *N. nuclear fraction: M, mitochondrial fraction; L. light mitochondrial tiaction: P, microsomal fraction; s, supematant; M + L denotes a combined mitochondrial and light mitochondrial fraction.

187

Assay of phospholipase

A activity:

non-radioactive

procedure

The same enzymatic procedure was also applied with non-labelled substrate. The free fatty acids released were extracted and estimated calorimetrically according to Duncombe [ 481. Assay of lysophospholipase

activity

Lysophospholipase activity was determined a substrate using the radioactive phospholipase

with lysophosphatidylcholine A procedure.

as

Results 1. Homogenate-supernatant

phospholipase

activity

In pilot enzyme assays, using non-labelled phosphatidylcholine as a substrate, phospholipase activity was found in the acid range (pH 4-6). No activity could be detected at neutral pH. The rather high blank values obtained with this technique pointed to a significant hydrolysis of endogenous substrate. No effect of Ca2+ or EDTA was noted. The positional specificity of this acid phospholipase activity was studied by using 1-acyl-2[ 1-l 4 C] linoleoyl-sn-glycero-3-phosphorylcholine. Phospholipase Al and phospholipase A2 activities were estimated by measuring the release of 2-[ ’ 4 C] lysophosphatidylcholine and [l-l 4 C] linoleic acid. The pH-curves of the homogenate-supernatant phospholipase activity (Fig. 1) show phospholipase A, -activity with pH-optima at 3.6 and 4.8 and phospholipase A2 -activity with pH-optima at 3.6 and 5.7. As these results (acid pH-optima and lack of Ca2’- and EDTA-effect) suggest that the enzymes are lysosomal, phospholipase

?o

.

. /

I; .

0

a \. 0%

00

\ I a I\) \

.

\

0

0

.

0-e

0

20

\\

! L 2

"\

I

J-4

I 3

4

. ' CG

I

I

1

5

6

7

PH

Fig. 1. Phospholipase A1 and A2 activities in homogenate-supernatant. Incubation (37°C. 60 min): 0.1 M citrate buffer (final vol. 0.5 ml) containing 4 mg protein. 2.5 wnol sonicated labelled phosphatidylcholine and 2 mg sodium taurocholate.

188

110

I

0

7

Fig. 2. Pbospholipase A1 and A;? activities in a M + L fraction as a function of PH. Incubation (37’C. 60 min): 0.1 M citrate buffer (final vol. 0.5 ml) containing 2,X mg protein from a M f L fraction sonicated during 30 s before use, 2.5 pmol sonicated labelled phosphatidylcholine (a) in the presence of 2 mg sodium taurocholate (b) in the absence of sodium taurocholate. Ordinate: nmol (e----l), [l-l 4Cl linoleic acid: (cio), [’ 4 Cl lysophosphatidylcholine released per mg protein and per h. Abscissa: pII.

activities were determined on M + L subcellular fractions. In the presence of sodium taurocholate (Fig. 2a) phospholipase A activity with A, (accumulation of [ 1 4C] lysocompound) and A2 (formation of [ ’ 4 C] linoleic acid) specificity could be demonstrated showing a single pH optimum around pH 4. No activity could be detected at neutral and alkaline pH. The phospholipase A, /phospholipase A2 ratio varies from one preparation to another or after freezing and thawing of the same sample. This could be an indication that one is eventually dealing with two enzymes. Using the same subcellular fraction in absence of sodium taurocholate yielded different phospholipase Ai /phospholipase AZ ratios, the phospholipase A, being considerable higher (Fig. Zb). Moreover, a small phospholipase AZ peak was found at pH 5.5. This taurocholate effect is compatible with the presence of different enzymes. The enzyme activities in function of time and protein concentration showed linearity up to 30 min and up to a protein concentration of about 1 mg/ml. In this subcellular fraction (in the absence of taurocholate) phosphatidyleth~olamine is a better substrate than phosphatidylcholine, resulting in a three times higher specific activity. Moreover with the use of phosphatidylethanolamine as a substrate an additional small phospholip~e AZ activity at pH 7.5 was observed. The homogenate-supernatant fraction also contains lysophospholipase activity with pH optimum 7.5 (Fig. 3a). In the presence of taurocholate (2 mg) the lysophospholipase activity decreased to 260% of the original value. To assess the extent to which lysophosph~lipase could cause erroneous results in the phospholipase radioactive assays through deacylation of the labelled lysocompound generated during the assay, control experiments were set up. Extracts

189

Fig. 3. a, Lysophospholipase activity in homogenate-supernatant. Incubation (37°C. 120 min): 0.1 M citrate buffer (final vol. 0.5 ml) containing 4 mg protein and 1.8 /.unol labelled lysophosphatidylcholine. Ordinate: nmol (o----1), [9.10-3Hlstearic acid released per h; Abscissa: PH. b, Phospholipase A2 activity in homogenate-sediment as a function of PH. Incubation (37°C. 30 min): 0.1 M ( -0) citrate or (-m) glycylglycine buffer (final vol. 0.5 ml) containing 1.36 mg protein, 2.5 fir1101 sonicated labelled phosphatidylcholine and 2 mg sodium taurocholate. Ordinate: nmol [l- 1 4Cl linoleic acid released per h and per mg protein. Abscissa: PH. c. phospholipase A2 activity in a crude microsomal fraction. Incubation (37’C. 30 min): 0.1 M glycylglycine buffer (final vol. 0.5 ml) containing 1.36 mg protein from microsomal fraction, 2.5 ~.tmol sonicated labelled phosphatidylcholine and 2 mg sodium taurocholate. Ordinate: nmol (aA). Cl- 1 4C1 linoleic acid and (A-A). [ ’ 4C1 lysophosphatidylcholine released per mg protein and per h. Abscissa: pH.

were incubated in presence of 20 nanomoles l-[ 9,10- 3H] -stearyl-lysophosphatidylcholine, 2.5 pmol of unlabelled phosphatidylcholine and detergent. In those conditions deacylation of the labelled lysocompound was negligible below pH 5.7. The decrease in susceptibility of the lysosubstrate when combined with phospholipid molecules is possible due to the fact that association of lysophosphatidylcholine with phosphatidylcholine makes the lyso-derivative inaccessible to the lysophospholipase [49]. From this finding one may conclude that the radioactive products formed from the diacyl substrates are due to hydrolysis by phospholipases A in the extract. This was further substantiated by the use of doubly labelled substrate incubated with a M + L fraction in the presence or absence of sodium taurocholate at pH 4 (Table I upper part). Although no strict stoichiometric relationship was obtained the recovery of fatty acid was nearly that of monoacylglycerophosphorylcholine indicating little or no interference from lysophospholipase under the experimental conditions. These data indicate that hydrolysis of substrate can proceed at both the 1 and 2 positions. The experiment also illustrates the different effect of the bile salt on the phospholipase Al and phospholipase A2 activities, while the overall hydrolytic activity was not drastically impaired. 2. Homogenate-sediment phospholipase activity The homogenate-sediment extract obtained from delipidated sediment showed phospholipase A2 activity giving optimal reaction at pH 6.5 (Fig. 3b) with phosphatidylcholine as a substrate. Phosphatidylcholine was found to be hydrolyzed to a much greater extent than phosphatidylethanolamine. The homogenate-sediment fraction also contains lysophospholipase activity (not plotted in Fig. 3b) showing a pH-optimum around pH 7.5. The controls for combined action of phospholipase A and lysophospholipase resulted at pH 7.5

I

tiaction

-

line)

A2

_

corresponds

93.9

expressed

_

to the combined

action

lysophospholipase.

activity

residual

hydrolytic

14C -

94.8

84.8

(W)

(3H-free

overall

of phospholipases

The

the sum Al

and

hydrolytic

fatty

5.2*

6.1*

by

0.6

5.5

4.6

15.8*

15.6*

0.6

11.7

4.1

AZ.

of

fraction

eventually

of

sum

in each

(%)

(3H-lysophosphatidylcholine

to the combined

94.6

95.9

89.9

89.8

choline

Phosphatidyl-

used).

compound

radioactivity

lysophospholipase.

the

corresponds

14c

3H

14C

3H

the

recovered 90%

Isotope

acid)

residual

expressed

fatty

60” C

treated

+ 14C-free

5 min

heat

sediment

Homogenate-

cholate

by

source

equaled

radioactivity

+ sodiumtauro-

M+L

Enzyme

experiment

activity

acid

(%)

Lysophosphati-

in the

dylcholine

3.5

12.1

fatty (%)

recovered

PHOSPHATIDYLCHOLINE

the amount

LABELLED

by comparing

DOUBLY

radioactivity

Free

total

acid

14c

3H

THE

Phosphatidyl-

83.8

+ eventually

* The overall

sediment

OF

was calculated

choline

3H

Isotope

experiment

(the

HYDROLYSIS

RECOVERED”

THE

in the

experiments

source

Homogenate-

M+L

Enzyme

OF

COMPOUND

recovered

of duplicate

activity

“PERCENT

STOICHIOMETRY

TABLE

the

4.6*

0.4

4.2

9.7*

4.5

5.2

Al + 14C-lysophosphatidylcho-

and

(%)

value

radio-

Lysophosphatidylcholine

of phospholipases

fatty

of average

amount are the

total

results

(W)

action

5.3s

5.0

0.3

10.6*

5.6

5.0

acid

Free

The

with

191

in 5% deacylation of the labelled lysocompound. The lack of phospholipase A,, suggested by the absence of radioactive lyso-derivatives in the phospholipase AZ assays, was confirmed using 1-[9,10-3H] -stearyl-2-acyl-sn-glycero-3phosphorylcholine as substrate. Radioactivity of the released stearic acid amounted to 5% of that of the corresponding lysocompound. The stoichiometric formation of lysophosphatidylcholine and fatty acid was further checked by incubating doubly labelled substrate with homogenate-sediment extract at pH 6.5. The results for free fatty acid and lysophosphatidylcholine ( [ ’ 4 Cllysophosphatidylcholine vs. 3H-free fatty acid (phospholipase Al ); 3H lysophosphatidylcholine vs. ’ 4C-free fatty acid (phospholipase A, )) are internally consistent, although a somewhat lower recovery for the lyso-derivatives was found (Table I, lower part). After heat treatment of the homogenate-sediment extract (5 min, 60°C: inactivating lysophospholipase) prior to incubation no increase in lysoderivative recovery was noted. Strict reproducibility of total and specific activity for phospholipase AZ was not achieved: a variation factor up to five was found among different preparations. Careful control of parameters (time, temperature, volume and removal of extraction fluids) during extraction did not solve this problem. Low activities could be increased by acidifying to pH 4 or by repeated freezing and thawing. These methods result in enhanced total and specific activity after elimination by centrifugation of the sediment formed during the activation procedures. Incomplete delipidation resulting in a non-controlled presence of endogenous substrates, could cause artificially low values for enzyme activity. All experimental data discussed below were obtained with enzyme preparations activated by acidifying and repeated freezing and thawing. The homogenate-sediment phospholipase activity was not affected by the addition of Ca’+-ions (concentrations up to 8 mM) or EDTA (5 mM). The influence of increasing amounts of sodium taurocholate on homogenate-sediment phospholipase AZ is shown in Table II, maximal activity being obtained when adding 2 mg taurocholate to 2 mg phosphatidylcholine. The stimulating effect of taurocholate can be ascribed to its amphiphilic nature [ 501 : the hydrophobic part of the molecule contacting the hydrophobic area of the enzyme and the hydrophylic head

TABLE

II

INFLUENCE A2

OF

SODIUM

TAUROCHOLATE

Incubation:

as in Fig.

3b except

that

pH

nmol

[ 14Cl

(mg/incubation

acid

released

mixture)

protein

Sodium

0 0.5

taurocholate

3.9 0.0

1.0

169.0

1.5

217.3

2.0

309.0

2.5

259.0

3.0

182.0

6.0

ON

THE

HOMOGENATE-SEDIMENT

ACTIVITY

87.0

and

was kept Linoleic per mg per hour

constant

at 6.5.

PHOSPHOLIPASE

192

groups conferring water solubility to the resulting complex. The reagent may also form mixed micelles with the substrate. For the decrease in activity of high amounts of taurocholate no proven explanation is yet available: a surface dilution phenomenon is likely to be important [51]. The time-activity curve was linear up to 30 min for a protein concentration of 1.36 mg in 0.5 ml of incubation mixture. The enzyme activity was not completely linear with respect to protein concentration (slightly convex curve, suggesting inhibition). Heat treatment for 5 min at 60°C did not result in any significant inactivation (