Regulation of phosphatidic acid ... - The FASEB Journal

Regulation during

of phosphatidic stimulation

acid phosphohydrolase

of human

activity

polymorphonuclear

leukocytes A, P. TRUETT III, S. B. BOCCKINO,t1 *J)parents of Medicine, Pharmacolog)c Medical

Center, Nashville,

Tennessee

J. J. MURBAY and

Phosphatidic acid phosphohydrolase (PPH) activity was determined in human polymorphonuclear leukocytes (PMNs) by measuring the hydrolysis of [32PJphosphatidic acid (PA) added to cell sonicates. Enzyme activity was localized primarily to a soluble fraction. Soluble and particulate activities required magnesium and were inhibited by calcium, N-ethylmaleimide, sphingosine, and propranolol. The activity in unstimulated PMNs was 0.64 ± 0.11 nmol of PA hydrolyzed. mg protein-’ mm ‘ in particulate and 4.20 ± 0.42 in soluble fractions. Stimulation of PMNs with 1 eM f-Met-Leu-Phe (FMLP) for 10 mm caused a slight decrease in soluble activity and a small increase in the activity of particulate fractions. Preincubation with 10 tiM cytochalasin B for 5 mm before FMLP stimulation markedly enhanced both of these changes. The effect of FMLP plus cytochalasin B was rapid (< 10 s), whereas the calcium ionophore A23187 (1 tiM) and phorbol myristate acetate (100 ng/ml) caused slower and smaller changes in enzyme activity. These results indicate that after chemoattractant stimulation; PPH activity decreases in the soluble fraction and increases in the particulate fraction suggesting that PPH may participate in signal transduction in the PMN. -Truett, A. P., III; Bocckino, S. B.; Murray, J. J. Regulation of phosphatidic acid phosphohydrolase activity during stimulation of human polymorphonuclear leukocytes. FASEBJ 6: 2720-2725; 1992.

neutrophil

diacyiglycerol .

phospholipase

D

.

superoxide

anion

signal transduction

STUDIES HAVE DEMONSTRATED THAT after chemoattractant stimulation, human polymorphonuclear leukocytes (PMNs)’ generate diacyiglycerol (DAG) by way of the hydrolysis of phosphatidylcholine (PC) by a phospholipase D (PLD) to yield phosphatidic acid (PA), which is then hydrolyzed by phosphatidic acid phosphohydrolase (PPH) (1-3). In addition to being a precursor to the intracellular signal DAG, evidence suggests that PA may also function in cellular signaling per se. These data include the observation that PA can directly activate the NADPH oxidase complex in cellfree preparations (4, 5). Furthermore, treatment of PMNs with propranolol, an inhibitor of PPH, produces a significant enhancement of superoxide anion production (6-8). We and others (6-9) have observed that propranolol blocks the rise in DAG mass without inhibition of PLD, suggesting that the dephosphorylation of PA is prevented. Because both PA and DAG may function as intracellular signals, changes in PPH activity may serve as a regulating point in the control of PMN function. MANY

2720

and

Biophysics,

Vanderbilt

University

37232-0111, USA

ABSTRACT

Key Words:

Physiology

Studies of liver, adipocytes, and pneumocytes have demonstrated that PPH is important in phospholipid and triacylglycerol synthesis (10-12). The enzyme has been best characterized in the liver, where approximately 70% of PPH activity was in the cytosol and 30% was membrane-associated (13). The enzyme required magnesium and was inhibited by calcium and propranolol. Prolonged incubations with dexamethasone, glucagon, and cAMP analogs have been shown to increase activity (14-16). More recently a magnesiumindependent form of PPH in liver has been described (16). There has been no evidence that change in the activity of either form of the enzyme functions in intracellular signaling. In this work, an assay for PPH in broken cell preparations based on the release of [‘2P]Pi from an exogenously added [32P]PA:PC vesicle substrate was adapted for use in PMNs (12). A rapid and significant decrease in soluble enzyme activity and an increase in particulate activity were observed in PMNs stimulated with the chemoattractant f-Met-LeuPhe (FMLP). These findings suggest a controlling function of this enzyme in the signal transduction mechanisms of PMNs.

METHODS Materials Unless otherwise noted, all reagents were from Sigma Chemical Co. (St. Louis, Mo.). Phorbol myristate acetate (PMA), FMLP, and A23187 were diluted in dimethylsulfoxide (DMSO) and stored as frozen stock solutions. Before addition to cells, the stimulants were diluted in HHBSS (Hank’s balanced salt solution with 10 mM HEPES and 4.2 mM NaCO,, pH 7.4). The final concentration of DMSO in the cell suspension was 0.01% and had no effect on any activity measured. Dioleoylglycerol and dioleoylphosphatidic acid were from Avanti Polar Lipid (Alabaster, Ala.). Diacylglycerol kinase was from LIPIDEX, Inc. (Westfield, NJ.). Silica gel 60 plates were from Whatman (Clifton, N.J.).

‘Current

Durham,

address:

Sphinx

Pharmaceuticals,

P.O.

Box

52330,

NC 27717, USA.

‘To whom correspondence should be addressed, at: Allergy Lab, 843 Light Hall, Vanderbilt University Medical Center, Nashville, TN 37232-0111, USA. 3Abbreviations: PMNs, polymorphonuclear leukocytes; DAG, diacylglycerol; PC, phosphatidylcholine; PLD, phospholipase D, PA, phosphatidic acid; PPH, phosphatidic acid phosphohydrolase;

FMLP, f-Met-Leu-Phe; PMA, phorbol myristate acetate; HHBSS, Hank’s balanced salt solution with 10 mM HEPES and 4.2 mM NaCO3, pH 7.4; NEM, N-ethylmaleimide; [3H]PA, [3H-palmitate]PA.

0892-6638/92/0006-2720/$01

.50. © FASEB

Leupeptin was from Transformation Research (Framingham, Mass.). PMNs were obtained from healthy donors and prepared as previously reported (17). Polymorphonuclear leukocytes were suspended in HHBSS. Preparation

of PPH

from

PMNs

The PMNs were untreated or treated with the cellular agonists for the indicated times in a final reaction volume of 400 gil, placed in an ice-cold water bath, and pelleted (400 g, 10 mm). This and subsequent procedures were performed at 4#{176}C. The cells were resuspended in ice-cold 100 mM potassium phosphate (pH 7.1) and 1 mM EGTA, and then disrupted by sonication (5 s at setting of 60 on Fisher Sonic Dismembrator using a microtip). After centrifugation of the homogenate (1000 g, 10 mm) to remove unbroken cells, the supernatant was removed. Soluble and particulate fractions were prepared from the supernatant by centrifugation at 40000 g for 60 mm. Fractions were kept at 4#{176}C until use. The protein concentration of each sample was determined by using the Pierce BCA protein assay reagent (Pierce, Inc., Rockford, Ill.). Measurement of every sample was necessary as the protein concentration between conditions (e.g., agonist treated or untreated) varied by as much as 10% in each experiment.

expressed as the mean ± SEM of the averages of the triplicatesin each experiment for the number of experiments indicated or the mean ± SD of the triplicates if a representative experiment is shown. Statistical significance (P < 0.05) was calculated by the paired I test comparing the average of the triplicates.

RESULTS Assay

of PPH

The duration or number of sonication cycles was found to be critical for maintaining PPH activity - increasing sonication time beyond 5 s or increasing the number of cycles beyond one caused a marked diminution of activity. The addition of protease inhibitors, including phenylmethylsulfonylfluoride and leupeptin, did not increase activity and 1 mM DTT caused a small decrease. EGTA (1 mM) added before sonication resulted in more stable and consistent activities within experiments. Triton X-100 (0.1%) added to either the potassium phosphate buffer or directly to the assay mixture was strongly inhibitory (data not shown). Fractions were stored at 4#{176}C for up to 48 h with little loss of enzyme activity. Validation

Assay

of PPH

activity

of PPH

activity

assay

activity

Because certain tissues (notably liver) express significant lipase activity (21) that may produce lysophosphatidic acid, The reaction mixture for assay of PPH activity consisted of glycero-3-phosphate, or other aqueous soluble products from 50 mM HEPES (pH 7.4), 1.25 mM EDTA, 3.25 mM MgC1,, PA, using the release of free phosphate from phosphatidic and 30-100 g of cellular protein in a final volume of 100 il. acid as an indicator of PPH activity may prove inexact. We The reaction was initiated by the addition of 50 nmol of a therefore performed experiments to confirm this assay of [‘2P]PA:PC substrate (1:1, mol:mol) (-0.2 itCi/nmol of [32P]Pi release as a valid measure of PPH activity in the PA), which had previously been sonicated to form vesicles. PMN in both stimulated and unstimulated PMNs. [32P]PA was prepared by phosphorylation of dioleoylglycerol An aliquot of the aqueous phase from the modified Bligh with diacylglycerol kinase in the presence of [32P]ATP (speand Dyer extract was removed and the free orthophosphate cific activity of - 200 Cilmol, ICN) (18) and separated on silwas extracted in the form of phosphomolybdate after the adica gel 60 plates in a system of ethyl acetate:iso-octane:acetic dition of ammonium molybdate (22). Greater than 95% of acid (45:15:10/v:v:v). Phosphatidic acid was eluted from silica with 10 ml of methanol and 100 l HC1 (1 N), dried under nitrogen, resuspended, and stored in chloroform (19). Protein fractions were warmed for I mm at 37#{176}C before the addition of vesicle substrate. The reactions were allowed to proceed for the specified time (typically 30 mm, during which time the assay was linear, data not shown) at 37#{176}C and were terminated by the addition of 750 d of chloroa) form:methanol (1:2/v:v). A modified Bligh and Dyer extrac‘I, tion was performed (20) by the addition of 350 ,tl of 0.1 N a) HC1 and 250 l of chloroform. After separation of phases, an a) aliquot of the aqueous phase was counted. ‘I’ a) [‘H-palmitate]PA ([3H]PA) was prepared as described (21) 0 except that PA was separated after organic extraction by seE quential chromatography on silica gel 60 plates in systems of 0 ethyl acetate:iso-octane:acetic acid (45:15:10/v:v) and the orC ganic phase of ethyl acetate:isooctane:acetic acid:water z (55:25:10:50/v:v). The specific activity of the PA substrate was calculated by adding a known amount of [“P]PA or [‘HIPA to dioleoylphosphatidic acid. The radiolabel accounted for - 1% of the Control Propranolol No Mg2 NEM total mass of the PA substrate. Phosphatidic acid phosphoFigure 1. Measurement of PPH activity by [32P]Pi release and hydrolase specific activity was determined by measuring the DAG formation. Soluble fractions were prepared from unstimu[32p]pi released from labeled PA divided by the specific aclated cells as described and PPH activity was determined. tivity of the PA substrate, the amount of protein in each Propranolol (1 mM) or N-ethylmaleimide (NEM) (4.2 mM) was tube, and the time. Phosphatidic acid phosphohydrolase added immediately before addition of [‘2P]PA substrate. The specific activity is expressed as nanomoles of PA hydroresults are the mean ± SD of triplicates of a representative experilyzed. mg protein’ min. ment for [32P}Pi release (black bars) and DAG formation (white Each experiment was performed in triplicate and data are bars).

PPH REGULATION

IN HUMAN

PMNs

2721

TABLE

1. Assay

of PPH

activity

by [32PJPi

release and [3HJDAG

fonnation [32P]Pi Fraction

nmol

Sonicate

Particulate Soluble

Control FMLP Control FMLP Control FMLP

release,

mg’

1.53 0.47 0.09 0.17 1.97 0.64

.

± ± ± ± ± ±

[5H]DAG nmol

min

0.27 0.06 0.00 0.01 0.25 0.03

mg’

1.88 0.42 0.06 0.13 2.41 0.61

formation, . min’

± ± ± ± ± ±

0.29 0.09 0.04 0.03 0.11 0.09

“PMNs were preincubated for 5 mm with 10 sM cytochalasin B and subsequently stimulated with 1 M FMLP for 10 mm. The results are expressed as nmol of PA hydrolyzed#{149} mg protein’ min’ and are the mean ± SD of triplicates from a representative of four similar experiments. No significant differences of the PPH activity using the two methods were detected in any fraction. The measurement of PPH activities in the fractions from stimulated and unstimulated cells of the 4 experiments comparing the two radiolabeled methods was highly correlated; r 0.968 (P = 0.001).

the radioactive counts in each aliquot could be found in the extracted phase, indicating that the radioactivity reflected the release of [32P]Pi. Assays for PPH activity were done with a nonradiolabeled PA:PC or a [‘H-palmitatejPA:PC substrate in the same preparation of PMNs as examined using the [32PJPA:PC substrate. After termination of the reaction, the DAG released from unlabeled substrate was measured by using DAG kinase (18) or the [‘H]DAG was separated from [3H]PA on silica gel 60 plates using a system of toluene:ether:ethanol: ammonium hydroxide (50:30:2:0.2/v:v). The amount of DAG released as measured by mass (Fig. 1) or [3H]-DAG (Table 1) closely agreed with the amount of [‘P1Pi released in control

100.

and FMLP-stimulated cells for sonicates, soluble, and particulate fractions. To confirm that lysophosphatidic acid was not formed to a significant degree, reactions were terminated with 1.25 ml of butan-1-ol (23). Lipid from the butanol phase was chromatographed on silica gel 60 thin layer chromatography plates with chloroform:methanol:acetic acid:acetone:water (10:2:2:4:1/v:v) for development. Authentic PA and lysoPA were run as standards and the plates were subjected to autoradiography. No formation of lysoPA could be detected (data not shown). Furthermore, glycerol-2-phosphate (5 mM), which competes with PA as a substrate for phospholipase A, (21), was without effect (data not shown). Charactermstics

100. C

0

0 C

1’ 40. 20. 1

I

100

200

Sphingosine

C

0 -C C C

a)

300

400

C C.) U

40.

a)

20.

500

-

4 Ca

(IIM)

i’o

t,

(mM)

C

80. 60. 40.

a) U

80. 60.

60. C

100.

activity

Optimized enzyme activity was 4.93 ± 0.48 nmol of PA hydrolyzed . mg protein min in sonicates (n = 10). Separation into cellular fractions demonstrated that most of the activitywas soluble-4.20 ± 0.42, soluble,and 0.64 ± 0.11, particulate (n = 21). Omission of magnesium from the assay decreased activity by approximately 5-fold in sonicates and 15-fold in soluble fractions (Fig. 1). Magnesium concentrations greater than 6.5 mM, as well as added calcium ions, strongly inhibited activity (Fig. 2). Addition of phosphatidylcholine to form vesicles augmented activity in the soluble fraction with no significant effect on activity in the particulate fraction (Fig. 2). Propranolol has been reported to increase superoxide production in intact PMNs, presumably by inhibition of PPH (6, 7). DL-propranolol, as well as the individual isomers, when added directly to the enzyme assay inhibited magnesium-dependent activity in a concentration-dependent fashion (Fig. 2). Because sphingosine has been reported to

A

o 80.

a) U C)

of PPH

20. A

ST!5 Propranolol

(mM)

l#{243}O260

300

Phosphatidylcholine

400

500

(pM)

Figure 2. Characteristics of soluble and particulate PPH activity. Soluble (#{149}) and particulate fractions (A) were prepared from unstimulated cells as described and PPH activity was determined by release of [“P]Pi. The results are the mean ± SD of triplicates of representative experiments. Panels A-C are expressed as the percentage inhibition of activity compared to control fractions whereas panel D is expressed by activity (nmol PA hydrolyzed mg protein’ . min’). A) Sphingosine; B) Ca’; C) propranolol; D) concentration of PC in PA:PC vesicles.

2722

Vol. 6

June 1992

The FASEB Journal

TRUETT ET AL.

100

Cytochalasin B alone had no effect on activity (data not shown). The changes in soluble and particulate fractions occurred within 10 s (Table 2). No significant change in activity occurred over 10 mm in untreated cells. As we have observed previously for the activation of PLD and generation of superoxide anion, A23187 and PMA caused similar but slower changes in PPH activity compared with FMLP (Fmg. 4). At 10 mm, soluble PPH activity decreased by 28 ± 5% and 58 ± 9% in response to 1 tiM A23187 (n = 4, P < 0.05) and 100 ng/ml PMA (n = 6, P < 0.05), respectively. The increases in particulate activity were also significant at 54 ± 41% and 61 ± 21%, respectively.

Cd-)

C

#{149}E (50

80

>.>
FMLP + cytochalasin B (32). These data indicate that A23l87 generates a large amount of PA by

Vol. 6

June 1992

DAG

despite

In

superoxide anion production relative to FMLP + cytochalasin B may therefore result from relatively lower PA levels with A23187 stimulation. Given these findings, a possible role of PPH in the signal transduction mechanisms of cells would be the inactivation of soluble PPH activity to varying degrees after cell stimulation with the different agonists, resulting in a differential inhibition of the hydrolysis of PA to DAG. For example, stimulation with A23187, which produces a smaller decrease in PPH activity than FMLP, would allow PA to be more readily converted to DAG and generate a smaller amount of superoxide anion. The dose response to FMLP demonstrating an approximate inverse relationship between superoxide anion generation and PPH