Phosphoinositide and calcium signalling responses in smooth muscle cells: Comparison between lipoproteins, Ang II, and PDGF

Vol.

188,

November

No. 16,

3, 1992

BIOCHEMICAL

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Pages

PHOSPHOINOSITIDE

AND CALCIUM

MUSCLE CELLS: COMPARISON Valery BOCHKOV*,

Department *Laboratory

Received

AND

1992

SIGNALLING

RESPONSES IN SMOOTH

BETWEEN LIPOPROTEINS,

Vsevolod TKACHUK*,

1295-1304

ANG II, AND PDGF

Frii BUHLER and Th&se

RESINKs

of Research, Base/ University Hospital, CH 4037 Basel, Switzerland of Molecular Endocrinology, Institute of Experimental Cardiology, Cardiology Research Center, Moscow, Russia September

18,

199.2

SUMMARY: The effects of low density lipoprotein (LDL) and high density lipoprotein (HDL& on second messenger systems were investigated in cultured human vascular smooth muscle cells (VSMC) and compared with those of angiotensin II (Ang II) and platelet-derived growth factor (PDGF-BB). Phosphoinositide metabolism was studied in myo-[2-3H]-inositol prelabelled VSMC using high performance liquid anionexchange chromatography. The spectra of inositol phosphate isomers increased after stimulation with either Ang II, LDL, HDL, or PDGF-BB were qualitatively identical. Major increases occurred in 4-IP,, 1,4-E’,, 1,3,4-IP, and 1,3,4,5IP,. These are metabolic conversion products of 1,4,5IP, for which only a minor increase was found. Thus lipoproteins, like Ang II and PDGF-BB, activate polyphosphatidylinositol-specific phospholipase C. Intracellular Ca2 ’ concentrations ([Ca2+li) were studied in furaloaded VSMC. In monolayer cultures LDL and HDb increased [Ca2+], with kinetics comparable to those for Ang II. Relative to the effects of these agonists, the PDGFBB-induced increase in [Ca2+]i was slower in onset and the decay from peak [Ca2’]i levels more gradual. Fluorescence recordings from single cells exposed to LDL and HDL, revealed a prolonged series of transient oscillations of [Ca2’], a phenomenon typical for calcium-mobilizing hormones. Additionally, as found for Ang II, preincubation of VSMC with either phorbol 12-myristate, 13-acetate, forskolin or 8bromo-cyclic GMP inhibited LDL- and HDtinduced accumulation of [3H]-inositol monophosphate. We propose that LDL and HD& stimulate signal transduction in VSMC via mechanisms analogous to those of Ca2+-mobilizing hormones. (01991 Academic Press,1°C.

$To whom correspondence should be addressed. ABBREVIATIONS

LDL and HDL, low and high density lipoprotein; VSMC, vascular smooth muscle cells; Ang II, angiotensin II; PDGF-BB, platelet-derived growth factor BB-homodimer; [Ca2+li, intracellular Ca2+ concentration; Fnrin and F,, minimal and maximal fluorescence; IP,, IP,,IP, and IP,, inositol mono-, bis, tris- and ten&is-phosphate; GroPI, glycerophosphoinositol; HPLC, high performance liquid chromatography; PMA, phorbol 12-myristate, 13-acetate; 8-Br-cGMP, 8-Bromo-cyclic GMP.

1295

Copyright 0 1992 All rights of reproduction

0006-291X/92 $4.00 by Academic Press. Im,. in any .form reserved

Vol.

188,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

High blood levels of low density lipoproteins (LDL) are a well recognized risk factor for the development mechanisms

of atherosclerosis [l] and probably also hypertension

underlying

the

association

between

cardiovascular

[2]. The

disease and

disturbances in cholesterol homeostasis have been under investigation for some time. However new insights into the role of LDL in cardiovascular pathologies have recently arisen through findings that LDL is capable of activating second messenger systems in a number of blood and vascular wall cell types [3,4], although

the molecular

mechanism whereby this occurs is uncertain. An increase in cellular cholesterol content can significantly

change the activity of a number

including Ca*+-transporters

of membrane-bound

enzymes

and phospholipase C in vascular smooth muscle cells [5]

and platelets [6]. However it is not clear as to whether lipoprotein-induced transduction is linked to the well known lipid-transporting

signal

function of lipoproteins

or

whether it represents an entirely new functional property for lipoproteins. This paper has therefore examined phosphoinositide

and intracellular

Ca*+ responses of cultured

human vascular smooth muscle cells (VSMC) to two types of lipoproteins

playing

opposite roles in cholesterol transport, namely LDL, which delivers cholesterol to cells [7], and high density lipoproteins

(HDLJ,

which plays an essential role in reverse

cholesterol transport from extrahepatic cells to the liver [S]. The effects of LDL and HDL, on these second messenger systems are compared with those elicited by other well recognized activators of phosphoinositide

signal transduction, namely angiotensin

II (Ang II), a vasoconstrictor peptide, and platelet-derived

growth factor (PDGF-BB),

a peptide growth factor. METHODS Cell culture: The isolation, phenotypic characterization and propagation of human VSMC from either microarterioles associated with omental tissue or from the lower thoracic aorta according to previously described pocedures [4]. Experiments described herein used diferent isolates of VSMC between passage 4 and 11, and unless otherwise specified, confluent cultures of VSMC were rendered quiescent by serum deprivation and maintenance in serum-free medium containing 0.1% (w/v) bovine serum albumin for 48 hr. before examination of signalling responses to agonists. Cell numbers were routinely determined using a Coulter counter following trypsinization of cell layers [4]. Isolation of lipoproteins: LDL and HD& were isolated from the plasma of healthy males using sequential ultracentrifugation and buoyant density centrifugation techniques with use of potassium bromide solutions for density adjustments [9,10]. 1 mM EDTA and 1 PM butylated hydroxytoluene were present during the entire isolation procedure. Lipoproteins were sterilized by filtration through 0.45 PM Gelman filters and stored in sterile plastic tubes at 4°C. Analysis of inositol phosphate isomers by high performance liquid chromatography: Confluent VSMC cultures were maintained for 48 hrs. in inositol-and serum-free medium containing 5 &?/ml myo-[2-3H]-inositol to prelabel inositol phospholipids. After removal of radiolabelled medium and washing of cell layers, VSMC were 1296

Vol.

188,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

preincubated in serum-free medium containing 15 mM LiCl for 30 min at 37°C. Thereafter agonists were added and incubations were terminated by aspiration of medium and addition of 1 ml of chloroform/methanol/HCl/phytin hydrolysate (1:3:0.04:5 Pi pg/ml). Phytin hydrolysate was prepared as described [ll]. Following lipid extraction and separation of organic and aqueous phases, 1 ml of the upper aqueous phase was transferred into eppendorf tubes, and after addition of 25 ~1 of 50 mM D-mannitol samples were vacuum dried in a SpeedVac centrifuge, and stored at 20” C until analysis by high performance liquid chromatography (HPLC). To control for variability between parallel samples a further 100 ~1 aliquot of the upper aqueous phase was applied to Dowex-AG l-X8 anion exchange columns and water soluble inositol phosphates resolved by sequential elution with increasing concentrations of ammonium formate/O.lM formic acid [4,12]. Before analysis of those samples prepared for separation on HPLC, parallel samples with less than 10% deviation from mean values determined after Dowex column chromatography were pooled by sequential solubilization in 200 ~1 H,O. Pooled samples were spiked with 5 ~1 of a solution containing 1 mM each of CAMP, GMP, ADP, GDP, ATP and GTP, and 100 ~1 aliquots were injected to the column. HPLC analysis of inositol phosphates was performed according to a previously described method [13] with slight modifications. The samples were separated on a Partisil 10 SAX strong anion exchanger column (Alltech Ass., U.S.A.), equipped with direct-connect in-line filter and guard column. The column was eluted at 1.25 ml/min with an increasing gradient of ammonium dihydrogen phosphate (pH 4.0) according to the following protocol: 0 to 0.15 M in 15 min, 0.15 to 1.0 M in 15 min and finally 1.0 M for 15 min. After collection of fractional volumes - 0.25 ml (5 - 15 min), 0.4 ml (19 - 32 min) or 0.8 ml (32 - 45 min) radioactivity was determined by liquid scintillation counting. Between runs the column was equilibrated with water for 20 min. Inositol 1-monophosphate (l-IP), inositol 1,4bisphosphate (1,4-IP,), inositol 1,4,5-trisphosphate (1,4,5-IP,) and inositol 1,3,4,5tetrakisphosphate (1,3,4,5-IP,) were identified by sequential separations of [3H]labelled cell extracts spiked with either a mixture of D-myo-[2-3H]inositol lmonophosphate,D-myo-[2-3H]inositol 1,4-bisphosphateandD-myo-[2-3H]inositol 1,4,5trisphosphate, or with D-myo-[2-3H]inositol 1,3,4,5-tetrakisphosphate. Authentic [3H]labelled standards were obtained from Amersham. Inositol 4-monophosphate (4-IP,) was tentatively identified as a product of D-myo-[2-3H]inositol 1,Cbisphosphate degradation, following incubation of the latter standard with a sonicate of VSMC; HPLC separation of the inositol monophosphate fraction generated thereof yielded two peaks, one of which was distinct from authentic myo-[2-3H]inositol lmonophosphate, and therefore may be assumed to represent inositol 4monophosphate. Measurements of [Ca’+],: Measurements of [Ca*+], were performed on subconfluent cultures of human aortic smooth muscle cells grown on glass coverslips (22 mm diameter). The cell permeant acetoxymethylester form of the Ca*+-sensitive probe fura- was used. After washing cell layers with physiological saline solution containing 140 mM NaCl, 10 mM KCl, 1.8 mM CaCl,, 1.0 mM MgCl,, 1.0 mM Na,HPG,, 5 mM D-glucose, 0.1 mg/ml (w/v) bovine serum albumin, 25 mM HEPES, pH 7.4 [14], VSMC were loaded with a 1 PM fura-2/AM solution for 15 min at 37°C. The cultures were washed with physiological saline solution to remove unincorporated furs-2/AM and incubated for a further 5 min at 37°C to allow diffusion of non-hydrolysed dye from the cell. The coverslip was mounted in a chamber and placed on the thermoregulated stage of a Nikon Diaphot inverted epifluorescence microscope, being part of a PhoCal fluorescence photon counting analyzer (Joyce Loebl, Gateshead, U.K.) Cells were illuminated with alternating 340 nm and 380 run light from a rotating filter wheel (6.25 Hz). Emission from either single cells (without any other cells adjacent) or the cell monolayer as a whole was monitored at 500 run. Data were 1297

Vol.

188,

No.

3,

BIOCHEMtCAL

1992

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

analyzed using the PhoCal software. Calibrations were performed by permeabilization of the cells with 5 x lo-%I digitonin to obtain F,, followed by the addition of 10 mM Ttis-EGTA to obtain Fmin. [Ca2+]i WAS calculated as described previously [14,15]. RESULTS AND DISCUSSION HPLC analysk of imitol

@saydata

genemted in iYpopmt&-, Ang II- and PDGF-BB-

&eat& cc&. Exposure of human omental VSMC to either LDL, HD&,

PDGF-BB resulted in the stimulation of phosphoinositide

Ang II or

catabolism (Fig. 1). Using

high performance liquid anion-exchange chromatography to analyse water-soluble 13H]inositol phosphate extracts from VSMC, complete separation of eight different isomers of inositol phosphates was achieved (Fig.1). Of these, l-IP,, 4-IP,, 1,4-IP,, 1,4,5-IP,, and 1,3,4,5IP, could be positively identified, but due to the lack of appropriate

T

? Gm-PI

E

a

8 11~- B 3 9% g .? s m -

I-IP,*

4-q*

0

n

LDL HDL3

standards

0 Angn lil PDGF-BB

1.4-IP2.

?-IP3

? 1.3.~IP3

I ,45-IP3*

I .3,4,5-IP4*

1,4-E+*’

?-IP3

? l3,4-Ip3

1.45~IP3*

1,3.4,5-lP4*

31

33

39

(mtmttmwim)

Em650.

MO350-

? Gm-PI Rctmtion timc(min)

9

1-q* 11

4-J&5

l

12.5

22

28

Fiare 1. Inositol phosphate isomers produced by VSMC in response to Ang II, LDL,HDL, and PDGF-BB. myo-[2-‘Hlinositol-prelabelled human omental VSMC were

incubated for 30 set (Panel A) or 5 min (Panel B) at 37°C either in the absence (control) or presence of LDL, HDL, (each at 100 pg/ml), Ang ll(2 nM) or PDGF-BB (2ng/ml). [‘HI-inositol phosphate isomers in cell extracts were separated on a Partisil 10 SAX column. All methodological procedures including positive identification of the different peaks of radioactivity (indicated with l ) are described in “Methods”. ? indicates peaks of uncertain or unknown identity. The results expressthe 13HH]-content in peak fractions eluting at times specified (retention time) as a percentage of that (100%) in the corresponding peaks from control samples. Values (mean 2 SD) were obtained from four separately performed experiments. Comparable results were obtained in a single experiment on human aortic VSMC. 1298

Vol.

188,

No.

3,

BIOCHEMICAL

1992

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

we are not certain of the identity of those isomers with retention times of 9 min. (possibly glycerophosphoinositol), IP,).

28 min. (unknown IP,) and 31 min. (possibly 1,3,4-

1,4,5-IP, is the isomer known to be rapidly and transiently elevated during the

first seconds of agonist-stimulated for LDL-, HD&-,

polyphosphoinositide

Ang II and PDGF-BB-stimulated

hydrolysis [16,17]. However

VSMC, we measured only small,

albeit significant (p < 0.05), increases in 1,4,5-IP, after 30 sec., and negligible increases after 5 min. Nevertheless, after both 30 sec. and 5 min. of agonist stimulation

there

were marked increases (ranging from 2 to 10 fold above control) in 4-IP,, 1,4-IP, and 1,3,4,5-IP, (Fig.l), inhibition

isomers of inositol phosphate which are more stable due to

of degradative enzymes by Li+ ions [17]. 1,4,5-IP, is metabolized via two

pathways, namely degradation to bis- and monophosphates, and phosphorylation 3-position with generation of tetrakisphosphate

at the

[ 16,171. Thus from the data in Figure

1 we infer that 1,4,5-IP, is the initial inositol phosphate generated in response to either LDL, HDb,

Ang II or PDGF-BB, and that it is subsequently rapidly metabolized to

4-IP,, 1,4-IP, and 1,3,4,5-IP,. Rapid metabolic conversion is thought to play a role in termination

of 1,4,5-IP3-induced Ca2’ release from internal pools [16].

It is interesting that the spectrum of inositol phosphate isomers generated by VSMC in response to Ang II and PDGF-BB were qualitatively

not different

since

vasoconstrictor peptides such as Ang II are known to activate the plasma membranebound l3 isoform of phospholipase C, whereas growth factor peptides such as PDGFBB activate the cytosolic y isoform of phospholipase via activation of intrinsic receptor tyrosine kinase [18]. Since the spectra of inositol phosphate isomers generated in response to LDL and HD& were qualitatively indistinguishable

from those produced

in response to Ang II and PDGF-BB, we conclude that lipoproteins phosphatidylinositol

4,5-bisphosphate-specific

presently determine

which isoform of phospholipase

also stimulate

phospholipase C. However we cannot C is stimulated by LDL and

HDb.

Regulahn ofqtosolic

C&concdnby

lipopmteinq

AngZZandPDGF-BB.

It is known that phospholipase C generates two second messengers: diacylglycerol, which activates protein kinase C, and 1,4,5-IP, which stimulates the release of Ca2’ from intracellular

stores [19,20]. We accordingly tested the ability of lipoproteins

raise [Ca2+], in monolayers of fura-Zloaded

to

human aortic VSMC. The fluorescence

recordings presented in Figure 2 demonstrate that both LDL and HDL, induced rapid and reversible elevations of [Ca2’],. After reaching peak levels within a few seconds, the [Ca”],

rapidly declined to a lower but sustained level that surpassed the initial 1299

Vol.

188,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

[CG+] i (nm)

mo1 min

o-

+

z

Aog U [&+I

PDGF-BB

i (nm) 375-

250125ii

+J-+

04

4

LDL

HDL3

Figure 2. Intracellular calcium response of VSMC to Ang II, LDL,HDL, and PDGFBB. [Ca”], was determined by recording the fluorescence signal from entire cultures of fixa- loaded human aortic VSMC as detailed in “Methods”. Representative tracings from VSMC stimulated with LDL, HDL, (each at 50 pg/ml), Ang II (2 nM) and PDGF-BB (2ng/ml) are presented. The experiment was repeated on at least seven other occasions with comparable results.

basal value. The initial rapid elevation of [Ca2’], was mainly due to mobilization intracellular

of

stores, since it was also observed in Ca2+-free medium (data not shown).

The sustained second phase of the [Ca2+]i response, which was sensitive to chelation of extracellular

CaZt (data not shown), could result from an activation of receptor-

operated Ca2’ channels [21,22], as well as from spontaneous [Ca2’], oscillations (see later, Figure 3). The effects of lipoproteins were compared with those of Ang II and PDGF-BB. There were marked similarities between the actions of lipoproteins

and Ang II with

respect to the kinetics of changes in [Ca2’]i (compare fluorescence tracings in Figure 2). The only consistent difference was that Ang II was a more powerful agonist, and at saturating concentrations

(dose profiles not shown) this peptide induced higher

elevations in [Ca2+li in comparison to lipoproteins. Kinetically, the effect of PDGF-BB on [Ca2’],

was very different from that of Ang II and lipoproteins. The [Ca2’]i only

began to rise after a lag phase of approximately 30 set, and this increase was more gradual, requiring

about 1 min to achieve maximum

(Fig. 2). The decay toward

baseline levels was also more gradual. The differences between Ang II and PDGF with respect to kinetics of changes in [Ca2’], have also been observed in Quin 24oaded rat

aortic VSMC [23]. 1300

Vol.

188,

No.

3,

BIOCHEMICAL

1992

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

[C&+1 i (nm)

O-

H I min

4 LDL

[&+I

i (nm) 300-

150-

O-

H

1

1 Ill,”

HDLJ

Figure 3. Spontaneous

oscillations

of [Caztli

induced

by LDL and HDL,

in VSMC.

[Ca’+], was determined in fura- loaded human aortic VSMC as detailed in “Methods”. The fluorescence signal from single cells was recorded. Typical tracings of single VSMC stimulated with either 50 pg/ml LDL or 50 pg/ml HDL are shown. The results are representative of five experiments with LDL and three with HDL. Recent data suggest that the [Ca2+],-response measured in a suspension or monolayer

of cells differs significantly

from that observed when fluorescence is

registered from a single cell [24]. Cazt -mobilizing

agonists characteristically

induce

oscillatory changes of [Ca’+], in single cells [24]. When the fluorescence signal from single human aortic VSMC was recorded, we observed that both LDL and HD& induced oscillations of [Ca’+]i with a period of about 1 min (Fig. 3). These periodic responses were observed in approximately 80% of single cell measurements and may be classified as transient oscillations [24], since after the peak of increase the [Ca2+], returned to basal levels. On some occasions (data not shown), during the decay phase of the first [Ca2+],-wave we also observed sinusoidal-type oscillations [24]. Thus, the effects of lipoproteins on [Ca2+], at the level of the single cell are very similar to those of Ca2+-mobilizing

hormones.

In addition to hormone-stimulated

mobilization

of intracellular Ca2’, elevations

of [Ca2’]i and subsequent physiological responses, such as aggregation of platelets and endothelium-dependent

relaxation

of vessels, can be induced by Ca2+-ionophores

[25,26]. A number of natural substances possess ionophoric properties [27,28]. Some of these, for example lysolecithin, are normal constituents of lipoprotein particles [29]. Therefore, [Ca’+],-elevating

effects of lipoproteins could be attributed to a nonspecific

increase in plasma membrane permeability. 1301

However, the periodic nature of [Ca2+],

Vol.

188, No. 3, 1992

BIOCHEMICAL 4co-

2 8 8 ES

3GQ-

AND BIOPHYSICAL

0

Control

= m m

PMA Forskolin 8-Br-cCMP

RESEARCH COMMUNICATIONS

1 jg 200.5 o .c G E s loo8 6 2 oLDL

HDL

Ang II

Figure 4. Influence of second messengerslanalogues on the stimulation of phosphoinositide catabolism by LDL, HDL3, and Ang IL Human omental VSMC were prelabelled for 48 hr with 1 pCi/ml myo-[2- 3 HI-inositol under inositol- and serumfree culture conditions. Non-incorporated isotope was removed and VSMC were preincubated for 4 hr at 37OC in serum-free medium either without or with inclusion of 10 PM forskolin, 100 PM 8-Br-cGMP or 1 pM PMA. LiCl (15 mM) was added during the final 30 min of preincubation. VSMC were then incubated for 7 min. in the absence and presence of 50 pg/ml LDL, 50 pg/ml HDL3, or 0.1 FM Ang II. Incubations were terminated by addition of trichloroacetic acid (final 4.6%). The trichloroacetic acid concentration was reduced to 0.8% prior to chromatography of cell lysates on Dowex AG l-X8 anion exchange columns [4, 121. Radioactivity in the inositol monophosphate fraction was measured. Values (mean + SEM, n=5) express the change in [3H] content of inositol monophosphate relative to that content (100%) in control VSMC (not treated with either second messenger compounds or agonists).

responses to LDL and HDb,

as observed in single cell measurements (Fig. 3), would

not support such a nonspecific action of lipoproteins. The oscillatory action of LDL and HDL,

on [Ca2’]i is quite distinct from the action of Ca2+-ionophores,

which

induce a monophasic, irreversible elevation of [Ca2’]i (data not shown). Regukhn

of qonist-stimurcltephphom-twnoverby

C and cy& nucleotidt-&pmdentprotein the hormone-like

activate

ofprvteh

kiime

k&rases. In order to obtain further support for

mechanisms of action of lipoprotein

on VSMC second mesenger

systems, we tested the well known property of hormonal responses to be attenuated by activators of protein kinase C and cyclic nucleotide-dependent 34]. Preincubation

protein kinases [30-

(4 hr.) of human omental VSMC with either phorbol lZmyristate,

13-acetate (PMA), an irreversible activator of protein kinase C [20], forskolin, known to activate adenylate cyclase and thereby increase cyclic AMP concentrations [35], or S-Bromo-cyclic GMP (&Br-cGMP), a cell-permeable analog of cGMP, resulted in a marked inhibition

of lipoprotein-induced

(Fig. 4). As expected, stimulatory

activation of phosphoinositide

catabolism

effects of Ang II on inositol phosphate generation

were also reduced (Fig. 4). 1302

Vol.

188,

No.

3,

1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

CONCLUSION The data presented herein provide evidence that LDL and HDb systems in VSMC via specific (receptor-mediated) mobilizing hormones. Furthermore,

stimulate signalling

pathway(s) analogous to Ca2’-

since the signalling responses elicited by LDL and

HDL, are comparable and occur with equivalent rapidity, we can assume that these events may be independent

of the cholesterol-transporting

properties

of these

lipoproteins. Acknowledgments: Financial support was provided by the Swiss National Foundation, grant nos. 31-29275.90 and 32-30315.90. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Brown WV. (1990) Eur. Heart. J. 11 (Suppl H):15-20 Williams RR, Hunt SC, Hopkins PN, Stults BM, Wu LL, Hasstedt SJ, Barlow GK, Stephenson SH, Lalouel JM and Kuida H. (1988) J. Amer. Med. Assoc. 259: 3579-3586 Block LH, Knorr M, Vogt E, Lecher R, Vetter W, Groscurth P, Qiao B-Y, Pometta D, James R, Regenass M and Pletscher A. (1988) Proc. Natl. Acad. Sci. USA. 85:885-889 Scott-Burden T, Resink TJ, Hahn AWA, Baur U, Box RJ and Buhler FR. (1989) J. Biol. Chem. 264:12582-12589 Gleason MM, Medow MS and Tulenko TN. (1991) Circ. Res. 69: 216-227 Sorisky A, Kucerd GL and Rittenhouse SE. (1990) Biochem. J. 265: 747-754 Brown MS and Goldstein JL. (1986) Sci. 232:34-47 Franceshini G, Maderna P and Sirtori CR. (1991) Atherosclerosis 88:99-107 Libby P, Miao P, Ordavas JM and Schaefer EJ. (1985) JCell. Physiol. 124:1-8 Have1 RJ, Eder HA and Bragdon HH. (1955) J. Clin. Invest. 34:1345-1353 Wregget KA, Howe LR, Moore JP and Irvine RF. (1987) Biochem. J. 245:933934 Resink TJ, Grigorian GY, Moldabaeva AK, Danilov SM and Buhler FR. (1987) Biochem. Biophys. Res. Commun. 144: 438-446 Dean NM and Moyer JD. (1987) Biochem. J. 242:361-366 Ambler SK, Poenie M, Tsien RY and Taylor P. (1988) J. Biol. Chem. 263: 1952-1959 Grynkiewin G, Poenie M and Tsien RY. (1985) J. Biol. Chem. 260~3440-3450 Berridge MJ and Irvine RF. (1989) Nature 341:197-205 Shears SB. (1989) Biochem. J. 260:313-324 Rhee SG, Suh PG, Ryu SH and Lee SY. (1989) Sci. 244: 546-550 Exton JH. (1986) Adv. Cycl. Nucl. Protein. Phosph. Res. 20:21 l-262 Nishizuka Y. (1988) Nature 334: 661-665 Nelson NT, Patlak JB, Worley JF and Standen NB. (1990) Am. J. Physiol. 259: C3-Cl8 Ruegg UT, Wallnofer A, Weir S and Cauvin C. (1989) J. Cardiovasc. Pharmacol. 14 (Suppl. 6): S49S58 Berk BC, Alexander RW, Brock TA, Gimbrone MA and Webb RC. (1986) Sci. 232~87-90 Berridge MJ. (1990) J. Biol. Chem. 165: 9583-9586 White JG, Rao GHR and Gerrard JM. (1974) Am. J. Pathol. 77:135-150 1303

Vol.

188, No. 3, 1992

26. 27. 28. 29. 30. 31. 32. 33. 34. 35.

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

Cohen RA, Zitnay KM, Haudenschild CC and Cunningham LD. (1988) Circ Res. 63: 903-910 Serhan C, Anderson P, Goodman E and Dunham P. (1981) J. Biol. Chem. 256: 2736-274 1 Lecher R, Weisser B, Mengden T, Bnmner C and Vetter W. (1992) Biochem. Biophys. Res. Commun. 183: 156-162 Kugiyama K, Kerns $4, Morrisett JD, Roberts R and Henry PD. (1990) Nature 8: 160-162 Billah MM, Lapetina EG and Cuatrecasas P. (1979) Biochem. Biophys. Res. Commun. 90: 92-98 Nakashima S, Tohmatsu T, Hattori H, Okano Y and Nozawa Y. (1986) Biochem. Biophys. Res. Commun. 135: 1099-1104 Macintyre DE, McNicol A and Drummond AH. (1985) FEBS Lett. 180: 160164 Exton JH. (1988) Rev. Physiol. Biochem. Pharmacol. 111: 116-224 Linden J and Deiahunty TM. (1989) Trends Pharmacol. Sci. 101114-120 Daly JW. (1984) Adv. Cycl. Nucl. Protein. Phosph. 17: 81-89

1304