Proc. Nati. Acad. Sci. USAVol. 91, pp. 1069-1073, February 1994Cell Biology

Lysophosphatidylcholine upregulates the level of heparin-bindingepidermal growth factor-like growth factor mRNA inhuman monocytes

(atherosclerosis, oxidized low density lipoprotein/phospholipase A2)

TORU NAKANO*, ELAINE W. RAINES*, JUDITH A. ABRAHAMt, MICHAEL KLAGSBRUNt, AND RUSSELL Ross*§*Department of Pathology, SM30, University of Washington, Seattle, WA 98195; tScios Nova Inc., Mountain View, CA 94043; and *Departments of Surgeryand of Biological Chemistry and Molecular Pharmacology, The Children's Hospital, Harvard Medical School, Boston, MA 02115

Communicated by Arno Motulsky, October 4, 1993

ABSTRACT Lysophosphatidylcholine is increased in theplasma of hypercholesterolemic patients, is a component ofoxidatively modified low-density lipoprotein, and, as such, mayplay an important role in atherosclerosis. Here we demonstratethat in human monocytes, lysophosphatidylcholine increasesthe level of mRNA encoding the heparin-binding epidermalgrowth factor-like growth factor (HB-EGF), a potent smoothmuscle mitogen. Lysophosphatidylcholine treatment also en-hances the release of heparin-binding mitogenic activity bythese cells in culture. The anti-inflammatory glucocorticoiddexamethasone inhibits the upregulation of HB-EGF mRNAinduced by either lysophosphatidylcholine or bacterial lipo-polysaccharide in cultured monocytes. However, the responsesinduced by lysophosphatidylcholine and by lipopolysaccharidediffer in their kinetics. In addition, the response to lysophos-phatidylcholine is resistant to the action of cycloheximide,whereas the response to lipopolysaccharide is not, suggestingthat the activation mechanisms induced by these two stimuli aredifferent. Since a nuclear run-on assay showed no effect oflysophosphatidylcholine on the transcription of the HB-EGFgene, we speculate that lysophosphatidylcholine may increasethe level of HB-EGF mRNA by altering the processing ordegradation of primary or mature transcripts. Lysophosphati-dylcholine enhancement of monocyte production of HB-EGFmay represent an important result of the interactions amongoxidized low-density lipoprotein and monocyte-derived mac-rophages and may play a role in initiation of smooth muscleproliferation in atherogenesis.

The advanced lesions of atherosclerosis result from intimalaccumulation of smooth muscle cells (SMCs), monocytes/macrophages, connective tissue, and lipids. The increasednumber of SMCs within the intima results from the responseof these cells to a number of growth-regulatory moleculesthat may be released by arterial cells or cells from thecirculation. Since attachment ofmonocytes to the artery walland monocyte infiltration into the intima are among theearliest events in the development of the lesions of athero-sclerosis (1), monocytes/macrophages may be importantearly sources of substances that can induce chemotaxis andproliferation of both monocytes and SMCs.One of the candidate monocyte/macrophage-derived

growth factors for SMCs is a macrophage-derived mitogen,heparin-binding epidermal growth factor-like growth factor(HB-EGF), which stimulates the proliferation of fibroblastsand SMCs and migration of SMCs in vitro. Although theactions of HB-EGF appear to be mediated by binding ofHB-EGF to the epidermal growth factor (EGF) receptor,HB-EGF is much more potent than EGF as a mitogen and

chemoattractant for SMCs (2, 3). Increased expression ofHB-EGF and the EGF receptor has been observed in humanatherosclerotic lesions (J. Miyagawa, S. Higashiyama, S.Kawata, Y. Inui, S. Tamura, K. Yamamoto, M. Nishida, T.Nakamura, S. Yamashita, Y. Matsuzawa, and N. Taniguchi,personal communication).Numerous epidemiologic studies have demonstrated that

chronically elevated levels of plasma lipoproteins, particu-larly low-density lipoprotein (LDL) and very-low-densitylipoprotein, are associated with an increased incidence ofatherosclerosis (4). Intensive lipid-lowering therapy has beenshown to enhance the regression of coronary artery disease(5). However, although hyperlipidemia is important in thedevelopment of atherosclerosis, the mechanisms which linkhyperlipidemia and SMC proliferation remain to be clarified.A large amount of data from experimental animals has

implicated oxidized LDL as a principal lipid mediator oflesion formation in atherosclerosis (6). However, in vitrostudies with oxidized LDL have demonstrated variability inbiological activities among laboratories and different prepa-rations (7). We therefore investigated whether the phospho-lipid lysophosphatidylcholine (lysoPC), a major componentof oxidized LDL, activates monocytes/macrophages andstimulates their production ofgrowth factors. Here we reportthat lysoPC elevates levels of HB-EGF mRNA and stimu-lates the release of heparin-binding mitogenic activity fromhuman monocytes. We discuss possible mechanisms for theproduction oflysoPC in the lesions of atherosclerosis and thepossible role of lysoPC in initiation of the proliferativeresponse by SMCs.

MATERIALS AND METHODSPreparation and Culture of Human Monocytes. Egg yolk

lysoPC, synthetic palmitoyl and stearoyl lysoPC, cyclohex-imide, and dexamethasone were obtained from Sigma. Hu-man monocytes were isolated from buffy coats obtained fromthe Puget Sound Blood Center. Buffy coats were diluted withendotoxin-free phosphate-buffered saline to 1 x 107 leuko-cytes and 3-4 x 109 erythrocytes per milliliter, and 10-mlaliquots transferred to LeukoPrep tubes (Becton Dickinson).The tubes were spun at 1000 x g for 25 min at roomtemperature. The cell layer containing monocytes was col-lected and washed. Cell preparations with <90% monocyteswere not used, and the preparations generally consisted of92% monocytes, 6% lymphocytes, and 2% neutrophils.Monocytes (2 x 107 cells) were cultured in 12 ml of RPMI1640 (BioWhittaker) containing 1o human plasma-derived

Abbreviations: SMC, smooth muscle cell; EGF, epidermal growthfactor; HB-EGF, heparin-binding EGF-like growth factor; LDL,low-density lipoprotein; lysoPC, lysophosphatidylcholine; LPS, li-popolysaccharide; LCAT, lecithin-cholesterol acyltransferase.§To whom reprint requests should be addressed.

1069

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

July

21,

202

1

Proc. Natl. Acad. Sci. USA 91 (1994)

serum. The plasma-derived serum and RPMI 1640 had <1endotoxin unit/ml (Limulus amebocyte lysate chromogenicassay BioWhittaker). To prevent attachment of the mono-cytes to plastic, the 10-cm dishes were coated with 2 ml of12% Hydron (Interferon Sciences, New Brunswick, NJ) inethanol.

Culture of SMCs. Human aortic SMCs were isolated fromnewborn human thoracic aortas (8) and cultured in Dulbec-co's modified Eagle's medium (DMEM) with 10%o fetalbovine serum. Cells were grown to confluence and changedto DMEM with 1% human plasma-derived serum (8) for 48 hrprior to experiments.

Assay of Heparin-Binding Mitogenic Activity with SwissMouse 3T3 Cells. Culture supernatants from 2 x 107 mono-cytes were collected, diluted 2-fold in 0.05 M Tris HCl (pH7.4), and loaded onto 0.3-ml heparin-Sepharose columns(Pharmacia). After extensive washing ofthe columns with thesame buffer containing 0.2 M NaCl, the samples were elutedwith 2 ml of 0.05 M Tris HCl, pH 7.5/3 M NaCl, desalted inCentricon-10 miniconcentrators (Amicon), and assayed formitogenic activity on Swiss 3T3 cells (9).

Radloreceptor Assay for EGF. The radioreceptor assay wasperformed by simultaneous incubation ofsample aliquots and125I-labeled EGF with A-431 human epidermoid carcinomacells (10).Northern Blotting. Total cellular RNA was extracted (11),

and samples (10 pg per lane) were run in 1% agarose/2.2 Mformaldehyde gels and transferred to nylon membranes (Ny-tran, Schleicher & Schuell) (12). After UV crosslinking,reversible RNA staining with methylene blue was performedto check the amount and quality ofRNA (13). After destain-ing, the membranes were subjected to hybridization (14) withprobes labeled with [t-32P]dCTP (Amersham) by means of arandom primer labeling system (Amersham). After hybrid-ization, the membranes were washed with 2x standard salinecitrate (SSC)/0.1% SDS at room temperature and with 0.2xSSC/0.1% SDS at 650C prior to autoradiography. For rep-robing, blots were stripped with two 30-min washes with0.02x SSC/0.1% SDS at 900C. All blots were rehybridizedwith (3-actin cDNA or glyceraldehyde-3-phosphate dehydro-genase cDNA probes to verify the mRNA load.

Transcriptional Analysis. A nuclear run-on assay was con-ducted (15) with some modifications. Monocytes (8 x 107cells per reaction mixture) were cultured for 3 or 24 hr in thepresence or absence of 75 mM stearoyl lysoPC and thencollected by centrifugation at 1000 x g for 10 min. Each cellpellet was suspended in 4 ml of lysis buffer [10mM Tris HCl,pH 7.4/10 mM NaCl/3 mM MgCl2/0.5% Nonidet P-40/2.75mM dithiothreitol with RNasin (Promega) at 20 units/ml],placed on ice for 5-8 min, and centrifuged at 270 x g for 5 miat 40C. The supernatants were removed, and each pellet wassuspended in 4 ml of lysis buffer without Nonidet P-40.Nuclei were spun down at 270 x g for 5 min at 40C. Thenuclear pellet from each sample was suspended in 200 pl ofstorage buffer [50 mM Tris HCl, pH 8.3/40%o (vol/vol) glyc-erol/5 mM MgCl2/0.1 mM EDTA] and stored at -70°C untiluse. Each nuclear run-on reaction mixture (400 pl) contained200 ,l of nuclei, 0.625 mM ATP, 0.312 mM GTP, 0.312 mMCTP, 250 mCi of [y-32P]UTP (>3000 Ci/mmol, New EnglandNuclear; 1 Ci = 37 GBq), 40 mM Tris HCl (pH 8.3), 150 mMNH4C1, 7.5 mM MgCl2 and RNasin at 200 units/ml. Thismixture was incubated for 30 min at 30'C. Radiolabeled RNAwas isolated by the same method used for Northern blotting,subjected to limited base hydrolysis (0.2 M NaOH on ice for3 min), ethanol precipitated, and suspended in 1 ml ofhybridization buffer at 107 cpm/ml. Unlabeled cDNA probes(5 pg per blot) for hybridization with run-on products weredenatured in 0.3 M NaOH at 650C for 1 hr, dot blotted ontoa Nytran membrane, hybridized for 45 hr and washed asdescribed for Northern blotting.

RESULTSLysoPC Upregulates mRNA Levels in Monocytes but not in

SMCs. Freshly isolated human monocytes were incubated insuspension in Hydron-coated dishes for 24 hr with variousconcentrations of egg lysoPC dissolved in ethanol. Expres-sion ofHB-EGFmRNAwas then examined byRNA blotting.Incubation of monocytes with egg lysoPC at 12.5-75 jg/mlfor 24 hr upregulated the level ofHB-EGF mRNA (Fig. 1). Inthis experiment, 37.5 ug/ml was the optimal concentration oflysoPC for induction of HB-EGF mRNA (4.1-fold, fromdensitometric scans of the films that included a film fromreprobing the same blots with (-actin cDNA to normalize forload differences). Although 1-actin mRNA levels were notaffected, there was some variation in the optimal concentra-tion, as well as the degree of mRNA upregulation, betweenexperiments (data not shown). This may be due in part tovariable uptake and modification of lysoPC by the cells,which may alter the concentration of lysoPC in the medium.Neither lysoPC nor the diluent (ethanol) had any effect on theviability of the cells as measured by trypan blue exclusion.Although HB-EGF was originally described as a mono-

cyte/macrophage-derived growth factor, it is also producedby SMCs (16, 17). In contrast to monocytes, the HB-EGFmRNA level in SMCs was not affected by lysoPC (Fig. 1).The Activity of LysoPC in Upreglatlon ofHB-EGF mRNA

Is Dependent on Its Fatty Acid Composition. The major fattyacids in mammalian phospholipids and in the egg yolk lysoPCused in our experiments are palmitic acid and stearic acid. Wecompared the activity of synthetic palmitoyl lysoPC andstearoyl lysoPC on the upregulation ofHB-EGF mRNA. Fig.2 is representative ofthe results offour experiments in whichstearoyl lysoPC upregulated the mRNA level ofHB-EGF butpalmitoyl lysoPC failed to do so. This was observed with twodifferent lots of the synthetic lysoPC.Screfton of HB-EGF-Like Activity from Monocytes Is In-

creased by LysoPC. To examine the release of growth-stimulatory molecules from monocytes, we used heparin-sepharose to fractionate the conditioned media of lysoPC-treated and control monocytes. Incubation of monocyteswith lysoPC significantly increased the heparin-binding mi-togenic activity in the conditioned medium, and the release ofmitogenic activity was well correlated with the HB-EGFmRNA level (Fig. 3). Since, in addition to binding heparin,HB-EGF competes with 125I-EGF for binding to the EGFreceptor, we also assayed EGF competitive activity in theheparin-binding fraction from monocyte-conditioned media.

M onocvtes

SMCS

.:.-X(2 37 5 v

L ysol(' (1grini)

FIG. 1. Effect of lysoPC on HB-EGF mRNA levels in humanmonocytes and arterial SMCs. Human monocytes or human arterialSMCs were incubated with various concentrations of egg lysoPC orwith vehicle for 24 hr. HB-EGF mRNA levels were analyzed asdescribed in Materials andMethods. No variations in P-actinmRNAlevels were observed. These data are representative offour replicateexperiments.

1070 Cell Biology: Nakano et al.

Dow

nloa

ded

by g

uest

on

July

21,

202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 1071

P-ActinNone LysoPC LPS

(A)

up

HB-EGF

_

DEX - + - + - +

0 12.5 25 37.5 12.5 25 37.5 (Wg/m1)Stearoyl PalmitoyllysoPC lysoPC

FIG. 2. Role of the fatty acid species in lysoPC stimulation ofHB-EGF. Human monocytes were incubated for 24 hr with syntheticstearoyl lysoPC, palmitoyl lysoPC, or vehicle. The data are repre-sentative of four replicate experiments. RNA load was evaluated byreprobing the same blot with P-actin cDNA.

The conditioned medium of monocytes incubated withlysoPC contained 12 times more heparin-binding EGF com-

petitive activity than the conditioned medium of monocytesincubated without lysoPC (basal, 0.011-0.019 ng per 106cells; lysoPC-treated, 0.14-0.23 ng per 106 cells).LysoPC and LUpopolysaccharide (LPS) Effects on HB-EGF

mRNA Levels Appear to Be Stimulated via Distinct SignalingMechanisms. HB-EGF mRNA was also increased in humanmonocytes treated with LPS, a bacterial endotoxin (Fig. 4).To further characterize the mechanism of mRNA increase,we examined the effects ofseveral agents on the upregulationof HB-EGF mRNA by either lysoPC or LPS. First, we

examined the effect of dexamethasone, a synthetic glucocor-ticoid, on HB-EGF mRNA levels. Dexamethasone com-

pletely inhibited both lysoPC- and LPS-induced upregulationofHB-EGF mRNA (Fig. 4A). Consistent with these results,dexamethasone has also been shown to downregulate theexpression of HB-EGF mRNA in both unstimulated andthrombin-stimulated cultured human SMCs (17). This sug-

gests that dexamethasone inhibits increases in HB-EGF

= 200000

O 100000 0 37.5 50

.=d p LysoPC (tg/ml))

0

NA CM 0 37.5 50

LysoPC (jg/ml)

FIG. 3. Release ofheparin-binding mitogenic activity from mono-cytes. Human monocytes were incubated for 24 hr with the indicatedconcentrations of egg lysoPC or with vehicle. Conditioned mediumand cells were separated by centrifugation. Heparin-binding mito-genic activity in the conditioned medium was measured as describedin Materials and Methods. Insert shows the expression of HB-EGFmRNA in the same cells. NA, pH]thymidine incorporation into 3T3cells with vehicle addition; CM, [3H]thymidine incorporation into3T3 cells stimulated with heparin-binding proteins extracted fromculture medium. Values plotted are means of triplicate determina-tions and represent data obtained in three replicate experiments.

CHX - + - + - +

FIG. 4. Effect of dexamethasone (DEX) and cycloheximide(CHX) on enhancement ofHB-EGFmRNA levels by lysoPC or LPS.Human monocytes were incubated for 24 hr with stearoyl lysoPC(37.5 ,u/ml) or LPS (Sigma, Cat. #L-2880; 100 ng/ml) in thepresence or absence of DEX (1 mM) (A) or CHX (1 pg/ml) (B).HB-EGF mRNA levels were analyzed. No alterations in P-actinmRNA levels were observed. Comparable changes in mRNA levelswere observed in three other experiments.

mRNA with distinct activation patterns in different celltypes.We next determined whether de novo protein synthesis

was necessary for the upregulation of HB-EGF mRNAobserved at 24 hr. Cycloheximide, an inhibitor of proteinsynthesis, did not inhibit the lysoPC enhancement of HB-EGF mRNA, whereas it completely blocked the LPS effect(Fig. 4B). The level ofHB-EGF mRNA was also enhanced byinterleukin 1P, and this effect was inhibited by cycloheximide(data not shown).To further examine the different cycloheximide suscepti-

bilities of the lysoPC and LPS effects, we studied the timecourse of regulation ofHB-EGF mRNA levels in lysoPC- orLPS-stimulated monocytes (Fig. 5). The level ofthe HB-EGFmRNA was relatively high just after the preparation ofmonocytes (at 0 hr) and remained relatively high. However,in the absence of lysoPC or LPS, the level decreased duringfurther incubation. The level of HB-EGF mRNA immedi-ately following isolation was variable among donors but didnot appear to be due to endotoxin contamination of reagents,which were all monitored. In the presence of lysoPC, upreg-ulation was observed at 3 hr and was maintained for at least24 hr. Cycloheximide did not affect the time course of thelysoPC-induced upregulation (data not shown). On the otherhand, in the LPS-stimulated monocytes, the mRNA level

None LysoPC

t .. 1..

0 3 8 16 24 3 8 16 24

LPS CHX + LPS

0 3 8 24 1 3 8 24Time (h)

FIG. 5. Time course of increase in HB-EGF mRNA. Humanmonocytes were incubated for the indicated times with stearoyllysoPC (37.5 pg/ml) or LPS (100 ng/ml) in the presence or absenceofcycloheximide (CHX, 1 1g/ml). HB-EGF mRNA was analyzed as

described and represents three separate experiments. P-ActinmRNA levels did not vary over the time course examined.

(B)

Cell Biology: Nakano et A

11 "m.in.l., :.

w:: ..TtI,.P w 00

Dow

nloa

ded

by g

uest

on

July

21,

202

1

Proc. Natil. Acad. Sci. USA 91 (1994)

reached a maximum at 1 hr, decreased to a low level at 3 hr,and increased again after 8 hr. In the presence of cyclohex-imide, LPS induced the first peak but did not induce thesecond, indicating that the second increase in HB-EGFmRNA was dependent on de novo protein synthesis.

Effect ofLysoPC on Transcription of the HBB-EGF Gene andStability ofBB-EGF mRNA. To examine the effect of lysoPCon transcription of the HB-EGF gene, we performed nuclearrun-on assays (Fig. 6). LysoPC did not enhance transcriptionat 3 hr or at 24 hr, whereas dexamethasone inhibited HB-EGFmRNA transcription.We also examined the effect of lysoPC on mRNA stability

in the presence ofan inhibitor ofRNA synthesis, actinomycinD (Fig. 7). LysoPC had no effect on the decrease ofHB-EGFmRNA observed in the presence of actinomycin D.

DISCUSSIONLysoPC and HB-EGF as Potential Mediators of the Devel-

opnment of Atherosclerotic Lesions. It has been reported thatlysoPC is a chemoattractant for human monocytes (18) andthat it enhances expression of the cell adhesion moleculesVCAM-1 and ICAM-1 in cultured human umbilical veinendothelial cells (19). Such an- effect could increase theadherence of monocytic cells to the endothelium if it oc-cuffed in vivo. The data reported here demonstrate thatlysoPC stimulates monocytes to release EGF-like heparin-binding mitogenic activity and to upregulate mRNA of HB-EGF, a potent mitogen for SMCs.

Heparin-binding EGF is a member ofa family of moleculesstructurally similar to EGF that include transforming growthfactor a, amphiregulin, and the poxvirus growth factors (20).In addition to sharing structural similarities, these distinctgene products all bind with high affinity to the EGF receptor.However, HB-EGF stimulates SMC proliferation to a levelcomparable to that observed in'response to platelet-derivedgrowth factor and is much m"re potent than other membersof the EGF family (3). This appears in part to be due to itsability to bind heparin, although this is also a property ofamphiregulin, which binds the EGF receptor with a loweraffinity than other EGF family members (21). The enhancedresponse to HB-EGF observed in SMCs appears to bemediated by a dual receptor system requiring interaction withboth cell surface glycosaminoglycans and the EGF receptor(22), similar to the dual receptor system required for activityofbasic fibroblast growth Actor (23). HB-EGF may also havea distinct receptor in addition to the EGF receptor.A recent survey of human thoracic and abdominal aortae

obtained from individuals from 2 months to 86 years of agedemonstrated regulated expression of HB-EGF protein (J.Miyagawa et al., personal communication). The medialSMCs of the aorta in neonates and infants constitutivelysynthesized HB-EG- protein, whereas HB-EGF-producingSMCs were rare in young and middle-aged adults. However,

(A) 3 h (B) 24 hNone IlvsoP(C None ILvsoPC

HB-EGiF

[-z-\ctin

(C) 3 h-DEX --DELX

K HB-EGF

ac ; GAPDH _ mm

FiG. 6. Effect of lysoPC on transcription of the HB-EGF gene.Human monocytes were incubated for 3 hr (A) or 24 hr (B) with orwithout stearoyl lysoPC (37.5 pg/ml) or for 3 hr with or withoutdexamethasone (DEX, 1 mM) (C). Transcription of the HB-EGF,,-actin, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH)genes was analyzed. In each expernient, enhancement of the HB-EGF mRNA level by lysoPC was confirmed by Northern blotting(data not shown). The data were coqnfimed in four separate exper-iments.

None LVSoP(`

FIG. 7. Effect of lysoPC on degradation of HB-EGF mRNA.Human monocytes were cultured for the indicated times with orwithout stearoyl lysoPC (37.5 yg/ml) in the presence of 1 mMactinomycin D. HB-EGF mRNA levels were analyzed. These dataare representative of three replicate experiments.

in atherosclerotic plaques, marked production of HB-EGFprotein was detected in both SMCs and macrophages. Inaddition, EGF receptors were detected on SMCs in thelesions, many of which were positive for proliferating-cellnuclear antigen, an index of cycling cells. Thus, the associ-ation of HB-EGF protein and the EGF receptor with lesionsof atherosclerosis is consistent with a potential role forHB-EGF in SMC migration and proliferation, characteristicof these developing lesions.Mechanism of Upregulation of HB-EGF mRNA. HB-EGF

mRNA in monocytes can be upregulated by lysoPC as wellas by LPS. However, the differences in time course andsusceptibility to cycloheximide suggest that lysoPC and LPSupregulate HB-EGF mRNA via different mechanisms. Inhi-bition of the second peak of LPS-induced HB-EGF mRNAexpression by cycloheximide suggests that the second re-sponse requires protein synthesis, whereas upregulation bylysoPC is independent of protein synthesis.

In evaluating the mechanism of lysoPC alteration of HB-EGF mRNA levels, we performed nuclear run-on assays toinvestigate transcriptional regulation. We could not detectenhancement of transcription of HB-EGF at 3 or 24 hr afterstimulation with lysoPC, although we did detect downregu-lation by the glucocorticoid dexamethasone (Fig. 6). As notedabove, lysoPC enhancement ofHB-EGF mRNA levels is notdependent upon protein synthesis, nor is the stability ofmature mRNA altered in the presence of the RNA synthesisinhibitor actinomycin D. It is possible that lysoPC may alterthe processing or degradation ofprimary or mature HB-EGFmRNA, but this effect may be inhibited by actinomycin D.This lysoPC inhibition of HB-EGF mRNA degradation orprocessing would depend upon the stability of some form ofRNA. Further experiments will be required to determine thenature of'the RNA and whether this suggestion is correct.Formation ofLysoPC in Atherosclerosis. Hydrolysis offatty

acid on the sn-2 position of phospholipids is a key step in theformation of lysophospholipids. Lecithin-cholesterol acyl-transferase (LCAT; phosphatidylcholine:sterol acyltransfer-ase, EC 2.3.1.43) can hydrolyze the sn-2 fatty acid of phos-phatidylcholine, transfer the fatty acid to cholesterol (24),and produce lysoPC. LCAT is present in plasma, and high-density lipoprotein enhances the catalytic activity of LCAT(25). Higher LCAT activity has been observed in the plasmaof atherosclerotic patients, who also have higher concentra-tions of lysoPC (26), suggesting that LCAT may be a majorfactor in the production of lysoPC in these patients.

Oxidative modification of LDL has been proposed as apossible contributor to the development of atherosclerosis(6). In vitro, oxidized LDL stimulates expression of colony-stimulating factors in endothelial cells (27) and acts as achemoattractant for human monocytes (28). However, invitro studies with oxidized LDL have demonstrated variabil-ity in biological activities among laboratories and differentpreparations (7). We have, therefore, focused these studieson a major component of oxidized LDL. Oxidation of LDLis a major source of phospholipids, especially in the plasma

1072 Cell Biology: Nakano et al.

Dow

nloa

ded

by g

uest

on

July

21,

202

1

Proc. Natl. Acad. Sci. USA 91 (1994) 1073

of hyperlipidemic individuals. During the oxidation of LDL,as much as 40o of the phosphatidyicholine of LDL may beconverted to lysoPC (29).The biological activities of lysoPC presented here and in

other papers (18, 19) suggest that lysoPC may contribute tothe atherogenic activities ofoxidized LDL. It has been shownthat apolipoprotein B in LDL possesses phospholipase A2activity. Thus, the increased lysoPC content ofoxidizedLDLmay be due to the hydrolysis of phosphatidylcholine in LDLby the phospholipase A2 activity of apolipoprotein B (30).Since LDL can be oxidatively modified by the cells in oraround the lesions of atherosclerosis (31-33), and sinceoxidatively modified LDL is present in the lesions (34-37),lysoPC in oxidized LDL may be important in the activationof monocytes/macrophages in the lesions.The Fatty Acid Chain Length of LysoPC Is Important. The

signal trasduction mechanisms responsible for the upregula-tion ofHB-EGF mRNA in human monocytes by lysoPC maybe an important issue to study. HB-EGF can act as achemoattractant for human monocytes, and it has beenpostulated that generation of diacylglycerol, an activator ofprotein kinase C, through a mechanism involving lyso-phospholipase C, may be involved in lysoPC-induced chemo-taxis (38). However, our preliminary experiments using asynthetic diacylglycerol did not demonstrate upregulation ofHB-EGF mRNA in human monocytes (data not shown). Thissuggests, therefore, that, different from chemotaxis, HB-EGF mRNA induction may not involve protein kinase C.Our data showed that a small difference in the fatty acid

chain length of lysoPC caused a profound difference inpotency of lysoPC in upregulating HB-EGF mRNA. StearoyllysoPC upregulated the mRNA, but palmitoyl lysoPC did not.Although the precise mechanism for the action of lysoPC isnot clear, a small difference in the physical property oflysoPC appears to be important for upregulation of HB-EGFmRNA in monocytes. LysoPC also induces expression of anadhesion-molecule gene in human endothelial cells (19). Inthis case, however, palmitoyl lysoPC induces the expressionof the adhesion molecule. This difference in effect of fattyacid chain length on these two human cells may suggest thatthe molecular mechanism for HB-EGF induction in mono-cytes is different from the mechanism that induces adhesion-molecule expression in endothelial cells.To further clarify the importance of lysoPC in vivo, it will

be necessary to develop specific agents that inactivatelysoPC or inhibit LCAT and phospholipase A2. Such agentsmay be capable of altering the development of lesions ofatherosclerosis as well as other inflammatory fibroprolifera-tive responses.

We thank Karen Engel, Li-Chuan Huang, and Jon Dickey fortechnical assistance. We gratefully acknowledge Shigeki Higash-iyama for generously providing data on the expression of HB-EGFin atherosclerotic lesions prior to publication. This work was sup-ported in part by National Institutes of Health Grant HL47151 toR.R. and E.W.R., an unrestricted grant for cardiovascular researchto R.R. from Bristol-Myers Squibb Company, and National Insti-tutes of Health Grant GM47397 to M.K.

1. Ross, R. (1986) N. Engl. J. Med. 314, 488-500.2. Besner, G., Higashiyama, S. & Klagsbrun, M. (1990) Cell

Regul. 1, 811-819.3. Higashiyama, S., Abraham, J. A., Miller, J., Fiddes, J. C. &

Klagsbrun, M. (1991) Science 251, 936-939.4. Inkeles, S. & Eisenberg, D. (1981) Medicine (Baltimore) 60,

110-123.5. Brown, G., Albers, J. J., Fisher, L. D., Schaefer, S. M., Lin,

J.-T., Kaplan, C., Zhao, X.-Q., Bisson, B. D., Fitzpatrick,V. F. & Dodge, H. T. (1990) N. Engl. J. Med. 323,1289-1298.

6. Steinberg, D., Parthasarathy, S., Carew, T. E., Khoo, J. C. &Witztum, J. L. (1989) N. Engl. J. Med. 320! 915-924.

7. Malden, L. T., Chait, A., Raines, E. W. & Ross, R. (1991) J.Biol. Chem. 266, 13901-13907.

8. Ross, R. & Kariya, B. (1980) in Handbook ofPhysiology-TheCardiovascular System II: Circulation. Vascular Smooth Mus-cle, eds. Bohr, D. F., Somlyo, A. P. & Sparks, H. V. (Am.Physiol. Soc., Bethesda, MD), pp. 69-91.

9. Raines, E. W. & Ross, R. (1985) Methods Enzymol. 109,749-773.

10. Madtes, D. K., Raines, E. W., Sakariassen, K. S., Assoian,R. K., Sporn, M. B.; Bell, G. I. & Ross, R. (1988) Cell 53,285-293.

11. Chomczynski, P. & Sacchi, N. (1987) Anal. Biochem. 162,156-159.

12. Wahl, G. M., Meinkoth, J. L. & Kimmel, A. R. (1987) Meth-ods Enzymol. 152, 572-581.

13. Herrin, D. L. & Schmidt, F. W. (1988) BioTechniques 6, 196-199.

14. Church, G. M. & Gilbert, W. (1984) Proc. Natd. Acad. Sci.USA 81, 1991-1995.

15. Coijay, M. H., Blank, R. S. & Owens, G. K. (1990) J. Cell.Physiol. 145, 391-397.

16. Dluz, S. M., Higashiyama, S., Damm, D., Abraham, J. A. &Klagsbrui, M. (1993) J. Biol. Chem. 268, 18330-18334.

17. Nakano, T., Raines, E. W., Abraham, J. A., Wenzel, F. G.,IV, Higashiyama, S., Klagsbrun, M. & Ross, R. (1993) J. Biol.Chem. 268, 22941-22947.

18. Quinn, M. T., Parthasarathy, S. & Steinberg, D. (1988) Proc.Natl. Acad. Sci. USA 85, 2805-2809.

19. Kume, N., Cybulsky, M. I. & Gimbrone, M. A., Jr. (1992) J.Clin. Invest. 90, 1138-1144.

20. Carpenter, G. & Cohen, S. (1990) J. Biol. Chem. 265, 7709-7712.

21. Shoyab, M., Plowman, G. D., McDonald, V. L., Bradley,J. G. & Todaro, G. J. (1989) Science 243, 1074-1076.

22. Higashiyama, S., Abraham, J. A. & Klagsbrun, M. (1993) J.Cell Biol. 122, 933-940.

23. Klagsbrun, M. & Baird, A. (1991) Cell 67, 229-231.24. Glomset, J. (1962) Biochim. Biophys. Acta 65, 128-135.25. Fielding, C. J., Shore, V. G. & Fielding, P. E. (1972) Biochem.

Biophys. Res. Commun. 46, 1493-1498.26. Wells, I. C., Peitzmeier, G. & Vincent, J. K. (1986) Exp. Mol.

Pathol. 45, 303-310.27. Rajavashisth, T. B., Andalibi, A., Territo, M. C., Berliner,

J. A., Navab, M., Fogelman, A. M. & Lusis, A. J. (1990)Nature (London) 344, 254-257.

28. Quinn, M. T., Parthasarathy, S., Fong, L. G. & Steinberg, D.(1987) Proc. Natd. Acad. Sci. USA 84, 2995-2998.

29. Parthasarathy, S., Steinbrecher, U. P., Barnett, J., Witztum,J. L. & Steinberg, D. (1985) Proc. Natl. Acad. Sci. USA 82,3000-3004.

30. Parthasarathy, S. & Steinberg, D. (1990) Proc. Natd. Acad. Sci.USA 87, 9741-9745.

31. Henriksen, T., Mahoney, E. M. & Steinberg, D. (1981) Proc.Natl. Acad. Sci. USA 78, 6499-6503.

32. Heinecke, J. W., Rosen, H. & Chait, A. (1984) J. Clin. Invest.74, 1890-1894.

33. Parthasarathy, S., Young, S. G., Witztum, J. L., Pittman,R. C. & Steinberg, D. (1986) J. Clin. Invest. 77, 641-644.

34. Haberland, M. E., Fong, D. & Cheng, L. (1988) Science 241,215-218.

35. Palinski, W., Rosenfeld, M. E., Yld-Herttuala, S., Gurtner,G. C., Socher, S. A., Butler, S. W., Parthasarathy, S., Carew,T. E., Steinberg, D. & Witztum, J. L. (1989) Proc. Natd. Acad.Sci. USA 86, 1372-1376.

36. Boyd, H. C., Gown, A. M., Wolfbauer, G. & Chait, A. (1989)Am. J. Pathol. 135, 815-825.

37. Rosenfeld, M. E., Palinski, W., Yla-Herttuala, S., Butler, S. &Witztum, J. L. (1990) Arteriosclerosis 10, 336-349.

38. Quinn, M. T., Kondratenko, N. & Parthasarathy, S. (1991)Biochim. Biophys. Acta 1082, 293-302.

Cell Biology: Nakano et A

Dow

nloa

ded

by g

uest

on

July

21,

202

1