Tobacco Documents Online

I I I I I I I I I I I I I I Ae+sPum Jo4rrul of Pmholoo. VbL 239, rb. 6s Datismt+n 1991 Cxg,.O.o."W%= An=WzWn CfPA6Dk8= WARNER-LAMBERT/PARKE-DAVIS AWARD LECTURE Viral Pathogenesis of Atherosclerosis Impact of Molecular Mimicry and Viral Genes David P. kajjar Prorrt dtt DE.pm-nea of BiocbernEspy and ParbohV, CrvxeH lfrriuastty Medtral Calkge, Nau York, New Yai Xurnan atibervgenaYs Lr a plafotsopic process uritb an wrdeflned catr.se. Several pathologic factors bave been lixkeid to tbe dLuase procesr, including asterial injury or activation of tbe endotbeliu"4 uticb may initiate proatberosclerotic eventa !ac Oe uessel u41L Atbeiosclerotic lesiosss aee rbaracterized in parS by tbe pracenoe of activated immune ul4 abnormal cell prollferardost, and alttreid cholesterpl metabo• ilsrn Tbese activated lmmunocompetent cells in Plaques produce vasoactive m.diators tbat can alter bomaostasic and maypromote the artsrlopotb}: Botb molecttlarand structtssrrl evidence irpserented thas berpesvirreses, by uaay of inductron of altened gene function and cellular cboGxterol metaboline, aou- pled wit#+ tbeir ability to activate coagulation and a monocyte r+eceptor on tbe infected eridotliellum, are inuolued in ma}orpatbogenic events assocYarrd rvr;tb atiKrosclerotis and d"rombaele Work from the au- tbor't taboratory, as unell as from otber researcb group; bavv sboum t6at avzan and buff&= bespetvi- rusas act rpectifiaally to induce alterations to the sur• ,Jizce and inner layers of the blood ueael u+a11 tfiat rnay pradispose to atbesarclerosis and its attendent dtnioal compii=tioss (AmJPadoi 1991. 13,9:1195- 1211) An important fsnure of the penesis of human atheroscle- rosis is its muttifacmrial nature. No single errtity has been tinked diroctty to its pathloQenesis in.,tMr animal models of tho disease or in the human art.riopaihy. Eviderlce is accurnulatirp that eady deveioprrwtW changes in spe- c'iflc vascular beds highly susceptlble to the disease are paramount to the laiter alterations of arterial metabolism that predisposes to subsequent atheromatous chanpas.'-'3 Utxloubtedly these changes are amplified by risk factors such as tlypercttiolesterolemia, hypenelr sion, smoiat~, and diabetes. For almost 20 years, interest in a viral cause of ath- erosderosis has continued, spurred by the oariy worlc of Bersditt and Bendrtf and Fabricant et a1.5° This perse- verance to define the vi4 cytopatfloJoyic mechanisms in4oived has certainly been enhanced by studies of an tulimal model ot viraily induced atherosclerosis under normocholesterolemic conditions,°"° and by the recog- nition of the association of this disease in patients who have underQore cardiac traalsplant surgery, butwho elso are infected with cyrprregalovirus,'a,' one of the seven herpesvin~ses now known to infect humans.12 Vascular injury purported to be invofved in early stages of the arteriopathy is also a corrrnon feature of acuta herpesvirus infection. Recent studies have identi- fied herpesvirus antigens and nucleic acids in the vas- cuiar vwaji"-'° after their passage through the endothe- lium. At the biochemical level, a variety of herpesvirus- induced paiflobiobgic changes have been documented within va,scular cells that impact on cellular lipid (choles- terol)° and connective tissue (proteoqlycans) metabo- lism.'° In fact, ttse harpesvirus-induced atteratiorls in lipid metabdism in vifro and in vivo paralW those patholcgic events rtininiscent of the artetiopatty found in the human disoase,cso4-2 Fi.rpecvinises induce an arterial lesion characteristic of leukocytodastic vasculibs. There is granuloeyte infiltra- tion in the artery, accompanied by ttuombin formadon AOCEQeO lor OubiQGm Au]ttQ ®.1491. AEeta=rsprirt r.4mrslo Dr. DariG P. M.pr. Prof®ordBbCtfar). iftry and P&p1?gy, COrT1d Unir.r.¢y MedP1! CObepe, 1300 YUk Aw- rue, nlea YoaK NY 10321. 1195 I

1196 HaiJar .SlpDet.nnbe+ 1.992. Vol. 139. No, 6 and fibrin deposition.2` Hence this veciar has the ability to influence not only atherosclerotic chatlges but also processes related to a thrombotic diathesis. Because the virus causes a fattyy-proliferative lesion in animals that oo- casionaly contains microthrombin,'such events support the hypothesis that surfaca expression of adhesion mof- ecules and intimal hyperpiasia may be initiated by a vi- relly induced event. This may trigQer cell proliferation by enhancement of growth factor or by cyrolune release from vascular cells or adherent macrophages, Or it alro may alter the balance between mitogen and growttt sup- pressor synthesis in the vascular well. Based on this background informaGon, both molecu- lar and structural evidence are highlighted to support the hypothesis that herpesviruses, by akering gene function and cellutar lipid metabolisrn, coupled with the abiliry to activate the coa8ulation system and a morxx,yte recep- tor on the endothefium, are invoived in pathogenic events associated with human atherosclerosis. Information about candidate virat or transforming genes potentialiy related to atherogenests a(so are presented. Epidemiologic Evidence Linking Herpesvirus and Atherosclerosis Herpesvirus infections are widespread in the general populatioon.2r-a' They are ubiquitous viruses found in most tissues of the body, even in blood ceils, By 10 years of age, more that15096 of children dertnnstrate antibod- ies to herpes simplex virus (H" type 1.25 The incidence of herpes simplex vinus, type 2. is more difficult to esti- mate, but two studies cite a prevalence of up to In the 35- to 44year-0id age group, substantially more than 1 t?96 demonstrate arrbbodies to HSV type 2,26-2' A more recent study demonstraied that the incidence of infection approaches 209'o in the population aged 15 to 74 years.27 It has now been documented that seven drf- ferent herpesvirus can infect hurnans.'2 Human atherosclerosis is not completeiy accounted for byall krown risk factors. Th;s suggests that tfle human risk factors may by synergistic with other, as of yet un- identi8ed, initiating factors.'•26 A direct seroepidernio- logic link between cytomegalovirus (CMV) infection and atherosclerosis has been suggosted.2° A cas¢.cantrof study was perforrrled wherein patients who underwent rardiovascular surQery were compared with a control group of subjects wittt similar cholesterol levels and epi- demiolog ic factors but who were not undergoing surgery. In approximatefy 160 pairs of patients, the prevalence of CMV antibodies was higher in the surgical group than in the control group (90% and 74%, respectively), and a greater percentage of surgical cases than controls had high trters of CMV antibodies (57% and 26%, respec- tively), There was no correlation between arrtibody titers and blood levels of cholesteroi or triplycerides. Interest ingly a recent Fremingham Heart Study failed to find an overall a5sociation between fevef blisters or cold sores, with a 6-year incidence of caronary heart disease in pa- tients 58 to 89 years of age.30 Previous work in this group of patienis had shown a strong correlation of a self- reported history with serologic evidence of previous herpes simplex virus type 1(I-iSV-1) infection. This study does not completely rule out tfze possibility of a relation- ship between HSV and atheroscierosis, particularty be- cause a subgroup ot women with recurrent cold sores had twice the risk of developing coronary heart disease. The strongest epidemiologic link between infection with herpesvin~ses and atherosclerosis in humans is in the heart transplant population. Several recent studies have demonstrated a strong correlation between CMV infecdon and accelerated atherosclerosis,'0" At Stan- ford University, 300 cardiac transplant pat;ents were treazed with immunosuppressan and were f0liowtd pro- spectrvely for the occurrence of CMV infection and the derelopment of atherosclerosis. Of those tested. 91 pa- tie„ts developed CMV infections, based on: 1) positive cultures for CMV. 2) demonstration of daacteristic CMV inclusion bodies in tissue samples, or 3) a fourfold rise in garrrra 3 imru.unoglobuiin (19G) CMV antibodies. After a 5-year folbvwup, the rate of graft bss due to acceserated aiherosderosis was 69% in CMV-infected patients, but only 37% in the non-CMV-infected group.10 Furthermore there was a 1t}fold greater incidence of patients who died with mare than 50% luminal obstruction of their cor- onary arteries. These individuals had CMV infections./0 Similar results wers reported from a sorne..what smaller study at the University of Minnesota." Thts study corn- pared rates of CMV infectiort and atherosclerosis in 102 immunosupprG;sed paisents who had received a car- diac transplant and survived for at least 1 year, At 2 years after transplant, 32% of the CMV-positive patients had corLxmiry artery disease, as opposed to 10°,6 of the CMV- negative patients. To date, these findings provide the strongest evidence in humans linking herpesvirus infeo- tion with atherosclerosis. Pathologic Evidence Linking Herpesviruses with Atherosclerosis After the publication in 1973 of a landmark study by Dr. Eari R Benditt on the possible rnonocknal nature of atf'~- erosclerotic Iesions.' his laboratory was one of the first groups to docvment the presence of herpesvirus nudeic acids in arterfal tissue removed during coronary bypass surgery. They used in sitU hybridizaibrt techniques." Approxirnaiel'y 10% of 160 tissue samples were positive in wit lea iiy ~ turt var I I I ticle da I Ing ~ sho I . DNI d em g. Cn diaiE c0uL Sug~ ataa senc mew Perh trof c thick F forrtx cal e VlCfin acid L-4 I I I I I I

Yra! Pathog.nesit of Ath.nosclerosis 1197 47 ORnels 1992, vcl 1,39, Na 6 I I I I r I a M I I I 1 F for herpeslirus. A second series of experiments exam- ined tissues that contained abnormally thickened intima Ffty percent of the samples reacted positively with an HSV•2 prabe. Postive staining occurred in cells bcated in discrete fod, characterized by increased cxllularity within or adjacent to the intima Melnick and his cot- lea9ues"•'s•'6 e)dended these studies, focusing primar- ily on CMV, which they were able to detect in cells cul- tured from arterial tissue derived from patiertis with ad- vanced (grade III) aiheroarteriosclerosis. In 1984, Gyorkey et al" reported the detection of vi- rians of the herpesvirus family in e(ec=m microscopic sections from aortae of 10 of 60 patients with atheroscie- rosis undergang cardiovascular surgery. These viral par- ticles in various stages of replication included empty nu- cleocapsids and virions with dense cores. They were present in smooth muscle and erxiothdial cells. Of 1360 grids that were examined, 35 show.d evidence of viral presence. Petrie et al,'s using in situ hybrid'¢aaion tech- niques, dernonstraied increased detection of CMV in cu4- tuned smooth muscle cells from arterial plaques by stain- ing for viral antigens, Most recentty, Hendridcs et al" showed that detection of CMV nuGeic acids by the poly- merase chain reaetien (PCR) was possible in 90% of samples obtained from patients with severe atheroscle- rosis as compared with only 50% in patients with minimal or no atherosclerosis. The presence of the complete viral genome was shown in these samples by botti dot blot DNA hybridization and PCR, using probes and primers derived from the immediate early and late genomic re- gions of CMV. Messenger RNA trarscribed from imme- diate-eariy CMV genes but not the late genomic regions could be demonstrated by in sltrj DNA hybridzation. This suggested that CMV exists in the vessel wall primarity in alatent state, wh.re ezpression of immediate-eary mes- senger RNA transcripts may occur withnn expression of messenger RNA coding for sttuctural capsid proteins." Perhaps this may lead to an alteration in factors that cor- trol growth in the vessel wall, predisposing to infimal thickening. Rnally, Yamashiroya and his colleagues' have per- formed both DNA hybrid'ization and immunohistochemk cal experiments on coronary vessels of young trauma victims to identify evidence of viral antigens or nucleic acid sequences at earlier stages of ttw art.riopattry, Ev- idence for HSV or CMV was detected in 8 of 20 speci- mens from coronary arteries. The viral DNA or ant;pens were found 'n celis of the irrtact luminal surface as well as in Jocal clusters of spindle-shaped or foam cells in the irtmal layer. Taken tngeti'w, this study and those rnen- tioned above support the hypodiesis that thase herpes- viruses may conm'bute to early events in ath.rogenesis, because the herp.svirus can be found in the vasculature at various stages of tr,e arteriopatrry. C3eariy herpesvin~ses can iMeci arterial calis. In faet. HSV types 1 and 2 can infect both bovine and human vascular endothelial cells" •`2 and vascular smooth mus- de cells.a Similarly hurnan CMV also can infect such cells.33" Infection of cells in vrbo with HSV resuhs in an inhibition of host cellular protein synthesis. This is de- picted in Figure 1. This 'shutoff' of host cell protein syn- thesis occurs in two stops. The first stage, occurring in the first few hours after infection. 'early shut4ff,' is medi- ated by a virion-associated protein and causes dissoG- ation of host celi mRfVAs from pofysomes. rendering the mRNA nonfunctional and enhancing mRNA degrada- tion.35m Kefalides and ZZaie" have demorutrated tttat this mechanism is relevant to the mRNAs encoding ex- tnscellular matrbc proteins in both human endothelial cells and human arterial smoottt muscle cells's infxted with HSV-1 fn vitfo. The delayed shut-off of host cell protein synthesis is thought to be mediated by a newly synthe- sized immediate eariy-viral gene product that subse- quently causes degradation of cellular messenger RNAs.9 Some of the viral genes responsible for the in- hibition of host cell protein synthesis have been mapped and are discussed elsewhere.'O 8iochemistry of Viral-induced Atherogenesis The concept that herpesviruses play a ma}or roie in the cause and pathogenesis of atherosderosis is supported by two observations:1) The findings of Paterson and Cot- tral, reported in 19b0.41 in wtuch res,rolymphomatosis and coronary scierosis induced by an infectious agent was linked to arteriosclerosis in chickens; and 2) intrecel- luiar and extracellular cholesterol accumulation was ob- served in cell cunures infected with feline herpesvfrus.s The work from this laboratory and those of our collab- ora¢ors have shown that atherosclerosis in chickens in- duced by an avian herpesvirus infection can provide con.siderable experimental evidence to support the car aept that herpesvfn1 ses are irnrolved in the pathogenesis of atherosclerosis." Pathogen-free normocholoster- olemic chickens were infected with 100 plaque-fomvnp units of an avian herpesvirus that causas Marek's dis- ease (MDV). Uninfecied pathogen-free chickens senved as a confrot. Mareks disease virus (MDV) is a herpesvi- rus that causes malignant lymphomas of T-cell origin in this animal moda!z Primarily nerve lesions and visceral tumors develop. An antigericafiy related herpesvirus to MDV, herpesvirus of turkeys (I-M'), has been used suo cessfully by the poultry industry as an effective vaccine against the neaplastic effects of MDV.`z Nonpathogenic for turkeys and chsckens, tttis naturaly oxurring herpes-

1198 Ha}lar VP Dea.+rtbe* 1991, Vo(. 1'J9, Na 6 I. Competition II. Dissociation nwinnlr tlt. Degradation I I Nost mRNA Namit trartslatian I I Djnooiatton of host a!l mRNA Irom poysomes Fgues 1. 1'flpadsttlCal seprnu of rfwpzbokgfc dotv+ga rlrs>ftg uadfj~trtdt~xf asbE•esdooSG~ 7bir srix+na desai6es dner srsps dw can l,ead m deviardproraie ry>v/xds in a'r manmaliart tell onat ClAafies'tgat uirb 6erpes v&uc PSm c'antpeaanas ocrras ubtra tha tsal n[RNit rut "eonyxte"aids ds6msndiNw in rfxearai nsRNApcnl in riie e.lldrav~ the rfmo~laeion.p~acrts yLt'srnesalhumr ojt~r'al btfcu1on wal mRNAS cros 6r dx p.edomistrW RNA e c4sQ{os t&a vr isantlrIIrd in rbe rnfrasd c.l; podua,g cswl P.ords+t !n addLrlon, dmwhaon of dw aas cetl mR1vAfm m mrpoly.+ernes em+ oms vt dx inf.ewd crllr. 3venupr>;,Q dse ne.mat aa,urarion of dw RN:, t+mcsaiprI fi,eat .br 3•3' dbw.Ytos on obupolyaome. WC rs dx ncaiarfnF codorr in eurQyvic celis jar vaulanon}. FbuUty, dagradarion of dw R1VA owuozpb b,v a va[o.i-+•rlaand o. nswlv syredis-ed uirally rnaMtd f=to•fs) (ddpiacd by -bw o+rsce>'tt moon fig:s) cat areoo 0169 baa a!{ prsdig,a" m dev.asuxi RNA rrisitstaao.: and b- poms'n rvndAset 7ba.O.r4ff ofP~owk OwuhesLc leads to aknurio.u bt a.rpta/Lelar maois and d.-4W+dprodr..arlo+t ef dcte eLMnW dw asr b¢oked ns ltptd aatabolrc acrativ. T!r cysopoaho/osio cl/eca indu& cyrtp4,wnlc lPid aaonulanort and aherariMec !n a'x eza+rwiluGb mrcri= Rep-trtmd rei& pembs,io.t" virus eliciLs an imrrtune response in chickens that prtt teCts aqainst subsequent tumor devetopment by MDY.`Z A mtsst interesting property of MDV is that it can cause atherosclerosis reproducibly in specific-pattlogen-free chickels"; and that imrnunizatlon of chickens with F VT can prevent MDV-indutxd atherosejerosis.a-' In tact, the distributionai, morphologic, and biochemical nature of the athernsclerotic lesions induced by MDV infection ciosely resembled human atherosclerosis' Coronary ar- tEries, the aorta, and its branches were affected, resulting in fatty or fatty-prvliferetivve-type lesions.' We showed that aortic tissue from the tlorrnocholes- terolernic, MDV-irtfected group had a significantiy higher content of free and esterffed chdesterd, triacyl9lycerots, and phosphdipids than did urUnfected controls.' Feed- inq ttle infected animal a high-cholesterol dfet produced a syrrergistic effect on Fipd accretion. Marek's disem virus infection also caused an abnormal rise in cho- 4oesteryl ester (CE) synthesis with a concomitant reduc- tion in CE hydrofysis in vi,ro, thus causing arterial lipid accumulatton.° Infecaon of cells with herpesviruses in vibv also produced profound effects on cellular eholes- tzrroi membolism2° 22 which paralleled our in vivo data Irtfection of avlan srrmooth musde cehs with MOV in vibe greatly incteased the accumuladon of choiesterol and CE20 This specific type of lipid accumulation, wf'tich atso oceurs durinq the human arteriopashy, was due to de- creased lysosomai and cytoplasmic CE hydrtalytie acov- ities. Detailed anatysis of enzyme actWaDOn showed that ttte CE cyc(e is aftered. nssufting in cytopiasmic CE ao- cumulariort"-~ (Figure 2). To defne some of the regulatory m®chanisms asso- ciated with the contrat of the CE cycle, including cyto- plasmic CE hydrolase, in herpesvirus-infected cells, the level of CE hydrolase activation in MDV-infected cells was examined in the presence of: 1) dibutyryl cyclic adenosine monophosphate (AMP), 2) dibutyryl cyclic AMP added together with protein kinase, or 3) agonists of adenyiate cyclase.1 Activation of cytoplasmic CE hydro- lase activity by cyclic AMP or protein kinase A was blocked in MDV-+nfeaed cells but not in uninfected celis or in cells infected with a control virus, turkey herpe.wirus (WT) 2i Furttnrnlore the rate of cholesterof efflux from arterial srmooth muscle cells challenged with dibutyryl cy- clic AMP was unchanged in MDV-infetrted cells as cort} pared with uninfected or HVT-infected ceils, in which ef- flux was actuaiy incteased.202' Hence it has been pro- posed that the reduced cytoplasmic CE hydroiase (NCEH) activity in lipid-laden, ht:rpesvirus-+nfected celis was due in part to the inability of the enzyme to be acti- vated by the cyrlic AMP-protein Idnase A mechanism.2' This may contribute to te patt+dogic changes seen in MOV-infected arteriai cells, such as intracellular CE ao- cunxtiatiotl. We e)dended these studies to human arterial snlooth muscle cells infected with herpes simplex virus (HSV) type 1: HSV also induced accumulation of saturated tri- acybycerois and CE in infected cells.22 The infected cvUs had roductions in the CE hydrolytic activities and in the enzyme itsetf, based on irtlmunopneciptrat+on data (Figure 3) and ac6vity assays because of decreased translaiion of the RNA that encodes the intracellular hy- W v c+i ~ t~.t ~ -,, W cn I I I I U c_ I I

Viral Patho9enesis of Atheroaclerosis 1199 yP Deeadr 1991, Vod 139. No. 6 I I I I I I I I Altered metabolic events Fi9ur. 2 Hnf+s ennapoxeaCan bva cdk 4adf~rg ro atre.ed mRaboL'c acraAty. T81r sclxrna depfca a byporberic aequenue by u6icb cvtulating berpc orracr nrars ceflr 6.vugli dx cav¢l wull m+d /wch to a(avsd merabollc ac Mry, 7ba isarYS cvadasrsrg ubw pnuams anekes b.v cauoV endodul;al uU vy~m, wee ebesr aelkecm be deJecadwirb be c:nsi Once dx s,t.ions me shed io cbe arrrsal ,vMcoat mrcV.L- ca14 bUo~btndtr{q ul1-0G• dsotertvyl es,cs (CF) rms mzs, .e.fib;a In deveatd CE bydrolase ~~ (ACEf1 and NYFFI). 7bis =m lead tb ino.aed CE a=w• vbn in OM e.d1 6s me forns of CEs.icb.ar 6jo1d rbvpkas uft* an a rbaacariuic fva. ~ o f~~ la~e~o cCe~ouc/am+t cdL Reg~•>y+ud droiases.?=' Because arachidonlc acid metabolitee such as prostacyclin (PGl2) and 124-IETE atter CE hydrolysis in arterial srrlooth musde cdts'. we rneesured these ei- casaraids in HSV-infected edls. We obsxrAd a reduc- tion in bolh spcnianeous t~basvline) and arachidanate- induced reiease of the cycbnxypenase producX PGI2 and a Iipnxypenase product, 12-HEfE': Hence we be- lieve that the herpesvirus-induced atherosckrosis re- sults, in parr, from attefations in metaboiic control of CE tmifficidng by eica~ in vasvJar cells. This concept is depicted in Figure 4. Role of Gytokines in Cholesterol Trafficking in HSV-infected Cells Recent studies have suggested that the immune system may participate in aMerosclerosis. because monocyres and T lymphocytes accumulate within the attlernsclerodc lesion.=Yral infectlon could induce mediator release from 'vrQrune and hernatopoietic cells, which may be respon- sibe for phenorypic changes in vascular cells. Of interest in this regard are recertt studies that show tttiat herpesri- ruses are capable of inducing messenger RNA expree- I

1200 NzJjar 4tP D.rmnbir 1991, voL 139, Na 6 I I I CTRL CTRL ' CELLS CELLS (IgG) CELLS Ad-2 ~ HSY A 6 C sion of marocyte-derived genes. For exarrtple, CMV in- fecacn in vibc induces macrophages to increase the ax- prestion of specific RNAs for intnrieLkin-1 0 (IL-1 P), tumor necrasis factor (1T4F), and colorry stimulatinq factor•43 The rde of TNF in the complex cytddne network in marn- maJian cells is not fuy understood, particularty regarding its andvirai ac&vity. Although TNF is an irxlucer of inter- feron messenger RNA transcripts, a well-documented cytoiane that has antiherpesviral activity, others have shown that TiVF's antiviral aetiviy is rxot abolished in the presence of anriserum to iraerferon," Wong and Go.d- deto have postulated that the antiviral activity of TlVF is attributab4e to lysis of viralty infected cells rather than to inductiort of interferon. Other poscibie mechanisnns to ex- plain the action of TNF irrvoh+e eitxosanaid metabdism, since TNF's y6c effects can be partialy abdistxd with cyck>exY9enase inhibaora Sudi studiea suggest that 0 Flqura 3. I e.f acid dbo- kst*yl ata' bydolase.4CFlifrorn bnprs rmc* ~a3eca ~ XSY-infectenf (mai = 0.7) SMC cYhr.ro. Bio synrbetic lat.JVrg of crll F.ot.ist uaas done uRm f35~ (40 uCi PW 5 Y 10•' Mllrl. Farh lane mnv.ifkd aiaDtu I mg of of! Praarn or a brnnogenan cwtslsrtng of 2 x 10e cells Gnnn.nobo6%t.e v.+re auhjeaed to SDyPAGE and autorad/oRfapbY uwa pw• fari/1Cd laae A, a cofool, m+mwtoColatt uas ohtained uub 8bc tcse of aormal rnosar 1RG. 7bis lovic siotta an arbarrm of .adtooc- tiurry at M, 3S•000 (arrour) (inolecu/ar reigi4a #aoeft at:e, x 10 "d), ubitb apprar- n.iarac ax molecutrr rscVbt of.tCES! Due to t6e Imge qumuiry ofprotein lonGed on dxs 8atx aestenW of thr soviceval p.ot.ft ,tueb a am (M, 43.0WA ndu1M msd purmcdlate JUa>rvma (M, 5?,lXJO-65,000.~ ond nryonn (1f. 200,000) pnripUared nonspecifttalty. lat.e a. bnmunoprec~ouatssf matera! +iro- taWfrwrs unrqf#=PA corraro! S+IGr wi.tg muGACEff nwrsoclolfal MmLaiy. (mSC) .t 6tsfd [s abott" by dx a•sotu carreparult»g to x, 55,000. lanu C, immursaprecipitared ACFFIfrvm adcna sur inf.cned cef(s fDM1 ui- rus aonrrol eroup.l larr D, rmmwiop•rcyo+- aand m-ial case~OrYlb~ to M, 55,000 fitr+c MV Infiiaed SMGs u crnuidnablp ltss imt tfrat fmtnd in Lmts B, l be lc.ls ef acsi+t antd wryatbi a8aort in latt R(and (ane C) K.+r b{q!a tlms ri5s trnotvsa atoun in laste 4 We aLso obsashd a prominsnt psotein hmid yop.~iar9 at b!, 150.000 in lanr D alty tr 6~6 mav Iz ojtx~P, w~m tspn. FI- W11y, we assayed AtM1 aah.try eo.ff ia.cnun nuprtUPttatlon m axocbe espeimant ta cDnfbrx that u.v ti+amsotoprecrpiixed ACEX• We fotaid tbai u'f could 6Mmfocpncfptrast about 60% of dm .1Lgf in dx all bttud on ow aaad,N,bw+d in ¢hr stpw,narasu mdpr/- 4Y Reprtnled unrrh pa•nutvon,2•1 cytdvnes, such as TNF, can promote synthesis of ei- cosanoid products that, in turn, can mediate specific bi- ofogic etfacts such as modulation of cholesterd merab- oiism, a concept originaly proposed by our laboratory.'b 8eaause HSV infectirn of vascular cells can produce a biochemical and cytopattabgic effect virwaly indistin- guishabie from atherosGerosis, we hypothesized that tfsese cytokines can prevent CE accretion in arterial srr*rth muscle cells that is associated wirh herpesvirus- induced atheroscierrsis `d Turnor neorosis factor and IL- 1, but not interferon, prevented CE accumularion in H5v- infected cells by induction of cyGic AMP-dependent CE hydrolysis. Tlvs effect was medisted through the arachi- donate 12-lipoxrygenase pattrway by 12-HETE, because prrtrP.Btrn€nt of aNis with a cocktail of lipoxygenase in- hibitors abotist,ed Me an6virai effect and 12-HETE pro- duction in the celi." 12-FiM is the rnajor lipoxygenase f- RVr dx 1 oni, met ove that turn hyd tirn kine virus Qaiiy tT1et caLL lipic ceU; I I I I I I I I Pot ~ anc AGr atte tt,ro assi I itiea I tein broz dicz acsc and dt2t I byr thro ever

Viral Pathogenesis of Ath.roscl.rosis 1201 NP Desonesr 1991. VoL 139. Na 6 I I 1.1 I I I I I I I Fgurs 4. Kvpa sara iRj.aion blaeiespb'r+d rrnrssduerion paebuavs in roeprioW onoot8 mu9cie cells 7hts modal sfiou4 6at bepcs tr• no• [n(OWon ieads to deae+zedlvsaeomad by. dro()k's f yF a. Fry ffie rtCfl•f ensVnu In a"- rion, lxfeaion leadr ro deoeasnd easn.iesion oj asartidonfc add X20.4,) to PGI, mnd 11- F= tttiio6 in rtan a9t /tad ro a reductd =knrron of adncylatr C•vclmr. Dwa.rzzrdc.v- a•!ir AMP prtxlutlion nuala in ds ull. Cws- drruing br ebis alicrvid mem6olie a.grstrwe of esmrs devxned ctietic &tfP nuf,.as lpss ac- mnarrae ofpvrebt lrtnrsv 14 atius na,ducs.ig rbe acruntiovt (rfsot.Rb pbopbo.Ylxxon) ol &e cyropfatmic CZ Lry+dnolan (Nffll) to Oe aicriee fo.m Jrom dx osnase fvm. 7laar meoatw/ic nv+wr Ot obr cyaoplasm, muplaL to dsaw rn dx+ hyaai-uK .aula bt r+uraeel/s<lm- CL• ae. ceria.i. CF101 ~C€ 4ACEN FA c2p:a ~ ~0,/ GH06+PA I P%.12•FETE +7a~ , AGT j M~s~ ~ I+Ca qe1a. AMP na~ PA~ metabdite found in arterial smoottz muscle celis-'° This overall conclusion is further supported by data that show tetTNFand 1L-1 enhance 12-E-IETE productlon. which in tum increases both intracellularcyclic AMP levels and CE hydroiysis." Coilectivdy Uiese findings identified for the first Ome a biochemical mechanism inwhved In a cyto- tdr~induced reduction of lipid accumulation in herpes- virus-infected ar[erial smooth muscle cells. This is pot&~ tially an important finding regarding control of chotesterol metabolism during virai-induced atherosclerosis, be- cause cytoldnes are importartt regulators of intreceliular lipid metabolism, not only in arterial smooth muscle ceHs,"g but also in monocyte-derived macrophages." Potential Links Between Atherosclerosis and Thrombosis A link between the coagufation system, fibrinotysis, and atherosclerosis has been proposed on the basis of thrombotc processes atthe level of the endothefum. The association between lesion progression, lipid abnom,al• ides invdving elevations of lipoproteins, including lipopro- tein (a) [Lp(a)], and the fibrin depasition is now being broadiy defired. Ear1y studes of Greenland Eakimos in- dicated that their diet enriched in pofyunsaturated fatty acids of fish oil origin, had an impact on p(ateiet function and eicosanoid production. It has been hypothesized that'rf the vessel wall is injured (or activated), for example, by herpesviruses, the endottielium can become pro- thrombotic, as evidenced by several cytopathologic events. These include enhanced thrombin generation and enhanced binding of platelets to endotheiium," de- lJCEt•tWML creased PG12 prvduction,`° enhanced tissue factor pro- duction, reduced thrombornodulin expression" and an inhibition of endothelial cell synthesis of heparan suifate proteogiycan," which is closely related structurally to heparin, a complex pofysaccharide that has anticoagu- lant activity. An example of molecular mimicry that relates throrn- bfltic processes to atherogenic events invotv@s the re- cent observadon of stnldng structurai homology benveen the apoprotein (a) component of human Lp(a) and plas- minogen,SO a major protein involved in fibrinoiysis.5' In- creased levels of Lp(a) have been associated with atti- enosclerasis in humans."L^-" Lipoprotein (a) as well as its apo(a) also have been shown to compete with plasrnino- gen for cellular binding sites,5t and Lp(a) has been as- sociated with enhanced functional, antigenic, and trarr script levels of plasminogen activator inhibitor (PAI) type I by the endothdium.S' Piasminogen activator inhibitor type 1 is the rapidiy acting physiologic inhibitor of both tissue-type and urokinaselike plasminogen act'rvators. The net effect of these events leads to down-fegulation of plasmin generation at endothelial cell surface.' Lipopro- tein (a) appears to function under some circumstances as an inhibitor of plasminogen activators, by competing vrith plasminogen for binding to streptokinase or by act- ing as a competitive inhibitor of tissue piasminogen acti- vator (t-PA) in the presence of fibnnogen. (KA Hajjar, per• sonal conmunication). Recendy we have observed that HSV infection may interfere with normal endothelial ceu flbrinoiysiss, as is the case with Lp(a), by inhibition of p41s- minogen binding and increasing PAl-1 activity, thereby preventing plasminogen activarion (unpublished obser- va>ions). The mechanisms of action of HSV infection and

1202 Hajjar 4fPLtsvnls 7997, vot 239, No, 6 virat protein production on processes related to fibnndy sis are currentty under investigation. Viral Acfivafi'on of the Coagulation Cascade An important event in the eary stages of atherosclerosis is the adhesion of bbod cells to altered endothetium. Cer- tainly both platelet (as well as neutrophil) and monocyte adhesion to the endothelium are processes attributable to the pathogenesis of thrombosis and atherosclerosis. Monocytes can migrate into ttie vessel media and begin accumuiating cholesterd, contributing to foam ceil for- mat;on? Herpes simplex virus-infected endothelium, as well as endothelium exposed to TNF orto IL•1, binds ptatelets,`6 granulocytas," and expresses tissue factor to a 9reater extent than it does to noninfected cells.`° Herpes simplex vinus-infected endothelium also expresses giyeoprvteins encoded by the HSV genome.s' These partiapate in mo- iecular mimicry because of their presumed furxaional role on the erxiothelium. For example, gtyzoprotein E (gE) can function as an Fc receptnr,55'5' and glycoprotein C (QC) can atso serve as a complement (C3b) receptor.s " A role for these proteins in the pathogenesis of endottie- ual injury is suggested by the observation that poryrr>or- phonuclear (PMN) celf adhesion to HSV-infected endot- he5um can be blocked by antiviral serum S8 Central to the pathogenesis of vascular injury is the localized activation of the coagulation cascade and the adhesion of circulating inflammatory cells to the exposed subendotheliaf vascular surface. Once adt*rent, tttiese cells can secrete growth factors, proteolytic enzymes, and cytoklnes that further activate the injured vessel sur• face.2-' For example, macrophage-derived cytokines such as TNF and IL-t induce endothelial expression of leukocyte adhesion molecules5° as well as tissue fac- tor.60 These events promote further leukocyte accumula- tion and locaJized tttrombin generation. Recently we demonstrated thar infection of endothe- lial cells prtxnotes enhanced monocyte adhssicn.c' En- hanced adhesion was blocked by rnonockmat arrtibod• fes to the viraliy encoded ceil surface giycoprotein g(C) but not by antibodies to g(D) or g(E). Adhesion nlsc was blocked by treating endothelial cells with specific throm- bin inhibitors or by growing ceus in prottirombin-depleted serum. This suggested that thrombin plays a role in the enhanced monocyte adhesion. Glycoprotein C bound and promoted activation of factor X on infected endothe- fal cells, thereby contributing to thrombin generation. To further support the hypothesis that g(C) was indeed irr volved in rnonocyte adhesion and factor X binding, we used cells (L cells) that do not normally express this gy coprotein but could be transfected with the gene for herpes-virus gtycoprotein C.6' We found ttiai factor X bound to transfected L cells thatwere induced to express g(C) by dezamethasone, a steroid horrnone that acti- vated the MMTV-LTR promoter region of an artificiat g(C) gene construct. Cross-linking and irnmunopreeipitation studies dernonstrated factor X-g(C) complex formation on the cell surface, suggesting that g(C}dependent thrombin generation by herpes-infected endothelium may be an important mediator of vascular pathology dur- ing viral infection. Thrombin itsstf can elicit a variety of cellular events. These include monocyte and neutrophil aGhesion, and cytokine refease and platelet activation on a damaged surface."-'61 Each one of these cytopatho• logic effects has been iinked to ft role of inflarnmaiion as it may present itself during the pathogenesis of athero- scterosis. Identificaf+on of a Monocyte Receptor on Herpesviral-infected Endob~elium Endottretial cells express several leukocyte receptors. in- duding GMP-140 (also known as PADGEM or CD 6Z), ELAM-1,ICAM-1 and 2, and VCAM-1 on their surfaces in response to cyrokine or exposure to other agonists.02 NorrnaltyGMP 140 is acytoptasmic protein found in rest- ing endothelial ceits found on the membrane of Weibel• Palade bodies ° After stimulation by thrombin, histamine or complement proteins. the Welbel-Pafade body is rap- idly franslocated, and its membrane becomes incorpo- rated into the plasma membrane, resulting in surface ez- pression of GMP 140.°0 This mechanism of new protein expression, ie, translocation from a preformed intracellu- lar membrane compartment to the cell surface, does not require de novo protein syrrttsP_sis. As described in the previous section, our recent data support tiie following model: HSV infection induces endothetial cell surface ex- pression of HSV g(C), which acts as a binding site for factor X. Concomitant generation of tissue factor converts bound factor X to an active prothrombinase, leading to generation of thrombin in tf* microenvironmerrt of the infection 6i Recenty we have exaended this hypothesis, because we now have evidence that thrombin can act in an autocrine manner to induce expression of the leuko- cyte receptor GMP1a0.a6' Our data indicate tftat mono- cyte adhesion induced by HSV infect;on is blocked by anti-GMP140. but not by anti-ELAM or antibodies to other adhesion molecules. This suggests that GMP-140 is a major receptor for monocytes on the HSV-infected endo- tltielium." These findings are summarized in Figure 5, which shows that HSV=utfected endotheiiat cells generate thrombin, which predisposes to increased rnonoc.yte ad- herence by ezpressing the monocyte receptor GMP-1a0. Thrombin generation by these cells is dependent on the 91 eXprf X bin As st, in ft (LFA ICAN cetis,' viralty spec Simii~ to stir syste bodiE I I I I I I I I I 1 TF HSV-; throrr weil i can n ~ platel tivatec may c injury, of ree ucts.` infecu endot Mc sites ii af fact, in9 to lappir gions ~ I I

Virat Pathogenesis of Ath.rosclerosis 1203 .VP Dasnlxr 1.9,91. VoL )-0, JJO, 6 I I I I I I I I I I I I I ...::::::.... HSV-infected EC SMC Fpun S. VbW aR reuwn of mr magulatlon ea=dr, HypodNffea! WttittW depiarng bau beprcs snPkr ivus fNSIV) ny4ztion G= lead m a p.or6rombon4 psarfi~rlwoeic ~rr o.r d*e,fdodwGar.a&faor by tnduang ihr ryr,mLes mid cpsaion o18lyKqo•ouv, (C) uticb ran ac+zr ar a 6indbig arrs)o• Faacr X a key pmvnrm•r of du encRularqn aataaxtr. 7bir tnoatfod birlQin,q ojFactor X and fa ruhseqirnt co>wasron jo dx aalxelvnS Facor \a, uar rtad ro dx aonarao3 ofpo&rowbvs to &onbrn rT,oao. l!a). Wben Facxor r1a it g-e.aed, d* = b+dtiaenwface eT.ezio.r of a nio+aocyrc mcep(or fGMP•14U ut~ frt non omfsomorr motacsYr ardbenorl ro dx andorb=m and +,uiu=~ af an isyrammeao.r atVo.iae by aaimrtns dw qyaoi- rirnevt Eace of d),,u cy,opzcorogrc fea- rm, psdipose ro a6ero- seL+osa ared dsombwiC expression of HSV g(C), which can act as a site for factor X binding and asscmblyoftfie protttrombinase complex as stated earlier, other receptors do not play a major role in this celt system. For ezample, the leukocyte integrirs (LFA-1. Mac-1, and p15o,95) mediate adhesion to ICAM-1 and iCAM-2 on cytokirtcstirnulated endotheliel cells,65 but do not appear to play a rore in adhesion to the virally infected celts, based on lack of inhibition seen with specific monocional antibodies or RGDS peptides.s` Simitarty VCAM-1, which mediates fympftocyte adhssion to stimulated endothelial cells,e6 is not functionai in this system, based on ladc of inhibbort with anti-VCAM anti- bodies." The significance of a procoagulant phenotype of HSV-infected cells is considerable: IocBi generation of thrombin at the site of infectdon may actt`rate platelets as weil as endothefial cells. Thcarnbin-stimulated platolets can adhere to monocytes, and ttterefore can recruit more platelets into ttv site of injury ?,6`-6' The presence of ao- tivated platelets and monocytes at the tocvs of infecrion ruy contribute further to the deYeiopmertt of vascular ;njury, chronic inflammation, and atherosckrosis by way of release of cytokines and lipoxygenase prod- ucts.","•6` Expression of adhesion motecules on HSV- infected cells thus may be an initiai step in v'rca}-rrxdiabsd endotheliaW injury and atherogen¢siz. Most recentiy, to identify the putative celMnteractlng sites in factor X, we have analyzed the stnucturat domains of factor X that are involved in the coordination of its bind- irsg to membrane receptors.6t A group of partially over- lapping synthetic peptides repre,sentatn+e of different re- gions of the factorX molemb Nwre used, Two unrelated surface membrane receptor regions that ceordinate this neaoqnition have been identified: CD11 bCD18 on mono- cytes and plycoprotein C(gC) on HSV-infected endothe- I'axn, ie, they recognize a comrnon strucaual motif in the catalytic domain of factor X.°° Each of the peptides we used blocked factot Xa-mediated monocyte procaagu• lant activity and suppressed monocyte adhesion, The stn.ctue of the catalytic domatn of factor X centaining the tfuee sites that are invofved in binding to herpesvirai g(C) and CD11 b/C018 was modelcd carnputationally. Factor X associates itself with herpesvirus-infected endoUheliaJ cell and monocyte rectptors using a similar mofecular reoognition motif. We were able to show that the vascular cell binding region of factor X is organ'ized into three dis- tinct interacting sites. The peptidyl anabgs prevent the consequences of vascular generation of tt,rombin such as PMN adhesion to endotheaial colls, chernotaxis, and monocyte-mediated deposition of insoluble 6brin. Our understanding of the cO biology of this system coupled with the molecuiar modeiing offers an unprecedented paradigm of the molecular compledty of this thrombotic process that can be linked to vira}-induced atherosclero- sis; it also directiy links the actinration of caagutariort pro- tens on vascular cells to typical irtAammato"rorrtbotic reactions. This was observed by corutituting three spa- tially distant surface Ioops that define a unique three- ddmensional celt-interacting network in the ligand, as shoNm in Figure 6. In summary, we observed that speciflc synthedc pep- tides inhibit facter X binding to maiocyrte CD11 b/CD18 and to g(C) expressed on herpesvirus-irdecteld endotfi-e iial cells. These peptides blocked rnonocyte generation I

1204 Hajjar AJP Dftv"rlas 1997, LbL 139. +No. 6 Fipure 6. Sss.a+sd n~orL! of dz Faaor x eardyKic do>+~ llomwlogy modd bLi4Gng tcdinlqua uwe usad ro eo.lsnT.ca a ntoder oJr6r crualyetr do.nairr oJJaaa-X frons a oyaal snArnor of aypcmogerz Snt.asirr[ly ebns&std re,qfont uWs vl.nrifstYt by etara! !n¢~e~.ziors of dhe=.ar-res of v3Ps>y; cGynroVyprr>R ekaroaie arrd ht1lbttnn, and rbr Faua' x myunrar uw aligrrard muraralh, uIrb dx sn7t+c.rcr of dw.e pomacec Avomc coa>dtNares fo. haaFbo.rr and eo,um,rdside drairu of fcaro• X awr amgsrrd daealy frri.n aVvcinoRen for dwe m~rwallv asuezie+d rMiom Gnordmata of noaroomxard arde dalnt uutn'rt sa r.cirsalty mnsmud rrgior~ ~ Od jor mininra= oe wlcrp ~ctib o~ osona Ato"ele aaadinaorr fEr ,~src7rsall,y c~scblr (loap) *Cgto.+s a~ere ra4en Jrvm a p.io.. ~rLdxd nec~dr.[ of Fvao. Xa rT. EaiqvsETOrs Lx a1.) 7be Eninal nsodrl sat.aure aar eneV mtnvniud tn sc{qaa fvst aUuuing all srde cf+alrzs to m-lca, w&d rben re(mnnq all IIde tl~evns togedxr uab rlar baek6unr atonrs of d t foop repsoru Baekhortt ef Facsur X uucih•ric dnmatm (rib8on rLicgrm++,d CaraA-r/c rnad btwr e.an ddr Wml mofaoe rcmuerj [r.yo p~ 1(G}DMFn) (67) r_PKlsofaeo (RYfuy Larip peprfde 3(IDRStDURG) (67), U'K riofau /eft cdcp peprlde 9 2YpnARr7w) (67J, ePIC sofao. (rop) ex•ardus rlsk ro Ilkrsr dxiva v+liow cPK (rou,ir orr+ar) rhr N•rn+.,isucs oJm. u4ui!}*c rfornain fs ltrwrd jur a6oco pepr/de 1, r6e o~bmwr bi+uiistg grmt r it apprpsi+nauly twrical and (o rbr nbht af dW. o-tarl and rbr Gu+mbuc: R ~(~~Id'L~ rderec oj Faaar x a,~rr.+~tes for mbic6 no sawsovJ enfor+rrarson ic atailable) c~ dr Garrrminta of~iri~L: 2. of thrombin and prevented monocyte adhesion to HSI- infected endokhelium, It is possib(etriai selective interrup- tion of coagulation by such symthetic analogs may impact positively to reduce vascular injury associated with morrocyte adhwrence and HSV infection of ehe endottie- (ium. Molecular Genetics of Atherosclerosis The monaclonal hypothesis of Benditt and Benditt` was based on the observation that 75% of ath.rosderotic plaques from human tissues removed at surgery or at autopsy derrronstrated a monocb)al pherotype of the gtucose-6 phosphate dehydrogenase (G-6-PD) enZyme, based on etecirophoretlc rrobiUty. This technique too+c advantage of the fact that the poiyrnorphic Eortns of this enzyme are expressed in an Xainked fashion. MediaJ cells from regions adjacent to atherosc{erotSc plaques frorn nonpiaque areas display equal arnounts of each pofymorphic form of the enzyme, as would be expected in an X-iinked gene, because one of the two X chromo- somes is inacvvated early in embryonic development, Smilar resuhs have been obtained by other laborato- ries °°'70 however the interpretation of these experiments in supporting the monoclonal hypothesis has been chai- lenged." The principal argument is that the G-frPD marker may be linked to sonle other gene that gives ttuse cells a selective growth advantage, ie, a'pheno- type-sefective advantage' wittw the proliferative lesion being trufy rna*ocfonel in origin. Tfw monoclonal hypottiesis purports that the srnooth muscle cell prel'rferacion seen in the ari,ernsclerotic lesion is similar to a benign tumor, for example, a uterine leio- n inft no I I in hu' rat ' _ (r1C - the nOr . D h ~ - nar Ma 3T; , forr re6 has pla I ~ 7, t: BLUt pla aic sior enc cer trar plac labc rYpE rec( with anti ang latc Thes gen anc cise unic f Iden rece I I I I I I , tervi mlh cedl: Cora taRz forrr I I