Monocyte chemoattractant protein-1 in human
atheromatous plaques.
N A Nelken, … , D Gordon, J N Wilcox
J Clin Invest.
1991;
88(4)
:1121-1127.
https://doi.org/10.1172/JCI115411
.
Monocytes appear to be central to atherogenesis both as the progenitors of foam cells and
as a potential source of growth factors mediating intimal hyperplasia, but the chemical
messages which stimulate the influx of monocytes into human atheroma remain unknown.
Monocyte chemoattractant protein-1 (MCP-1) is a recently described molecule with powerful
monocyte chemotactic activity expressed by monocytes, vascular endothelial cells, and
smooth muscle cells in culture. To begin to address the role of MCP-1 in vivo, we examined
10 normal arteries and 14 diseased human arteries for MCP-1 expression by in situ
hybridization. MCP-1 mRNA was detected in 16% of 10,768 cells counted in human carotid
endarterectomy specimens with highest expression seen in organizing thrombi (33%) and in
macrophage rich areas bordering the necrotic lipid core (24%) as compared to the fibrous
cap (8%) and the necrotic lipid core itself (5%). Based on immunohistochemical staining of
serial sections and on cell morphology, MCP-1 mRNA appeared to be expressed by
vascular smooth muscle cells (VSMC), mesenchymal appearing intimal cells (MICs), and
macrophages. By contrast, few cells expressing MCP-1 mRNA were found in normal
arteries (less than 0.1%). These data suggest a potential role for MCP-1 in mediating
monocytic infiltration of the artery wall.
Research Article
Monocyte
Chemoattractant Protein-1
in
Human Atheromatous Plaques
Nicolas A. Nelken,*Shaun R.Coughlin,*DavidGordon,9andJosiah N.Wilcoxil
Cardiovascular Research Institute and *Division of Vascular Surgery and
*Department of
Medicine, Universityof
California, San Francisco,SanFrancisco, California 94143; University of Washington, Departmentof
Pathology, Seattle,Washington 98195;andI1GenentechCorporation, South San Francisco, California 94080
Abstract
Monocytes appear tobe central toatherogenesis bothasthe
progenitors of foam cells and as a potential source of growth factors mediating intimalhyperplasia, but thechemical mes-sages whichstimulatetheinflux ofmonocytesintohuman ath-eromaremainunknown. Monocyte chemoattractantprotein-i
(MCP-1)is arecently described moleculewithpowerful
mono-cytechemotactic activity expressed bymonocytes,vascular
en-dothelialcells,and smoothmuscle cells inculture.To
begin
toaddress the role of MCP-1 invivo, we examined 10 normal
arteriesand 14diseased human arteriesfor MCP-1 expression byinsituhybridization.MCP-1 mRNA wasdetectedin16%of
10,768cells counted inhumancarotidendarterectomy
speci-menswithhighest expressionseeninorganizing thrombi (33%)
and inmacrophage richareasborderingthenecroticlipidcore
(24%)ascompared to thefibrouscap(8%) andthenecroticlipid
coreitself (5%).Based onimmunohistochemical staining of
se-rialsectionsand oncell morphology, MCP-1mRNAappeared
tobeexpressedbyvascular smoothmuscle cells
(VSMC),
mes-enchymal appearing intimalcells (MICs),andmacrophages.
Bycontrast,fewcellsexpressing MCP-1mRNA werefoundin normalarteries(< 0.1%). Thesedata suggest apotential role forMCP-1 in mediating monocytic infiltration of the artery wall.(J. Clin. Invest. 1991.
88:1121-1127.)
Keywords: Insitu hybridization-vascularsmoothmuscle cells *vasculardisease-macrophages*cytokines
Introduction
Monocyte-derived
foam cellsare amajor
component ofather-oma,comprisingasmuch as 60% of cellsfoundinthenecrotic
lipidcore,and 10-20%ofcells in thefibrouscap(1). Thesecells havebeendemonstratednotonlyin matureplaques but also in early
fatty
streaks in animal models ofatherogenesis
and in humandisease
(2). Ithas been suggested that growth factors produced by plaque macrophages may beimportant
interms ofplaque progression (3).ThesignalswhichattractmonocytesReprint requests should be addressedtoJosiah N.Wilcox,Division of Hematology/Oncology, Emory UniversitySchool ofMedicine, 308 Woodruff MemorialBuilding,DrawerAR, Atlanta, GA30322.
Receivedforpublication8November1990and in revisedform16
April1991.
1.Abbreviations usedinthispaper:MCP-1, monocytechemoattractant
protein-i;
MIC, mesenchymal intimalcells;PCR, polymerasechainreaction; VSMC,vascularsmooth muscle cell.
into these lesionsremain unknown. Monocyte
chemoattrac-tantprotein-l
(MCP-1)I
isasecreted molecule withpotent che-motactic activity for monocytes(4). MCP-l is expressed byculturedgliomacells(4),fibroblasts(5), monocytes(6),
lym-phocytes (7), endothelial cells (8), and smooth muscle cells (9). TNFa (5, 8, 10), IL-1j (5, 8, 10),
yIFN
(8), and endotoxin (5,10) have all beenshowntoincrease MCP-l mRNAlevels in culturedendothelial cells and fibroblasts. In addition, increases in MCP-l-like activity and MCP-l mRNA have been docu-mented in supernatants of human umbilical vein endothelial cells (HUVE's) stimulated with serum enriched with mini-mallyoxidized LDL and
ft
VLDL (1 1, 12). These in vitro stud-ies suggest a potential role for MCP-l in therecruitment ofmonocytesinto the vessel wall in the setting ofinflammationor
hyperlipidemia.
Toaddress whether MCP-l may play a role in recruiting
monocytes tothe vessel wall in vivo, we set out toidentify
MCP-l mRNA production in human atheroma usingin situ
hybridization. Atherosclerotic plaques exhibited marked hy-bridization, especially in macrophage-rich inflammatory
re-gions. Bycomparison, minimal expressionwas noted in con-trol arteries without atherosclerotic disease.
To identify possible regulators ofMCP-1 production in cellsderived from human atheroma, vascular smooth muscle cells (VSMCs) grown from human carotid endarterectomy
specimenswerestimulated with variouscytokinesand MCP- 1 mRNA levels were determinedbyNorthern blot. Tumor
ne-crosis factor (TNFa),gammainterferon
(yIFN),
and superna-tants from mixedlymphocyteculture were shown to stimulateproductionofMCP-l mRNA, whereas 10%FCSdidnot. These data suggest a potential in vivo role for MCP-l in mediating monocytic infiltration of the artery wall, and a role forinduction of MCP-lexpression byinflammatorymediators in modulating this process.
Methods
MCP-1probe isolation.The cDNA for MCP-Iwas clonedfrom stimu-lated humanlung fibroblastsusingpolymerase chain reaction(PCR) amplificationof cDNA made fromcytoplasmicRNA. Human lung
fibroblastswereobtained fromexplantsobtainedatautopsy. Cultures weremaintained in DMEM supplemented with 10%FCS, 1% penicil-lin/streptomycin,and25 mM Hepes, buffer pH 7.4. To increase MCP-1 mRNA levels, cellswerestimulated withphorbolmyristicacetate
(100ng/ml for 48h),thenwith E. coliendotoxin(10pg/ml,Sigma Corp.,St.Louis, MO)in the presence of 1 zg/mlcycloheximide (Sigma
ChemicalCorp.)for 6 h(13).TotalRNA wasisolatedusing guanidium thiocyanatelysisandCsCldensity gradientcentrifugation (14).cDNA wasthen producedusing12-18meroligodTprimers (PharmaciaFine
Chemicals, Piscataway, NJ) and reverse transcription followed by
RNAseAdigestion.30cyclesofPCRwereperformed(15)using33mer
oligonucleotideprimerswhichprimedthe 5' and 3' ends of thecoding regionfor MCP-1 and encodedprotrudingXhoI(5'primer)and Eco RI(3'primer)restrictionsites.The PCRproductwassubcloned into
Monocyte ChemoattractantProtein-i in Human Atheroma 1121 J.Clin. Invest.
© The American Society for ClinicalInvestigation,Inc.
0021-9738/91/10/1121/07 $2.00
.
*::.: " \ s.:<ws )
OX SC}*: :....
P ..
to .*.:.
..E :.
i ..
A
./p
d
id__4.5
-'Na
,e:+
.,:* > . .,j3
I
-4-A..
.1
.W.'. :., I ... 4.,A: .-%%
:v
.:W:l
-40.
%
i-J%
IL'.
..X
... :..
fr.
'It.
the Xho I and EcoRIsites of Bluescript SK-plasmid and its sequence determined by dideoxy nucleotide sequencing (16). Primer sequences were 5' CCC CTC GAG TCC AGC ATG AAA GTCTCT GCC CTT for the 5' sequence, and 5' CCC GAA TTC TGT TCA AGT CTTCGG
AGT TTGGGT for the 3' complementary sequence.
In situhybridization. Human carotid endarterectomy specimens wereobtained directly from the operating room, fixed for 2 h at4VCin
4% paraformaldehyde immediately upon surgical excision, cryopre-served in 15% sucrosefor3 h andfrozen in OCT at -70'C. To con-struct riboprobes, Bluescript KS+ plasmid containing a full-length MCP-Iinsert was made linear by restriction digestion with Eco RI or Xho I(forsenseandantisense probes, respectively), then transcribed using T3 or T7 RNA polymerase (Stratagene Corp., La Jolla, CA) in the presenceof"S-UTP.Insituhybridizationwasperformed on 10Itm
frozen histologic sections as previouslydescribed (17, 18). After coating withphotographic emulsion, sections were exposed for 4-12 wk. Ex-posedsections were counterstained with hematoxylin and eosin.
The proportion of cells expressing MCP-1 was determined by counting 10,768 individual cells identified by nuclear staining at 400x in the14atherosclerotic specimens hybridized to theMCP-lriboprobe (Fig. 1, eandf).Acell was considered positive if the number of silver grains associated with stained nuclei exceeded five times background
overequivalentcell-sizedareaswithoutnuclei.Formostsectionswe
screened, cells withseven or moregrains were considered positive.
Immunohistochemistry.Serial microscopic frozen sections immedi-atelyadjacent to sections hybridized with theMCP-l probe were used forimmunohistochemical identification ofcells in three atheroscle-rotic specimens. Primary monoclonal antibodies included HAM56 (Dako Corp.,Carpinteria, CA) which reacts with - 95%ofthe human monocyte/macrophage cellpopulation ( 19). Cross-reactivity ofthis an-tibody has been shown to capillary endothelium, hence, all slides stained with HAM56 were compared with sequential slides stained with Ulex europaeus lectin (Vector Laboratories, Inc., Burlingame, CA), which selectively stainsendothelium (20). Vascular smooth
mus-cle cellswere identified with HHF35 (Enzo Pharmaceuticals, New York, NY)tomuscle actin whichspecificallyreactswithcardiac, skele-tal, and smooth muscle cells, but not fibroblasts, endothelium, neural
tissue,orepithelium (21).Tlymphocyteswereidentifiedwith UCHL1 (Dako Corp., Santa Barbara, CA) which reacts with CD45RO
ex-pressedbyprestimulated andasubset of restingTcells (22). This
anti-bodydoesnotreactwithBcellsorNKcells,althoughthereissome
cross-reactivitywith cellsofmyeloidand monocytic lineage(PMNs wererarein thesespecimens). Appropriate biotin-conjugated second-ary antibodieswereused anddetected with avidin-alkaline
phospha-tase(ABC-AP kit;VectorLaboratories,Inc.)accordingtothe
manufac-turer's instructionsas reportedpreviously (17, 18). Counterstaining was donewith 0.4% methyl green (Sigma Corp.).
Cell culture. VSMCs were culturedin collagen matrix (Vitrogen; Collagen Corp., Palo Alto, CA) from human atheroma obtained at carotid endarterectomy using published techniques (23). Medium was changed twice weekly, and cells passed before reaching confluence. All cellswereusedwithin the first six passages.
Smooth muscle cell phenotype wasconfirmed by immunohisto-chemical staining using HHF35 andfluoresceinsecondary antibody. Cultured cells stained positivefor actin in a spiculated pattern typical for smooth muscle cells in 83±5% (95% confidence interval) of 195 cellscounted (not shown).
Northern blotting. VSMCs grown to 90% confluence in 150 cm2 flasksweremadequiescent by incubatingfor 48 h with serum-free mediumafter washing 3x with PBS. Thereafter, theywereeither stimu-latedfor4or 24 h with the cytokines or growthfactors used in each
specific experiment.Total cellular RNAwasisolatedusingthe guanid-iumisothiocyanatemethod (14). 10 ugof glyoxal-treated RNA was thenrunina1%agarosegel and transferedtonylon reinforced nitro-cellulose (Schleicher & Schuell Corp., Keene, NH)asdescribed (14).
32P-labeled MCP-l probewasproducedusing the random primer tech-nique andT7DNA polymerase (StratageneRandom Primer kit; Strat-ageneCorp., La Jolla, CA). Nitrocellulose membraneswere prehybri-dized for4h andhybridizedat420overnightat aconcentrationof 106 cpm/ml in buffercontaining50%formamide. Two washesusing2X SSC with 0.1% SDSat550 andtwohigh-stringencywashesusing0.1 x SSC with 0.1% SDSat550wereperformedfor 15 min each. Nitrocellu-lose was then exposed to Kodak X-OMAT film with anintensifying
screen at -70° for variableperiodsof timedependingonestimated
radioactivity of the nitrocellulose membranes.
Results
Insitu hybridization. 14 carotid endarterectomy specimens, eightnormal coronaryarteries, and two normal internal mam-maryarterieswereexamined byinsitu hybridization.Positive hybridization for MCP-l was clearly present in all atheroma examined(Fig. 1, a, c,e).No cellularhybridizationwasseen in control slides probed with MCP-l sense riboprobe (data not
shown).
Expression of MCP-I in histologically
defined
regions of human atheroma. Positive hybridization was detected in 16% of10,768 cells counted. Random high-power fieldsamongrep-resentativehistologic regionswerescoredusingacounting grid
Figure 1. (a)Aportion of atherosclerotic plaqueshowingextensivehybridizationof MCP-I antisense probeat8 wk exposure.Areaof inflam-matory cellinfiltrate inorganizingthrombushybridizesmoststrongly in this section. Darkfieldpolarized lightepiluminescenceviewenhances silvergrains correspondingto areasofhybridization (8-wk exposure;magnification, 100). (b)Samefieldas ainbrightfieldstained with
hema-toxylinand eosinshowing fibrouscaparea(right) adjacenttoorganizingthrombus inintraplaque hemorrhage(left).Infiltration ofinflammatory
cellscanbeseenbetween thesetwo areas(magnification, 100). (c) Darkfield polarized-light epiluminescenceviewof carotid atherosclerotic plaqueshowinghybridizationfor MCP-Iin theinflammatorycell-richareaunderlyingthenecroticlipidcoreabuttingthearterial media(above) (8-wkexposure; magnification, 100).(d) Adjacent frozen sectionto cstainedwithantimacrophagemonoclonalantibody (HAM56) showing
correlation ofimmunohistochemicalstaining for macrophageswithareasof intensehybridizationseeninc. Positivestainingappears red
com-paredwithmethylgreencounterstain(magnification,100). (e)High-power darkfieldpolarized light epiluminescenceview ofanatherosclerotic plaqueshowing MCP-1 hybridizationinaVSMC in thefibrous capregion.Discrete vascular smooth muscle cellswereshowntohybridize
strongly andwereoften surroundedby morphologicallysimilar cells which didnothybridize (8-wkexposure,magnification, 500).
(f)
Same fieldas einlightfieldview forcomparison, showinganareawithahighpercentage of smooth muscle cells(positiveHHF35stainingonserialsection notshown). Note typical elongated nucleus and smooth musclemorphologyinhybridizedcellseenine.(g)Section of normal internalmammary
artery obtained from coronarybypasssurgical specimens hybridizedwith MCP-Iantisenseprobe. Verylowlevels ofhybridizationwere seen
in normal vessels.Examinationof this section under high power revealed only four cellshybridizingto
MCP-l
probe (arrows). The otherbrightareasin theadventitiarepresentrefractile fibrous tissue,anormal feature of the adventitia when viewedbyepiluminescence(8-wkexposure;
magnification, 100). (h)Section of control coronary arteryobtainedfromcardiomyopathy patient duringhearttransplantation(native heart)
hybridizedwith
MCP-l
antisense probe. Hybridizationwas rareinthesespecimens.One smoothmuscle-appearingcellcanbeseenin themedia withhigherthanbackgroundhybridization(arrow).Concentric fibrous intimalthickeningwas auniversalfindingin adult coronaryspecimens,however, itwasnotassociated withhybridizationof probe (8-wk exposure;magnification, 125).
14
vW
'p
.%... A
i
4'-rS
/W
-*~~~~~~~~~~~~
.:I A/
r
JOY
t: ~ ~ ~ ~ ~F:.~
Ak'
-1
i
4ko /
4.B;
.0
4s
*:0F : l;.
,wt
A.
:%
044
:
*Si; !. .:t
*t ?a@
.:.,
41,
*:<:
.. i::
*. iris
S
?
4w
-_
b.
:t
A..
:
`.
, .1
W. :,.:
..:. "
i.:.: Ak.
.41 z,
\w
..I
A .,*.
'IN:.
".,"
--:
TableL ComparisonofPercentage Nuclei Associated wtInS
Hybridizationto MCP-I mRNA
Positivecells/totalcells
Morphologicsubdivision
Fibrous cap (uniform portion) 76/1471 (5%)
Fibrous cap(pleomorphic portion) 288/2871 (10%) Inflammatory area underlying necrotic core 482/1999 (24%)
Necrotic core 28/475 (6%)
Organizingthrombus 342/1035 (33%) Immunohistochemical subdivisic
HAM56positive cells 439/1163 (38%)
HHF35positivecells 82/ 1754 (5%)
Total 1737/10768 (16%)
Acell wasconsidered positive if the number of silver grains associated with stained nuclei exceeded five times background over equivalent cell-sized areas. For mostsections we screened, cells with seven or moregrains were considered positive. Both morphologic and immu-nohistochemical criteria were used for subdivision of tissues. For the morphologic subdivisions, uniform and pleomorphic portions of the fibrous cap were seen to have different morphologic features and to hybridize to different degrees (see text for details), and were separated onthosegrounds. For immunohistochemical subdivisions, HAM56 identified cells of monocyte lineage (primarily macrophages), and HHF35identified VSMCs. Results show a preponderance of hybrid-ization to macrophages, inflammatory areas underlying the necrotic
lipidcore, andorganizingthrombus.Vascular smooth muscle, how-ever,wasalso seentohybridizebuttoalesserdegree.
and thepercentageof hybridizedcellstallied (TableI).For the purposesof this analysiswecategorizedareasofthe 14
athero-sclerotic plaques into five histologically identifiable regions:(a)
fibrous cap containing uniform elongated "pancake" nuclei arranged inaprimarily unidirectional orientation, (b) fibrous
cap containing pleomorphic nuclei without ordered
orienta-tion,
(c)inflammatory
areasborderingthenecroticlipidcore,(d)
necrotic lipid core, characterized by cellular debris, and cholesterolclefts,
and (e)organizing
intraplaque thrombus characterized by the presenceofmorphologically identifiable fibrindeposits
orhemosiderin containing
macrophages.Fibrouscap. Inthefibrouscap,8%of4,342 cells hybridized
tothe MCP-l riboprobe (Fig. 1, eandf). Significantly more
MCP-
1-positive
cells were found in areas of the fibrous capregions predominantly containing cells with pleomorphic
nu-clei arranged inarandom orientation(10% of 2,871 cells)
com-pared with fibrouscapregions
containing predominantly
uni-formspindle
shaped cellstypical of VSMCs (5% of1,471cells,P<0.001, chi square). The in situ signal intensitywas generally
lowerovercellsinthefibrouscapcomparedwith the intense
signals
seenin thesamesectionsmostnotably in inflammatoryareasbordering the necrotic
lipid
core(seebelow).Regions of the fibrous cap with uniform spindle-shaped
cellsgenerally showed uniform staining with the HHF35 anti-body. Fibrous cap regions with pleomorphic nuclei often exhib-ited fewer VSMCs staining with HHF35 and frequently
con-tained HAM56and afewUCHLl-positive cellsaswell. MCP-1-positive cells in the fibrous cap included a population of
mesenchymalappearing intimal cells (MIC) often in pleomor-phic zones which did not co-stain with any of the antibodies used in thisstudy (Fig. 2). These cells had large pale nuclei
distinct from the typical spindle-shaped smooth muscle cell nucleus, a stellate cytoplasm, and were morphologically similar to theplatelet-derived growth factor (17) ortissue factor(18)
mRNA-containing MICs described previously in
atheroscle-rotic plaques. Manyareas within the atherosclerotic plaques
examined containednumerouscells meeting thesecriteria espe-cially in the pleomorphic regions ofthe fibrous cap and areas of
organizing thrombi.Itisdifficulttoestimate theproportion of positive MICsbecausetheir detection is bya processof
exclu-sion rather thandirect identification.
Organizing thrombus. Many cells
hybridized
strongly to theMCP-l probe inareasoffocal organizing thrombi(33%
of1,035 cells; Fig. 1, a andb). These regions often contained
many HAM56-positive cells with some UCHLl-positive T
cells. Spatial correlationof HAM56-positive cells and cells
hy-bridizing for MCP-l strongly suggeststhat macrophages ex-press MCP-1. We cannot exclude thepossibilitythat some
ex-pression in theseareasisby MICs becausethese cellsareoften found inareasof focalorganizing thrombus (24, 25).
Necrotic
lipid
core. Discrete cells in the necroticlipidcore were markedlypositive forMCP-1 mRNA (6% of475cells) andwereoftenadjacenttononhybridizingcells. The necrotic coresexhibited diminished cellulardensity
andcontainedpre-dominantly necrotic debris, foam cell deposits, cholesterol
clefts,
and calcification. In agreement with previous studies macrophageswerethe predominantcell typefound in thene-croticcorealthough UCHLl-positive cellswerealsoidentified in theseregions (1).
Inflammatory
regionsbordering the
necroticlipid
core. Of all theregions
studied the mostintensely hybridizing
MCP-l-positive
cellswerefound ininflammatory regions
bor-dering
the necrotic core. Positive cellswererelatively
abun-dant,
representing
24%of 1,999
cellscounted. Theseinflamma-tory
regions
containedmanyHAM56-positive
cellswithafewscattered
UCHLl-positive
cells.Histologically
identified foam cells wereoften observedtohybridize
totheMCP-l
probe in theseareas.Expression
of
MCP-Jby
cell type.Toidentify specificcell typesassociated with MCP-lproduction,
serialmicroscopic
sections
immediately adjacent
tosectionshybridized
withtheMCP-1 probe were used for
immunohistochemical
staining
with HAM56, UCHL1, HHF35, and Ulexlectin inthree
repre-sentative
specimens. Carefully
chosenregions
which stainedexclusively
with HAM56werefoundtobestrongly
positive
forFigure 2.(a) Brightfield epiluminescentviewof cellular area withahighproportionofimmunohistochemicallyunstainedmesenchymal
ap-pearingintimal cells(MICs)incarotidatheroscleroticplaque (arrows). Nucleiarevisibleasthedarkhematoxylin-stainedbodies surroundedby
brightsilvergrains. Hematoxylinand eosinsection;magnification,200.(b)Adjacentsectionto astainedwithHAM56,showinglackof tissue
macrophagesin this section(typical cross-reactivitytoendotheliumby thisantibodyis alsoillustrated). (c)Serial section stained with HHF35 foractin-containingVSMCs. Somepositivecellswerenoted, but donotappeartoberelatedtothosehybridizingforMCP-1inthissection.
(d)
Serial section stained withUCHL1 showingabsenceof T cells in this section. Positivestainingfor each monoclonalantibodyappearsred,
MCP- 1 (38%of1,754 cells). Cells meeting histologic criteria for macrophages, including large pale nucei with foamy
cyto-plasm, clearly hybridizedwithMCP-1 riboprobe, whereas
he-mosiderin containingmacrophages did not. Fewer MCP-l
posi-tive cells were found in regions staining exclusively with HHF35 (5% of1,163 cells). These areas showed no HAM56
staining. Despite
the low percentageof
cells hybridizing in theseareas, individual cells clearly meeting histologiccriteria for VSMCs,includingspindle-shapednuclei and elongatedcy-toplasm, hybridized for MCP-l (Fig. 1, eandf).Itisnot possi-bleto comment on the colocalization ofTcell staining with MCP-l hybridization because no region was found to stain
exclusively with UCHL1 in the absence ofHAM56 staining.
These data suggest that both VSMCs and macrophages can express MCP-l in vivo. In addition, cells not staining with HAM56orHHF35, meetingthehistological criteria forMICs were also seen to hybridize strongly for MCP-1.
We failed to detect
MCP-l
hybridization to endothelial cellslining thevasavasorum or onthe lumenal surface oftheplaques. Given the factthatthetotalamountofendothelium seenin thesespecimenswassmall duetotransversesectioning, lack
of
endothelium inportions
ofthe plaque and insome cases loss of endothelium in preparation, conclusions regardingMCP-l
expressed by the endothelium in vivocannotbe drawn fromthisstudy.Normal, nondiseased internalmammary arteryspecimens obtained fromcoronarybypass
operations
exhibited littlepro-duction
ofMCP-l mRNA asdeterminedby in situhybridiza-tion.
Hybridizing
cells, ifpresent, tendedto be foundin the adventitia(Fig. 1g).Occasional
macrophage stainingwas seenin theadventitia suggestingthathybridizingcells may be
scat-tered tissue macrophages. Like the internalmammaryarteries, normalcoronary vessels
obtained
fromexplanted hearts also showedonly minimal MCP-Iexpression (Fig.
1h).
Thesespeci-mens wereobtained from
nonatherosclerotic
coronaryarteries inpatients
withidiopathic cardiomyopathies undergoing
hearttransplantation.
Asmallamountof concentric intimal thick-ening whichwasinvariably
presentin thesespecimens
butnoadvanced atherosclerotic lesionswereseen.
Hybridization
wasfound
in <0.1% of cells countedinnormal vessels.Cytokine stimulation of
VSMCsinculture.
Thefinding
that VSMCs expressedMCP-l
invivosuggestedapossible
rolefor smooth muscle-produced MCP-l ininitiating
orenhancing
monocytic
infiltrationof
thevessel wall. Tobegin
toexamine factorspotentially responsible for regulation
ofMCP-1
expres-sion by VSMCs,
Northern blotanalysis
wasperformed
oncyto-kine-treated human VSMCs in culture. Northern blot
analysis
of10
gg
of total cellularRNA persample
revealedaverylow levelof detectableMCP-l
RNAinresting
cellsaswellasthose stimulated with10%FCS.Supernatants
fromamixedlympho-cyteculture(donated byBruce
Hall, Stanford University,
Palo Alto, CA) however, produced a marked increase in MCP-1 mRNAdetectedat 4 and 24 h(data
notshown). Inflammatory
regulatorsTNFa(200 U/ml) and yIFN (50 U/ml)werethen
investigated
alone and in thepresence of5%FCS. Bothcyto-kines stimulated marked increases inthelevelof MCP-1 RNA
(Fig.
3). TNFaelicited
greaterincreasesat24 hthan at 4h,
whereasyIFNgenerated maximalresponse at 4 hwith diminu-tionovertime. The
differing
temposof MCP-1induction afterTNFa and yIFN
stimulation
presumably
reflect different mechanismsof
action. Therewasnoappreciable difference
inOC)-\~~~~~~~~~~~~~~~.
-\I
288-188
Figure 3. Northern blot of whole cell
lysate
RNA (10ytg
per lane) hybridized with 32p labeledMCP-l1
probe. Lanes include duplicate sample control cells incubated in serum-free medium for 48 h;-yIFN,
50 U/ml for 4 and 24 h;yIFN,
50 U/ml with 10% FCS for 4 and 24 h;TNFa, 200 U/ml for 4 and 24 h; TNFa, 200 U/ml with 10% FCS for 4 and 24 h. Very low level baseline production is seen in control lanescompared with peaks at 4 h foreyIFN
and 24 h for TNFa. FCS had no effect on production of MCP-1 mRNA either in cytokinetreated
cells (above) or otherwise unstimulated cells (not shown).theresponses to cytokines seen in the presence or absence of fetal calf serum.
Discussion
Human monocyte chemoattractant
protein-l1
(MCP- 1) is a se-creted peptide which exhibits powerful monocyte chemoat-tractant properties at nanomolar concentration. It has been detected in smooth muscle cells, endothelium, stimulated monocytes, lymphocytes, fibroblasts, and various tumor cell lines (6). Because the signals which attract monocytes to the atherosclerotic plaque remain unknown, we examined arteries forMCP-l
production in vivo. We have shown that there is easily detectableMCP-l
expression by multiple cell types in various regions of the atherosclerotic plaque, but little expres-sion in normal vessels.The majority of MCP- 1 production was seen in macro-phages. MCP- 1 was noted in all regions ofthe plaque with high concentrations of macrophages, especially underlying necrotic lipid cores and in organizing thrombi. The production of MCP-1 by macrophages may represent an amplification mechanism forthe recruitment ofadditional macrophages to the atheroscle-rotic plaque.
bothwith andwithout HAM56 staining.Becausesignalswhich
stimulate MCP-l production in VSMCs are not known and because MCP-l production by VSMCs may beinvolvedin the
initial recruitment ofmacrophages to the vessel wall, we
ex-posed cultured atheroma-derived VSMCs TNFa and
yIFN,
thendetermined MCP- I mRNAlevelsby Northern blot. The increase inMCP-1 expressioninduced by
yIFN
and superna-tantsof mixed lymphocyte culturesuggest apossible rolefor activated T lymphocytes in stimulation ofvascularsmoothmuscle MCP-l production.
Theproduction ofMCP- 1 by MICs adds to the increasing evidencesuggestingthat thesecellssecrete anumberof bioac-tive factors (17,18). Theorigin ofMICsand their rolein
ather-ogenesis
remain unknown.Theidentification ofVSMCs and MICs as sourcesof MCP-1in atherosclerotic plaquesuggests thatthismoleculemay po-tentiallyplay a roleinmonocyteinfiltration inthe early devel-opment of the plaque. Mediators stimulatingmonocyte infil-tration into early fatty streaksareunknown, but theselesions have beenidentified aspossibleprecursors to mature
athero-sclerotic plaques (3). It istemptingto speculate thatMCP-l production by VSMCsorMICsmaybeelicitedbylipoproteins (11)orotherstimuli in early atherosclerotic lesions. With infil-tration of monocytes, a positive feedback mechanism might exist whereby
infiltrating
monocyte/macrophages secrete MCP- 1 in largeamounts, thereby attractingmoreperipheral bloodmonocytes tothedeveloping atherosclerotic plaque.An-tagonists
of MCP-1action will be neededtoevaluateany causalrole forMCP- I in vivo.
In summary, we have shown that
MCP-l
mRNA is ex-pressedin atherosclerotic plaques in vivo.To ourknowledge,this is thefirst studytodemonstrate theproduction ofaspecific
monocytechemotacticagentin human atheroma.
Acknowledgments
We thank R.Stoney, W.Ehrenfeld,J.Goldstone, and J. Rappof the Divisionof Vascular Surgery,UniversityofCalifornia,SanFrancisco,
D. E. Strandnessof the Division of VascularSurgery, Universityof
Washington, forprovidinguswith thetissues used in thisstudy,and Dr. LoisEpsteinof theUniversityofCalifornia, SanFrancisco,fora
generousgiftof gamma interferon.
N. A.Nelkenwassupportedbyagrant from the Pacific Vascular
Foundation and National Institutes of Health National Research Ser-vice Award grant F32HL-08241.D. Gordonwassupportedby
ROl
HL-42119.This workwassupportedinpart by
ROl HL-43322.
References
1.Jonasson,L., J. Holm,0.Skalli,G.Bondjers,andG. K. Hannson. 1986.
Regional accumulations ofTcells,macrophages,and smoothmuscle cells in the
humanatherosclerotic plaque. Arteriosclerosis.6:131-138.
2.Gerrity,R.1981. The roleofthemonocyte inatherogenesis.I.Transitionof
blood-borne monocytes into foam cells infattylesions.Am.J. Pathol. 103:181-190.
3. Ross, R. 1986. Thepathogenesis ofatherosclerosis. N. Engl.J. Med. 314:488-500.
4.Robinson,E. A.,T.Yoshimura,E. J.Leonard,S.Tanaka,P. R.Griffin,J.
Shabanowitz,D.F.Hunt,and E.Appella.1989.Completeaminoacid sequence
of a human monocyte chemoattractant, a putative mediator of cellular immune reactions. Proc.Natl. Acad. Sci. USA. 86:1850-1854.
5.Strieter, R. M., R. Wiggins, S. H. Phan, B. L. Wharram, H. J. Showell, D. G.
Remick,S. W.Chensue,and S. L. Kunkel. 1989.Monocyte chemotacticprotein
geneexpressionby cytokine-treated human fibroblasts and endothelial cells. Bio-chem. Biophys. Res. Commun. 162:694-700.
6.Yoshimura, T., N. Yuhki, S. K. Moore, E. Appella, M. I. Lerman, and E. J. Leonard. 1989. Human monocyte chemoattractant protein-l (MCP-1). Full-length cDNAcloning, expression in mitogen-stimulated blood mononuclear leu-kocytes, and sequencesimilarity to mouse competence gene JE. FEBS(Fed. Eur. Biochem. Soc.) Lett. 244:487-493.
7. Kaczmarek,B., B. Calabretta, and R. Baserga. 1985. Expression of cell-cy-cle-dependent genes in phytohemagglytinin-stimulated human lymphocytes. Proc. Nail. Acad. Sci. USA. 82:5375-5379.
8.Rollins, B. J., T. Yoshimura, E. J. Leonard, and J. S. Pober. 1990. Cyto-kine-activated human endothelial cells synthesize and secrete a monocyte che-moattractant,MCP-1/JE. Am. J. Pathol. 136:1229-1233.
9. Valente, A. J., D. T. Graves, C. E. Vialle-Valentin, R. Delgado, and C. J. Schwartz. 1988.Purification of a monocyte chemotactic factor secreted by non-human primate vascular cells in culture. Biochemistry. 27:4162-4168.
10.Sica, A., J. M. Wang, F. Colotta, E. DeJana, A. Mantovani, J. J. Oppen-heim, C. G. Larsen, C. 0. C. Zachariae, and K. Matsushima. 1990. Monocyte chemotactic and activating factor gene expression induced in endothelial cells by IL-I and tumor necrosis factor. J. Immunol. 144:3034-3038.
11.Berliner, J. A., M. C. Territo, A. Sevanian, S. Ramin, J. A. Kim, B. Bamshad, M. Esterson, and A. M. Fogelman. 1990.Minimally modified low density lipoprotein stimulates monocyte endothelialinteractions.J.Clin.Invest. 85:1260-1266.
12.Cushing, S. D., J. A. Berliner, A. J. Valente, M. C. Territo, M. Navab, F. Parhami, R.Gerrity, C. J. Schwartz, and A. M. Fogelman. 1990. Minimally modified low density lipoprotein induces monocytechemotactic protein1 in humanendothelial cells and smooth muscle cells. Proc.Nail. Acad.Sci. USA. 87:5134-5138.
13. Furutani,Y., H. Nomura, M. Notake, Y. Oyamada, T. Fukui, M. Ya-mada, C. G. Larsen, J. J. Oppenheim, and K. Matsushima. 1989. Cloning and sequencingof the cDNA for human monocyte chemotactic and activating factor (MCAF). Biochem.Biophys.Res.Commun. 159:249-255.
14.Sambrook,J., E. F. Fritsch, and T. Maniatis. 1981. MolecularCloning.A
Laboratory Manual. ColdSpring Harbor Laboratory Press, Cold Spring Harbor, NY. 7.19-7.22.
15.Saiki, R. K., D. H. Gelfand, S. Stoffel, S. J. Scharf, R. Higuchi, G. T. Horn, K. B.Mullis, and H. A.Erlich. 1988. Primer-directed enzymatic amplification of DNAwith a thermostable DNA polymerase. Science (Wash. DC).
233:1076-1078.
16.Sanger, F., S.Nicklen, andA.R.Coulson. 1977. DNAsequencingwith
chain-terminatinginhibitors. Proc.Natl. Acad.Sci. USA. 74:5463-5469.
17.Wilcox,J. N., K. M.Smith, L. T. Williams,S. M. Schwartz,andD.
Gordon. 1988. Platelet-derived growth factor mRNA detection in human athero-scleroticplaques byinsituhybridization.J.Clin.Invest.82:1134-1143.
18.Wilcox, J. N.,K.M.Smith, S. M. Schwartz, andD.Gordon. 1989. Local-ization of tissue factor in the normal vessel wall and in theatherosclerotic plaque.
Proc.Natl. Acad.Sci. USA.86:2839-2843.
19. Gown,A.W.,T.Tsukada, and R. Ross. 1986. Humanatherosclerosis.II. Immunocytochemicalanalysis of the cellularcomposition ofhuman atheroscle-roticlesions. Am. J. Pathol. 125:191-207.
20.Holthofer, H., 1.Virtanen,A.L.Karimiemi,M.Hormia,E.Linder,and A.Miettinen. 1982.UlexeuropaeusIlectinas amarkerfor vascular endothelium in humantissues.Lab.Invest. 47:60-66.
21. Tsukada, T., D.Tippens,D. Gordon,and R. Ross. 1986.HHF35,a
muscle-actin-specificmonoclonalantibody.Am.J.Pathol. 126:51-60.
22. Terry, L. A., M. H. Brown, and P. C. L.Beverley. 1988. The monoclonal
antibodyUCHLIrecognizesa180 kD component ofthe humanleucocyte
com-mon antigen, CD45.Immunology.64:331-339.
23.Libby, P., S. J. Warner, R. N.Salomon,and L.K.Birinyi.1988. Produc-tion ofplatelet-derived growthfactor-likemitogen bysmooth-muscle cells from human atheroma. N.Engl.J. Med.318:493-498.
24.Wilcox, J.N.1991.Analysisoflocal geneexpressionin human atheroscle-rotic plaques by in situhybridization. Trends Cardiovasc. Med.1:17-24.
25.Wilcox, J. N., S. M.Schwartz,and D. Gordon.1990. Localproductionof platelet-derived growth factor in the human atheroscleroticplaque.InGrowth Factors, DifferentiationFactors, andCytokines.A.Habenicht,editor.
Springer-Verlag, Berlin.305-323.