Intact human ceruloplasmin oxidatively
modifies low density lipoprotein.
E Ehrenwald, … , G M Chisolm, P L Fox
J Clin Invest. 1994;93(4):1493-1501. https://doi.org/10.1172/JCI117127.
Ceruloplasmin is a plasma protein that carries most of the copper found in the blood.
Although its elevation after inflammation and trauma has led to its classification as an acute phase protein, its physiological role is uncertain. A frequently reported activity of
ceruloplasmin is its ability to suppress oxidation of lipids. In light of the intense recent interest in the oxidation of plasma LDL, we investigated the effects of ceruloplasmin on the oxidation of this lipoprotein. In contrast to our expectations, highly purified, undegraded human ceruloplasmin enhanced rather than suppressed copper ion-mediated oxidation of LDL. Ceruloplasmin increased the oxidative modification of LDL as measured by
thiobarbituric acid-reacting substances by at least 25-fold in 20 h, and increased electrophoretic mobility, conjugated dienes, and total lipid peroxides. In contrast, ceruloplasmin that was degraded to a complex containing 115- and 19-kD fragments inhibited cupric ion oxidation of LDL, as did commercial preparations, which were also degraded. However, the antioxidant capability of degraded ceruloplasmin in this system was similar to that of other proteins, including albumin. The copper in ceruloplasmin responsible for oxidant activity was not removed by ultrafiltration, indicating a tight association. Treatment of ceruloplasmin with Chelex-100 removed one of seven copper atoms per molecule and completely blocked oxidant activity. Restoration of the copper to ceruloplasmin also restored oxidant activity. These data indicate […]
Research Article
Find the latest version:
Intact
Human
Ceruloplasmin Oxidatively Modifies
Low
Density Lipoprotein
Eduardo Ehrenwald, Guy M. Chisolm, and PaulL.Fox
DepartmentofCell Biology, Cleveland Clinic Research Institute, Cleveland, Ohio 44195
Abstract
Ceruloplasminisaplasmaproteinthatcarriesmostof the cop-perfound in the blood. Although its elevation after inflamma-tion andtraumahas ledtoits classificationas an acutephase protein, its physiological role is uncertain. A frequently re-portedactivity ofceruloplasmin is itsabilitytosuppress oxida-tion oflipids.Inlight oftheintenserecentinterestinthe oxida-tion of plasma LDL,weinvestigatedtheeffectsof
ceruloplas-min on the oxidation of this lipoprotein. In contrast to our expectations, highly purified, undegraded human ceruloplas-minenhancedratherthansuppressedcopperion-mediated oxi-dation ofLDL.Ceruloplasmin increased the oxidative modifica-tionofLDL asmeasured by thiobarbituricacid-reacting sub-stances by at least 25-fold in 20 h, and increased electrophoreticmobility,conjugateddienes,and totallipid per-oxides.In contrast,ceruloplasminthatwasdegradedto a
com-plexcontaining 115-and19-kD fragments inhibitedcupricion
oxidationof LDL,asdid commercialpreparations,whichwere
alsodegraded.However,theantioxidantcapability ofdegraded
ceruloplasmininthissystem wassimilartothatof other pro-teins,includingalbumin.The copperinceruloplasmin
responsi-bleforoxidantactivitywasnotremovedby
ultrafiltration,
indi-cating a tight association. Treatment of ceruloplasmin with Chelex-100 removedoneofsevencopperatoms permolecule and completely blocked oxidant activity. Restoration of the copper toceruloplasminalso restoredoxidant activity.These dataindicatethatceruloplasmin,dependingontheintegrityof itsstructure andits boundcopper, can exert a potentoxidant rather thanantioxidant actionon LDL.Our results invite specu-lationthatceruloplasminmaybein partresponsiblefor oxida-tionofLDLinbloodorinthearterial wallandmay thus haveaphysiological role that isquitedistinct from whatiscommonly
believed. (J. Clin. Invest. 1994. 93:1493-1501.) Keywords:
ceruloplasmin* copper *oxidizedLDL*oxidants*
metallopro-teinases
Introduction
Ceruloplasmin is an abundant plasma protein that carries
- 95%ofthetotalcirculatingcopper.Itisa monomerof 132
kDcomposedalmostentirely of three 42-45-kD domains that arehighly homologoustoeachother(1)andtothe three
do-Aportion ofthis work has appeared in abstractform(1993.
Circula-tion. 88:180.[Abstr.]).
Address correspondenceto Dr. Paul L.Fox, Department of Cell
Biology,Cleveland Clinic ResearchInstitute,NC10,9500 Euclid
Ave-nue,Cleveland, OH 44195.
Receivedfor publication 30August 1993 and in revisedform17 November 1993.
mains that makeupthe large, active proteolytic subunits of factors V andVIIIof thecoagulationcascade (2). Ceruloplas-minisan acute phaseproteinwitha responseof intermediate magnitude compared with other acute phase proteins; the plasmaconcentration isincreasedtwo- tothreefold during in-flammation, pregnancy, andafter traumatic injury, including surgery(seereferences 3 and4for review).
The physiological function of ceruloplasmin has been a subject of considerableinvestigation, speculation, and contra-diction. Multiple biochemical activities of ceruloplasmin have been described, including copper transport, oxidation of various amines, oxidation ofFe2+toFe3+ for subsequent
up-take bytransferrin andferritin,and antioxidantactivityagainst
lipidperoxidation(seereferences4and 5 forevidenceforand
againstthese functions). Reported cellularactionsof cerulo-plasminincludestimulationof cell proliferation (6) and
angio-genesis (7). Ceruloplasmin's antioxidant property has been
consistently observed by many laboratories (8-13), and is thought bysome to beresponsible fora largeportion of the antioxidant capacity ofserum(8). The antioxidant action has been proposed as acrucial function ofceruloplasmin during inflammatoryandacutephaseresponses( 14).Multiple mecha-nisms have been proposed to explain ceruloplasmin
antioxi-dant activity, including scavenging ofsuperoxide and other
reactiveoxygenspecies (10), andinhibitingthe Fenton reac-tion by conversion of Fe2+toFe3+(ceruloplasmin is also called
"ferroxidase")( 12, 15). The latter mechanism is supported by aconsiderablebody ofevidence,buttheability of ceruloplas-mintoblock Cu2 -mediatedlipid oxidationsuggeststhat alter-nateantioxidant mechanismsmustalso pertain(16). Thereis evidence thatceruloplasminas anantioxidant blocks protein ( 17) and DNA damage (18), and that itaffords protection
against free radical-initiatedcellinjuryandlysis ( 19).
Ceruloplasminhas been showntoinhibittheoxidationof
tissue homogenates(8), tissueextracts of lipids(18),
micro-somalmembranes(1 1,12, 20), and vesicles ofpurified polyun-saturatedfatty acids(9) andphospholipids(10), butitsrole in theoxidation ofplasmalipoproteinshasreceived little atten-tion.This activityofphysiologicalantioxidants has becomea
subject ofintense interest since there is convincing evidence thatlipoproteinsare modified by oxidation in vivo. Lipopro-teinparticles resembling oxidizedforms havebeenobserved in normal human plasma (21, 22), inplasmaofdiabetic subjects (23), and in theserumof patients with cardiovascular diseases
(24). The datademonstratingthepresence ofoxidized LDL within arterial lesionsare particularly strong(25). Evidence thatlipidoxidation has acausative role in lesion formation is largely circumstantialand based onin vitro studies.However, recentdatashow that severalantioxidantsreduce atherosclero-sisprogression in variousanimalmodels(26-29), and some are associated with reduced risk of cardiovascular disease in humans(30, 31 ).
The cellular andmetabolicpathways by whichlipids found inatherosclerotic lesionsareoxidizedare not yetknown, but arelikely to involve eitherpathological increases in oxidant
J.Clin. Invest.
©TheAmericanSocietyfor ClinicalInvestigation,Inc.
0021-9738/94/04/1493/09 $2.00
activity, possibly via cell-mediated free radical processes, or deficienciesintheantioxidant activity of host defensesystems. We therefore examined the ability of ceruloplasmin,as a puta-tive plasma antioxidant, to protectLDLagainst modification bycopper ion-mediated free radical oxidation. Our results indi-cate that ceruloplasmin can exhibit a potentoxidant, rather than antioxidant, activity that dependson specific structural properties of the molecule.
Methods
Materials. LDL and ceruloplasmin were isolated from fresh human plasma (< 48 h)obtained fromtheAmericanRed Cross. Rabbit poly-clonal antibodies againsthuman ceruloplasminand human a1-anti-trypsinwereobtained fromAccurateChemicalCo.(Westbury,NY).
Chromatographic media werepurchasedfrom Pharmacia LKB
Bio-technology AB(Uppsala, Sweden),andChelex-l00wasfromBio-Rad Laboratories (Richmond,CA). Other reagents were from Sigma
Chem-icalCo.(St. Louis,MO) unlessotherwisenoted.
Preparation ofhuman ceruloplasmin. Ceruloplasminwaspurified fromhumanplasma byamodification (3la)ofthemethodofNoyer et al.(32). Inbrief, QAE-SephadexA50 batch chromatography,
ammo-nium sulfate precipitation, hydroxyapatitechromatography, Sephadex G50chromatography, ultrafiltration,and MonoQ fastprotein liquid chromatographystepswereapplied sequentially. Hydroxyapatite chro-matographyyieldedtwopeaksof ceruloplasmin differingonly in
car-bohydratecontent(33).Themajor fraction,referredtoas
ceruloplas-min I(orfor brevity, just ceruloplasmin),hasthreeN-linked oligosac-charides, while the lesser fraction, ceruloplasmin II, has only two
oligosaccharides(4).Eachofthese fractionswasseparately subjected
tothesubsequentpurificationsteps. Theprocedurewasmonitored by
ceruloplasmin oxidase activitymeasured essentially asdescribed by Sunderman and Nomoto(34)withp-phenylenediamineassubstrate. The term"oxidase activity"is used here to refertotheability of
cerulo-plasmin to oxidize p-phenylenediamine, while "oxidant activity"
refers to the oxidationofLDLbyceruloplasmin(see below).
Cerulo-plasmin proteinwasdetermined byabsorbance(OD'm,6i0nm=0.66)
andtotalproteinwithbicinchoninicacid(Bio-Rad Laboratories) using bovineserumalbuminasstandard.Thepurificationprocedure yielded ceruloplasmin preparationsthatwereessentiallypureasdeterminedby ceruloplasmin oxidasespecific activity(34), byanabsorbance ratio (610/280 nm)>0.045(4, 35),andbyhomogeneity accordingto SDS-PAGE and silverstaining. Purifiedceruloplasminwasprimarily the intact 132-kDmonomer,but alsocontainedasmallamount(<10% bysilver stainofSDS-PAGE) ofthe 115- and 19-kDproteolytic
frag-mentsof ceruloplasminalso present inserum(36).The 115-kD
frag-menthas been observed inmostpreparationsofceruloplasmin (unless
it is furtherdegradedto 50- and 67-kDfragments)(37, 38).
Cerulo-plasmin prepared bythis procedure was completelystable for pro-longedperiods(> 1mo at370C)duetothe removal ofanendogenous metalloproteinasethatspecificallycleavedintact, 132-kD
ceruloplas-min into15-and 19-kDfragments.Thematerialwasstoredat-70°C.
Humanceruloplasminwasalso obtained fromSigmaChemical Co. and Serva Biochemicals(Westbury,NY).Ceruloplasmincopper
con-tent wasdetermined by flame atomicabsorption spectrophotometry (3030B;Perkin-ElmerCorp.,Norwalk, CT).
Electrophoresis. Samples were dissolved in Laemmli buffer (39) containing 10mMdithiothreitolforSDS-PAGE,orinLaemmli buffer without SDS for nondenaturing PAGE, and subjectedto electrophore-sis on4-12% polyacrylamide gels (Novex, Encinitas, CA). Protein bandsweredetectedby silverstaining (Integrated Separation Systems,
Natick, MA) and molecular weights derived by comparisontoMark 12 (Novex)orRainbowprotein standards(AmershamCorp., Arlington Heights, IL). Flat-bed agaroseelectrophoresiswasdoneusinga Tris-acetate buffer system in the absence of EDTA.
Lipoprotein methods. Human LDL was prepared from freshly
drawn, citratednormolipemic plasmatowhichEDTAwasadded
be-foreultracentrifugation. LDL (density = 1.019-1.063 g/ml) was iso-lated by sequential ultracentrifugation asdescribedpreviously (40). Immediately before use, LDL was extensively dialyzed at4VCagainst saline. LDL (1 mg/ml cholesterol, final concentration) was oxidized by incubation with CuS04(0.7-1.0 MM) and other testmaterials in 0.05 ml PBS for 20-24 h at37"C.Lipoprotein oxidation was measured as formation of thiobarbituric acid-reacting substances (TBARS)' by a modification (41 ) of the method of Schuh et al. (42). Duplicate
lipoprotein samples were compared with a standard curve prepared withmalondialdehyde(MDA), and TBARS, detected by fluorescence at 515 nmexcitation/553 nm emission, were expressed as nanomoles ofMDAequivalentsper milligram LDL cholesterol. In some
experi-ments lipoprotein oxidation was continuously monitored as the forma-tion of conjugated dienes by absorbance at 234 nm (43). The total lipid peroxide content of modified lipoproteins was measured asdescribed
byEl-Saadani et al. (44). Relative electrophoretic mobility of modified LDL was determined by electrophoresis on 1% agarose gels (Ciba Corning, Palo Alto, CA) as described by Morel et al. (40), except that the electrophoresis was extended to 45 min to improve separation.
Results
Oxidant activity ofceruloplasmin. To investigatethe hypothe-sisthat human ceruloplasmin protects lipoproteins against oxi-dative modification, we examined its effect on copper ion-in-duced oxidation ofhuman plasma LDL. CUSO4 alone in-creased LDL TBARS by - 15-fold to 16 nmol MDA
equivalents/mg cholesterol, astimulation consistentwith pre-vious reports (45, 46). Surprisingly, the addition ofpurified, intact humanceruloplasmin did not suppress the oxidation of LDL byCuS04, but ratheraugmentedit in adose-dependent manner to as much as 25 nmol MDA/mgcholesterol(Fig. 1). Totest if the oxidant activity of ceruloplasmin required exoge-nous copper ions, the effect of ceruloplasmin by itself was tested. Incubation of LDL with ceruloplasmin for 24 h in-creased oxidation levels by - 40-fold to 39 nmol MDA/mg
cholesterol at the normal (nonevoked) physiological plasma concentration of 300 ug/ml (Fig. 2). Potent oxidant activity was observed in all 4 preparations ofceruloplasmin (isolated from differentpoolsof plasma) that were tested, and using at least 10separate preparationsof LDL. Ceruloplasminoxidant activity wascomparableto thatofCuS04(when expressed as moles ofceruloplasminprotein and molesofCuS04),with the activity ofdifferent ceruloplasmin preparations rangingfrom about half as potent to twice as potent as the soluble salt (not shown). Ifceruloplasmincopper isinvolvedin oxidant activity (see below), then it is likely that not all copper atoms partici-pate in the reaction or that theinvolvedatoms are less effective than free copper. As a control to showwhether ceruloplasmin alteredthemeasurementofTBARS, ceruloplasminwasadded toLDLimmediatelybefore the assay and was found to have no influence on the TBARS measured.
Toverify that LDL was indeedoxidizedbyceruloplasmin, other physical andchemical parametersoflipid and lipopro-tein oxidation were measured. Since ceruloplasmin contains between six and seven atoms of copper permolecule(35, 47), we speculated that the chemical characteristics ofLDL oxi-dized byceruloplasminwould be similar to those of copper-ox-idized LDL. Concentrations ofceruloplasmin (100
gg/ml,
30 | X ! |
C
XA
0.225
0
0 20
0
(AU 15
0
0 40 80 120 160 200
Ceruloplasmin(iglml)
Figure 1.Oxidant activity of human ceruloplasmin. LDL (1 mg/ml
cholesterol,finalconcentration)wasoxidized by incubation with
CuS04 (0.7AM)inthepresenceofintact,purified human cerulo-plasmin I (a) for 20 hat370C inafinal volume of 0.05ml.Oxidation levels of treated and untreated LDL (o)weredeterminedasTBARS
andexpressedasnanomoles MDAequivalentspermilligramLDL
cholesterol.
equivalentto 0.75 MM) andCuS04(0.7
gM)
werechosen togive similar values of TBARS after24 hof incubationat370C. The initialrateandextentof formation of TBARS (Fig. 3 A), totallipid peroxides (Fig. 3 B), and conjugateddienes(Fig. 3
C)werenearly identical for both oxidants.As furtherevidence
oflipoprotein oxidation, incubation ofLDL with the same
concentrations ofceruloplasmin orCuS04 for20 h at 370C
increased its relative electrophoretic mobility by 63 and 56%, respectively (not shown). The similar responsesconfirm the oxidant nature ofceruloplasmin and suggest similar
mecha-nisms of action for ceruloplasmin and freecopperion.
The oxidant activities of the differentially glycosylated iso-formsofceruloplasminwereevaluated.CeruloplasminIand II
exhibited nearly identical oxidant potencies with respect to
modification of LDL (Fig. 4 A). Both isoformsgave
half-maxi-mal oxidationataconcentrationnear25 Mg/mland maximal
oxidation at - 200,ug/ml.Although ceruloplasmin prepared
60
8] 50
TIi:
Con
10
0
0 50 100 150 200 250
Ceruloplasmin(RImQ)
Figure2.Concentrationdependenceof LDL
plasmin.LDL(1 mg/mlcholesterol)wasoxid 24 hat370Cwithceruloplasmin I(-).Lipop
determinedasTBARS andexpressedas nanor permilligramLDLcholesterol.
C
a
.2
0
'R
0
CL 03
-I
2 50
40
i 40
20
B 10 E
0
I
2 E 3
7
6
5 4 3 2
E
i N
8
TII
0 4 8 12 16 20 24 28 32
Time(h)
Figure 3. Timecourseof oxidation of LDL byCuS04and
cerulo-plasmin. LDL (1 mg/ml cholesterol)wasoxidized in thepresenceof 100gg/mlceruloplasmin I (a), 0.7AMCUS04(o),orbuffer (xn) for
30 hat370C inafinalvolume of 0.05 ml. Aliquotsweretakenat
indicated times andlipid oxidation determined (A)asTBARS and
expressedasnanomoles MDA equivalentspermilligram LDL
cho-lesterol, (B)astotallipid peroxides (LPO) and expressedas micro-molespermilligram cholesterol,or(C)asconjugateddienes and
ex-pressedasabsorbanceat234nm(aftera1:10dilution withPBS).
inourlaboratory ishighly homogeneous, its oxidant activity
could have beenduetoaminor contaminant. To examine this
possibilitytwoapproachesweretaken. First,ceruloplasmin
ox-idantactivitvwas.testtedafter filrthermirifirntinn hv
Peletrnnhn-lJQllLa%,IIzW1LJ vvaa WkLvau%aLW~l IU1LI.61 JUinn~aLIVII USAVlskL;LVjJ11V
resisonagarosegels,followed by elution of the ceruloplasmin
band. Fig.4Bshows that after elution, both glycosylated iso-forms hadhighlypotentactivities thatwereonlyslightly
dimin-ished compared with the original material. In a second
ap-proach, ceruloplasminwasremoved byimmunoadsorption
us-ing
aspecific polyclonal anticeruloplasmin antibody.
The antibody wasshown tobe mono-specific since onlya single protein (whichcomigrated with authentic human ceruloplas-min)wasrecognized inhumanserumby Western blotanalysis(not shown). Protein A-Sepharose 4B beadswere
preincu-bated withrabbitantibodies raised against either human
ceru-loplasminora1-antitrypsin, and the antibody-bead complexes
collected after low-speed centrifugation. Purified ceruloplas-minI(40 ig)wasincubatedwith the beads for 12 hat4VC, and
300 350 400 the low-speedsupernatantswererecovered andtestedfor
oxi-dantactivity. Measurement of ceruloplasmin oxidase activity
oxidation by cerulo- in the precipitate confirmed that immunoadsorption with anti-lized by incubation for ceruloplasmin, but not anti-a1-antitrypsin, removed at least
proteinoxidationwas 90% ofthe ceruloplasmin. The anticeruloplasmin antibody
re-molesMDAequivalents moved - 85% ofthe oxidantactivitywhileonly 15%was
re-movedbyanirrelevant antibody (Fig. 5). These results
indi-B
0
-A
25 20
15
110
15
10
S0 25
1
1510
5
0
0 50 100 150 200
Ceruloplasmin(pg/ml)
Figure 4. Oxidant activity oftwoisoforms ofhumanceruloplasmin differingincarbohydratecontent.(A)LDL(Img/mlcholesterol)
wasoxidizedbyincubation with the purified, undegradedhuman
isoforms,ceruloplasminI(o) and11(e), for 20h at37°C ina final volumeof0.05ml. (B) Bothisoforms (500 ,ug)werefurther resolved
on1%agarosegels in Tris-acetatebufferin the absence ofEDTA. The
ceruloplasminbands(visiblewithoutstaining)wereexcised,cutinto smallpieces,and eluted by incubation with PBS for 10hat25°C.
Agarosepieceswereremoved bycentrifugation( 15,000gfor 10min) andceruloplasminwasdialyzed extensively againstPBS. Eluted cerulo-plasminwasincubated withLDL asdescribed inA.Lipoprotein oxi-dationwasdeterminedasTBARS andexpressedasnanomolesMDA
equivalentspermilligramLDLcholesterol.
catethat it is ceruloplasminrather thanaminorcontaminant that isresponsiblefor oxidantactivity.
Regulation ofceruloplasmin
oxidant activity. Since the oxi-dant action of ceruloplasminwas notconsistentwithprevious
reportsof antioxidant activity, wetested theeffect of cerulo-plasmin from commercial suppliersused in mostprevious
re-portsonoxidation of
lipids.
Commercial ceruloplasminhadessentiallynocapacitytooxidizeLDL in the absence of exoge-nous copperion; themaximum oxidationlevel was 1-3nmol MDA equivalents/mg cholesterol, compared with 20-40 for
intact,
purified
ceruloplasmin(Fig. 6).
Whenadded in the pres-enceofCuSO4,
commercial ceruloplasminwasfoundtobeanantioxidant, afindingconsistentwith previousreports. This result contrasted with that forintact, 132-kD
ceruloplasmin,
whichaugmentedcopperion-mediatedLDLoxidation (Fig. 1). According to SDS-PAGE,
purified
ceruloplasmin I washighly homogeneousand presentprimarilyasanintact132-kD
proteinwith onlyasmallamount(< 10%)ofthe 1 5-kDdegra-dationproduct. Thisproportion ofintactproteintothe 115-kDfragmentissimilarto that observed by Western blot analy-sisofwholeplasma (36,and datanotshown).
Electrophoresis
on anondenaturing gel confirmedthepurityof
ceruloplasmin
I(Fig.6, inset); the 115-andl9-kD degradation products
re-mained complexedunder these conditions and
cochromato-graphedwith the intact
protein.
The 132-kDproteinwasen-tirely missingfrom the commercial preparationand was
re-25
c
0-.
a
20
C-o ij 15
0 110
5
0
0 50 100 150 200 Ceruloplasmin
(ILg/mi)
Figure 5.Removal of oxidant activity of ceruloplasminby immunoad-sorption with anticeruloplasmin. Protein A-Sepharose4B was
pre-treated with 0.2% bovineserumalbumintoreducenonspecific bind-ing.After extensive washing withPBS, 0.2 ml of beads wasincubated overnightat4VCwith0.2mlof rabbit antibody againsthuman
ceru-loplasminor
a1-antitrypsin
as a control fornonspecific binding.Theantibody/proteinA-Sepharosecomplexes wereprecipitated,washed with PBS, and incubated with 40;tgceruloplasminIat4VCovernight. LDL(1mg/ml cholesterol)wasincubated for20 h at
370C
(final volumeof0.05 ml) withceruloplasminI (o), thesupernatantad-sorbed with anticeruloplasmin antibody
(in),
orthe supernatant simi-larly treated withanti-aI-antitrypsin
antibody(o).Lipoprotein oxi-dationwasdeterminedasTBARSandexpressedas nanomolesMDA-equivalentspermilligramLDLcholesterol.
placedby lower molecular massfragments(Fig. 6, inset) and,
possibly, nonceruloplasmin contaminants.Similar antioxidant
capacitieswerefoundusing ceruloplasmin from several
com-mercialsources, allofwhich were highly degraded but retained normal levels ofceruloplasmin oxidase-specific activity that were comparable to that of purified ceruloplasmin I (not
shown).
35
._
0 0
05O=aCo E. 2025
qPA
00
us.. 10 .CO
C-0
cco
0
a
o
-.?
0 20 40 60 80 100
Ceruloplasmin
(Rg/mI)
Figure 6. Effect of commercially obtained ceruloplasmin on LDL ox-idation. LDL(1mg/ml cholesterol) was oxidized by incubation for 20h at37°C with commercial ceruloplasmin in the absence (o) or presence (a) of 0.7
MM
CuSO4. Lipoprotein oxidation was deter-minedasTBARSand expressedasnanomoles MDA equivalents permilligramLDLcholesterol. The insert shows
100
ng of purifiedhu-manceruloplasminI(lane
1)
and commercial humanceruloplasmin(lane2)resolved by SDS-PAGE and detected by silver stain. The
samepreparationswerealso resolvedby nondenaturingPAGE(lanes
To begin to understand the molecular mechanisms in-volved, we examined the role of ceruloplasmin structure in
oxidantactivity. We first examined whether the difference in structuralintegrityof theceruloplasmin preparationscould ac-count for the observed difference in oxidantactivity. Proteoly-tically modified ceruloplasmin was prepared by incubating
highlypurified, intact ceruloplasminIwiththeplasma metal-loproteinase that wasisolatedduringtheceruloplasmin prepa-ration and partiallypurified(3 la).After incubationfor72 h at
370C, the 132-kD ceruloplasmin molecule was completely
converted to the 115-kDfragment (Fig. 7A, inset). Both de-graded andundegradedforms ofceruloplasminwereincubated
with LDL for 24 h and the level of oxidation measured as TBARS. Proteolytically modified ceruloplasmin lost about three-fourths oftheoxidant activity oftheintact molecule(Fig.
25
20
15
C
-c£0
L.
. O.
it E
._ a
:z
cz2 m
10
5
0
40 35 30 25
20
15 10 5 0
0 20 40 60 80 100 120
Ceruloplasmin
(gg/ml)
140 160
0 10 20 30 40
Protein
(jig/ml)
Figure 7.Effectofproteolytic modification on ceruloplasmin oxidant andantioxidant activity.(A)CeruloplasminI(200,g)was preincu-bated with(designated "proteolytically modifiedceruloplasmin") or without(designated"intactceruloplasmin"), a partially purified
plasma metalloproteinase preparation(. 10 ng)for 72 h at370Cin 200glof25 mMTris-HCl,10 mMCaC12,pH6.6. The 132-kD
ceru-loplasmin (inset,lane1)wascompletelymodifiedtothe 115-kDform (inset, lane2)asshown by silver-stained SDS-PAGE(100ng
pro-tein/lane).Thearrowindicates132-kDceruloplasmin. LDL(1mg/
mlcholesterol) was incubated for 20 h at370Cwith intact
cerulo-plasmin(o), proteolytically modified ceruloplasmin (a), or intact
ceruloplasminwith thesame amountof metalloproteinase added just before the incubationwithLDL(c).(B) LDL(1mg/ml cholesterol)
wasincubated for 20 h at370Cwith
0.8MgM
CuSO4 in the presence of commercialceruloplasmin (.),proteolytically modifiedcerulo-plasminI(m), orbovine serum albumin (o). Lipoprotein oxidation wasdetermined as TBARS and expressed as nanomoles MDA equiv-alents permilligramLDLcholesterol.
7A). Addition of theproteasejust beforetheincubation with
LDLdid not substantially reduce the oxidant activity
ofcerulo-plasmin, showing that the smallamountofprotease present did notexhibit nonspecificantioxidantactivity(see below). Struc-tural differencesbetween native and proteolyticallymodified
ceruloplasminarelikelytobe subtle since theproteins cochro-matograph onnondenaturinggels(indicatinga stablecomplex
of the fragmentsunder theseconditions), they contain equal
amountsofcopperbyatomic absorptionspectroscopy(6.6and 6.7 atoms per ceruloplasmin for native and proteolytically modified protein, respectively),andthedegradedprotein
re-tains 90% of the oxidaseactivity of intact
ceruloplasmin
(notshown). Inadditiontothe lossinoxidantactivity, proteolytic modification of ceruloplasmin induces antioxidant activity
whenincubated with LDL in the presenceofCuSO4(Fig.7B). Theantioxidant activity ofproteolyticallymodified
ceruloplas-min was similar tothatof bovine serum albumin and only
moderately less than commercialceruloplasmin. These
find-ings are consistent with those ofGutteridge ( 16), who has
shownthat, inacupric ionsystem, theantioxidant
activity
of (degraded)ceruloplasmin isnotunique,but ratherisanonspe-cific propertyofmostproteins. KalantandMcCormick (48)
havesimilarly shown thatalbumin and otherserumproteins
suppress metalion-mediatedoxidation ofLDL.
We compared theabilityof variousproteins toinhibit
ceru-loplasmin-mediated as well as CuSO4-mediated LDL oxida-tion.Allproteinsandpeptidestestedsuppressedboth CuSO4-dependent(Fig. 8A) andceruloplasmin-dependent(Fig. 8 B) oxidation in a concentration-dependent
fashion,
and with a rank orderofpotencyof lactalbuminhydrolysate>albuminc
.2
0^
11
la
-E
Z
isIC
8 <
0
0 200 400 600 800 1000
Protein(iglml)
Figure 8. Inhibitionof copper- andceruloplasmin-mediatedLDL
oxidation byproteins. (A)LDL(1 mg/ml cholesterol)wasoxidized
byincubation for 20 hat370Cwith 0.7MM CuSO4in the presence of lactalbuminhydrolysate (o),bovineserumalbumin(.),gelatin (o),ortype Icollagen(A).(B)SameasinA,but with 50Mg/ml ceruloplasminIinsteadofCuSO4.Lipoproteinoxidationwas
deter-minedasTBARS andexpressedasnanomolesMDAequivalentsper
proteolytically
modifiedceruloplasmin
>gelatin
> type I collagen. All proteins inhibited LDL oxidation by CuSO4 and by ceruloplasminwith similar potencies, suggesting that themechanismsof oxidant activity by both agents may be related. The results show that protein contaminants can interfere with
ceruloplasmin oxidant activityand, in conjunction with
degra-dation,may account fordifferencesin oxidant activity of ceru-loplasmin preparations. They also show that in cupric
ion-me-diatedor in intactceruloplasmin-mediated oxidation, the
an-tioxidant activity of degraded or commercial sources of
ceruloplasmin isa nonspecific effectin common with many proteins.
Boundceruloplasmin copper is responsiblefor oxidant
activ-ity. Theparallel timecoursesof oxidationof LDL by
cerulo-plasmin and by CuSO4 described above (and parallel
inhibi-tion byproteins)suggest that copper may be responsible for
ceruloplasmin oxidant activity. Sincethe purification of
ceru-loplasmin involved multiple column elutions (some under
conditions of high ionicstrength) and multiple filtration steps, we did not believe it likely that significant amounts of free copper could becopurifying withtheprotein.Totestthe possi-bility directly,theoxidant activity of ceruloplasminwas
exam-ined after thorough washingbyultrafiltration using I0-kD
Mi-crocon filters(Amicon Corp.,Danvers, MA).Approximately 90% oftheoxidantactivitywasretainedbythe filter and< 10% passed through, confirmingthe association of the oxidant cop-per atomswithceruloplasmin. Tofurthertestthehypothesis
thatceruloplasmincopper wasresponsible forLDLoxidation,
we examined the ability ofachelating agent, Chelex-l00, to
inactivate ceruloplasmin. Ceruloplasminwas incubated with Chelex-100beadsfor4h at room temperature and the super-natantassayedforitsabilitytooxidizeLDL. Thistreatment
decreased ceruloplasmin copper content from - 6.6 to 5.5
atomsper moleculeofprotein,aresultconsistent withthe re-moval of 1 of7 copper atoms as reported previously (47). Incubation of ceruloplasmin with Chelex-100completely sup-pressed theoxidantactivity oftheprotein (Fig. 9) without sig-nificantly altering oxidase activity (not shown). As was the casewithprotease treatment,
suppression
ofceruloplasmin
ox-idant activity bytreatment withChelex-100wasaccompanied by the appearance of antioxidant activity as determinedby inhibition ofLDL oxidation in the presence of
CuSO,4
(not shown). Restoration ofcopperbyincubation withCuSO4 (fol-lowedby ultrafiltration) almostcompletely restored oxidant activity to Chelex-treatedceruloplasmin.
This demonstrates that theinactivation by Chelex-100wasnotduetoirreversiblechangesintheoxidantcapacity ofthe
protein
andconfirmsthecriticalrole of copper in
ceruloplasmin
oxidantactivity.
Discussion
Theseexperimentsdemonstrate that
purified,
undegraded hu-manceruloplasminhas potentoxidantactivity
towards LDL. We have several lines of evidence that thisactivityis duetoceruloplasmin
itself andnot to aminorcontaminant:(a)
theceruloplasminused in these
experiments
ishomogeneousby
silver staining of SDS-PAGE, by
ceruloplasmin
oxidase spe-cificactivity,andbyanabsorbance ratio(610/280
nm)ofatleast 0.045; (b)both glycosylated isoforms of
ceruloplasmin
have equal oxidant capability; (c) ceruloplasmineluted from
nondenaturingagarose gels ishighly active; (d)two treatments knowntospecificallyalter
ceruloplasmin
structure,i.e.,
incu-bation withaceruloplasmin-degrading
protease and withChe-35 jZ30
25
1
202.
I
10 51
o 50100 200 50 100200 50 100 200 50 100200 Untreated Chelex.Treted Clx-Tratezd, CopperOnly
Cerulopesmin(orequivalnt)
(pgml)
Figure 9. Effectofcopperdepletiononceruloplasminoxidant activ-ity. Purified ceruloplasminI(200
Mg)
wasincubated withChelex-100(150
Ml
of beads washed withmetalion-freePBS)for4hat23°C.Analiquot (100
Mg)
ofthesupernatantwasincubated withexcessCuSO4(
10MM)
for2 h at room temperaturefollowedbywashingbyrepeatedultrafiltration(6-10times)toremoveunboundcopper. LDL(Img/
mlcholesterol)wasoxidized by incubated for 20hat370Cwith
un-treatedceruloplasmin(lighthatchedbars),withChelex-treated
ceru-loplasmin (black bars),withChelex-treatedceruloplasminthat
un-derwentCuSO4replacement (openbars), or,as acontrolto make sure thatfreeCuSO4wasremovedbythe washprocedure,withan
ultra-filteredCuSO4solution(ceruloplasmin-free) equivalenttothat used inthecopperreplacementtreatment(darkhatchedbars).
Lipo-protein oxidationwasdeterminedasTBARS andexpressedas
nano-moles MDAequivalentspermilligramLDLcholesterol.
lex-
100, also moderate oxidant activity; and (e) immunoprecip-itation by a ceruloplasmin-specific antibody removedessen-tiallyalloxidantactivity.
Theoxidant activitywas unanticipated in view of reports
from many laboratoriesthat ceruloplasmin is a potent lipid
antioxidant (see references14and 49 for review). We believe that the mostlikelysourceofthe different results is the purity and structural integrity ofceruloplasminitself. The purified
ceruloplasminusedin thisstudy is prepared from fresh plasma
accordingto amodifiedprotocol that removes a contaminating
metalloproteinase and yields an intact 1 32-kD monomer with
minimal degradationtothe115-kD or smaller fragments (3 1 a). The ceruloplasmin from commercial sources is highly
de-graded with the 132-kD monomer entirely missing and
re-placedby smallerfragments.We have shown that specific
deg-radation of ceruloplasminto the
115-
and l9-kD fragments isaccompaniedby the disappearance of oxidant activity and ap-pearance of antioxidant activity. Similarly, highly degraded
commercial ceruloplasminlacksoxidantactivity and is an ap-parentantioxidantincupric ion-mediated oxidationof LDL. While ourresults suggest a new physiological role for
ceru-loplasminthatis counter to most current ideas, the seemingly
disparate findingsare notincompatible.Theantioxidant
activ-ity of ceruloplasmin canbeascribed to twoofits
properties.
The first isa
nonspecific,
antioxidant effect thatcan be ob-served for almost anyprotein in,
forexample,
acupric
ion-me-diated lipidperoxidation
reaction(Fig.
7 and reference16).
Second, theoxidaseactivityof
ceruloplasmin,
inherent in the intact,purified proteinas wellasin thedegradedprotein
that dominatescommercialpreparations, participatesintheredoxmodel"(50), drives the iron equilibrium from an actively oxi-dizing mix of both valencestates tothe inactive ferric state, thusaccounting for its antioxidant activity. These two proper-tiescould explain the previous reports showing ceruloplasmin tobe an antioxidant.
Our results clearly implicate bound copper in ceruloplas-minoxidant activity. Thefirstevidence isthe resemblance of
ceruloplasmin-oxidized LDLand copper-oxidizedLDL with respect tomultiple chemical and physical indicators of oxida-tion. When aconcentrationof copperischosen so that the rate of formation of LDL TBARS matches that causedbyagiven concentration ofceruloplasmin,other measures of LDL oxida-tion that we have tested also matchclosely.Furthermore, sev-eralproteins block the oxidation of LDL mediatedbyboth free copper and ceruloplasmin, suggesting similar mechanisms. Most importantly, treatment of ceruloplasmin with Chelex-100, which removes asingle copperfrom ceruloplasmin, al-mostcompletely suppresses theoxidantactivity,whichcan be
subsequentlyrestoredby readdition of copper. However,just the presenceoftheoxidantcopperatom(s) isnotsufficientfor
oxidant activity since protease treatment completely sup-pressestheactivity without removingany copper.Thisresultis consistent with a structural modification ofthe molecule that blocks theaccessibilityoftheoxidantcopper site.Interestingly, neither chelation nor proteolysis alterceruloplasmin oxidase activity, indicating that the active sites and specific copper atomsrequired for oxidaseandoxidantactivities may not be thesame.
Little is known about the structure ofceruloplasmin in vivo. We and others (36) have shown by Western blot analysis thatceruloplasmin in plasmaconsistsprimarily oftheintact
132-kD monomeraccompanied by a small amount ofthe 1 5-kDdegradationfragment. Wethereforeargue thatcirculating ceruloplasmin ispredominantlyin theforminwhichthe
oxi-dant rather thanantioxidant activitymay beexpressed.Recent electronparamagneticresonancestudies ofplasmahaveshown that the spectrum of ceruloplasmin in plasma is essentially
identical to purified ceruloplasmin (51). In addition to the
ceruloplasmin isoforms used in these experiments, other iso-forms have been observed in plasma,includingan
apocerulo-plasminthatis completely devoid ofcopper and a 200-kD
mol-ecule,but theamountsandrolesof theseproteinsareuncertain
(36,52,53).
The mechanisms causing LDLoxidation in vivo are not known and we canonlyspeculateonthe roleofceruloplasmin
in this process. Thereissomeevidence for oxidized LDLin plasma (21, 54). A process by whichceruloplasminmay
oxi-dizecirculatingLDLthatis then taken upbythe artery wall may beenvisioned.Theaccumulationof oxidatively modified
LDLin the plasma may belimitedbyplasmaantioxidantsand
by rapidremoval by theliver (55). In an alternatescenario, ceruloplasminmayoxidizeLDLwithinthearterialwallitself.
There isasyet noevidence for ceruloplasminineithernormal orlesioned blood vessels. However,ceruloplasmincould accu-mulateinthevessel wall byatleasttwopathways: uptakeand
synthesis.Insupport of the firstmechanism, specific high
affin-ity receptors for ceruloplasmin have been described in liver endothelium (56) and in aortic extracts (57). Alternatively,
conditionsorprocedures thatcausevascularinjury(surgeryor angioplasty), in which ceruloplasmin levels are elevated (58) andendothelialpermeabilitytoplasma-borne substancesis in-creased, mayresult in high interstitial concentrations of both ceruloplasmin and lipoproteins. With respect to the second
mechanism,while theliveris believedtobe the primary site of
synthesis of ceruloplasmin in the adult, activation of alveolar
macrophages by inflammatory agents stimulates
ceruloplas-min gene expression and protein production (59). A similar activation and response by macrophages in vascular lesions is possible.
Vascular endothelial cells and smooth muscle cells oxidize LDLby mechanisms requiring transition metals (40, 60), and activatedhuman monocytes oxidize LDL even in metal ion-poor media(61, 62). The role of ceruloplasmin in these pro-cesseshasnotbeeninvestigated. Much of theresearchonthe mechanisms of oxidation of LDL by cells has been done with soluble metalion-dependent systems; however, the existence of freeiron and copper ions in vivo is uncertain. Our results invite speculation that the participation of copper in cerulo-plasmin-bound form could be an in vivo analogue of metal ion-dependent oxidation of LDL, which could pertain in ei-ther cell-free and cell-dependent systems. Oei-ther transition metal-containing proteins may also participate in lipoprotein oxidation and, in fact, Balla et al. (63) have shown that the iron-binding protein hemin can oxidize LDL in vitro. Their observations differ from our own in that hemin-mediated oxi-dation requires a source ofperoxidesand the addition of exoge-nous metal ion.
The physiological role of ceruloplasmin has not yet been identified with certainty. Menkes' and Wilson's diseases are inherited human disorderscharacterizedbymarkedly reduced plasmaceruloplasmin levels (64). These diseases, however, are notparticularly informativeabout thephysiological or patho-logical functions ofceruloplasmin, since the primary defect does notreside within theceruloplasmingene, but rather in-volves defective cellular copper transport, which is likely to affect all copper proteins (65). Our findings recommend a cautiousapproachwhenconsideringthephysiologicalrole of
ceruloplasmin.Although muchprevious evidencesupports its role as anantioxidant that protectslipidsand cellsagainstfree radical-induced injury, ceruloplasmin may also have a
damag-ing, oxidantactivity. The structureofthe molecule invivomay be of paramount importance indetermining its action. Our dataindicate that theproportion of ceruloplasminin the unde-graded 132-kDform (compared with degraded) and the pres-ence orchemical coordination of complexed copper atoms are
criticalparameters indeterminingthe netoxidantactivity.The
importance ofoxidantactivityin the physiological role
ofceru-loplasmin isnotknown, butitmaybe centralinits primary
function, e.g., as a participant in the host defense system through an injurious oxidant action on invasive organisms. Thisactivityis consistent with ceruloplasmin's role in the acute phase responseoccurring afterinfectionorinflammation and, infact,abactericidalactivity has been recently reported (66).
Inactivation of ceruloplasminby ametalloproteinasemay rep-resent aphysiologicalmechanism to limit the scopeof itsacute oxidant action. The ability ofceruloplasmin to oxidatively modifyLDLmay help toexplain recent studies showingthat serumceruloplasminisanindependentrisk factorin the pre-diction ofmyocardial infarction and cardiovascular disease (67-70).
Acknowledgments
Health.P. L. Fox isanEstablished Investigatorof theAmericanHeart
Association.
References
1.Ortel, T. L., N.Takahashi, andF.W. Putnam. 1984.Structuralmodelof
humanceruloplasminbased oninternal triplication, hydrophilic/hydrophobic character, and secondary structure of domains. Proc. Nat!. Acad. Sci. USA.
81:4761-4765.
2.Church,W. R., R. L.Jernigan,J. Toole, R. M. Hewick, J. Knopf, G. J. Knutson, M. E.Nesheim,K.G.Mann, and D. N. Fass. 1984.Coagulationfactors Vand VIII andceruloplasmin constituteafamily ofstructurally related proteins. Proc.Nat!.Acad.Sci. USA.81:6934-6937.
3.Gutteridge, J. M.,and J. Stocks. 1981. Caeruloplasmin:physiologicaland
pathological perspectives.Crit.Rev.Clin. Lab. Sci. 14:257-329.
4.Ryden, L. 1984.Ceruloplasmin.In CopperProteinsand Copper Enzymes, vol.III.R.Lontie,editor.CRCPress, Inc., Boca Raton, FL. 37-100.
5.Frieden,E.1986. Perspectivesoncopperbiochemistry. Clin. Physiol.
Bio-chem. 4:11-19.
6.Alcain,F., H. Ldw,andF. L.Crane. 1991.Ceruloplasmin stimulatesthymi-dineincorporationby CCL-39 cells in the absence ofserum orgrowth factors. Biochem. Biophys.Res.Commun. 180:790-796.
7.Alessandri, G.,KRaju,and P. M.Gullino.1983.Mobilizationofcapillary
endothelium in vitroinducedbyeffectors of angiogenesisin vivo. Cancer Res.
43:1790-1797.
8.Stocks,J., J. M.C.Gutteridge,R. J.Sharp,and T. L. Dormandy. 1974.The inhibition of lipid autoxidation byhuman serum andits relationtoserum
pro-teins and a-tocopherol. Clin. Sci. Mol. Med. 47:223-233.
9.Al-Timimi,D.J., and T. L. Dormandy.1977.Theinhibitionoflipid
autox-idationby humancaeruloplasmin. Biochem.J. 168:283-288.
10.Goldstein,I.M., H. B. Kaplan, H. S.Edelson,andG.Weissmann.1979.
Ceruloplasmin. A scavenger ofsuperoxide anion radicals. J. Biol. Chem. 254:4040-4045.
11.Yamashoji, S., and G. Kajimoto. 1983. Antioxidant effectof
caeruloplas-min onmicrosomallipidperoxidation.FEBS(Fed.Eur.Biochem.Soc.)Lett.
152:168-170.
12. Samokyszyn, V.M., D. M.Miller, D. W.Reif, and S.D. Aust. 1989. Inhibitionofsuperoxide and ferritin-dependent lipidperoxidationby
ceruloplas-min.J. Biol. Chem. 264:21-26.
13.Lykens,M.G.,W. B.Davis, andE. R. Pacht.1992. Antioxidant activity of bronchoalveolar lavage fluid intheadultrespiratory distress syndrome.Am.J.
Physiol.262:L169-L175.
14.Dormandy,T.L.1981. Caeruloplasmin: acute-phase antioxidant.Agents
Actions. 8:185-197.
15.Gutteridge,J. M. 1985. Inhibition ofthe Fentonreaction bytheprotein caeruloplasminandothercoppercomplexes.Assessmentofferroxidase and
radi-calscavengingactivities. Chem. Biol.Interact.56:113-120.
16. Gutteridge,J.M.C. 1983. Antioxidantpropertiesofcaeruloplasmin
to-wardsiron- and copper-dependentoxygenradicalformation. FEBS(Fed.Eur. Biochem.Soc.)Lett. 157:37-40.
17.Krsek-Staples,J. A.,andR.0. Webster. 1993. Ceruloplasmin inhibits carbonyl formation in endogenous cellproteins. FreeRadical Biol. & Med. 14:115-125.
18.Gutteridge, J.M.C.,R.Richmond,and B.Halliwell. 1980.Oxygen free-radicalsandlipidperoxidation:Inhibitionbytheproteincaeruloplasmin.FEBS
(Fed.Eur.Biochem.Soc.)Lett. 112:269-272.
19. Lovstad,R. A.1982.Theprotectiveactionofceruloplasminoncopperion stimulatedlysisofraterythrocytes.Int. J.Biochem. 14:585-589.
20.Nakano, H.,K.Ogita,J. M.Gutteridge,and M. Nakano.1984. Inhibition bytheprotein ceruloplasminoflipid peroxidationstimulatedbyan
Fe'+-ADP-adriamycincomplex. FEBS(Fed.Eur.Biochem.Soc.)Lett. 166:232-236.
21. Avogaro,P., G.B.Bon, and G. Cazzolato. 1988. Presence ofamodified
lowdensity lipoproteinin humans.Arteriosclerosis. 8:79-87.
22. Bowry, V. W., K. K.Stanley,and R.Stocker.1992.High density
lipopro-tein is themajor carrier oflipid hydroperoxidesin humanbloodplasma from fastingdonors. Proc.Nat!.Acad.Sci. USA.89:10316-10320.
23.Chisolm,G. M., K. C.Irwin,andM.S. Penn. 1992.Lipoprotein oxidation
andlipoprotein-inducedcellinjuryindiabetes. Diabetes. 41:61-66.
24.Salmon, S.,C. Mazibre, L.Theron, I. Beucler,M. Ayrault Jarrier,S.
Goldstein,and J.Polonovski. 1987.Immunologicaldetection oflow-density
li-poproteins modifiedbymalondialdehydein vitro or in vivo.Biochim.Biophys.
Acta. 920:215-220.
25.Yla-Herttuala,S. 1991.Macrophagesand oxidized lowdensity
lipopro-teins in thepathogenesisofatherosclerosis.Ann.Med. 23:561-567.
26.Kita,T., Y. Nagano, M.Yokode,K.Ishii,N.Kume,A.Ooshima,H.
Yoshida,andC. Kawai. 1987.Probucolprevents theprogressionof atherosclero-sis in Watanabe heritablehyperlipidemicrabbit,ananimal model forfamilial
hypercholesterolemia.Proc.Nat!.Acad.Sci. USA.84:5928-5931.
27.Carew,T.W.,D.C.Schwenke,and D.Steinberg.1987.Antiatherogenic
effect ofprobucolunrelated to itshypocholesterolemiceffect: evidence that
an-tioxidantsinvivo canselectivelyinhibit low densitylipoprotein degradationin
macrophage-richfatty streaks and slow theprogressionofatherosclerosisin the Watanabe heritablehyperlipidemicrabbit. Proc.Natd.Acad. Sci. USA. 84:7725-7729.
28.Steinberg,D.,and WorkshopParticipants. 1992.Antioxidantsin the pre-vention of humanatherosclerosis: summary of theproceedingsof a National Heart, Lung, and BloodInstitute workshop: September5-6, 1991, Bethesda,
MD.Circulation. 85:2337-2344.
29.Chisolm,G. M. 1991.Antioxidantsandatherosclerosis:a current assess-ment.Clin. Cardiol.14:1-25-I-30.
30. Gaziano,J. M., J. E. Manson, P. M. Kidker,J. E. Buring,and C. H.
Hennekens. 1990. Beta carotene therapy for chronic stable angina.Circulation. 82:201. (Abstr.)
31.Stampfer,J. J., C. H.Hennekens,J. E.Manson,G. A. Colditz, B.Rosner,
and W. C. Willett. 1993. Vitamin Econsumptionand the riskofcoronarydisease in women. N.Engl. J. Med. 328:1444-1449.
31a. Ehrenwald,E., and P. L. Fox. 1994.Isolationofhumanceruloplasminby
chromatographic removal of 9 plasma metalloproteinase. Arch. Biochem. Biophys.309: in press.
32. Noyer, M., F. E. Dwulet, Y. L. Hao, and F. W. Putnam. 1980.Purification
and characterization ofundegraded human ceruloplasmin. Anal. Biochem. 102:450-458.
33.Broman,L. 1958.Separationandcharacterizationoftwocoeruloplasmins
from humanserum.Nature(Lond.). 182:1655-1657.
34.Sunderman,F. W., and S.Nomoto.1970.Measurementof human serum
ceruloplasminbyitsp-phenylenediamineoxidaseactivity.Clin.Chem.
16:903-910.
35.Ryden,L., andI. Bjork.1976.Reinvestigationof somephysicochemical
and chemical properties of humanceruloplasmin (ferroxidase). Biochemistry. 15:3411-3417.
36. Sato, M., M. L. Schilsky,R. J. Stockert, A. G. Morell, and I.Sternlieb.
1990.Detectionof multiple forms of humanceruloplasmin.Anovel
M,
200,000form. J.Biol. Chem. 265:2533-2537.
37.Kingston,I. B., B. L.Kingston,and F. W.Putnam.1977.Chemical evi-dence thatproteolyticcleavage causes theheterogeneity presentin human cerulo-plasmin preparations.Proc.Natl.Acad. Sci. USA.74:5377-5381.
38.Takahashi,N., T. L. Ortel, and F. W.Putnam. 1984.Single-chain struc-ture of humanceruloplasmin:the complete amino acid sequence of the whole molecule. Proc.Natl. Acad.Sci. USA.81:390-394.
39.Laemmli,U. K. 1970.Cleavageofstructuralproteins during theassembly
of the headof bacteriophageT4. Nature(Lond.). 227:680-685.
40. Morel, D. W., P. E.DiCorleto,and G. M. Chisolm. 1984.Endothelialand smooth muscle cells alter low density lipoprotein in vitro by free radical oxida-tion.Arteriosclerosis. 4:357-364.
41.Kosugi, K., D. W. Morel, P. E.DiCorleto, and G. M. Chisolm. 1987. Toxicity ofoxidized low-density lipoproteintocultured fibroblastsisselectivefor S phaseofthe cell cycle.J.Cell. Physiol. 130:311-320.
42. Schuh, J., G. F.Fairclough,and R. H.Haschemeyer.1978.
Oxygen-me-diatedheterogeneity of apo-low-density lipoprotein.Proc.Natl.Acad. Sci. USA.
75:3173-3177.
43.Esterbauer,H., G. Striegl, H. Puhl, and M.Rothender.1989.Continuous monitoringof in vitro oxidation of human low densitylipoprotein.FreeRadical.
Res.Commun. 6:67-75.
44.El-Saadani,M., H.Esterbauer,M.El-Sayed,M. Goher, A. Y.Nassar,and G.Jtlrgens.1989. Aspectrophotometricassayforlipid peroxides in serum lipo-proteins using acommerciallyavailable reagent.J.Lipid Res.30:627-630.
45.Steinbrecher,U. P.,S. Parthasarathy,D. S. Leake, J. L.Witztum,and D.
Steinberg. 1984. Modification oflowdensity lipoproteinbyendothelialcells in-volves lipidperoxidationanddegradationof low densitylipoprotein
phospho-lipids. Proc.NatLAcad.Sci. USA.81:3883-3887.
46. Fox, P. L., G. M. Chisolm, and P. E.DiCorleto.1987.
Lipoprotein-me-diatedinhibition of endothelialcellproductionofplatelet-derivedgrowth factor-likeprotein dependsonfree radical lipid peroxidation.J.Biol. Chem.
262:6046-6054.
47.Huber, C. T., and E.Frieden. 1970. Substrateactivationandthekinetics of ferroxidase. J.Biol. Chem. 245:3973-3978.
48.Kalant,N., and S.McCormick.1992.Inhibitionbyserumcomponentsof oxidation andcollagen-bindingoflow-densitylipoprotein. Biochim. Biophys.
Acta. 1128:211-219.
49.Halliwell,B.,and J. M. C.Gutteridge.1986. Oxygen freeradicalsand iron in relation to biology andmedicine: Someproblemsandconcepts.Arch.
Bio-chem. Biophys. 246:501-514.
50.Samokyszyn, V.M., D. M. Miller, D. W. Reif, and S. D. Aust. 1990.
Iron-catalyzedlipidperoxidation.InMembraneLipidOxidation,vol. 1. C. Vigo-Pelfrey, editor. CRC Press, Inc., Boca Raton, FL.10l-127.
51. Musci, G., M. C.Bonaccorsidi Patti, U. Fagiolo, and L. Calabrese. 1993.
Age-related changesin humanceruloplasmin.Evidenceforoxidative
modifica-tions. J.Biol. Chem. 268:13388-13395.
53.Sato, M.,andJ.D.Gitlin. 1991. Mechanisms of copper incorporation
during the biosynthesis of humanceruloplasmin. J.Biol.Chem.266:5128-5134. 54.Morel,D.W., and G. M. Chisolm. 1990. Antioxidanttreatmentof dia-beticratsinhibitslipoprotein oxidation and cytotoxicity.J.LipidRes. 30:1827-1833.
55. van Berkel, T. J. C., Y. B. De Rijke, and J. K. Kruijt. 1991. Different fate in vivoofoxidatively modifiedlowdensity lipoproteinand acetylated low density lipoprotein in rats.Recognitionby various scavenger receptors on Kupffer and
endothelialliver cells. J.Biol. Chem.266:2282-2289.
56.Kataoka, M., and M. Tavassoli.1985.The role ofliver endothelium in the
bindinganduptakeofceruloplasmin:studieswith colloidal goldprobe. J.
Ultra-struct. Res.90:194-202.
57.Stevens,M.D.,R. A.DiSilvestro,and E. D.Harris. 1984.Specificreceptor
forceruloplasminin membranefragments from aorticand heart tissues.
Biochem-istry.23:261-266.
58. Hallbook, T., and H.Hedelin.1980. Changes in serum copper and serum
ceruloplasmin concentration inducedby surgical trauma. Acta Chir. Scand.
146:371-373.
59. Fleming, R. E., I. P. Whitman, and J. D. Gitlin. 1991. Induction
ofcerulo-plasmingeneexpression inratlungduring inflammationandhyperoxia. Am.J. Physiol.260:L68-L74.
60.Heinecke,J.W., H. Rosen, L. A. Suzuki, and A. Chait. 1987. The role of
sulfur-containing amino acidsinsuperoxide productionandmodificationoflow
densitylipoprotein by arterialsmoothmuscle cells. J.Biol.Chem. 262:10098-10103.
61.Cathcart, M. K., A. K. McNally, D. W. Morel, and G. M. Chisolm. 1989.
Superoxide anion participationin humanmonocyte-mediated oxidation of
low-densitylipoprotein and conversion of low-density lipoproteinto acytotoxin. J. Immunol. 142:1963-1969.
62.McNally,A.K.,G.M.Chisolm,D. W. Morel, and M. K. Cathcart. 1990.
Activatedhuman monocytes oxidizelow-densitylipoprotein by a
lipoxygenase-dependent pathway.J.Immunol. 145:254-259.
63. Balla,G.,H.S. Jacob,J. W.Eaton, J.D.Belcher, and G.M.Vercellotti.
1991.Hemin:Apossible physiological mediator oflowdensitylipoprotein
oxida-tion and endothelialinjury.Arterioscler.Thromb. 11: 1700-1711.
64.Matsuda,I., T.Pearson,and N.A.Holtzman. 1974. Determination of
apoceruloplasmin by radioimmunoassay in nutritional copper deficiency,
Menkes' kinky hair syndrome, Wilson'sdisease,and umbilical cord blood.
Pe-diatr.Res.8:821-824.
65.Vulpe, C.,B.Levinson, S.Whitney,S.Packman,and J. Gitschier. 1993.
Isolation ofacandidategeneforMenkesdisease andevidence that it encodesa
copper-transportingATPase. Nature Genetics. 3:7-13.
66.Klebanoff, S.J.1992. Bactericidal effect ofFe2",ceruloplasmin,and
phos-phate.Arch.Biochem.Biophys. 295:302-308.
67.Belch, J. J., M.Chopra,S.Hutchison,R.Lorimer,R.D.Sturrock, C. D. Forbes, and W. E. Smith.1989.Freeradical pathologyinchronic arterialdisease. FreeRadicalBiol.Med. 6:375-378.
68.Kok, F. J., C. M. VanDuijn,A.Hofman, G.B. Vander Voet, F. A. de Wolff, C. H. Paays, and H. A. Valkenburg. 1988. Serum copper and zinc and the riskofdeathfromcancer and cardiovascular disease. Am. J.Epidemiol. 128:352-359.
69.Salonen, J. T., R. Salonen, H. Korpela, S.Suntioinen,and J.Tuomilehto.
1991. Serum copper and the risk of acute myocardial infarction: A prospective population study in men in Eastern Finland. Am. J.Epidemiol. 134:268-276.
70. Reunanen, A., P. Knekt, and R.-K. Aaran. 1992. Serum ceruloplasmin
level and therisk ofmyocardial infarction and stroke. Am. J. Epidemiol.
