Immunolocalization of Native Antioxidant Scavenger Enzymes ... · Immunolocalization of Native Antioxidant Scavenger Enzymes in Early Hypertensive and Atherosclerotic Arteries Role

403

Immunolocalization of Native AntioxidantScavenger Enzymes in Early Hypertensive and

Atherosclerotic ArteriesRole of Oxygen Free Radicals

Ramesh C. Sharma, Donald W. Crawford, Dieter M. Kramsch, Alex Sevanian, and Qun Jiao

To elucidate the role of oxygen free radicals and lipid peroxidation in the pathogenesis of earlyhypertension and atherosclerosis, we studied the native distribution of three primary arterial antioxidantenzymes (AEs). Specific immunohistochemical localization of superoxide dismutase (Cu-Zn SOD),glutathione peroxidase (GSH-Px), and catalase (CAT) was examined in the arterial wall of New ZealandWhite rabbits: six sham-operated normotensive/normolipidemics (NT/NL), seven coarctation-inducedhypertensive/normolipidemics (HT/NL), eight normotensive diet-induced hyperlipidemics (NT/HL), andsix hypertensive/hyperlipidemics (HT/HL). All three AEs were confined primarily to the endothelium inNT/NL rabbit aortas. However, in HT and HL rabbits a greater proportion of the arterial wall, includingthe endothelium, inner media, and middle media, displayed immunolocalization of three AEs. Multiplelinear-regression analysis revealed that more than 70% of the total variability in the depth ofimmunolocalization of arterial AEs could be explained by changes in blood pressure and/or totalcholesterol. Also, levels of plasma and arterial cholesterol oxides were significantly different (p<0.05) inHT and HL rabbits compared with controls, with twofold increases in NT/HLs, threefold increases inHT/NLs, and fourfold increases in HT/HLs. We conclude that intense free-radical activity in the arterialwall of HT and HL animals is one possibility and that this occurs despite the presence of abundant AEs.(Arteriosclerosis and Thrombosis 1992;12:403-415)

KEY WORDS • hypertension • oxygen free radicals • cholesterol oxides • antioxidant enzymes •glutathione peroxidase • superoxide dismutase • catalase • hypercholesterolemia • polyclonalantibodies

While it is known that hypercholesterolemiaand hypertension are two major indepen-dent risk factors involved in the etiopatho-

genesis of atherosclerosis, many of the fundamental cellprocesses involved are yet unclear. Oxygen-derived freeradicals such as superpxide (O2), hydroxyl ions (OH),and lipid peroxide (RO) and the toxic oxygen metabo-lite hydrogen peroxide (H2O2) have been discussed as apotential source of the deleterious effects on the vascu-lar endothelium directly or indirectly through the oxi-dative modification of low density lipoprotein (LDL).1

In vitro studies have indicated that monocyte-derivedmacrophages cannot take up unmodified LDL rapidlyenough to produce lipid loading, suggesting that un-modified LDL may not play a significant role in fattystreak formation in vivo.2 However, other studies sug-gest that oxidative modification of LDL may occur invivo, potentially converting this lipoprotein to a form

From the Atherosclerosis Research Institute, University ofSouthern California, Los Angeles.

Supported by the National Heart, Lung, and Blood InstituteProgram Project, Bethesda, Md.

Address for correspondence: Ramesh C. Sharma, MD, PhD,Atherosclerosis Research Institute, USC School of Medicine, 2011Zonal Avenue, HMR 804, Los Angeles, CA 90033.

Received February 1, 1991; revision accepted December 12,1991.

that is recognized by the macrophage scavenger recep-tors.3 Both endothelial cells and arterial smooth musclecells (SMCs) have been shown in vitro to actively oxidizeLDL and could be a potential source of its oxidativemodification in vivo.4 The cytotoxicity of oxidized LDLhas been recently demonstrated in vitro with culturedfibroblasts,5 human umbilical vein endothelial cells,6

and arterial SMCs.7

Direct evidence to support the existence of oxidizedLDL in vivo has only recently begun to emerge. Usingimmunohistochemical methods and specific monoclonalantibodies, Haberland et al8 demonstrated malondial-dehyde-tysine residues in Watanabe heritable hyperlip-idemic rabbit atherosclerotic lesions. In addition, sev-eral different oxidation-specific epitopes, including4-hydroxynonenal,9 apolipoprotein B, and LDL oxi-dized in the presence of the promoter transition metalcopper10 have been reported recently. Autoantibodiesto malondialdehyde-tysine in human and rabbit plasmahave also been described.11

From an oxidative stress point of view, the arterialendothelium and intima are unique tissues. The partialpressure of oxygen (P02) is high in the intima, poten-tially facilitating lipid peroxidation.12 The observationthat P02 rapidly falls to low levels in the media of theaorta in rabbits with early short-term hypertension,alongside evidence of extensive lipid peroxidation in the

by guest on June 13, 2017http://atvb.ahajournals.org/

Dow

nloaded from

http://atvb.ahajournals.org/

404 Arteriosclerosis and Thrombosis Vol 12, No 4 April 1992

well-oxygenated intima and subintima, is interesting butis scientifically unexplained.13 However, the oxygendistribution in the sequestered subintimal microenvi-ronment above an area of hypoxia deeper in the mediamay facilitate widespread oxygen free-radical formationleading to peroxidation of lipids and thus, formation oflesions. Increased mural oxygen consumption also hasbeen reported by others in experimental hypertension.14

Furthermore, evidence is mounting that in the hyper-tensive artery, increased metabolism as well as diffu-sional limitation may play a significant role in theformation of intimal thickening.15 Therefore, the heter-ogeneous mural oxygen distribution in the sequesteredsubintimal microenvironment may facilitate extensiveoxygen free-radical generation.

On the other hand, to combat such a hostile hyper-oxidative stress, living cells possess an elaborate systemof defense. They possess three primary antioxidantenzymes (AEs), namely superoxide dismutase (Cu-ZnSOD), glutathione peroxidase (GSH-Px), and catalase(CAT), which have the capacity to destroy the highlytoxic oxidant species O2, organic hydroperoxides(ROOH), and H2O2, respectively.16 These powerfulscavenger enzymes play a protective and regulatory rolein the defense of cells against a variety of exogenous andendogenous oxidants.17

Despite the progress made in modern technology, ithas been difficult to unequivocally show the involvementof oxygen free-radical species directly in biologic tis-sues. However, electron spin resonance spectroscopy18

and spin trapping methods offer indirect evidence toconfirm the presence of these short-lived labile oxygenfree-radicals.19 Del Boccio et al,20 using biochemicalassay methods, have studied time-related changes inaortic AEs in cholesterol-fed rabbits, but the preciselocalization of the AEs by immunohistochemical meth-ods in atherosclerotic or hypertensive arteries has notbeen reported to our knowledge. Therefore, we deter-mined to study the potential role of oxygen free radicalsin the pathogenesis of hypertensive as well as athero-sclerotic intimal lesions by examining the distribution ofnative AEs in the states of hypertension, hyperlipid-emia, or both conditions combined, using specific im-munocytochemical techniques. In another series of ex-periments to further elucidate the role of oxygen freeradicals, we analyzed the cholesterol oxide content inrabbit aortic arch sections and plasma.

MethodsStudy Design

All rabbit experiments were performed according tothe rules and regulations approved by the Department ofHealth, Education, and Welfare and the Animal Careand Use Committee of the University of Southern Cali-fornia, Los Angeles. Twenty-eight New Zealand Whiteadult male rabbits weighing 2.7-3.5 kg were obtainedfrom Irish Farms, Corona, Calif., and randomly dividedinto four groups. Two of these groups were kept nor-motensive (NT), and the other two groups were renderedhypertensive (HT) by surgical coarctation of the upperabdominal aorta as previously described.13 All the NTanimals were sham-operated and subjected to the samesurgical procedure except for actual ligation and that thesuture remained loosely tied around the aorta without

causing constriction. One NT group and one HT groupwere maintained normolipidemic (NL) by feeding thecontrol diet (Purina Rabbit Chow, Purina Mills Inc., St.Louis, Mo.) while one NT group and one HT group weremade hyperlipidemic (HL) by feeding an atherogenicdiet containing 2% cholesterol and 10% peanut oil asdescribed by Kritchevsky et al.21 The number of animalsper group were NT/NL, six; HT/NL, seven; NT/HL,eight; and HT/HL, six. All animals were housed in steelcages with easy access to water and diet ad libitum for anaverage period of 6 weeks. Animals were provided withlight between the hours of 6 AM and 6 PM daily in atemperature-controlled environment (76-78°F).

Blood Samples

Blood samples were collected from the central earartery of each animal after an 18-hour fast at baselineand thereafter at biweekly intervals. Samples weredrawn into glass Vacutainer tubes containing 1.5 mgEDTA/ml blood and placed on ice immediately. Plasmawas separated in a centrifuge set at 4°C and spun at3,000 rpm for 20 minutes.

For the measurement of cholesterol oxides, 2-mlplasma samples were stored under argon in the dark at—70°C until analysis, as described by Sevanian andMcLeod.22 Total plasma cholesterol measurementswere done by standard enzymatic methods.23

Blood Pressure Measurements

Routine weekly blood pressures were measured in adark room by the ear capsule transillumination methodof Grant and Rothschild24 after the animals had beenwarmed and were well rested. A mean of at least fivesuccessive measurements was recorded, and the averagestandard deviation of these groups of measurementswas ±4 mm Hg. In the coarcted animals the bloodpressure was recorded 2 weeks postoperatively and wasfollowed by weekly recordings until the time the animalswere killed. It has been demonstrated that pressuremeasured by this method is linearly correlated withmean intra-arterial pressure.25

Autopsy and Tissue Preparation

All animals were anesthetized by intramuscular ad-ministration of ketamine (100 mg/ml) and xylazine (20mg/ml) and were killed by an intravenous overdose ofsodium pentobarbital. The aortic arch was dissected andcleaned in ice-cold normal saline. One 5-mm ringsegment of the midaortic arch was embedded in cryo-mold containing OCT medium (Tissue Tek, FisherScientific, Tustin, Calif.) and snap frozen in liquidnitrogen. The tissues were stored at — 70°C in anultrafreezer until they were ready for cryostat section-ing. Positive control sections for the AEs were obtainedfrom the liver of the same rabbit and processed in thesame manner. Nine-micron-thick cryostat ring sectionsof the aortic arch were cut and placed on microscopeslides that had been precoated with 10% poly L-lysineadhesive subbing solution (Sigma Diagnostics, St. Louis,Mo.) and then air dried. The sections were fixed inacetone for 10 minutes, air dried, and subjected toimmunohistochemical staining. In addition, other adja-cent 5 -mm ring sections of aortic arch were excised andimmediately immersed in Bouin's solution, postfixed in


Dow

nloaded from


Sharma et al Antloxidant Enzymes in Hypertension/Atherosclerosis 405

70% methanol, embedded in paraffin, and sectioned at5 fim thickness. Consecutive sections were stained withhematoxylin and eosin for routine analysis of morpho-logical detail by light microscopy and with VerhoefFs-Sirius red for detection of elastin and collagen.

Immunohistochemistry

Specific poryclonal antibodies against Cu-Zn SOD,GSH-Px, and CAT were employed (kindly provided byN. Rao, Department of Ophthalmology, University ofSouthern California School of Medicine, Los Ange-les).26-28 The specificity of the immunostaining for eachantibody was vigorously tested by performing tests witha variety of different controls. The antibody activity wasdemonstrated by the Ouchterlony gel-diffusion tech-nique. The enzymatic activity of the antisera was as-sayed according to the method described by Abei.29 Thepurity of the enzyme was verified by silver-stained7-5% sodium dodecyl sulfate-poh/acrylamide gel elec-trophoresis. The selective reaction of the antiserum toCu-Zn SOD, GSH-Px, and CAT was determined byimmunoblot technique.

Immunohistochemical staining was performed ac-cording to the method of Atalla et al.27'28 In brief, thepositive control slides and the test slides were treatedwith polyclonal antibodies against the three AEs at afinal dilution of 1:64 for 30 minutes. The negativecontrol slides were reacted with 1) preimmune plasmafrom the same animal that generated the primaryantibody and 2) phosphate-buffered saline (PBS). Allthe slides were incubated in a temperature-controlledwork station (Stainplate, CRL, Cambridge, Mass.) con-comitantly. The sections were washed with 0.1 M PBSfor 10 minutes and were reacted with biotinylated goatanti-rabbit secondary antibody (Vector Labs, Burlin-game, Calif.). Another negative control was done byomitting the secondary antibody and substituting it withPBS. The sections were washed for 10 minutes in PBSand reincubated with avidin-biotin-peroxidase complex(Vector Labs); this was followed by development of3-amino-9-ethylcarbazol H2O2 substrate reaction. Someof the slides were counterstained with hematoxylin.

Image Analysis Microscopy

To determine the relative depth of immunolocaliza-tion of AEs as well as total intimal-medial wall thick-ness of the aorta of different experimental animalgroups, we used a computerized image analysis/lightmicroscopy system. Multiple immunostained arterialcross sections for each AE were evaluated through aZeiss universal microscope with a x20 objective lens.The image was scanned by a Hitachi VK-C 2000 videocamera and digitized by a data translation frame-grab-ber in a Micro-VAX computer, using an algorithmdeveloped by Paul Lee at Jet Propulsion Laboratory,Pasadena, Calif., resulting in a 480x480 image with aresolution of 0.8 /im/pixel. The immunohistochemicalstaining of AEs in the arterial cross section of HT andatherosclerotic animals was not microscopically homo-geneous and varied from case to case. Therefore, wedetermined only the maximum depth of positive AEimmunostaining from the endothelium through the me-dia by manually defining the end points with a cursor.

This was done in three consecutive readings, and themean was calculated (Figure 5).

Analysis of Cholesterol Oxides

Arterial tissue. Freshly excised adjacent minced seg-ments of the aortic arch (approximately 500 mg wetweight) were homogenized in 6 ml chloroform/methanol(2:1, vol/vol) containing 0.01% butylated hydroxytolu-ene, using a Tekman tissuemizer (Tekman, Cincinnati,Ohio). Samples were homogenized and rinsed with twosuccessive washings of the homogenizer probe with theaforementioned solvents. All procedures were carriedout in an ice-cold water bath under an atmosphere ofargon. After allowing the homogenate to stand in asealed tube for 1 hour at room temperature, the sampleswere centrifuged at 2,500 rpm for 15 minutes and thesupernatant was recovered. The solvent was evaporatedto dryness under nitrogen and the residue dissolved andreextracted according to the method of Bligh and Dyer,30

with modifications as described previously.22 The organicphase was evaporated and the residue dissolved in chlo-roform and associated to diol extraction columns (VWRScientific, Cerritos, Calif.), hydrolyzed, methylated withdiazomethane, and converted to o-trimethylsiryl etherderivatives, and 1-yxl aliquots were analyzed by gas chro-matography equipped with a DB-1 capillary column anda flame ionization detector. Chromatographic analyseswere performed with Axxi-chrom 747 analytical chroma-tography software (Axxiom Chromatography, Inc.,Moorpark, Calif.). Oxysterol identity was confirmed bygas chromatography and mass spectrometry.

Plasma. Briefly, six volumes of chloroform/methanol(2:1, vol/vol) containing 0.01% butylated hydroxytolu-ene were added to 1 ml plasma, mixed vigorously for 5minutes, and centrifuged to separate the organic phase,which was collected and saved. The aqueous phase wasreextracted as above with three volumes of solvent, andthe organic phase was pooled. The lipids were evapo-rated to dryness under argon and then redissorved in 1ml argon-saturated toluene/ethyl acetate (1:1, vol/vol).As an internal standard, 100 fA 5a-cholestane (5 mg/mlin toluene) was added to each sample. The sampleswere applied to diol extraction columns as describedpreviously.

The amount of each oxidation product was calculatedas either micrograms per aorta or micrograms perdeciliter plasma. Alternately, the levels of each choles-terol oxide were calculated as a percentage of totalcholesterol in the sample.

Statistical Analysis

In all tests of significance the null hypothesis wasrejected when its probability was 5% or less. Because ofthe nonnormality of the data, nonparametric proce-dures were applied. For more general comparison ofcholesterol and blood pressure determinations amongNT/NL, NT/HL, NL/HT, and HT/HL groups, an analy-sis of variance of the Kruskal-Wallis test was used,coupled with application of the Bonferroni adjustmentwhen appropriate. Within a group changes from base-line were tested using the Wilcoxon signed-rank test forrepeated measures. The relation of blood pressurechange and total cholesterol change to the depth ofimmunohistochemical localization of AEs was analyzed


Dow

nloaded from



TABLE 1. Data From In Vivo Studies of Baseline Versus 6-Week Systolic Blood Pressures and Total Plasma Choles-terol Levels in Different Groups

Parameter

Blood pressure (mm Hg)NT/NL'HT/NLNT/HLHT/NL

Total cholesterol (mg/dl)NT/NL'HT/NLNT/HLHT/HL

n

6786

6786

Baseline

63±367±363±265±2

39±548±449±783±11

6-Week

64±2

83±3t63±2

76±3t

43±533±3

593±66§804±100§

Change

1±116±3t

-2±2l l±7t

4±1-15±3544±98t721 + 120$

Values are mean±SEM.Blood pressures were determined weekly for all animals, and a mean of at least five concordant measurements was

recorded. Plasma cholesterol levels were obtained biweekly. Group comparison of systolic blood pressure data at 6weeks revealed statistically significant difference (/><0.01).

NT, normotensive; NL, normolipidemkr, HT, hypertensive; HL, hyperlipidemic.'Sham operated.tp<0.05, Wilcoxon rank-sum test, compared with NT animals in NL and HL groups.tp<0.05, Wilcoxon signed-rank test for repeated measures; one-sided.§p<0.001, Wilcoxon rank-sum test, compared with NL animals in NT and HT groups.

by calculating simple correlation coefficients followed byapplying multiple linear-regression analysis to test theinteractions between the independent variables and tofind a model that best fit the data. All analyses were donewith the aid of the Statistical Analysis System (SAS).

ResultsThe animals remained in good health and consumed

their respective diets well throughout the experimentalperiod except for one rabbit of the HT/HL group thatdied of congestive heart failure due to excessive coarcta-tion. The amount of food consumed and the weight gainwere not significantly different among the experimentalgroups, nor was there any characteristic difference in thegrowth pattern. At the termination of the experiment,the animals had gained between 568 and 683 g.

Blood Pressure and Plasma Cholesterol

Table 1 lists the recordings of mean systolic bloodpressures and mean total plasma cholesterol levels inthe four experimental groups of rabbits during a periodof 6 weeks. The hypertension was mild but statisticallysignificant (p<0.01) in the coarcted rabbits at the timethe animals were killed compared with preoperativebaseline (as well as with NT animals; see below) withthe following mean values±SEM: 83±3 versus 67±3mm Hg for the HT/NL group and 76±3 versus 65 ±2mm Hg in the HT/HL group, respectively. By contrast,in the noncoarcted sham-operated animals the bloodpressure remained within the normal range at the startand during the experiment with the following groupmeans: 63±3 versus 64±2 mm Hg in NT/NL and 63±2versus 63 ±2 mm Hg in NT/HL rabbits.

The mean total plasma cholesterol levels in theanimals on the standard rabbit chow diet were withinthe normal range at baseline and remained so duringthe experiment with the following group means ±SEM(in milligrams per deciliter): 39±5 versus 43±5 inNT/NL and 48±4 versus 33±3 in HT/NL groups. How-ever, in the rabbits on the atherogenic diet there was a

moderate but statistically significant rise in total plasmacholesterol levels with the following group means (inmilligrams per deciliter) before and after feeding theatherogenic diet, respectively: 49±7 and 593±66 in theNT/HL and 83±11 and 804±100 in the HT/HL group.A relatively moderate increase in the total cholesterollevels, despite feeding the animals a 2% cholesterol and10% peanut oil-rich diet for 6 weeks, could partially bedue to the irregular eating pattern of the rabbits duringthe postoperative convalescence period.

Cholesterol Oxide

Table 2 shows the amounts of cholesterol and cho-lesterol oxides in the plasma and aortic arch of NewZealand White rabbits representing the four experi-mental groups at the time of autopsy. The levels ofcholesterol oxides, particularly cholesterol epoxides andcholestane triols, were significantly elevated both in theplasma and aortic arch of HT and atherosclerotic ani-mals. The increase in the level of plasma cholesteroloxides was more than twofold in NT/HL, threefold inHT/NL, and fourfold in HT/HL compared with that ofthe control NT/NL rabbits. Of particular interest is thefinding of cholesterol-5/3,6/3-epoxides as the major cho-lesterol oxide in nearly all samples, suggesting that amajor portion of the cholesterol oxides were generatedthrough free-radical chain-propagation reactions.31 Al-though the level of circulating cholesterol oxides inHT/NL and HT/HL groups (5.4±0.1% versus 6.1±0.1%,respectively) appears to be similar as shown in Table 2,this observation can easily be explained on the basis thatcholesterol oxides were computed as a percentage oftotal cholesterol, which is much higher in HT/HL. Sim-ilarly, as revealed in the upper portion of Table 2, ahighly significant increase in cholesterol content of thewhole aorta (micrograms per aorta) occurred in both HLgroups compared with the NL groups, NT/HL=231±11and HT/HL=413±8 versus NT/NL=151±4, whereas theslight rise of aortic cholesterol content in the HT/NLgroup (181 ±15) was not significant.


Dow

nloaded from


Sharma et al Antioxidant Enzymes in Hypertension/Atherosclerosis 407

TABLE 2. Concentrations of Cholesterol and Cholesterol Oxide Levels in Serum and Aortic Arch of Rabbits in Dif-ferent Experimental Groups at Sacrifice

ParameterNT/NL*(«=6)

HT/NL(n=7)

NT/HL(n=8)

HT/HL(«=6)

CholesterolSerum (mg/dl)Aorta (ftg/aorta)

Cholesterol oddestSerum (% cholesterol)Aorta (% cholesterol)

91±5151±4

1.5±0.41.9±0.8

% ± 8

181±15

5.4±0.1t5.2±0.7t

667±33*231±11*

3.2±0.2t7.1±0^

918±55*413±8t

6.1±0.1*3.6±0.3$

Values are mean±SEM.Severity of disease in hypertensive as well as in hyperlipidemic rabbits correlated with sharply greater amounts of

cholesterol epoxides (expressed as a percentage of cholesterol) compared with NT/NL control animals. It is of interestthat cholesterol oxides are substantially elevated in HT/NL rabbits despite the fart that cholesterol levels are within thenormal range. Levels of cholesterol and cholesterol oxides in the aortic arch essentially reflect those measured in serum.

NT, normotensive; NL, normolipidemic; HT, hypertensive; HL, hyperlipidemic.'Sham operated.tCholesterol oxides consisted of cholesterol-5ft60-epoxide, cholesterol-5a,6o-epoxides, cholesterol triol, 7-lceto-

cholesterol, and 7-hydroxycholesterol.$p<0.001 vs. NT/NL animals, tested by Wilcoxon rank-sum test. a=0.02, Bonferroni adjustment.

Morphological and Immunohistochemical Results

Gross examination of the aortic arch from all theanimals did not reveal any visible plaques. Immediatelyfrozen sections were examined in this study becauseparaffin-embedded sections revealed a suboptimal im-munohistochemical staining, possibly due to poor anti-gen preservation. Multiple serial cryosections of aorticarch (three sections per slide) of each animal in differ-ent groups were evaluated under optimally dennedconditions for the immunohistochemical demonstrationof primary AEs. Sections incubated with preimmunesera, secondary antibodies, or sera adsorbed withCu-Zn SOD, GSH-Px, or CAT revealed no immuno-staining. Sections from liver showed intense positiveimmunoreactivity of all three AEs.

Normotcnsivelnormolipidemic group. Figure 1A exhib-its the architecture of a normal arterial wall. Immunore-active native Cu-Zn SOD was predominantly localized inthe aortic endothelium of NT/NL rabbits (Figure IB). Athigher magnification, granular aggregates of the immu-nopositive Cu-Zn SOD conjugates were seen in thecytoplasm of the endothelial cells and in the superficialSMC layer of the inner media. However, the rest of thetunica media appeared devoid of the immunostain. Anidentical distribution pattern of enzyme reactivity wasobserved for GSH-Px and CAT (Figures 1C and ID).

Hypertensivelnormolipidemic group. The arterial wallappeared to be considerably thickened in the HT/NLrabbits compared with that of the NT/NL group. Focalintimal thickenings were noted at several areas aroundthe intimal circumference (Figure 2A). The early HTlesions displayed thicker collagen bundles and someintimal neoelastica formation as revealed by VerhoefFs-Sirius red staining. The internal elastic lamina was frag-mented and deranged. Increased cellularity was seen inthe HT lesions. The arterial media also showed appar-ently thicker collagen bundles. These intimal and medialchanges appeared to contribute to the substantiallythicker appearance of aortic wall in HT animals.

By comparison, the antioxidant immunohistochemi-cal staining of the adjacent representative cross sectionsof HT/NL rabbit midaortic arch revealed a greater

proportion of arterial wall with intense Cu-Zn SOD,GSH-Px, and CAT immunoreactivity (Figures 2B-2D).Diffuse dense intracellular as well as extracellular gran-ular deposits of the specific immunostains were evidentin the endothelial cells and in SMCs deep into thearterial media, especially at HT intimal thickenings.Some staining around the internal elastic lamina wasalso noted. The shoulders of the lesion revealed com-paratively decreasing cellular and extracellular stainingof the AEs. The non-lesion-adjacent areas of theHT/NL arterial wall also exhibited a relatively greaterdistribution of AEs compared with NT/NL rabbit aorta.The staining pattern of CAT, although enhanced whencompared with that of NT/NL animals, was relativelyweaker (Figure 2D).

Normotensivelhyperlipidemic group. The morphologi-cal structure of the aortic arch in NT/HL animalsexhibited a typical foam cell lesion formation (Figure3A). The intima was markedly raised because of accu-mulation of a large number of foam cells and someminor deposits of collagen and elastin as revealed by theVerhoefFs—Sirius red stain.

The subsequent consecutive section showed intenseimmunoreactivity of Cu-Zn SOD and GSH-Px (Figures3B and 3C). The positive immunostain deposits wereassociated with cells as well as with the extracellularmatrix of the lesion. However, in contrast to HT/NLanimals, the arterial media under the lesion also showedscattered AE activity. The nonlesion areas revealed ageneral increase in immunostaining compared withthose of NT/NL rabbit aortas. The CAT immunostain-ing again was relatively less intense (Figure 3D).

HypertensivelhyperUpidemic group. Figure 4A exhibits acharacteristic fibrous-fatty lesion as seen in HT/HL rabbitaortas. Increased collagen bundles were evident in theregion as demonstrated by Sirius red stain, while foamcells were strikingly few compared with the fatty streaklesions seen in NT/HL animals. The internal elastic lam-ina was totally fragmented and deranged, and the forma-tion of neoelastica was evident in the lesion intima.

Immunohistochemical localization of the AEs in theadjacent serial cryostat sections revealed extensive en-


Dow

nloaded from



FIGURE 1. Serial cryosections ofmidaortic arch of a hypertensive, normolipidemic rabbit. Panel A: Section shows normal thintunica intima, intact internal elastic lamina, and underlying orderly arrangement of smooth muscle cells and elastic lamellae intunica media. Note the normal width of the arterial wall (Verhoeff's-Sirius red stain). Panel B: Superoxide dismutase; panel C:Glutathione peroxidase; and panel D: Catalase. Avidin-biotin-immunoperoxidase staining in adjacent cryosections showscomparative staining pattern of the three antioxidant enzymes. Note that immunostaining is predominantly localized at theendothelium and that the staining pattern is similar with all antioxidant enzymes. Bar=100 fim. L, lumen; Ad, adventitia; M,media; E, endothelium.

zyme reactivity of Cu-Zn SOD and GSH-Px in thefibrocellular plaque (Figures 4B and 4C). In the lesion,dense granular cytoplasmic enzyme deposits were seenin the cells of the lesion intima, including endothelialcells and medial SMCs. Some positive immunostainproducts also were seen associated with the extracellu-lar lesion matrix. In some lesion areas the localization ofCu-Zn SOD and GSH-Px enzymes was sharply demar-cated, whereas in other areas massive accumulationswere diffuse and confluent. In the nonlesion areas therewas an overall increase of the AE distribution through-

out the tunica media and adventitia. The staining pat-tern of CAT immunoreactivity once again seemedmilder in intensity compared with Cu-Zn SOD andGSH-Px, but CAT was much stronger than that seen inthe normal arteries of NT/NL rabbits (Figure 4D).

Morphometric Image Analysis

The results of morphometric image analysis for thedepth of immunolocalization of AEs across the arterialwall (Figure 5) in the four experimental groups are


Dow

nloaded from



B

FIGURE 2. Comparative staining patterns of consecutive cryostat sections of midaortic arch of a representative hypertensive,normoUpidemic rabbit. Panel A: Section exhibits a typical hypertensive lesion, showing thickened intima and tunica media. Notethe deranged internal elastic lamina (IEL) and neoelastica formation in the thickened intima as well as wider elastic interiamellarspaces in the media due to increased collagen formation (Verhoeff's-Sirius red stain). Panel B: Superoxide dismutase; panel C:Glutathione percaddase; and panel D: Catalase, avidin-biotin-immunoperoxidase staining. These sections reveal the distributionpatterns of the three antiaxidant enzymes. Note the markedly increased depth of immunostaining of the three antioxidant enzymesin the hypertensive lesion. Arrow shows extracellular pools of immunostaining. The intensity but not the depth of catalaseimmunostaining was relatively milder than that of the other antioxidant enzymes. Bar=100 \im.

summarized as a bar graph in Figure 6. In each animalgroup the mean value of the depth of immunohistochem-ical staining as well as total intimal-medial thickness ofthe arterial wall was determined. Compared with normalcontrols (NT/NL), the relative proportion of the depth ofintimal-medial immunohistochemical localization of thethree AEs was much greater in all HT and HL groups

with the following order of depth of involvement in thearterial wall: HT/HL>NT/HL>HT/NL>NT/NL.

Statistical Analysis

Correlations between differences in absolute depth ofintimal-medial AE immunostaining and blood pressure


Dow

nloaded from



FIGURE 3. Serial cryostat sections through midaortic arch of a normotensive, hyperlipidemic rabbit Panel A: Section shows atypical large foam cell/fatty streak lesion, displaying marked thickening ofintima mainly due to increased ceUularity of fat-ladencells and very few collagen strands. Tunica media revealed normal collagen deposits, which were relatively less marked comparedwith those from the hypertensive aortic arch (Verhoeffs-Sirius red stain). Panel B: Superoxide dismutase; panel C: Ghitathionepenvddasc; and panel D: Catalase, avidin-biotin-immunoperoxidase staining. These serial cross sections reveal the extensiveincrease in the immunohistochemical expression of the three antkuddant enzymes across the arterial wall Note the increase inimmunostaining throughout the media as compared with the hypertensive, normolipidemic rabbit aortic arch. Catalaseimmunostaining is relatively less intense. Bar=100 fim. IEL, internal elastic lamina.

change in treatment subgroups at 6 weeks are summa-rized in Table 3. A statistically significant positivecorrelation was observed between blood pressurechange and the depth of immunolocalization of AEs at6 weeks of induction of hypertension (p<0.05). Simi-larly, a much stronger positive correlation (p<0.0001)was observed between the change of total cholesteroland the depth of immunolocalization of the three AEsacross the arterial wall. Obviously, the greater depth ofAE immunostaining in the aortic arch of HL animals

was due to the relatively larger foam cell/fatty streaklesions compared with smaller HT lesions.

Multiple linear-regression analysis was done to eval-uate the relation between change of blood pressure andchange of cholesterol with the depth of intimal-medialimmunolocalization of three AEs, and the results aresummarized in Table 4. We found that more than 70%of the variability in the depth of immunohistochemicallocalization of three AEs could be explained by theincrease in blood pressure and elevation in total plasma


Dow

nloaded from



FIGURE 4. Sequential cryostat sections through the midaortic arch of a representative hypertensive, hyperUpidemic rabbit. PanelA: Section reveals a typical fibrous-fatty lesion showing marked ceUularity due to foam cell accumulation and abundant depositsof collagen. The fragmented and deranged internal elastic lamina and substantially thickened tunica media are evident(Verhoeffs-Sirius red stain). Panel B: Superoxide dismutase; panel C: Glutathione peroxidase; and panel D: Catalase,avidin-biotin-immunoperoxidase staining. Sections exhibit an extensive immunoexpression of the three antioxidant enzymes inearfy stages of combined hypertension and hyperlipidemia. Simultaneous diffuse extracellular (open arrows) and cell-associatedstaining (solid arrow) in the lesion is noted Bar=100 um.

cholesterol. In addition, significant interaction betweenchange of blood pressure and change of total choles-terol in relation to the depth of AE distribution acrossthe arterial wall was observed.

Discussion

The vascular endothelium clearly faces an oxidativechallenge of relative luminal hyperoxia and abluminalhypoxia. In addition to this generally hostile oxidative

environment of arteries, reactive oxygen intermediatessuch as O2, H2O2, OH, RO, and singlet molecularoxygen (O2) are generated, as in any biologic system,even under normal physiological conditions via severalpathways, both enzymatic and nonenzymatic.32 Endog-enous sources of free radicals include those producedduring oxidative phosphorylation in the mitochondria,from the activity of cyclooxygenase, lipoxygenase per-oxidases and dehydrogenase, auto-oxidation, flavins,


Dow

nloaded from



FIGURE 5. Digitized image-analysis photomicrograph of asuperoxide dismutase-immunostained cryostat arterial crosssection from a representative hypertensive, hyperlipidemicrabbit used for quantitative morphometry. The depth ofantioxidant enzyme immunostaining was determined at thesite of maximum distribution across the intima and media bytaking the mean of three consecutive readings. Bar=100 fim.

thiols, and xanthine oxidases.33-34 Toxic oxyradicals arereleased extracellularry during the respiratory burst ofphagocytes.35

100* -t

80%-

60%-

40%

20%

NT/NL HT/NL NT/HL HT/HL

Treatment Group

FIGURE 6. Bar graph showing the relative proportion of thedepth of intimal-medial immunostaining of three antioxidantenzymes across the arterial cross sections of four groups ofrabbits after 6 weeks of diet. An average percentage of totalintimal-medial thickness showing antioxidant enzyme immu-nostaining was used in each treatment group. NT, normoten-sive; NL, normolipidemic; HT, hypertensive; HL, hyperlipid-emic; SOD, superoxide dismutase; GSH, glutathioneperoxidase; CAT, catalase.

To adapt to this oxidative stress, not only endothelialcells but also possibly all cells have evolved a complexAE defense system.36 The in vivo immunohistochemicaldemonstration of Cu-Zn SOD, GSH-Px, and CAT inthe endothelial cells of the NT/NL control rabbit aorticarch suggests that the vascular endothelium is wellequipped in its armamentarium to defend against thehighly toxic oxygen free radicals. The importance of theglutathione redox cycle as a potent antioxidant defensesystem has previously been demonstrated by in vitrostudies in human endothelial cells.37

Our finding of an identical immunohistochemicalexpression of the three AEs reflects a harmoniousinterrelation among Cu-Zn SOD, GSH-Px, and CAT forthe detoxification of oxygen metabolites under normalphysiological conditions. We postulate that the delicatebalance between pro-oxidants and antioxidants could beeasily altered during the early stages of hypertensionand/or hypercholesterolemia.

In the HT/NL rabbits a greater proportion of thearterial wall displayed Cu-Zn SOD and GSH-Px immu-nostaining in the early HT lesions, around the lesionshoulders as well as in the nonlesion areas. The CATreactivity, although increased compared with that ofNT/NL, was relatively less intense. We found intracel-lular as well as extracellular immunohistochemical dis-tribution of the AEs: over the cells of the lesion intimaas well as over the extracellular matrix of the lesion,including along the deranged internal elastic lamina.These results may be interpreted to suggest that exten-sive oxygen free-radical formation occurs during theearly stages of hypertension followed by membrane lipidperoxidation, leading to the release of lysosomal andhydrolytic enzymes.3* These hydrolytic enzymes notonly bring about cell necrosis but also induce drasticmodifications in the chemistry of the extracellular ma-trix.39 Membrane lipid peroxidation has been exten-sively reviewed.40-43 The unsaturated fatty acids andphospholipids present in the membrane and the oxidiz-able amino acids present in transmembrane proteins arethe potential targets of oxyradical damage.16-17-32-33 Sig-nificant changes in membrane chemistry impair thefunctional characteristics, including membrane perme-ability, fluidity, the activity of membrane-associatedenzymes, and finally, cell death. Moreover, it must beemphasized that cell membrane lipids are not the onlycomponents at risk from injury by the endogenouslygenerated free radicals. In addition, the extracellularmatrix such as grycosaminoglycan and especially colla-gen also is sensitive to free-radical damage.33

There is ample evidence that HT stimuli in experi-mental animals lead to early arterial wall hypermetab-olism and augmented oxygen consumption.14-15-4445

Crawford and Kramsch13 have demonstrated relativesubintimal mural hypoxia in the HT rabbit aorta. Simi-larly, Seidel and Strong14 observed a clear increase inoxygen consumption in aortic rings from rats withspontaneous, renal, and deoxycorticosterone salt hyper-tension. These changes in oxygen distribution in the HTarterial wall, with areas of relative hypoxia adjacent toareas of normal oxygen tension, could favor the forma-tion of oxygen free radicals in a fashion analogous toreperfusion after myocardial infarction.46-47

Another condition in which AEs were found distrib-uted in the intima, subintima, and media was in athero-


Dow

nloaded from



TABLE 3. Relation Between Differences in Absolute Depth of Intimal-Medial Antioxidant Enzyme Immunostaining and Blood PressureChange in Treatment Subgroups

Parameter

Cu-Zn SODGSH-PxCAT

NT/NL'depth (/un)

(«=6)6.8±0.78.2±1.38.5±1.0

Normolipidemic

HT/NLdepth (/im)

(«=7)66.4±8.967.6±8.669.4±9.0

Pt0.0030.0030.003

Corrcoefft

0.670.670.67

NT/HLdepth (fun)

(n=8)

134.8±6.3137.6+7.0136.8+5.7

Hyperlipidemic

HT/HLdepth (;un)

(11=6)

321.0±26.3320.8±24.8321.8±25.6

Pt0.0020.0020.002

Corrcoeffj

0.630.660.67

Depth values are mean±SEM of antioxidant enzyme immunostaining.NT, normotensive; NL, normolipidemic; HT, hypertensive; HL, hyperlipidemic; BP, blood pressure; Cu-Zn SOD, superoxide dismutase;

GSH-Px, glutathione peroxidase; CAT, catalase.'Sham operated.tWilcoxon rank-sum test, comparing the depth of arterial wall antioxidant enzyme immunostaining between NT and HT animals in NL

and HL groups, respectively.^Correlation coefficient (Corr coeff) between BP change and depth of arterial wall antioxidant enzyme immunostaining in normolipidemic

and hyperlipidemic groups. All are statistically significant (p<0.05).

sclerotic aortas of NT animals fed an atherogenic dietfor 6 weeks. We observed an even greater depth of theimmunohistochemical distribution of three AEs in theatherosclerotic arterial wall of NT/HL rabbits com-pared with HT/NL animals. One explanation of thisfinding could be that the intimal lesions in purely HLrabbits are large fatty streaks composed of many layersof lipid-laden foam cells, while the purely HT lesionsare relatively thin fibromuscular thickenings of theintima. As recently demonstrated in our laboratory,these HT intimal thickenings contain, apart from SMCs,large numbers of monocytes/macrophages, which appar-ently invade the lesion from the bloodstream.48 It is ofinterest to note that macrophages also are capable ofoxygen free-radical formation and peroxidation of lip-id.7-8 Our results are in agreement with those of DelBoccio et al,20 who have recently reported an increase inCu-Zn SOD and GSH-Px levels in plasma as well as in

TABLE 4. Multiple Regression Models Unking Blood PressureIncrease and Total Cholesterol Elevation to Extent of Immunolo-calization of Antioxidant Enzymes Across the Arterial Wallat 6 Weeks

Dependentvariables

Cu-Zn SOD

GSH-Px

CAT

Independentvariables'

TC (mg/dl)

BP (mm Hg)

TCxBPfIntercept

TC (mg/dl)BP (mm Hg)TCxBPtIntercept

TC (mg/dl)BP (mm Hg)TCxBPtIntercept

Parameterestimate

0.1151.1130.010

44.08

0.1171.2000.010

44.01

0.1181.4280.009

43.25

SEM

0.0461.7490.004

0.0431.6670.004

0.0451.7070.004

P

0.020.530.03

0.010.480.02

0.010.410.03

TotalR1

0.71

0.73

0.72

Cu-Zn SOD, superoxide dismutase; GSH-Px, glutathione per-oxidase; CAT, catalase; TC, total cholesterol; BP, blood pressure.

'All the independent variables are changes from baseline to 6weeks.

•(•Interaction term between blood pressure increase and totalcholesterol elevation in relation to intimal-medial depth of antiox-idant enzyme immunostaining.

the arterial tissue of rabbits fed a hypercholesterolemicdiet for 10, 30, and 60 days. These authors, usingbiochemical assays, also reported a relatively smallerincrease in CAT activity than that found with Cu-ZnSOD and GSH-Px. This finding is further supported bythe observation that GSH-Px protects the endothelialcells more efficiently than does CAT.37 Similarly, Hen-riksen et al2 demonstrated an increase in the activity ofCu-Zn SOD in whole aortas of cholesterol-fed rabbits.Bjornheden and Bondjers12 concluded that, in earlyrabbit atherosclerosis, aortic mural oxygen consumptionis increased because of foam cell activity, therebycausing abluminal hypoxia. Heughan et al49 reportedthat oxygen tension in the arterial wall of HL rabbitswas 10 mm Hg compared with 30-40 mm Hg in normalcontrols.

The statistical analysis of the data using simple cor-relation coefficients revealed that the change in totalcholesterol was relatively more significant (p<0.0001)than the change in blood pressure (/><0.05) in relationto the depth of immunohistochemical localization ofthree AEs across the arterial wall at the 6-week timepoint of the experiment. The results of multiple linear-regression analysis revealed that more than 70% of thetotal variability in the depth of the intimal-medialdistribution of the three AEs in the arterial wall couldbe explained by the elevation in blood pressure and therise of total cholesterol. In addition, significant interac-tion between the change in blood pressure and thechange in cholesterol in relation to the depth of AEdistribution suggests that these two variables have astrong predictive capacity during the early stages of HTand atherosclerotic lesion development.

There is now considerable experimental evi-dence343-49-50 indicating the importance of the roles ofoxidized LDL and recruitment of monocytes/macro-phages in the pathogenesis of atherosclerosis. Addi-tional evidence that oxidized LDL could be generated invivo comes from studies of the immunohistochemicalcolocalization of malondialdehyde-conjugated LDL8 or4-hydroxynonenal LDL9 with unmodified LDL. In lightof the available data, we can interpret our results ofincreased AE activity as evidence of increased free-radical formation in the altered sequestered microenvi-ronment of the arterial intima and subintima in HT andalso hypercholesterolemic rabbits, with a compounding


Dow

nloaded from



of this effect when both conditions are present simulta-neously. Indirect in vitro evidence suggests that oxida-tion of LDL may be mediated by endothelial cells,SMCs, and macrophages.5-8 Native LDL appears to benontoxic to arterial cells.2-3 However, one of the newerhypotheses is that LDL is trapped in the arterial wall byforming complexes with several matrix proteins, andthen it could become peroxidated during a prolongedresidence time in the vessel.33'51-52 Oxidative modifica-tion of LDL has been considered a key step in thegeneration of macrophage-derived foam cells and in theinitiation of the atherosclerotic process; however, theprecise mechanisms and the exact sequence of eventsremain speculative. Steinbrecher et al53 reported a clearconnection between lipid peroxidation and oxidativemodification of LDL. We postulate that increased oxy-gen free-radical formation occurs in the subintima andinner media of HT arteries as well as in arteries exposedto hyperlipidemia, which could be responsible for theoxidative modification of LDL and the subsequentgeneration of oxidation-specific lipid protein adductsderived from LDL in the atherosclerotic lesions.

Therefore, it becomes clear that the surroundingendothelial cells and SMCs must acquire abundant AEsto shield against the cytotoxic effects of oxyradicals,lipid hydroperoxides, and oxidized LDL. The invasionof macrophages to scavenge oxidized LDL through theirscavenger receptors3-50 appears to be a reasonably be-nign function in the early stages of lesion formation.However, later on when the macrophages are over-loaded with oxidized LDL due to continuous hyperlip-idemia and/or hypertension, they become potentiallypathogenic, producing a vicious cycle of respiratoryburst, oxyradicals, lipid peroxidation,54 cytotoxicity, ne-crosis, and chemotaxis.55'56 Steinberg et al43 contendthat oxidative modification is not likely to occur in thecirculation and seems to be a local event in the microen-vironment of the vascular wall. We found, biochemi-cally, a threefold increase in the levels of cholesteroloxides in the arteries and sera of hypertensive rabbits aswell as in rabbits with hypercholesterolemia. This pro-vides additional supportive evidence that the arterialwall is a site of formation of lipid peroxidation products,some of which may have escaped into the bloodstreamto account for their presence in the sera.

We conclude that intense oxygen free-radical gen-eration is one possible process in the arterial wall ofHT and atherosclerotic animals and that this occurseven in the presence of considerable amounts ofantioxidant mechanisms.

AcknowledgmentsThe authors wish to thank Dr. Narsing Rao, Professor of

Pathology and Ophthalmology, Dr. Atalla of the Doheny EyeInstitute for providing the polyclonal antibodies; Dr. Paul Leeand Ms. Helen Kwong-Fu of the Jet Propulsion Laboratory,Pasadena, Calif., for helping with the operation of the image-analysis light microscopy system; Ms. Stephanie Beavon forher technical assistance; and Ms. Rosalinda Baca for assistingwith the preparation of this manuscript.

References1. Belch JJ, Chopra M, Hutchison S, Lorimer R, Sturrock RD,

Forbes CD, Smith WE: Free radical pathology in chronic arterialdisease. Free Radic BM Med 1989;6:375-378

2. Henriksen T, Mahoney EM, Steinberg D: Enhanced macrophagedegradation of low density lipoprotein previously incubated withcultured cndothclial cells: Recognition by receptors for acetylatedlow density lipoproteins. Proc Nad Acad Sa U S A 1981 ;78:6499-6503

3. Goldstein JL, Ho YK, Basu SK, Brown MS: Binding site onmacrophages that mediates uptake and degradation of acetylatedlow density lipoprotein, producing massive cholesterol deposition.Proc Nad Acad Sci U S A 1979;76:333-377

4. Morel DW, Hessler JR, Chisolm GM: Low density lipoproteincytotoxicity induced by free radical peroxidation of lipid. J LipidRes 1983;24:1070-1076

5. Evenson SA, Galdal KS, Nilsen E: LDL-induced cytotoxicity andits inhibition by anti-oxidant treatment in cultured human endo-tbelial cells and fibroblasts. Atherosclerosis 1983;49:23-30

6. Heinecke JW, Baker L, Rosen H, Chait A: Superoxide-mediatedmodification of low density lipoprotein by arterial smooth musclecells. J Clin Invest 1986;77:757-761

7. Quinn MT, Parthasarathy S, Fong LG, Steinberg D: Oxidativerymodified low density lipoproteins: A potential role in recruitmentand retention of monocyte/macrophages during atherogenesis.Proc Nad Acad Sci U S A 1987;84:2995-2998

8. Haberland ME, Fong D, Cheng L: Malondialdehyde-altered pro-tein occurs in atheroma of Watanabe Heritable Hypertipidemicrabbits. Science 1988;241:215-218

9. Palinski W, Yla-Herttuala S, Rosenfeld ME, Butler SW, SocherSA, Parthasarathy S, Curtiss LK, Witztum JL: Antisera andmonoclonal antibodies specific for epitopes generated duringoxidative modification of low density lipoprotein. Arteriosclerosis199O;10:325-335

10. Rosenfeld ME, Palinski W, Yla-Herttuala S, Butler S, Witztum JL:Distribution of oxidation specific lipid-protein adducts and apo-lipoprotein B in atherosclerotic lesions of varying severity fromWHHL rabbits. Arteriosclerosis 1990;10:336-349

11. Yla-Herttuala S, Palinski W, Rosenfeld ME, Parthasarathy S,Carew TE, Butler S, Witzrum JL, Steinberg D: Evidence for thepresence of oxidatively modified low density lipoprotein in athero-sclerotic lesions of rabbits and man. J Clin Invest 1989;84:1086-1095

12. BJornheden T, Bondjers G: Oxygen consumption in aortic tissuefrom rabbits with diet-induced atherosclerosis. Arteriosclerosis1987;7:238-247

13. Crawford DW, Kramsch DM: The oxygen environment of thearterial media in early rabbit hypertension. Exp Mol Padwl 1988;49:215-233

14. Seidel CL, Strong R: Metabolic characteristics of aorta fromspontaneously hypertensive and renal and deoxycorticosteroneacetate-salt hypertensive rats. Hypertension 1986;8:103-108

15. Schwartz SM, Ross R: Cellular proliferation in atherosclerosis andhypertension. Prog Cardiovasc Dis 1984;26J55-372

16. Davis conference: Oxygen radicals and human disease. Ann InternMed 1987;107:526-545

17. Forman HJ, Boveris A; Superoxide radical and hydrogen peroxidein mitochondria, in PTyor WA (ed): Free Radicals in Biology. NewYork, Academic Press, Inc, 1982, pp 65-91

18. Zweier JL, Kuppusamy P, Lutty GA: Measurement of endothelialceD free radical generation: Evidence for a central mechanism of freeradical injury in postischemic tissues. Proc Nad Acad Sci U S A1988;85:4046-4050

19. Janzen EG, Stanks HJ, Dubose CM, Poyez JL, McCay PB:Chemistry and biology of spin-trapping radicals associated withhalocarbon metabolism in vitro and in vivo. Environ Health Penpect1985;64:151-170

20. Del Boccio G, Lapenna D, Porreca E, PenneUi A, Salvini F,Feliciani P, Ricci G, Cuccrullo F: Aortic antioxidant defensemechanisms: Time related changes in cholesterol-fed rabbits.Atherosclerosis 199O;81:127-135

21. Kritchevsky D, Tepper SA, Vesselinovitch D, Wissler RW: Cho-lesterol vehicle in experimental atherosclerosis: Part II. (peanutoil). Atherosclerosis 1971;14:53-64

22. Sevanian A, McLeod LL: Cholesterol autoxidation in phospholipidmembrane bilayers. Lipids 1987^22:627-636

23. AUain CC, Poon LS, Chan CSG, Richmond W, Fu PC: Enzymaticdetermination of total serum cholesterol. Clin Chem 1974;20:470-475

24. Grant BT, Rothschild P: A device for estimating blood pressure inthe rabbit. / Physiol 1934;81:265-269

25. Walmsley JG, Granter SR: Changes in rabbit ear vascular resis-tance with aortic coarctation hypertension (abstract). Fed Proc1985;44:1645


Dow

nloaded from



26. Rao NA, Tboete LG, Delmage M, Sevanian A: Superoxide dismu-tase in ocular structures. Invest Ophthalmol Vis Sd 1985;26:1778-1781

27. Atalla LR, Fernandez MA, Rao NA; Immunohistochemical local-ization of catalase in ocular tissue. Curr Eye Res 1987;6:1181-1187

28. Atalla LR, Sevanian A, Rao NA; Immunohistochemical localiza-tion of glutathione peroxidase in ocular tissue. Curr Eye Res1988;7:1023-1027

29. Abei DM: Catalase in vitro, in Colorich SP, Kaplan N (eds):Methods in Enzymoiogy. New York, Academic Press, Inc, 1984, vol29, pp 121-126

30. Bligh EG, Dyer WT: A rapid method of total lipid extration andpurification. Can J Biochem Physiol 1959^7:911-917

31. Smith LL; Cholesterol auto-oxidation: 1981-1986. Chem Phys Lipid1987;44:87-125

32. Pryor WA: The role of free radical reactions in biological systems,in Pryor WA (ed): Free Radicals in Biology. New York, AcademicPress, Inc, 1976, pp 1-50

33. Freeman BA, Crapo JD: Free radicals and tissue injury. Lab Invest1982;47:412-426

34. Fridovich I: The biology of oxygen radicals. Science 1978;201:875-880

35. Rister M, Baehner RL: The alteration of superoxide dismutase,catalase, glutathione peroxidase, and NAD(P)H cytochrome creductase in guinea pig polymorphonuclear leukocytes and alveo-lar macrophages during hyperoxia. J Clin Invest 1976^8:1174-1184

36. Suttorp N, Toepfer W, Roka L: Antioxidant defense mechanism ofendothelial cells: Glutathione redox cycle versus catalase. Am JPhysiol 1986;251:C671-C680

37. Harlan JM, Levine JD, Callahan KS, Marker LA; Glutathioneredox cycle protects cultured endothelial cells against lysis byextracellularry generated hydrogen peroxide. / Clin Invest 1984;73:706-713

38. Huber H, Ledochowski M, Michlmayr G: The role of macrophagesas effector cells, in Schmalzl F, Hukn D, Schaefer HE (eds):Disorders of the Monocyte-Macrophage System. Berlin, SpringerVerlag, 1981, pp 39-48

39. Kagan HM, Milbury PE, Kramsch DM: A possible role for elastinligands in the proteoh/tic degradation of arterial elastic lamellae inthe rabbit. Ore Res 1979;44:95-103

40. Harland WA, Gilbert JD, Brooks CJW: Iipids of human ather-oma: Oxidised derivatives of cholesteryl linoleate. Biochim BiophysActa 1973316:378-385

41. Fogelman AM, Shechter I, Seager J, Hokom M, Child FS, EdwardsPA: Malondialdehyde alteration of low density lipoproteins leadsto cholesteryl ester accumulation in human raonocyte-macro-phages. Proc NatI Acad Set U S A 1980;77:2214-2218

42. Halliwell B: Lipid pcroxidation in vivo and in vitro in relation toatherosclerosis: Some fundamental questions. Fourth Cologne Ath-erosclerosis Conference. Basel, Switzerland, Birkhauser Verlag,1988, pp 223-231

43. Steinberg D, Parthasarathy S, Carew TE, Khoo JC, Witztum JL:Beyond cholesterol: Modification of low-density lipoprotein thatincreases its atherogenicity. N Engl J Med 1989^20:915-924

44. Bevan RD: An autoradiographic and pathological study of cellularproliferation in rabbit arteries correlated with an increase inarterial pressure. Blood Vessels 1976;13:100-128

45. Loeb AL, Mandel MG, Straw JA, Bean BL: Increased aortic DNAsynthesis precedes renal hypertension in rats: An obligatory step?Hypertension 1986;8:754-761

46. Meerson FZ, Kagan VE, Kozlov YP, Belkina LM, Arkhipenko YV:The role of lipid peroxidation in pathogenesis of ischemic damageand the antioxidant protection of the heart. Sane Res Cardioi1982;77:465-485

47. Manning AS, Hearse DJ: Reperfusion-induced arrhythmias:Mechanisms and prevention. J Mol Cell Cardioi 1984;16:497-518

48. Kramsch DM, Sharma RC, Crawford DW, Sevanian A, Chin HP:Arterial connective tissue, lipid peroxides and macrophages inshort-term hypertension and/or hyperlipidemia (abstract). Arterio-sclerosis 1989;9:748a

49. Heughan C, Niinikoski J, Hunt TK: Oxygen tension in lesions ofexperimental arteriosclerosis of rabbits. Atherosclerosis 1973;17:361-367

50. Cozzi PJ, Lyon RT, Davis HR, Sylora J, Glagov S, Zarins CK:Aortic wall metabolism in relation to susceptibility and resistanceto experimental atherosclerosis. J Vase Surg 1988;7:706-714

51. Camejo G: The interaction of Iipids and lipoproteins with theintercellular matrix of arterial tissue; Its possible role in athero-genesis. Adv Lipid Res 1982;19:l-53

52. Srinivasan SR, Yost C, Bhandaru RR, Radhakrishnamurthy B,Berenson GS: Lipoprotein-grycosaminogrycan interactions in aor-tas of rabbits fed atherogenic diet containing different fats. Ath-erosclerosis 1982;43:289-301

53. Steinbrecher UP, Parthasarathy S, Leake DS, Witztum JL, Stein-berg D: Modification of low density lipoprotein by endothelial cellsinvolves lipid peroxidation and degradation of low density lipopro-tein phospholipids. Proc Nati Acad Sci U S A 1984;81:3883-3887

54. Brown MS, Ho YK, Goldstein JL: The cholesteryl ester cycle inmacrophage foam cells. J Biol Chem 1980^55:9344-9352

55. Gerrity RG: The role of the monocyte in atherogenesis: I. Tran-sition of blood-borne monocytes into foam cells in fatty lesions. AmJPathol 1981;103:181-190

56. Gerrity RG: The role of the monocyte in atherogenesis: II.Migration of foam cells from atherosclerotic lesions. Am J Pathol1981;103:191-200


Dow

nloaded from


R C Sharma, D W Crawford, D M Kramsch, A Sevanian and Q Jiaoatherosclerotic arteries. Role of oxygen free radicals.

Immunolocalization of native antioxidant scavenger enzymes in early hypertensive and

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1992 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.12.4.4031992;12:403-415Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/12/4/403World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


Dow

nloaded from

http://atvb.ahajournals.org/content/12/4/403

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


Documents

Immunolocalization of Native Antioxidant Scavenger Enzymes ... · Immunolocalization of Native Antioxidant Scavenger Enzymes in Early Hypertensive and Atherosclerotic Arteries Role