ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 341, No. 2, May 15, pp. 287–294, 1997Article No. BB979975

The Mechanism of Apolipoprotein B-100 Thiol Depletionduring Oxidative Modification of Low-Density Lipoprotein

Eric Ferguson, Ravinder Jit Singh, Neil Hogg, and B. Kalyanaraman1

Biophysics Research Institute, Medical College of Wisconsin, Milwaukee, Wisconsin 53226

Received January 8, 1997, and in revised form February 24, 1997

onset of atherosclerosis (1–6). In vivo data show thatOxidation of low-density lipoprotein (LDL) is recog- a variety of antioxidant regimens inhibit atherogenesis

nized to be a key step in atherogenesis. Previous stud- in experimental animals (7–12). Under in vitro condi-ies show that LDL contains low-molecular-weight anti- tions, supplementation of LDL with phenolic antioxi-oxidants such as vitamin E, b-carotene, and ubiquinol, dants, such as butylated hydroxytoluene (BHT), inhib-which can retard oxidative modification. In this re- its dramatically Cu2/-induced modification of LDL toport, we have evaluated the antioxidant potential of an atherogenic form (13–15).apolipoprotein B-100 (apo-B) thiols during LDL oxida- Recently, peroxynitrite has been implicated as an intion. Both apo-B thiols and vitamin E were depleted vivo oxidant which may play a role in atherosclerosisconcomitantly during the lag phase of Cu2/-mediated (16–19). The finding by Beckman et al. that nitrotyro-LDL oxidation. The rate of thiol depletion was signifi- sine, a product of the reaction between peroxynitritecantly inhibited by the lipophilic spin trap N-tert-bu- and protein tyrosinyl residues, is present in humantyl-a-phenylnitrone (PBN) but not by the water-solu- atherosclerotic lesions supports the idea that peroxyni-ble spin trap a-(4-pyridyl-1-oxide)-N-tert-butylnitrone trite formation may have pathophysiological effects in(POBN). Blocking apo-B thiols with sulfhydryl modi-

vivo (20). It has been shown that rapid depletion offying agents increased the oxidizability of LDL. Asendogenous vitamin E occurs during peroxynitrite-in-with Cu2/, peroxynitrite also caused depletion of apo-duced oxidative modification of LDL (21) and that na-B thiols, and again thiol depletion was inhibited bytive LDL treated with low levels of peroxynitrite isPBN but not by POBN. A PBN/lipid-derived radical ad-more susceptible to Cu2/-mediated oxidation (22).duct was observed by the electron spin resonance tech-

LDL is a highly compartmentalized particle, con-nique during oxidation of LDL with peroxynitrite. Wesisting of a monolayer of phospholipids, a cholesterylconclude that apo-B thiol depletion is mediated byester core, and lipophilic antioxidants (e.g., vitamin E,lipid peroxidation, prior to the onset of the propaga-b-carotene, and ubiquinol) distributed throughout thetion phase of LDL oxidation. The implications of apo-

B thiols as intrinsic antioxidants of LDL are discussed. particle (23, 24). LDL also has a single molecule ofq 1997 Academic Press apolipoprotein B-100 (apo-B), containing a total of 25

Key Words: low-density lipoprotein; lipid peroxida- cysteinyl residues. Of these cysteinyl residues, 5–6 res-tion; peroxynitrite; protein thiols; electron spin reso- idues (corresponding to 10–12 nmol/mg) exist as thiolsnance. located in hydrophobic regions of LDL, and the remain-

der form disulfide linkages (25, 26). Thiols are knownscavengers of peroxyl radicals (27). Oxidants such as

There is compelling evidence that links oxidative hypochlorite are reported to cause apo-B thiol depletionmodification of low-density lipoprotein (LDL)2 to the (28); however, data on the mechanism of apo-B thiol

depletion during LDL oxidation do not exist. The mech-1 To whom correspondence should be addressed at Biophysics Re- anism of apo-B thiol depletion has been examined fullysearch Institute, Medical College of Wisconsin, 8701 Watertownin this report using lipophilic thiol blockers, spin traps,Plank Road, Milwaukee, WI 53226-0509. Fax: (414)-266-8515.

E-mail: balarama@post.its.mcw.edu. and spin labels (see Fig. 1 for structures). The results2 Abbreviations used: LDL, low-density lipoprotein; TBARS, thio-

barbituric acid-reactive substances; LOOH, lipid hydroperoxide;apo-B, apolipoprotein B-100; MAL-6, 4-malemido-tetramethyl-1-pip- electron spin resonance; PBN, N-tert-butyl-a-phenylnitrone; POBN,

a-(4-pyridyl-1-oxide)-N-tert-butylnitrone; PBS, phosphate-bufferederdinyloxy; DTT, dithiothreitol; NEM, N-ethylmaleimide; BHT, bu-tylated hydroxytoluene; REM, relative electrophoretic mobility; ESR, saline; DTPA, diethylenetriaminepentaacetic acid.

2870003-9861/97 $25.00Copyright q 1997 by Academic PressAll rights of reproduction in any form reserved.

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

288 FERGUSON ET AL.

FIG. 1. Spin traps and sulfhydryl modifying agents.

Modifications of apo-B cysteinyl residues. The cysteinyl residuesobtained from this study indicate that apo-B thiolof apo-B in LDL (2–3 mg/ml) were labeled with MAL-6 (1 mM) by adepletion occurs as a result of lipid peroxidation. Themodified technique of the method described by Singh et al. (26). Theimplications for apo-B thiols as antioxidants during reaction mixture was incubated for 12 h at 47C and then dialyzed

LDL oxidation are discussed. seven times against Chelex-treated PBS at 47C over a 24-h period.Vitamin E and lipid hydroperoxide (LOOH) concentrations weremeasured following dialysis. LDL was incubated with cystamine and

MATERIALS AND METHODS DTT using the same technique.Materials. 4-Maleimido-tetramethyl-1-piperidinyloxy (MAL-6) Thiol determinations. Protein thiol groups in LDL samples were

was purchased from Aldrich Chemical Co. (Milwaukee, WI). 5,5*- assayed using Ellman’s reagent (32). An extinction coefficient (e ÅDithio-bis-(2-nitrobenzoic acid) (Ellman’s reagent), cystamine, N- 11,000 M01 cm01) was determined from a standard curve generatedethylmaleimide (NEM), dithiothreitol (DTT), N-tert-butyl-a-phenyl- using reduced cysteine and was used for all calculations. High con-nitrone (PBN), a-(4-pyridyl-1-oxide)-N-tert-butylnitrone (POBN), centrations of LDL (minimum of 2 mg/ml) were required in order to2,5,6-tetramethyl-3-pyrroline-1-oxyl-carboxymide, potassium bro- attain sufficient sensitivity and reproducibility in these measure-mide, thiobarbituric acid, Folin and Ciocalteu’s phenol reagent, gua- ments. This is due to both the small concentrations of endogenousnidine hydrochloride, sodium dodecyl sulfate (SDS), ethylenedi- thiols contained in LDL (10–12 nmol/mg) and the low sensitivity ofaminetetraacetic acid (EDTA), BHT, and materials for phosphate- the Ellman’s reagent.buffered saline [PBS, sodium phosphate (25 mM), sodium chloride Conjugated diene measurement. LDL (125 mg/ml) was incubated(125 mM), pH 7.4] were obtained from Sigma Chemical Co. (St. Louis, with Cu2/ (20 mM) in PBS at 377C. Changes in absorbance at 234MO). Sodium nitrite, sodium tartrate, sodium hydroxide, cupric sul- nm were monitored continuously by following the formation of conju-fate, and diethylenetriaminepentaacetic acid (DTPA) were pur- gated dienes as described by Gieseg and Esterbauer (33).chased from Fisher Scientific (Itasca, IL). Chelex was obtained from

Thiobarbituric acid-reactive substances (TBARS) measurement.Bio-Rad Laboratories (Hercules, CA). High-performance liquid chro-Aliquots of LDL (5 mg/ml) were collected at various time intervalsmatography (HPLC)-grade hexane, chloroform, and methanol wereduring Cu2/-mediated oxidation and BHT (500 mM in ethanol) andobtained through Baxter Healthcare Corp. from the Burdick andDTPA (500 mM) were added in order to quench the reactions. TBARSJackson Division (McGraw Park, IL). Ethanol was purchased fromwere measured by incubating LDL (10–20 mg) with thiobarbituricQuantum Chemical Corp. (Tuscola, IL). Nitrogen and argon gasesacid (0.5% w/v in H2SO4, 50 mM) for 30 min at 1007C (30). Sampleswere supplied by AIRCO Gas (Murray Hill, IL).were centrifuged for 5 min, and the difference in absorbanceLDL was isolated from human plasma by sequential ultracentrifu-(A532 nm 0 A580 nm) was determined. Concentrations of TBARS weregation through a potassium bromide density gradient (29). LDL pro-calculated using an extinction coefficient of 150,000 M01 cm01 astein concentrations were measured by the Lowry assay (30). Thedescribed previously (27).freshly prepared LDL was stored under argon at 47C in PBS con-

taining EDTA (10 mM). Peroxynitrite was prepared as described pre- Vitamin E measurement. The concentrations of vitamin E weredetermined by reverse-phase HPLC as described previously (34).viously (31), and concentrations were determined spectrophotometri-

cally at 302 nm (e Å 1670 M01 cm01) (22, 31). Samples were extracted into heptane, dried under nitrogen, and re-

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

289MECHANISM OF APOLIPOPROTEIN B-100 THIOL DEPLETION

dissolved in methanol before injection onto a Partisil 10 ODS-3 re-verse-phase HPLC column. The mobile phase consisted of metha-nol:water (95:5) for 10 min, a linear gradient to 100% methanol overthe next 5 min, and an additional 5 min of 100% methanol. VitaminE was monitored by fluorescence (lex Å 275 nm and lem Å 320 nm).Compounds were quantified using a standard curve generated withknown concentrations of vitamin E.

LOOH measurement. LOOH concentrations of native LDL andof Cu2/-oxidized LDL were assayed by the iodometric method asdescribed previously (35).

Apo-B modification. Changes in the net negative charge of apo-Bduring LDL oxidation were measured by agarose gel electrophoresis(Paragon lipo gel electrophoresis kit, Beckman). Relative electropho-retic mobility (REM) was calculated as the ratio of the distance mi-grated by the samples and the distance migrated by native LDL (36).

Electron spin resonance (ESR) studies. ESR spin trapping wasperformed using PBN and POBN. PBN or POBN (50 mM) was prein-cubated with LDL (6.0 mg/ml) for 12 h at 47C. After addition ofperoxynitrite (4 mM), samples were transferred immediately to anaqueous flat cell, and ESR spectra were recorded using a Varian E-109 Century Series spectrometer. Alternatively, following peroxyni-trite addition, the lipid phase was isolated first by extraction withchloroform:methanol (2:1 v/v), dried under N2, and resolubilized in0.25 ml of N2-purged ethanol. The sample was transferred to anaqueous flat cell and spectra were recorded at room temperature. TheESR spectra were analyzed using Sumspec 92 (software developed atthe National Biomedical ESR Center, Milwaukee, WI). FIG. 2. The effect of MAL-6 treatment on Cu2/-mediated LDL

oxidation. LDL (125 mg/ml, native LDL or MAL-6-treated LDL) wasincubated with Cu2/ (20 mM) in PBS at 377C. The absorbance at

RESULTS 234 nm was monitored continuously. Data are representative of 12independent experiments.Apo-B thiol concentrations in fresh and aged LDL.

Freshly prepared LDL contained typically between 7and 9 nmol thiols per milligram of LDL protein. LDL Thiol content was measured from these LDL samples.preparations were stored under argon at 47C in PBS A time-dependent decrease in thiol concentration wascontaining the metal ion chelator EDTA (1 mM) in order observed. At 2 weeks, thiol concentration had de-to prevent transition metal-catalyzed LDL oxidation. creased by 5%, and at 6 weeks, concentrations had

fallen by 18%. In LDL from other preparations whichhad aged over 120 days, 100% thiol depletion was ob-served. For all experiments in this study, the LDL wasTABLE Iused immediately after isolation.

Effect of Apo-B Cysteinyl Residue Modification onEffect of apo-B thiol modification on LDL oxidizabil-Conjugated Diene Formation during

ity. A decrease in LDL thiol concentration was ob-Cu2/-Mediated Oxidation of LDLserved after a 12-h incubation at 47C with either MAL-6 or cystamine, known sulfhydryl blocking agents (Ta-Apo-B Thiolble I). Conversely, a similar incubation of LDL withcysteine Lag time (% of concentration Vitamin E

modification native LDL) (nmol/mg) (nmol/mg) DTT, a disulfide reducing agent, increased LDL thiolconcentrations (Table I). Neither LDL vitamin E con-

Control 100 { 6.66 8.99 { 1.51 8.07 { 0.97 tent (Table I) nor LOOH content (data not shown) wasMAL-6 65.24 { 16.45* 5.83 { 0.69* 8.22 { 1.14affected by the thiol modification procedure, indicatingCystamine 65.41 { 10.92* 4.05 { 0.47* 8.57 { 1.10that the LDL was not significantly oxidized during thisDTT 117.40 { 10.31* 12.81 { 0.90* 7.98 { 0.99process.

Note. LDL was treated with MAL-6 (1 mM), NEM (1 mM), cysta- The formation of conjugated dienes during Cu2/-me-mine (1 mM), or DTT (1 mM). Following a 12-h incubation time at diated oxidation of LDL was measured by an increase47C, LDL was dialyzed exhaustively in Chelex 100-treated, EDTA-

in absorption at 234 nm (37). LDL in which apo-B thiolsfree PBS (pH 7.4) for 24 h, after which both thiols and vitamin Ewere blocked with MAL-6 oxidized more rapidly thanwere measured. LDL was oxidized at 377C with Cu2/ (20 mM Cu2/,

125 mg/ml LDL, in PBS). Absorbance at 234 nm was monitored con- freshly prepared native LDL as shown in Fig. 2. Cysta-tinuously. Control lag time was 55.03{ 3.66 min. Data are expressed mine treatment also caused a decrease in the lag timeas a percentage of the lag time preceding the rapid chain propagation of Cu2/-mediated oxidation with respect to native LDL.phase of LDL oxidation. Statistical analysis was done using a paired

In contrast, DTT, a disulfide reducing agent, increasedt test (n Å 12 for each cysteine modification).* Significantly different than control (P õ 0.05). the lag time (Table I).

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

290 FERGUSON ET AL.

Figure 3 shows the relationship between apo-B thiolconcentrations and lipid peroxidation. LDL concentra-tions in this experiment were higher than in otherassays in order to measure thiol concentrations withsufficient sensitivity. After a lag time of approximately75 min, a rapid increase in TBARS and LOOHs wasobserved. During this 75-min lag time, apo-B thiolswere depleted with a half-life of approximately 20 min.TBARS, LOOHs, and apo-B thiol concentrations re-mained unchanged in LDL that was incubated at 377Cin the absence of Cu2/ (data not shown).

In native LDL, the kinetics of vitamin E depletionduring Cu2/-mediated oxidation were followed (Fig. 4).LDL (125 mg/ml) was incubated with Cu2/ for varioustime intervals. A time-dependent decrease in vitaminE levels was observed. After 20 min of oxidation, vita-min E concentrations dropped by approximately 30%in native LDL, as shown in Fig. 4. Blocking apo-B thiolswith MAL-6 enhanced the rate of vitamin E depletion FIG. 4. Kinetics of vitamin E loss during Cu2/-mediated oxida-

tion of LDL. LDL (125 mg/ml) of either native LDL (l) or MAL-6-during Cu2/-mediated oxidation (Fig. 4). In MAL-6-treated LDL (s) was incubated with Cu2/ (5 mM) in PBS at 377C.treated LDL, vitamin E decreased by more than 50%Aliquots of LDL were collected at various time intervals andduring the first 5 min of oxidation, and by 12 min, quenched with BHT (500 mM) and DTPA (500 mM), and vitamin E

vitamin E was totally depleted (Fig. 4). These results concentrations in LDL were measured by HPLC. Data points repre-suggest that vitamin E competes with protein thiols sent means { SE (n Å 3).for the scavenging of peroxyl radicals.

It has been reported previously that PBN but notthe trapping of LDL lipid-derived radicals (38, 39). ToPOBN gives rise to an immobilized ESR signal frominvestigate the possibility that apo-B thiol depletionoccurs by a radical mechanism, LDL (2.5 mg/ml) waspreincubated with the lipophilic PBN (50 mM), the hy-drophilic POBN (50 mM), or PBS alone for 12 h at 47C.Subsequently, the LDL was subjected to Cu2/-medi-ated oxidation, and at selected time intervals, apo-Bthiol concentration was measured. The rate of apo-Bthiol depletion was inhibited in PBN-treated LDL butnot in POBN-treated LDL as shown in Fig. 5. Theseresults indicate that lipid-derived radicals are involvedin the mechanism of apo-B thiol depletion during Cu2/-mediated oxidation of LDL.

Apo-B thiol depletion during peroxynitrite-mediatedoxidation of LDL. It has been reported previously thatperoxynitrite is able to oxidatively modify LDL re-sulting in an increase in electrophoretic mobility (18).Figure 6 shows the increase in REM as a function ofperoxynitrite concentration. A parallel depletion of apo-B thiols was also observed (Fig. 6).

The differential hydrophobicity of PBN and POBNwas used to determine if peroxynitrite-dependent thiol

FIG. 3. Kinetics of apo-B thiol depletion, TBARS formation, and depletion was mediated by lipid peroxidation. LDL (2.5LOOH formation during Cu2/-mediated oxidation of LDL. (Top) na-mg/ml) was incubated with the lipophilic spin trap PBNtive LDL (5 mg/ml in PBS, pH 7.4, at 377C) was incubated with Cu2/

(50 mM), the water-soluble spin trap POBN (50 mM),(150 mM). Aliquots of LDL were collected at various time intervals,and both the TBARS content (j) and the thiol concentration (l) were or PBS alone for 12 h. Subsequently, increasing concen-measured. TBARS were measured using an extinction coefficient of trations of peroxynitrite were added to aliquots of LDL,150,000 M01 cm01. Thiols were measured using Ellman’s reagent. and thiol concentrations were measured. As shown in(Bottom) Same conditions as (top), except that aliquots of LDL were

Fig. 7, PBN treatment but not POBN treatment sig-collected at various time intervals and LOOHs were measured byiodometric assay (m). Data points represent means { SE (n Å 3). nificantly inhibited peroxynitrite-dependent thiol de-

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

291MECHANISM OF APOLIPOPROTEIN B-100 THIOL DEPLETION

quots of the LDL/POBN incubation, and spectra wererecorded immediately (Fig. 8B). This spectrum is com-posed of three species with the following hyperfine cou-pling constants: aN Å 16.3 G; aN Å 16.3 G, abH Å 20.6G; aN Å 15.7 G. The relative concentrations of thesethree species are 36, 31, and 33%, respectively. Thespectrum is characteristic of a freely tumbling ni-troxide in aqueous solution. The spectral intensity wasgreatly attenuated when ‘‘decomposed’’ peroxynitritewas added to the POBN/LDL mixture (Fig. 8B, bottom).ESR data indicate that PBN/lipid-derived adducts arelocalized in the lipid phase, and the POBN adducts areformed in the aqueous phase.

DISCUSSION

Locations of apo-B thiols. Previously, we have usedESR spin labeling and paramagnetic relaxing agentsto characterize the environment of apo-B thiols withinLDL (26). These studies show unequivocally that apo-B thiols lie in hydrophobic environments. This leavesthe possibilities that apo-B thiols are buried in the lipid

FIG. 5. Effect of PBN and POBN pretreatment on thiol depletion or hydrophobic protein phases of LDL. We hypothesizeduring Cu2/-mediated oxidation of LDL. LDL (2.5 mg/ml) was incu- that apo-B thiols are situated such that lipid-radicalbated with PBN (50 mM) (l), POBN (50 mM) (X), or PBS (s) for 18

scavenging reactions involving these thiols occur dur-h. The samples were subsequently oxidized with Cu2/ (150 mM) ating oxidative modification of LDL.377C. Thiol concentrations were measured using Ellman’s reagent.

Data points represent means { SE (n Å 3). To examine the effect of apo-B thiols on the oxidiza-bility of LDL, thiols were blocked with sulfhydryl modi-fying agents (see Fig. 1 for structures). These modifica-tions increased the oxidizability as measured by thepletion. PBN and POBN, however, had identical effects

on thiol depletion in bovine serum albumin. EitherPBN (50 mM) or POBN (50 mM) was incubated withBSA (1 mg/ml) in PBS for 12 h. After addition of peroxy-nitrite (500 mM), protein thiol concentrations had de-creased by 48.7 { 8.5% in the PBN-treated group and49.9 { 5.79% in the POBN-treated group (not signifi-cantly different, P ú 0.2). These results suggest thatlipid peroxidation mediates peroxynitrite-dependentthiol depletion in LDL.

ESR spin trapping of radicals formed during peroxy-nitrite-mediated oxidation of LDL. Addition of perox-ynitrite to a mixture of LDL and PBN resulted in anESR spectrum (Fig. 8A, top). This spectrum consists ofan immobilized and mobile component characteristicof a rotationally restricted long-chain nitroxide in lipidmembrane. This spectrum has been previously attrib-uted to a PBN/LDL lipid-derived radical adduct (37).A more resolved spectrum was obtained after lipid ex-traction of these aliquots and solubilization in ethanol,as described previously (38, 39). Decomposed peroxyni-trite did not form any spin adduct, as shown in Fig.

FIG 6. Effect of peroxynitrite on apo-B thiol concentration and on8A (bottom). This indicates that the PBN-derived spinelectrophoretic mobility. LDL (2–3 mg/ml) was oxidized with varyingadduct did not result from any contaminants (e.g., ni-concentrations of peroxynitrite at 257C in PBS. Thiol concentrationstrite, H2O2) formed during synthesis of peroxynitrite. (s) were measured by Ellman’s reagent, and relative electrophoretic

LDL (8 mg/ml) was incubated for 18 h with POBN mobility (l) was measured by agarose gel electrophoresis. Datapoints represent means { SE (n Å 3).(50 mM). Peroxynitrite (4 mM) was added to 200-ml ali-

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

292 FERGUSON ET AL.

mechanisms occurring in the hydrophobic regions ofLDL, as illustrated below (Eqs. [10] and [11]).

Apo-B-SH/ L•/OLOO•r Apo-B-S•/ LH/LOOH [10]

Apo-B-S•r oxidation products [11]

Radical-mediated apo-B thiol depletion by peroxyni-trite. As shown in Fig. 6, apo-B thiols were depletedby peroxynitrite in a concentration-dependent manner.In addition, changes in electrophoretic mobility, whichprovide an additional means of assessing oxidativemodifications of LDL, were also examined. In agree-ment with previously reported data (18), peroxynitritealso caused a concentration-dependent increase inREM of LDL (Fig. 6). Haberland et al. suggested thatcharge modifications of apo-B are an important conse-quence of reactions between apo-B lysinyl residues andshort-chain aldehydes (3, 41, 42). Aldehydes such asmalondialdehyde are generated during the lipid peroxi-dation process (1, 43). It has been suggested by Grahamet al. that peroxynitrite may modify apo-B lysinyl resi-dues indirectly via lipid peroxidation, thereby increas-ing the net negative charge of LDL (18). This type ofmodification would account for the peroxynitrite-de-pendent increase in electrophoretic mobility.

FIG. 7. Effect of pretreatment with PBN and POBN on thiol To assess further the possibility that peroxynitrite-depletion during peroxynitrite-mediated LDL oxidation. (A) LDL (2.5 dependent depletion of apo-B thiols occurs as a conse-mg/ml) was treated without (s) or with PBN (50 mM, l) and incu-

quence of lipid peroxidation, we again used the twobated for 18 h after which peroxynitrite was added. The samples werethen immediately assayed for thiol concentrations using Ellman’s spin traps PBN and POBN, which have differences inreagent. (B) As in (A) except that POBN was used instead of PBN. lipophilicity (see Fig. 1 for structures). Apo-B thiolsData points represent means { SE (n Å 3). were spared during peroxynitrite-mediated oxidation

when LDL was preincubated with the lipophilic PBN,as shown in Fig. 7A. POBN, however, had no effect (Fig.

decreased lag time for the formation of conjugated 7B). These data suggest strongly that the mechanism ofdienes during Cu2/-mediated oxidation (Table I and peroxynitrite-mediated depletion of apo-B thiols is aFig. 2). Treatment with the disulfide reducing agent consequence of lipid oxidation (Eq. [12]).DTT increased thiol concentrations to maximum values

ONOO0 / LH / H/r L• / •NO2 / H2O [12]which are reported to exist in native LDL. DTT treat-

ment also increased the lag time during Cu2/-mediated Further evidence in support of this mechanism wasoxidation (Table I). These results demonstrate that pro- provided by ESR studies using the lipophilic spin traptein thiols protect LDL against Cu2/-mediated lipid PBN (Fig. 8A). As shown in Fig. 8A, the ESR spectrumperoxidation. of the PBN-derived spin adduct was characteristic of

The mechanism by which apo-B thiols protect trapping long-chain lipid-derived radicals in LDL lipid.against Cu2/-mediated oxidation, a lipid hydroperox-

Implications of the antioxidant effect of apo-B thiols.ide-dependent process (40), was investigated by com-The oxidation of LDL is implicated in atherogenesis,paring the effects of lipophilic and hydrophilic spinalthough the exact mechanisms by which this processtraps PBN and POBN, respectively. The lipophilic PBNoccurs in vivo remain elusive (1–6). Thiols have pre-has been shown previously to inhibit lipid peroxidationviously been implicated in radical-scavenging reactions(39) by stopping propagating reactions in the lipid(27). Mass spectrometric characterization of Cu2/-oxi-phase of LDL (Eqs. [8] and [9]).dized apo-B reported by Burlet et al. shows that a spe-

PBN / L•r PBN/L• [8] cific cysteic acid residue is formed during oxidation

(44). It is suggested that this modification may signifi-PBN / OLOO•r PBN/OLOO• [9] cantly affect the physical properties of LDL, for exam-

ple, the propensity of LDL to form intermolecular inter-PBN retarded thiol depletion, whereas the hydrophilicspin trap POBN had no effect (Fig. 5). This implies actions via disulfide bridges.

In general, LDL, which is supplemented with lipid-that apo-B thiol depletion occurs as a result of radical

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

293MECHANISM OF APOLIPOPROTEIN B-100 THIOL DEPLETION

FIG. 8. (A) ESR spectra of PBN spin adducts in peroxynitrite-oxidized LDL. (Top) LDL (6 mg/ml) was incubated with PBN (50 mM) inPBS for 18 h after which peroxynitrite (4 mM) was added to the incubation. Spectrometer conditions: modulation amplitude, 3.2 G; microwavepower, 20 mW; scan range, 100 G; time constant, 0.25 s; scan time, 2 min. (Bottom) Same as (top) but peroxynitrite was decayed at 257Cbefore addition to the PBN/LDL mixture. Arrows a and b indicate low-field line positions of the immobilized and the mobile components ofthe LDL/PBN spin adduct, respectively. (B) ESR spectra of POBN spin adducts in peroxynitrite-oxidized LDL. (Top) LDL (6 mg/ml) wasincubated with POBN (50 mM) in PBS for 18 h after which peroxynitrite (4 mM) was added to the incubation. Spectrometer conditions:modulation amplitude, 1.0 G; microwave power, 5 mW; scan range, 100 G; time constant, 0.250 s; scan time, 2 min. (Bottom) Same as (top)but peroxynitrite was decayed before addition to the LDL/POBN mixture. Dashed line represents a computer-simulated spectrum of threespecies with hyperfine coupling constants and relative concentrations as referenced in the text.

soluble chain-breaking antioxidants such as vitamin of LDL. This hypothesis would help to explain the lackof consistent vitamin E data which have been reportedE, butylated hydroxytoluene, or probucol, is markedly

resistant to oxidation (for a review, see Ref. 24). Stein- in the literature (46). In conclusion, we have shownthat endogenous apo-B thiol depletion occurs as a re-brecher et al. show that exogenous vitamin E at high

concentrations (100 mM, 900 nmol/mg LDL protein) in sult of lipid peroxidation. Data suggest that endoge-nous thiols are an additional key component of the totalthe culture medium of endothelial cells prevents oxida-

tive modifications of LDL (45). Vitamin E has been antioxidant capacity of LDL.shown to protect LDL from oxidative modification invitro; however, the correlation between vitamin E con- ACKNOWLEDGMENTStent and susceptibility of LDL to oxidation has not al-

This research has been supported by NIH Grants HL45048 andways been consistent (46). We have observed thiol con- RR01008. The authors thank Dr. Jimmy B. Feix for advice on apo-centration to be low in aged LDL. In addition, we have B labeling, Mrs. Karen Hyde for assistance in preparation of theobserved that thiols can significantly influence the rate figures, and Dr. Sampath Parthasarathy for his continued interest

and valuable suggestions in this work.of vitamin E consumption during Cu2/-mediated LDLoxidation (Fig. 4). It is of interest to consider that theconcentration of endogenous vitamin E contained in REFERENCESLDL is similar to that of apo-B thiols. A maximum of

1. Steinberg, D., Parthasarathy, S., Carew, T., Khoo, J., and Witz-6 to 10 vitamin E molecules are incorporated into each tum, J. (1989) N. Engl. J. Med. 320, 915–924.LDL molecule (47). We propose that apo-B thiols pos- 2. Witztum, J., and Steinberg, D. (1991) J. Clin. Invest. 88, 1785–sess lipid-derived radical-scavenging abilities similar 1792.to vitamin E and therefore should be considered to be 3. Haberland, M., Fong, D., and Cheng, L. (1988) Science 241, 215–

218.an additional variable in the total antioxidant capacity

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca

294 FERGUSON ET AL.

4. Palinski, W., Rosenfeld, M., Yla-Herttuala, S., Gurtner, G., 24. Esterbauer, H., Gebicki, H., Puhl, H., and Jurgens, G. (1990)Free Radical Biol. Med. 13, 341–390.Socher, S., Butler, S., Parthasarathy, S., Carew, T., Steinberg,

D., and Witztum, J. (1989) Proc. Natl. Acad. Sci. USA 86, 1372– 25. Yang, C. Y., Kim, T. W., Weng, S. A., Lee, B., Yang, M., and1376. Gotto, A., Jr. (1990) Proc. Natl. Acad. Sci. USA 87, 5223–5527.

5. Yla-Herttuala, S., Palinski, W., Rosenfeld, M., Parthasarathy, 26. Singh, R., Feix, J., Mchaourab, H., Hogg, N., and Kalyanaraman,S., Carew, T., Butler, S., Witztum, J., and Steinberg, D. (1989) B. (1995) Arch. Biochem. Biophys. 320, 155–161.J. Clin. Invest. 84, 1086–1095. 27. Lim, Y. S., Cha, M. K., Kim, H. K., Uhm, T. B., Park, J. W., Kim,

K., and Kim, I. H. (1993) Biochem. Biophys. Res. Commun. 192,6. Parthasarathy, S. (1994) Modified Lipoproteins in the Pathogen-273–280.esis of Atherosclerosis, Landes, Austin, TX.

28. Hazell, L. J., and Stocker, R. (1993) Biochem. J. 290, 165–172.7. Keaney, J., Xu, A., Cunningham, D., Jackson, T., Frei, B., and29. Hatch, F. T., and Lees, R. S (1968) Adv. Lipid Res. 6, 2–68.Vita, J. (1995) J. Clin. Invest. 95, 2520–2529.30. Lowry, O. H., Rosenbrough, N. J., Farr, A. L., and Randall, R. J.8. Carew, T., Schwenke, D., and Steinberg, D. (1987) Proc. Natl.

(1951) J. Biol. Chem. 193, 265–275.Acad. Sci. USA 84, 7725–7729.31. Swain, J. A., Darley-Usmar, V., and Gutteridge, J. M. C. (1994)9. Supari, F., Ungerner, T., Harrison, D. G., and Williams, J. K.

FEBS Lett. 342, 49–52.(1995) Circulation 91, 1123–1128.32. Riddles, P. W., Blakeley, R. L., and Zerner, B. (1979) Anal. Bio-10. Bjorkhem, I., Henriksson-Freyschuss, A., Breuer, O., Dicfalusy,

chem. 9, 75–81.U., Berglund, L., and Henriksson, P. (1991) Arterioscler. Thromb.33. Gieseg, S. P., and Esterbauer, H. (1994) FEBS Lett. 343, 188–11, 15–22.

194.11. Mao, S., Yates, M., Parker, R., Chi, E., and Jackson, R. (1991)34. Hogg, N., Joseph, J., and Kalyanaraman, B. (1995) Arch. Bio-Arterioscler. Thromb. 11, 1266–1275.

chem. Biophys. 314, 153–158.12. Reaven, P. D., and Witztum, J. L. (1996) Annu. Rev. Nutr. 16,

35. Girotti, A. W. (1985) J. Free Radical Biol. Med. 1, 87–95.51–71.36. Hogg, N., Darley-Usmar, V., Wilson, M. T., and Moncada, S.

13. Reaven, P. D., Ferguson, E., Navab, M., and Powell, F. L. (1994) (1993) FEBS Lett. 326, 199–203.Arterioscler. Thromb. 14, 1163–1169.

37. Esterbauer, H., Striegl, G., Puhl, H., and Rotheneder, M. (1989)14. Kaneko, T., Kaji, K., and Matsuo, M. (1994) Free Radical Biol. Free Radical Res. Commun. 6, 67–75.

Med. 16, 405–409. 38. Thomas, C. E., Ku, G., and Kalyanaraman, B. (1994) J. Lipid15. Mao, S. J. T., Yates, M. T., and Jackson, R. L. (1994) Methods Res. 35, 610–619.

Enzymol. 234, 505–513. 39. Kalyanaraman, B., Joseph, J., and Parthasarathy, S. (1991)FEBS Lett. 280, 17–20.16. Beckman, J., Beckman, T., Chen, J., Marshall, P., and Freeman,

B. (1990) Proc. Natl. Acad. Sci. USA 87, 1620–1624. 40. Thomas, C. E., and Jackson, R. L. (1991) J. Pharmacol. Exp.Ther. 256, 1182–1188.17. Darley-Usmar, V., Hogg, N., O’Leary, V., Wilson, M., and Mon-

41. Haberland, M. E., Olch, C. L., and Folgelman, A. M. (1984) J.cada, S. (1992) Free Radical Res. Commun. 17, 9–20.Biol. Chem. 259, 11305–11311.18. Graham, A., Hogg, N., Kalyanaraman, B., O’Leary, V., Darley-

42. Haberland, M. E., Fogelman, A. M., and Edwards, P. A. (1982)Usmar, V., and Moncada, S. (1993) FEBS Lett. 330, 181–185.Proc. Natl. Acad. Sci. USA 79, 1712–1716.

19. Moore, K., Darley-Usmar, V., Morrow, J., and Roberts, L. (1995)43. Steinbrecher, U. P., Zhang, H., and Lougheed, M. (1990) FreeCirc. Res. 77, 335–341.

Radical Biol. Med. 9, 155–168.20. Beckman, J., Yao, Y., Anderson, P., Chen, J., Accavitti, M., Tar- 44. Burlet, O., Chao-Yuh, Y., Guyton, J. R., and Gaskell, S. J. (1995)

pey, M., and White, R. (1994) Biol. Chem. Hoppe Seyler 375, 81– Am. Soc. Mass Spectrom. 6, 242–247.88.

45. Steinbrecher, U. P., Parthasarathy, S., Leake, D. S., Witztum,21. Hogg, N., Darley-Usmar, V., Wilson, M. T., and Moncada, S. J. L., and Steinberg, D. (1984) Proc. Natl. Acad. Sci. USA 81,

(1993) FEBS Lett. 326, 199–203. 3383–3387.22. Hogg, N., Darley-Usmar, V., Graham, A., and Moncada, S. (1993) 46. Dieber-Rotheneder, M., Puhl, H., Waeg, G., Striegl, G., and Es-

Biochem. Soc. Trans. 21, 358–361. terbauer, H. (1991) J. Lipid Res. 32, 1325–1332.47. Esterbauer, H., Dieber-Rotheneder, M., Waeg, G., Striegl, G.,23. Esterbauer, J., Dieber-Rothender, M., Waeg, G., Striegl, G., and

Jurgens, G. (1990) Chem. Res. Toxicol. 3, 77–92. and Jurgens, G. (1990) Chem. Res. Toxicol. 3, 77–91.

AID ABB 9975 / 6b33$$$481 04-14-97 16:15:54 arca