Tyrosine Modification Is Not Required for Myeloperoxidase-induced Loss of Apolipoprotein A-I Functional Activities*

Apolipoprotein A-I (apoAI), the major protein of high density lipoprotein, plays an important role in reverse cholesterol transport via its activity as an ABCA1-dependent acceptor of cellular cholesterol. We reported recently that myeloperoxidase (MPO) modification of apoAI inhibits its ABCA1-dependent cholesterol acceptor activity (Zheng, L., Nukuna, B., Brennan, M. L., Sun, M., Goormastic, M., Settle, M., Schmitt, D., Fu, X., Thomson, L., Fox, P. L., Ischiropoulos, H., Smith, J. D., Kinter, M., and Hazen, S. L. (2004) J. Clin. Invest. 114, 529–541). We also reported that MPO-mediated chlorination preferentially modifies two of the seven tyrosines in apoAI, and loss of parent peptides containing these residues dose-dependently correlates with loss in ABCA1-mediated cholesterol acceptor activity (Zheng, L., Settle, M., Brubaker, G., Schmitt, D., Hazen, S. L., Smith, J. D., and Kinter, M. (2005) J. Biol. Chem. 280, 38–47). To determine whether oxidative modification of apoA-I tyrosine residues was responsible for the MPO-mediated inactivation of cholesterol acceptor activity, we made recombinant apoAI with site-specific substitutions of all seven tyrosine residues to phenylalanine. ApoAI and the tyrosine-free apoAI were equally susceptible to dose-dependent MPO-mediated loss of ABCA1-dependent cholesterol acceptor activity, as well as lipid binding activity. MPO modification altered the migration of apoAI on SDS gels and decreased its α-helix content. MPO-induced modification also targeted apoAI tryptophan and lysine residues. Specifically, we detected apoAI tryptophan oxidation to mono- and dihydroxytryptophan and apoAI lysine modification to chlorolysine and 2-aminoadipic acid. Thus, tyrosine modification of apoAI is not required for its MPO-mediated inhibition of cholesterol acceptor activity.

The levels of high density lipoprotein (HDL) 2 and its major protein component, apolipoprotein AI (apoAI), are inversely correlated to the risk of coronary artery disease in developed countries (1). Their protective effects may be mediated by several activities, including their ability to remove excess cholesterol from peripheral tissues for delivery to the liver via the reverse cholesterol transport pathway. Cellular lipid efflux to apoAI, as well as HDL biogenesis, is mediated by the transporter ABCA1, which is defective in Tangier disease (2)(3)(4)(5). Although apoAI may have additional anti-atherogenic properties, recent studies in the mouse have proved that apoAI can stimulate the removal of the preloaded cholesterol from macrophages and the appearance of this cholesterol in the feces (6). ApoAI is a 243-residue polypeptide containing an N-terminal domain followed by ten 11-or 22-residue amphipathic ␣-helical domains (7). ApoAI adopts two distinct physiological conformations, lipid-free and lipid-bound. In the absence of lipid, apoAI is thought to exist as a helical bundle or hairpin that is stabilized by hydrophobic helix-helix interactions (8 -12). These helical bundles undergo a conformational change during lipid association that results in the substitution of helix-helix interactions for helix-lipid interactions (13).
Myeloperoxidase (MPO) is a neutrophil and monocyte/macrophage enzyme that utilizes hydrogen peroxide and chloride to generate the chlorinating oxidant HOCl to kill pathogens (14). MPO can also utilize hydrogen peroxide and nitrite to generate nitrating oxidants capable of initiating lipid peroxidation and protein nitration (15). MPO can promote both protein nitration and lipid peroxidation in vivo (16 -18), and the blood levels of MPO and nitrotyrosine are predictors of an increased risk of cardiovascular disease (19 -21). We recently described that plasma and atherosclerotic lesion-derived apoAI are selective targets for MPO modification resulting in the appearance of specific nitro-and chlorotyrosine residues, and that this modification dose-dependently decreased the cholesterol acceptor and lipid binding activities of apoAI (22,23). Mass spectroscopy was used to map the sites of MPO-mediated apoAI nitration and chlorination in vitro and on apoAI isolated from human atheroma. After in vitro modification, we could detect MPOmediated chlorination of four out of the seven tyrosines, in order of preference at position 192 Ͼ 166 Ͼ Ͼ 29 -236 (23). The preference for tyrosine 192 modification by MPO and the loss of cholesterol acceptor activity of apoAI by MPO-mediated chlorination were independently confirmed (24).
Although apoAI tyrosine chlorination was associated with the MPOmediated loss of the cholesterol acceptor activity of apoAI, tyrosine chlorination might only serve as a molecular fingerprint for apoAI modification by MPO, and not the cause of apolipoprotein loss of cholesterol efflux function, as other changes in apoAI have been found after oxidative modification (25)(26)(27)(28)(29)(30). To address the specific role of tyrosine modification in MPO-dependent inactivation of apoAI, we used site-directed mutagenesis to create a tyrosine-free recombinant human apoAI. WefoundthatexposuretotheMPO/H 2 O 2 /Cl Ϫ systeminhibitedABCA1dependent cholesterol acceptor activity to an equal extent in the tyrosine-containing and tyrosine-free apoAI acceptors. MPO-mediated modification decreased the ␣-helix content of apoAI, a physical feature related to lipid binding activity of apoAI (13). ApoAI tryptophan fluo-rescence and lysine content were decreased by MPO modification in an H 2 O 2 dose-dependent manner. Tryptophan and lysine oxidation were confirmed by mass spectrometry. Thus, modification of residues other than tyrosine are responsible for MPO-induced loss of the cholesterol acceptor activity of apoAI.

Generation and Isolation of Recombinant Human ApoAI and ApoAI
Mutants-cDNA encoding His 6 -tagged recombinant human apoAI (rh-apoAI) in the pET-20b bacterial expression vector was gift from Dr. Oda (see Ref. 31). Point mutations of tyrosine to phenylalanine in rh-apoA-I cDNA at positions 18,29,100,115,166,192, and 236 (rh7YF-apoAI) were made by using Quick-Change mutagenesis kit (Stratagene). All the mutations were confirmed by DNA sequencing. The plasmids were transformed into Escherichia coli strain BL21(DE-3) pLysS, and apoAI expression was induced with 0.5 mM isopropyl ␤-Dthiogalactopyranoside by overnight incubation at room temperature. The cell pellet was dissolved in B-PER lysis buffer (Pierce) followed by centrifugation to sediment cell debris. The supernatant was diluted into phosphate-buffered saline containing 3 M guanidine HCl, pH 7.0, loaded onto a nickel-chelating histidine-binding resin column, and then specifically eluted with imidazole as described previously (31). Recombinant apoAI was dialyzed extensively in 60 mM sodium phosphate buffer, pH 7.0, containing 100 M diethylenetriamine pentaacetic acid. ApoAI samples were analyzed by electrophoresis on a 14% SDS-polyacrylamide gel under reducing conditions. The gels were either stained for protein with 0.25% Coomassie Blue or transferred to a PVDF membrane and probed with goat anti-human apoA1 (1:1000, Diasorin, Saco, ME) in the presence of casein blocker (Pierce). Detection was performed with a rabbit anti-goat antibody linked to horseradish peroxidase (1:10,000) followed by enhanced chemiluminescence.
MPO Modification of ApoAI-The MPO-mediated modification reactions were carried out with 3.5-7 M apoAI in 60 mM phosphate buffer, pH 7.0, containing 100 M diethylenetriamine pentaacetic acid, 57 nM purified human MPO, prepared as described previously (22), and 100 mM sodium chloride. The MPO reactions were initiated by adding hydrogen peroxide at varying concentrations (0 -150 M) in 4 aliquots at 15-min intervals at 37°C and continuing incubation for a total of 90 min. When indicated, 2 mM L-methionine was added to quench the generated HOCl. These reaction conditions include concentrations of MPO, chloride, and hydrogen peroxide that range from physiologic to pathologic. The modified protein was stored at Ϫ20°C, unless indicated otherwise.
Determination of Cholesterol Acceptor Activity-Cholesterol efflux experiments were performed according to established procedures (32). Briefly, RAW264.7 cells in 24-well dishes were cholesterol-loaded and labeled overnight in 0.5 ml of DGGB (DMEM supplemented with 50 mM glucose, 2 mM glutamine, and 0.2% bovine serum albumin), containing [ 3 H]cholesterol-labeled acetylated low density lipoprotein. The day after labeling, the cells were washed three times in DMEM containing 0.2% bovine serum albumin and incubated with 0.5 ml of DMEM with or without 0.3 mM 8-Br-cAMP for 16 h. The following day, 5 g/ml apoAI in 0.5 ml of DMEM with or without 8-Br-cAMP was added to each well. After a 4-h incubation at 37°C, the medium was removed and centrifuged, and 100 l of medium was counted for radioactivity as a measure of the cholesterol released into the media. The respective cells from each well were extracted with hexane/isopropyl alcohol (3:2, v/v), and the radioactivity was determined as a measure of the cholesterol retained in the cell. The percent cholesterol efflux was calculated as the radioactivity in the medium divided by the total radioactivity (medium radioactivity plus cell radioactivity).
ApoAI Lipid Binding Activity-An LDL aggregation assay was used to test apoAI lipid binding as described (23). Briefly, in a 96-well assay plate, 75 g of LDL was mixed with or without 3 g of native or modified apoAI in a final volume of 200 l of reaction buffer (50 mM Tris-HCl, pH 7.4, with 150 mM NaCl, and 2 mM CaCl 2 ). The plate was incubated in an absorbance plate reader at 37°C for 10 min. A 20-l aliquot of diluted phospholipase C, derived from Bacillus cereus (Sigma catalog number P7147), sufficient to induce LDL aggregation in 1 h, or a buffer control, was added to each well to hydrolyze the phospholipid polar head groups to make the LDL surface hydrophobic and to initiate aggregation. Aggregation at 37°C was monitored by absorbance at 478 nm, read every 2 min for a period of 1 h. ApoAI lipid binding activity resulted in the inhibition of the LDL aggregation and was calculated from the aggregation rate (⌬OD/⌬min) during the rapid linear phase of aggregation that occurred after a short time lag. The lipid binding activity for the modified apoAI was normalized to the activity of unmodified apoAI.
Spectral Studies of ApoAI-Far-UV circular dichroism spectra were recorded on a Pistar180 spectropolarimeter (Applied Photophysics, Surrey, UK). Standardization was performed with an aqueous solution of 0.06% ammonium d-(ϩ)-10-camphor sulfonate at a wavelengths of 260 to 200 nm and with a path length of 10 mm. Unmodified and modified apoAI samples were analyzed at ambient temperature in continuous scan mode with a 1-nm bandwidth (100000 counts/step). The spectra were normalized to mean residue ellipticity with the use of a mean residue molecular mass of 115.2 Da for apoAI. Fractional ␣-helix contents were calculated using the neural network-based K2d program (33) or by use of the mean residue ellipticity at 222 nm as described previously (34). Tryptophan fluorescence spectra were obtained using apoAI diluted to 25 g/ml in 60 mM phosphate buffer, pH 7.4, 100 mM NaCl. Tryptophan fluorescence emission spectra from 300 to 450 nm were collected at 25°C with an excitation at 295 nm to avoid tyrosine fluorescence. The wavelength of maximum fluorescence (WMF) and the fluorescence area under the curve were determined from the spectra after subtraction of the buffer base-line spectrum.
ApoAI Mass Spectrometry Analysis-Intact protein mass of unmodified and modified apoAI were determined by LC ESI-MS analysis with an LCQ Deca MS (Thermo, CA) equipped with a ultra plus II HPLC system (Micro-tech Scientific, Vista, CA) and a Vydac C4 column (5 m, 300 Å, 1 mm ϫ 5 cm). For these analyses, aqueous trifluoroacetic acid/acetonitrile solvents were used with a gradient of 30 -70% acetonitrile over 20 min and a flow rate of 30 l/min. The deconvolution of mass spectra of intact proteins was performed using Magtran software. For protein sequence data, apoAI gel bands were digested in situ with the protease Glu-C (Promega, Madison, WI) or trypsin, as described previously (22). The peptides were extracted from the gel, evaporated to dryness, and reconstituted in either 1% acetic acid or 0.1% formic acid for analysis by capillary column HPLC-electrospray ionization mass spectrometry. The detection of the oxidation sites was carried out using an LCQ Deca ion trap mass spectrometer system (ThermoFinnigan, San Jose, CA) equipped with a nanospray ionization source (Protana, Odense, Denmark). The source was operated under microspray conditions at a flow rate of 200 nl/min. The Glu-C-digested apoAI was analyzed by reversed phase capillary HPLC using a 50-m inner diameter. column with a 15-m inner diameter tip (New Objective Corp., Woburn, MA). The column was packed with ϳ6 cm of C18 packing material (Phenomenex, Torrence, CA) and eluted using a 45-min acetonitrile gradient (2-70%) in 50 mM acetic acid. The data were acquired in the data-dependent mode, recording a mass spectrum and three col-lision-induced dissociation (CID) spectra in repetitive cycles. The program Sequest was used to compare all CID spectra recorded to the sequence of protein and considering the appropriate mass changes in the residues (35).
Analysis of ApoAI Lysine Modification-ApoAI lysines were assayed using the sensitive fluorogenic o-phthaldialdehyde (OPA) reagent (36). Two-g aliquots of apoAI or MPO-modified apoAI were mixed with 250 l of complete OPA reagent (Sigma, catalog number P0532) and incubated for 5 min in 96-well plates with moderate shaking at room temperature. The fluorescence was determined in a fluorescence plate reader with excitation of 360 nm and emission of 460 nm. Fluorescence was corrected by subtraction of the MPO reaction buffer blank. All values are given relative to the unmodified apoAI sample. Control reactions in 3 M guanidine or 1% SDS revealed that Ͼ99% of apoAI lysine residues was accessible to the OPA reagent in aqueous buffer. In an attempt to reverse the MPO-mediated apoAI lysine modification (H 2 O 2 /apoAI ϭ 7.1:1), MPO/H 2 O 2 /Cl Ϫ -modified apoAI (7 M) was treated with NaBH 4 (on ice) or NaI (at room temperature), and both were performed with 350 M final concentration with stock reagents added in three intervals over a 30-min period. Acetic acid treatment of 7 M MPO-modified apoAI was performed at room temperature at a 1 M HAc final concentration, with stock reagent added in three intervals over a 30-min period. Methionine treatment of 7 M MPO-modified apoAI was performed by adding methionine to a final concentration of 350 M at 37°C for 30 min. All of the samples were then dialyzed against 50 mM sodium phosphate buffer, pH 7.0, at 4°C. Chloramine formation following MPO modification was quantified using 5-thio-2-nitrobenzoic acid (TNB) as described by Firuzi et al. (37). Briefly, after MPOmediated modification of apoAI, Trolox was added to 0.4 mM to remove excess HOCl. Five minutes later, TNB was added to 75 M, and the samples were incubated for 5 min. The absorbance of the samples was measured at 412 nm. The decrease of absorbance at 412 nm reflects the formation of chloramines, and the concentration of chloramines formed was calculated using an NaOCl standard.
Protein carbonyls were detected by Western blot using OxyBlot protein oxidation detection kit (Chemicon). The unmodified and modified apoAI were reacted with DNPH for 15 min followed by separation on 4 -20% SDS-polyacrylamide gel under reducing conditions. Protein was transferred to a PVDF membrane and probed with rabbit anti-2,4-dinitrophenol antibody. Detection was performed by enhanced chemiluminescence.
Lysine and 2-aminoadipic acid in oxidized apoAI were quantified by stable isotope dilution LC MS/MS on a triple quadrupole mass spectrometer (Ionics, Concord, Canada) interfaced with an Aria LX Series HPLC multiplexing system (Cohesive Technologies Inc., Franklin, MA). After hydrolysis in 6 N HCl under argon atmosphere for 24 h, samples were passed through a cation exchange solid-phase extraction column (Discovery DSC-SCX minicolumn; 1 ml; Supelco, Bellefone, PA) and washed with 0.8% formic acid. Lysine and 2-aminoadipic acid were then eluted with 5% NH 3 solution containing 70% methanol. After drying under vacuum and resuspension in 0.2% formic acid, samples were injected and separated on a Rexchrom phenyl column (5 m, 100 Å, 4.6 mm ϫ 25 cm, Bodman Industries, Aston, PA) at a flow rate of 0.8 ml/min. Mass spectrometric analyses were performed using electrospray ionization in positive ion mode with multiple reaction monitoring of precursor and characteristic fragment ions of lysine and 2-aminoadipic acid. The stable isotope [ 13 C 6 , 15  Analysis of ApoAI Net Charge-Isoelectric focusing (IEF) was performed to determine the isoelectric point (pI) of apoAI and modified apoAI. IEF was performed using precast pH 3-7 gels and IEF buffers (Invitrogen). The pH values on the gel were calibrated using an IEF standard protein marker (Invitrogen) and Coomassie staining. After electrophoresis at 100 V for 1 h, 200 V for 1 h, and 500 V for 30 min, the protein was transferred to a PVDF membrane for Western blot using a goat anti-human apoAI antibody, as described above.
Statistical Analyses-All data are presented as the mean Ϯ S.D. of triplicate determinations, unless stated otherwise. Statistical comparison of three or more groups was performed by using analysis of variance and the Newman-Keuls multiple comparison post-test.

RESULTS
Tyrosine-free ApoAI Is Susceptible to MPO-induced Loss of Function -We prepared recombinant human His-tagged wild type apoA-I (rh-apoAI) and a mutant form in which all seven tyrosine residues were substituted with phenylalanine residues (rh7YF-apoAI), as described under "Experimental Procedures." Both of the recombinant apoAI proteins migrated on reducing SDS-FIGURE 1. Gel electrophoresis and cholesterol acceptor activity of plasma and recombinant apoAI preparations. A, Coomassie staining of 1 g each of plasma apoAI, rh-apoAI, and tyrosine-free rh-apoAI (rh7YF-apoAI) run on a 14% SDS-polyacrylamide gel under reducing conditions. B, cholesterol acceptor activity of 5 g/ml each of apoAI, rh-apoAI, and rh7YF-apoAI incubated for 4 h with cholesterol-loaded RAW264.7 cells in the absence (ABCA1-independent) or presence (ABCA1-dependent) of 8-Br-cAMP pretreatment (n ϭ 3, Ϯ S.D.).
PAGEslightlylarger(ϳ29kDa)thanplasma-derivedapoAI(ϳ28kDa)because of the N-terminal His 6 tag (Fig. 1A). Coomassie Blue staining of the recombinant apoAI proteins revealed that they were at least 95% pure (Fig. 1A). To determine whether the His tag affected the lipid acceptor activity of the recombinant apoAI proteins, we compared their ABCA1-dependent and -independent cholesterol acceptor activity with plasma-derived apoAI (Biodesign International, Saco, ME) using the RAW264.7 murine macrophage cell line in which ABCA1 expression is inducible by cAMP analogues (38)(39)(40). All three apoAI preparations at 5 g/ml yielded robust and comparable ABCA1-dependent cholesterol acceptor activity, with virtually no ABCA1-independent cholesterol acceptor activity (Fig. 1B). The rh-apoAI and the tyrosine-free rh7YF-apoAI were subjected to MPO/H 2 O 2 /Cl Ϫ modification (H 2 O 2 /apoAI molar ratio of 28.6:1) and control incubations with MPO/Cl Ϫ or H 2 O 2 /Cl Ϫ . Only the full MPO/H 2 O 2 /Cl Ϫ system led to virtually complete loss of the ABCA1-dependent cholesterol acceptor activity for both recombinant proteins ( Fig. 2A). MPO/H 2 O 2 /Cl Ϫ modifications of rh-apoAI and rh7YF-apoAI were performed with varying H 2 O 2 /apoAI molar ratios and demonstrated similar dosedependent losses of ABCA1-dependent cholesterol acceptor activity (Fig. 2B). Thus, apoAI tyrosine chlorination is not required for MPO-mediated loss of the cholesterol acceptor activity of apoAI.
We have shown previously that the MPO-induced loss of the cholesterol acceptor activity of apoAI was associated with a decrease in the lipid binding activity of apoAI (23). The tyrosine-free rh7YF-apoAI and rh-apoAI were equally susceptible to MPO-induced loss of lipid binding activity (Fig. 3). Thus, tyrosine modification is not responsible for MPOmediated loss of the lipid binding activity of apoAI.
MPO-induced Changes in ApoAI Gel Migration Are Not Tyrosine-dependent-The rh-apoAI and rh7YF-apoAI proteins were incubated in the presence or absence of the complete MPO/H 2 O 2 /Cl Ϫ system at an H 2 O 2 /apoAI molar ration of 7.1:1 and were assessed by SDS-PAGE separation under reducing conditions. MPO modification led to a decrease in Coomassie Blue staining intensity, but other changes were too subtle to readily detect by this method (Fig. 4A). When the remaining half of the same gel was used for a Western blot and probed for apoAI, several MPO-mediated changes in apoAI migration were apparent, and all of these changes were observed for both the wild type and tyrosine-free recombinant proteins and thus were not dependent upon the presence of tyrosine (Fig. 4B). MPO modification induced the appearance of two distinct bands flanking the 28-kDa position of the unmodified protein. MPO induced a minor amount of fragmentation leading to a band at a molecular mass of Ͻ20 kDa. MPO also led to the appearance of two higher molecular mass bands of ϳ55 and 65 kDa, which most likely represent covalent dimers of the heterogeneous  . MPO modification alters electrophoretic mobility of both rh-apoA1 and rh7YF-apoA1. rh-apoA1 and rh7YF-apoA1 treated with or without the complete MPO/ H 2 O 2 /Cl Ϫ modification system were separated on a 14% SDS-polyacrylamide gel under reducing conditions. A, one-half of the gel was stained with Coomassie Blue. B, the other half of the gel was used for a Western blot that was probed with a goat anti-human apoA1 antibody, which revealed apoAI migration heterogeneity, putative dimerization, and fragmentation. monomeric forms. The presence of MPO-induced dimer bands for both rh-apoAI and tyrosine-free rh7YF-apoAI indicates that dityrosine formation is not the primary mechanism of apoAI dimerization.
Other Effects of MPO Modification on ApoAI Structure-Because tyrosine modification is not required for the MPO-mediated loss of apoAI function, we examined other features of apoAI after MPO modification to gain insight into the potential mechanism of the loss of function for apoAI. The effects of MPO-mediated modification on apoAI secondary structure were assayed by CD at increasing molar ratios of H 2 O 2 /apoAI (Fig. 5A). ApoAI ␣-helix content was calculated by two methods as described under "Experimental Procedures," one based on the molar residue ellipticity at 222 nm and the other used curve fitting for the spectra from 200 to 240 nm. MPO modification led to a dose-dependent decrease in apoAI ␣-helix content (TABLE ONE), and a similar finding was obtained previously using NaOCl treatment of apoAI (30).
Tryptophan residues are known to be susceptible to HOCl and/or MPO modification (41)(42)(43)(44). Fluorescence spectroscopy was performed to examine MPO-mediated spectral changes in the four tryptophan residues of apoAI. MPO modification of apoAI led to an H 2 O 2 dose-dependent loss of tryptophan fluorescence (Fig. 5B and TABLE TWO). In addition, there was a red shift of the tryptophan WMF from 342.5 nm for unmodified apoAI to 351 nm at an H 2 O 2 /apoAI ratio of 4.3:1. This WMF red shift is indicative of tryptophan movement from a shielded hydrophobic environment to an aqueous exposed environment, similar to the shift that occurs during apoAI denaturation by guanidine (11). At a 7.1:1 ratio of H 2 O 2 /apoAI, the WMF started to reverse, although the total fluorescence area under the curve was less than 20%, decreasing the confidence in the WMF assignment. We observed a similar loss of tryptophan fluorescence and a biphasic shift in the WMF when the tryrosine-free rh7YF-apoAI was dose-dependently modified with the complete MPO/H 2 O 2 /Cl Ϫ system (TABLE TWO).
Protein mass of unmodified and MPO-modified apoAI was determined by mass spectrometry, as described under "Experimental Procedures," after isolation of monomeric apoAI by fast protein liquid size exclusion chromatography of guanidine-denatured apoAI on a Superdex 200 column. The mass of unmodified apoAI was 28,077 Da. With increasing H 2 O 2 /apoAI molar ratio in the modification reaction, the apoAI mass became heterogeneous, but a major peak could still be identified up to an H 2 O 2 /apoAI molar ratio of 14.3, but beyond this ratio no major mass was evident due to excessive mass heterogeneity. There was a dose-dependent increase in apoAI mass in response to MPO modification (Fig. 6).
ApoAI Tryptophan Oxidation-Because apoAI tryptophan fluorescence was decreased and protein mass was increased due to MPO modification, we used tandem mass spectrometry to identify specific tryptophan modifications in MPO/H 2 O 2 /Cl Ϫ -treated apoAI. CID spectra were obtained for the apoAI peptide Phe 71 -Glu 78 . We identified two    OCTOBER 7, 2005 • VOLUME 280 • NUMBER 40

ApoAI Tyrosines Do Not Mediate MPO-induced Functional Loss
higher mass forms of this peptide after MPO modification, with increases of 16 and 32 atomic mass units compared with the unmodified peptide (Fig. 7). The CID spectra revealed this was due to modification of tryptophan 72, consistent with mono-and dihydroxylation. Additional CID spectra revealed that the three other tryptophan residues, 8, 50, and 108, were also detected as hydroxylated (data not shown). ApoAI Lysine Modification-MPO-or HOCl-induced chlorination is known to convert lysine ⑀-amines in protein into semi-stable chloramine derivatives, which can further react with other protein moieties (45)(46)(47)(48)(49)(50). There are 22 free amino groups in apoAI, one at the N terminus and 21 lysine ⑀-amines, many of which are highly conserved among species (7). We first examined the overall charge of apoAI after modification by MPO, and we found that the complete MPO/H 2 O 2 /Cl Ϫ system led to an acidic shift in the apoAI's charge revealed by isoelectric focusing gel analysis (Fig. 8). The native apoAI and the apoAI treated only with H 2 O 2 each had a pI of 5.6 to 5.7, whereas the MPO-modified apoAI was more heterogeneous with a wider pI range extending from ϳ5.0 to 5.6. This result is consistent with the hypothesis that the MPO modification of apoAI converted lysine ⑀-amino-positive charges to the uncharged chloramine derivatives or to some other uncharged species.
Using the OPA reagent to quantify apoAI free amino groups, there was a dose-dependent loss of apoAI lysine residues in response to MPO modification at increasing molar ratios of H 2 O 2 /apoAI, with ϳ33% of the lysine residues lost at an H 2 O 2 /apoAI ratio of 14.3 (Fig. 9A).
The immediate chlorinating product of the lysine ⑀-amino group is lysine mono-and dichloramine. The formation of chloramines following MPO modification of apoAI was assayed using the TNB reagent. The result showed a H 2 O 2 dose-dependent increase of chloramine formation (Fig. 9B). However, the chloramine content of MPO-modified apoAI was not stable, as overnight storage at 4°C led to chloramines loss (Fig. 9C). But the same sample did not show recovery of lysine ⑀-amino groups with sample storage (Fig. 9D), suggesting that lysine chloramines decomposed to something other than lysine. Treatments of this modified and stored apoAI with potassium iodide, sodium borohydride, methionine, or acetic acid, all reported to regenerate lysines after HOCl modification of peptides or protein (49 -51), failed to regenerate the primary amines, nor did they lead to the recovery of ABCA1-dependent cholesterol acceptor activity of apoAI (data not shown). We also determined that neither acetic acid nor methionine treatment could reverse the MPO-mediated apoAI dimer observed upon Western blot (data not shown). Thus the loss of apoAI lysine residues in the presently used MPO/H 2 O 2 /Cl Ϫ modification reaction was associated with modification beyond the reversible chloramine derivative.
Lysine chloramines can decompose into the carbonyl aminoadipic semialdehyde, which can be further oxidized to 2-aminoadipic acid (Fig.  10A) (52)(53)(54). The presence of apoAI carbonyls was detected by Western blot using anti-2,4-dinitrophenol antibody after reaction of apoAI with DNPH. Carbonyls were found on all species of modified apoAI (at H 2 O 2 /apoAI molar ratio of 14.3) but not on unmodified apoAI (Fig.  10B); however, this assay is not specific for aminoadipic semialdehyde. Mass spectrometry after protein acid hydrolysis was used in order to specifically detect and quantify lysine residues and 2-aminoadipic acid, the end product of lysine oxidation. MPO modification at an H 2 O 2 / apoAI molar ratio of 14.3 led to a 17% loss of lysine compared with unmodified apoAI, accompanied by the de novo formation of 2-aminoadipic acid, which could account for 44% of lysine loss (Fig. 10C). The 17% loss of lysine detected by this method is less than the 33% lysine loss observed by the OPA assay (Fig. 9A) under similar modification condi-  tions. This could be because of the strong acid hydrolysis procedure used for amino acid analysis, which can convert chloramine back to lysine (49).

DISCUSSION
The oxidative modification of LDL is widely believed to be proatherogenic, as oxidized LDL uptake by cellular scavenger receptors leads to foam cell formation and due to the inherent inflammatory properties of oxidized lipids (55). Oxidative modification of HDL may also be proatherogenic by reducing the protective properties of HDL. The study of dysfunctional HDL is of growing interest, and recent findings point toward MPO-catalyzed oxidation as a likely mechanism for the formation of dysfunctional HDL. However, the precise underlying modifications leading to apoAI/HDL dysfunction remain unclear. The present study unequivocally shows that tyrosine modification is not required for the MPO-mediated loss of the function of apoAI.
There is an extensive literature on HDL and apoAI modification that serves as the context for the current study. Copper-mediated oxidation of HDL was shown to lead to altered HDL migration on agarose gels, apoAI proteolysis, and decreased ability of HDL to unload cholesterol esters from cholesterol-loaded macrophages (56). Malondialdehyde modification of HDL was shown to lead to similar changes as copper oxidation in regard to HDL structure and function, and it was associated with the loss of lysine and tryptophan residues and apoAI polymerization (57), changes that are similar to those we observed in the current study by MPO modification of lipid-free apoAI. Subsequently, incubation of HDL with activated polymorphonuclear leukocytes was shown to render apoAI more negatively charged and to decrease its cellular cholesterol acceptor activity (58,59).
In 1999, Bergt et al. (60) demonstrated that MPO/H 2 O 2 /Cl Ϫ or reagent HOCl modification of HDL led to a loss of unsaturated fatty acids in phospholipids and cholesterol esters and the loss of cholesterol acceptor activity from cholesterol-loaded J774 macrophages. As this study was performed prior to the discovery of ABCA1 as the Tangier disease gene and the appreciation of its preference for mobilizing the intracellular cholesterol pool (61,62), the conditions under which these investigators assessed cholesterol efflux (in the absence of cAMP induction of ABCA1 (63), and labeling the plasma membrane pool with cholesterol) strongly favors scavenger receptor-BI-mediated lipid efflux (64). Indeed, the Ͼ100:1 ratio of HOCl/apoAI that Bergt et al. (60) used to modify HDL also led to loss of phospholipid fatty acids, thus providing a mechanism for decreased SR-BI-mediated lipid efflux, which is correlated with HDL phospholipid content (64). Bergt et al. (60) also showed that modification of lipid-free apoAI at an HOCl/apoAI ratio of 25:1 led to the loss of apoAI methionine residues, consistent with the high sensitivity of methionine to MPO/H 2 O 2 /Cl Ϫ (65). Subsequently, it was shown that selective apoAI methionine oxidation does not disrupt its structure or function and actually led to an increase in lipid binding and cellular lipid acceptor activities (29). Bergt et al. (60) using an HOCl/apoAI ratio of 100:1 also observed the loss of tyrosines and phenylalanines and the partial loss of lysines and arginines. The cholesterol acceptor activity of recombinant discs made by cholate dialysis using apoAI and dipalmitoylphosphatidylcholine was much more sensitive to the HOCl treatment, with loss of activity starting at an HOCl/apoAI ratio of 5:1 (60), more similar to the sensitivity than we observed for the modification of lipid-free apoAI in the current study. In addition to decreased lipid acceptor activity, Panzenboeck et al. (28) also showed that MPO or HOCl-modified HDL was more susceptible to uptake and degradation by macrophages, thus turning HDL from a lipid-removing lipoprotein to a lipid-loading lipoprotein.
In 2000, Bergt et al. (30) used mass spectroscopy after trypsin digestion and HPLC separation in order to identify specific alterations caused FIGURE 9. MPO modification of apoAI effects on lysine and chlorolysine content. A, apoAI was incubated with the complete MPO/H 2 O 2 /Cl Ϫ system at the indicated H 2 O 2 /apoAI molar ratios. The lysine content of these samples decreased in an H 2 O 2 dose-dependent manner, assayed using the OPA reagent as described under "Experimental Procedures." The data are expressed as the % lysine reactivity relative to the unmodified sample. B, apoAI was modified with MPO/H 2 O 2 /Cl Ϫ system at the indicated molar ratio followed by addition of Trolox to quench HOCl. Chloramine concentrations were assayed using the TNB reagent as described under "Experimental Procedures," based on a standard curve of NaOCl. C and D, MPOmodified apoAI, in the absence of methionine post-treatment, was assayed directly or stored at 4°C overnight. TNB (C) and OPA (D) assays were performed to measure chloramine and lysine content of the modified and unmodified apoAI samples, respectively. by reagent-or MPO-generated HOCl; these modifications caused the disappearance of the methionine-containing tryptic peptides and decreases in the abundance of other peptides. Mass spectroscopy identified new five tryptic fragments consistent with the formation of methionine sulfoxide for three fragments, and chlorination of an unknown residue for two fragments (residues 46 -59 and residues 62-77) (30). These two peptides were isolated and subjected to further HOCl modification and amino acid analysis, which revealed the quantitative loss of phenylalanine and the partial loss of lysine and tryptophan. In addition, after HBr hydrolysis of these peptides, 4-chlorophenylalanine was identified by mass spectroscopy (30). We suspect that this and many subsequent mass spectroscopy studies may have suffered because of the prevalent use of trypsin to isolate peptides for mass analysis. Indeed, these two studies by Bergt et al. (30,60) and the current study show that lysine residues are modified by MPO or HOCl, which by itself would lead to the loss of trypsin reactivity.
In the past year, there has been a major shift in the focus of HDL and apoAI modification studies with work from our group and by Heinecke and co-workers (22, 24, 66 -68) showing that this modification is physiologically relevant by demonstrating the following. 1) ApoAI in plasma and more so in arterial lesions is a selective target of MPO modification that leads to the nitration and chlorination of specific apoAI tyrosine residues. 2) Plasma apoAI nitro-and chlorotyrosine levels are higher in coronary artery disease patients than in control subjects. 3) ApoAI tyrosine chlorination, whether in endogenous plasma or after in vitro MPOmediated modification, is associated with the specific loss of ABCA1mediated cholesterol acceptor activity. 4) The levels of chlorotyrosine in apoAI recovered from lesions are higher than those associated with the loss of the cholesterol acceptor activity from apoAI through in vitro modification, suggesting that this modification is pathophysiologically relevant. In a mixed cohort of cardiovascular disease subjects and controls, we demonstrated that the level of apoAI chlorotyrosine was a better predictor of cardiovascular disease than the levels of total plasma protein-bound chloro-or nitrotyrosine (22). Both groups using similar mass spectroscopy methods identified tyrosine 192, within helix 8, as the most sensitive site to MPO-mediated chlorination (23, 24). We demonstrated selective modification of tyrosine 192 as a preferred target of apoAI modification in lipoprotein recovered from human atheroma (23). Parallel in vitro studies from Heinecke and co-workers (66) has shown through use of short synthetic peptides that tyrosine chlorination is maximally stimulated when the tyrosine is three residues away from a lysine residue, which is the case for the sensitive tyrosine 192. These collective findings inspired us to determine whether tyrosine modification was responsible for the loss of apoAI function. We demonstrate in the current study that tyrosine-free apoAI is equally susceptible as native apoAI to MPO-mediated loss of function and gel mobility changes observed by Western blot. Thus, the levels of plasma apoAI nitro-and chlorotyrosine may serve as physiological useful and specific markers of MPO activity in vivo; however, tyrosine modification itself is not responsible for turning the "good" HDL cholesterol "bad." The discovery of which apoAI residues when modified by MPO/H 2 O 2 /Cl Ϫ or HOCl are responsible for its loss of function must await further sitedirected mutagenesis studies to replace other sensitive residues such as tryptophan, phenylalanine, methionine, arginine, histidine, and lysine.
We have extended and confirmed several prior observations on the effects of MPO-or reagent-generated HOCl on apoAI structure, specifically apoAI dimerization, the loss of ␣-helix content, and the loss of tryptophan and lysine residues. By using tandem mass spectrometry, we also made the novel observation that all four apoAI tryptophan residues can be identified as their hydoxylated derivatives. Furthermore, we have demonstrated that apoAI lysine ⑀-amines can be converted to semistable chloramines, which decompose to yield 2-aminoadipic acid.
The protective effects of HDL may be mediated by several mechanisms, including its role in reverse cholesterol transport and via its antiinflammatory properties. Navab and co-workers (69 -71) have demonstrated several conditions in which HDL can lose its anti-inflammatory properties, including during the acute phase response when serum amyloid A partially replaces apoAI in HDL, during influenza A infection, and in apolipoprotein AII transgenic mice. Ansell et al. (72) demonstrated that HDL from coronary heart disease subjects promoted LDLinduced monocyte chemotactic activity in a co-culture assay, whereas HDL from control subjects inhibited this activity, although there was a large inter-individual variation in this activity of HDL. Previously, we assessed the ABCA1-dependent cholesterol acceptor activity of plasmaderived apoAI immunoisolated from a cohort of human subjects, and we found that this cholesterol acceptor activity was inversely correlated with apoAI nitro-and chlorotyrosine levels (22). Thus, it is possible that apoAI modification in vivo by MPO may play a role not only in reverse cholesterol transport but may also explain some of the inter-individual variation in the anti-inflammatory activity of HDL.