UBC Faculty Research and Publications

Lipid-bound apolipoproteins in tyrosyl radical-oxidized HDL stabilize ABCA1 like lipid-free apolipoprotein… Hossain, Mohammad A; Ngeth, Sereyrath; Chan, Teddy; Oda, Michael N; Francis, Gordon A Jan 16, 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12858_2011_Article_376.pdf [ 350.78kB ]
JSON: 52383-1.0223226.json
JSON-LD: 52383-1.0223226-ld.json
RDF/XML (Pretty): 52383-1.0223226-rdf.xml
RDF/JSON: 52383-1.0223226-rdf.json
Turtle: 52383-1.0223226-turtle.txt
N-Triples: 52383-1.0223226-rdf-ntriples.txt
Original Record: 52383-1.0223226-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessLipid-bound apolipoproteins in tyrosyl radical-oxidized HDL stabilize ABCA1 like lipid-freeapolipoprotein A-IMohammad A Hossain1, Sereyrath Ngeth1, Teddy Chan1, Michael N Oda2 and Gordon A Francis1*AbstractBackground: ATP-binding cassette transporter A1 (ABCA1) mediates the lipidation of exchangeableapolipoproteins, the rate-limiting step in the formation of high density lipoproteins (HDL). We previouslydemonstrated that HDL oxidized ex vivo by peroxidase-generated tyrosyl radical (tyrosylated HDL, tyrHDL) increasesthe availability of cellular cholesterol for efflux and reduces the development of atherosclerosis when administeredto apolipoprotein E-deficient mice as compared to treatment with control HDL.Results: In the current study we determined that tyrHDL requires functional ABCA1 for this enhanced activity. Likelipid-free apolipoprotein A-I (apoA-I), tyrHDL increases total and cell surface ABCA1, inhibits calpain-dependent and-independent proteolysis of ABCA1, and can be bound by cell surface ABCA1 in human skin fibroblasts.Additionally, tyrHDL apoproteins are susceptible to digestion by enteropeptidase like lipid-free apoA-I, but unlikelipid-bound apoA-I on HDL, which is resistant to proteolysis.Conclusions: These results provide the first evidence that lipid-bound apolipoproteins on the surface of sphericalHDL particles can behave like lipid-free apoA-I to increase ABCA1 protein levels and activity.BackgroundHigh density lipoprotein cholesterol (HDL-C) levels inhuman plasma correlate generally with protectionagainst coronary heart disease, an effect believed to bedue to multiple protective actions of HDL including sti-mulating the removal of excess cholesterol from cellsand reducing inflammation in the artery wall [1]. Theinitial formation of HDL particles requires the mem-brane lipid transporter ATP-binding cassette transporterA1 (ABCA1) to mediate delivery of cellular lipids toHDL apolipoproteins. This reduces excess cholesterolstores in cells including arterial wall macrophages [2]. Inaddition to receiving cellular lipids, lipid-free apolipo-protein A-I (apoA-I) binds to and inhibits the degrada-tion of ABCA1, further enhancing the formation ofHDL particles [3,4]. Lipidated apoA-I on the surface ofdiscoidal or spherical HDL particles, however, has areduced affinity for ABCA1 and does not inhibitABCA1 degradation [3,5]. These findings suggest struc-tural motifs present in lipid-free but not lipid-boundapoA-I participate in ABCA1 binding and enhancementof ABCA1 cell surface stability.In addition to increasing ABCA1 expression transcrip-tionally, therapies that increase ABCA1 cell surface sta-bility represent a potential means to increase HDLproduction. We previously showed that oxidation ofHDL ex vivo with peroxidase-generated tyrosyl radical(tyrosylated HDL or tyrHDL) increases the ability ofHDL to deplete the regulatory pool of intracellular cho-lesterol [6], increase cholesterol available for subsequentefflux to apoA-I and other lipid acceptors [7], andreduce the development of atherosclerosis in apoE-defi-cient mice [8]. Enhanced depletion of intracellular cho-lesterol was seen only when tyrHDL apoproteins andspecifically the apoAI-apoAII heterodimer in tyrHDLwere bound to the surface of spherical HDL particles,but not in their lipid-free form. This suggests a particu-lar conformation of tyrHDL apos when lipid-boundmediates the enhanced effect [9]. In the current study* Correspondence: gordon.francis@hli.ubc.ca1Department of Medicine, UBC James Hogg Research Centre, Heart andLung Institute, St. Paul’s Hospital, Vancouver, British Columbia, Canada V6Z1Y6Full list of author information is available at the end of the articleHossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1© 2012 Hossain et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.we tested the hypothesis that tyrHDL has these benefi-cial effects by enhancing ABCA1 activity, either byincreasing ABCA1 expression or cell surface stability.Our results indicate that, like lipid-free apoAI, tyrHDLhas no effect on ABCA1 mRNA levels, but increasestotal and cell surface ABCA1, reduces calpain-depen-dent and -independent degradation of ABCA1, and canbe crosslinked directly to ABCA1. Also like lipid-freeapoA-I but not HDL apos, apos on the surface oftyrHDL are more susceptible to enteropeptidase diges-tion. These results suggest that tyrosyl radical oxidationinduces changes in apolipoprotein conformation on thesurface of spherical HDL that allows lipid-boundtyrHDL apos to interact with ABCA1 like lipid-freeapoA-I.MethodsMaterialsCholesterol, L-tyrosine, horseradish peroxidase, hydro-gen peroxide (30%, ACS grade), dietheylenetriaminepen-taacetic acid (DTPA; free acid form), essentially fattyacid-free bovine serum albumin (BSA), N-acetyl-Leu-Leu-norleucinal (ALLN), LXR agonist T0901317 andfetal bovine serum were purchased from Sigma. [14C]Oleate (55 mCi/mmol) was from GE Healthcare. Dul-becco’s modified Eagle’s medium (DMEM) was pur-chased from Hyclone. PE-SIL G plastic backed flexibleplates used for thin-layer chromatography analysis werefrom Whatman. Nitrocellulose membranes, sodiumdodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) reagents, pre-stained protein molecular massmarkers and Chelex 100 resin were from Bio-Rad. μ-Calpain was from Calbiochem, and Trizol from Invitro-gen. Cross-linking agent dithiobis(succinimidyl) propio-nate (DSP) was from Pierce.Preparation of HDL, apolipoprotein A-I, and tyrosylatedHDLHDL3 was isolated from pooled plasma from healthyfasting donors by density gradient ultracentrifugation[10]. Apo A-I was purified from human plasma usingQ-Sepharose Fast Flow chromatography as previouslydescribed [11]. Tyrosylation of HDL was carried out at37°C for 24 h in 66 mM potassium phosphate buffer,pH 8.0, which had been passed over Chelex 100 resin toremove transition metal ions. The reaction mixture con-tained a final concentration of 1 mg/mL HDL protein,100 μM diethylenetriaminepentaacetic acid (to inhibitmetal ion-catalyzed oxidation), 100 nM horseradish per-oxidase (250 units/mg), 100 μM H2O2, and 100 μM L-tyrosine. The reaction mixture was subjected to sizeexclusion chromatography as previously described [7]prior to use in cell culture experiments.Cell cultureHuman skin fibroblasts were cultured in DMEM supple-mented with 10% FBS containing 50 units/ml penicillin-streptomycin solution (Invitrogen) and grown in humi-dified 95% air and 5% CO2 at 37°C. Confluent cells wererinsed twice with phosphate-buffered saline (PBS) con-taining 1 mg/ml BSA (PBS-BSA) and incubated for 24 hin DMEM containing 2 mg/ml BSA with 30 μg/ml non-lipoprotein cholesterol. To allow equilibration of addedcholesterol, cells were rinsed twice with PBS-BSA andincubated for an additional 24 h in DMEM containing 1mg/ml BSA (DMEM-BSA).Cholesterol esterification assayCholesterol-loaded cells were incubated for 16 h inDMEM-BSA and the indicated additions, washed oncewith PBS, and incubated for 1 h at 37°C with DMEMcontaining 9 mM [14C] oleate bound to 3 mM BSA[12]. Cells were chilled on ice and rinsed twice with ice-cold PBS-BSA, twice with PBS, and stored at -20°C untilextraction. Cellular lipids were extracted with hexane:isopropanol (3:2, v/v). Sterol species were separated bythin layer chromatography (TLC) on PE SIL G plastic-backed plates (Whatman) developed in hexane/diethylether/acetic acid (130:40:1.5 v/v/v). Lipid spots wereidentified by staining with I2 vapor and co-migrationwith standard. After allowing I2 stain to evaporate, cho-lesteryl ester spots were taken for determination ofradioactivity by liquid scintillation counting. Cell proteinwas extracted by incubation of cells with 0.5 mL of 0.1N NaOH for 1 h on a rotary shaker. Total cell proteinwas determined by the Lowry assay [13] using albuminas a standard.Reverse transcriptase-PCR analysis of ABCA1 mRNATotal RNA was isolated from cells using Trizol, follow-ing manufacturer’s protocols. Single-strand cDNA wassynthesized by a SuperScript pre-amplification system(Invitrogen) from 2 μg of the total RNA. ABCA1 DNAamplification was performed by initial denaturation at95°C for 3 mins. Thereafter, denaturing was at 95°C for20 seconds, annealing at 58°C for 20 seconds, andextension at 72°C for 40 seconds for at total of 40 cycles(10). SYBR Green (Quanta Biosciences) was used todetect PCR products in real-time using a Realplex2 Mas-tercycler thermocycler (Eppendorf). The human house-keeping gene cyclophilin cDNA was amplified using thesame conditions. ABCA1 mRNA levels were calculatedusing the comparative CT method relative to cyclophilin.The following primers were used: human ABCA1, 5’-GAC ATC CTG AAG CCA ATC CTG (forward), 5’-CCT TGT GGC TGG AGT GTC AGG T (reverse);human cyclophilin, 5’- ACC CAA AGG GAA CTGHossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 2 of 11CAG CGA GAG C (forward), 5’-CCG CGT CTC CTTTGA GCT GTT TGC AG (reverse).Gel electrophoresis and immunoblottingCells were scraped into N-dodecyl-b-D-maltoside–con-taining lysis buffer (20 mM tris, 5 mM EDTA, 5 mMEGTA, 0.5% N-dodecyl-b-D-maltoside, pH 7.5) withcomplete protease inhibitor (Roche Molecular Biochem-icals) and stored at -80°C. Cells were homogenized bysonication, centrifuged to remove the cellular debris,and the lysate collected for ABCA1 protein detection.The cell lysate was run on a 5-15% gradient SDS-PAGEgel under reducing conditions and transferred to nitro-cellulose membrane for 16 h at 4°C. Immunoblottingwas performed using a polyclonal rabbit anti-humanABCA1 antibody from Novus Biologicals (#NB400-105,1:1000 dilution) and a goat anti-rabbit IgG horseradishperodixase-conjugated secondary antibody from Sigma(#A-0545, 1:10,000 dilution). Immunoblots were re-probed with rabbit polyclonal anti-actin (Abcam,#AB8227-50, 1:2000) as loading control. SDS-PAGE wasalso performed for the detection of apoA-I and apoA-IIusing 12% polyacrylamide gels under reducing condi-tions followed by staining with 0.25% Coomassie Brilli-ant Blue. Immunoblot analysis of apoA-I, HDL andtyrHDL were performed using rabbit anti-human apoA-I polyclonal antibody (Calbiochem #178422, 1:10,000)and goat anti-human apoA-II monoclonal antibody (Cal-biochem, #178464, 1:20,000).Calpain and ALLN studiesSensitivity of ABCA1 to digestion by exogenous calpainfollowing incubation with apoA-I, HDL, or tyrHDL wasperformed using digitonin permeabilization and treat-ment with 0.5 μM μcalpain for 20 min as previouslydescribed [4]. To assess cellular levels of ABCA1 follow-ing inhibition of endogenous calpain-dependent degra-dation by ALLN, cells were incubated for 16 h with theindicated conditions plus 50 μM ALLN as previouslydescribed [3]. Cell lysates were then analyzed by ABCA1immunoblotting as described above.Biotinylation of cell-surface proteinsProteins on the surface of cells treated with the indi-cated additions were biotinylated, harvested, and puri-fied on streptavidin resin using a Pierce Cell SurfaceProtein Isolation Kit (Pierce) following manufacturer’sprotocols [14]. Briefly, after surface biotinylation andquenching, cells were lysed and debris was removed bycentrifugation at 10,000 × g for 10 min. Biotinylatedproteins in the supernatant (500 μl) were bound to astreptavidin column and eluted in 250 μl of SDS samplebuffer. The total and cell surface fraction were treatedwith 50 mM dithiothreitol to remove biotin, resolved bySDS-PAGE, and immunoblotted using antibodies againstABCA1, b-actin (Abcam), and ERK (New EnglandBiolabs).Crosslinking of apoA-I, HDL or tyrHDL to ABCA1Chemical cross-linking was performed as described byWang et al. [15] with minor modifications as follows. Cellswere grown to confluence in 100 mm culture dishes, cho-lesterol loaded with non-lipoprotein cholesterol, equili-brated, and then treated with 5 μM LXR Agonist(T0901317) in DMEM-BSA for 24 h to up-regulateABCA1 expression. The cells were washed twice with PBSand then treated with DMEM/BSA or the same mediacontaining 10 μg/mL apoA-I, HDL, or tyrHDL in DMEM-BSA for 2 h at 37°C. Cells were then placed on ice for 15min and washed three times with PBS. Cross linking agentDSP was dissolved immediately before use in dimethylsulfoxide (Sigma) and diluted to 250 μM with PBS. 10 mlof DSP solution was added to the cells for 1 h at roomtemperature, the medium removed, and dishes werewashed twice with ice-cold PBS. Cells were lysed at 4°C inbuffer containing 20 mM Tris, 0.5 mM EDTA, 0.5 mMEGTA, 1% Triton X-100 (BDH Chemicals, pH 7.5), withcomplete protease inhibitor (Roche). by aspiration with afine-tipped needle, and the mixture was left to rotate at 4°C for 30 min. Cell lysates were centrifuged at 1,500 rpmfor 10 min at 4°C to remove cellular debris The superna-tant was collected and incubated with loading buffer inthe absence or presence of b-mercaptoethanol before load-ing onto 4-20% gradient SDS-PAGE gels for electrophor-esis followed by electrotransfer to a nitrocellulose sheet forimmunoblot analysis. ApoA-I cross-linked to ABCA1 wasdetected by blotting the membranes with rabbit polyclonalanti-human apoA-I antibody and goat anti-rabbit IgGhorseradish peroxidase-conjugated secondary antibody asnoted above. To detect ABCA1, the membrane wasstripped and reprobed with rabbit polyclonal antibody toABCA1 and goat anti-rabbit IgG horseradish peroxidase-conjugated secondary antibody as noted above.Enteropeptidase digestionLipid-free apoA-I, HDL and tyrHDL were incubatedwith 0.13 U enteropeptidase (Calbiochem) per g of pro-tein for 6 h at 37°C as previously described for apoA-I[16]. Cleaved products were run on 12% SDS-PAGEunder reducing conditions and stained with 0.25% Coo-massie brilliant blue. Immunoblot analysis of apoA-Iand apoA-II were performed using rabbit anti-humanapoA-I polyclonal antibodies and goat anti-humanapoA-II monoclonal antibodies.StatisticsResults for Figures 1 and 2A were analyzed usingGraphPad Prism version 5.0 and are presented as theHossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 3 of 11mean ± S.D. Significant differences between experimen-tal groups were determined using the Student’s t test.ResultstyrHDL requires ABCA1 to deplete cholesterol availablefor esterificationLipid-free apoA-I depletes the regulatory pool of choles-terol available for esterification by acyl-CoA:cholesterolacyltransferase (ACAT) much more readily than HDL[17], due to mobilization of this cholesterol pool byABCA1 and delivery to apoA-I [2]. tyrHDL mobilizesACAT-accessible cholesterol much more readily thanHDL, an effect that was not found to be due to directinhibition of ACAT or stimulation of cholesteryl esterhydrolysis [6,7]. To investigate the role of functionalABCA1 in this activity, HDL and tyrHDL were incu-bated with cholesterol-loaded fibroblasts from controland two unrelated patients with Tangier Disease, andthe subsequent production of cholesteryl[14C]oleatequantified. As reported previously [6], tyrHDL showed amarkedly increased ability to deplete ACAT-accessiblecholesterol from normal fibroblasts in comparison tocontrol HDL (Figure 1A). tyrHDL did not have anenhanced ability to mobilize ACAT-accessible choles-terol compared to HDL from Tangier Disease fibro-blasts, and depletion of this pool was reduced to similarlevels by both forms of HDL in comparison to normalcells (Figure 1B and 1C). This suggests tyrHDL requiresABCA1 for its enhanced cholesterol-mobilizing effect.tyrHDL increases ABCA1 protein but not mRNA levelsTo test whether tyrHDL has an effect on ABCA1 tran-scription, cholesterol-loaded fibroblasts were incubatedwith either fatty acid-free bovine serum albumin (BSA)alone or with the addition of 10 μg/mL apoA-I, HDL, ortyrHDL for 2 h or 16 h prior to determination ofABCA1 mRNA level by real-time PCR. At 2 h, ABCA1mRNA level was reduced similarly by all treatmentswhen compared to cholesterol-loaded cells at 0 h (Fig-ure 2A). Following a 16 h incubation, ABCA1 mRNA0 10 20 30 400204060801000 10 20 30 40HDL Protein (—g/mL)Cholesterol Esterification (% Control) Normal TD10 10 20 30 40020406080100TD2HDLtyrHDLA. C. B. * * Figure 1 tyrHDL requires ABCA1 to enhance depletion of ACAT-accessible cholesterol. Human skin fibroblasts from a normal donor and 2unrelated Tangier Disease (TD) subjects were grown to confluence, loaded with non-lipoprotein cholesterol for 24 h, equilibrated for 24 h inDMEM containing 1 mg/ml BSA, and then incubated in the same medium plus the indicated concentration of HDL or tyrHDL for 16 h. Tomeasure residual ACAT-accessible cholesterol, cells were then incubated with [14C]oleate for 1 h, and cholesteryl [14C]esters present in cellularlipid extracts were determined. Results are expressed as percentage of cholesterol esterification in cells untreated with HDL or tyrHDL. Avg ± SDof quadruplicates, representative of 3 experiments. *, p < 0.01 when compared with tyrHDL-treated cells.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 4 of 11Chol (-)Chol (+)BSA (2h)ApoAI (2h)HDL (2h)tyrHDL (2h)BSA (16h)ApoAI (16h)HDL (16h)tyrHDL (16h)ABCA1 mRNA Expression(normalized to cyclophilin)012345ABCA1 PDI Cholesterol ApoAI HDL tyrHDL 0 hr 2 hrs 16 hrs - + + + + + + + + + - - - + - - - + - - - - - - + - - - + - - - - - - + - - - + * * *   0.4        1        0.8      0.9      0.7     1.3      1.1     1.9      1.6      3.0  p<0.01 * Figure 2 tyrHDL increases ABCA1 protein but not mRNA levels. Fibroblasts grown to confluence, cholesterol loaded, and equilibrated as inFigure 1 were treated with medium containing 1 mg/ml BSA alone or plus 10 μg/ml apoA-I, HDL or tyrHDL for 2 or 16 h. (A) mRNA wascollected using Trizol™ extraction at the indicated time and qPCR was performed with cyclophilin used as the internal control. Data representAvg ± SD for triplicates of each sample and are representative of three experiments with similar results. *, p < 0.01 when compared with BSA-treated cells at 16 h. (B) Cells were collected in N-dodecyl-b-D-maltoside lysis buffer with proteinase inhibitors, and ABCA1 protein wasdetermined following separation of proteins on 5-15% gradient gels and immunoblotting. Values at the top of lanes represent the ratio ofdensitometer readings of ABCA1 protein normalized for protein disulfide isomerase (PDI) loading control, with this ratio in cholesterol-loadedcells at 0 h set as 1. The data are representative of three experiments with similar results. Chol- indicates cells not loaded with cholesterolloading prior to analyses.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 5 of 11remained elevated in cells treated with BSA alone, butwas reduced similarly to low levels by apoAI, HDL, andtyrHDL, consistent with depletion of cellular cholesterolregulating ABCA1 expression by all of these treatments.These results are consistent with previous results show-ing no effect of apoA-I on ABCA1 mRNA level [3], andsuggest no differential effect of tyrHDL on ABCA1 tran-scription when compared to apoA-I or HDL.ApoA-I has previously been shown to increase ABCA1protein levels in cultured cells and in the liver of apoA-I-injected mice [3,4]. Consistent with these findings,total cell homogenates obtained from cholesterol-loadedfibroblasts incubated for 16 h with 10 μg/mL apoA-Iexhibited higher levels of ABCA1 protein than cells trea-ted with albumin alone or similar levels of HDL (Figure2B). Cells treated with 10 μg/mL tyrHDL showedincreased levels of ABCA1 protein when compared toapoA-I, HDL or BSA treatments at both 2 and 16 h.These results suggest tyrHDL has an enhanced ability tostabilize ABCA1 protein post-translationally when com-pared to apoA-I and control HDL.tyrHDL protects ABCA1 from calpain-dependent and-independent proteolysisCalpain-dependent proteolysis of ABCA1 is mediatedthrough a proline, glutamic acid, serine, and threonine-rich (PEST) sequence [4], and is inhibited by lipid-freeapoA-I but not HDL [3,4]. To determine whethertyrHDL inhibits calpain-dependent proteolysis ofABCA1, cholesterol-loaded fibroblasts were incubatedwith BSA alone or plus apoA-I, HDL or tyrHDL for 2or 16 h, followed by treatment with 0.5 μM μ-calpainfor 20 min [4]. ABCA1 protein degradation by μ-calpainwas observed in BSA and HDL treated fibroblasts atboth time points, with a 50-60% reduction in ABCA1protein when compared to the absence of μ-calpain(Figure 3). No appreciable degradation of ABCA1 bycalpain was observed in apoA-I- or tyrHDL-treated cellsat either time point (Figure 3).Lipid-free apoA-I also increases ABCA1 protein levelsby interfering with calpain-independent proteolysis ofABCA1 [3]. Increased ABCA1 protein seen in the pre-sence of the calpain inhibitor ALLN and BSA alone at16 h was increased further by apoA-I, and most mark-edly by incubation with tyrHDL (Figure 4). These resultsindicate tyrHDL inhibits calpain-dependent and -inde-pendent proteolysis of ABCA1 to a level similar orgreater than lipid-free apoA-I.tyrHDL increases cell surface ABCA1 like lipid-free apoA-IApoA-I-dependent reduction of ABCA1 proteolysisresults in increased cell surface as well as total cellularABCA1 [18]. To determine the effect of tyrHDL on cellsurface ABCA1, cells incubated with either BSA aloneor with addition of 10 μg/mL apoA-I, HDL, or tyrHDLfor 16 h were then biotinylated to label cell surface pro-teins. Total and biotinylated cell surface ABCA1 wereisolated and probed for ABCA1, ERK (a cytoplasmicprotein), and b-actin proteins by immunoblot. Increasedcell surface ABCA1 was seen with both apoA-I andtyrHDL treatment, paralleling the increase in total cellu-lar ABCA1, and to a greater extent than seen with HDLtreatment (Figure 5).tyrHDL forms crosslinks with ABCA1 similar to lipid-freeapoA-ISeveral studies using the chemical cross-linker dithiobis(succinimidyl) propionate (DSP) have suggested apoA-Ican directly bind to ABCA1, whereas HDL does not[15,19,20]. These findings indicate conformational fea-tures present in lipid-free but not lipid-bound apoA-Iallow the lipid-free form to bind to ABCA1. To deter-mine whether tyrHDL might also bind to ABCA1 likelipid-free apoA-I, cholesterol-loaded fibroblasts weretreated with LXR agonist T0901317 to upregulate cellsurface ABCA1 protein expression, and then incubatedwith either BSA alone or with addition of apoA-I, HDL,or tyrHDL for 2 h, followed by incubation with DSP foran additional hour, and separation of whole cell homo-genates for immunoblotting as described [15]. ApoA-Icolocalized with a band corresponding to ABCA1 undernon-reducing conditions (Figure 6). In the presence ofthe reducing agent b-mercaptoethanol, free apoA-I wasobserved at its expected molecular weight of 28 kDa.Like apoA-I, when tyrHDL was DSP crosslinked, a pro-minent anti-apoA-I antibody reactive band colocalizedwith ABCA1 under non-reducing conditions, and wasreleased in the presence of reducing agent to a ladder oflower molecular weight apoA-I-containing bands. Incontrast to control HDL, which showed very minimalbinding to ABCA1, these results suggest conformationalchanges are present in lipid-bound tyrHDL apolipopro-teins that allow them to bind to ABCA1 like lipid-freeapoA-I.tyrHDL undergoes enteropeptidase digestion similar tolipid-free apoA-ITo investigate the possibility that lipid-bound apos ontyrHDL can assume a lipid-free conformation, the sus-ceptibility of tyrHDL to enteropeptidase digestion wascompared to unmodified HDL and apoA-I. Enteropepti-dase cleaves lipid-free apoA-I at Arg188 to form a majorfragment of about 22 kDa, but has no ability to cleavelipid bound apoA-I present either on reconstituted ornative HDL (15). Enteropeptidase treatment of lipid-freeapoA-I converted a majority of the protein to a lowermolecular weight fragment (~22 kDa), while HDL wasresistant to proteolysis (Figure 7). Exposure of tyrHDLHossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 6 of 11ABCA1 b-actin Cholesterol HDL apoAI tyrHDL —-Calpain - + + + + + + + + + + + + + + + + + - - - + - + - + - + - + - + - + - + - - - - + + - - - - - - + + - - - - - - - - - - + + - - - - - - + + - - - - - - - - - - + + - - - - - - + +  0.6     1    1.0    0.4    1.2   1.1   0.5    0.2    0.9  0.9        0.9   0.4   1.2   1.2   1.1   0.6   1.6   1.6     0 h 2 h 16 h Figure 3 tyrHDL protects ABCA1 from calpain-mediated degradation. Cholesterol-loaded fibroblasts pretreated with BSA plus 10 μg/mlapoA-I, HDL, or tyrHDL for 2 h or 16 h were then incubated with or without 0.5 μM μ-calpain following digitonin permeabilization. ABCA1protein was analyzed by 5-15% SDS-PAGE gradient gels followed by immunoblotting. Results of duplicate incubations with apoA-I, HDL, andtyrHDL are shown. The data are representative of three experiments with similar results.ABCA1 b-Actin Cholesterol apoAI HDL tyrHDL ALLN - + + + + + + + + + - - - - - - + + + + - - - + - - - + - - - - - - + - - - + - - - - - - + - - - + 1     1.4   0.8   1.9   2.5   4.3   3.0  5.9 0 h 16 h Figure 4 tyrHDL increases cellular ABCA1 in the presence of ALLN. Cholesterol-loaded fibroblasts were incubated with BSA and 10 μg/mlapoA-I, HDL, or tyrHDL with or without protease inhibitor ALLN (50 μM) to inhibit endogenous calpain for 16 h. ABCA1 was determined byimmunoblotting as in Figure 3. The data are representative of three experiments with similar results.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 7 of 11to enteropeptidase resulted in loss of the majority of a37 kDa band consistent with an apoA-I-A-II monomerheterodimer, both on Coomassie blue-stained SDSPAGE gel and by immunoblotting with anti-apoAI andanti-apoAII. This apoA-I-apoA-II heterodimer has beenidentified as the protein element of tyrHDL that isresponsible for enhanced cholesterol efflux when com-pared to HDL [9]. Nearly all of the apoA-I in this het-erodimer on tyrHDL was digested, whereas the apoAIIcomponent was only partially digested, possibly due toapoAII multimers also running at this molecular weight.The apoAI monomer band of tyrHDL was also mainlyABCA1 Total Cellular  Cell Surface  ERK BSA apoAI HDL tyrHDL BSA apoAI HDL tyrHDL E-Actin         1            1.5           0.5            2.0                 1           11.8          5.0         11.1      Figure 5 tyrHDL increases cell surface ABCA1. Cholesterol-loaded fibroblasts were incubated with BSA alone or with 10 μg/ml apoA-I, HDL ortyrHDL for 16 h. Total cellular ABCA1 protein was determined from whole cell lysates. Cell surface ABCA1 was determined following biotinylationand precipitation of biotinylated cell surface proteins. Total cellular and cell surface samples were run on the same 5-15% SDS-Page gradient gelfollowed by immunoblotting. Data are representative of three separate experiments with similar results.Anti-ABCA1 Anti-apoA1 250 KD 25 KD ȕ-ME: -    +    -     +    -    +     -    +    -    +    -     +     -    +     -    +    BSA   ApoAI    HDL   tyrHDL BSA    ApoAI    HDL   tyrHDL Figure 6 Crosslinking of tyrHDL to ABCA1. Cholesterol-loaded fibroblasts were equilibrated in the presence of 5 μM T0901317 for 24 h toupregulate ABCA1 expression. The cells were then treated with BSA alone or plus 10 μg/ml apoA-I, HDL and tyrHDL for 2 h, washed, and thenincubated with 250 mM of DSP for 1 h. Whole cell lysates were isolated, run on a 5-15% SDS-PAGE gradient gel and probed for the presence ofABCA1 and apoA-I by immunoblot. 5% beta mercaptoethanol (b-ME) was used as a reducing agent to cleave crosslinks formed by DSP. Resultsare representative of three separate experiments with similar results.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 8 of 11digested, to fragments smaller than 22 kDa, suggestingapoAI in tyrHDL is susceptible to enteropeptidase butproduces different fragments than native apoAI (Figure7A and 7B). These results suggest that unlike lipid-bound proteins on the surface of native HDL, the lipid-bound apoAI-AII heterodimer, and also tyrosylatedapoAI, on the surface of tyrHDL are able to assume atleast partial lipid-free conformations, making them sus-ceptible to enteropeptidase digestion.DiscussionThe current study provides several lines of evidence thatoxidation of HDL by peroxidase-generated tyrosyl radicalinduces conformational changes in lipid-bound HDL apo-lipoproteins that allows them to interact with ABCA1 likelipid-free apoA-I. Incubation of ABCA1-expressing humanfibroblasts with tyrHDL increases total cellular and cellsurface ABCA1 and inhibits calpain-dependent and -inde-pendent proteolysis of ABCA1, as also mediated by lipid-free apoA-I but not HDL [3,4]. Additionally, tyrHDL canbe crosslinked to ABCA1 and is susceptible to proteolysisby enteropeptidase like lipid-free apoA-I. These findings,therefore, represent the first demonstration of apoproteinson the surface of spherical HDL particles behaving likelipid-free apoAI and capable of interacting with and stabi-lizing ABCA1 against proteolysis although lipid-bound.The tyrHDL-induced increase in ABCA1 protein is consis-tent with our previously observed ability of tyrHDL tomarkedly deplete ACAT-accessible cholesterol [6],increase cholesterol available for removal by apoA-I [7],and to reduce atherosclerosis development in apoE-defi-cient mice [8] in comparison to native HDL.   -    +   -   +   -   + Enteropeptidase 37kD 25kD 15kD 10kD  ApoAI      HDL     tyrHDL Anti-ApoAI Anti-ApoAII Enteropeptidase   -   +   -   +   -   +  -   +    -    +   -    + 37kD 25kD 15kD 10kD ApoAI     HDL    tyrHDL ApoAI        HDL    tyrHDL A) B) C) Figure 7 Enteropeptidase digestion of tyrHDL. Lipid-free apoA-I, HDL, and tyrHDL were treated with or without enteropeptidase 0.13 U perμg of protein for 6 h at 37°C. Samples were run on 15% SDS-PAGE under reducing conditions and stained with Coomassie Blue (A) ortransferred to nitrocellulose for immunoblotting with anti-apoA-I (B) or anti-apoA-II (C). The data is representative of 5 separate experiments withsimilar results.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 9 of 11Apoproteins on the surface of spherical tyrHDL increaseABCA1 protein level and activityOur findings indicate that tyrHDL increases cholesterolmobilization from cells by interacting directly withABCA1 and enhancing ABCA1 protein stability. Likelipid-free apoA-I, no effect was seen by tyrHDL onABCA1 mRNA levels. Consistent with this, previousstudies demonstrated the lipid extract of tyrHDL had noability to enhance cholesterol mobilization from cellswhen reconstituted with control HDL apoproteins, sug-gesting oxysterols present in tyrHDL lipids were notdriving ABCA1 expression at a transcriptional level [9].Those studies also found the lipid-free protein fractionof tyrHDL had a similar but slightly lower ability todeplete ACAT-accessible cholesterol than equivalentconcentrations of lipid-free proteins from native HDL[9]. tyrHDL proteins reconstituted into discoidal HDLshowed a modestly increased ability to deplete this cho-lesterol pool in comparison to discoidal rHDL madewith native HDL proteins. Only when tyrHDL proteinswere reconstituted onto the surface of spherical HDLwas the full differential activity of tyrHDL compared tonative HDL observed [9]. These results suggest theenhanced activity of tyrHDL versus native HDL is mostclearly seen when tyrHDL proteins are bound to a sphe-rical rather than a discoidal HDL surface. They also sug-gest that it is the particular conformation of tyrHDLapoproteins when present on the surface of sphericalHDL, and not tyrHDL apos that dissociate off the parti-cle surface, that are responsible for this effect.Apoproteins on the surface of tyrHDL appear to assume apartially lipid-free conformationThe combined results suggest HDL apoproteins oxidizedby tyrosyl radical and specifically the apoAI-AII hetero-dimer created by this oxidation binds to the surface ofthe lipoprotein differently than apoproteins on HDL,and that the modifications induced by tyrosyl radicaloxidation produce a partially lipid-free conformation inlipid-bound apoA-I. We propose a possible modelwherein the more lipophilic apoA-II component of theapoAI-AII heterodimers on tyrHDL [9] tethers apoA-Ito the HDL particle surface, allowing apoA-I to beattached to HDL while it also assumes a partially lipid-free conformation. This conformation would facilitatebinding of the tethered apoA1 to ABCA1, and inhibitionof ABCA1 degradation. As a result, exposure of cells totyrHDL causes ABCA1 to be retained on the cell surfaceand protected from calpain-dependent (cell surface) aswell as calpain-independent (intracellular) proteolysis.The increased ABCA1 protein level leads to theenhanced mobilization of excess intracellular cholesterolthat would otherwise be esterified, and makes thischolesterol available for efflux to apoA-I and new HDLparticle generation.The nature of the conformation of tyrHDL apoAI-AIIheterodimers on the surface of tyrHDL requires furtherinvestigation. Particular regions of apoA-I thought to becritical to the apoA-I-ABCA1 interaction have been sug-gested but not yet confirmed [21-23]. The similarities inbinding to and stabilization of ABCA1 between lipid-bound tyrHDL apos and lipid-free apoA-I suggeststyrHDL represents a useful tool to investigate the natureof this apoprotein-ABCA1 interaction critical to the for-mation of new HDL particles.Whether or not peroxidase-generated tyrosyl radicaloxidation of HDL represents a significant mechanism ofHDL oxidation in vivo is an unanswered question, andis not the purpose of these studies. Myeloperoxidase isknown to generate tyrosyl radical as one of its oxidizingproducts in addition to hypochlorous acid [24], andLDL containing products of tyrosyl radical oxidationhave been isolated from human atherosclerotic plaques[25]. The relative importance of tyrosyl radical oxidationof HDL in vivo, which ex vivo appears to enhanceABCA1 activity, as compared to other oxidative modifi-cations of HDL and apoA-I that impair ABCA1-depen-dent cholesterol efflux [26,27], remains to bedetermined. Regardless, the ability of tyrHDL generatedex vivo to enhance ABCA1 activity represents an impor-tant model for the development of peptide therapiesthat retain this ability when in the lipid-bound form.Since nearly all HDL apoproteins are lipid-bound invivo, this signifies a major difference from apoA-I,which only shows ABCA1-enhancing activity whenlipid-free.ConclusionsThe results presented here provide the first demonstra-tion of lipid-bound HDL apolipoproteins capable ofenhancing ABCA1 protein level and activity like lipid-free apoA-I, and an explanation for the previously iden-tified salutary effects of tyrHDL.List of abbreviationsThe abbreviations used are: HDL: high density lipoproteins; apo:apolipoprotein; ABCA1: ATP-binding cassette transporter A1; tyrHDL: tyrosylradical-oxidized HDL; BSA: fatty acid-free bovine serum albumin; DMEM:Dulbecco’s modified Eagle’s medium; ALLN: N-acetyl-Leu-Leu-norleucinal;ACAT: acyl-CoA:cholesterol acyltransferase.AcknowledgementsThis work was funded by CIHR operating grant MOP-12660 to GAF.Author details1Department of Medicine, UBC James Hogg Research Centre, Heart andLung Institute, St. Paul’s Hospital, Vancouver, British Columbia, Canada V6Z1Y6. 2Children’s Hospital Oakland Research Institute, Oakland, CA, USA 94609.Hossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 10 of 11Authors’ contributionsMAH participated in the design of experiments and writing of themanuscript, and performed the majority of the experiments. SN performedthe initial experiments suggesting an interaction of tyrHDL with ABCA1. TCperformed experiments for the study and assisted with figure generationand writing of the manuscript. MNO provided intellectual input andassistance with the writing of the manuscript. GAF designed theexperiments, oversaw the project and data analysis, and wrote most of themanuscript. All authors read and approved the final manuscript.Received: 21 October 2011 Accepted: 16 January 2012Published: 16 January 2012References1. Francis GA: The complexity of HDL. Biochim Biophys Acta 2010,1801:1286-1293.2. Oram JF, Heinecke JW: ATP-binding cassette transporter A1: a cellcholesterol exporter that protects against cardiovascular disease. PhysiolRev 2005, 85:1343-1372.3. Arakawa R, Yokoyama S: Helical apolipoproteins stabilize ATP-bindingcassette transporter A1 by protecting it from thiol protease-mediateddegradation. J Biol Chem 2002, 277:22426-22429.4. Wang N, Chen W, Linsel-Nitschke P, Martinez LO, Agerholm-Larsen B,Silver DL, Tall AR: A PEST sequence in ABCA1 regulates degradation bycalpain protease and stabilization of ABCA1 by apoA-I. J Clin Invest 2003,111(1):99-107.5. Denis M, Haidar B, Marcil M, Bouvier M, Krimbou L, Genest J Jr: Molecularand cellular physiology of apolipoprotein A-I lipidation by the ATP-binding cassette transporter A1 (ABCA1). J Biol Chem 2004,279(9):7384-7394.6. Francis GA, Mendez AJ, Bierman EL, Heinecke JW: Oxidative tyrosylation ofhigh density lipoprotein by peroxidase enhances cholesterol removalfrom cultured fibroblasts and macrophage foam cells. Proc Natl Acad SciUSA 1993, 90(14):6631-6635.7. Francis GA, Oram JF, Heinecke JW, Bierman EL: Oxidative tyrosylation ofHDL enhances the depletion of cellular cholesteryl esters by amechanism independent of passive sterol desorption. Biochemistry 1996,35:15188-15197.8. Macdonald DL, Terry TL, Agellon LB, Nation PN, Francis GA: Administrationof tyrosyl radical-oxidized HDL inhibits the development ofatherosclerosis in apolipoprotein E-deficient mice. Arterioscler ThrombVasc Biol 2003, 23(9):1583-1588.9. Wang WQ, Merriam DL, Moses AS, Francis GA: Enhanced cholesterol effluxby tyrosyl radical-oxidized high density lipoprotein is mediated byapolipoprotein AI-AII heterodimers. J Biol Chem 1998, 273:17391-17398.10. Chung BH, Wilkinson T, Geer JC, Segrest JP: Preparative and quantitativeisolation of plasma lipoproteins: rapid, single discontinuous densitygradient ultracentrifugation in a vertical rotor. J Lipid Res 1980,21(3):284-291.11. Boadu E, Choi HY, Lee DW, Waddington EI, Chan T, Asztalos B, Vance JE,Chan A, Castro G, Francis GA: Correction of apolipoprotein A-I-mediatedlipid efflux and high density lipoprotein particle formation in humanNiemann-Pick type C disease fibroblasts. J Biol Chem 2006,281:37081-37090.12. Inazu A, Jiang XC, Haraki T, Yagi K, Kamon N, Koizumi J, Mabuchi H,Takeda R, Takata K, Moriyama Y, et al: Genetic cholesteryl ester transferprotein deficiency caused by two prevalent mutations as a majordeterminant of increased levels of high density lipoprotein cholesterol.The Journal of Clinical Investigation 1994, 94:1872-1882.13. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ: Protein measurement withthe Folin phenol reagent. J Biol Chem 1951, 193:265-275.14. Feng B, Tabas I: ABCA1-mediated cholesterol efflux is defective in freecholesterol-loaded macrophages. Mechanism involves enhanced ABCA1degradation in a process requiring full NPC1 activity. J Biol Chem 2002,277:43271-43280.15. Wang N, Silver DL, Costet P, Tall AR: Specific binding of ApoA-I, enhancedcholesterol efflux, and altered plasma membrane morphology in cellsexpressing ABC1. J Biol Chem 2000, 275:33053-33058.16. Safi W, Maiorano JN, Davidson WS: A proteolytic method fordistinguishing between lipid-free and lipid-bound apolipoprotein A-I. JLipid Res 2001, 42(5):864-872.17. Francis GA, Knopp RH, Oram JF: Defective removal of cellular cholesteroland phospholipids by apolipoprotein A-I in Tangier Disease. J Clin Invest1995, 96:78-87.18. Martinez LO, Agerholm-Larsen B, Wang N, Chen W, Tall AR:Phosphorylation of a pest sequence in ABCA1 promotes calpaindegradation and is reversed by ApoA-I. J Biol Chem 2003,278(39):37368-37374.19. Fitzgerald ML, Morris AL, Chroni A, Mendez AJ, Zannis VI, Freeman MW:ABCA1 and amphipathic apolipoproteins form high-affinity molecularcomplexes required for cholesterol efflux. J Lipid Res 2004, 45(2):287-294.20. Oram JF, Lawn RM, Garvin MR, Wade DP: ABCA1 is the cAMP-inducibleapolipoprotein receptor that mediates cholesterol secretion frommacrophages. J Biol Chem 2000, 275:34508-34511.21. Panagotopulos SE, Witting SR, Horace EM, Hui DY, Maiorano JN,Davidson WS: The role of apolipoprotein A-I helix 10 in apolipoprotein-mediated cholesterol efflux via the ATP-binding cassette transporterABCA1. J Biol Chem 2002, 277(42):39477-39484.22. Natarajan P, Forte TM, Chu B, Phillips MC, Oram JF, Bielicki JK: Identificationof an apolipoprotein A-I structural element that mediates cellularcholesterol efflux and stabilizes ATP binding cassette transporter A1. JBiol Chem 2004, 279(23):24044-24052.23. Vedhachalam C, Chetty PS, Nickel M, Dhanasekaran P, Lund-Katz S,Rothblat GH, Phillips MC: Influence of apolipoprotein (Apo) A-I structureon nascent high density lipoprotein (HDL) particle size distribution. J BiolChem 2010, 285(42):31965-31973.24. Heinecke JW, Li W, Francis GA, Goldstein JA: Tyrosyl radical generated bymyeloperoxidase catalyzes the oxidative cross-linking of proteins. Journalof Clinical Investigation 1993, 91:2866-2872.25. Leeuwenburgh C, Rasmussen JE, Hsu FF, Mueller DM, Pennathur S,Heinecke JW: Mass spectrometric quantification of markers for proteinoxidation by tyrosyl radical, copper, and hydroxyl radical in low densitylipoprotein isolated from human atherosclerotic plaques. Journal ofBiological Chemistry 1997, 272(6):3520-3526.26. Bergt C, Pennathur S, Fu X, Byun J, O’Brien K, McDonald TO, Singh P,Anantharamaiah GM, Chait A, Brunzell J, et al: The myeloperoxidaseproduct hypochlorous acid oxidizes HDL in the human artery wall andimpairs ABCA1-dependent cholesterol transport. Proc Natl Acad Sci USA2004, 101(35):13032-13037.27. Zheng L, Nukuna B, Brennan ML, Sun M, Goormastic M, Settle M, Schmitt D,Fu X, Thomson L, Fox PL, et al: Apolipoprotein A-I is a selective target formyeloperoxidase-catalyzed oxidation and functional impairment insubjects with cardiovascular disease. J Clin Invest 2004, 114(4):529-541.doi:10.1186/1471-2091-13-1Cite this article as: Hossain et al.: Lipid-bound apolipoproteins in tyrosylradical-oxidized HDL stabilize ABCA1 like lipid-free apolipoprotein A-I.BMC Biochemistry 2012 13:1.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitHossain et al. BMC Biochemistry 2012, 13:1http://www.biomedcentral.com/1471-2091/13/1Page 11 of 11


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items