Partial e haracterizat ion of I i poprotei ns containing apo [ a ] in human atherosclerotic lesions

Previously we quantified the amounts of immunoreactive apo[ a] found in human atherosclerotic lesions extracted sequentially with phosphate-buffered saline (PBS) and guanidine hydrochloride (GuHCI). In this study we have attempted to characterize lipoproteins containing apo[a] in such PBS and GuHCl fractions, obtained from autopsy samples, in order to eventually determine their structure-function relationships critical for evaluating the mechanisms that make them atherogenic. Apo[a] in the PBS extracts migrated slightly ahead of plasma Lp[a] on agarose electrophoresis. Although apo[a] in extracts showed the same isoforms as in plasma in SDS-PAGE, it was also highly fragmented. When a d < 1.10 g/ml ultracentrifugation fraction of the PBS extract was subjected to gel filtration, a major part of the immunoreactive apo[a] in this fraction co-isolated with plasma Lp[a]. When the Lp[a]-sized fraction was further separated by density gradient ultracentrifugation, a subpopulation was isolated containing apo[a] in the 1.06 < d < 1.08 g/ml density range that was free of lesion-derived low density lipoprotein (LDL) (A-LDL). This fraction contained immunoreactive apo[a] and apoB, had a total cholesterol to protein ratio of about 1, and demonstrated increases in fluorescence (360 ex/430 em) and conjugated dienes that were even greater than values obtained for the corresponding A-LDL sample. The void volume fraction following gel exclusion chromatography of the d < 1.10 g/ml fractions contained both apo[a] and apoB that comigrated on nondenaturing PAGE, suggesting that they were present on the same particle. Apo[a] in GuHCl extracts comigrated with plasma Lp[a] on agarose electrophoresis and contained apo[a] isoforms of similar molecular weights as those found in corresponding plasma samples. When the GuHCl extract was subjected directly to gel filtration in the presence of 6 M GuHCI, two included peaks of apo[a] immunoreactivity were present, one eluting slightly ahead of plasma Lp[a], the other slightly ahead of plasma LDL. Collectively, these data indicate that apo[a] is present in human atherosclerotic lesions in forms resembling intact but oxidized plasma Lp[a], as larger particles possibly representing Lp[a] complexed to itself or other plaque components, and as slightly smaller particles possibly representing degraded Lp[a].-Hoff, H. F., J. O'Neil, and A. Yashiro. Partial characterization of lipoproteins containing apo[a] in human atherosclerotic lesions. J. Lipid Res. 1993. 34: 789-798. Supplementary key words chromatography immunoblotting Lp[a] apoB arteries plaques affinity Lp[a] is a cholesterol-rich plasma lipoprotein that shares many structural and chemical properties with plasma LDL (1-3). It has attracted attention in recent years, as numerous clinical-chemical correlative studies have indicated that Lp[a] is an independent risk factor for cardiovascular diseases. Elevations of plasma Lp[a] were shown to be associated with increased atherosclerosis in the coronary arteries leading to myocardial infarction (4-7), in intraand extracranial arteries leading to stroke (7), and in saphenous vein grafts after coronary artery bypass surgery (8). Although both LDL and Lp[a] share apoB-100, and have similar relative lipid compositions, Lp[a] also possesses a unique protein designated apo[a], whose structural and chemical characteristics have been extensively described (1-3, 9, 10). In an effort to better understand the atherogenicity of Lp[a], we recently undertook a study to quantify the amounts of Lp[a] accumulating in human atherosclerotic lesions using immunoreactive apo[a] as a measure of Lp[ a] accumulation (11). We also measured the accumulation of apo[a] and of apoB as indicators of the accumulation of Lp[a] and LDL in plaques, as had also been done by Cushing et al. (12) and Rath et al. (13). All these studies showed that plaque and plasma apo[a] contents correlated positively in contrast to plaque and plasma apoB. When normalized to equivalent plasma concentrations of Lp[a] and LDL, both we (11) and Cushing et al. (12) showed that apo[a] accumulated in plaques to a greater degree than apoB. This roughly translated into a greater accumulation of Lp[a] than LDL for equivalent plasma concentrations. We also extracted lesions sequenAbbreviations: Lp[a],lipoprotein[a]; LDL, low density lipoprotein; apola], apolipoprotein[a]; apoB, apolipoprotein B; PBS, phosphatebuffered saline; BSA, bovine serum albumin; REM, relative electrophoretic mobility; PVDF, polyvinylidene difluoride; BCA, bicinchoninic acid; TEM, transmission electron microscopy; RIA, radioimmunoassay; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; GuHCI, guanidine hydrochloride. 'To whom correspondence and reprint requests should be addressed. Journal of Lipid Research Volume 34, 1993 789 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 1, 2013 w w w .j.org D ow nladed fom tially with phosphate-buffered saline (PBS) and 6 M GuHCl. We found that, although on an absolute scale apoB greatly exceeded apo[a] in both the PBS extract as well as in the GuHCl extract, when assessing the relative distributions in these two fractions, major differences were found. Although 68% of the total apoB in lesions was extractable with PBS, only 18% of the total apo[a] in lesions was extractable with PBS, the remainder being extractable with GuHCl. This result suggested that Lp[a] was more tightly bound in lesions than LDL. Because characterization of the structure-function properties of Lp[a] will be necessary to eventually elucidate the mechanisms responsible for the atherogenicity of Lp[a], we initiated a study to chemically and structurally characterize lipoproteins containing immunoreactive apo[a] in human atherosclerotic lesions. We now report initial results on the partial characterization of such particles. Our results indicate that modified forms of intact Lp[a]-like particles are present in lesions, some chemically altered such as by oxidation, some possibly degraded, and some possibly aggregated. MATERIALS AND METHODS Carrier-free N a l T was obtained from ICN Pharmaceuticals, Inc. (Irvine, CA). Guanidine HC1, aprotinin, leupeptin, pepstatin, vitamin E, BHT, EDTA, NaBr, and NaN3 were purchased from Sigma Chemical Co. (St. Louis, MO). Plasminogen and bovine serum albumin (fatty acid-free) were purchased from Boehringer Mannheim (Indianapolis, IN). CNBr-activated Sepharose used for affinity chromatography, Sephacryl 400 H R used for gel filtration, and premade 4-15% gradient polyacrylamide Phast Gel for SDS PAGE were obtained from Pharmacia LKB Biotech Inc. (Piscataway, NJ). SDS gradient polyacrylamide gels (3-8%) were prepared using diacrylylpiperazine (Integrated Separation Systems, Hyde Park, MA) as crosslinker. Pre-made 1% agarose gels and Fat Red 7B were purchased from Corning (Palo Alto, CA). Nitrocellulose membranes were obtained from Schleicher and Schuell, (Keene, NH), Immobilon-P from Millipore Corp. (Bedford, MA), and Spectrapor 2 and 3 dialysis tubing from Spectrum Medical Industries Inc. (Los Angeles, CA). Goat anti-human LDL and goat antihuman Lp[a], normal goat and rabbit sera, and rabbit anti-goat IgG were purchased from Bethyl Laboratories (West Montgomery, TX). Gold-conjugated rabbit antigoat IgG, as well as the silver enhancement kit, were purchased from Amersham (Arlington Heights, IL). Silver stain for PAGE was obtained from Daiichi (Tokyo, Japan). Bicinchoninic acid (BCA) protein kit was from Pierce (Rockford, IL), while the microprotein kit Quantagold was from Diversified Biotech (Newton Centre, MA). Isolation of plasma LDL and Lp[a] LDL was isolated from fresh plasma, obtained from the Cleveland Clinic Blood Bank, by sequential ultracentrifugation as a 1.019 < d < 1.063 g/ml fraction using the procedure of Hatch and Lees (14) and stored in 0.15 M NaCl containing 0.5 mM EDTA, pH 8.5. Lp[a] was isolated by plasmaphoresis from fresh plasma of healthy donors with Lp[a] levels of > 30 mg/dl. Aprotinin (100 K.I.U./ml), leupeptin (0.5 pg/ml), pepstatin (0.7 pg/ml), EDTA (1 mg/ml), and NaN3 (0.01%) were added immediately after plasmaphoresis. Lp[a] was then isolated by sequential ultracentrifugation as a 1.050 < d < 1.120 g/ml fraction, concentrated by dialyzing against 30% PEG (MW 20,000) using a Spectropor 3 membrane (3500 MW) in 10 mM Tris, pH 7.2, containing 1 mg/ml E m A . It was further separated by gel filtration chromatography on Sephacryl 400 H R (2.5 x 95 cm) in 10 mM Tris, pH 7.2, containing 1 mg/ml EDTA and 0.01% NaN3. Only those fractions completely free of LDL and HDL were pooled. Purity of Lp[a] was evaluated by using the PhastGel Automated System (Pharmacia-LKB Biotech) and premade SDS polyacrylamide 4-15 % gradient under nonreducing conditions followed by silver staining for visualization of protein. Lp[a] was stored in 10 mM Tris, pH 7.2, containing 1 mg/ml EDTA. Isolation of Lp[a] from aortic plaques Lp[a] was extracted from aortic plaques as described earlier (11) except that human aortas with advanced atherosclerotic plaques and their corresponding plasma were obtained at autopsy within 12 h of death and only from lesions derived from individuals with plasma levels > 20 mg/dl Lp[a]. The tunica intima was stripped from the underlying media at a natural cleavage plane. The data shown in this report were obtained from six representative cases, although an additional eight autopsy cases and four surgery cases were used to compare the relative electrophoretic mobility (REM) on agarose electrophoresis of LDL-like and Lp[a]-like fractions in plaque extracts and corresponding plasma. A loosely bound lipoprotein fraction was isolated from minces of the tunica intima by extraction at 4OC with PBS, (0.1 M phosphate, 0.3 mM EDTA, 0.15 M NaCl, pH 7.4), at a ratio of 1 g wet weight tissue to 5 ml PBS for 18 h at 4OC (11). Additives were included as described for Lp[a] isolated from plasma except that 20 pM vitamin E and 50 pg/ml gentamicin were added instead of NaN3. The resulting suspensions were centrifuged at 20,000 g for 1 h at 4°C to obtain a supernatant fraction. The remaining pellet fraction was extracted at 4OC for 6 days with 6 M GuHCl at a ratio of 1 g tissue wet weight to 3 ml GuHCl which included all the additives present in the PBS buffer. The supernatant fraction of the GuHCl extract after centrifugation at 20,000 g for 1 h at 4OC was then chromatographed on 790 Journal of Lipid Research Volume 34, 1993 at P E N N S T A T E U N IV E R S IT Y , on F ebuary 1, 2013 w w w .j.org D ow nladed fom Sephacryl 400 HR (1.5 x 95 cm) in the presence of 6 M GuHCI. A portion of the supernatant fraction of PRS extracts was subjected to ultracentrifu.gation at 15OC for 18 h at 300,000 gat a density of d 1.10 glml. The floating fraction of lipoproteins was then subjected to gel filtration on a Sephacryl 400 HR column of dimensions 2.5 x 95 cm. An Lp( a]-sized fraction containing immunoreactive apola] was then further fractionated by density gradient ultracentrifugation according to R e d p v e et al. (15) with minor modifications. Fractions in the 1.06 < d < 1.08 glml density range were monitored for increases in conjugated dienes and fluorescence (360 ed430 em) as described previously (16). In select experiments a GuHCl tissue extract was dialyzed against 20 mM phosphate, 0.15 mM NaCI, 0.3 mki EIXA, pH 7.4, for 12 h and then applied to anti-apoR-Sepharose at a ratio of 1 ml of extract to 0.5 ml of anti-apoR-Sepharose. The anti-apoB used for this chromatomphy was derived from p a t antiLpIa] as described earlier (11). Elution was performed usins saline EDTA adjusted to pH 11 with NH,OH as previously described (17). Protein content of lipoproteins was measured by the RCA assay as described by Smith et al. (18) except that a 60-min. 60°C heating step was used with RSA as a standard. A microprotein assay was also used to measure bound and eluted lipoprotein after affinity chromatography (Procedure Sheet, Diversified Riotech, Newton Centre, MA). Transmission electron microscopy (TEM) was performed on negatively stained lipoprotein samples as reported earlier (19). Measurements of conjugated dienes and fluorescence at 360 ex1430 em were performed as described previously (16, 20). Polyclonal antibodies raised in p a t s against human plasma Lpla] and plasma LDL were obtained as described earlier (11) and used for radioimmunoassays (RIAs) and immunoblotting for apo(a] and apoR. Recause anti-Lpla] contained both anti-apola] and antiapoR, we quantitatively removed all anti-apoR by applying an I& fraction of anti-Lpla] on an LDL-Scpharose column as reported earlier (17). Iodination of LDL and Lpla] for RIAs was performed using the iodine monochloride (21) method. A brief description of the RIAs used in this study for Lp(a] and LDL has been previously made by Pepin, O'Neil, and Hoff (11). Electrophoretic mobility of Lp( a] relative to LDL (REM) was determined on premade 1% agarose gels following the manufacturer's instructions except that electrophoresis was performed at 90 V for 75 min. RSA (1%) was added to ensure reproducible migration distances. Gels were stained for lipid using 0.025% Fat Red 7R in 60% methanol. Immunoblotting was performed after transferring lipoproteins to nitrocellulose by simple diffusion which required only 2-4 h. After blocking nitrocellulose membranes with nonfat kDa