Metabolism of Human Intermediate and Very Low Density Lipoprotein Subtractions from Normal and Dysbetalipoproteinemic Plasma

Transcription

1 Metabolism of Human Intermediate and Very Low Density Lipoprotein Subtractions from Normal and Dysbetalipoproteinemic Plasma In Vivo Studies in the Rat Thomas A. Musliner, Kathleen M. McVicker, Joe F. losefa, and Ronald M. Krauss Downloaded from by on September 6, 2018 Subtractions of radlolodinated d < gml llpoproteins were Isolated by nonequillbrlum density gradient ultracentrlfugatlon (DGU) from normal and dysbetalipoproteinemic human plasma and were Injected into rats. Size and density (d) of lipoprotein products formed over 8 hours were assessed by gradient gel electrophoresis and equilibrium DGU, respectively. Subtractions containing a subspecies of very low density lipoprotelns (VLDL) of particle diameter > 350 A were cleared rapidly from the plasma and formed only small amounts of low density lipoprotelns (LDL). Fractions containing VLDL subspecies of smaller diameter (300 to 350 A) were cleared much more slowly, and formed greater amounts of a discrete LDL product with the characteristics of human LDL-II (peak particle diameter 255 to 265 A, d = to gml). A similar LDL product was formed from subtractions containing intermediate density lipoprotelns (IDL). Cholesterol-enriched subspecies within the smaller, denser portion of the IDL spectrum, however, yielded two additional products. One had size and density characteristic of the major human LDL-I subclass reported previously (265 to 275 A, d = to gml), while the other was yet larger (275 to 285 A) and overlapped normal IDL In size and density. In dysbetalipoproteinemic plasma, the metabolic precursors of the largest product were shifted from the IDL to the small VLDL (B-VLDL) particle distribution. Since B-VLDL are known to predispose to accelerated atherosclerosis in dysbetallpoproteinemla, it may be that metabollcally homologous cholesterol-enriched IDL subspecies In other subjects have similar atherogenlc properties. (Arteriosclerosis 7: , JulyAugust 1987) Low density lipoproteins (LDL) are believed to originate from the sequential catabolism of very low density lipoprotein (VLDL) precursors. As VLDL trtglyceride is hydrolyzed by lipoprotein lipase, smaller, cholesterol-enriched intermediate density lipoproteins (IDL) are formed and are subsequently converted to LDL by mechanisms that are poorly understood. In normal humans, nearly all of the protein moiety of plasma LDL derives from VLDL through IDL intermediates, 1 " 3 although in patients with familial hypercholesterolemia 4 ' 5 and in some primate species 6 independent production of LDL has been reported. A portion of the apolipoprotein (apo) B contained in VLDL, moreover, is removed from the circulation before conversion to LDL. This is the fate of 10% to 60% of the VLDL apo B in normal human subjects and a still greater proportion of the VLDL apo B in hypertriglyceridemic subjects 2 * 3S>7 " 10 and in From the Clinical Investigation Center, Naval Hospital, Oakland and the Donner Laboratory, Lawrence Berkeley Laboratory, University of California, Berkeley, California. This work was performed under the United States Navy Clinical Investigation Program study number , and was also supported in part by NIH Grant HL All opinions are those of the authors and not necessarily those of the United States Navy. Address for reprints: Thomas A. Musliner, M.D., Lawrence Berkeley Laboratory, 1 Cyclotron Road, 1-213, Berkeley, California Received January 27,1986; revision accepted March 26,1987. some animal species. 11 ' 12 This direct catabolism of VLDL or VLDL remnants is thought to occur in the liver Human VLDL 13 " 18 and LDL 19 " 24 are known to represent heterogeneous populations of particles with respect to size, physical properties, and chemical composition. The techniques of density gradient ultracentrifugation (DGU) and gradient gel electrophoresis (GGE), moreover, have revealed multiple discrete subspecies within the conventional LDL density range (1.019 to gml) Recent studies of human d < gml lipoproteins in our laboratory 25 have also demonstrated multiple subspecies in the IDL (1.006<d< gml) as well as the VLDL (d< gml) density ranges. The existence of this heterogeneity raises the question whether subtractions of VLDL and IDL differ in their capacity to form LDL and, further, whether they give rise to the same or different LDL products. In the studies reported here, we have investigated these questions using a new electrophoretic "blotting" technique which exploits the resolving power of GGE and permits accurate measurement of size transformations of radiolabelled lipoprotein subspecies as they are metabolized in vivo. Series of radiolabelled d< gml lipoprotein subtractions of sequentially smaller particle diameter were isolated by nonequilibrium DGU from the plasma of normolipidemic subjects and from a patient with dysbetalipoproteinemia. 408

2 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al. 409 Downloaded from by on September 6, 2018 The clearance and metabolic transformations of these subfractions were then studied in the rat, a species in which the presence of an active lipolytic pathway 12 and the absence of significant cholesteryl estertriglyceride exchange activity 26 makes it possible to identify direct precursor-product relationships in an in vivo model. Despite potential interspecies differences in lipoprotein metabolism, our findings in rats support the view that specific precursors are capable of generating different LDL products. Methods Isolation and lodlnatlon of Llpoprotelns Blood was collected from normolipidemic and hyperlipidemic adult male volunteers, after a 12 to 14-hour fast, into tubes containing disodium ethylene diaminetetraacetate (EDTA) at a final concentration of 1.5 mgml. The tubes were immediately placed on ice and plasma was separated at 4 C by low-speed centrifugation. The plasma total cholesterol of the three normal donors ranged from 179 to 251 mgdl and the triglycerides, from 59 to 164 mgdl. The patient with dysbetalipoproteinemia had a ratio of VLDL cholesterol to plasma triglycerides of >0.30, 3-VLDL demonstrable on agarose electrophoresis of the d < gml lipoproteins, and the apo E22 phenotype by isoelectric focusing. A cysteine residue at position 158 of the amino acid sequence of his apo E had previously been demonstrated. 27 At the time of the study, this patient was taking no lipid-altering drug, and his plasma cholesterol ranged from 268 to 289 mgdl and his triglyceride, from 123 to 279 mgdl. The d < gml lipoproteins were isolated from 30 to 120 ml of plasma by preparative ultracentrifugation under standard conditions. 28 Where necessary, lipoprotein solutions were concentrated by dry dialysis using Aquacide I-A (Calbiochem, La Jolla, California) or by ultrafiltration using Amicon (Danvers, Massachusetts) XM50 filters. Lipoproteins (5 to 8 mg protein) were iodinated using the iodine monochloride method One to three atoms of iodine were substituted for each mole of protein, assuming a molecular weight of 350,000 for apo VLDL. Free iodine was removed by gel chromatography with Sephadex G25, followed by sequential dialysis against 0.1 M potassium iodide in normal saline (ph 7.4), normal saline containing 0.01% EDTA (ph 7.4), and normal saline alone (ph 7.4), with seven changes over a period of 18 hours. Approximately 95% of the label remained bound to protein after delipidation. Separation of VLDL and IDL Subfractions The iodinated d< gml lipoproteins (0.5 to 1.0 mgml protein, specific activity 28 to 62 ficumg protein) were subjected to nonequilibrium DGU, as described in detail elsewhere. 25 Briefly, the density of the lipoprotein solution was adjusted to 1.21 gml and a 4.5 ml sample was placed in a 916 x 3 Vb-inch Ultraclear tube (Beckman Instruments, Mountain View, California) and was sequentially overlayered with 3.0, 3.0, and 1.5 ml, respectively, of solutions of densities 1.020, 1.010, and gml. The tubes were centrifuged for 6 hours in a Beckman SW41 rotor at 40,000 rpm at 17 C. The contents of the tube were withdrawn by pipetting four consecutive 1.0 ml fractions followed by four 0.5 ml fractions, designated 1 through 8 in order from the top of the tube. Frequently, the first and second 0.5 ml portions of fraction 1 were pipetted separately and designated 1a and 1b, respectively. In Vivo Studies Male Sprague-Dawley albino rats (Simonson Laboratories, Gilroy, California) weighing between 280 and 350 grams were used for all studies. The rats had free access to water and standard Purina lab chow. Iodinated VLDL or IDL subfractions (50 to 100 fig of protein) from individual donors were injected into an exposed jugular vein with the animal under light methoxyflurane anesthesia. Samples were dialyzed against normal saline and were diluted before infusion so that each rat received the same quantity of 125 l-apo B in an identical volume (ranging from 0.5 to 1.2 ml). Blood samples of 0.3 to 0.5 ml were withdrawn from the opposite jugular vein for analysis. At 8 hours, the rats were exsanguinated from the inferior vena cava or abdominal aorta. Blood samples were anticoagulated with EDTA at a final concentration of 1.5 mgml of blood. Tubes were immediately placed on ice and plasma was separated by low-speed centrifugation at 4 C. Determination of Apo B Radioactivity In Plasma and Lipoprotein Fractions The isopropanol precipitation method of Egusa et al. 31 was employed to quantify 125 l-apo B radioactivity. In brief, plasma or lipoprotein samples in a volume of 1 ml were mixed well with an equal volume of isopropanol and were left overnight at 4 C. The precipitate was isolated by lowspeed centrifugation and was washed once with 50% (vol vol) isopropanolsaline. The pellet and supernatant radioactivity were counted in a Packard gamma scintillation counter, and the percent 125 l-apo B was calculated as the ratio of precipitable 125 ltotal 125 lx100. Control experiments demonstrated that identical results were obtained whether the 125 l-labelled VLDL or LDL were precipitated in ultracentrifugally separated lipoprotein fractions as originally described by Egusa et al. 31 or in mixtures with plasma. When isopropanol precipitation was performed on lipoprotein fractions containing less than 100 jigml of protein, 50 fi\ of unlabelled normal human plasma or normal human LDL were added to increase the size of the pellet and minimize nonspecific losses. The percent 125 l-apo B remaining in rat plasma was calculated relative to the 12S I- apo B present in a sample drawn between 60 and 90 seconds after injection of the tracer. While this method may underestimate the decay of very rapidly cleared lipoproteins, errors associated with the calculation of rat plasma volumes are avoided. Polyacrylamlde Gradient Gel Electrophoresis Electrophoresis of plasma samples or lipoprotein fractions was performed on 2% to 16% polyacrylamide gradient gels (Pharmacia PAA 216, Piscataway, New Jersey) for 24 hours at 125 V in Tris (0.09 M)boric acid (0.08

3 410 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST 1987 Downloaded from by on September 6, 2018 M)Na 2 EDTA (0.003 M) buffer (ph 8.3) as described elsewhere. 24 Gels were stained for proteins in a solution containing 0.1% Coomassie brilliant blue R-250,50% ethanol, and 9% acetic acid (volvol), or for lipid with oil red O or Sudan black in 60% ethanol at 55 C. 24 Gels were scanned at 555 nm for Coomassie blue staining or 530 nm for oil red O staining, using a Transidyne RFT densitometer or at 633 nm using an LKB Instruments Ultroscan Laser Densitometer. Migration distances for each absorbance peak were determined and the molecular diameters corresponding to each peak were calculated from a calibration curve generated from the migration distances of standards of known diameter. These included carboxylated latex beads (Dow Chemical), thyroglobulin dimer, thyroglobulin, and apoferritin (Pharmacia) with molecular diameters of 380 ± 7.5 A, 236 A, 170 A, and 122 A, respectively. Gradient gel electrophoresis in the presence of sodium dodecyl sulfate (SDS) was performed using 4% to 30% polyacrylamide gradient gels (Pharmacia PAA 430, Piscataway, New Jersey) in a buffer system containing 0.04 M sodium acetate, 2 mm EDTA, and 0.2% SDS (ph 7.4). The sample buffer consisted of 10 mm Tris, 1 mm EDTA (ph 8.0) containing 2.5% SDS and 5% (3-mercaptoethanol. LJpoproteins were delipidated before electrophoresis using methanol and diethylether as described by Herbert et al., 32 were dissolved in the sample buffer, and were heated for 5 minutes at 100 C before application to the gel. Electrophoresis was carried out for approximately 4 hours at 150 V, 12 C. Gels were stained for protein as described above. Electrophoretlc Blotting and Autoradiography Polyacrylamide gradient gels were electrophoretically blotted onto nitrocellulose paper, as described in greater detail elsewhere. 33 In brief, following the initial electrophoresis, the gradient gel was transferred to 0.05% acetic acid (volvol) and was gently shaken at room temperature for 90 to 120 minutes with three changes of solution. The gel was then placed in a "sandwich" between nitrocellulose paper and Whatman no. 1 filter paper soaked in blotting buffer (2% Triton X-100, 0.05% acetic acid) and was placed in a Pharmacia GD 4 destainer containing the same buffer, with the nitrocellulose paper facing the cathode. Electrophoresis was performed for 24 hours at 24 V, with the chamber temperature maintained at 19 to 20 C. The nitrocellulose blot was then rinsed in 0.05% acetic acid and dried. The boundaries of the gel, as well as the visualized thyroglobulin and ferritin standards, were marked with traces of radioactive protein, after which the blots were subjected to autoradiography at -70 C using Kodak XAR-5 film. As shown elsewhere in detail, 34 GGE lipoprotein banding was well preserved following electrophoretic transfer of apolipoproteins to nitrocellulose, permitting resolution of lipoprotein subspecies by autoradiography. Standard peaks were marked by aligning the autoradiograms with the original gel stained with Coomassie blue and re-equilibrated with the blotting buffer. Autoradiograms were scanned at 555 nm using a Transidyne RFT densitometer. Particle diameters corresponding to each peak were calculated from a calibration curve as described above. Equilibrium Density Gradient Ultracentrlfugatlon The products of lipoprotein catabolism in the rat were analyzed by equilibrium DGU, as previously described. 21 In brief, the d = to gml LDL fraction was isolated by preparative ultracentrifugation and dialyzed to d = gml. The LDL (2 ml) were layered above a NaBr solution of d = gml (2.5 ml) in a cellulose nitrate tube. This was overlayered with 2.5 ml of a NaBr solution of d = gml. Tubes were centrifuged to equilibrium at 40,000 rpm in a Beckman SW 45 rotor at 22 C. The tube contents were then withdrawn by sequentially pipetting the top 0.5 ml, six 1-ml fractions, and the bottom 0.5 ml. A background salt tube was included for determination of the salt densities of the pipetted fractions. Results Characterization of Radlolodlnated VLDL and IDL Subtractions Nonequilibrium DGU of d< gml lipoprotein fractions provided a series of fractions of progressively smaller mean particle diameter. The chemical and physical properties of the VLDL and IDL subspecies contained in these fractions from normolipidemic and dysbetalipoproteinemic subjects have been previously described in detail. 25 Representative gradient gel electrophoretograms of nonequilibrium DGU fractions 1a through 8 from one of the normolipidemic subjects (total cholesterol 179 mgdl, triglyceride 134 mgdl) whose plasma was used in subsequent experiments are shown in Figure 1A. Fraction 1a included nearly all of the VLDL of particle diameter >350 A. A smaller subspecies of VLDL was also found in small amounts in fraction 1, and extended as a broad distribution of particles through fraction 4. The particles of IDL density contained in fractions 5 to 7, as previously described, were seen to include two subspecies of overlapping buoyant density. The predominant IDL species in this subject displayed a peak particle diameter extending from 291 A in fraction 5 to 283 A in fraction 7. Lipoprotein subspecies in corresponding fractions from all of the normal donors were within 4 A in size of those shown for the subject in Figure 1 A, and variation in particle diameters for each subfraction from individual subjects at differing sampling times during the study were no greater than 5 A. Lipoproteins of similar size were also found in the corresponding nonequilibrium DGU fractions from the plasma of the patient with dysbetalipoproteinemia (total cholesterol 274 mgdl, triglycerides 279 mgdl) (Figure 1B) used in subsequent experiments, with differences from the normal pattern as previously described. 25 Agarose electrophoresis showed the presence of a 2 -VLDL only in fraction 1 a, while I3-VLDL was found in fractions 1a through 4 (data not shown). The d < gml lipoproteins from these two subjects were radioiodinated, and the distribution of apo B-associated (isopropanol-precipitable) radioactivity among the nonequilibrium DGU subfradions was determined (Table 1). Compared with the normal subject, the patient with dysbetalipoproteinemia showed a greater proportion of counts in the fractions containing small VLDL and IDL. The

4 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al. 411 Fractal ill I Figure 1. Densitometric scans of 2% to 16% gradient gels of subfractlons obtained by nonequilibrium density gradient ultracentrifugation of the d < gml lipoprotelns from a normal male subject (A) and a male subject with dysbetalipoproteinemia (B). The gels were stained for protein with Coomassie blue R-250. Particle diameters of the major peaks (in A) were determined from calibration curves as described in Methods. Downloaded from by on September 6, 2018 percentage of radioactivity precipitable by isopropanol progressively increased in fractions 1a through 7, as shown in Table 1. Measurement of the distribution of radioactivity among apolipoproteins by densitometric scanning of SDS gradient gel blot autoradiograms (data not shown) revealed similar proportions of 125 l-apo B and non-b apolipoprotein as found by isopropanol precipitation. The percent isopropanol-precipitable radioactivity in VLDL fraction 1a for all of the preparations used ranged from 50% to 62% for normals and 51% to 66% for type 3. Dysbetalipoproteinemic VLDL and IDL subfractions contained somewhat greater amounts of apo E than normals, but in both cases 125 l-apo E represented only a minor component (less than 2%) of the total labelled apolipoprotein, in agreement with previous reports of dissociation of apo E from lipoproteins during ultracentrifugation 35 as well as inefficient labelling of this apolipoprotein. 38 Nonequilibrium DGU was shown to result in substantial losses of apo E from the d < subfractions, with greater than 90% of the recoverable apo E (as measured by radioimmunoassay in the laboratory of John Albers) found in the lipoprotein-free fraction at the bottom of the tube. A band corresponding to apo B-48 was sometimes seen in fraction 1, with negligible amounts in subsequent fractions. In the patient with dysbetalipoproteinemia, the amount of 125 I associated with this band was minimal when plasma triglycerides were maintained below 150 mgdl by dietary treatment. Plasma Decay of I-Apo B Radloiodlnated nonequilibrium DGU fractions isolated from the d < gml lipoproteins from individual normolipidemic donors were infused into rats in separate experiments, and their catabolism was followed over an 8-hour period. The total amount of protein injected ranged from 50 to 100 xg per rat. At intervals from 1 minute to 8 hours after infusion, residual 125 l-apo B was measured and lipoproteln products were analyzed by GGE. The behavior of corresponding subfractions from the three normal donors was similar. Figure 2A illustrates plasma decay curves of 125 I- apo B for fractions 1 a, 3, 5, and 7, averaged for these three experiments. Fraction 1a, containing the larger VLDL species, was cleared much more rapidly than the other fractions and showed a biphasic exponential decay curve. Greater than 50% of the 125 l-apo B present at 1 minute was cleared within 30 minutes. Fraction 3 containing small VLDL and fraction 5 containing predominantly the larger Table 1. Distribution of Apo B-Assoclated Radioactivity among VLDL and IDL Subfractlons Normal Type 3 SW41 fraction 1a 1b Percent isopropranol preclpitability Distribution of apo B in d<1.0i9 gml fractions* Percent Isopropranol preciplt ability 'Values represent percent of total isopropanol percipitable counts in the d < gml fraction. Distribution of apo B in d<1.019 gml fractions*

5 412 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST 1987 Downloaded from by on September 6, eo 50 I :, Hours Fraction 3 l 5 r LDL B Hours Figure 2. Plasma decay of 12S l-apo B after infusion into rats of lipoprotein subfractions Isolated from normolipidemic human subjects: A. Radioiodinated nonequilibrium DGU fractions 1a ( ), 3 ( ), 5 ( ), and 7 (D). B. Radioiodinated LDL (d = to gml). The percent remaining 125 l-apo B (measured as outlined in Methods) was calculated relative to the '^l-apo B counts present in a plasma sample drawn 60 to 90 seconds after injection of the labelled subtraction. Each point represents the mean ± SEM for three experiments using fractions from three subjects In A and eight experiments with LDL from two subjects in B. IDL species also displayed an early phase of more rapid clearance, quantitatively much less pronounced than for fraction 1a. The decay curve for fraction 7, which was enriched in the smaller, denser IDL species, was monoexponential. The rates of decay for all of the fractions after the first 2 hours were similar and roughly paralleled that of radlolabelled human LDL (d = to gml) infused into rats under similar conditions (Figure 2B). Analysis of Residual Lipoprotein Species In Plasma Lipoprotein species in the plasma samples from the above studies were characterized by 2% to 16% GGE. Gradient gels were blotted onto nitrocellulose paper, as described in Methods, and the lipoprotein bands were visualized by autoradiography. The findings for the three subjects were similar. Densitometric scans of representative gel-blot autoradiograms for fractions 1 a, 3, 5, and 7 from one of the subjects (cholesterol 251 mgdl, triglycerides 164 mgdl) are shown in Figure 3. The residual radioactive tracer in the rat injected with fraction 1a moved rapidly into the IDL size range, with a peak particle diameter of 302 A at 30 minutes and 285 A at 1 hour. By 2 hours, the major band was reduced in size to 280 A and a discrete shoulder was evident with a particle diameter of 261 A. At 4 hours, most of the radioactivity was confined to the peak at 260 A. The amount of this species appeared to gradually decrease over the following 4 hours, but its particle diameter changed little (257 A at 8 hours). The major LDL subspecies found in the plasma of the donor subject had a peak particle diameter of 259 A. The small VLDL species contained in fractions 2 through 4 formed a similar LDL-slzed particle, as shown for fraction 3 in Figure 3. The starting material in fraction 3 consisted of a sharp band with particle diameter 322 A, compared to the VLDL of peak particle diameter 372 A in fraction 1a. One hour after infusion, the bulk of the material appeared as a band in the size range of IDL (287 A peak particle diameter) with a shoulder at 266 A. Over the following hours, the IDL-sized material diminished and eventually disappeared, while the relative magnitude of an LDL peak of final diameter 258 A increased. The amount of label appearing in the LDL peak was much greater for fractions 2 through 4 compared with fraction 1a (Figure 2). Fraction 5 contained primarily the larger IDL species, with a peak particle diameter of 297 A. At 1 hour after infusion, the bulk of the lipoprotein remaining in the plasma was found in a peak at 287 A, with a shoulder at 264 A. Over the ensuing hours, the relative intensity of the smallest band again increased; by 6 to 8 hours, however, there remained two closely overlapping bands at 280 A and 268 A in addition to the more prominent peak at 258 A. For fraction 7 containing a larger proportion of smaller, denser IDL a similar pattern was seen. Three overlapping peaks were present at 6 to 8 hours; however, in this case the largest lipoprotein subspecies, with particle diameter measuring 284 A, was the predominant species. To further assess the density distribution of the to gml 125 l-labelled lipoprotein products at 8 hours, they were subjected to equilibrium DGU, performed as described in Methods. Figure 4 illustrates the distribution

6 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al. 413 Normal Subject FRACTION TIME 1a min 264 2hr I 1261 L_A 8hr Downloaded from by on September 6, 2018 Figure 3. Densitometric scans of autoradiograms of 2% to 16% gradient gel-nitrocellulose blots showing the distribution of 125 l-labelled lipoprotein products after infusion into rats of radioiodinated nonequillbrlum DQU fractions 1a, 3, 5, and 7 from a normolipidemic subject. Particle diameters of the major peaks (in A) were determined from calibration curves as described in Methods. of 12S l-apo B among the equilibrium DGU fractions at 8 hours for starting substrates 1 a, 3,5, and 7. The bulk of the LDL formed from fraction 1a was found in the equilibrium DGU fractions 1 and 2 (d = to gml). The LDL formed from nonequilibrium DGU fraction 3 were somewhat denser, peaking in equilibrium DGU fractions 2 and 3, corresponding to the density range of to gml. The LDL deriving from nonequilibrium DGU fraction 5 was again found predominantly In equilibrium DGU fractions 2 and 3. A larger portion of 12s l-apo B, however, was now seen in equilibrium DGU fractions 0 and 1 (d < gml). For the product formed from nonequilibrium DGU fraction 7, equilibrium DGU fraction 0 (d< gml) contained the largest amount of 12S l-apo B, with the remainder distributed among fractions 1 to 3. Fraction 0 has been shown to include IDL In fact, GGE autoradiograms (not shown) of fraction 0 derived from rats injected with nonequilibrium DGU fractions 5 and 7 contained a peak with particle diameter similar to the largest of the species present in whole plasma 8 hours after the injection of these two fractions (280 and 284 A, respectively; Figure 4). On the other hand, the peak particle diameters of the lipoproteins in equilibrium fraction 1 from rats injected with the same two nonequilibrium fractions corresponded to the middle peaks seen on GGE autoradiograms of the 8-hour plasma (269 and 273 A, respectively; Figure 3). Finally, lipoproteins in equilibrium fractions 2 and 3, deriving from all of the nonequilibrium DGU fractions, had peak particle diameters s 263 A, values similar to those for the smallest of the catabolic products observed on GGE autoradiograms of the uncentrifuged 8-hour plasma (Figure 3). Particles of similar diameter (260 A) were also found in equilibrium fractions 0 and 1 from rats infused with nonequilibrium DGU fraction 1a, in contrast to the larger particles in these equilibrium DGU fractions deriving from the other nonequilibrium DGU fractions. Thus, a more buoyant, but not larger, LDL product was formed from large VLDL. A separate equilibrium DGU run using fractions derived from a different normolipidemic subject gave essentially identical results. Greater than 88% of the non-isopropanol-precipitable radioactivity in the 8-hour samples derived from fraction 1 a was found in the d<1.019 and d> gml fractions. Estimation of the distribution of label among apolipoproteins by densitometric scanning of SDS-gel blot autoradiograms of delipidated d = to gml lipoproteins at 8 hours revealed detectable residual 125 I-C apolipoproteins only in the products derived from fractions 1 a and 2, where they comprised 13% and 3%, respectively, of the total radioactivity. SDS-gel blot analyses of the equilibrium DGU fractions, however, showed that most of this radioactivity was actually confined to fractions of d> gml where increasing amounts of rat high density lipoprotein, (HDL,) may be found. 37 Less than 2% of the radioactivity in the equilibrium DGU fractions containing the major labelled LDL products migrated with C-apolipoproteins, the remainder consisting of 12S l-apo B.

7 414 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST Fx la y Fx a 10 CD i in 2 50 co s *- u i i i i r"*h-fci i i i i i i i Fx ^"T*"i 1 i i i Fx c) Downloaded from by on September 6, 2018 ^ 1 Volume (ml) Fraction No 0 *~1 2~ T" ~4~ 5 6 i i i i i ^^TT J Rgure 4. Density distribution of the 12S l-apo B-containing llpoprotem products (d = to gml) 8 hours after infusion into rats of radioiodinated nonequilibrium DGU fractions 1 a, 3, 5, and 7 from a normolipidemic subject. Equilibrium density gradient ultracentrifugation was performed as described in Methods. Catabollsm of VLDL and IDL Subtraction from a Patient with Dysbetallpoprotelnemla The d< gml lipoproteins from a patient with dysbetalipoproteinemia were radioiodinated, fractionated by nonequilibrium DGU, and infused into rats, as described for the normal subjects above. Plasma decay of 125 l-apo B for the different fractions averaged for three separate experiments is shown in Figure 5. The clearance of 125 l-apo B from fractions 1a through 4 was considerably more rapid than observed with the corresponding fractions from normal subjects (Figure 2). Clearance of 50% of the 125 l-apo B following injection of dysbetalipoproteinemic fraction 3 averaged 50 minutes, compared with 6.5 hours for the corresponding fraction from normal subjects. Fraction 5 decayed at a slower rate, but still more rapidly than for this fraction from normal subjects. The more rapid clearance of these dysbetalipoproteinemic fractions was attributable to an early phase of rapid decay. After the first 2 hours, the rates of decay were comparable to those observed with normal fractions or normal human LDL (Figure 2B). Clearance of 12S l-apo B after infusion of fraction 7 from the dysbetalipoproteinemic patient was monoexponential and similar to that seen with this fraction from normolipidemic subjects (compare Figures 2A and 5). 6~ 102 Analysis by GGE of the lipoprotein products after infusion of dysbetalipoproteinemic precursors is shown in Figure 6 for a representative experiment. Fraction 1 a formed a relatively homogeneous LDL product measuring 259 A in peak particle diameter, resembling in size the product formed from the corresponding starting fraction isolated from normal subjects. As expected from the rapid 125 l-apo B decay curve for this fraction (Figure 5), only very small amounts of this material were present at 8 hours. Fractions 2 through 4 from dysbetalipoproteinemic plasma formed only minimal amounts of 260 A LDL, with most of the lipoprotein product found in two overlapping bands in the 270 to 290 A size range. This pattern resembled that seen with fraction 7 from normal subjects, as shown in the bottom panel of Figure 6. The major product formed from dysbetalipoproteinemic fraction 5 in this experiment was a somewhat broader intermediate-sized peak at 272 A, although in another experiment a 280 to 288 A band similar to that seen in the product of fraction 3 was the predominant product formed from this fraction as well. Dysbetalipoproteinemic fraction 7, however, gave rise to a more homogeneous peak with particle diameter 261 A at 8 hours with little, if any, of the larger lipoproteins that were found after injection of fraction 7 from normal subjects (Figure 3).

8 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al. 415 m s c Fraction in the different density intervals is shown in Figure 7. The LDL product arising from nonequilibrium DGU fraction 1a had a similar equilibrium DGU profile to the product formed from fraction 1 a isolated from normolipidemic plasma (Figure 4). The product formed from fraction 3, in contrast, was strikingly different; most of the 125 l-apo B in the case of the dysbetalipoproteinemic precursor was found in the d < gml density range (fraction 0), as opposed to the d = to gml interval (fractions 2 and 3) for the corresponding fraction from the normolipidemic subject. The product from dysbetalipoproteinemic fraction 5 also contained greater amounts of more buoyant Iipoproteins with smaller amounts of d = to gml Iipoproteins compared with the same fraction from the normal subject. Fraction 7 from the dysbetalipoproteinemic subject, in contrast, gave rise almost exclusively to d = to gml LDL, without the larger amounts of more buoyant material that were seen to accumulate with normolipidemic fraction 7. Thus, the major products of fractions 2 through 4, which contained most of the G-VLDL, were similar in size and density to the LDL-I species formed from IDL contained in fraction 7 in normal subjects. The smaller IDL found in fraction 7 from dysbetalipoproteinemic plasma, in contrast, formed a fairly homogeneous product corresponding in size and density to LDL-II. Downloaded from by on September 6, 2018 HOUT8 Figure 5. Plasma decay of 125 l-apo B after infusion into rats of lipoprotein subtractions isolated from a subject with dysbetalipoproteinemia: radlolodlnated nonequilibrium DGU fractions 1a ( ), 2 (O), 3 ( ), 4 (A), 5 ( ), 6 ( ), and 7 (D). The percent remaining 1Z5 l-apo B was calculated as described in the legend for Figure 2. Each point represents the mean for three rats from three separate experiments. Although most of the 125 l-apo B was found in the d = to gml fraction by 8 hours with all of the fractions from dysbetalipoproteinemic plasma, a larger amount remained in the d < gml fraction for fractions 2 through 5 compared to the normal subjects (8% to 11 % vs. 1.6% to 2.5%, respectively). GGE autoradiograms of these d < gml fractions revealed primarily a lipoprotein species with peak particle diameter of 290 A, overlapping with similar-sized Iipoproteins (288 A peak diameter, Figure 6) that were the predominant residual particles in whole plasma and the d = to gml fractions. Thus, the distribution of these products overlapped the conventional IDL and LDL density boundaries. Greater than 97% of the 125 l-apo B at 8 hours in rats infused with dysbetalipoproteinemic fraction 7, however, was found in the d = to density range, as was the case with the corresponding fractions from normal subjects. Labelled non-b apolipoproteins were detected only in the 8- hour d = to gml fractions after injection of type 3 fractions 1a or 2. The d = to gml Iipoproteins isolated from the plasma of rats 8 hours after infusion of dysbetalipoproteinemic nonequilibrium DGU fractions 1 a, 3,5, and 7 were analyzed by equilibrium DGU. The distribution of 125 l-apo B Discussion The technique of gradient gel electrophoresis has provided a tool for identifying multiple distinct lipoprotein subspecies in human plasma in the low 24 and intermediate 25 density ranges. By combining GGE with a method for electrophoretic transfer of apolipoproteins from lipoprotein bands on gradient gels to nitrocellulose paper, it has been possible to follow the size transformations of subtractions of radiolabelled Iipoproteins as they are metabolized in vivo. The rat provides a useful model for the study of in vivo lipoprotein catabolism. Previous studies have shown that LDL are formed from catabolism of human VLDL injected into rats. 38 ' M Our findings demonstrate that this LDL product closely resembles in size, density, and apolipoprotein composition the LDL found within human plasma. Nevertheless, rat and human lipoprotein metabolism differ in several ways that could influence our results. In the course of hydrolysis of VLDL in the rat, as much as 90% of the associated apo B is removed from the circulation by the liver prior to formation of LDL. 39 ' * Removal of VLDL remnants by a similar pathway also occurs in humans but is quantitatively less important. 5 ' 7 ~ 10 Some laboratories report more rapid catabolism of rat vs. human LDL in rats 41 and apparent higher affinity of purified rat liver LDL-receptors for rat compared with human LDL. 42 However, these studies used rat LDL preparations containing molar ratios of apo E to apo B close to or greater than 1, obscuring the interpretation of observed differences. In other studies, similar clearance of rat and human LDL in both estradioltreated and control rats has been reported. 43 The low levels of cholesteryl estertriglyceride transfer activity in rat plasma 26 may limit lipid "remodeling" of LDL, believed to occur in vivo in humans. On the other hand, low

9 416 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST 1987 Type 3 Subject FRACTION 1a TIME mtn u 2hr 3*4 282 I nr 2S I 261 I Downloaded from by on September 6, 2018 Figure 6. Oensitometric scans of autoradiograms of 2% to 16% gradient gel-nitrocellulose blots showing the distribution of 125 l-labelled lipoprotein products after infusion into rats of radioiodinated nonequilibrium DGU fractions 1a, 3, 5, and 7 isolated from a subject with dysbetalipoproteinemia. Particle diameters of the major peaks (in A) were determined from calibration curves as described In Methods. lipid transfer protein activity makes it possible to investigate the direct conversion of subtractions of triglyceriderich lipoproteins to LDL subspecies in an in vivo setting free from this complicating factor. The question of whether the precursor-product relationships defined in this study and in parallel in vitro studies 44 correspond to physiologic human metabolic pathways, however, will require direct testing in humans. We report here that a subpopulation of VLDL-containing particles of 300 to 330 A mean peak diameter from normolipidemic human plasma were preferentially converted to LDL, while larger VLDL (>350 A mean particle diameter) were cleared from the plasma much more rapidly and gave rise to only small amounts of an LDL-sized product. Since nonequilibrium DGU fraction 1a also contains variable amounts of 300 to 330 A-sized VLDL in addition to the predominant larger VLDL, 25 it is uncertain whether the small amounts of LDL that were formed derive from the large, the small, or possibly both VLDL subspecies. It is also uncertain to what extent large VLDL are degraded to smaller particles prior to rapid clearance. Our findings are in agreement with those of Packard et al. 45 who reported that small and large VLDL differ in the extent to which they give rise to LDL in normal and hypertriglyceridemic human subjects. This phenomenon may have a variety of metabolic consequences. It may well contribute to the reported differences in the extent to which VLDL is transformed to LDL in different animal species and in hypertriglyceridemic states in humans where large VLDL predominate. IDL are generally thought of as transition products of VLDL metabolism. Our findings indicate that small VLDL do give rise to particles of intermediate size which, in turn, form products with size and density characteristic of the LDL-II subclass (Figure 3). Based on the observation that IDL-containing fractions also yield LDL-II products, it is likely that these IDL include metabolic intermediates of small VLDL catabolism. IDL-containing fractions, however, also yielded two additional subspecies not clearly identified in the product formed from catabolism of small VLDL. One of these corresponded in size and density to the LDL-I subclass previously reported in normal human plasma 21 ' 24 of diameter 260 to 275 A and density to gml. The second of these products was still larger (275 to 285 A) and overlapped the conventional LDL and IDL density intervals. In a previous report 25 we described two major forms of IDL in normal subjects: IDL-1, found predominantly in nonequilibrium DGU fraction 5 by the procedure used here; and IDL-2, isolated primarily in fractions 6 and 7, which include smaller, denser, and cholesterol-enriched particles. We have here shown that lipoproteins in fractions 6 and 7 yield higher proportions of the LDL-I and IDL-sized products than the larger, more triglyceride-enriched IDL in

10 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al I I I I I I I Fxia I 1 1 I Fx 5 v ^ ~ CD a 10 CO ID CM 50 CD ^ 40 C u <D 30 i i i i i r*-^ i i i i i i i Fx 3 y \ ^**f>^.^^^j Fx 7 j -103 I I 102 \.5? 106 C Downloaded from by on September 6, I^fc-L ^)^ Volume (ml) Fraction No "~] 2~ T" ~4~ "T Figure 7. Density distribution of the 125 l-apo B-containing lipoprotein products (d = to gml) 8 hours after infusion into rats of radioiodinated nonequllibrium DGU fractions 1 a, 3, 5, and 7 isolated from a subject with dysbetallpoproteinemia. Equilibrium density gradient uhracentrifugation was performed as described in Methods. fraction 5. It would therefore seem plausible that IDL-2 is metabolically related to at least one of these products. Since IDL-2 include particles that are smaller than the IDLsized product, 25 we surmise that IDL-2 form LDL-I, while the larger products may arise from a larger form of cholesterol-enriched IDL that is obscured within the smaller, denser end of the IDL-1 particle distribution. Firm conclusions as to these precursor-product relationships will require improved methodology for separating IDL-1 and IDL- 2 subspecies. The precursor-product relationships among VLDL, IDL, and LDL subpopulations suggested by our data are represented schematically in Figure 8. The gml density boundary conventionally separating IDL from LDL must be viewed as arbitrary. Thus, although the cholesterol-enriched IDL that give rise to LDL-I and the larger products are diagrammed as LDL "precursors" in Figure 8, they could also be viewed as particles contained within the more buoyant extremes of particle distributions overlapping the d = gml boundary, which can be converted to particles in the conventional LDL density range (LDL-I or larger species) by hydrolysis of contained lipids. The origin of the cholesterol-enriched IDL precursors of LDL-I and the larger products remains uncertain. They could be secreted 102 directly by the liver; alternatively, they could derive from subpopulations of VLDL representing a small part of the total VLDL. It is also not known to what extent the small VLDL precursor of LDL-II is directly secreted by the liver or formed in the circulation from larger VLDL precursors. Recent studies of in vitro lipolysis of the VLDL and IDL subtractions studied here using bovine milk lipoprotein lipase have shown similar precursor-product relationships gml- Smal VLDL DL-1 P-VLDL DL gml [ Larger Species--- LDL-n LDL^ Figure 8. Hypothetical precursor-product relations among VLDL, IDL, and LDL subpopulations.

11 418 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST 1987 Downloaded from by on September 6, 2018 to those found in vivo in the rat. 44 This suggests that the formation of LDL-I and LDL-II from VLDL and IDL precursors in vivo in the rat may occur by direct lipolysis. While LDL-I has been shown to be polydisperse, 24 a distinct LDL subspecies corresponding to the largest product observed here was not detected in previous gradient gel electrophoretic analyses. 24 This may be a consequence of the fact that it overlaps the size and density range of IDL. It is also possible that large amounts of this product are not found in plasma because it is formed in small amounts or catabolized more rapidly in humans than other LDL species. In this regard, it may be relevant that human plasma contains substantial concentrations of LDL subclasses smaller and denser than the LDL species formed in these studies, 2124 and it may be that in normal humans, further catabolic processing of the larger primary products may give rise to one or more of the smaller, denser subclasses. Deckelbaum et al. 46 have suggested that cycles of cholesterol-triglyceride exchange and subsequent lipolysis are responsible for formation of the small, dense LDL found in hypertriglyceridemic patients, and this process may be involved with formation of smaller LDL subspecies from LDL-I, LDL-II, or larger cholesterol-rich precursors in humans. As discussed above, the low levels of cholesteryl estertriglyceride exchange protein activity in rat plasma could limit "remodeling" of lipoproteins that normally results from lipid transfer during lipolysis. The role of this process in the formation of LDL subspecies in the setting of normotriglyceridemia, however, is speculative at the present time. Our findings that particles with properties of human LDL-I and LDL-II are readily produced from VLDL and IDL precursors in the rat as well as in vitro 44 indicate that plasma lipid transfer activity is not required for the initial formation of these larger LDL subspecies from specific precursor lipoproteins. The question arises whether the observed metabolic heterogeneity within the IDL might reflect an artifact introduced by methods of isolation. Ultracentrifugation may result in selective dissociation of apolipoproteins from lipoprotein particles This might be particularly likely to occur with dysbetalipoproteinemic, compared with normal, d< gml lipoproteins where initial apo E content is greater than normal. Could the loss of apo C-ll, apo E, or other apolipoproteins from a portion of the IDL have given rise to a subpopulation of particles for which further hydrolysis by lipoprotein lipase or other possible metabolic transformations are impaired? Although difficult to exclude completely, it seems unlikely that this accounts for the metabolic heterogeneity observed because an ample pool of endogenous rat apo C-ll, apo E, and other exchangeable apolipoproteins exists in rat plasma. Transfer of rat apolipoproteins must be considered in interpreting the fate of the injected human subtractions. Although it is not possible to directly quantitate the apo E associated with the labelled subspecies after injection into rats, apo E is known to exchange rapidly between lipoproteins. Reassociation of apo E with the larger VLDL subspecies is suggested by the rapid clearance of these particles despite prior large preparative losses of apo E. The still more rapid clearance of both large and small VLDL from the dysbetalipoproteinemic subject presumably also reflects exchange or transfer of rat apo E to these lipoproteins, permitting uptake mediated by rat liver receptors that do not interact with the E2 isoform. 48 It would appear that the observed differences in plasma decay of different subspecies injected into rats probably reflect intrinsic properties of these particles determining their capacity to associate with apo E and that losses of this apolipoprotein during ultracentrifugal isolation are rapidly replenished in vivo by transfer of rat apo E. Our data show that LDL subspecies by all measured criteria identical to the major LDL-I and LDL-II subspecies found in normal human plasma are formed from specific human precursors in vivo in the rat despite depletion in apo E prior to the injection. Nevertheless, both loss of apo E during preparation of lipoproteins, as well as acquisition of rat apolipoproteins following injection, could have influenced these results. Both normal and dysbetalipoproteinemic plasma contain lipoprotein precursors of particles with characteristics of LDL-I and LDL-II, as well as larger IDL-sized products. In normals, the precursors of the latter are found predominantly in the IDL size and density range (nonequilibrium DGU fraction 7), while in dysbetalipoproteinemia these precursors are in fractions which contain 13-VLDL. Despite some resemblance to 13-VLDL in size, composition, and density, 25 the small VLDL species in normals, as noted above, are preferentially converted to an LDL-ll-like product after injection into rats. It seems likely that the (3-VLDL in dysbetalipoproteinemic plasma that form IDL-sized products correspond metabolically to the smaller and denser IDL subpopulation in normals which behaves similarly. The fact that, in contrast to normals, only small proportions of the largest product are formed from dysbetalipoproteinemic nonequilibrium DGU fractions 6 and 7 is consistent with the idea that the precursors for this product have shifted to larger size and lower buoyant density in dysbetalipoproteinemia (Figure 8). This shift may be a consequence of lipid enrichment of lipoproteins whose residence time in plasma is abnormally prolonged as a result of the impaired apo E-receptor interaction characteristic of this disorder. The present findings are consistent with the hypothesis that the cholesterol content of a given precursor (relative to protein or apo B) is the major determinant of the size of the final LDL product formed in the rat. It is noteworthy in this regard that the cholesterolprotein ratio of dysbetalipoproteinemic (3-VLDL is considerably higher than that of small VLDL from normals, but only slightly higher than the cholesterol-enriched IDL found in equilibrium DGU fraction 7 from normals. 25 The resemblance of the product formed from type 3 13-VLDL to that formed from a component of normal IDL, therefore, might result from similar degrees of cholesterol enrichment of a homologous lipoprotein particle. Other investigators have observed that a subpopulation of the d < gml lipoproteins from subjects with dysbetalipoproteinemia is resistant to conversion to an LDL product both in vivo 45 and in vitro upon incubation with lipoprotein lipase. 49 Ehnholm et al. 50 recently confirmed the latter result specifically for an apo B-100-containing 6- VLDL subfraction isolated from the plasma of the same subject (D.R.) studied in this report. They also found that

12 METABOLISM OF IDL AND VLDL SUBSPECIES Musliner et al. 419 Downloaded from by on September 6, 2018 when this fraction of B-VLDL was hydrolyzed by milk lipoprotein lipase in the presence of d > 1.21 gml lipoproteindeficient plasma and apolipoproteins C-ll and E3 (but not E2), a product in the LDL density range (d = to gml) was formed. Our data show that in vivo in the rat, a subpopulation of dysbetalipoproteinemic small VLDL is converted to a product outside the size and density range of most normal LDL, regardless of the presence of normal rat apo E, while other dysbetalipoproteinemic triglyceriderich subspecies (e.g., large VLDL or subpopulations of IDL) which also contain apo E2, form products of LDL-II size and density. A variety of clinical and experimental data suggest that lipoproteins in the IDL size and density range are associated with atherosclerosis. This is most evident in patients with dysbetalipoproteinemia. 51 Increased levels of IDL and cholesterol enrichment of VLDL have also been reported in populations of myocardial infarction survivors in Italy 52 ' M and Japan. 54 ' K Evidence for a strong association between Sj 10 to 20 lipoproteins and atherosclerosis was first presented by Gofman et al. 56 More recently, analytic ultracentrifugation measurements of lipoproteins from patients in the NHLBI Type II Intervention Trial have shown that changes in IDL of Sj 10 to 20, particularly Sj 10 to 14, are most strongly associated with coronary artery disease progression, compared with other lipoprotein subtractions. 57 The IDL species shown here to have metabolic characteristics analogous to 3-VLDL are found in an ultracentrifugal fraction (nonequilibrium DGU fraction 7) that contains lipoproteins of Sj 10 to Thus, it seems likely that these cholesterol-enriched particles isolated as IDL subspecies in normal and hypercholesterolemic subjects and as 13- VLDL in dysbetalipoproteinemia have properties that render them relatively atherogenic. Acknowledgments The authors appreciate the expert technical assistance of Sharon Balcom, M. Heidi Clough, and Janet Jacobson, and the graphics work of Joan Carter. We also thank Thomas Bersot for assistance in obtaining plasma samples from the dysbetalipoproteinemic subject and John Albers tor performance of apo E radioimmunoassays. References 1. Elsenberg S, Bllhelmer DW, Levy Rl, Llndgren FT. On the metabolic conversion of human plasma very low density lipoprotein to low density lipoprotein. Biochim Biophys Ada 1973;326: Berman M, Hall M, Levy Rl, et al. Metabolism of apo B and apo C lipoproteins In man: kinetic studies in normal and hyperiipoproteinemic subjects. J LJpid Res 1978;19: Slgurdsson GA, Lewis N, Lewis B. Conversion of very low density lipoprotein to low density lipoprotein. J Clin Invest 1975:56: Soutar AK, Myant NB, Thompson QR. Simultaneous measurement of apolipoprotein B turnovers in very-low and lowdensity lipoproteins in familial hypercholesterolemia. Atherosclerosis 1977:28: Janus ED, Nlcoll A, Wooton R, Turner PR, Maglll PJ, Lew- Is B. Quantitative studies of very low density lipoprotein conversion to low density lipoprotein in normal controls and primary hyperlipoproteinemic states and the role of direct secretion of low density lipoprotein in heterozygous familial hypercholesterolemia. Eur J Clin Invest 1980;10: Goldberg IJ, Le NA, Ginsberg HN, Paternltl JR, Brown WV. Metabolism of apoprotein B in cynomologous monkey: evidence for independent production of low density lipoprotein apoprotein B. Am J Physiol 1983:244 (Endocrinol Metab 7):E196-E Reardon MF, Fldge NH, Nestel PJ. Catabolism of very low density lipoprotein B apoprotein in man. J Clin Invest 1978:61: Klssebah AH, Alfarsl S, Adams PW, Wynn V. The metabolic fate of plasma lipoproteins in normal subjects and in patients with insulin resistance and endogenous hypertriglyceridemia. Diabetologia 1976:12: Eaton RP, Allen RC, Schade DS. Overproduction of a kinetic subclass of VLDL-apoB, and direct catabolism of VLDL-apoB in human endogenous hypertriglyceridemia: an analytical model solution of tracer data. J Lipid Res 1983;24: Fisher WR, Zech LI, Bardalaye P, Warmke G, Berman M. The metabolism of apolipoprotein B in subjects with hypertriglyceridemia and polydisperse LDL. J Lipid Res 1980;21: Faergeman O, Sala T, Kane JP, Havel RJ. Metabolism of apoprotein B of plasma very low density lipoproteins in the rat. J Clin Invest 1979:56: Elsenberg S, Rachmllewltz D. Metabolism of rat plasma very low density lipoprotein. II. Fate in circulation of apoprotein subunits. Biochim Biophys Acta 1973:326: Patsch W, Patsch JR, Kostner GM, Sailer S, Braunstelner H. Isolation of subtractions of human very low density lipoproteins by zonal ultracentrifugation. J Biol Chem 1978:253: Pagnan A, Havel RJ, Kane JP, Kottte L. Characterization of human very low density lipoproteins containing two electrophoretic populations: double pre-beta lipoproteinemla and primary dysbetalipoproteinemia. J Lipid Res 1977;18: Sata T, Havel RJ, Jones AL. Characterization of subpopulations of triglyceride-rich lipoproteins separated by gel chromatography from blood plasma of normolipemic and hyperiipemic humans. J Lipid Res 1972;13: Quariordt SH, Nathans A, Dowdee M, Hllderman HL. Heterogeneity of human very low density lipoproteins by gel filtration chromatography. J LJpid Res 1972:13: Kuchlnsklene Z, Carlson LA. Composition, concentration, and size of low density lipoproteins and of subtractions of very low density lipoproteins from serum of normal men and women. J Lipid Res 1982:23: Trezzl E, Calvl C, Roma P, Catapano AL. Subfractionation of human very low density lipoproteins by heparin-sepharose affinity chromatography. J LJpid Res 1983:24: Llndgren FT, Jensen LC, Wills RD, Freeman NK. Flotation rates, molecular weights, and hydrated densities of the low density lipoproteins. Lipids 1969:4: Lee DM. Isolation and characterization of low density lipoproteins. In: Day CE, Levy RS, eds. Low density lipoproteins. New York: Plenum Press, 1976: Shen MS, Krauss RM, Llndgren FT, Forte TM. Heterogeneity of serum low density lipoproteins in normal human subjects. J LJpid Res 1981:22: Adams GH, Schumaker VN. Equilibrium banding of low density lipoproteins. III. Studies on normal individuals and the effects of diet and heparin-induced lipase. Biochim Biophys Acta 1970:210:462^* Hammond MG, Mengel MC, Warmke GL, Fisher WP. Macromolecular dispersion of human plasma low density lipoproteins in hyperiipoproteinemia. Metabolism 1977;26: Krausa RM, Burke DJ. Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J LJpid Res 1982:23: Musliner TA, Glotas C, Krauss RM. Presence of multiple subpopulations of lipoproteins of intermediate density in normal subjects. Arteriosclerosis 1986;6: Barter PJ, Gooden JM, Rajaram OV. Species differences in the activity of a serum triglyceride transferring factor. Atherosclerosis 1979:33:

13 420 ARTERIOSCLEROSIS VOL 7, No 4, JULYAUGUST 1987 Downloaded from by on September 6, Rail SC, Weisgraber KH Jr, Mahley RW. Human apolipoprotein E. J Biol Chem 1982;257: Lindgren FT, Jensen LC, Hatch FT. The isolation and quantitative analysis of serum lipoproteins. In: Nelson GJ, ed. Blood lipids and lipoproteins: quantitation, composition, and metabolism. New York: John Wiley and Sons, 1972: MacFarlane AS. Efficient trace labeling of proteins with iodine. Nature London 1958; 182: Bilheimer DW, Eisenberg S, Levy Rl. The metabolism of very low density lipoprotein proteins. 1. Preliminary in vitro and in vivo observations. Biochim Biophys Acta 1972;260: Egusa G, Brady DW, Grundy SM, Howard BV. Isopropanol precipitation method for the determination of apolipoprotein B specific activity and plasma concentrations during metabolic studies of very low density lipoprotein and low density lipoprotein apolipoprotein B. J Lipid Res 1983;24: Herbert PN, Bausserman LL, Henderson LO, et al. Apolipoprotein quantitation. In: Peeters H, ed. The lipoprotein molecule. London: Plenum 1978; Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. In: Segrest JP, Albers JJ, eds. Methods in enzymology: Plasma lipoproteins, vol 128. New York: Academic Press, 1986: Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci 1979;76: Gibson JC, Rubinstein A, Bukberg PR, Brown WV. Apolipoprotein E-enriched lipoprotein subclasses in normolipidemic subjects. J Lipid Res 1983;24: Van't Hooft F, Havel RJ. Metabolism of chromatographically separated rat serum lipoproteins specifically labeled with 125 I- apolipoprotein E. J Biol Chem 1981;256: Lusk LT, Walker LF, DuBien LH, Getz GS. Isolation and partial characterization of high density lipoprotein HDLT from rat plasma by gradient centrifugation. Biochem J 1979; 183: Fidge N, Poulis P. Studies on the metabolism of rat serum very low density lipoprotein. J Lipid Res 1975;16: Eisenberg S, Windmueller HG, Levy Rl. Metabolic fate of rat and human lipoprotein apoproteins in the rat. J Lipid Res 1973;14: Stein O, Rachmilewitz D, Sanger L, Eisenberg S, Stein Y. Metabolism of iodinated very low density lipoprotein in the rat. Autoradiographic localization in the liver. Biochim Biophys Acta 1974;360: Van Tol A, Van't Hooft FM, Van Gent T. Discrepancies in the catabolic pathways of rat and human low density lipoproteins as revealed by partial hepatectomy in the rat. Atherosclerosis 1978;29: Windier EET, Kovanen PT, Chao YS, Brown MS, Havel RJ, Goldstein JL. The estradiol-stimulated lipoprotein receptor of rat liver. J Biol Chem 1980;255: Chao YS, Windier EE, Chen GC, Havel RJ. Hepatic catabolism of rat and human lipoproteins in rats treated with 17aethinyl estradiol. J Biol Chem 1979;254: Krauss RM, Musliner TA, Giotas C. Differing content of low density lipoprotein precursors in subclasses of triglyceriderich lipoproteins. Arteriosclerosis 1983;3:510a-511a 45. Packard CJ, Munro A, Lorimer AR, Gotto AM, Shepard J. Metabolism of apolipoprotein B in large triglyceride-rich very low density lipoproteins of normal and hypertriglyceridemic subjects. J Clin Invest 1984;74: Deckelbaum RJ, Eisenberg S, Oschry Y, Butbul E, Sharon I, Olivecrona T. Reversible modification of human plasma low density lipoprotein toward triglyceride-rich precursors. J Biol Chem 1982;257: Herbert PN, Forte TM, Schulman RS, et al. Structural and compositional changes attending the ultracentrifugation of very low density lipoproteins. Prep Biochem 1975;5: Schneider WJ, Kovanen PT, Brown MS, et al. Familial dysbetalipoproteinemia: abnormal binding of mutant apoprotein E to low density lipoprotein receptors of human fibroblasts and membranes from liver and adrenal of rats, rabbits, and cows. J Clin Invest 1981 ;68: Chung BH, Segrest JP. Resistance of a very low density lipoprotein subpopulation from familial dysbetalipoproteinemia to in vitro lipolytic conversion to the low density lipoprotein density fraction. J Lipid Res 1983;24: Ehnholm C, Mahley RW, Chappell DA, Weisgraber KH, Ludwig E, Witztum JL. Role of apolipoprotein E in the lipolytic conversion of B-very low density lipoproteins to low density lipoproteins in type III hyperlipoproteinemia. Proc Natl Acad Sci USA 1984,81: Mahley RW. Atherogenic hyperlipoproteinemia. Med Clin North Am 1982;66: Avogaro P, Cazzolato G, Bittolo Bon G, Belussi F. Levels and chemical composition of HDL 2, HDL 3, and other major lipoprotein classes in survivors of myocardial infarction. Artery 1979:5: Avogaro P, Bittolo Bon G, Cazzolato G, Rorai E. Relationship between apolipoproteins and chemical components of lipoprotein in survivors of myocardial infarction. Atherosclerosis 1980:37: Kameda K, Matsuzawa Y, Kubo M, et al. Increased frequency of lipoprotein disorders similar to type III hyperlipoproteinemia in survivors of myocardial infarction in Japan. Atherosclerosis 1984;51: Tatami R, Mabuchi H, Ueda K, et al. Intermediate-density lipoprotein and cholesterol-rich very low density lipoprotein in angiographically determined coronary artery disease. Circulation 1981:64: Gofman JW, Lindgren F, Elliott H, et al. The role of lipids and lipoproteins in atherosclerosis. Science 1950:111: Krauss RM, Lindgren FT, Williams PT, et al. Changes in lipoprotein subtractions associated with cholestyramine treatment and coronary artery disease progression. Arteriosclerosis 1984;4:558a Index Terms: lipoproteins lipoprotein subspecies lipoprotein metabolism very low density lipoproteins intermediate density lipoproteins low density lipoproteins dysbetalipoproteinemia atherosclerosis