THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1985 by The American Society of Biological Chemists, Inc

Vol. 260, No. 14, Issue of July 15, pp. 8509-8513, 1985 Printed in U.S.A.

Structure and Function of Human Low Density Lipoproteins STUDIES USING PROTEOLYTIC CLEAVAGE BY PLASMA KALLIKREIN*

(Received for publication, January 14, 1985)

Masaharu Yamamoto, Subramanian Ranganathan, and Bruce A. Kottke From the Atherosclerosis Research Unit, Mayo Clinic and Foundation, Rochester, Minnesota 55905

Kallikrein digestion of human low density lipopro- teins (LDL) has recently been shown to result in the degradation of apolipoprotein B (apo-B) into four ma- jor fragments, two of them being B-26 and B-74, which have been reported to be present in the LDL of some individuals. We studied the binding of kallikrein- treated LDL to human fibroblasts; digestion did not affect binding. Digested LDL was not taken up by macrophages, showing that it behaved like normal LDL. The activation of acyl-CoA cholesterol acyltrans- ferase by LDL in fibroblasts was also not altered by kallikrein digestion. When delipidated LDL was treated with kallikrein, apo-B was digested into very small fragments, indicating that kallikrein can cleave apo-B at sites other than those which result in the formation of B-26 and B-74. The partial delipidation of LDL with heptane also resulted in more extensive digestion of apo-B, although binding to cells was un- affected. These studies suggest that the cholesterol core maintains the proper orientation of apo-B in the LDL particle and that kallikrein may be used as a tool to elucidate the association of apo-B and lipids in the LDL particle.

The significance of low density lipoproteins (LDL’) in the transport and delivery of cholesterol to cells in the body has been documented by the studies of Goldstein and Brown (1, 2). They have also elucidated the mechanism by which cho- lesterol metabolism is regulated by LDL in the LDL pathway. This pathway involves the high affinity binding of LDL to specific receptors on the cell surface. However, the structure of apolipoprotein B (apo-B), the major apolipoprotein of LDL, is not clearly understood primarily because of its large size and insolubility in aqueous buffers after delipidation (3-5).

The apparent molecular weight of apo-B by sodium dodecyl sulfate (SDS) electrophoresis has been reported to be 550,000 (6). Kane et al. (6) have reported the presence of smaller fragments of apo-B in LDL and have designated these frag- ments as B-100, B-74, and B-26. This designation is based on a centile system in which each apoprotein is assigned a number which represents the ratio of its molecular weight relative to that of the largest species. Significant amounts of

*These studies were supported in part by Training Grant HL- 07329 from the National Heart, Lung, and Blood Institute and by a grant from the Whirlpool Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

’ The abbreviations used are: LDL, low density lipoproteins; apo- B, apolipoprotein B; DMEM, Dulbecco’s modified Eagle’s minimum essential medium; SDS, sodium dodecyl sulfate; HEPES, N-2-hy- droxyethylpiperazine-N’-2-ethanesulfonic acid; Tricine, N-[2-hy- droxy-l,l-bis(hydroxymethyl)ethyl]-glycine.

B-74 and B-26, which may be complementary constituents of B-100, have been detected in the LDL of many individuals (6).

Recently Cardin et al. (7) reported that the appearance of apo-B-74 and apo-B-26 in LDL can be prevented by the addition of protease inhibitors to plasma at the time of collection, suggesting that the smaller fragments are produced by the degradation of apo-B-100 by proteolytic enzymes. They also found that kallikrein can degrade apo-B to produce B-74 and B-26. However, the physiological significance of apo-B- 14 and -B-26 is not understood. In this paper, we present data which show that kallikrein digestion does not alter the prop- erties of LDL with respect to receptor binding and activation of cholesterol esterification in cells. Although apo-B in intact LDL was cleaved by kallikrein, resulting in the formation of four major fragments, including B-74 and B-26, delipidated apo-B was degraded by kallikrein to form fragments of very low molecular weight.

When partially delipidated LDL (8), from which cholesterol had been extracted with heptane leaving the phospholipids unchanged, was incubated with kallikrein, apo-B was cleaved with formation of smaller amounts of B-74 and B-26 and larger amounts of fragments with lower molecular weights. These results show that the cleavage of apo-B by kallikrein depends upon the nature of its association with lipids in the LDL particle.

EXPERIMENTAL PROCEDURES

Materiak-Sodium [‘251]iodide (carrier free) was obtained from Amersham Corp. [“C)Oleic acid and [3H]cholesteryl oleate were obtained from New England Nuclear. Human plasma kallikrein, bovine serum albumin (essentially fatty acid free), phenylmethane- sulfonyl fluoride, sodium oleate, and cholesteryl oleate were purchased from Sigma. Sodium decyl sulfate was obtained from Eastman. All tissue culture supplies were purchased from Grand Island Biological.

Isolation of LDL and Lipoprotein-deficient Serum-Normal human plasma was obtained from the blood bank and phenylmethanesulfonyl fluoride (0.5 mM), sodium azide (0.01%), EDTA (1 mM), and genta- mycin (0.1 mg/ml) were added. LDL was isolated by sequential preparative ultracentrifugation at 4 “C between densities of 1.019 and 1.063 and washed by recentrifugation at a density 1.063. Solid KBr was used to adjust the densities. The resultant LDL was dialyzed against 4 liters of buffer containing 0.15 M NaCl, 10 mM Tris-HC1 (pH 7.41, 1 mM EDTA, and 0.01% NaN3 with four changes. Lipopro- tein-deficient serum was prepared by the method described by Red- ding and Steinberg (9).

Preparation of Partially Delipidated and Reconstituted LDL-Par- tial delipidation was carried out by the method described by Krieger et al. (10). Briefly, LDL was dialyzed against 0.3 mM EDTA, pH 7.0, and lyophilized in the presence of potato starch. The lipids were extracted with heptane at -10 “C. This procedure selectively removes cholesteryl esters and triglycerides from LDL but not the phospho- lipids (8). To reconstitute the LDL, cholesteryl oleate was added. Heptane was evaporated and partially delipidated, or reconstituted LDL was extracted with Tricine buffer.

Acetylation and Radioiodination of LDL-Acetylation of LDL was

8509

8510 Proteolytic Cleavage of LDL by Kallikrein

carried out as described by Basu et al. (11). Iodination of LDL using sodium ["'I]iodide and iodine monochloride was carried out according to the method of Bilheimer et al. (12). The iodinated LDL was separated from unreacted sodium iodide by fractionation on a Bio- Gel P-2 column.

Preparation of Apo-B-Soluble apo-B from human plasma LDL was prepared according to the method described by Cardin et al. (13). In this procedure, lyophilized LDL was delipidated with diethyl ether:ethanol (3:1, v/v), and apo-B was dissolved in sodium decyl sulfate. After dialysis to remove decyl sulfate, apo-B was precipitated with ethanol and redissolved in 6 M guanidine hydrochloride. This solution was dialyzed successively against 6 M urea, 1 M urea, and a buffer containing 10 mM Tris-HC1, pH 8.4, and 1 mM EDTA. The yield of soluble apo-B was about 40% of the LDL protein. The solubility of apo-B in this buffer was no more than 1 mg/ml.

Digestion of LDL by Kallikrein-The digestion of LDL by kallikrein was carried out by incubating 1 ml of LDL (10 mg of protein) with 0.1 unit of kallikrein (10 units/mg protein) for 48 h at 37 "C. At 24 h of incubation, another 0.1 unit of kallikrein was added.

Fibroblasts-Human skin fibroblasts derived from normal subjects were obtained from the Human Genetic Mutant Cell Repository, Camden, NJ. The cells were grown in monolayers and used for experiments between the 8th and 20th passages. The cultures were maintained in Dulbecco's modified Eagle's medium (DMEM) supple- mented with penicillin (100 units/ml), streptomycin (100 pglml), HEPES (25 mM, pH 7.4), and fetal calf serum (10%) in 75-cm2 tissue culture flasks. Cells were incubated in a humidified incubator with 5% CO, at 37 "C. Subcultures for the experiments were obtained by trypsin treatment (37 "C, 3-4 min) of the confluent stock monolayers using a solution of 0.05% trypsin in Dulbecco's phosphate-buffered saline containing 0.5 mM EDTA. The cells were suspended in the growth medium, and they were seeded in culture dishes (60 X 15 mm) at a concentration of 1.0 X lo5 cells/dish in 3 ml of medium. On alternate days the medium was removed and fresh medium added.

Macrophages-Mouse peritoneal macrophages were isolated using calcium-free phosphate-buffered saline (14). After washing, the cells were suspended in culture medium containing 20% fetal calf serum and seeded in culture dishes (35 X 15 mm). After 24 h, the monolayer was washed with phosphate-buffered saline and incubated in DMEM supplemented with 5% lipoprotein-deficient serum for 24 h and used for the experiments.

LDL Binding to Fibroblasts-The binding of IZ5I-LDL to fibroblasts was studied according to the method described by Brown and Gold- stein (15). In short, on the 5th day of subculture, the cells were washed and incubated in DMEM containing 5% lipoprotein-deficient serum for 48 h. At the end of this period, the medium was replaced with fresh DMEM containing 5% lipoprotein-deficient serum. LDL binding was studied a t 4 "C for 3 h. After incubation, the monolayers were washed extensively and dissolved in 0.2 M NaOH. The radioac- tivity was determined in a y counter, and cell protein was determined by the method of Lowry et al. (16).

Acyl-CoA Cholesterol Acyltransferase-After 48-h incubation of the monolayers in DMEM containing lipoprotein-deficient serum, they were incubated with 0.1 mM [1-"Cloleate bound to albumin for 20 h at 37 "C, and radioactivity incorporated in cholesteryl oleate was determined as described by Goldstein et al. (17).

Agarose and SDS-Polyacrylamide Gel Electrophoresis-LDL was prestained with Sudan Black B and electrophoresed on 0.9% agarose gels in 50 mM barbital buffer, pH 8.6. SDS-polyacrylamide gel elec- trophoresis was carried out in 3-15% gradient slab gels containing 0.25 M Tris-HC1, pH 9, and 0.1% SDS. The electrophoresis buffer was made with 25 mM Tris-HC1, 0.2 M glycine, and 0.1% SDS, pH 8.4. After electrophoresis a t 15 mA, the gel was stained with Coomas- sie Brilliant Blue R.

Presentation of Results-The values presented in the figures and the table are arithmetic means of triplicate determinations. The variation among the triplicates was less than 10%.

RESULTS

When LDL was incubated with human plasma kallikrein and electrophoresed on SDS-polyacrylamide gel, fragments of apo-B with apparent molecular weights corresponding to B- 74, B-26, and two other major fragments were observed (Fig. 1). This observation was similar to that reported by Cardin et al. (7). Two other minor fragments with molecular weights in

.E" 00 -8-74

4-2 6

1- 2-

3- 4- 5-

6-

a b c d e f FIG. 1. SDS-polyacrylamide gel electrophoresis of kalli-

krein-treated LDL and apo-B. LDL and apo-B were treated with kallikrein for 48 h at 37 "C. Approximately 10 pg of protein were subjected to electrophoresis. Lune a, molecular weight standards ( I , 78,000; 2, 66,000; 3, 45,000; 4, 30,000; 5, 17,000; 6, 12,000); lane b, LDL; lane c, kallikrein-treated LDL; lane d, kallikrein-treated LDL reisolated by ultracentrifugation; lane e, apo-B; lane f, kallikrein- treated apo-B.

the range of 80,000 were also observed. In order to find if all the fragments were still associated with the LDL particle, kallikrein-treated LDL was reisolated by ultracentrifugation at a density 1.063. On electrophoresis no change was observed in the distribution of the fragments (Fig. l), nor were frag- ments detected in the infranatant (data not shown). Purified apo-B was treated with kallikrein to find out if the enzyme cleaves apo-B at the same sites as in intact LDL. The results shown in Fig. 1 indicate that apo-B is degraded with formation of fragments with the molecular weights ranging from 10,000 to 50,000. This observation suggests that kallikrein action is not limited to those which result in the formation of B-74 and

LDL, acetylated LDL, and kallikrein-treated LDL were subjected to agarose gel electrophoresis to determine differ- ences in charge. The results shown in Fig. 2 indicate that there is no alteration of charge after kallikrein treatment. Acetylated LDL was included for comparison, as it is known to have a different charge.

The binding capacity of kallikrein-treated LDL was inves- tigated in human skin fibroblasts. The results shown in Fig. 3 indicate that specific binding to receptors increased with increasing concentration of kallikrein-treated LDL and was saturable, reaching a maximum a t about 25 pg/ml. Nonspe- cific binding determined in the presence of a 10-fold excess of unlabeled kallikrein-treated LDL was less than 15% of the total binding. This indicates that kallikrein cleavage of LDL protein does not alter the receptor-binding properties, al- though most of the apo-B in the LDL was cleaved to smaller fragments (Fig. 1). Next, the effect of kallikrein-treated LDL on the binding of '*'I-LDL was studied. Fig. 4 shows that kallikrein-treated LDL can compete with '*'I-LDL as effec- tively as unlabeled LDL, suggesting that the binding charac-

B-26.

Proteolytic Cleavage of LDL by Kallikrein 851 1

a b C 150

100

50

FIG. 2. Agarose gel electrophoresis of kallikrein-treated LDL. Lipoproteins were prestained with Sudan Black B and electro- phoresed for 4 h at 200 V. Lane a, LDL; lane b, kallikrein-treated LDL; lane c, acetylated LDL.

0 c 0 3

a 300

In N - 200

P z i c

m c a -

0 5 10 15 20 25 Kallikrein treated 251-LDL

( p g protein/rnl) FIG. 3. Binding of kallikrein-treated "'1-LDL to human

skin fibroblasts. After the fibroblasts were incubated with medium containing lipoprotein-deficient serum for 48 h, the medium was removed, and fresh medium containing lipoprotein-deficient serum was added. The cells were cooled to 4 "C and incubated with indicated amounts of kallikrein-treated lZ5I-LDL (275 dpm/ng) for 3 h. The bound LDL was determined according to the method described by Brown and Goldstein (15). Nonspecific binding was determined by the addition of a 10-fold excess of unlabeled kallikrein-treated LDL. The specific binding was obtained by subtracting the nonspecific binding from the total binding. The average cell protein in each plate was 0.2 mg.

teristics of kallikrein-treated LDL are identical to those of the untreated LDL.

To confirm that the physiological function of LDL is pre- served after kallikrein treatment, activation of cholesterol esterification by LDL and kallikrein-treated LDL was studied in human fibroblasts and mouse peritoneal macrophages. It is known that acetylated LDL can activate cholesterol ester- ification in macrophages and not in fibroblasts (1). The results presented in Table I show that LDL and kallikrein-treated LDL activate cholesterol esterification in fibroblasts to the same extent, whereas they do not activate cholesterol esteri- fication in macrophages. Only acetylated LDL can activate esterification of cholesterol in the macrophages.

0- 0 20 40 60 00

Unlabelled lipoprotein (pg/rnl)

FIG. 4. Effect of unlabeled LDL and kallikrein-treated LDL on '"I-LDL binding to human skin fibroblasts. After the fibro- blasts were incubated with medium containing lipoprotein-deficient serum for 48 h, the medium was removed, and fresh medium contain- ing lipoprotein-deficient serum was added. The cells were cooled to 4 "C and incubated with 15 pgof lZ5I-LDL (141 dpm/ng) and indicated concentrations of unlabeled LDL or kallikrein-treated LDL for 3 h at 4 "C. The cell-associated 1251-LDL was determined after extensive washing as described by Brown and Goldstein (15). ."-., LDL; A-A, kallikrein-treated LDL. The average cell protein in each plate was 0.2 mg.

TABLE I Effect of LDL, kallikrein-treated LDL, and acetylated LDL on the

incorporation of ["Cloleate into cholesteryl esters in fibroblasts and macrophages

The fibroblasts and macrophages were incubated with medium containing lipoprotein-deficient serum for 48 and 24 h, respectively. At the end of this incubation, each plate received 2 ml of fresh medium containing 5% lipoprotein-deficient serum and 50 pg of LDL, kallikrein-treated LDL, or acetylated LDL. After incubation for 2 h at 37 "C, 0.1 mM [1-"Cloleate (2300 dpm/nmol) was added, and the dishes were incubated for 20 h. The medium was then removed, and the cells were washed and processed as described in the text.

Additions Fibroblasts Macrophages nmol ["Cloleate

esterified/mgprotein/20 h None 4.8 ND" LDL 53.0 10.7 Kallikrein-treated LDL 60.7 6.4 Acetylated LDL ND 350.8 ND, not determined.

From these results it appears that degradation of apo-B by kallikrein in LDL does not alter the physiological character- istics. These studies also suggest that the action of kallikrein on apo-B depends on the association of apo-B with the lipids in the LDL particle. Attempts were made to find how LDL receptor binding and digestion by kallikrein are altered when the lipid organization of LDL is modified. Soluble apo-B obtained by complete delipidation of LDL using the methods described by Cardin et al. (13) lost its binding capacity for receptors on human skin fibroblasts (Fig. 5). This suggests that apo-B has to be associated with lipids in order to have the proper configuration for binding. When the cholesteryl esters and triglycerides were selectively extracted from LDL with heptane, leaving the phospholipids behind, the binding capacity was preserved, although it was slightly less than that of intact LDL (Fig. 5). Addition of cholesteryl oleate to the depleted particles did not significantly alter the binding prop- erties (Fig. 5).

Since detergents are known to alter the orientation of proteins in lipid-protein complexes (18) we tested the effect

8512 Proteolytic Cleavage of LDL by Kallikrein

150 I 1

0 0 20 40 60 80

Unlabelled lipoprotein (pg/ml)

FIG. 5. Effect of unlabeled lipoproteins and apo-B on "'1- LDL binding to human skin fibroblasts. The experimental details are as described in the legend for Fig. 4. X-X, apo-B; A-A, decyl sulfate-treated LDL; U, cholesterol-depleted LDL recon- stituted with cholesteryl oleate; A-A, cholesterol-depleted LDL; o"-o, LDL. The average cell protein in each plate was 0.2 mg.

a b c d e f 9 h I

FIG. 6. SDS-polyacrylamide gel electrophoresis of intact and modified LDL. Modifications of LDL were carried out as described in the text. Approximately 10 pg of protein were subjected to electrophoresis. Lane a, LDL; lane b, kallikrein-treated LDL; lane c, cholesterol-depleted LDL; lane d, cholesterol-depleted LDL treated with kallikrein; lune e, cholesterol-depleted LDL reconstituted with cholesteryl oleate; lune f, cholesterol-depleted LDL reconstituted with cholesteryl oleate and treated with kallikrein; lane g, sodium decyl sulfate-treated LDL; lane h, sodium decyl sulfate-treated LDL treated with kallikrein; lune i, molecular weight standards ( I , 330,000; 2, 78,000; 3, 66,000; 4, 45,000; 5, 17,000; 6, 12,000).

of sodium decyl sulfate on the physiological function of LDL. Fig. 5 shows that LDL, when incubated with 20 mM dodecyl sulfate for 16 h at 37 "C and dialyzed extensively to remove the detergent, lost about 40% of its binding capacity, suggest- ing that a specific orientation of apo-B is required for maximal binding. Shorter incubation (1 h) did not affect the binding (data not shown).

Degradation of apo-B in LDL by kallikrein also appears to be determined by specific association of apo-B with lipids. Fig. 6 shows that when cholesterol-depleted LDL was incu- bated with kallikrein, the fragments obtained were different from those resulting from the incubation of intact LDL with

kallikrein. It can be noticed that a small degree of degradation took place during partial delipidation. Cholesterol-depleted LDL was degraded to yield fragments with molecular weights of less than 80,000. However, when cholesteryl oleate was reincorporated into the depleted particles, the degradation pattern was different. Apo-B was cleaved, resulting in the formation of a major band corresponding to the molecular weight of B-26 and minor bands corresponding to B-74 and other products. Other significant bands with lower molecular weights were also found. Decyl sulfate-treated LDL was de- graded by kallikrein with formation of a larger number of fragments in addition to B-74 and B-26.

DISCUSSION

Kallikrein degradation of LDL results in the formation of four major fragments, including B-74 and B-26, as reported by Cardin et al. (7). The results presented in this paper show that, in spite of cleavage of B-100 at points which can lead to smaller fragments, the binding properties of LDL to its recep- tors were not changed. This may indicate that the entire apo- B molecule is not involved in binding. LDL treated with kallikrein was able to activate cholesterol esterification in fibroblasts as efficiently as intact LDL but did not activate it in macrophages. These results indicate that the physiological properties of LDL are not altered by kallikrein treatment.

Although kallikrein digestion of LDL produced specific fragments, apo-B gave a large number of fragments with molecular weights ranging from 50,000 to 10,000. Apo-B in- corporated into a dimyristoylphosphatidylcholine complex was also degraded to very small fragments by kallikrein (data not shown). The degradation of apo-B in LDL to specific fragments may thus be due to limited accessibility to the enzyme because of association with lipids. The experiments in which cholesteryl esters and triglycerides were specifically removed leaving phospholipids unchanged also suggest that kallikrein does not have specific cleavage sites on apo-B. Steinberg et al. (19) have shown that cholesterol-free LDL is smaller than native LDL and has a disc-like appearance (19). These particles also had a tendency to associate with each other to form stacks. This may be due to increased exposure of apo-B. However, this modified LDL did not lose the capac- ity to bind to receptors, and kallikrein digestion yielded a large number of fragments smaller than B-26. When choles- terol oleate was reincorporated, kallikrein digestion resulted in the production of a significant amount of B-26, traces of B-74, and many smaller fragments, probably due to dimin- ished exposure of apo-B to the enzyme in the presence of cholesteryl esters. All of these results suggest that the pattern of digestion of apo-B by kallikrein is determined by the extent of accessibility of apo-B to the enzyme.

In another set of experiments, alteration of LDL by sodium decyl sulfate was studied because detergents are known to affect the orientation of proteins in lipid-protein complexes (18). When LDL was treated with sodium decyl sulfate, di- alyzed extensively to remove the detergent, and incubated with kallikrein, apo-B was cleaved forming many fragments with the molecular weights mostly in the range from 400,000 to 150,000. This pattern of digestion is entirely different from that obtained with native LDL. The random formation of fragments of high molecular weight is possibly due to exposure of different sites on apo-B to the enzyme. This modified LDL lost about 40% of its capacity to bind to receptors possibly because of alteration in the orientation of apo-B. Although Shireman et al. (20, 21) reported that apo-B complexed with bovine serum albumin bound to cellular receptors, we could not reproduce these results. When apo-B was complexed with

Proteolytic Cleavage of LDL by Kallikrein 85 13

phospholipid using sonication or detergents it was not able to bind to receptors? This suggests that random association of phospholipids with apo-B does not give the proper configu- ration required for binding. It is possible that expression of receptor-binding activity requires a defined structure of apo- B, as in the case in its immunoreactivity with certain mono- clonal antibodies reported by Marcel et al. (22).

We have investigated the effect of kallikrein digestion of LDL on its physiological properties, and the potential use of this enzyme in the elucidation of the molecular structure of LDL has been discussed. Kallikrein is involved in the initia- tion of the intrinsic pathway of the blood coagulation system. Although the specific site of the cleavage is not clearly under- stood, Pisano (23) has reported that this enzyme cleaves at arginine in kininogen, one of the physiological substrates for this enzyme. Although kallikrein formation is activated under certain conditions (24) and kallikrein can degrade apo-B in LDL, the present studies show that such a degradation does not seem to cause any changes of physiological importance.

Acknowledgments-We thank Laurie K. Bale for excellent techni- cal assistance and Susan M. Woychik for assistance in the prepara- tion of the manuscript.

REFERENCES 1. Goldstein, J. L., and Brown, M. S. (1983) The Metabolic Basis of

Inherited Disease, 5th Ed., pp. 672-712, McGraw-Hill, New York

2. Goldstein, J. L., and Brown, M. S. (1977) Annu. Reu. Biochem.

3. Shore, B., and Shore, V. (1967) Biochem. Biophys. Res. Commun.

4. Lee, D. M., Valente, A. J., Kuo, W. H., and Maeda, H. (1981)

46,897-930

28,1003-1007

Biochim. Biophys. Acta 666, 133-146

M. Yamamoto, S. Ranganathan, and B. A. Kottke, unpublished data.

5. Watt, R. M., and Reynolds, J. A. (1980) Biochemistry 19, 1593- 1598

6. Kane, J. P., Hardman, D. A., and Paulus, H. E. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 2465-2469

7. Cardin, A. D., Witt, K. R., Chao, J., Margolius, H. S., Donaldson, V. H., and Jackson, R. L. (1984) J. Bwl. Chem. 259 , 8522- 8528

8. Gustafson, A. (1965) J. Lipid Res. 6, 512-517 9. Redding, C. M., and Steinberg, D. (1960) J . Clin. Inuest. 39 ,

10. Krieger, M., Brown, M. S., Faust, J. R., and Goldstein, J. L.

11. Basu, S. K., Goldstein, J. L., Anderson, R. G. W., and Brown, M. S. (1976) Proc. Natl. Acad. Sci. U. S. A. 73, 3178-3182

12. Bilheimer, D. W., Eisenberg, S., and Levy, R. I. (1972) Biochim.

13. Cardin, A. D., Witt, K. R., Barnhart, C. L., and Jackson, R. L.

14. Edelson, P. J., and Cohn, Z. A. (1974) J. Exp. Med. 140 , 1364-

15. Brown, M. S., and Goldstein, J. L. (1974) Proc. Natl. Acad. Sci.

16. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275

17. Goldstein, J. L., Faust, J. R., Dygos, J. H., Chorvat, R. J., and Brown, M. S. (1978) Proc. Natl. Acad. Sci. U. S. A . 75 , 1877- 1881

1560-1569

(1978) J . Biol. Chem. 253,4093-4101

Bi0ph.y~. Acta 260, 212-221

(1982) Biochemistry 21,4503-4511

1386

U. S. A. 7 1 , 788-792

18. Lodish, H. F., and Rothman, J. E. (1979) Sci. Am. 240,48-63 19. Steinberg, D., Nestel, P. J., Weinstein, D. B., Remaut-Desmeth,

M., and Chang, C. M. (1978) Biochim. Biophys. Acta 528,199- 212

20. Shireman, R. B., Kilgore, L. L., and Fisher, W. R. (1977) Proc. Natl. Acad. Sci. U. S. A . 74, 5150-5154

21. Shireman, R. B., and Fisher, W. R. (1979) J. Lipid Res. 20,594- 598

22. Marcel, Y . L., Hogue, M., Weech, P. K., and Milne, R. W. (1984) J . Biol. Chem. 259,6952-6957

23. Pisano, J. J. (1975) in Proteases and Biological Control (Reich, E., Rifiin, D. B., and Shaw, E. eds) pp. 199-222, Cold Spring Harbor Press, New York

24. Nolly, H., and Lama, M. C. (1982) Clin. Sci. (Lord.) 6 3 , 2498- 251s