INVESTIGATION OF STRUCTURAL DOMAINS IN HUMAN SERUM LOW DENSITY LIPOPROTEIN APOLIPOPROTEIN 8100.

INVESTIGATION OF STRUCTURAL DOMAINS IN HUMAN SERUM LOW DENSITY LIPOPROTEIN APOLIPOPROTEIN 8100

INTRODUCTION

Human serum low density lipoprotein (LDL) apolipoprotein 8100 (apo-8100) plays a major role in lipid transport and cholesterol metabolism by mediating lipoprotein binding to specific cell surface receptors (the LDL receptor).1 Amino acid mutations in either the LDL receptor or in the apo-8100 molecule can disrupt efficient binding of LDL to the receptor and result in dyslipidemia.2,3 To understand fully how this protein functions at the molecular level, detailed knowledge of its structure must be obtained. The primary structure of apo-8100 has been established indicating that the protein is a sinole polypeptide chain composed of 4536 amino acid residues.4 ,5,o There is, however, little detailed information regardinQ the secondary and tertiary structure of the protein. The modulat1on of these different levels of protein structure must have important consequences for both lipoprotein structure and function.

Proteins which contain more Jh.an 200 residues are frequently organized into structural domains.? ,8 Although the word domain is frequently used to signify quite different characteristics of a protein (for example, lipid binding or receptor binding domains), the definition of a structural domain can be rather precise. As defined by Wetlaufer9, structural domains are contiguous stretches of sequences which are also contiguous in three dimensions (implying compactness). Domains may have physical stability as separate structures isolated from the rest of the protein10 and they have important functional roles.11 The identification of structura domains within a protein is essential for understanding the fashion in which a protein folds. Structural domains may represent intermediates along the folding pathway. Domains may fold independently and then be further assembled to Qive a fully folded protein. Finally, the fine-structure of a domam may be further modulated by tertiary or quaternary type interactions. 1 2 Although structural domains have been described for a wide variety of proteins, they have not been extensively investigated in the case of serum apolipoproteins Recently, two structural domains were described for apo-E3, 13,14 an ammo-terminal and a carboxyl-terminal domain. The amino-terminal domain, which has been crystallized, 1 5 contains the information necessary for interaction with the LDL receptor whereas the carboxyl-terminal domain may function more effectively in interactiog with lipids. In the case a.f apo-81 00, previous X-ray1 6 and electron microscopic 17 studies have suggested that the protein may contain

Figure 1. Denaturation curve of apo-81 00 in human serum low density lipoprotein. The fraction of apo-8100 in the native state is plotted as a function of the concentration of guanidine hydrochloride.

a number of structural domains. Indeed, the large size of apo-8100 suggests that the protein could contain several structural domains. Digestion with several enzymes has shown that apo-81 00 in LDL contains three large regions resistant to limited proteolysis separated by two protease susceptible regions.1 B

One of the objectives in our studies of low density lipoprotein is to look for, to isolate and to study the organization and folding of structural domains in apo-8100. Several techniques are potentially useful for demonstrating the presence of structural domains 1n proteins. One approach that is particularly powerful is the investigation of the stability of the protein to denaturation.

DENATURATION OF APOLIPOPROTEIN 8100

The denaturation of apo-8100 with guanidine hydrochloride was assessed by far ultraviolet circular dichroism. The denaturation of the protein is quantitated by following the loss of secondary structure at a given wavelength, for example at 218 nanometers. The fraction of the amount of ordered secondary structure at 218 nm with respect to the native protein is plotted in figure 1 as a function of the guanidine hydrochloride concentration. As the concentration of guanidine hydrochloride in solution is increased, there is a progressive loss of ordered structure as indicated by the decrease in the fraction of native structure. It is clear that the curve is asymmetric with a number of inflection points. These results suggest that the denaturation of the protein IS a weakly cooperative process which proceeds in an independent fashion for different parts of the protein. This type of behavior is consistent with the presence of several structural domains.

The denaturation of apo-8100 dissociated from its lipid moiety was also studied. Apo-81 00 was prepared by h1gh performance liquid chromatography in the presence of the non-ionic detergent C1 2 E 8 (n-octaethylene glycol monoether). The denaturation by guanidine hydrochloride of the apo-81 00-detergent complex was also found to be asymmetric w1th a number of inflection points and this behavior is consistent with the presence of structural domains. However, as compared to the lipid bound form, the denaturation occurs at lower guanidine hydrochloride concentrations.

LIMITED PROTEOLYSIS OF APOLIPOPROTEIN-81 00

In order to obtain further evidence for the existence of structural domains in apo-8100 and to tentatively identify which parts of the apo-81 00 sequence are involved in domain structures, limited proteolysis was used as an independent technique. The rate of enzymatic hydrolysis of a peptide bond depends not only upon the chemical environment of the bond but also upon the physical stability of the bond which is determined by the tertiary structure of the protein.? ,8, 19,20 Exposed or loosely folded polypeptide segments are rapidly hydrolyzed.21 Structural domains appear relatively resistant to enzymatic hydrolysis presumably because most potentially hydrolyzable peptide bonds are not exposed to the enzyme or are stabilized by other forces--hydrophobic interactions, hydrogen bonds. Further, domains are relatively stable to unfolding which would otherwise make them increasingly susceptible to hydrolysis. Thus, the goals of the limited proteolysis experiments are, on the one hand, to identify stable mtermediates during the course of proteolysis and, on the other hand, to locate reg1ons highly susceptible to proteolysis.

The experimental approach used was to determine the time course of hydrolysis by incubating the lipoprotein and the proteolytic enzyme at room temperature. At different time points, aliquots of the reaction mixture were removed and the appropriate protease inhibitor was added to terminate the reaction. Peptide fragments were identified in terms of their molecular weights and immunoreactivities toward monoclonal antibodies having known epitopes. Certain fraQments were further characterized with respect to their biophysical characteristics, such as their Stokes rad1i and were studied by electron microscopy to appreciate their structural characteristics.

Figure 2. Time course of hydrolysis of human serum low density lipoprotein apo-81 00 by various proteolytic enzymes as assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. LDL (5 mg/ml) in 10 mM Tris , 1 mM EDTA, pH 7.4 was incubated at room temperature with the indicated enzyme. At various times, 7 ug of protein was removed , the enzymatic reaction was terminated with phenylmethanesulfonyl fluoride and the reaction products analyzed by sodium dodecylsulfate-polyacrylamide gel electrophoresis. The mobility of peptides of known molecular weight is shown at the left. A. chymotrypsin digestion (E:S=1 :100, w/w). B. elastase digestion (E:S=1 :50, w/w) .

Several enzymes having different specificities were used in the limited proteolysis experiments. Two examples serve to illustrate the important points derived from the limited proteolysis experiments. The time course of hydrolysis of LDL apo-8100 by chymotrypsin is shown in figure 2A. It is clear that in the initial states of proteolysis there are several classes of stable fragments of large molecular weight. It is important to emphasize the large size of these stable intermediates which are of the order of 145,000 and 240,000. Similar results were obtained with several enzymes. Stable classes of fragments of molecular wei hts around 145,000 and 240,000 were consistently noted. In add1tion to these large stable intermediates other classes of stable intermediates of somewhat smaller molecular weight have been observed as illustrated in figure 28 by the time course of subtilisin hydrolysis of LDL. In addition to the previously mentioned classes, there is another class of fragments of about 100,000 daltons mass.

In order to assess the location of these stable fragments within the apo-8 sequence, their immunoreactivity was assessed by Western blottinQ with three monoclonal antibodies (figure 3).

Antibody L3 recogn1zes an epitope in the vicinity of res1due 4355 situated at the carboxyl-termmal region of apo-8.2 2 Antibody L7 recognizes an epitope in the vicinity of residue 2331 which is located in the central part of apo-8.2 2 Finally, antibodly SC3 recognizes an epitope situated at the amino-terminal of apo-8 (unpublished results). Using these antibodies it was determined that stable intermediates produced during the course of limited proteolysis are situated in all three regions of the apo-81 00 molecule. This is consistent with the existence of several structural domains in the protein.

Since there are stable intermediates derived from several parts of the apo-81 00 sequence, model fragments from each of the regions were investigated in more detail. One group of stable fragments of 145,000 daltons mass has its origin in the am1no-terminal portion of the molecule. Therefore, the thrombolytic fragment T4 of apo-8 was used as a model for a structural domain situated in this region of the molecule. Thrombin cleaves apo-81 00 at residues 1297 and 3249 yielding three peptide fragments: T4 of about 140 kilodaltons mass, T3 of about 240 kilodaltons and T2 of 170 kilodaltons (figure 3). T3 and T2 are joined by a disulfide bond4. The fragment, T4, was isolated in the presence of a non-ionic, non-denaturing detergent to preserve its structural characteristics. LDL was hydrolyzed by thrombin and the apo-81 00 fragments were prepared by high performance liquid chromatography in the presence of tne detergent C :1 2 E 8. As shown in f1gure 4, two protein fractions labelled I and Tl were obtained. Each fraction was further purified, after concentration, by rechromatography.

The identity of the two protein fractions obtained by high performance liquid chromatography was verified by polyacrylamide gel electrophoresis in the presence of SDS as shown 1n figure 5. Fraction I was shown to contain mainly the fragments T3-molecular weight 240000-and T2-molecular weight 170000-with a small amount of unhydrolyzed apo-81 00 and T4. The fragments T3 and T2 are distinguishable since the gel is run under reducing conditions. Fraction II contains the fragment T4 with little contamination.

The purification of these fragments by high performance liquid chromatography proyide.s.., as well, a measure of their Stokes radii, and thus their form23,2<+,25. From the partition coefficients of T4, T32 and apo-8100, their Stokes radii could be estimated to be 52 A for T4, 110 A for T32 and 132 A for apo-81 00. These values suggest a compact structure for T4 whereas T32 and apo-81 00 are highly asymmetric.

Figure 4. Preparation of thrombolytic fragments of apo-81 00 by high performance liquid chromatography. LDL was proteolyzed with thrombin, incubated with an excess of C12Es and 500 ug of LDL was injected onto a TSK 5000 column equilibrated at 0.5 ml/min. with 20 mM TAPS, 300 mM NaCI, 1 mM C12Es, pH 10 at 20°C.

Electron microscopy was also used to assess the structure of the purified fragments of apo-8100. The images of T4 (fraction II) obtamed by low angle rotary shad.s>wi.Dg of the molecules dried from glycerol containing solutions26,21 is shown in figure 6. The fragment T4 is compact and globular. The size of the fragment, as

assessed by electron microscopy, is in good agreement with the Stokes radius measured by high performance liqu1d chromatography .

Figure 5. Sodium dodecylsulfate-polyacrylamide gel electro­ phoresis of the thrombolytic fragments of apo-8100 separated by high performance liquid chromatography . Peaks I and II, obtained by chromatography of LDL digested with thrombin as described in figure 4, were analyzed by electrophoresis in 5% polyacrylamide gels containing SDS . The mobility of peptides of known molecular weight is indicated on the left.

Figure 6. Electron microscopy of the fragment T4 of apo-8100 . The fragment T4 produced by thrombin digestion of LDL and isolated by high performance liquid chromatography was examined by low angle rotary shadowing of samples dried from glycerol containing solutions . The fragment appears compact but slightly elongated.

The form of T4 was also studied by freeze-drying electron microscopy. In this technique, the sample is rapidly frozen to preserve its structure and the ice is then sublimed away gradually prior to shadowing with a heavy metal. The form of the particle mdicates a compact globular form. The size agrees well w1th that obtained by rotary shadowing electron microscopy and high performance liquid chromatography.

CONCLUSIONS

In conclusion, our experiments involving denaturation of apo-B by guanidine hydrochloride and limited proteolysis suggest the presence of structural domains in apo-B100. To further study the characteristics of one of these domains we used the thrombolytic fragment T4 of apo-B as a model amino-terminal domain. After isolation in a non-ionic, non-denaturing detergent, T4 appeared to be compact and globular. Apo-B1 00 and the fragment T32 were highly asymmetric and flexible. Further studies of structural domains in apo-B100 such as these will help to define the physical organization of the protein and the relationship between 1ts st ructure and its function.

ACKNOWLEDGEMENTS

The authors thank Dr. J. Yon-Kahn, University of Paris XI, Orsay, for use of the circular dichroism instrument and J. C. Dedieu for graphic arts. This work was supported in part by Research Grant HL 18577-11 from the National Institutes of Health.

REFERENCES

1. . M. S. Brown and J. L. Goldstein, A receptor-mediated pathway for cholesterol homeostasis, cjence 232: 34-47 (1986).

2. J. L. Goldstein, M. S. Brown, R. W. Anderson, D. W. Russell and

W. J. Schneider, Receptor-mediated endocytosis: Concepts emerging from the LDL receptor system, Ann. Rev. Cell Bioi. 1: 1-39 (1985).

3. L. F. Soria, E. H. Ludwig, H. R. G. Clarke, G. L. Vega, S. M. Grundy

and B. J. McCarthy, Association between a specific apolipoprotein B mutation and familial defective apolipoprotein B-100, Proc. Natl. Acad. Sci. U.S.A. 86: 587-591 (1989).

4. i. J. Knott, R. J. Pease, L. M. Powell, S. C. Wallis, S. C. Rail, Jr., T. L. lnnerarity, B. Blackhart, W. H. Taylor, Y. Marcel, R. Milne, D. Johnson, M. Fuller, A. J. Lusis, B. J. McCarthy, R. W. Mahley, B. Levy-Wilson and J. Scott, Complete protem sequence and identification of structural domains of human apolipoprotein B, Nature 323: 734-738 (1986).

5. C.-Y. Yang, S.-H. Chen, S. H. Gianturco, W. A. Bradley, J. T.

Sparrow, M. Tanimura, W.-H. Li, D. A. Sparrow, H. DeLoof, M.

Rosseneu, F.-S. Lee, A. M. Gotto and L. Chan, Sequence, structure receptor-binding domains and internal repeats of human apolipoprotein B-100, Nature 323: 738-742.

6. S. W. Law, S. M. Grant, K. Higuchi, A. Hospattankar, K. Lackner,

N. Lee and H. B. Brewer, Jr., Human liver apolipoprotein B-100 eDNA: Complete nucleic acid and derived amino acid sequence,

Proc. Natl. Acad. Sci. U S.A 83: 8142-8146 (1986). M. G. Rossmann and P. Argos, Protein folding, Annu. Rev. Bjochem 50: 497-532 (1981).

P. S. Kim and R. L. Baldwin, Specific intermediates in the folding reactions of small !roteins and the mechanism of protein folding, Annu. Rev. !;liochem. 51: 459-489 (1982).

D. B. Wetlaufer, Nucleation, rapid folding and globular intrachain regions in proteins, Proc, Natl. Acad. Sci. U S.A. 70: 697-701 (1973).

P. L. Privalov, Stability of proteins. Small globular proteins, Adv. Protein Chem. 33: 167-241 (1979).

J. Janin and S. J. Wodak, Structural domains in proteins and their role in the dynamics of protein function, Prog. Bjophys. Mol. Bioi. 42: 21-78 (1983).

J. S. Richardson, lhe anatomy and taxonomy of protein structure, AdY. Protein Chem. 34: 167-339 (1981).

J. R. Wetterau, L. P. Aggerbeck, S. C. Rail, Jr., and K. H.

Weisgraber, Human apofipoprotein E3 in aqueous solution. I. Evidence for two structural domains, J. Bioi. Chem. 263: 6240-6248 (1988).

L. P. Aggerbeck, J. R. Wetterau, K. H. Weis9raber, C.-S. C. Wu and F. T. Lindgren, Human apolipoprotem E3 in aqueous solution. II. Properties of the ammo- and carboxyl-terminal domains, J. Bioi. Qhem. 263: 6249-6258 (1988).

L. P. Aggerbeck, J. R. Wetterau, K. H. Weisgraber, R. W. Mahley and D. A. Agard, Crystallization and preliminary X-ray diffraction stud1es on the amino-terminal (receptor binding) domain of human apolipoprotein E3 from serum very low density lipoproteins, .L Mol. Bioi. 202: 179-181 (1988).

V. Luzzat1, A. Tardieu and L. P. AgQerbeck, Structure of serum low density lipoprotein. I. A solut1on X-ray scattering study of a hyperlipidemic monkey low density lipoprotein, J. Mol.

.fu.Q.L. 131: 435-473 (1979).

T. Gulik-Krzywicki, M. Yates and L. P. Aggerbeck, Structure of serum low density lipoprotein. II. A freeze-etching electron microscopic study, J. Mol. Bioi. 131: 475-484 (1979).

G. C. Chen, S. Zhu, D. A. Hardman, J. W. Schilling, K. Lau and J. P.

Kane, Structural domains of human apolipoprotein B-100. Differential accessibility to limited proteolysis of B-100 in low density and very low density lipoproteins, J. Bioi. _Cham. 264: 14369-14375 (1989) 0 E. Mihalyi, "Application of Proteolytic Enzymes to Protein Structure Studies , 2nd Ed., Vol. 1, CRC Press, Inc., Boca Raton, FL. (1978).

E. Mihalyi, "Application of Proteolytic Enzymes to Protein Structure Studies , 2nd Ed., Vol. 2, CRC Press, Inc., Boca Raton, FL. (1978).

T. E. Creighton, "Proteins. Structures and Molecular Principles", Freeman Publications, San Francisco (1983).

R. J. Pease, R. W. Milne, W. K. Jessup, A. Law, P. Provost, J.-C. Fruchart, R. T. Dean, Y. L. Marcel and J. Scott, Use of bacterial expression cloning to localize the epitopes for a series of monoclonal antibodies against apolipoprotein B100, J. Bioi. Ghem. 265: 553-568 (1990).

M. IeMaire, E. Rivas and J. V. Moiler, Use of gel chromatography for the determination of size and molecular weight of proteins: Further caution, AnaL Bjochem 106: 12-21 (1980).

M. IeMaire, L. P. Aggerbeck, C. Monteilhet, J. P. Andersen and J.

V. Moiler, The use of high performance liquid chromatography for the determination of size and molecular weight of proteins: A caution and a list of membrane proteins suitable as standards, Anal. Biochem. 154: 525-535 (1986).

25. M. IeMaire, A. Ghazi, J. V. Moiler and L. P. Aggerbeck, The use of gel chromatography for the determination of sizes and relative molecular masses of proteins: Interpretation of calibration curves in terms of gel-pore-size distribution, Biochem. J. 243: 399-404 (1987). A. Gulik, L. P. Aggerbeck, J. C. Dedieu and T. Gulik-Krzywicki, Freeze-fracture electron microscopic analysis of solutions of biological molecules, J. Mjcroscopd' 125: 207-213 (1982).

J. E. B. Fox, L. P. Aggerbeck an M. C. Berndt, Structure of ycoprotein lb-IX complex from platelet membranes, J. Bioi.

hem. 263: 4882-4890 (1988).