BINDING OF APOLIPOPROTEINS A TO ADIPOSE CELLS : ROLE OF RECEPTOR SITES IN CHOLESTEROL EFFLUX AND PURIFICATION OF BINDING PROTEIN(S).

BINDING OF APOLIPOPROTEINS A TO ADIPOSE CELLS : ROLE OF RECEPTOR SITES IN CHOLESTEROL EFFLUX AND PURIFICATION OF BINDING PROTEIN(S)

Epidemiological studies have shown a relationship between low concentrations of high density lipoprotein (HDL) cholesterol and the incidence risk of cardiovascular diseases 1'2 . Recent pharmacological studies3 have clearly demonstrated the protective role of HDL and their involvment in reverse cholesterol transport in vi vo4, 5. In that respect apo E-free HDL has been long known to bind to a variety of cells and to promote cholesterol efflux6. Among peripheral tissues, adipose tissue is recognized both in man and rodents for its ability to accumulate, store and, when needed, mobilize a large pool of unesterified cholesterol?' 8. Thus adipose cells represent a cell type suitable to study the first step in reverse cholesterol transport, i.e. cholesterol efflux. Unfortunately adipocytes isolated from adipose tissue loose their viability within a few hours, preventing the analysis of middle-term and long-term responses. During the last decade have been established in our laboratory preadipocyte cell lines from adipose tissue of genetically­ obese ob/ob mice9 and their lean counterpartlO. The validity of these cellular models is supported by i) the biochemical properties of differentiated cells which are similar, if not identical, to those of adipocytes isolated from fat tissue and ii) the ability of undifferentiated cells to differentiate in vivo within a few weeks into fully mature fat cells after their injection into athymic mice,· under conditions where these cells could be unambiguously demonstrated not to be fat cells originating from the host animalll. Most of the studies, if not otherwise stated, were performed with Obl771 cells, a subclone of Obl7 cells established from ob/ob mice12, 13.

CHARACTERIZATION OF LDL AND HDL BINDING SITES AND CHOLESTEROL FLUX/EFFLUX IN OB1771 CELLS

The binding of human apo AI, apo AI! and apo AIV to mouse adipose cells and the study of their functional properties were made feasible owing to extensive homologies existing between rat, mouse and human apolipoproteins 14-16. In addition important homologies do exist between rat (and likely mouse) and human apo B, including the consensus region of apo B and apo E which should be involved in the binding to the apo B,E receptor17-20. The binding of 125r-LDL was competitively inhibited by LDL > VLDL > total HDL ; human LDL and mouse LDL were equipotent in competition assays. Methylated LDL and apo E-free HDL were not competitors. In contrast, the binding of 125r-apo E-free HDL was competitively inhibited by apo E­ free HDL > total HDL and that of 125r-HDL3 by mouse HDL. Thus mouse adipose cells possess distinct apo B,E and apo E-free HDL binding sites which can recognize heterologous or homologous lipoproteins. Further studies of apo E-free HDL binding sites revealed that the binding of 125r-HDL3 was competitively inhibited by apo AI/dimyristoylphosphatidyl­

choline complexes > mouse HDL > HDL3. To explore the possibility that apo AI, apo AI! and apo AIV bind to the same sites, competition experiments were performed in which binding of either of the three apolipoproteins was performed in the presence of the two other unlabeled apolipoproteins. The results suggest strongly that apo AI, apo AI! and apo AIV bind to common receptor sites21. This hypothesis is supported by the finding that a highly purified protein from Ob1771 cells remains able to bind the three apolipoproteins (ref.22 and vide infra) . The observation that - 1 mol of apo AI! is bound per 2 mol of apo AI or apo AIV (Table I) could be explained if one assumes that the receptor site is a dimeric structure (see Fig.l) which recognizes each monomer of dimeric apo AI! in the same way that it recognizes two molecules of monomeric apo AIV or apo AI. In any event, the stoichiometry of apo AI (or apo AIV) versus apo AI! binding has been consistently observed in intact cells and in homogenates after detergent solubilization as well as after extensive purification of binding proteins of 80 and 92 kDa (ref.22 and vide infra).

During the course of these studies, it was observed that the endogenous cholesterol synthesis was nil12 but the most striking observation was the fact that long-term exposure of adipose cells to LDL and HDL3 did not affect the number of apo B,E receptor sites and that of apo E-free HDL receptor sites. In other words, the "buffering" capacity of adipose cells seems limited with respect to the regulation of cholesterol content. This lack of cholesterol homeostasis would explain the rather unique ability of adipose tissue in vivo to accumulate and mobilize a large pool of unesterified cholesterol 7,B. Since differentiated Ob1771 cells were able to find, internalize and degrade LDL12, it appeared that adipose cells did not show an efficient cholesterol homeostasis in vitro and thus, as a first prediction, should accumulate cholesterol. The second prediction was that cholesterol­ preloaded cells should mobilize cholesterol when exposed to appropriate lipoprotein particles. Both predictions were fullfilled. As shown in Table I, it is of interest to note

that comparisons between apparent Kct values for binding of apo AI and apo AIV and the ECso values for cholesterol efflux are within the same range of concentrations2l-2 4 . These results suggest that specific binding to these distinct sites was a pre-requisite to cholesterol accumulation and subsequently to cholesterol mobilization. It is also of interest to note that cholesterol accumulation was taking place in the presence of LDL under the form of unesterified cholesterol only, in agreement with the fact that, at least in rat fat tissue, the majority (75-95%) of adipocyte cholesterol is unesterified and associated with central oil (triacylglycerol) droplet7,8.

RELATIONSHIPS IN ADIPOSE CELLS BETWEEN RECEPTOR SITES FOR APO AI, APO AI I AND PROMOTION OF CHOLESTEROL EFFLUX

In order to establish whether receptor sites for HDL were indeed required for the promotion of cholesterol efflux, use was made of Obl7 cells in which have been induced genetically defined alterations of the growth control mechanism by transferring cloned oncogenes25. Obl 7PY cells were obtained after transfer of the complete early region of polyoma virus whereas Obl 7MT cells were obtained after transfer of a modified genome encoding only the middle T protein. The broad range of phenotypes thus generated has also offered us unique opportunities to study cholesterol efflux in adipose cells as cells of the Obl7MT18 subclone had a 3-fold higher number of HDL receptor sites than cells of the parental Obl7 clone whereas growing Obl 7PY cells did not have any detectable sites (see below) .

As a pre-requisite to study the critical role, if any, of HDL receptor sites and to undertake their purification, conditions for their visualisation were searched and found using bivalent cross-linking reagent discuccimidyl suberate at 4°C in the presence of apo AI-containing liposomes and intact Obl771 cells or derived crude membranes26. The existence of two specific cell-surface protein components of Mr 100,000 and 130,000 was demonstrated. It is possible that two different proteins of Mr -70,000 and -100,000 able to bind one molecule of proteins of apo AI of Mr 28,000 are indeed present in adipose cells. Alternatively the possibility of either a single glycoprotein able to bind one molecule of apo AI but having different degrees of glycosylation, or a single glycoprotein able to cross-link one or two molecules of apo AI, could be envisionned. The key observation in our study on the role of HDL receptor sites in the promotion of cholesterol efflux was that no binding of HDL3, apo AI, apo AI! or apo AIV was observed in growing Obl7PY cells and derived crude membranes, in contrast to growing or growth­ arrested Obl771 cells (see Table I) or Obl7MT18 cells (not shown). After thymidine block, growth-arrested Obl7PY cells became able to recover in parallel binding activities for HDL3, apo AI, apo AI! and apo AIV. The possibility that this recovery was an event common to various cell surface receptors is not very likely since apo B,E and transferrin receptor sites were both present in growing and growth-arrested Obl 7PY cells as well as in Obl 771 cells. The recovery of HDL receptor sites in growth-arrested Obl 7PY cells was rapid (16 h) and prevented in actinomycin D- or cycloheximide­ treated cells, adding further support to the conclusion that these sites are protein component (s) . When experiments of cholesterol efflux were performed, the results showed that, after cholesterol accumulation taking place in the presence of LDL cholesterol, subsequent exposure to HDL3 or apo AI (but again not apo AI!) led to cholesterol efflux from Obl771 cells and growth- arrested Obl7PY cells but not from growing Obl7PY cells26. Thus it appears that the presence of high-affinity receptor sites for HDL in intact adipose cells is required for the promotion of cholesterol efflux. The existence of cell surface binding sites which recognize apolipoproteins A is supported by recent experiments showing that the binding of apo AI/DMPC complexes to intact Obl771 cells was followed within 1-2 minutes by the formation of diacylglycerol from phosphatidylcholine as substrate ; it is of interest that apo AII/DMPC complexes were inactive in that respect, supporting the view that apo AI! was playing the role of an antagonist21,24,27,28. Altogether, these observations led us to attempt in purifying apo A binding proteins by using Obl7MT18 cells, a transformed cell line enriched 3-fold in apo A binding sites as compared to the parental Obl7 cells.

The purification scheme is shown in Figure 2. An 1,400- fold purification over the starting crude homogenate was achieved22. The purified material contained two proteins that were both able to bind apolipoproteins AI, AI! and AIV but not LDL. Glycopeptidase F treatment showed the existence of a single protein bearing either N-linked high-mannose or complex oligosaccharide chains. The purified material showed an apparent molecular mass of 80 ± 9 kDa by high-pressure liquid chromatography on TSKG 3000 SW column. Rabbit polyclonal antibodies directed against the purified material revealed two protein bands of 80 and 92 kDa after sodium dodecyl sulfate polyacrylamide gel electrophoresis under reducing conditions and immunoblotting. These bands were undetectable in growing Obl7PY cells previously shown not to bind the various apo As or not to undergo cholesterol efflux, whereas they were conspicuous in growth-arrested Obl7PY cells which recovered these properties. It is of utmost importance to recall that these binding sites are present at the cell surface of intact cells but more than 90% of apo A and apo B,E (LDL) binding sites were shown to be present intracellularlyl2. This situation is similar to that observed in skin fibroblasts where a large proportion of LDL binding activity is also present within the cells. Therefore both cell surface and intracellular binding sites were purified in the present study, but it must be recalled that the affinities of these binding sites for their ligands were very similar in intact Obl7 cells and derived crude membranesl2 and that the binding parameters were found to be very similar for intact Obl7MT18 cells and the fraction purified from these cells by DEAE­ Trisacryl chromatography (Fig.2). Thus it is assumed that cell surface binding sites and intracellular binding sites are identical and that a receptor recognizing apolipoprotein A has been purified. Figure 1 summarizes our working hypothesis : it is possible that the functional apolipoprotein A receptor, required for cholesterol efflux but not for

binding activity (Fig.1), is a dimer of two single polypeptide chains. This dimeric structure would be able to recognize either one mole of the dimeric apo AI! or two moles of the monomeric apo AI! or AIV. If so, we envision that the binding of one molecule of apo AI (or apo AIV), but not that of apo AII, might induce a conformational change allowing the binding of a second molecule of apo AI (or AIV) and the formation of an activated receptor. It is suggested that, within the apo AI (or apo AIV)/DMPC complexes, the apolipoprotein plays the role of a ligand triggering the PKC pathway (cholesterol translocation to the cell surface) whereas the liposomal structure per se plays the role of a cholesterol acceptor (cholesterol efflux from the cell surface) . Recent experiments indicate that it is indeed the case and that a distinction between both events can be made experimentaly (N. Theret-Bidoui et al., unpublished work).

ACKNOWLEDGEMENTS

The authors wish to thank Miss V. Boivin and Mrs. B. Barhanin for expert technical help, Dr. J. Barhanin for helpful advice in cross-link experiments and Mrs.G. Oillaux for expert secretarial assistance. This work was supported by the "Centre National de la Recherche Scientifique" (CNRS UPR 7300), by the "Fondation pour la Recherche Medicale Francaise" (Nice) and by "Institut Pasteur" (Lille).

REFERENCES

1. T. Gordon, W.P. Castelli, M. C. Hjortland, W. B. Kannel and T. R. Dawber, High density lipoprotein as a protective factor agains coronary heart disease: the

Frammingham Study, Am· J. Med, 62:707 (1977).

2. N. E. Miller, 0. H. Forde, D. S. Thelle and 0. D. Mjos, The Tromso Heart Study: high-density lipoprotein and coronary heart disease: a prospective cas control study, Lancet 1:965 (1977).

3. M. Heikki Frick et al., Helsinki heart study: Primary­ prevention trial with gemfibrozil in middle-aged men with dyslipidiema: safety or treatment, changes in risk factors, and incidence of coronary heart disease, N. Engl. J Med. 317:1237 (1987).

4. N. E. Miller, A. La Ville and D. Crook, Direct evidence that reverse cholesterol transport is mediated by high­ density lipoprotein in rabbit, Nature 314:109 (1985).

5. Th. J.C. van Berkel, H. F. Bakkeren, F. Kuipers and R. J. Vonk : Proceedings of Xth International Symposium on Drugs Affecting Metabolism, p.47 (1989).

6. A. R. Tall and D. M. Small, Body cholesterol removal: role of plasma high-density proteins, Adv. Lipid Res. 17:1 (1980).

7. R. K. Krause and A. D. Hartman, Adipose tissue and cholesterol metabolism, J. LiPid Res. 25:97 (1984).

8. A. Angel and B. Fang, Lipoprotein interactions and cholesterol metabolism in human fat cells, in: "The Adipocyte and Obesity: Cellular and Molecular Mechanisms", A. Angel, C. H. Hollenberg and D. A. K. Roncari, eds., Raven Press, New York (1983).

9. R. Negrel, P. Grimaldi and G. Ailhaud, Establishment of preadipocyte clonal line from epididymal fat pad ob/ob mouse that responds to insulin and to lipolytic hormones, Proc. Natl. Acad. Sci. USA 75:6054 (1978).

10. C. Forest, A. Doglio, L. Casteilla, D. Ricquier and G. Ailhaud, Expression of the mitochondrial uncoupling protein in brown adipocytes. Absence in brown preadipocytes and BFC-1 cells. Modulation by isoproterenol in adipocytes, Exp. Cell Res. 168:233 (1987).

11. D. Gaillard, P. Poli and R. Negrel, Characterization of ouabain-resistant mutants of the preadipocyte Obl 7 clonal line. Adipose conversion in vitro and in vivo, Exp Cell Res. 156:513 (1985).

12. R. Barbaras, P. Grimaldi, R. Negrel, and G. Ailhaud, Binding of lipoproteins and regulation of cholesterol synthesis in cultured mouse adipose cells, Biochim. Biophys. Acta 845:492 (1985).

13. R. Barbaras, P. Grimaldi, R. Negrel, and G. Ailhaud, Characterization of high-density lipoprotein binding and cholesterol efflux in cultured mouse adipose cells, Biochim. BioPh Acta 888:143 (1986).

14. P. Forgez, M. J. Chapman, S. C. Rall Jr. and M. C. Camus, The lipid transport system in the mouse, Mus musculus: isolation and characterization of apolipoproteins B, AI, AII, and CIII, J. Lipid Res. 25:954 (1984).

15. C. G. Miller, T. D. Lee, R. C. LeBoeuf and J. E. Shively, Primary structure of apolipoprotein AI! from inbred mouse strain BALB/c, J. Lipid Res. 28:311 (1987)

16. S. C. Williams, S. M. Bruckheimer, A. J. Lusis, R. C. LeBoeuf and A. J. Kinniburgh, Mouse apolipoprotein AIV gene: nucleotide sequence and induction by a high-lipid diet, Mol. Cell. Bjol. 6:3807

(1986).

17. T. L. Innerarity, E. J. Friedlander, S. C. Rall, K. H. Weisgarber and R. W. Mahley, The receptor-binding domain of human apolipoprotein E, J. Biol. Chern.

258:12341 (1983).

18. T. J. Knott, R. J. Pease, L. M. Powell, S. C. Wallis, S. C. Rall Jr., T. L. Innerarity, B. Blankhart, W. H. Taylor, Y. Marcel, R. W. Mahley, B. Levy-Wilson and J. Scott, Complete protein sequence and identification of structural domains of human apolipoprotein B, Nature 323:734 (1986).

19. C. Y. Yang, S. H. Chen, S. H. Gianturco, W. A. Bradley, J. T. Sparrow, M. Tanimura, W. K. Li, D. A. Sparrow, H. DeLoof, M. Rosseneu, F. S. Lee, Z. W. Gu, A. M. Gotto Jr. and L. Chan, Sequence, structure, receptor binding domains and internal repeats of human apolipoprotein B-100, Nature 323:738 (1986).

20. A. J. Lusis, R. West, M. Mehrabian, M. A. Reuben, R. C. LeBoeuf, J. S. Kaptein, D. F. Johnson, V. N. Schumaker, M. P. Yuhasz, M. C. Schatz and J. Elovson, Cloning and expression of apolipoprotein B, the major protein of low and very low density lipoproteins, Natl. Acad. Sci. USA 82:4597 (1985).

21. A. Steinmetz, R. Barbaras, N. Ghalim, V. Clavey, J.C. Fruchart, and G. Ailhaud, Human apolipoprotein AIV binds to apolipoprotein AI/AI! receptor sites and promotes cholesterol efflux from adipose cells, J. Biol. Chern. 265:7859 (1990)

22. R. Barbaras, P. Puchois, J.C. Fruchart, A. Pradines­ Figueres and G. Ailhaud, Purification of the apolipoprotein A receptor from mouse adipose cells, Biochem. J., in press.

23. R. Barbaras, P. Puchois, J.C. Fruchart and G. Ailhaud, Cholesterol efflux from cultured adipose cells is mediated by LpAI particles and not by Lpai :AI! particles, Biochem. Bjophys Res. Commun. 142:63 (1987) .

24. R. Barbaras, P. Puchois, P. Grimaldi, A. Barkia, J.C. Fruchart and G. Ailhaud, .in.: "Eicosanoids, Apolipoproteins, Lipoprotein Particles, and Atherosclerosis", C. L. Malmendier and P. Alaupovic, eds., Plenum Press, New York (1988).

25. P. Grimaldi, D. Czerucka, M. Rassoulzadegan, F. Cuzin and G. Ailhaud, Obl 7 cells transformed by middle-T-only gene of polyoma virus differentiate in vitro and in vivo into adipose cells, Proc Natl Acad. Sci. USA 81:5440 (1984).

26. R. Barbaras, P. Puchois, P. Grimaldi, A. Barkia, J.C. Fruchart and G. Ailhaud, Relationship in adipose cells between the presence of receptor sites for high density lipoproteins and the promotion of reverse cholesterol transport, Biochem. Biophys. Res. Commun.

149:545 (1987).

27. C. Delbart, N. Theret, G. Ailhaud, J.C. Fruchart, Phosphatidylcholine breakdown during receptor binding of HDL3, Circulation 80:487 (1989).

28. A. Barkia, P. Puchois, N. Ghalim, R. Barbaras, G. Ailhaud and J.C. Fruchart, Differential role of apolipoprotein AI-containing particles in cholesterol efflux from adipose cells, submitted.