Prog. Lipid ICes.Vol.30, No. 2/3,pp. 259-266,1991 Printed in GreatBritain.All fightsreserved

0163-7827/91/$0.00+ 0.50 © 1991PergamonPressplc

APOLIPOPROTEIN D--AN ATYPICAL APOLIPOPROTEIN PmLIP K. WEECH,*~ PIERRE PROVOST,~ NATHALIEM. TREMBLAY,~

RrNo N. CAMATO,§ROSSW. MILNE,§ YVESL. MARCEL§ and Emc RASSART~ tMerck Frosst Centre for Therapeutic Research, Merck Frosst Canada Inc., Pointe Claire-Dorval, Quebec, Canada Universite du Quebec d Montreal, Montreal §Institut de recherches cliniques de Montreal, Montreal, Canada

I. II. III. IV. V.

VI. VII.

CONTENTS ABB~C~rIONS INTRODUCTION STRUCTUREANDPROPERTIESOFApoD IMMUNOOmMICALSTUDmS A. Distribution of ApoD in plasma B. Monoclonal antibodies and antigen heterogeneity At,nMALSTUDmS SITESOVGErmEXPp.~SSION A. Organs and tissues B. Cells C. Regulation of expression CLINICALSTUDIES SUMMARY ACKNOWLEDGEMENTS ~NCES

259 259 260 261 261 261 262 263 263 264 265 265 265 265 265

ABBREVIATIONS Apo--apolipoprotein HDL--high density lipoproteins LDL--low density lipoproteins VLDL--very low density lipoproteins LCAT--lecithin:cholesterol acyl transferase CETP-----cholesterylester transfer protein I. INTRODUCTION Petar Alaupovic, Walter McConathy and their colleagues discovered and pioneered studies o f Apolipoprotein D (ApoD). Therefore, it could easily have been a case of the English expression "bringing coals to Newcastle" to report studies of ApoD at the Oklahoma Medical Research Foundation. However, during the past six years, there has been a remarkable growth in our understanding of ApoD. It was an exceptional pleasure for Philip Weech, therefore, to return and present our studies and this review of ApoD at the symposium honoring Dr Petar Alaupovic and his contributions to lipoprotein research. Nevertheless, 18 years after the first report of ApoD, the goal remains to discover the role o f this protein in normal tissue metabolism, and its function in the plasma lipoprotein system. It has been obvious, since it was first observed in the late 1960s and its partial characterization in 1973, that ApoD is a distinct component o f the human plasma lipoprotein system, but that its properties are unlike those of the better-known apolipoproteins. The lack o f a statement of its physiological role hindered widespread research on ApoD, but gave us an incentive to tackle the following problem: given good evidence for the existence o f this protein, can we discover its role and function using systematic structural studies as the basis? *Corresponding address: Merck Frosst Centre for Therapeutic Research, Merck Frosst Canada Inc., CP/PO 1005, Pointe Claire-Dorval, Que., H9R 4P8, Canada. 259

260

P.K. WEECHet al.

The opposite, straightforward approach is to begin with an observation of metabolism and then proceed to elucidate the mechanism and proteins that effect it. This was not possible with ApoD, since none of the early studies identified a metabolic event or disorder that depended on it. Our understanding of its role has come so far from observations and correlation of (1) the structure of the protein, its gene, and homology with known proteins, (2) where the protein is synthesized and under what conditions, and (3) where it is subsequently found, and with which other proteins and molecules it associates. This now helps us to place ApoD into a metabolic framework as a local lipid transporter, and experiments testing relevant hypotheses of its function should be forthcoming. II. STRUCTURE AND PROPERTIES OF ApoD The discovery of ApoD started with the observations of the "thin-line polypeptide," referring to the sharp immunoprecipitin reaction found in double immunodiffusion experiments with antisera raised against human high density lipoproteins (HDL) or impure ApoA-I. 2'7'26 Several investigators that observed the thin-line polypeptide subsequently described other novel apolipoproteins and it is possible that some other proteins could have been separately responsible for the narrow immunoprecipitin line. The discovery is more properly ascribed to the isolation and partial characterization of ApoD that was reported by McConathy and Alaupovic3~'32 and the protein called apolipoprotein A-III by Kostner. 27 They described a novel glycoprotein (with 18% carbohydrate) present in the HDL either on its own family of lipoprotein particles 3~ or on a family including other ApoA proteins. 27 The apparent Mr of ApoD on SDSpolyacrylamide gel electrophoresis is about 30,000 but this is greater than the true Mr of the protein (19,300) due to the effect of the carbohydrate. ApoD is polymorphic in charge5 with four major isomorphs of pI 4.7-5.2, mainly due to variation in sialic acid content. 45 The partial protein sequence of human ApoD was determined by Drayna et al. in 1986 with the complete sequence inferred from the cDNAfl Both this structure and the exon-intron structure of its gene 19 showed that the protein was homologous to retinol-binding protein and what was then called the ~t2-microglobulin superfamily of proteins. This family has grown rapidly over the past few years with the discovery of other proteins having homologous sequences, they are now called the "ligand-carrier protein" or Lipocalin family. The homologous sequences form the motifs Asn-hyd-acidichyd-X-basic-hyd-X--Gly-X-Trp-aro-X-hyd-hyd-hyd-hyd and aro-X-hyd-hyd-XThr-Asp-Tyr-Asp-X-aro-hyd-hyd-hyd (where acidic is Asp or Glu; basic is Arg or Lys; hyd is a hydrophobic residue Ala, Ile, Leu, Met, Phe, Tyr, Val, or Trp; aro is aromatic, Phe or Tyr; and X is any residue).35 The three-dimensional structure of some members of the Lipocalin family (retinolbinding protein, beta-lactoglobulin, insecticyanin) has been determined by X-ray crystallography. They conform to an 8-stranded beta-barrel structure surrounding a single molecule of the lipophilic ligand. This suggests that the members of the protein family serve the role of transporters, each molecule of protein transporting or protecting a single ligand molecule. Each protein is likely to be specific for a single ligand, or a chemically related class of ligands. Peitsch and Boguski u recently compared the primary structure of ApoD with other Lipocalins and concluded that ApoD was most similar to insecticyanin from the Tobacco Hornworm, M a n d u c a sexta. Using the coordinates of the latter, they hypothesized a molecular model for the three-dimensional structure of the ligand-binding pocket of ApoD. They concluded that the ligand most likely to fit was a heme-related compound, and less likely to be cholesterol. A preliminary experiment showed that ApoD could bind bilirubin in a one-to-one molar ratio. One of the most informative observations appeared in the report of the primary structure of a major protein component from the fluid of human breast gross cystic disease, GCDFP-24. 8 Peptides were isolated from a trypsin digest of GCDFP-24 and their amino

Apolipoprotein D

261

acid sequences were determined. Twelve peptides and the unbloeked N-terminal structure were identical to sequences within ApoD. Therefore, it was concluded that GCDFP-24 is ApoD. GCDFP-24 had been previously identified as a progesterone-binding protein of cyst fluid, and its steroid-binding specificity has been reported. ~ The greatest affinity was found with pregnenolone, progesterone, 5-beta-pregnan-3-ol-20-one, 5-alpha and 5-beta-pregnan-3,20-dione, and some synthetic steroid analogs. However, binding affinity was relatively low (about 1 x 106 l/mol). Thus, Balbin et aL 8 reported a novel site of accumulation of ApoD (about 30 mg/ml) and a potential transport role for another ligand. However, the competitive binding of progesterone and heme compounds has not been reported. ApoD appears to associate as a complex with Mr 81,000 in breast cyst fluid. 33Cyst fluid also contained a protease activity that was coisolated with GCDFP-24/ApoD. 24'25It remains to be determined whether ApoD is a protease itself, if it associates with a protease, or if the protease simply has similar chromatographic properties. III. IMMUNOCHEMICAL STUDIES A. Distribution of ApoD in Plasma There is no doubt that ApoD can be classified as a plasma apolipoprotein in man, the rabbit and baboon, and it is likely to be an apolipoprotein in other mammals, except perhaps the rat and sheep. 9'14'3~'43'45 Immunoreactive ApoD was found mainly in the dense HDL, but a few percent were found in the low density- (LDL) and very low density-lipoproteins (VLDL), and essentially no free ApoD was found with the plasma proteins. A possible insight into the role of ApoD came from observations of the association of ApoD with the enzyme LCAT lecithin:cholesterol acyl transferase in blood plasma. 5,~s,2°,2~,4~ The function of the ApoD-LCAT complex remains to be elucidated, as Kostner did not find a clear cofactor function for ApoD, although he found that LCAT activity was stabilized over time by ApoD. 4° It can be conjectured that one function of ApoD would be to present one or both of the substrates, phosphatidylcholine and cholesterol, to LCAT or accept the products of the reaction, lysophospholipid and cholesteryl ester. However, the Lipid Transfer Protein CETP is believed to be responsible for all of the transfer of cholesteryl esters between lipoproteins ~2 and albumin could bind the lysophospholipid. Thus, the function of ApoD with LCAT remains a mystery, although its association with the H D L and this enzyme presumably has some significance. B. Monoclonal Antibodies and Antigen Heterogeneity We prepared and characterized monoclonal antibodies against ApoD. 45 With these antibodies, we were able to deafly measure the qualitative and quantitative distribution of ApoD antigen in the plasma lipoprotein and protein fractions of man and the rabbit. 14'36'45 The physico-chemical characteristics of ApoD detected with the antibodies agreed with those that were known for the isolated pure protein, and the latter crossreacted with both the monoclonal antibodies and an antiserum prepared by McConathy. 45 In addition to ApoD in man, we detected up to seven additional proteins that crossreacted with the antibodies. 14'45 All normolipidemic subjects analyzed had at least some of these crossreacting proteins present in their plasma, t4 All of the ApoD in plasma was found entirely in the lipoprotein fractions, and similarly all of the crossreacting proteins were lipoprotein bound, except for some of the Mr 94,000 protein. Thus, there was no free ApoD in the plasma protein fraction, and the crossreacting proteins were also apolipoproteins. There seemed to be a progressive enrichment of the higher Mr crossreacting proteins with increasing density of the lipoprotein. ~4

262

P . K . WEEC'net al.

FIG. 1. The presence of a heterodimer of ApoB-ApoD and free ApeD in human L D L LDL proteins (nonreduced) were separated by electrophoresis with SDS in 4% polyaerylamide gels, and after electrotransfer to nitrocellulose proteins were detected with monoclonal anti-ApeD 4El I or anti-ApoB ID1.

A small proportion of human plasma ApeD was found in the LDL and some of it was present as a mixed disulphide complex between either ApoB-100 or ApoB-74 and ApeD 14 (Fig. 1). This is analogous to the complex that forms between ApoB and ApoLp (a). 42 There is also some ApeD in the VLDL.14'43 It would be interesting to know if ApeD affects the function of ApoB and vice versa in these complexes. We isolated the human Mr 38,000 protein, that crossreacted with monoclonal antibody 4Ell, and in collaboration with Dr Cynthia Wadsworth (University of Wisconsin Biotechnology Center) determined what we believe to be 40% of its primary structure. The sequences that we found were homologous to ApeD and ApoA-II, suggesting that a complex of these two proteins could constitute the Mr 38,000 protein (manuscript in preparation). IV. A N I M A L

STUDIES

ApeD was described as a component of the plasma lipoproteins of the baboon) The HDL of rabbit, pig, dog, cow, goat, sheep, cynomolgous and rhesus monkeys contained a protein that crossreacted with our monoclonal antibody 5G10 (anti-human ApeD), having Mr 29,000 like human ApeD. 36 The distribution of this ApeD in rabbit plasma, primarily in HDL, was similar to that in man except that the crossreacting proteins were not f o u n d . 36 Rabbit ApeD also had a similar charge heterogeneity to human ApeD, Fig. 2.

Apolipoprotein D

pI

5.19

263

4.66

1 MP --- 9/,000 ,-- 67000

43000 30000 20100 "- 1 4 4 0 0

FIG.2. Heterogeneityof isoelectricpoint of rabbit ApoD. Rabbit HDL proteins were separated by 2-dimensional isoelectric focusing-SDS-polyacrylamidegel electrophoresis, transferred to nitrocelluloseand ApoD was detectedusingmonoelonalantibody5G10, ~25I-anti-mouseIgG and autoradiography. We determined the nucleotide sequence of rabbit ApoD cDNA following molecular cloning from a testis library using an RNA probe derived from the human ApoD cDNA. 36 Eighty percent of the protein sequence inferred for rabbit ApoD was homologous with human ApoD and its mRNA was present in many organs, like human mpoD. 36Thus, we concluded that the rabbit is a good model animal with which to study tissue and organ expression of ApoD synthesis. In contrast, in the rat we did not find any ApoD in the plasma that crossreacted with our monoclonal antibody 5G10, 36 although Spreyer et al. reported that 5G10 crossreacted with ApoD in lipoproteins isolated from regenerating rat sciatic nerve 39 and they cloned and sequenced rat ApoD cDNA from that site. While the animal species that have been investigated have been important to our understanding of ApoD in tissues, it seems that they have not yet helped us to understand the functions of ApoD in the plasma lipoproteins. It is possible that these functions are different in man from those in other mammals, since it is only man that has been found to have the higher molecular weight crossreacting proteins, so far. V. SITES OF GENE EXPRESSION A. Organs and Tissues

Given the similarity of rabbit and human ApoD described above, we used the rabbit to quantify the relative levels of ApoD mRNA in organs. The striking finding from this study was that the distribution of ApoD mRNA was unlike any of the other apolipo-

264

P.K. W~CHet al.

proteins. ApoD mRNA was found in many different organs and not only in the liver and intestine. ]s'36 ApoE, too, is synthesized in organs other than liver and intestine. However, unlike ApoE, many organs had a level of ApoD mRNA that was much higher than that of the fiver: the spleen being 59 times that of the liver, and is probably more important as a site of synthesis even after allowing for organ weight. ~ Organs rich in ApoD mRNA were spleen, adrenal gland, lung, brain, testis, kidney and heart. Drayna et a l f l had noted a similar distribution in some organs in nonquantitative Northern analysis of human ApoD mRNA. B. Ce//s In situ hybridization of ApoD mRNA has been used by Smith et al., 3s Boyles et al., 11 Spreyer et al. 39 and ourselves to determine which cell types contain the most

ApoD mRNA. First, although all organs that we examined contain some ApoD mRNA, not all cells do. Second!y, the cell type that appears positive in most studies is fibroblast-like, although this cell-type has not been noted to be a producer of other apolipoproteins. In the Rhesus monkey, ApoD mRNA was found in neuroglial cells, cells in the subarachnoid space of the brain, including pial and perivascular cells, and scattered neurones in the brain. 3s In other organs, it was seen in interstitial and connective tissue fibroblasts, often associated with blood capillaries. In regenerating rat sciatic nerve, ApoD mRNA was found in endoneurial fibroblasts, but not in Schwann cells, macrophages, perineurial or epineurial cells. 39 ApoD was found to be synthesized by segments of regenerating rat sciatic nerve. H Our results of in situ hybridization in the rabbit (manuscript in preparation) are in general agreement with the reports above. In particular, the highly vascularized connective tissue supporting the seminiferous tubules of the testis, the ductus epididymis and the ductili efferentes were notable sites of expression in male genital organs. Capillaryassociated cells of the adrenal glands expressed ApoD mRNA. Peribronchiolar and periarteriolar connective tissues or fibroblasts of the lung were positive. Scattered glial cells in the white matter throughout the central nervous system expressed ApoD mRNA, but the number of positive cells in the grey matter was very low compared with the white matter. The red pulp of the spleen showed a very strong hybridization, most probably in the recticular cells of the spleenic cords. A weaker level of hybridization was seen with the tissues of the capsule, trabeculae and white pulp. Evidence was reviewed recently for recognizing a specialized form of fibroblastic cell in the spleen, referred to as a Barrier cell. ~ This cell-type was often found adjacent to blood vessels, and most closely resembled reticular cells. Barrier cells were dense, dendritic, fibroblast-like and proliferated in response to Interleukin-1, infectious disease and imperfect blood cells. It would be interesting to know if these barrier cells express ApoD mRNA, given their similar anatomical distribution to the ApoD mRNA-containing cells that we observed in rabbit spleen. The results above, from in situ hybridization with ApoD mRNA, show that synthesis of ApoD is probably restricted to a few cell types, including some, but not all, cells of the fibroblast lineage. In contrast, reports of immunocytochemistry with antibodies against ApoD m°.m2.3° have found that many cell-types contain ApoD protein, notably those that we believe to be weak or negative for mRNA by in situ hybridization. This suggests that many cells accumulate ApoD, perhaps by endocytosis from the extracellular fluid. The role of ApoD, therefore, could be either (1) the local transport of a ligand from one cell to another within an organ, (2) the scavenging of a ligand within the organ for transport to the blood circulation, or (3) the local transport of a ligand that is delivered to tissues by the blood circulation, but is transported in a controlled fashion from the perivascular cells within the tissue. Each of these three possibilities could include the biochemical modification of the ligand in the cells that synthesize ApoD.

Apolipoprotein D

265

C. Regulation o f Expression

Two reports H,39have described 40 to 500-fold increases in ApoD expression in the sciatic nerve of the rat, rabbit and marmoset following crush injury and regeneration. Noninjured mature nerve and nerve that was prevented from regenerating had lower levels of ApoD mRNA. The endoneurial fibroblasts synthesized ApoD, which then accumulated on lipoproteins in the endoneurial extracellular space. The biochemical stimulus for increased ApoD synthesis in the nerve has not been reported yet. In cultured breast cancer cell lines, estrogen decreased, but androgen increased ApoD secretion. Cell proliferation was regulated in the opposite sense by estrogen and androgen. 37 vI. CLINICAL STUDIES It is likely that the most interesting clinical studies of ApoD have not been performed yet. It is only recently that ApoD was recognized as a major component of human breast gross cystic disease fluid and as a major protein in animal models of nerve regeneration, and these observations have not yet been exploited. Most studies have reported that normal plasma ApoD concentration is about 12 rag/100 ml. 14,t6Among studies of the dyslipoproteinemias and patients at increased risk of ischemic heart disease there were no clear indications of any physiologic or pathophysiologic role of ApoD in lipoprotein metabolism. Low plasma levels of ApoD (about one-half of normal) were found in disorders that produce low HDL-cholesterol levels with or without hypertriglyceridemia. These included Tangier disease, a-beta-lipoproteinemia, LCAT deficiency, Lipoprotein Lipase deficiency, glucose-6-phosphatase deficiency, and hypertriglyceridemia at birth 1'3'4'13'23'2s and treatment with Stanozolol.6 However, no complete deficiencies, nor supranormal levels, of plasma ApoD have been reported. VII. SUMMARY The structure of ApoD and its sites of synthesis have been discovered. These characteristics differ from those of the other apolipoproteins. The role of ApoD in the plasma lipoprotein system remains to be discovered, but the recent, rapid increase in our knowledge of this protein suggests that it plays an important role in the homeostasis or housekeeping of probably all organs. One of its functions is likely to be the transport of a hydrophobic ligand (a lipid) in a one-to-one molar ratio with itself. This transport is likely to occur unidirectionally between neighboring cells in an organ, and between perivascular cells and the blood circulation. The chemical structure of the natural ligand, or ligands, of ApoD in normal cells in vivo or in culture is not known, but ApoD has been shown to bind some steroids and bilirubin. Remarkable upregulation of synthesis of ApoD has been observed during regeneration of injured peripheral nerves. Perhaps the physiologic role of ApoD will prove to be more interesting and of equal importance in biology to the roles of the other apolipoproteins in cardiovascular disease. Acknowledgements--The studiesby the authors were supportedby grants from the Medical ResearchCouncil

of Canada (PG-27, MT-9880)and the QuebecHeart Foundation,and wereconductedmainlyat the Institutde recherches cliniques de Montrtal and Universit6 du Quebec a Montrtal. Louise Charlton gave excellent secretarial help. (Received 28 M a y 1991) REFERENCES

1. At~upovIc,P. and FERNANDES, J. Pediatr. Res. 19, 380-384 (1985). 2. AI.AUPOVlC,P., LEE,D. M. and McCONATHY,W. J. Biochim. Biophys. Acta 260, 689-707 (1972). 3. ALAUPOVlC,P., MCCONATWLW. J., Cun~Y,M. D., MAON~, H. N., TOnSVlK,H., B~o, R. and GJONE,E. Scand. J. Clin. lab. Invest. 33 (Suppl. 137): 83-87 (1974). 4. ALAUPOVlC,P., SCHA~reK,E. J., MCCONATHY,W. J., Festal, J. D. and BREwI~,H. B., JR Metabolism 30, 805-809 (1981).

266

P. K. W~CH et al.

5. ~ , J. J., ~ o , M. C., EWENS,S. L. and TOLL~FSON,J. H. Atherosclerosis 39, 395--409 (1981). 6. ALiaS, J. J., TAOOAaT,H. M., API'LJmAuM-B^oWD~,~,D., HAFFN~, S., Cm~NUT, C. H. and H^ZZAaD, W. R. Biochim. Biophys. Acta 795, 293-296 0984). 7. A~,AULT-JARm~, M., L~vY, (3. and POLONOVSrd,J. Bull. Soc. Chim. Biol. 4.5, 703-713 (1963). 8. B ~ a ~ , M., Fala~, J. M. P., F u ~ o , A., S ~ c ~ z , L. M. and LOPEz-OTrN,C. Biochem. J. 271, 803-807 (1990). 9. BOJ~OVSKI,D., A~ln~owc, P., MCCONA~rr~,W. J. and KELLY,J. L. FEBS Left. 112,, 251-254 (1980). 10. B o u t , M. E., DE BS~DT, J. P., A ~ . A U L T - J ~ , M., BUaDIN,J., VERTm~, N. and RSdSSO~, A. Scnnd. J. Gastroenterol. 23, 477-483 (1988). II. BoYI~S,J. K., NOTT~PEK, L. M. and A ~ s o N , L. J. J. Biol. Chem. 2,65, 17805-17815 (1990). 12. BoY, s, J. K., N c r r r ~ , L. M., WAaD~L, R. B. and R~L, S. R., JR J. Lipid Res. 31, 2243-2256 0990). 13. B~CK~mDOE, W. C., Lrrr~, J. A., A~uPowc, P., WANG,C. S., KUKSm,A., KAKIS,(3., L~TDOm~, F. and GARD~'~, (3. A therosclerosis 4.5, 161-179 (1982). 14. CANTO, R. N., MARCEL,Y. L., M1LNE,R. W., LUSS~R-CAcA~,S. and WEECH, P. K. J. Lipid Res. 30, 865-875 (1989). 15. CI~uNO, M. C., WOLF,A. C., LUM,K. D., TOLU~SON,J. H. and ALB~XS,J. J. J. Lipid Res. 27, 1135-1144 (1986). 16. CURRY,M. D., McCoNAT~, W. J. and ALAUPOVlC,P. Biochim. Biophys. Acta 491, 232-241 0977). 17. D n ~ , W. (3, HAAO~SEN,D. E., COX, C. E. and WELLS,S. A. JR Breast Cancer Res. Treatment 16, 253-260 (1990). 18. DRAYNA,D., FIELDINO,C., MCLEAn, J., B ~ , B., CASTRO, (3, CHL~, E., COMSTOCK,L., H~Z~L, W., Kox-m, W., Km~, L, WloN, K. and LAWN,R. J. Biol. Chem. 261, 16535-16539 (1986). 19. DRAMA, D., McLEAN, J., WION, K., TRENT,J., DP.AaKIN,H. and LAWN,R. DNA 6, 199-204 (1987). 20. FI~LD~, P. E. and Fmt,DING,C. J. Proc. Natl. Acad. Sci. USA. 77, 3327-3330 (1980). 21. FRANCONI~,O. L, GURAKAR,A. and FmLDI~G,C. J. Biol. Chem. 264, 7066-7072 (1989). 22. HmXJ~R,C. B., TALL,A. R., SWE~SOn,T. L., W~C'H, P. K., MARCEL,Y. L. and MIL~r~,R. W. J. Biol. Chem. 263, 5020-5023 (1988). 23. ILLINGWORTH,D. R., CONNOR,W. E. and ALAuPOv~C,P. Ann. Nutr. Metab. 25, 1-10 (1981). 24. Kest,n~R,L., Yu, W. S. and BSad3LOW,H. L. Ann. N.Y. Acad. Sci. 586, 198-203 (1990). 25. KessleR, L., Yu, W. S., BRADLOW,H. L., BREED,C. W. and F~JSHER, M. Cancer Res. 48, 6379-6383 (1988). 26. KooK, A. K., ECKHAUS,A. S. and RUBENS~I~,P. Can. J. Biochem. 48, 712 (1970). 27. KOST~, G. M. Biochim. Biophys. Acta 336, 383-395 (1974). 28. LANE,D. M. and MCCONATHY,W. J. Pediatr. Res. 17, 83 (1983). 29. LEA,O. A. Steroids 52, 337-338 (1988). 30. MAZOUJIAN,G. Am. J. Dermatopath. 12, 452-457 (1990). 31. MCCOSATHY,W. and ALAUPOWC,P. FEBS Lett. 37, 178-182 (1973). 32. MCCOSATI-IY,W. and ALAUPOVIC,P. Biochemistry 15, 515-520 (1976). 33. PSARLMA~,W. H., PE~G, L. H., MAZOUaA~,G., H~AGENSE~,D. E., JR, WELLS,S. A., JR and KmTER,S. J. J. Endocrinol. 75, 19-20 (1977). 34. l~i~cH, M. C. and BOOUSKI,M. S. New Biologist 2, 197-206 (1990). 35. I~VS~R, J., REED, R. R., b'~InST~JN,P. G. and S~r~DER,S. H. Science 241, 336-339 (1988). 36. I~tOVOST,P. R., W~CH, P. K., TR~MBLAY,N. M., MARCEL,Y. L. and RASSART,E. J. Lipid Res. 31, 2057-2065 (1990). 37. SI~a~D, J., DAUVOm,S., HAA~.~S~N,D. E., LEWS~UE,C., MER~ND,Y. and LAB~, F. Endocrinology 126, 3223-3231 (1990). 38. SMITH,K. M., LAWN,R. M. and WILCOX,J. N. J. Lipid Res. 31, 995-1004 (1990). 39. S P ~ , P., SCAXL, H,, KU~N, G., ROT~, T., UNTERnECK,A., Ot~K, R. and MULLER, H. EMBO J. 9, 2479-2484 (1990). 40. STEYR~R,E. and KOSTNER,G. M. Biochim. Biophys. Acta 958, 484-491 (1988). 41. UT~RMANN,G., MENZEL,H. J., ADLER,G., DmKE~,P. and WEBER,W. Eur. J. Biochem. 107, 225-241 (1980). 42. U'I~aMAN~,(3. and WEBER,W. FEBS Lett. 154, 357-361 (1983). 43. VEZINA,C. A., MIL~E, R. W., W~CH, P. K. and MARCEL,Y. L. J. Lipid Res. 29, 573-585 (1988). 44. Wvass, L. Immunol. Today 12, 24-29 (1991). 45. WEECH,P. K., CAMATO,R. N., MILNE,R. W. and MARCEL,Y. L. J. BioL Chem. 261, 7941-7951 (1986).