Mapping a Gene Involved in Regulating Cholesterol Absorption 1041 The Journal of Clinical Investigation Volume 102, Number 5, September 1998, 1041–1044 http://www.jci.org Mapping a Gene Involved in Regulating Dietary Cholesterol Absorption The Sitosterolemia Locus Is Found at Chromosome 2p21 Shailendra B. Patel,* Gerald Salen, ‡ Hideki Hidaka, § Peter O. Kwiterovich, Jr., i Anton F.H. Stalenhoef, ¶ Tatu A. Miettinen,** Scott M. Grundy,* Mi-Hye Lee,* Jeffrey S. Rubenstein, ‡‡ Mihael H. Polymeropoulos, ‡‡ and Michael J. Brownstein §§ *Center for Human Nutrition, University of Texas Southwestern Medical Center, Dallas, Texas 75235-9052; ‡ Department of Medicine, UMD-New Jersey Medical School, Newark, New Jersey 07103, and Gasteroenterology Research Laboratory, New Jersey Veterans Health Care System, East Orange, New Jersey 07018; § Third Department of Medicine, Shiga University of Medical Science, Seta-Tsukinowa-cho, Otsu, Shiga 520-21, Japan; i Department of Pediatrics, The Johns Hopkins Hospital, The Children’s Medical and Surgical Center, Baltimore, Maryland 21287-3654; ¶ Department of Medicine, Division of General Internal Medicine, University Hospital Nijmegen, 6500 HB Nijmegen, Nijmegen, The Netherlands; **Department of Internal Medicine, Helsinki University Central Hospital, 00290 Helsinki, Finland; ‡‡ Gene Mapping Unit, Laboratory of Genetic Disease Research, National Human Genome Research Institute, and §§ Section on Genetics: People Investigating Genes, National Institute of Mental Health/National Human Genome Research Institute, Bethesda, Maryland 20892 Abstract The molecular mechanisms regulating the amount of di- etary cholesterol retained in the body as well as the body’s ability to selectively exclude other dietary sterols are poorly understood. Studies of the rare autosomal recessively inher- ited disease sitosterolemia (OMIM 210250) may shed some light on these processes. Patients suffering from this disease appear to hyperabsorb both cholesterol and plant sterols from the intestine. Additionally, there is failure of the liver’s ability to preferentially and rapidly excrete these non-cho- lesterol sterols into bile. Consequently, people who suffer from this disease have very elevated plasma plant sterol lev- els and develop tendon and tuberous xanthomas, acceler- ated atherosclerosis, and premature coronary artery dis- ease. Identification of this gene defect may therefore throw light on regulation of net dietary cholesterol absorption and lead to an advancement in the management of this impor- tant cardiovascular risk factor. By studying 10 well-charac- terized families with this disorder, we have localized the ge- netic defect to chromosome 2p21, between microsatellite markers D2S1788 and D2S1352 (maximum lodscore 4.49, u 5 0.0). ( J. Clin. Invest. 1998. 102:1041–1044.) Key words: genetics • sitosterolemia • linkage analyses • chromosomal localization Introduction Sitosterolemia is a lipid disorder first described by Bhatta- charyya and Connor in 1974 (1). The disease is characterized by the presence of tendon and tuberous xanthomas, premature atherosclerotic disease, absence of a family history of prema- ture coronary artery disease, and normal to occasionally ele- vated plasma cholesterol levels (2). Premature fatal myocar- dial infarction was the presenting feature in the next three families reported with this condition (2). Sitosterolemia shares several clinical characteristics with the well-characterized ho- mozygous familial hypercholesterolemia (FH), 1 including ex- pression of tendon xanthomas in the first ten years of life, and development of aortic stenosis and premature atherosclerosis in the first several decades. However, homozygotes for sito- sterolemia usually have normal to moderately elevated total sterol levels, in contrast to the profound hypercholesterolemia in FH homozygotes. Sitosterolemia is further distinguished from FH by its autosomal recessive mode of inheritance and the diagnostic elevation of plasma phytosterols. This condition may be underreported, as the detection of plasma phytosterols requires the use of HPLC or capillary GLC (3). Sitosterol is clearly the major plant sterol species, hence the name sitoster- olemia, although many other plant sterols are also significantly elevated in the plasma. The term phytosterolemia may there- fore be preferable. Segregation analyses of a large Amish ped- igree showed that the trait showed an autosomal recessive pat- tern of inheritance (4). The true prevalence of this disorder is not known, but . 40 individuals with this condition have been reported worldwide (3). Under normal circumstances, our diets contain almost equal amounts of cholesterol and plant sterols. However, only 30–60% of total dietary cholesterol and , 5% of total plant sterols are retained by the body (5). The liver excretes most of the absorbed noncholesterol sterols rapidly into bile, almost unchanged, such that the net absorption of these sterols is al- most negligible (5). In sitosterolemia, affected individuals show an increase in the absorption of total dietary sterols, with failure to discriminate between different sterol species and a failure to excrete absorbed noncholesterol species rapidly into bile (6–10). This leads to greatly expanded body pool size of both cholesterol and sitosterol. Clinical studies of affected in- dividuals show that, in addition to the defects of absorption and excretion of sterols, whole body cholesterol biosynthesis is depressed (6, 11, 12). Treatments that normally result in an in- crease in cholesterol biosynthesis in normal individuals fail to do so in affected individuals (12). Additionally, analyses of liver and intestinal biopsies from affected individuals show that activities of many of the enzymes involved in cholesterol biosynthesis pathway are depressed (9, 11–13), suggesting that intracellular sterol pools are being sensed as “replete.” How- ever, the expression of the LDL receptor in liver and intestinal biopsies was found to be elevated, suggesting discordant regu- Address correspondence to Shailendra B. Patel, Y3.208, Center for Human Nutrition, UT Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75235-9052. Phone: 214-648-8734; FAX: 214- 648-7150; E-mail: spatel@crcdec.swmed.edu Received for publication 8 May 1998 and accepted in revised form 9 July 1998. 1. Abbreviations used in this paper: FH, familial hypercholester- olemia; SREBP, sterol regulatory element binding protein. 1042 Patel et al. lation (11). The structural genes for LDL receptor and some of the enzymes along the cholesterol biosynthetic pathway have been excluded as sites of the defect (14). Similarly, transcrip- tional factors that are thought to play a major role in regulat- ing many of these genes, sterol regulatory element binding proteins (SREBPs), have also been excluded (14). To identify the genetic defect in sitosterolemia, a genome-wide scan and linkage analyses of ten well-characterized pedigrees was un- dertaken. Methods Pedigrees. All probands were identified on the basis of clinical find- ings of tendon and tuberous xanthomas, the suspicion of sitoster- olemia as a diagnosis, confirmed by the diagnostic elevation of the plasma sitosterol levels determined by either HPLC or capillary gas- liquid chromatography as previously described. Pedigrees 100, 200, 500, 700, 800, 2200, and 2300 have been previously described (4, 9, 15–19). Informed consent was obtained from all participants. Al- though individuals 100.1 and 2200.75 are dead, their DNA was avail- able from fibroblast cultures established before death. Individuals 500.24, 500.27, 2200.76, 2200.79, 2200.81, 2200.82, 2200.83, 2200.84, and 2200.87 were not available for genotyping. All of these, except 2200.84, are unaffected siblings, as determined by plasma sitosterol levels. Genotyping. Genomic DNA was extracted from whole-blood fi- broblast cultures or transformed lymphoblastoid cell lines, as previ- ously described (14). Genome-wide linkage analyses were performed using primer sets from the Weber version 8 screening set (Research Genetics, Birmingham, AL). Reverse primers from each set were end-labeled with g - 32 P-ATP, using standard techniques. PCR amplifi- cation of the microsatellite repeats was performed as previously de- scribed (14), and the products separated by denaturing urea-acryl- amide gels, dried and analyzed by phosphorimager. Additional markers were analyzed using FAM-labeled primers (Research Ge- netics) and the PCR products analyzed by gel separation and fluori- metric detection using a fluorimager (Molecular Dynamics, Sunny- vale CA). Allele sizing was determined using CEPH family DNA, 1331-01 and 1331-02. Linkage analyses. Linkage analyses were performed using the computer package LINKAGE (20, 21). The PREPLINK function was recompiled as a macro to run on a Macintosh using a Microsoft Excel worksheet. This checks the pedigrees for misinheritances of the alleles and generates an appropriately formatted output file. Two- point linkage was performed using MLINK, running on a DEC sta- tion 5000/200, Ultrix v4.1 operating system. Autosomal recessive in- heritance and a disease prevalence of 1:5000 were specified. Allele frequencies were obtained from the CEPH family database. Multi- point analysis was performed using GENEHUNTER (22). Results We have assembled 10 families with the diagnosis of sitoster- olemia, some of which have been reported previously (Fig. 1). There are two Japanese families (pedigrees 700 and 800), one Amish family (pedigree 2200), one Asian Indian family (pedi- gree 500), one Finnish family (pedigree 400), and one Dutch family (pedigree 2500); the remaining families are from the United States, and are of Caucasian origins (pedigrees 100, 200, and 300). DNA from 22 affected individuals, 11 unaf- fected siblings, and 17 parents was available for genotyping. The plasma sitosterol levels in affected individuals, their obli- gate heterozygous parents, unaffected siblings (from all except pedigree 700), and unrelated normal controls are shown in Fig. 2. Sitosterol levels in pedigree 700 have been reported previ- ously (23). We have also included plasma sitosterol levels from other known sitosterolemia individuals on file in our data- bases. As can be seen, the diagnosis of affected status can be made definitively. Based solely on the plasma sitosterol levels, the heterozygous and normal control states are indistinguish- able. Total plasma sterol levels in affected individuals may only be marginally elevated relative to control populations and do not correlate with the presence of accelerated atherosclero- sis, suggesting that elevated cholesterol per se is not responsi- ble for the increased cardiovascular risk. A genome-wide scan was performed using a microsatellite screening set at an average interval of 25 cM, using families 100, 200, 500, 700, 800, and 2200. All of the autosomal chromo- somes were examined. A significant two-point lodscore (3.49, u 5 0.1) was obtained with marker D2S1788. For the remain- ing autosomes, the highest maximum lodscore obtained was 0.74, u 5 0.2, for one telomeric marker on chromosome 19 (data not shown). Mapping markers at an average density of 10-cM intervals, using all of the families, gave a maximum lod- score of 4.49, u 5 0.0, with the marker D2S1346 (Table I). A multipoint analysis is shown in Fig. 3. By comparing the shared haplotypes within sibling pairs and examining for informative recombination events, we located the disease locus between markers D2S1788 and D2S1352 (Fig. 3). This region spans al- most 15 cM. Family 800 showed no informative recombination events within the area of interest. Family 400 had one affected child, and his other two siblings did not share any haplotypes with the affected individual. Areas of haplotype sharing were made within pedigrees only and are shown by the closed areas. The unshared regions, by virtue of a recombination event on the maternal or the paternal chromosome, are shown in open Figure 1. Sitosterolemia pedigree trees. In the pedigrees assembled for this study, affected individuals are shown by filled symbols. Only those individuals for whom both se- rum sitosterol values and DNA were available were included in the analyses. A total of 17 parents, 22 affected individuals, and 11 unaf- fected siblings were genotyped. Mapping a Gene Involved in Regulating Cholesterol Absorption 1043 areas. In families 300 and 500, two recombination events were detected. These were inferred for family 300 because the mother was not available for genotyping. For family 500, dis- tinct haplotypes could be constructed for both parents. For all of the pedigrees analyzed, haplotype analyses showed that all of the affected individuals share the region between D2S1788 and D2S1352. Markers D2S1346 and D2S1348 are very close to each other ( , 0.5 cM), and their order has not been deter- mined accurately. No evidence of genetic heterogeneity has been detected in our pedigrees. As sitosterolemia appears to be very rare and the disease locus in pedigrees drawn from dif- ferent racial backgrounds maps to the same region, the likeli- hood that mutations in different genes along a common meta- bolic pathway are responsible for causing the same or similar disease is reduced. Discussion The disease locus for sitosterolemia maps to 2p21, to an inter- val no larger than 15 cM, spanned by the microsatellite mark- ers D2S1788 and D2S1352. Because the families used for these analyses are of different racial origins and no evidence of ge- netic heterogeneity was found, these results would suggest that there is only one locus that is responsible for causing this rare disease. Although this region is too large for application of po- sitional cloning, a number of cDNAs have been mapped to this region and could be considered as candidate genes. For exam- ple, the genes for 3-hydroxyanthranilate 3,4-dioxygenase, di- oxin-inducible cytochrome P450 ( CYP1B1 ), sodium/calcium exchanger 1 precursor, lutropin-choriogonadotropic hormone receptor and FSH receptor, amino acid transporter protein, phosphatidylinositol-glycan biosynthesis class F protein, a 130- KD leucine-rich protein, genes homologous to the Drosophila Figure 2. Plasma sitosterol levels in affected individuals, their obli- gate heterozygous parents, unaffected siblings, and normal controls. Plasma sitosterol levels were determined by capillary GLC or HPLC for all of our pedigree volunteers shown above as well as for 20 nor- mal individuals chosen at random. Additionally, data on file on other families not included in this study are also shown. Note the break in the y-axis. In general, most unaffected individuals had plasma sito- sterol levels , 1 mg/dL. Of the three parents and three siblings who had values higher than 1 mg/dL, all of these were measured by HPLC; these values may reflect the different technique used. None was higher than 2 mg/dL. All of the affected individuals had plasma sitosterol values . 8 mg/dL, and many of these values were obtained while the patients were in treatment. Table I. Two-point Lodscores between Microsatellite Markers and Sitosterolemia Recombination fraction, u Locus 0.00 0.02 0.10 0.20 0.30 Z max u max D2S1325 - ∞ 1.36 1.89 1.4 0.74 1.91 0.08 D2S1788 - ∞ 3.25 3.49 2.57 1.42 3.59 0.08 D2S1348 4.41 4.19 3.31 2.19 1.14 4.41 0.00 D2S1346 4.49 4.27 3.37 2.23 1.16 4.49 0.00 D2S1352 - ∞ 1.83 2.62 1.93 1.02 2.65 0.08 D2S1337 - ∞ 2 0.46 0.89 0.84 0.48 0.95 0.14 D2S441 - ∞ 2 6.27 2 1.90 2 0.54 2 0.11 0.00 0.5 D2S1328 - ∞ 2 5.36 2 1.30 2 0.23 0.03 0.03 0.3 Two-point lodscores were calculated with an autosomal recessive pat- tern of inheritance, a phenocopy rate of 0, and a disease prevalence of 1:5,000. Significant lodscores ( . 3.0) are in boldface. Figure 3. Multipoint and haplotype analyses of infor- mative markers to localize the sitosterolemia gene. Mul- tipoint analysis between markers D2S1325 and D2S441 was performed as described in Methods. The multipoint analysis is shown graphically on the right and the haplotype analysis on the left. Only those haplotypes that had an informative re- combination are shown. The allele assignments are as indi- cated. The maximal area of localization is indicated by the shaded rectangle. The upper border is marked by the marker D2S1788, which is recombinant in pedigrees 300 and 2300. The lower border is defined by recombination events in pedigrees 200 and 2200 for marker D2S1352. For family 300, the mother was not available for genotyping and her haplotypes were therefore inferred. 1044 Patel et al. melanogaster dosage compensation regulator, serum protein MSE55, TIS11B protein, and calmodulin all map to this re- gion. More than 100 unidentified transcripts also have mapped to this region. Of the known genes that map to this region, none seem likely candidates genes based on their known and/ or proposed functions, which are based on homology. Normal individuals show almost no net dietary absorption of non-cholesterol sterols. This may result from a highly selec- tive intestinal absorptive process, allowing cholesterol, but not other sterols, to enter the body, or by a very rapid and prefer- ential clearance of these noncholesterol sterols by the liver into bile, giving the appearance of selective sterol absorption. Equally feasible is that both of these mechanisms may be oper- ative. The identification of the gene in sitosterolemia is there- fore likely to be very informative of these processes. The other pathological features of sitosterolemia are in- creased foam cell formation, leading to widespread tendon and tuberous xanthoma formation with accelerated atherosclero- sis. Although the mechanisms by which atherosclerosis is ac- celerated are not fully worked out, sitosterol and cholesterol per se are unlikely to be the key sterols that lead to the cascade of differential gene expression changes and foam cell forma- tion. Sitosterol infusions in rats or man do not recapitulate the discordant gene regulation (24), and the plasma levels of cho- lesterol in many affected individuals with accelerated athero- sclerosis are frequently not elevated (2, 3). However, as many other sterol species such as shell-fish sterols (8), are also ab- sorbed, and many of these also are partially metabolized by the body to oxy-sterols (2, 3), any one of these compounds could be responsible for these changes (25). It is known that cholesterol needs to be metabolized before it can be “sensed” by the cell. Considerable advances have been made in identify- ing transcriptional factors that regulate intracellular choles- terol homeostasis, namely SREBPs, as well as the novel mechanisms by which SREBPs are activated. However, the in- tracellular sterol “sensor” has not been definitively identified. Although a number of sterol species have been shown to act as potent repressors of this pathway, both the exact sterol species that is sensed and the intracellular compartment that this oc- curs in have not been elucidated. While we believe that the gene product defective in sitosterolemia is primarily involved in selective transport of cholesterol across the plasma mem- brane, the pattern of discordant gene regulation observed sug- gests that these two pathways are linked. The elucidation of the gene defective in sitosterolemia may therefore shed some light on these processes. Acknowledgments We thank John Poindexter for help with linkage analyses. We also thank all of our pedigree members for their continued cooperation. This work was supported by the Medical Research Service, De- partment of Veterans Affairs, Washington, DC, US Public Health Service grant HL-17818 (G. Salen); by a grant-in-aid (10671064) from the Ministry of Education, Science, and Culture, Japan (H. Hidaka); by an American Heart Association Scientist Development Award 9730087N (S.B. Patel); by the Southwestern Medical Foundation, Dallas, TX, and the Moss Heart Foundation, Dallas, TX. References 1. Bhattacharyya, A.K., and W.E. Connor. 1974. Beta-sitosterolemia and xanthomatosis. A newly described lipid storage disease in two sisters. J. Clin. Invest. 53:1033–1043. 2. Salen, G., S. Shefer, L. Nguyen, G.C. Ness, G.S. Tint, and V. Shore. 1992. Sitosterolemia. [Review]. J. Lipid Res. 33:945–955. 3. Bjorkhem, I., and K.M. Boberg. 1995. Inborn errors in bile acid biosyn- thesis and storage of sterols other than cholesterol. In The Metabolic Basis of Inherited Disease. Vol. 2. C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle, editors. McGraw-Hill, Inc., New York. 2073–2102. 4. Beaty, T.H., P. Kwiterovich, Jr., M.J. Khoury, S. White, P.S. Bachorik, H.H. Smith, B. Teng, and A. Sniderman. 1986. Genetic analysis of plasma sito- sterol, apoprotein B, and lipoproteins in a large Amish pedigree with sitoster- olemia. Am. J. Hum. Genet. 38:492–504. 5. Salen, G., E. Ahrens, Jr., and S.M. Grundy. 1970. Metabolism of beta- sitosterol in man. J. Clin. Invest. 49:952–967. 6. Miettinen, T.A. 1980. Phytosterolaemia, xanthomatosis and premature atherosclerotic arterial disease: a case with high plant sterol absorption, im- paired sterol elimination and low cholesterol synthesis. Eur. J. Clin. Invest. 10: 27–35. 7. Bhattacharyya, A.K., W.E. Connor, D.S. Lin, M.M. McMurry, and R.S. Shulman. 1991. Sluggish sitosterol turnover and hepatic failure to excrete sito- sterol into bile cause expansion of body pool of sitosterol in patients with sito- sterolemia and xanthomatosis. Arterioscler. Thromb. 11:1287–1294. 8. Gregg, R.E., W.E. Connor, D.S. Lin, and H. Brewer, Jr. 1986. Abnormal metabolism of shellfish sterols in a patient with sitosterolemia and xanthomato- sis. J. Clin. Invest. 77:1864–1872. 9. Salen, G., V. Shore, G.S. Tint, T. Forte, S. Shefer, I. Horak, E. Horak, B. Dayal, L. Nguyen, A.K. Batta, et al. 1989. Increased sitosterol absorption, de- creased removal, and expanded body pools compensate for reduced cholesterol synthesis in sitosterolemia with xanthomatosis. J. Lipid Res. 30:1319–1330. 10. Salen, G., G.S. Tint, S. Shefer, V. Shore, and L. Nguyen. 1992. Increased sitosterol absorption is offset by rapid elimination to prevent accumulation in heterozygotes with sitosterolemia. Arterioscler. Thromb. 12:563–568. 11. Nguyen, L.B., S. Shefer, G. Salen, G.C. Ness, G.S. Tint, F.G. Zaki, and I. Rani. 1990. A molecular defect in hepatic cholesterol biosynthesis in sitoster- olemia with xanthomatosis. J. Clin. Invest. 86:923–931. 12. Nguyen, L.B., G. Salen, S. Shefer, G.S. Tint, V. Shore, and G.C. Ness. 1990. Decreased cholesterol biosynthesis in sitosterolemia with xanthomatosis: diminished mononuclear leukocyte 3-hydroxy-3-methylglutaryl coenzyme A reductase activity and enzyme protein associated with increased low-density li- poprotein receptor function. Metab. Clin. Exp. 39:436–443. 13. Nguyen, L.B., G. Salen, S. Shefer, J. Bullock, T. Chen, G.S. Tint, I.R. Chowdhary, and S. Lerner. 1994. Deficient ileal 3-hydroxy-3-methylglutaryl co- enzyme A reductase activity in sitosterolemia: sitosterol is not a feedback inhib- itor of intestinal cholesterol biosynthesis. Metab. Clin. Exp. 43:855–859. 14. Patel, S.B., A. Honda, and G. Salen. 1998. Sitosterolemia: exclusion of genes involved in reduced cholesterol biosynthesis. J. Lipid Res. 39:1055–1061. 15. Salen, G., P. Kwiterovich, Jr., S. Shefer, G.S. Tint, I. Horak, V. Shore, B. Dayal, and E. Horak. 1985. Increased plasma cholestanol and 5 alpha-saturated plant sterol derivatives in subjects with sitosterolemia and xanthomatosis. J. Lipid Res. 26:203–209. 16. Berger, G.M., W.M. Deppe, A.D. Marais, and M. Biggs. 1994. Phy- tosterolaemia in three unrelated South African families. Postgrad. Med. J. 70: 631–637. 17. Berger, G.M., R.J. Pegoraro, S.B. Patel, P. Naidu, L. Rom, H. Hidaka, A.D. Marais, A. Jadhav, R.P. Naoumova, and G.R. Thompson. 1998. 3-hydroxy 3-methylglutaryl coenzyme A reductase is not the site of the primary defect in phytosterolemia. J. Lipid Res. 39:1046–1054. 18. Hidaka, H., T. Nakamura, T. Aoki, H. Kojima, Y. Nakajima, K. Kosugi, I. Hatanaka, M. Harada, M. Kobayashi, A. Tamura, et al. 1990. Increased plasma plant sterol levels in heterozygotes with sitosterolemia and xanthomato- sis. J. Lipid Res. 31:881–888. 19. Hidaka, H., H. Sugiura, T. Nakamura, H. Kojima, M. Fujita, N. Sugie, H. Okabe, Y. Nishio, H. Maegawa, A. Kashiwagi, et al. 1997. b -Sitosterolemia with generalized eruptive xanthomatosis. Endocr. J. 44:59–64. 20. Lathrop, G.M., J.M. Lalouel, C. Julier, and J. Ott. 1984. Strategies for multilocus linkage analysis in humans. Proc. Natl. Acad. Sci. USA. 81:3443– 3446. 21. Terwillinger, J.D., and J. Ott. 1994. Handbook of Human Genetic Link- age. Johns Hopkins University Press, Baltimore. 22. Kruglyak, L., M.J. Daly, M.P. Reeve-Daly, and E.S. Lander. 1996. Para- metric and nonparametric linkage analysis: a unified multipoint approach. Am. J. Hum. Genet. 58:1347–1363. 23. Kwiterovich, P., Jr., P.S. Bachorik, H.H. Smith, V.A. McKusick, W.E. Connor, B. Teng, and A.D. Sniderman. 1981. Hyperapobetalipoproteinaemia in two families with xanthomas and phytosterolaemia. Lancet. 1:466–469. 24. Shefer, S., G. Salen, J. Bullock, L.B. Nguyen, G.C. Ness, Z. Vhao, P.F. Belamarich, I. Chowdhary, S. Lerner, A.K. Batta, et al. 1994. The effect of in- creased hepatic sitosterol on the regulation of 3-hydroxy-3-methylglutaryl- coenzyme A reductase and cholesterol 7 alpha-hydroxylase in the rat and sito- sterolemic homozygotes. Hepatology. 20:213–219. 25. Smith, L.L., and B.H. Johnson. 1989. Biological activities of oxysterols. Free Radical Biol. Med. 7:285–332.