This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rigotti, A. Articles by Krieger, M. Search for Related Content PubMed PubMed Citation Articles by Rigotti, A. Articles by Krieger, M.

The Role of the High-Density Lipoprotein Receptor SR-BI in the Lipid Metabolism of Endocrine and Other Tissues

Departamento de Gastroenterología (A.R.), Facultad de Medicina, Pontificia Universidad Católica, Santiago, Chile; Department of Medicine (H.E.M.), University of Helsinki, Helsinki 00290, Finland; and Biology Department (M.K.), Massachusetts Institute of Technology, Cambridge, Massachusetts 02139

	Abstract

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
Because cholesterol is a precursor for the synthesis of steroid hormones, steroidogenic tissues have evolved multiple pathways to ensure adequate supplies of cholesterol. These include synthesis, storage as cholesteryl esters, and import from lipoproteins. In addition to endocytosis via members of the low-density lipoprotein receptor superfamily, steroidogenic cells acquire cholesterol from lipoproteins by selective lipid uptake. This pathway, which does not involve lysosomal degradation of the lipoprotein, is mediated by the scavenger receptor class B type I (SR-BI). SR-BI is highly expressed in steroidogenic cells, where its expression is regulated by various trophic hormones, as well as in the liver. Studies of genetically manipulated strains of mice have established that SR-BI plays a key role in regulating lipoprotein metabolism and cholesterol transport to steroidogenic tissues and to the liver for biliary secretion. In addition, analysis of SR-BI-deficient mice has shown that SR-BI expression is important for -tocopherol and nitric oxide metabolism, as well as normal red blood cell maturation and female fertility. These mouse models have also revealed that SR-BI can protect against atherosclerosis. If SR-BI plays similar physiological and pathophysiological roles in humans, it may be an attractive target for therapeutic intervention in cardiovascular and reproductive diseases.

I. Introduction

II. Lipoprotein Metabolism

III. Lipoprotein Receptors

IV. Structure and Subcellular Distribution of SR-BI

V. Ligand Binding Activities of SR-BI

VI. Lipid Transfer Activity of SR-BI

A. Selective cholesterol uptake

B. Cholesterol efflux

VII. Tissue Expression, Regulation, and Function of SR-BI

A. Liver

B. Adrenal gland

C. Ovary

D. Testis

E. Gut

F. Lung

G. Nervous system

H. Mammary gland

I. Endothelium

J. Macrophages

K. Skin

L. Embryonic and extraembryonic tissues

VIII. Mechanisms Underlying the Regulation of SR-BI Expression

IX. SR-BI and Lipoprotein Metabolism

A. HDL metabolism

B. Metabolism of other lipoproteins

X. Role of SR-BI in Physiology and Pathophysiology

A. Reproductive biology: female fertility

B. Atherosclerosis

C. Cholesterol gallstone disease

D. Intestinal cholesterol absorption

E. Embryogenesis and fetal development

F. Red blood cell (RBC) maturation

G. NO metabolism

H. Vitamin E transport

XI. SR-BI in Humans

XII. Concluding Remarks

	I. Introduction

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
CHOLESTEROL METABOLISM IS intimately associated with the proper functioning of the endocrine system. In addition to the housekeeping functions of cholesterol in all mammalian cells as a key contributor to membrane structure and function, it is the precursor for cytochrome P450 cholesterol side chain cleavage enzyme-initiated synthesis of steroid hormones. (Cholesterol can also be converted to bile acids in the liver and vitamin D in the skin and kidney.) Thus, it is not surprising that endocrine, especially steroidogenic, tissues and other cell types have evolved multiple pathways to ensure the adequate provision of this crucial lipid.

Cells can obtain cholesterol by endogenous synthesis from mevalonate and isoprenoid precursors (1). An important rate-controlling step in biosynthesis is catalyzed by the enzyme hydroxymethylglutaryl coenzyme A (HMG CoA) reductase. Hydrolysis of cholesteryl esters stored in cytoplasmic lipid droplets can also provide cholesterol. Although most body cells have relatively small cholesteryl ester pools, cytoplasmic cholesteryl ester stores are especially abundant in steroidogenic cells, providing a ready supply of cholesterol for rapidly inducible steroidogenesis. Additionally, cells use surface receptors to extract exogenous cholesterol from the lipoprotein particles that transport this water-insoluble molecule through the aqueous circulatory system (2). In fact, circulating lipoproteins are a major physiological source of cholesterol for steroidogenic cells. Plasma lipoproteins play a critical role not only in normal cholesterol homeostasis, but also in pathophysiological states such as atherosclerotic cardiovascular disease [increased risk associated directly with the plasma concentrations of low-density lipoprotein (LDL) and inversely with those of high-density lipoprotein (HDL)] (3 ,4). Here, we will review the role of the scavenger receptor class B type I (SR-BI), the first HDL receptor to be characterized, in the cellular metabolism of lipoproteins and its relevance for body lipid homeostasis and disease.

	II. Lipoprotein Metabolism

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
Plasma lipids, i.e., mainly cholesterol, triglycerides, and phospholipids along with low levels of other hydrophobic compounds such as vitamin E, are carried in the blood by water-soluble lipoproteins. Lipoproteins have a polar outer shell of protein and phospholipid and inner core of neutral lipid. The largest lipoprotein particles are chylomicrons, which are synthesized in the intestine. Dietary fat and lipid secreted via bile into the gastrointestinal lumen are absorbed and packaged into chylomicrons containing triglycerides as their major core lipid component and apolipoproteins (apo)B₄₈ and apoE as key major protein constituents. Chylomicrons are secreted into lymph, and, after their entrance into the bloodstream, their core triglycerides are hydrolyzed by lipoprotein lipase (LPL), an enzyme bound to the luminal surfaces of capillary endothelial cells located in fat and muscle tissue. Hydrolysis of triglycerides reduces the volume of the lipoprotein core. Hydrolysis also results in the release of excess surface shell components, i.e., phospholipids, unesterified cholesterol, and exchangeable apolipoproteins, and their transfer to other circulating lipoproteins, especially to circulating HDL. The residual chylomicron particles remaining in the circulation after lipolysis are called chylomicron remnants, which are rapidly taken up by the liver by receptor-mediated endocytosis (reviewed in Ref.5).

Very LDL (VLDL) are triglyceride-rich lipoproteins whose major apolipoprotein components include apoB₁₀₀ (in rodents, also apoB₄₈) and apoE. VLDL is synthesized by hepatocytes and secreted into the blood. As is the case for chylomicrons, VLDL triglycerides are hydrolyzed by LPL, giving rise to smaller intermediate-density lipoprotein (IDL) particles and excess surface components that can be transferred to other lipoproteins. IDL is either cleared from the circulation by hepatic endocytosis or, through the action of hepatic lipase and cholesteryl ester transfer protein (CETP), further converted to LDL (reviewed in Ref.5). LDLs are the major carriers of cholesterol (as cholesteryl esters) in humans, but normally there is very little LDL in rodents that lack CETP. LDL is removed from the circulation by LDL receptor (LDLR)-mediated endocytosis, in which LDL is internalized and delivered to lysosomes for degradation (see below). Most LDLR activity, and thus clearance of plasma LDL, occurs in the liver.

HDLs are not a single macromolecular entity, but rather a group of lipoprotein particles containing nearly equal amounts of lipid and protein that has been subclassified according to physical properties and apolipoprotein composition (5, 6, 7, 8). The major apolipoproteins of HDL are apoA-I and apoA-II (with the minor components apoA-IV, apoC, apoD, and apoE; Refs.6, 7, 8). Both of these apolipoproteins play structural roles and serve as regulators of HDL metabolism (6, 7, 8). In some cases, the apolipoproteins are involved in the recognition of HDL by cell surface receptors. There are three major classes of HDL particles: 1) discoidal particles comprising phospholipids and apoA-I, which are present only transiently in the plasma; 2) small, very lipid-poor apoA-I/phospholipid particles, called pre-ß HDL because of their electrophoretic mobilities, that are present in low concentrations in the plasma; and 3) the most abundant, -HDLs, which are large spherical particles with neutral lipid cores and two or more apolipoproteins on their outer shells. -HDLs are comprised of two major subpopulations differing in their relative contents of apoA-I and apoA-II; these are lipoprotein A (LpA)-I HDL, which contains apoA-I only, and LpA-I:A-II HDL, which contains both apoA-I and apoA-II.

HDL metabolism is complex, involving several HDL-remodeling enzymes, lipid transfer proteins, and cell surface receptors (reviewed in Refs.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18). HDL originates as discoidal particles secreted by the liver and intestine or as byproducts of triglyceride-rich lipoprotein metabolism (release of surface components during lipolysis). These nascent HDL particles contain mostly apoA-I and phospholipids. They apparently mature into spherical HDL particles primarily by incorporation of unesterified cholesterol and phospholipids from cells followed by esterification of cholesterol. The transfer of cellular lipids to nascent HDL requires activity of the ATP-binding cassette (ABC) A1 transporter, and its absence results in the HDL deficiency disease called Tangier disease. The unesterified cholesterol incorporated into plasma HDL is converted to cholesteryl esters by the plasma enzyme lecithin:cholesterol acyltransferase (LCAT). Esterification creates an unesterified cholesterol concentration gradient between HDL and cell membranes and is considered critical for efficient cholesterol efflux from cells to HDL. In species that express plasma CETP, a significant fraction of HDL cholesteryl ester is transferred to other plasma lipoproteins (e.g., VLDL, IDL, and LDL) for further transport, primarily to the liver (9, 10). Thus, a substantial fraction of the plasma HDL cholesterol is indirectly delivered to the liver via hepatic endocytic receptors for IDL and LDL. In rodents and rabbits, and probably in humans, there are additional pathways for HDL lipid transport in which HDL directly delivers its cholesteryl esters to cells.

Although cellular endocytic uptake of HDL has been described (reviewed in Ref.19), both in vitro and in vivo studies have established that a key mechanism for the direct delivery of HDL cholesterol to cells is fundamentally different from the classic receptor-mediated endocytic pathway initially described for the LDLR (Refs.20, 21, 22 ; also reviewed in Ref.23). Instead, receptor-mediated HDL binding to the cell surface results in the transfer of its cholesteryl esters to the cell and the subsequent release of the lipid-depleted HDL back into the extracellular fluid rather than the lysosomal degradation of the particle. This novel cellular mechanism, which can extract cholesteryl esters from LDL as well as HDL, is called selective lipid uptake. This pathway appears to involve temperature-dependent reversible incorporation of HDL cholesteryl ester into the plasma membrane followed by a process in which cholesteryl ester molecules are irreversibly internalized and subsequently hydrolyzed through a nonlysosomal pathway (24, 25, 26, 27, 28, 29, 30, 31, 32). The lipid-depleted HDL particles released from cells after selective lipid uptake have the potential for reacquiring cellular cholesterol from peripheral tissues as described above, or they can be catabolized by the kidney (33) as a consequence of glomerulofiltration and subsequent reabsorption and degradation in the proximal tubules by a cubilin-associated endocytic pathway (34, 35). In rats, the liver removes 60–70% of HDL cholesteryl esters from plasma via the selective uptake pathway (Refs.20 and33 ; also reviewed in Ref.23), a process that appears to be efficiently coupled to biliary lipid secretion (reviewed in Ref.36). At least in rodents, selective uptake of HDL cholesterol also plays an important role in the transport of cholesterol to steroidogenic tissues (adrenal gland, ovary, and testes; Refs.20, 21, 22 and37, 38, 39), providing cholesterol as substrate for steroid hormone synthesis and for storage in cytoplasmic cholesteryl ester droplets. For example, in the rodent adrenal gland, selective uptake accounts for 90% or more of the cholesterol destined for steroid hormone production (40).

The pathway by which cholesterol is removed from peripheral cells and transported via plasma lipoproteins, such as HDL, into the liver for further metabolism (e.g., VLDL formation, bile acid synthesis, and biliary secretion) is essential for cholesterol homeostasis. This is because peripheral cells (apart from low levels of catabolism in steroidogenic and skin cells) are not able to degrade cholesterol. This pathway, known as reverse cholesterol transport (41), is likely to be responsible for some of the antiatherogenic activity of HDL.

	III. Lipoprotein Receptors

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
To secure an adequate supply of cholesterol to cells, multiple mechanisms for cholesterol uptake from plasma lipoproteins have evolved. Through the work of many investigators, but especially the pioneering studies of Brown and Goldstein (42) and their colleagues, we have a detailed picture of the cellular metabolism of LDL via the LDLR endocytic pathway and the role of LDLRs in body cholesterol transport in health and disease. Other members of the LDLR family of proteins (e.g., LDLR-related protein) participate in the hepatic metabolism of chylomicron and VLDL remnants, and they also appear to have functions independent of lipid metabolism (43, 44, 45). All of these receptors process lipoproteins via clathrin-coated pit-mediated endocytosis and lysosomal degradation, which releases cholesterol for subsequent cellular metabolism. The autosomal recessive hypercholesterolemia protein, a cytosolic polypeptide with a phosphotyrosine-binding domain, was recently identified by genetic analysis of patients with defects in LDL metabolism as a regulator of LDLR-mediated endocytosis (46). It presumably influences LDLR activity by binding to the NPYX endocytic motif of the cytoplasmic region of the receptor (46).

A series of metabolic regulatory responses linked to the LDLR pathway helps ensure cellular cholesterol homeostasis. For example, increase of the intracellular cholesterol pool due to the endocytosis of LDL leads to decreased cholesterol synthesis, increased cholesterol esterification for storage, and reduced LDLR levels to prevent excessive intracellular cholesterol accumulation. Substantial progress has been made in understanding the mechanisms by which these processes are regulated via membrane-bound transcription factors known as sterol regulatory element (SRE)-binding proteins (SREBPs; Ref.47). Furthermore, knowledge of these regulatory mechanisms contributed to the development of statin drugs, which inhibit HMG CoA reductase and consequently lower plasma LDL cholesterol levels by stimulating, via the SREBP pathway, LDLR expression. The statins represent a major breakthrough in the treatment and prevention of atherosclerotic cardiovascular disease.

About 70% of the LDLR activity is concentrated in the liver, where LDLRs remove LDL and, working together with LDLR-related protein, internalize LDL, VLDL, and chylomicron remnants from the circulation. Based on both in vitro and in vivo studies, all steroidogenic tissues from human and nonhuman mammals appear to express the LDLR pathway of lipoprotein cholesterol uptake for steroid hormone synthesis (48). If the LDLR was the sole source of sufficient amounts of cholesterol for steroidogenesis, then subjects with abnormalities in LDLR-mediated endocytosis would be expected to exhibit significantly abnormal steroid hormone production. This does not appear to be the case. Abetalipoproteinemia, which is characterized by the absence of circulating LDL, is not associated with abnormal basal plasma and urinary corticosteroids, but the response of these patients to ACTH stimulation is somewhat impaired (49, 50). Post-ACTH aldosterone and androstenedione secretion is not affected in these patients. Ovarian, testicular, and placental function of abetalipoproteinemic patients has not been exhaustively studied; however, abetalipoproteinemic females, who exhibit decreased plasma progesterone during menstrual luteal phase, have normal pregnancies (51). In addition, various reports of familial hypercholesterolemia (FH), due to LDLR deficiency, have not described major adrenal, ovarian, or testicular steroidogenic dysfunction. Plasma cortisol levels in the basal state are within normal range in heterozygous FH patients; however, plasma cortisol is reduced in homozygous FH subjects relative to controls (52, 53, 54) after ACTH administration (however, see Ref.55). In addition, LDLR-deficient Watanabe rabbits have impaired functional adrenal (56) and ovarian (57) reserves. Although fecundity of the Watanabe rabbit is somewhat impaired (58), no apparent other major endocrine abnormalities have been reported in these rabbits, as is the case for homozygous null LDLR knockout (KO) mice. Furthermore, female homozygous FH patients can get pregnant and give birth to normal offspring. Interestingly, adverse effects on adrenal or gonadal function have not been reported in patients with FH receiving cholesterol synthesis inhibitors (53, 59, 60). Taken together, these data suggest that alternative pathways for lipoprotein cholesterol delivery to cells might be available to supply functionally adequate amounts of lipoprotein-derived cholesterol to sustain hormone synthesis in various steroidogenic tissues.

The LDLR-independent accumulation of cholesterol in macrophages in artery walls is one of the earliest steps in atherosclerosis. It is generally thought that macrophage surface receptors that are not under cholesterol-feedback regulation and that recognize and endocytose modified lipoproteins (e.g., aggregated or oxidized LDL) are essential for massive cholesteryl ester lipid droplet formation in these cells. The multiligand receptors that bind and internalize modified lipoproteins are called scavenger receptors (43, 61, 62, 63). There are now many members of a superfamily of scavenger receptors that have diverse structures, expression patterns, and functions (reviewed in Refs.12 and61, 62, 63, 64, 65, 66). In vivo experiments have implicated scavenger receptor class A types I and II (SR-AI/II) and the class B scavenger receptor CD36 in atherogenesis (12, 66, 67). There is strong support for the proposal that some scavenger receptors also participate in host defense as multiligand pattern recognition receptors for pathogens (65, 66).

Using either direct binding or ligand blotting assays and a variety of preparations from tissues of different species, many laboratories have reported HDL binding activities (6, 7, 19, 68, 69). Serendipitously, SR-BI, originally identified by expression cloning because of its ability to bind modified LDL (70), was the first molecularly well characterized HDL receptor to be discovered (71). Its properties and functions are described below.

	IV. Structure and Subcellular Distribution of SR-BI

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
The predicted sequences of SR-BI proteins from different mammalian species share 70–80% sequence identity over their 509-amino-acid lengths (Fig. 1). An alternatively spliced mRNA of SR-BI and its corresponding protein have been identified and designated SR-BII (72). SR-BI is a member of a family of structurally related proteins, called the CD36 superfamily (73), which includes the mammalian proteins CD36 (74) and LIMPII (a lysosomal integral membrane protein; Ref.75); two Drosophila melanogaster proteins, emp (76) and croquemort (a hemocyte/macrophage receptor; Refs.77 and78); SnmP-1 (a silk moth olfactory neuron membrane protein; Ref.79); a putative Caenorhabditis elegans protein (GenBank accession no. Z54270), and others. There are several similarities in the intron/exon organization of the SR-BI (80) and CD36 (81) genes. Human SR-BI was first cloned as a homolog of CD36 and LIMPII of unknown function that mapped to chromosome 12 (73, 80) and was called CLA-1 (82). In addition, phylogenetic analysis of the CD36 gene family indicated that LIMPII, SR-BI, emp, and CD36 diverged from a common ancestor gene very early during evolution (73).

View larger version (71K): [in this window] [in a new window]   Figure 1. Model of SR-BI. SR-BI is a 509-amino-acid glycoprotein with a large extracellular domain anchored to the plasma membrane at both the N and C termini by hydrophobic transmembrane regions that have short extensions into the cytoplasm. SR-BI is highly glycosylated on all its potential N-glycosylation sites (84 85 ). SR-BI is also palmitoylated on the cysteines located at the C-terminal cytoplasmic and transmembrane domains (84 87 ). SR-BI contains a well conserved set of cysteines in its carboxy-terminal half. [Reproduced with permission from M. Krieger: J Clin Invest 108:793–797, 2001 (372 ).]

As predicted from the presence of consensus motifs for N-linked glycosylation sites in the cDNA, SR-BI is heavily N-glycosylated (84). This accounts for the difference between its mass predicted from the amino acid sequence (57 kDa; Ref.81) and that observed using immunochemical methods (82 kDa; Refs.71 and84). In cultured cells, SR-BI is cotranslationally modified on all of its potential N-glycosylation sites (84, 85). Some, but not all, of these N-glycosylated oligosaccharides are transformed into complex carbohydrate chains during transit through the Golgi apparatus (84). Furthermore, metabolic labeling studies have established that SR-BI is fatty acylated (84), a structural feature shared with CD36 (86). Analysis of point mutants has shown that the major sites of palmitoylation of SR-BI are cysteine residues 462 and 470 located in its carboxy-terminal transmembrane and cytoplasmic domains (87).

In cultured cells, SR-BI and SR-BII appear to be concentrated in microdomains that under certain growth conditions correspond to cholesterol and sphingolipid-enriched plasma membrane microdomains (lipid rafts) called caveolae (see below) (84, 88). For some proteins, fatty acylation can serve as a sorting signal for concentration in caveolae. However, inhibition of fatty acylation of SR-BI by simultaneous mutation of cysteines 462 and 470 has no apparent effect on SR-BI surface localization in transfected cultured cells (87). Currently, the motifs in SR-BI that determine its subcellular localization are unknown.

SR-BI is expressed in a variety of polarized epithelial cells, including hepatocytes, colangiocytes, and enterocytes. Adenovirus-mediated in vivo overexpression of SR-BI in murine hepatocytes results in SR-BI expression on both the apical (canalicular) and basolateral (sinusoidal) surfaces of the cells (89). Examination of normal murine liver sections for expression of endogenous SR-BI established that SR-BI is primarily located in the basolateral domain of hepatocytes (90, 91). It is also seen in the basolateral membranes of SR-BI-transfected MDCK cells (92, 93). In contrast, SR-BI is predominantly found in the apical domain of enterocytes (94, 95), the epithelial intestinal cell line CaCo2 (94), and colangiocytes (96). The marked differences in the subcellular distribution of SR-BI in polarized cells are likely to be related to important differences in sorting pathways, and possibly in the expression and function of adapter/chaperone proteins. Candidate modulators of the localization, stability, and function of SR-BI have been isolated from rat liver membrane homogenates (97). These proteins interact with the carboxy-terminal cytoplasmic domain of SR-BI. Molecular characterization of one of these proteins established that it is a soluble protein composed of multiple PDZ domains (97). This protein, called CLAMP (carboxy-terminal linking and modulating protein), recognizes SR-BI via its N-terminal PDZ domain. Coexpression of CLAMP with SR-BI in transfected cells affects the stability of SR-BI as well as the efficiency of conversion of HDL cholesteryl esters taken up via SR-BI to intracellular unesterified cholesterol (97). Furthermore, CLAMP has been proposed to be involved in determining the basolateral distribution of SR-BI in hepatocytes (97). In fact, a recent study using carboxy-terminal deletions of the SR-BI protein indicates that the CLAMP-interacting domain of SR-BI is essential for its cell surface expression in liver (98). Further studies will be required to identify the polarity and subcellular localization targeting signals in SR-BI, the role of SR-BI associating proteins such as CLAMP, and the functional consequences of the differential plasma membrane localizations of SR-BI.

	V. Ligand Binding Activities of SR-BI

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
SR-BI is a multiligand receptor (11, 12, 14). It was initially discovered as a scavenger receptor, because of its modified (acetylated and oxidized) LDL binding activity (70, 99). SR-BI also binds hypochlorite-modified HDL (100). However, SR-BI does not bind the wide array of polyanionic molecules (e.g., fucoidin, polyguanosine, carrageenan) that are all ligands of other classes of scavenger receptors (11, 12, 14, 61, 62, 64, 65, 66, 67, 70). In addition to lipoprotein ligands (see below), cell culture ligand binding studies show that SR-BI recognizes a diverse set of ligands. SR-BI binds maleylated BSA (70), advanced glycation end product modified proteins (101), and anionic, but not zwitterionic, phospholipids in the form of liposomes (102, 103). Indeed, SR-BI and CD36 were the first molecularly well defined cell surface receptors for phosphatidylserine to be identified (102). These findings suggested that SR-BI might be able to bind senescent, damaged, and apoptotic cells with high levels of phosphatidylserine on the exofacial surface of the plasma membrane, and thus might play a role in the recognition and clearance of these cells during development, tissue remodeling, and disease (102). In fact, transfected cultured cells expressing exogenous SR-BI can bind apoptotic cells (103, 104, 105, 106, 107). Furthermore, SR-BI mediates binding of ß-amyloid to transfected Chinese hamster ovary cells (108) and cultured microglia and brain vascular smooth muscle cells (109, 110), suggesting a potential role in Alzheimer’s disease. More recently, the hepatitis C virus envelope glycoprotein E2 was reported to be a ligand for human SR-BI, suggesting its role as hepatitis C virus receptor (111). Analysis of the mechanisms by which multiple ligands with very different structures bind SR-BI indicates very complex interactions, suggesting multiple classes of binding sites (11, 12, 112).

Most of the studies of ligand binding to SR-BI have focused on native lipoproteins. The first native lipoprotein shown to bind to SR-BI with high affinity (K_d 10 nM) was LDL (70). Before that finding, the only molecularly defined receptor for LDL was the classic LDLR (1, 2). SR-BI-mediated internalization and degradation of LDL was, however, far less efficient than that by the classic LDLR (70). Similarly, inefficient SR-BI-mediated endocytosis and degradation was observed for modified LDL (Ref.70 ; however, see Ref.99). SR-BI has also been shown to bind to VLDL (113).

Although the potential of SR-BI to function as a new "second" LDLR was intriguing, most investigations of SR-BI have been influenced by the unanticipated discovery that SR-BI was the first molecularly well defined and physiologically relevant HDL receptor to be identified (71). SR-BI binds HDL with high affinity in a calcium- and apoE-independent fashion. The ability of SR-BI to bind to several native lipoproteins suggests that it may be responsible for some of the multilipoprotein binding activities previously reported in a wide variety of cells types and tissues (6, 7, 19, 68, 69, 114). There may, however, be additional cell surface HDL receptors, and they may have diverse structures, tissue expressions, and functional activities.

As is the case for some of the other classes of scavenger receptors (43, 115), SR-BI exhibits nonreciprocal cross competition, wherein one ligand effectively blocks the binding of a second ligand, but the reverse is not the case (112). For example, HDL effectively blocks all LDL binding, whereas LDL inhibition of HDL binding is poor, relative to the K_d for LDL binding to SR-BI (71). Because of nonreciprocal cross-competition, LDL is not expected to substantially interfere with the binding of HDL to SR-BI in vivo. Using a method called "retroviral library-based activity dissection," Gu et al. (112) have identified double mutations in murine SR-BI (⁴⁰²GluArg and ⁴¹⁸GluArg) that together block most of the binding of HDL to SR-BI without significantly altering LDL binding. The same method also showed that a single methionine to arginine mutation at position 158 disrupts binding to SR-BI of both HDL and LDL, but not acetylated LDL (116).

SR-BI binds spherical -HDL. The lower the density of the -HDL particles (larger, higher lipid content), the more tightly they bind to SR-BI (117, 118). Pre-ß-HDLs are poor substrates of SR-BI (117). Several reports (117, 119) suggest that lipid-free apoA-I is also a poor substrate for SR-BI (however, see Ref.120), whereas reconstituted discoidal complexes of phospholipid/cholesterol/apolipoprotein (i.e., apoA-I, apoA-II, apoE, or apoC-III) bind more tightly to SR-BI than native spherical -HDLs (117, 119, 121). SR-BI appears to be able to bind some lipid-free apoE, which has been reported to facilitate lipoprotein cholesterol uptake (122). The receptor binding domain is located in the amino-terminal 1–165 region of the apoE protein (121). The relative amounts of apoA-I and apoA-II in spherical HDL particles apparently influence their interactions with SR-BI, although the precise nature of the interactions is not yet clear. One report suggests that enriching plasma HDL with additional copies of apoA-II enhances its affinity for SR-BI but inhibits cholesteryl ester uptake (123). Another study reported that reconstituted apoA-I/apoA-II-containing particles exhibit decreased binding relative to particles containing apoA-I only (124). The binding of reconstituted discoidal apoA-I complexes by SR-BI requires either the amino- or carboxy-terminal amphipathic helices of apoA-I (117). These (117, 118, 119) and other (125) studies strongly suggest that the conformation/organization of apoA-I in HDL particles critically affects the nature of the interaction of the particle with SR-BI. Using phospholipid particles reconstituted with fragments of apoA-I and a model peptide, Williams et al. (120) have provided strong evidence that amphipathic helices play critical roles in the binding of apolipoproteins to SR-BI. This may explain the ability of SR-BI to bind to a wide variety of lipoproteins.

	VI. Lipid Transfer Activity of SR-BI

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
A. Selective cholesterol uptake
SR-BI mediates the transfer of a variety of lipids between cells and lipoproteins. Unlike the interaction of LDL with the LDLR, the binding of HDL to SR-BI does not lead to its lysosomal degradation (71). Instead, SR-BI mediates selective lipid uptake from native lipoproteins (71). As noted above, selective lipid uptake involves the net transfer of lipids from lipoproteins to cells and the subsequent release of the lipid-depleted lipoprotein into the extracellular medium. SR-BI mediates selective uptake from HDL of cholesteryl esters (71), unesterified cholesterol (126, 127, 128, 129), phospholipids (130), triglycerides (131), -tocopherol (132, 133), DiI (71), BODIPY-cholesteryl ester (134), pyrene-labeled phospholipids (130), and cholesteryl ethers (often used as nonhydrolyzable analogs of cholesteryl esters; Ref.71). The rate constant analyses of lipid transfer from reconstituted HDL to the plasma membrane of SR-BI-expressing cells have shown that the more hydrophobic lipids (e.g., unesterified cholesterol, cholesteryl esters, and triglycerides) can transfer more easily than the more polar phospholipid molecules (135). The physiological relevance of SR-BI-mediated selective lipid uptake in cultured cells was first directly established using steroidogenic cells, which express high levels of SR-BI (see below), HDL binding and selective cholesterol uptake, and use HDL cholesterol for steroidogenesis. An anti-SR-BI antibody simultaneously inhibited HDL binding, selective HDL lipid uptake, and HDL cholesterol-dependent steroidogenesis in murine adrenocortical cells (136). Similar results have been reported for rat ovarian granulosa cells in primary culture (137).

Several investigators have established that selective cholesteryl ester uptake is not dependent on a single class of lipoprotein. HDL (20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 37, 38, 39, 40, 138, 139, 140); reconstituted HDL-like particles prepared with apoA-I, apoC, or native or modified apoE (141); IDL (141, 142); and LDL (38, 140, 143, 144, 145, 146) have all been shown to be sources of cholesterol for selective lipid uptake in vitro and/or in vivo. It seems likely that SR-BI is responsible for many of these activities. For example, SR-BI mediates selective uptake of cholesteryl esters from LDL (128, 146, 147, 148) and the uptake of fluorescent lipids from VLDL (113); however, the fractional transfer of surface-bound LDL-cholesteryl esters to cells via SR-BI was significantly lower compared with HDL-cholesteryl esters (149). With regard to HDL subclasses, SR-BI-mediated selective cholesteryl uptake is greater from LpA-I compared with LpA-I:A-II (150). However, it is noteworthy that recent studies provide support for the existence of additional, SR-BI-independent selective lipid uptake mechanisms, such as one facilitated by cell surface proteoglycans (151, 152, 153, 154, 155).

SR-BI (71, 104, 113) and its homolog CD36 (87, 148, 156) can both bind HDL with high affinity, but of the two, only SR-BI mediates efficient selective lipid uptake (87, 148). Thus, the cell surface binding of HDL is not sufficient for selective lipid uptake. SR-BI apparently directly facilitates a lipid transfer step that follows binding (two-step mechanism of selective uptake; Ref.87). Analysis of SR-BI/CD36 chimeras established that sequences in the extracellular domain of SR-BI are sufficient to confer efficient selective lipid uptake activity on the cytoplasmic and transmembrane domains of CD36 (87, 148). However, the carboxy-terminal cytoplasmic domain of the receptor can influence the efficiency of selective lipid uptake, because SR-BII, the isoform of SR-BI with an alternative carboxy-terminal cytoplasmic domain, facilitates selective uptake less well than SR-BI (88). Analysis of the interaction of discoidal HDL particles prepared with wild-type and mutant apoA-Is with wild-type and mutant forms of SR-BI established that, somewhat analogously to enzyme/substrate interactions, there must be productive binding of an HDL particle to SR-BI to permit subsequent efficient lipid transport (125). Inhibition of SR-BI-mediated selective lipid uptake using small molecules identified in a chemical library screen (blockers of lipid transport) also supports a mechanistic coupling between HDL binding and lipid transport (157). Furthermore, apoA-I-deficient HDL binds with high affinity to SR-BI, but does not mediate efficient lipid uptake, indicating that the HDL features required for binding to SR-BI are distinct from those properties necessary for the selective cholesterol uptake from HDL to cells (158). The molecular details of the interactions between HDL and SR-BI that result in productive binding and lipid transport remain to be defined. Similarly, the precise mechanism of lipid transport has not been elucidated. The discovery of chemical inhibitors of the selective transfer of lipids mediated by SR-BI should provide new mechanistic insights into this lipid uptake pathway (157).

Just as specialized regions in the plasma membrane (clathrin-coated pits) are required for LDLR-mediated endocytosis (1, 25), specialized surface structures may sometimes be involved in SR-BI-dependent selective lipid uptake. Under certain culture conditions, SR-BI has been localized in cell surface caveolae (84, 88). Interestingly, SR-BI stabilizes caveolin-1 protein, independently of its transcriptional control, in cultured cells but not vice versa (159). The apparent clustering of SR-BI in caveolae in cultured cells under some culture conditions raises the possibility that these or similar lipid raft-like plasma membrane domains might be relevant for SR-BI-mediated cholesterol trafficking between lipoproteins and cells (84). Indeed, caveolae appear to be the sites where the initial steps of SR-BI-mediated selective cholesterol uptake into the plasma membrane take place before irreversible internalization of cholesterol into intracellular compartments (87, 160). Furthermore, selective uptake activity in certain cells appears to correlate with levels of caveolin-1 (161), which has cholesterol binding activity (162) and has been implicated in intracellular cholesterol transport (163, 164). One study has suggested that a caveolin-containing multiprotein complex may facilitate movement of SR-BI-derived cholesteryl esters from the cell surface to intracellular compartments (165). However, other studies have shown that: 1) caveolin-1 negatively regulates SR-BI-mediated selective uptake of HDL cholesteryl esters (166); 2) SR-BI-mediated selective uptake of HDL cholesteryl esters is not affected by caveolin-1 expression (167); and 3) localization of SR-BI in caveolae or raft-like domains is not required for selective cholesteryl ester uptake (168). In murine adrenocortical cells, SR-BI was found in numerous circular and oval structures of the cell surface that appear to represent cross-sections through previously described microvilli-rich intercellular channels (Fig. 2; and Ref.169). HDL has been shown to accumulate in these channels (170, 171, 172), and they may play an important role in the selective lipid uptake process (134, 169, 170, 171, 172). SR-BI has been reported to: 1) be present in microvillar channels in vivo (39, 137); 2) induce microvillar channel formation when overexpressed in insect cells (134); and 3) play a role in the formation of microvillar channels in murine steroidogenic cells in vivo (173). The important question as to whether the selective uptake of lipid from HDL occurs at the cell surface, in some internal compartment followed by retroendocytosis (secretion) of the lipid-depleted lipoprotein particle, or both has not yet been unequivocally resolved (11, 12, 14, 38, 39, 93, 137, 169, 170, 171, 174). Recent studies in which purified SR-BI was reconstituted into phosphatidylcholine/cholesterol liposomes have established that SR-BI can mediate HDL binding and selective lipid uptake without the required intervention of other proteins or specialized cellular structures or compartments (175).

View larger version (58K): [in this window] [in a new window]   Figure 2. Regulation of SR-BI expression in rodent steroidogenic tissues. Left, In murine adrenal glands, SR-BI immunostaining (brown) was found in the adrenal cortex but not in the medulla (Med.) of saline-treated control animals [top panels (169 182 )]. SR-BI staining was strongest in the outer zona fasciculata of the adrenal cortex and decreased toward the inner zona reticularis. ACTH treatment markedly increased SR-BI immunostaining in the cortex (bottom panels). At higher magnification, SR-BI was most abundantly detected in the plasma membrane of zona fasciculata cells, and the increased intensity of staining in the ACTH-treated mice was seen at both the sinusoidal surface and the intercellular junctions of steroidogenic cells. SR-BI was found in numerous circular and oval structures of the cell surface that appear to represent cross-sections through microvilli-rich intercellular channels (169 ). Right, In rat testes, little immunodetectable SR-BI was found in testicular steroidogenic Leydig cells of control animals [top (182 )], but not in the seminiferous tubules (S.T.). After human chorionic gonadotropin (hCG) hormone administration, a dramatic increase in SR-BI expression was exclusively localized to Leydig cells [bottom (182 )]. [Reproduced or derived with permission from A. Rigotti et al.: J Biol Chem 271:33545–33549, 1996 (169 ); and K. Landschulz et al.: J Clin Invest 98:984–995, 1996 (182 ).]

The precise mechanisms for the intracellular transport of the HDL lipids taken up by SR-BI-mediated selective uptake remain mostly unknown (31). Current evidence indicates that certain lipid transfer proteins, an intact Golgi apparatus, and some cytoskeletal elements are not required for intracellular traffic of cholesterol derived from plasma lipoproteins by selective lipid uptake. However, this process seems to be somewhat energy-dependent and mediated by N-ethylmaleimide-sensitive protein(s) that are yet to be identified (31). The role of the hepatic neutral cholesteryl ester hydrolase (reviewed in Ref.190) as well as cholesterol-rich carrier vesicles or protein complexes (165) in controlling the availability and traffic of unesterified cholesterol derived from the uptake of HDL cholesteryl esters by SR-BI deserves to be explored. Indeed, the generation of unesterified cholesterol after delivery of cholesteryl esters from HDL to cells via SR-BI involves cell type-specific neutral cholesteryl ester hydrolases (32). In addition, CLAMP modulates the efficiency of conversion of HDL cholesteryl esters taken up via SR-BI to intracellular unesterified cholesterol (97). A caveolin and annexin II-containing complex may in some cases be involved in the movement of SR-BI-derived cholesteryl esters from the plasma membrane to intracellular sites (165). Further studies of the detailed molecular mechanism(s) of intracellular cholesterol traffic linked to selective cholesterol uptake will provide new insights to the understanding of this pathway for lipoprotein metabolism.

B. Cholesterol efflux
The discovery of SR-BI-mediated selective lipid uptake (influx) raised the possibility that SR-BI might be able to mediate another step in the reverse transport process, cholesterol efflux from cells to appropriate acceptors in the extracellular fluid (11, 89). Consistent with this proposal were the observations that: 1) caveolae, the plasma membrane domains in which SR-BI appears, at least under some conditions, to be concentrated in cultured cells, have been proposed to be major sites of cholesterol efflux from cells (191); and 2) SR-BI expression levels in different cell lines correlate with rates of unesterified (free) cholesterol efflux to HDL in those cells (126). The ability of SR-BI to mediate cholesterol efflux from cells was directly established by studies in which SR-BI-transfected cultured cell lines were shown to exhibit increased radiolabeled unesterified cholesterol efflux to HDL and phosphatidylcholine liposomes, but not lipid-free apoA-I (127). SR-BI has also been reported to influence cholesterol efflux from cells to exogenously added apoE (192). SR-BI-mediated transfer of unesterified cholesterol between cells and HDL is bidirectional (188). Because SR-BI-mediated transport does not appear to be dependent on ATP, net flux of lipids is likely driven by the concentration gradient between cells and extracellular donor/acceptor particles. Although it mediates efflux to lipid-poor HDL from macrophages loaded with nonmetabolizable cholesterol, SR-BI apparently facilitates selective cholesterol uptake from cholesterol-rich particles to the plasma membrane of hepatocytes and steroidogenic cells down a gradient determined by further intracellular cholesterol metabolism.

In all studies performed to date, SR-BI-mediated cholesterol efflux has been correlated with the ability of this receptor to mediate selective uptake (116, 193), suggesting that the mechanisms of influx and efflux may be linked. Analysis of mutant forms of SR-BI that have lost the ability to bind and mediate selective uptake from either HDL alone or HDL and LDL (but not acetylated LDL), combined with the analysis of the effect of antibodies that block lipoprotein binding to SR-BI, strongly suggests that lipoprotein binding is required for SR-BI-dependent cholesterol efflux (116). Furthermore, decreased SR-BI-mediated cholesterol efflux to HDL particles reconstituted with mutant apoA-I forms suggests that the formation of a productive complex between HDL and SR-BI is required for efficient cholesterol transport activity (125). However, an alternative view that cholesterol-acceptor binding to SR-BI is not required for SR-BI-dependent cholesterol efflux has been proposed (194, 195). Interestingly, it has been shown that SR-BI expression in cultured cells alters the accessibility of cellular cholesterol to modification by cholesterol oxidase added to the culture medium (194, 195). The relationship between this change in surface cholesterol accessibility and efflux is not yet certain, but this important observation is likely to lead to new insights into the mechanisms underlying SR-BI-mediated lipid transport.

SR-BI-dependent cholesterol efflux activity is modulated by enzymatic and nonenzymatic changes in HDL lipid composition (188). Both phosphatidylcholine and sphingomyelin enrichment of HDL increase net cholesterol efflux from SR-BI-expressing cultured cells. Cholesterol efflux mediated by SR-BI is significantly correlated with HDL phosphatidylcholine content, suggesting that the relative phospholipid/cholesteryl ester composition of HDL influences the efficiency of cholesterol efflux. This may be due to a dependence on the cholesterol gradient between HDL particles and cell plasma membrane (reviewed in Refs.68 and196).

Tangier disease, characterized by impaired cholesterol efflux from cells to lipid-free or lipid-poor apoA-I and dramatically reduced plasma HDL levels, is a consequence of mutations in the gene encoding the ABCA1 transporter (197, 198, 199, 200). This discovery has had a major impact on the study of cellular cholesterol efflux. The potential structural and/or functional interaction between ABCA1 and SR-BI in regulating cellular cholesterol efflux has only just begun to be explored (201). It is also interesting to note that although ABCA1 facilitates efflux of cholesterol, it may be a primary transporter for phospholipid efflux that secondarily drives cholesterol transport out of cells (202, 203, 204). It is conceivable that SR-BI and ABCA1 might act coordinately in cellular cholesterol efflux because nascent HDL particles formed from lipid-free apoA-I by ABCA1-mediated phospholipid efflux (202, 203, 204) could serve as an acceptor for further cholesterol efflux mediated by SR-BI. In this regard, it is interesting that ABCA1 binds lipid-free apoA-I in preference to spherical HDL particles (203, 204, 205, 206), and the opposite is true for SR-BI (117, 118, 119). A study involving coexpression of high levels of SR-BI and ABCA1 suggested that SR-BI and ABCA1 may have distinct and, under some circumstances, opposite effects on unesterified cholesterol flux between HDL and cells (201).

The SR-BI-mediated transfer of unesterified cholesterol between cells and the extracellular fluid need not always involve lipoproteins or phospholipid carriers. SR-BI can facilitate the cellular uptake of unesterified cholesterol added directly to culture medium in the absence of lipoproteins (147, 207). This lipoprotein-independent activity of SR-BI may be related to the recent report that SR-BI is indeed a cholesterol binding protein (206).

	VII. Tissue Expression, Regulation, and Function of SR-BI

Top Abstract I. Introduction II. Lipoprotein Metabolism III. Lipoprotein Receptors IV. Structure and Subcellular... V. Ligand Binding Activities... VI. Lipid Transfer Activity... VII. Tissue Expression,... VIII. Mechanisms Underlying the... IX. SR-BI and Lipoprotein... X. Role of SR-BI... XI. SR-BI in Humans XII. Concluding Remarks References

 
Immunochemical methods have shown that SR-BI is most highly expressed in adult mammal tissues that are the principal sites of selective lipid uptake in vivo, i.e., the liver and steroidogenic tissues (71, 182). In addition, SR-BI has been observed at lower levels in intestine (94, 95, 182, 207), mammary glands of pregnant rats (71, 182), trophoblast, yolk sac and placenta during intrauterine development (208, 209), and uterine endometrium (210). The presence of SR-BI has also been reported in lung (211), biliary tree (96), macrophages (104, 126, 212, 213, 214), endothelial cells (132, 208, 215, 216, 217), smooth muscle cells (217), neuroglia (109, 110, 218), retinal pigmental epithelial cells (219), and keratinocytes (220). Substantial amounts of SR-BI mRNA have been detected in 3T3-L1 adipocytes and murine adipose tissue (70); however, the relationship between the mRNA and protein levels of SR-BI in fat is not yet resolved (71, 182). In addition, regulation of SR-BI protein expression is not always correlated with variations in mRNA levels (91, 211, 221, 222, 223). Thus, evaluation of the absolute levels and changes in SR-BI mRNA levels may not necessarily reflect the levels of SR-BI protein and SR-BI activity in tissues or cells.

A. Liver
The liver expresses the highest levels of total tissue SR-BI protein, a finding consistent with the major role of the liver in selective HDL-cholesterol uptake. The liver accounts for 90% of selective HDL cholesteryl ester uptake as well as 50% of total HDL cholesteryl ester clearance in non-CETP-expressing rodents (rats and mice; Refs.20, 21, 22, 23), and it mediates approximately 20% of total HDL cholesteryl ester clearance in CETP-expressing animals (224). SR-BI may account for much of the previously described hepatic lipoprotein binding site activity (225), because both SR-BI (70, 71, 113, 128, 146) and lipoprotein binding site (114, 140, 142, 143, 226) exhibit multilipoprotein binding as well as selective cholesterol uptake activities for different classes of lipoproteins.

Under basal conditions, most hepatic SR-BI expression is in parenchymal cells (90, 91, 212). Although SR-BI can be detected in the canalicular domains of hepatocytes in mice overexpressing hepatic SR-BI (89) and in cultured hepatocyte couplets (93), immunofluorescence analysis indicates that SR-BI is almost exclusively expressed on the sinusoidal surface of hepatocytes in wild-type animals (90, 91). In addition to its expression in liver parenchymal cells, SR-BI protein expression has been reported in Kupffer (212, 227) and liver endothelial cells (227).

Hepatic SR-BI expression can be regulated by a series of dietary, hormonal, and pharmacological manipulations. In the hamster, dietary plant-derived polyunsaturated fatty acids have been shown to stimulate hepatic SR-BI expression and HDL cholesteryl ester uptake (228). In contrast, dietary myristic acid decreased liver SR-BI levels in association with increased plasma HDL levels in hamsters (229). SR-BI expression is inversely regulated by cellular -tocopherol concentrations in the HepG2 human hepatocyte cell line and by dietary vitamin E supply in the mouse liver (223). In hamsters, a novel acyl-coenzyme A:cholesterol acyltransferase inhibitor has been reported to increase hepatic SR-BI, but not LDLR, expression (230). In contrast, lipopolysaccharide (LPS), TNF, and IL-1 decreased hepatic SR-BI mRNA and protein levels in the Syrian hamster (231). Streptozotocin administration in genetically hypercholesterolemic RICO rats increased SR-BI protein levels, which correlated with lower plasma HDL (232). On the other hand, insulin-treated diabetic hamsters exhibited lower levels of SR-BI compared with similar hamsters injected with saline only (233). In rats, estrogen administration at pharmacological levels suppresses total hepatic SR-BI expression (182, 212, 234, 235), which also correlates with decreased hepatic selective cholesterol uptake (236). Interestingly, estrogens increase hepatic SR-BII expression (235). The decrease in hepatic SR-BI expression was an indirect effect that was abolished by hypophysectomy and dependent on the estrogen-induced increase in LDLR activity (90). Other studies established that administration of high levels of estrogen increases hepatic Kupffer cell (macrophage) expression while lowering parenchymal cell expression (212). This study also demonstrated similar reciprocal changes in parenchymal and Kupffer cell expression after cholesterol feeding (212). However, studies with hamsters and mice did not find that hepatic SR-BI levels and/or HDL cholesteryl ester transport were regulated by changes in dietary cholesterol (237, 238, 239). In addition, a deficiency of plasma HDL in apoA-I KO mice does not increase hepatic SR-BI expression (240). Taken together, these studies indicate that, at least in some species, in vivo hepatic levels of SR-BI are not under sensitive feedback control by hepatic cholesterol content and/or plasma HDL-cholesterol concentration.

A recent study has shown that prolonged adrenal stimulation by ACTH in rats and mice decreased hepatic SR-BI protein expression associated with increased plasma HDL-cholesterol levels (241). This effect was not reproduced by exogenous corticosteroid administration and was abolished by adrenalectomy (241). These findings suggest that under chronic ACTH stimulation, adrenals release a factor that represses hepatic SR-BI levels as a potential mechanism for preferential channeling of plasma lipoprotein cholesterol to the adrenal tissue during stress.

In addition, the effects of pregnancy and lactation on hepatic SR-BI expression in female rats has been described (242). Hepatic levels of SR-BI increase in late gestation, are maintained at those levels early after birth, return to nonpregnant levels by 3 d postpartum, drop under control levels in late lactation, and revert to basal levels in postlactating females. Interestingly, the regulation of SR-BI and LDLR expression was not the same, suggesting that the roles of these receptors during pregnancy and lactation differ substantially.

The regulation of hepatic SR-BI expression by nuclear receptors (91, 243, 244, 245) and the roles of hepatic SR-BI expression in lipoprotein metabolism and associated diseases are considered below.

B. Adrenal gland
The highest expression levels of SR-BI per gram of tissue have been found in rodent adrenal glands (71, 182). High levels of SR-BI expression also have been detected in the human adrenal gland (80). Immunofluorescence and immunohistochemical methods have established that SR-BI is expressed primarily on the surfaces of steroidogenic parenchymal cells, such as in the zona fasciculata and zona reticularis cells of the adrenal cortex (Refs.169 and182 ; Fig. 2). High resolution immunohistochemistry established that in murine adrenocortical cells, SR-BI was found in numerous circular and oval structures on the cell surface that appear to represent cross-sections through previously described structures rich in microvillar channels (Ref.182 and Fig. 2). These channels accumulate HDL, and they seem to play an important role in the selective lipid uptake process (38, 170, 172). The formation of similar cell surface double-membrane channels has been observed in insect cells expressing high levels of recombinant SR-BI (134). In addition, hormonal regulation of adrenal SR-BI expression is closely associated in vivo with changes in microvillar mass and microvillar channel formation in this rat steroidogenic tissue (246). By ultrastructural analysis of adrenal glands obtained from SR-BI-deficient mice, it has been established that SR-BI expression is essential for microvillar channel formation and the localization of HDL particles on the plasma membrane of adrenocortical cells (173). This finding indicates that expression of SR-BI can induce channel formation, probably directly as a consequence of the high-density expression of the receptor on the surfaces of cells.

ACTH treatment of mice, which induces adrenal steroidogenesis in concert with enhanced selective uptake of HDL cholesterol (reviewed in Ref.247), increases SR-BI protein expression in adrenocortical cells (Ref.169 ; Fig. 2). It seems likely that ACTH acts directly on the adrenocortical cells, rather than indirectly, because it also dramatically stimulates SR-BI protein and mRNA levels in cultured murine adrenal cells (169, 177, 248). The effect of ACTH on SR-BI expression is likely mediated by the second messenger cAMP (80, 249, 250). SR-BI expression in cultured human adrenal cells and its regulation by ACTH (249, 250) are similar to those observed in murine cells (169, 177, 248). The ACTH-mediated regulation of human SR-BI in vivo is probably similar to that of the mouse, because levels of SR-BI mRNA in normal adrenal tissue adjacent to Cushing adenoma are low, likely due to the reduced plasma ACTH levels found in these patients (249).

In vivo administration of the glucocorticoid dexamethasone, which suppresses ACTH secretion and thus adrenocortical steroidogenesis, dramatically suppresses SR-BI protein expression in the murine adrenal gland, suggesting that the relatively high basal levels of SR-BI in this tissue may be due to endogenous basal ACTH production (169). Taken together, these studies of hormonal regulation of adrenal SR-BI expression strongly support the hypothesis that in this tissue SR-BI mediates physiologically relevant selective uptake of cholesteryl esters to supply substrate for steroid hormone synthesis.

In addition to the activity of the hypothalamic-hypophyseal-adrenal axis, the adrenal cholesterol content can independently regulate adrenal SR-BI levels, although this cholesterol-dependent effect is overridden by ACTH regulation in vivo and in vitro (80, 248). Increased adrenal SR-BI expression has also been reported in apoA-I-deficient (177), hepatic lipase-deficient (177), and LCAT-deficient (251) mice, suggesting that the expression of these lipoprotein metabolism-related genes directly or indirectly influences the levels of SR-BI in the adrenal gland. However, another study of apoA-I KO mice detected no changes in SR-BI mRNA or protein in the adrenal gland when compared with apoA-I-expressing animals (240). Further studies are required to established the basis for these divergent results.

In vivo estrogen treatment of rats: 1) induces adrenocortical expression of SR-BI (182); 2) enhances binding of HDL (apoA-I) to the surface of adrenal cells (182); 3) stimulates uptake of a fluorescent lipophilic dye surrogate for cholesteryl esters, DiI (71), from circulating DiI-labeled HDL (182); and 4) increases adrenocortical selective uptake of HDL cholesteryl ester (252). The relationship in the rat between estrogen’s stimulation of adrenocortical expression and suppression of hepatic expression of SR-BI is not established. Hypophysectomy dramatically decreases adrenal SR-BI expression (182), and chronic estrogen therapy affects ACTH secretion (253), strongly suggesting that trophic hormones play a key role in the estrogen-induced regulation of this receptor. In fact, dexamethasone administration, which inhibits endogenous ACTH production, blocked the increase in SR-BI expression in the adrenals of estrogen-treated hypophysectomized rats, demonstrating that ACTH is a critical factor involved in the estrogen-associated increase in adrenal SR-BI expression (90).

The functional significance of adrenocortical SR-BI expression has been established in vitro and in vivo. Anti-SR-BI antibodies that block HDL binding inhibit HDL-dependent steroidogenesis in cultured murine adrenocortical cells (136). Furthermore, SR-BI can provide HDL cholesteryl esters for steroidogenesis in cultured human adrenocortical cells, suggesting a potential role for SR-BI in human adrenals (254). Under normal physiological conditions, however, cultured human adrenocortical cells may derive more cholesterol from the LDLR pathway than the HDL pathway (250), suggesting that in humans LDL may be the primary lipoprotein delivering cholesterol to the adrenals via LDLRs. Nevertheless, in FH patients or patients with abetalipoproteinemia, plasma cortisol responses to ACTH are normal except under prolonged maximal stimulation (see above), indicating the existence of LDLR-independent pathways for adrenal lipoprotein cholesterol uptake. Furthermore, treatment of FH patients with mevinolin, a statin that inhibits cholesterol synthesis, did not result in impairment in adrenal function (53, 59). These results suggest that normal LDL uptake via LDLRs is not required for essentially normal adrenal function and that SR-BI might be able to supply cholesterol for steroidogenesis in human adrenals, at least when the LDLR pathway is impaired.

Analysis of mice with targeted null mutations in the SR-BI gene has established that SR-BI plays a role in murine adrenocortical cholesterol metabolism. Biochemical and histochemical studies established that these mice exhibit depleted adrenal cholesterol stores (cytoplasmic cholesteryl ester storage droplets) in a gene-dose-dependent manner with reductions of 42% in heterozygotes and 72% in homozygote null SR-BI mutants (SR-BI KO mice; Ref.255 ; Fig. 3). Similar lipid depletion is also observed in the ovaries (see below). Comparable adrenal cholesterol depletion has been observed in other murine mutants with disturbances in HDL metabolism, including apoA-I-deficient (172) and LCAT-deficient (251) mice, both of which exhibit abnormally high adrenal SR-BI mRNA levels (177, 251) that presumably represent regulatory responses made to compensate for reduced uptake of HDL-derived cholesterol. Furthermore, apoA-I KO mice also had blunted adrenal steroidogenic response to ACTH stimulation (172). Although SR-BI KO mice show no gross phenotypic indications of adrenal insufficiency, additional studies will be required to determine whether there are more subtle alterations in adrenal steroidogenesis in these mice. It is important to note that, except under conditions of very high demand, endogenous cholesterol synthesis via the HMG CoA reductase pathway would be expected to provide adequate amounts of substrate cholesterol for steroidogenesis in the adrenal glands, ovaries, and testes, even when LDLR- or SR-BI-mediated cholesterol import was not available.

View larger version (116K): [in this window]