Liver-specific overexpression of scavenger receptor BI decreases levels of very low density lipoprotein ApoB, low density lipoprotein ApoB, and high density lipoprotein in transgenic mice.

Scavenger receptor BI (SR-BI) is known to mediate the selective uptake of high density lipoprotein (HDL) cholesteryl ester (CE) in liver and steroidogenic tissues. To evaluate the role of SR-BI in plasma lipoprotein metabolism, we have generated transgenic mice with liver-specific overexpression of murine SR-BI. On a chow diet SR-BI transgenic (SR-BI Tg) mice have decreased HDL-CE, apoA-I, and apoA-II levels; plasma triglycerides, low density lipoprotein (LDL) cholesterol, and very low density lipoprotein (VLDL) and LDL apoB were also decreased, compared with control mice. Turnover studies using non-degradable CE and protein labels showed markedly increased total and selective uptake of HDL-CE in the liver and increased HDL protein catabolism in both liver and kidney. To evaluate the changes in apoB further, mice were challenged with high fat, high cholesterol diets. In SR-BI Tg mice plasma apoB levels were only 3-15% of control levels, and the dietary increase in VLDL and LDL apoB was virtually abolished. These studies show that steady state overexpression of hepatic SR-BI reduces HDL levels and increases reverse cholesterol transport. They also indicate that SR-BI can play a role in the metabolism of apoB-containing lipoproteins. The dual effects of increased reverse cholesterol transport and lowering of apoB-containing lipoproteins that result from hepatic SR-BI overexpression could have anti-atherogenic consequences.

tor BI (SR-BI), mediates high affinity binding of HDL and the selective uptake of HDL cholesteryl ester (CE) (5), a process for delivery of cholesteryl ester into cells without degradation of HDL proteins (6). Furthermore, SR-BI mRNA and protein levels are highest in adrenal gland, ovary, testis, and liver, tissues that display greatest selective cholesteryl ester uptake from HDL (7)(8)(9). SR-BI expression in steroidogenic cells is regulated by hormones and mutations that alter cholesterol supply or metabolism in those tissues in vivo (8 -11). More recently, strong support for the role of SR-BI in HDL metabolism has been provided by studies of mice with a targeted mutation resulting in decreased SR-BI gene expression (12,13). These mice demonstrate increased plasma HDL cholesterol, decreased adrenal cholesterol content (12,13), and decreased hepatic fractional clearance rate (FCR) for HDL-CE (13), suggesting that SR-BI is the major molecule mediating HDL-CEselective uptake in the liver. By contrast, adenovirus-mediated, hepatic overexpression of SR-BI in mice results in depletion of plasma HDL and an increase in biliary cholesterol concentration (14). Although these studies nicely demonstrate the effect of acute overexpression of SR-BI on HDL levels (14), they do not necessarily demonstrate plasma lipoprotein changes that would accompany steady state overexpression of SR-BI.
In this paper we report an in depth study of transgenic mice with hepatic overexpression of murine SR-BI. These studies were designed to understand better the role of SR-BI in HDL metabolism and reverse cholesterol transport. During the initial characterization of these animals on a chow diet, we observed decreased LDL cholesterol and apoB levels. Whereas the mouse model studies to date have focused on HDL changes, SR-BI was originally identified as a receptor recognizing both native and modified LDL (15). Thus, further studies were performed on high fat, high cholesterol diets in order to delineate the effects of SR-BI on plasma apoB levels.

MATERIALS AND METHODS
Generation of Transgenic Mice-A 1.5-kilobase cDNA fragment of murine SR-BI (16) was cloned into the HpaI site of the pLIV-7 plasmid (17), kindly provided by Dr. John M. Taylor (Gladstone Institute of Cardiovascular Disease, University of California, San Francisco). A linearized fragment of the construct containing the promoter, first exon, first intron, and part of the second exon of the human apoE gene, the murine SR-BI cDNA, and the polyadenylation sequence, and hepatic control region of the apoE/C-I gene locus was used to generate transgenic mice by standard procedures. Founder animals were backcrossed to C57Bl/6J mice and two transgenic mouse lines, SR-BI Tg(1) and SR-BI Tg(2), were established. Studies in this paper were performed using 8 -10-week-old SR-BI Tg(1) or SR-BI Tg(2) N2 or N3 mice positive for both SR-BI transgene genotype and phenotype (decreased plasma total cholesterol) versus control littermates negative for the SR-BI transgene.
For studies of responses to high fat diets, mice were fed either a Western type diet containing 20% hydrogenated coconut oil and 0.15% cholesterol (Research Diets, Inc.) or a very high cholesterol diet containing 1.25% cholesterol, 7.5% cocoa butter, 7.5% casein, and 0.5% sodium cholate for 2 weeks.
Plasma Lipoprotein Analysis-Total plasma cholesterol, free cholesterol, phospholipids, and triglycerides were determined using commercial enzymatic assays (Wako, Japan) (8). Determination of plasma apoB levels was carried out using an enzyme-linked immunosorbent immunoassay with an affinity purified polyclonal antibody against murine apoB. For SDS-PAGE, VLDL (d Ͻ1.006 g/ml), IDL ϩ LDL (d ϭ 1.006 -1.055 g/ml), and HDL (d ϭ 1.055-1.21 g/ml) were separated by sequential density ultracentrifugation of pooled mouse plasma. In some experiments VLDL ϩ IDL (d Ͻ1.006 -1.019) and LDL (d ϭ 1.019 -1.055) were separated as indicated. Denaturing polyacrylamide gel analysis of isolated lipoproteins was performed using 4 -20% SDS-PAGE gradient gels from Bio-Rad. Gels were stained with Coomassie Brilliant Blue R, and the identity of individual apolipoproteins was confirmed by Western analysis.
HDL Catabolism-HDL was prepared in the density range 1.063-1.21 g/ml from plasma of C57BL/6 wild type mice, dialyzed against phosphate-buffered saline containing 0.3 mM EDTA and 0.02% NaN 3 , and radiolabeled in the protein moiety with 125 I-N-methyltyramine cellobiose ( 125 I-NMTC) (18) Experiments to determine plasma decay of both HDL tracers and their tissue sites of uptake were carried out (19,20). Food was removed from five female control and SR-BI Tg mice 4 h before tracer injection, and animals were fasted throughout the 24-h study period but had free access to water. Doubly radiolabeled HDL was injected at 10:00 a.m. in an iliac vein, and blood samples were drawn from the tail vein of each animal at 0.08, 0.5, 2.0, 5.0, 9.0, and 24.0 h post-injection. Plasma samples were directly radioassayed for 125 I and analyzed for [ 3 H] after lipid extraction (19). 24 h after tracer injection the animals were anesthetized and perfused with saline (50 ml per animal), and organs were collected, weighed, homogenized, and radioassayed. Tissue content of 125 I radioactivity was directly assayed and that of [ 3 H] was analyzed after lipid extraction. Based on plasma decay of both HDL tracers, plasma FCRs were calculated using a two-compartment model (21). Organ FCRs, representing the fraction of the plasma pool of the traced HDL component cleared per h by an organ, were calculated as the plasma FCR ϫ fraction of total tracer (%) recovered in a specific organ (19,20).
Miscellaneous-Western blot analysis for SR-BI and Southern and Northern analysis were performed as described (8,22). Dot blot was carried out with a 900-base pair cDNA fragment of mouse apoB to determine hepatic apoB mRNA levels. Hepatic and adrenal cholesterol contents were measured by gas-liquid chromatography (23). Plasma relative LCAT activity was determined by examining conversion of endogenous plasma free cholesterol to cholesteryl ester. Plasma-specific LCAT activity was determined using reconstituted discoidal HDL as exogenous substrate, which was prepared with the sodium cholate method at an initial molar ratio of 80:4:1:160, egg phosphatidylcholine: cholesterol:apoA-I:sodium cholate (34). Briefly, each assay mixture contained reconstituted HDL (24 g of apoA-I), 4% defatted bovine serum albumin, and 4 mM ␤-mercaptoethanol in a total volume of 0.5 ml. The reaction was initiated by adding the indicated amount of plasma and carried out at 37°C for 20 min. LCAT activity was determined by the percentage conversion of [ 14 C]cholesterol to CE.

SR-BI Expression in SR-BI Tg
Mice-Two separate lines of SR-BI Tg mice, SR-BI Tg(1) and SR-BI Tg(2), were established. The SR-BI Tg(1) mice demonstrated a pattern of marked liverspecific overexpression of SR-BI mRNA (Fig. 1A); there was no appreciable expression in the kidney (Fig. 1A, lane 1). Western analysis showed a 12-fold increase in hepatic membrane SR-BI levels in transgenic mice (Fig. 1, B and C). Similar levels of expression were observed for both lines of SR-BI Tg mice.
Plasma Lipids, Lipoproteins, and Apolipoproteins of Mice on Chow Diet-Analysis of plasma lipids on a chow diet revealed that female SR-BI Tg mice (both lines) had a profound 92-94% decrease of plasma total cholesterol (TC) (Table I), with de-creases in both free cholesterol (FC) (ϳ80%) and cholesteryl ester (CE) (96%). There was also a significant but less pronounced decrease in plasma phospholipids (PL) (ϳ75%) and triglycerides (TG) (45-58%) ( Table I). Similar results were obtained for male mice (not shown).
When plasma was analyzed by fast protein liquid chromatography (FPLC), most of the CE and FC were in the HDL fraction in the control mice on the chow diet (Fig. 2, A and B). By contrast HDL-CE and FC were almost undetectable in SR-BI Tg mice. HDL phospholipids were also markedly decreased (Fig. 2C). There was also a decrease in lipids in VLDL and LDL region, although these were also low in control mice.
Assessment of apolipoprotein composition of centrifugally isolated lipoproteins by reducing SDS-PAGE gels revealed a marked decrease of HDL apoA-I, apoA-II, and apoE levels in SR-BI Tg(1) mice (Fig. 3A). The VLDL and LDL apoB and apoE levels also were decreased. The results were confirmed by Western analysis using antisera specific for murine apoA-I, apoA-II, and apoE. Similar results were obtained in four separate analyses of pooled plasma from a total of 9 SR-BI Tg(1) mice and 10 control mice and were also confirmed in the SR-BI Tg(2) line (data not shown).
HDL Metabolism-The changes in HDL in SR-BI Tg mice resemble these occurring 3 days after adenovirus-mediated expression of SR-BI (14) where HDL turnover was evaluated using 125 I and DiI labels (14). Next we carried out HDL turnover studies using non-degradable radiolabels (18,19). In the control mice, the higher rate of removal from plasma of the lipid ([ 3 H]CEt), relative to protein ( 125 I-NMTC), represents whole body selective uptake of HDL-CE (Fig. 4A). There was a significantly accelerated rate of clearance for both tracers in SR-BI Tg mice. The plasma FCRs calculated from these decay curves showed a 370% increase in protein catabolism and 330% increase in lipid tracer catabolism (Fig. 4B). The selective removal of HDL-CE from plasma, calculated as the difference between CE and protein FCRs, was increased by 260% in SR-BI Tg mice.
Tissue sites of tracer uptake from doubly radiolabeled HDL were determined, and results are expressed as the organ FCRs (Table II). The liver was the predominant organ for both HDL lipid and protein catabolism (19,20). The higher liver FCR for lipid, relative to protein, indicates selective uptake of HDL [ 3 H]CEt in the liver. In contrast, a negative value of 3 H minus 125 I was derived for kidney FCR (Table II), indicating this organ is a major site for selective HDL protein catabolism (20). SR-BI Tg mice showed a substantial increase in hepatic FCRs for both HDL protein (7.5-fold) and lipid (6.4-fold). The renal FCR of HDL proteins was also markedly increased (6.6-fold) in SR-BI Tg mice, whereas this organ contributed little to the clearance of HDL lipid in wild type or SR-BI Tg mice. Adrenal FCRs for HDL protein and lipid were increased 10.5-and 6.3fold, respectively, and the adrenal-selective uptake of HDL-CE was increased 5.6-fold. This finding may have reflected an up-regulation of endogenous adrenal SR-BI expression secondary to reduced HDL levels and depletion of adrenal cholesterol stores (see below). Other organs (heart, spleen, and stomach), with minor contribution to HDL lipid and protein uptake, did not display any major changes in SR-BI Tg mice.
The HDL turnover studies demonstrated that the reduced plasma HDL lipids and apolipoproteins were at least in part due to accelerated HDL catabolism in SR-BI Tg mice. We also measured hepatic apoA-I mRNA levels and found no difference between the control and SR-BI Tg animals (not shown). Hepatic free cholesterol was increased by 43% (p Ͻ 0.001) in SR-BI Tg mice (Table III). By contrast adrenal cholesteryl ester virtually disappeared in SR-BI Tg mice, and free cholesterol was also decreased (Table III), probably reflecting the decreased plasma HDL-CE levels that result from hepatic overexpression of SR-BI (8). There was a 3.5-fold increase in adrenal SR-BI protein (not shown), probably secondary to decreased adrenal cholesterol content (8). The hepatic LDL receptor mRNA levels in SR-BI Tg mice were not altered, as determined by Northern analysis using poly(A) ϩ RNA from liver (data not shown).
Plasma Lipoprotein Responses to High Fat, High Cholesterol Diets-An unexpected finding in SR-BI Tg mice on the chow diet was the marked decreases in LDL apoB levels (Fig. 3A). To evaluate apoB changes further, mice were challenged with a Western type high fat diet (20% hydrogenated coconut oil and 0.15% cholesterol) or a very high cholesterol diet (1.25% cholesterol, 7.5% cocoa butter, 7.5% casein, 0.5% sodium cholate) for 2 weeks. In response to these diets VLDL and LDL apoB levels were increased in control mice, but changes in apoB were much smaller in SR-BI Tg mice (Fig. 3, B and C). The decrease in apoB was more pronounced for SR-BI Tg(2) mice than SR-BI Tg(1) mice. Quantitative determination of plasma apoB levels by immunoassay revealed that plasma apoB levels were markedly reduced by 89 -94% on the chow diet and 85-97% on the high fat, high cholesterol diets in SR-BI Tg mice compared with control mice (Table IV). The decreases in apoB appeared to include both apoB100 and apoB48 in VLDL and LDL (Fig. 3). Hepatic apoB mRNA levels were not significantly altered in SR-BI Tg mice relative to the control mice on the chow diet (not shown). Similar to the chow diet, apoA-I levels were much lower in SR-BI Tg mice than the control mice on high fat, high cholesterol diets (Fig. 3, B and C). However, apoE levels in HDL were not decreased on the high fat, high cholesterol diets.
On the Western type diet plasma TC was decreased by 27-36% and CE by 77-80% in SR-BI Tg mice compared with control mice (Table I). Surprisingly, plasma FC was approximately 2-fold higher in SR-BI Tg mice than in the control mice, and phospholipids were only slightly decreased (Table I). Plasma triglycerides were moderately decreased (36 -37%) in SR-BI Tg mice. On the very high cholesterol diet plasma TC, CE, FC, and PL were all decreased, but the changes were less pronounced than on the chow diet in SR-BI Tg mice (Table I).
FPLC analysis of plasma lipoprotein lipids of mice on the Western type diet showed that HDL-CE, FC, and PL were substantially depressed in SR-BI Tg mice (Fig. 2, A--C). Free cholesterol eluting in the VLDL region, however, was markedly increased, and free cholesterol eluting in the IDL/LDL region was moderately increased and shifted toward larger particle size. CE eluting in the VLDL region was only slightly increased in SR-BI Tg mice. VLDL-eluting free cholesterol accounted for Ͼ90% of total VLDL cholesterol, whereas IDL/LDL fractions contained ϳ70% total cholesterol as free cholesterol. Consistent with these alterations, phospholipids eluting in the VLDL and IDL/LDL region were also markedly increased in SR-BI Tg mice (Fig. 2C). The high content of free cholesterol and phospholipids, low cholesteryl ester, and low apoB (Fig. 2B) suggested that the VLDL/IDL fractions might contain lipoproteins consisting of apoB-free lamellar-free cholesterol/phospholipid particles, i.e. lipoprotein X-like particles of vesicular structure (24). Subsequently, SR-BI Tg mice were found to have functional LCAT deficiency (see below) in which lipoprotein X accumulates especially on a high fat diet (25,26).
The percentage composition of HDL fractions isolated from plasma of mice on Western type diet is shown in Table V. In SR-BI Tg mice, HDL was essentially devoid of cholesteryl ester, whereas triglyceride was increased and became the major neutral lipid. Phospholipids were moderately increased and so was free cholesterol. LCAT Activity-The pronounced accumulation of FC and PL in VLDL and LDL on the Western type diet suggested that SR-BI Tg mice might have a defect in plasma CE formation secondary to the markedly decreased HDL levels, as has been shown previously in apoA-I knock-out mice (27). Therefore, we determined the plasma cholesteryl ester formation (Fig. 5). SR-BI Tg mice displayed a markedly depressed fractional plasma CE formation rate as compared with the control animals on both chow and Western type diets. In SR-BI Tg mice, the marked depression of plasma apoA-I might affect LCAT activity, since apoA-I activates LCAT (28). To evaluate further the basis of decreased CE formation, LCAT levels were assessed, using addition of exogenous discoidal PL/apoA-I substrates to small amount of plasma. This revealed similar levels of LCAT in the control and SR-BI Tg plasma (for control 0.42, 0.74, and 1.20% of FC to CE conversion; for SR-BI Tg 0.39, 0.65, and 1.31% of FC to CE conversion by incubating with 0.1, 0.5, and 2 l of plasma, respectively), suggesting that the defect in plasma cholesterol esterification is related to the reduced levels of apoA-I. Thus, SR-BI Tg mice appear to have functional LCAT deficiency secondary to depletion of plasma apoA-I. DISCUSSION Kozarsky et al. (14) showed that the acute adenovirus-mediated overexpression of SR-BI in the liver resulted in a marked decrease in HDL cholesterol and apoA-I levels, enhanced clearance of HDL protein from plasma, increased hepatic uptake of DiI label from HDL, and increased biliary cholesterol levels. Our studies show a major decrease in HDL cholesterol, apoA-I, and apoA-II as a result of sustained hepatic overexpression of SR-BI in a transgenic mouse model. They further demonstrate increased selective uptake of HDL-CE in the liver, and increased uptake of HDL protein in both liver and kidney. Moreover, we observed a profound decrease in VLDL and LDL CE , SR-BI Tg(1) (n ϭ 10), or SR-BI Tg(2) (n ϭ 11) mice fed either the chow diet, the Western type diet, or the very high cholesterol diet for 2 weeks were analyzed by FPLC using two Superose columns in series. The same mice were fed these diets in sequence. B, FPLC free cholesterol profiles. C, FPLC phospholipid profiles. Data are shown for one of two independent FPLC analyses in which similar profiles were obtained. ࡗ, control; □, SR-BI Tg(1); E, SR-BI Tg(2). and apoB levels in SR-BI Tg mice compared with controls, and in one line of mice a failure to increase plasma apoB levels when challenged with high fat, high cholesterol diets. This provides the first in vivo evidence that SR-BI can play a role in the determination of plasma lipoprotein apoB levels, and this suggests that there will be important consequences of hepatic SR-BI overexpression on VLDL and LDL metabolism that may influence the outcome of atherosclerosis studies.
SR-BI was originally described as a receptor that bound both native LDL and acetyl-LDL with high affinity (15). CLA-1, the human homolog of SR-BI, has been shown to bind VLDL in addition to HDL and LDL (29). The ability of SR-BI to mediate the cellular uptake and degradation of LDL and VLDL has not been reported, and its role in apoB metabolism in vivo is unknown. One explanation for the decrease in VLDL and LDL apoB in SR-BI Tg mice is that hepatic SR-BI directly mediates the removal of apoB-containing lipoproteins from plasma. However, there are several alternative explanations. For example, SR-BI overexpression could mediate increased binding of VLDL and LDL to hepatocytes, and this might lead to increased particle catabolism via the LDL receptor or proteoglycan pathways (35). Another explanation is that SR-BI overexpression leads to decreased secretion of apoB from liver cells. However, the transgenic mice have increased uptake of HDL cholesterol and esterified fatty acids into the liver which would be more likely to increase apoB secretion (30). Another factor involved in reduced apoB levels could be the state of partial LCAT deficiency, since LCAT knock-out mice have reduced apoB levels (31). However, the moderate 30% reduction of apoB levels in knock-out mice with complete LCAT deficiency on a very high cholesterol diet (31) is unlikely to explain the profound 85-97% decrease of apoB levels in SR-BI Tg mice (Table  IV). The changes in apoB were observed in the context of about 12-fold overexpression of SR-BI in the liver. In contrast, mice with decreased SR-BI expression in the liver, as a result of disruption of SR-BI gene expression, do not display increased apoB in plasma lipoproteins compared with wild type mice on FIG. 3. Protein composition of plasma lipoproteins of control and SR-BI Tg mice. A, an aliquot of lipoproteins isolated by sequential density centrifugation from equal amount (500 l) of pooled plasma of female mice (n ϭ 5-7) were loaded onto reducing 4 -20% SDS-PAGE gels and stained with Coomassie Brilliant Blue. B, similar to A except that mice were fed the Western type diet for 2 weeks before samples were taken. C, similar to A except that mice were fed the very high cholesterol diet for 2 weeks. Data shown are representative of two to three different experiments with similar results.

FIG. 4. Plasma decay curves and FCRs for 125 I-NMTC and [ 3 H]Cet-labeled HDL.
A, plasma decay curves. Mice were injected with labeled HDL, and blood was collected periodically over 24 h, and plasma content of both tracers was determined as described under "Materials and Methods." The values are means Ϯ S.D. of the female control and SR-BI Tg mice (n ϭ 5). B, plasma FCRs for the labeled HDL. FCRs were calculated as described under "Materials and Methods" using the data from the plasma decay curves. Histogram represents means Ϯ S.D. and denotes the fraction of the plasma pool cleared per h ϫ 1000. FCR for selective uptake of HDL-CE was calculated by subtracting the FCR value of protein uptake from lipid uptake. chow diets (13). In the knock-out mice it is possible that decreased SR-BI expression is compensated by LDL receptor activity. The decrease in VLDL and LDL CE in SR-BI Tg mice (Fig. 2) might be due both to increased selective uptake of lipid from VLDL and LDL and to increased particle removal.
Paradoxically, even though plasma triglyceride levels and VLDL and LDL CE and apoB levels were decreased, SR-BI Tg mice showed increases in VLDL and LDL free cholesterol on Western type diet. In response to adenovirus-mediated overexpression of SR-BI (14), there was also a marked increase in VLDL and LDL cholesterol levels, which peaked 7 days after the nadir of HDL cholesterol. A potential explanation for these findings was provided by our discovery that mice with hepatic SR-BI overexpression have impaired plasma cholesteryl ester formation (total and fractional). On the Western type diet the particles accumulating in VLDL were found to contain free cholesterol and phospholipids and to be almost devoid of cholesteryl esters and apoB. Thus they are likely to represent lipoprotein X-like particles that are non-apoB-containing vesicular lipoproteins that accumulate in animals with LCAT deficiency, especially in response to high fat diets, where they probably represent surface remnants of triglyceride-rich lipoproteins (25,26). It is not clear why these particles were more prominent on the Western type diet than the very high cholesterol diet, although it could be related to a more pronounced rise in plasma Tg levels on the Western type diet    5. Plasma cholesteryl ester formation in control and SR-BI Tg mice. Plasma LCAT activity was determined according to procedures described under "Materials and Methods." Plasma from female control and SR-BI Tg(1) mice was incubated at 37°C, and an aliquot (15 l) of plasma was taken at the times indicated. Total and free cholesterol were determined, and cholesteryl ester formation was expressed as percentage decrease of plasma free cholesterol. The bar represents means Ϯ S.D., n ϭ 3. E, control; q, SR-BI Tg. (Table I).
An important feature of our study was the quantitative measurements of organ uptake of HDL cholesteryl esters and protein, using non-degradable radiolabeled lipid and protein.
One potential caveat to the turnover data is that the results could reflect the markedly decreased pool size of HDL in the SR-BI Tg mice. However, comparably reduced pool size of HDL in apoA-I knock-out mice does not affect the fractional catabolism of HDL-CE (23). We found that 12-fold overexpression of SR-BI in the liver led to a 6-fold increase in selective uptake of HDL-CE in the same organ, showing a remarkably increased selective uptake capacity in the liver (13). The SR-BI transgenic mice also have increased biliary free cholesterol content and decreased dietary cholesterol absorption. 2 Interestingly, we also observed an increase in HDL protein uptake in both liver and kidney (Table II). In the kidney where there was negligible expression of SR-BI (Fig. 1), these changes must reflect modifications of HDL size or composition that result from overexpression of SR-BI in the liver. 3 These results contrast with decreased expression of SR-BI where there were no changes in HDL protein uptake in the liver (13). Although SR-BI overexpression could be directly responsible for increased uptake of HDL proteins in the liver, we hypothesize that the marked modification of HDL secondary to increased SR-BI activity leads to entry of HDL into distinct HDL protein catabolic pathways that are active in the liver and kidney.
Our studies show that in addition to stimulating reverse cholesterol transport, SR-BI overexpression leads to marked decreases in VLDL and LDL CE and apoB levels. While an elucidation of the physiological role of SR-BI in the removal of apoB from plasma must await further studies with SR-BI knock-out mice, pharmacological overexpression of SR-BI is likely to have similar consequences to transgenic overexpression of SR-BI. The results of stimulating reverse cholesterol transport by SR-BI overexpression are uncertain, because they will also lead to reduced HDL levels. However, it seems likely that reductions in plasma apoB levels (and the associated decreases in VLDL and LDL cholesterol) will have anti-atherogenic consequences (33).