Endocytic adaptors Arh and Dab2 control homeostasis of circulatory cholesterol

High serum cholesterol (hypercholesterolemia) strongly associates with cardiovascular diseases as the atherogenic LDLs promote atheroma development in arteries (atherosclerosis). LDL clearance from the circulation by LDL receptor (LDLR)-mediated endocytosis by hepatic and peripheral tissues and subsequent feedback regulation of endogenous synthesis of cholesterol is a key determinant of serum LDL level. Human mutation analysis revealed that autosomal recessive hypercholesterolemia (ARH), an LDLR endocytic adaptor, perturbs LDLR function and thus impacts serum cholesterol levels. In our genetic analysis of mutant mice, we found that deletion of another LDLR endocytic adaptor, Disabled-2 (Dab2), only slightly affected serum cholesterol levels. However, elimination of both arh and dab2 genes in mice resulted in profound hypercholesterolemia similar to that resulting from ldlr homozygous deletion. In the liver, Dab2 is expressed in sinusoid endothelial cells but not in hepatocytes. When deleting both Dab2 and Arh, HMG-CoA reductase level increased to the level similar to that of ldlr knockout. Thus, in the absence of Arh, Dab2 in liver endothelial cells regulates cholesterol synthesis in hepatocytes. We conclude that the combination of Arh and Dab2 is responsible for the majority of adaptor function in LDLR endocytosis and LDLR-mediated cholesterol homeostasis.

Cholesterol is a major component of animal cell membranes as well as the starting substrate of steroid hormones. The homeostasis of this important and essential lipid is maintained by feedback regulation and transportation between tissues through the circulation (1). Serum cholesterol level is Since the original arh mutant allele contains a neomycin-resistant cassette flanked by two LoxP sites (19), the genotyping procedure was modified to distinguish the original arh KO as well as arh deltaflox , in which two LoxP sites were removed by Cre recombinase. Genotyping PCRs were performed with all primers at 1 µM concentration using the master mix from Denville Scientific. Four primers were used in the PCR reactions: mArhF3 (5′ GTG TTC AGA GGA GAG GAA CT 3′); mArhR3 (5′ CTG GTG TGC AGT TAG GTT CA 3′); XU2010-01 (5′ CTG CAG ACA AGA TGC ACG AC 3′); and Neo-36 (5′ CAG GAC AGC AAG GGG GAG GAT TGG G 3′). The parameters for the genotyping PCR program were as follows: 5 min at 94°C; 30 cycles (30 s at 94°C, 30 s at 57°C, 3 min at 72°C); 10 min at 72°C; and 4°C overnight. The predicted sizes of the amplicons are the following: wild-type, 650 bp; arh KO , 380 bp; and arh deltaflox , 300 bp. In this report, the genotypes of the mutant mice and derived cells are designated as arh (+/) or arh (/).
For the nomenclature of dab2 mutant mice, "fl" and "df" were used to indicate conditional and deleted alleles, respectively. Mutant mice and tissues derived from Sox2-Cre-mediated deletion are denoted as "HET" and "CKO." Mouse primary embryonic fibroblast (MEF) cells were genotyped directly following expansion in cultures and are noted as dab2 (+/+), (+/), or (/).

Profiling of serum lipids
Mice were fasted for 4-6 h and then anesthetized prior to tissue collection. Blood was collected by cardiac puncture, and the tissues and fats to be examined were dissected and collected. To obtain serum, blood samples were allowed to clot at room temperature for 15 min, and the serum was collected following centrifugation for 10 min at maximum speed in a bench-top clinical centrifuge and stored at 80°C until use. Lipid chemistry analysis was performed using an automated Vitros 250 chemistry analyzer (Johnson and Johnson) by the University of Miami Comparative Pathology Laboratory Core. The levels of serum cholesterol, triglyceride, and HDL were determined colorimetrically. LDL and VLDL levels were estimated with the following formulae: VLDL = 1/5 of triglycerides; LDL = total cholesterol  (HDL + VLDL).

Extraction and measurement of cholesterol in tissues
Cholesterol was extracted from liver samples according to the Folch method (29). Briefly, liver tissue was lysed in RIPA buffer and homogenized with glass beads. A total of 10 mg protein was added to 10 vol of an extract solvent containing chloroform-methanol (2:1, v/v). Monoacylglycerols (100 pmol) and acylethanolamides (100 pmol) were added to each sample as internal standards. Following vortexing to extract the lipids, the mixtures were centrifuged at 5,000 g for 10 min. The lipid extracts in the chloroform (lower phase) were collected, and chloroform was vaporized under vacuum with a gentle flow of nitrogen gas. The samples were analyzed by LC/MS in the Lipidomics Core of Rutgers University (http:// dixonlab.rutgers.edu/pages/lc_ms_facility).

Preparation of primary cells
MEFs were isolated from E12.5 embryos generated from designated genotypes (24). Briefly, the head of dissected embryos was first removed and used for PCR genotyping, and the remainder was minced using a scalpel. After incubation for 10 min in 0.25% trypsin at 37°C, the cells were pelleted by a brief centrifugation and plated in a T75 flask in DMEM media supplemented with 10% FBS. The cells from the embryonic outgrowths were harvested and expanded. Western blot analysis of the cells was used to confirm the genotype obtained from PCR of the head tissues.

Western blots and immunofluorescence microscopy
For Western blot analysis, cells were washed twice with cold PBS and collected in RIPA buffer (20 mM Tris pH 7.5, 50 mM NaCl, mutations were also identified in individuals with peculiarly low serum cholesterol (15). Small-molecule and antibody blockage of PCSK9 have now been developed and tested, showing promise as effective LDL cholesterollowering drugs (16,17).
The role of ARH in recruiting LDLR into clathrincoated vesicles in endocytosis has been studied in cultured cells (18), and the impact on circulating cholesterol has been investigated in mutant mice (19)(20)(21). Another endocytic clathrin adaptor, Disabled-2 (Dab2), is potentially an LDLR-selective adaptor (22). The activity and ability of Dab2 to mediate LDLR endocytosis have been demonstrated in cultured cells (23). However, Dab2-deficient mice have only minor changes in serum cholesterol level (24), and the involvement of Dab2 in cholesterol metabolism in intact animals is uncertain.
The autosomal dominant familiar hypercholesterolemia caused by inactivating mutations in the LDLR gene exhibits a greater serum LDL level and higher degree of severity of the symptoms than found with mutations in ARH gene (11). Analysis of mutant mice also showed that serum cholesterol is much higher in LDLR-null than in arh-deleted mice (19)(20)(21)25). Presumably, additional endocytic adaptors other than Arh may support LDLR endocytosis. Conventional dab2 gene deletion in mice results in early embryonic death (26)(27)(28). We generated a conditional knockout line and were able to produce Dab2-null mice to study their postnatal phenotypes (24). Here, we investigated the relative importance of Dab2, another endocytic adaptor capable of binding to LDLR (22), in LDLR-mediated clearance of circulating LDL and the regulation of cholesterol synthesis.
Mice were routinely maintained on a regular enriched diet, 5K20 chow (10% fat; LabDiets Inc.). A high-sucrose diet (no. 960237; MP Biomedicals) containing 60.2% sucrose and 20% casein, 15% cellulose, 0.3% DL-methionine, 3.5% minerals and vitamins, and <0.2% fat was used as a high-caloric-diet regimen for 3 months to induce hypercholesterolemia. Animals were housed on a 12:12 h light-dark cycle with ad libitum access to food and water. The Institutional Animal Care and Use Committee reviewed and approved all the experimental procedures using laboratory mice described in this study.

Genotyping of mutant mice
The PCR genotyping of ldlr knockout (25) and dab2 conditional knockout (24) mice was performed following the previously published protocols.
Duration Substrate (Pierce Biotech) was used for chemoluminescence detection of specific proteins.

Assay for LDL uptake by cells in cultures
MEFs were prepared from control and mutant mice with various genotypes as described previously (24). Cells were maintained in DMEM media supplemented with 10% FBS. The cells were switched to the DMEM media containing 5% lipoprotein-deficient serum for 12 h prior to assay. A total of 25 g/ml human LDL particles labeled with BODIPY dye (Sigma Inc.) were added to the cells. After 5 h incubation, cells were washed twice with PBS and fixed in 4% paraformaldehyde. DAPI (4',6-diamidino-2-phenylindole) was used as a nuclear counterstain. Images were acquired using a Zeiss Z1 inverted microscope operated by AxioVision software.

Quantitative RT-PCR for the analysis of gene expression
Total RNA was harvested from MEF cells using the Qiagen RNAeasy kit. cDNAs were prepared from the extracted RNA by reverse transcription with iScript™ cDNA Synthesis Kit according to the manufacturer's instructions (BioRad). Quantitative RT-PCR was performed using CFX Connect™ Real-Time PCR Detection System with 2× SYBR® Green Supermix and specific primers. Target gene quantification is reported with normalization to the signals obtained from reference gene GAPDH. The primers used were as follows: LDLR forward sequence, 5′ CTG GAC CGG AGC GAG TAC AC 3′; LDLR reverse sequence, 5′ GAC GCC GTG GGC 0.1% SDS, 0.5% sodium deoxycholate). The samples were then lyzed in SDS gel loading buffer and boiled for 10 min. The proteins were separated on 8% SDS-polyacrylamide gels and transferred from the gels to the nitrocellulose membranes. The membranes with proteins samples were used for Western blot analysis.
For immunofluorescence microscopy, tissue were fixed in 4% paraformaldehyde in PBS, cryopreserved in 30% sucrose, and embedded in OCT. Cryosections of 10 m thickness were subjected to antigen retrieval with 1% SDS and then incubated with specific primary antibodies. Fluorescently conjugated antibodies directed against the appropriate species of primary antibodies were used as secondary antibodies (24,30).
The secondary antibodies for Western blotting were conjugated with HRP and were used (1/5,000 dilution) following the instructions from the manufacturer (rabbit and mouse secondary antibodies were from BioRad; goat secondary antibodies were from Jackson ImmunoResearch). SuperSignal West Extended Fig. 1. Expression spectrum of Dab2, Arh, and LDLR mRNA in murine tissues. A: Microarray data sets were acquired from an online public data set GeneAtlas MOE430, gcrma (http://biogps.org/#goto = genereport and id = 13132). The high-throughput gene expression profiling for Dab2, Arh, and LDLR was generated from a diverse array of normal tissues, organs, and cell lines from mice. The final readouts were averages of four different probe sets with SEM shown as error bars, normalized to housekeeping genes, and ranked according to the abundance of mRNA expression. B: Selective tissues were harvested from 3-month-old mice, and lysates were prepared for Western blot analysis. white adipose, adrenal, lactating mammary glands; however, the expression, although present widely, is much lower in magnitude in other tissues including the liver (Fig. 1A). The presence of proteins were validated by Western blot of tissues obtained from experimental mice, and the results indicate that LDLR and Arh proteins are indeed abundant in liver, whereas Dab2 is much lower in the liver compared with other tissues analyzed (Fig. 1B).

RESULTS
We first surveyed the mRNA expression profile of LDLR, Arh, and Dab2 in mouse tissues and cell lines by analysis of GeneAtlas public data set. Liver is the top site among all tissues for which both LDLR and Arh are highly expressed, though both LDLR and Arh are widely expressed among other tissues albeit at lower levels (11) (Fig. 1). This is consistent with that liver is the main site for LDLR-mediated clearance of LDL particles, and Arh acts as the main endocytic adaptor. In comparison, Dab2 is expressed highly in several tissues and cell types including macrophages, kidney, Fig. 2. Generation of Dab2 and Arh double-knockout mice. A: Mice from colonies of dab2 flox, Sox2-Cre transgenic, and arh null were crossed to obtain compound Dab2-and arh-null mice. The arh mutant allele is noted as "*". A three-step strategy was designed to produce the double-knockout mice: dab2 (fl/df); Sox2-Cre; arh (/) (indicated with a red box). Only male mice carrying the Sox2-Cre transgene were used in the crosses to avoid transmission of the Cre protein through oocytes. B: MEFs were prepared from embryos with dab2 (fl/df), dab2 (df/df), dab2 (fl/df):arh (/), dab2 (df/df):arh (/), and ldlr (/) genotypes. The cells were expanded and genotyped to determine whether dab2 statues were (+/) or (/). Cell lysates were prepared for Western blotting to assay for Dab2 and Arh proteins. C: Hepatic tissues were harvested from the mutant mice of 3 months of age. Total cell lysates were prepared and analyzed by Western blotting. The dab2 genotypes from Sox2-Cre-mediated deletion were noted as HET or CKO. D: Expression of LDLR and VLDLR mRNA in dab2and/or arh-deleted MEF cells. Quantitative real-time RT-PCT analysis of LDLR and VLDLR mRNA isolated from control and mutant MEFs was performed in triplicate and reported as means plus SEMs. the genotypes on serum cholesterol level, following the experimental conditions used in the previous studies in Arh and LDLR knockout mice (19,20). In such a situation, the source of circulatory cholesterol was solely from endogenous synthesis. On a normal chow diet, there were no noticeable differences between the body weights of wild-type, mutant, and double-knockout mice. However, on highsucrose diet, mice with dab2 deletion gained less body weight than dab2 heterozygous (Fig. 3A). We found that the dab2 and arh double-knockout mice had serum cholesterol elevated to a level similar to that of the LDLR knockout (Fig.  3B). The high serum cholesterol in the double-knockout mice appears to be a combination of increased cholesterol in LDL, HDL, and VLDL (Fig. 3B). Interestingly, the serum triglyceride level in the double-knockout mice increased, but not to the same level as in LDLR knockout mice (Fig.  3B). Increased accumulation of cholesterol in the livers of the double-knockout mice was also found quantitatively comparable to those of LDLR knockout (Fig. 3C).
The MEFs from control and knockout genotypes were assayed for their ability to uptake BODIPY-labeled LDL particles. Compared with dab2 (+/) cells, Dab2-or Arhnull cells had significant reduced ability to uptake LDL particles (Fig. 4A). Remarkably, the Arh and Dab2 doubleknockout cells had little ability to uptake LDL, of a level similar to that of LDLR null (Fig. 4). We quantified LDL embryos by the E9.5 stage (24). Previously, we found that dab2 wild-type and heterozygous mice are essentially indistinguishable in all physiological parameters tested (24); thus, dab2 heterozygous siblings generated from the colonies were used as controls and substituted for true wild-type mice. Remarkably, Dab2 and Arh double-knockout mice were produced in the expected Mendelian ratio and were grossly normal. Western blot analysis of MEFs and liver tissues derived from the control and mutant mice confirmed the complete absence of Dab2 and Arh proteins in the compound mutant mice (Fig.  2B, C). Elimination of Dab2 or LDLR elicited a compensatory increase in Arh protein in MEFs (Fig. 2B) but not in liver tissues (Fig. 2C). The increased Arh in MEFs also contains a modified form showing an increased apparent molecular mass (Fig. 2B). It was uncertain if these variants are related to the posttranslational modifications such as S-nitrosylation of Arh (31). In contrast, arh deletion did not cause a compensatory expression of Dab2 or LDLR proteins in either MEFs or hepatic tissues (Fig. 2B, C). In MEFs, we also observed an increased LDLR mRNA in Arh-null cells and reduced VLDLR upon dab2 deletion (Fig. 2D). Thus, a transcriptional regulatory network between the receptors and endocytic adaptors exists that warrants future investigation.
The matching control, ldlr (/), dab2 (df/df) (or CKO), arh (/), and Dab2-Arh double-knockout mice were kept on a high-sucrose diet to compare the impact of Fig. 3. Serum lipid profiles of control and mutant mice fed a high-sucrose diet. Control and mutant mice (12 mice in each category) were maintained according to identical procedures and conditions. For cholesterol measurement, the mice were provided with a cholesterolfree, high-sucrose diet (No. 960237; MP Biomedical: 60.2% sucrose, 20% casein, 15% cellulose, 0.3% DL-methionine, 3.5% mineral and vitamins, and 0.2% fat) at 3 weeks of age and continued to 3 months of age. A: Body weights were measured at the end of the experiments for each animal, the values from each cohort (6 mice each) of genotype were averaged, and SEMs were calculated. B: Serum from each mouse from a panel of 10 for each genotype was collected for quantification of lipid content, including total serum cholesterol, LDL cholesterol, triglycerides, HDL cholesterol, and VLDL. The average values with SEM are presented. The differences between values from Dab2 or Arh individual knockout to those of double knockout were determined to be statistically significant by Student's t-test (P < 0.005). C: Liver tissues from individual mice from a panel of 10 for each genotype were collected for quantification of cholesterol content, and averages with SEM are shown. The differences between all the categories were determined to be statistically significant (P < 0.005). (Fig. 6A). Under higher magnification (Fig. 6A), the Dab2 signal appeared localized to most, but not all, endothelial cells in the liver. The liver tissue was LDLR positive throughout, with a slightly higher intensive along the sinusoidal endothelial edges (Fig. 6B). Additionally, we compared Dab2 with F4/80 staining, a marker of macrophages and Kupffer cells (Fig. 6C). The staining patterns were clearly dissimilar between Dab2 and F4/80 in the same section, and the two immunofluorescent signals were often seen adjacent but not colocalized (Fig. 6C), suggesting that Dab2 is not expressed in Kupffer cells. Thus, we conclude that Dab2 is expressed in the sinusoid endothelial cells but not in hepatocytes or Kupffer cells in the liver. This would suggest that Dab2 mediates a noncell autonomous activity to regulate hepatic HMGCR activity and cholesterol homeostasis. uptake by either determining the number of BODIPYpositive cells (Fig. 4B) or by measuring the total BODIPY signal in a defined field (Fig. 4C). Interestingly, the values obtained from either measurement were similar. We further determined that Dab2 was heterogeneously expressed in the cell population, and only 30% of the MEFs were Dab2 positive (Fig. 4D). Thus, cells positive for either Dab2 and/or Arh are sufficient to support LDLRmediated uptake of BODIPY-labeled LDL and were fluorescently labeled to a similar extent; only the cells negative for both Dab2 and Arh remained unlabeled. The observation suggests that the dosage of LDLR adaptor protein is not a limiting factor for LDL uptake, and the presence of either Dab2 or Arh is sufficient, yet the presence of both adaptors is not additive in the same cell. Thus, the combination of Dab2 and Arh accounts for the majority of adaptor activity responsible for LDLR endocytosis and LDL uptake by MEFs, by presenting individual cells with either Arh and/or Dab2 as the endocytic adaptor for LDLR.
Cholesterol in these mice was exclusively endogenously produced because the high-sucrose diet lacked exogenous cholesterol. Thus, the increased serum cholesterol would be either because of a reduced uptake or an augmented synthesis/secretion. We measured HMGCR, the ratelimiting enzyme for cholesterol synthesis in liver tissues (Fig. 5). Arh and Dab2 double-knockout liver tissues had a greatly increased HMGCR level, comparable to that of LDLR knockouts (Fig. 5).
We also examined the cell type in which Dab2 is present in the liver. Surprisingly, Dab2 is not expressed in hepatocytes. The pattern of Dab2 staining by immunofluorescence microscopy resembles that of PECAM-1, also noted as CD31, a marker of liver sinusoidal endothelial cells Fig. 4. LDL uptake by control and mutant MEFs. BODIPY-conjugated LDL particles at a concentration of 25 g/ml were incubated for 5 h with MEFs of control and mutant genotypes that were cultured in medium with 5% lipoprotein-depleted serum. The Dab2-deficient MEFs derived from Sox2-Cre-mediated deletion are denoted as dab2 (/) or (+/). A: BODIPY fluorescence images (green) of the cells were acquired using a Zeiss Z1 inverted microscope operated by AxioVision software. B: The fraction of BODIPY-positive cells of a defined area was quantified in triplicate, and the averages with SEM are shown. All categories had statistically significant differences (P < 0.005). C: The total BODIPY fluorescence signal in a defined field was quantified using Image J program in triplicate, and all categories had statistically significant differences. D: Wild-type MEFs were stained with antibodies to Dab2 (red) and Arh (green) for analysis using immunofluorescence microscopy. Fig. 5. Induction of liver HMGCR in Dab2 and Arh doubleknockout mice. Hepatic tissues were harvested from the mutant mice of 3 months of age. Total cell lysates were prepared and analyzed by Western blotting for the expression of HMGCR. The signals from the blot were quantified using Image J program and a relative value for HMGCR was estimated with normalization using -actin signal.

DISCUSSION
The experiments described in this study show that double knockout of the endocytic adaptors Arh and Dab2 mirrored the phenotypes of the LDLR knockout, in the similar elevated level of circulating and liver cholesterol and increased hepatic HMGCR. Thus, we conclude that LDLR endocytic adaptors Arh and Dab2 control the LDLRdependent regulation of the homeostasis of circulatory cholesterol, including both synthesis and removal (1). A model is proposed that the combined roles of Arh and Dab2 account for the adaptor function in LDLR endocytosis and cholesterol homeostasis in intact animals. We suggest that Arh is the main LDLR endocytic adaptor for clearance of circulating LDL in both liver and extrahepatic tissues. In the liver, Dab2 participates in LDLR-mediated At a greater magnification, the concentrated and punctate LDLR staining near the sinusoidal edge did not overlap with PECAM staining that marks the sinusoidal endothelial cells, as examples indicated by arrows (Fig. 7A). This is consistent with the notion that the majority of LDLR locates within hepatocytes rather than the endothelial cells. Similarly, Dab2 staining is distinct from the intense LDLR signals near the sinusoidal edge (Fig. 7B). However, partial overlapping of Dab2 and LDLR signals was observable (Fig.  7B, arrows). Thus, the majority of liver LDLR located within the hepatocytes is inaccessible to Dab2, which associates only with LDLR of the sinusoidal endothelial cells. Therefore, we conclude that Dab2-mediated LDLR endocytosis in the liver sinusoidal endothelial cells, which contain only a much lower level of LDLR, is able to regulate hepatic HMGCR expression and cholesterol synthesis.  6. Expression of Dab2 in the sinusoidal endothelial cells of the liver. Hepatic tissues were harvested from wild-type mice of about 3 months of age. The samples were prepared for cryosectioning, and the sections were placed on glass slides for immunofluorescence analysis using confocal microscopy. A: Liver sections were stained for Dab2 (red), PECAM (green), and DAPI (blue). An image at higher magnification is also shown (right panel). B: Liver sections were stained for LDLR (red), PECAM (green), and DAPI (blue). C: Liver sections were stained for Dab2 (red), F4/80 (green), and DAPI (blue). cholesterol but not hepatic HMGCR, Arh-mediated, but not Dab2-mediated, LDLR endocytosis is the main activity in the removal of circulating cholesterol. Hepatic cholesterol synthesis is only activated either in the absence of LDLR or by deletion of both Arh and Dab2.
The results from the double-knockout mice suggest that Dab2 plays a partially redundant or supplementary role to Arh in mediating LDLR endocytosis and is likely more relevant in regulating hepatic cholesterol synthesis through its function in sinusoid endothelia. Nevertheless, we did observe a significant increase in liver cholesterol content in Dab2 knockout animals, suggesting that Dab2 does impact LDL uptake through the liver sinusoidal endothelial cells and cholesterol synthesis despite the presence of Arh. The absence of both Dab2 and Arh in liver leads to greatly increased liver cholesterol to the level comparable to those of LDLR knockout, likely as a result of unsuppressed synthesis. Thus, the combined roles of Arh and Dab2 account for the majority of the requirement for LDLR endocytosis and clearance of circulating LDL in intact animals. Although extensive studies on the roles of Dab2 and Arh in LDLR endocytosis have been performed using cultured cells, the current study determined their roles in intact animals, particularly their impact on circulating cholesterol. This initial study limited the investigation to animals placed on a highsucrose diet in the absence of dietary cholesterol. The mechanism would be expected to be much more intricate due to additional factors when the mice are on a diet of high fat and exogenous cholesterol (19)(20)(21).
This study brought new information to light regarding the roles of LDLR cellular endocytic trafficking in a whole organism and advanced the understanding of this complex but important issue of cholesterol metabolism.
Several key genes including ARH, PCSK9, and Dab2 affect serum cholesterol levels by modulating the endocytosis and cellular trafficking of LDLR. The increasing LDL uptake only through the sinusoid endothelial cells, and the derived cholesterol resides in a cellular pool that is able to regulate cholesterol synthesis in hepatocytes. We conclude that the concerted roles of Arh and Dab2 are responsible for the majority of adaptor function for LDLR endocytosis and LDLR-mediated cholesterol clearance from the circulation, as well as the regulation of cholesterol synthesis through controlling HMGCR expression.
One unexpected finding is that Dab2 is expressed only in sinusoid endothelial cells but not in hepatocytes or Kupffer cells. However, the majority of liver LDLR locates within the hepatocytes, and endothelial cells likely only contribute to a very minor level of receptor-mediated LDL uptake. In the absence of Arh, LDLR-mediated LDL uptake in liver presumably occurs only in the sinusoid endothelia supported by Dab2 but not in hepatocytes, and only at a very low level. Because Arh knockout mice have low hepatic HMGCR protein, cholesterol derived from Dab2and LDLR-mediated uptake of LDL in the sinusoid endothelial cells must be able to suppress hepatic HMGCR. Despite an increased liver cholesterol level in either LDLRnull or Arh/Dab2 double-knockout mice, we observed that liver HMGCR expression was greatly induced, to a similar extent in both mutant mice. This is consistent with the finding that cellular cholesterol is compartmentalized, and only a certain pool of cholesterol is able to regulate HMGCR (32). Therefore, the endothelial cholesterol derived from LDL uptake likely is exchangeable with the hepatic pool of cholesterol that is capable in the feedback regulation of its de novo synthesis.
Dab2 gene ablation alone in mice has a relatively minor impact on serum cholesterol, and the results suggest that Arh is the main LDLR endocytic adaptor involved in the clearance of circulating LDL in both liver and extrahepatic tissues and in the regulation of HMGCR for cholesterol synthesis. Because Arh knockout greatly affects serum Fig. 7. Colocalization of Dab2 and LDLR in the sinusoidal endothelial cells of the liver. Cryosections of liver tissues from wild-type mice were analyzed by confocal immunofluorescence microscopy. A: Liver sections were stained for LDLR (red), PECAM (green), and DAPI (blue). B: Liver sections were stained for LDLR (red), Dab2 (green), and DAPI (blue). Enlarged images with overlaid signals are shown on top.
understanding of genes and cellular mechanisms impacting circulating cholesterol likely will be helpful to meet the challenge of the development of new drugs, biology-based practice of medicine, and formulation of guidelines for management of dyslipidemias and strategies to reduce cardiovascular diseases (3,17,33).