Expression of the VLDL Receptor in Endothelial Cells
Abstract In this article we describe the cellular distribution of the very low density lipoprotein receptor (VLDLR), a transmembrane protein that is expressed at high concentrations in skeletal muscle, heart, adipose tissue, and brain but in only trace amounts in the liver. Indirect immunofluorescence localization studies were performed in murine and bovine tissues using a rabbit polyclonal anti-human VLDLR antibody. Immunoreactive VLDLR protein was detected in the endothelium of capillaries and small arterioles but not in veins or venules of bovine skeletal muscle, heart, ovary, and brain. In the liver, there was intense staining of the capillaries and arterioles that supply the capsule and hepatic vessels but no staining of the sinusoidal surfaces. We failed to detect any signal from nonendothelial cells in the liver or peripheral organs. The VLDLR was also expressed at high levels on the endothelial surface of bovine coronary arteries; in contrast, little or no staining was seen in aortic endothelium. Antibody staining of cultured bovine coronary artery endothelial cells demonstrated punctate cell-surface staining, as well as staining of large and small cytoplasmic vesicles. This tissue and cell pattern of expression suggests that the VLDLR plays a role in the transport of VLDL or another plasma constituent from the vascular compartment to adjacent tissues.
- Received October 12, 1995.
- Accepted December 7, 1995.
In 1992 Yamamoto and colleagues (Takahashi et al1 ) cloned a cDNA from rabbit heart that encoded a protein strikingly similar in sequence and structure to the LDLR. Expression of this cDNA in cultured cells produced a protein that bound rabbit VLDL, but not LDL, with high affinity.1 On the basis of these ligand-binding characteristics, this newly discovered member of the LDLR gene family was named the VLDLR. Characterization of the human VLDLR gene revealed that its structure, including the locations of intron-exon junctions, was almost identical to that of the human LDLR gene.2 Both genes encode a protein with six domains: (1) a signal sequence followed by (2) a ligand-binding (3) EGF precursor homology domain, (4) O-linked sugar, (5) transmembrane, and (6) cytoplasmic domains. Interestingly, human and rabbit VLDLR cDNAs share a higher degree of sequence identity than do human and rabbit LDLR cDNAs (96% versus 76%).2
The structural feature that most clearly distinguishes the VLDLR from the LDLR is an additional ligand-binding repeat.1 The ligand-binding domain of the LDLR comprises seven ≈40–amino acid Cys-rich repeats, each of which contains the highly conserved amino acid triplet Ser-Asp-Glu.3 The VLDLR has eight rather than seven ligand-binding repeats, and when the recombinant protein is expressed at high levels in cultured cells, the receptor binds rabbit VLDL and β-VLDL but not LDL or human VLDL with high affinity.1 4 The VLDLR also binds and internalizes a 39-kD protein, which has been referred to as RAP,5 and LPL-enriched β-VLDL4 in cultured cells. These two ligands also bind LRP, another LDLR gene family member.6 7 LRP, like the VLDLR, also binds β-VLDL; however, unlike the VLDLR, LRP binds β-VLDL only when the latter particle has been supplemented with exogenous apoE.1 8
VLDL is a TG-rich lipoprotein that is synthesized in the liver and secreted into the bloodstream. The TGs of VLDL are hydrolyzed by LPL, which resides on the endothelial surfaces of capillaries, and the resulting free fatty acids are delivered to adjacent tissues. Yamamoto and colleagues9 have proposed that the VLDLR mediates the uptake of VLDL particles by cells in peripheral tissues, and the tissue distribution of expressed VLDLR is consistent with this scenario. The highest levels of the VLDLR occur in muscle, heart, adipose tissue, and brain, all of which utilize lipoprotein-derived free fatty acids as an energy source.1 However, for the VLDLR to play an important role in the delivery of TGs to adipocytes, myocytes, or other cells in peripheral tissues, it must be expressed on the capillary endothelium because the capillary endothelium in these tissues is continuous and thus impermeable to particles as large as VLDL.10
The cell-specific pattern of VLDLR expression in tissues has not been determined owing to the lack of antibodies that recognize the native VLDLR. Development of such antibodies has been hampered by the very high degree of sequence conservation of the VLDLR between species. For example, there is 94% sequence identity between the proteins encoded by rat and human VLDLR cDNAs.11 In addition, the human VLDLR shares 84% of its sequence with the chicken vitellogenin receptor,12 which also binds VLDL and is required for the delivery of nutrients to the developing avian oocyte. The importance of this receptor was revealed by identification of a naturally occurring, sex-linked mutation in the VLDLR gene of chickens: hens that lack the VLDLR are hyperlipidemic and do not lay eggs.13
We now report the development of a rabbit anti-human VLDLR antibody that reacts specifically with the murine and bovine VLDLRs. This antibody was used to analyze the distribution of immunoreactive VLDLR in bovine liver, muscle, brain, ovary, coronary arteries, and aorta, as well as in cultured BCAECs. We found that the VLDLR is highly expressed on the endothelium of capillaries, small arterioles, and coronary arteries but is present at low levels, if at all, in peripheral cells. These findings are consistent with the proposed role for the VLDLR in mediating the delivery of VLDL-derived lipids to peripheral tissues.
A 4.0-kb BamHI fragment from human VLDLR cDNA (designated pHKY and obtained from Dr Tokuo Yamamoto, Tohoku University, Sendai, Japan) was subcloned downstream of the cytomegalovirus promoter in expression vector pCB6 (provided by Dr David Russell, University of Texas Southwestern Medical Center [UTSWMC], Dallas).14 A stable cell line expressing the human VLDLR cDNA (TR-2037) was developed by transforming Chinese n>amster ovary cells that lacked LDLR activity (ldlA7 cells provided by Dr Monty Kreiger, Massachusetts Institute of Technology, Boston) with pCB6-VLDLR.15 The BCAECs were provided by Dr Phillip Thorpe (Department of Pharmacology, UTSWMC, Dallas).
β-VLDL was purified from cholesterol-fed NZW rabbits by density gradient ultracentrifugation16 and was labeled with DiI as previously described.17 Mice in which the VLDLR had been inactivated by homologous recombination and that had no immunodetectable VLDLR protein (VLDLR−/−) were obtained from Dr Joachim Herz (UTSWMC, Dallas).18
Preparation of Rabbit Anti-Human VLDLR Antibody
A 2281-bp fragment that included the sequences contained in exons 1 through 15 of the human VLDLR was amplified from pHKY by the PCR and two oppositely oriented oligonucleotides: VLDLH-5 (5′-GTTGTGGTTGGCGGGATCCTATAAATATGCCGACGTCCGCGCTCTGGG-3′) and VLDLH-6 (5′-TCTGCCATCGAATTCCTAGTGGTGGTGGTGGTGGTGACTTTGACAGTCTCGGCCATTTTC-3′).19 VLDLH-6 encodes for six His at its 3′ end. The PCR product was digested with EcoRI and BamHI, purified with the Promega Wizard DNA purification system, and subcloned into baculovirus expression vector pBacPAK8 (Clontech) to form pBacPAK8-hVLDLR. The plasmid was cotransfected with linearized Bsu36I-digested BacPAK6 viral DNA (Clontech) into Spodoptera frugiperda (Sf 9) cells by lipofection.20 Positive clones were isolated by plaque assay exactly as described.20 Single plaques were picked and recombinant virus was isolated after three rounds of amplification and reinfection. High Five insect cells (Invitrogen) were infected with recombinant baculovirus at a multiplicity of infection of 8 to 10. The media were collected and subjected to a 75% (wt/vol) (NH4)2SO4 precipitation. The precipitate was resuspended in buffer A (0.5 mol/L NaCl, 20 mmol/L Tris HCl, pH 8.0), applied to an Ni2+-Sepharose affinity column, and eluted with a linear gradient of imidazole (25 to 400 mmol/L) in buffer A. A protein peak at 100 mmol/L imidazole was collected, dialyzed against buffer B (5 mmol/L Na2EDTA, 0.1 mmol/L NP-40, 20 mmol/L Tris HCl, pH 8.0), concentrated using Centriprep 30 concentrators (Amicon), and applied to a MonoQ column (Pharmacia). The proteins were eluted with a linear NaCl gradient (0 to 1 mol/L) in buffer B. A 77-kD fragment that ligand-blotted with a 125I-radiolabeled fusion protein that contained glutathione S-transferase and the entire coding sequence of rat RAP6 eluted from the column at 0.4 mol/L NaCl. This protein was >90% pure, as estimated by size fractionation on an SDS-polyacrylamide gel and staining with Coomassie Brilliant Blue R (data not shown). Initially a total of 300 μg and subsequently 150 μg of the recombinant protein were injected into two NZW rabbits every 4 weeks for a total of 5 months. The IgG from 5 mL rabbit plasma was purified on Protein A–Sepharose and the antibody was named IgG-M404.
Immunoblot Analysis of VLDLR in Cultured Cells and Bovine Tissue
TR-2037 cells were harvested in ice-cold PBS. The cell pellet was lysed in 50 mmol/L Tris HCl, 2 mmol/L CaCl2, 80 mmol/L NaCl, 1% (vol/vol) Triton X-100, 1 mmol/L PMSF, 0.5 mmol/L leupeptin, 20 μg/mL aprotinin, and 5 μg/mL pepstatin A, pH 8.0. The crude lysate was subjected to centrifugation at 3000g at 4°C for 10 minutes. The supernatant was collected and the protein concentration determined by Coomassie Plus protein assay reagent (Pierce). Aliquots of cell lysate were analyzed by immunoblotting as described below.
Extracts of bovine tissue and BCAECs were prepared, and immunoblotting with IgG-6A6 was performed exactly as described.11 The same protocol was used for immunodetection with IgG-M404, except that the washes were performed with 1% (vol/vol) NP-40 (Calbiochem), 0.5% (wt/vol) deoxycholic acid, and 0.05% (wt/vol) SDS in Tris-buffered saline, pH 7.4. The blots were developed with chemiluminescent agents (ECL, Amersham) and subjected to autoradiography.
Immunolocalization of the VLDLR in Murine and Bovine Tissues
C57BL/6J and 129Sv hybrid mice (weight, ≈25 g) were anesthetized with 0.4 mg pentobarbital sodium (Abbott Laboratories) and perfused via the left ventricle with PBS followed by fixative A (3% [wt/vol] p-formaldehyde, 3 mmol/L trinitrophenol, 5 mmol/L MgCl2 in 50 mmol/L Sorenson’s phosphate buffer, pH 7.4). The tissues were subsequently removed and refixed for 24 hours in fixative A. Unperfused bovine tissues were diced into small pieces and washed several times in PBS to remove blood; the tissues were then fixed for 24 hours in fixative A. For cryosectioning, aldehyde-fixed murine tissues were rinsed three times with PBS before cryoprotection by sequential incubation for 24 hours each with 15% (wt/vol) sucrose, 30% (wt/vol) sucrose, and 30% (wt/vol) sucrose in 20% (vol/vol) OCT (Miles), all diluted in PBS. Finally, the tissues were embedded in OCT and frozen in liquid N2–chilled (−100°C) isopentane. Cryoembedded tissues were either stored at −80°C or cut on a Leica cryomicrotome. Sections (10 to 12 μm thick) were mounted on ProbeOn Plus (Fisher Scientific) slides and processed for indirect immunofluorescence as described below. Bovine tissues were processed for conventional paraffin embedding and sectioning. Deparaffinized sections were either stained with hematoxylin and eosin or immunostained using the same protocol as described.
Localization of the VLDLR in Cultured Cells
Immunolocalization studies were performed in ldlA7 cells, TR-2037 cells, and primary cultured BCAECs. BCAECs were harvested at passages 4 through 7 and grown to ≈80% confluence on glass coverslips in DMEM (GIBCO BRL) containing 15% (vol/vol) fetal calf serum (GIBCO BRL), 18 mmol/L HEPES, 0.1 mmol/L nonessential amino acids (Sigma Chemical Co), 1 mmol/L sodium pyruvate, 2 mmol/L Glu, 100 U/mL penicillin, and 100 μg/mL streptomycin (Bio-Whittaker) in an 8% to 9% CO2 incubator.
All three cell lines were grown on glass coverslips, washed in PBS, fixed in fixative B (3% [wt/vol] p-formaldehyde in 3 mmol/L MgCl2, 3 mmol/L KCl, 3 mmol/L trinitrophenol, and 37.5 mmol/L Sorenson’s phosphate buffer, pH 7.8) for 1 hour, and processed for indirect immunofluorescence as described.21 The cells were either left nonpermeabilized or permeabilized by incubation for 2 minutes in PBS containing 0.1% (vol/vol) Triton X-100 before incubation for 30 minutes in buffer C (0.2 mol/L NaCl, 1% [wt/vol] BSA, 0.01% [wt/vol] NaN3, and 20 mmol/L Tris HCl, pH 9.0) to block nonspecific sites. The cells were then incubated overnight with either preimmune IgG (100 μg/mL); IgG-M404 (100 μg/mL); 2EI, a mouse anti-rat LRP monoclonal antibody (100 μg/mL)6 ; S713, a rabbit anti-bovine LDLR polyclonal antibody (100 μg/mL)22 ; or a rabbit anti-human von Willebrand factor polyclonal antibody (1:100 dilution; Dako).23 The coverslips were rinsed twice with buffer D (0.2 mol/L NaCl, 0.1% [wt/vol] BSA, 0.01% [wt/vol] NaN3, 0.01% [vol/vol] Triton X-100, 20 mmol/L Tris HCl, pH 9.0) for 10 minutes and then incubated for 2 hours with 25 μg/mL of either a goat anti-rabbit or goat anti-mouse IgG conjugated to FITC (Zymed). The cells were washed for 10 minutes with buffer D, rinsed in distilled water, and dried before they were mounted in 1,4-diazobicyclo[2.2.2.]octane. The slides were photographed using a Zeiss Photomicroscope equipped with a filter package for fluorescein.
DiI-Labeled β-VLDL Binding and Uptake
ldlA7 or TR-2037 cells were washed with DMEM in 25 mmol/L HEPES (pH 7.4) that had been supplemented with 1% (wt/vol) BSA and chilled to 4°C for 30 minutes. Cells were incubated with 30 μg/mL of DiI-labeled β-VLDL in HEPES-buffered DMEM for 60 minutes on ice and then washed in the same buffer. The cells were either fixed in fixative B or warmed to 37°C for 1 hour to allow internalization of receptor-bound β-VLDL before washing and fixation. After fixation, the cells were washed in PBS and mounted on glass slides as described above.
The extracellular domain of the human VLDLR, including the ligand-binding and EGF-precursor homology domains, was expressed as a recombinant protein in a baculovirus expression system as described in “Methods.” The resultant 77-kD protein was used to immunize NZW rabbits. Purified serum IgG from an immunized rabbit (IgG-M404) was used for immunoblot analysis of membranes from various murine tissues, and the tissue-specific detection pattern closely resembled that observed with a mouse monoclonal anti-VLDLR antibody, as previously reported.18
To confirm the specificity of the antibody and examine the cellular distribution of bound and internalized DiI-labeled β-VLDL, immunofluorescence localization and DiI-labeled β-VLDL uptake studies were performed in nontransfected ldlA7 cells, which do not express VLDLR, and in TR-2037 cells, which express large amounts of recombinant human VLDLR (Fig 1⇓). After permeabilization and incubation with IgG-M404 there was no staining of the ldlA7 cells (Fig 1A⇓), but an intense fluorescence signal was evident on the surface and in some internal compartments of the TR-2037 cells (Fig 1B⇓, 1C⇓, and 1E⇓). In nonpermeabilized TR-2037 cells, staining was restricted to the cell surface (Fig 1C⇓). A similar cell-surface staining pattern appeared when the TR-2037 cells were incubated with DiI-labeled β-VLDL at 4°C (Fig 1D⇓); no staining was seen with the ldlA7 cells under the same conditions, however (data not shown). When permeabilized TR-2037 cells were incubated with IgG-M404, large amounts of VLDLR staining were seen in some intracellular vesicles (Fig 1E⇓). A similar pattern of intracellular staining was also found in cells that had been incubated with DiI-labeled β-VLDL at 4°C and then warmed to 37°C to permit internalization of the bound ligand (Fig 1F⇓).
To examine the cellular distribution of the VLDLR and further confirm the specificity of the antibody, immunofluorescence localization studies were performed in skeletal muscle tissue of wt C57BL/6J-129Sv hybrid mice and mice of the same genetic background in which the VLDLR gene had been inactivated by homologous recombination (VLDLR−/−) (Fig 2⇓).18 Skeletal muscle was used because it had high levels of immunodetectable VLDLR by immunoblotting.18 When IgG-M404 was applied to skeletal muscle sections, a punctate staining pattern was seen at the periphery of the transversely cut fiber bundles of the wt (Fig 2A⇓) but not the VLDLR−/− (Fig 2B⇓) mice. The regions in wt mouse muscle that were stained (arrows, Fig 2A⇓) corresponded to the locations of capillaries or small arterioles, as determined by examination of the same sections either under phase-contrast (arrows, Fig 2C⇓) or after staining of an adjacent section with hematoxylin and eosin (data not shown). The absence of immunoreactivity from capillaries or arterioles in skeletal muscle sections of VLDLR−/− mice (Fig 2B⇓ and 2D⇓) confirms that IgG-M404 was specific for VLDLR protein because this antibody immunoreacted with the receptor in wt but not VLDLR−/− mice. The preimmune antibody did not stain frozen sections of muscle from either wt or VLDLR−/− mice (data not shown).
The IgG-M404 used in these studies was raised against the human VLDLR. Although IgG-M404 reacted with the mouse VLDLR, this reactivity was relatively weak; however, it showed much stronger reactivity with the bovine VLDLR. Fig 3⇓ shows an immunoblot analysis of membranes from various bovine tissues, which compares IgG-M404 and IgG-6A6, a monoclonal antibody to the cytoplasmic tail of the VLDLR, which has been shown by immunoblotting to recognize the mouse and rat VLDLR.11 18 An identical band pattern was seen with both antibodies in muscle, heart, and cultured BCAECs. There were no detectable bands in liver tissue with either antibody. In our previous studies, we have shown that the sizes of the immunoreactive bands differ in the different tissue types due to differential splicing of the VLDLR gene2 and the resultant differences in protein glycosylation.11
Next we examined the distribution of the VLDLR by indirect immunofluorescence in bovine tissues (Fig 4⇓), including muscle (Fig 4A⇓ and 4B⇓), heart (Fig 4C⇓), brain (Fig 4D⇓), liver (Fig 4E⇓), ovary (Fig 4F⇓), aorta (Fig 4G⇓), and coronary artery (Fig 4H⇓). We observed intense fluorescence in the capillaries or small arterioles in skeletal muscle after staining the sections with IgG-M404 (Fig 4B⇓) but not with preimmune IgG (Fig 4A⇓). The capillaries or small arterioles also stained intensely in the heart and brain (Fig 4C⇓ and 4D⇓). In contrast, liver tissue showed only a small amount of localized staining in the capillaries or arterioles that were associated with hepatic arterial vessels and the capsule (data not shown). The hepatic arterioles or capillaries (arrows, Fig 4E⇓) adjacent to the portal vein (asterisk, Fig 4E⇓) also were stained. Hepatocytes, Kupffer cells, and the endothelium of sinusoidal surfaces, central veins, portal veins, biliary duct, and hepatic veins did not stain with the anti-VLDLR antibody. In the ovary, there was staining of arterioles (arrows, Fig 4F⇓) that emanated from small arteries (asterisk, Fig 4F⇓). The connective tissue cells of the ovarian cortex, stroma, primary follicles, immature ovum, and follicular cells did not stain with IgG-M404 (data not shown). No staining was associated with the aortic endothelium (Fig 4G⇓) from the descending aorta, but there was intense staining of the endothelial surface of a distal coronary artery (Fig 4H⇓). Sections from the aortic root showed trace amounts of immunoreactive staining on the endothelial surface. We were unable to determine the distribution of VLDLR expression in adipose tissue because of difficulties in preparation and intense autofluorescence in this tissue.
Cultured BCAECs were permeabilized and stained with a variety of antibodies (Fig 5⇓). The anti-VLDLR antibody (IgG-M404) stained a large number of cells and produced a punctate pattern of fluorescence (Fig 5B⇓). When the cells were examined under high magnification, at least two distinct staining patterns were recognized: large, intracellular vacuoles (Fig 5C⇓) and linear arrays of punctate dots that appeared to be on the cell surface (Fig 5D⇓). When IgG-M404 was applied to nonpermeabilized BCAECs, only the linear arrays of punctate dots could be seen (data not shown). The anti-LDLR antibody (S713) also revealed mostly punctate staining (Fig 5E⇓) in a pattern similar to that in cultured human skin fibroblasts.21 No staining was visible when an anti-LRP–specific antibody was used (data not shown), consistent with the fact that LRP is not expressed in endothelial cells.24 An antibody to von Willebrand factor stained ≈30% of the cells and revealed small, fluorescently labeled vesicles in a distribution consistent with that of Weibel-Palade bodies25 (Fig 5F⇓). No staining was seen when nonpermeabilized cells were stained with this antibody (data not shown). A control IgG from a nonimmunized rabbit did not stain any cellular structures in these cells (Fig 5A⇓).
Previously it was proposed that the VLDLR mediated the delivery of VLDL TGs to peripheral tissues9 because it bound VLDL with high affinity in cultured cells and was expressed at high concentrations in tissues that take up VLDL-derived lipids, including muscle, heart, fat, and brain.1 11 26 The problem with this scenario is that circulating VLDL does not have free access to the cells in these tissues, because it is too large a particle to pass through the nonfenestrated endothelium. This apparent paradox is explained by our current finding that the VLDLR is present at high levels in the endothelium of capillaries and small arterioles in these tissues but is present at much lower levels, if at all, on parenchymal cells. Thus, the VLDLR is in a position to interact directly with VLDL particles that are circulating in the blood.
Another clue to the function of the VLDLR came from analysis of the type of endothelium in which it was expressed. We did not find the VLDLR on the sinusoidal endothelium of the liver or the fenestrated capillaries of the intestines and kidneys (data not shown), consistent with prior studies that revealed very low levels of VLDLR mRNA and protein in these tissues.1 11 26 Because fenestrated capillary endothelium allows free access of macromolecules to adjacent tissues,27 there would be no need for the VLDLR in this type of endothelium, as VLDL could freely diffuse through the fenestrations into the tissue. The VLDLR is preferentially expressed on nonfenestrated capillaries, where the endothelial cells are tightly interconnected and VLDL particles in the bloodstream have no direct contact with peripheral cells. This class of endothelium, the so-called “continuous endothelium,” provides a selective barrier between the blood in the vascular compartment and the interstitial fluid in tissues.28 Furthermore, we did not find any immunoreactive VLDLR on the endothelial lining of the venules. Taken together, these findings suggest that the VLDLR plays a role in the delivery of VLDL or another blood-borne nutrient into tissues rather than vice versa.
Immunolocalization studies in cultured BCAECs revealed VLDLR expression on the cell surface and in intracellular vacuoles/vesicles, which would also be anticipated if the VLDLR participated in the transport of VLDL or other plasma constituents into endothelial cells and/or peripheral tissues. There are two major pathways by which plasma constituents are transported through a continuous endothelium into tissues. One pathway involves plasmalemmal vesicles, which may fuse to form channels.28 These vesicles do not usually deliver their contents to lysosomes but rather transport them to the underlying interstitial space by transcytosis.28 The other pathway uses specific receptors in clathrin-coated pits that take up plasma proteins into cells by receptor-mediated endocytosis.16 The cytoplasmic domain of the VLDLR contains the sequence NPXY, which is required for localization of the LDLR and other endocytotic receptors to the coated pits.29 Electron microscopy studies are now being performed to determine whether the VLDLR resides in the coated pits and/or plasmalemmal vesicles in BCAECs.
If the VLDLR is found associated with plasmalemmal vesicles, it would strongly suggest that this receptor plays a role in the transport of VLDL or another plasma constituent into peripheral tissues. VLDLR−/− mice have a normal lipoprotein profile, even after ingesting a high-fat or high-carbohydrate diet, which suggests that the VLDLR is not essential for lipolysis of TGs in murine peripheral tissue.18 This observation does not rule out a role for the receptor in VLDL metabolism. The receptor may increase the efficiency of lipolysis of TG-rich lipoproteins by “tethering” the particles to endothelial surfaces and optimizing their interaction with LPL. Alternatively, the VLDLR may bind LPL and participate in its transport from tissues to the endothelial surface or facilitate delivery of a component of VLDL other than TGs to tissues.
It is also possible that VLDL is not the physiological ligand for the receptor. There is precedent for nonlipoprotein ligand binding to members of the LDLR family. For example, LRP, another LDLR family member, has multiple ligands that are unrelated to lipoprotein metabolism, including α-2 macroglobulin, numerous protease-antiprotease inhibitor complexes, Pseudomonas aeruginosa exotoxin A, as well as others.30 In chickens the VLDLR mediates uptake of riboflavin-binding protein, vitellogenin, and lactoferrin, as well as VLDL into the developing oocyte.12 It is unlikely that the VLDLR plays a parallel role in nutrient delivery to the developing mammalian fetus, despite its presence on placental trophoblasts,31 because female VLDLR−/− mice have normal fertility and their offspring are the same size as control mice.18
In support of a role for the VLDLR in the delivery of some blood-borne nutrient(s) to peripheral tissues is the finding that VLDLR−/− mice have a 15% to 20% lower body mass index, which is due to a lower adipose tissue mass.18 The VLDLR may mediate delivery of excess calories to peripheral tissues, especially adipose tissue, for storage. We have previously shown that in rats the VLDLR level is not regulated acutely by the nutritional status of the animal.11 The VLDLR may be rate limiting in the clearance of TG-rich particles only in the chronically fed state and thus is physiologically unimportant unless the mice are challenged with a high-fat, hypercaloric diet for prolonged periods. Alternatively, the VLDLR may become physiologically important only during periods of caloric restriction to ensure maintenance of efficient delivery of nutrients to tissues. The reason for lack of a more dramatic phenotype in VLDLR−/− mice may be that the nutrient whose transport is facilitated by the VLDLR may be provided in vast excess in the diet fed to laboratory mice.
Finally, we cannot exclude the possibility that the VLDLR is expressed in other cell types at a concentration below the detection limits of the immunofluorescence methods that were used in these studies. In all tissues we have studied thus far (excluding the bovine coronary artery), immunoreactive VLDLR was seen only in association with the endothelial cells of selected capillaries and small arteries.
The presence of high concentrations of the VLDLR on the surface of bovine coronary arteries suggests that plasma lipoproteins may be transported into the arterial wall even if the endothelium is intact.32 Recent isolation and characterization of apoB-containing lipoproteins from human atherosclerotic lesions have revealed that one third of apoB-associated cholesterol occurs either in the VLDL or IDL fraction.33 Interestingly, the apoB associated with these particles was only modestly fragmented, suggesting that the particles entered the lesions while still intact. The VLDLR may participate in the delivery of VLDL to the arterial wall and contribute to the development of atherosclerosis, and the availability of VLDLR−/− mice will allow us to directly test whether this receptor contributes to the development of vascular lesions.
Selected Abbreviations and Acronyms
|BCAEC(s)||=||bovine coronary artery endothelial cell(s)|
|DMEM||=||Dulbecco’s modified Eagle’s medium|
|EGF||=||epidermal growth factor|
|LRP||=||LDL receptor–related protein|
|NZW||=||New Zealand White|
This work was supported by National Institutes of Health grant HL20948 (Bethesda, Md) and the Perot Family Fund. Dr Wyne is supported by NIH 2 T32 DK07307-17. We thank Melissa Christiansen, Richard Gibson, Lisa Beatty, Laura Deane, and Sadeq Kharzai for their excellent technical assistance and Tommy Hyatt and Philip Frykman with assistance in the animal studies. We thank Kathy Landschulz, Richard Anderson, Michael Brown, and Joseph Goldstein for helpful discussions.
Takahashi S, Kawarabayasi Y, Nakai T, Sakai J, Yamamoto T. Rabbit very low density lipoprotein receptor: a low density lipoprotein receptor-like protein with distinct ligand specificity. Proc Natl Acad Sci U S A. 1992;89:9252-9256.
Sakai J, Hoshino A, Takahashi S, Miura Y, Ishii H, Suzuki H, Kawarabayasi Y, Yamamoto T. Structure, chromosome location, and expression of the human very low density lipoprotein receptor gene. J Biol Chem. 1994;269:2173-2182.
Südhof TC, Goldstein JL, Brown MS, Russell DW. The LDL receptor gene: a mosaic of exons shared with different proteins. Science. 1985;228:815-822.
Takahashi S, Suzuki J, Kohno M, Oida K, Tamai T, Miyabo S, Yamamoto T, Nakai T. Enhancement of the binding of triglyceride-rich lipoproteins to the very low density lipoprotein receptor by apolipoprotein E and lipoprotein lipase. J Biol Chem. 1995;270:15747-15754.
Battey FD, Gåfvels ME, Fitzgerald DJ, Argraves WS, Chappell DA, Strauss JF III, Strickland DK. The 39-kDa receptor-associated protein regulates ligand binding by the very low density lipoprotein receptor. J Biol Chem. 1994;269:23268-23273.
Herz J, Goldstein JL, Strickland DK, Ho YK, Brown MS. 39-kDa protein modulates binding of ligands to low density lipoprotein receptor-related protein/α2-macroglobulin receptor. J Biol Chem. 1991;266:21232-21238.
Beisiegel U, Weber W, Bengtsson-Olivecrona G. Lipoprotein lipase enhances the binding of chylomicrons to low density lipoprotein receptor-related protein. Proc Natl Acad Sci U S A. 1991;88:8342-8346.
Kowal RC, Herz J, Goldstein JL, Esser V, Brown MS. Low density lipoprotein receptor-related protein mediates uptake of cholesteryl esters derived from apoprotein E-enriched lipoproteins. Proc Natl Acad Sci U S A. 1989;86:5810-5814.
Yamamoto T, Takahashi S, Sakai J, Kawarabayasi Y. The very low density lipoprotein receptor. Trends Cardiovasc Med. 1993;3:144-148.
Simionescu N. Cellular aspects of transcapillary exchange. Physiol Rev. 1983;63:1536-1579.
Jokinen EV, Landschulz KT, Wyne KL, Ho YK, Frykman PK, Hobbs HH. Regulation of the very low density lipoprotein receptor by thyroid hormone in rat skeletal muscle. J Biol Chem. 1994;269:26411-26418.
Nimpf J, Radosavljevic MJ, Schneider WJ. Oocytes from the mutant restricted ovulator hen lack receptor for very low density lipoprotein. J Biol Chem. 1989;264:1393-1398.
Brewer CB. Cytomegalovirus plasmid vectors for permanent lines of polarized epithelial cells. Methods Cell Biol. 1994;43:233-245.
Kingsley DM, Krieger M. Receptor-mediated endocytosis of low density lipoprotein: somatic cell mutants define multiple genes required for expression of surface-receptor activity. Proc Natl Acad Sci U S A. 1984;81:5454-5458.
Pitas RE, Innerarity TL, Weinstein JN, Mahley RW. Acetoacetylated lipoproteins used to distinguish fibroblasts from macrophages in vitro by fluorescence microscopy. Arteriosclerosis. 1981;1:177-185.
Frykman PK, Brown MS, Yamamoto T, Goldstein JL, Herz J. Normal plasma lipoproteins and fertility in gene-targeted mice homozygous for a disruption in the gene encoding very low density lipoprotein receptor. Proc Natl Acad Sci U S A. 1995;92:8453-8457.
Saiki RK, Scharf S, Faloona F, Mullis KB, Horn GT, Erlich HA, Arnheim N. Enzymatic amplification of β-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science. 1985;230:1350-1354.
Cremers FPM, Armstrong SA, Seabra MC, Brown MS, Goldstein JL. REP-2, a rab escort protein encoded by the choroideremia-like gene. J Biol Chem. 1994;269:2111-2117.
Pathak RK, Anderson RGW. Use of dinitrophenol IgG conjugates: immunogold labeling of cellular antigens on thin sections of osmicated and epon-embedded specimens. In: Hayat MA, ed. Colloidal Gold: Principles, Methods, and Applications. San Diego, Calif: Academic Press; 1991;3:223-241.
Sehested M, Hou-Jensen K. Factor VIII related antigen as an endothelial cell marker in benign and malignant diseases. Virchows Arch (Pathol Anat). 1981;391:217-225.
Wagner DD, Olmsted JB, Marder VJ. Immunolocalization of von Willebrand protein in Weibel-Palade bodies of human endothelial cells. J Cell Biol. 1982;95:355-360.
Webb JC, Patel DD, Jones MD, Knight BL, Soutar AK. Characterization and tissue-specific expression of the human “very low density lipoprotein (VLDL) receptor” mRNA. Hum Mol Genet. 1994;3:531-537.
Fawcett DW. Blood and lymph vascular systems. In: Fawcett DW, ed. Bloom and Fawcett: A Textbook of Histology. Philadelphia, Pa: WB Saunders Co; 1986:367-391.
Bruns RR, Palade GE. Studies on blood capillaries, I: general organization of blood capillaries in muscle. J Cell Biol. 1968;37:244-276.
Chen WJ, Goldstein JL, Brown MS. NPXY, a sequence often found in cytoplasmic tails, is required for coated pit-mediated internalization of the low density lipoprotein receptor. J Biol Chem. 1990;265:3116-3123.
Wittmaack FM, Gåfvels ME, Bronner M, Matsuo H, McCrae KA, Tomaszewski JE, Robinson SL, Strickland DK, Strauss JF III. Localization and regulation of the human very low density lipoprotein/apolipoprotein-E receptor: trophoblast expression predicts a role for the receptor in placental lipid transport. Endocrinology. 1995;136:340-348.
Rapp JH, Lespine A, Hamilton RL, Colyvas N, Chaumeton AH, Tweedie-Hardman J, Kotite L, Kunitake ST, Havel RJ, Kane JP. Triglyceride-rich lipoproteins isolated by selected-affinity anti-apolipoprotein B immunosorption from human atherosclerotic plaque. Arterioscler Thromb. 1994;14:1767-1774.