Megalin is a member of the LDL receptor gene family that plays an important role in forebrain development and in cellular vitamin D metabolism through endocytic uptake of vitamin D metabolites. Similar to other receptors in this gene family, megalin is believed to functionally interact with intracellular proteins through adaptors that bind to the receptor tail and regulate its endocytic and signal transducing activities. Using yeast two-hybrid screens, we identified a novel scaffold protein with tetratrico peptide repeats, the megalin-binding protein (MegBP) that associates with the receptor. The binding site of MegBP was mapped to an N-terminal region on the receptor tail harboring a proline-rich peptide element. MegBP binding did not block the endocytic activity of the receptor; however, overexpression resulted in cellular lethality. In further screens, we identified proteins that bound to MegBP and thus might be recruited to the megalin tail. MegBP-interacting partners included several transcriptional regulators such as the SKI-interacting protein (SKIP), a co-activator of the vitamin D receptor. These finding suggest a model whereby megalin directly participates in transcriptional regulation through controlled sequestration or release of transcription factors via MegBP.
The low-density lipoprotein (LDL) receptor gene family constitutes a unique group of cell-surface receptors that fulfill many cellular functions. In particular, members of this gene family act both as endocytic and as signal transducing receptors, activities previously attributed to separate receptor classes (Nykjaer and Willnow, 2002; Herz and Bock, 2002). As endocytic receptors for lipoproteins, lipophilic vitamins and protease/inhibitor complexes, they regulate the cellular lipid and protease metabolism (Nykjaer and Willnow, 2002; Herz and Strickland, 2002). Through binding of extracellular signaling molecules such as platelet-derived growth factor, reelin, or sonic hedgehog, they may transmit extracellular signals into cells (D'Arcangelo et al., 1999; Hiesberger et al., 1999; Boucher et al., 2002; Herz and Bock, 2002; Loukinova et al., 2002; McCarthy et al., 2002).
Remarkably, endocytic as well as signaling functions of LDL receptor gene family members are controlled by adaptor proteins that recognize binding epitopes in the cytoplasmic tail of the receptors. Because these adaptors simultaneously interact with other components of the cellular machinery, they assemble multimeric protein complexes that determine receptor activities. For example, binding of the adaptor complex AP-2 couples receptors to the clathrin coat of endocytic vesicle and facilitates endocytosis (Kirchhausen et al., 1997). Interaction with the adaptor disabled (Dab)1 may link receptors with intracellular tyrosine kinases of the Src and Abl family and modulates downstream signal transduction pathways (D'Arcangelo et al., 1999; Hiesberger et al., 1999).
The molecular mechanisms that control interaction of the receptors with adaptor proteins and the consequences for endocytosis and signal transduction are far from being fully understood. Specific epitopes in the receptor tails serve as high-affinity binding site for cytoplasmic scaffold or adaptor proteins. Among others, NPxY motifs, (S/T)xY elements, as well as SH3 binding sites have been identified in the receptor tails (Gotthardt et al., 2001; Rader et al., 2001). Interestingly, each receptor species harbors a unique combination of binding motifs that may even be changed by alternative splicing, further supporting the concept that adaptor interactions play a crucial role in the regulation of individual receptor functions (Stockinger et al., 2002). As far as interacting partners are concerned, a number of adaptor proteins have been uncovered that bind to the LDL receptor gene family. They represent three main types of proteins: proteins with phosphotyrosine-binding domain (PTB) (Hiesberger et al., 1999; Gotthardt et al., 2001; Oleinikov et al., 2001), with PDZ domain (Gotthardt et al., 2001; Patrie et al., 2001) or with an ankyrin repeat (Rader et al., 2001). The significance of most of these adaptors for receptor function remains unknown.
To better understand the molecular interactions of adaptors with LDL receptor gene family members, we have performed new yeast two-hybrid screens to identify novel interaction partners for megalin, a member of the gene family that plays an important role in brain development and in renal vitamin D metabolism; functions that probably involve signal transduction (Willnow et al., 1996; Nykjaer et al., 1999; McCarthy et al., 2002). We identified a novel class of adaptor molecule with tetratrico peptide repeat domains, designated megalin-binding protein (MegBP) that functionally interacts with megalin and links this receptor to transcription regulation pathways.
Materials and Methods
Materials and standard procedures
Cell lines were obtained from the American Type Culture Collection (www.atcc.org) or generated in-house. Experiments involving recombinant DNA technology were performed according to standard protocols (Sambrook and Russel, 2001). The full-length murine MegBP cDNA sequence was assembled from an EST clone obtained from the German Resource Center (RZPD, Berlin, Germany; IMAGE 17691) (see Fig. 1B for detail). During the course of the studies, another murine cDNA sequence was submitted to the database that included additional 21 amino acids at the N-terminus of the original clone (XM_134130).
Yeast two-hybrid screen
The cDNA sequence of the human megalin tail or truncations thereof were produced by RT-PCR from total RNA of human embryonic kidney cell line 293 and confirmed by sequencing. The following primers were used: 5′-GATCCTCATCATATGCACTATAGAAGGACCGGCTC-3′ and 5′-GGTGGTGGGATCCCTATTACTATACTTCAGAGTCTTCTTTAACAAGATTTGCGGTGTCTTT-3′ (full-length tail), 5′-GATCCTCATCATATGCACTATAGAAGGACCGGCTC-3′ and 5′-GGTGGTGGGATCCCTATTACACTTTGACAGCACTGCTCTG-3′ (fragment AB), 5′-GATCCTCATCATATGAAAGTGGTTCAGCCAATCC-3′ and 5′-GGTGGTGGGATCCCTATTACTATCTTCAGAGTCTTCTTTAACAAGATTTGCGGTGTCTTT-3′ (fragment CD), 5′-GATCCTCATCATATGCACTATAGAAGGACCGGCTC-3′ and 5′-GGATCCCTATTACACTCCAATATCCATGTTAAGATC-3′ (fragment A) and 5′-CATATGGGAGTGTCTGGTTTTGGACCT-3′ and 5′-GGTGGTGGGATCCCTATTACACTTTGACAGCACTGTCTCTG-3′ (fragment B). The sequences were cloned into vector pAS2-1 (Clontech, www.clontech.com) via EcoRI and BamHI restriction sites and used as bait to screen a GAL4 Matchmaker library from human brain tissue (HY4028AH, Clontech) according to the manufacturers' recommendations. Positive clones were isolated from the yeast strain Y190, sequenced, and their interaction with the megalin tail confirmed by retransformation of the purified plasmids in the absence or presence of the megalin bait vector. Similarly, truncations of the megalin tail sequence were generated by PCR cloning approach, introduced into vector pAS2-1, and their interaction with MegBP tested by transformation of the respective constructs into yeast strain Y190.
Expression of recombinant proteins
GST and GST-fusions of full-length murine MegBP or the protein interacting domain of Dab1 (Gotthardt et al., 2001) were obtained by cloning of the respective gene sequences into expression vector pGEX-4T-1 (Amersham, www.amershambiosciences.com) and expression in DH5α bacteria. Recombinant proteins were purified according to standard procedures using glutathione-agarose affinity chromatography. An N-terminal fusion of the megalin tail with a hexahistidine epitope (His-megalin tail) was produced by cloning of the receptor sequence into vector pET14b (Novagen, www.novagen.com) and by purification of the fusion protein from BL21 bacteria by routine Ni-NTA affinity chromatography. A fusion protein of GST, Discosoma red (DsRed) fluorescent protein and rat RAP (GST-DsRed-RAP) was generated by introducing the DsRed sequence (from vector pDsRed, Clontech) in-frame into a pGEX expression construct containing the GST-rat RAP sequence (provided by J. Herz). The fusion protein was purified from DH5α bacteria by glutathione-agarose affinity chromatography. The purity of all protein preparations was confirmed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE).
In vitro protein interaction
For ligand blot analysis, equal amounts of purified GST or full-length GST-fusion proteins were subjected to reducing 10% SDS-PAGE and transferred onto nitrocellulose membranes. Membranes were blocked for 2 hours at room temperature in buffer A containing 100 mM Tris/HCl, pH 8.0, 0.9% (w/v) NaCl, and 2% (v/v) Tween-10 (binding buffer). Thereafter, the membranes were incubated for 16 hours at 4°C in buffer B (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 2 mM CaCl2, 2 mM MgCl2, 0.5% bovine serum albumin) containing 125I-labeled His-megalin tail protein (1×106 cpm/ml), washed four times for 15 minutes in buffer B at room temperature and exposed to X-ray film. Iodination of the His-megalin tail protein followed the protocol of Fraker and Speck (Fraker and Speck, 1978). For coprecipitation experiments, 110 μg of anti-GST IgG (G-7781; Sigma, www.sigma.com) was incubated with either 35 μl GST-MegBP (0.869 mg/ml) or 35 μl binding buffer for 2 hours at 4°C. Then, 455 μl renal membrane extracts (0.824 mg/ml) were added to each sample and incubated for an additional 2 hours at 4°C. IgG-bound proteins were recovered using the Seize Classic Immunoprecipitation Kit (#45213; Pierce, www.piercenet.com). BIAcore analysis of the interaction of adaptor proteins with immobilized rabbit megalin or the megalin tail protein has been described in detail before (Nykjaer et al., 1999).
Cell culture studies
Sequences encoding the full-length mouse MegBP or the partial human clone isolated from the GAL4 Matchmaker library were introduced into vector pEGFP-C1 (Clontech) using XhoI and EcoRI restriction sites and expressed as a C-terminal fusion with the enhanced green fluorescent protein (EGFP). For protein expression, cell lines 293 (human embryonic kidney cells), BN16 (rat choriocarcinoma cells), LLCPK-1 (Lewis lung carcinoma porcine kidney cells), L2 (rat yolk sac cells) and MEF (murine embryonic fibroblasts) were grown in standard DMEM containing 10% fetal calf serum (FCS). Cells were transfected with 0.5-1 μg/ml of plasmids encoding EGFP or EGFP-MegBP by biolistic particle delivery (BN16; Biolistic PDS-100 delivery system, www.biorad.com) or liposomal transfection technology (Fugene 6; Roche, www.roche.com). Expression of the proteins was confirmed by fluorescence microscopy (Olympus BX51 microscope) or confocal fluorescence microscopy (Leica TCS SP2). For immunodetection of megalin, cells were fixed in 4% (w/v) paraformaldehyde, blocked for 1 hour at 37°C with 10% FCS/PBS and incubated for 1 hour at 37°C with a goat anti-rabbit megalin antibody (1:1000 in 10%FCS/PBS). Bound IgG was detected by subsequent incubation for 1 hour at 37°C with a Cy5-conjugated rabbit anti-goat antibody (10 μg/ml in 10% FCS/PBS; #81-1616, ZYMED; www.zymed.com). The endocytic uptake of GST-DsRed-RAP was tested by incubation of the cells in phenol red-free DMEM containing 30 μg/ml of purified fusion protein followed by fluorescence microscopy.
We performed yeast two-hybrid screens of a human brain library to identify adaptor proteins that interact with the megalin tail and may play a role in regulation of receptor function. A number of proteins were identified in these screens that specifically interacted with the cytoplasmic domain of the receptor (Table 1). The Glut 1 C-terminal-binding protein (GLUT1CBP), also known as SEMCAP-1, and the Jun N-terminal kinase-interacting proteins (JIP)-1 and -2 have been uncovered as adaptors of megalin in previous screens (Gotthardt et al., 2001). E2a-Pbx 1-activated protein (EB-1), a PTB domain protein that is upregulated in pre-B-cell acute lymphocytic leukemia (Fu et al., 1999) has not been identified as interaction partner before. In addition, a partial cDNA sequence, tentatively titled megalin-binding protein (MegBP), was isolated (Table 1; Fig. 1A). By sequence alignment, a complete cDNA sequence was assembled from mouse and human ESTs (Fig. 1B). Sequence analysis predicted a cytoplasmic protein of 350 amino acids showing 86% conservation at the amino-acid level between human and mouse. The polypeptide sequence encoded several potential phosphorylation sites for protein kinase C, casein kinase II, and tyrosine kinase (Fig. 1B). The protein structure was characterized by the presence of two tetratrico peptide repeats (TPR-1 and -2) (Fig. 1C), a degenerate 34 amino acid motif that forms an amphipathic α-helix (Lamb et al., 1995). TPRs are proposed to be involved in protein-protein interaction (Lamb et al., 1995; Das et al., 1998). No structural elements commonly found in known adaptors of the LDL receptor gene family, such as PTB, PDZ, or ankyrin repeat domains were present in MegBP. MegBP is ubiquitously expressed. High levels of expression are found in ovary, testis, brain and kidney (The RIKEN Consortium, 2001).
In the following experiments, we focused our attention on MegBP because TPR-containing proteins have previously not been considered as adaptors of the LDL receptor gene family. We confirmed the interaction of MegBP with the receptor employing ligand blot and surface plasmon resonance (BIAcore) analysis. To do so, a fusion protein of glutathione S-transferase (GST) and full-length murine MegBP (GST-MegBP) was constructed and expressed in bacteria. The intact protein could be recovered from bacterial extracts by glutathione-agarose affinity chromatography (Fig. 2, lane 1). Additional lower molecular weight proteins in the GST-MegBP preparation represented degradation products of the fusion protein that included the GST moiety as demonstrated by western blot analysis using an anti-GST antibody (Fig. 2, lane 4). As a positive control for the binding assays, we expressed and purified a GST fusion of the protein interacting domain of Dab1, an established adaptor of megalin (Gotthardt et al., 2001) (Fig. 2, lanes 3 and 6). By ligand blot analysis, we demonstrated strong binding of the iodinated megalin tail to full-length GST-MegBP and to GST-Dab1 but not to GST or GST-MegBP degradation products (Fig. 2, lanes 7 to 9). By coprecipitation experiments, we specifically precipitated endogenous megalin from renal mouse extracts using the GST-MegBP fusion protein and glutathione agarose beads (Fig. 3). Finally, by BIAcore analysis, we tested binding of a dilution series of GST-MegBP and GST-Dab1 preparations to the full-length receptor protein and to the recombinant tail domain immobilized on the sensor chip surface (Fig. 4). GST-MegBP reversibly interacted with the full-length receptor and the receptor tail domain with a Kd of 1.1 and 0.8 μM, respectively (Fig. 4A,B). Similar affinities were obtained when binding of GST-Dab1 to the receptor (0.25 μM) or to the tail fragment was analyzed (0.15 μM) (Fig. 4C,D). GST alone did not interact with the receptor polypeptides (Fig. 4B,D). As GST-MegBP bound with the same affinity to native megalin and to the recombinant tail, these data indicated a specific binding site for MegBP in the cytoplasmic but not in the extracellular receptor domain.
Next, we mapped the binding site of MegBP on the receptor tail to gain information about recognition motifs for TPRs. We generated truncations of the human megalin tail sequence and tested their interaction with MegBP in the yeast two-hybrid system. As summarized in Fig. 5, MegBP specifically bound to the N-terminal half of the megalin tail (fragment AB) that encompasses one NPxY and a PxxP motif. No binding to the C-terminal region (fragment CD), which includes the second NPxY motif and the (S/T)xY element of the receptor, was seen. In further fine mapping, the binding site of MegBP was narrowed down to stretch of 45 amino acids adjacent to the membrane anchor of the receptor (fragment A). This polypeptide sequence included the PxxP motif but lacked the NPxY element.
To test whether there is a functional interaction between megalin and MegBP in vivo, we tried to transiently express MegBP as a fusion with the enhanced green fluorescent protein (EGFP) in mammalian cell lines. Surprisingly, no or very faint EGFP-MegBP expression could be detected 24 to 48 hours after transfection of cell lines BN16 and L2 (Fig. 6). These cells expressed large amounts of megalin as demonstrated by western blot analysis (Fig. 7). By contrast, the cell lines readily produced significant amounts of EGFP as shown by fluorescence microscopy (Fig. 6). Contrary to BN16 and L2 cells, strong expression of EGFP-MegBP was achieved in a number of cell lines that did not produce endogenous megalin, including 293 cells, LLCPCK1 cells or murine embryonic fibroblasts (MEF) (Figs 6 and 7). In these cells, MegBP was seen as a punctuate and vesicular staining pattern. The signal was distinctly different from the cytoplasmic staining seen for EGFP in all the cell lines (Fig. 6).
The data obtained so far indicated that high levels of MegBP may be incompatible with megalin expression in BN16 and L2 cells. Therefore, we tested expression of MegBP in BN16 cells at shorter time points after transfection. Consistent with our hypothesis, low levels of MegBP expression could be seen in BN16 cells 4 hours after transfection (Fig. 8). The protein was detected in a vesicular staining pattern close to the plasma membrane and throughout the cytoplasm (Fig. 8B). A localization close to the plasma membrane partially overlapped with the localization of megalin (Fig. 8A,C), suggesting interaction of the proteins in cells. Consistent with an impaired viability, all MegBP-expressing BN16 cells died within 24 hours of transfection. Cell death was not caused by apoptosis as was demonstrated by the absence of cells positive for double-stranded low molecular weight DNA (TUNEL reaction). Rather, MegBP-expressing cells were characterized by disintegration of cellular architecture and necrosis (data not shown).
Previously, it was speculated that binding of adaptors to the tail of megalin may block binding of the AP-2 complex to the receptor and, thus, interfere with endocytosis (Rader et al., 2001). To test whether MegBP binding to megalin has such an effect we established an in vivo fluorescence assay to analyze endocytic uptake of ligands via megalin. As a model ligand we used the receptor-associated protein (RAP), a high-affinity ligand of the receptor (Christensen et al., 1992). RAP was expressed as a fusion protein with Discosoma red (DsRed) fluorescent protein (GST-DsRed-RAP). GST-DsRed-RAP bound to megalin in BIAcore assays (data not shown) and was taken up via this receptor pathway in BN16 cells as shown by immunofluorescence microscopy (Fig. 9A,B). We tested uptake of this ligand in BN16 cells that were transiently transfected with the EGFP-MegBP construct. Four hours after transfection, internalization of GST-DsRed-RAP was seen in all cells, including those that expressed the adaptor (Fig. 9C,D). These findings demonstrated that MegBP expression interfered with the viability of megalin-positive cells; however, it did not inhibit the endocytic activity of the receptor.
The presence of TPRs in MegBP suggested a role for them in protein-protein interaction and perhaps in the assembly of protein complexes at the cytoplasmic domain of megalin. To identify candidates that may interact with MegBP and thereby be recruited to the receptor tail, we used the MegBP sequence as a bait to screen a human brain library. A number of clones was identified by yeast two-hybrid screening that specifically bound to this adaptor (Table 2). In keeping with the proposed role of megalin in signaling processes, potential interaction partners for MegBP could be grouped into two distinct physiological pathways: (i) transcriptional regulators such as SKI-interacting protein (SKIP), transforming growth factor (TGF)-β-stimulated clone 22 homologous gene (THG-1) and Hlark; or (ii) components of signal transduction cascades including protein kinase C α-binding protein (PRKCABP), mitogen-activated protein kinase kinase kinase 10 (MAP3K) and brain-specific angiogenesis inhibitor 2 (BAI2).
Here, we have identified the TPR-containing adaptor MegBP as a novel megalin-interacting protein. MegBP binds to the N-terminal region of the receptor tail. Binding does not affect the endocytic activity of megalin but overexpression of MegBP seems to interfere with cellular viability through receptor interaction. The ability of MegBP to sequester transcriptional regulators suggests a model whereby megalin may directly participate in the regulation of gene transcription.
Recently, a surprising role for the LDL receptor gene family in cellular signal transduction has been uncovered (Trommsdorff et al., 1999). This function has been characterized in detail for the very low-density lipoprotein (VLDL) receptor and the apolipoprotein (apo) E receptor-2, two family members expressed in migrating neurons in the embryonic brain. These neurons receive positional information through binding of the extracellular factor reelin to both receptors. Binding of reelin results in phosphorylation of Dab1 bound to the receptor tails and in activation of downstream signaling pathways (Hiesberger et al., 1999; D'Arcangelo et al., 1999). Genetic defects in the VLDL receptor and in the apo E receptor-2 in mice result in abnormal layering of neurons in the cortex, hippocampus and cerebellum (Trommsdorff et al., 1999). The same phenotype is seen in mice with Dab1 or reelin gene defects (D'Arcangelo et al., 1995; Sheldon et al., 1997).
As well as the VLDL receptor and the apoE receptor-2, megalin also plays a crucial role in brain development as judged by the forebrain formation defects seen in knockout mice (Willnow et al., 1996). In addition, the receptor is essential for the retrieval of filtered vitamin D metabolites in the proximal tubules of the kidney and for regulation of the systemic vitamin D and bone metabolism (Nykjaer et al., 1999; Hilpert et al., 2002). Several lines of evidence suggest that adaptors may contribute to these receptor functions. Thus, a patient was identified that suffered from holoprosencephalic syndrome in combination with renal tubular resorption defects and vitamin D deficiency (Muller et al., 2001). Although this syndrome did not involve the megalin gene directly, as shown by haplotype analysis, the close phenotypic similarity with murine megalin deficiency suggested receptor malfunction caused by an adaptor defect as the underlying disease mechanism (Muller et al., 2001). A similar finding was obtained by Hobbs and colleagues who detected mutations in the adaptor ARH (autosomal recessive hypercholesterolemia) as the reason for LDL receptor dysfunction in a kindred with familial hypercholesterolemia (Garcia et al., 2002). Additional support for a role of megalin in signaling processes through adaptor proteins came from studies that uncovered megalin as a receptor for sonic hedgehog, a central factor in forebrain development (McCarthy et al., 2002), and from the identification of a number of proteins that interacted with the receptor tail. These adaptors included the scaffold proteins JIP-1, JIP-2 and Dab1 in the brain (Gotthardt et al., 2001), as well as Dab2 (Oleinikov et al., 2001), ANKRA (Rader et al., 2001) and MAGI-1 (Patrie et al., 2001) in the kidney.
Although little is known about the physiological significance of the identified adaptors for megalin function, these studies have already provided some insights into the molecular structures required for receptor and adaptor interaction. So far, three major types of adaptors have been shown to bind to the receptor tail. PTB-containing proteins (e.g. Dab1) bind to NxPY motifs in the receptor tail, whereas PDZ domain proteins interact with the motif (S/T)xY (e.g., MAGI-1). A third group of adaptors, represented by the ankyrin-repeat-containing protein ANKRA, may recognize a proline-rich PxxP element (Rader et al., 2001). MegBP constitutes a novel class of megalin adaptor with TPR structure. It binds to a 45 amino-acid motif in the N-terminal region of the receptor tail characterized by a proline-rich PxxP motif. The very same element is recognized by ANKRA and may also serve as a docking site for factors with SH3 domain (Rader et al., 2001). Because TPRs are the only obvious structural element in MegBP, these repeats probably represent the binding domain of the adaptor. Whether one or both repeats contribute to receptor recognition remains to be shown. The inability of the megalin tail to bind to C-terminal GST-MegBP truncation products (Fig. 2) suggests that both repeats may be required to do so.
TPRs are found in a number of nuclear, cytoplasmic and mitochondrial proteins. It is assumed that they serve as interaction sites with non-TPR sequences and facilitate the assembly of multimeric protein complexes (Lamb et al., 1995; Das et al., 1998). Examples of such protein complexes include the mitochondrial import complex, peroxisomal import receptor complex and transcription regulation complexes (Lamb et al., 1995; Das et al., 1998). In support of a role of TPR proteins in transcription regulation, proteins that bound to MegBP in the yeast two-hybrid system included those involved in RNA binding (Hlark) (Jackson et al., 1997) and in transcriptional activation (SKIP) (Zhang et al., 2002) or repression (THG-1) (Kester et al., 1999). This observation raises the intriguing hypothesis that megalin may directly regulate gene transcription through controlled binding or release of transcription factors. Overexpression of MegBP in megalin producing cells may thus result in uncontrolled sequestration of transcriptional regulators (or other MegBP interaction partners) on the receptor tail and in impaired cellular viability.
What is the physiological relevance of the megalin and MegBP interaction? Two of the identified MegBP-binding partners provide some insights into that question. TGF-β-stimulated clone 22 (TSC-22) and its homologue the TGF-β-stimulated clone 22 homologous gene 1 (THG-1) are two related leucine zipper proteins (Kester et al., 1999). When fused to a GAL4 DNA-binding domain, they inhibit transcription from a heterologous promoter, suggesting that these proteins act as endogenous transcription repressors when sequestered at the DNA (Kester et al., 1999). A link to megalin function is provided by the fact that TSC-22 is highly upregulated in megalin-deficient kidneys as uncovered by expression profiling (Hilpert et al., 2002). The second interaction partner with obvious relevance for megalin function is SKIP. SKIP, also known as NcoA-62, is a co-activator of the vitamin D receptor (VDR). It binds to the hormone receptor and stimulates its activity in a ligand-independent manner (Baudino et al., 1998; Zhang et al., 2002). A role for megalin in cellular vitamin D metabolism is well established. It binds vitamin D metabolites complexed with the carrier vitamin-D-binding protein (DBP) and mediates their uptake into cells (Nykjaer et al., 1999; Nykjaer et al., 2001). Thus, endocytosis of vitamin D metabolites through megalin may be directly linked to activation of the nuclear hormone receptor. In such a model, SKIP may be sequestered via MegBP to the receptor tail and released upon endocytic uptake of vitamin D metabolites. Unbound SKIP then interacts with the VDR and primes it for incoming ligands. Although speculative, this model incorporates data obtained in independent experimental systems (including receptor-deficient mice) and provides a working hypothesis.
We are indepted to D. Bischof, H. Schulz and D. Vetter for expert technical assistance and to M. Gotthardt for helpful discussions. Studies presented here were funded by a grant from the Deutsche Forschungsgemeinschaft and by a fellowship of the Alfred Benson Foundation to H.H.P.
- Accepted October 29, 2002.
- © The Company of Biologists Limited 2003