Evidence for the Role of Megalin in Renal Uptake of Transthyretin*

The kidney is a major organ for uptake of the thyroid hormone thyroxine (T 4 ) and its conversion to the active form, triiodothyronine. In the plasma, one of the T 4 carriers is transthyretin (TTR). In the present study we observed that TTR, the transporter of both T 4 and reti-nol-binding protein, binds to megalin, the multiligand receptor expressed on the luminal surface of various epithelia including the renal proximal tubules. In the kidney, megalin plays an important role in tubular uptake of macromolecules filtered through the glomerulus. To evaluate the importance of megalin for renal uptake of TTR, we performed binding/uptake assays using im-mortalized rat yolk sac cells with high expression levels of megalin. Radiolabeled TTR,

Proximal tubules of the kidney serve an important function for the uptake of macromolecules passing the glomerular filtration barrier. Therefore, despite the massive influx of protein in the proximal tubules, human urine is virtually devoid of significant amounts of protein. By this way the proximal tubules salvage amino acids and essential protein-bound components such as lipids, hormones, and vitamins.
The receptor megalin plays an important role in the reuptake mechanism. Megalin is a multiligand endocytic receptor expressed in clathrin-coated pits at the apical surface of a number of absorptive epithelia, including those of the proximal tubule (1) and yolk sac (2). Megalin is a member of the low density lipoprotein receptor family (3) and binds, as do the other members of this family, the ϳ40-kDa receptor-associated protein (RAP) 1 (4), which functions as a specialized chaperone/ escort protein during the biosynthesis of some of the members of the low density lipoprotein receptor family and in their delivery to the cell surface (5,6).
Megalin ligands include vitamin carriers known to be filtered, such as transcobalamin (vitamin B 12 -binding protein) (7), vitamin D-binding protein (8), and retinol-binding protein (RBP) (9). The general importance of megalin was supported by the findings that knockout mice for the megalin gene result in high mortality, developmental abnormalities (10), and tubular reabsorption deficiency with excretion of low molecular weight plasma proteins in the urine (low molecular weight proteinuria) (11).
In the plasma, holo-RBP strongly interacts with transthyretin (TTR), and approximately 50% of TTR circulates as a 1:1 molar TTR⅐RBP complex (12). The formation of the TTR⅐RBP complex prevents to a certain extent the RBP⅐retinol complex from being filtered in the glomeruli. However, 4 -5% of the circulating RBP⅐retinol is not bound to TTR (13) and is taken up by means of megalin in the proximal tubule (9). Apart from transporting retinol via binding to RBP, plasma TTR, a tetramer of four identical subunits of approximately 14 kDa (14), acts as a transport protein for the thyroid hormone thyroxine (T 4 ). Most, if not all, of the active form of T 4 , triiodothyronine, is generated by deiodination of T 4 mainly in the liver and in the proximal tubules of the kidney (15).
Despite no detection of TTR mRNA in the adult kidney (16), which is one of the major extrahepatic sites of TTR degradation (18), positive staining was reported in the epithelium of the renal proximal tubules (17). However, the mechanism of TTR internalization and degradation remains to be elucidated, although a receptor-mediated uptake has been suggested (19). Because megalin has been implicated in the renal reuptake of plasma proteins that carry lipophilic compounds, we investigated the possibility that this receptor might also play a role in renal uptake of TTR.

EXPERIMENTAL PROCEDURES
Proteins and Antibodies-Recombinant TTR was purified from Escherichia coli D1210 transformed with plasmids carrying either wild type TTR (pINTRwt) or the suitable mutant TTR cDNA (pINTR30 and pINTR119) according to Almeida et al. (20). Serum RBP was isolated by affinity in a TTR column and saturated with a molar excess of all-trans retinol (Sigma) by incubation at 37°C in the dark for 1 h; excess retinol was separated from RBP by gel filtration in 10-ml Biogel P-6 DG columns (Bio-Rad). Recombinant RAP was expressed and purified as described previously (21). Megalin was purified by RAP affinity chromatography from human kidney cortex according to standard procedures (22). Purified sheep polyclonal antibodies against rat megalin have been described, and their specificity has been characterized (7). Purified sheep non-immune IgG was used as a negative control in binding experiments.
Protein Iodination-TTR and RAP were iodinated following the iodogen method. Briefly, to reaction tubes coated with 10 g of iodogen (Sigma), 100 l of 0.25 M phosphate buffer and 1 mCi (37 MBq) of Na 125 I (Amersham Pharmacia Biotech) were added, followed by 10 g of protein in phosphate-buffered saline (PBS). The reaction was allowed to proceed in an ice bath for 10 min. Labeled protein was separated from free iodide in a 5-ml Sephadex G50 column (Amersham Pharmacia Biotech). For TTR, specific activities were determined after each iodination by a quantitative enzyme-linked immunosorbent assay using polyclonal anti-human TTR (Dako) as the coating antibody and peroxidase-conjugated anti-human TTR (The Binding Site) as secondary antibody. 125 I-labeled TTR ( 125 I-TTR) concentration was calculated from a standard curve ranging from 5 to 200 ng/ml. Characteristic specific activities were of 10 4 cpm/ng. Uptake of TTR in Cultured Yolk Sac Cells-Megalin-expressing Brown Norway Rat yolk sac epithelial cells transformed with mouse sarcoma virus (BN cells) (24) were grown to confluence in 24-well plates (Nunc A/S) in Dulbecco's modified Eagle's medium containing 10% (v/v) fetal calf serum, 100 units/ml penicillin, and 100 g/ml streptomycin. Before incubation cells were washed with ice-cold PBS. Incubation with 125 I-TTR was carried out in serum-free Dulbecco's modified Eagle's medium supplemented with 0.1% (w/v) ovalbumin for the indicated periods of time, either at 4 or 37°C. In some experiments 125 I-TTR was added in the presence of RAP (1 M), IgG antibody against megalin (200 g/ml), IgG antibody against cubilin (200 g/ml), or sheep non-immune IgG (200 g/ml). Degradation of labeled protein was measured by precipitation of the incubation medium in 10% trichloroacetic acid. In all experiments, a control was included in which the amount of degradation was assessed in the absence of cells. Cell-associated radioactivity was determined by measuring radioactivity of the washed cell layer in ice-cold PBS followed by cell solubilization in 0.1 M NaOH. Total cellular protein was measured with the Bio-Rad protein assay kit (Bio-Rad), using bovine serum albumin as a standard. Cell association of 125 I-TTR measured at a saturating concentration of unlabeled ligand (1 mg/ml) was considered nonspecific and subtracted from all values.
Surface Plasmon Resonance (SPR) Analysis-Receptor-ligand inter-actions were assessed by SPR analysis on a BIAcore 2000 instrument (Biacore) as described (25). Megalin was immobilized onto a CM5 sensor chip, using the amine-coupling kit as described by the supplier, at indicated densities. A control channel was routinely activated and blocked in the absence of protein. Binding to coated channels was corrected for binding to noncoated channels. SPR analysis was assessed in 150 mM NaCl, 2 mM CaCl 2 , 0.005% (v/v) Tween 20, and 20 mM Hepes (pH 7.4) at 20°C. Preparation of Renal Tissue-Megalin-deficient mice were produced by gene targeting as described (11). Wild type littermates were used as controls. Mouse megalin knockout and control kidneys were fixed by perfusion through the heart with 4% paraformaldehyde in 0.1 M sodium cacodylate buffer. The tissue was trimmed into small blocks, further fixed by immersion for 1 h in 1% paraformaldehyde, infiltrated with 2.3 M sucrose containing 2% paraformaldehyde for 30 min, and frozen in liquid nitrogen.
Immunohistochemistry-For light microscopy, 0.8-m cryosections were obtained at Ϫ80°C with an FCS Reichert Ultracut cryoultramicrotome as described previously (1). For immunolabeling, the sections were incubated with rabbit anti-rat TTR primary antibody (26), diluted 1:200, at room temperature for 1 h after preincubation in PBS containing 0.05 M glycine and 1% bovine serum albumin. The sections were subsequently incubated with peroxidase-conjugated secondary antibodies (Dako), and the peroxidase was visualized with diaminobenzidine. As control, sections were incubated with secondary antibodies alone or with nonspecific rabbit IgG. The sections were subsequently counterstained with Meier's stain for 2 min and examined in a Leica DMR microscope equipped with a Sony 3CCD color video camera attached to a Sony Digital still recorder. Images were processed using Adobe Photoshop 4.0.
Urine Samples-Overnight urine samples from six patients with Fanconi syndrome, four of whom had Dent's disease (27) due to mutations of the renal chloride channel (28), and from three healthy control subjects were refrigerated immediately after collection and stored at Ϫ80°C until processed for immunoblotting.
SDS-Polyacrylamide Gel Electrophoresis of Urine Samples-10-l urine samples were electrophoresed in 8 -16% SDS-polyacrylamide gel electrophoresis gels and subsequently transferred to nitrocellulose. Blots were blocked in 5% milk in PBS containing 0.1% Tween 20 for 1 h and incubated for 1 h at room temperature with anti-human TTR (Dako) in PBS containing 0.1% Tween 20. After washing in PBS containing 0.1% Tween 20, blots were incubated for 1 h with alkaline phosphatase-conjugated anti-rabbit IgG (Dako). Antibody binding was visualized using nitro blue tetrazolium and 5-bromo-1-chloro-3-indolyl phosphate color development substrates (Promega).

RESULTS
To investigate the hypothesis that megalin might be responsible for the tubular uptake of T 4 by internalizing TTR, we analyzed TTR binding/uptake using an established immortalized rat yolk sac epithelial cell line with high expression of megalin (24). 125 I-TTR bound efficiently and in a saturable manner to the cells at 4°C (Fig. 1a), consistent with the existence of a TTR receptor. The apparent K d for the 4°C, 4-h binding to the cells was estimated to be approximately 500 nM when assuming one class of binding sites (Fig. 1b). 125 I-TTR was rapidly taken up, and in accordance with an endocytic process, radiolabeled degradation products appeared in the medium after a lag time of approximately 30 min (Fig. 1c). Saturating concentrations of polyclonal antibodies against megalin showed a 63% inhibition in uptake, whereas no significant effect was seen with anti-cubilin antibody nor with nonimmune IgG (Fig. 2a). Cubilin, an apical receptor expressed in the same tissues as megalin (29), is another apical receptor present in the BN cell line. The absence of significant inhibition of TTR uptake by the anti-cubilin antibody rules out the possibility of a nonspecific steric effect produced by the anti-megalin antibody that could be responsible for the observed inhibition of TTR uptake. Therefore, the BN cell line presents specific megalin-mediated TTR degradation.
A strong inhibition in uptake was also observed with RAP (67%) (Fig. 2a). These data further suggest that the mechanism of TTR uptake presents features of a low density lipoprotein receptor-mediated mechanism, namely, RAP sensitivity. BN cells were analyzed under non-reducing conditions by antimegalin immunoblotting and ligand blotting with radiolabeled RAP (Fig. 2b). The only RAP-binding protein observed was megalin, thus showing that in the BN cells, megalin is the prime RAP-binding receptor. Although these cells express cubilin, which has modest affinity for RAP, the involvement of this receptor in TTR internalization had been ruled out by competition experiments using the anti-cubilin antibody as described above.
SPR (data not shown) confirmed binding of purified TTR to immobilized megalin with an approximate K d of 5 M at 20°C, comparable with the low affinity of the interaction of RBP and megalin (9). The influence of T 4 on TTR binding to megalin could not be assessed by SPR due to the well known hydrophobicity of the hormone, resulting in nonspecific binding to the SPR sensor chip.
The influence of TTR ligands on the interaction of the protein with megalin was further studied by uptake experiments of 125 I-TTR complexed with either RBP or T 4 in the rat yolk sac epithelial cell line. No significant difference was observed, however, by the presence of TTR ligands on the degradation and cellular association of TTR (Fig. 3a).
Several point mutations in TTR have been described (30), and most of them are associated with the occurrence of familial amyloidotic polyneuropathy, a disease characterized by the extracellular deposition of TTR amyloid fibrils in various tissues (31). To study the influence of TTR mutations on the uptake by megalin, we tested TTR V30M, an amyloidogenic mutant, and TTR T199M, a non-amyloidogenic variant. The uptake of the different TTRs was corrected for their specific activities. TTR V30M was the mutant with the highest uptake, whereas TTR T119M was the variant with the lowest uptake (Fig. 3b). These data suggested that the conformation of TTR influences the recognition by megalin.
To evaluate the physiological importance of an endocytic mechanism for TTR uptake in the proximal tubules, we analyzed the urine of patients with Fanconi syndrome and Dent's disease. Dent's disease is known to be associated with tubular failure, which is probably caused by a defect in receptor-mediated endocytosis in the proximal tubules. The molecular basis of this disorder has been defined to be due to inactivating chloride channel 5 (CLC-5) mutations (28). Furthermore CLC-5 has been shown to be present in the early endosomes of the receptor-mediated tubular endocytic pathway, in which megalin is the prime receptor (32). Western blotting analysis clearly identified TTR in the urine of patients with Fanconi syndrome, whereas it was absent in control individuals (Fig. 4). This suggested that in vivo TTR might be taken up in the proximal tubules by an endocytic mechanism.
To further demonstrate that the receptor responsible for tubular uptake of TTR in vivo is megalin, we compared proximal tubules of megalin knockout mice and control animals by TTR immunohistochemistry. Light microscope immunohistochemistry revealed a granular staining pattern for TTR in renal proximal tubules of control mice (Fig. 5a). The staining was observed only in segment 1 of the proximal tubule and probably only the initial part of segment 1, indicating that the protein under physiological conditions is removed very efficiently in the early part of the proximal tubule after glomerular filtration. Megalin-deficient mice (Fig. 5b) presented no staining in kidneys, suggesting absence of TTR uptake in the proximal tubules. In both wild type and knockout animals, erythrocytes were stained because of endogenous peroxidase activity (Fig. 5). The absence of TTR vesicular labeling in the proximal tubules of megalin-deficient mice demonstrates that these animals lack the endocytic mechanism of TTR tubular uptake present in the control animals. TTR therefore represents a novel megalin ligand.

DISCUSSION
The present study reveals that megalin is a receptor for tubular uptake of TTR. Megalin interaction with TTR was demonstrated in vitro by SPR analysis and by uptake studies of 125 I-TTR in cells with high expression levels of megalin. Although the affinity of TTR for this receptor is relatively low, it is the same order of magnitude previously reported for other megalin ligands (11). Further evidence pointing to the in vivo relevance of this interaction came from the observations that patients with renal tubule failure excrete TTR in the urine and also that megalin knockout mice do not present lysosomal accumulation of TTR in renal tubules when compared with control wild type littermates. Because TTR is a carrier of T 4 and retinol (in the latter case by the formation of a complex with RBP), it is possible that the presented mechanism is of potential importance in the transepithelial transport of retinol or thyroxine or both. Because RBP is also a megalin ligand (9), TTR might be more important for the renal uptake of T 4 .
Thyroid hormones are synthesized in the thyroid gland and are important in regulating basal metabolism and in controlling cellular growth and differentiation (33). T 4 represents the majority of the hormone synthesized and the major circulating form in the plasma. Triiodothyronine, the biologically active form of the hormone, derives mostly from T 4 deiodination in the peripheral tissues including the kidney (15). Once secreted, more than 99% of the thyroid hormones in circulation are bound to plasma proteins. In human plasma, TTR is one of the three proteins responsible for the transport of T 4 , the main carrier being thyroxine-binding globulin; albumin is the third T 4 -binding protein and the one that presents the lowest affinity for the hormone (14). In rodent serum, TTR is the major carrier of T 4 . Interestingly, albumin is known to be taken up in the proximal tubules by two receptors: megalin and the co-localizing receptor, cubilin (34).
Plasma TTR derives mostly from the liver and transports about 15% of T 4 , which may be reabsorbed via megalin. However, in transthyretin null mice (35) it was shown that T 4 and triiodothyronine tissue content is normal in the case of the kidney and that the amount of deiodinase mRNA, which is directly correlated with the enzymatic activity, remains unaltered in this tissue (36). However, the putative importance of TTR in the normal physiological uptake of T 4 should not be disregarded, because it is possible that a redundant mechanism accounts for T 4 uptake in TTR knockout mice. The thyroid hormone status of megalin knockouts should be addressed in the future for the possibility that these animals excrete T 4 in the urine as a result of a lack of TTR uptake in the proximal tubules.
Several point mutations have been described in TTR (30), the most common being a Val for Met substitution at position 30 of the protein (30). Most of these mutations are related to the occurrence of familial amyloidotic polyneuropathy, which is an autosomal dominant disorder that is characterized by the deposition of amyloid fibrils in several tissues, particularly in the peripheral nervous system (31). Some of these TTR mutants were used to evaluate binding to megalin, and in accordance with the binding data, TTR structure seems to be important for the uptake by megalin.
It is interesting to note that TTR plasma levels are decreased in familial amyloidotic polyneuropathy V30M patients (37), despite an equal expression of the variant and normal TTR in the liver. These facts suggested that TTR metabolism could be involved in the amyloid formation process. Comparative clearance studies of TTR V30M and TTR T119M have been performed (38) and showed a slower clearance for TTR T119M and a faster one for TTR V30M. This led to the hypothesis that, at least in part, the different clearances could account for the differences in circulating plasma levels observed for each of the mutations. It was speculated that one of the factors that might be involved in the existence of differences in clearance would be the existence of cellular receptors for TTR with different affinities for the two mutant forms of the protein. This has been demonstrated here in the case of megalin. TTR has also been described to interact with a variety of compounds; the significance of these interactions is, in most cases, still unexplained. These minor TTR ligands include noradrenaline oxidation products (39), pterins (40), chicken lu-tein, hemin and hemoglobin (41), polyhalogenated compounds (42), and retinoic acid (43). The megalin-mediated TTR uptake in the proximal tubules may also be important in the uptake of these TTR ligands. In conclusion, the findings here presented show TTR as a novel megalin ligand with potential importance in T 4 transepithelial transport and reinforce the concept that megalin is a general endocytic receptor for protein in the proximal tubule (23), with a multifaceted role in retaining and capturing vital substances from the tubular fluid after glomerular filtration.