Megalin binds and mediates cellular internalization of folate binding protein Henrik Birn 1 , Xiaoyue Zhai 1 , Jan Holm 2 , Steen I. Hansen 2 , Christian Jacobsen 3 , Erik I. Christensen 1 and Søren K. Moestrup 3 1 Department of Cell Biology, Institute of Anatomy, University of Aarhus, Denmark 2 Department of of Clinical Chemistry, Hillerød Hospital, Denmark 3 Department of Medical Biochemistry, University of Aarhus, Denmark Folate is a vitamin involved in essential biological processes such as the synthesis of nucleic acid and the metabolism of amino acids. Humans are unable to synthesize this vitamin and thus rely on intestinal reabsorption. Distribution into tissues thereafter is dependent on specific uptake mechanisms involving the reduced folate carrier and ⁄ or the 30–40 kDa folate binding proteins (FBPs) [1–3]. At least three isoforms of FBPs have been identified including both glycosylphosphatidylinositol (GPI)-linked, membrane associated folate receptors highly expressed in certain epithelial tissues and soluble proteins present in serum and other biological fluids [1,2] including the secretory fluids: milk, saliva, and semen [4–9]. The significance of the soluble folate binders is largely unresolved, but it has been suggested that milk FBP may play a role in folate uptake [2], i.e. by protect- ing against oxidation, by promoting folate absorption in the suckling animal [10–12], or protect against bacterial utilization of secreted folates [13]. Based on Keywords absorption; endocytosis; intestine; kidney; vitamins Correspondence H. Birn, Department of Cell Biology, Institute of Anatomy, University of Aarhus, University Park Building 234, DK-8000 Aarhus C, Denmark Fax: +45 86198664 Tel: +45 89423051 E-mail: hb@ana.au.dk (Received 15 March 2005, revised 4 July 2005, accepted 11 July 2005) doi:10.1111/j.1742-4658.2005.04857.x Folate is an essential vitamin involved in a number of biological processes. High affinity folate binding proteins (FBPs) exist both as glycosylphospha- tidylinositol-linked, membrane associated folate binding proteins and as soluble FBPs in plasma and some secretory fluids such as milk, saliva and semen. The function and significance of FBPs are unresolved, however, it has been suggested that they may facilitate folate uptake, e.g. during suck- ling. The present study shows that megalin, a large, multiligand endocytic receptor and member of the low-density lipoprotein-receptor family, is able to bind and mediate cellular uptake of FBP. Surface plasmon resonance analysis shows binding of bovine and human milk FBP to immobilized megalin, but not to low density lipoprotein receptor related protein. Bind- ing of 125 I-labeled folate binding protein (FBP) to sections of kidney proxi- mal tubule, known to express high levels of megalin, is inhibitable by excess unlabeled FBP and by receptor associated protein, a known inhibitor of binding to megalin. Immortalized rat yolk sac cells, represent- ing an established model for studying megalin-mediated uptake, reveal 125 I-labeled FBP uptake which is inhibited by receptor associated protein and by antimegalin antibodies. Microinjection of 125 I-labeled FBP into renal tubules in vivo shows proximal tubular uptake by endocytosis. Megalin is expressed in several absorptive epithelia, including intestine and kidney proximal tubule, and thus the present findings provide a mechanism for intestinal and renal endocytic uptake of soluble FBP. Abbreviations 5-MTHF, 5-methyltetrahydrofolate; CPM, counts per minute; FBP, folate binding protein; GPI, glycosylphosphatidylinositol; LRP, low-density lipoprotein receptor related protein; RAP, receptor associated protein. FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS 4423 studies of folate-FBP uptake in suckling rats it was suggested that intestinal uptake of FBP-bound folate resembles the endocytic absorption of other macro- molecules in the neonate intestine [12], however, the mechanism of this process has not been identified. We have previously characterized different receptors involved in the endocytic uptake of proteins and nutrients in the kidney and other tissues [14]. Meg- alin, an  600 kDa member of the low-density lipoprotein-receptor family [15], is a multiligand, endocytic receptor expressed in a number of absorp- tive epithelia including kidney, yolk sac, choroid plexus, and intestine [14,16,17]. So far more than 40 different ligands have been identified representing a wide variety of substance including lipoproteins, hor- mones, carrier proteins, enzymes, and drugs. Megalin is involved in the endocytic uptake of a number of vitamin carrier proteins [18], including retinol binding protein [19], transcobalamin [20], and vitamin D binding protein [21,22] and its role for the recovery of vitamins and carrier proteins filtered in the renal glomerulus is well established [18,23]. In addition, megalin seems important for normal expression of the intestinal intrinsic factor-vitamin B 12 -receptor cubilin [16,24]. Inspired by these observations we have examined a possible interaction between FBP and megalin. The presented data shows that megalin is able to bind and mediate the endocytosis of soluble FBP providing a potential mechanism for intestinal and renal uptake of soluble FBP-bound folate. Results Megalin binds FBP Surface plasmon resonance analysis showed binding of purified, human and bovine milk apo-FBP to immo- bilized rabbit megalin. Binding was also observed with bovine FBP saturated with folic acid, 5-methyl- tetrahydrofolate (5-MTHF), or methotrexate (Fig. 1). Using a biaevaluation program K d was estimated to  0.3–0.5 lm for apo-FBP and saturated FBP (Fig. 1). No binding was identified to low-density lipoprotein receptor related protein (LRP), the most closely related member of the LDL-receptor family [25]. To confirm binding of FBP to megalin in tissues we used autoradiography on sections of rat kidney cortex showing very high expression of megalin in kidney proximal tubule brush border membranes. Incubation with 125 I-labeled human or bovine milk FBP resulted in accumulation of autoradiographic grains along the apical part of proximal tubule cross-sections similar to the localization of megalin (Fig. 2). This was inhibited by coincubation with excess unlabeled FBP or with receptor associated protein (RAP), an established inhibitor of the uptake of most ligands by megalin [14], suggesting specific binding to megalin. Endocytosis of FBP Microinjection of 125 I-labeled bovine milk FBP in rat kidney nephrons in vivo followed by autoradiography A B Fig. 1. Binding of soluble FBP to immobilized megalin and LRP by surface plasmon resonance analysis (BIAcore). Soluble human or bovine milk FBP purified by a combination of ion-exchange and affinity chromatography was passed over the sensor chips with immobilized, purified megalin, or LRP and the SPR signal (RU) representing bound protein was recorded. (A) SPR sensorgram showing binding of bovine milk apo-FBP, folic acid-FBP, 5-MTHF-FBP, and methotrexate-FBP to immobilized megalin. Using the BIAEVALUATION program K d was estimated between 0.3 and 0.5 l M. Similar binding of human apo-FBP was observed (not shown). (B) No binding of bovine milk apo-FBP, folic acid- FBP, 5-MTHF-FBP, or methotrexate-FBP was observed to purified LRP. Megalin-mediated uptake of FBP H. Birn et al. 4424 FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS revealed uptake of FBP within proximal tubule seg- ments (Fig. 3), known to express high levels of meg- alin. In contrast, distal nephron segments and collecting ducts revealed no significant labeling. The label could be identified in endocytic vesicles and vacu- oles 15–30 min after microinjection (Fig. 3B,C). Megalin mediates cellular uptake of FBP Megalin expressing, rat yolk sac BN-16 cells internal- ized human milk 125 I-labeled FBP (Fig. 4). Total uptake was calculated as the sum of degraded, non tri- chloroacetic acid-precipitable label and cell associated label. The time course of uptake is comparable to pre- vious published uptake of the vitamin B 12 carrier transcobalamin in similar cells [20]. The total uptake was significantly inhibited by RAP and antimegalin antibodies, showing the involvement of megalin in uptake. A small, however, significant inhibition with nonspecific IgG was observed at 4 h explained by low- affinity binding of immunoglobulin light chains to megalin [26]. While degradation was clearly inhibited, the cell associated amount of label was less affected by RAP. Discussion FBP is present in different biological, secretory fluids, including saliva and semen, and in particular milk [4–9]. The significance of these folate binders is largely unknown, but it has been suggested that milk FBP, shown to be resistant to gastric digestion [27], may play a role in folate uptake in the neonate intestine [10–12]. The present study shows that megalin, a large, Fig. 2. Autoradiography showing RAP-inhibitable binding of soluble 125 I-labeled FBP to proximal tubules of rat kidney cortex. Sections of per- fusion fixed rat kidney cortex were incubated with 125 I-labeled human or bovine milk FBP. Binding to sections was identified by autoradiogra- phy following 36–47 days of exposure. For inhibition studies sections were coincubated with either excess unlabeled FBP or RAP (50 lgÆmL )1 ). Labeled bovine (A) and human (C) milk FBP is concentrated along the apical part of proximal tubule cross-sections similar to the localization of megalin (insert , 2C). Binding to sections is inhibited by excess unlabeled bovine FBP (B, compare with A) and RAP (D, compare with C) suggesting binding to megalin. Bars equal 10 lm. H. Birn et al. Megalin-mediated uptake of FBP FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS 4425 multiligand, endocytic receptor expressed in the intes- tine and other absorptive epithelia binds and mediates the internalization of FBP. This is based on the obser- vations that: (a) purified bovine and human milk FBP binds to immobilized megalin when analysed by sur- face plasmon resonance analysis. The K d can be esti- mated to  0.3 lm comparable to the affinity of several other ligands binding to megalin; (b) milk FBP binds to proximal tubules in sections of rat kidney cortex known to express high levels of megalin. This binding is inhibited by excess unlabeled FBP and by RAP, a chaperone and known inhibitor of the uptake of most ligands by megalin; (c) in vivo microinjection of 125 I-labeled milk FBP into rat kidney nephrons reveal endocytic uptake in megalin-expressing proximal tubule cells only; (d) BN-16 cells known to express megalin and representing an established model for megalin-mediated uptake, internalize 125 I-labeled milk FBP, and internalization is strongly inhibited by RAP and antimegalin antibodies. These findings suggest a potential role for megalin in mediating internalization of FBP-bound folate, providing a candidate mecha- nism for both intestinal and renal tubular uptake of soluble FBP. Megalin is expressed in the intestine [16,17] and megalin-mediated endocytosis provides a mechanism for the previously suggested [12] endocytic absorption of FBP-bound folate during suckling. It was noted that in contrast to free folate, FBP-bound folate is absorbed more avidly in the ileum than in the jejunum correlating with the observed expression of megalin in purified apical brush-border membranes from distal, but not proximal, rat intestine [17]. The kinetics of FBP transport in neonatal goat intestinal brush border have revealed that K m ¼ 0.39 lm for unsaturated FBP [11], comparable to our findings using isolated proteins. Small differences in K d were calculated comparing the response curves for apo- FBP, folic acid-FBP, 5-methyltetrahydrofolate-FBP, and methotrexate-FBP, with apo-FBP having the highest affinity compared to saturated FBP. While this in line with the observation of [11], the differ- ences are small and probably within the methodological variation associated with surface plasmon resonance analysis. Renal proximal tubular epithelium expresses abun- dant megalin serving an important role mediating reabsorption of vitamin-carrier protein complexes fil- tered in the glomeruli thus preventing excessive urinary loss [18,23]. These carrier complexes include retinol binding protein [19], transcobalamin [20], and vitamin D binding protein [21,22]. Although most folate in plasma is filtered in the renal glomerulus as unbound folate, a soluble FBP is present in plasma at a concen- tration of about 0.6 nm in humans [28]. The molecular weight of  35 kDa suggests that this protein to a large extent is filtered, and FBP has been detected in human urine at a concentration of 0.3 nm [29]. Assu- ming a glomerular filtration rate of 180 L ⁄ 24 h and urinary excretion rate of 2 L ⁄ 24 h it may be calculated that only < 1% of the plasma FBP actually filtered in the glomeruli is excreted suggesting efficient tubular reabsorption. Thus,  108 nmoles or  48 lg of folate may be recovered daily by tubular uptake of FBP which may be important in individuals with very low-folate intake. The present data provides a mecha- Fig. 3. Uptake of 125 I-labeled bovine milk FBP microinjected into rat nephrons in vivo. Uptake of 125 I-labeled FBP is visualized by autoradio- graphy on sections from fixed kidney cortex. Grains are located over proximal tubule profiles (A; PT) only revealing selective uptake in this part of the nephron characterized by heavy expression of megalin (Fig. 2C, inset). No labeling is oberved in distal tubule profiles (A; DT). Labeling is concentrated in the subapical part of the proximal tubule cells localized close to vacuolar structures (arrows in B) which by electron microscopy can be identified as apical, endocytic vesicles or vacuoles (C; E). MV, microvilli. Bars equal 10 lm (A and B) and 0.5 lm(C). Megalin-mediated uptake of FBP H. Birn et al. 4426 FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS nism for this involving megalin-mediated endocytosis similar to the uptake of other vitamin carrier proteins. Free folate is reabsorbed within the renal proximal tubule by a mechanism dependent on GPI-anchored FBP expressed in the luminal plasma membrane of proximal tubule cells [30,31]. However, in particular at low plasma folate concentrations an additional mechanism of tubular folate uptake may operate [31]. Megalin-mediated uptake of filtered folate bound to plasma FBP may constitute such an alternative mechanism. Megalin-mediated uptake of FBP in BN-16 cells results in the degradation of FBP. Most likely folate bound to FBP is released from the binder in the acidic environment of the endocytic compartment followed by transport across the vesicular membrane into the cyto- sol. Similarly, the uptake of free folate mediated by GPI-anchored FBPs expressed in the apical membrane is suggested to involve binding of folate to the GPI- linked FBPs, followed by endocytosis, release of folate from the receptor within an internal compartment and recycling of the receptor [32–34]. Thus, it is possible that folate internalized either by megalin-mediated endocy- tosis of soluble FBP or by binding of free folates to membrane-associated folate receptors is released within the same endocytic compartment and may be further transported by a common pathway. However, while the GPI-linked FBP is recycled to the plasma membrane, FBP bound to megalin seems to be degraded. In conclusion megalin is able to bind and internalize soluble FBP by endocytosis providing a potential mechanism for intestinal and renal uptake of soluble FBP-bound folate in milk or the ultrafiltrate. Further studies are needed to determine the importance of folate uptake via this pathway. Experimental procedures Purification of proteins High-affinity FBPs were purified from bovine and human milk by a combination of cation exchange chromatography and ligand (methotrexate) chromatography on a column desorbed with a pH-gradient [9,35,36]. Bovine FBP was purified from cow’s whey powder and consisted of 222 amino-acid residues with a molecular mass of 30 kDa based on amino-acid composition and carbohydrate content [37]. Two FBPs were purified from Triton X-100 solubilized human raw milk obtained from voluntary donors with their full understanding and full consent. One desorbed at pH 5 had a molecular mass of 30 kDa based on its amino-acid composition and sequence homology with bovine milk FBP [9], while one desorbed at pH 3 was GPI-linked and extre- mely hydrophobic only existing in a micellar form with Triton X-100 [36]. Both FBPs had identical N-terminal sequence for 39 cycles and were immunologically identical [9,36]. The enzyme phosphatidyl inositol specific phospholi- pase C cleaved the hydrophobic GPI residue of micellar FBP and converted it to soluble FBP [38]. Human and bovine milk FBP was iodinated by the chloramine-T method [39]. Receptor associated protein, an established inhibitor of the uptake of most ligands by megalin [14], was prepared Fig. 4. Uptake of soluble 125 I-labeled FBP in megalin expressing BN-16 cells (percent of total added activity). Megalin expressing, yolk sac BN-16 epithelial cells were grown to confluence and incu- bated with 125 I-labeled human milk FBP [9000 counts per minute (CPM)]. Total uptake represents the sum of degraded (non-trichloro- acetic acid-precipitable) and cell-associated activity. For inhibition studies cells were coincubated with either RAP (1 l M, A), sheep anti-megalin IgG (200 lgÆmL )1 , B), or nonspecific sheep IgG (200 lgÆmL )1 , B). Labeled FBP is internalized and degraded in BN-16 cells (A). Uptake is significantly inhibited by RAP (A), and anti-megalin IgG (B). A minor, however, significant inhibition with nonspecific IgG was observed at 4 h. Data represent mean ± SD of four experiments. In panel A the RAP induced difference in total uptake was significant (P < 0.001, t-test) at all time points. In (B), significance is indicated by * (ANOVA followed by posthoc t-test using Bonferroni correction at each time point). H. Birn et al. Megalin-mediated uptake of FBP FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS 4427 and used for purification of rabbit megalin by affinity chromatography as described [40]. LRP was purified from solubilized human placental membranes as described [41]. Binding of milk FBP to megalin by surface plasmon resonance analysis (SPR) For the surface plasmon resonance analyses, the BIAcore sensor chips (type CM5; Biosensor, Uppsala, Sweden) were activated with a 1 : 1 mixture of 0.2 m N-ethyl-N¢-(3-dimeth- ylaminopropyl) carbodiimide and 0.05 m N-hydroxysuccini- mide in water according to the manufacturer. Purified megalin or LRP were immobilized on the sensor chip in 10 mm sodium acetate, pH 4.5, and the remaining binding sites were blocked with 1 m ethanolamine, pH 8.5. The resulting receptor densities were in the range of 23–40 fmol receptor per mm 2 . A control flow cell was made by perform- ing the activation and blocking procedures and by using immobilized receptor proteins reduced by injection of 0.5% dithiothreitol in 6 m guanidine hydrochloride, 5 mm EDTA, and 50 m m Tris, pH 8.0, into the flow cell. Purified FBP, with or without the addition of 2.5 lm folic acid, 5-MTHF or methotrexate was dissolved in 10 mm Hepes, 150 mm NaCl, 2 mm CaCl 2 , and 0.005% Tween 20, pH 7.4. Sample and running buffers were identical. The regeneration of sen- sor chips after each analysis cycle was performed with 1.6 m glycine-HCl buffer, pH 3.0. The BIAcore response is expressed in relative response units and represents the bind- ing response using the native receptor corrected for the response registered with the control flow cell. K d for binding was estimated using a biaevaluation program. Binding of milk 125 I-labeled FBP to tissue cryosections Binding of FBP to megalin in tissue was studied by auto- radiography on 1 lm cryosections of rat kidney cortex, fixed in 4% (v ⁄ v) paraformaldehyde in 0.1 m sodium cacodylate buffer, pH 7.4, by retrograde perfusion through the abdom- inal aorta. Sections were cut at 190–200 K using a Reichert Ultracut S cryoultramicrotome and placed on gelatin coated glass slides, preincubated in 0.01 m NaCl ⁄ P i , 0.05 m glycine, 0.15 m NaCl, 0.1% (w ⁄ v) skimmed milk and 0.02 m NaN 3 and incubated with 125 I-labeled human or bovine milk FBP (2Æ 10 6 CPMÆmL )1 ) in 0.01 m NaCl ⁄ P i , 0.05 m Tris buffer, 0.15 m NaCl, 1 mm CaCl 2 , 0,1% BSA and 0.02 NaN 3 . For inhibition studies excess unlabeled bovine milk FBP (10 lm) or receptor associated protein (RAP, 1.2 lm), was added to the 125 I-labeled FBP incubation buffer. Sections were washed, fixed in 1% (v ⁄ v) glutaraldehyde in 0.1 m sodium cacodylate buffer, pH ¼ 7.4 and prepared for light micro- scope autoradiography using Ilford emulsion. After 8 days of exposure the sections were developed and observed in a Leica LMR microscope (Wetzlar, Germany). Uptake of FBP in kidney proximal tubule and cultured yolk sac cells To study uptake 125 I-labeled bovine milk FBP was micro- injected into kidney proximal tubules from anaesthetized male Wistar rats (207–237 g) placed on a thermostatically controlled heated table. A tracheostomy was performed, and the jugular vein was catherized and infused with saline, 3.8 mLÆh )1 . The left kidney was exposed by flank incision, placed in a stabilized cup and covered with paraffin oil maintained at 37–38 °C. Single surface proximal tubules were injected with 52 nL of 125 I-labeled FBP in 0.15 m NaCl, 1 mm CaCl 2 and lissamine green and fixed by micro- injection of 1% glutaraldehyde 15–30 min after microinjec- tion with 125 I-labeled FBP. Small tissue blocks containing the microinfused tubules were postfixed, dehydrated and embedded into Epon 812. Sections were processed for light microscope or electron microscope autoradiography using Ilford emulsion K2 or L4, respectively, and observed in a Leica LMR microscope or Philips EM208 or CM100 elec- tron microscope (Eindhoven, the Netherlands). All animal experiments were carried out to minimize pain and discom- fort and in accordance with the provisions for the animal care license provided by the Danish National Animal Experiments Inspectorate. In addition, megalin-mediated uptake was studied in a rat yolk sac BN-16 epithelial cell-line previously shown to express megalin and representing an established model for megalin-mediated uptake [19,20,42]. Cells were grown to confluence in 24 well cell culture plates using RPMI 1640 medium (Life Technologies, Gaithersburg, MD, USA) with 5% (v ⁄ v) fetal bovine serum, 50 UÆmL )1 penicillin, and 50 lgÆmL )1 streptomycin (Bio-Whittaker, Wokingham, UK). At confluence cells were incubated with 125 I-labeled human milk FBP ( 9000 CPM) in 0.5 mL RPMI 1640 with 0.1% fetal bovine serum at 37 °C. Following 2, 4 or 8 h incu- bation, the medium was recovered, cells were washed once in warm medium, and harvested by trypsinization for 20 min. An equal volume of 1% (w ⁄ v) BSA solution was added to the collected medium along with the washing medium and followed by precipitation with 10% (v ⁄ v) trichloroacetic acid. The activity of the precipitate, the supernatant, and the cells was counted separately in a Packard Cobra 5002 gamma-counter. The degraded amount of 125 I-labeled FBP in the medium was estimated as non-trichloroacetic acid pre- cipitated activity in the medium corrected for the non-trichlo- roacetic acid precipitated activity in the medium of wells incubated for the same time without cells. Total uptake of 125 I-labeled FBP was calculated as the sum of cell associated activity and degraded 125 I-labeled FBP in the medium and expressed in percent of total activity added. For inhibition studies cells were coincubated with either RAP (1 lm), purified sheep anti-rat megalin IgG (200 lgÆmL )1 [43]), or purified, nonspecific sheep IgG (200 lgÆmL )1 , DAKO, Megalin-mediated uptake of FBP H. Birn et al. 4428 FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS Denmark). Antibodies were purified by protein A-agarose affinity chromatography according to manufacturer’s instructions (Pierce, Rockford, IL, USA). Data represent mean ± SD of four experiments, and statistical analysis was performed using unpaired t-test or ANOVA. Immunocytochemistry Normal Wistar rats were fixed by retrograde fixation through the abdominal aorta using 4% (v ⁄ v) paraformalde- hyde. For light microscope immunocytochemistry semithin cryosections were cut as described above. Sections were incubated 1 h with polyclonal sheep anti-rat megalin IgG (1 : 50 000 [43]), in 10 mm NaCl ⁄ P i , 0.15 m NaCl, 0.1% skimmed milk and 20 mm NaN 3 , followed by HRP-conju- gated goat anti-sheep IgG, and visualization by incubation with diaminobenzidine and 0.03% (v ⁄ v) H 2 O 2 for 10 min. All incubations were performed at room temperature and sections were counterstained with Meiers before examina- tion in the light microscope as described above. Acknowledgements The work was supported in part by the Danish Medical Research Council, the University of Aarhus, the NOVO-Nordisk Foundation, Fonden til Lægevidenska- bens Fremme, the Biomembrane Research Center, and the Birn-Foundation. The skillful technical assistance by Pia K. Nielsen, Hanne Sidelmann, and Inger Kristoffer- sen is greatly appreciated. The study was in part presen- ted at the ASN Annual Meeting, Philadelphia, PA, November 1–4, 2002 and at the 13th International Sym- posium on Chemistry & Biology of Pteridines & Folates, Egmond aan Zee, the Netherlands, June 20–24, 2005, and published in part as abstract. References 1 Antony AC (1996) Folate receptors. Annu Rev Nutr 16, 501–521. 2 Henderson GB (1990) Folate-binding proteins. Annu Rev Nutr 10, 319–335. 3 Matherly LH & Goldman DI (2003) Membrane trans- port of folates. Vitam Horm 66, 403–456. 4 Ghitis J (1967) The folate binding in milk. Am J Clin Nutr 20, 1–4. 5 Salter DN, Ford JE, Scott KJ & Andrews P (1972) Iso- lation of the folate-binding protein from cow’s milk by the use of affinity chromatography. FEBS Lett 20, 302– 306. 6 Selhub J, Arnold R, Smith AM & Piccano MF (1984) Milk folate binding protein (FBP): a secretory protein for folate? Nutr Res 4, 181–187. 7 Verma RS & Antony AC (1992) Immunoreactive folate-binding proteins from human saliva. Isolation and comparison of two distinct species. Biochem J 286, 707–715. 8 Holm J, Hansen SI & Høier Madsen M (1991) A high- affinity folate binding protein in human semen. Biosci Report 11, 237–242. 9 Svendsen I, Hansen SI, Holm J & Lyngbye J (1982) Amino acid sequence homology between human and bovine low molecular weight folate binding protein iso- lated from milk. Carlsberg Res Comnun 47 , 371–376. 10 Colman N, Hettiarachchy N & Herbert V (1981) Detec- tion of a milk factor that facilitates folate uptake by intestinal cells. Science 211, 1427–1429. 11 Salter DN & Blakeborough P (1988) Influence of goat’s-milk folate-binding protein on transport of 5-methyltetrahydrofolate in neonatal-goat small intestinal brush-border-membrane vesicles. Br J Nutr 59, 497–507. 12 Mason JB & Selhub J (1988) Folate-binding protein and the absorption of folic acid in the small intestine of the suckling rat. Am J Clin Nutr 48, 620–625. 13 Ford JE (1974) Some observations on the possible nutri- tional significance of vitamin B12-and folate-binding proteins in milk. Br J Nutr 31, 243–257. 14 Christensen EI & Birn H (2002) Megalin and cubilin: multifunctional endocytic receptors. Nat Rev Mol Cell Biol 3, 256–266. 15 Saito A, Pietromonaco S, Loo AK & Farquhar MG (1994) Complete cloning and sequencing of rat gp330 ⁄ ’megalin’, a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci USA 91, 9725–9729. 16 Birn H, Verroust PJ, Nexø E, Hager H, Jacobsen C, Christensen EI & Moestrup SK (1997) Characterization of an epithelial 460 kDa protein that facilitates endocy- tosis of intrinsic factor-vitamin B 12 and binds receptor- associated protein. J Biol Chem 272, 26497–26504. 17 Yammani RR, Seetharam S & Seetharam B (2001) Cubilin and megalin expression and their interaction in the rat intestine: effect of thyroidectomy. Am J Physiol Endocrinol Metab 281, E900–E907. 18 Christensen EI & Willnow TE (1999) Essential role of megalin in renal proximal tubule for vitamin homeo- stasis. J Am Soc Nephrol 10, 2224–2236. 19 Christensen EI, Moskaug JO, Vorum H, Jacobsen C, Gundersen TE, Nykjær A, Blomhoff R, Willnow TE & Moestrup SK (1999) Evidence for an essential role of megalin in transepithelial transport of retinol. J Am Soc Nephrol 10, 685–695. 20 Moestrup SK, Birn H, Fischer PB, Petersen CM, Verroust PJ, Sim RB, Christensen EI & Nexø E (1996) Megalin-mediated endocytosis of transcobalamin-vita- min-B 12 complexes suggests a role of the receptor in H. Birn et al. Megalin-mediated uptake of FBP FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS 4429 vitamin-B 12 homeostasis. Proc Natl Acad Sci USA 93, 8612–8617. 21 Nykjær A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F, Christensen EI & Willnow TE (1999) An endocytic pathway essential for renal uptake and acti- vation of the steroid 25-(OH) vitamin D 3 . Cell 96, 507– 515. 22 Nykjær A, Fyfe JC, Kozyraki R, Leheste JR, Jacobsen C, Nielsen MS, Verroust PJ, Aminoff M, Chapelle A, Moestrup SK, Ray R, Gliemann J & Christensen EI (2001) Cubilin dysfunction causes abnormal metabolism of the steroid hormone 25-(OH) vitamin D 3 . Proc Natl Acad Sci USA 98, 13895–13900. 23 Christensen EI & Birn H (2001) Megalin and cubilin: synergistic endocytic receptors in renal proximal tubule. Am J Physiol Renal Physiol 280, F562–F573. 24 Moestrup SK, Kozyraki R, Kristiansen M, Kaysen JH, Rasmussen HH, Brault D, Pontillon F, Goda FO, Christensen EI, Hammond TG & Verroust PJ (1998) The intrinsic factor-vitamin B12 receptor and target of teratogenic antibodies is a megalin-binding peripheral membrane protein with homology to developmental proteins. J Biol Chem 273, 5235–5242. 25 Willnow TE, Nykjaer A & Herz J (1999) Lipoprotein receptors: new roles for ancient proteins. Nat Cell Biol 1, E157–E162. 26 Birn H, Lebeolleux M, Moestrup SK, Ronco P, Aucou- turier P & Christensen EI (2002) Receptor mediated uptake of light chains in kidney proximal tubules. In Monoclonal Gammapathies and the Kidney (Aucouturier P, Hermine P, Ronco P & Touchard G, eds). Klu ¨ wer, the Netherlands. 27 Salter DN & Mowlem A (1983) Neonatal role of milk folate-binding protein: studies on the course of digestion of goat’s milk folate binder in the 6-d-old kid. Br J Nutr 50, 589–596. 28 Holm J, Hansen SI & Lyngbye J (1980) High-affinity binding of folate to a protein in serum of male subjects. Clin Chim Acta 100, 113–119. 29 Hansen SI, Holm J & Lyngbye J (1980) High-affinity protein binding of folate in urine. IRCS Med Sci 8, 846–847. 30 Selhub J, Emmanouel D, Stavropoulos T & Arnold R (1987) Renal folate absorption and the kidney folate binding protein. I. Urinary clearance studies. Am J Physiol 252, F750–F756. 31 Birn H, Spiegelstein O, Christensen E, I & Finnell RH (2005) Renal tubular reabsorption of folate mediated by folate binding protein 1. J Am Soc Nephrol 16, 608–615. 32 Hjelle JT, Christensen EI, Carone FA & Selhub J (1991) Cell fractionation and electron microscope stu- dies of kidney folate-binding protein. Am J Physiol 260, C338–C346. 33 Birn H, Selhub J & Christensen EI (1993) Internalization and intracellular transport of folate-binding protein in rat kidney proximal tubule. Am J Physiol 264, C302–C310. 34 Birn H, Nielsen S & Christensen EI (1997) Internaliza- tion and apical to basolateral transport of folate in rat kidney proximal tubule. Am J Physiol 272, F70–F78. 35 Svendsen I, Martin B, Pedersen TG, Hansen SI, Holm J & Lyngbye J (1979) Isolation and characterization of the folate-binding protein from cows milk. Carlsberg Res Comnun 44, 89–99. 36 Holm J, Hansen SI & Hoier-Madsen M (2002) A com- bination of cation exchange and ligand-affinity chroma- tography for purification of two molecular species of the folate binding protein in human milk, one equipped with a hydrophobic glycosyl phosphatidylinositol tail: characterization of hydrophobicity and electrical charge. Biosci Report 22, 443–454. 37 Svendsen I, Hansen SI, Holm J & Lyngbye J. (1984) The complete amino acid sequence of the folate binding protein from cow’s milk. Carlsberg Res Commun 49, 123–131. 38 Hansen, SI & Holm, J (1992) Conversion of an appar- ent 100 kDa folate binding protein from human milk, choroid plexus and semen to a 25 kDa molecular species by phosphatidylinositol-specific phospholipase C. Biosci Report 12, 87–93. 39 Greenwood FC, Hunter WM & Glover JS (1963) The preparation of I-131-labelled human growth hormone of high specific radioactivity. Biochem J 89, 114–123. 40 Moestrup SK, Nielsen S, Andreasen P, Jørgensen KE, Nykjær A, Roigaard H, Gliemann J & Christensen EI (1993) Epithelial glycoprotein-330 mediates endo- cytosis of plasminogen activator-plasminogen activator inhibitor type-1 complexes. J Biol Chem 268, 16564– 16570. 41 Moestrup SK, Kaltoft K, Sottrup-Jensen L & Gliemann J (1990) The human alpha 2-macroglobulin receptor contains high affinity calcium binding sites important for receptor conformation and ligand recognition. J Biol Chem 265, 12623–12628. 42 Le Panse S, Galceran M, Pontillon F, Lelongt B, van de Putte M, Ronco PM & Verroust PJ (1995) Immuno- functional properties of a yolk sac epithelial cell line expressing two proteins gp280 and gp330 of the inter- microvillar area of proximal tubule cells: inhibition of endocytosis by the specific antibodies. Eur J Cell Biol 67, 120–129. 43 Sahali D, Mulliez N, Chatelet F, Laurent Winter C, Citadelle D, Sabourin JC, Roux C, Ronco P & Verroust P (1993) Comparative immunochemistry and ontogeny of two closely related coated pit proteins. The 280-kDa target of teratogenic antibodies and the 330-kDa target of nephritogenic antibodies. Am J Pathol 142, 1654– 1667. Megalin-mediated uptake of FBP H. Birn et al. 4430 FEBS Journal 272 (2005) 4423–4430 ª 2005 FEBS . Megalin binds and mediates cellular internalization of folate binding protein Henrik Birn 1 , Xiaoyue Zhai 1 , Jan Holm 2 , Steen I. Hansen 2 , Christian Jacobsen 3 , Erik I. Christensen 1 and. number of vitamin carrier proteins [18], including retinol binding protein [19], transcobalamin [20], and vitamin D binding protein [21,22] and its role for the recovery of vitamins and carrier proteins. stu- dies of kidney folate- binding protein. Am J Physiol 260, C338–C346. 33 Birn H, Selhub J & Christensen EI (1993) Internalization and intracellular transport of folate- binding protein in