Mutually-dependent localization of megalin and ... - Renal Physiology
Mutually-dependent localization of megalin and ... - Renal Physiology
Mutually-dependent localization of megalin and ... - Renal Physiology
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Articles in PresS. Am J Physiol <strong>Renal</strong> Physiol (May 3, 2005). doi:10.1152/ajprenal.00292.2004<br />
<strong>Mutually</strong>-<strong>dependent</strong> <strong>localization</strong> <strong>of</strong> <strong>megalin</strong> <strong>and</strong> Dab2 in the renal<br />
proximal tubule<br />
Nagai J 1,2 , Christensen EI 2 , Morris SM 3 , Willnow TE 4 , Cooper JA 3 , Nielsen R 2<br />
1 Department <strong>of</strong> Pharmaceutics <strong>and</strong> Therapeutics, Hiroshima University, Hiroshima 734-8551, Japan, 2<br />
Cell Biology, Institute <strong>of</strong> Anatomy, University <strong>of</strong> Aarhus, 8000 Aarhus, Denmark, 3 Fred Hutchinson Cancer<br />
Research Center, Seattle, Washington 98109-1024, USA, <strong>and</strong> 4 Max-Delbrueck-Center for Molecular<br />
Medicine, 13125 Berlin, Germany<br />
Running title: Dab2 <strong>localization</strong> in the renal proximal tubule<br />
Correspondence: Rikke Nielsen M.Sc., Ph.D.<br />
Cell Biology, Institute <strong>of</strong> Anatomy, University <strong>of</strong> Aarhus,<br />
University Park, Building 234, DK-8000 Aarhus C,<br />
Denmark<br />
Telephone number: +45-8942-3030<br />
FAX number: +45-8619-8664<br />
E-mail address: rn@ana.au.dk<br />
Copyright © 2005 by the American Physiological Society.
Abstract<br />
Disabled-2 (Dab2) is a cytoplasmic adaptor protein that binds to the cytoplasmic tail <strong>of</strong><br />
the multilig<strong>and</strong> endocytic receptor <strong>megalin</strong>, abundantly expressed in renal proximal<br />
tubules. Deletion <strong>of</strong> Dab2 induces urinary increase in specific plasma proteins such as<br />
vitamin D binding protein <strong>and</strong> retinol binding protein (Morris et al. EMBO J., 21: 1555-<br />
1564, 2002). However, the subcellular <strong>localization</strong> <strong>of</strong> Dab2 in the renal proximal tubule<br />
<strong>and</strong> its function has not been fully elucidated yet. Here we report the characterization<br />
<strong>of</strong> Dab2 in the renal proximal tubule. Immunohistocytochemistry revealed co-<br />
<strong>localization</strong> with <strong>megalin</strong> in coated- pits <strong>and</strong> vesicles, but not in dense apical tubules<br />
<strong>and</strong> the brush border. Kidney-specific <strong>megalin</strong> knock-out almost abolished Dab2<br />
staining, indicating that Dab2 subcellular <strong>localization</strong> requires <strong>megalin</strong> in the proximal<br />
tubule. Reciprocally, knock-out <strong>of</strong> Dab2 led to a redistribution <strong>of</strong> <strong>megalin</strong> from<br />
endosomes to microvilli. In addition, there was an overall decrease in levels <strong>of</strong> <strong>megalin</strong><br />
protein observed by immunoblotting, but no decrease in clathrin or -adaptin protein<br />
levels or in <strong>megalin</strong> mRNA. In rat yolk sac epithelial BN16 cells Dab2 was present<br />
apically <strong>and</strong> co-localized with <strong>megalin</strong>. Introduction <strong>of</strong> anti-Dab2 antibody into BN16<br />
cells decreased the internalization <strong>of</strong> 125 I-RAP, substantiating the role <strong>of</strong> Dab2 in<br />
<strong>megalin</strong>-mediated endocytosis. The present study shows that Dab2 is localized in the<br />
apical endocytic apparatus <strong>of</strong> the renal proximal tubule <strong>and</strong> that this <strong>localization</strong><br />
requires <strong>megalin</strong>. Furthermore, the study suggests that the urinary loss <strong>of</strong> <strong>megalin</strong><br />
lig<strong>and</strong>s observed in Dab2 knock-out mice is caused by suboptimal trafficking <strong>of</strong> <strong>megalin</strong><br />
leading to decreased <strong>megalin</strong> levels.<br />
Keywords: receptor mediated endocytosis, kidney, adaptor proteins
Introduction<br />
Recent research has demonstrated that <strong>megalin</strong> (35) is one <strong>of</strong> the most important<br />
receptors for protein reabsorption in the renal proximal tubules (reviewed in 4,5). The<br />
giant endocytic receptor (600 kDa) is abundantly expressed in the endocytic apparatus<br />
<strong>of</strong> the renal proximal tubule, including the brush-border, clathrin-coated pits, clathrin-<br />
coated vesicles <strong>and</strong> dense apical tubules. Its lig<strong>and</strong>s include Ca 2+ , vitamin binding<br />
proteins, apolipoproteins, hormones, enzymes, <strong>and</strong> drugs as well as receptor-<br />
associated protein (RAP), a chaperone for the low density lipoprotein receptor (LDLR)<br />
gene family (38). Megalin has a large amino-terminal extracellular region, a single<br />
transmembrane domain <strong>and</strong> a short carboxy-terminal cytoplasmic tail (11,35). The<br />
extracellular region contains four clusters <strong>of</strong> cysteine-rich complement-type repeats<br />
forming the lig<strong>and</strong>-binding regions characteristic <strong>of</strong> the LDLR gene family. The<br />
sequence <strong>of</strong> the cytoplasmic tail has little similarity to that <strong>of</strong> other members <strong>of</strong> the<br />
LDLR gene family except for the NPXY motifs, which are known to serve as the sorting<br />
signal for rapid endocytosis <strong>of</strong> the LDLR via clathrin-coated pits (2,11,35).<br />
Dab2 was isolated as a phosphoprotein involved in colony stimulating factor-1<br />
signal transduction (40). Interestingly, Dab2 expression is down-regulated during<br />
prostate degeneration <strong>and</strong> in many human carcinomas <strong>and</strong> accordingly is also called<br />
DOC-2, deleted in ovarian cancer-2 (6,25,26,37). In addition, Dab1, a related protein<br />
expressed predominantly in the brain, binds to the NPXY motif(s) in the cytoplasmic tail<br />
<strong>of</strong> very low density lipoprotein receptor (VLDLR) <strong>and</strong> apolipoprotein E receptor2<br />
(ApoER2) through the phosphotyrosine binding (PTB) domain <strong>of</strong> Dab1 (9,12-16,22).<br />
This interaction between the members <strong>of</strong> LDLR gene family <strong>and</strong> Dab1 plays an<br />
important role in neuronal positioning (12,13,15). Thus, some <strong>of</strong> the members <strong>of</strong> LDLR<br />
gene family might serve as transmembrane signal transduction molecules (9,16,32,39).<br />
Much attention has been focused on the involvement <strong>of</strong> the cytoplasmic tail<br />
domain <strong>of</strong> <strong>megalin</strong> in subcellular <strong>localization</strong>, endocytic trafficking <strong>and</strong> signaling (5,32-
34,36,39). The cytoplasmic tail domain <strong>of</strong> <strong>megalin</strong> has binding sites for some<br />
intracellular protein-protein interaction domains including Src homology 3 (SH3) binding<br />
motifs (PXXP) <strong>and</strong> NPXY motifs, which bind to SH3 domains <strong>and</strong> PTB domains,<br />
respectively (11,35). Recently, it has been found that Dab2 binds to the cytoplasmic tail<br />
<strong>of</strong> <strong>megalin</strong> by using in vitro methods such as co-immunoprecipitation <strong>and</strong> binding<br />
assays (33). Furthermore, Morris et al. (29) observed that the deletion <strong>of</strong> the Dab2<br />
gene in mice induced an increase in urinary excretion <strong>of</strong> vitamin D binding protein<br />
(DBP) <strong>and</strong> retinol binding protein (RBP), lig<strong>and</strong>s <strong>of</strong> <strong>megalin</strong> (5). These observations<br />
may indicate an involvement <strong>of</strong> Dab2 in <strong>megalin</strong>-mediated endocytosis in renal proximal<br />
tubules under in vivo conditions. However, the precise <strong>localization</strong> <strong>of</strong> Dab2 in the renal<br />
proximal tubules has not been determined <strong>and</strong> the role <strong>of</strong> Dab2 in <strong>megalin</strong>-mediated<br />
endocytosis has not been fully characterized.<br />
The present study was performed to investigate the subcellular <strong>localization</strong> <strong>of</strong><br />
Dab2 in the renal proximal tubules by immunohistocytochemistry <strong>and</strong> to approach the<br />
mechanism underlying increased urinary excretion <strong>of</strong> <strong>megalin</strong> lig<strong>and</strong>s in Dab2 knock-<br />
out mice by investigation <strong>of</strong> Dab2 <strong>and</strong> <strong>megalin</strong> levels in knock-out mouse models.
Materials <strong>and</strong> Methods<br />
Animals<br />
Dab2 conditional knockout mice were prepared by breeding Dab2 fl/fl to Dab2 +/- Meox2<br />
cre/+ This circumvents the embryonic lethality caused by Dab2 deletion, <strong>and</strong> allows the<br />
generation <strong>of</strong> mice lacking Dab2 genes in most cells (29). Young adult Dab2 -/- <strong>and</strong><br />
Dab2 -/+ were anaesthetized, perfused with 4% paraformaldehyde in PBS, <strong>and</strong> kidneys<br />
were further processed (3).<br />
The generation <strong>of</strong> mice carrying a kidney-specific <strong>megalin</strong> gene deletion has<br />
been described before (19). Cre recombinase-mediated inactivation <strong>of</strong> the <strong>megalin</strong><br />
gene in the kidney results in a complete loss <strong>of</strong> receptor expression in approximately<br />
80-90% <strong>of</strong> all cells in the renal proximal tubule.<br />
Antibodies<br />
Rabbit anti-Dab2 polyclonal antibody (H-110) <strong>and</strong> goat anti-clathrin HC polyclonal<br />
antibody (C-20) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA,<br />
USA). Sheep anti-<strong>megalin</strong> polyclonal antibody <strong>and</strong> rabbit <strong>megalin</strong> polyclonal antibody<br />
were obtained as described previously (3,24,41). Mouse anti- -adaptin monoclonal<br />
antibody was from Affinity BioReagents Inc. (Golden, CO, USA). Horseradish<br />
peroxidase-conjugated goat anti-rabbit Ig (P448), rabbit anti-sheep Ig (P163), rabbit<br />
anti-goat Ig (P449), goat anti-mouse Ig (P447) <strong>and</strong> TRITC coupled to swine anti-rabbit<br />
Ig (R156) were from Dako A/S (Copenhagen, Denmark). Alexa Fluor488 donkey anti-<br />
sheep Ig (A11015) was obtained from Molecular Probes (Leiden, the Netherl<strong>and</strong>s).<br />
Immunocytochemistry<br />
For light microscopy, semithin (0.8 µm) cryosections were cut at –80 o C on an FCS<br />
Reichert Ultracut S cryomicrotome (Reichert-Jung, Vienna, Austria) <strong>and</strong> placed on
gelatin-coated glass slides. The sections were incubated with rabbit anti-Dab2 antibody<br />
(1:20 – 1:80) or sheep anti-<strong>megalin</strong> antibody (1:20,000) at room temperature for 1 hour<br />
following preincubation in 10 mM phosphate-buffered saline (PBS) containing 50 mM<br />
glycine <strong>and</strong> 0.1% skimmed milk. After incubation for 1 hour with horseradish<br />
peroxidase-conjugated secondary antibodies, the peroxidase was visualized with<br />
diaminobenzidine, <strong>and</strong> sections were counterstained with Meier’s hematoxyline stain<br />
<strong>and</strong> examined in a Leica DMR microscope equipped with a Sony 3CCD color video<br />
camera <strong>and</strong> a Sony Digital Still recorder (Sony Corp., Tokyo, Japan). For electron<br />
microscopy, ultrathin (70 to 90 nm) cryosections were obtained, <strong>and</strong> the sections were<br />
incubated with rabbit anti-Dab2 antibody (1:80 - 1:240) <strong>and</strong>/or sheep anti-<strong>megalin</strong><br />
antibody (1:20,000) followed by incubation with gold conjugated secondary antibodies<br />
(British BioCell International, Cardiff, UK). The cryosections were embedded in 2%<br />
methylcellulose containing 0.3% uranyl acetate <strong>and</strong> were examined in a Phillips CM<br />
100 electron microscope.<br />
Morphometric determination <strong>of</strong> <strong>megalin</strong> <strong>localization</strong> in kidneys from normal <strong>and</strong><br />
Dab2 knock-out mice<br />
Sections from control <strong>and</strong> Dab2 knock-out mice were prepared for electron microscopy<br />
as described above. The sections were stained with sheep anti <strong>megalin</strong> antibody<br />
1:20,000 <strong>and</strong> developed with a secondary antibody conjugated to 10 nm gold particles.<br />
Eight r<strong>and</strong>omized pictures were taken from 6 control <strong>and</strong> 6 knock-out mice at a<br />
magnifiacation <strong>of</strong> 25,000X. All pictures were from tubules which were sectioned in the<br />
longitudinal axis <strong>of</strong> the microvilli. Areas encompassing the microvilli <strong>and</strong> cytoplasm were<br />
measured as well as gold particles were counted by the use <strong>of</strong> AnalySIS.
Immunoblotting<br />
Crude kidney membranes for immunoblotting were prepared as described previously<br />
(1,21). Briefly, the excised kidneys were homogenized in an ice-cold buffer (0.3 M<br />
sucrose, 25 mM imidazole, 1 mM ethylenediaminetetraacetic acid (EDTA), 8.5 µM<br />
leupeptin, <strong>and</strong> 1 mM phenylmethylsulfonyl fluoride, pH 7.2) for 30 sec with an IKA<br />
ULTRA-TURRAX T8 homogenizer (IKA LABORTENIK, Germany). The homogenate<br />
was centrifuged at 4,000 g for 15 min at 4 o C. The supernatant was transferred to an<br />
ULTRA-CLEAR tube (Beckman Instruments, Fullerton, CA, USA) <strong>and</strong> centrifuged at<br />
17,000 g for 30 min at 4 o C in an L8-70 M ultracentrifuge (Beckman Instruments) with<br />
rotor 70.1 T1. The pellet was suspended in the ice-cold buffer used for the<br />
homogenization <strong>and</strong> was heated for 5 min at 95 o C in Laemmli sample buffer containing<br />
2.5% sodium dodecyl sulfate (SDS) <strong>and</strong> 1% 2-mercaptoethanol. The samples (5 µg <strong>of</strong><br />
protein) were run on 3 to 16% SDS polyacrylamide gradient gels <strong>and</strong> were then<br />
transferred to a nitrocellulose membrane for 60 min at 4 o C. Subsequently, the<br />
membrane was blocked in 5% skimmed milk in phosphate-buffered saline (PBS-T; 80<br />
mM Na2HPO4, 20 mM NaH2PO4, 100 mM NaCl, 0.1% Tween 20, pH 7.5) for 1 hour.<br />
The membranes were washed for 15 min in PBS-T followed by washing for 5 min twice<br />
<strong>and</strong> incubated overnight at 4 o C with primary antibody in PBS-T with 1% bovine serum<br />
albumin (BSA) [rabbit anti-Dab2 antibody (1:100), rabbit anti-rat <strong>megalin</strong> antibody<br />
(1:1,000), goat anti-clathrin HC antibody (1:400) or mouse anti- -adaptin antibody<br />
(1:100)]. After washing for 15 min in PBS-T followed by washing twice for 5 min, the<br />
blots were incubated for 1 hour with the horseradish peroxidase-conjugated secondary<br />
antibody, washed 3 times in PBS-T, <strong>and</strong> visualized with enhanced chemiluminescence<br />
(ECL; Amersham International, Buckinghamshire, United Kingdom).
Cell culture<br />
Rat yolk sac carcinoma cells BN16 were cultured as described previously (7,20).<br />
Briefly, BN16 cells were grown in 25 cm 2 plastic culture flasks (Corning Costar,<br />
Badhoevedrop, Holl<strong>and</strong>), in Eagle’s Minimal Essential Medium (Bio-Whittaker,<br />
Welkersville, MD, USA) supplemented with 10% fetal calf serum (Biological Industries,<br />
Fredensborg, Denmark), 2 mM L-glutamine, 50 U/ml penicillin, <strong>and</strong> 50 µg/ml<br />
streptomycin (Bio-Whittaker) in a humidified atmosphere <strong>of</strong> 5% CO2 <strong>and</strong> 95% air at 37 o<br />
C. The cells were subcultured every fourth day with a split ratio <strong>of</strong> 1:5 by using 0.02%<br />
EDTA <strong>and</strong> 0.05% trypsin (Bio-Whittaker). Experiments were carried out with confluent<br />
monolayers <strong>of</strong> BN16 cells cultured in 24-well plates (Nagle Nunc International) for<br />
quantitative uptake studies.<br />
Immun<strong>of</strong>luorescence microscopy<br />
Co-<strong>localization</strong> <strong>of</strong> Dab2 <strong>and</strong> <strong>megalin</strong>. Immunohistochemistry was performed on cells<br />
grown confluent in flasks, fixed in 2% paraformaldehyde in 0.1 M cacodylate (Cac)<br />
buffer, washed 3 times in Cac buffer <strong>and</strong> scraped <strong>of</strong>f in 1% gelatin in Cac buffer, 37º C.<br />
The cells were centrifuged twice in 1% gelatin <strong>and</strong> finally a few drops <strong>of</strong> 12% gelatin<br />
were added. The gelatin containing the cells was cooled on ice <strong>and</strong> infiltrated with 2.3 M<br />
sucrose in 0.01 M PBS, pH 7,4 for two hours. The blocks were frozen on nails. Sections<br />
were made <strong>and</strong> labeled with a mixture <strong>of</strong> primary antibodies: rabbit polyclonal anti-Dab2<br />
1:10 (H110) <strong>and</strong> sheep polyclonal anti-rat <strong>megalin</strong> 1:4,000 <strong>and</strong> then incubated with a<br />
mixture <strong>of</strong> secondary antibodies: TRITC coupled to swine anti-rabbit Ig 1:20 <strong>and</strong> Alexa<br />
Fluor488 donkey anti-sheep IgG(H+L) 1:300).<br />
Effect <strong>of</strong> Anti-Dab2 antibody transfection on iodinated RAP uptake. Anti-Dab2 antibody<br />
was transfected into BN16 cells with the BioPorter protein delivery reagent from<br />
BioCarta (CA, USA). Briefly, BN16 cells cultured on 24-well plate for 3 days were<br />
incubated with the protein delivery reagent (5 µl) containing rabbit polyclonal anti-Dab2
antibody (H-110, 1 µg <strong>of</strong> protein) in a 5% CO2 incubator at 37 o C for 1 hour. As control<br />
the same amount <strong>of</strong> rabbit Ig (X903, DAKO A/S, Denmark) was used. After washing<br />
three times, the cells were preincubated with serum-free EMEM containing 0.5%<br />
ovalbumin for 10 min at 37 o C <strong>and</strong> then were incubated with the above medium<br />
including iodinated RAP for 30 min in the 5% CO2 incubator. At the end <strong>of</strong> the<br />
incubation, the uptake buffer was removed <strong>and</strong> 200 µl <strong>of</strong> 0.1 M NaOH was added to<br />
dissolve the cells. The amount <strong>of</strong> the lig<strong>and</strong> taken up by the cells was measured by<br />
counting the radioactivity. Protein was determined by the Bio-Rad Protein Assay ® with<br />
bovine serum albumin as the st<strong>and</strong>ard.<br />
Northern blotting<br />
Dig-labeled RNA probes for <strong>megalin</strong> <strong>and</strong> actin were produced by RT-PCR <strong>and</strong> in vitro<br />
transcription. RT reaction: Total RNA was purified from mouse kidney cortex by TRIzol<br />
reagent (InVitrogen, CA). Purified RNA (0.1 µg) was incubated in a reverse<br />
transcriptase reaction mixture for 30 min at 42° C in a Genius thermocycler (Techne).<br />
PCR reaction: The PCR reaction was performed with Hotstartaq master mix kit (Qiagen,<br />
UK), 1 µl <strong>of</strong> RT reaction <strong>and</strong> 10 pmol <strong>of</strong> each primer. The identity <strong>of</strong> the products was<br />
confirmed by Big Dye termination sequencing <strong>and</strong> analyzed on an ABI PRISM 310<br />
Genetic Analyzer (Perkin Elmer, USA). Megalin primers had the following sequences<br />
<strong>and</strong> <strong>localization</strong> in the rat sequence (gi 561852): sense 5’ cacaggtgactgtaccagaa 3’,<br />
position 13762-13781; antisense 5’ gtgagtgtctaaatgttccc 3’, position 14141-14122 giving<br />
a 380 bp product. The PCR reaction was conducted with an annealing temperature <strong>of</strong><br />
60° C. Actin primers had the following sequences <strong>and</strong> <strong>localization</strong> in the rat sequence<br />
(gi 57573): sense 5’ tacgtgggtgatgaggccca 3’, position 180-199; antisense 5’<br />
tagccctcatagatgggcac 3’, position 529-510 giving a 350 bp product. The PCR reaction<br />
was conducted with annealing temperature <strong>of</strong> 65° C. PCR products containing the T7<br />
promoter sequence were produced from the purified PCR products <strong>and</strong> the primers
described above except that the sense primer included a T7 sequence in the 5’ end.<br />
The T7 DNA was used for in vitro transcription in a mixture containing dig -11-UTP<br />
(Roche) to produce the dig labeled RNA probes.<br />
Blotting: 10 µg <strong>of</strong> total RNA was separated on a 1% agarose-2% formaldehyde gel.<br />
The loaded amount <strong>of</strong> RNA was adjusted according to the optical density <strong>of</strong> the purified<br />
RNA. After electrophoresis the integrity <strong>of</strong> the RNA was confirmed <strong>and</strong> the RNA was<br />
transferred to a nylon membrane (Roche) by capillary blotting <strong>and</strong> finally the RNA was<br />
linked to the membrane by UV-linking (Hoefer UVC 500). The blots were prehybridized<br />
<strong>and</strong> hybridized with dig-labeled RNA probes (<strong>megalin</strong> 20-24 µg, actin 8 µg) over night<br />
at 68° C. Afterwards the blots were washed <strong>and</strong> blocked. Anti-digoxin-AP conjugate<br />
(SIGMA) 1:10,000 was applied <strong>and</strong> incubated with CSPD (Roche). The developed<br />
b<strong>and</strong>s were quantified by the Fluor-S TM MultiImager system (BioRad). The intensity <strong>of</strong><br />
the actin b<strong>and</strong>s was used to st<strong>and</strong>ardize the <strong>megalin</strong> b<strong>and</strong>s.
Results<br />
Localization <strong>of</strong> Dab2 in the endocytic apparatus <strong>of</strong> the renal proximal tubule<br />
In kidneys from Dab2 knock-out mice, Dab2 labeling could not be detected in proximal<br />
tubules by light microscopic immunohistochemistry confirming the complete loss <strong>of</strong><br />
Dab2 <strong>and</strong> the specificity <strong>of</strong> the antibody used for the detection <strong>of</strong> Dab2 (Figure 1A).<br />
Kidneys from control mice showed a b<strong>and</strong> <strong>of</strong> Dab2 labeling below the brush-border as<br />
well as labeling <strong>of</strong> endocytic vacuoles were observed in the cytoplasm <strong>of</strong> the renal<br />
proximal tubules from control mice (Figure 1B). Electron microscopic<br />
immunocytochemistry <strong>of</strong> cryosections from control mice confirmed labeling <strong>of</strong> the<br />
endocytic apparatus. More specifically, labeling was detected in apical coated pits <strong>and</strong><br />
apical vesicles, but virtually not in dense apical tubules (DAT, defined as membrane<br />
limited longitudinal structures with diameters slightly smaller than the diameter <strong>of</strong> a<br />
microvillus) (Figure 2).<br />
Absence <strong>of</strong> Dab2 in <strong>megalin</strong> deficient cells from kidney specific <strong>megalin</strong> knock-<br />
out mice<br />
Kidney-specific <strong>megalin</strong> knock-out mice lack <strong>megalin</strong> in most but not all kidney cells<br />
(19). Investigation <strong>of</strong> Dab2 in kidney sections from these mice by fluorescence<br />
microscopy revealed Dab2 only in cells that still expressed <strong>megalin</strong> (Figure 3). Most <strong>of</strong><br />
the proximal tubule cells apparently lacked both <strong>megalin</strong> <strong>and</strong> Dab2. Immunoelectron<br />
microscopy for <strong>megalin</strong> <strong>and</strong> Dab2 also demonstrated that Dab2 labeling was virtually<br />
absent in <strong>megalin</strong>-deficient cells (Figure 4). In the renal proximal tubular cells that did<br />
express <strong>megalin</strong>, Dab2 co-localized with <strong>megalin</strong> in the apical endocytic apparatus such<br />
as coated pits <strong>and</strong> small endocytic vacuoles, but was almost undetectable in dense<br />
apical tubules, where <strong>megalin</strong> is present. The reduced staining <strong>of</strong> Dab2 in <strong>megalin</strong> -/-<br />
cells could be due to a decrease in Dab2 protein levels or due to decreased sensitivity<br />
<strong>of</strong> detection owing to diffuse <strong>localization</strong> within the cell. These possibilities could not be
conclusively distinguished because <strong>of</strong> the mosaic nature <strong>of</strong> the kidney-specific<br />
knockout. However, immunoblotting <strong>of</strong> kidney homogenates from mice with <strong>megalin</strong><br />
knock-out in all cells did not reveal any decrease <strong>of</strong> Dab2 compared to controls (result<br />
not shown). This suggests that Dab2 accumulation in apical endocytic pits <strong>and</strong> vesicles<br />
requires <strong>megalin</strong>.<br />
Decreased <strong>megalin</strong> levels in Dab2 knock-out mice<br />
Immunohistocytochemistry for <strong>megalin</strong> in the proximal tubule from Dab2-knock out mice<br />
was examined <strong>and</strong> compared with <strong>megalin</strong> in control mouse kidney. As shown in<br />
Figure 5, there was apparently a decrease in the subcellular compartments containing<br />
<strong>megalin</strong> <strong>of</strong> the renal proximal tubule in addition to an apparent reduction in the number<br />
<strong>of</strong> coated pits <strong>and</strong> coated vesicles as described previously (29). Decreased levels <strong>of</strong><br />
<strong>megalin</strong> in the apical cytoplasm were supported by morphometric analysis <strong>of</strong> electron<br />
microscope sections (110 gold particles/µm 2 in controls versus 72 gold particles/µm 2 in<br />
knock-outs, Table 1). Furthermore, this analysis showed that <strong>megalin</strong> levels were<br />
increased in the microvilli (96 gold particles/µm 2 <strong>and</strong> 150 gold particles/µm 2 in controls<br />
<strong>and</strong> knock-outs respectively, Table 1). This suggests a redistribution <strong>of</strong> <strong>megalin</strong> from<br />
vesicles <strong>of</strong> the apical cytoplasm to microvilli, consistent with decreased endocytosis,<br />
when Dab2 is absent.<br />
Surprisingly, immunoblotting <strong>of</strong> crude membranes from Dab2 knock-out mice <strong>and</strong><br />
control mice revealed an overall decrease in <strong>megalin</strong> levels when Dab2 was absent<br />
(Figure 6). In contrast, no obvious differences in protein levels <strong>of</strong> clathrin <strong>and</strong> -<br />
adaptin, the subunit <strong>of</strong> clathrin adaptor protein AP-2, were found between Dab2<br />
knock-out mice <strong>and</strong> control mice. As expected, Dab2 was not detected in kidney from<br />
knock-out mice. Northern blot analysis did not reveal any differences in <strong>megalin</strong> mRNA<br />
expression between kidneys from Dab2 knock-out mice <strong>and</strong> control mice (Figure 7).
This suggests that decreases in <strong>megalin</strong> protein levels are due to a decreased<br />
rate <strong>of</strong> protein synthesis or increased <strong>megalin</strong> protein turnover.<br />
Involvement <strong>of</strong> Dab2 in <strong>megalin</strong>-mediated endocytosis<br />
To examine the role <strong>of</strong> Dab2 in receptor-mediated endocytosis, uptake studies were<br />
performed in BN16 cells, a cell line derived from the rat yolk sac highly expressing<br />
<strong>megalin</strong>. First, we performed double labeling immun<strong>of</strong>luorescence on confluent BN16<br />
cell monolayers with anti-Dab2 antibody <strong>and</strong> anti-<strong>megalin</strong> antibody. The labeling<br />
pattern for Dab2 was present apically <strong>and</strong> partly co-localized with <strong>megalin</strong> (Figure 8).<br />
Next, we introduced anti-Dab2 antibody into the confluent BN16 cell monolayers in<br />
order to inhibit the function <strong>of</strong> Dab2 <strong>and</strong> then uptake <strong>and</strong> binding <strong>of</strong> iodinated RAP in<br />
the cells was measured (Figure 9). Uptake <strong>and</strong> binding <strong>of</strong> 125 I-RAP was slightly but<br />
significantly reduced in the transfected cells.
Discussion<br />
Dab2 is present in various tissues including heart, lung, liver <strong>and</strong> skeletal muscle, <strong>and</strong> is<br />
highly abundant in kidney (6,29), where it has been localized to the renal proximal<br />
tubule by immunohistochemistry (29,33). Dab2 has been demonstrated to bind to the<br />
cytoplasmic tail <strong>of</strong> <strong>megalin</strong> by several approaches such as two-hybrid screening, protein<br />
binding assay <strong>and</strong> co-immunoprecipitation (33). Recently, an increase in urinary<br />
excretion <strong>of</strong> specific plasma proteins such as DBP <strong>and</strong> RBP in Dab2 knock-out mice<br />
was reported (29). These proteins represent lig<strong>and</strong>s to <strong>megalin</strong>. Thus, Dab2 may play<br />
an important role in <strong>megalin</strong>-mediated endocytosis in the renal proximal tubule. The<br />
present study has been performed to clarify the cellular <strong>localization</strong> <strong>of</strong> Dab2 in the renal<br />
proximal tubule <strong>and</strong> its role in <strong>megalin</strong>-meditated endocytosis.<br />
We found that Dab2 is abundantly expressed in the apical endocytic apparatus<br />
such as coated pits <strong>and</strong> coated vesicles, but almost absent in the brush border <strong>and</strong><br />
dense apical tubules <strong>of</strong> the renal proximal tubules from normal mice. The low Dab2<br />
staining in the brush border is in contrast to immun<strong>of</strong>luorescence data obtained by<br />
Oleinikov et al. (33). The reason for this discrepancy is unknown <strong>and</strong> difficult to<br />
comment further as micrographs were not shown in the paper <strong>of</strong> Oleinikov et al.<br />
However, in fibroblasts <strong>and</strong> non-polarized epithelial cells (HeLa), Dab2 <strong>localization</strong><br />
depends on a region that binds to adaptin <strong>and</strong> clathrin (23,27), which are present in<br />
coated structures <strong>and</strong> not associated with the microvillar membrane. Immunoelectron<br />
microscope analysis revealed that labeling for Dab2 was present at the cytoplasmic side<br />
<strong>of</strong> the vesicular membrane <strong>of</strong> the endocytic apparatus, suggesting that Dab2 serves as<br />
a cytoplasmic adaptor protein mediating protein-protein interactions. In addition, co-<br />
<strong>localization</strong> <strong>of</strong> <strong>megalin</strong> <strong>and</strong> Dab2 was observed in apical coated pits <strong>and</strong> vesicles, but<br />
not in dense apical tubules <strong>of</strong> the renal proximal tubules. Considering that the<br />
cytoplasmic tail <strong>of</strong> <strong>megalin</strong> has been shown to interact with Dab2 (33), Dab2 may be<br />
involved in <strong>megalin</strong>-mediated endocytosis in the renal proximal tubular cells by directly
interacting with the cytoplasmic tail <strong>of</strong> <strong>megalin</strong> during endocytosis, but dissociating from<br />
<strong>megalin</strong> during recycling <strong>of</strong> the receptor. Similarly, Dab2 was absent from uncoated<br />
endosomes in fibroblasts (27).<br />
Analysis <strong>of</strong> kidney-specific <strong>megalin</strong> knock-out mice, which show a severe<br />
impairment in renal absorption <strong>of</strong> protein owing to a significant decrease in the number<br />
<strong>of</strong> proximal tubular cells expressing <strong>megalin</strong> (8,19), revealed that Dab2 <strong>localization</strong> to<br />
the endocytic apparatus is strictly <strong>dependent</strong> on <strong>megalin</strong>. An almost complete loss <strong>of</strong><br />
Dab2 in the endocytic apparatus was observed in <strong>megalin</strong>-deficient cells <strong>of</strong> the renal<br />
proximal tubule, whereas Dab2 was expressed normally <strong>and</strong> partly co-localized with<br />
<strong>megalin</strong> in normal tubular cells expressing <strong>megalin</strong>. These results suggest that Dab2<br />
association with coated pits in polarized cells may require cooperative association with<br />
the <strong>megalin</strong> cargo as well as with clathrin <strong>and</strong> AP-2, whereas in fibroblasts, association<br />
with coated pits appears to be in<strong>dependent</strong> <strong>of</strong> cargo (27).<br />
Several recent studies (10,23,27) have demonstrated that Dab2 interacts directly<br />
with phosphoinositides, clathrin <strong>and</strong> -adaptin subunit <strong>of</strong> the clathrin adaptor protein<br />
AP-2, which are components <strong>of</strong> clathrin-coated pits <strong>and</strong> vesicles. The subcellular<br />
<strong>localization</strong> <strong>of</strong> Dab2 in non-polarized cells is not due to the PTB domain but due to the<br />
central region that binds to the -adaptin subunit <strong>of</strong> AP-2 (23,27). In addition, a C-<br />
terminal region <strong>of</strong> Dab2 binds to myosin VI, an actin-based motor protein involved in<br />
cargo movement such as vesicular trafficking (17,28). Thus it is possible that none <strong>of</strong><br />
the interactions is tight enough on its own to recruit Dab2, <strong>and</strong> the absence <strong>of</strong> <strong>megalin</strong><br />
is enough to reduce formation <strong>of</strong> clathrin-, AP-2- <strong>and</strong> Dab2-containing structures.<br />
Alternatively, the absence <strong>of</strong> Dab2 in the endocytotic compartments <strong>of</strong> <strong>megalin</strong>-deficient<br />
cells may be due to a lack <strong>of</strong> other proteins than <strong>megalin</strong> that could regulate the<br />
<strong>localization</strong> <strong>of</strong> Dab2. In fact, clathrin <strong>and</strong> -adaptin are less concentrated in apical<br />
endocytic structures in <strong>megalin</strong> knock-out mice as judged by immunohistochemistry<br />
(results not shown).
Megalin protein levels <strong>and</strong> subcellular <strong>localization</strong> also depend on Dab2.<br />
Electron immunohistochemical analysis showed a decrease in <strong>megalin</strong> in endosomes,<br />
<strong>and</strong> an increase in <strong>megalin</strong> in microvilli, in renal proximal tubular cells from Dab2 knock-<br />
out mice, consistent with a role for Dab2 in <strong>megalin</strong> endocytosis. Surprisingly, total<br />
<strong>megalin</strong> levels were reduced in Dab2 knock-out mice, as observed by immunoblotting.<br />
However, there was no significant change in <strong>megalin</strong> mRNA level between Dab2 knock-<br />
out mice <strong>and</strong> control mice, suggesting that the decrease in <strong>megalin</strong> protein is due to<br />
decreased protein synthesis or increased protein turnover. The first <strong>and</strong> third NPXY<br />
motifs in the cytoplasmic tail <strong>of</strong> <strong>megalin</strong> are responsible for efficient endocytosis,<br />
whereas the second NPXY-like motif is essential for the apical sorting <strong>of</strong> <strong>megalin</strong> (36).<br />
In addition, the third NPXY motif has been suggested to be involved in the interaction<br />
between the PTB domain <strong>of</strong> Dab2 <strong>and</strong> the cytoplasmic tail <strong>of</strong> <strong>megalin</strong> (33). Therefore,<br />
decrease in <strong>megalin</strong> levels in Dab2 knock-out mice may result from an impaired<br />
trafficking <strong>of</strong> <strong>megalin</strong> through the endocytic/recycling pathway, leading to increased<br />
<strong>megalin</strong> shedding or degradation.<br />
Even though Dab2 interacts with clathrin <strong>and</strong> AP-2, <strong>and</strong> there is a decrease in<br />
the number <strong>of</strong> apical coated pits <strong>and</strong> apical coated vesicles in the renal proximal tubular<br />
cells from Dab2 knock-out mice (29) <strong>and</strong> <strong>megalin</strong> knock-out mice (18), the levels <strong>of</strong><br />
clathrin <strong>and</strong> -adaptin in Dab2 knock-out mouse kidneys were normal.<br />
Immunohistochemistry for clathrin also revealed equal levels in Dab2 knock-out mice<br />
<strong>and</strong> control mice (data not shown).<br />
To investigate the impact <strong>of</strong> decreased <strong>megalin</strong> levels induced by Dab2 knock-<br />
out, we transfected anti-Dab2 rabbit antibodies into BN16 cells. This slightly but<br />
significantly decreased internalization <strong>of</strong> receptor-associated protein (RAP), a high<br />
affinity lig<strong>and</strong> for <strong>megalin</strong>, as compared with that <strong>of</strong> control rabbit Ig. This is in<br />
accordance with the observations obtained by immunohistochemistry in Dab2 knock-out
mice, where deletion <strong>of</strong> Dab2 does not completely abolish endocytosis <strong>of</strong> RBP <strong>and</strong><br />
DBP (results not shown).<br />
Dab2 knock-out mice present an increase in urinary excretion <strong>of</strong> DBP <strong>and</strong> RBP,<br />
but the increases are not as severe as in full <strong>megalin</strong> knock-out mice, which exhibit<br />
vitamin D deficiency <strong>and</strong> bone formation defects (31). Indeed, there was a decreased<br />
endosomal level <strong>of</strong> <strong>megalin</strong> in Dab2 knock-out mice, <strong>and</strong> <strong>megalin</strong> protein levels were<br />
reduced but not completely disrupted by the absence <strong>of</strong> Dab2. Thus, another adaptor<br />
protein binding to the cytoplasmic tail <strong>of</strong> <strong>megalin</strong>, such as ARH (30), may substitute for<br />
Dab2 in Dab2 knock-out mice.<br />
In conclusion these experiments show the presence <strong>of</strong> Dab2 in coated structures<br />
<strong>of</strong> the endocytic apparatus <strong>of</strong> the proximal tubule. Here its presence seems to be<br />
<strong>dependent</strong> on <strong>megalin</strong> or other factors associated with <strong>megalin</strong>. On the other h<strong>and</strong><br />
knock-out <strong>of</strong> the Dab2 gene decreases <strong>megalin</strong> levels <strong>and</strong> alters the subcellular<br />
distribution <strong>of</strong> <strong>megalin</strong>. These results indicate that Dab2 <strong>and</strong> <strong>megalin</strong> mutually regulate<br />
each others <strong>localization</strong> in renal proximal tubule cells.
Acknowledgments<br />
The authors thank Inger Blenker Krist<strong>of</strong>fersen, Hanne Sidelmann, Pia Kamuk Nielsen,<br />
Anne Merete Hass <strong>and</strong> Priscilla Kronstadt O’Brien for excellent technical assistance.<br />
This study was founded by Novo Nordisk, Karen Elise Jensen Foundation, NIH<br />
(GMO66257), The European Commission (EU Framework Program 6, EureGene,<br />
contract number 05085) <strong>and</strong> The Carlsberg Foundation.
References<br />
1. Birn H, Vorum H, Verroust PJ, Moestrup SK, <strong>and</strong> Christensen EI. Receptor-<br />
associated protein is important for normal processing <strong>of</strong> <strong>megalin</strong> in kidney proximal<br />
tubules. J Am Soc Nephrol 11: 191-202, 2000.<br />
2. Chen WJ, Goldstein JL, <strong>and</strong> Brown MS. NPXY, a sequence <strong>of</strong>ten found in<br />
cytoplasmic tails, is required for coated pit-mediated internalization <strong>of</strong> the low density<br />
lipoprotein receptor. J Biol Chem 265: 3116-3123, 1990.<br />
3. Christensen EI, Nielsen S, Moestrup SK, Borre C, Maunsbach AB, de Heer E,<br />
Ronco P, Hammond TG, <strong>and</strong> Verroust P. Segmental distribution <strong>of</strong> the endocytosis<br />
receptor gp330 in renal proximal tubules. Eur J Cell Biol 66: 349-364, 1995.<br />
4. Christensen EI, <strong>and</strong> Birn H. Megalin <strong>and</strong> cubilin: synergistic endocytic receptors in<br />
renal proximal tubule. Am J Physiol <strong>Renal</strong> Physiol 280: F562-F573, 2001.<br />
5. Christensen EI, <strong>and</strong> Birn H. Megalin <strong>and</strong> cubilin: Multifunctional endocytic<br />
receptors. Nat Rev Mol Cell Biol 3: 256-266, 2002.<br />
6. Fazili Z, Sun W, Mittelstaedt S, Cohen C, <strong>and</strong> Xu XX. Disabled-2 inactivation is an<br />
early step in ovarian tumorigenicity. Oncogene 18: 3104-3113, 1999.<br />
7. Gburek J, Verroust PJ, Willnow TE, Fyfe JC, Nowacki W, Jacobsen C, Moestrup<br />
SK, <strong>and</strong> Christensen EI. Megalin <strong>and</strong> cubilin are endocytic receptors involved in renal<br />
clearance <strong>of</strong> hemoglobin. J Am Soc Nephrol 13: 423-430, 2002.<br />
8. Gburek J, Birn H, Verroust PJ, Goj B, Jacobsen C, Moestrup SK, Willnow TE,<br />
<strong>and</strong> Christensen EI. <strong>Renal</strong> uptake <strong>of</strong> myoglobin is mediated by the endocytic receptors<br />
<strong>megalin</strong> <strong>and</strong> cubilin. Am J Physiol <strong>Renal</strong> Physiol 285: F451-F458, 2003.<br />
9. Gotthardt M, Trommsdorff M, Nevitt MF, Shelton J, Richardson JA, Stockinger<br />
W, Nimpf J, <strong>and</strong> Herz J. Interactions <strong>of</strong> the low density lipoprotein receptor gene family<br />
with cytosolic adaptor <strong>and</strong> scaffold proteins suggest diverse biological functions in<br />
cellular communication <strong>and</strong> signal transduction. J Biol Chem 275: 25616-25624, 2000.
10. Hirst J, <strong>and</strong> Robinson MS. Clathrin <strong>and</strong> adaptors. Biochim Biophys Acta 1404:<br />
173-193, 1998.<br />
11. Hjalm G, Murray E, Crumley G, Harazim W, Lundgren S, Onyango I, Ek B,<br />
Larsson M, Juhlin C, Hellman P, Davis H, Akerstrom G, Rask L, <strong>and</strong> Morse B.<br />
Cloning <strong>and</strong> sequencing <strong>of</strong> human gp330, a Ca 2+ -binding receptor with potential<br />
intracellular signaling properties. Eur J Biochem 239: 132-137, 1996.<br />
12. Howell BW, Gertler FB, <strong>and</strong> Cooper JA. Mouse disabled (mDab1): a Src binding<br />
protein implicated in neuronal development. EMBO J 16: 121-132, 1997.<br />
13. Howell BW, Hawkes R, Soriano P, <strong>and</strong> Cooper JA. Neuronal position in the<br />
developing brain is regulated by mouse disabled-1. Nature 389: 733-737, 1997.<br />
14. Howell BW, Lanier LM, Frank R, Gertler FB, <strong>and</strong> Cooper JA. The disabled 1<br />
phosphotyrosine-binding domain binds to the internalization signals <strong>of</strong> transmembrane<br />
glycoproteins <strong>and</strong> to phospholipids. Mol Cell Biol 19: 5179-5188, 1999.<br />
15. Howell BW, Herrick TM, <strong>and</strong> Cooper JA. Reelin-induced tyrosine phosphorylation<br />
<strong>of</strong> disabled 1 during neuronal positioning. Genes Dev 13: 643-648, 1999.<br />
16. Howell BW, <strong>and</strong> Herz J. The LDL receptor gene family: signaling functions during<br />
development. Curr Opin Neurobiol 11: 74-81, 2001.<br />
17. Inoue A, Sato O, Homma K, <strong>and</strong> Ikebe M. DOC-2/DAB2 is the binding partner <strong>of</strong><br />
myosin VI. Biochem Biophys Res Commun 292: 300-307, 2002.<br />
18. Leheste JR, Rolinski B, Vorum H, Hilpert J, Nykjaer A, Jacobsen C,<br />
Aucouturier P, Moskaug JO, Otto A, Christensen EI, <strong>and</strong> Willnow TE. Megalin<br />
knockout mice as an animal model <strong>of</strong> low molecular weight proteinuria. Am J Pathol<br />
155: 1361-1370, 1999.<br />
19. Leheste JR, Melsen F, Wellner M, Jansen P, Schlichting U, Renner-Muller I,<br />
Andreassen TT, Wolf E, Bachmann S, Nykjaer A, <strong>and</strong> Willnow TE. Hypocalcemia<br />
<strong>and</strong> osteopathy in mice with kidney-specific <strong>megalin</strong> gene defect. FASEB J 17: 247-<br />
249, 2003.
20. Le Panse S, Verroust P, <strong>and</strong> Christensen EI. Internalization <strong>and</strong> recycling <strong>of</strong><br />
glycoprotein 280 in BN/MSV yolk sac epithelial cells: a model system <strong>of</strong> relevance to<br />
receptor-mediated endocytosis in the renal proximal tubule. Exp Nephrol 5: 375-383,<br />
1997.<br />
21. Marples D, Knepper MA, Christensen EI, <strong>and</strong> Nielsen S. Redistribution <strong>of</strong><br />
aquaporin-2 water channels induced by vasopressin in rat kidney inner medullary<br />
collecting duct. Am J Physiol 269: C655-C664, 1995.<br />
22. Margolis B, Borg JP, Straight S, <strong>and</strong> Meyer D. The function <strong>of</strong> PTB domain<br />
proteins. Kidney Int 56: 1230-1237, 1999.<br />
23. Mishra SK, Keyel PA, Hawryluk MJ, Agostinelli NR, Watkins SC, <strong>and</strong> Traub LM.<br />
Disabled-2 exhibits the properties <strong>of</strong> a cargo-selective endocytic clathrin adaptor. EMBO<br />
J 21:4915-4926, 2002.<br />
24. Moestrup SK, Nielsen S, Andreasen P, Jorgensen KE, Nykjaer A, Roigaard H,<br />
Gliemann J, <strong>and</strong> Christensen EI. Epithelial glycoprotein-330 mediates endocytosis <strong>of</strong><br />
plasminogen activator-plasminogen activator inhibitor type-1 complexes. J Biol Chem<br />
268: 16564-16570, 1993.<br />
25. Mok SC, Wong KK, Chan RK, Lau CC, Tsao SW, Knapp RC, <strong>and</strong> Berkowitz RS.<br />
Molecular cloning <strong>of</strong> differentially expressed genes in human epithelial ovarian cancer.<br />
Gynecol Oncol 52: 247-252, 1994.<br />
26. Mok SC, Chan WY, Wong KK, Cheung KK, Lau CC, Ng SW, Baldini A, Colitti<br />
CV, Rock CO, <strong>and</strong> Berkowitz RS. DOC-2, a c<strong>and</strong>idate tumor suppressor gene in<br />
human epithelial ovarian cancer. Oncogene 16: 2381-2387, 1998.<br />
27. Morris SM, <strong>and</strong> Cooper JA. Disabled-2 colocalizes with the LDLR in clathrin-<br />
coated pits <strong>and</strong> interacts with AP-2. Traffic 2: 111-123, 2001.<br />
28. Morris SM, Arden SD, Roberts RC, Kendrick-Jones J, Cooper JA, Luzio JP,<br />
<strong>and</strong> Buss F. Myosin VI binds to <strong>and</strong> localizes with Dab2, potentially linking receptor-<br />
mediated endocytosis <strong>and</strong> the actin cytoskeleton. Traffic 3: 331-341, 2002.
29. Morris SM, Tallquist MD, Rock CO, <strong>and</strong> Cooper JA. Dual roles for the Dab2<br />
adaptor protein in embryonic development <strong>and</strong> kidney transport. EMBO J 21: 1555-<br />
1564, 2002.<br />
30. Nagai M, Meerloo T, Takeda T, <strong>and</strong> Farquhar MG. The adaptor protein ARH<br />
escorts <strong>megalin</strong> to <strong>and</strong> through endosomes. Mol Biol Cell 14: 4984-4996, 2003.<br />
31.Nykjaer A, Dragun D, Walther D, Vorum H, Jacobsen C, Herz J, Melsen F,<br />
Christensen EI, <strong>and</strong> Willnow TE. An endocytic pathway essential for renal uptake <strong>and</strong><br />
activation <strong>of</strong> the steroid 25-(OH) vitamin D3. Cell 96: 507-515, 1999.<br />
32. Nykjaer A, <strong>and</strong> Willnow TE. The low-density lipoprotein receptor gene family: a<br />
cellular Swiss army knife? Trends Cell Biol 12: 273-280, 2002.<br />
33. Oleinikov AV, Zhao J, <strong>and</strong> Makker SP. Cytosolic adaptor protein Dab2 is an<br />
intracellular lig<strong>and</strong> <strong>of</strong> endocytic receptor gp600/<strong>megalin</strong>. Biochem J 347: 613-621, 2000.<br />
34. Rader K, Orl<strong>and</strong>o RA, Lou X, <strong>and</strong> Farquhar MG. Characterization <strong>of</strong> ANKRA, a<br />
novel ankyrin repeat protein that interacts with the cytoplasmic domain <strong>of</strong> <strong>megalin</strong>. J Am<br />
Soc Nephrol 11: 2167-2178, 2000.<br />
35. Saito A, Pietromonaco S, Loo AK, <strong>and</strong> Farquhar MG. Complete cloning <strong>and</strong><br />
sequencing <strong>of</strong> rat gp330/"<strong>megalin</strong>," a distinctive member <strong>of</strong> the low density lipoprotein<br />
receptor gene family. Proc Natl Acad Sci USA 91: 9725-9729, 1994.<br />
36. Takeda T, Yamazaki H, <strong>and</strong> Farquhar MG. Identification <strong>of</strong> an apical sorting<br />
determinant in the cytoplasmic tail <strong>of</strong> <strong>megalin</strong>. Am J Physiol Cell Physiol 284: C1105-<br />
C1113, 2003.<br />
37. Tseng CP, Ely BD, Li Y, Pong RC, <strong>and</strong> Hsieh JT. Regulation <strong>of</strong> rat DOC-2 gene<br />
during castration-induced rat ventral prostate degeneration <strong>and</strong> its growth inhibitory<br />
function in human prostatic carcinoma cells. Endocrinology 139: 3542-3553, 1998.<br />
38. Willnow TE, Rohlmann A, Horton J, Otani H, Braun JR, Hammer RE, <strong>and</strong> Herz<br />
J. RAP, a specialized chaperone, prevents lig<strong>and</strong>-induced ER retention <strong>and</strong>
degradation <strong>of</strong> LDL receptor-related endocytic receptors. EMBO J 15: 2632-2639,<br />
1996.<br />
39. Willnow TE, Nykjaer A, <strong>and</strong> Herz J. Lipoprotein receptors: new roles for ancient<br />
proteins. Nat Cell Biol 1: E157-E162, 1999.<br />
40. Xu XX, Yang W, Jackowski S, <strong>and</strong> Rock CO. Cloning <strong>of</strong> a novel phosphoprotein<br />
regulated by colony-stimulating factor 1 shares a domain with the Drosophila disabled<br />
gene product. J Biol Chem 270: 14184-14191, 1995.<br />
41. Zhai XY, Nielsen R, Birn H, Drumm K, Mildenberger S, Freudinger R, Moestrup<br />
SK, Verroust PJ, Christensen EI, <strong>and</strong> Gekle M. Cubilin- <strong>and</strong> <strong>megalin</strong>-mediated uptake<br />
<strong>of</strong> albumin in cultured proximal tubule cells <strong>of</strong> opossum kidney. Kidney Int 58: 1523-<br />
1533, 2000.
Figure legends<br />
Figure 1 Immunohistochemistry using anti-Dab2 antibody on cryosections <strong>of</strong> kidney<br />
cortex from a Dab2 knock-out mouse (A) <strong>and</strong> a control mouse (B). Magnification:<br />
X1,500.<br />
Figure 2 Electron micrograph <strong>of</strong> a renal proximal tubule cell from a control rat<br />
labeled with anti-Dab2 antibody. Magnification: X50,000; Insert (coated vesicle):<br />
X160,000. CV = coated vesicle ; D = dense apical tubule; BB = brush border.<br />
Figure 3 Immun<strong>of</strong>luorescent <strong>localization</strong> <strong>of</strong> Dab2 (A) <strong>and</strong> <strong>megalin</strong> (B) in kidney<br />
specific <strong>megalin</strong> knock-out mouse. Labeling for Dab2 was seen in some renal proximal<br />
tubular cells coinciding with cells containing <strong>megalin</strong>, whereas lack <strong>of</strong> Dab2 was<br />
observed in neighboring cells also lacking <strong>megalin</strong>. A section with an unusually high<br />
number <strong>of</strong> <strong>megalin</strong>-expressing cells is shown. Merged picture is shown in C.<br />
Magnification: X 3,000<br />
Figure 4 Double immunogold labeling for Dab2 (10 nm gold particles, arrows) <strong>and</strong><br />
<strong>megalin</strong> (5 nm gold particles, arrowheads) on ultrathin frozen sections <strong>of</strong> renal cortex<br />
from kidney specific <strong>megalin</strong> knock-out mice. Magnification: X54,000.<br />
Figure 5 Immunohistochemistry using anti-<strong>megalin</strong> antibody on cryosections <strong>of</strong><br />
kidney cortex from a Dab2 knock-out mouse (A) <strong>and</strong> a control mouse (B). Arrows<br />
indicate the cytoplasmic vesicles labeled with anti-<strong>megalin</strong> antibody. Magnification:<br />
X1,200.
Figure 6 Immunoblotting <strong>of</strong> Dab2, <strong>megalin</strong>, cubilin, clathrin heavy chain (HC) <strong>and</strong><br />
-adaptin in crude membrane fractions <strong>of</strong> kidney from Dab2 knock-out mice <strong>and</strong> control<br />
mice.<br />
Figure 7 Northern blot analysis for <strong>megalin</strong> mRNA in kidneys from Dab2 knock-out<br />
mice <strong>and</strong> control mice.<br />
Figure 8 Immun<strong>of</strong>luorescent <strong>localization</strong> <strong>of</strong> Dab2 (A) <strong>and</strong> <strong>megalin</strong> (B) in BN16 cells.<br />
The merged image (C) indicates that Dab2 <strong>and</strong> <strong>megalin</strong> are partly co-localized (yellow).<br />
Magnification: X1,000 in A-C.<br />
Figure 9 Effect <strong>of</strong> transfection <strong>of</strong> anti-Dab2 antibody on 125 I-RAP uptake in<br />
confluent BN16 cell monolayers. Confluent monolayers <strong>of</strong> BN16 cells were incubated in<br />
the presence <strong>of</strong> rabbit anti-Dab2 antibody or normal rabbit Ig. Each column is the mean<br />
± SE <strong>of</strong> five monolayers. *P
Figures<br />
Figure 1<br />
a) b)<br />
Figure 2
c)<br />
Figure 3<br />
a) b)
Figure 4
Figure 6<br />
Figure 5<br />
a) b)
Figure 7<br />
Figure 8<br />
a) b)<br />
c)
Figure 9<br />
Table 1. Morphometric determination <strong>of</strong> <strong>megalin</strong><br />
<strong>localization</strong> in microvilli <strong>and</strong> cytoplasm <strong>of</strong> control<br />
<strong>and</strong> Dab2 knock-out mice.<br />
Gold/µm2 Controls Dab2 k.o.<br />
Cytoplasm 110±14 72±22<br />
Microvilli 96±14 150±55<br />
Values are means ± S.D., n=6.