THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2000 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 275, No. 16, Issue of April 21, pp. 12003–12008, 2000
Printed in U.S.A.
Megalin Acts in Concert with Cubilin to Mediate Endocytosis of
High Density Lipoproteins*
(Received for publication, December 13, 1999, and in revised form, February 3, 2000)
Samar M. Hammad‡§, Jeremy L. Barth‡¶, Christian Knaak, and W. Scott Argraves储
From the Department of Cell Biology and Anatomy, Medical University of South Carolina,
Charleston, South Carolina 29425-2204
Cubilin is a recently identified receptor that mediates binding and endocytosis of intrinsic factor-vitamin B12 complex (1),
immunoglobulin light chains (2), and high density lipoproteins
(HDL)1 (3, 4). The primary sequence of the ⬃460-kDa cubilin
polypeptide lacks a discernible membrane-spanning element
(1, 5). However, a sequence located in the amino-terminal region having amphipathic helical characteristics has been implicated in plasma membrane anchoring/association (6). It is
* This work was supported by Grant 9950344N from the American
Heart Association (AHA) (to W. S. A.). The costs of publication of this
article were defrayed in part by the payment of page charges. This
article must therefore be hereby marked “advertisement” in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ These two authors contributed equally to this work.
§ Recipient of a fellowship from the AHA.
¶ Recipient of National Institutes of Health Fellowship HL07710.
储 To whom correspondence should be addressed: Cell Biology and
Anatomy Dept., Medical University of South Carolina, 171 Ashley Ave.,
Charleston, SC 29425-2204. Tel.: 843-792-5482; Fax: 843-792-0664;
E-mail: [email protected].
1
The abbreviations used are: HDL, high density lipoprotein(s); LDL,
low density lipoprotein(s); LDLR, LDL receptor; FITC, fluorescein isothiocyanate; PCR, polymerase chain reaction; DiI, 1,1⬘-dioctadecyl3,3,3⬘,3⬘-tetramethylindocarbocyanine perchlorate; RAP, receptor-associated protein; ODN, oligodeoxynucleotide; LRP, LDLR-related protein;
uPA, urokinase-type plasminogen activator; uPAR, uPA receptor;
PAI-1, plasminogen activator inhibitor-1; GAPDH, glyceraldehyde-3phosphate dehydrogenase.
This paper is available on line at http://www.jbc.org
not clear how such a “peripheral” association allows interaction
of cubilin with cytosolic components of the endocytic machinery. A plausible hypothesis is that an integral membrane protein(s) functions in concert with cubilin to facilitate the endocytic activities. The principal candidate for such an accessory
protein is megalin (also known as gp330 and LRP-2), a member
of the low density lipoprotein receptor family. Megalin binds
cubilin and colocalizes with it in apical endocytic invaginations
and endosomes of renal proximal tubule cells and yolk sac
epithelial cells (1). In the present study, the role of megalin in
cubilin-mediated endocytosis of HDL is examined.
EXPERIMENTAL PROCEDURES
Proteins—Human apolipoprotein A-I (apoA-I) was obtained from Dr.
Bryan Brewer (Molecular Disease Branch, NHLBI, National Institutes
of Health, Bethesda, MD). Megalin was purified from porcine kidney as
described previously (7).
Lipoproteins—Human DiI-HDL was purchased from Biomedical
Technologies. Human HDL (density ⫽ 1.063–1.21 g/ml) and LDL were
provided by Dr. Bryan Brewer. Delipidated HDL (apoHDL) was prepared as described (8). DiI-HDL and HDL were depleted of apoE-HDL
and other heparin-binding particles according to Oram (9), dialyzed
against 150 mM NaCl, 50 mM Tris, pH 7.4 (Tris-buffered saline) containing 0.3 mM EDTA and filter-sterilized. Lipoprotein concentration
was determined by BCA protein assay (Pierce).
Antibodies—Rabbit polyclonal (rb239 and rb6286) and mouse monoclonal megalin (6C53C3) antibodies have been described previously (10,
11). Rabbit anti-megalin IgG was purified by protein-G-Sepharose and
then affinity-selected by chromatography on megalin-Sepharose (12).
Rabbit anti-cubilin serum was provided by Dr. Pierre Verroust (Hospital Tenon, Paris, France), and IgG was purified by protein G-Sepharose
chromatography. A rabbit polyclonal antiserum was raised against an
amino-terminal segment of rat cubilin (residues 21– 473) expressed in
bacteria as a fusion protein with glutathione S-transferase using the
pGEX6P vector (Amersham Pharmacia Biotech). The resulting antiserum was preabsorbed on glutathione S-transferase-Sepharose, and
IgG was purified using protein G-Sepharose chromatography. Rabbit
anti-megalin (rb6286) and rabbit anti-cubilin amino-terminal segment
IgGs were labeled with fluorescein isothiocyanate (FITC) and tetramethylrhodamine, respectively, according to the instructions of the manufacturer (Pierce). Rabbit anti-LDLR (711) was obtained from Dr.
Joachim Herz (University of Texas Southwestern Medical Center, Dallas, TX). Horseradish peroxidase-conjugated anti-IgGs were obtained
from Amersham Pharmacia Biotech.
Cells—Murine sarcoma virus-transformed Brown Norway rat yolk
sac cells (BN cells) were provided by Dr. Pierre Verroust (Hospital
Tenon, Paris, France). Mouse embryonal teratocarcinoma F9 cells
(ATCC CRL1720) were differentiated by treatment with retinoic acid
and dibutyryl cyclic AMP for 6 days as described previously (13).
HDL-Sepharose Affinity Chromatography—HDL-Sepharose affinity
chromatography was performed as described previously (4), with the
exception that detergent extracts of BN cells were used.
Enzyme-linked Immunosorbent Assays—Enzyme-linked immunosorbent assays were performed as described previously (11). Megalin (150
nM in Tris-buffered saline, 0.05% Tween 20) was incubated with microtiter wells coated with LDL, HDL, apoHDL, apoA-I, or ovalbumin (3
␮g/ml). Bound receptor was detected using the monoclonal megalin
antibody 6C53C3, sheep anti-mouse IgG-horse radish peroxidase, and
the chromogenic substrate o-phenylenediamine (Sigma).
Reverse Transcription and PCR—Total RNA was isolated from
12003
Downloaded from www.jbc.org at MUSC Library on September 11, 2007
Cubilin has recently been shown to function as an
endocytic receptor for high density lipoproteins (HDL).
The lack of apparent transmembrane and cytoplasmic
domains in cubilin raises questions as to the means by
which it can mediate endocytosis. Since cubilin has
been reported to bind the endocytic receptor megalin,
we explored the possibility that megalin acts in conjunction with cubilin to mediate HDL endocytosis. While
megalin did not bind to HDL, delipidated HDL, or
apoA-I, it was found to copurify with cubilin isolated by
HDL-Sepharose affinity chromatography. Cubilin and
megalin exhibited coincident patterns of mRNA expression in mouse tissues including the kidney, ileum, thymus, placenta, and yolk sac endoderm. The expression of
both receptors in yolk sac endoderm-like cells was inducible by retinoic acid treatment but not by conditions
of sterol depletion. Suppression of megalin activity or
expression by treatment with either megalin antibodies
or megalin antisense oligodeoxynucleotides resulted in
inhibition of cubilin-mediated endocytosis of HDL. Furthermore, megalin antisense oligodeoxynucleotide
treatment resulted in reduced cell surface expression of
cubilin. These data demonstrate that megalin acts together with cubilin to mediate HDL endocytosis and
further suggest that megalin may play a role in the intracellular trafficking of cubilin.
12004
Megalin and Cubilin Are HDL Co-receptors
TABLE II
Megalin antisense oligodeoxynucleotides
TABLE I
PCR primers
Target gene
Mouse megalina
b
Mouse cubilin
Mouse GAPDHc
Rat megalind
Rat cubiline
Rat LDLRf
Sequence (5⬘–3⬘)
Position in cDNA
sequence
Oligodeoxynucleotide
Sequence (5⬘–3⬘)
Position in
cDNA sequencea
CCTTGCCAAACCCTCTGAAAAT
CACAAGGTTTGCGGTGTCTTTA
CAACATGGAACACAAACACTTT
AGCTATTGAATGTACGTCCACA
CGGTGTGAACGGATTTGGC
GCAGTGATGGCATGGACTGT
ACACCGCTTCTGCCGTCT
TCTGAGCACTCCCGAGGAAC
GCCTGCCCCATTTATCTCTTC
CGCCGTTTCTTACCTCCAA
GTTCCGAGAGAAAGGGTCCAG
TGCGTGACGTTGTGAAACAG
10–31
571–550
30–51
422–401
58–86
588–569
10775–10792
11410–11391
9723–9743
10375–10357
1686–1706
2132–2113
A1
A2
A3
A4
CAGGCAGGTTACTCCGCT
CTGCATGGGGTCTGT
CATCGCGGAGACGGCCC
GCCAATTTCATCTGTGT
58–41
1454–1440
125–109
305–289
a
GenBank™ accession no. L34049.
a
GenBank™ accession no. AF197160.
GenBank™ accession no. AF197159.
c
GenBank™ accession no. M32599.
d
GenBank™ accession no. L34049.
e
GenBank™ accession no. AF022247.
f
GenBank™ accession no. X13722.
b
FIG. 1. Megalin copurifies with cubilin isolated by HDL-Sepharose chromatography. A, silver-stained PAGE profile of cubilin-containing fractions sequentially eluted from HDL-Sepharose using 8 M
urea. B, immunoblot analysis using anti-cubilin IgG of an aliquot from
the peak fraction (*) eluted from HDL-Sepharose shown in A. C, immunoblot analysis using anti-megalin IgG of the peak fraction (*) eluted
from HDL-Sepharose shown in A.
FIG. 2. Megalin does not bind to HDL or apoA-I in enzymelinked immunosorbent assays. Microtiter wells coated with LDL,
apoE-free HDL, delipidated HDL (apoHDL), apoA-I, or ovalbumin (3
␮g/ml) were incubated with megalin (150 nM) in Tris-buffered saline
containing 0.05% Tween 20 for 18 h at 4 °C. Bound receptor was
detected using monoclonal antibodies to megalin. The plotted values are
means of duplicate determinations with the range indicated by bars.
ferred to polyvinylidene difluoride membranes. Membrane-bound proteins were detected with receptor-specific antibodies, horseradish peroxidase-conjugated anti-IgG, and ECL-plus (Amersham Pharmacia
Biotech).
Analysis of the Effect of Sterol on Receptor Expression—BN cells were
cultured at 1–1.5 ⫻ 105 cells/cm2 in complete medium for 6 h. The
medium was replaced with serum-free medium (minimal essential medium, ITS; Roche Molecular Biochemicals), penicillin/streptomycin)
containing either 1) mevalonic acid lactone (50 ␮M) (Sigma) and sterols
(25-hydroxycholesterol (1 ␮g/ml) and cholesterol (10 ␮g/ml)); 2) mevalonic acid lactone (50 ␮M) and compactin/mevastatin (50 ␮M) (Calbiochem); 3) sterols; or 4) compactin/mevastatin (50 ␮M). The cells were
cultured for 72 h at 37 °C, 5% CO2 (15). The medium was then removed,
and the cells were extracted with detergent and centrifuged at
100,000 ⫻ g. Equal amounts (5 ␮g) of protein from each extract were
analyzed by immunoblotting as described above.
Antisense Oligodeoxynucleotide Treatment of Cells and Analysis of
Receptor Activity and Expression—BN cells were plated in 12-well
plates (Corning Glass, Corning, NY) at 1–1.5 ⫻ 105 cells/cm2 and
allowed to attach for 6 h in complete medium. The cells were washed
with serum-free medium; serum-free medium containing synthetic oligodeoxynucleotides (5 ␮M) was then added and incubated with the cells
for 20 h at 37 °C, 5% CO2. Oligodeoxynucleotides used for antisense
Downloaded from www.jbc.org at MUSC Library on September 11, 2007
mouse tissues and from BN cells using RNA Stat-60 (Tel-Test, Inc.)
according to the manufacturer’s specifications. Reverse transcription of
total mRNA using random hexamer oligodeoxynucleotide primers and
cDNA synthesis were performed as described previously (14). For the
design of mouse cubilin amplification primers, a GenBankTM search
was used to identify an orthologous mouse cubilin cDNA (clone ID
C0026H06) having 87.6% similarity to rat cubilin cDNA and encoding a
polypeptide with 87.3% identity to the rat cubilin carboxyl terminus.
For the design of mouse megalin primers, a 686-base pair cDNA sequence (GenBankTM accession no. AF197160) was assembled from 14
overlapping mouse expressed sequence tags (GenBankTM accession
numbers C80829, AI047614, AU043137, AA574617, AU043459,
AA109222, AA082958, AA109769, AU045327, AU045353, AU045179,
AI436683, AA617362, and AI025985). This 686-base pair cDNA is
92.6% identical to rat megalin cDNA and encodes a polypeptide 93.4%
similar to the rat megalin carboxyl terminus. The PCR primer sequences used in this study are shown in Table I. Primer pairs and cDNA
templates were tested over a range of amplification cycles to determine
an optimal cycle number for exponential phase of production. Reactions
were performed with Taq polymerase (0.2 units/␮l) (Qiagen), 1⫻ PCR
buffer (Qiagen), 1.6 mM total dNTP, 0.2 ␮M primers, and 2.5 mM MgCl2.
Annealing temperatures and cycles of amplification for each primer
pair are as follows: mouse cubilin, 53 °C, 28 cycles; mouse megalin,
54 °C, 28 cycles; mouse GAPDH, 53 °C, 22 cycles; rat cubilin, 52 °C, 20
cycles; rat megalin, 61 °C, 23 cycles; rat LDLR, 52 °C, 27 cycles.
Confocal Microscopic Analysis of Cubilin and Megalin in Cultured
Yolk Sac Endoderm-like Cells—Cells were plated into wells of plastic
chamber slides (Nalge Nunc) (4 ⫻ 104 cells/cm2) and incubated for 18 h
at 37 °C in complete medium (minimal essential medium containing
10% fetal bovine serum, 100 units/ml penicillin, 100 ␮g/ml streptomycin). The cells were washed once with serum-free medium (minimal
essential medium, ITS (5 ␮g/ml insulin, 5 ␮g/ml transferrin, 5 ng/ml
sodium selenite; Roche Molecular Biochemicals), penicillin/streptomycin) and incubated for 4 h in serum-free medium at 37 °C. Fluorescently
labeled IgGs (FITC-anti-megalin and rhodamine-anti-cubilin aminoterminal segment (residues 21– 473) were added (final concentration,
50 ␮g/ml each) in the presence or absence of HDL (final concentration,
200 ␮g/ml) or RAP (final concentration, 2 ␮M) and incubated for 1 h at
37 °C. The cells were washed with Dulbecco’s phosphate-buffered saline, fixed in 3% formaldehyde in Dulbecco’s phosphate-buffered saline
for 20 min, and analyzed by confocal microscopy.
Analysis of the Effect of Retinoic Acid and Dibutyryl Cyclic AMP on
Receptor Expression—Mouse embryonal teratocarcinoma F9 cells were
seeded onto 0.1% gelatin-coated plates at 1 ⫻ 102 cells/cm2 and cultured
for 6 days in Dulbecco’s modified Eagle’s medium (Mediatech), 10%
iron-fortified bovine calf serum (Hyclone), 100 units/ml penicillin, and
100 ␮g/ml streptomycin in the presence or absence of 0.1 ␮M retinoic
acid (Sigma) or 0.2 ␮M dibutyryl cyclic AMP (Sigma). The medium was
then removed, and the cells were extracted with detergent buffer (1%
Triton X-100, 0.5% Tween 20, 0.5 M NaCl, 50 mM Hepes, pH 7.5,
containing a protease inhibitor mixture (EDTA-free; Roche Molecular
Biochemicals)). Protein concentrations in the extracts were determined
by the BCA protein assay (Pierce). Equal amounts (20 ␮g) of protein
were run on Novex gels (4 –12% polyacrylamide gradient) and trans-
Megalin and Cubilin Are HDL Co-receptors
12005
FIG. 3. Cubilin and megalin mRNAs
are coexpressed in mouse tissues. Cubilin and megalin mRNA expression was
evaluated by PCR using cDNA prepared
from the indicated mouse tissue mRNA
preparations. GAPDH amplifications
were performed as an experimental
control.
RESULTS
Megalin Copurifies with Cubilin—Recently, we have demonstrated that cubilin can be isolated from cultured yolk sac
endoderm-like cells using HDL-Sepharose affinity chromatography (4). To determine whether megalin co-purifies with cubilin, immunoblot analysis was performed on cubilin-containing fractions eluted from HDL-Sepharose. As shown in Fig. 1,
the peak cubilin-containing fraction also contained megalin,
detectable with a monoclonal antibody. In enzyme-linked immunosorbent assays, megalin did not bind to HDL, delipidated
HDL (apoHDL), or to purified apoA-I, the major apolipoprotein
constituent of HDL (Fig. 2). This supports the concept that the
presence of megalin in the HDL-Sepharose column eluates was
due to an association with cubilin rather than a direct interaction with the affinity matrix. These results are consistent with
the previously reported finding that cubilin and megalin interact with high affinity (1).
Cubilin and Megalin Expression Analysis—The expression of
cubilin and megalin mRNAs in mouse tissues was evaluated by
reverse transcription-PCR. As shown in Fig. 3, the overall
patterns of cubilin and megalin transcript expression are overlapping. Cubilin and megalin transcripts were co-expressed in
the kidney, thymus, ileum, placenta, and yolk sac. The only
exception was in the lung, where megalin mRNA expression
was detectable but cubilin mRNA was not.
The expression of both cubilin and megalin has been shown
to be stimulated in mouse F9 cells treated with a combination
of retinoic acid and dibutyryl cyclic AMP (4, 13). To determine
the effect of these agents separately on the expression of cubilin
and megalin, F9 cells were cultured in the presence of each
agent, and the level of receptor expression was evaluated immunologically. As shown in Fig. 4, retinoic acid treatment
stimulated expression of both cubilin and megalin, whereas
dibutyryl cyclic AMP treatment alone had no effect on expression of either receptor (Fig. 4).
The fact that both cubilin and megalin are involved with
lipoprotein catabolism also prompted us to determine whether
the expression of either receptor was sterol-responsive, similar
to the LDLR. Neither sterol-depleted conditions nor treatment
FIG. 4. Retinoic acid promotes the expression of both cubilin
and megalin. Immunoblot analysis was used to evaluate the expression of cubilin, megalin, and LDLR in detergent extracts of mouse F9
cells cultured in the presence or absence of retinoic acid (RA) and/or
dibutyryl cAMP (Bt2cAMP).
with the 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor, compactin, increased the accumulation of cubilin or
megalin mRNA or protein as was observed for the LDLR (Fig. 5).
The similarity in response of the two receptors to a regulatory stimulus (Fig. 4), together with the similarity in tissue
expression (Fig. 3), supports the hypothesis that they may have
a functional relationship. Further support for such a relationship comes from confocal microscopic studies of the subcellular
distribution of cubilin and megalin in cultured yolk sac
endoderm-like cells (Fig. 6). Cubilin and megalin displayed a
coincident pattern of subcellular distribution consistent with
an endocytic vesicle localization (Fig. 6, upper row). Treatment
of the cells with RAP, an antagonist of ligand-binding to megalin, leads to a reduction in the number of punctate foci stained
by cubilin antibody (Fig. 6). In contrast, RAP treatment caused
a reduction in the intensity of the vesicular staining pattern of
megalin but not a major reduction in the number of punctate
foci per cell (Fig. 6). These findings suggest that RAP prevents
cubilin from becoming efficiently internalized into endocytic
vesicles while not affecting the internalization of megalin. The
observed reduction in intensity of megalin staining caused by
RAP could be the result of masking of epitopes or destruction of
epitopes through alteration of megalin conformation. Administration of HDL abolished the vesicular staining patterns for
both receptors. Since HDL does not bind to megalin (Fig. 6),
these latter findings suggest that binding of HDL to cubilin
sterically or allosterically prevents antibody binding to megalin, a possibility made more plausible if the two receptors
existed in a complex.
Megalin Antibodies Inhibit Cubilin-mediated Uptake of
HDL—To determine whether megalin plays a role in endocytosis of cubilin ligands, function-blocking megalin antibodies
were tested for their ability to perturb HDL uptake by cultured
endoderm-like cells. Megalin antibodies were found to effectively inhibit the uptake of DiI-HDL (Fig. 7). The inhibitory
effect of the megalin antibodies was apparent as a ⬎40% reduction of the number of cells in the population that internalized DiI-HDL (Fig. 7A) as well as a ⬎30% reduction in the
amount of DiI-HDL that the cells internalized (Fig. 7B). Cubi-
Downloaded from www.jbc.org at MUSC Library on September 11, 2007
treatment (Table II) were synthesized with phosphorothioate ester linkages (the first two and the last three linkages) and with phosphodiester
linkages elsewhere. Immunoblot analysis was performed on cell extracts as described above. Cell surface receptor levels were evaluated by
flow cytometry after releasing the cells from culture plates with 2 mM
EDTA in Dulbecco’s phosphate-buffered saline and then washing the
cells with serum-free medium, 0.1% sodium azide, 2% normal goat
serum and incubating the cells with receptor antibodies and FITCconjugated goat anti-IgG. For flow cytometry analysis, the cells were
washed with serum-free medium and then incubated with serum-free
medium containing DiI-HDL (2 ␮g/ml) and oligodeoxynucleotide for
1.5 h at 37 °C, 5% CO2. The cells were then washed, stripped of cell
surface-associated DiI-HDL by treatment with 0.5 mg/ml trypsin, 50
␮g/ml proteinase K, 0.53 mM EDTA in Dulbecco’s phosphate-buffered
saline and subjected to flow cytometry (FACStar Plus, Becton
Dickinson).
12006
Megalin and Cubilin Are HDL Co-receptors
FIG. 5. Cubilin and megalin expression is not sterol-regulated. BN cells
were cultured in the presence of sterol
(25-hydroxycholesterol and cholesterol)
and/or the 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitor compactin
and/or mevalonate for 72 h, and the level
of LDLR, cubilin, and megalin protein (A)
and mRNA (B) expression was evaluated
by immunoblotting and reverse transcriptase-PCR, respectively.
lin antibodies had similar effects on the uptake of DiI-HDL.
Megalin Antisense Oligodeoxynucleotides Inhibit HDL Uptake—An antisense oligodeoxynucleotide (ODN) approach was
also employed to abrogate megalin expression and evaluate the
consequence on cubilin-mediated HDL uptake. Cultured
endoderm-like cells were treated with a series of ODNs complementary to megalin mRNA. As shown in Fig. 8, treatment of
the cells with one of the antisense ODNs (A2) greatly inhibited
uptake of DiI-HDL in comparison with the sense sequence of
this ODN (S2) or other megalin sense and antisense ODNs
(Fig. 8A). The specificity of antisense ODN A2 treatment on
megalin protein production was confirmed by immunoblot
analysis of detergent extracts of the treated cells. Treatment of
cells for 20 h with A2 reduced the level of megalin in cell
extracts by ⬎50% (Fig. 8B). Analysis of the level of megalin
expressed on the surface of the A2-treated cells was performed
by flow cytometry. As shown in Fig. 8C, cells treated with the
megalin antisense ODN A2 for 20 h had cell surface levels of
megalin ⬃50% that of cells treated with the megalin sense
ODN S2. Immunoblot analysis of the conditioned culture medium from cells treated with A2 and S2 revealed no overt
difference in cubilin levels (not shown).
Analysis of the effects of antisense ODN A2 treatment on
cubilin and LDLR protein levels in total cell extracts was also
performed. As shown in Fig. 8B, A2 treatment slightly increased (⬃5%) the levels of cubilin and LDLR in total cell
extracts as compared with cells treated with the sense ODN S2.
Flow cytometry analysis of cubilin expression on the surface of
the A2-treated cells showed ⬃20% reduction in levels of cubilin
as compared with the control (Fig. 8C). By contrast, there was
no difference in cell surface levels of LDLR in cells treated with
A2 and S2.
Downloaded from www.jbc.org at MUSC Library on September 11, 2007
FIG. 6. Laser scanning confocal microscopic analysis of the subcellular
localization of cubilin and megalin.
Cultured yolk sac endoderm-like cells
(BN) were incubated with tetramethylrhodamine-labeled rabbit anti-cubilin
amino-terminal segment IgG and FITClabeled rabbit anti-megalin IgG (upper
row) or with both antibodies plus RAP (2
␮M) (middle row) or HDL (200 ␮g/ml) (bottom row).
Megalin and Cubilin Are HDL Co-receptors
DISCUSSION
The findings reported in this study add new dimensions to
the function of megalin. First, the findings indicate that megalin can act in concert with another cell surface protein to
mediate ligand endocytosis. Similar cooperative relationships
have been reported for a close relative of megalin, low density
lipoprotein receptor-related protein (LRP) (16). LRP cooperates
with cell surface proteoglycans to mediate endocytosis of several ligands including thrombospondin-1 (17, 18) and hepatic
lipase (19). Furthermore, LRP and the urokinase receptor
12007
REFERENCES
FIG. 7. Megalin antibodies inhibit HDL uptake. Differentiated
F9 cells were incubated with DiI-HDL (3 ␮g/ml) alone or in the presence
of HDL (300 ␮g/ml), normal rabbit IgG (200 ␮g/ml), anti-megalin IgG
(200 ␮g/ml), or anti-cubilin (Verroust) IgG (200 ␮g/ml) for 1.5 h. Flow
cytometry was used to measure the percentage of cells having internalized DiI-HDL (A) and the level of DiI-HDL internalized (B). The plotted
data are mean values derived from analysis of 1 ⫻ 104 cells. Approximately 20% of the cells in differentiated F9 cell cultures show measurable levels of internalized DiI-HDL in these assays.
1. Moestrup, S. K., Kozyraki, R., Kristiansen, M., Kaysen, J. H., Rasmussen,
H. H., Brault, D., Pontillon, F., Goda, F. O., Christensen, E. I., Hammond,
T. G., and Verroust, P. J. (1998) J. Biol. Chem. 273, 5235–5242
2. Batuman, V., Verroust, P. J., Navar, G. L., Kaysen, J. H., Goda, F. O.,
Campbell, W. C., Simon, E., Pontillon, F., Lyles, M., Bruno, J., and
Hammond, T. G. (1998) Am. J. Physiol. 275, F246 –F254
3. Kozyraki, R., Fyfe, J., Kristiansen, M., Gerdes, C., Jacobsen, C., Cui, S.,
Christensen, E. I., Aminoff, M., de la Chapelle, A., Krahe, R., Verroust,
P. J., and Moestrup, S. K. (1999) Nat. Med. 5, 656 – 661
4. Hammad, S. M., Stefansson, S., Twal, W. O., Drake, C. J., Fleming, P.,
Remaley, A., Brewer, H. B., Jr., and Argraves, W. S. (1999) Proc. Natl. Acad.
Sci. U. S. A. 96, 10158 –10163
5. Kozyraki, R., Kristiansen, M., Silahtaroglu, A., Hansen, C., Jacobsen, C.,
Tommerup, N., Verroust, P. J., and Moestrup, S. K. (1998) Blood 91,
FIG. 8. Megalin antisense oligodeoxynucleotide treatment inhibits HDL uptake. A, flow cytometry analysis of DiI-HDL uptake by BN
cells that had been pretreated for 20 h with antisense megalin ODNs (A1, A2, A3, and A4) (5 ␮M each) or with the corresponding sense megalin
ODNs (S1, S2, S3, and S4) (5 ␮M each). Controls included BN cells cultured in the absence of ODNs incubated with DiI-HDL (2 ␮g/ml) alone or
DiI-HDL plus unlabeled HDL (200 ␮g/ml). Internalization of the DiI-HDL was measured by flow cytometry, and the mean fluorescence intensity
of 1.5 ⫻ 104 cells is plotted. B, BN cells treated for 20 h with antisense megalin ODN A2 or the corresponding sense ODN S2 were extracted, and
the levels of megalin, cubilin, and LDLR evaluated by immunoblot analysis. Plotted values in B were derived from densitometry measurements
made using the computer program NIH Image 1.62. C, analysis of surface receptor levels by flow cytometry of BN cells treated for 20 h with
antisense megalin ODN A2 or the corresponding sense ODN S2.
Downloaded from www.jbc.org at MUSC Library on September 11, 2007
(uPAR) cooperate in the process of endocytosis of urokinasetype plasminogen activator-plasminogen activator inhibitor-1
complexes (uPA䡠PAI-I) (20). Similar to cubilin, uPAR lacks a
membrane-spanning element, but it is glycosylphosphatidylinositol-anchored to the plasma membrane. Since megalin can
mediate uptake of uPA䡠PAI-I (21), it is plausible that megalin
can also function in concert with uPAR to mediate uPA䡠PAI-I
uptake. Such a process would, however, be distinct from megalin-mediated uptake of HDL in that uPA䡠PAI-I binds to megalin
(21) but HDL and its apolipoprotein component apoA-I do not
(Fig. 2). Therefore, in the process of HDL/apoA-I uptake, a
transfer of HDL/apoA-I from cubilin to megalin does not seem
feasible. Alternatively, megalin may mediate endocytosis of a
complex of cubilin and HDL/apoA-I.
An additional dimension to the functionality of megalin is
highlighted by the observation that megalin antisense oligodeoxynucleotide treatment leads to reduced cell surface expression of cubilin. This result suggests that megalin may play a
role in the biosynthetic trafficking of cubilin. Impaired cubilin
trafficking to apical brush border membranes of ilieum and
renal proximal tubule epithelial cells has been described in a
canine model for the human autosomal recessive ImerslundGrasbeck syndrome (22, 23). The molecular basis for the canine
disorder, which involves defective intestinal absorption of intrinsic factor-vitamin B12 complex, was speculated to involve
mutations within the cubilin gene as was the case for human
Imerslund-Grasbeck syndrome (24). However, a recent study
has demonstrated that the canine disorder is not caused by a
defect in the cubilin gene (25). In light of our findings, it is
possible that a mutation within the gene for megalin or its
biosynthetic trafficking chaperone protein, RAP, accounts for
the impaired cell surface expression of cubilin observed in the
affected dogs.
12008
Megalin and Cubilin Are HDL Co-receptors
3593–3600
6. Kristiansen, M., Kozyraki, R., Jacobsen, C., Nexo, E., Verroust, P. J., and
Moestrup, S. K. (1999) J. Biol. Chem. 274, 20540 –20544
7. Kounnas, M. Z., Stefansson, S., Loukinova, E., Argraves, K. M., Strickland,
D. K., and Argraves, W. S. (1994) Ann. N. Y. Acad. Sci. 737, 114 –123
8. Osborne, J. C. (1986) Methods Enzymol. 128, 213–222
9. Oram, J. F. (1986) Methods Enzymol. 129, 645– 659
10. Kounnas, M. Z., Chappell, D. A., Strickland, D. K., and Argraves, W. S. (1993)
J. Biol. Chem. 268, 14176 –14181
11. Hammad, S. M., Ranganathan, S., Loukinova, E., Twal, W. O., and Argraves,
W. S. (1997) J. Biol. Chem. 272, 18644 –18649
12. Kounnas, M. Z., Haudenschild, C. C., Strickland, D. K., and Argraves, W. S.
(1994) In Vivo 8, 343–351
13. Stefansson, S., Chappell, D. A., Argraves, K. M., Strickland, D. K., and
Argraves, W. S. (1995) J. Biol. Chem. 270, 19417–19421
14. Barth, J. L., Argraves, K. M., Roark, E. F., Little, C. D., and Argraves, W. S.
(1998) Matrix Biol. 17, 635– 646
15. Nohturfft, A., Brown, M. S., and Goldstein, J. L. (1998) Proc. Natl. Acad. Sci.
U. S. A. 95, 12848 –12853
16. Hussain, M. M., Strickland, D. K., and Bakillah, A. (1999) Annu. Rev. Nutr. 19,
141–172
17. Godyna, S., Liau, G., Popa, I., Stefansson, S., and Argraves, W. S. (1995) J. Cell
Biol. 129, 1403–1410
18. Mikhailenko, I., Kounnas, M. Z., and Strickland, D. K. (1995) J. Biol. Chem.
270, 9543–9549
19. Kounnas, M. Z., Chappell, D. A., Wong, H., Argraves, W. S., and Strickland,
D. K. (1995) J. Biol. Chem. 270, 9307–9312
20. Kounnas, M. Z., Henkin, J., Argraves, W. S., and Strickland, D. K. (1993)
J. Biol. Chem. 268, 21862–21867
21. Stefansson, S., Kounnas, M. Z., Henkin, J., Mallampalli, R. K., Chappell, D. A.,
Strickland, D. K., and Argraves, W. S. (1995) J. Cell Sci. 108, 2361–2368
22. Fyfe, J. C., Ramanujam, K. S., Ramaswamy, K., Patterson, D. F., and
Seetharam, B. (1991) J. Biol. Chem. 266, 4489 – 4494
23. Fyfe, J. C., Giger, U., Hall, C. A., Jezyk, P. F., Klumpp, S. A., Levine, J. S., and
Patterson, D. F. (1991) Pediatr. Res. 29, 24 –31
24. Aminoff, M., Carter, J. E., Chadwick, R. B., Johnson, C., Grasbeck, R.,
Abdelaal, M. A., Broch, H., Jenner, L. B., Verroust, P. J., Moestrup, S. K.,
de la Chapelle, A., and Krahe, R. (1999) Nat. Genet. 21, 309 –313
25. Xu, D., Kozyraki, R., Newman, T. C., and Fyfe, J. C. (1999) Blood 94,
3604 –3606
Downloaded from www.jbc.org at MUSC Library on September 11, 2007