Molecular Biology of the Cell
Vol. 18, 2667–2677, July 2007
Cholesterol Controls Lipid Endocytosis through Rab11□
Miwa Takahashi,*†Motohide Murate,* Mitsunori Fukuda,‡§Satoshi B. Sato,*?
Akinori Ohta,†and Toshihide Kobayashi*¶#
*Frontier Research System,‡Fukuda Initiative Research Unit, and¶Lipid Biology Laboratory, RIKEN, Wako,
Saitama 351-0198, Japan;†Department of Biotechnology, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657,
Japan;§Department of Developmental Biology and Neurosciences, Graduate School of Life Sciences, Tohoku
University, Miyagi 980-8578, Japan;?Department of Biophysics, Graduate School of Science, Kyoto University,
Kyoto 606-8502, Japan; and#Institut National de la Sante ´ et de la Recherche Me ´dicale U870, Institut National
de la Recherche Agronomique U1235, Institut National des Sciences Appliquées de Lyon, University Lyon 1
and Hospices Civils de Lyon, 69621 Villeurbanne, France
Submitted October 17, 2006; Revised April 20, 2007; Accepted April 24, 2007
Monitoring Editor: Robert Parton
Cellular cholesterol increases when cells reach confluency in Chinese hamster ovary (CHO) cells. We examined the
endocytosis of several lipid probes in subconfluent and confluent CHO cells. In subconfluent cells, fluorescent lipid
probes including poly(ethylene glycol)derivatized cholesterol, 22-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-23,24-
bisnor-5-cholen-3?-ol, and fluorescent sphingomyelin analogs were internalized to pericentriolar recycling endo-
somes. This accumulation was not observed in confluent cells. Internalization of fluorescent lactosylceramide was not
affected by cell confluency, suggesting that the endocytosis of specific membrane components is affected by cell
confluency. The crucial role of cellular cholesterol in cell confluency–dependent endocytosis was suggested by the
observation that the fluorescent sphingomyelin was transported to recycling endosomes when cellular cholesterol
was depleted in confluent cells. To understand the molecular mechanism(s) of cell confluency– and cholesterol-
dependent endocytosis, we examined intracellular distribution of rab small GTPases. Our results indicate that rab11
but not rab4, altered intracellular localization in a cell confluency–associated manner, and this alteration was
dependent on cell cholesterol. In addition, the expression of a constitutive active mutant of rab11 changed the
endocytic route of lipid probes from early to recycling endosomes. These results thus suggest that cholesterol
controls endocytic routes of a subset of membrane lipids through rab11.
Little is known about the mechanisms of internalization of
plasma membrane lipids and lipid domains. Recent studies
using various lipid probes revealed different endocytic
pathways of plasma membrane lipids (Kok et al., 1991;
Mukherjee et al., 1999; Puri et al., 2001; Hao et al., 2002;
Maxfield and Wustner, 2002; Pagano, 2003; Sato et al.,
2004). Targeted trafficking of lipid probes might be due to
the differential partitioning preference of these lipids and
lipid analogs in coexisting lateral membrane domains of
varying chemical composition and physical properties
(Mukherjee et al., 1999). Alteration of cellular cholesterol
level affects the endocytosis of lipid analogs. The addition
of cholesterol stimulates the uptake of fluorescent lacto-
sylceramide (LacCer) analog (Sharma et al., 2004). It is also
reported that different lipid probes are differently affected
by cholesterol. In normal skin fibroblasts, both fluorescent
LacCer and fluorescent sphingomyelin are internalized
into the Golgi apparatus. However, fluorescent LacCer,
but not sphingomyelin, is transported to the late endosomes/
lysosomes in Niemann-Pick fibroblasts or when cellular cho-
lesterol is elevated in normal skin fibroblasts (Puri et al., 1999,
2001). When cholesterol is depleted, long-chain 1,1?-dioctade-
cyl-3,3,3?,3?-tetramethylindocarbocyanine perchlorate (DiI) an-
alogs are sorted to the recycling pathway instead of going to
the late endosomes/lysosomes, whereas the endocytosis of
short chain DiI analog is inhibited under these conditions (Hao
et al., 2004).
Rab proteins also regulate lipid transport. Overexpression
of rab11 results in accumulation of cholesterol, pyrene-la-
beled sphingomyelin analog, and sulfatide in rab11-positive
organelles (Holtta-Vuori et al., 2002). Esterification of choles-
terol is inhibited under these conditions. Transfection of
dominant-negative form of rab7 or rab9 in normal human
fibroblasts disrupts Golgi targeting of the fluorescent LacCer
analog (Choudhury et al., 2002). Interestingly, overexpres-
sion of wild-type rab7 and rab9 in Niemann-Pick fibroblasts
results in correction of fluorescent LacCer trafficking and
reductions in intracellular cholesterol stores (Choudhury
et al., 2002; Walter et al., 2003; Narita et al., 2005). Similarly,
overexpression of rab8 rescues the late endosomal choles-
This article was published online ahead of print in MBC in Press
on May 2, 2007.
DThe online version of this article contains supplemental material
at MBC Online (http://www.molbiolcell.org).
Address correspondence to: Toshihide Kobayashi (email@example.com).
Abbreviations used: PEG-cholesterol, poly(ethylene glycol)-derivat-
ized cholesterol; NBD-SM, N-((6-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)
amino)hexanoyl)sphingosylphosphorylcholine; BODIPY-SM, N-(4,4-
sphingosylphosphorylcholine; NBD-cholesterol, 22-(N-(7-nitrobenz-2-
© 2007 by The American Society for Cell Biology2667
terol deposition and sphingolipid mistargetting in Niemann-
Pick fibroblasts (Linder et al., 2007). It is also reported that the
microinjection of rab-guanine nucleotide dissociation inhibitor
(GDI) inhibits late endosomal/lysosomal cholesterol mobiliza-
tion (Holtta-Vuori et al., 2000).
Recent results indicate that one of the consequences of
cholesterol effects on lipid traffic is the alteration of intracel-
lular distribution of rab proteins. The cholesterol-accumu-
lating drug U18666A increases the amounts of membrane-
associated rab7 and inhibits rab7 membrane extraction by
GDI (Lebrand et al., 2002). The levels of rab4 were increased
1.5–2-fold in membrane fractions from Niemann-Pick fibro-
blasts, compared with normal human skin fibroblasts
(Choudhury et al., 2004). It is also shown that the capacity of
rab5 and rab9 for GDI-mediated extraction from membranes
is reduced in Niemann-Pick fibroblasts (Ganley and Pfeffer,
2006). The authors also demonstrated that cholesterol stabi-
lizes rab proteins directly in liposome experiment.
It is intriguing that different rab proteins are affected by
cholesterol-accumulating disease and cholesterol-accumu-
lating reagent. Although Niemann-Pick mutations and
U18666A accumulate cholesterol in late endosomes/lyso-
somes, U18666A causes pleiotropic effects on cellular cho-
lesterol homeostasis (Liscum and Underwood, 1995). These
results suggest that different rab proteins are affected by
different modification of cellular cholesterol. These results
also suggest that a new method of modifying cholesterol
might uncover a new role of cholesterol in regulating rab
proteins as well as lipid traffic. It is reported that cholesterol
increase three to four times upon reaching confluence in
cultured endothelial cells (Cansell et al., 1997; Corvera et al.,
2000). In the present study, we found that the content and
distribution of cellular cholesterol were altered when cells
approach confluency in Chinese hamster ovary (CHO) cells.
Comparison of subconfluent and confluent cells provides
another system to examine the role of cholesterol on lipid
traffic without using drugs or mutations involved in choles-
terol homeostasis. We examined the internalization of sev-
eral lipid probes in subconfluent and confluent CHO cells.
Interestingly, the internalization of a subset of lipid probes
was altered by cell confluency. This alteration was accom-
panied by redistribution of rab11. Cholesterol depletion
from confluent cells diminished these alterations. These re-
sults suggest that cholesterol controls endocytic routes of a
subset of lipid components through rab11.
MATERIALS AND METHODS
Chemicals and Antibodies
Poly(ethylene glycol)-derived cholesterol (PEG-cholesterol) was prepared as
described (Sato et al., 2004). Filipin was purchased from Polysciences (War-
rington, PA). 22-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-23,24-bisnor-
5-cholen-3?-ol (NBD-cholesterol), N-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-
rylcholine (NBD-SM), N-(4,4-difluoro-5,7-dimethyl-4-bora-3a,4a-diaza-s-in-
dacene-3-pentanoyl)sphingosyl 1-?-d-lactoside (BODIPY-LacCer), Alexa Fluor
546–conjugated transferrin and Alexa Fluor-conjugated second antibodies were
from Molecular Probes (Eugene, OR). Rabbit anti-?-tubulin antibody and li-
poprotein-deficient bovine serum (LPDS) were from Sigma (St. Louis, MO), goat
anti-aldolase A polyclonal IgG was from Santa Cruz Biotechnology (Santa Cruz,
CA). All other antibodies were form BD Transduction Laboratories (San Diego,
CA). Methyl-?-cyclodextrin (M?CD) was from Cyclolab (Budapest, Hungary).
[32P]orthophosphate was from Perkin Elmer (Boston, MA). Phosphate-free Dul-
becco’s modified Eagle’s medium (DMEM) and dialyzed fetal bovine serum
were from GIBCO Invitrogen (Carlsbad, CA).
pDsRed-Monomer-Golgi (human galactosyltransferase) vector (DsRed-GalT)
was purchased from BD Clontech (Palo Alto, CA). pEGFP-C1-rab4, -rab5, -rab7,
and -rab11 were constructed as described previously (Fukuda, 2003; Tsuboi and
Fukuda, 2006). A mutant rab11 containing an S25N or Q70L substitution was
produced by conventional or two-step PCR techniques using the following
mutagenic oligonucleotides with a restriction enzyme site (underlined) and sub-
stituted nucleotides (in boldface type) as described previously (Fukuda et al.,
primer, sense), 5?-CTCGAGCCCTGCTGTGTCCCATAT-3? (Q70L primer 1, an-
tisense), and 5?-CTCGAGCGGTACAGGGCTATAACG-3? (Q70L primer 2,
sense). The mutant rab11 fragments were subcloned into the BglII/EcoRI site of
the pEGFP-C1 vector (BD Clontech) or the BamHI/NotI site of the pmRFP-C1
(BD Clontech) and verified by DNA sequencing as described previously (Tsuboi
and Fukuda, 2005, 2006).
cDNA encoding mouse GDI? was amplified from Marathon-Ready adult
mouse brain cDNA (BD Clontech) by PCR using the following pairs of
oligonucleotides with a BamHI linker (underlined) or a stop codon (italics) as
described previously (Fukuda et al., 1999): 5?-GGATCCATGGATGAGGAAT-
ACGATGT-3? (GDI?-Met primer; sense) and 5?-TCACTGATCAGCTTCTC-
CAA-3? (GDI?-stop primer; antisense). Purified PCR products were directly
inserted into the pGEM-T Easy vector (Promega, Madison, WI) as described
previously (Fukuda et al., 1999). After verification by DNA sequencing, cDNA
inserts were transferred to the pEF-T7 tag mammalian expression vectors
(modified from pEF-BOS; Fukuda et al., 1999), and the resultant plasmids are
referred to as pEF-T7-GDI?.
The CHO-K1 cell line was obtained from the American Type Culture Collec-
tion (ATCC; Manassas, VA). Cells were maintained in Ham’s F-12 medium
supplemented with 10% fetal calf serum (FCS), 100 U/ml penicillin, and 100
?g/ml streptomycin at 37°C.
Cellular Content of Cholesterol and Cholesteryl Ester
Cells, 1.9 ? 105, were cultured in 140-mm dishes (to prepare subconfluent cell
culture) and in 12-well plates (?23 mm, to obtain confluent cells) for 2 d. Cells,
1.28 ? 106, were cultured in 60-mm dishes for 2 d to prepare cells treated with
M?CD and LPDS.
Cells were then washed with phosphate-buffered saline (PBS) and scraped.
A part of cell suspension was used for the determination of protein concen-
trations using a Bradford protein assay kit (Bio-Rad, Richmond, CA) or BCA
protein assay kit (Pierce, Rockford, IL). Total lipids were extracted by the
method of Bligh and Dyer (1959) and then separated by TLC with a solvent of
hexane/diethyl ether/acetic acid (80:20:2, vol/vol). The lipids were visual-
ized with phosphomolybdic acid (Nacalai Tesque, Kyoto, Japan), and choles-
terol and cholesteryl ester were quantified by using a LAS1000plus lumines-
cent imaging analyzer (Fuji Film, Tokyo, Japan).
For filipin staining, 8.0 ? 104cells were cultured on coverslips in 90-mm
dishes (subconfluent) and 24-well plates (?15 mm; confluent) for 2 d. The
coverslips were rinsed in ice-cold PBS and fixed with 3% paraformaldehyde
for 10 min at room temperature, followed by treatment with 50 mM NH4Cl in
PBS for 10 min and then were permeabilized with 50 ?g/ml digitonin in PBS
for 5 min. Cells were then doubly labeled with 50 ?g/ml filipin and rabbit
anti-?-tubulin antibody in PBS followed by Alexa 546–conjugated anti-rabbit
IgG. rab11 or rab4 staining was performed by the method of Chen et al. (1998)
with a modification. Cells were washed with ice-cold PBS and permeabilized
with 0.05% saponin (Sigma, St. Louis, MO) in 80 mM PIPES-KOH (pH 6.8), 5
mM EGTA, and 1 mM MgCl2for 5 min on ice and washed once with ice-cold
PBS. Cells were fixed, followed by NH4Cl treatment and then labeled with
mouse anti-rab11 or -rab4 mAb overnight at 4°C. After the excess anti-rab11
and -rab4 antibodies were washed away, cells were incubated with Alexa 488–
conjugated anti-mouse IgG. For double staining with anti-rab11 and -GM130
antibodies, cells were labeled with anti-rab11 antibody and Alexa Fluor-conju-
gated secondary antibody, followed by labeling with fluorescein isothiocyanate–
conjugated anti-GM130 antibody.
Microscopy and Quantitation of Colocalization
The confocal images of cells doubly labeled with BODIPY-LacCer and DsRed-
GalT were acquired on Fluoview FV1000 confocal microscope equipped with
PLAPO 60XOLSM (1.1 NA) objective (Olympus, Tokyo, Japan). Other speci-
mens were observed under LSM 510 confocal microscope equipped with
C-Apochromat 63XW Korr (1.2 NA) objective (Carl Zeiss, Oberkochen, Ger-
many). To quantitate colocalization, the Pearson correlation coefficient (r) was
determined using LSM Image Examiner software version 3.2 (Carl Zeiss).
?ch1i? ch1aver? ? ?ch2i? ch2aver?
M. Takahashi et al.
Molecular Biology of the Cell 2668
where ch1i and ch2i are the red and green intensities of voxel i, respec-
tively, and ch1aver and ch2aver the average value of ch1i and ch2i, respec-
tively. There is a positive correlation when r is higher than 0.2, whereas r
between ?0.2 and 0.2 indicates no correlation.
Incorporation of Fluorescent PEG-Cholesterol
Incorporation of fluorescent PEG-cholesterol was performed as described
(Sato et al., 2004) with a modification. Cells, 1.3 ? 103, were grown in two-well
chambers (4.2-cm2area; to obtain subconfluent cells; Lab-Tek Chambered
Coverglass, Nalge Nunc International, Rochester, NY), or 8.6 ? 103cells were
grown in eight-well chambers (0.8-cm2area; confluent cells) for 3 d. In Figure
7, 8.0 ? 104cells were grown in glass-bottom 24-well plates (Iwaki, Tokyo,
Japan) for 2 d. Cells were incubated with 0.5 ?M fluorescent PEG-cholesterol
in DMEM/Nutrient mixture Ham’s F-12, without phenol red (DMEM F-12,
Sigma) for 1 min (subconfluent) or 5 min (confluent) at 15°C. Five minutes is
required for confluent cells because of the low efficiency of insertion of
PEG-cholesterol. After excess PEG-cholesterol was washed with DMEM F-12,
cells were incubated for 15 min at 37°C. The specimens were treated with 10
mM NH4Cl in DMEM F-12 medium before acquiring image. When doubly
labeled with BODIPY-SM or Alexa 546–conjugated transferrin (Tf), cells were
preincubated with 4 ?M BODIPY-SM or 50 ?g/ml Alexa-546 Tf in serum-free
Ham’s F-12 medium supplemented with 0.2% BSA for 30 min at 15°C. After
washing, cells were labeled with PEG-cholesterol and incubated for a further
15 min at 37°C.
Incorporation of NBD-Cholesterol
Cells were incubated with 5 ?g/ml NBD-cholesterol in 10% FCS containing
Ham’s F-12 medium. After 30 min at 10°C, cells were washed with DMEM
F-12, followed by further incubation for 5 min at 37°C.
Incorporation of BODIPY-SM, NBD-SM,
Cells were incubated with 4 ?M BODIPY-SM or NBD-SM in 10% FCS con-
taining Ham’s F-12 medium. When cells were treated with M?CD or LPDS,
BODIPY-SM was dispersed in 0.03 mg/ml fatty acid–free bovine serum
albumin (BSA; Sigma) containing DMEM F-12. After 30 min at 10°C, cells
were washed with DMEM F-12, followed by further incubation for 15 min at
37°C. Incorporation of BODIPY-LacCer was performed by the method of
Sharma et al. (2004) with a modification. Cells were incubated with 2 ?M
BODIPY-LacCer in 10% FCS containing Ham’s F-12 medium for 30 min at
10°C, washed with DMEM F-12, and further incubated for 15 min at 37°C,
followed by back-exchange with 5% fatty acid free BSA.
Recycling Assay of NBD-SM
Cells, 8.0 ? 104, were seeded in 90-mm dishes (subconfluent) and 24-well
plates (confluent) and grown for 2 d. Internalization of NBD-SM at 37°C
was performed by incubating cells with 4 ?M NBD-SM for 30 min at 10°C,
followed by washing with DMEM F-12 at 10°C and further incubation with
prewarmed DMEM F-12 medium for 10 min at 37°C. For 16°C internal-
ization, cells were labeled with NBD-SM for 60 min at 16°C and were
washed with DMEM F-12 at 16°C. After internalization, cells were treated
with 50 mM sodium dithionite (Nacalai Tesque) to quench the fluorescent
lipids at the cell surface (Kobayashi et al., 1992), followed by washing.
Then 1 ml of 37°C chase medium (5% fatty acid free BSA in DMEM-F-12
medium) was added. This was defined as chase time 0. Cells were chased
in the chase medium at 37°C for up to 60 min. At each time point, chase
medium was collected, and prewarmed fresh chase medium was added to
cells. At the end of a 60-min chase, cells were washed and scraped. The
fluorescent lipids of chase medium and harvested cells were extracted by
the method of Bligh and Dyer (1959). The fluorescent lipids were dissolved
with ethanol, and fluorescence measurements were performed using a
FP-6500 spectrofluorometer (Jasco, Tokyo, Japan).
Cells grown to 90% confluence in 60- or 90-mm dishes were transiently
transfected using Lipofectamine 2000 (Invitrogen) with 2 or 4 ?g plasmid
DNA, respectively. Transfection was carried out according to the manufac-
turer’s instructions. After transfection, cells were incubated overnight at 37°C
and then replated.
Analysis of GDP and GTP-bound to rab11
Cells, 2.0 ? 105, transfected with EGFP-rab11 were seeded in 140-mm dishes
or 12-well plates, and cells were grown for 2 d. Analysis of GDP and
GTP-bound to rab11 was performed by the method of Satoh et al. (1988) with
a modification. Cells were labeled for 3 h at 37°C with 0.33 mCi/ml
[32P]orthophosphate in phosphate-free DMEM supplemented with 10% dia-
lyzed fetal bovine serum. The cells were washed, scraped, and lysed with
immunoprecipitation buffer (50 mM Tris-HCl, pH 7.5, 20 mM MgCl2, 150 mM
NH4Cl, 0.5% NP-40) containing protease inhibitor cocktail (Calbiochem, La-
Jolla, CA). The cell lysate was immunoprecipitated using rabbit anti-GFP
antibody (Molecular Probes) and protein G-Sepharose, and the immuno-
precipitate was resolved by polyethyleneimine-cellulose TLC (Merck, Rah-
way, NJ) with a solvent of 1.0 M LiCl/1.8 M formic acid. Radiolabeled GDP and
GTP were detected with BAS2000 Bio-imaging analyzer (Fuji Film). The molar
ratio of rab-bound GTP was calculated as [GTP]/([GDP] ? 1.5 ? [GTP]) ? 100.
M?CD and LPDS Treatment
Cells were rinsed in serum-free medium three times followed by the incuba-
tion with 10 mM M?CD in serum-free medium for 15 min at 37°C. For LPDS
treatment, cells grown for 1 d were incubated with Ham’s F-12 medium
supplemented with 5% LPDS for 20 h at 37°C.
Subcellular Fractionation and Immunoblotting
Cells, 1.28 ? 106, were seeded in 60-mm dishes and grown for 2 d. Cells were
treated with M?CD or LPDS as described above. For fractionation of mem-
brane and cytosol and for immunoblot, cells were washed three times with
PBS, scraped in ice-cold homogenization buffer (10 mM HEPES, 250 mM
sucrose containing protease inhibitor cocktail), and homogenized using a
Handy microhomogenizer (NS-310E, MICROTEC Co., Chiba, Japan). Then
fractionation and immunoblot were performed as described (Chen et al.,
Content and Distribution of Cholesterol Alter in a Cell
Confluency–associated Manner in CHO Cells
We prepared subconfluent and confluent cultures of CHO
cells by seeding the same number of cells to dishes of dif-
ferent sizes. As observed in endothelial cells (Cansell et al.,
1997; Corvera et al., 2000), both cholesterol and cholesteryl
ester increased as cells reached confluency (Table 1). The
free cholesterol–specific antibiotic filipin (Sokol et al., 1988;
Sato et al., 2004) weakly labeled the pericentriolar compart-
ment, which colocalized with a centriole marker, ?-tubulin,
in subconfluent cells (Figure 1). The pericentriolar staining
pattern is reminiscent of the recycling endosome of CHO
Table 1. Cholesterol and cholesteryl ester content of subconfluent
and confluent cells
25.6 ? 1.8
46.5 ? 8.7
3.5 ? 0.3
9.2 ? 1.1
Numbers are mean of triplicate experiments ? SE, expressed as
Subconfluent and confluent cells were doubly labeled with filipin
and anti-?-tubulin. Bar, 10 ?m. Images were obtained under the
same acquisition conditions (exposure time 1.97 s with laser power
of 364 nm:5% and 543 nm:25%).
Cholesterol increases when CHO cells reach confluence.
Cholesterol Regulates Rab11
Vol. 18, July 2007 2669
cells (Ullrich et al., 1996). Filipin staining was much brighter
in confluent cells. In confluent cells, in addition to pericen-
triolar fluorescence, the cell surface as well as perinuclear
region, was intensely labeled with filipin.
Cell Confluency Alters Internalization of Fluorescent
The above results indicate that the content and distribution
of cellular cholesterol are significantly altered by cell con-
fluency in CHO cells. In this study, we examined the inter-
nalization of several fluorescent lipid probes under these
conditions (Figures 2 and 3). PEG derivative of cholesterol
(PEG-cholesterol) preferentially partitions to specific mem-
brane domains (Sato et al., 2004). The bulk PEG moiety of
PEG-cholesterol prevents transbilayer movement of the mol-
ecule. This inhibits the transport of PEG-cholesterol to the
cytoplasmic leaflet and the nonspecific diffusion of the mol-
ecule to the cytoplasm when the molecule is inserted to the
cell surface. Thus PEG-cholesterol could selectively monitor
the endocytic pathway of plasma membrane subdomains.
Subconfluent and confluent CHO cells were briefly labeled
with the fluorescein ester of PEG-cholesterol (fluorescein-
PEG-cholesterol), followed by washing excess compound
and incubation at 37°C for 15 min. In subconfluent cells,
internalized fluorescein-PEG-cholesterol was accumulated
at the pericentriole region of cells, suggesting the transport
of cell surface fluorescein-PEG-cholesterol to recycling en-
dosomes. In contrast, in confluent cells, fluorescein-PEG-
cholesterol was distributed in small vesicles throughout the
cytoplasm. Although PEG-cholesterol selectively inhibits clath-
rin-independent endocytosis at high concentration, under our
experimental conditions the presence of PEG-cholesterol did
not inhibit internalization of fluorescent dextran, which is in-
ternalized via clathrin-independent mechanism (Sabharanjak
et al., 2002; Supplementary Figure 1). The recycling endosomes
are characterized by the accumulation of transferrin internal-
tion of fluorescent PEG-cholesterol and NBD-
cholesterol. (A) Subconfluent and confluent
CHO cells transfected with human transferrin
receptor were incubated with Alexa 546–con-
jugated transferrin (Tf) and fluorescein-PEG-
cholesterol (PEG-Chol) for 15 min at 37°C.
Arrows indicate pericentriolar recycling endo-
somes where incorporated transferrin was ac-
cumulated. Bar, 10 ?m. (B) The correlation
coefficients between the localization of inter-
nalized PEG-cholesterol and Alexa 546–con-
jugated Tf were calculated in subconfluent
and confluent cells. ?, subconfluent cells; E,
confluent cells. Dotted lines indicate the aver-
age values of subconfluent (0.262) and conflu-
ent cells (0.059). (C) Confluent CHO cells
transfected with GFP-rab5 (rab5) or GFP-rab7
(rab7) were incubated with TRITC-PEG-cho-
lesterol (PEG-Chol) for 15 min at 37°C. Arrow-
heads indicate colocalization of rab5 and
TRITC-PEG-cholesterol. The enclosed areas
are enlarged below. Bar, 10 ?m. (D) The cor-
relation coefficients between the localization of
TRITC-PEG-cholesterol and GFP-rab5 or GFP-
rab7 were calculated. ?, rab5 and PEG-cho-
lesterol; E, rab7 and PEG-cholesterol. Dotted
lines indicate the average values of rab5 and
PEG-cholesterol (0.297), or rab7 and PEG-cho-
lesterol (0.185). (E) Subconfluent and confluent
CHO cells were incubated with NBD-choles-
terol for 5 min at 37°C as described in Materials
and Methods. Bar, 10 ?m.
Cell confluency alters internaliza-
M. Takahashi et al.
Molecular Biology of the Cell2670
whether fluorescein-PEG-cholesterol was accumulated in recy-
cling endosomes, we transfected human transferrin receptor to
CHO cells, followed by doubly labeling cells with fluorescein-
PEG-cholesterol and Alexa-546–labeled transferrin. Fluores-
cent transferrin was accumulated in the central region of
the cells both in subconfluent and confluent culture. In sub-
confluent cells, internalized transferrin was colocalized with
endocytosed fluorescein-PEG-cholesterol (Figure 2A). We cal-
culated Pearson correlation coefficient to quantitate the colocal-
ization between fluorescein-PEG-cholesterol and transferrin in
subconfluent and confluent cells (Figure 2B). Clearly there is a
higher correlation in subconfluent cells than in confluent cells,
indicating that fluorescein-PEG-cholesterol is colocalized with
transferrin in subconfluent cells, but not in confluent cells
(Figure 2B). This result suggests that the endocytic route of
fluorescein-PEG-cholesterol, but not of transferrin, alters in a
cell confluency-associated manner.
rab5 and rab7 are known to localize in the early and late
endosomes, respectively (Chavrier et al., 1990; Soldati et al.,
1995). In Figure 2C, confluent cells transfected with green
fluorescent protein (GFP)-labeled rab5 or rab7, were incu-
bated with tetramethylrhodamine B isothiocyanate (TRITC)-
labeled PEG-cholesterol. TRITC-PEG-cholesterol was inter-
nalized similarly to fluorescein-PEG-cholesterol (Sato and
Kobayashi, unpublished results). The colocalization of GFP-
rab5 and TRITC-PEG-cholesterol indicates that PEG-choles-
terol is internalized into early endosomes in confluent cells
(Figure 2C, arrowheads). In contrast to rab5, transfected
rab7, a late endosome marker, was not significantly colocal-
ized with internalized PEG-cholesterol (Figure 2C). Correla-
tion coefficients between PEG-cholesterol and rab5 or rab7
show that positive correlation occurs only between rab5 and
PEG-cholesterol (Figure 2D).
In Figure 2E, subconfluent and confluent CHO cells
were labeled with 22-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-
at 10°C, followed by washing excess compound and in-
cubation at 37°C for 5 min. It is suggested that NBD-
cholesterol is internalized by fast lipid binding protein–
mediated process and by slow endocytic pathway (Frolov
et al., 2000). When cells were incubated at low tempera-
ture, fluorescence was observed throughout cytoplasm in
both subconfluent and confluent cells (not shown). After
warming up, NBD-cholesterol was also accumulated at
the pericentriole region in subconfluent cells. In contrast,
pericentriole labeling was not observed in confluent cells.
Quantitation indicates that 76.4% (185 of 242) of subcon-
fluent cells internalized NBD-cholesterol in pericentriolar
region, whereas this value is decreased to 28.0% (122 of
436 cells) in confluent cells.
In Figure3A,we compared
BODIPY-SM with that of TRITC-PEG-cholesterol in subconflu-
ent and confluent CHO cells. The colocalization of fluorescence
indicates the internalization of both BODIPY-SM and PEG-
cholesterol alters in a cell confluency-dependent manner. We
also examined the internalization of BODIPY-LacCer in pD-
sRed-galactosyltransferase (DsRed-GalT) transfected CHO
cells. GalT is localized in the Golgi apparatus (Sciaky et al.,
1997). BODIPY-LacCer was rapidly incorporated to the DsRed-
GalT–positive region in both subconfluent and confluent cells
(Figure 3B). These results indicate that cell confluency affects
the internalization route of limited subset of molecules.
Fluorescent SM undergoes rapid recycling in CHO cells
(Koval and Pagano, 1989; Mayor et al., 1993; Hao and
Maxfield, 2000). We compared the recycling kinetics of
fluorescent SM in subconfluent and confluent cells. In this
experiment, we used NBD-SM. NBD fluorescence is
readily removed from the cell surface by serum albumin
or water soluble quencher (Koval and Pagano, 1989;
Kobayashi et al., 1992). It is reported that the bulk of the
molecules internalized at 37°C are recycled via the recy-
cling endosomes, whereas those internalized at 16°C are
recycled back to the plasma membrane from early endo-
somes without passing through the recycling endosomes
(Ren et al., 1998). In Figure 3C, NBD-SM was internalized
either at 16 or 37°C, and the recycling of fluorescent lipid
to the plasma membrane was monitored at 37°C. The
fraction of lipid that was cell-associated at each chase time
point was fit with a double exponential decay. Theoretical
fits were then plotted as solid lines. As previously de-
fluorescent SM but not the internalization of fluorescent LacCer. (A)
Subconfluent and confluent cells were treated with TRITC-PEG-
cholesterol (PEG-Chol) and BODIPY-SM for 15 min at 37°C. Bar,
10 ?m. (B) Subconfluent and confluent cells transfected with DsRed-
galactosyltransferase (GalT) were treated with BODIPY-LacCer
(LacCer) for 15 min at 37°C. Bar, 10 ?m. (C) Recycling of NBD-SM
in subconfluent and confluent CHO cells. Recycling was measured
as described in Materials and Methods. The percentage of cell-asso-
ciated fluorescence was calculated for each time point. Double ex-
ponential decay fits are plotted. ?, subconfluent cells; E, confluent
cells. Data points were derived from an average of triplicate exper-
iments. Kinetic parameters from Figure 3 are listed in Table 2.
Cell confluency alters internalization and recycling of
Cholesterol Regulates Rab11
Vol. 18, July 20072671
scribed (Hao and Maxfield, 2000), a double exponential
decay accounts well for the data. When NBD-SM was
internalized at 37°C, the decay curves were slightly but
significantly different between subconfluent and confluent
cells. Although the rapid recycling component accounted
for 64% of the recycled population in subconfluent cells,
this component decreased to 41% in confluent cells (Table
2). These results suggest that the confluency-dependent
transport of NBD-SM to the different destinations results
in the altered efficiency of lipid recycling. The decay
curves were not significantly altered between subconflu-
ent and confluent cells when the fluorescent lipid was
internalized at 16°C (Figure 3C and Table 2).
Rab11 Changes Intracellular Distribution in Cell
Rab proteins play crucial roles in intracellular membrane
traffic (Zerial and McBride, 2001). Localization to distinct
compartments is a prerequisite for rab proteins to catalyze
proper vesicular transport. We investigated whether cell
confluency affects the intracellular distribution of rab4 and
rab11, which are involved in the recycling of internalized
lipids and proteins (van der Sluijs et al., 1992; Daro et al.,
1996; Choudhury et al., 2004; Maxfield and McGraw, 2004).
Cell confluency did not significantly alter intracellular dis-
tribution of rab4 (Figure 4A). In contrast, intracellular dis-
tribution of rab11 was dramatically changed when cells
reached confluence. In subconfluent cells, endogenous rab11
was diffusely distributed throughout the cell with partial
enrichment in the perinuclear region which is partially co-
localize with the Golgi marker, GM130 (Figure 4, A and B).
Transfected GFP-rab11 was similarly distributed, whereas
the enrichment in the perinuclear region was more promi-
nent. In marked contrast, both endogenous and GFP-rab11
are concentrated in the pericentriolar region and colocalize
with ?-tubulin in confluent cells (Figure 4, A and B). Quan-
titation indicates that 13.5% (n ? 163) of subconfluent cells
show pericentriolar localization of endogenous rab11,
whereas 74.3% (n ? 261) of confluent cells accumulate rab11
in pericentriolar region.
Constitutive Active Form of rab11 Alters Internalization
of Lipid Probes in Confluent Cells
Our results indicate simultaneous change of intracellular
distribution of rab11 and the endocytosis of fluorescent lipid
probes when cells approaching confluence. These results
suggest that observed alteration of internalization of lipid
probes is caused by the alteration of the rab11 activity.
Previously it was shown that the transport of transferrin
from the early endosomes to the recycling endosomes is
slowed down by the overexpression of rab11S25N (domi-
nant negative). In contrast, overexpression of rab11Q70L
(constitutively active) induces marked accumulation of
transferrin in the recycling compartment (Ullrich et al., 1996;
Ren et al., 1998). Expression of GFP-rab11Q70L results in the
accumulation of the protein in pericetriolar recycling endo-
somes as described previously (Ullrich et al., 1996; Ren et al.,
1998). In GFP-rab11Q70L transfected cells, internalized
TRITC-PEG-cholesterol was colocalized with GFP-rab11Q70L
(Figure 5A, arrows). Among 14 transfected cells, GFP-
rab11Q70L colocalized with internalized PEG-cholesterol in 13
cells. This is a marked contrast to the nontransfected cells in
which TRITC-PEG-cholesterol was accumulated in small ves-
icles distributed in the cytoplasm (Figure 5A, small arrows).
Incorporation of TRITC-PEG-cholesterol was not significantly
affected by the expression of GFP-rab11S25N. Among 13 trans-
fected cells, only one cell showed pericentriolar distribution of
PEG-cholesterol. Similar to PEG-cholesterol, expression of
RFP-rab11Q70L caused the accumulation of NBD-cholesterol
to the pericentriolar compartment (Figure 5B, arrow). All 29
RFP-rab11Q70L–positive pericentriolar compartment. Intracel-
lular distribution of NBD-cholesterol was not affected by ex-
pression of RFP-rab11S25N as observed in the incorporation of
PEG-cholesterol. Twelve cells showed the accumulation of
NBD-cholesterol in centriolar compartment in 60 transfected
cells (20%). This is close to the value (28%) of nontransfected
cells (Figure 2E). These results suggest that rab11 regulates the
fate of the internalization of lipid probes and the activity of
rab11 is modified when cells reach confluence.
Intracellular Distribution of rab11 and Internalization of
Fluorescent Sphingomyelin Are Changed by Partial
Depletion of Cholesterol from Confluent Cells
Accumulation of cholesterol in confluent cells suggests that
cholesterol may play a role in the redistribution of rab11 and
the alteration of the endocytic pathway of lipid probes.
Cellular cholesterol content is modified by acute cholesterol
depletion using M?CD (Rothblat et al., 1999; Sato et al., 2004)
or by growing cells in LPDS. We then investigated the
intracellular distribution of rab11 and the endocytic path-
way of fluorescent SM after confluent cells were treated with
M?CD and LPDS (Figure 6A). M?CD treatment resulted in
redistribution of rab11 to diffuse localization in the cyto-
plasm and partial enrichment in the perinuclear region (top
row, arrows). This was accompanied by the internalization
of BODIPY-SM to the pericentriolar region of the cell (bot-
tom row, arrows). This internalization is similar to that
observed in subconfluent cells. Internalized BODIPY-SM
was observed in pericentriolar region in 49.5% (44 of 200) of
M?CD-treated cells. In contrast, 6.3% (10 of 158) of control
Table 2. Kinetic parameters of the recycling of NBD-SM in subconfluent and confluent CHO cells
a b (min?1)
c d (min?1)
0.59 ? 0.07
0.62 ? 0.07
0.16 ? 0.01 (4.4)
0.17 ? 0.01 (4.2)
0.41 ? 0.07
0.41 ? 0.07
0.060 ? 0.005 (11.6)
0.063 ? 0.005 (11.0)
0.64 ? 0.02
0.41 ? 0.04
0.19 ? 0.01 (3.7)
0.33 ? 0.13 (2.1)
0.36 ? 0.02
0.59 ? 0.04
0.048 ? 0.002 (14.2)
0.043 ? 0.002 (16.1)
Kinetic parameters were obtained by fitting data points to a double exponential decay equation: l ? ae?bt? ce?dt. Numbers are mean of
triplicate experiments ? SD. Half-time values (min) are provided in parentheses.
M. Takahashi et al.
Molecular Biology of the Cell2672
cells accumulated fluorescent SM in pericentriolar region.
When cells were grown in LPDS medium, rab11 was dif-
fusely distributed throughout cytoplasm. Under these con-
ditions, fluorescent lipid was accumulated in the Golgi
marker DsRed-GalT positive perinuclear region (Figure 6, A
and B). The correlation coefficient (Figure 6B) shows pos-
itive correlation between internalized NBD-SM and
DsRed-GalT. These results indicate that cellular level of
cholesterol affects both intracellular distribution of rab11
and the internalization of fluorescent SM.
Rab proteins cycle between the membrane and cytosol.
We then investigated whether cellular level of cholesterol
affects the solubility of rab11. In Figure 6C, the distribution
of rab11 in membrane and cytosol fractions was analyzed by
Western blotting after M?CD or LPDS treatment. We veri-
fied that GS28 (Golgi SNARE), a membrane marker, and
aldolase A, a cytosol marker, were concentrated in mem-
brane and cytosol fractions, respectively (Figure 6C). The
results were quantified and summarized in Table 3. Choles-
terol contents after the treatments are also shown in Table 3.
The soluble form of rab11 was increased from 48.8 ? 3.2%
(mean of several experiments ? SE) to 54.2 ? 2.9 and 63.1 ?
1.2% by treating with M?CD and LPDS, respectively (Figure
6C, Table 3). This result suggests that cholesterol affects the
membrane-cytosol cycle of rab11.
Cholesterol and Cell Confluency Do Not Affect GTP/GDP
Ratio of rab11
Rab GTPases are in equilibrium between the inactive GDP-
bound and activated GTP-bound states. GDI extracts only
GDP-bound form of rab proteins and forms a complex with
rab proteins in GDP-bound form in the cytosol (Alory, 2001;
Pfeffer and Aivazian, 2004). It is possible that cholesterol
increases the soluble form of rab11 by changing GTP/GDP
ratio of rab11. We thus investigated whether GTP/GDP
ratio of rab11 is changed in a cell confluency– or cholesterol-
dependent manner (Table 3, Supplementary Figure 2). In
subconfluent cells, GTP-rab11 comprised 55.0 ? 1.2% (mean
of triplicate ? SE) and in confluent cells, this value was
57.6 ? 1.1%. Similarly, there was also no significant differ-
ence in GTP/GDP ratio of rab11 between control cells and
M?CD treated confluent cells. These results suggest that
difference of the distribution, rather than the GTP-bound
form/GDP-bound form ratio of rab11, is the major conse-
quence of the difference in cell confluency and M?CD-
tion of rab11 but not rab4. (A) Localization of endoge-
nous and GFP-rab4 and -rab11 in subconfluent and
confluent CHO cells. Bar, 10 ?m. (B) Subconfluent and
confluent cells were labeled with anti-rab11, anti-
GM130, and anti-?-tubulin. Bar, 10 ?m.
Cell confluency alters intracellular localiza-
Cholesterol Regulates Rab11
Vol. 18, July 20072673
Overexpression of GDI Redistributes rab11
and Fluorescent Lipid Probes in Confluent Cells
Our results suggest that intracellular distribution of rab11
determines the fate of fluorescent lipid probes. The present
results also suggest that cholesterol controls the distribution
of rab11. One possible explanation of our results is that the
accumulation of cholesterol in confluent cells diminishes the
extraction of rab11 by GDI, as observed in other rab proteins
(Lebrand et al., 2002; Choudhury et al., 2004; Ganley and
Pfeffer, 2006). Excess GDI added in vitro has been demon-
strated to release GDP-bound rab proteins from membranes
by forming a soluble complex (Dirac-Svejstrup et al., 1994;
Peter et al., 1994; Ullrich et al., 1994; Lebrand et al., 2002;
Choudhury et al., 2004; Ganley and Pfeffer, 2006). Chen et al.
(1998) demonstrated that rab11 is selectively released from
membranes by the overexpression of GDI. In Figure 7, we
overexpressed GDI? and examined the intracellular distri-
bution of rab11 and the distribution of endocytosed fluores-
cein-PEG-cholesterol in confluent CHO cells. Overexpres-
sion of GDI? significantly altered the distribution of rab11.
Fluorescence was observed throughout the cytoplasm with
partial enrichment in the perinuclear region. Overexpression
of GDI? slightly increased cytosolic rab11 (42.6 ? 1.9 to
48.0 ? 0.9%, n ? 3). Internalized fluorescein PEG-cholesterol
was observed in pericentriolar region in 5.9% (11 of 186) of
nontransfected cells. This value was increased to 15.0% (32
of 213) by the transfection of GDI?. In addition to PEG-
cholesterol, overexpression of GDI? (40.6%, n ? 653) in-
creased the number of cells with pericentriolar staining pat-
tern of NBD-cholesterol compared with mock-transfected
cells (23.4%, n ? 788).
of PEG-cholesterol and NBD-cholesterol in confluent cells. (A) Con-
fluent cells were transiently transfected with GFP-rab11Q70L (con-
stitutive active) or GFP-rab11S25N (dominant negative). Cells were
then incubated with TRITC-PEG-cholesterol for 15 min at 37°C.
Arrows indicate pericentriolar recycling endosomes where GFP-
rab11Q70L is accumulated, whereas small arrows indicate small
vesicles in cytoplasm. Bar, 10 ?m. (B) Confluent cells transfected
with RFP-rab11Q70L or RFP-rab11S25N were incubated with NBD-
cholesterol for 5 min at 37°C. Arrows indicate pericentriolar recy-
cling endosomes where GFP-rab11Q70L is accumulated. Bar, 10 ?m.
Constitutive active form of rab11 alters the endocytosis
and the internalization of fluorescent SM in confluent cells. (A) Cells
were treated with M?CD or LPDS as described in Materials and
Methods. Cells were then fixed, labeled with anti-rab11 antibody,
BODIPY-SM was measured as described in Materials and Methods.
Arrows indicate diffuse localization of rab11 in cytoplasm and peri-
centriolar accumulation of BODIPY-SM. Bar, 10 ?m. (B) Cells trans-
fected with DsRed-GalT were treated with LPDS. Then, the inter-
nalization of NBD-SM was measured as described in Materials and
Methods. The correlation coefficients between the localization of
NBD-SM and DsRed-GalT were calculated. Dotted line indicates the
average value (0.270). Bar, 10 ?m. (C) After M?CD or LPDS treat-
ment, cells were homogenized and cytosol (C) and membrane (M)
fractions were prepared as described in Materials and Methods. Frac-
tions were analyzed by Western blotting with antibody against
rab11. The amount of GS28 and aldolase A were analyzed as mark-
ers of membrane and cytosol, respectively. The data are calculated
and displayed in Table 3.
(A) M?CD and LPDS treatment alter rab11 distribution
M. Takahashi et al.
Molecular Biology of the Cell2674
Present results indicate that both content and distribution of
cholesterol alters in CHO cells when cells approach conflu-
ency. This is accompanied by the altered internalization of
several lipid probes. Present study suggests that the inter-
nalization of fluorescent lipid probes is dependent on intra-
cellular distribution of rab11, which is modified by changing
cellular level of cholesterol.
Previously it was shown that NBD-SM is internalized into
recycling endosomes together with transferrin in CHO cells
(Koval and Pagano, 1989; Mayor et al., 1993). Our results also
indicate fluorescent SM analogs were internalized together
with fluorescent PEG-cholesterol (Sato et al., 2004) into peri-
centriolar recycling endosomes in subconfluent cells. In ad-
dition, NBD-cholesterol was accumulated into the pericent-
riolar compartments. However, in confluent cells, these
fluorescent lipid probes did not reach the pericentriolar
endosomes, whereas transferrin was transported to pericen-
triolar compartments. BODIPY-LacCer was transported to
the Golgi apparatus both in subconfluent and confluent
cells. These results indicate that cell confluency affects the
endocytic pathways of limited number of plasma membrane
components. It is reported that during endocytosis cargo
moves through distinct domains on endosomes (Sonnichsen
et al., 2000). BODIPY-LacCer is internalized almost exclu-
sively by a clathrin-independent mechanism, whereas
BODIPY-SM is taken up approximately equally by clathrin-
dependent and -independent pathways in human skin fibro-
blasts (Puri et al., 2001). Transferrin is a well-known marker
of clathrin-dependent endocytosis (Hanover et al., 1984). The
above observations suggest that these cargos are distributed
in different membrane domains in early endosomes. The
results thus suggest that the transport of specific lipid do-
mains from early endosomes to recycling endosomes is de-
layed when cells reach confluence.
We demonstrated that the recycling rate of NBD-SM was
delayed in confluent cells when the fluorescent lipid was
internalized at 37°C. However, there was no significant dif-
ference of lipid recycling when NBD-SM was internalized at
16°C. It is reported that the bulk of the molecules internal-
ized at 16°C are recycled back to the plasma membrane from
early endosomes without passing through the recycling
endosomes (Ren et al., 1998). Our results thus suggest that
cell confluency affects the transport of NBD-SM from early
endosomes to the recycling endosomes but not from the
early endosomes to the plasma membrane. Rab11 regulates
recycling through the pericentriolar recycling endosomes
(Ullrich et al., 1996; Ren et al., 1998), whereas rab4 is sug-
gested to play a role in fast recycling from early endosomes
back to the plasma membrane (van der Sluijs et al., 1992;
Sheff et al., 1999). Our results suggest that rab11-mediated
recycling is selectively modified during cell confluency.
Consistent with this, our results indicated that intracellular
distribution of rab11, but not rab4, altered in a cell conflu-
ency–dependent manner. In subconfluent cells, rab11 was
dispersed throughout the cytoplasm in addition to some
enrichment in the perinuclear region. In contrast, rab11 was
concentrated in the pericentriolar recycling endosomes in con-
fluent cells. It has been reported that the distribution of rab11
varies in different cell types (Ullrich et al., 1996; Green et al.,
1997; Casanova et al., 1999; Brown et al., 2000; Sonnichsen et al.,
2000; van Ijzendoorn et al., 2003). Our results suggest that this
is in part because of the cell confluency-dependent redistribu-
tion of rab11.
When constitutive active rab11Q70L was expressed in
confluent cells, the endocytosed fluorescent PEG-cholesterol
and NBD-cholesterol were accumulated in the recycling en-
dosomes. In contrast, dominant negative rab11S25N did not
alter the internalization of these lipid probes. These results
also support the idea that the endocytosis of examined lipid
probes to the recycling endosomes is under the control of
rab11. Holtta-Vuori et al. (2002) demonstrated that the tran-
sient overexpression of rab11 resulted in prominent accu-
mulation of free cholesterol in rab11-positive organelles and
inhibited cellular cholesterol esterification. It is suggested
that in rab11-overexpressing cells, deposition of cholesterol
in recycling endosomes results in its impaired esterification,
presumably due to defective recycling of cholesterol to the
cent cholesterol probes in confluent cells. Confluent CHO cells were
transiently transfected with mock or GDI?. Cells were then fixed
and labeled with anti-rab11. Internalization of fluorescent PEG-
cholesterol or NBD-cholesterol was measured as described in Ma-
terials and Methods. Arrows indicate pericentriolar accumulation of
fluorescence. Bar, 10 ?m.
Overexpression of GDI redistributes rab11 and fluores-
Table 3. Cellular cholesterol content and soluble and GTP-form of rab11 in different conditions
Cholesterol content (nmol/mg protein)
% of soluble form rab11
% of GTP-bound form rab11
57.1 ? 5.3 (6)
48.8 ? 3.2 (8)
59.5 ? 0.7 (3)
50.5 ? 4.2 (5)
54.2 ? 2.9 (5)
56.4 ? 2.1 (3)
46.4 ? 4.0 (4)
63.1 ? 1.2 (4)
% of GTP-bound form rab1155.0 ? 1.2 (3) 57.6 ? 1.1 (3)
Numbers are mean ? SE of several experiments, with n values in parentheses.
Cholesterol Regulates Rab11
Vol. 18, July 20072675
plasma membrane. Our results showed that PEG-cholesterol
and slowly internalized NBD-cholesterol were accumulated
where constitutive active rab11 was expressed. Previously
we showed that in vitro PEG-cholesterol is preferentially
partitions to cholesterol-rich membranes (Sato et al., 2004). It
is shown that NBD-cholesterol is internalized into both mi-
tochondria, where endogenous cholesterol is not accumu-
lated, and recycling endosomes. In recycling endosomes,
NBD-fluorescence was colocalized with the fluorescence of
internalized dehydroergosterol (DHE; Mukherjee et al., 1998).
Precise distribution of PEG-cholesterol and NBD-cholesterol
on the plasma membrane is not known. However, our results,
together with the results of Holtta-Vuori, suggest that PEG-
cholesterol and slowly internalized NBD-cholesterol may rep-
resent, at least in part, the fate of cell surface cholesterol. It has
been reported that the Golgi targeting of internalized BODIPY-
LacCer is dependent on rab7 and rab9 but is independent of
rab11 (Choudhury et al., 2002). This is consistent with our
results that BODIPY-LacCer was transported to the Golgi ap-
paratus in both subconfluent and confluent cells.
Present results indicate that the partial depletion of cho-
lesterol from confluent cells alters intracellular distribution
of rab11, suggesting that cholesterol controls the distribution
of rab11 in a cell confluency–associated manner. M?CD
treatment did not alter GTP-bound form/GDP-bound form
ratio of rab11. However, the significant increase in soluble
form of rab11 was accompanied by cholesterol depletion by
M?CD or LPDS in confluent cells in vivo. It has been re-
ported that accumulation of cholesterol inhibits GDI-depen-
al., 2002; Choudhury et al., 2004; Ganley and Pfeffer, 2006).
When cells were treated with hydrophobic amine, U18666A,
cholesterol is accumulated in late endosomes (Liscum and
Faust, 1989; Kobayashi et al., 1999). This increases the
amounts of late endosome membrane-associated rab7 and
inhibits the extraction of membrane-bound rab7 by GDI
(Lebrand et al., 2002). This results in the loss of bidirectional
mobility of late endosomes. In vitro extraction of rab4, rab9,
and rab5 with GDI is severely attenuated in endosomal
fractions from human skin fibroblasts derived from Niemann-
Pick patients (Choudhury et al., 2004; Ganley and Pfeffer,
2006). In confluent CHO cells, cholesterol was accumulated
in both pericentriolar compartments and the plasma mem-
brane. This may lead to the attenuation of the extraction of
rab11 from the pericentriolar recycling endosomes.
M?CD treatment not only altered the intracellular distribu-
tion of rab11 but changed the internalization of BODIPY-SM in
confluent cells. After treating cells, BODIPY-SM was incorpo-
rated into pericentriolar organelle as observed in subconfluent
cells. These results suggest that cholesterol controls cell conflu-
ency–dependent endocytic pathway of a subset of lipid probes
via rab11. PEG-cholesterol and NBD-cholesterol were accumu-
lated in the recycling endosomes in confluent cells by overex-
pression of GDI, suggesting that importance of cellular distri-
bution of rab11 on the endocytic pathway of these lipid probes.
Choudhury et al. (2004) reported that the extraction of rab11 by
GDI was not affected by cholesterol depletion in Niemann-Pick
A and C fibroblasts in vitro. The difference could be explained
that in Niemann-Pick cells, cholesterol is mainly accumulated
in early and late endosomes, whereas preferential accumula-
tion of cholesterol in recycling endosomes may occur when
cells approach confluence. Furthermore, these differences in
cholesterol distribution could account for different rabs inhibi-
tion by high levels of cellular cholesterol between Niemann-
Pick cells and confluent CHO cells.
In summary, present study suggests that cholesterol con-
trols the endocytosis of a limited subset of lipid probes by
modulating rab11 distribution in a cell confluency–depen-
dent manner. This is caused by the inhibition of extraction of
rab11 by GDI under high cellular cholesterol condition. The
mechanism(s) for cholesterol-regulated inhibition of GDI
extraction of rab proteins is not well understood. Whereas
the involvement of specific membrane constituents are pro-
posed (Choudhury et al., 2004), recent results using choles-
terol containing liposomes suggest that cholesterol alone
can influence rab retrieval from membranes (Ganley and
Pfeffer, 2006). Further studies are required to understand
the molecular mechanisms of cholesterol-dependent reg-
ulation of rab proteins.
We are grateful to T. Hayakawa for the kinetic analyses. We thank members
of Kobayashi laboratory and A. Yamaji-Hasegawa for critical reading of the
manuscript and technical advice. We also thank members of Nakano labora-
tory of RIKEN, especially A. Nakano and R. Hirata for their support. This
work was supported by grants from the Ministry of Education, Science,
Sports, and Culture of Japan (16044247 and 17390025), grants from RIKEN
Frontier Research System, Bioarchitect Project of RIKEN, and International
HDL Award Program to T.K.
Alory, C., and Balch, W. E. (2001). Organization of the Rab-GDI/CHM su-
perfamily: The functional basis for choroideremia disease. Traffic 2, 532–543.
Bligh, E. G., and Dyer, W. J. (1959). A rapid method of total lipid extraction
and purification. Can. J. Biochem. Physiol. 37, 911–917.
Brown, P. S., Wang, E., Aroeti, B., Chapin, S. J., Mostov, K. E., and Dunn,
K. W. (2000). Definition of distinct compartments in polarized Madin-Darby
canine kidney (MDCK) cells for membrane-volume sorting, polarized sorting
and apical recycling. Traffic 1, 124–140.
Cansell, M., Gouygou, J. P., Jozefonvicz, J., and Letourneur, D. (1997). Lipid
composition of cultured endothelial cells in relation to their growth. Lipids 32,
Casanova, J. E., Wang, X., Kumar, R., Bhartur, S. G., Navarre, J., Woodrum,
J. E., Altschuler, Y., Ray, G. S., and Goldenring, J. R. (1999). Association of
Rab25 and Rab11a with the apical recycling system of polarized Madin-Darby
canine kidney cells. Mol. Biol. Cell 10, 47–61.
Chavrier, P., Parton, R. G., Hauri, H. P., Simons, K., and Zerial, M. (1990).
Localization of low molecular weight GTP binding proteins to exocytic and
endocytic compartments. Cell 62, 317–329.
Chen, W., Feng, Y., Chen, D., and Wandinger-Ness, A. (1998). Rab11 is
required for trans-golgi network-to-plasma membrane transport and a pref-
erential target for GDP dissociation inhibitor. Mol. Biol. Cell 9, 3241–3257.
Choudhury, A., Dominguez, M., Puri, V., Sharma, D. K., Narita, K., Wheatley,
C. L., Marks, D. L., and Pagano, R. E. (2002). Rab proteins mediate Golgi
transport of caveola-internalized glycosphingolipids and correct lipid traf-
ficking in Niemann-Pick C cells. J. Clin. Invest. 109, 1541–1550.
Choudhury, A., Sharma, D. K., Marks, D. L., and Pagano, R. E. (2004).
Elevated endosomal cholesterol levels in Niemann-Pick cells inhibit rab4 and
perturb membrane recycling. Mol. Biol. Cell 15, 4500–4511.
Corvera, S., DiBonaventura, C., and Shpetner, H. S. (2000). Cell confluence-
dependent remodeling of endothelial membranes mediated by cholesterol.
J. Biol. Chem. 275, 31414–31421.
Daro, E., van der Sluijs, P., Galli, T., and Mellman, I. (1996). Rab4 and
cellubrevin define different early endosome populations on the pathway of
transferrin receptor recycling. Proc. Natl. Acad. Sci. USA 93, 9559–9564.
Dirac-Svejstrup, A. B., Soldati, T., Shapiro, A. D., and Pfeffer, S. R. (1994).
Rab-GDI presents functional Rab9 to the intracellular transport machinery
and contributes selectivity to Rab9 membrane recruitment. J. Biol. Chem. 269,
Fukuda, M. (2003). Distinct Rab binding specificity of Rim1, Rim2, rabphilin,
and Noc2. Identification of a critical determinant of Rab3A/Rab27A recogni-
tion by Rim2. J. Biol. Chem. 278, 15373–15380.
Frolov, A., Perescu, A., Atshaves, B. P., So, P.T.C., Gratton, E., Serrero, G., and
Schroeder, F. (2000). High density lipoprotein-mediated cholesterol uptake
and targeting to lipid droplets in intact L-cell fibroblasts. A single and
multiphoton fluorescence approach. J. Biol. Chem. 275, 12769–12780.
M. Takahashi et al.
Molecular Biology of the Cell2676
Fukuda, M., Kanno, E., and Mikoshiba, K. (1999). Conserved N-terminal Download full-text
cysteine motif is essential for homo- and heterodimer formation of synapto-
tagmins III, V, VI, and X. J. Biol. Chem. 274, 31421–31427.
Ganley, I. G., and Pfeffer, S. R. (2006). Cholesterol accumulation sequesters
Rab9 and disrupts late endosome function in NPC1-deficient cells. J. Biol.
Chem. 281, 17890–17899.
Green, E. G., Ramm, E., Riley, N. M., Spiro, D. J., Goldenring, J. R., and Wessling-
Resnick, M. (1997). Rab11 is associated with transferrin-containing recycling
compartments in K562 cells. Biochem. Biophys. Res. Commun. 239, 612–616.
Hanover, J. A., Willingham, M. C., and Pastan, I. (1984). Kinetics of transit of
transferrin and epidermal growth factor through clathrin-coated membranes.
Cell 39, 283–293.
Hao, M., Lin, S. X., Karylowski, O. J., Wustner, D., McGraw, T. E., and
Maxfield, F. R. (2002). Vesicular and non-vesicular sterol transport in living
cells. The endocytic recycling compartment is a major sterol storage organelle.
J. Biol. Chem. 277, 609–617.
Hao, M., and Maxfield, F. R. (2000). Characterization of rapid membrane
internalization and recycling. J. Biol. Chem. 275, 15279–15286.
Hao, M., Mukherjee, S., Sun, Y., and Maxfield, F. R. (2004). Effects of choles-
terol depletion and increased lipid unsaturation on the properties of endo-
cytic membranes. J. Biol. Chem. 279, 14171–14178.
Holtta-Vuori, M., Maatta, J., Ullrich, O., Kuismanen, E., and Ikonen, E. (2000).
Mobilization of late-endosomal cholesterol is inhibited by Rab guanine nu-
cleotide dissociation inhibitor. Curr. Biol. 10, 95–98.
Holtta-Vuori, M., Tanhuanpaa, K., Mobius, W., Somerharju, P., and Ikonen, E.
(2002). Modulation of cellular cholesterol transport and homeostasis by
Rab11. Mol. Biol. Cell 13, 3107–3122.
Kobayashi, T., Beuchat, M. H., Lindsay, M., Frias, S., Palmiter, R. D., Sakuraba,
H., Parton, R. G., and Gruenberg, J. (1999). Late endosomal membranes rich in
lysobisphosphatidic acid regulate cholesterol transport. Nat. Cell Biol. 1, 113–118.
Kobayashi, T., Storrie, B., Simons, K., and Dotti, C. G. (1992). A functional
barrier to movement of lipids in polarized neurons. Nature 359, 647–650.
Kok, J. W., Babia, T., and Hoekstra, D. (1991). Sorting of sphingolipids in the
endocytic pathway of HT29 cells. J. Cell Biol. 114, 231–239.
Koval, M., and Pagano, R. E. (1989). Lipid recycling between the plasma
membrane and intracellular compartments: transport and metabolism of flu-
orescent sphingomyelin analogues in cultured fibroblasts. J. Cell Biol. 108,
Lebrand, C., Corti, M., Goodson, H., Cosson, P., Cavalli, V., Mayran, N.,
Faure, J., and Gruenberg, J. (2002). Late endosome motility depends on lipids
via the small GTPase Rab7. EMBO J. 21, 1289–1300.
Linder, M. D., Uronen, R. L., Holtta-Vuori, M., van der Sluijs, P., Peranen, J., and
Ikonen, E. (2007). Rab8-dependent recycling promotes endosomal cholesterol
removal in normal and sphingolipidosis cells. Mol. Biol. Cell 18, 47–56.
Liscum, L., and Faust, J. R. (1989). The intracellular transport of low density
lipoprotein-derived cholesterol is inhibited in Chinese hamster ovary cells
cultured with 3-beta-[2-(diethylamino)ethoxy]androst-5-en-17-one. J. Biol.
Chem. 264, 11796–11806.
Liscum, L., and Underwood, K. W. (1995). Intracellular cholesterol transport
and compartmentation. J. Biol. Chem. 270, 15443–15446.
Maxfield, F. R., and McGraw, T. E. (2004). Endocytic recycling. Nat. Rev. Mol.
Cell Biol. 5, 121–132.
Maxfield, F. R., and Wustner, D. (2002). Intracellular cholesterol transport.
J. Clin. Invest. 110, 891–898.
Mayor, S., Presley, J. F., and Maxfield, F. R. (1993). Sorting of membrane
components from endosomes and subsequent recycling to the cell surface
occurs by a bulk flow process. J. Cell Biol. 121, 1257–1269.
Mukherjee, S., Soe, T. T., and Maxfield, F. R. (1999). Endocytic sorting of lipid
analogues differing solely in the chemistry of their hydrophobic tails. J. Cell
Biol. 144, 1271–1284.
Mukherjee, S., Zha, X., Tabas, I., and Maxfield, F. R. (1998). Cholesterol
distribution in living cells: fluorescence imaging using dehydroergosterol as a
fluorescent cholesterol analog. Biophys. J. 75, 1915–1925.
Narita, K., Choudhury, A., Dobrenis, K., Sharma, D. K., Holicky, E. L., Marks,
D. L., Walkley, S. U., and Pagano, R. E. (2005). Protein transduction of Rab9
in Niemann-Pick C cells reduces cholesterol storage. FASEB J. 19, 1558–1560.
Pagano, R. E. (2003). Endocytic trafficking of glycosphingolipids in sphingo-
lipid storage diseases. Philos. Trans. R. Soc. Lond. B Biol. Sci. 358, 885–891.
Peter, F., Nuoffer, C., Pind, S. N., and Balch, W. E. (1994). Guanine nucleotide
dissociation inhibitor is essential for Rab1 function in budding from the
endoplasmic reticulum and transport through the Golgi stack. J. Cell Biol. 126,
Pfeffer, S., and Aivazian, D. (2004). Targeting Rab GTPases to distinct mem-
brane compartments. Nat. Rev. Mol. Cell Biol. 5, 886–896.
Puri, V., Watanabe, R., Dominguez, M., Sun, X., Wheatley, C. L., Marks, D. L.,
and Pagano, R. E. (1999). Cholesterol modulates membrane traffic along the
endocytic pathway in sphingolipid-storage diseases. Nat. Cell Biol. 1, 386–
Puri, V., Watanabe, R., Singh, R. D., Dominguez, M., Brown, J. C., Wheatley,
C. L., Marks, D. L., and Pagano, R. E. (2001). Clathrin-dependent and -inde-
pendent internalization of plasma membrane sphingolipids initiates two
Golgi targeting pathways. J. Cell Biol. 154, 535–547.
Ren, M., Xu, G., Zeng, J., De Lemos-Chiarandini, C., Adesnik, M., and Saba-
tini, D. D. (1998). Hydrolysis of GTP on rab11 is required for the direct
delivery of transferrin from the pericentriolar recycling compartment to the
cell surface but not from sorting endosomes. Proc. Natl. Acad. Sci. USA 95,
Rothblat, G. H., de la Llera-Moya, M., Atger, V., Kellner-Weibel, G., Williams,
D. L., and Phillips, M. C. (1999). Cell cholesterol efflux: integration of old and
new observations provides new insights. J. Lipid. Res. 40, 781–796.
Sabharanjak, S., Sharma, P., Parton, R. G., and Mayor, S. (2002). GPI-anchored
proteins are delivered to recycling endosomes via a distinct cdc42-regulated,
clathrin-independent pinocytic pathway. Dev. Cell 2, 411–423.
Sato, S. B., Ishii, K., Makino, A., Iwabuchi, K., Yamaji-Hasegawa, A., Senoh,
Y., Nagaoka, I., Sakuraba, H., and Kobayashi, T. (2004). Distribution and
transport of cholesterol-rich membrane domains monitored by a membrane-
impermeant fluorescent polyethylene glycol-derivatized cholesterol. J. Biol.
Chem. 279, 23790–23796.
Satoh, T., Endo, M., Nakamura, S., and Kaziro, Y. (1988). Analysis of guanine
nucleotide bound to ras protein in PC12 cells. FEBS Lett. 236, 185–189.
Sciaky, N., Presley, J., Smith, C., Zaal, K. J., Cole, N., Moreira, J. E., Terasaki,
M., Siggia, E., and Lippincott-Schwartz, J. (1997). Golgi tubule traffic and the
effects of brefeldin A visualized in living cells. J. Cell Biol. 139, 1137–1155.
Sharma, D. K., Brown, J. C., Choudhury, A., Peterson, T. E., Holicky, E.,
Marks, D. L., Simari, R., Parton, R. G., and Pagano, R. E. (2004). Selective
stimulation of caveolar endocytosis by glycosphingolipids and cholesterol.
Mol. Biol. Cell 15, 3114–3122.
Sheff, D. R., Daro, E. A., Hull, M., and Mellman, I. (1999). The receptor
recycling pathway contains two distinct populations of early endosomes with
different sorting functions. J. Cell Biol. 145, 123–139.
Sokol, J. et al. (1988). Type C Niemann-Pick disease. Lysosomal accumulation
and defective intracellular mobilization of low density lipoprotein cholesterol.
J. Biol. Chem. 263, 3411–3417.
Soldati, T., Rancano, C., Geissler, H., and Pfeffer, S. R. (1995). Rab7 and Rab9
are recruited onto late endosomes by biochemically distinguishable processes.
J. Biol. Chem. 270, 25541–25548.
Sonnichsen, B., De Renzis, S., Nielsen, E., Rietdorf, J., and Zerial, M. (2000).
Distinct membrane domains on endosomes in the recycling pathway visual-
ized by multicolor imaging of Rab4, Rab5, and Rab11. J. Cell Biol. 149,
Tsuboi, T., and Fukuda, M. (2005). The C2B domain of rabphilin directly
interacts with SNAP-25 and regulates the docking step of dense core vesicle
exocytosis in PC12 cells. J. Biol. Chem. 280, 39253–39259.
Tsuboi, T., and Fukuda, M. (2006). Rab3A and Rab27A cooperatively regulate
the docking step of dense-core vesicle exocytosis in PC12 cells. J. Cell Sci. 119,
Ullrich, O., Horiuchi, H., Bucci, C., and Zerial, M. (1994). Membrane associ-
ation of Rab5 mediated by GDP-dissociation inhibitor and accompanied by
GDP/GTP exchange. Nature 368, 157–160.
Ullrich, O., Reinsch, S., Urbe, S., Zerial, M., and Parton, R. G. (1996). Rab11
regulates recycling through the pericentriolar recycling endosome. J. Cell Biol.
van der Sluijs, P., Hull, M., Webster, P., Male, P., Goud, B., and Mellman, I.
(1992). The small GTP-binding protein rab4 controls an early sorting event on
the endocytic pathway. Cell 70, 729–740.
van Ijzendoorn, S. C., Mostov, K. E., and Hoekstra, D. (2003). Role of rab
proteins in epithelial membrane traffic. Int. Rev. Cytol. 232, 59–88.
Walter, M., Davies, J. P., and Ioannou, Y. A. (2003). Telomerase immortaliza-
tion upregulates Rab9 expression and restores LDL cholesterol egress from
Niemann-Pick C1 late endosomes. J. Lipid Res. 44, 243–253.
Zerial, M., and McBride, H. (2001). Rab proteins as membrane organizers.
Nat. Rev. Mol. Cell Biol. 2, 107–117.
Cholesterol Regulates Rab11
Vol. 18, July 2007 2677