Journal of Cell Science 108, 3349-3358 (1995)
Printed in Great Britain © The Company of Biologists Limited 1995
JCS9377
3349
The rab7 GTPase resides on a vesicular compartment connected to
lysosomes
Stéphane Méresse, Jean-Pierre Gorvel and Philippe Chavrier*
Centre d’Immunologie INSERM-CNRS de Marseille-Luminy, Case 906, 13288 Marseille Cedex 9, France
*Author for correspondence
SUMMARY
Rab GTPases belong to the Ras GTPase superfamily and
are key regulators of membrane traffic. Among them, rab7
has been localized on late endosomes of NRK cells but its
function remains unknown. In order to investigate its role,
we generated stable HeLa cell lines that express either wild
type or a GTPase-defective mutant of rab7 in an inducible
manner. A morphological analysis of the intracellular
localization of these proteins was performed by confocal
laser microscopy. Here we show that, in HeLa cells, rab7
is present on a vesicular compartment that extends from
the perinuclear area to the cell periphery and shows only
a partial colocalization with the cation-independent
mannose 6-phosphate receptor, a marker for late
endosomes. The topology of this compartment is dependent
on the microtubule network since nocodazole treatment
results in its scattering throughout the cytoplasm. In
addition, we observed that, in contrast to the wild-type
protein, a rab7 mutant with a reduced GTPase activity is
in part associated with lysosomal membranes. This observation was confirmed by subcellular fractionation in a
Percoll gradient. Our data implicate rab7 as the first
GTPase functioning on terminal endocytic structures in
mammalian cells.
INTRODUCTION
pathway by acting either positively on the retrograde intraGolgi transport or negatively on the anterograde transport
(Martinez et al., 1994). Rab4 and rab5 are both localized to
early endosomes. Rab5 regulates endocytosis from the plasma
membrane and early endosome fusion (Gorvel et al., 1991;
Bucci et al., 1992). By contrast, rab4 controls the recycling
from early endosomes to the cell surface (van der Sluijs et al.,
1992).
Rab7 and rab9 are both associated with late endocytic compartments. While the role of rab9 in the mannose 6-phosphate
receptor recycling from late endosomes to the trans-Golgi
network has been clearly established (Lombardi et al., 1993;
Riederer et al., 1994) the function of rab7 remains elusive.
More is known about Ypt7p, the yeast homologue of rab7.
Deletion of the YPT7 gene leads to the fragmentation of the
vacuole and to a delay in the processing of vacuolar proteins
(Wichmann et al., 1992). Another study has suggested that
Ypt7p regulates transport between late endosomes and the
vacuole (Schimmöller and Riezman, 1993).
Therefore, it was of importance to examine the role of rab7
and to establish whether its function in the late endocytic
pathway of mammalian cells is similar to that of Ypt7p. Like
p21-ras, rab-proteins cycle between GDP- and GTP-bound conformations (Bourne et al., 1990). Mutant proteins analogous to
the oncogenic forms of p21-ras with a reduced GTPase activity
or with an altered affinity for guanine nucleotides are functionally impaired (Walworth et al., 1992; Tisdale et al., 1992;
Bucci et al., 1992; van der Sluijs et al., 1992; Riederer et al.,
In eukaryotic cells vesicles convey material from one compartment to another. Upon budding from the donor organelle,
vesicles deliver their cargo by fusing with the acceptor
membrane. The shuttling mechanism requires a sophisticated
cellular machinery to control the processes of vesicle budding,
targeting and fusion with the acceptor compartment. A combination of genetic and biochemical approaches has been used to
identify some of the components of this cellular machinery (for
review see Rothman, 1994).
Ras-related GTPases of the rab family have been identified
as key regulators of membrane transport (Novick and Garett,
1994; Pfeffer, 1994). Evidence for the crucial role of these
proteins was first provided by the identification of sec4 and
ypt1 mutants in yeast that were blocked at distinct steps of the
secretory pathway (Goud et al., 1988; Segev et al., 1988).
Sec4p and Ypt1p GTPases belong to the rab family that
includes more than thirty gene products conserved from yeast
to man (for reviews see Goud and McCaffrey, 1991; Zerial and
Stenmark, 1993). These proteins are localized at the cytoplasmic face of distinct organelles (Chavrier et al., 1990a; Goud et
al., 1990; van der Sluijs et al., 1992) suggesting that they may
regulate individual steps of vesicular transport by controlling
vesicle targeting and/or fusion. For example rab1a, rab1b and
rab2 are regulating components of the vesicular transport
between early compartments of the secretory pathway (Tisdale
et al., 1992). Rab6 controls transport steps along the secretory
Key words: rab7, late endosome, lysosome, mannose 6-phosphate
receptor, lamp
3350 S. Méresse, J.-P. Gorvel and P. Chavrier
1994; Stenmark et al., 1994; Martinez et al., 1994). Thus, it
appears that the GDP/GTP cycle is crucial for rab function. We
chose to look for specific alterations in cells overexpressing
wild-type or a mutant rab7 protein with reduced GTPase
activity. This study was performed on stable transformed cells
that express proteins of interest in an inducible, tightly regulated
manner. Here, we present evidence that, in HeLa cells, rab7 is
marginally localized in the late endosomal cation-independent
mannose 6-phosphate receptor (CI-MPR) rich compartment.
Strikingly, rab7 is mainly associated with a very extended
vesicular compartment whose architecture is dependent on the
microtubule network. By combining morphological and biochemical analyses we show that the GTPase-defective mutant
of rab7 is partially redistributed to lysosomes. These results
suggest that rab7 plays a role in a vesicular traffic step
connected to lysosomes.
MATERIALS AND METHODS
Materials
Rabbit anti-human lamp-1 and mouse anti-human lamp-2 were
generous gifts of Dr Minoru Fukuda (La Jolla Cancer Research Foundation, La Jolla, CA, USA). Mouse 9E10 cells secreting the anti-myc
antibody were obtained from the ATCC (Rockville, Maryland). FITCand Texas Red-conjugated secondary antibodies were from Jackson
Immunoresearch Laboratories (Immunotech, France). Lipofectamine
and TRIzol reagent were obtained from Gibco-BRL (Cergy-Pontoise,
France). Acrylamide was purchased from National Diagnostic
(Atlanta, USA). Anhydrotetracycline was from Janssen Chimica (Geel,
Belgium). Protein was assayed using the BCA kit (Pierce Chemical,
Interchim, Montluçon, France). Unless otherwise indicated, reagents
were from Sigma Chimie (St Quentin-Fallavier, France).
Preparation of cDNA constructs
An EcoRI-PvuII fragment from pGEM-1rab7 (Chavrier et al., 1990a)
was introduced into the BamHI site of pUHD10.3 after blunt-ending
(Gossen and Bujard, 1992). Rab7Q67L was created by mutagenesis
with a PCR-based approach (Landt et al., 1990) with the primer 5′GGAACCGTTCAAGGCCTGCTGTGTCC-3′ and two outer primers
derived from the sequence of pUHD10.3, UP: 5′-CTCGGTACCCGGGTCGAG-3′ and DW: 5′-CATCAATGTATCTTATCATGTC3′. The myc-tagged constructs were generated by PCR by introducing a NdeI site overlapping the initiation codon of rab7 with the primer
5′-GATCGATCATATGACCTCTAGGAAGAAAG-3′. This NdeI
site was used to ligate the sequence coding for the myc epitope (Evan
et al., 1985) with rab7 to give pGEM-1myc-rab7. After partial
digestion an EcoRI fragment from pGEM-1myc-rab7 encompassing
the complete ORF was inserted into the corresponding site of
pUHD10.3, to give pUHDmyc-rab7. Plasmid pUHDmyc-rab7Q67L
was obtained by direct mutagenesis by PCR as described above.
Generation of cell lines expressing recombinant rab7
HeLa cells expressing the tetracycline regulated transactivator
(HtTA1 cells; Gossen and Bujard, 1992) were kindly provided by H.
Bujard (Heidelberg, Germany). They were grown in DMEM supplemented with 10% fetal calf serum and antibiotics. Confluent HtTA1
cells were seeded at 1/10 (surface/surface) in 25 cm2 flask. After 20
hours, cells were transfected with 2.5 µg of pUHDmyc-rab7wt or
pUHDmyc-rab7Q67L and 50 ng of pSV2gpt. Transfection was
performed using lipofectamine as described by the manufacturer.
After 24 hours cells were trypsinized and seeded at 1/20 in 10 cm
dishes. Selection was performed by growing cells in regular medium
supplemented with 5 µg/ml micophenolic acid, 250 µg/ml xanthine,
15 µg/ml hypoxanthine and 0.1 µg/ml anhydrotetracycline (ATc;
Gossen and Bujard, 1993). ATc was added every other day. Single
clones were isolated and confirmed for recombinant protein
expression by indirect immunofluorescence and western blot analyses.
For this purpose and subsequent experiments, cells were trypsinized,
seeded at 1/10 and grown for 3 days in the presence (non induced) or
absence (induced) of ATc. Medium was changed after 24 hours.
Indirect immunofluorescence
Cells were plated on glass coverslips and grown to ~50% confluency
in the presence or absence of ATc. Cells were fixed with 3%
paraformaldehyde in PBS, pH 7.4, for 30 minutes, extensively washed
with 50 mM NH4Cl in PBS and washed twice with 0.05% saponin in
PBS. Cells were incubated for 30 minutes with primary antibodies in
10% horse serum, 0.05% saponin in PBS, extensively washed with
0.05% saponin and incubated for 30 minutes with secondary antibodies. Coverslips were then washed, mounted in Mowiol and viewed
under a Leica TCS 4DA confocal microscope. Series of 2 plane
sections of 0.3 µm thickness were monitored. For double-staining
experiments identical optical sections are presented. Superimposed
images were treated with a pseudocolor scale. Colocalized structures
are seen in a yellow color.
Percoll gradients
For each gradient, 150 cm2 of 80% confluent cells were usually used.
Dishes (10 cm) were chilled on ice, washed 3 times in ice-cold PBS
and scraped. They were pelleted for 5 minutes at 800 rpm in a clinical
centrifuge, overlaid with 3 ml of homogenization buffer (250 mM
sucrose, 10 mM triethanolamine, 10 mM acetic acid, 1 mM EDTA,
pH 7.4) and centrifuged for 5 minutes at 3,000 rpm. Cells were resuspended in 1 ml homogenization buffer, containing 1 mM PMSF and
homogenized by 5 to 10 passages through at 22G1S needle. After centrifugation for 10 minutes at 3,000 rpm the post nuclear supernatant
was loaded on a 9 ml cushion of 27% Percoll in homogenization
buffer, 1 mM PMSF. The tube was centrifuged in a 50Ti rotor for 65
minutes at 23,000 rpm, 4°C and 0.5 ml fractions were collected from
the top. Fractions were supplemented with 0.25% NP-40 and centrifuged for 25 minutes at 70,000 rpm in a TL 120.1 rotor. Percoll free
supernatants were saved and used for western blots and enzymatic
activity assays.
In some experiments cells were loaded with horseradish peroxidase
(HRP) (1.5 mg/ml in DMEM, 10 mM Hepes, 0.2% BSA) for 20
minutes at 37°C. Cells were either rapidly washed twice with PBS
and returned at 37°C for the indicated time of chase or chilled on ice
and washed 5 times for 5 minutes in ice cold DMEM, 10 mM Hepes,
0.2% BSA. HRP on the gradient was assayed according to the method
of Gruenberg and Howell (1986). β-hexosaminidase activities was
determined as described by Ozawa et al. (1993).
Western blots
Cells were washed twice in PBS and lysed by addition of 20 µl/cm2
of lysis buffer (150 mM NaCl, 50 mM Tris-HCl, 1 mM EDTA, 1%
NP-40, 1 mM PMSF, pH 7). Samples were centrifuged at 13,000 rpm
in a table-top centrifuge. Supernatants were saved and protein assays
were performed using BCA reagent. Proteins were loaded on 12%
SDS-PAGE gels and transferred onto Immobilon-P membranes
(Millipore). Blocking, incubation with antibodies and washing were
done in PBS with 4% dry milk and 0.05% Tween-20. Detection was
performed using either peroxidase-conjugated goat anti-rabbit or antimouse antibodies with the enhanced chemiluminescence system
(ECL, Amersham, UK). Western blots were photographed using a
CCD camera (The Imager, Appligene, Illkrich, France) and quantified with the NHI Image software.
RESULTS
It has been shown for several GTPases that mutation equiva-
The rab7 GTPase cycles to lysosomes 3351
lent to the Q61L substitution in p21-ras alters both intrinsic and
GAP stimulated GTPases activities (Der et al., 1986; Walworth
et al., 1992; Tanigawa et al; 1993; Stenmark et al., 1994). An
analogous mutant of rab7 was generated. Using PCR-based
mutagenesis a substitution of leucine for glutamine in position
67 (rab7Q67L) was made. Both wild-type and mutant proteins
were tagged N-terminally with the 14 residues of the c-myc
epitope (myc-rab7wt and myc-rab7Q67L) (Evan et al., 1985).
Inducible expression of myc-rab7wt and mycrab7Q67L
In order to prevent possible toxic effects caused by long-term
expression of recombinant proteins we used a recently
developed system in which gene expression is regulated by a
tetracycline-controlled transactivator (tTA) (Gossen and
Bujard, 1992). cDNAs were subcloned into the pUHD vector
downstream of a tTA-responsive promoter and transfected into
HeLa cells expressing constitutively the tTA transactivator
(HtTA1 cells). In these cells, tTA binding upstream of the
promoter efficiently activates transcription. In the presence of
ATc which complexes tTA with high affinity and inhibits tTA
binding to DNA, transcription does not take place.
HtTA1 cells were cotransfected with plasmids encoding
either myc-rab7wt or myc-rab7Q67L and pSV2gpt conferring
resistance to mycophenolic acid in the presence of xanthine.
After selection, subclones were screened by immunofluorescence for expression of recombinant proteins. Positive clones
expressing either myc-rab7wt or myc-rab7Q67L were further
analyzed by northern blot using a rab7 cDNA probe. As shown
Fig. 1. Northern and western blot analyses of inducible expression in
HtTA1 cells. HtTA1 cells (Ctrl) or selected clones expressing either
myc-rab7wt (WT) or myc-rab7Q67L (Q67L) proteins were grown for
72 hours in the presence (+) or absence (−) of 0.1 µg/ml
anhydrotetracycline (ATc). (A) Total RNA (20 µg) was resolved on
a 1% denaturing agarose gel, transferred onto GeneScreen Plus
membrane and probed with a 32P-labeled rab7 canine cDNA. The
positions of 28 S and 18 S ribosomal RNAs are indicated. (B) Cells
were washed in PBS and lysed in 1% NP-40 containing buffer.
Samples (20 µg) of proteins were separated by 12% SDS-PAGE,
transferred onto Immobilon-P and analyzed by western blot using
affinity purified anti-rab7 (upper panel) or mouse anti-myc (lower
panel) antibodies. The positions of the molecular mass markers 31
and 21 kDa are indicated.
in Fig. 1A, HtTA1 cells and selected clones grown in the
presence of ATc expressed two transcripts corresponding to
endogenous rab7 mRNAs (Chavrier et al., 1990b). When
cultures were depleted of ATc for 72 hours an additional transcript of the expected size, expressed from the transfected
construct was detected. Western blot analyses were performed
using either polyclonal anti-rab7 or monoclonal anti-myc antibodies. Endogenous rab7 was detected as a 25 kDa protein in
HtTA1 cells and selected clones, regardless of whether ATc
was present or not in the culture medium (Fig. 1B). Three days
after removal of ATc, anti-rab7 antibody revealed a strong
induction of myc-rab7wt and myc-rab7Q67L which exhibited a
3 kDa shift in their apparent molecular mass due to the
A
B
C
D
E
Fig. 2. Wild-type rab7 and Q67L mutant have distinct intracellular
localizations. HtTA1 cells (A), myc-rab7wt- (B,C) or myc-rab7Q67L(D,E) expressing cells were grown for 3 days on coverslips in the
absence of ATc. Cells were processed for single (A) or double (B-E)
indirect immunofluorescence using rabbit anti-rab7 (A,B,D) or
mouse anti-myc (C,E) antibodies. Slides were viewed using a Leica
TCS 4D microscope and by confocal laser scanning. Optical section,
0.3 µm. Bar, 10 µm.
3352 S. Méresse, J.-P. Gorvel and P. Chavrier
presence of the N-terminal c-myc epitope (Fig. 1B, upper
panel). This was confirmed using anti-myc monoclonal
antibody (Fig. 1B, lower panel). Quantitation of western blots
indicated that myc-rab7wt and myc-rab7Q67L were expressed
at 4- and 5-fold higher levels than the endogenous rab7, respectively. No morphological or growth behavior changes of the
cells were observed upon expression of recombinant proteins.
Taken together, these data demonstrate a tight control of gene
expression in the selected clones.
The cytosol versus membrane partition of myc-tagged
proteins was determined. Rab proteins are post-translationally
modified by the addition of a carboxy-terminal isoprenyl group
(Peter et al., 1992). This modification is necessary for
membrane attachment and function (Magee and Newman,
1992). As for the endogenous rab7, more than 80% of overexpressed proteins were found associated with membranes
(data not shown) suggesting that both wild type and GTPase
mutant of rab7 were correctly isoprenylated.
myc-rab7wt and myc-rab7Q67L have overlapping
but distinct intracellular localizations
The intracellular distribution of endogenous and recombinant
rab7 proteins was studied by indirect immunofluorescence
confocal microscopy. In HtTA1 cells, endogenous rab7 was
associated with vesicular-tubular structures that developed as
a reticulum from the perinuclear area to the cellular periphery
(Fig. 2A). A similar distribution of rab7 was detected in
selected clones grown in the presence of ATc (not shown). In
the absence of ATc, the intracellular localization of mycrab7wt revealed either by the polyclonal anti-rab7 or by the
anti-myc antibodies was very similar to that of the endogenous
protein (Fig. 2B,C). Previous studies have localized rab7 in
A
D
B
E
late, CI-MPR rich, endosomes of NRK cells (Chavrier et al.,
1990a). In HtTA1 cells, the CI-MPR staining was mainly
restricted to perinuclear structures that are likely to be late
endosomes and to a certain extent the TGN (Fig. 3B) (Griffiths
et al., 1988; Geuze et al., 1988). Double-staining immunofluorescence performed on myc-rab7wt cells showed that rab7
colocalized only partially with the CI-MPR in perinuclear
structures (Fig. 3A-C). Rab7 was rather present on a compartment much larger than CI-MPR rich endosomes. In this
respect, rab7 and the CI-MPR patterns overlapped but were
clearly different.
The membrane localization of the Q67L mutant protein (Fig.
2D-E) differed strikingly from that of the wild-type rab7. In
addition to the reticular pattern previously described for the
wild-type protein, myc-rab7Q67L was present on large vesicular
structures spread throughout the cytoplasm. In these cells, the
CI-MPR distribution appeared unchanged and colocalized with
myc-rab7Q67L to a similar extent to overexpressed wild-type
protein (Fig. 3D-F).
We can rule out the possibility that observed phenotypes are
related either to a clonal effect or are due to the presence of the
N-terminal tag since independent clones overexpressing the
same recombinant proteins or non-tagged version of rab7wt and
rab7Q67L presented identical intracellular distributions (data not
shown). These data are consistent with a report where a myctagged version of rab5 was phenotypically indistinguishable
from the non-tagged protein (Stenmark et al., 1994).
Presented results indicate that rab7 is mainly present in a very
expanded vesicular compartment and only partially colocalizes
with the CI-MPR in the perinuclear area. In addition, rab7Q67L,
the GTPase-defective mutant, is localized on large vesicles we
intend to identify.
C
F
Fig. 3. Wild type and Q67L mutant of
rab7 colocalize only partially with the
CI-MPR. After 3 days of induction mycrab7wt- (A-C) or myc-rab7Q67L- (D-F)
expressing cells were processed for
double indirect immunofluorescence
using mouse anti-myc (A,D) or rabbit
anti-CI-MPR (B,E) antibodies.
Superimposed images of recombinant
rab7 proteins and CI-MPR are shown
(C,F). Optical section, 0.3 µm. Bar,
10 µm.
The rab7 GTPase cycles to lysosomes 3353
myc-rab7Q67L is partly localized to lysosomes
In order to investigate the nature of the large myc-rab7Q67Lpositive vesicles, we compared the intracellular distribution of
rab7 with several markers of the endocytic pathway. Rab7wt
and rab7Q67L did not colocalize with internalized FITC-conjugated transferrin, clathrin or rab5, three markers of early
endosomal compartments (data not shown). As described
above, only a partial colocalization was found with the CIMPR, a marker for late endosomes (Fig. 3A-F). Lysosomes
were labeled with either a mouse anti-lamp-2 or a rabbit antilamp-1 antibodies. These two integral highly glycosylated
proteins are mainly localized to lysosomes (Kornfeld and
Mellman, 1989; Carlsson et al., 1988) and double labeling of
HtTA1 with anti-lamp-1 and -lamp-2 antibodies revealed
identical intracellular distributions (data not shown). No
obvious overlapping could be detected between myc-rab7wt
and the lysosomal proteins lamp-1 (Fig. 4A,C,E) and lamp-2
(data not shown). Nevertheless double positive vesicles were
occasionally detected at the cell periphery (see arrows in Fig.
A
B
C
D
E
F
4A,C). By contrast, rab7-positive vesicular structures found in
myc-rab7Q67L-expressing cells were heavily decorated with
antibodies specific for lamp-1 (Fig. 4B,D,F) or for lamp-2 (data
not shown). Likewise, myc-rab7Q67L was also found to colocalize with the lysosomal enzyme cathepsin D and with the
fluid-phase marker Lucifer Yellow, internalized for 15 minutes
and chased for 1 hour in order to label terminal endocytic compartments (data not shown). It is noteworthy that the overall
aspect and distribution of lamp-1/lamp-2 positive compartments in these cells appeared unchanged (compare Fig. 4C and
D). Altogether these morphological data strongly suggest that
myc-rab7Q67L is in part distributed on lysosomes.
Subcellular fractionation was then performed to analyze the
distribution of rab7 in the selected clones. Percoll gradients
(27%) have been successfully employed to separate lysosomes
from other endocytic compartments (Davidson et al., 1990;
Griffiths et al., 1990). We used horseradish peroxidase (HRP),
a fluid phase marker, to assess the separation efficiency of this
gradient. HRP was internalized in HtTA1 cells for 20 minutes.
Fig. 4. Rab7Q67L colocalizes with
lamp-1. Expression of myc-rab7wt
(A,C,E) or myc-rab7Q67L (B,D,F)
was induced by growing cells for 72
hours in the absence of ATc. After
paraformaldehyde fixation, double
indirect immunofluorescence was
performed using mouse anti-myc
(A,B) or rabbit anti-lamp-1 (C,D)
antibodies. Superimposed images of
recombinant rab7 proteins and lamp1 are shown (E,F). Arrows in A and
C show rare double positive vesicles
in myc-rab7wt expressing cells.
Optical section, 0.3 µm. Bar, 10 µm.
3354 S. Méresse, J.-P. Gorvel and P. Chavrier
5A, fractions 11-16). Cells expressing myc-rab7wt were fractionated and western blot analyses of the gradient were
performed. As shown in Fig. 5B, endogenous rab7 (25 kDa)
and myc-rab7wt (28 kDa) cosedimented and were totally
recovered in top light fractions together with rab5, a marker of
early endosomes. By contrast, the lysosomal enzyme cathepsin
D partitioned in fractions 11 to 16. Taken together, these data
indicated that 27% Percoll gradients clearly separated
lysosomes from a peak of membranes containing early/late
endosomes and the rab7-positive compartment.
The same procedure was applied to cells expressing the mycrab7Q67L mutant. As expected, rab5 and the cathepsin D were
found in fractions 7-10 and 11-16, respectively (Fig. 5B). Interestingly, the distribution of myc-rab7Q67L differed quite
extensively from those of myc-rab7 wt and endogenous protein.
Although myc-rab7Q67L was also found in the light fractions,
a significant part of the recombinant protein clearly distributed
in fractions 11 to 16 (Fig. 5C). Quantification of the western
blot indicated that about 30% of myc-rab7Q67L was present in
dense, cathepsin D-positive, fractions. By contrast, 25 kDa
endogenous rab7 remained totally associated with membranes
of the light fractions.
Analysis of subcellular fractionation data show that a significant part of rab7Q67L is present in lysosomes enriched
fractions whereas distribution of endogenous and overexpressed wild-type rab7 or rab5 remain unchanged. These
results are consistent with morphological data and strongly
suggest that the GTPase-defective mutant of rab7 is partly distributed to lysosomes.
Fig. 5. myc-rab7Q67L cofractionates with lysosomes in a Percoll
gradient. Cells grown for 3 days in the absence of ATc were
fractionated on a 27% Percoll gradient as described in Materials and
Methods. Fractions were collected from the top (1 and 20 are the
lowest and highest density fractions, respectively). (A) Fractions
were analyzed for their content in β-hexosaminidase (n). HRP
activity was measured after HtTA1 cells were incubated for 20
minutes in the presence of HRP (1.5 mg/ml), washed and processed
for subcellular fractionation either immediately (s) or after 1 hour of
chase in the absence of HRP at 37°C (d). Results are expressed as a
percentage of total enzymatic activities. Refractive index is indicated
(+). Cells expressing either myc-rab7wt (B) or myc-rab7Q67L (C)
were analyzed by western blotting. Equal volumes of fractions 6 to
17 (30 µl) were loaded on a 12% SDS-PAGE, gel transferred onto
Immobilon-P and analyzed by immunoblotting using either anticathepsin D or affinity purified anti-rab5 or anti-rab7 antibodies.
In these conditions HRP is known to distribute both in early
and late endocytic compartments (Gruenberg et al., 1989). In
some experiments, HRP was washed away and cells further
incubated for 1 hour at 37°C in order to chase the fluid-phase
marker to lysosomes. Postnuclear supernatants were prepared
and fractionated on 27% Percoll gradients. HRP internalized
for 30 minutes was mainly detected in the light fractions (Fig.
5A, fractions 7-10). After 1 hour chase, HRP activity was
almost completely shifted in dense fractions and cosedimented
with the β-hexosaminidase activity, a lysosomal marker (Fig.
Spatial distribution of rab7 depends on the
microtubule network
Late endosomes and lysosomes are clustered in the perinuclear region in a microtubule-dependent manner (Scheel et
al., 1990). In order to test whether the same holds true for the
rab7-compartment, a double immunofluorescence analysis of
rab7 and tubulin was performed on myc-rab7wt expressing
cells. It revealed that rab7 colocalizes with the microtubules
(Fig. 6A-C). This was especially clear in cellular extensions
where rows of rab7-positive vesicles appeared aligned on
microtubules (Fig. 6D). Similar overlapping of endogenous
rab7 with tubulin was observed in HtTA1 cells (not shown).
The relation between rab7-vesicles and microtubules was
further confirmed by treating cells for 30 minutes with the
microtubule-disrupting drug, nocodazole (10 µM). Staining
with an anti-tubulin antibody indicated that, in these conditions, microtubules had totally disappeared (not shown).
Nocodazole treatment strongly affected the overall organization of the rab7-compartment (Fig. 7A). Interestingly, in
addition to random scattering of rab7 vesicles throughout the
cytoplasm, the GTPase was also present on large lamp-2
positive structures (Fig. 7B). These unusually large vesicles
were cathepsin D-positive but rab5- and CI-MPR-negative
and could therefore be identified as lysosomes (not shown).
A similar phenotype was observed in cells expressing mycrab7Q67L (Fig. 7C,D).
These experiments suggest that the spatial distribution of rab7
depends on the microtubule network providing an explanation
for the reticular aspect of the rab7 compartment in HeLa cells.
In addition, these data suggest that microtubule disruption
triggers a partial redistribution of rab7 on lysosomes.
The rab7 GTPase cycles to lysosomes 3355
DISCUSSION
The fact that the small GTPase rab7 was found on late
endosomes in NRK cells (Chavrier et al., 1990a), together with
the crucial function of Ypt7p in the endocytic traffic towards
the vacuole in yeast (Schimmöller and Riezman, 1993) led us
to hypothesize a similar role for rab7 in mammalian cells. In
this paper, we have developed complementary approaches to
analyze the localization of rab7. By expressing a mutant rab7,
we have found that the overall distribution of this protein
depends on its nucleotide state and on the integrity of the
microtubule network.
Based on recent studies, a model for rab cycling coupled to
guanine nucleotide exchange has been refined (Ulrich et al.,
1994; Soldati et al., 1994; Horiuchi et al., 1995). It predicts
that a GDP dissociation inhibitor (GDI) delivers cytosolic rab
to the appropriate donor compartment in a reaction coupled to
the exchange of GDP for GTP. The GTP-bound protein is
recruited on a nascent vesicle and GTP hydrolysis occurs after
the vesicle has fused with the acceptor membrane. The
resulting GDP-bound protein is then recycled by GDI back to
the donor compartment through the cytosol (for review see
Novick and Garett, 1994). If this model is correct, the
membrane association of the rab7Q67L mutant, which has a
reduced GTPase activity, would be predicted to differ
somehow from that of the wild-type protein. Indeed, we
observed that, in contrast to the wild-type protein, the GTPasedefective mutant of rab7 is present on lysosomes. This strongly
suggests that the lysosome is one of the compartments to which
rab7 cycles.
Similar approaches have been recently used to address the
function of several rab proteins. A well documented example
is given by rab5. This small GTP-binding protein is required
for the transport between the plasma membrane and early
endosomes and regulates the dynamic of early endosomes
fusion (Bucci et al., 1992). Expression of a GTP-bound mutant
rab5Q79L increases the rate of endocytosis and leads to the
appearance of very large endocytic structures (Stenmark et al.,
1994). As predicted by the current model for rab cycling, the
GTPase mutant of rab5 was mainly found on the acceptor
membrane, i.e. the early endosome. Therefore, considering the
localization of the Q67L mutant of rab7, it is tempting to
propose that lysosomes represent the acceptor compartment
for this GTPase. Nevertheless, the fact that wild-type rab7 is
hardly detected on lysosomes, contrasting with the presence
of rab5 both on its donor and acceptor membranes is difficult
to interpret. A likely explanation would be that upon GTP
hydrolysis, the wild-type protein is very rapidly recycled to
the donor membrane. On the contrary, the Q67L mutant in the
GTP-bound form would be less efficiently removed by rab
GDI and would reside longer on the acceptor compartment,
i.e. the lysosome. Alternatively, rab7 may have a higher
affinity for GDP than GTP resulting in the accumulation of
the protein on the donor compartment. This would contrast
with the equal affinity for both guanyl nucleotides reported
for rab5 and rab9 (Stenmark et al., 1994; Riederer et al., 1994).
The presence of rab7 on lysosomal membranes is consistent
with data obtained in unicellular eukaryotes. It was suggested
that Ypt7p, the Saccharomyces cerevisiae homologue of rab7,
regulates transport from late endosomes to the vacuole, the
lysosome-like compartment in yeast (Schimmöller and
Riezman, 1993). More recently, using subcellular fractionation, rab7- and rab4-like GTPases were found on lysosomal
membranes of Dictyostelium discoideum (Temesvari et al.,
1994). Interestingly, the lysosome is not the terminal endocytic
compartment in D. discoideum. A post-lysosomal compartment from which fluid phase components are egested was
recently identified (Padh et al., 1993; Aubry et al., 1993).
Therefore, the distribution of rab7 to pre-terminal or terminal
endocytic compartments appears as a general rule.
In HeLa cells, rab7 is localized on a very extended, lampnegative, reticular compartment predominantly located in the
perinuclear area and stretched to the cellular periphery.
Detailed observations revealed that the protein is present on
coalescent vesicles in close contact with microtubules. While
rab7 is generally considered as a marker for late endosomes,
the observed pattern and the limited overlapping with the CIMPR are quite puzzling. What could this vesicular reticulum
be? Our observations are consistent with a study by Rabinowitz
et al. (1992) performed in peritoneal macrophages. In these
cells, rab7 uniformly labels the membranes of a compartment
referred to as a late endosomal tubulo-reticular compartment
whereas the CI-MPR is restricted to specialized regions of this
complex membranous structure. This may account for the
limited overlap of rab7 and CI-MPR labelings we observed.
However, it is presently unclear whether, in HeLa cells, rab7
is present on a tubulo-reticular compartment with regions
enriched in CI-MPR or in part present on vesicles devoid of
CI-MPR.
Our results show that rab7 is present on a compartment
connected to lysosomes. The precise sequence of events
leading to the delivery of materials destined to be degraded
to lysosomes is not clear. As suggested by Cohn et al. (1966),
lysosome biogenesis may result from the fusion of late
endosomes with dense, hydrolases-rich lysosomes. Another
possibility is that late endosomes progressively mature into
lysosomes by fusing with Golgi derived vesicles carrying
newly synthesized lysosomal enzymes and by recycling
MPRs to the trans-Golgi network (for review see Kornfeld
and Mellman, 1989). Finally, considering our data, it is
possible that lysosomes are generated by the continuous
translocation and fusion of late endosome-derived vesicles.
Indeed, it is conceivable that rab7-vesicles represent transport
intermediates that have budded from perinuclear, CI-MPRrich, late endosomes and that migrate along microtubules en
route to lysosomes. However, the existence of vesicular
transport from late endosomes to lysosomes remains to be
demonstrated. A less likely possibility is that rab7-vesicles
originate from the lysosomes. Experimental data supporting
the existence of a retrograde microtubule-dependent transport
to late endosomes have been recently published (Jahraus et
al., 1994). However, this hypothesis is not consistent with the
presence of the GTPase mutant on what would be the donor
membrane.
Rab 5 is known to regulate the dynamics of early endosome
fusion (Gorvel et al., 1991; Bucci et al., 1992). We noticed that
the overall distributions of the CI-MPR and lamps appear
unchanged upon expression of either rab7wt or rab7Q67L.
Thus, our data do not favor the view that rab7 plays a role in
the dynamics of late endosomes or lysosomes fusion. However,
subtle changes in the morphology of these compartments might
not be detected at the light microscopy level and this question
3356 S. Méresse, J.-P. Gorvel and P. Chavrier
A
B
C
D
A
C
Fig. 6. Co-staining of rab7
and the microtubule
network. Cells expressing
myc-rab7wt were grown for
72 hours in the absence of
ATc. After
paraformaldehyde fixation,
double indirect
immunofluorescence was
performed using rabbit antirab7 (A) and mouse antitubulin (B) antibodies.
Superimposed images of
rab7 and tubulin (C-D) are
presented. Identical cells are
shown in A-C; higher
magnification of another cell
in D. Bar: A-C, 10 µm; D,
20 µm.
B
D
will need to be addressed in more detail by immunoelectronic
microscopy.
Another question arose from the distribution of endoge-
Fig. 7. Nocodazole treatment causes rab7 to
redistribute to lysosomal membranes. HtTA1 cells
expressing either myc-rab7wt (A,B) or mycrab7Q67L (C,D) were incubated with 10 µM
nocodazole for 30 minutes at 37°C. Double
indirect immunofluorescences of rab7 (A,C) and
lamp-2 (B,D) were performed. Bar, 10 µm.
nous rab7 in nocodazole treated cells. While rab7wt was
nearly undetectable on lysosomes of drug-free cells, it was
clearly present on these organelles upon nocodazole
The rab7 GTPase cycles to lysosomes 3357
treatment. We observed that microtubule disruption strongly
affects the intracellular distribution of the CI-MPR. However,
no colocalization of lamp-2 with this receptor was observed.
Thus, it rules out the hypothesis that lysosomes and late
endosomes had simply fused together triggering the accumulation of rab7 on these collapsed membranes. The redistribution of rab7 in nocodazole-treated cells might also reflect a
simple inhibition of the GDI-mediated removal of GTPases,
thereby leading to the accumulation of the protein on the
acceptor membrane. However, this is very unlikely since it
was shown that nocodazole does not affect early stages of
endocytosis, which also require functional cycling of other
rab proteins (Gruenberg et al., 1989). We rather propose
another explanation. It is known that terminal endocytic
structures are very dynamic organelles that are capable of
undergoing fusion (Deng et al., 1991). In addition, the microtubule network appeared to control spatial organization of
both the rab7-compartment and lysosomes. Therefore, it is
conceivable that disruption of the microtubule network
increases the frequency of encounters and consequently the
fusion of intermediate rab7-positive vesicles with lysosomes
thereby leading to an accumulation of rab7 on these compartments. This would imply that the transport towards
lysosomes is limited by the kinetics of transport of rab7positive vesicles along microtubules. Finally, this phenomenon clearly confirms that the lysosome is a compartment on
which rab7 cycles.
In conclusion, we have shown that rab7 is localized on
putative intermediates of the endocytic pathway towards
lysosomes. Our observations suggest that these intermediates
are transported on microtubules to lysosomes and eventually
fuse with them. This hypothesis is supported by the lysosomal
localization of rab7 in its GTP-bound conformation and by the
lysosomal redistribution of this protein in nocodazole-treated
cells. Further biochemical characterization of rab7-vesicles
together with the analysis of other mutants will help to define
more precisely the localization and the function of this GTPbinding protein in the late endocytic pathway.
We are indebted to Drs Manfred Gossen and Hermann Bujard for
their generous gift of the tetracycline-controlled gene expression
system and to Dr Minoru Fukuda for providing us with anti-lamp-1
and -lamp-2 antibodies. We thank Dr Jean Davoust and Mohamed
Fathallah for helpful advice on confocal microscopy. We are grateful
to Dr Jean Gruenberg for critically reading the manuscript, and
members of the laboratory for helpful discussions. This work was
supported by CNRS, INSERM and a grant from the Association pour
la Recherche sur le Cancer.
REFERENCES
Aubry, L., Klein, G., Martiel, J. L. and Satre, M. (1993). Kinetics of
endosomal pH evolution in Dictyostelium discoideum amoebae. Study by
fluorescence spectroscopy. J. Cell Sci. 105, 861-866.
Bourne, H. R., Sanders, D. A. and McCormick, F. (1990). The GTPase
superfamily: a conserved switch for diverse cell functions. Nature 348, 125132.
Bucci, C., Parton, R. G., Mather, I. H., Stunnenberg, H., Simons, K.,
Hoflack, B. and Zerial, M. (1992). The small GTPase rab5 function as a
regulatory factor in the endocytic pathway. Cell 70, 715-728.
Carlsson, S. R., Roth, J., Piller F. and Fukuda, M. (1988). Isolation and
characterization of human lysosomal membrane glycoproteins, h-lamp1 and
h-lamp2. J. Biol. Chem. 263, 18911-18919.
Chavrier, P., Parton, R. G., Hauri, H. P., Simons, K. and Zerial, M.
(1990a). Localization of low molecular weight GTP binding proteins to
exocytic and endocytic compartments. Cell 62, 317-329.
Chavrier, P., Vingron, M., Sander, C., Simons, K. and Zerial, M. (1990b).
Molecular cloning of YPT1/SEC4-related cDNAs from an epithelial cell
line. Mol. Cell. Biol. 10, 6578-6585.
Cohn, Z. A., Hirsch, J. G. and Fedorko, M. E. (1966). The in vitro
differentiation of mononuclear phagocytes. V. The formation of
macrophages lysosomes. J. Exp. Med. 123, 757-766.
Davidson, H. W., West M. A. and Watts, C. (1990). Endocytosis,
intracellular trafficking, and processing of membrane IgG and monovalent
antigen/membrane IgG complexes in B lymphocytes. J. Immunol. 144, 41014109.
Deng, Y., Griffiths, G. and Storrie, B. (1991). Comparative behavior of
lysosomes and the pre-lysosome compartment (PLC) in in vivo cell fusion
experiments. J. Cell Sci. 99, 571-582.
Der, C. J., Finkel, T. and Cooper, G. M. (1986). Biological and biochemical
properties of human ras gene mutated at codon 61. Cell 44, 167-176.
Evan, G. I., Lewis, G. K., Ramsay, G. and Bishop, J. M. (1985). Isolation of
monoclonal antibodies specific for human c-myc proto-oncogene product.
Mol. Cell. Biol. 5, 3610-3616.
Geuze, H. J., Stoorvogel, W., Strous, G. J., Slot, J. W., Bleekemolen, J. E.
and Mellman, I. (1988). Sorting of mannose 6-phosphate receptors and
lysosomal membrane proteins in endocytic vesicles. J. Cell Biol. 107, 24912501.
Gorvel, J. P., Chavrier, P., Zerial, M. and Gruenberg, J. (1991). Rab5
controls early endosomes fusion in vitro. Cell 64, 915-925.
Gossen, M. and Bujard, H. (1992). Tight control of gene expression in
mammalian cells by tetracycline-responsive promoters. Proc. Nat. Acad. Sci.
USA 89, 5547-5551.
Gossen, M. and Bujard, H. (1993). Anhydrotetracycline, a novel effector for
tetracycline controlled gene expression systems in eukaryotic cells. Nucl.
Acids Res. 21, 4411-4412.
Goud, B., Salminen, A., Walworth, N. C. and Novick, P. J. (1988). A GTPbinding protein required for secretion rapidly associates with secretory
vesicles and the plasma membrane in yeast. Cell 53, 753-768
Goud, B., Zahraoui, A., Tavitian, A. and Saraste, J. (1990). Small GTPbinding protein associated with Golgi cisternae. Nature 345, 553-556.
Goud, B. and McCaffrey, M. (1991). Small GTP-binding proteins and their
role in transport. Curr. Opin. Cell Biol. 3, 626-633.
Griffiths, G., Hoflack, B., Simons, K., Mellman, I. and Kornfeld, S. (1988).
The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell 52,
329-341.
Griffiths, G., Matteoni, R., Back, R. and Hoflack, B. (1990).
Characterization of the cation-independent mannose 6-phosphate receptor
enriched compartment in NRK cells. J. Cell Sci. 95, 441-461.
Gruenberg, J. and Howell, K. E. (1986). Reconstitution of vesicle fusion
occurring in endocytosis with a cell free system. EMBO J. 5, 3091-3101.
Gruenberg, J., Griffiths, G. and Howell, K. E. (1989). Characterization of
the early endosome and putative endocytic carrier vesicles in vivo and with
an assay of vesicle fusion in vitro. J. Cell Biol. 108, 1301-1316.
Horiuchi, H., Giner, A., Hoflack, B. and Zerial, M. (1995). A GDP/GTP
exchange-stimulatory activity for the rab5-RabGDI complex on clathrincoated vesicles from bovine brain. J. Biol. Chem. 270, 11257-11262.
Jahraus, A., Storrie, B., Griffiths, G. and Desjardins, M. (1994). Evidence
for retrograde traffic between terminal lysosomes and the prelysosomal/late
endosome compartment. J. Cell Sci. 107, 145-157.
Kornfeld, S. and Mellman, I. (1989). The biogenesis of lysosomes. Annu. Rev.
Cell Biol. 5, 483-525
Landt, O., Grunert, H.P. and Hahn, U. (1990). A general method for rapid
site-directed mutagenesis using the polymerase chain reaction. Gene 96, 125128.
Lombardi, D., Soldati, T., Riederer, M. A., Goda, Y., Zerial, M. and
Pfeffer, S. R. (1993). Rab9 functions in transport between late endosomes
and the trans Golgi network. EMBO J. 12, 677-682.
Magee, H. R. and Newman, C. (1992). The role of lipid anchors for small G
proteins in membrane trafficking. Trends Cell Biol. 2, 318-323.
Martinez, O., Schmidt, A., Salaméro, J., Hoflack, B., Roa, M. and Goud, B.
(1994). The small GTP-binding protein rab6 functions in intra-Golgi
transport. J. Cell Biol. 127, 1603-1616.
Novick, P. and Garrett, M. D. (1994). No exchange without receipt. Nature
369, 18-19.
Ozawa, K., Szallasi, Z., Kazanietz, M. G., Blumberg, P. M., Mischak, H.,
Mushinski, J. F. and Beaven, M. A. (1993). Ca(2+)-dependent and Ca(2+)-
3358 S. Méresse, J.-P. Gorvel and P. Chavrier
independent isozymes of protein kinase C mediate exocytosis in antigenstimulated rat basophilic RBL-2H3 cells. Reconstitution of secretory
responses with Ca2+ and purified isozymes in washed permeabilized cells. J.
Biol. Chem. 268, 1749-1756.
Padh, H., Ha, J., Lavasa, M. and Steck, T. L. (1993). A post-lysosomal
compartment in Dictyostelium Discoideum J. Biol. Chem. 268, 6742-6747.
Pfeffer, S. R. (1994). Rab GTPases: master regulators of membrane trafficking.
Curr. Opin. Cell Biol. 6, 522-526.
Peter, M., Chavrier, P., Nigg, E. A. and Zerial, M. (1992). Isoprenylation of
rab proteins on structurally distinct cysteine motifs. J. Cell Sci. 102, 857-865.
Rabinowitz, S., Horstmann, H., Gordon, S. and Griffiths, G. (1992).
Immunocytochemical characterization of the endocytic and phagolysosomal
compartments in peritoneal macrophages. J. Cell Biol. 116, 95-112.
Riederer, M. A., Soldati, T., Shapiro, A. D., Lin, J. and Pfeffer, S. R. (1994).
Lysosome biogenesis requires rab9 function and receptor recycling from
endosomes to the trans-Golgi network. J. Cell Biol. 125, 573-582.
Rothman, J. E. (1994). Mechanism of intracellular transport. Nature 372, 5563.
Segev, N., Mulholland, J. and Botstein, D. (1988). The yeast GTP-binding
YPT1 protein and a mammalian counterpart are associated with the secretion
machinery. Cell 52, 915-924.
Soldati, T., Shapiro, A. D., Svejstrup, A. B. and Pfeffer, S.R. (1994).
Membrane targeting of the small GTPase Rab9 is accompanied by nucleotide
exchange. Nature 369, 18-19.
Stenmark, H., Parton, R. G., Steele-Mortimer, O., Lütcke, A., Gruenberg,
J. and Zerial, M. (1994). Inhibition of rab5 GTPase activity stimulates
membrane fusion and endocytosis. EMBO J. 13, 1287-1296.
Scheel, J., Matteoni, R., Ludwig, T., Hoflack, B. and Kreis, T. E. (1990).
Microtubule depolymerization inhibits transport of cathepsin D from the
Golgi apparatus to lysosomes. J. Cell Sci. 96, 711-720.
Schimmöller, F. and Riezman, H. (1993). Involvement of Ypt7p, a small
GTPase, in traffic from late endosome to the vacuole in yeast. J. Cell Sci.
106, 823-830.
Tanigawa, G., Orci, L., Amherdt, M., Ravazzola, M. and Rothman, J. E.
(1993). Hydrolysis of bound GTP by ARF proteins triggers uncoating of
Golgi-derived COP-coated vesicles. J. Cell Biol. 123, 1365-1371.
Temesvari, L., Rodriguez-Paris, J., Bush, J., Steck, T. L. and Cardelli, J.
(1994). Characterization of lysosomal membrane proteins of Dictyostelium
discoideum. J. Biol. Chem. 269, 25719-25727.
Tisdale, E. J., Bourne, J. R., Khosravi-Far, R., Der, C. J. and Balch, W. E.
(1992). GTP-binding mutants of rab1 and rab2 are potent inhibitors of the
vesicular transport from the endoplasmic reticulum to the Golgi complex. J.
Cell Biol. 119, 749-761.
Ullrich, O., Horiuchi, H., Bucci, C. and Zerial, M. (1994). Membrane
association of Rab5 mediated by GDP-dissociation inhibitor and
accompanied by GDP/GTP exchange. Nature 368, 157-160.
van der Sluijs, P., Hull, M., Webster, P., Mâle, P., Goud, B. and Mellman, I.
(1992). The small GTP-binding protein rab4 controls early sorting event on
the endocytic pathway. Cell 70, 729-740.
Walworth, N. C., Brennwald, P., Kabcenell, A. K., Garret, M. and Novick,
P. (1992). Hydrolysis of GTP by Sec4 protein plays an important role in
vesicular transport and is stimulated by a GTPase-activating protein in
Saccharomyces cerevisiae. Mol. Cell. Biol. 12, 2017-2028.
Wichmann, H., Hengst, L. and Gallwitz, D. (1992). Endocytosis in yeast:
Evidence for the involvement of a small GTP-binding protein (Ypt7). Cell
71, 1131-1142.
Zerial, M. and Stenmark, H. (1993). Rab GTPases in vesicular transport.
Curr. Opin. Cell Biol. 5, 613-620.
(Received 3 May 1995 - Accepted 2 August 1995)