Flow cytometric sorting and biochemical characterization of the late endosomal rab7-containing compartment

2682 S. Mdresse rt a/. Eleclrophowsis 1991, 18, 2682-2688

Stbphane MCresse' Pascale Andre' Zohair Mishal' Marc Barad' Nicole Brun' Michel Desjardins3 Jean-Pierre Gorvel'

'Centre d'Immunologie INSERM-CNRS de Marseille-Luminy, Marseille, France 'Centre de Soutien pour la Recherche sur le Cancer, Villejuif, France 'FacultC de MCdicine, DCpartment d'hatomie, UniversitC de MontrCal, Montrkal, Quebec, Canada

Flow cytometric sorting and biochemical characterization of the late endosomal rab7-containing compartment

Rab7 is a small molecular weight GTPase that is known to be associated with late endocytic compartments. Studies in which wild-type or mutant forms of this protein have been overexpressed in mammalian cells have indicated that rab7 plays a role in controlling membrane transport between late endocytic compartments. However, both the precise site(s) of action and localization of rab7 remain unclear. In the present study, we have used density-gradient centrifugation in combination with a new epitope-specific flow cytometric sorting method to isolate rab7-containing vesicles from baby hamster kidney (BHK) cells. Electron-micrographs of sorted elements showed a homogeneous population of vesicles that resembles late endosomes. The polypeptide composition of rab7-containing vesicles was then analyzed by two-dimensional (2-D) :gel electrophoresis. Rab7-containing vesicles were enriched in the cation-independent mannose 6-phosphate receptor and especially in the precursor forms of cathepsin D. Taken together, these results show that the rab7-containing vesicles are a component of the endocytic pathway that connects late endosomes and lysosomes and in which precursor forms of lysosomal hydrolases, segregated from their receptor, might be included.

1 Introduction

After a ligand binds to its cell-surface receptor, ligand- receptor complexes are internalized via coated vesicles that rapidly uncoat and fuse with early endosomes. The acidic environment in this compartment triggers the dissociation of most ligand-receptor complexes [ 1, 21. Some receptors, like the transferrin receptor, are sorted into recycling vesicles that mediate their transport back to the plasma membrane. Others, like the LDL-receptor, are transported to late endosomal compartments by carrier vesicles that bud from peripheral early endosomes and migrate along microtubules to a perinuclear location where they fuse with late endosomes [3, 41. Molecules destined for degradation are eventually transferred to lysosomes by a process that remains enigmatic. In order to maintain the integrity of intracellular compartments the specificity of the multiple interactions which occur between them must be strictly regulated. This regulation requires a complex molecular machinery in which rab proteins may confer vectoriality and fidelity [5, 61. These small GTPases are anchored by prenylation at the cytoplasmic face of distinct organelles and it has been pro- posed that they control the correct partnering of vesicles with their target membranes [7, 81. The function of these proteins is dependent on their ability to cycle between GDP- and GTP-bound conformations. The results of expression of GTPase-deficient mutants, both in yeast and mammalian cells, suggest that the GTP-bound form

Correspondence: Dr. J. P. Gorvel, CIML, Parc Scientifique de Luminy, case 906, F-13288 Marseille Cedex 9, France (Tel: +33-4-9126-9466; Fax: +33-4-9126-9430; E-mail: [email protected])

Nonstandard abbreviations: BHK, baby hamster kidney; CI-MPR, cation-independent mannose 6-phosphate receptor; FSC, forward angle light scatter; SSC, side angle light scatter; VSV-G, spike glyco- Drotein C of vesicular stomatitis virus

Keywords: Rab7 / Late endosomes / Lysosomes / Flow cytometric sorting / Two-dimensional polyacrylamide gels electrophoresis

is the active conformation. The concept emerging from recent studies is that rab GTPases serve as guanine nucleotide-dependent membrane anchors for down- stream effectors [9] and ensure a spatial and temporal control of membrane docking and fusion events [lo].

Several rab GTPases are associated with the endocytic pathway. Among them, rab5 and rab4 are localized to specific compartments of the early endosomal network. Rab5 regulates endocytosis via clathrin-coated vesicles and early endosome fusion [11, 121, while rab4 controls recycling from early endosomes to the plasma membrane [13]. Three GTPases, rdb7, rab9 and rab24, have been found on late endosomes [14-161. The role of rab9 in the recycling of the cation-independent mannose 6-phosphate receptor (CI-MPR) from late endosomes to the trans-Golgi network has been established [15, 171 and rab24, which is also localized in the endoplasmic reticulum, might play a role in the autophagy process [16]. However, the function of rab7, both in yeast and mammalian cells, still gives rise to much debate. Ypt7, the yeast homolog of rab7, regulates the integrity of the vacuole [18] and transport from late endosomes to the vacuole. Similarly, in mammalian cells, rab7 clearly exerts its effect later in the endocytic pathway than rab.5, though it remains unclear at exactly which steps i t is involved [19]. We have previously shown that a GTPase- defective mutant of rab7 is partially redistributed from a prelysosomal compartment to lysosomes. A lysosomal redistribution of this protein was also observed upon depolymerization of microtubules, suggesting that rab7 plays a role in the vesicular traffic to lysosomes [20]. In contrast, the accumulation of spike glycoprotein Gr of vesicular stomatitis (VSV-G) proteins in transferrin-positive vesicles, and delayed cleavage of the paramyxovirus SV5 hemagglutinin-neuraminidase (SV5 HN) in cells expressing dominant negative forms of rab7 led Feng et at. [21] to favor a more upstream effect: regulating early to late endosome transport. Finally, rab7 has recently been found to be involved in the modulation of the

0 WILEY-VCH Verlag G m b H , 69451 Weinheim, 1997 0173-0835/97/1414-2682 $17.50+.50/0

li‘lecfrophorrsis 1997, 18. 2682-2688 Surting and characterization o f rah7-containing vesicles 2683

homotypic fusion between late endosomes in a cell-free system [22].

Thus, while rab7 clearly plays an important role in controlling late steps of endocytosis, its precise localization and function must still be clarified. Although it is gene- rally considered as a marker for late endosomes, we [20] and more recently others [19] have shown that rab7 is not restricted to the CI-MRP-rich compartment. Indeed, it appears to localize to a very extensive compartment stretching from the perinuclear region to the cell per- iphery. In the current study, we combined density gradient centrifugation and flow cytometry using rab7- specific antibodies to purify rab7-containing vesicles from baby hamster kidney (BHK) cells. Isolated vesicles were analyzed by electron microscopy, two-dimensional (2-D) gel electrophoresis, and Western blotting.

2 Materials and methods

2.1 Materials

Cell culture reagents were purchased from Gibco-BRL (Cergy Pontoise, France). FITC-conjugated secondary antibody was from Jackson Immunoresearch Laborato- ries (Immunotech, France). Molecular weight standards were from Amersham (Les Ulis, France). Acrylamide was purchased from National Diagnostic (Atlanta, GA, USA). Keyhole limpet hemocyanin was from France Bio- chem (Meudon, France). Immobilon-P was from Milli- pore Corporation (St. Quentin-Yvelines, France). Unless otherwise indicated, all other reagents were obtained from Sigma Chimie (St. Quentin-Fallavier, France).

2.2 Cells

BHK cells were grown and maintained in BHK-21 medium supplemented with 10% fetal calf serum and 10% tryptose phosphate broth as described by Gruen- berg et al. [3]. Cells were metabolically labeled overnight in methionine-free RPMI, containing 50 pCi/mL [35S]methionine, 5% dialyzed fetal calf serum, and 5% dialyzed tryptose phosphate broth.

2.3 Antibodies

Antibodies were obtained by repeated immunizations of rabbits with synthetic peptides covalently coupled to keyhole limpet hemocyanin. Anti-rab7 antibodies were raised against a synthetic peptide corresponding to the carboxy terminus of the protein (residues 172-204). To obtain anti-lamp1 antibodies, the 11 residue cytoplasmic domain of this protein (RKRSHAGYQTI), which is highly conserved among species [23], was used as an immunogen. Both antibodies were affinity-purified using the appropriate peptide immobilized on an Affigel 10 (Bio-Rad) column. Anti-rab5 antibody was a gift from Dr. Philipe Chavrier (CIML, Marseille, France). Anti- CI-MPR antibody was kindly provided by Dr. Bernard Hoflack (IBL, Lille, France). Anti-cathepsin D antibody was a gift from Dr. Stuart Kornfeld (St. Louis, MO, USA).

2.4 Preparation of early and late endosome-enriched

For each experiment, 10 X 10 cm Petri dishes of BHK cells were plated the day before use and grown overnight at 37°C. All subsequent manipulations were done at 4°C. Early and late endosome-enriched fractions were prepared following a previously described procedure [ 121. Briefly, monolayers were washed in PBS, then scraped and gently homogenized in homogenization buffer (HB), containing 250 mM sucrose, 3 mM imidazole, pH 7.4, and a protease inhibitor cocktail (0.5 mM EDTA, 05 mM PMSF, 1 pg/mL antipain, 1 vg/mL pepstatin, 15 pg/mL benzamidin, and 1 yg/mL leupeptin). A postnuclear supernatant (PNS) was prepared by centrifugation for 10 min at 1000 g, brought to 40.6% sucrose and 1 mM MgCl, and overlaid with 1.5 mL of 35% sucrose in 3 mM imidazole/l mM MgCl,, 1 mL 30% sucrose in 3 mM imi- dazolell mM MgCI, and 1 mL HB. Step gradients were run for 60 min at 124 000 g in an SW 60 rotor. Early and late endosomes were collected in 1 mL from the 35- 30% and 30%-HB interfaces, respectively.

membrane fractions

2.5 Flow cytometric analysis and sorting

For flow cytometry, 50 yL of membranes were diluted in 250 pL of PBS containing 2% horse serum in the presence of 5 pg/mL of affinity-purified anti-rab5 or anti-rab7 antibodies. FITC-conjugated donkey anti-rabbit antibody was then added at a 11300 final dilution. After 10 min, samples were diluted to 1 mL with PBS-2% horse serum and immediately analyzed using a Becton Dickinson FACStar Plus flow cytometer. The cell sorter was cali- brated with microspheres of 0.1, 0.2, 0.5 and 1.5 pm diameter (Polysciences Inc, Warrington, PA, USA). We excluded by threshold in the side angle light scatter channel (SSC) electronic noise and particles present in PBS-2% horse serum. The filter setting for FITC was 525/530 nm Oriel band-pass. The laser excitation was 488 nm. The sheath pressure was 11 PSI and the fre- quency 24 kHz. The diameter of the nozzle tip was 70 ym. Sorting was performed in the ‘coincidence out’ mode with a sort envelope of 3. Sorted elements were harvested in PBS-2 O/o horse serum and immediately pelleted by centrifugation at 100 000 g in a TLlOO table-top centrifuge (Beckman).

2.6 Confocal and electron microscopy

A drop of sorted elements was included between a microscope slide and a coverslip and immediately viewed under a Leica TDS 4DA confocal microscope. Series of two plane sections of 0.2 ym thickness were monitored. For electron microscopy, membrane pellets of sorted vesicles were fixed in 0.1 M sucrose, 0.1 M caco- dylate, 2% glutaraldehyde, pH 7.2. Pellets were dehy- drated in an ethanol series before embedding in Epon 812. Thin sections were cut randomly and examined with a Zeiss 901 electron microscope.

2.7 2-D gel electrophoresis

Membrane pellets were dissolved in 9.8 M urea, 4% Nonidet P-40, 2% Ampholines pH 7-9 and 100 mM

2684 S . Meresse et al. Electrophoresis 1997. 18, 2682-2688

dithiothreitol. A combination of IEF and SDS-PAGE was used to resolve proteins in two dimensions. Tube gels used for the IEF gels were 25 cm long and 2.5 mm in internal diameter. IEF gels were run at 1200 V for 17 h. The pH gradient was linear between pH 4.5 and pH 7.4. The second-dimensional resolving gels were 12% w/v acrylamide and prepared for autoradiography using Intensify (Dupont, New England Nuclear, Boston, MA, USA).

2.8 Western blotting

Early/late endosome-enriched membrane fractions and sorted rab7-containing vesicles were resuspended in Laemmli sample buffer, separated on 8-12 O/o SDS- PAGE and transferred onto Immobilon-P. Blocking, incu- bation with antibodies, and washings were performed in PBS with 4% w/v dry milk powder and 0.05% v/v Tween 20. Detection was performed using a peroxidase-conjugated goat anti-rabbit IgG with the enhanced chemilumi- nescence system (ECL, Amersham, UK). Blots were scanned and then quantified using the NIH Image software. Sorted rab7-positive vesicles contained a very low amount of proteins impairing precise protein quantita- tion. Therefore, fractions were normalized for equivalent amounts of rab7. Aliquots of rab7-containing vesicles and various dilutions of early and late endosome-enriched fractions were separated by SDS-PAGE and transferred to Immobilon-P. The area of the membrane corresponding to the 21-30 kDa region was probed with the anti-rab7 antibody. The dilution of endosome fractions which yielded the same intensity of signal for rab7 as the rab7-containing vesicles was selected for further analysis.

3 Results

3.1 Flow cytometric analysis of rab7-containing vesicles

In a previous study, we developed multiparametric flow cytometric analyses and a method of early endosome sorting in order to better characterize the mechanisms of early endosome membrane dynamics in v i m [24]. In par- ticular, we investigated the roles of rab4 and rab5 and showed that early endosomal membrane fusion is effi- ciently stimulated by recombinant rab5 whereas the fis- sion process is stimulated by rab4. In order to gain new insights into the function of rab7 in endocytosis, we have now developed a method for the sorting of rab7- containing vesicles based on the binding of specific anti- rab7 antibodies. First, we analyzed the distribution of vesicles, which were labeled by affinity-purified anti-rab7 antibodies. Crude PNS analyzed by flow cytometry dis- played a very low specific labeling with the anti-rab7 antibody. Therefore, in order to increase the percentage of endosomal vesicles, the PNS was first fractionated on a sucrose step flotation gradient [12]. Early and late endo- soma1 membrane fractions, previously shown to be enriched in rab5 and rab7, respectively, were prepared from BHK cells [12] and analyzed by flow cytometry as described in Section 2.5. Rab7-containing vesicles accounted for about 10% of total elements present in the late endosomal fraction. SSC versus forward angle light scatter (FSC) dot-blot analysis of this fraction (Fig. 1A)

showed that most of the unlabeled particles exhibited a low SSC value. An R1 region (Fig. 1A) which excluded these elements was designed in order to increase the ratio of rab7-positive versus rab7-negative elements. After gating, 40-70% (range for n = 8) of elements from the late endosome-enriched fraction were shown to express rab7 (Fig. 1D) whereas only 5-20°/o were positive for rab5 (Fig. 1C). Conversely, analysis of early endosome-enriched membranes showed that 50-65% and 20-30% of elements were positive for rab5 and rab7, respectively (not shown). This is in agreement with the biochemical partition of rab5 and rab7 in early and late endosome fractions, which has been demonstrated using this sucrose gradient [12].

3.2 Flow cytometric sorting of rab7-containing vesicles

The highly fluorescent rab7-containing vesicles of the R2 region (Fig. 1D) were sorted by flow cytometry. Sample flow rate was optimized to allow the sorting of 150-;!00 positive events per second. Between 3 X lo5 and 6 X lo5 rab7 elements were sorted from the equivalent of 500 cm2 of confluent BHK cells. Re-analysis of sorted vesicles by flow cytometry showed that this fraction contained over 95 "10 fluorescent rab7-positive particles (Fig. 1E).

3.3 Microscopic analyses of sorted rab7-containing vesicles

Fluorescence microscopy indicated that the sorted population was composed of discrete vesicles (Fig. 2a). By electron microscopy, this fraction corresponded to a homogeneous population of vesicles of 0.48 t 0.17 Mm (n = 100) average diameter (Fig. 2b). Most vesicles contained internal membranes and exhibited the typical morphology of BHK late endosomes as observed both in vivo or after subcellular fractionation [3, 121.

3.4 Polypeptide composition of rab7-containing vesicles

The protein composition of rab7-containing vesicles was investigated by 2-D gel electrophoresis. BHK cells were metabolically labeled to equilibrium with ['5S]niethio- nine. Late endosome-enriched membranes were prepared and rab7-containing vesicles were isolated by flow cytometry as described above. Patterns of polypeptide from late endosome-enriched membranes and rab7- containing vesicles were analyzed and compared. Both samples exhibited complex patterns with approximately 100 major polypeptides. Although the majority of polypeptides present in rab7-rich vesicles (Fig. 3B) were also detected in late endosomes (Fig. 3A), more than 20 polypeptides present in the late endosome fraction were not detected in the rab7-containing vesicles (Fig. 3C, gray spots). In contrast, about 30 polypeptides showed a high enrichment in rab7-containing vesicles (Fig. 3D, gray spots). This indicates that there are differences in polypeptide composition between the rab7-containing vesicles and the late endosome-enriched fraction, accounting for at least 50 different proteins. We compared our 2-D gel maps with those of endosomal compartments of BHK cells published by Aniento et al. [4]. The polypep-

Sorting and characterization of rab7-containing vesicles 2685

0 0 E 0 0 co

0 0 rC1

0 0 V

0 0 +I

_. . . , . . .. . .

SSC- Height

E

1 0" 10" FlTC fluorescence

Figure 1. Flow cytometric analysis and sorting of rab7-containing vesicles. Late endosome-enriched membranes were collected from a sucrose flotation gradient (see Section 2.4) and analyzed. (A) S S U F S C dot plot analysis. The R1 region was designed to optimize the ratio of rab7-positive (gray dots) versus negative elements (black spots). (B)-(E) Histogram plots of the gated (Rl) population showing the fluorescein emission in a logarithmic scale versus the number of events using the secondary FITC-labeled antibody (B), anti-rab5 (C) or anti-rab7 (D, E) pri- mary antibodies. The dotted line shows the fluorescence of unlabeled membranes (B) or the fluorescence resulting from the nonspecific binding of the secondary antibody (C), (D). Events of the R2 region (D) that correspond to a homogeneous population of highly fluorescent vesicles were sorted. (E) Independent experiment in which nonfluo- rescent (dotted line) and fluorescent (solid line) elements were sorted and subsequently re-analyzed.

tide pattern of rab7-containing vesicles was clearly distinct from that of endosomal carrier vesicles which mediate transport from early to late endosomes. In contrast, similarities were observed with late endosomes since at least 50 common polypeptides could be identi- fied. Interestingly, the two acidic glycoproteins 4A1 and 2A5, shown by Aniento et al. [4] to be abundant in late endosomes, were not detected in sorted rab7-containing vesicles.

lular compartments by Western blotting (Fig. 4). For these experiments aliquots of early or late endosome- enriched fractions and rab7-containing vesicles which contained equivalent amounts of rab7 were loaded onto 8-12% SDS-PAGE (see Section 2.7). Rab5, a marker for early endosomes, was present in low amounts in late endosome-enriched fractions and barely detectable in rab7-containing vesicles. The levels of the CI-MPR and cathepsin D, a lysosomal hydrolase known to be transported in an MPR-dependent manner, were examined. These two proteins were already noticeably enriched in late versus early endosomal fractions of the sucrose step gradient. The CI-MPR showed a 3- to 6-fold (n = 4)

3.5 Rab7-containing vesicles are enriched in CI-MPR and precursor forms of cathepsin D

To further characterize rab7-containing vesicles, we enrichment over late endosomes in sorted vesicles. The looked for the presence of specific markers of subcel- situation appeared to be more complex for cathepsin D.

2686 S . Meresse ct a/ Electruptruresis 1997, 18, 2682-2688

Figure 2. Morphological characterization of sorted rab7-containing vesicles. Sorted vesicles were either immediately observed by (A) confocal laser scanning microscopy or (B) pelleted by centrifugation and prepared for electron microscopy. Fluorescence confocal microscopy (A) revealed the presence of discrete fluorescent vesicles of a homogeneous size. By electron-microscopy (B) rdb7-containing vesicles are formed or intramembrane vesicles which are typical of BHK late endosomes. Bar: A, 1 vm; B, 0.3 pm.

This hydrolase is synthesized as a 53 kDa precursor which acquires mannose 6-phosphate residues during its transport through the Golgi apparatus. The 53 kDa precursor form is then transported to an endosomal compartment where its processing into the 47 kDa form starts. Further proteolytic cleavage to the 31 kDa mature form is thought to occur in lysosomes. While the late endosome fraction contained almost exclusively cathepsin D in its intermediate and mature forms, rab7- containing vesicles showed a striking enrichment in the 53 kDa precursor form (> 25-fold over late endosome- containing fractions). This 53 kDa form was almost unde- tectable in the PNS (not shown) and in the late endosomes. Interestingly, rab7-positive vesicles were also enriched in another intermediate form of 49 kDa [25]. The intermediate (47 kDa form) and mature (31 kDa) forms of cathepsin D accounted only for 25% and 5% in rab7- containing vesicles, respectively. These results are con- sistent with studies suggesting that the first cleavage of cathepsin D occurs in a prelysosomal compartment [26, 271 or showing that late endosomal membranes contain only low amounts of mature cathepsin D [28]. Finally, the presence of the lysosomal protein lamp1 was examined. Only trace amounts of this protein were found in the early endosome fraction and lamp1 was slightly less abundant in the purified rab7-containing vesicles than in the late endosomes.

4 Discussion In the present study, we have developed a new method to highly purify rab7-containing vesicles that combines a

sucrose step flotation gradient with epitope-specific flow cytometric sorting. We show that sorted vesicles have the morphological and biochemical characteristics that one would expect of a prelysosomal compartment. In addition, we established the two-dimensional electrophoresis map of this intracellular compartment. The sorting procedure was based on the recognition of rab7 located at the cytoplasmic face of endosomal vesicles by indirect im- munofluorescence using affinity-purified anti-rab7 antibodies. Flow cytometric sorting was highly efficient, as demonstrated by the re-analysis of sorted material (Fig. 1E) and by the striking enrichment (> 25-fold) of this fraction in the 53 and 49 kDa precursor forms of cathepsin D (Fig. 4). Compared to standard techniques currently used to investigate endocytosis, such as differ- ential centrifugation, the major advantage of flow cytometry is that vesicle separation is based on the presence of a marker protein, i.e. rab7, rather than on biophysical cri- teria. Although flow cytometry was previously used for measuring the acidification activity of endocytic compartments (for a review see [29]) or for the analysis/sorting of endocytic organelles from cells that had internalized a fluorescent fluid phase probe [24, 30, 311, the originality of this study is to combine an epitope-based flow cytometric sorting together with a biochemical analysis of sorted elements.

Irnmuno-isolation of subcellular compartments on1 a solid support [4, 28, 32, 331 and epitope-based flow cytometric sorting share a similar rationale. In both techniques, organelles that have bound specific antibodies are selected. Late endosomal compartments have been immuno-isolated using various epitopes. Aniento et al. [4] used the cytoplasmic domain of the spike VSV-G protein as an antigen for immuno-isolation after this prolein was implanted into the plasma membrane of BHK cells, internalized, and chased into various endocytic compartments. We found striking similarities for the total protein composition of rab7-containing vesicles when compared with late endosomes but not with endosomal carrier vesicles nor with early endosomes. These two compartments were also similar in size and morphology, indi- cating that rab7-containing vesicles are late endocytic compartments.

It has been shown that newly synthesized cathepsin ID is transported from the trans-Golgi network to the prelyso- soma1 compartment bound to its receptor, the Cl-MPR [34]. Here we show by Western blotting that there is a significant enrichment of both CI-MPR and cathepsin D precursor (53 and 49 kDa) forms in rab7-containing vesicles when compared with the late endosome-enriched fractions. The 49 kDa form of cathepsin D, which is sig- nificantly increased in rab7-containing vesicles, was barely detectable in late endosome-enriched membranes (Fig. 4). This form may correspond to the high molecular weight intermediate form which has been shown to accumulate in rabbit fibroblasts grown in the presence of protease inhibitors [25]. Interestingly, this limited proteolysis is the result of autocatalysis whereas processing to the 47 kDa form requires at least one other protease, possibly cathepsin B [35]. In addition, rab7-rich vesicles contain low amounts of intermediate and mature forms of cathepsin D. Taken together, these observations sug-

t ' l rc r rophor~s is 1997, 18. 2682-2688

I F F

Sorting and characterization of' rab7-containing vesicles 2687

Figurv 3. 2-D gel analysis of the protein composition of sorted rab7-containing veyicles. Late endosome-enriched membranes were prepared from ['iS]methionine-labeled BHK cells. Rab7-containing vesicles were sorted by flow cytometry, analyzed by 2-D electrophoresis (B), (D) and compared with the late endosome-enriched membranes (A), (C). Directions of IEF in the first dimension and SDS-PAGE in the second dimension are indicated. Gels were treated for fluorography and auloradiographed (A), (B). Autoradiographies were scanned and majors spots were outlined using the Adobe Photoshop software (C), (D). Spots which are absent in rab7-containing vesicles and present in late endosomes (C) or shown to be highly enriched in rab7-containing vesicles (D) are in gray. Molecular mass markers are indicated (200, 97, 69, 46, 30, and 14 kDa).

gest that, soon after dissociation from the C1-MPR, a limited maturation of the lysosomal enzyme occurs in this compartment.

A similar enrichment in the 53 kDa precursor form of cathepsin D has been reported for prelysosomal membranes immuno-isolated from HepG2 cells using antibodies directed against the cytoplasmic domain of the CI-MPR [28]. Although the presence of rab7 was not analyzed in that study, the combination of data strongly sug- gests that rab7- and the CI-MPR-rich membranes are related compartments. Interestingly, sorted vesicles were 4-8 times more enriched in precursor forms of cathepsin D than in Cl-MPR. This may suggest that lysosomal pro- teases accumulate in rab7-containing vesicles before being transported to lysosomes. However, the limited proteolysis of cathepsin D indicates that this compart-

ment does not contain the high proteolytic activities that would result from such an accumulation. Another expla- nation would be that the bulk of procathepsin D resides in rab7-positive transport intermediates from which the CI-MPR would be excluded. This would correlate with the limited colocalization of both molecules previously observed in macrophages [36 ] and HeLa cells [20]. Finally we cannot exclude the possibility that a subpopu- lation of rab-7 containing vesicles was selected by the gating function of the flow cytometer.

Altogether, these data provide strong evidence that rab7- containing vesicles are part of the late endocytic pathway which includes the CI-MPR-rich compartment. This is in good agreement with our model which pro- poses that this small GTPase is present on late endosomes and on late endosome-derived vesicles en route

2688 S. MCresse pf 01. Electrophoresis 1997, 18. 2682-26188

Figure 4. Western blotting analysis of sorted rab7-containing vesicles. Early (EE) and late endosome (LE)-enriched fractions and sorted rab7-containing vesicles (S) were pelleted by centrifugation and resuspended in sample buffer. Equivalent amounts of rab7 from each fraction were loaded on 8-12"/0 SDS-PAGE, transferred onto Immobilon- P and analyzed by Western blotting using various antibodies. Positions of the 53 kDa precursor (P), 49 kDa intermediate (arrow), 47 kDa intermediate ( I ) , and 31 kDa mature (M) forms of cathepsin D and molecular mass markers are indicated.

to lysosomes [20]. This hypothesis, previously based on the intracellular localization of rab7, is now supported by the biochemical characterization of the rab7-containing vesicles. In conclusion, we have developed a method to rapidly isolate epitope-containing organelles by flow cytometry. This technique will now be extended in order to highly purify other intracellular compartments using antibodies raised against specific cytoplasmic determinants.

We wish to thank Dr. Philippe Chavrier, Dr. Bernard Hoflack and Dr. Stuart Kornfeld for providing antibodies. We are grateful to Christiane Rondeau .for expert technical assistance and to Drs. Jonathan Ewbank and Olivia Steele- Mortimer for their critical review of the manuscript. This work was supported by CNRS, INSERM and a grant from the Association pour la Recherche sur le Cancer.

Received July 17, 1997

5 References

Mellman, I . , Ann. Rev. Cell Dev. Rial . 1996, 12, 575-625. Gruenberg, J . , Griffiths, G., Howell, K. E., J . Cell B i d . 1989, 108,

Aniento, F., Emans, N., Griffiths, G., Gruenberg, J. , J . Cell Biol.

Pfeffer, S. R., Curr. Opin. Cell Biol. 1994, 6 , 522-526. Nuoffer, C., Balch, W. E., Annu. Rev. Biochem. 1994, 63, 949-990. Sollner, T., Whiteheart, S. W., Brunner, M.. Erdjument-Bromage, H., Geromanos, S., Tempst, P., Rothman, J. E.. Nature 1993, 362,

Novick, P., Brennwald, P., Cell 1993. 75, 597-601. Stenmark, H., Vitale, G., Ullrich, O., Zerial, M., Cell 1995, 83,

Rybin, V., Ullrich, O., Rubino, M., Alexandrov, K., Simon, I., Seabra, C., Goody, R., Zerial, M., Nature 1996, 383, 266-269. Bucci, C., Parton, R. G., Mather, 1. H., Stunnenberg, H., Simons, K., Hoflack, B., Zerial, M., Cell 1992, 70, 715-728. Gorvel, J. P., Chavrier, P., Zerial, M., Gruenberg, J. , Cell 1991, 64,

Van der Sluijs, P., Hull, M., Webster, P., Male, P., Goud, B., Mellman, I., Cell 1992, 70, 729-740. Chavrier, P., Parton, R. G . , Hauri. €1. P., Simons, K., Zerial, 'M., Cell 1990, 62, 317-329.

1301-1316.

1993, 123, 1373-1387.

318-324.

423-432.

9 15-925.

[15] Lombardi, D., Soldati, T., Riederer, M. A,. Goda, Y., Zerial, M.,

[16] Olkkonen, V. M., Dupree, P., Killisch, I., Lutcke, A,, Zerial, M.,

[17] Riederer, M. A., Soldati, T., Shapiro, A. D., Lin, J., Pfeffer, S. R.,

Pfeffer, S. R., EMBO J . 1993, 12, 677-682.

Simons, K., J . Cell Sci. 1993, 106, 1249-1261.

J . Cell Biol. 1994, 125, 573-582. Wichmann, H., Hengst, L., Gallwitz, D., Cell 1992, 71, 1131-1142. Vitelli, R., Santillo, M., Lattero, D., Chiariello, M., Bifulco, iM., Bruni, C. B., Bucci, C., J. B i d . Chem. 1997, 272, 4391-4397. Meresse, S., Gorvel, J. P., Chavrier. P., J . Cell Sci. 1995, 108,

Feng, Y., Press, B., Wandinger-Ness, A,, J . Cell B i d . 1995, 131,

Papini, E., Satin, B., Bucci, C., de Bernard, M., Telford, J. L., Manetti, R., Rappuoli, R., Zerial, M., Montecucco, C., EMBO J .

Hunziker, W., Geuze, H. J., Bioessays 1996, 18, 379-389. Chavrier, P., Van der Sluijs, P., Mishal, Z., Nagelkerken, B . , Gorvel, J.-P., Cytometry 1997, 29, 41-47. Samarel, A. M., Ferguson, A. G., Decker, R. S., Lesch, M., Am. J . Physiol. 1989, 257, C1069-C1079. Gieselmann, V., Pohlmann, R., Hasilik, A,, von Figura, K., J . Cel l Bid . 1983, 97, 1-5. Rijnboutt, S. , Stoorvogel, W., Geuze, H. J., Strous, G . J., J. B id . Chem. 1992, 267, 15665-15672. Delbruck, R., Desel, C., von Figura, K., Hille-Rehfeld, A,, Eur. J . Cell Biol. 1994, 64, 7-14. Wilson, R. B., Murphy, R. F., Methods Cell B i d . 1989, 31, 293-317. Murphy, R. F., Proc. Natl. Acad. Sci. USA 1985, 82, 8523-8526. Harris, H. W., Jr., Zeidel, M. L., Jo, I., Hammond, T. G., L Biol . Chem. 1994, 269, 11993-12000. Howell, K. E., Schmid, R., Ugelstad, J., Gruenberg, J., Methods Cell Biol. 1989, 31, 265-292. Messner, D. J . , Griffiths, G. , Kornfeld, S . , J . Cell Biol. 1989, 108,

Kornfeld, S., Annu. Rev. Biochem. 1992, 61, 307-330. Samarel, A. M., Worobec, S. W., Ferguson, A. G., Decker, R. S., Lesch, M., Am J . Physiol. 1986, 250, C589-C596.

3349-3358.

1435-1452.

1997, 16, 15-24.

2149-2162.

[36] Rabinowitz, S. , Horstmann, H., Gordon, S . , Griffiths, G., J . Cell Biol. 1992, 116, 95-112.

[ l ] Mellman, I . , Fuchs, R., Helenius, A,, Annu. Rev. Biochem. 1986, 55, 663-700.

Documents

Flow cytometric sorting and biochemical characterization of the late endosomal rab7-containing compartment