that the plasma membrane microdomains, caveolae (Schnitzer et al., 1995), may also ... Cazaubon and Parker, 1993; Orr and Newton, 1994). ...... 200, 805-810.
2575
Journal of Cell Science 113, 2575-2584 (2000) Printed in Great Britain © The Company of Biologists Limited 2000 JCS1166
Protein kinase Cα actively downregulates through caveolae-dependent traffic to an endosomal compartment Corinne Prevostel1, Vallentin Alice2, Dominique Joubert2 and Peter J. Parker1,* 1Protein Phosphorylation Laboratory, Imperial Cancer Research Fund, 44 Lincoln’s 2INSERM U469, 141 rue de la Cardonille, 34094 Montpellier cedex 05, France
Inn Fields, London WC2a 3PX, UK
*Author for correspondence
Accepted 25 April; published on WWW 22 June 2000
SUMMARY Receptor desensitization occurs through receptor internalization and targeting to endosomes, a prerequisite for sorting and degradation. Such trafficking processes may not be restricted to membrane associated receptors but may also play an important role in the downregulation of cytoplasmic transducers such as protein kinase C (PKC). It is demonstrated here that acute TPA exposure induces the transport of activated PKCα from the plasma membrane to endosomes. This process requires PKC activity and catalytically competent PKC can even promote a similar process for a truncated regulatory domain PKCα protein. It is established that PKCα is targeted to the endosome compartment as an active kinase, where it colocalizes with annexin I, a substrate of PKC. Thus, PKCα downregulation shares features with plasma membrane
associated receptor sorting and degradation. However, it is shown that PKCα delivery to the endosome compartment is not a Rab5 mediated process in contrast to the well characterised internalisation of the transferrin receptor. An alternative route for PKCα is evidenced by the finding that the cholesterol binding drugs nystatin and filipin, known to inhibit caveolae mediated trafficking, are able to block PKCα traffic and down regulation. Consistent with this, the endosomes where PKCα is found also contain caveolin. It is concluded that the initial step in desensitisation of PKCα involves active delivery to endosomes via a caveolae mediated process.
INTRODUCTION
endocytosis (Moore et al., 1995; van der Goot, 1997; reviewed by Anderson, 1993; Anderson et al., 1992). Since caveolae can contain not only receptors but also many other proteins involved in signal transduction such as protein kinases C (PKCs; Mineo et al., 1998; Oka et al., 1997; Smart et al., 1995), it is possible that some desensitisation processes acting on cytoplasmic transducers may be mediated through this compartment. Although the mechanisms involved in the down regulation of signal transducers, including members of PKC family, are still poorly understood, it is well established that their subcellular localization closely regulates the biological function and activity of these proteins. In response to agonists, classical and novel PKCs (c/nPKC) are recruited to membranes through the acute production of diacylglycerol (and for conventional isoforms upon increased calcium concentrations). This is triggered through activation of the phosphoinositide specific phospholipase C (Noh et al., 1995), as well as from phosphatidylcholine during prolonged diacylglycerol generation (Exton, 1990). Diacylglycerol exerts an allosteric control on PKC and permits a concomitant membrane compartmentalisation stabilised by targeting proteins known to interact exclusively with the active enzyme
The transduction of external signals in cells is turned off through both membrane receptor desensitization and down regulation of their downstream signal transducers. Vesicle trafficking appears to play an important role in the control of signal transduction, since receptor desensitization involves the internalization and subsequent targeting of the receptors to early endosomes (Mellman, 1996). This endocytic process is regulated by several rab GTPases (Novick and Zerial, 1997) such as Rab5 which appears to play a critical role in the regulation of the endocytic rate. Indeed, Rab5 promotes the budding of coated vesicles from the plasma membrane (McLauchlan et al., 1998) and the targeting of these vesicles to early endosomes (Bucci et al., 1992). Rab5 has also been shown to be involved in homotypic endosome fusion (Barbieri et al., 1998a; Gorvel et al., 1991). This small GTPase is in turn regulated by upstream factors including phosphatidylinositol3 kinase (Jones and Clague, 1995; Li et al., 1995; Spiro et al., 1996) and ras (Li et al., 1997). There is in addition evidence that the plasma membrane microdomains, caveolae (Schnitzer et al., 1995), may also play a role in membrane associated receptor desensitization through a Rab5 independent
Key words: Protein kinase C, Caveolus, Downregulation, Vesicle traffic, Dephosphorylation
2576 C. Prevostel and others through specific PKC epitopes (Jaken, 1992; Mochly-Rosen, 1995). As a consequence of this spatial control the enzyme can be sequestered in the vicinity of its substrates and thus fulfils its biological function (Yedovitzky et al., 1997). Prolonged stimulation leads to PKC proteolysis which may occur through polyubiquitination and subsequent degradation by proteasomes (Lee et al., 1997; Lu et al., 1998). As a prelude to degradation, PKC becomes dephosphorylated in an agonist dependent manner (Hansra et al., 1996, 1999; Lee et al., 1996) and this dephosphorylation appears to be necessary for the subsequent degradation of the enzyme (Hansra et al., 1996). Among the three priming phosphorylation sites identified within the cPKC members α, βΙ, βΙΙ (Bornancin and Parker, 1997; Bornancin and Parker, 1996; Keranen et al., 1995; Tsutakawa et al., 1995), phosphorylation of the activation loop site (T497 in PKCα), was shown to be essential for cPKCs catalytic activity (Cazaubon et al., 1994; Cazaubon and Parker, 1993; Orr and Newton, 1994). Phosphorylation of the two carboxy-terminal sites (T638 and S657 in PKCα) plays a more complex role in the control of cPKC phosphorylation rate and seems to be involved in the accumulation of an active enzyme by keeping the kinase in a phosphatase resistant state (Bornancin and Parker, 1996, 1997). Dephosphorylation of these phosphorylation sites is therefore expected to promote a desensitisation, by switching off the kinase activity as well as possibly increasing the kinase sensitivity to proteolysis. Compartmentalization seems to be intimately involved in this process (Hansra et al., 1999), suggesting that dephosphorylation does not occur at the plasma membrane and that like membrane associated receptors, agonist induced vesicle trafficking may be required to carry the kinase into specific cellular organelles for desensitization. The desensitization of PKC elicited by tetra-decanoyl phorbol acetate (TPA) and related tumour promoters, is a consequence of an ability to activate PKC in a chronic fashion. While activation plays a central role in some of the biological effects of these promoters, long-term responses including tumour promotion itself, are inevitably tied to desensitization. A key issue is whether gain or loss of function is important in effecting tumour promotion. The tight interrelationship between the gain and loss means that a resolution of this question may only be facilitated by elucidation of the desensitization pathway. In order to analyze the involvement of TPA induced vesicle trafficking on PKC down regulation we have investigated here the effects of acute and chronic TPA exposures on the cellular compartmentalization and the associated dephosphorylation and degradation of a GFP-tagged PKCα transiently transfected in MCF7 cells. We establish that PKCα down regulation requires the transport of the kinase from the plasma membrane to cytoplasmic vesicles, where it accumulates as a phosphorylated and active enzyme. This process requires PKC activity, which can even promote in trans the traffic of a PKCα truncated form lacking the kinase domain (RDα). Immunostaining experiments clearly identified PKCα accumulating in a Rab5 positive endosomal compartment, although we demonstrate that PKCα traffic and delivery to these structures is not a Rab5 mediated process. The finding that the PKCα associated endosomes colocalize with caveolin and that cholesterol binding drugs are able to inhibit both PKCα traffic and down regulation, leads to the conclusion that
PKCα delivery to the endosome is a caveolae mediated process.
MATERIALS AND METHODS Plasmid constructions GFP-tagged full length PKCα and PKCα regulatory domain (RDα) were obtained by subcloning the PKCα coding sequence in the polylinker EcoRI and KpnI sites of the GFP-N1 vector from Clontech. In order to fuse the coding sequence of the green fluorescent protein (GFP) in frame with the C terminus of PKCα and RDα, the p-Babe vector (Alvaro et al., 1997) containing human PKCα was used as a template in a PCR reaction using the 5′ primer 5′GGAATTCCGGAGCAAGAGGTGGTT3′ and the two following 3′ primers 5′GGGGTACCCCTACTGCACTCTGTAAGATG3′ (for full length PKCα) and 5′GGGGTACCCCCGTGAGTTTCACTCGGTC3′ (for RDα). PCR amplified full length PKCα (2062 bp) and RDα (1056 bp) were digested at the created (unique) restriction sites EcoRI and KpnI and inserted into the corresponding GFP-N1 sites. The integrity of the PCR amplified PKCα (full length) and regulatory domain was verified by sequencing. Similarly, the N-terminal myc-tagged full length human PKCα was obtained by subcloning the human PKCα coding sequence in the polylinker EcoRI site of a modified pcDNA3 vector (Clontech) in which the region between BamHI and EcoRI has been replaced by a sequence encoding the myc epitope 5′ATGGAGCAGAAGCTCATATCGGAGGAGGACCTAGGGCCCGAATTC3′. The Rab5 constructs coding for the wild type (WT), the constitutively active Q79L and the S34N dominant negative mutants were a generous gift from Dr P. D. Stahl (Washington University, School of Medicine, US). MCF7 cell transfections and treatments MCF7 cells were maintained at 37°C in a 10% CO2 atmosphere in Eagles medium (E4) supplemented with 10% foetal calf serum and insulin (10 µg/ml). Eighteen hours before transfections, cells were seeded at 2.0×105 cells in 35 mm plates (for western blot experiments) or on coverslips at 0.5×105 cells in 24 well plates (for immunocytochemistry experiments) and were grown in supplemented E4 medium. DNA transfections (10 µg of GFPPKCα, GFP-RDα or myc-PKCα constructs and/or 60 µg of each Rab5 (or GFP-Rab5) constructs per 0.5×105 cells) were carried out by calcium phosphate precipitation, as previously (Olivier and Parker, 1991). Briefly, appropriate DNA concentrations were mixed with the calcium phosphate solution (140 mM NaCl, 5 mM KCl, 0.75 mM Na2HPO4 2H2O), incubated at room temperature for 30 minutes in order to allow the precipitate to form and directly added to the culture medium. Six hours later, cells were transferred to fresh E4 supplemented medium and incubated for a further 16 hours at 37°C. Treatments with TPA (100 nM) in the presence or absence of bisindolymaleimide I (10 µM; Fig. 1B) LY 294002 (50 µM; Fig. 5A) nystatin (50 µM; Fig. 5A and C) or filipin (5 µg/ml) were performed at 37°C in an E4 medium supplemented only with foetal calf serum for the times indicated in the corresponding figures. The effects of LY 294002 treatment or Rab5 (WT, Q79L, S34N) overexpression on the Rab5 dependent transferrin receptor internalization were monitored on MCF7 cells serum starved for 30 minutes and incubated with Cy3.5 labeled transferrin (50 µg/ml; see below) in the presence or absence of TPA (100 nM; Fig. 5B and 5D). Cy3.5 labeled transferrin internalization was quantified using the NIH programme on 15 pictures obtained by confocal analysis. The values obtained for MCF7 cells overexpressing either wt-Rab5, the Q79L mutant, or the S34N mutant were normalized by the values obtained for the adjacent untransfected cells present on the same pictures (internal control).
PKCα downregulation via caveolae 2577 Immunofluorescence microscopy Indirect immunostaining: treated (or untreated) MCF7 transfected cells (see above) were washed twice in PBS and fixed in 4% paraformaldehyde. Fixed cells were permeabilized in 0.2% Triton X100 for 5 minutes and incubated for 10 minutes in 1 mg/ml of freshly prepared sodium borohydride solution. Non specific sites were blocked in 1% BSA and coverslips were then incubated for 1 hour with the appropriate antibodies at the following concentrations: mouse anti-myc (1/200; ICRF production), mouse anti-Rab5 (1/100; Transduction Laboratories), Rabbit anti-Rab7 (1/100; gift from Dr Zerial, EMBL, Germany), Rabbit anti-caveolin I (1/100; Santa Cruz Biotechnology, Inc), mouse anti-annexin I (1/200; Transduction Laboratories). Coverslips were washed three times in PBS and incubated with an anti-mouse Texas red labelled molecular probe (Jackson Laboratories; 1/500) or a Cy3 anti-rabbit antibody (1/500; Amersham). Cells were washed three more times in PBS and mounted in Mowiol containing 2.5% 1,4 diazabicyclo [2,2,2] octane (DABCO; Sigma Chemical Co.). Direct immunostaining Antibodies against the phosphorylated T497, S657 and T250 sites as well as transferrin (Sigma) were directly labelled with the CyTM 3.5 monofunctional dye (Amersham International). Briefly, 2 ml of antisera were applied onto a centricon 100 column (Amicon Inc.). After a 2500 rpm centrifugation for 1 hour, 2 ml of a 50 mM bicine solution (Sigma Chemical Co.) were added to the column which was centrifuged again for 1 hour at 2500 rpm. The resulting concentrated antibody on the top of the column was then recovered in 100 µl of 50 mM bicine and was used for the subsequent Cy 3.5 labelling. CyTM3.5 was reconstituted in 20 µl of N,N dimethylformamide and used for coupling to 0.5 mg of IgG. The labelled antibody was then separated from free dye by applying it to a Bio-Rad Econo pac 10DG chromatography column (Bio-Rad) previously equilibrated with TBS (10 mM Tris, 100 mM NaCl). Elution was performed, by adding TBS to the column. The lower band corresponding to the IgG bound dye was collected. Immunocytochemistry using Cy 3.5 labelled antibodies (1/200) was performed as descibed above except that coverslips incubated with the primary labelled antibodies were washed 5 times in PBS and directly mounted in Mowiol containing 2.5% DABCO. MCF7 living cells treated with Cy3.5 transferrin were washed twice in PBS, fixed in 4% paraformaldehyde and mounted in Mowiol containing 2.5% DABCO. Confocal analysis Images were acquired using a confocal laser scanning microscope (LSM410, Carl Zeiss Inc.) equiped with a ×100/1.4 PlanAPOCHROMAT oil immersion objective. Cy2 (GFP) was excited with the 488 nm line, whereas Cy3 and Cy3.5 were excited with the 543 nm line of the of the Ar laser. Each image represents a 2dimensional projection of sections in the Z-series taken at 0.25 µm intervals across the depth of the cell. Western blot analysis For western blotting, transfected MCF7 cells (2×105) were treated as described in the figure legends, before being harvested in 100 µl of Laemmli sample buffer (Laemmli, 1970) and boiled for 10 minutes. Equal protein samples (25 µg in 25 µl) were loaded on 10% polyacrylamide gels. Separated proteins were transferred to an ImmobilonTM-P membrane (Millipore); uniform transfer was checked by staining with ponceau S. Membranes were then incubated with either a mouse monoclonal antibody against GFP (1/1000) or the rabbit antisera raised against the phosphorylated T497, S657 or T250 PKCα phosphorylation sites (Le Good et al., 1998; Ng et al., 1999). Blots were then incubated with peroxidaselinked anti-mouse IgG or anti-rabbit IgG antibodies (1/2000). Immunoblots were revealed using the ECL western blotting detection reagents (Amersham).
RESULTS PKCα induced vesicle trafficking requires the kinase activity and preludes the subsequent degradation of the enzyme Previously, our laboratory has reported that the down regulation of an overexpressed mammalian PKC in Schizosaccharomyces pombe parallels an accumulation of cytoplasmic vesicular structures (Goode et al., 1994). More recently, we have shown that acute TPA stimulation induces PKCα targeting to cytoplasmic vesicles in COS-7 cells, whereas prolonged TPA exposure leads to the enzyme accumulation in a polarized perinuclear region (Hansra et al., 1999, and see Fig. 1A). These data led us to consider whether PKC’s ability to induce vesicle trafficking may provide the context for PKC down regulation. MCF7 cells transiently transfected with GFP tagged PKCα, were TPA treated for various times in the presence or absence of bisindolylmaleimide I, a known inhibitor of cPKC and nPKC kinase activities. As shown in Fig. 1A, the PKCα induced vesicle trafficking, as well as the PKCα down regulation normally observed after 4 hours of TPA treatment, were completely inhibited by bisindolylmaleimide I. These results demonstrate that movement of activated PKCα from the plasma membrane, for sorting and subsequent degradation, requires PKCα kinase activity. A prediction of this model is that inactive forms of the enzyme would not induce vesicle trafficking, and therefore would not down regulate. Thus we investigated the consequences of short and prolonged TPA exposures on MCF7 cells transiently transfected with a truncated GFP tagged PKCα lacking the kinase domain (GFPRDα)*. Results presented in Fig. 1A show that the GFP-RDα fusion protein is localized predominantly at the plasma membrane regardless of TPA treatment (up to 1 hour). However, a longer TPA exposure (4 hours) induced an accumulation of GFP-RDα in a perinuclear region (Fig. 1A) similar to the full length GFP-PKCα. These results suggest that TPA induced vesicle trafficking seems only to be delayed for a PKCα lacking kinase activity. However the complete inhibition of RDα traffic induced by bisindolylmaleimide I treatment (Fig. 1A), suggests that the lack of kinase activity may be overcome by endogenous PKC which is able to promote the vesicle associated transport of a PKCα lacking kinase activity. It is notable that endogenous PKC is however not able to promote RDα degradation upon prolonged TPA exposure (Fig. 1B). These data demonstrate that TPA induced PKCα traffic is a PKCα kinase activity dependent process, although not necessarily requiring an intrinsic kinase activity; this can be supplied in trans by an active PKC. However, an intrinsic kinase domain seems to be required for subsequent enzyme degradation, presumably through recognition by a component(s) in the preinuclear compartment. TPA stimulated PKCα internalizes and accumulates in cytoplasmic vesicles as a phosphorylated enzyme Evidence has indicated that PKC dephosphorylation, observed upon prolonged TPA stimulation, is a prerequisite for PKC *Kinase inactive point mutants of PKCα are predominantly insoluble and can have profound dominant negative properties (Garcia-Paramio et al., 1998). The regulatory domain construct by contrast expresses well as a soluble protein (see text).
2578 C. Prevostel and others proteolysis which can occur through proteasomes (Lee et little phospho-T250 staining was detected in the PKCα al., 1996). Using PKC antibodies raised against PKCα containing vesicles after 10 minutes of TPA exposure (Fig. priming phosphorylation sites, we showed recently that 2B). Stronger PKCα phospho-T250 staining appeared after dephosphorylation at these sites was temperature sensitive, one hour of TPA treatment in the perinuclear compartment consistent with a requirement for vesicle trafficking (Hansra, (data not shown). 1999). In order to determine if PKCα induced targeting to Together these data demonstrate that no major changes in cytoplasmic vesicles is related to a change in the kinase phosphorylation state, MCF7 cells transiently transfected with GFP-PKCα were TPA treated for various times and the phosphorylation of the T497 and S657 priming sites was assayed by western blotting. As shown in Fig. 2A, the GFP tagged PKCα dephosphorylation on T497 and S657 occurs after one hour of TPA exposure. Because PKCα traffic occurs between 5 and 30 minutes of TPA treatment, it is likely that PKCα is translocated to vesicles as a phosphorylated enzyme. Indeed, the PKCα containing vesicles were stained strongly with the T497 and the S657 antisera (Fig. 2B). This staining was lost after four hours of TPA treatment when PKCα accumulates in the perinuclear compartment (data not shown). Interestingly, previous reports have established that PKCα is phosphorylated upon activation (Lu et al., 1998). The T250 phosphorylation site recently identified in our laboratory is an autophosphorylation site which could account for this PKC activity dependent phosphorylation and could therefore be considered as a marker of the PKCα activated state (Ng et al., 1999). In order to determine if the T250 phosphorylation state was modified for PKCα accumulated in vesicles, antisera raised against the T250 TPA-induced autophosphorylation site was used in western Fig. 1. PKCα induced vesicle trafficking is a kinase activity dependent process required for the blotting experiments as well as subsequent degradation of the enzyme. (A) The relationship between PKCα down regulation and in direct immunocytochemical PKCα induced traffic and targeting to cytoplasmic vesicles was investigated in MCF7 cells transiently experiments. As shown in Fig. transfected with a GFP tagged PKCα or a truncated GFP tagged PKCα lacking the kinase domain 2A, T250 phosphorylation (RDα). Cells were stimulated with TPA for the indicated times, in the presence or absence of 10 µM occurs after one hour upon bisindolylmaleimide I (BIM). The inset illustrates the vesicle heterogeneity (see text). (B) The GFPTPA treatment. This is RDα protein content following prolonged TPA treatment was determined by western blot using an anti-GFP antibody. consistent with the fact that
PKCα downregulation via caveolae 2579 Fig. 2. TPA stimulated PKCα is targeted to cytoplasmic vesicles as a phosphorylated enzyme. MCF7 cells transiently transfected with a GFP tagged PKCα were TPA stimulated for the indicated times. (A) The GFP-PKCα phosphorylation state was assayed by western blotting using antisera raised against the phosphorylation sites T497, S657 and the autophosphorylation site T250; the GFP-PKCα protein was detected using an antiGFP antibody. (B) Directly Cy3.5 labelled phosphorylation site antisera were employed to follow the phosphorylation state of the internalised GFP-PKCα protein by immunocytochemistry.
the phosphorylation state of the enzyme occur in the PKCα associated vesicles after 10 minutes of TPA treatment and establish that PKCα is targeted to these vesicles as a phosphorylated enzyme. Moreover, the finding that T250 autophosphorylation is able to occur subsequently to PKCα transport to these vesicles strongly suggests that the accumulated PKCα in that compartment is still active. It is concluded that PKCα traffics through these endosomal compartments in an active state and that the dephosphorylation that effects desensitisation and preludes degradation takes place in a distal compartment. TPA stimulated PKCα accumulates in an endosomal compartment coated with caveolae A time course of TPA treatment showed that GFP-PKCα initially moved from the plasma membrane to a vesicle compartment characterised initially by small, 500 nm, vesicles. This behaviour was asynchronous such that after 10 minutes of TPA treatment, there were cells still displaying predominantly small GFP-PKCα positive vesicles as well as cells with large GFP-PKCα positive vesicles (see inset in Fig. 1A). The temporal relationship between these compartments indicated that the initially observed smaller vesicles underwent homo/heterotypic fusion to accumulate as larger vesicles. In order to characterize the PKCα accumulating vesicles observed in response to TPA stimulation, GFP-PKCα transiently transfected MCF7 cells were TPA treated for 10 minutes and immunostained with antibodies raised against vesicular compartment associated markers. These experiments showed that the TPA induced PKCα associated vesicles colocalize with the small G-protein Rab5 (Fig. 4A). This colocalisation was visible between 5 and 30 minutes of TPA treatment (data not shown). The heterogeneity of cells with large and small vesicles is clearly illustrated in this figure (top right panel) and it is notable that while the GFP-PKCα/Rab5 staining is homogeneous within the large vesicles, there is considerable non-overlap in the small vesicles, with some polarity evident within specific vesicles (see inset). It is supposed that the fusions occurring to effect the conversion of small vesicles to large vesicles may include heterotypic events involving Rab5-positive and separate PKCα-positive vesicles. No colocalisation could be seen with other vesicle markers such as the late endosome marker Rab7 or the lysosome marker, LysoTracker Molecular probe (data not shown)*. Consistent with the finding that PKCα is targeted to the Rab5 associated endosomes, PKCα accumulating vesicles also colocalized with annexin I (Fig. 4B), which has been shown to be associated with endosomal compartments (Seemann et al., *At later times, >1 hour, we observed some colocalisation of PKCα with lysosomes.
2580 C. Prevostel and others 1996). PKCα was also associated with caveolin I in these structures (Fig. 4C). Note that the caveolin I staining on these vesicles is punctate and is not precisely coincident with GFPPKCα (see inset in Fig. 4C). While the association with Rab5 is consistent with a role for this protein in the endocytosis of PKCα, the association with caveolae-like structures questions whether PKCα is actually trafficked in a Rab5 dependent manner. PKCα traffic and delivery to endosomes is a Rab5 independent process which is mediated through caveolae Among the increasing number of the Rab proteins which are known to play a critical role in endocytosis (Novick and Zerial, 1997), Rab5 appears to be an essential regulator of the endocytic rate. Indeed, overexpression of the wild-type (WT) Rab5 or its constitutively active mutant Q79L, lacking GTP hydrolysis activity, greatly increases the endocytic rate. In contrast, overexpression of the dominant negative Rab5 mutant S34N abolishes endocytosis (Li and Stahl, 1993). Using Rab5 and GFP-Rab5 constructs, we investigated whether Rab5 could influence the behaviour of PKCα following TPA stimulation. MCF7 cells were transfected with GFP-PKCα together with either the WT-Rab5, the Q79L constitutively active Rab5 mutant or the S34N dominant negative Rab5 mutant. Neither the S34N dominant negative Rab5, nor WT-Rab5 had an effect on the extent of TPA-induced GFP-PKCα down regulation (Fig. 5A). Coexpression of either construct (or the constitutively active Rab5) did affect the efficiency of expression of PKC itself, however it is clear that the extent of downregulation is not influenced. By using GFP tagged Rab5 constructs to monitor Rab5 transfected cells specifically, it was evident that the dominant negative Rab5 blocked the transferrin receptor internalisation while not affecting the traffic of a myc tagged PKCα (Fig. 5B,C,D). However, it is notable that the GFP-S34N reduced the extent of large PKCα positive vesicle accumulation. This suggests that Rab5 activity is not necessary for PKCα traffic from the plasma membrane but may however participate in homotypic and/or heterotypic fusion events between PKCα positive vesicles/endosomes and/or endosomes (Barbieri et al.,
1998a; Gorvel et al., 1991; Fig. 5B). Direct evidence for the functionality of the expressed Rab5 proteins was provided by quantitative analysis of the extent of transferrin receptor internalisation (Fig. 5D) as well as the GFP-Rab5 vesicle dimensions. WT-Rab5 was on vesicles up to 400 nm, Rab5 Q79L on vesicles from 200-2000 nm and Rab5 S34N on vesicles
