The von Hippel-Lindau Tumor Suppressor Protein Mediates Ubiquitination of Activated Atypical Protein Kinase C*

The von Hippel-Lindau tumor-suppressor protein (pVHL) forms a protein complex (VCB-Cul2) with elongin C, elongin B, Cul-2, and Rbx1, which functions as a ubiquitin-protein ligase (E3). The (cid:1) -subunits of the hy-poxia-inducible factors have been identified as targets for the VCB-Cul2 ubiquitin ligase. However, a variety of cellular defects caused by the depletion of pVHL cannot be explained solely by the ubiquitin-mediated degradation of hypoxia-inducible factor- (cid:1) . We show here that a member of the atypical protein kinase C (PKC) group, PKC (cid:2) , is ubiquitinated by the pVHL-containing E3 enzyme. An active PKC (cid:2) mutant is ubiquitinated more extensively than wild-type PKC (cid:2) in HEK293 cells, and the ubiquitination is further enhanced by the overexpression of pVHL. The activation of wild-type PKC (cid:2) by serum stimulation of cells the reaction mixtures electropho- resed 6% SDS-PAGE gels an anti-PKC (cid:4) (cid:3) (Transduction Laboratories). C , immunoprecipitated endogenous PKC (cid:3) is ubiquitinated by VCB-Cul2 in vitro. In vitro ubiquitination reactions for immunoprecipitated PKC (cid:3) from NIH3T3 cells were performed as described under “Experimental Procedures” except for the following.

The von Hippel-Lindau (VHL) 1 tumor-suppressor gene is responsible for the inherited VHL cancer syndrome, and mutations of the VHL gene accompanied by loss of heterozygosity are also found in more than 50% of sporadic clear cell renal carcinomas and hemangioblastomas (1). The VHL protein (pVHL) has been shown to associate with elongin B, elongin C, Cullin-2 (Cul2), and Rbx1 to form the VCB-Cul2 complex. This complex is structurally analogous to SCF (Skp-1, Cullin-1, Fbox protein) ubiquitin-protein ligase (2,3) and has been shown to function as a ubiquitin ligase; pVHL is a substrate recognition subunit of the ligase complex as F-box proteins are in SCF (4). Ubiquitination is a multistep process that conjugates ubiquitin moieties to internal lysine residues of proteins. Successive conjugation of ubiquitin molecules generates polyubiquitin chains. Polyubiquitinated proteins are then degraded by the 26 S proteasome. Among the molecules involved in ubiquitin conjugation reactions, the E3 ubiquitin-protein ligases play a pivotal role in substrate recognition (5).
As a consequence of VHL gene inactivation, the expression level of vascular endothelial growth factor is increased and subsequent stimulation of angiogenesis is observed (1). Hypoxia-inducible factor-1 and -2 (HIF-1, -2) are key transcription factors for the induction of vascular endothelial growth factor gene expression. HIFs are composed of ␣ and ␤ subunits. The activity of HIFs are regulated by the oxygen-dependent degradation of the ␣ subunit of HIF-1 and HIF-2 (HIF-␣) via the ubiquitin-proteasome pathway (6 -10). However, various phenomena observed in VHL-deficient cells, including growth factor-independent cell growth (11), loss of contact inhibition (12), and abnormal organization in cytoskeletal proteins (13), cannot be attributed solely to the accumulation of HIFs. Thus, it is supposed that there are other targets for the VCB-Cul2 ubiquitin ligase. Although fibronectin (14), Sp1 (15), and atypical PKC (aPKC) (16,17) have been shown to associate with pVHL, it is not clear whether these proteins are substrates of VCB-Cul2 for ubiquitination. Moreover, the physiological significance of these interactions remains to be elucidated.
Atypical PKC comprises two members, PKC/ (PKC is the human homologue of mouse PKC) and PKC, that are implicated in signaling through lipid metabolites, including phosphatidylinositol 3-phosphates (PIP 3 ) (18). A series of studies suggests that aPKC plays an important role in various cellular processes, including proliferation (19), survival (20), and establishment and maintenance of cell polarity (21)(22)(23)(24)(25)(26). In mammalian epithelial cells, aPKC forms a complex with ASIP and mPAR-6, mammalian homologues of Caenorhabditis elegans PAR-3 and PAR-6, and localizes at tight junctions, a structural cue of cell polarity (25). Because modifications of cell polarity and cell-cell contact are fundamental steps in the tumorigenesis of epithelial cells, the interaction of pVHL and aPKC suggests the intriguing possibility that aPKC is a target of VCB-Cul2 for ubiquitination. Our previous finding that aPKC directly binds to the ␤-domain of pVHL, a target recognition site, further supports this possibility (17).
To test whether aPKC can be a substrate of VCB-Cul2 ubiquitin ligase, we monitored the ubiquitination of recombinant PKC in vivo and in vitro. In experiments using transfected HEK293T cells, we show that pVHL mediates the ubiquitination of the activated form of PKC through an interaction with the regulatory domain of PKC. By using an in vitro reconstitution system for ubiquitination, we could clearly show that the ubiquitination of PKC is mediated by pVHL-containing E3.
Cell Culture-Human embryonic kidney cells, HEK293T, were cultured in Dulbecco's modified Eagle's medium (DMEM) (Life Technologies, Inc.) containing 10% (v/v) fetal bovine serum (FBS). For the activation of wild-type PKC, cells were exposed to DMEM containing 0.5% (v/v) FBS for 20 h and then supplemented with 20% (v/v) FBS for the appropriate time period. A mouse fibroblast cell line, NIH3T3, was cultured in DMEM containing 10% (v/v) calf serum.
Transfection and Immunoprecipitation-Transfection was performed following a standard calcium phosphate co-precipitation procedure. The amounts of expression vector DNA used were 5 g for PKC, and 1 g for HA-Ub and Myc-Cul2; an appropriate amount of T7-VHL was selected for each experiment. For the ubiquitination study, cells were treated with 10 M MG132 for 6 h before harvest. After transfection, cells were cultured for 24 or 48 h for the PKC activation experiments, and then harvested. The cells were suspended in 500 l of lysis buffer composed of 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 10 mM N-ethylmaleimide, 50 mM ␤-glycerophosphate, 10 g/ml leupeptin, 2 mM phenylmethylsulfonyl fluoride, 2 mM vanadate, 20 mM NaF, and 10 g/ml aprotinin. After a 20-min incubation on ice, 0.1% cysteine was added and the suspension was incubated on ice for an additional 10 min. After centrifugation at 15,000 ϫ g for 20 min, the supernatants were incubated with an anti-PKC antibody (28) and Protein G-Sepharose 4FF (Amersham Pharmacia Biotech) at 4°C for 2 h. The immunocomplexes were washed four times with lysis buffer and twice with final wash buffer containing 50 mM Tris-HCl (pH 8.0), 1% Nonidet P-40, 0.5% deoxycholate, 50 mM ␤-glycerophosphate, 10 g/ml leupeptin, 2 mM phenylmethylsulfonyl fluoride, 2 mM vanadate, and 20 mM NaF.
Purification of Recombinant Proteins and in Vitro Ubiquitination-UbcH5b, Ubc12, and Nedd8 were expressed and purified from Escherichia coli as described previously (4). Infection of baculoviruses to Hi Five cells and purification were performed as described (4). Purified or immunoprecipitated PKC was incubated together with E1, UbcH5B, ATP regeneration system (0.5 mM ATP, 10 mM creatine phosphate, and 5 g of creatine phosphokinase), Nedd8 conjugation system (300 ng of APP-BP1-Uba3, 200 ng of Ubc12, and 1 g of Nedd8), and the indicated amount of VCB-Cul2 complex (see Fig. 6 legend) in the presence or absence of 10 g of ubiquitin at 37°C for 2 h. After stopping the reaction by adding 4ϫ SDS sample buffer, the reaction mixtures were electrophoresed on to 6% SDS-PAGE and detected by anti-PKC/ antibody (Transduction Laboratories).
Western Blotting-Cell extracts or immunocomplexes were separated by SDS-PAGE on 12% polyacrylamide gels and transferred electrophoretically to Immobilon-P polyvinylidene difluoride membranes (Millipore). The blotted membranes were soaked in phosphate-buffered saline containing 5% skim milk overnight at 4°C. The membranes were blocked with 1% normal goat serum in Tris-buffered saline containing 0.05% Tween 20 (TBST) for 1 h, and then incubated with the indicated antibody in TBST containing 0.1% bovine serum albumin for 1 h at 37°C. After washing with TBST, the membranes were incubated with a horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG antibody (KPL) and then visualized using an ECL system (Amersham Pharmacia Biotech). Densitometric analysis was performed with Molecular Analyst (Bio-Rad). The antibodies used were: anti-PKC (28), anti-PKC (Transduction Lab-oratories), anti-HA (Roche Molecular Biochemicals), anti-T7 Tag (Novagen), and anti-human VHL (PharMingen).

The Activated Form of PKC Is Preferentially Ubiquitinated in HEK293T
Cells-The identification of the VCB-Cul2 complex as an ubiquitin-protein ligase prompted us to examine whether PKC is ubiquitinated by the pVHL-containing complex. We transiently expressed PKC in HEK293T cells together with T7 epitope-tagged VHL (T7-VHL) and HA-tagged ubiquitin (HA-Ub) and monitored the conjugation of HA-Ub to immunoprecipitated PKC. As shown in Fig. 1A, faint smear bands of higher molecular mass were recognized by the anti-HA antibody in the immunoprecipitates from cells transfected with WT PKC, indicating that WT PKC is ubiquitinated (lane 3). We have observed previously that pVHL binds to the constitutively active PKC mutant (PKC4RA) more tightly than to wild-type PKC (17). Therefore, we predicted that this mutant would be ubiquitinated more efficiently by the pVHLdependent ubiquitination system. In fact, we were able to show relatively clear ubiquitination of PKC4RA in HEK293T cells (Fig. 1A, lane 2).
PKC interacts directly with pVHL (17), and the interaction depends on the N-terminal ␤-domain of pVHL, which is predicted to be a binding site for proteins targeted by VCB-Cul2 for ubiquitination (29). Taken together with the results described above, these observations strongly suggest that pVHL recognizes PKC for ubiquitination in vivo. To prove this point, we first tested the dose dependence of overexpressed pVHL (T7-VHL) on the ubiquitination of PKC4RA by transfecting different amounts of T7-VHL expression vector (Fig. 1B). As determined by densitometric analysis, the ubiquitination of PKC4RA was most prominent when 0.1 g of T7 expression vector was used (Fig. 1B, lanes 4 -6). It should be noted that the ubiquitination of PKC4RA was also observed without the expression of T7-VHL; this may depend on endogenous VHL protein in 293T cells (Fig. 1B, lane 3). To confirm the contribution of VCB-Cul2 to the ubiquitination of PKC, we employed a dominant-negative mutant of pVHL that is predicted to suppress the interaction of wild-type pVHL with PKC4RA. For the evaluation of the dominant-negative effect on ubiquitination, we used a C-terminal deleted pVHL (T7-VHL122), which binds to PKC but cannot form a VCB complex (17). As shown in Fig. 1C, the ubiquitination of PKC4RA was greatly diminished by co-expression of T7-VHL122, both in the presence and absence of exogenous pVHL.
pVHL Is Involved in the Ubiquitination of Physiologically Activated PKC-Although we suspected that activated PKC is preferentially ubiquitinated by the pVHL-containing ligase, the efficient ubiquitination of mutant PKC by VCB-Cul2 might be because of its abnormal protein structure. To detect the ubiquitination of physiologically activated PKC, we monitored the ubiquitination of wild-type PKC before and after the serum stimulation of cells, which induces the activation of PKC in vivo (30). The amount of ubiquitinated PKC increased as early as 5 min after the addition of serum and was sustained for at least 120 min ( Fig. 2A). The ubiquitinated PKC decreased to the basal level at 360 min after serum stimulation, very similar to the activation of serum-induced activation of aPKC in HepG2 cells (30). These observations strongly support the notion that the active form of PKC is preferentially ubiquitinated.
To test the contribution of VCB-Cul2 to the ubiquitination of activated PKC, we monitored the effect of T7-VHL122, which acts as a dominant-negative mutant of pVHL, on the ubiquitination level. As shown in Fig. 2B, the ubiquitination of PKC in serum-stimulated cells was greatly diminished by co-expres-pVHL-mediated Ubiquitination of PKC sion of the T7-VHL122 both in the presence and absence of exogenous pVHL, as had been observed with PKC4RA. Taken together, these observations strongly suggest that the ubiquitination of activated PKC is mediated by the VCB-Cul2 ubiquitin ligase.
Degradation of the Ubiquitin-conjugated Active Form of PKC Is Blocked by a Proteasome Inhibitor in Vivo-The most common physiological significance of ubiquitin conjugation is the marking of proteins to be degraded by the 26 S proteasome (31). To detect the ubiquitin-conjugated PKC in the experi-  4 -6). C, ubiquitination of the constitutively activated form of PKC4RA by endogenous or exogenous VHL (lanes 1 and 3) is inhibited in a dominant-negative fashion by the co-expression of T7-VHL122, a Cterminal deletion mutant lacking the ability to form the VCB complex (lanes 2 and 4). 293T cells were transfected with the indicated expression vectors. 0.5 g of DNA of the T7-VHL WT expression vector and 15 g of DNA of the T7-VHL122 expression vector were used. Immunoprecipitates obtained with an anti-PKC antibody were analyzed by immunoblotting with the indicated antibodies.

FIG. 2. Participation of VHL in the ubiquitination of physiologically activated wild-type PKC.
A, activation of wild-type PKC by serum stimulation induces its ubiquitination by VHL. 293T cells were transfected with the T7-VHL, PKCWT, and HA-Ub expression vectors. After transfection, cells were exposed to 0.5% FBS containing DMEM for 20 h for serum starvation, and then finally stimulated by adding 20% FBS for the indicated period, and analyzed as in Fig. 1. B, ubiquitination of physiologically activated PKC by the endogenous or exogenous VHL (lanes 1 and 3) is inhibited in a dominant-negative fashion by the co-expression of T7-VHL122, which lacks the ability to form a VCB complex (lanes 2 and 4). 293T cells were transfected with the indicated expression vectors. 0.1 g of DNA of the T7-VHL WT expression vector and 15 g of DNA of the T7-VHL122 expression vector were used. Transfected 293T cells were serum-stimulated as in A for 30 min and then immunoprecipitated and analyzed as in Fig. 1. pVHL-mediated Ubiquitination of PKC ments shown above, we added MG132, a proteasome inhibitor, to protect it from degradation ( Figs. 1 and 2). The ubiquitinated PKC was certainly detectable in the presence of MG132, and was more prominent in cells overexpressing pVHL (Fig. 3,  lanes 3 and 4). In contrast, ubiquitin-conjugated PKC was hardly detectable in the absence of the proteasome inhibitor (Fig. 3, lanes 1 and 2). As shown in the middle panel of Fig. 3, the total amount of immunoprecipitated PKC did not change significantly in the presence or absence of the proteasome inhibitor. This indicates that only a small fraction of the PKC activated by serum stimulation is targeted for ubiquitination and degradation by proteasomes. These observations support the notion that activated PKC is eliminated by the ubiquitinproteasome protein degradation system.
The Regulatory Domain of aPKC Is Essential for Its Ubiquitination by pVHL-We formerly reported that pVHL binds to the regulatory domain of aPKC (17). To explore whether the regulatory domain is also required for the ubiquitination of PKC, we examined the conjugation of HA-tagged ubiquitin to wild-type PKC, the regulatory domain of PKC (aa 1-242), or the kinase domain of PKC (aa 191-586), in HEK293T cells. Because the cells were not stimulated by serum in this experiment, ubiquitination of the wild-type PKC was rather limited even in the cells expressing exogenous pVHL, whereas a relatively high level of ubiquitination was observed when only the regulatory domain was present (Fig. 4, lanes 1-4). On the other hand, no ubiquitination was observed with the kinase domain alone, as expected from its inability to bind to pVHL (Fig. 4,  lanes 5 and 6). These results indicate that the regulatory domain is necessary and sufficient for ubiquitination as well as for pVHL interaction. However, we cannot rule out the possibility that the kinase domain of wild-type PKC also contains ubiquitin conjugation sites.
Both the ␣-Domain and ␤-Domain of pVHL Are Essential for Ubiquitination of aPKC-As shown in Figs. 1C and 2B, an excess amount of pVHL lacking the C-terminal ␣-domain shows a strong dominant-negative effect on the ubiquitination of PKC, indicating that VCB-Cul2 complex formation is required for the ubiquitination of PKC. We wished to verify that the N-terminal ␤-domain as well as the C-terminal ␣-domain of pVHL is required for the ubiquitination of PKC. To this end, we expressed the VHL⌬87-130 mutant bearing a deletion of amino acids 87-130 (including a large part of the ␤-domain), as well as wild-type pVHL or the VHL122 mutant bearing a deletion of the C-terminal 91 amino acids (including the ␣-domain) in HEK293T cells, and the ubiquitination of PKC was examined in these cells after serum stimulation. PKC ubiquitination was obvious in cells expressing wild-type pVHL (Fig.  5, lane 1), as already shown in Fig. 2. In contrast, the ubiquitination of PKC was quite limited in cells expressing VHL⌬87-130 or VHL122 (Fig. 5, lanes 2 and 3), and was much weaker than that in cells without the overexpression of the VHL mutants (data not shown), suggesting a dominant-negative effect of the mutants. Considering that VHL⌬87-130 and VHL122 can bind to elongin C and PKC, respectively (17), these results indicate that the formation of the VCB-Cul2 enzymatic complex is definitely required for the ubiquitination of PKC.
PKC Is Ubiquitinated in Vitro by VCB-Cul2-Results from in vivo studies strongly suggest that VCB-Cul2 ubiquitin ligase can recognize and ubiquitinate PKC. However, direct evidence that the VCB-Cul2 complex acts as a ubiquitin ligase for PKC is still missing. Therefore, an in vitro ubiquitination assay for PKC was established using purified components. All the recombinant components for the ubiquitination of PKC were either expressed in E. coli or baculovirally expressed in insect cells and purified as shown in Fig. 6A, followed by testing for ubiquitination of PKC by VCB-Cul2 in vitro. As shown in Fig. 6B, a mobility shift of PKC was observed in a ubiquitinand VCB-Cul2-dependent manner. That ubiquitin was conjugated to the mobility-shifted PKC was demonstrated by Western blot analysis using an anti-ubiquitin antibody (data not shown). Because ubiquitination of baculovirally expressed PKC might be because of the presence of malfolded proteins caused by the overexpression, we next examined whether en-  1 and 2) as a vehicle for 6 h before harvest. Immunoprecipitates obtained with the anti-PKC antibody and total cell lysates were analyzed by immunoblotting. In the absence of a proteasome inhibitor, ubiquitin-PKC conjugates were not detected (lanes 1 and 2). pVHL-mediated Ubiquitination of PKC dogenous PKC is ubiquitinated by VCB-Cul2 in vitro. As shown in Fig. 6C, the endogenous PKC immunoprecipitated from NIH3T3 cells is indeed ubiquitinated by VCB-Cul2, and it is not ubiquitinated in the absence of E2, ATP, or VCB-Cul2, suggesting that endogenous PKC is the substrate of VCB-Cul2 for ubiquitination. The ability of the VCB complex to ubiquitinate PKC in vitro strongly suggests that pVHL contributes to the regulation of aPKC activity in vivo by inducing ubiquitination and consequent proteolysis.

DISCUSSION
In this report, we showed that pVHL mediates ubiquitination of PKC in vitro and in vivo and that the activated form of PKC is preferentially targeted for ubiquitination. We also showed that both the ␣-domain and ␤-domain of pVHL are required for the VCB complex to function as a ubiquitin ligase for PKC. These results replicate the results of our previous binding study of pVHL and aPKC (17), which showed that aPKC directly associates with the ␤-domain of pVHL, and that this interaction occurs preferentially with the active form of aPKC. From these results, we propose that the VCB complex is responsible for the down-regulation of activated aPKC via the ubiquitin-proteolytic pathway, which terminates aPKC-mediated cellular signaling.
The activity of the atypical types of PKC isozymes, including PKC and -, is induced by an interaction with PIP 3 , a lipid derivative produced by phosphoinositide 3-kinase (32), or by phosphorylation of a threonine residue catalyzed by PDK1 (33). The activated enzymes may be rapidly down-regulated to terminate unnecessary PKC activity in cells. Proteolysis is a common mechanism of down-regulation reported for both conventional and novel PKC isozymes (cPKC and nPKC). The ubiquitination and the degradation of cPKC␣ and nPKC␦ and -⑀ has been reported to be triggered by the induction of their activation by 12-O-tetradecanoylphorbol-13-acetate treatment of cells (34). However, the ubiquitin ligase which is responsible for the ubiquitination of c/nPKC has not been identified. As for aPKC, this is the first report dealing with the down-regulation of the activated enzyme and the identification of the VCB-Cul2 complex as an active aPKC-specific ubiquitin ligase.  (14 kDa)). 400 ng of APP-BP1-Uba3 complex (lane 5) exhibited two bands, for APP-BP1 (65 kDa) and Uba3 (50 kDa). All lanes demonstrated that there was no serious contamination of the proteins. B, the ubiquitination of PKC is mediated by VCB-Cul2 E3 ubiquitin-protein ligase. Purified PKC was incubated together with E1, UbcH5B, the ATP regeneration system (0.5 mM ATP, 10 mM creatine phosphate, and 5 g of creatine phosphokinase), the Nedd8 conjugation system (300 ng of APP-BP1-Uba3, 200 ng of Ubc12, and 1 g of Nedd8), and the indicated amount of VCB-Cul2 complex in the presence (lanes 1-6) or absence (lanes 7-12) of 10 g of ubiquitin at 37°C for 2 h. The amount of VCB-Cul2 complex added to the assay mixture was 200 ng (lanes 2 and 8), 400 ng (lanes 3 and 9), 800 ng (lanes 4 and 10), 1600 ng (lanes 5 and 11), and 3200 ng (lanes 6 and 12). No VCB-Cul2 was added to lanes 1 and 7. After stopping the reaction by adding 4ϫ SDS sample buffer, the reaction mixtures were electrophoresed on 6% SDS-PAGE gels and detected by an anti-PKC/ antibody (Transduction Laboratories). C, immunoprecipitated endogenous PKC is ubiquitinated by VCB-Cul2 in vitro. In vitro ubiquitination reactions for immunoprecipitated PKC from NIH3T3 cells were performed as described under "Experimental Procedures" except for the following.

pVHL-mediated Ubiquitination of PKC
Ubiquitination and degradation of proteins often depend on post-translational modifications. Target proteins for the SCF complex-type ubiquitin ligases, such as ␤-catenin, p27 kip , and IB, are recognized by F-box proteins only when phosphorylated at defined amino acid residues (35,36). This molecular mechanism allows phosphorylation-dependent quick degradation of proteins, which plays crucial roles in the regulation of the cell cycle, growth, and differentiation. Degradation of HIF-␣, a target of VCB-Cul2 ubiquitin ligase, depends on oxygen, and the degradation is blocked under hypoxic conditions (6,7,9). Recent studies revealed that the hydroxylation of a proline residue located in the oxygen-dependent degradation domain is essential for the binding to pVHL (37,38). Then how does pVHL recognize active PKC? Protein kinases of the PKC family consist of a regulatory domain and a kinase domain, and the regulatory domain contains a cysteine-rich lipid binding region and a pseudosubstrate region. In nonactivated PKC, the pseudosubstrate region is predicted to interact with the kinase domain, and this intramolecular interaction may inhibit the kinase activity. It is supposed that the binding of a certain second messenger, such as PIP 3 , to the regulatory domain may lead to the dissociation of the pseudosubstrate region from the kinase domain and to the interaction with substrate molecules. This process involves a conformational change of the aPKC molecule; during this change, a specific region of regulatory domain that faces the inside of the molecule in the inactive form might be exposed to the outer environment. We showed that the free regulatory domain binds to pVHL in the absence of the kinase domain (17) and is ubiquitinated in vivo (Fig. 4). Neither the interaction of pVHL nor the ubiquitination was observed with kinase domains from which the regulatory domains have been removed (17) (Fig. 4). These observations support the idea that pVHL specifically recognizes a part of the regulatory domain that is masked in inactive aPKC. This allows the selective recruitment of the activated form of aPKC into the degradation pathway.
Although several lines of evidence have shown that the aPKC isotypes, PKC and PKC, play fundamental roles in cellular events such as proliferation (19) and survival (20), any influence of an aPKC signal transduction disorder on tumor development is rather unclear. Recent studies have elucidated the important role of aPKC in the establishment or maintenance of cell polarity (21)(22)(23)(24). Atypical PKC forms a complex with ASIP and mPAR-6, and the complex localizes in tight junctions in epithelial cells. Activated small G-proteins, Cdc42/ Rac1, also interact with this protein complex through mPAR-6, which modifies actin cytoskeleton structures and cell polarity. Therefore, one can predict that aPKC influences cell-cell and cell-substrate adhesion and actin filament organization in epithelial cells through this protein complex. In fact, the contribution of aPKC to the formation of tight junction and actin filament organization has been reported (25,39). It should be noted that pVHL has also been shown to be involved in such cellular events. Ectopic expression of pVHL in VHL-deficient renal cell carcinoma cells induces the formation of focal adhesions and stress fibers, which reduce cell motility (13). pVHL is also involved in the establishment and maintenance of contact inhibition of cell growth (12). Taken together, these observations support the idea that some of pVHL's functions as a tumor suppressor are attained at least in part by the control of aPKC activity. Given the preferential pVHL interaction and ubiquitination/degradation of the active form aPKC, one can predict that the loss of pVHL function leads to the endurance of activated aPKC, which may cause disorders in cytoskeletal organization and cell-substrate and cell-cell interaction as are commonly observed in tumorigenic cells.
When we compared the steady-state level of aPKC between VHL-deficient renal cell carcinoma cells and their derivatives with an introduced ectopic VHL gene, no significant difference was observed (data not shown). This observation is in apparent contradiction to the prediction that pVHL plays an important role in the degradation of aPKC. However, because the inactive form of aPKC, a main component of the cellular aPKC protein, is expected to be resistant to the pVHL-driven degradation pathway, the steady-state protein level of aPKC is predicted to be independent of the protein level of pVHL. The activation may occur in a small fraction of the aPKC pool, and this restricted activation allows the activation of second and third fractions of aPKC soon after the activation and degradation of the first fraction. pVHL may involved in such quick and repeatable on/off control of the aPKC activity, and a disorder of this switching system may result in tumorigenesis of epithelial cells.