Tumor suppressor protein VHL is induced at high cell density and mediates contact inhibition of cell growth

Abstract

In spite of the general recognition of von Hippel-Lindau (VHL) as a tumor suppressor gene, the physiological and pathological importance of VHL protein in cell growth regulation and tumorigenesis remains unclear. Here we show that in normal human renal proximal tubule epithelial cells (RPTEC), the steady-state amount of VHL protein is strictly regulated by cell density. The cellular VHL content is more than 100-fold higher in dense cultures than in sparse cultures. The increase in VHL protein at high cell density was also observed for NIH3T3 fibroblasts, suggesting the generality of the phenomenon. The growth rates of renal cell carcinoma cells lacking an intact VHL gene and their derivatives with wild-type or mutant VHL expression vector do not differ significantly when they are growing in log-phase. Importantly, however, there is a difference when they reach confluency: cells lacking wild-type VHL grew continuously, while cells expressing exogenous VHL protein showed relatively limited cell growth. Using an ecdysone-inducible VHL expressing cell line, we also show that the growth inhibition at high cell density can be released by attenuating the VHL expression. Taken together, we propose that VHL protein functions as a growth suppressor at high cell density, and this might be the basis of the tumor suppressor function of VHL.

Introduction

Inactivation of the von Hippel-Lindau (VHL) gene is associated with the development of von Hippel-Lindau disease, which is characterized by a predisposition to develop tumors or cysts in many tissues including the eyes, central nervous system, kidneys, adrenal glands, and pancreas (Latif et al., 1993; Melmon and Rosen, 1964). The VHL gene is also inactivated in the majority (80%) of sporadic clear cell type renal cell carcinomas (RCC), the most common type of kidney cancer that develops in the renal proximal tubule (Foster et al., 1994; Gnarra et al., 1994; Shuin et al., 1994). These observations, together with experimental data showing that the re-introduction of a normal VHL gene into a VHL-deficient RCC cell line suppresses its tumorigenicity (Iliopoulos et al., 1995), indicate that VHL is a tumor suppressor gene. Nevertheless, the molecular basis of tumor formation by the inactivation of VHL gene remains obscure.

Recently, biochemical analysis of the VHL protein revealed that this tumor suppressor gene product forms a stable complex with elongin C, elongin B, and Cul-2, called the VCB-Cul-2 complex (Lonergan et al., 1998; Pause et al., 1997), which shows structural and functional homology with a ubiquitin ligase E3 enzyme, the SCF complex (Iwai et al., 1999; Kamura et al., 1999; Lisztwan et al., 1999). One of the potential targets of the VHL-containing E3 enzyme is hypoxia-inducible factor-1α (HIF-1α), a transcription factor responsible for the induction of the VEGF gene (Cockman et al., 2000; Maxwell et al., 1999; Ohh et al., 2000). Therefore, it is conceivable that a decelerated breakdown of HIF-1α caused by the loss of the VHL gene results in the acceleration of VEGF production and tumor angiogenesis. In fact, hyper-vascularization is a typical feature of tumors in VHL diseases such as hemangioblastomas and clear cell type renal cell carcinoma. However, angiogenesis itself may not be an initial event in tumorigenesis, while it is supposed to be essential for tumor outgrowth (Juda, 1995). This is supported by the fact that VHL disease causes the formation of poorly vascularized cysts in various tissues (Lamiell et al., 1989) as well as hyper-vascularized tumors. Therefore, VHL must also be involved directly in the growth regulation of epithelial cells.

In normal epithelial cells, growth rate is tightly regulated by different mechanisms depending on growth factors and cell–cell or cell–substrate contact, and most of these mechanisms are lost or disordered in tumor forming cells. Renal cell carcinoma 786-O cells, which lack an intact VHL gene, have been shown to acquire the ability to grow independently on serum in the culture medium (Pause et al., 1998). Moreover, 786-O cells can grow as spheroids independent of cell–substrate contact (Lieubeau-Teillet et al., 1998). The ability to grow as spheroids may also mean a disturbance of contact inhibition in 786-O cells because cell density is extremely high for spheroid cells. However, it has been shown that DNA synthesis is greatly reduced in 786-O cells growing on culture dishes at high cell density (Pause et al., 1998) indicating the retention of contact inhibition at least in part. Further investigations are required to clarify the function of VHL in contact inhibition and to understand the function of VHL in cell growth regulation.

The VHL protein has been detected in several human tissues by immuno-histochemical staining (Corless et al., 1997; Los et al., 1996). However, biochemical identification of endogenous VHL proteins by immuno-precipitation and Western blot analysis has been reported only in restricted cultured cell lines (Gao et al., 1995; Iliopoulos et al., 1995; Kibel et al., 1995; Schoenfeld et al., 1998), and no quantitative analysis to describe dynamic changes in the amount of VHL protein has been reported. Such analysis may help to reveal the biological function of VHL. To understand the function of VHL in the growth regulation of normal epithelial cells, we first tried to monitor the amount of endogenous VHL protein in human renal proximal tubule epithelial cells (RPTEC) at different phases of cell growth. Here we show that the amount of VHL protein is tightly regulated by the density of RPTEC and significantly increases at high cell density. We also show that VHL functions as a suppressor of cell growth only at high cell density, by using VHL deficient RCC cell lines and their derivatives carrying an exogenous VHL gene.

Results

Identification of the long and short forms of VHL proteins in human renal proximal tubule epithelial cells and mouse NIH3T3 cells

We detected endogenous VHL protein in human renal proximal tubule epithelial cells (RPTEC) and mouse NIH3T3 cells by immuno-precipitation followed by Western blot analysis using antibodies against human VHL protein. Human VHL mRNA can be translated from either the first or the second methionines and gives two proteins with apparent molecular masses of 28–30 kD (long form) and 18–19 kD (short form) respectively (Blankenship et al., 1999; Iliopoulos et al., 1998; Schoenfeld et al., 1998). Despite the absence of the N-terminal 53 amino acids in the short form of the protein, no significant differences in their biochemical properties or tumor suppressor function have been reported. We identified two proteins expressed in RPTEC as VHL gene products, whose apparent molecular weights on SDS–PAGE are identical to those of the long and short forms of human VHL proteins over-expressed in COS-1 cells (Figure 1). We used a mouse monoclonal antibody (Ig32) against human VHL protein to immuno-precipitate and an affinity purified rabbit antibody for Western blotting. Therefore, these proteins were recognized by two independent antibodies against the VHL protein, supporting the accuracy of the identification. Notably, the small molecular size protein was much more abundant than the large one, which gave only a slight signal on Western blotting (Figure 1, lane 6). These antibodies also react with proteins of 24 kD (indicated by asterisk in Figure 1, lane 6). They are most likely a degradation product of VHL long form, because pulse-label/chase analysis of over expressed human VHL long form revealed the production of a partially degraded protein of the same molecular mass (data not shown). Taken together, we conclude that the short form of VHL protein is predominantly expressed in RPTEC.

Figure 1
figure1

Detection of mouse and human endogenous pVHL. NIH3T3 cells (lanes 3 and 4) and human normal renal proximal tubular cells (RPTEC; lanes 5 and 6) were harvested and lysed in lysis buffer (RIPA). The cell lysates containing 80 μg of proteins were immuno-precipitated with anti-human VHL monoclonal antibody (Ig32; lanes 4 and 6) or anti-T7 monoclonal antibody (T7; lanes 3 and 5) as a control. The immuno-precipitated proteins were applied to SDS-polyacrylamide gel (15%) electrophoresis. Then, Western blot analysis was performed with affinity purified anti-VHL polyclonal antibody. Lysates of COS-1 cells expressing the human pVHL long form (1–213 a.a.; lane 1) or human pVHL short form (54–213 a.a.; lane 2) were also analysed by Western blotting. The pVHLlong form is seen at about 28 kD, although its expected molecular mass is 24 kD. An asterisk indicates the position of a putative degraded product of human pVHL long form. In mouse NIH3T3 cells, two different protein sizes, 21 kD and 19 kD, were detected using the same set of VHL (lane 4). The arrowhead indicates the position of mouse pVHL short (19 kD), which is only slightly visible in this figure

In mouse NIH3T3 cells, two different protein sizes, 21 kD and 19 kD, were detected using the same set of VHL antibodies as above (Figure 1, lane 4). Mouse VHL mRNA can also be translated from either the first or the second methionine and gives two proteins, 21 kD (long form) and 19 kD (short form) (Gao et al., 1995). Therefore, we identified these proteins detected by anti-human VHL antibodies as mouse VHL proteins expressed in NIH3T3 cells. In contrast to the VHL proteins in human cells, the long form of the VHL protein is the major species in mouse NIH3T3 cells.

Dynamic changes in the steady-state amount of VHL protein in cultured cells including human RPTEC

During our attempts to detect cellular VHL protein expressed in a variety of cultured cells, we came to suspect that the amount of VHL might depend on the culture conditions employed, especially on cell density. In order to clarify this point, we used human RPTEC and monitored the cellular VHL protein after seeding the cells in culture dishes at extremely low density, 8×104 cells in a 6 cm dish. As shown in Figure 2a, the VHL short form was clearly detected in the Ig32 immuno-precipitate, while the VHL long form signal was very faint at all time points (data not shown, see Figure 1, lane 6). Importantly, the amount of VHL short form shows a drastic increase in parallel with growth of the cell culture (the increase in cell density). The relative amount of VHL protein short form in cells in the sub-confluent state reached more than 100-fold of that in cells 1 day after the seeding, and this high level expression was maintained thereafter. VHL protein long form was not detectable at early days and it was relatively less abundant as shown in Figure 1, lane 6 even when the amount of the short form increased dramatically (data not shown).

Figure 2
figure2

Endogenous pVHL is induced when the cell density increases. RPTECs maintained in a contact naive state were seeded at 8×104 cells in 6 cm dishes one day before the start of the assay. At each time point, three dishes were trypsinized and cell numbers were determined with a hemocytometer (a, upper panel), and some dishes were harvested for immunoprecipitation of pVHL. Cell lysates were normalized for protein concentration and 65 μg of protein was used for the quantification of pVHL by immuno-precipitation and Western blotting (a, lower panel). The intensity of chemiluminescence was directly measured by a Luminescent Image Analyzer LAS-1000 (Fuji Film) and the results are indicated as relative values to the intensity at 1 day after seeding (see bottom of the lower panel in a). PVHL long was not detectable under the exposure conditions used here. Essentially the same results were obtained from two independent experiments. To monitor the cell cycle and protein concentrations, p27kip1, cyclinA and actin in the cell lysate were analysed by Western blot analysis (b). Cells were photographed 1 day or 13 days after seeding (c). To eliminate the possibility that some cytokines released from the cells induce the expression of pVHL, the conditioned medium of a confluent culture (sp. + C.M.) or fresh medium (sp.) was supplied to the sparse cultures of RPTEC for 48 h, and pVHL was quantified by immuno-precipitation and Western blotting. (d). RPTECs were seeded under sparse conditions (1×105 cells/15 cm-dish) or confluent conditions (5×105 cells/3.5 cm-dish). After 2 days, the same number of cells at different density were lysed in SDS–PAGE sample buffer and VHL protein was detected by direct Western blotting using affinity purified anti-VHL polyclonal antibody (the upper panel in (e). Actin was also detected by Western blotting (the lower panel of e)

Figure 2a also shows that the growth rate of RPTEC gradually decreases along with the increase in cell density. Cell growth was almost arrested 9–13 days after seeding, when most of the cells made contact with each other (Figure 2c). The amount of VHL completely paralleled the decrease in growth rate. The amount of p27kip1 also increased as the cell density increased, while cyclin A decreased (Figure 2b). The amounts of p27kip1 and cyclin A most likely reflect the state of the cell cycle (Lehner and O'Farrell, 1989; Pines and Hunter, 1990; Poon et al., 1995). These results suggest that the amount of VHL protein is induced upon cell density-dependent growth suppression. To exclude the possibility that the increase in VHL protein depends on the increased concentration of certain factor(s) secreted from RPTEC in the culture medium, conditioned medium from sub-confluent RPTEC cultures or fresh culture medium was supplied to a sparse culture of RPTEC and the amount of VHL protein was compared after 2 days. As show in Figure 2d, the amount of VHL was not induced but rather reduced in cells supplied with conditioned medium. The increase in the amount of VHL protein was not due to senescence of RPTEC, because the cells started to grow again and the VHL level was reduced when sub-confluent cells were sparsely reseeded (data not shown). Moreover, the amount of VHL protein was also significantly higher in sub-confluent cells than in dispersed cells when the cells were seeded concurrently 2 days before preparing cell lysates of different densities (Figures 2e and 4 uppermost panel). The increase in the amount of VHL protein was also detectable by direct Western blot analysis without immuno-precipitation (Figure 2e). Therefore, the difference was not likely due to the changes in the efficiency of immuno-precipitation. Taken together, these observations strongly support the notion that the VHL protein level is positively regulated by cell–cell contact.

Figure 4
figure4

pVHL increases as the VCB-complex at high cell density. RPTEC cells were seeded under sparse conditions (1×105 cells/10 cm-dish) or confluent conditions (1×106 cells/6 cm-dish). After 2 days, cell lysates were prepared, and lysates containing equal amounts of protein were used for immuno-precipitation with anti-human VHL monoclonal antibody (Ig32). After SDS-polyacrylamide gel electrophoresis, pVHL short, Cul-2, elonginB, elonginC, and Rbx1 in the immuno-complex were detected by Western blot analysis using specific antibodies. Western blot analysis of actin is shown as a control for the adjustment of the amount of protein in the cell lysates. Note that each component of the VCB complex (Cul-2, elonginB, elonginC, Rbx1) immuno-precipitated in proportion to the amount of immuno-precipitated pVHL. The same experiments were performed more than three times and a typical result is shown

The cell density-dependent increase in VHL was also observed for mouse NIH3T3 fibroblasts. NIH3T3 cells were seeded sparsely (1×105 cells/10 cm-dish) or to confluently (1×107 cells/10 cm-dish), and lysates containing equal amounts of protein were prepared after 2 days for immuno-precipitation with anti-human VHL monoclonal antibody (Ig32). Western blot analysis using affinity purified anti-VHL polyclonal antibody gave a band migrating at 21 kD corresponding to the mouse VHL prtein long form (Figure 3, see also Figure 1, lane 4). Western blotting of the cell lysates also revealed that the difference in VHL amount between sparse and dense cultures parallels that of p27kip1 and is in clear contrast to cyclin A, which shows a completely opposite pattern of signal intensity, whereas the amount of actin remains constant (Figure 3). Similar experiments on rat fibroblast 3Y1 cells gave very similar results (data not shown), suggesting the generality of the phenomenon.

Figure 3
figure3

pVHL levels in fibroblasts increased at high cell density too. NIH3T3 cells were seeded under sparse conditions (1×105 cells/10 cm-dish) or confluent conditions (1×107 cells/10 cm-dish). After 2 days, cell lysates were prepared, and lysates containing equal amounts of protein were used for immuno-precipitation with anti-human VHL monoclonal antibody (Ig32). Mouse pVHL long was detected by Western blot analysis using affinity purified anti-VHL polyclonal antibody (upper most panel). The amounts of p27kip1 and cyclin A were estimated by direct Western blotting using cell lysates containing equal amounts of proteins. CBB staining of a 46 kD protein (actin) is shown as a control. The same experiments were performed more than three times and a typical result is shown

VHL protein induced at high cell density forms a VCB-Cul2 complex

The VHL protein is reported to form a complex with elongin C, elongin B, Cul-2, and Rbx-1 (Iwai et al., 1999; Kamura et al., 1999; Lisztwan et al., 1999; Lonergan et al., 1998; Pause et al., 1997). To determine whether the increase in VHL protein leads to an increase in the amount of the protein complex, we examined the amount of these protein components co-precipitated with the VHL protein from RPTEC. As shown in Figure 4, VHL-associated proteins increased in close correlation with VHL protein induction. Therefore, the cell density-dependent increase in the amount of VHL should lead to an increase in the amount of the VHL-containing protein complex that has been suggested to be an ubiquitin ligase E3 enzyme.

Ectopic expression of VHL suppresses cell growth at high cell density

Given the increase in the amount of VHL protein complex upon contact inhibition of cell growth, the question arises whether VHL is required for the establishment of contact inhibition. To address this question, we compared growth profiles of renal carcinoma cell lines lacking an intact VHL gene, UMRC-6 and 786-O, and their stable transformants expressing exogenous VHL or its mutant. The growth profiles of UMRC-6 and its derivatives expressing the VHL protein long form (VHL long) or VHL protein short form (VHL short) did not differ from each other until they reached the sub-confluent state, 1×106 cells in 6 cm dish. However, they showed considerable differences in growth rate thereafter (Figure 5a). Parental UMRC-6 cells grew at the same rate until they reached a density of 4×106 cells in a 6 cm dish and grew continuously thereafter reaching 1×107 cells in a 6 cm dish 14 days after seeding. In contrast, the growth rate of UMRC-6 cells expressing the long or short form VHL protein started to decrease when the cells reached a density of state, 1×106 cells in a 6 cm dish, and their growth was mostly arrested before 14 days after seeding. Statistical analysis indicates that the difference in cell number is not significant among each cell lines until 4 days after seeding, while it becomes significant (P<0.01) after 5 days. It should be noted that only a limited number of cells if any, detached from dish during that period. Therefore, the counted cell numbers directly show the state of cell proliferation. The number of cells in a 6 cm dish at this time was 4×106 cells for cells expressing VHL short and 7×106 cells for cells expressing VHL long (Figure 5a). The differences in the growth profiles between these two cell lines could be due to the different expression levels of the long or short forms of VHL (Figure 5a, insert) and/or to differences in the biological activity of the long and short forms of the VHL protein. In any case, these observations suggest that VHL functions in the growth arrest induced by cell–cell contact.

Figure 5
figure5

pVHL acts as a growth suppressor only at high cell density. Cell counts were obtained for renal cell carcinoma cell lines lacking an intact VHL gene and their stable transformants expressing wild-type VHL or mutant VHL. Parental UMRC-6 cells, UMRC6 (−/−), and stable transformant expressing human pVHL long form, UMRC6 (VHL long) or human pVHL short form, UMRC (VHL short), and cell number were counted 1 to 14 days after seeding (a). The expression of exogenous pVHL was examined by Western blot analysis using an affinity-purified polyclonal antibody against human pVHL (insert in a). Parental 786-O cells, 786-O (−/−) and stable transformants expressing truncated mutants of pVHL long form, 786-O (VHL1-115a.a.) or wild-type pVHL long form, 786-O (VHL long) were seeded at a density of 1×105 cells per 6 cm dish, and cell number was determined 1–11 days after seeding (b). In both experiments (a) and (b), there were few cells which were detached and floating. So, the cell number indicated here reflects the cell growth rate

This notion was further supported by the growth profile of another renal carcinoma cell line, 786-O, and its derivatives expressing exogenous VHL long or VHL truncated mutants lacking the C-terminal elongin C binding domain (VHL 1–155 a.a.) (Iliopoulos et al., 1995). The growth rates of all these cell lines decreased after the cell number exceeded 1×106 cells in a 6 cm dish. However, the decreased second phase growth rate of parental 786-O cells and mutant VHL expressing cells was significantly higher than that of cells expressing wild-type VHL long (P<0.01) (Figure 5b). Since we analysed only one wt VHL-expressing 786-O cell line, one can argue that the difference in the growth rate may be explained by a clonal effect. To diminish this possibility, we also compared the growth rate of 786-O cells transfected with adenovirus vectors for human VHL short form or lacZ. Again we observed a significant growth suppression by the expression of VHL protein only at high cell density (data not shown).

Figure 6 shows a typical microscopic view of the above cells at the over-confluent state, 14 days after seeding for UMRC-6 derived cell lines and 11 days after seeding for 786-O derived cell lines. Parental UMRC-6 cells formed numerous large foci (Figure 6a) covering most of the culture dish, while UMRC-6 cells expressing the long or short forms of VHL grew as a monolayer showing the cobblestone-like feature characteristic of epithelial cells (Figure 6b,c). Parental 786-O cells and the mutant VHL-expressing cells also formed numerous large foci covering most of the culture dish (Figure 6d,e). In contrast, 786-O cells expressing VHL long formed only small foci, and most of the culture dish were covered by cells growing as a monolayer (Figure 6f). There were few cells stained by Trypan blue in the all dishes. These observations strongly support the notion that VHL plays a key role in contact inhibition of cell growth.

Figure 6
figure6

pVHL suppresses focus formation in monolayer cultures. Parental UMRC-6 cells, UMRC-6 (−/−) (a), and stable transformants expressing human pVHL long form, UMRC-6 (VHL long) (b) or human pVHL short form, UMRC-6 (VHL short) (c) were allowed to grow for 14 days to an over-confluent state. Parental 786-O cells, 786-O (−/−) (d) and stable transformants expressing truncated mutants of pVHL long form, 786-O (VHL1–115 a.a.) (e) or wild-type pVHL long form, 786-O (VHL long) (f) were seeded at a density of 1×105 cells per 6 cm dish, and allowed to grow for 11 days to an over-confluent state. Cells were fixed with methanol and stained with Giesma. Un-fixed cells were stained with Trypan blue to certify that most of the cells were alive (data not shown). Bars, 30 μm

VHL withdrawal causes an acceleration of cell growth at high cell density

The amount of VHL in RPTEC remained high for 4 days after the cells reached sub-confluency (see Figure 2a, 13 days after seeding), and the VHL level did not decline before another 4 days. This suggests that continuous expression of the VHL protein is required for the establishment of contact inhibition. To test this possibility, we produced a stable transformant of UMRC-6 cells that can express VHL short in a muristerone/RXR-dependent manner (see insert of Figure 7). Cells were cultured in the presence of muristerone until they reached confluency, and then muristerone was depleted from the culture medium to reduce VHL expression, while keeping muristerone in the culture medium of the control group. As shown in Figure 7, cell growth was accelerated by the depletion of muristerone, which causes a reduction in the amount of VHL. In contrast, control cells retaining VHL expression showed only limited, if any, growth, and their growth profile was quite similar to that of VHL short expressing cells as described in Figure 5a. The differences in the cell numbers after the depletion of muristerone are statistically significant (day 16: P <0.05, day 18: P<0.01). It should be noted that muristerone itself does not affect the growth profile of UMRC-6 cells expressing only RXR, a muristerone-dependent transcription activator (data not shown). Taken together, these observations clearly indicate that the continuous expression of VHL protein is required to maintain growth inhibition and further support the notion that VHL plays a key role in the contact inhibition of cell growth.

Figure 7
figure7

pVHL functions in the maintenance of contact inhibition. A stable clone of UMRC-6 expressing pVHL short (UMRC-6/PIND-VHL) in a muristerone A-dependent manner was pre-cultured in medium containing 1 mg/ml of Zeocin, 1 mg/ml of G418 and 10 μM muristerone A for 24 h. For cell-count experiments, cells were seeded at 6×105 cells/6 cm dish and allowed to grow in the presence of muristerone A for 14 days until the cells reached confluency. Then muristerone A was removed from the culture medium of the experimental group while muristerone A remained in the culture medium of the control group. Cell growth was accelerated by the depletion of muristerone (day 16 P<0.05, day 18 P<0.01), which causes a reduction in the amount of VHL. The expression of pVHL short in the cells cultured for 2 days in the presence (+) or absence (−) of muristerone A was examined by immuno-precipitation with Ig32 followed by Western blot analysis using affinity purified anti-VHL polyclonal antibody. Note no pVHL short form expression in UMRC-6/PIND-VHL in the absence of muristerone A (lane 2) and the induced expression of pVHL short form in the presence of muristerone A (lane 3)

Discussion

In the present study, we demonstrate that the amount of VHL protein is strictly regulated by cell density in cells, including primary cultures of human epithelial cells (RPTEC) and mouse fibroblast (NIH3T3). This provides a novel basis for understanding the physiological and pathological roles of the VHL protein. For example, it suggests that the VHL protein functions at high cell density. Taken together with the tumor suppressor function of VHL, our present demonstration suggests the involvement of VHL in growth suppression at high cell density. This notion is strongly supported by two additional observations described in this report. One is that the growth rate of VHL-deficient cells at high cell density is reduced considerably by the re-introduction of the VHL gene. Second is that the attenuation of VHL expression at high cell density results in an acceleration of cell growth rate.

Quantitative analysis of endogenous VHL protein may provide a critical information for understanding its function. However, little is known about the quantitative changes in endogenous VHL proteins, and this is the first, to our knowledge, description of these changes. Since VHL mRNA is not significantly induced at high cell density (Baba, unpublished data), the increase in amount of protein may depend mainly on the increase in the translation rate or the stability of the VHL protein. In any case, the drastic induction of the VHL protein upon the increase in cell density strongly suggests that the VHL protein functions at high cell density, and this allowed us to investigate the function of VHL in growth suppression at high cell density.

Up to now, the function of VHL in cell growth regulation is only poorly understood. It has been reported that the re-introduction of the VHL gene in renal cell carcinoma 786-O cells does not cause any changes in growth rate in culture dishes (Pause et al., 1998). We also could not detect any significant effect of VHL expression in 786-O cells or UMRC-6 cells at low cell density (Figure 5a), while the amount of VHL protein in these cells are several-fold higher than that of RPTEC at high cell density (Baba, unpublished observation). These observations indicate that VHL is not a functional growth suppressor in cells growing on dishes at low density. However, the re-introduction of the VHL gene caused a considerable reduction of growth rate in cells at high density (Figure 5a,b). In other words, VHL functions as a growth suppressor only in cells at high density, even when the VHL protein is forced to express at all times at any cell density. Growth suppression by VHL expression has also been reported for cells grown as multicellular tumor spheroids (Lieubeau-Teillet et al., 1998). Under these conditions, each cell maintains contacts in three dimensions and cell density is extremely high. Therefore, cells growing in such culture conditions could be in a situation similar to that of cells growing on culture dishes at high cell density, where VHL can function as a growth suppressor. On the other hand, it has been reported that renal cell carcinoma 786-O cells, which lack an intact VHL gene, still show contact inhibition of cell growth (Pause et al., 1998). Our results show that the growth rate is in fact reduced at high cell density when the cells are approaching confluency. However, the reduction is not as strict as that observed with VHL expressing cells (Figure 5a,b). Together with the quantitative changes in the endogenous VHL protein, these observations indicate that the VHL protein functions as a growth suppressor at high cell density.

How does the VHL protein suppress cell growth? Since suppression was observed only at high cell density, VHL is most likely involved in so-called contact inhibition. The molecular mechanism of contact inhibition is not fully understood; however, the importance of E-cadhelin and β-catenin for the establishment of contact inhibition has been suggested in several kinds of cells including fibroblasts, myoblasts, and epithelial cells (Kandikonda et al., 1996; Orford et al., 1999; Soler et al., 1999; Watabe et al., 1994). Recently β-catenin has been shown to function as a partner of transcription factor TCF to induce the expression of growth regulating genes such as c-myc (He et al., 1998). The loss of free β-catenin to form a complex with TCF may cause the suppression of cell growth that may induced by the trapping of β-catenin by E-cadherin or by the phosphorylation and subsequent degradation of β-catenin induced by Wnt signaling (Behrens et al., 1996; Hinck et al., 1994; Korinek et al., 1997; Morin et al., 1997; Orford et al., 1997). The degradation of β-catenin is accelerated by ubiquitination catalysed by a ubiquitin ligase E3 enzyme containing FWD1/β-Trcp (Hart et al., 1999; Kitagawa et al., 1999). Since the VHL protein induced at high cell density is associated with other components of the ubiquitin ligase E3 enzyme, elongin B/C, Cul-2, and Rbx-1, they might catalyse the ubiquitination of β-catenin or another component of the signaling pathway. Notably, a CDK inhibitor, p27kip1, is induced by the over-expression of VHL (Kim et al., 1998) as well as by contact inhibition (St. Croix et al., 1998), suggesting that VHL acts through p27kip1 in growth suppression. In addition, a direct interaction between PKCs and the VHL protein has been reported, and is known to be involved in the regulation of cell growth or apoptosis (Okuda et al., 1999; Pal et al., 1997). VHL also acts as a Nedd8 ligase for Cul-2, which may be involved in cell cycle regulation (Feng et al., 1999; Liakopoulos et al., 1999; Wada et al., 1999). However, the importance of these proteins in VHL-dependent growth suppression remains to be determined.

Renal proximal tubule epithelial cells are one of the major origins of tumors developed by VHL gene inactivation. Abnormal cell growth may be caused by the loss of normal features of epithelial cells, including contact inhibition of cell growth, anchorage- and growth factor-dependency of cell growth, and defined cell polarity. The observations reported here suggest that the deficiency in contact inhibition is a trigger for VHL tumorigenesis in renal proximal tubules. The hyper vascularization and loss of other epithelial cell features may be involved in the acceleration of tumorigenesis. However, the initial event caused by the loss of VHL in normal epithelial cells is not yet identified. Further analysis of epithelial cell features dependent on the VHL protein, as well as a search for the target molecule of VHL, may help to identify this initial event, and provide new aspects to the physiological function of VHL and the mechanism of tumorigenesis.

Materials and methods

Cell culture

Human normal renal proximal tubuler epithelial cells (RPTEC) were purchased from Clonetics, and maintained in the recommended medium, REGM(Clonetics). NIH3T3 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 7% calf serum (7% CS/DMEM). COS-1 and 3Y1 cells were maintained in DMEM supplemented with 10% fetal calf serum (10% FCS/DMEM).

Renal cell carcinoma cell lines, which lack wild-type VHL, and stable transfectants of VHL were gifts. 786-O, 786-O (HA-VHL1–213 a.a.: VHLlong), 786-O (HA-VHL1–115 a.a.) were from Dr William G Kaelin. UMRC-6, UMRC-6 (VHL1–213 a.a.: VHLlong), UMRC-6 (VHL54–213 a.a.: VHLshort) were from Dr Igor Kuzmin. These cell lines were maintained in 10% FCS/DMEM at 37°C in a humidified, 5% CO2 atmosphere. Clonal selection was done in 10% FCS-DMEM with 1 mg/ml G418 for 786-O stable transfectants, or 0.4 mg/ml hygromycin B for UMRC-6 stable transfectants.

Inducible stable clones of pVHL short form were established using the UMRC-6 cell line and an ecdysone expression system based on pIND and pVgRXR (Invitrogen). First, pVgRXR stable transfectants were selected with 1 mg/ml of Zeocin (Invitrogen). No leaking transfectants were selected with transient transfection of pIND-luciferase and following luciferase assay. Cells of the selected clone were then transfected with VHL short form expression vector containing a cis -acting element induced by RXR, and double stable transfectants, pIND-VHLshort and pVgRXR, were selected using 1 mg/ml of Zeocin and 1 mg/ml of G418. These stable clones were maintained in 10% FCS/DMEM with 1 mg/ml of Zeocin and 1 mg/ml of G418. Expression of pVHL was induced with 10 μM muristeroneA for 24 h. No leaking pVHL without muristeroneA were detectable after immuno-precipitation followed by Western blot analysis.

Adenovirus vector for human pVHL short was generated using a cosmid vector pAxCAwt (Miyake et al., 1996). For cell number count experiments 786-O cells were infected with Ax-lacZ or Ax-VHL short at MOI=10, a condition sufficient for nearly 100% infection of the cells.

Antibodies

The anti-human VHL monoclonal antibody (Ig32) was purchased from Pharmingen. The anti-actin polyclonal antibody (BT560) was from Biomedical Technologies. The anti p-27 and -cyclinA polyclonal antibodies were from Santa Cruz. The anti-elonginC antibody was from Transduction laboratory. The anti-T7 monoclonal antibody (Novagen) was used for control immuno-precipitation. The anti-elonginB polyclonal antibody was a gift from Dr Teijiro Aso. To generate rabbit polyclonal antibodies against Cul-2 and Rbx1, peptides corresponding to the C-terminal sequence of Cul-2 (CIDKQYIERSQASADEYSYVA) and Rbx1 (CKTRQVCPLDNREWEFQKYGH) were used as antigens. To generate rabbit polyclonal anti-VHL antibody, GST-human VHL (54–213 a.a.) was expressed in Escherichia coli, purified by SDS polyacrylamide gel electrophoresis, and used as an antigen. The antibodies were affinity purified with thioredoxin-fused VHL protein produced in E. coli, following the method described previously (Talian et al., 1983).

Immuno-precipitation

Cells were lysed in lysis buffer containing 20 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 10 μg/ml of leupeptine, 10 μg/ml of aprotinin, 1 mM PMSF, 1 mM vanadate, 50 mM NaF, 1.0% TritonX-100, 0.5% deoxycholate, and 0.1% SDS (RIPA) on ice for 30 min and vortexed. The protein concentration of the cell lysate was measured by Bradford's method, and lysates containing equal amounts, 60–80 μg, of protein were immuno-precipitated at 4°C for 3 h with 2 μg of anti-VHL monoclonal antibody Ig32 (Pharmingen) or anti-T7 monoclonal antibody (Novagen) pre-fixed on 10 μl protein G-Sepharose4 Fast Flow (Pharmacia Biotech). The Sepharose resin was washed five times with lysis buffer, and the immuno-precipitated proteins were eluted with SDS-sample buffer for Western blot analysis.

Western blot analysis

After SDS-polyacrylamide gel (15%) electrophoresis, the separated proteins were electrophoretically transferred to a polyvinylidene difluoride membrane. The blotted membrane was soaked in PBS containing 5% skim milk overnight at 4°C. The membrane was blocked with 5% calf serum in Tris-buffered saline containing 0.05% Tween-20 (TBST) for 1 h, and then incubated with appropriately diluted antibody in TBST containing 0.1% bovine serum albumin for 1 h at 37°C. After washing with TBST, the membrane was incubated with peroxidase-conjugated antibody against rabbit or mouse IgG antibody (Amersham) in TBST containing 5% skim milk. The membrane was washed again and the immuno-reactions were visualized with an ECL or ECL Plus chemiluminescence system (Amersham). The intensity of the luminescence was directly quantified using a CCD camera combined with image analysing system (LAS-1000, Fuji Film).

Cell number

Cells were cultured in 15 cm-dishes maintaining the contact naive state, and 1×105 or 8×104 cells were seeded in 6 cm-dishes and maintained in the appropriate culture medium without G418 or hygromycin. Half of the culture medium was changed every 24 h during cell growth in the log phase. After the cells reached confluency, the total medium was changed every 24 h. At each time point, three dishes were trypsinized and cell numbers were determined with hemocytometers. Statistical analysis was performed with Fisher's PLSD or Student's t-test.

Abbreviations

RCC:

renal cell carcinoma

SDS-PAGE:

sodium dodecyl sulfate polyacrylamide gel electrophoresis

VEGF:

vascular endothelial growth factor

VHL:

von Hippel-Lindau

References

  1. Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D, Grosschedl R, Birchmeier W . 1996 Nature 382: 638–642

  2. Blankenship C, Naglich JG, Whaley JM, Seizinger B, Kley N . 1999 Oncogene 18: 1529–1535

  3. Cockman ME, Masson N, Mole DR, Jaakkola P, Chang GW, Clifford SC, Maher ER, Pugh CW, Ratcliffe PJ, Maxwell PH . 2000 J. Biol. Chem. 275: 25733–25741

  4. Corless CL, Kibel AS, Iliopoulos O, Kaelin Jr WG . 1997 Hum. Pathol. 28: 459–464

  5. Feng H, Zhong W, Punkosdy G, Gu S, Zhou L, Seabolt EK, Kipreos ET . 1999 Nat. Cell. Biol. 1: 486–492

  6. Foster K, Prowse A, van den Berg A, Fleming S, Hulsbeek MM, Crossey PA, Richards FM, Cairns P, Affara NA, Ferguson-Smith MA, Buys CHCM, Maher E . 1994 Hum. Mol. Genet. 3: 2169–2173

  7. Gao J, Naglich JG, Laidlaw J, Whaley JM, Seizinger BR, Kley N . 1995 Cancer Res. 55: 743–747

  8. Gnarra JR, Tory K, Weng Y, Scmidt L, Wei MH, Li H, Latif F, Liu S, Chen F, Duh FM, Lubensky I, Duan DR, Florence C, Pozzati R, Walther, MM, Bander NH, Grossman BH, Brauch H, Pomer S, Brooks D, Isaacs WB, Lerman ML, Zbar B, Linehan WM . 1994 Nat. Genet. 7: 85–90

  9. Hart M, Concordet JP, Lassot I, Albert I, del los Santos R, Durand H, Perret C, Rubinfeld B, Margottin F, Benarous R, Polakis P . 1999 Curr. Biol. 9: 207–210

  10. He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT, Morin PJ, Vogelstein B, Kinzler KW . 1998 Science 281: 1509–1512

  11. Hinck L, Nelson WJ, Papkoff J . 1994 J. Cell. Biol. 124: 729–741

  12. Iliopoulos O, Kibel A, Gray S, Kaelin Jr WG . 1995 Nat. Med. 1: 822–826

  13. Iliopoulos O. Ohh M, Kaelin Jr WG . 1998 Proc. Natl. Acad. Sci. USA 95: 11661–11666

  14. Iwai K, Yamanaka K, Kamura T, Minato N, Conaway RC, Conaway JW, Klausner RD, Pause A . 1999 Proc. Natl. Acad. Sci. USA 96: 12436–12441

  15. Judah F . 1995 Nat. Med. 1: 27–31

  16. Kamura T, Koepp DM, Conrad MN, Skowyra D, Moreland RJ, Iliopoulos O, Lane WS, Kaelin Jr WG. Elledge SJ, Conaway RC, Harper JW, Conaway JW . 1999 Science 284: 657–661

  17. Kandikonda S, Oda D, Niederman R, Sorkin BC . 1996 Cell Adhes. Commun. 4: 13–24

  18. Kibel A, Iliopoulos O, DeCaprio JA, Kaelin Jr WG . 1995 Science 269: 1444–1446

  19. Kim M, Katayose Y, Li Q, Rakkar AN, Li Z, Hwang SG, Katayose D, Trepel J, Cowan KH, Seth P . 1998 Biochem. Biophys. Res. Commun. 253: 672–677

  20. Kitagawa M, Hatakeyama S, Shirane M, Matsumoto M, Ishida N, Hattori K, Nakamichi I, Kikuchi A, Nakayam K . 1999 EMBO J. 18: 2401–2410

  21. Korinek V, Barker N, Morin PJ, van Wichen D, de Weger R, Kinzler KW, Vogelstein B, Clevers H . 1997 Science 275: 1784–1787

  22. Lamiell JM, Salazar FG, Hsia YE . 1989 Medicine (Baltimore) 68: 1–29

  23. Latif F, Tory K, Gnarra J, Yao M, Duh FM, Orcutt ML, Stackhouse T, Kuzmin I, Modi W, Geil L, Schmidt L, Zhou F, Li H, Wei MH, Chen F, Glenn G, Choyke P, Walther MM, Weng YR, Dean M, Glavac D, Richards FM, Crossey PA, Ferguson-Smith MA, Pasilier LD, Chumakov I, Cohen D, Chinault AC, Maher ER, Linehan WM, Zbar B, Lerman MI . 1993 Science 260: 1317–1320

  24. Lee S, Chen DY, Humphrey JS, Gnarra JR, Linehan WM, Klausner RD . 1996 Proc. Natl. Acad. Sci. USA. 93: 1770–1775

  25. Lehner CF, O'Farrell PH . 1989 Cell 56: 957–968

  26. Liakopoulos D, Busgen T, Brychzy A, Jentsch S, Pause A . 1999 Proc. Natl. Acad. Sci. USA 96: 5510–5515

  27. Lieubeau-Teillet B, Rak J, Jothy S, Iliopoulos O, Kaelin Jr WG, Kerbel RS . 1998 Cancer Res. 58: 4957–4962

  28. Lisztwan J, Imbert G, Wirbelauer C, Gstaiger M, Krek W . 1999 Genes Dev. 13: 1822–1833

  29. Lonergan KM, Iliopoulos O, Ohh M, Kamura T, Conaway RC, Conaway JW, Kaelin Jr WG . 1998 Mol. Cell. Biol. 18: 732–741

  30. Los M, Jansen GH, Kaelin WG, Lips CJ, Blijham GH, Voest EE . 1996 Lab. Invest. 75: 231–238

  31. Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME, Wykoff CC, Pugh CW, Maher ER, Ratcliffe PJ . 1999 Nature 399: 271–275

  32. Melmon KL, Rosen SW . 1964 Am. J. Med. 36: 595–607

  33. Miyake S, Makimura M, Kanegae Y, Harada S, Sato Y, Takamori K, Tokuda C, Saito I . 1996 Proc. Natl. Acad. Sci. USA 93: 1320–1324

  34. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B, Kinzler KW . 1997 Science 275: 1787–1790

  35. Ohh M, Park CW, Ivan M, Hoffman MA, Kim T-Y, Huang LE, Pavletich N, Chau V, Kaelin Jr WG . 2000 Nature Cell Biol. 2: 423–427

  36. Okuda H, Hirai S, Takaki Y, Kamada M, Baba M, Sakai N, Kishida T, Kaneko S, Yao M, Ohno S, Shuin T . 1999 Biochem. Biophys. Res. Commun. 263: 491–497

  37. Orford K, Crockett C, Jensen JP, Weissman AM, Byers SW . 1997 J. Biol. Chem. 272: 24735–24738

  38. Orford K, Orford CC, Byers SW . 1999 J. Cell Biol. 146: 855–868

  39. Pal S, Claffey KP, Dvorak HF, Mukhopadhyay D . 1997 J. Biol. Chem. 272: 27509–27512

  40. Pause A, Lee S, Lonergan KM, Klausner RD . 1998 Proc. Natl. Acad. Sci. USA 95: 993–998

  41. Pause A, Lee S, Worrell RA, Chen DY, Burgess WH, Linehan WM, Klausner RD . 1997 Proc. Natl. Acad. Sci. USA 94: 2156–2161

  42. Pines J, Hunter T . 1990 Nature 346: 760–763

  43. Poon RY, Toyoshima H, Hunter T . 1995 Mol. Biol. Cell 6: 1197–1213

  44. Schoenfeld A, Davidowitz EJ, Burk RD . 1998 Proc. Natl. Acad. Sci. USA 95: 8817–8822

  45. Shuin T, Kondo K, Torigoe S, Kishida T, Kubota Y, Hosaka M, Nagashima Y, Kitamura H, Latif F, Zbar B, Lerman MI, Yao M . 1994 Cancer Res. 54: 2852–2855

  46. Soler C, Grangeasse C, Baggetto LG, Damour O . 1999 FEBS Lett. 442: 178–182

  47. St Croix B, Sheehan C, Rak JW, Florenes VA, Slingerland JM, Kerbel RS . 1998 J. Cell. Biol. 142: 557–571

  48. Talian JC, Olmsted JB, Goldman RD . 1983 J. Cell. Biol. 97: 1277–1282

  49. Wada H, Yeh ET, Kamitani T . 1999 J. Biol. Chem. 274: 36025–36029

  50. Watabe M, Nagafuchi A, Tsukita S, Takeichi M . 1994 J. Cell. Biol. 127: 247–256

  51. Ye Y, Vasavada S, Kuzmin I, Stackhouse T, Zbar B, Williams BR . 1998 Int. J. Cancer 78: 62–69

Download references

Acknowledgements

We thank Dr William G Kaelin for 786-O cells and their VHL expressing subclones, and Dr Igor Kuzmin and Berton Zbar for UMRC-6 cells and their VHL expressing subclones, We thank Dr Teijiro Aso for anti elonginB antibody. We also thank Dr Kazuhiro Iwai for helpful discussion during the course of the work. This work was supported in part by Grants in Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan and Grants from the Japan Society for the Promotion of Science 12219215 (to S Ohno) and 11160216 (to S Hirai). M Baba received support from the Yokohama Foundation for Advancement of Medical Science.

Author information

Affiliations

Authors

Corresponding author

Correspondence to Shigeo Ohno.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Baba, M., Hirai, Si., Kawakami, S. et al. Tumor suppressor protein VHL is induced at high cell density and mediates contact inhibition of cell growth. Oncogene 20, 2727–2736 (2001). https://doi.org/10.1038/sj.onc.1204397

Download citation

Keywords

  • VHL
  • tumor suppressor
  • contact inhibition
  • renal cell carcinoma
  • epithelial cells

Further reading

Search

Quick links