Abstract
Reptin is overexpressed in most human hepatocellular carcinomas. Reptin is involved in chromatin remodeling, transcription regulation, or supramolecular complexes assembly. Its silencing leads to growth arrest and apoptosis in cultured hepatocellular carcinoma cells and stops hepatocellular carcinoma progression in xenografts. Reptin has an ATPase activity linked to Walker A and B domains. It is unclear whether every Reptin function depends on its ATPase activity. Here, we expressed Walker B ATPase-dead mutants (D299N or E300G) in hepatocellular carcinoma cells in the presence of endogenous Reptin. Then, we silenced endogenous Reptin and substituted it with siRNA-resistant wild-type (WT) or Flag-Reptin mutants. There was a significant decrease in cell growth when expressing either mutant in the presence of endogenous Reptin, revealing a dominant negative effect of the ATPase dead mutants on hepatocellular carcinoma cell growth. Substitution of endogenous Reptin by WT Flag-Reptin rescued cell growth of HuH7. On the other hand, substitution by Flag-Reptin D299N or E300G led to cell growth arrest. Similar results were seen with Hep3B cells. Reptin silencing in HuH7 cells led to an increased apoptotic cell death, which was prevented by WT Flag-Reptin but not by the D299N mutant. These data show that Reptin functions relevant for cancer are dependent on its ATPase activity, and suggest that antagonists of Reptin ATPase activity may be useful as anticancer agents. Mol Cancer Res; 11(2); 133–9. ©2012 AACR.
Introduction
Reptin (RUVBL2, TIP48) is a member of the AAA+ (ATPases Associated with various cellular Activities) family (1). Reptin and its homolog and partner protein Pontin (RUVBL1, TIP49a) are essential for many cellular functions. They are involved in the remodeling of chromatin, the regulation of transcription, and in DNA-damage sensing and repair (2, 3). They function as chaperones and are required for the assembly of ribonucleoprotein complexes such as telomerase (4), snoRNPs (5), or, as part of the R2TP complex, for the assembly and/or stabilization of the RNA polymerase II complex (6) and of all members of the PIKK family including ATM, ATR, DNA-PKcs, SMG-1, TRRAP, and mTOR (7).
We have previously shown an overexpression of Reptin in hepatocellular carcinoma (8) and that Reptin silencing reduced tumor cell growth and viability in vitro (8, 9), a finding confirmed by others in a number of cell lines of different origins (10–12). Remarkably, Reptin silencing in vivo led to regression of hepatocellular carcinoma tumor xenografts, in association with the induction of tumor cell senescence (9). These results, which are in agreement with the reported functions of Reptin, suggest that it could be an interesting target for cancer therapy. We have recently reported, as a proof of principle, that it is possible to inhibit the ATPase activity of the related protein Pontin with small molecules (13). The same applies to Reptin (P Lestienne; unpublished data) and it is tempting to speculate that drugs inhibiting Reptin ATPase activity could have antitumor effects.
The ATPase activity of AAA+ proteins relies on the presence of the conserved Walker A and B domains, responsible for ATP binding and hydrolysis, respectively. Several studies have used mutants of these domains, designed according to the homology of Reptin and Pontin with the bacterial ATPase and DNA helicase RuvB (14), to probe the role of their ATPase activity. Studies in mammalian cells are scarce and have been mainly conducted for Pontin. Overexpression of the Walker B mutant Pontin D302N in the presence of endogenous Pontin, thus, inhibited cell transformation by several oncogenes such as E1A (15), c-myc (16), or β-catenin (17). With regard to Reptin, the Walker B mutant D299N was not able to complement the effect of the loss of endogenous Reptin on the amount of PIKK proteins in HEK293 cells (7), suggesting that the ATPase activity of Reptin is indeed required for the regulation of PIKK levels. However, the same mutant was as efficient as wild-type (WT) Reptin for the repression of the influenza A virus polymerase (18) or of the transcriptional activity of ATF2 (19). In the latter case, the C-terminus part of Reptin lacking both Walker domains was almost as efficient as the complete protein. Taken together, these data suggest that some functions of Reptin are independent of its ATPase activity, and may rather depend on protein–protein interactions. Particularly, it remains to be shown that the effects of Reptin underlying its role in tumor cell growth and viability are dependent on its ATPase activity.
Thus, in this study, we have used Reptin Walker B mutants to systematically probe the role of the ATPase activity on phenotypic properties of hepatocellular carcinoma cells relevant for carcinogenesis.
Materials and Methods
Cell culture and siRNA transfection
The human hepatocellular carcinoma cell lines HuH7 and Hep3B were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum in a 5% CO2 atmosphere at 37°C. Cells were authenticated using short tandem repeat analysis and tested for mycoplasma contamination on a regular basis.
The siR2 siRNA-targeting Reptin mRNA and the control siGL2 siRNA were described previously (8). They were bought from Eurofins MWG Operon (Ebersberg, Germany) and transfected by reverse transfection using Lipofectamine RNAiMAX (Invitrogen) at a final concentration of 16 nmol/L.
Constructs and lentiviral transduction
The D299N mutant Flag-Reptin cDNA was a gift from M Cole (Lebanon, USA). The E300G mutant was constructed with the QuikChange site-directed mutagenesis kit from Agilent Technologies (Massy, France) and verified by sequencing. Mutants and WT Flag-Reptin were inserted in the previously described lentiviral vector pRRLSIN-MND-MCS-WPRE (8). Titers were determined by the transduction of HEK 293T cells through the serial dilution of the lentiviral supernatant and were analyzed for p24 protein expression 5 days later.
Cell proliferation
Cells were seeded at a density of 50,000 per well in 12-well plates. Adherent cells were counted at various times with a Coulter counter (Beckman Coulter) in triplicate wells.
Propidium iodide staining and flow cytometry
For cell-cycle analysis, cells were collected by trypsinization and fixed in 70% ethanol at 4°C for a minimum of 30 minutes. After fixation, cells were washed with PBS, and the DNA was stained with propidium iodide (PI) at a final concentration of 0.05 mg/mL in 0.1% NaCitrate-0.2% Triton X-100. The samples were analyzed by flow cytometry using a BD LSRFortessa flow cytometer (BD Biosciences).
Caspase 3 activity assay
Cells were transduced as described earlier. After 6 days, 2 × 106 cells were solubilized in 100 μL of buffer and assayed with the caspase-3 colorimetric activity assay from Chemicon international (Millipore) according to the manufacturer's instructions.
Western blot and immunoprecipitation
Western blot was done as described previously (20). All blots were analyzed with the Odyssey system (Li-Cor Biosciences). We used anti-Reptin mouse monoclonal (BD Biosciences, Pharmingen), anti-Pontin rabbit polyclonal (ProteinTech), anti-FLAG-M2, and anti-β-actin mouse monoclonal (Sigma-Aldrich) antibodies.
For immunoprecipitation, cells were washed in PBS. Lysis buffer (20 mmol/L Tris-HCl pH 7.5, 1% Triton X-100, 1 mmol/L NaVO3, and 100 mmol/L NaF) supplemented with protease inhibitor cocktail (Roche) was added to the pellet. After 20 minutes of incubation in ice, the pellet was vortexed and centrifuged at 15,000g for 5 minutes. Cell extracts were incubated overnight at 4°C with monoclonal anti-Flag M2 beads (Sigma-Aldrich). The beads were washed 5 times with lysis buffer and eluted with Laemmli sample buffer. Eluates were separated on a 10% SDS-PAGE for Western blotting.
Senescence assay
Senescence-Associated Heterochromatin Foci (SAHF; ref. 21) were visualized with DAPI staining (0.5 μg/mL; Sigma) and quantified by counting using a Zeiss Axioplan microscope. Representative pictures were obtained with a Leica SP5 scanning laser microscope (Leica Microsystems GmbH).
Results
Expression of a Reptin ATPase-dead mutant impairs hepatocellular carcinoma cell growth
Following the observation that the D113N mutation in bacterial RuvB impairs ATP hydrolysis, it was shown that the homologous D299N point mutation in the Walker B domain of human Reptin led to the loss of the ATPase activity of Reptin–Pontin complexes (22). We, thus, used this point mutant as a tool to investigate the role of Reptin ATPase activity in hepatocellular carcinoma cells. We first made sure that this mutation did not affect key properties of Reptin, such as its subcellular localization and its ability to interact with Pontin. Nuclear and cytoplasmic repartition of WT Flag-Reptin or Flag-tagged mutants was examined by Western blotting after cell fractionation. As observed for WT Flag-Reptin, the D299N and the E300G mutants were both expressed in the nucleus as well as in the cytoplasm (Supplementary Fig. S1). In addition, we examined the capacity of the D299N mutant to interact with Pontin, which is the major protein partner of Reptin. As shown on Fig. 1A, Pontin was co-immunoprecipitated with both Flag-Reptin and Flag-Reptin D299N. Furthermore, Flag-Reptin D299N was also able to interact with endogenous Reptin as observed for WT Flag-Reptin.
Expression of Flag-Reptin D299N and its effect on cell growth in HuH7 cells. A, Reptin D299N mutant interacts with Pontin and endogenous Reptin. HuH7 were left either nontransduced (NT, Ctl), or were transduced with Flag-Reptin or Flag-Reptin D299N lentiviral vectors. Twenty days after transduction, whole-cell extracts were used for anti-Flag immunoprecipitation. Reptin and Pontin were detected by Western blot. The lower amount of D299N Reptin as compared with wild-type (WT) Reptin in the input is because of a counterselection of cells at the time of the experiments. B, HuH7 cells were transduced with lentiviral vectors coding Flag-Reptin or Flag-Reptin D299N, or left NT. Six days later, they were plated at an initial density of 50,000/well. Expression of WT or D299N Flag-Reptin was controlled by Western blot on whole-cell extracts 9 days after lentiviral transduction. C, quantification of endogenous Reptin and Flag-Reptin expression. The graph shows the mean ± SD of 3 independent experiments. Total Reptin amount was not significantly changed in any condition. Note that Flag-Reptin D299N systematically migrates slightly slower than WT Flag-Reptin. D, adherent cells numbers were monitored at the indicated time points with a Coulter counter (n = 3; ***, P < 0.001 for day 4 and day 7 by 2-way ANOVA followed by Bonferroni test). Note that SDs are not always visible because of their small size.
Taken together, these results show that the D299N mutation does not alter Reptin localization or its interaction with Pontin. These experiments also showed that the D299N mutant was counter-selected over time (in Fig. 1A, compare Input of WT Flag-Reptin and D299N Flag-Reptin). For all subsequent experiments, we thus carefully selected a time window where mutant and WT Reptin were expressed at similar levels.
We previously showed that silencing of endogenous Reptin through RNA interference induced a decrease in growth of hepatocellular carcinoma cells (8, 9). With the aim of determining the role of Reptin ATPase activity in hepatocellular carcinoma cell growth, we first expressed the ATPase dead mutant D299N in the presence of endogenous Reptin, using lentiviral vectors. In agreement with our previous results showing a tight posttranslational control of Reptin expression (20), this strategy did not allow for an overexpression of Flag-tagged Reptin, either WT or mutated. Indeed, the total amount of Reptin was similar between the 3 conditions, with Flag-tagged Reptin representing approximately 50% of the total amount in transduced cells (Fig. 1B and C). Cells expressing WT Flag-Reptin had a normal growth pattern as compared with nontransduced cells, emphasizing that the Flag tag did not have an impact on Reptin function. On the contrary, we observed a significant diminution of cell growth in HuH7 cells expressing the D299N Reptin mutant (Fig. 1D). This reveals a dominant negative effect of the ATPase dead mutant on hepatocellular carcinoma cell growth.
In order to determine the effects of Reptin ATPase dead mutant in the absence of endogenous Reptin, we used a substitution strategy as previously described (9). Briefly, we previously generated a Flag-tagged Reptin cDNA harboring silent mutations that made the mRNA insensitive to the anti-Reptin siR2 siRNA. By directed mutagenesis, we then inserted the D299N point mutation in this sequence. Hepatocytes carcinoma cells were first transduced with siR2-resistant Flag-Reptin (FRrR2) or Flag-Reptin D299N (FRD299NrR2) and then transfected with the siR2 siRNA. As shown on Fig. 2, this led to the quantitative replacement of endogenous Reptin by resistant WT or D299N mutant Flag-tagged Reptin, respectively (Fig. 2A and B). We previously showed that Reptin silencing led to a codepletion of Pontin through a posttranslational mechanism (20). Here, we show that substitution of endogenous Reptin with siRNA-resistant Flag-tagged Reptin preserved Pontin expression, whether Reptin was WT or D299N (Fig. 2A). When we carried out quantitative substitution of Reptin by WT Flag-Reptin, cell growth of HuH7 was normal (Fig 2C). On the other hand, quantitative substitution of Reptin by Flag-Reptin D299N led to growth arrest in HuH7 cells from 3 days after transfection by siR2 (Fig. 2C). Similar effects were seen in the Hep3B cell line (Supplementary Fig. S2).
Substitution of endogenous Reptin with D299N mutant leads to HuH7 cell growth arrest. Seven days after lentiviral transduction by siR2-resistant Flag-Reptin (FRrR2) or siR2-resistant Flag-Reptin D299N (FRD299NrR2), HuH7 cells were transfected with a control (siGL2) or an anti-Reptin siRNA (siR2). A, quantitative substitution of endogenous Reptin by Flag-Reptin or Flag-Reptin D299N was controlled by Western blot on whole-cell extracts 3 days after siRNA transfection. B, quantification of endogenous Reptin and Flag-Reptin expression. The graph shows the mean ± SD of 3 independent experiments. Total Reptin or Pontin amount was not significantly changed in any condition. C, adherent cells numbers were monitored at indicated time points with a Coulter counter. The graph shows the mean ± SD of 3 independent experiments (n = 3; ***, P < 0.001 for day 3 and day 6 by 2-way ANOVA followed by Bonferroni test). Note that SDs are not always visible because of their small size.
We then tested the effects of Reptin on cell-cycle progression using PI staining and flow cytometry. The most striking effect was a large accumulation of cells in G2/M phase on Reptin silencing, which was rescued by complementation with siRNA-resistant Flag-tagged Reptin, but not with the D299N mutant (Supplementary Fig. S3).
In order to rule out that our results were linked to the specific mutation used, we checked the effects of another point mutation in the Walker B domain of human Reptin. The mutation was based on the yeast Reptin E297G ATPase dead mutant previously used to show the requirement of the Walker B domain for yeast viability (23). As with the D299N mutant, we were able to achieve a quantitative substitution of endogenous Reptin, and to preserve Pontin levels on silencing of endogenous Reptin (Fig. 3A and B). As shown on Fig. 3C, the E300G Reptin mutant imposed a dominant negative effect on cell growth in the presence of endogenous Reptin. Furthermore, and as observed for the D299N mutant, quantitative substitution of endogenous Reptin by Flag-Reptin E300G led to cell growth arrest in HuH7.
Effect of Reptin E300G mutant expression and endogenous Reptin replacement on HuH7 cell growth. Seven days after lentiviral transduction by siR2-resistant Flag-Reptin (FRrR2) or siR2-resistant Flag-Reptin E300G (FRE300GrR2), HuH7 were transfected with a control (siGL2) or anti-Reptin siRNA (siR2). A, expression of Flag-Reptin (WT or E300G) and quantitative substitution of Reptin was controlled by Western blot on whole-cell extracts 3 days after transfection. B, the graph shows the mean ± SD of 3 independent experiments. Total Reptin or Pontin amount was not significantly changed in any condition except when endogenous Reptin was silenced without being replaced. C, adherent cells numbers were monitored at indicated time points with a Coulter counter (n = 3; ***, P < 0.0001 by 2-way ANOVA followed by Bonferroni test).
Role of Reptin ATPase activity on hepatocellular carcinoma cells viability and senescence
We previously showed that Reptin silencing resulted in increased apoptotic cell death (8). We thus examined whether an intact Reptin ATPase activity was required to prevent cells from apoptotic death. As a marker of apoptotic death, we quantified cells with less than 2n DNA using PI staining and flow cytometry. The increased cell death consequent to Reptin silencing was completely prevented in HuH7 cells expressing WT Flag-Reptin resistant to the siRNA. Interestingly, the simple expression of Flag-Reptin D299N in the presence of endogenous Reptin led to a 2-fold increase in cell death as compared with nontransduced HuH7 (Fig. 4). Moreover, quantitative substitution of Reptin by the D299N mutant dramatically increased cell death. Similar results were obtained when cell death was assayed through the measurement of caspase 3 activity (Supplementary Fig. S4).
Measurement of cell death by PI staining. HuH7 NT or transduced by siR2-resistant Flag-Reptin (FRrR2) or Flag-Reptin D299N (FRD299NrR2) were transfected with control (siGL2) or anti-Reptin (siR2) siRNA. Five days after siRNA transfection, cells were stained with PI and the percentage of subG1 cells was quantified with flow cytometry. The top panels show graphs from a representative experiment and the graph shows the mean of 3 experiments (***, P < 0.0001 by 1-way ANOVA followed by Bonferroni test).
We have reported the induction of senescence following Reptin silencing in hepatocellular carcinoma cells (9). In order to test the requirement of Reptin ATPase activity for the prevention of senescence, we quantified senescence-associated heterochromatin foci (SAHF) as a marker (21). As expected, Reptin silencing induced a large increase in SAHF formation that was rescued when expressing the siRNA-resistant Reptin. However, silencing of endogenous Reptin in cells expressing 1 or the other mutant did not induce a significant increase in SAHF. Similar results were obtained at either 5 (Fig. 5) or 6 days after transduction (Supplementary Fig. S5).
Measurement of senescence by counting senescence-associated heterochromatin foci (SAHF). HuH7 NT or transduced by siR2-resistant Flag-Reptin (FRrR2) or Flag-Reptin D299N (FRD299NrR2) were transfected with a control (siGL2) or an anti-Reptin (siR2) siRNA. Five days after siRNA transfection, cells were fixed and nuclei were stained with DAPI. The top panel shows representative pictures for each condition obtained using confocal microscopy. The percentage of cells with SAHF was quantified by counting (n = 4; ***, P < 0.0001 by 1-way ANOVA followed by Bonferroni test).
Discussion
Previous studies have suggested that Reptin was an attractive target for cancer therapy. In this respect, antagonizing its ATPase activity appears as an attractive option and could be feasible because we have been recently able to discover a small molecule antagonizing the ATPase activity of the Reptin homolog, Pontin (13). However, because it is uncertain that every Reptin function is dependent on its ATPase activity, we investigated in depth this key issue.
Here, we show that expression of 2 different Walker B mutants in the presence of endogenous Reptin, or substitution of endogenous Reptin by those mutants, had effects reminiscent of those evoked by Reptin silencing with respect to cell growth and viability. Identical effects were observed in 2 different hepatocellular carcinoma cell lines. Reptin mutants retained the same subcellular distribution as WT Reptin, and also their ability to partner with Pontin, suggesting that they had no gross defects besides the loss of ATPase activity that could explain their phenotypic effects. We and others have previously shown that Reptin silencing induced the codepletion of Pontin, via a posttranslational mechanism (4, 7, 20). However, we show here that cells reconstituted with the Reptin Walker B mutants expressed normal levels of Pontin such as those reconstituted with WT siRNA-resistant Reptin (20), thus ruling out that some of the effects seen with the mutants might be due to the loss of Pontin. Although the issue is not completely settled, the most recent evidence suggests that human Reptin and Pontin are organized in heterohexamers (24–26). It is thus likely that, when expressed in the presence of endogenous Reptin and Pontin, Walker B mutants will replace WT Reptin within hexamers and can thus exert a dominant negative effect. This was shown in the case of bacterial RuvB where the D113N mutant (homologous to the D299N Reptin mutant) associated in heterohexamers with WT RuvB (14). This resulted in a dose-dependent loss of ATPase activity of the hexamers (14). Similarly, Puri and colleagues showed that the ATPase activity of Reptin–Pontin complexes was lost when a Reptin Walker B mutant was associated with WT Pontin (22). In our experiments, immunoprecipitation of the mutants using the Flag epitope pulled down endogenous Pontin and Reptin together, a finding compatible with the incorporation of mutant Reptin in mixed heterohexamers.
Unexpectedly, cells expressing the D299N or the E300G mutant were apparently protected against senescence induced by silencing of endogenous Reptin almost as efficiently as cells expressing WT siRNA-resistant Reptin. We propose that the mutants exhibit per se a toxic activity that drives cells toward death, preventing them from entering senescence. This is supported by our observations showing that mutant-expressing cells exhibit a larger reduction in growth (see Figs. 2C and 3C) and a higher rate of death (Fig. 4) as compared with cells only silenced for Reptin. Thus, although those cells exhibit somehow less senescence, the net effect of expressing the mutants is still reduced cell growth and increased cell death.
Taken together, our data show that the ATPase activity of Reptin is required for its effects on the growth and viability of hepatocellular carcinoma cells. Several known functions of Reptin may explain why its silencing would reduce cell growth and increase cell death. Indeed, together with Pontin, Reptin is involved in the assembly of telomerase and their silencing leads to a decreased telomerase activity (4, 9). It also controls the levels of all members of the phosphatidylinositol 3-kinase–related protein kinases family, notably mTOR, with Reptin silencing leading to decreased mTOR level and a consequent defect in signaling through S6 kinase (7).
Our data also lead us to conclude accordingly that antagonists of Reptin ATPase activity may be of benefit for the therapy of hepatocellular carcinoma. Besides hepatocellular carcinoma, an overexpression of Reptin has been shown in several types of tumors (2, 11, 12, 27), and Reptin silencing also affects growth and viability of a variety of tumor cells (10–12). We thus propose that targeting Reptin ATPase activity may have broad applications in oncology.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Authors' Contributions
Conception and design: A. Grigoletto, J. Rosenbaum
Development of methodology: A. Grigoletto
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): V. Neaud, N. Allain-Courtois, A. Grigoletto
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Grigoletto, P. Lestienne, J. Rosenbaum
Writing, review, and/or revision of the manuscript: A. Grigoletto, J. Rosenbaum
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N. Allain-Courtois
Study supervision: J. Rosenbaum
Grant Support
This work was funded by grants from the Equipe Labélisée Ligue Contre le Cancer 2011, Institut National du Cancer (PLBIO10-155) and Association pour la Recherche sur le Cancer (#1126). A. Grigoletto was the recipient of fellowships from the French Ministry of Research and Fondation pour la Recherche Médicale.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Acknowledgments
The authors thank Michael Cole for plasmids, Eric Chevet for many helpful discussions, Frédéric Saltel for help with confocal microscopy, and the Vectorology platform of University Bordeaux Segalen for lentiviral constructs.
Footnotes
Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/).
- Received July 31, 2012.
- Revision received November 5, 2012.
- Accepted November 23, 2012.
- ©2012 American Association for Cancer Research.