Skip to main content
NCBI home page
EMBO Reports logoLink to EMBO Reports
. 2018 Sep 13;19(11):e45587. doi: 10.15252/embr.201745587

SIRT2‐mediated inactivation of p73 is required for glioblastoma tumorigenicity

Kosuke Funato 1,2,†, Tomoatsu Hayashi 1,†, Kanae Echizen 1, Lumi Negishi 1, Naomi Shimizu 1, Ryo Koyama‐Nasu 1, Yukiko Nasu‐Nishimura 1, Yasuyuki Morishita 3, Viviane Tabar 2, Tomoki Todo 4, Yasushi Ino 4, Akitake Mukasa 4, Nobuhito Saito 4, Tetsu Akiyama 1,✉
PMCID: PMC6216266  PMID: 30213795

Abstract

Glioblastoma is one of the most aggressive forms of cancers and has a poor prognosis. Genomewide analyses have revealed that a set of core signaling pathways, the p53, RB, and RTK pathways, are commonly deregulated in glioblastomas. However, the molecular mechanisms underlying the tumorigenicity of glioblastoma are not fully understood. Here, we show that the lysine deacetylase SIRT2 is required for the proliferation and tumorigenicity of glioblastoma cells, including glioblastoma stem cells. Furthermore, we demonstrate that SIRT2 regulates p73 transcriptional activity by deacetylation of its C‐terminal lysine residues. Our results suggest that SIRT2‐mediated inactivation of p73 is critical for the proliferation and tumorigenicity of glioblastoma cells and that SIRT2 may be a promising molecular target for the therapy of glioblastoma.

Keywords: cancer stem cells, glioblastoma, p73, SIRT2

Subject Categories: Cancer; Post-translational Modifications, Proteolysis & Proteomics; Transcription

Introduction

Glioblastoma is the most malignant form of glioma. Despite great efforts, the median survival of glioblastoma patients has remained at around 1 year for the past decade 1. Several hallmark features of glioblastoma contribute to its aggressive phenotypes, such as uncontrolled proliferation, diffuse infiltration into surrounding brain tissue, and poor response to therapeutic agents. Furthermore, it has been reported that glioblastoma stem cells (GSCs), subsets of glioblastoma cells that possess the capability of self‐renewal and exhibit extensive tumorigenicity, are resistant to both chemotherapy and radiotherapy and thus are responsible for the poor prognosis of glioblastoma 2, 3.

Integrative genomic analyses have revealed that the p53, RB, and RTK pathways, termed core signaling pathways, are commonly deregulated in glioblastoma 2, 4, 5. Genetic alterations in these pathways enable glioblastoma cells to escape from cell‐cycle checkpoints, apoptosis, and senescence, resulting in uncontrolled proliferation and enhanced survival. It has also been reported that mutations and deletions in neurofibromatosis type 1 (NF1) gene occur in 23% of glioblastoma 5. Furthermore, many glioblastomas have a heterozygous deletion of the NF‐κB inhibitor α (NFKB1A) gene, which encodes IκB, a negative regulator in the NF‐κB pathway 6, and/or mutations in the isocitrate dehydrogenase 1 (IDH1) gene 4.

Sirtuins [class III (NAD‐dependent) histone deacetylases (HDACs)] are conserved from bacteria to human and have been implicated in aging and longevity, resulting from their regulation of genomic stability and metabolism 7. There are seven sirtuin members in mammals, sirtuin 1–7 (SIRT1–7), and these exhibit diverse functions, including transcriptional silencing, metabolic regulation, and apoptosis. SIRT2 is localized to both the nucleus and the cytoplasm 8, 9, 10, 11 and deacetylates α‐tubulin, FOXO1, FOXO3a, p300, Lys16 of histone H4, Slug, CDH1, CDC20, and p53 8, 9, 10, 11, 12, 13. SIRT2 also deacetylates the transcriptional repressor Slug to prevent its degradation and thereby controls malignancy of basal‐like breast cancer 11. Furthermore, a SIRT2‐specific inhibitor, a thiomyristoyl lysine compound, promotes c‐Myc ubiquitination and degradation and exhibits broad anticancer activity 14. It has also been reported that SIRT2 regulates cell‐cycle progression and genome stability by modulating the mitotic deposition of H4K20 methylation 15. In addition, it has been reported that SIRT inhibitors induce cell death and p53 acetylation through targeting both SIRT1 and SIRT2 13.

In this study, we show that SIRT2 is required for the proliferation and tumorigenicity of glioblastoma cell. We further demonstrate that the suppression of glioblastoma proliferation by knockdown of SIRT2 requires p73. We find that SIRT2 regulates the transcriptional activity of p73 by deacetylation of its C‐terminal lysine residues. Our results suggest that SIRT2 regulates the survival of glioblastoma via repression of p73 function.

Results and Discussion

Knockdown of SIRT2 induces the growth arrest and apoptosis of glioblastoma cells

Using cells isolated from human glioblastoma, we have established a series of glioblastoma neurospheres (GB1–GB13 and GB16). We identified loss‐of‐function mutations in the DNA‐binding domain of p53 in two of these glioblastoma neurospheres, GB2 and GB16, which result in Ser241 and His193 being replaced with Phe and Arg, respectively (Appendix Table S1). In addition, we identified EGFR amplification in GB13. Moreover, we previously showed that GB2 possesses the capability of self‐renewal and exhibits extensive tumorigenicity 16. To identify novel therapeutic targets for glioblastoma cells, we performed an RNA interference (RNAi) screen using GB2, which is easy to culture and possesses high tumorigenic activity. GB2 cells were transduced with an siRNA library targeting 246 genes commonly expressed in glioblastoma neurospheres (Appendix Table S2) and then assayed for CD133 expression by quantitative RT–PCR (qRT–PCR) (Appendix Fig S1). CD133 has been successfully used as a stem cell marker for some glioblastomas 3, 17, 18, and it was previously shown that CD133 can be used as a stem cell marker for the glioblastoma spheres which were derived from the same cell specimen as GB2 19. Candidate genes that modulated CD133 expression more than twofold (Appendix Fig S1 and Appendix Table S2) were further validated for their effects on CD133 and/or nestin expression.

From this screen, we identified SIRT2 as a candidate modulator of these properties of GB cells (Figs 1A and EV1A, Appendix Fig S1, Appendix Tables S2 and S3). In these experiments, knockdown of SIRT2 led to an increase in the acetylation of α‐tubulin, a known substrate of SIRT2 9, indicating that SIRT2 was functionally suppressed in these cells (Fig EV1B). We also found that knockdown of SIRT2 resulted in significant inhibition of sphere formation in other primary glioblastoma neurospheres (GB4, GB11, GB13, and GB16) and glioblastoma cells isolated freshly from tumor samples (GB15) (Fig EV1A and B). Furthermore, limiting dilution assays confirmed that knockdown of SIRT2 caused inhibition of primary glioblastoma sphere formation (GB16) (Figs 1B and EV1C). In addition, we examined the effects of eight out of the top 10 candidate genes on the expression of Sox2, EZH2, and Olig2. We found that EHMT1, PTPRO, PTCH1, and TAL1 as well as SIRT2 suppressed the expression of Sox2, EZH2, and Olig2 (Appendix Table S3).

Figure 1. Knockdown of SIRT2 using siRNA or treatment with AGK2 induces growth arrest and apoptosis of glioblastoma cells.

Figure 1

Open in a new tab
  1. Sphere formation of GB2 cells transfected with an shRNA targeting SIRT2 was analyzed by an In Cell Analyzer 2000. Primary spheres were re‐plated to evaluate secondary sphere formation. Bars indicate mean ± SD of 10 wells.
  2. Knockdown of SIRT2 causes a decrease in the sphere formation capacity of GB16. The figure shows a representative result of three independent experiments.
  3. mRNA levels of the indicated genes in GB2, GB4, and GB16 cells infected with a lentivirus expressing an shRNA targeting SIRT2 were measured by qRT–PCR. The results were normalized with the values for GAPDH. Bars indicate mean ± SD (n = 3–4).
  4. The sphere formation capacity of CD133‐positive and CD133‐negative cells sorted by FACS directly from a tumor sample. GB17 was infected with a control (Empty) or shSIRT2‐expressing (shS2 #1) lentivirus. Bars indicate mean ± SD of eight wells.
  5. The sphere formation capacity of CD133‐positive and CD133‐negative cells sorted by FACS directly from a tumor sample. (Left panel) GB18 was treated with AGK2 (10 μM) or DMSO. (Right panel) Secondary sphere formation of GB18 was examined in the absence of AGK2. Bars indicate mean ± SD of eight wells.
Data information: Statistical significance was evaluated using the likelihood ratio test (for panel B) or unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.

Figure EV1. Knockdown of SIRT2 induces growth arrest in glioblastoma cells (related to Fig 1).

Figure EV1

Open in a new tab
  1. Sphere‐forming capacities of GB cells transfected with an shRNA targeting SIRT2. GB cells were plated on 96‐well plates at the indicated cell numbers. After 10 days of incubation, the spheres were analyzed by microscopy or using an In Cell Analyzer 2000. For GB15 and GB16, primary spheres were re‐plated to evaluate secondary sphere formation. Bars indicate mean ± SD of 10 wells.
  2. GB cells infected with a lentivirus expressing an shRNA targeting SIRT2 were subjected to immunoblotting analysis using the indicated antibodies.
  3. Knockdown of SIRT2 causes a decrease in the sphere formation capacity of GB16. Estimated stem cell frequencies were determined from the data shown in Fig 1B by extreme limiting dilution analysis (http://bioinf.wehi.edu.au/software/elda).
  4. GB cells transfected with a lentivirus expressing an shRNA targeting SIRT2 were plated on laminin‐coated 12‐well plates (1 × 105 cells per well). After 5 days of incubation, the number of viable cells was counted at the indicated time points by Trypan blue staining. Bars indicate mean ± SD (n = 3).
  5. GB2 cells were infected with a lentivirus expressing an shRNA targeting SIRT2 and plated in 96‐well plates. After 12 days of incubation under the non‐adherent condition, the number of viable cells was counted. Bars indicate mean ± SD (n = 3).
  6. GB2 cells cultured in serum‐free or serum‐containing medium for more than 10 passages were infected with a control or shSIRT2‐expressing lentivirus. After 11 days, cells were analyzed for sub‐G0 DNA content. Bars indicate mean ± SD (n = 3).
  7. Knockdown of SIRT2 causes an increase in the levels of PUMA expression and cleavage of caspase‐3 in GB2 cells cultured in the presence or absence of serum. Five days after viral infection, cells were subjected to qRT–PCR analysis for SIRT2 expression (left panel) and to immunoblotting analysis using the indicated antibodies (right panel). Bars indicate mean ± SD (n = 3).
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01; N.S., not significant.Source data are available online for this figure.

We also found that knockdown of SIRT2 significantly decreased the number of viable GB2, GB4, and GB16 cells (Fig EV1D and E). Furthermore, SIRT2 knockdown in GB2 cells resulted in an increase in the sub‐G0 DNA fraction and an increase in the cleavage of caspase‐3, indicating that SIRT2 knockdown induces apoptosis of GB2 cells (Fig EV1F and G). Consistent with these results, knockdown of SIRT2 resulted in an increased expression of multiple apoptotic and cell‐cycle‐regulating genes, such as PUMA, NOXA, and GADD45 in GB2, GB4, and GB16 cells (Figs 1C and EV1G).

It is known that glioblastoma cells lose many of their original properties when cultured in serum‐containing medium, including their stem cell‐like characteristics 19. In contrast to GB2 cells cultured in serum‐free medium, knockdown of SIRT2 in GB2 cells cultured in serum‐containing medium barely caused increases in the sub‐G0 DNA fraction (Fig EV1F) and the level of PUMA expression or cleavage of caspase‐3 (Fig EV1G). Of particular note, the expression of SIRT2 was significantly decreased in GB2 cells cultured in serum‐containing medium.

It has been reported that GSCs can be enriched using several cell‐surface markers, including CD133 3, 17, 18, 19, 20. To further investigate the role of SIRT2 in GSCs, we sorted CD133‐positive cells by FACS from GB17 and GB18 samples. Consistent with previous reports 19, CD133‐positive cells had higher sphere formation capacity than CD133‐negative cells (Fig 1D and E). We found that knockdown of SIRT2 resulted in the suppression of sphere formation of CD133‐positive cells (Fig 1D). These results suggest that SIRT2 is required for the growth of GSCs.

The SIRT2‐specific inhibitor AGK2 induces growth arrest and apoptosis of glioblastoma cells

AGK2 is a SIRT2‐specific inhibitor that holds promise as an agent for the treatment of Parkinson's and other neurodegenerative diseases 21, 22. When GB2, GB4, GB11, or GB16 cells were treated with AGK2, their sphere formation capacities were reduced compared to vehicle‐treated cells (Fig EV2A and B). Treatment with AGK2 caused an increase in α‐tubulin acetylation, indicating that SIRT2 deacetylase activity was functionally inhibited (Fig EV2C). In addition, the reduction in GB2 cell number caused by lentivirus expression of an shRNA against SIRT2 was not further reduced by AGK2 treatment, suggesting that the effect of AGK2 is specific to SIRT2 (Fig EV2D). Furthermore, AGK2 treatment suppressed sphere formation of CD133‐positive cells freshly isolated from a patient sample (GB18) (Figs 1E and EV2E).

Figure EV2. The SIRT2‐specific inhibitor AGK2 induces growth arrest of glioblastoma cells (related to Fig 1).

Figure EV2

Open in a new tab
  1. GB2 cells cultured in the presence of AGK2 at the indicated concentration or DMSO were subjected to sphere formation assays. Bars indicate mean ± SD of 10 wells.
  2. Treatment of GB glioblastoma neurospheres with SIRT2 inhibitors, but not with EX527, causes a decrease in the number of spheres. Cells were plated in 96‐well cell culture plates at the indicated cell numbers. After incubation with the indicated inhibitors for 8 days, the number of spheres was counted under a microscope. Bars indicate mean ± SD of at least 8 wells.
  3. GB2 cells were treated with 20 μM AGK2 or DMSO for 6 h and subjected to immunoblotting analysis using the indicated antibodies.
  4. AGK2 treatment is not additive with SIRT2 knockdown in suppressing GB2 cell proliferation. Cells infected with a control or shSIRT2‐expressing (shS2 #1) lentivirus were treated with AGK2 (10 μM) or DMSO for 24 h. The number of viable cells was counted by Trypan blue staining. Bars indicate mean ± SD of triplicate technical repeats (n = 2).
  5. CD133‐positive and CD133‐negative cells sorted by FACS directly from a tumor sample (GB18) were subjected to sphere formation assays in the presence of AGK2 (10 μM) or DMSO. Scale bar, 200 μm.
  6. Treatment of GB2 cells with AGK2 increases the expression of PUMA, BAX, NOXA, and GADD45, but decreases the expression of CD133. GB2 cells (1.5 × 105 cells) were treated with 20 μM AGK2 or DMSO for 72 h and subjected to qRT–PCR to determine the mRNA expression levels of the indicated genes. Bars indicate mean ± SD (n = 3–4).
  7. p53 mutant GB cells do not respond to treatment with doxorubicin or camptothecin. 2 × 104 cells were treated with 20 μM AGK2, 5 μM doxorubicin (DXR), 1 μM camptothecin (CPT), or DMSO (control), respectively. After 4 h of incubation, the mRNA expression of the indicated genes was measured by qRT–PCR. Bars indicate mean ± SD (n = 3).
  8. AGK2 treatment induces apoptosis of GB2 cells. GB2 cells (5 × 105 cells) were treated with 20 μM AGK2 or DMSO for the indicated times and subjected to sub‐G0 assays (n = 3).
  9. Treatment of GB2 cells with AGK2 increases the amount of cleaved caspase‐3. GB2 cells (1.5 × 105 cells) were treated with 20 μM AGK2 or DMSO for the indicated times and subjected to immunoblotting with anti‐caspase antibody.
  10. GB2 cells were treated with AGK2 as indicated, and caspase‐3/7 activation was measured using the caspase‐3/7 Glo assay. Bars indicate mean ± SD (n = 4).
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.Source data are available online for this figure.

Consistent with the results of SIRT2 knockdown experiments, AGK2 treatment increased the expression of the p53‐inducible genes, PUMA, NOXA, and GADD45, in GB2 cells (Fig EV2F). By contrast, treatment with camptothecin or doxorubicin failed to induce p53 target genes (Fig EV2G). Furthermore, we detected an increase in the sub‐G0 DNA fraction and increased cleavage of caspase‐3 in AGK2‐treated GB2 cells (Fig EV2H and I). We also measured caspase‐3/7 activation using the caspase‐3/7 Glo assay and confirmed these results (Fig EV2J). Thus, AGK2 treatment induces apoptosis of GB2 cells. Taken together, these findings suggest that the deacetylase activity of SIRT2 is required for the proliferation and survival of glioblastoma cells.

Among the family of sirtuins, SIRT1 is known to be involved in the development of cancer by blocking senescence and apoptosis 7. It has also been reported that SIRT1 inhibition enhances radiosensitivity of CD133‐positive glioblastoma cells 23. We therefore examined the effects of SIRT1 inhibition on the proliferation and survival of glioblastoma cells. Salermide is an inhibitor of both SIRT1 and SIRT2 24, while EX527 is a specific inhibitor of SIRT1 25. Similar to AGK2, salermide treatment decreased the sphere formation and proliferation of GB2, GB4, GB11, and GB16 cells and induced the expression of PUMA and GADD45 in GB2 cells (Figs EV2B and EV3A and B). By contrast, EX527 barely suppressed proliferation and sphere formation even at a concentration (20 μM) 500 times higher than its IC50 (38 nM) (Figs EV2B and EV3B–D). EX527 treatment did not induce the expression of PUMA or GADD45 (Fig EV3E). Furthermore, knockdown of SIRT1 using shRNA also barely affected the proliferation and sphere formation of GB2 cells (Fig EV3F). These results suggest that SIRT2, but not SIRT1, plays an important role in the proliferation and survival of glioblastoma cells.

Figure EV3. The SIRT2‐specific inhibitor AGK2 induces growth arrest of glioblastoma cells (related to Fig 1).

Figure EV3

Open in a new tab
  1. Salermide treatment decreases the proliferation of adherent GB2 cells. 5 × 104 cells were treated with 20 μM salermide or DMSO, and the number of viable cells was counted at the indicated time points. Bars indicate mean ± SD (n = 3).
  2. Dose–response curves for AGK2, salermide, or EX527 treatment on GB2 cells. 5 × 104 cells were treated with each reagent, respectively, at the indicated concentrations. After 72 h, the number of viable cells was counted. Bars indicate mean ± SD (n = 3).
  3. EX527 does not inhibit proliferation of GB2 cells. 5 × 104 cells were treated with AGK2 (20 μM), salermide (20 μM), or EX527 (1 or 20 μM) for 72 h. Bars indicate mean ± SD of triplicate technical repeats.
  4. EX527 does not inhibit sphere formation of GB2 cells. Cells were plated in 96‐well plate (1.0 × 103 cells per well) and treated with AGK2 (20 μM), salermide (20 μM), or EX527 (1 or 20 μM). After 7 days of incubation, the number of spheres was counted. Bars indicate mean ± SD of 10 wells.
  5. EX527 does not increase the expression of PUMA or GADD45 mRNA. 1.5 × 105 cells were treated with AGK2 (20 μM), salermide (20 μM), or EX527 (1 or 20 μM), respectively, and subjected to qRT–PCR. Bars indicate mean ± SD (n = 3–4).
  6. GB2 cells were transfected with the indicated shRNAs. (Left panel) Immunoblotting analysis of the effect of shRNA on SIRT1 expression. (Middle panel) The number of viable cells was counted. (Right panel) The expression of PUMA was measured by qRT–PCR. Bars indicate mean ± SD (n = 3).
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01; N.S., not significant.Source data are available online for this figure.

Knockdown of SIRT2 suppresses the tumorigenicity of glioblastoma cells

To examine the effect of SIRT2 knockdown on the tumorigenicity of glioblastoma cells, we transplanted GB2 cells in which SIRT2 expression had been knocked down into the frontal lobe of immunocompromised mice. All of the mice transplanted with control GB2 cells started to decline after 3 months and died within 4 months (Fig 2A). By contrast, all except for one of the mice transplanted with SIRT2‐knockdown GB2 cells survived for more than 5 months after transplantation and only two mice showed obvious clinical symptoms of declining health (Fig 2A). We confirmed these results by a quantitative qRT–PCR measure of tumor burden using primers specific for human versus mouse β‐actin (Fig 2B). Consistent with the results in GB2 cells, knockdown of SIRT2 also significantly suppressed the tumorigenicity of GB16 (mutated p53) cells (Fig 2A). On the other hand, SIRT2 knockdown had modest effect on GB13 (wild‐type p53). Histological studies revealed that all mice transplanted with GB16 cells had developed tumors with diffuse infiltration into surrounding brain tissues, one of the hallmark features of glioblastoma (Fig 2C). By contrast, we found no obvious pathological lesions in the brains of the mice transplanted with SIRT2‐knockdown GB16 cells.

Figure 2. Knockdown of SIRT2 suppresses the tumorigenicity of GB2 cells.

Figure 2

Open in a new tab
  1. Kaplan–Meier overall survival curves of mice transplanted with GB2 cells (left panel), GB13 cells (Middle panel) or GB16 cells (right panel) infected with the indicated lentivirus (1 × 104 cells, 7 mice per group). The y‐axis indicates the percent survival.
  2. Mice were transplanted with the indicated number of GB2 cells infected with a control (empty) or shSIRT2‐expressing (shS2 #1) lentivirus. Six weeks after transplantation, mice (3 or 4 animals, see number of dots) were sacrificed and the expression levels of human β‐actin mRNA were quantified by qRT–PCR.
  3. H&E staining of tumors that were developed in mice implanted with GB16 cells that had been infected with a control (empty) or shSIRT2‐expressing (shS2#1) lentivirus. Scale bar, 1 mm. An image with higher magnification is shown on the right (scale bar, 50 μm).
  4. The effect of AK7 on the deacetylation of acetyl‐tubulin (left panel) and the proliferation of GB2 cells (right panel). (Left panel) Immunoblotting analysis of the effect of AK7 (20 μM) on deacetylation of acetyl‐tubulin/tubulin was performed on day 3 in right panel. Bars indicate mean ± SD (n = 3).
  5. Ten days after intracranial transplantation of GB2 cells (1.0 × 104 cells), AK7 was intraperitoneally administrated for 4 weeks (15 mg/kg, twice/week). After 8 weeks, mice (4 or 5 animals, see number of dots) were sacrificed and the expression levels of human β‐actin mRNA were quantified by qRT–PCR.
Data information: Statistical significance was evaluated using the log‐rank test (for panel A) or unpaired two‐tailed t‐test. **P < 0.01.Source data are available online for this figure.

We next investigated whether a small‐molecule SIRT2 inhibitor could suppress the tumorigenicity of GB2 cells. Since AGK2 is toxic to mice, we examined the effect of AK7, a derivative of AGK2 that passes through the blood–brain barrier 26. Similar to AGK2, treatment with AK7 inhibited deacetylation of acetyl‐tubulin (Fig 2D, left panel) and suppressed the proliferation of GB2 cells in vitro (Fig 2D, right panel). When AK7 was intraperitoneally administrated twice a week for 4 weeks into mice transplanted with GB2 cells, the proliferation and tumorigenicity of the transplanted cells were significantly suppressed (Fig 2E). No toxicity was observed with AK7 treatment, as judged by body weight and food consumption. These results suggest that the function of SIRT2 is critical for the formation and progression of glioblastoma in these mice models.

p73 is required for SIRT2 knockdown‐induced suppression of glioblastoma proliferation

p53 and its family members are known to be important mediators of cell‐cycle arrest and apoptosis through the transactivation of target genes such as PUMA, NOXA, and GADD45 27, 28. It has been reported that SIRT2 binds to and deacetylates p53 29. Because p53 is mutated in GB2 and GB16 cells (Appendix Table S1), we examined whether p73, a close homolog of p53, is involved in the growth arrest and apoptosis of glioblastoma cells resulting from SIRT2 knockdown. It has been reported that p73 has a number of isoforms 30, 31 (schematic representation of the four isoforms studied are shown in Fig 3A). For example, the ΔN isoform of p73 lacks the N‐terminal transactivation domain and thereby acts as a dominant negative regulator, suppressing the activities of the full‐length TA isoforms of p73 and p53 27, 30, 31, 32, 33. In addition to the N‐terminal variation, there are several types of splicing variants, including the α‐ and β‐isoforms, which differ in their C‐termini.

Figure 3. p73 is required for SIRT2 knockdown‐induced suppression of glioblastoma proliferation.

Figure 3

Open in a new tab
  1. Schematic representation of p73 isoforms. TAD, transactivation domain; DBD, DNA‐binding domain; OD, oligomerization domain; SAM, sterile alpha motif domain.
  2. Immunoblotting analysis of p73 in GB2 cells infected with a control (Empty) or shTP73‐expressing lentivirus for 96 h.
  3. Growth curve of GB16 cells infected with the indicated lentiviruses. Bars indicate mean ± SD (n = 3).
  4. Ectopic expression of the ΔN isoform of p73β suppresses the expression levels of PUMA and GADD45 induced by SIRT2 knockdown in GB2 cells. Four days after lentiviral infection, the expression levels of PUMA and GADD45 mRNA were measured by qRT–PCR. Bars indicate mean ± SD (n = 3).
  5. GB2 and GB4 cells [5 × 104 and 2 × 104 cells, respectively (indicated by the dashed lines)] were infected with the indicated lentiviruses. After 96 h, the number of viable cells was counted. Bars indicate mean ± SD (n = 4).
  6. HEK293 cells were transfected with TAp73α along with ΔNp73β. PUMA mRNA was measured by qRT–PCR analyses. Bars indicate mean ± SD (n = 3).
  7. GB2 cells (5 × 104 cells, dashed line) were infected with the indicated lentiviruses, respectively. After 96 h, the number of viable cells was counted. shTP73 corresponds to both α‐ and β‐isoforms of p73. Bars indicate mean ± SD (n = 4).
  8. GB2 cells were transfected with TAp73α along with NLS‐SIRT2 (nuclear‐localizing mutant of SIRT2) or 3mut (deacetylase‐inactive, nuclear‐localizing mutant of SIRT2) and a reporter construct consisting of the promoter region of PUMA fused to a luciferase gene (left panel). Reporter activities were determined by dual‐luciferase assays (right panel). Bars indicate mean ± SD (n = 3).
  9. GB2 cells were electroporated with TAp73α along with NLS‐SIRT2 (nuclear‐localizing mutant of SIRT2) or 3mut (deacetylase‐inactive, nuclear‐localizing mutant of SIRT2). The expression of PUMA and GADD45 was measured by qRT–PCR analysis. Bars indicate mean ± SD (n = 3).
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.Source data are available online for this figure.

We found that the TA isoform of p73 was expressed in GB2 cells and that its expression was suppressed by an shRNA targeting p73 (Fig 3B). Immunostaining experiments using an antibody against p73 and two distinct antibodies against SIRT2 revealed that SIRT2 and p73 were localized in the nucleus in patient samples (Fig EV4A and B). In addition, proximity ligation assay confirmed that SIRT2 and p73 were colocalized in the nucleus (Fig EV4C). When the ΔN isoform of p73β (ΔNp73β), a dominant negative isoform of p73, was ectopically expressed in GB16 cells, the proliferation of GB16 cells was not inhibited by knockdown of SIRT2 (Fig 3C). We could confirm these results by Ki67 staining (Fig EV4D). Consistent with the results obtained with GB16 cells, ectopic expression of ΔNp73β in GB2 or GB4 cells suppressed the expression of PUMA and GADD45 and rescued the growth suppression induced by SIRT2 knockdown (Fig 3D and E). In addition, we confirmed that ΔNp73β suppressed p73α‐mediated upregulation of PUMA in a dose‐dependent manner (Figs 3F and EV4E). We also observed that knockdown of p73 by a specific shRNA in GB2 cells rescued the growth suppression induced by SIRT2 knockdown (Fig 3G). These findings suggest that p73 is required for SIRT2‐mediated proliferation of glioblastoma cells.

Figure EV4. Nuclear expression and the role of p73 and SIRT2 in glioblastoma (related to Fig 3).

Figure EV4

Open in a new tab
  1. Tissue sections were stained with H&E, anti‐SIRT2 antibody, or anti‐TAp73 antibody. Scale bars, 50 μm.
  2. Histological examination of patients' samples. Tissue sections were stained with H&E, anti‐SIRT2 antibody, or anti‐TAp73 antibody, respectively. Scale bars, 50 μm.
  3. GB2 cells were incubated with the indicated antibodies and subjected to proximity ligation assay. Scale bars, 10 μm.
  4. GB16 cells were infected with the indicated lentiviruses and plated in laminin‐coated 24‐well tissue culture plates. Three days after lentiviral infection, cells were fixed and stained with anti‐Ki67 antibody. The percentage of Ki67‐positive cells was assessed by immunofluorescence. Bars indicate mean ± SD of five randomly selected fields.
  5. Immunoblotting analysis of p73 protein related to Fig 3F (white arrowhead, TAp73α; black arrowhead, ΔNp73β; *, non‐specific band).
  6. The nuclear and cytoplasmic fractions were extracted from GB2 cells and subjected to immunoblotting with the indicated antibodies. Lamin and α‐tubulin were used as nuclear and cytoplasmic markers, respectively (*, non‐specific band).
  7. Representative images of subcellular localization of SIRT2. GB2 cells were cultured in the presence of laminin. Adherent GB2 cells were immunostained with anti‐SIRT2 antibody and stained with TO‐PRO‐3. Scale bars, 20 μm.
  8. GB2 cells were electroporated with TAp73α along with NLS‐SIRT2 (nuclear‐localizing mutant of SIRT2) or 3mut (deacetylase‐inactive, nuclear‐localizing mutant of SIRT2). Immunoblotting analysis of the indicated proteins related to Fig 3I.
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.Source data are available online for this figure.

Suppression of p73 transcriptional function requires SIRT2 deacetylase activity

To elucidate how SIRT2 modulates the transcriptional activity of p73, we generated a reporter construct consisting of the promoter region of PUMA fused to a luciferase gene (Fig 3H left panel). Co‐transfection of the TAp73α isoform (TAp73α) with this reporter construct into GB2 cells resulted in a significant enhancement of reporter activity (~5‐fold), confirming that this promoter region is regulated by p73 (Fig 3H right panel). Although endogenous SIRT2 is mainly localized in the nucleus in GB2 cells and patient samples (Fig EV4A–C, F and G), ectopically expressed wild‐type SIRT2 was localized predominantly in the cytoplasm 34. We therefore used a nuclear‐localizing mutant of SIRT2 8 for the following experiments. Co‐transfection of the nuclear‐localizing SIRT2 along with TAp73α into GB2 cells resulted in significantly lower reporter activity compared to transfection of TAp73α alone (Fig 3H). By contrast, co‐transfection of a deacetylase‐inactive, nuclear‐localizing mutant of SIRT2 (3mut) failed to suppress TAp73α reporter activity. In addition, we examined the effects of SIRT2 and TAp73α on PUMA and GADD45 mRNA expression and obtained similar results (Figs 3I and EV4H). These results suggest that SIRT2 suppresses the transcriptional activity of p73 in a deacetylase activity‐dependent manner.

SIRT2 deacetylates p73α

We next investigated whether SIRT2 deacetylates p73. HEK293T cells were transfected with each p73 isoform, treated with AGK2 for 6 h, and then, acetylation levels were determined by immunoprecipitation. Treatment of 293T cells with AGK2 resulted in a significant increase in the acetylation levels of TAp73α and a slight increase in that of the ΔNp73α isoform (Fig 4A). By contrast, acetylation of the β isoforms was not increased by AGK2 treatment. We obtained similar results by shRNA suppression of SIRT2 expression (Fig EV5A and B).

Figure 4. SIRT2 deacetylates p73 and suppresses its transcriptional activity.

Figure 4

Open in a new tab
  1. Cells transfected with FLAG‐tagged p73 isoforms were treated with 20 μM AGK2 or DMSO for 6 h and subjected to immunoprecipitation with anti‐FLAG antibody followed by immunoblotting with antibody against acetylated lysine or FLAG.
  2. Three conserved lysine residues are located in the most C‐terminal region of p73. Amino acid sequences of p73 of the indicated species and human p53 and p63 are aligned. *, conserved and acetylated lysine residues, black frame, highly conserved residues, gray frame, conserved residues.
  3. Schematic representation of TAp73α and K3R. K3R is a mutant TAp73α in which three lysine (K) residues in the C‐terminal region are replaced with arginine (R).
  4. Cells transfected with the indicated constructs were treated with 20 μM AGK2 or DMSO, respectively, for 6 h and subjected to immunoprecipitation with anti‐FLAG antibody followed by immunoblotting with antibody against acetylated lysine or FLAG.
  5. HEK293T cells transfected with FLAG‐tagged TAp73α were treated with 20 μM AGK2 (pre‐AGK2+) or DMSO (pre‐AGK2−) for 6 h. p73 was purified by immunoprecipitation and incubated with recombinant SIRT2 (10 U), NAD (1 mM), and/or AGK2 (3.5 μM) as indicated.
  6. qRT–PCR analysis of PUMA and GADD45 mRNA in GB2 cells infected with the indicated lentiviruses. Bars indicate mean ± SEM (n = 4–5).
  7. Expression of p73 in (F) was determined by immunoblotting with anti‐p73 antibody. RFP was used as a control.
  8. GB glioblastoma neurospheres [5 × 104 cells (indicated by the dashed line)] were infected with the indicated lentiviruses. After 96 h, the number of viable cells was counted. Bars indicate mean ± SD (n = 4).
  9. GB2 cells infected with the indicated lentiviruses were intracranially transplanted into immunocompromised mice. After 8 weeks, mice (3–5 animals, see number of dots) were sacrificed and the expression levels of human β‐actin mRNA were quantified by qRT–PCR.
  10. SIRT2‐mediated inactivation of p73 is critical for the proliferation and tumorigenicity of glioblastoma cells. SIRT2 regulates the transcriptional activity of the tumor suppressor p73 by deacetylating its C‐terminal lysine residues.
Data information: Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.Source data are available online for this figure.

Figure EV5. SIRT2 deacetylates lysine residues located in the C‐terminal region of p73 (related to Fig 4).

Figure EV5

Open in a new tab
  1. Knockdown of SIRT2 causes an increase in the acetylation levels of α isoforms of p73. HEK293T cells were infected with a control or shSIRT2‐expressing lentivirus, respectively. Four days after lentiviral infection, cells were transfected with the indicated constructs. After 20 h, cells were lysed and subjected to immunoprecipitation with anti‐FLAG antibody followed by immunoblotting with the indicated antibodies.
  2. Knockdown of SIRT2 does not cause an increase in the acetylation levels of the β isoforms of p73.
  3. FLAG‐tagged TAp73α expressed in HEK293 cells was subjected to Nano LC‐MS/MS analysis to identify acetylation sites.
  4. Overexpression of K3R barely induces the expression of PUMA and GADD45. GB16 cells were infected with the indicated lentiviruses. Four days after lentiviral infection, the expression levels of PUMA and GADD45 mRNA were measured by qRT–PCR. Bars indicate mean ± SD (n = 3). Statistical significance was evaluated using unpaired two‐tailed t‐test. *P < 0.05; **P < 0.01.

Source data are available online for this figure.

There is a cluster of three lysine residues in the C‐terminal region of the p73α isoform, but not in the other isoforms, including the p73β isoform (Fig 4B). To examine whether these residues are substrates for SIRT2, we generated a mutant p73 (K3R), in which these lysine residues were replaced with arginine (Fig 4C). The acetylation level of K3R was barely increased by AGK2‐treatment or SIRT2 knockdown (Figs 4D and EV5A). Moreover, Nano LC‐MS/MS analysis revealed that Lys620, 623, and 627 of p73 (Fig 4B) were indeed acetylated (Fig EV5C). These results suggest that the acetylated lysine residues located in the C‐terminal region of p73α are substrates for SIRT2.

To examine whether deacetylation of p73 is directly catalyzed by SIRT2, we performed in vitro deacetylation assays using recombinant SIRT2 and the TAp73α purified from HEK293 cells that had been treated with AGK2. We found that acetylation of TAp73α was significantly decreased in an NAD‐dependent manner when incubated with SIRT2 (Fig 4E). Furthermore, addition of AGK2 inhibited the SIRT2‐mediated acetylation of TAp73α. Thus, deacetylation of p73 may be directly catalyzed by SIRT2.

The transcriptional activity of p73 is regulated by acetylation of the lysine residues located in the C‐terminal region

The transcriptional activity of p53 is known to be enhanced by acetylation of lysine residues in its C‐terminal region 35, 36. We therefore asked whether the transcriptional activity of p73 is similarly regulated by acetylation and deacetylation of the lysine residues in its C‐terminal region. When GB2 cells were infected with a lentivirus expressing TAp73α, the expression levels of PUMA and GADD45 were significantly enhanced (Fig 4F and G). By contrast, the K3R mutant barely induced the expression of these genes. Similar results were obtained with GB16 cells, which have a mutated p53 (Fig EV5D). Furthermore, we also found that ectopic expression of TAp73α, but not of the K3R mutant, decreased proliferation of GB2, GB4, and GB16 cells (Fig 4H). Moreover, ectopic expression of TAp73α, but not of the K3R mutant, in GB2 cells resulted in a decrease in the tumorigenicity of these cells (Fig 4I). These results suggest that acetylation of the lysine residues located in the C‐terminal region of p73 enhances its transcriptional activity and thereby negatively regulates the proliferation and survival of glioblastoma cells.

In this study, we showed that SIRT2‐mediated inactivation of p73 is required for the proliferation and tumorigenicity of glioblastoma cells. We used GB2 and GB16 cells, both of which contain a mutant p53 with an inactivating mutation in its DNA‐binding domain, but which possess wild‐type p73. It has been reported that the transcriptional targets of p73 and p53 overlap and that p73 plays important roles in the induction of growth arrest and apoptosis 27. We showed that overexpression of p73 or inhibition of SIRT2 in GB2 or GB16 cells results in the transactivation of PUMA and the induction of apoptosis. Furthermore, we showed that SIRT2 deacetylates p73 and represses its transcriptional activity. These findings suggest that SIRT2 allows glioblastoma cells containing mutated p53 to escape p73‐mediated growth arrest and apoptosis. Since SIRT2 is also known to bind and deacetylate p53 29, SIRT2 may exert its effect by inactivating p53 in glioblastoma cells containing wild‐type p53.

It is well known that most tumor cells contain genetic alterations that typically result from genome instability, including mutations, amplification, rearrangements, and aneuploidy 37. Furthermore, it is also well known that the p53 family of tumor suppressors plays critical roles in regulating stress‐induced gene expression 38. Thus, we speculate that p73 is activated in glioblastoma cells, presumably due to stress caused by aberrant genomic structures, and that SIRT2 is required in these cells for their survival.

It has been reported that the lysine residues located in the C‐terminal region of p53 are acetylated in response to DNA damage by acetyltransferases such as p300, CBP, and PCAF 35, 36, 39. Transcriptional co‐activators are recruited to these acetylated residues and enhance p53‐mediated transactivation. Furthermore, it has been suggested that acetylation stabilizes p53 by preventing MDM2‐mediated ubiquitination at these lysine residues 40, 41. However, the C‐terminal regions of p73 and p53 are not homologous (Fig 4B), and it has been reported that the lysine residues in the DNA‐binding domain of p73, a domain located in the central part of p73, are acetylated by p300 in response to DNA damage 42. These findings imply that the transcriptional activities of p53 and p73 might be regulated by somewhat different mechanisms.

In conclusion, we have demonstrated that SIRT2‐mediated inactivation of p73 is critical for the proliferation and tumorigenicity of glioblastoma cells. Administration of AK7 suppressed the tumorigenicity of glioblastoma cells without showing any adverse effects (Fig 2D and E). These observations suggest that SIRT2 may be a promising molecular target for the therapy of glioblastoma.

Materials and Methods

Tumor specimen and cell cultures

Tumor samples classified as primary glioblastoma were obtained from patients undergoing surgical treatment at the University of Tokyo Hospital as approved by the Institutional Review Board. All the patients provided written informed consent. Tumors were washed and mechanically and enzymatically dissociated into single cells. Tumor cells were cultured in neurobasal or DMEM/F12 (Invitrogen) containing B27 supplement minus vitamin A (Invitrogen), EGF and FGF2 (20 ng/ml each, Wako Pure Chemical Industries). Some glioblastoma neurospheres did not proliferate when cultured in vitro or inoculated into mice. Thus, the glioblastoma spheres that grew well both in vitro and in vivo were used for experiments. For serum‐induced differentiation, cells were cultured in DMEM/F12 containing 10% FBS. HEK293T and 293FT cells were cultured in DMEM (Nissui) containing 10% FBS.

siRNA screen

An siRNA library targeting 246 genes commonly expressed in GB glioblastoma neurospheres (Appendix Table S2) was purchased from Invitrogen. GB2 cells were trypsinized and transfected with a pool of two siRNA duplexes targeting each gene using RNAiMAX (Invitrogen). After incubation for 72 h, CD133 expression was assessed by qRT–PCR. Knockdown efficiencies were evaluated for randomly selected 31 genes. Efficiency was more than 50% for 23 genes (74%) and more than 70% for 17 genes (55%).

Antibodies

Mouse monoclonal antibody (mAb) to M2 FLAG‐tag and rabbit polyclonal antibody (pAb) to SIRT2 (S8447) (used for Fig EV4A) were obtained from Sigma‐Aldrich. Mouse mAb to p73α/β (Clone ER‐15) for immunoblotting was from Thermo Scientific. Rabbit monoclonal antibody to acetylated lysine, acetylated α‐tubulin Lys40 (D20G3), and rabbit pAb to cleaved caspase‐3 (Asp175) were from Cell Signaling Technology. Rabbit mAb to SIRT2 (EPR1667, AJ1718a) was purchased from Abgent (The representative blot of SIRT2 using this antibody is shown in Appendix Fig S2). Mouse mAb to SIRT2 (A‐5, sc‐28298) (used for Fig EV4C) and rabbit pAb to GFP (sc‐8334), SIRT2 (H‐95, sc‐20966) (used for Figs EV4B and 4F), and p73 (H‐79, sc‐7957) (used in Fig EV4C) were from Santa Cruz Biotechnology. Mouse mAb to α‐tubulin (DM1A) was from Calbiochem. Mouse mAb to GAPDH and rabbit pAb to p73 (Ab‐4) (used for Fig EV4A and B) was from Millipore. Mouse mAb to Ki67 was from Dako.

Inhibitors

AGK2 and AK7 were obtained from ChemBridge, dissolved in dimethyl sulfoxide (DMSO), and stored as aliquots at −80°C. EX527 (sc‐203044), doxorubicin–HCl (sc‐200923), and camptothecin (sc‐200871) were from Santa Cruz Biotechnology. The concentrations of inhibitors used were as follows: AGK2, 20 μM; salermide, 20 μM; EX527, 1 or 20 μM; doxorubicin, 5 μM; camptothecin, 1 μM.

Sphere‐forming assay

Cells were labeled with EGFP by lentiviral infection, trypsinized, and plated in 96‐well tissue culture plates at the indicated cell number (5–1,000 cells). After incubation for 10–16 days at 37°C, sphere number and diameter were analyzed using an In Cell Analyzer 2000 (GE Healthcare).

Immunoblotting

Cells were lysed in lysis buffer (20 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% NP‐40, 0.5% deoxycholate, 0.1% SDS, 50 mM NaF, 1 mM dithiothreitol). Lysates were fractionated by SDS–polyacrylamide gel electrophoresis and transferred to a PVDF membrane (Immobilon P, Millipore). The membrane was subjected to immunoblot analysis using horseradish peroxidase‐conjugated donkey anti‐rabbit immunoglobulin G (IgG) (GE Healthcare) or sheep anti‐mouse IgG (GE Healthcare) as a secondary antibody. Visualization was performed using the Enhanced Chemiluminescence Plus Western Blotting Detection System (GE Healthcare) and a LAS‐4000EPUVmini Luminescent Image Analyzer (Fujifilm). For reprobing, blots were stripped for 15 min at room temperature in Restore Western Blot Stripping Buffer (Thermo Scientific).

Immunoprecipitation

Cells were lysed with lysis buffer (10 mM Tris–HCl, pH 8.0, 140 mM NaCl, 1 mM EDTA, 1% Triton X‐100). Lysates were centrifuged to remove cell debris, and the supernatants were incubated with antibodies with constant rotation. A 50% slurry of protein A‐sepharose was added and incubated for 30 min. Beads were washed 5 times with wash buffer (10 mM Tris–HCl, pH 8.0, 140 mM NaCl, 1 mM EDTA, 0.1% Triton X‐100) and once with wash buffer without Triton X‐100.

Quantitative RT–PCR

Total RNA was extracted using the NucleoSpin RNA Clean‐up kit (Macherey‐Nagel) and reverse‐transcribed into cDNA using the ReverTra Ace qPCR RT Kit (Toyobo). Real‐time PCR was performed using LightCycler480 SYBR Green I Master and a LightCycler480 Instrument (Roche). The results were normalized against the values detected for GAPDH. Primers used in RT–PCR are described in Supplemental Information.

Lentivirus production and RNA interference

Lentiviral vectors (CS‐Rfa‐CG, CS‐RfA‐CMV‐mRFP1) harboring a shRNA driven by the H1 promoter or a gene driven by the CMV promoter were transfected with the packaging vectors pCAG‐HIV‐gp and pCMV‐VSV‐G‐RSV‐Rev into 293FT cells using Lipofectamine 2000 Transfection Reagent (Invitrogen). All plasmids were kindly provided by H. Miyoshi (RIKEN BioResource Center, Japan). Virus supernatant was purified by ultracentrifugation at 106,800 g for 90 min (SW28 rotor, Beckman). The target sequences for shRNA are as follows:

  • human SIRT2 #1, 5′‐GTCGCAGAGTCATCTGTTTGG‐3′;

  • human SIRT2 #2, 5′‐GCTACACGCAGAACATAGATA‐3′;

  • human SIRT2 #3, 5′‐GAGATCAGCTATTTCAAGAAA‐3′;

  • human p73, 5′‐GCGTGGAAGGCAATAATCTCT‐3′;

The infection efficiency of the lentiviruses was more than 95%, as judged by GFP or RFP fluorescence.

Intracranial xenografts

One week after lentivirus infection, RNA was extracted from a portion of the infected cells and the knockdown efficiency was measured by quantitative RT–PCR. The knockdown cells (1.0 × 103 or 1.0 × 104 cells) were injected stereotactically into the right frontal lobe of 5‐week‐old nude mice (nu/nu, Charles River), following administration of general anesthesia. The injection coordinates were 2 mm to the right of the midline, 1 mm anterior to the coronal suture, and 3 mm deep. Mice were monitored for up to 6 months. Survival of mice was evaluated by Kaplan–Meier analysis. P‐value was calculated using a log‐rank test. Tumors were histologically analyzed after staining with hematoxylin and eosin (H&E). Tumor distribution was analyzed by GFP immunostaining. For the quantification of tumors, whole brains were homogenized in TRIsure (Bioline) and RNA was extracted following the manufacturer's instructions. After reverse transcription, the amount of human β‐actin mRNA was quantified by real‐time RT–PCR using the species‐specific primers described in Supplemental Information. The results were normalized against the values detected for mouse β‐actin. AK7 was intraperitoneally administrated for 4 weeks (15 mg/kg, twice/week). Animals were randomly assigned to cages (six mice per cage). No statistical methods were used to predetermine the sample size of animals. All animal experimental protocols were performed in accordance with the politics of the Animal Ethics Committee of the University of Tokyo.

Flow cytometry

Cells were trypsinized and resuspended in PBS. After blocking with FcR‐blocking reagent (Miltenyi Biotec), cells were incubated with FITC‐conjugated anti‐CD133 antibody (BD Biosciences) for 15 min at 4°C. Cells were sorted using FACSAria (BD Biosciences), and the sorted cells were reanalyzed using FACS Caliber (BD Biosciences). For sub‐G0 DNA assay, cells were trypsinized, collected in PBS and fixed in 70% cold ethanol. After RNase treatment, cells were stained with propidium iodide (50 μg/ml) in PBS and analyzed using FACS Calibur (Becton Dickinson).

Caspase‐3/7 glo assay

Caspase‐Glo 3/7 assay was performed according to manufacturer's instructions (Promega). Briefly, GB2 cells were plated on 96‐well plate and treated with AGK2 (10 μM or 20 μM) or DMSO for 3, 6 or 24 h, respectively. After 30 min of incubation with the caspase‐Glo 3/7 reagent, the luminescence of each well was measured using a Synergy H1 Hybrid Multi‐Mode Microplate Reader (BioTek).

Immunohistochemistry

Immunostaining of paraffin‐embedded tissue sections was performed according to the manufacturer's instructions (Vector Laboratories). Briefly, sections were deparaffinized in xylene and then rehydrated into distilled water using graded ethanol. Antigens were retrieved by microwaving the slides in citrate buffer (pH 6.0) for 15 min. Sections were incubated with the indicated antibodies overnight at 4°C. Staining patterns obtained with the antibodies to GFP were visualized with VECTASTAIN Elite ABC Rabbit IgG Kit (Vector Laboratories). The sections were visualized and photographed under a microscope.

In situ PLA

Proximity ligation assay was performed using Duolink in situ PLA kit (Sigma‐Aldrich). Briefly, GB2 cells were plated on cover slips coated with laminin (R&D Systems) 48 h prior to assay. Cells were fixed with 4% paraformaldehyde in PBS and permeabilized in 0.3% Triton X‐100, 0.1% bovine serum albumin in PBS (1 h, room temperature). After 1 h of blocking in Duolink blocking solution, cells were incubated with anti‐SIRT2 mouse monoclonal antibody (Santa Cruz A‐5) and anti‐p73 rabbit polyclonal antibody (Santa Cruz H‐79) for overnight at 4°C. A protein–protein interaction was detected using Duolink in situ PLA probes, followed by ligation and amplification steps according to the manufacturer's instruction. Texas Red and DAPI signals were visualized using a fluorescence microscope (Olympus BX51 equipped with Hamamatsu C4742‐95 digital CCD camera).

Luciferase assay

Cells were plated in 24‐well dishes 24 h prior to transfection. Transfections were performed with Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. Briefly, 0.3 μg of pcDNA3.1‐TAp73α, 0.4 μg of pcDNA3.1‐NLS‐SIRT2 (WT or Mutant), 0.1 μg of pGL3 PUMA reporter, and 0.005 μg of pRL‐TK Renilla Luciferase Control Vector were incubated with 2 μl of Lipofectamine 2000 (Invitrogen) and co‐transfected into GB2 cells. pcDNA3.1 was used as a vector control. Luciferase assays were performed using the dual‐luciferase reporter assay system (Promega) according to the manufacturer's protocol. A nuclear‐localizing mutant of SIRT2 mutant was created as described 8.

Electroporation

GB2 cells (1 × 106 cells/sample) were transfected by NEPA21 electroporator (Nepagene) with 125V poring pulse voltage, 7.5 ms at 50‐ms intervals, two pulses, and 10% decay rate, and a transfer pulse of 20 V, 50 ms at 50‐ms intervals, five pulses, and a 40% decay rate.

In vitro deacetylation assay

293T cells ectopically expressing FLAG‐tagged p73 were treated with AGK2 for 6 h, and p73 was immunoprecipitated. The precipitants were resuspended in deacetylation buffer (50 mM Tris–HCl pH 8.0, 140 mM NaCl, 2 mM MgCl2) and incubated with recombinant SIRT2 protein (10 U) for 2 h at room temperature. The reaction was terminated by boiling with SDS sample buffer, and acetylation levels were determined by immunoblotting.

Mass spectrometry

Excised gel bands were incubated with 10 mM DTT in 100 mM NH4HCO3 for 1 h at 56°C, followed by 55 mM iodoacetamide in 100 mM NH4HCO3 for 45 min at room temperature in the dark. Gel pieces were incubated with 25 ng/ml of trypsin gold (Promega) in 50 mM NH4HCO3 at 37°C for 6 h. Eluted peptides were subjected to Nano LC‐MS/MS analysis. The Nano LC‐MS/MS analysis was conducted using a LTQ‐Orbitrap Velos mass spectrometer (Thermo Fisher Scientific, Bremen, Germany) equipped with a Nano LC‐MS/MS interface (AMR, Tokyo, Japan), a Nano LC‐MS/MS system (Michrom Paradigm MS2), and an HTC‐PAL autosampler (CTC, Analytics, Zwingen, Switzerland). The in‐gel digested samples were loaded onto a trap column (0.3 mm ID × 5 mm, 5 mm, L‐column; CERI) and directly connected to a Zaplous a Pep‐C18 packed column (3 μm, 0.1 × 150 mm) (AMR, Tokyo, Japan). MS data were searched against the Uniprot protein sequence database (Homo sapiens) using the search program Proteome Discoverer 1.4 (Thermo Fisher Scientific). Cys carbamidomethylation was searched as a fixed modification, whereas oxidized Met and acetylation of Lys were searched as variable modifications.

Statistical analysis

Unpaired two‐tailed t‐test (Welch's t‐test) was performed for statistical analysis. Normality of data was evaluated by Shapiro–Wilk normality test. Survival of mice inoculated with glioblastoma neurospheres was evaluated by Kaplan–Meier analysis. P‐value was calculated using a log‐rank test.

Data availability

The raw data of the mass spectrometry were deposited at jPOST (jPOST ID: JIST000414, PXID: PXD010411).

Author contributions

KF, TH, KE, NS, RK‐N, YN‐N, and YM performed the experiments. LN performed mass spectrometry analysis. VT, TT, YI, AM, and NS prepared glioblastoma specimens. KF, TH, and TA analyzed the data and wrote the paper.

Conflict of interest

The authors declare that they have no conflict of interest.

Supporting information

Appendix

Click here for additional data file. (943.1KB, pdf)

Expanded View Figures PDF

Click here for additional data file. (1.2MB, pdf)

Source Data for Expanded View

Click here for additional data file. (6.8MB, zip)

Review Process File

Click here for additional data file. (385.1KB, pdf)

Source Data for Figure 2

Click here for additional data file. (755.5KB, pdf)

Source Data for Figure 3

Click here for additional data file. (208.9KB, pdf)

Source Data for Figure 4

Click here for additional data file. (1.7MB, pdf)

Acknowledgements

This work was supported by P‐CREATE (Project for Cancer Research and Therapeutic Evolution, no. 17cm0106103h0002) and P‐DIRECT (Project for Development of Innovative Research on Cancer Therapeutics, no. 16cm0106103h0001) grants from the Japan Agency for Medical Research and Development, and Grants‐in‐Aid for Scientific Research on Innovative Areas (Integrative Analysis and Regulation of Cellular Diversity, no. 17H06325) from MEXT, Japan.

EMBO Reports (2018) 19: e45587

References

  • 1. Fine HA (2015) New strategies in glioblastoma: exploiting the new biology. Clin Cancer Res 21: 1984–1988 [DOI] [PubMed] [Google Scholar]
  • 2. Brennan CW, Verhaak RG, McKenna A, Campos B, Noushmehr H, Salama SR, Zheng S, Chakravarty D, Sanborn JZ, Berman SH et al (2013) The somatic genomic landscape of glioblastoma. Cell 155: 462–477 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3. Lathia JD, Mack SC, Mulkearns‐Hubert EE, Valentim CL, Rich JN (2015) Cancer stem cells in glioblastoma. Genes Dev 29: 1203–1217 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Parsons DW, Jones S, Zhang XS, Lin JCH, Leary RJ, Angenendt P, Mankoo P, Carter H, Siu IM, Gallia GL et al (2008) An integrated genomic analysis of human glioblastoma multiforme. Science 321: 1807–1812 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Chin L, Meyerson M, Aldape K, Bigner D, Mikkelsen T, VandenBerg S, Kahn A, Penny R, Ferguson ML, Gerhard DS et al (2008) Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature 455: 1061–1068 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Bredel M, Scholtens DM, Yadav AK, Alvarez AA, Renfrow JJ, Chandler JP, Yu ILY, Carro MS, Dai FP, Tagge MJ et al (2011) NFKBIA deletion in glioblastomas. New Engl J Med 364: 627–637 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Sebastian C, Satterstrom FK, Haigis MC, Mostoslavsky R (2012) From sirtuin biology to human diseases: an update. J Biol Chem 287: 42444–42452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Luthi‐Carter R, Taylor DM, Pallos J, Lambert E, Amore A, Parker A, Moffitt H, Smith DL, Runne H, Gokce O et al (2010) SIRT2 inhibition achieves neuroprotection by decreasing sterol biosynthesis. Proc Natl Acad Sci USA 107: 7927–7932 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. North BJ, Marshall BL, Borra MT, Denu JM, Verdin E (2003) The human Sir2 ortholog, SIRT2, is an NAD(+)‐dependent tubulin deacetylase. Mol Cell 11: 437–444 [DOI] [PubMed] [Google Scholar]
  • 10. Kim HS, Vassilopoulos A, Wang RH, Lahusen T, Xiao Z, Xu X, Li C, Veenstra TD, Li B, Yu H et al (2011) SIRT2 maintains genome integrity and suppresses tumorigenesis through regulating APC/C activity. Cancer Cell 20: 487–499 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Zhou WH, Ni TK, Wronski A, Glass B, Skibinski A, Beck A, Kuperwasser C (2016) The SIRT2 deacetylase stabilizes slug to control malignancy of basal‐like breast cancer. Cell Rep 17: 1302–1317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Roth M, Chen WY (2014) Sorting out functions of sirtuins in cancer. Oncogene 33: 1609–1620 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Peck B, Chen CY, Ho KK, Di Fruscia P, Myatt SS, Coombes RC, Fuchter MJ, Hsiao CD, Lam EW (2010) SIRT inhibitors induce cell death and p53 acetylation through targeting both SIRT1 and SIRT2. Mol Cancer Ther 9: 844–855 [DOI] [PubMed] [Google Scholar]
  • 14. Jing H, Hu J, He B, Abril YLN, Stupinski J, Weiser K, Carbonaro M, Chiang YL, Southard T, Giannakakou P et al (2016) A SIRT2‐selective inhibitor promotes c‐Myc oncoprotein degradation and exhibits broad anticancer activity. Cancer Cell 29: 297–310 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Serrano L, Martinez‐Redondo P, Marazuela‐Duque A, Vazquez BN, Dooley SJ, Voigt P, Beck DB, Kane‐Goldsmith N, Tong Q, Rabanal RM et al (2013) The tumor suppressor SirT2 regulates cell cycle progression and genome stability by modulating the mitotic deposition of H4K20 methylation. Gene Dev 27: 639–653 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Koyama‐Nasu R, Nasu‐Nishimura Y, Todo T, Ino Y, Saito N, Aburatani H, Funato K, Echizen K, Sugano H, Haruta R et al (2013) The critical role of cyclin D2 in cell cycle progression and tumorigenicity of glioblastoma stem cells. Oncogene 32: 3840–3845 [DOI] [PubMed] [Google Scholar]
  • 17. Singh SK, Hawkins C, Clarke ID, Squire JA, Bayani J, Hide T, Henkelman RM, Cusimano MD, Dirks PB (2004) Identification of human brain tumour initiating cells. Nature 432: 396–401 [DOI] [PubMed] [Google Scholar]
  • 18. Brescia P, Ortensi B, Fornasari L, Levi D, Broggi G, Pelicci G (2013) CD133 is essential for glioblastoma stem cell maintenance. Stem Cells 31: 857–869 [DOI] [PubMed] [Google Scholar]
  • 19. Ikushima H, Todo T, Ino Y, Takahashi M, Miyazawa K, Miyazono K (2009) Autocrine TGF‐beta signaling maintains tumorigenicity of glioma‐initiating cells through Sry‐related HMG‐Box factors. Cell Stem Cell 5: 504–514 [DOI] [PubMed] [Google Scholar]
  • 20. Son MJ, Woolard K, Nam DH, Lee J, Fine HA (2009) SSEA‐1 is an enrichment marker for tumor‐initiating cells in human glioblastoma. Cell Stem Cell 4: 440–452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21. Gomes P, Outeiro TF, Cavadas C (2015) Emerging role of sirtuin 2 in the regulation of mammalian metabolism. Trends Pharmacol Sci 36: 756–768 [DOI] [PubMed] [Google Scholar]
  • 22. Outeiro TF, Kontopoulos E, Altmann SM, Kufareva I, Strathearn KE, Amore AM, Volk CB, Maxwell MM, Rochet JC, McLean PJ et al (2007) Sirtuin 2 inhibitors rescue alpha‐synuclein‐mediated toxicity in models of Parkinson's disease. Science 317: 516–519 [DOI] [PubMed] [Google Scholar]
  • 23. Chang CJ, Hsu CC, Yung MC, Chen KY, Tzao C, Wu WF, Chou HY, Lee YY, Lu KH, Chiou SH et al (2009) Enhanced radiosensitivity and radiation‐induced apoptosis in glioma CD133‐positive cells by knockdown of SirT1 expression. Biochem Biophys Res Commun 380: 236–242 [DOI] [PubMed] [Google Scholar]
  • 24. Lara E, Mai A, Calvanese V, Altucci L, Lopez‐Nieva P, Martinez‐Chantar ML, Varela‐Rey M, Rotili D, Nebbioso A, Ropero S et al (2009) Salermide, a Sirtuin inhibitor with a strong cancer‐specific proapoptotic effect. Oncogene 28: 781–791 [DOI] [PubMed] [Google Scholar]
  • 25. Solomon JM, Pasupuleti R, Xu L, McDonagh T, Curtis R, DiStefano PS, Huber LJ (2006) Inhibition of SIRT1 catalytic activity increases p53 acetylation but does not alter cell survival following DNA damage. Mol Cell Biol 26: 28–38 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Taylor DM, Balabadra U, Xiang ZM, Woodman B, Meade S, Amore A, Maxwell MM, Reeves S, Bates GP, Luthi‐Carter R et al (2011) A brain‐permeable small molecule reduces neuronal cholesterol by inhibiting activity of sirtuin 2 deacetylase. ACS Chem Biol 6: 540–546 [DOI] [PubMed] [Google Scholar]
  • 27. Melino G, De Laurenzi V, Vousden KH (2002) p73: friend or foe in tumorigenesis. Nat Rev Cancer 2: 605–615 [DOI] [PubMed] [Google Scholar]
  • 28. Menendez D, Inga A, Resnick MA (2009) The expanding universe of p53 targets. Nat Rev Cancer 9: 724–737 [DOI] [PubMed] [Google Scholar]
  • 29. van Leeuwen IM, Higgins M, Campbell J, McCarthy AR, Sachweh MC, Navarro AM, Lain S (2013) Modulation of p53 C‐terminal acetylation by mdm2, p14ARF, and cytoplasmic SirT2. Mol Cancer Ther 12: 471–480 [DOI] [PubMed] [Google Scholar]
  • 30. Vikhreva P, Melino G, Amelio I (2018) p73 alternative splicing: exploring a biological role for the C‐terminal isoforms. J Mol Biol 430: 1829–1838 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. De Laurenzi VD, Catani MV, Terrinoni A, Corazzari M, Melino G, Costanzo A, Levrero M, Knight RA (1999) Additional complexity in p73: induction by mitogens in lymphoid cells and identification of two new splicing variants epsilon and zeta. Cell Death Differ 6: 389–390 [DOI] [PubMed] [Google Scholar]
  • 32. Ishimoto O, Kawahara C, Enjo K, Obinata M, Nukiwa T, Ikawa S (2002) Possible oncogenic potential of delta Np73: a newly identified isoform of human p73. Cancer Res 62: 636–641 [PubMed] [Google Scholar]
  • 33. Stiewe T, Theseling CC, Putzer BM (2002) Transactivation‐deficient triangle TA‐p73 inhibits p53 by direct competition for DNA binding – implications for tumorigenesis. J Biol Chem 277: 14177–14185 [DOI] [PubMed] [Google Scholar]
  • 34. Skoge RH, Dolle C, Ziegler M (2014) Regulation of SIRT2‐dependent alpha‐tubulin deacetylation by cellular NAD levels. DNA Repair 23: 33–38 [DOI] [PubMed] [Google Scholar]
  • 35. Gu W, Roeder RG (1997) Activation of p53 sequence‐specific DNA binding by acetylation of the p53 C‐terminal domain. Cell 90: 595–606 [DOI] [PubMed] [Google Scholar]
  • 36. Liu L, Scolnick DM, Trievel RC, Zhang HB, Marmorstein R, Halazonetis TD, Berger SL (1999) p53 sites acetylated in vitro by PCAF and p300 are acetylated in vivo in response to DNA damage. Mol Cell Biol 19: 1202–1209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 144: 646–674 [DOI] [PubMed] [Google Scholar]
  • 38. Vousden KH, Lu X (2002) Live or let die: the cell's response to p53. Nat Rev Cancer 2: 594–604 [DOI] [PubMed] [Google Scholar]
  • 39. Barlev NA, Liu L, Chehab NH, Mansfield K, Harris KG, Halazonetis TD, Berger SL (2001) Acetylation of p53 activates transcription through recruitment of coactivators/histone acetyltransferases. Mol Cell 8: 1243–1254 [DOI] [PubMed] [Google Scholar]
  • 40. Rodriguez MS, Desterro JM, Lain S, Lane DP, Hay RT (2000) Multiple C‐terminal lysine residues target p53 for ubiquitin‐proteasome‐mediated degradation. Mol Cell Biol 20: 8458–8467 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Nakamura S, Roth JA, Mukhopadhyay T (2000) Multiple lysine mutations in the C‐terminal domain of p53 interfere with MDM2‐dependent protein degradation and ubiquitination. Mol Cell Biol 20: 9391–9398 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Costanzo A, Merlo P, Pediconi N, Fulco M, Sartorelli V, Cole PA, Fontemaggi G, Fanciulli M, Schiltz L, Blandino G et al (2002) DNA damage‐dependent acetylation of p73 dictates the selective activation of apoptotic target genes. Mol Cell 9: 175–186 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Appendix

Click here for additional data file. (943.1KB, pdf)

Expanded View Figures PDF

Click here for additional data file. (1.2MB, pdf)

Source Data for Expanded View

Click here for additional data file. (6.8MB, zip)

Review Process File

Click here for additional data file. (385.1KB, pdf)

Source Data for Figure 2

Click here for additional data file. (755.5KB, pdf)

Source Data for Figure 3

Click here for additional data file. (208.9KB, pdf)

Source Data for Figure 4

Click here for additional data file. (1.7MB, pdf)

Data Availability Statement

The raw data of the mass spectrometry were deposited at jPOST (jPOST ID: JIST000414, PXID: PXD010411).

Articles from EMBO Reports are provided here courtesy of Nature Publishing Group

RESOURCES