Fig 1.
(A) Bacterially expressed and purified GST-tagged ING5 was incubated with 12.5 fkatal of the indicated purified cyclin/CDK complexes in the presence of 32P-γ-ATP. The proteins were resolved by 7–17% SDS-PAGE and visualized by autoradiography (32P, upper panel) or Coomassie blue staining (CB, lower panel). (B) HEK293 cells were transiently transfected with plasmids encoding HA-ING5 (10 μg), cyclin E/CDK2 (500 ng each), a catalytically inactive form of CDK2 (dnCDK2, 5 μg) and p27KIP1 (5 μg) as indicated. The cells were metabolically labeled with 32P-orthophosphate and ING5 was then immunoprecipitated by virtue of its HA-tag. The top panel shows an autoradiograph, the bottom panel a Western blot using mAb 3F10 (HA-tag). (C) HA-ING5 was expressed in HEK293 cells and immunoprecipitated using the ING5-specific polyclonal serum (pAb #3716) or a pre-immune serum. In addition aliquots of total cell lysates of transiently transfected HEK293 cells or of two tumor cell lines (HCT116 and U2OS) were analyzed as indicated. Parallel Western blots were developed using mAb 7A11 (ING5) and mAb 3F10 (HA-tag). (D) HEK293 cells were left untreated or treated with 200 μM hydroxyurea (HU), 400 ng/ml nocodazole (Noc) or 200 ng/ml colcemid (Col) for 18 hrs and then metabolically labeled with 32P-orthophosphate. Endogenous ING5 protein was immunoprecipitated from RIPA buffer lysates using the ING5-specific polyclonal serum (#3716). The top panel shows an autoradiograph, the bottom panel a Western blot using mAb 7A11 (ING5). The FACS profiles of the cells treated with the different cell cycle inhibitors are shown below.
Fig 2.
ING5 is phosphorylated at Thr152 by cyclin E/CDK2.
(A) Schematic drawing of the ING5 protein. The numbers refer to amino acid positions. LZL, leucine-zipper-like region (blue); NCR, novel conserved region (orange); NLS, nuclear localization signals (green, a bipartite NLS from amino acid 128–171 and a basic NLS at the C-terminus); PHD, C-terminal plant homeodomain zinc finger (yellow); the two potential CDK phosphorylation sites at S57 and at T152, and three RXL motifs, potential cyclin binding sites, are indicated [35,36,37].(B) Recombinant GST-ING5, GST-ING5-S57A, and GST-ING5-T152A were incubated with baculovirus-derived cyclin E/CDK2 and cyclin A/CDK2 in the presence of 32P-γ-ATP. For control kinase assays were performed with histone H1. CB, Coomassie blue stained gels; 32P, autoradiographies.(C) GST or GST-ING5 were incubated with cyclin E/CDK2 in the presence or absence of ATP. Western blot analysis was performed using mAb 3H5, which was generated against a phospho-peptide containing phosphorylated T152. CB, Coomassie blue stained gel.(D) ING5 or ING5-T152A were expressed transiently in HEK293 cells. ING5 proteins were immunoprecipitated using a rat monoclonal HA-antibody (3F10) or for control a rat monoclonal ASH2L antibody (4C5), treated with our without shrimps alkaline phosphatase in the presence or absence of phosphatase inhibitor as indicated. Phosphorylation at T152 was measured on Western blots using mAb 3H5. Total levels of ING5 were determined with mAb 7A11.(E) Endogens ING5 from HEK 293 cells was immunoprecipitated using an ING5 antibody (7A11) or for control an ASH2L antibody (4C5). Phosphorylation at T152 was measured on Western blots using mAb 3H5. Total levels of ING5 were determined with mAb 7A11 (left panel). ING5 and ING5-T152A were expressed in HEK293 cells as indicated and immunoprecipitated using mAb 7A11 and stained with mAb 3H5 for ING5-T152-P and with mAb 7A11 for ING5 (right panel).(F) Endogenous ING5 was immunoprecipitated from RIPA buffer lysates of HEK293 cells using a rabbit polyclonal serum (#3716), treated and analyzed as described in panel (D).
Fig 3.
Phosphorylation of ING5 at T152 in cells.
(A) The indicated proteins were expressed transiently in HEK293 cells. T152 phosphorylation was assessed using mAb 3H5. The other proteins were detected as detailed in the Material and Methods section. (B) ING5 or ING5-T152A were expressed transiently in HEK293 cells and treated with 25 μM roscovitine 2 h prior to harvesting. ING5 and ING5-P were analyzed using mAb 7A11 and 3H5, respectively. (C) ING5 or ING5-T152A were expressed transiently in HEK293 cells. The cells were then reseeded and transfected with control siRNA (siC) or siRNA targeting cyclin E (siE), cyclin A (siA) or both cyclins (siE/A). ING5 proteins were immunoprecipitated using mAB 7A11 and T152-P and total amounts of ING5 were analyzed using mAB 3H5 and 7A11, respectively. Functionality of the siRNAS were analyzed in total cell lysates (right panel). (D) The cells of the glioblastoma cell line T98A were arrested in G0 with 0.1% FCS. Cell cycle re-entry was stimulated by reseeding and stimulating with 10% FCS. The expression of the indicated proteins was analyzed at the indicated time points after G0 release. T152 phosphorylation was measured using mAb 3H5.(E) MCF7 cells were released from a fulvestrant block (1 μM, 72 h) by adding a 10-fold excess of tamoxifen. The expression of the indicated proteins was analyzed at the indicated time points after addition of tamoxifen. Endogenous ING5 was immunoprecipitated (pAb 3716) and analyzed using mAB 3H5 and 7A11.
Fig 4.
Phosphorylation of T152 does not affect subcellular localization of ING5.
HA-tagged ING5 and ING5 phospho-site mutants, with and without a deleted C-terminal basic NLS (see Fig 2), were expressed in HeLa cells. ING5 was detected by indirect immunofluorescence staining using mAb 3F10.
Fig 5.
Overexpression of ING5 has only a minor effect on cell proliferation.
(A) HCT116 cells were co-transfected with plasmids expressing the indicated proteins and a puromycin resistance plasmid. The cells were selected in puromycin and the growth of colonies evaluated 10 days after transfection by staining with methylene blue. Mean values and standard deviations of 5 independent experiments is depicted on the right.(B) As in panel (A) but with HCT116-p53-/- cells. (C) ING5 and mutants were transiently expressed in HCT116 cells and the ING5 protein levels measured with mAb 7A11. Actin is analyzed for control. (D and E) As in panel (A) but with U2OS and HeLa cells. Three independent experiments were quantified for both cell lines. (F) HCT116 cells were co-transfected with plasmids expressing the indicated proteins and a plasmid encoding GFP-tubulin. Two days after transfection, the cells were stained with Vybrant and the cell cycle distribution of the transfected cells measured by FACS. Quantification of 3 experiments is shown at the bottom. (G) As in panel (F) but with HCT116-p53-/- cells. * p<0.05, ** p<0.01.
Fig 6.
Knockdown of ING5 inhibits cell proliferation in tumor cells independent of p53.
(A) HEK293 cells were transfected with the indicated pSuper constructs. The expression of ING5 was analyzed using mAb 7A11. Actin is shown for control. (B) HCT116 cells were co-transfected with plasmids expressing the indicated shRNAs and a puromycin resistance plasmid. The puromycin selected cells were evaluated 10 days after transfection by staining with methylene blue. (C-F) Quantification of 3 independent experiments with HCT116 (C), HCT116-p53-/- (D), U2OS (E) and HeLa (F) cells. ** p<0.01.
Fig 7.
Knockdown of ING5 induces apoptosis independent of p53.
(A) HCT116 cells were co-transfected with the indicated pSuper constructs and a plasmid encoding GFP-tubulin. The cells were treated with or without Z-VAD. The cells were stained for annexin V and with propidium iodide and analyzed by FACS 48 h after transfection. GFP-positive cells were gated. (B) Quantification of experiments with HCT116 cells. Mean values and standard deviations of 3 independent experiments are displayed for the shRNA experiments. For Z-VAD mean values of 2 experiments are shown.(C) Quantification of experiments as in panel (B) but with HCT116-p53-/- cells. * p<0.05, ** p<0.01.