Effect of inositol hexakisphosphate kinase 2 on transforming growth factor beta-activated kinase 1 and NF-kappaB activation

Abstract

We previously showed that inositol hexakisphosphate kinase 2 (IHPK2) functions as a growth-suppressive and apoptosis-enhancing kinase during cell stress. Overexpression of IHPK2 sensitized ovarian carcinoma cell lines to the growth-suppressive and apoptotic effects of interferon beta (IFN-beta), IFN-alpha2, and gamma-irradiation. Expression of a kinase-dead mutant abrogated 50% of the apoptosis induced by IFN-beta. Because the kinase-dead mutant retained significant response to cell stressors, we hypothesized that a portion of the death-promoting function of IHPK2 was independent of its kinase activity. We now demonstrate that IHPK2 binds to tumor necrosis factor (TNF) receptor-associated factor (TRAF) 2 and interferes with phosphorylation of transforming growth factor beta-activated kinase 1 (TAK1), thereby inhibiting NF-kappaB signaling. IHPK2 contains two sites required for TRAF2 binding, Ser-347 and Ser-359. Compared with wild type IHPK2-transfected cells, cells expressing S347A and S359A mutations displayed 3.5-fold greater TAK1 activation following TNF-alpha. This mutant demonstrated a 6-10-fold increase in NF-kappaB DNA binding following TNF-alpha compared with wild type IHPK2-expressing cells in which NF-kappaB DNA binding was inhibited. Cells transfected with wild type IHPK2 or IHPK2 mutants that lacked S347A and S359A mutations displayed enhanced terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling staining following TNF-alpha. We believe that IHPK2-TRAF2 binding leads to attenuation of TAK1- and NF-kappaB-mediated signaling and is partially responsible for the apoptotic activity of IHPK2.

FIGURE 1

IHPK2-TRAF2 interaction.a, interaction of recombinant proteins in vitro. Presence and absence of proteins indicated by + and–, respectively. The reaction was subject to immunoprecipitation (IP) followed by immunoblot (IB) with the indicated antibodies. b, wild type NIH-OVCAR-3 cells were treated with IFN-β (100 units/ml × 24 h). Lysates were subject to immunoprecipitation followed by immunoblot with the indicated antibodies. Anti-p50 NF-κB and anti-TAK1 were included as negative controls. As additional controls, lysis buffer (LB) alone, without cell lysate, was immunoprecipitated. c, untransfected (WT) cells were treated with PBS (0h) or IFN-β as above and harvested at 0.5 and 24 h. Immunoblots indicate induction of IHPK2 and STAT1 (at 24 h) and induction of pSTAT1 (at 30 min). Cells were transiently transfected with siRNA directed against IHPK2 (siIHPK2) or scrambled siRNA (siSCR) by nucleofection. Two days later cells were treated with PBS or IFN-β (0.5 and 24 h) and subjected to immunoblot analysis. d, transiently transfected cells were treated with PBS, IFN-β (100 units/ml), or TNF-α (10 ng/ml) for 24 h and subjected to TUNEL staining. e, transiently transfected cells were grown in the presence of IFN-β (0–100 units/ml). After 4 days of growth, cells were fixed and stained, and cell number (proportional to intensity of retained SRB dye) was expressed as a percentage of PBS-treated control cells (n = 8 replicates/data point).

FIGURE 2

Effect of IHPK2 mutation on binding of IHPK2 to TRAF2.a, NIH-OVCAR-3 cells were co-transfected with IHPK2 and TRAF2 by nucleofection. Pluses and minuses indicate the presence or absence of indicated construct, respectively. Total cell protein (100 μg) from each transfected cell line was immunoprecipitated with IHPK2 mouse monoclonal antibody. Precipitate was subjected to immunoblot analysis with TRAF2 polyclonal antibody (bottom row). Immunoblot analysis demonstrated that the IHPK2 S3&4 mutations in combination (but not single mutations) caused reduced TRAF2 binding when compared with the interaction of wild type IHPK2 and all other IHPK2 mutations. b, immunoblot to demonstrate protein levels of IHPK2, TRAF2, and GAPDH in lysates of transfected cells. The experiment was performed three times with similar results.

FIGURE 3

Effect of IHPK2 mutation on phosphorylation of TAK1 and AKT.a, untransfected (WT), vector-transfected (pCXN2), and mutant-transfected cells were treated with PBS (–) or TNF-α, 15 ng/ml (+) for 1 h. Lysates were subjected to immunoblot with anti-phospho-TAK1 followed by stripping and reprobing with anti-TAK1. Expression of IHPK2 transgene is indicated, and GAPDH served as loading control. Bands were quantitated by densitometry, and -fold induction was calculated (n = 3). b, using the same lysates, similar studies were performed with anti-phospho-AKT and anti-AKT.

FIGURE 4

IKK activity. IKK activity was assessed using recombinant GST-IκBα-(1–54) and [γ-³²P]ATP as substrates. The phosphorylated GST fusion protein was detected by autoradiography. a, IKK activity was determined in NIH-OVCAR-3 cells co-transfected with IHPK2 (wild type or mutants) and TRAF2, followed by TNF-α stimulation for 15 min. Cells expressing the double mutant did not show significant differences in IKK activity compared with cells expressing wild type IHPK2. Untreated cells typically exhibit 5–10% the activity of stimulated cells (31). Expression of IHPK2 transgene is indicated, and GAPDH served as loading control. The experiment was performed three times with similar results. b, Coomassie blue-stained gel demonstrated equal loading of GST-IκBα-(1–54) substrate utilized in the kinase assay.

FIGURE 5

Effect of IHPK2 mutation upon NF-κB DNA binding activity.a, wild type untransfected NIH-OVCAR-3 cells were treated with TNF-α, and EMSA was performed. Lysates were incubated with anti-GAPDH or anti-NF-κB(p50) prior to EMSA to demonstrate supershift (ss). b, cells transfected with IHPK2 S3&4 mutations (S347A and S359A) displayed enhanced NF-κB DNA binding activity induced by TNF-α. Cells transfected with wild type IHPK2 or IHPK2 mutants that lacked S3&4 mutations had suppressed NF-κB DNA binding activity. Cells were stimulated with TNF-α (20 ng/ml) for 15 min. Expression of IHPK2 transgene is indicated, and GAPDH served as loading control. EMSA band intensities were quantified with a Phosphorimager, and -fold induction was calculated. Cells expressing IHPK2 mutants S3&4, S1&3&4, S2&3&4, and S1&2&3&4 were 7- to 11-fold more effective at activation of NF-κB compared with cells transfected with wild type IHPK2, whereas cells expressing IHPK2 and single mutants had <2-fold induction. Data are expressed as mean ± S.E. of three separate experiments.

FIGURE 6

Effect of IHPK2 mutation on XIAP levels. Immunoblotting for XIAP was performed with lysates harvested 24 h after TNF-α (15 ng/ml). Expression of IHPK2 transgene is indicated, and GAPDH served as loading control. Cells expressing IHPK2 mutations S3&4 had up to 2.4-fold increase in XIAP. XIAP was not induced in cells expressing IHPK2, or IHPK2 single mutations, demonstrating -fold increases of -0.1 to -0.9. -Fold increase <1.0 corresponds to decreased XIAP expression following TNF-α (n = 3).

FIGURE 7

Effect of IHPK2 mutation on apoptosis induction by TNF-α. Untransfected (WT), vector-transfected (pCXN2), and mutant-transfected cells were treated with PBS or TNF-α, 15 ng/ml for 16 h. Induction of apoptosis was determined by TUNEL staining.

FIGURE 8

Effect of IHPK2 mutation on IFN-β antiproliferative activity. NIH-OVCAR-3 cells were stably transfected with IHPK2 constructs and grown in the presence of 5–500 units/ml IFN-β. After 4 days, cells were fixed and stained with sulforhodamine B. Absorbance of bound dye was expressed as percent of untreated controls (n = 8, each data point). Overexpression of IHPK2 sensitized cells to IFN-β. Cells expressing IHPK2 mutations S3&4 in combination conferred relative resistance to IFN-β. The dose response curve of cells expressing irrelevant IHPK2 mutations (S1, S2, S3, S4, S1&2, S1&3, S1&4, S2&3, S2&4, S1&2&3, S1&2&4) lay in a narrow band between pCXN2 (vector) and IHPK2 and is represented as a single curve (IRREL).