Inositol hexakisphosphate kinase 2 sensitizes ovarian carcinoma cells to multiple cancer therapeutics

Abstract

We recently identified inositol hexakisphosphate kinase 2 (IP6K2) as a positive regulator of apoptosis. Overexpression of IP6K2 enhances apoptosis induced by interferon-beta (IFN-beta) and cytotoxic agents in NIH-OVCAR-3 ovarian carcinoma cells. In this study, we contrast and compare IFN-beta and radiation-induced death, and show that IP6K2 expression sensitizes tumor cells. Unirradiated NIH-OVCAR-3 cells transfected with IP6K2 formed fewer colonies compared to unirradiated vector-expressing cells. IP6K2 overexpression caused increased radiosensitivity, evidenced by decreased colony forming units (CFU). Both IFN-beta and radiation induced caspase 8. IFN-beta, but not gamma-irradiation, induced TRAIL in NIH-OVCAR-3 cells. Gamma irradiation, but not IFN-beta, induced DR4 mRNA. Apoptotic effects of IFN-beta or gamma-irradiation were blocked by expression of a dominant negative mutant death receptor 5 (DR5Delta) or by Bcl-2. Caspase-8 mRNA induction was more pronounced in IP6K2-expressing cells compared to vector-expressing cells. These data suggest that overexpression of IP6K2 enhances sensitivity of some ovarian carcinomas to radiation and IFN-beta. IP6K2 may function to enhance the expression and/or function of caspase 8 and DR4 following cell injury. Both IFN-beta and gamma-irradiation induce apoptosis through the extrinsic, receptor-mediated pathway, IFN-beta through TRAIL, radiation through DR4, and both through caspase 8. The function of both death inducers is positively regulated by IP6K2.

Figure 1

Effect of γ-irradiation and IFN-β on IP6K2 protein levels. (a) NIH-OVCAR-3 cells growing at 75% confluence were trypsinized, suspended in 1 ml PBS, and subject to 0, 1 (white bars) or 2 Gy (gray bars) irradiation and replated in complete medium. At 8 and 16 h after plating, cells were harvested, lysed and subject to Western blot analysis using monoclonal anti-IP6K2 antibodies. To control for loading, blots were probed with anti-actin antibodies; bands were digitized and quantified. IP6K2 signal intensity was normalized to actin, then expressed as fold induction, with unirradiated cells representing a protein level of 1. Purified rIP6K2 was included as a positive control (lane 1). Numbers to right of blots indicate MW markers. Only relevant portion of the blots are shown. (b) NIH-OVCAR-3 cells were grown in the presence of 200 U/ml IFN-β for 0 – 48 h and then subject to Western blot analysis and quantified as above

Figure 2

IP6K2 constructs. (a) Diagrammatic representation of IP6K2 protein. The putative inositol phosphate binding domain (IPBD) is shown in gray. SUB, the IP6K2 dominant negative substitution mutant, contains seven point mutations within the IPBD. SUB expression inhibits cell death and IP6K enzymatic activity in NIH-OVCAR-3 cells (Morrison et al., 2001). (b) Western blot of myc-tagged IP6K2 proteins. Stably transfected NIH-OVCAR-3 cells were probed with anti-myc to verify expression of full length (FL) IP6K2 or mutants depicted above. No myc signal was detected in vector-transfected cells (V). Numbers to left of blot indicate MW markers

Figure 3

Clonogenic assay: SUB mutant, Bcl-2, and DR5Δ are radioprotective. (a) NIH-OVCAR-3 cells stably expressing pCXN2 vector or IP6K2 received 0, 2 or 4 Gy irradiation and were plated at a density of 5000 cells per 10 cm dish and grown for 30 days. Colonies were fixed with TCA and stained with sulforhodamine B. Representative plates from each treatment group are shown. (b) Colonies were identified and counted using the Kodak 1DS system. Average colony size was smaller in IP6K2 expressing cells compared to vector transfected cells at equivalent dose levels (not shown). CFU were determined for the indicated NIH-OVCAR-3 cells stably expressing various IP6K2 mutants and indicated proteins. Compared to pCXN2 vector-expressing cells, IP6K2 and B-region expressing cells displayed greatly reduced CFU. In contrast, SUB, Bcl-2, and DR5Δ expression caused increased CFU, whereas A-region, C-region CFU were no different from vector control

Figure 4

Antiproliferative effects of IFN-β are TRAIL-dependent. (a) NIH-OVCAR-3 cells stably expressing pCXN2, IP6K2, Bcl-2, or DR5Δ were grown in 96-well plates for 7 days in the presence of 0–500 U/ml IFN-β. At the end of the growth period, cells were fixed and stained with sulforhodamine B. Absorbance (570 nm) of bound dye was measured and expressed as a per cent of untreated controls. Each data point represents mean ± s.e. of eight replicates. Negative values on the γ-axis indicate death of initially plated cells. (b) Combination γ-irradiation and IFN-β treatment. Cells received 0.2, 2, or 4 Gy and were grown in the presence of 20, 200, or 400 U/ml IFN-β. After 7 days cells were fixed and growth was expressed as a percentage of control (untreated) cells. Inset graph represents combination index (C.I.) plotted as a function of cell fraction affected (FA). No growth inhibition corresponds to FA = 0 and total growth inhibition corresponds to FA = 1. C.I. is calculated for each of the three dose levels. C.I. = 1 indicates that antiproliferative effects exerted by γ-irradiation and IFN-β treatment are additive

Figure 5

Ribonuclease protection assay for death-associated genes. (a) Ribonuclease protection assay for caspase 8 and TRAIL mRNAs. Untransfected NIH-OVCAR-3 cells were cultured in the presence of 20 or 200 U/ml IFN-β for 8 h. Some cells received 10 µg/ml actinomycin D for 1 h prior to IFN-β treatment. Total RNA was harvested and analysed by RPA. L32 ribosomal RNA indicates equal amounts of RNA have been loaded. Only the relevant portions of the autoradiogram are shown. (b) NIH-OVCAR-3 cells expressing pCNX2 or IP6K2, and Hey cells, received 0, 2, or 4 Gy prior to harvesting total RNA. Transfected plasmids, dose of radiation, and times of harvest are indicated at the top of gel. Equal amounts of RNA (10 µg) were subject to ribonuclease protection assay after hybridization to a set of commercially available death associated gene probes labeled with ³²P. Protected RNAs were resolved on polyacrylamide gels, subject to autoradiography and digitized by phosphorimager. The positions of death-associated genes and controls (ribosomal protein L32, and glyceraldehyde-3-phosphate dehydrogenase, GAPDH) are indicated. A negative control reaction (yeast tRNA) showed no evidence of these bands (data not shown). (b) Bands representing induced caspase 8 and DR4 mRNAs were quantitated by normalizing their intensities to L32 band densities. These band intensity ratios are expressed as fold induction for vector-transfected (white bars), and IP6K2-transfected (gray bars) cells

Figure 6

Caspase 8 activity is increased by γ-irradiation and IFN-β. NIH-OVCAR-3 cells expressing pCNX2 (white bars) or IP6K2 (gray bars) were exposed to (a) 2 Gy or (b) 200 U/ml IFN-β and were assayed for caspase 8 activity at 0 – 72 h. Bars represent caspase 8 activity ± s.e.m. of three independent experiments. Enzyme activities were normalized to values obtained with untreated cells (0 h) and expressed as fold induction for each cell type. Column 1 in each graph represents caspase 8 activity in the presence of IETD-fmk (Inhibitor)