Effect of CDIP inhibition on genotoxic stress- and p53-mediated apoptosis. (A) CDIP repression compromises p53-mediated apoptosis in response to genotoxic stress. U2OS cells were transfected with an shRNA vector targeting CDIP or a control vector of identical scrambled sequence (scrambled CDIP-shRNA), and treated with ETO (30 μM, 12 and 24 h). Northern blotting was performed to determine the levels of CDIP and 36B4 (upper panel). U2OS or IMR90-E1A cells were transfected with either CDIP-shRNA, scrambled CDIP-shRNA or a control shRNA (luciferase) for 24 h, and treated with ETO (30 μM) for the indicated time points. Cells were then assessed for apoptosis by TUNEL, DNA fragmentation assay or PI staining to detect the sub-G1 cell population. Both TUNEL-positive and sub-G1 PI-stained cells were enumerated by flow cytometry. The middle left panel shows a representative experiment; apoptotic cells containing 3′-hydroxyl DNA strand breaks are detected in the FITC fluorescence range (FL1-H: 515–565 nm). In the right panel, the percentage of TUNEL-positive cells is shown as the mean±s.e.m. (n=3). U2OS cells and IMR90-E1A used for the DNA fragmentation (lower left panels) are represented as the amount of extranuclear or extracellular DNA fragments relative to untreated controls. Error bars indicate ±s.d. of three independent experiments with each sample in triplicate. (B) Combined effects of Puma and CDIP depletion, and Noxa and CDIP depletion on genotoxic stress-induced apoptosis. U2OS and U2OS/shCDIP cells were transfected with siRNAs directed against Puma or Noxa, and apoptosis was assessed by TUNEL assay 40 h after treatment with 30 μM ETO (right panel, mean percent of TUNEL-positive cells, ±s.e.m. n=3). Western blot analysis of Puma and Noxa was performed for cells similarly treated to establish effective knockdown of protein expression by siRNA (right panel). (C) Effect of CDIP depletion on p53-dependent apoptosis. Saos2 cells were transfected with CDIP-shRNA or control GFP-shRNA, followed by infection with adenovirus expressing GFP (Ad-GFP) or p53 (Ad-p53) at an MOI of 20. The apoptotic response of Ad-infected Saos2 was similarly assessed by DNA fragmentation assay and PI staining, as described (lower panels).

Caspase-8 is required for CDIP-dependent apoptosis. (A) U2OS/CDIP-tet-on cells were grown in the absence (−) or presence (+) of dox and harvested at the time intervals shown. Whole-cell lysates were either directly assessed for caspase-8, caspase-9, PARP and β-actin expression by Western blot analysis (left panel), or fractionated. Active caspase-8 species detected are consistent with three N-terminal Asp cleavage fragments. Fractionates were assessed for the presence of cytochrome c (cyt.-c), oxidative complex II (OxPhos II) and α-tubulin, by Western blotting; the latter were used as mitochondria- and cytoplasm-specific markers, respectively (upper right panel). Caspase-8 enzyme activity was measured by incubating cell lysates from similarly treated U2OS/CDIP-tet-on cells with the caspase-8 substrate, Ac-IETD-pNA. DNA fragmentation was also determined by cell death ELISA in dox-treated U2OS/CDIP-tet-on cells in the absence (+DMSO) and presence of the caspase-8 inhibitor II (lower right panel). (B) Inhibition of caspase-8 by siRNA directed against caspase-8 blocks CDIP-induced apoptosis. Caspase-8 siRNA and control siRNAs were transfected into U2OS/CDIP-tet-on cells. After 24 h, cells were grown in the absence or presence of dox and harvested at the time intervals shown. CDIP, caspase-8, PARP and β-actin (as a loading control) expression were assessed by Western blot analysis. Apoptosis was assessed in U2OS/CDIP-tet-on cells similarly transfected by TUNEL assay (40 h post-dox treatment) and trypan blue exclusion (24 and 48 h post-dox treatment). The percentages of TUNEL- and trypan blue-positive cells are shown as the mean±s.e.m. (n=3). (C) Inhibition of caspase-8 by siRNA attenuates apoptosis induced by genotoxic stress. ETO induces p53, caspase-8 and PARP cleavage 24 h post-ETO (30 μM) treatment (upper left panel). U2OS cells were transfected with siRNAs directed against caspase-8 and apoptosis was assessed by TUNEL assay 40 h after ETO treatment (middle panel is a representative experiment; lower panel is the mean percent of TUNEL-positive cells, ±s.e.m. n=3). Western blot analysis of p53 and caspase-8 cells similarly treated, establishes effective knockdown of caspase-8 protein expression by siRNA (right panel).

Inhibition of TNF-α attenuates apoptosis induced by DNA damage. (A) U2OS cells were transfected with shTNF-α constructs (A and B, where indicated, otherwise an equal amount of A combined with B) and treated with ETO (30 μM, 36 h). Northern blots were performed using ³²P-labeled probes (p21^Waf1, TNF-α and 36B4). Apoptosis was assessed by TUNEL assay. A representative experiment is shown in the lower left panel; apoptotic cells are positive for dUTP-FITC (FL1-H: 515–565 nm). In the lower right panel, the percentage of TUNEL-positive cells is shown as the mean±s.e.m. (n=3). Cell viability was also determined by trypan blue exclusion (upper right panel); average values±s.e.m. (n=3). (B) IMR90-E1A cells were transfected with either control (Cont.-) or CDIP-shRNA and treated with CPT (1 μM, 24 h). Western blot analysis confirmed the activation of p53 (s15p) and upregulation of CDIP by CPT. In contrast, CDIP-shRNA-transfected cells did not upregulate CDIP. TNF-α expression was assessed by qRT-PCR in cells similarly treated (upper middle panel). CPT-induced apoptosis was assessed by TUNEL assay in control (Cont.-), CDIP- and TNF-α-shRNA-transfected cells; apoptotic cells are positive for dUTP-TMR red (lower left, FL2-H: 570–620 nm). The percentage of TUNEL-positive cells is shown as the mean±s.e.m. (n=3; lower right panel). Long-term cell viability was determined by crystal violet staining of IMR90-E1A cells similarly transfected and treated with DMSO or CPT (1 μM) for 48 h (upper right panel).

Identification of CDIP as a direct target of p53 transcriptional activation. (A) p53-dependent induction of CDIP. The left panels show induction of CDIP mRNA after infection with recombinant adenovirus expressing p53 (Ad-p53) in Saos2 and EJ cells. Northern blots were performed sequentially using a ³²P-labeled probe against CDIP, p53, p21^Waf1 and 36B4 (loading control). (B) p53-dependent induction of CDIP in response to DNA damage. Upper panel shows MCF7, IMR90-E1A, U2OS and Saos2 (p53 null) cells that were either left untreated (−) or treated with the DNA damaging agent ETO (25 μM) for 1 day (+1) or 2 days (+2). Western blot analysis with a custom CDIP antibody also shows p53-associated CDIP upregulation in U2OS cells treated with ETO (upper right panel). The lower panel shows MCF7 cells transfected with either siRNA oligos targeting wild-type p53, or control siRNA oligos (GFP), followed by exposure to ETO for 24 h. Northern blots were performed sequentially using a ³²P-labeled probe against p53, CDIP and 36B4 (loading control); Western blotting with an anti-p53 antibody confirmed that siRNA targeting of p53 resulted in the knockdown of p53 protein (lower panels). (C) A putative p53 recognition site located in the promoter region of the CDIP gene (p53-BS, set in bold, from −1618 to −1639) is shown; four such sites were identified within the CDIP promoter region. Three constructs containing these putative p53-BSs or a mutated copy of the 5′ p53-BS, as indicated, were cloned into a luciferase reporter vector (pGL3-promoter vector). Saos2 cells were cotransfected with either of the four CDIP promoter constructs (T1, T2, T3 or M1, plus pGL3 basic), and either wild-type p53, or mutant p53 (V143A) expression constructs, as well as a control pcDNA3.1 empty vector. Luciferase assay results for each construct are shown. Additionally, U2OS cells were transfected with the various CDIP promoter constructs described, treated with ETO to induce endogenous p53, and the relative luciferase activity was similarly determined. (D) ChIP assay of p53 DNA binding activity at the CDIP promoter in U2OS and Saos2 cells. U2OS and Saos2 cells were infected with Ad-p53 or Ad-GFP for 24 h (upper panel). ChIP assays were performed using an anti-p53 antibody or an anti-HA antibody as a negative control. U2OS cells were also treated with ETO for 24 h ChIP assays were performed with an anti-p53 antibody or a control anti-IgG antibody. A sequence map of the CDIP promoter region containing the ChIP primers (shown as arrows) used to amplify the p53-responsive element at −1618 to −1639 is shown (bold); the amplification product was 229 bp. Primers amplifying a 417-bp product of the p21^Waf1 promoter were used as positive control.

Induction of apoptosis by CDIP expression. (A) Apoptosis analysis of U2OS/CDIP-tet-on cells grown in the absence (−) or presence (+) of the tetracycline analogue, doxycyclin (dox). Western blots were performed using an anti-HA antibody to detect CDIP; β-actin antibody was used as a loading control (upper panel). U2OS/CDIP-tet-on cells were grown ±dox for the indicated time periods, and apoptosis was assessed by three independent cell death assays: TUNEL, DNA fragmentation by cell death ELISA and PI staining to detect the sub-G1 population. (B) Viability of U2OS/CDIP- and U2OS/GFP-tet-on cells by crystal violet staining, 40 h after the administration of dox (2 μg/ml). (C) Western blotting and cell death analyses of MCF7 cells infected with a recombinant adenovirus expressing GFP (Ad-GFP) or CDIP (Ad-CDIP) for 12 and 24 h, as indicated. The upper panel shows anti-HA detection of CDIP, and the lower panel shows β-actin levels as a loading control. Apoptotic MCF-7 cells infected with Ad-CDIP or Ad-GFP were detected by DNA fragmentation assay (lower left panel) and flow cytometric analysis of PI-stained cells (lower right panel). Error bars indicate ±s.d. of three independent experiments.

CDIP and p53 activate TNF-α, and p53 induction of TNF-α requires CDIP. (A) TNF-α mRNA increases upon induction of CDIP. U2OS/CDIP-tet-on cells grown in the absence (−) or presence (+) of dox, and MCF7 cells infected with a recombinant adenovirus expressing GFP (Ad-GFP) or CDIP (Ad-CDIP), were subjected to Northern blot analysis sequentially using ³²P-labeled probes that detect CDIP, TNF-α and 36B4, or 18S rRNA (loading controls). (B) CDIP induces TNF-α protein. U2OS/CDIP-tet-on cells were grown in the presence or absence of dox for the indicated times, and TNF-α protein was quantified by ELISA. (C) CDIP binds and activates the TNF-α promoter. CDIP binding to the TNF-α promoter at the indicated regions was assessed by ChIP in U2OS/CDIP-tet-on cells in the absence (−) and presence (+) of dox, 24 h post-treatment. Bound chromatin was immunoprecipitated with anti-HA antibody to detect CDIP and anti-rabbit IgG as a negative control. Negative (−) and positive (+) PCR controls consisted of H₂O and a pGL3-TNF-α promoter plasmid using the same amplification mixture, respectively. U2OS cells were transfected with a pcDNA3.1-CDIP expression vector (and empty vector control), together with either the pGL3-basic-promoter-luciferase construct or pGL3-TNF-α promoter-driven luciferase constructs with a wild-type promoter (wt) or a –421 to –178 deletion mutant (mut). Luciferase activity is shown relative to pGL3-luciferase basal activity. pGL3-basic-, -TNF-α (wt)- and -TNF-α (mut)-transfected cells were additionally stimulated with ETO to activate endogenous CDIP (lower panel). (D) p53- and genotoxic stress-mediated TNF-α induction is inhibited by shRNA-directed knockdown of CDIP. CDIP-shRNA vector and control shRNA vectors (scrambled CDIP-shRNA) were transfected into U2OS cells, followed by ETO treatment for 24 h. p53-null Saos2 cells were transfected with CDIP-shRNA or control scrambled CDIP-shRNA, and infected with adenovirus expressing wt-p53. The expression of CDIP, TNF-α, p21^Waf1 and 36B4 (loading control) was assessed by sequential Northern blotting with ³²P-labeled probes specific for these transcripts. (E) CDIP stabilizes TNF-α mRNA. U2OS/CDIP-tet-on cells grown ±dox were treated with actinomycin D (Act D, 8 μg/ml), and harvested at the indicated time intervals. TNF-α, TTP and 36B4 (loading control) mRNA expression was assessed by Northern blot analysis. Saos2 cells were also transfected with CDIP-shRNA or control scrambled CDIP-shRNA, and infected with adenovirus expressing wt-p53 (Ad-p53) or Ad-GFP (−). After 24 h, the actinomycin D chase and Northern blot experiments were performed as described above. (F) CDIP expression leads to an increase in TNF-α half-life. Cells were cultured ±dox, and TNF-α transcript levels were assessed by quantitative real-time RT-PCR (qRT-PCR), after treatment with actinomycin D. The amount of TNF-α transcript relative to that present after 24 h ±dox (i.e., t=0 for actinomycin treatment), was quantitated, and the half-life was determined by regression analysis.

Induction of apoptosis by CDIP involves the TNF-α death receptor signaling pathway. (A) TNF-α neutralizing antibodies block CDIP-mediated apoptosis. U2OS/CDIP-tet-on cells (±dox) were incubated with either a neutralizing monoclonal antibody to human TNF-α (8 μg/ml) or Rb IgG (8 μg/ml), and assessed for apoptosis using DNA fragmentation (cell death ELISA) and TUNEL assays; cell viability was additionally measured by trypan blue exclusion. The percentage of TUNEL-positive cells is shown as the mean±s.e.m. (n=3). (B) The effect of FADD inhibition on CDIP-mediated cell death. U2OS/CDIP-tet-on cells were transfected either pcDNA3.1-DN-FADD or empty vector, and grown ±dox for 24 h, and then analyzed for cell death using trypan blue exclusion and DNA fragmentation/cell death ELISA. Error bars indicate ±s.d. of three independent experiments.