Non-apoptotic function of caspases in a cellular model of hydrogen peroxide-associated colitis

Abstract

Oxidative stress, caused by reactive oxygen species (ROS), is a major contributor to inflammatory bowel disease (IBD)-associated neoplasia. We mimicked ROS exposure of the epithelium in IBD using non-tumour human colonic epithelial cells (HCEC) and hydrogen peroxide (H2 O2 ). A population of HCEC survived H2 O2 -induced oxidative stress via JNK-dependent cell cycle arrests. Caspases, p21(WAF1) and γ-H2AX were identified as JNK-regulated proteins. Up-regulation of caspases was linked to cell survival and not, as expected, to apoptosis. Inhibition using the pan-caspase inhibitor Z-VAD-FMK caused up-regulation of γ-H2AX, a DNA-damage sensor, indicating its negative regulation via caspases. Cell cycle analysis revealed an accumulation of HCEC in the G1 -phase as first response to oxidative stress and increased S-phase population and then apoptosis as second response following caspase inhibition. Thus, caspases execute a non-apoptotic function by promoting cells through G1 - and S-phase by overriding the G1 /S- and intra-S checkpoints despite DNA-damage. This led to the accumulation of cells in the G2 /M-phase and decreased apoptosis. Caspases mediate survival of oxidatively damaged HCEC via γ-H2AX suppression, although its direct proteolytic inactivation was excluded. Conversely, we found that oxidative stress led to caspase-dependent proteolytic degradation of the DNA-damage checkpoint protein ATM that is upstream of γ-H2AX. As a consequence, undetected DNA-damage and increased proliferation were found in repeatedly H2 O2 -exposed HCEC. Such features have been associated with neoplastic transformation and appear here to be mediated by a non-apoptotic function of caspases. Overexpression of upstream p-JNK in active ulcerative colitis also suggests a potential importance of this pathway in vivo.

Keywords: ATM degradation; DNA-damage checkpoints; JNK-dependent cell cycle arrests; hydrogen peroxide-associated colitis; inflammation; neoplastic transformation; non-apoptotic caspase function; γ-H2AX.

Fig. 1

H₂O₂ treatment activates DNA-damage checkpoints through the JNK pathway. (A) Cell cycle analysis of H₂O₂-treated HCEC showed S arrest after 24 hrs, G₂/M arrest after 48 and 72 hrs and apoptosis after 72 hrs. The data are representative of three independent experiments. (B) Activation of the JNK pathway with p-JNK and p-c-jun (Ser63/Ser73) and up-regulation of p21^WAF1. (C) H₂O₂ did not lead to significant activation of the MAPK's p38 and ERK in HCEC. (D) Abrogation of cell cycle arrest following JNK inhibition after 24 to 72 hrs.

Fig. 2

Identification of cellular JNK-regulated proteins in H₂O₂-exposed HCEC. (A) Immunoblot analysis following JNK inhibition by the JNK inhibitor SP600125 revealed p21^WAF1, γ-H2AX, as well as caspases 3, 8 and 9 as cellular JNK-regulated proteins. The effective inhibition of JNK was ensured through missing phosphorylation of the transcription factor c-jun at serine residues 63 and 73. (B) Immunoblot analysis showed an overexpression of the death receptor TRAIL-R1/DR4 after H₂O₂. (C) Immunoblot analysis of subcellular fractionated cellular proteins of H₂O₂-treated HCEC. Each extract was analysed using specific antibodies against proteins from various cellular compartments, including cytoplasmic (GAPDH, HSP90), plasma membrane (TRAIL-R1/DR4), nuclear soluble (Sp1) and chromatin-bound (histone 3). (D) Analysis of the membrane extracts of HCEC and H₂O₂-exposed HCEC by immunoblotting showed accumulation of the TRAIL-R1/DR4 in H₂O₂-exposed cells.

Fig. 3

Non-apoptotic function of caspases through their role in cell cycle regulation. (A) Cell cycle analysis of H₂O₂-treated HCEC following pre-incubation with the pan-caspase inhibitor Z-VAD-FMK showed more cells in the G₁-phase (24 hrs) and fewer cells in the G₂/M-phase, but more cells in the G₁- and S-phase after 48 hrs. After 72 hrs, an apoptotic cell population (cell debris) increased significantly. The data are representative of three independent experiments. (B) Immunoblot analysis of H₂O₂-exposed HCEC, pre-treated with the pan-caspase inhibitor Z-VAD-FMK, revealed a caspase-mediated negative regulation of γ-H2AX and a positive regulation of the JNK pathway, including p-JNK and p-c-jun (Ser63/Ser73), after 24 hrs.

Fig. 4

Caspase-dependent degradation of the ATM-kinase. (A) Immunoblot analysis showed H₂O₂-induced cleavage of ATM after 24, 48 and 72 hrs. (B) This cleavage was shown to be reversible after 24 hrs following caspase inhibition (Z-VAD-FMK).

Fig. 5

Undetected DNA-damage is linked to γ-H2AX down-regulation. (A) Immunoblot analysis of altered HCEC^patH2O2C3 showed a decreased ratio of γ-H2AX to H2AX, as well as a decreased ATM level in altered HCEC compared with the control. (B) Comet assay revealed damaged DNA of HCEC^patH2O2C3 as enlarged nuclei and comet tails were detected. Untreated HCEC served as control. (C) Altered HCEC showed down-regulation of the negative cell cycle regulator p21^WAF1 and up-regulation of the positive cell cycle regulators CDK6 and cyclin D2. Furthermore, an up-regulation of the oncogenic transcription factors c-myc and c-fos as well as of β-catenin and TRAIL-R1/DR4 was detected. Caspases 3, 8 and 9 were found to be down-regulated. (D) The repeated exposure of HCEC to H₂O₂ led to increased cell proliferation: after 7 days, the number of cells was higher in HCEC^patH2O2C1 to HCEC^patH2O2C3 compared with HCEC. Data indicate mean ± SEM and were obtained from four independent experiments. ***P < 0.001.

Fig. 6

Restoration of the normal HCEC phenotype through caspase inhibition. (A) Phase contrast micrographs showed that treatment of HCEC^patH2O2C3 with the caspase inhibitor Z-VAD-FMK led to the restoration of the morphological HCEC phenotype after 24 hrs. (B) FITC-Phalloidin-staining of HCEC^patH2O2C3, HCEC^patH2O2C3 treated with Z-VAD-FMK and HCEC. Filopodia are marked. (C) HCEC^patH2O2C3 were treated with caspase inhibitor Z-VAD-FMK, and protein expression of γ-H2AX and p21^WAF1 was analysed after 72 hrs. An increase in both γ-H2AX and p21^WAF1 expression was detected. (D) Cell numbers of HCEC, HCEC^patH2O2C3 and Z-VAD-FMK-treated HCEC^patH2O2C3 after 72 hrs. The data are representative of three independent experiments. (E) Determination of percental filopodia-containing cells of HCEC and HCEC^patH2O2C3 and of Z-VAD-FMK-treated HCEC^patH2O2C3 after 24 and 48 hrs. Data indicate mean ± SEM and were obtained from two independent experiments. *P < 0.05, **P < 0.01. (F) Immunohistochemical analysis of p-JNK in normal colonic mucosa, active ulcerative colitis (UC) and UC in complete remission.

Fig. 7

Proposed model of how the non-apoptotic function of caspases allows survival and proliferation of colonic epithelial cells in H₂O₂-associated colitis during oxidative stress. (A) H₂O₂ leads to the induction of the receptor TRAIL-R1/DR4, which, in turn, activates the JNK pathway including p21^WAF1, γ-H2AX and caspases. This results in S and G₂/M cell cycle arrest. However, caspases are involved in ATM degradation, and this leads to γ-H2AX down-regulation. Caspase activity circumvents G₁/S- and intra-S checkpoint control by override and progression of cells through the G₁- and S-phase (survival). (B) Inhibition of caspase activity cause γ-H2AX up-regulation and accumulation of cells in G₁, S, as well as in debris (apoptosis). (C) Inflammation-associated ROS cause activation of DNA-damage checkpoints. However, caspase activity causes reduced total γ-H2AX, which leads to undetected accumulation of DNA-damage and increased proliferation of altered HCEC.