Integral role of Noxa in p53-mediated apoptotic response

Abstract

The tumor suppressor p53 exerts its versatile function to maintain the genomic integrity of a cell, and the life of cancerous cells with DNA damage is often terminated by induction of apoptosis. We studied the role of Noxa, one of the transcriptional targets of p53 that encodes a proapoptotic protein of the Bcl-2 family, by the gene-targeting approach. Mouse embryonic fibroblasts deficient in Noxa [Noxa(-/-) mouse embryonic fibroblasts (MEFs)] showed notable resistance to oncogene-dependent apoptosis in response to DNA damage, which was further increased by introducing an additional null zygosity for Bax. These MEFs also showed increased sensitivity to oncogene-induced cell transformation in vitro. Furthermore, Noxa is also involved in the oncogene-independent gradual apoptosis induced by severe genotoxic stresses, under which p53 activates both survival and apoptotic pathways through induction of p21(WAF1/Cip1) and Noxa, respectively. Noxa(-/-) mice showed resistance to X-ray irradiation-induced gastrointestinal death, accompanied with impaired apoptosis of the epithelial cells of small intestinal crypts, indicating the contribution of Noxa to the p53 response in vivo.

Figure 1.

Effect of deficiency in Noxa on p53-mediated apoptotic response. (A,C,D) Values shown are mean ± S.D. from triplicate samples. (A) DNA damage-induced apoptosis of E1A-expressing MEFs. MEFs were infected with control (pBabe) or E1A-expressing retrovirus (pBabe-E1A), and were treated with adriamycin (0.25 μg/mL) for 24 h. Wild-type (WT) and Noxa^-/- MEFs (left), as well as Noxa^-/- and Noxa^-/-/Bax^-/- MEFs (right), were each prepared from the same litter. Results are representative of three independent experiments. (B) Cytochrome c release in E1A-expressing MEFs after adriamycin treatment. E1A-expressing MEFs were treated with adriamycin (0.25 μg/mL) for 12 h. Wild-type (WT) and Noxa^-/- MEFs were prepared from the same litter. cyt. c indicates cytochrome c. (C) Caspase activation in E1A-expressing MEFs after adriamycin treatment. E1A-expressing MEFs were treated with adriamycin (0.25 μg/mL) for indicated times. Wild-type (WT) and Noxa^-/- MEFs were prepared from the same litter. (D) Radiation-induced apoptosis of thymocytes. Thymocytes were X-ray irradiated, and 18 h later, cell death was quantitated. Wild-type (WT) and Noxa^-/- thymocytes (left) as well as Noxa^-/- and Noxa^-/-/Bax^-/- thymocytes (right) were each prepared from mice from the same litter. Results are representative of three independent experiments.

Figure 2.

Contribution of Noxa to activation of Bax and Bak. (A) Bax membrane insertion during E1A-dependent apoptosis. Control and E1A-expressing MEFs were treated with DMSO control buffer or adriamycin (0.25 μg/mL) for 6 h. A protein sample (10 μg) was separated into alkali-sensitive supernatant (S) and alkali-resistant pellet (P), and was loaded into the corresponding lane. Without stimulation, Bax molecules reside in the cytoplasm or are weakly attached to the surface of mitochondria, which are readily detached by alkali treatment. Thus, they are separated into the S fraction upon alkali extraction. When inserted into the mitochondrial membrane, Bax molecules acquire alkali resistance and are separated into the P fraction (Goping et al. 1998). The level of Bax molecules that acquired alkali resistance is lower in E1A-expressing Noxa^-/- MEFs after adriamycin treatment than in similarly treated wild-type (WT) MEFs. IB indicates immunoblot. (B) Bax oligomerization during E1A-dependent apoptosis. Control and E1A-expressing MEFs were treated with DMSO control buffer or adriamycin (0.25 μg/mL) for 6 h, and subsequently, Bax oligomerization was analyzed through BMH cross-linking. A cross-linked protein sample (40 μg) was loaded into each lane. Markers M, D, and Tr correspond to the sizes of the monomer, dimer, and trimer, respectively. (C) Bakoligomerization during E1A-dependent apoptosis analyzed as in B. Markers M and D correspond to the sizes of the monomer and dimer, respectively. (ast;) An intramolecularly oligomerized Bak monomer. (D) Colony formation of MEFs expressing E1A and Ras on medium containing methylcellulose. Wild-type (WT) and Noxa^-/- MEFs (left) as well as Noxa^-/- and Noxa^-/-/Bax^-/- MEFs (right), were each prepared from the same litter. Values shown are means ± S.D. from triplicate samples. Results are representative of three independent experiments. (^**) Noxa^-/- MEFs generated a significantly increased number of colonies compared with wild-type MEFs (P < 0.01).

Figure 3.

p21-mediated survival pathway and Noxa-mediated apoptotic pathway in MEFs with severe DNA damage. (B-D) Values shown are mean ± S.D. from triplicate samples, and results are representative of three independent experiments. (A) Noxa, p21, and Bax mRNA expressions after treatment with apoptogenic agents. MEFs were treated with etoposide (60 μM for 6 h), UV (60 J/m², 6 h prior to RNA preparation), or staurosporin (300 nM for 6 h) or were left untreated. (Etp) Etoposide; (STS) staurosporin. (B) Apoptosis of MEFs induced by DNA-damaging agents. MEFs were treated with etoposide (for 36 h), adriamycin (for 72 h), cisplatin (for 72 h), or UV (36 h prior to the cell viability assay). Wild-type (WT) and Noxa^-/- MEFs were prepared from the same litter. Contrary to the data shown here, apoptosis caused by staurosporin or an anti-Fas antibody normally occurred in Noxa^-/- MEFs (data not shown). (C) Effect of ectopic expression of p21 on etoposide- and UV-induced apoptosis. MEFs were infected with control (Ad-empty) or p21-expressing adenovirus (Ad-p21) at multiplicity of infection (m.o.i.) of 100. Twelve hours after infection, MEFs were treated with etoposide (50 μM, for 36 h) or UV (50 J/m², 36 h prior to the cell viability assay) or were left untreated. (D) Effect of ectopic expression of Noxa in the presence or absence of DNA-damaging agents. MEFs were infected with control (Ad-empty) or Noxa-expressing adenovirus (Ad-Noxa). Twelve hours after infection, MEFs were treated with etoposide (50 μM, for 36 h) or UV (50 J/m², 36 h prior to the cell viability assay) or were left untreated.

Figure 4.

Resistance to DNA damage-induced apoptosis in vivo in the absence of Noxa. (A) Radiation-induced apoptosis of cells in small intestine. Sections of the jejunum from X-ray irradiated (10 Gy) mice were analyzed. In the TUNEL assay (top), TUNEL-positive cells are identified by the green fluorescence. In HE staining (bottom), apoptotic cells in the lower third (cell position 1-7, encompassing the stem cell region) of crypts are indicated by black arrowheads, and those in the lamina propria are indicated by red arrowheads. (B) Histograms of apoptotic cells in the lower third of crypts. Approximately 150 half-crypts were scored per section. Values shown are mean ± S.D. from three sections, each obtained from different mice. (C) Frequency histograms of apoptotic cells in the lamina propria. Approximately 100 crypt-villus units were scored per section. Data shown are means from three sections each obtained from different mice. (D) Survival curve of mice after whole-body irradiation. Seven groups of littermates including 15 wild-type (WT) mice and 13 Noxa^-/- mice together with 10 p53^-/- mice were subjected to X-ray irradiation (10 Gy), and the survival of these mice for 20 d after irradiation are shown. The survival of mice was scored up to 30 d after irradiation. The survival periods (mean ± S.D.) of wild-type (WT), Noxa^-/-, and p53^-/- mice were 9.1 ± 3.3 d, 17.5 ± 9.5 d, and 10.7 ± 7.9 d, respectively. The survival periods of Noxa^-/- mice was significantly longer than those of wild-type mice or p53^-/- mice (P < 0.01 vs. wild-type, P < 0.05 vs. p53^-/-).