ASK1 and ASK2 differentially regulate the counteracting roles of apoptosis and inflammation in tumorigenesis

Abstract

Apoptosis and inflammation generally exert opposite effects on tumorigenesis: apoptosis serves as a barrier to tumour initiation, whereas inflammation promotes tumorigenesis. Although both events are induced by various common stressors, relatively little is known about the stress-induced signalling pathways regulating these events in tumorigenesis. Here, we show that stress-activated MAP3Ks, ASK1 and ASK2, which are involved in cellular responses to various stressors such as reactive oxygen species, differentially regulate the initiation and promotion of tumorigenesis. ASK2 in cooperation with ASK1 functioned as a tumour suppressor by exerting proapoptotic activity in epithelial cells, which was consistent with the reduction in ASK2 expression in human cancer cells and tissues. In contrast, ASK1-dependent cytokine production in inflammatory cells promoted tumorigenesis. Our findings suggest that ASK1 and ASK2 are critically involved in tumorigenesis by differentially regulating apoptosis and inflammation.

Figure 1

ASK2 deficiency promotes chemically induced skin tumorigenesis. (A) Normal tissue distribution of ASK2 in mice. Cell lysates were obtained from the indicated mouse organs and levels of protein expression of ASK1 and ASK2 were detected by immunoblotting. (B) Expression of ASK2 mRNA in mouse skin. ASK2 mRNA was detected by in situ hybridisation of adult mouse skin with anti-sense (AS) and sense (S) riboprobes. Scale bar=100 μm. (C) ASK2 protein is highly expressed in primary keratinocytes. ASK2 protein was detected by immunoblotting in various cultured mouse cells. RAW264.7, a macrophagic cell line; Keratinocytes, primary keratinocytes; MSF, mouse skin fibroblasts. (D) Chemically induced skin tumorigenesis is promoted in ASK2^−/− mice. Wild-type (WT; n=15), ASK2^+/− (n=7) and ASK2^−/− mice (n=22) were treated once with DMBA (100 μg in 200 μl of acetone) and then continually treated with TPA (10 μg in 200 μl of acetone) twice a week for 20 weeks. The percentage of mice with papillomas (left) and the average number of papillomas per mouse (right) are shown.

Figure 2

ASK2 facilitates DMBA-induced apoptosis in primary keratinocytes. (A) ASK2 is activated in response to DMBA. HEK293 cells transfected with HA-ASK2 in combination with (+) or without (−) HA-FKP3-ASK1-kinase-negative (KN) were treated (+) or not (−) with 100 μM DMBA for 1 h. Cell lysates were subjected to immunoblotting. (B) Immunoblotting of primary keratinocytes treated with 100 μM DMBA for the indicated periods. Phospho-ASK antibody recognised activation states of both ASK1 and ASK2 (p-ASK1/2). (C) ASK2^−/− primary keratinocytes were more resistant to the toxic effects of DMBA than WT cells. WT and ASK2^−/− keratinocytes were treated with DMSO or 100 μM DMBA for 8 h. Live and dead cells were simultaneously stained with 0.5 mM calcein AM and 0.5 mM EthD-1, respectively. Percentage of dead cells (at least 200 cells were counted for each group) is shown. Data are mean±s.e.m. (n=3). ^**P<0.01, compared with DMBA-treated WT cells. (D) DMBA-induced apoptosis is suppressed in ASK2^−/− primary keratinocytes. DNA fragmentation in WT and ASK2^−/− primary keratinocytes treated with 100 μM DMBA for the indicated periods was quantified. Data are presented as the fold increases relative to the value of unstimulated cells and are mean±s.e.m. (n=3). ^*P<0.05, compared with WT cells treated with DMBA for 8 h.

Figure 3

ASK2 facilitates DMBA-induced apoptosis in the epidermis. (A) Immunohistochemical staining of DMBA-treated skin sections with an antibody against active caspase-3. Skin samples were taken 24 h after the application of acetone (DMBA (−)) or 100 μg DMBA. Images are representatives of three independent experiments. Arrows indicate active caspase-3-positive cells. Scale bar=100 μm. (B) Quantification of active caspase-3-positive cells. The number of active caspase-3-positive cells in the hair follicles and interfollicular epidermis was counted over a linear distance of approximately 15 mm, and averaged for each 1-mm interval. Values are the mean±s.e.m. (n=3). ^**P<0.01, compared with DMBA-treated WT mice. (C) TUNEL staining of sections as in (A). Arrows indicate TUNEL-positive cells. Scale bar=100 μm. (D) Quantification of TUNEL-positive cells. The number of TUNEL-positive cells was counted and averaged as in (B). Values are the mean±s.e.m. (n=3). ^**P<0.01, compared with DMBA-treated WT mice.

Figure 4

ASK2 is critically involved in the signalling evoked by ROS-producing carcinogenic stimuli. (A) DMBA induces intracellular ROS production. HaCaT cells were pretreated with 1 mM N-acetylcysteine (NAC) or 20 μM propyl gallate (PG) for 15 min and then stimulated with DMSO or 100 μM DMBA for 30 min. ROS production was measured by FACS analysis and the percentage of ROS-producing cells was calculated. Values are the mean±s.e.m. (n=5). ^**P<0.01. (B) DMBA-induced activation of ASK2 is inhibited by antioxidants. 293-ASK2/1KN cells were pretreated with (+) or without (−) 20 μM PG for 15 min and then stimulated with 100 μM DMBA or 0.3 mM H₂O₂. Cell lysates were subjected to immunoblotting. (C) Immunoblotting of WT primary keratinocytes pretreated with antioxidants (20 μM PG or 1 mM NAC) for 15 min and stimulated with 100 μM DMBA or 0.3 mM H₂O₂ for the indicated periods. (D) DMBA-induced apoptosis in primary keratinocytes is reduced by antioxidants. DNA fragmentation in WT primary keratinocytes pretreated with 20 μM PG or 1 mM NAC for 15 min and stimulated with (+) or without (−) 100 μM DMBA for 8 h. Data are presented as the fold increases relative to the value of untreated cells and are mean±s.e.m. (n=4). ^*P<0.05. (E) UVA-induced activation of ASK2 is mediated by ROS. 293-ASK2/1KN cells were pretreated with (+) or without (−) 20 μM PG for 15 min and then stimulated with 80 kJ/m² UVA for the indicated periods. Cell lysates were subjected to immunoblotting. (F) UVA-induced activation of JNK and p38 is reduced in ASK2^−/− primary keratinocytes. WT and ASK2^−/− primary keratinocytes were treated with 80 kJ/m² UVA for the indicated periods. Cell lysates were subjected to immunoblotting. (G) UVA-induced cell death is reduced in ASK2^−/− primary keratinocytes. Viability of WT and ASK2^−/− primary keratinocytes treated with 80 kJ/m² UVA was measured by MTT assay. Data are presented as the percentage of viable cells and are mean±s.e.m. (n=3). ^*P<0.05, ^**P<0.01, compared with WT cells.

Figure 5

ASK2 expression is reduced in human cancer cell lines and tissues. (A) Quantification of levels of ASK2 mRNA expression in various human cancer cell lines. Real-time RT–PCR of ASK2 mRNA in cancer cell lines and respective normal tissues was carried out. After normalisation to GAPDH, levels of ASK2 mRNA expression in each human cancer cell line are indicated as fold increase relative to that in the respective normal tissue. (B) mRNA and protein expression of ASK2 in esophageal cancer cell lines, KYSE30, KYSE110, KYSE850 and KYSE1250. Data of mRNA expression of ASK2 was extracted from Figure 5A. Protein expression of ASK2 was detected by immunoblotting. (C, D) Immunohistochemical staining of ASK2 protein in esophageal cancer specimens. Representative ASK2-negative (#22) and ASK2-positive (#84) cases are shown in (C). The boundary of the normal tissue (N) and ESCC in an ASK2-negative case (#16) is shown in (D). LP, lamina propria. Scale bar=100 μm.

Figure 6

Tumour-promoting activity of ASK1 counteracts with the proapoptotic activity of ASK1 and ASK2 in keratinocytes. (A) DMBA-induced activation of JNK and p38 is reduced in ASK1^−/− primary keratinocytes. Immunoblotting of WT and ASK1^−/− primary keratinocytes treated with 100 μM DMBA for the indicated periods. (B) DMBA-induced apoptosis is suppressed in ASK1^−/− primary keratinocytes to a similar level to that in ASK2^−/− cells. DNA fragmentation in WT, ASK1^−/− and ASK2^−/− primary keratinocytes treated with 100 μM DMBA for 8 h was quantified. Data are presented as the fold increases relative to the value of respective unstimulated cells and are mean±s.e.m. (n=4). ^**P<0.01, compared with WT cells. (C) Two-stage skin tumorigenesis induced by DMBA and TPA in WT (n=14), ASK2^−/− (n=10), ASK1^−/− (n=9) and ASK1^−/−; ASK2^−/− (n=11) mice. The average number of papillomas per mouse is shown.

Figure 7

ASK1-dependent inflammatory response is required for tumour promotion. (A) TPA-induced inflammatory response is reduced in ASK1^−/− mice. The dorsal skin of WT, ASK2^−/− and ASK1^−/− mice was treated twice with acetone (TPA (−)) or TPA (10 μg each) with 24 h interval. Mice were killed 48 h after the latter treatment, and skin sections were processed for hematoxylin-eosin staining. Images are representatives of nine mice for each genotype. Scale bar=100 μm. Macroscopic observations of the dorsal skin of TPA-treated mice are also shown (left top panels). Quantification of epidermal thickness of these mice is shown in a graph on the right. Values are the mean±s.e.m. (n=9). ^*P<0.05, ^**P<0.01, compared with TPA-treated WT mice. (B) TPA-induced activation of p38 is attenuated in ASK1^−/− skin. Acetone and 10 μg TPA were independently applied to separate areas of the dorsal skin of the same mouse. Mice were killed at 24 h after treatment, and each treated region of skin was excised and lysed and then subjected to immunoblotting. (C) Induction of TNF-α and IL-6 is attenuated in TPA-treated skin of ASK1^−/− and ASK1^−/−;ASK2^−/− mice. The dorsal skin of WT, ASK2^−/−, ASK1^−/− and ASK1^−/−;ASK2^−/− mice was treated with acetone or 10 μg TPA. Mice were killed 24 h after the treatment, and RNA was extracted from the treated skin. mRNA expression of TNF-α and IL-6 was quantified using real-time RT–PCR. Data are mean±s.e.m. (n=4 for TPA-treated mice and n=3 for acetone-treated mice). (D) H₂O₂-induced activation of p38 is reduced in bone marrow-derived macrophages (BMDMs) from ASK1^−/− mice, but not from ASK2^−/− mice. WT, ASK2^−/− and ASK1^−/− BMDM were treated with 0.3 mM H₂O₂ for the indicated periods. Cell lysates were subjected to immunoblotting. (E) Comparison of ASK1 and ASK2 protein expression among primary cultured mouse cells. Cell lysates from primary keratinocytes, mouse skin fibroblasts (MSF) and BMDM were subjected to immunoblotting.

Figure 8

A schema of the roles of ASK family kinases in two-stage skin tumorigenesis. See Discussion for details.