Protein kinase Ciota is required for Ras transformation and colon carcinogenesis in vivo

Abstract

Protein kinase C iota (PKCiota) has been implicated in Ras signaling, however, a role for PKCiota in oncogenic Ras-mediated transformation has not been established. Here, we show that PKCiota is a critical downstream effector of oncogenic Ras in the colonic epithelium. Transgenic mice expressing constitutively active PKCiota in the colon are highly susceptible to carcinogen-induced colon carcinogenesis, whereas mice expressing kinase-deficient PKCiota (kdPKCiota) are resistant to both carcinogen- and oncogenic Ras-mediated carcinogenesis. Expression of kdPKCiota in Ras-transformed rat intestinal epithelial cells blocks oncogenic Ras-mediated activation of Rac1, cellular invasion, and anchorage-independent growth. Constitutively active Rac1 (RacV12) restores invasiveness and anchorage-independent growth in Ras-transformed rat intestinal epithelial cells expressing kdPKCiota. Our data demonstrate that PKCiota is required for oncogenic Ras- and carcinogen-mediated colon carcinogenesis in vivo and define a procarcinogenic signaling axis consisting of Ras, PKCiota, and Rac1.

Figure 1.

PKCι is elevated mouse and human colon tumors. (a) Total protein lysates and (b) total RNA extracts were prepared from AOM-induced mouse colon tumors and uninvolved colonic epithelium from the same animals as described previously (Gökmen-Polar et al., 2001). (a) Immunoblot analysis for PKCι and actin. (b) RT-PCR analysis for mouse PKCι and actin mRNA. Lanes N1–N4 indicate uninvolved mouse colonic epithelium; T1–T4 indicate mouse colon tumors. (c) Immunoblot analysis of lysates from colon tumor tissue and matched, uninvolved colonic epithelium from five patients with colon carcinoma. Equal amounts of protein (50 μg) were subjected to immunoblot analysis for PKCι and actin. Lanes N1–N5 indicate uninvolved human colonic epithelium; lanes T1–T5 indicate matched human colon tumors.

Figure 2.

PKCι expression is elevated in AOM-induced colon tumors. Immunohistochemical analysis of sections from (a and c) normal, uninvolved epithelium and (b and d) an AOM-induced colon tumor from the same animal was performed using a specific PKCι antibody in the (a and b) absence or (c and d) presence of competing PKCι peptide as described in Materials and methods. Bars, 50 μm.

Figure 3.

Transgenic caPKCι mice are highly susceptible to AOM-induced colon carcinogenesis. (a and b) Total protein lysates from scraped colonic epithelium from nontransgenic (Ntg) and transgenic (a) caPKCι (CA) or (b) kdPKCι (KD) mice were analyzed for PKCι protein (a and b, top) and PKCι activity by immunoprecipitation histone kinase assay (a and b, bottom). (c) Colons from AOM-treated mice were scored for ACF. CA/CA, homozygous caPKCι mice; CA/+, heterozygous caPKCι mice; Ntg, nontransgenic mice; KD/KD, homozygous kdPKCι. Results represent the average ACF/animal ± SEM (n = 4–9; *P = 0.05 vs. Ntg; **P = 0.02 vs. Ntg). (d) H&E-stained section of a tubular adenoma from a nontransgenic mouse. (e) H&E-stained section of a carcinoma in situ from a caPKCι mouse. (d and e) Bars, 100 μm.

Figure 4.

PKCι is required for oncogenic Ras-induced Rac1 activation and invasion in vitro. (a) RIE cells were stably transfected with empty vector (RIE), Ras (RIE/Ras), Ras and wtPKCι (RIE/Ras/wtPKCι), or Ras and dnPKCι (RIE/Ras/kdPKCι). Total cell lysates were subjected to immunoblot analysis for PKCι (top), oncogenic V12 Ras (second from top) and β-actin (third from top). Immunoprecipitates using a specific PKCι antibody were analyzed for PKCι expression (fourth from top) and PKCι activity (fifth from top). Anti-FLAG immunoprecipitates were analyzed PKCι expression (second from bottom) and PKCι activity (bottom). (b) Anchorage-dependent growth of RIE cells and RIE cell transfectants. Data represent the mean ± SD from three independent determinations. (c) Active (GTP bound) Rac1 was isolated from the indicated RIE cell transfectants: control empty vector; Ras; Ras and RacN17; Ras and kdPKCι; and Ras and kdPKCι and RacV12. Immunoblot analysis was performed for active Rac1 (top), total cellular Rac1 (middle), and actin (bottom). The asterisk indicates the migration of Myc-tagged, virally expressed Rac1 mutants. (d) The indicated RIE transfectants were assayed for invasion. Data represent the average number of cells invading into the bottom chamber ± SD from three independent experiments. *P = 0.02 versus RIE + control vector; **P < 0.02 versus RIE/Ras; ***P = 0.005 versus RIE/Ras/kdPKCι.

Figure 5.

Expression of dnPKCι blocks Ras-mediated transformation of the intestinal epithelium in vitro and in vivo. (a and b) RIE cells were stably transfected with control empty vector (RIE), Ras (RIE/Ras), Ras and wtPKCι (RIE/Ras/wtPKCι), or Ras and kdPKCι (RIE/ Ras/kdPKCι), and evaluated for growth in soft agar. (a) Representative experimental results. Numbers in parentheses represents number of colonies/dish. (b) Values represent the average of three independent experiments ± SEM. *P < 0.002 versus RIE/Ras. (c) The indicated RIE cell transfectants were analyzed as described in panel a. Values represent the average of five determinations ± SEM. *P = 0.008 versus RIE/Ras; **P = 0.0001 versus RIE/Ras/kdPKCι. (d) 12-wk-old K-Ras^LA2 and K-Ras^LA2/kd PKCι mice were analyzed for ACF in the proximal colon. Average number of ACF per mouse is plotted ± the SEM; n = 5; *P = 0.04.